]]
Publications
Supervising Scientist Report 164
van Dam RA, Walden DJ and Begg GW
Supervising Scientist, 2002
ISSN 1325-1554
ISBN 0 642 24370 0
A hard copy of the full report can be ordered from Publications, Supervising Scientist Division. A full list of SSD publications including prices is available on the publications page.
An order form can be downloaded and faxed or posted back. Alternatively, you can email your order to publications_ssd@environment.gov.au or phone +61 08 8920 1100.
1. Background and approach
Cane toads (Bufo marinus) entered the Northern Territory (NT) in 1980 from Queensland and are rapidly approaching Kakadu National Park (KNP), having recently been reported in the upper Mann River and Snowdrop Creek, approximately 15-30 km to the east of Kakadu National Park. Concern about the invasion of cane toads in Kakadu National Park has been highlighted on a number of occasions, and in 1998 participants at a workshop on the potential impacts and control of cane toads in Kakadu National Park conceded that a strategic approach for assessing and possibly minimising cane toad impacts should be developed. The first stage would be an ecological risk assessment to predict the likely extent of impacts of cane toads in Kakadu National Park and identify key vulnerable habitats and species. This information could be used to develop new monitoring programs and assess existing ones. This assessment is a direct result of Environment Australia's concern about the potential impacts of cane toads in Kakadu National Park.
The wetland risk assessment framework developed by eriss for the Ramsar Convention was used to predict key habitats and the species most at risk. The majority of the assessment involved identifying the problem, the potential extent and effects of the problem, the risk, and subsequently making recommendations on monitoring. Major information gaps relevant to predicting impacts and developing appropriate monitoring programs were also identified.
The risk assessment was based on information from published and unpublished scientific and anecdotal reports. Information on Kakadu National Park was derived from relevant research projects undertaken in the Park since the early 1980s. A number of relevant Territory and Commonwealth agencies were consulted, as were relevant cane toad, native fauna and/or wildlife management experts from around Australia. Discussions were held with community members in the Borroloola and Mataranka regions to gain an indigenous/cultural perspective of the cane toad issue.
2. Identification of the problem
Since their introduction to Australia in 1935 to control sugar cane pests in Queensland, cane toads have spread naturally and with human assistance throughout much of Queensland, northern NSW and the Top End of the NT. The cane toad's preference for certain disturbed areas means that areas of degraded natural habitat have probably helped their spread. They eat a wide variety of prey, breed opportunistically, have a far greater fecundity than native anurans, and develop rapidly particularly in warmer waters. They tolerate a broad range of environmental and climatic conditions, can occupy many different habitats and compete for resources with many native species. Most significantly, they possess highly toxic chemical predator defences, with many experimental and anecdotal reports of deaths of native predators that have attempted to consume cane toads.
It is accepted that the cane toad will establish and spread rapidly in Kakadu National Park - a World Heritage area with Ramsar listed wetlands, well known for its spectacular wilderness, nature conservation values, rich diversity of habitats, flora and fauna, and cultural significance. There is serious concern that the World Heritage status of Kakadu National Park could be diminished if any of these attributes were adversely affected by cane toads.
3. The potential extent of cane toads in Kakadu National Park
Cane toads are likely to colonise almost every habitat type within Kakadu National Park. The saline regions of the coastal plains and deltaic estuarine floodplains will most likely support some cane toads at various times, although they are not likely to use these habitats on a permanent basis. Other less suitable areas include deep open water and/or flowing channel habitats and tidal regions of larger rivers (excluding riparian zones) which extend 70 to 80 km inland during the Dry season. The steady range expansion over the last ten years indicates that most wetland habitats are probably suitable as breeding habitat and also as Dry season refuges.
Patterns of dispersal within Kakadu will probably rely on the transport corridors and the major rivers and creeks. Dispersal rates within a catchment could be up to 100 km y-1. The current location of cane toads would indicate an initial progression down the South Alligator River catchment via its sub-catchments (eg Jim Jim Creek, Deaf Adder Creek). Invasion of other areas of the Park will likely depend on which waterways' headwaters are colonised first (eg Mary River, East Alligator River).
Maximum population densities of various cane toad life stages for limited areas of suitable habitat in Kakadu could be expected to be in the order of: 4000 to 36 000 eggs per metre of shoreline; ~15 to 60 m-2 for tadpoles; 2.5 m-2 for metamorphlings; and 2000 ha-1 for adults, depending on temporal and spatial factors.
The Dry season will see a gradual retreat of many cane toads from seasonally inundated wetlands. The vegetation and cracks in the black soils on the floodplains should offer sheltered, moist habitat during the mid Dry season. In the late Dry season, adult cane toads will congregate near permanent water with adequate shelter. Few cane toads would be present in the drier areas of the tall, open eucalypt forest and woodland habitats of the lowland plains. The first rains of the Wet season will stimulate dispersal and increased breeding activity. With the progression of the Wet season, cane toads will disperse into terrestrial habitats, namely the open forests and woodlands. When large areas of the floodplains are inundated, cane toads will be concentrated on the remaining dry ground, which may make them highly visible to Park visitors.
4. The potential effects of cane toads on Kakadu National Park
The potential effects of cane toads upon Kakadu National Park are outlined in six sections of this report: effects on predator species; effects on prey species; effects of resource competition; cultural effects; economic effects; and other potential effects.
4.1 Predators
The majority of information on cane toad impacts relates to toxic effects on predators. A substantial amount of literature exists on effects on individuals, but little scientific information is available on population effects. The degree of susceptibility of potential cane toad predator species in Kakadu National Park was determined using three criteria:
Definite: documented adverse effects upon populations of this species have been reported in the literature;
Probable: documented in the literature as having eaten cane toads or their early life stages and adverse effects on individuals reported, but not on populations;
Possible: documented in the literature or through expert consultation as eating, or thought likely to eat, native frogs or their early life stages, but effects of eating cane toads unknown.
A total of 151 species or species groups were identified under these criteria, covering a broad taxonomic range including aquatic invertebrates, fish, frogs, lizards, snakes, birds and mammals. Eleven species were considered definitely susceptible to cane toads, comprising 5 lizard, 3 snake and 3 mammal species. Sixteen species or species groups were considered probably susceptible to cane toads, while 124 species or species groups were considered possibly susceptible to cane toads.
4.2 Prey
Little information was available on effects of cane toads on prey species. Cane toad tadpoles have been observed preying on the eggs of some native frogs, though they are thought not to be significant predators of native anuran early life stages. Rather, cane toad tadpoles have been observed to feed mainly on cane toad eggs, algae and detritus, as well as scavenging upon dead animals and animal material which they will consume in preference to plant material. Juvenile and adult cane toads are generalist feeders, consuming almost any type of terrestrial animals, with ground-dwelling ants, termites and beetles usually dominating the diet. Some small mammals, birds, reptiles and frogs are consumed in very small numbers. No study has specifically investigated the impact of cane toads on communities of ground dwelling arthropods. One general impact study reported a decline in beetle (Coleoptera) numbers, possibly due to cane toads. It is impossible to determine how many of the undescribed invertebrate species in Kakadu, many of which may be endemic, could be affected by cane toads.
4.3 Competition
Little information was available on competition between cane toads and native animals for resources such as food, shelter and breeding sites. The potential for competition between cane toad tadpoles and native frog tadpoles (eg the ornate burrowing frog) appears to exist, although, several reports suggest considerable segregation of breeding sites. Competition between adult cane toads and frogs appears to be minimal, with the pattern of habitat and food exploitation differing markedly. The major factor separating resource use is the cane toad's heavy reliance on ground-dwelling ants, termites and beetles as major food sources. There has been some indication from the Roper River region of the NT of competition effects. In particular, some species of small reptile were found to decline in areas colonised by cane toads. A competition effect was suspected, but not confirmed. Two frog species (the brown tree frog and green tree frog) have possibly been linked to competition-related declines, although the evidence is not strong. It is possible that many other species within Kakadu, including endemic aquatic invertebrates, could be subject to competition by cane toads.
4.4 Cultural effects
Concerns for the decline in numbers of bush tucker species such as monitor lizards, snakes and turtles have already been noted by several Aboriginal communities in the NT. This decline is likely to have very significant impacts upon Aboriginal communities within Kakadu. Some traditional ceremonies in the Borroloola region have been altered to request the spirits to return these foods, and in some cases, totem species (eg freshwater crocodile). From experience elsewhere in the NT, it appears that Aboriginal people, by necessity, eventually grow accustomed to the presence of cane toads, although this does not necessarily diminish the underlying concerns of these people. Areas of human habitation in Kakadu including the township of Jabiru, Aboriginal communities, Ranger stations, tourist accommodation and camping grounds are expected to have high densities of cane toads. This will impact on outdoor recreational activities and, in some areas, increase the likelihood of pets being poisoned from mouthing or ingesting cane toads.
4.5 Economic effects
Cane toads are unlikely to have an adverse impact on the general economy and tourism income of Kakadu National Park. The reactions to cane toads in the NT have ranged from disinterest to dismay. International tourists do not recognise toads as an invasive species, while visitors from Queensland are well accustomed to toads. However, tourists from other states express deep concern about cane toads, especially in World Heritage sites such as Kakadu. Tour operators in Kakadu share a similar concern. However, the major attributes of Kakadu continue to attract tourists, and are likely to overshadow any concerns about adverse economic impacts of cane toads.
Cane toads do have an economic value as dissecting specimens for research and education purposes, and as a supply for medicinal and leather products. Such industries exist in Queensland and will probably become established in the NT once cane toads are present in sufficient numbers.
4.6 Other potential effects
Another potential effect is the contamination of water supplies with rotting toad carcasses and the subsequent release of the toxins. There have been many reports of the poisoning of pets and poultry from drinking contaminated water. Experimental water-borne exposure of the toxin to various organisms has resulted in toxicity, but generally only at high concentrations.
The issue of potential impacts of cane toads on granivorous prey insects and resultant repercussions on Kakadu's native plants has been raised, although this is highly speculative. There is evidence, for example, that high densities of harvester ants can significantly reduce the density of speargrass (Sorghum intrans). In terms of plant-animal interactions, it is possible that subtle ecological changes could occur amongst other biota, and other flow-on effects.
Feral cats and pigs have been known to die from mouthing or ingesting cane toads. These animals cause damage to the native fauna and landscape of Kakadu, and any decline in their numbers would be considered a benefit. The reduction in numbers of predators such as varanids (goannas) and snakes could be of benefit to the several species of ground-dwelling/nesting birds in Kakadu, in addition to crocodiles and turtles whose eggs are preyed upon by other large reptiles.
Cane toads are known to feed on human faeces, and as a result they may harbour human strains of Salmonella and other bacteria. The eggs of human parasites are also spread via toad faeces. In areas where modern sanitation practices are lacking, the presence of large numbers of cane toads could represent a health hazard. Another health-related issue is the potential for substance abuse of the cane toad toxin, a habit forming practice that is established in northern Queensland and in countries such as Fiji.
5. Identification of the risk
The data on cane toad effects, distribution and densities are mostly inconclusive and/or show great variability. In addition, information on distributions and abundance of Kakadu animal species are deficient. Nevertheless, it is still possible to identify key habitats and also prioritise particular species based on the likelihood that they will be at greater risk from cane toads than other species, and their importance to the ecological and/or cultural values of Kakadu.
5.1 Identification of key habitats
5.1.1 Aquatic stages
In Kakadu, cane toads will breed in both temporary and permanent waterbodies and so their aquatic stages will be found in a variety of aquatic habitats. They will concentrate their breeding activity during the wetter periods, although they are also known to breed during the Dry season. During the Wet season, when many of the major wetland habitats are inundated, cane toad breeding may be concentrated in the wetland habitats associated with the open forests and woodlands of the lowland plains.
5.1.2 Terrestrial stages
As the Dry season progresses, cane toads will move progressively from sites of temporary water to permanent water. The floodplains and sheltered habitats on the margins of floodplains and temporary or shallow billabongs will provide ideal cane toad habitat during the early to mid Dry season. The late Dry season will see high densities of cane toads near permanent water or moisture, including permanent billabongs and patches of monsoon rainforest.
The Wet season will probably see the highest numbers of cane toad metamorphlings, mainly around the moist margins of the waterbodies they emerged from. Wet season inundation of the major wetland habitats will see the majority of adult cane toads dispersing into the woodlands and open forests of the lowland plains. The vegetation within the woodlands will provide suitable shelter for cane toads during the Wet season.
5.2 Identification of species at risk
5.2.1 Predators
The initial susceptibility ranking of each of the 151 predator species identified as being definitely, probably or possibly susceptible to cane toads was further refined to a ranking of risk using exposure (ie available habitat overlap, feeding ecology, behaviour) and ecological/cultural importance status information. Four risk categories - likely, possible, uncertain and unlikely - were defined, being adapted from the original susceptibility criteria. Within these categories, different priorities were assigned.
The original 151 predator species were allocated a risk ranking accordingly. Ten species were considered likely to be at risk of experiencing population level effects, with the northern quoll being assigned the highest priority. The 9 remaining species including 5 lizards, 3 snakes, and one mammal were assigned high priority. Twelve species or species groups were considered to be at possible risk of experiencing population level effects, although none were listed as endangered or vulnerable, or thought to be notable (rare, or have restricted range, outstanding taxonomic interest, or uncertain or declining status) or flagship (ecological/cultural importance to Kakadu) species. Thus, all species were assigned moderate priority status. Represented in this category were two groups of aquatic invertebrates, 3 frogs, one lizard, 3 snakes, freshwater crocodile and 2 birds. Due to a lack of information, the risk of population level effects was considered to be uncertain for 98 species or species groups, although 21 of these were assigned high priority. These species include 3 fish, 3 frogs, 6 lizards, one snake, 4 birds and 4 mammals. The remaining species in this risk category were assigned moderate priority. These include two groups of invertebrates, 4 fish, 17 frogs, 9 snakes, 42 birds and 3 mammals. A total of 31 species were considered unlikely to be at risk of experiencing population level effects (based on relevant ecological, feeding or behavioural information) and were assigned low priority. These included 11 fish, 18 birds and 2 mammals.
5.2.2 Prey
Quantitative data on impacts to prey species are scant, and very little can be concluded about the species or species groups at risk. Cane toads occasionally consume small vertebrates, but populations of these are not likely to be at risk. There is little doubt that termites, beetles and ants will be heavily exploited by cane toads in Kakadu. Due to the potentially high cane toad densities, and an individual cane toad's ability to consume up to hundreds of prey items in one night, ground-dwelling arthropods are at greatest risk. The potential impact of cane toads on endemic invertebrates is unknown. The only species known to suffer long-term population decline or extinction from the impact of cane toads is a tapeworm found in the intestines of a snake.
5.2.3 Competitors
The available experimental information suggests that some native frog tadpoles (eg L. ornatus) may be at risk through competition with cane toad tadpoles. However, observations suggest that native frogs rarely share breeding habitats with cane toads. Although adult native frogs do not appear to compete with cane toads, the potential risk to native tadpoles represents a risk to native frog populations. Some of the smaller insectivorous reptile species of Kakadu may be at risk from competition for food resources by cane toads, but nothing more can be concluded.
5.3 Cultural, socio-economic and other risks
The major impacts on Aboriginal communities within Kakadu National Park will be a decline in some traditional foods and, in some situations, the alteration of ceremonies following declines of food and totem species. Aboriginal people elsewhere in the NT have accepted the presence of cane toads but still express concern regarding the impacts. Aboriginal communities within Kakadu may also become accustomed to cane toads albeit most likely sharing the same concerns. Cane toads will congregate in areas of human habitation within Kakadu, and will be of nuisance value in these places, and will also represent a risk to domestic and semi-domestic dogs. (No domestic cats are allowed to be kept in the Park.)
Tourism, the major economic activity of Kakadu, is not at risk from the presence of cane toads, and visitor numbers will not decrease as a result. With predicted high numbers in Kakadu, there may be an opportunity to harvest them for commercial benefit.
Other potential effects of cane toads have been hypothesised, including the contamination of water supplies, secondary effects on vegetation communities, the spread of human diseases, and the substance abuse of cane toad toxin. Details of these potential effects and hence the risks posed by them are essentially unknown.
6. Uncertainty and information gaps
This assessment has highlighted that there are major information gaps contributing to a large degree of uncertainty about the potential extent and impacts of cane toads in Kakadu. These include: uncertainty about densities of cane toads in Kakadu, effects of fire and burning regimes, degree of land/habitat disturbance, and the extent to which the Arnhem Land escarpment and plateau will act as a barrier and/or be colonised; the lack of quantitative data on the impacts on animal populations, particularly in the long-term, quantitative data on Kakadu fauna populations and distributions as well as dietary information; incomplete knowledge of Kakadu's invertebrate fauna, many being undescribed and possibly endemic; unknown response and susceptibility of most Kakadu fish species; unknown competitive interactions with native frogs; unknown chemoreceptive response in snakes and their ability to detect cane toad toxins; conflicting and unclear information on freshwater turtles; insufficient information on conservation listed species; the lack of experimental or anecdotal evidence regarding effects on bats; and impacts to as yet unidentified endemic species.
7. Recommendations for additional surveys and monitoring
7.1 Priority habitats for monitoring
Seven major habitat types were identified for future monitoring: floodplain communities; swamp communities; monsoon forest; riparian communities; woodland and open forest communities; springs, soaks and waterholes; and escarpment/plateau pools.
7.2 Priority species for monitoring
The species of most concern, and therefore a priority for monitoring, include the northern quoll, sandstone antechinus, red-cheeked dunnart, brush-tailed phascogale, dingo, all of the varanid lizards, northern death adder, king brown snake, western brown snake, ghost bat, black-necked stork, comb-crested jacana, Oenpelli python and freshwater crocodile. These are based on their risk rating, notability or listing as vulnerable, and also importance to Aboriginal people.
Given that many species assigned to risk category 3 were done so due to a lack of information about effects of cane toads, it is possible that further information could result in the re-prioritisation of some species.
Although risks to prey species are unknown, beetles, termites and ants should be considered for inclusion in monitoring programs.
Monitoring the possible effects of competition between cane toads and native aquatic invertebrates and vertebrates should be given high priority, particularly in escarpment/plateau pools where endemic species are known to exist. Similarly, monitoring for competitive effects between adult cane toads and insectivorous reptiles should also have high priority.
7.3 Priorities for addressing information gaps
A number of information gaps require addressing before more confident estimates of risks can be derived. Monitoring programs assessing the effects of cane toads upon Kakadu species will allow greater understanding of the risks. There is a need for appropriate baseline data, not just for cane toads but to monitor and assess other management issues that will arise in the future (eg other invasive species, fire and tourism). In addition, surveys should be conducted to identify and map the distribution of the endemic species of Kakadu, particularly in the escarpment and sandstone regions. All survey and/or monitoring programs should concurrently measure cane toad abundances and habitat preferences. Other information gaps that could be addressed but are less of a priority, include the effects of fire on cane toads and the lack of information for particular species or species groups (eg freshwater turtles, red goshawk).
7.4 Evaluation of past and present monitoring programs
As it may be several years before all of Kakadu is occupied by cane toads (eg some escarpment/plateau habitats), it is possible that some new monitoring programs may have sufficient time to accumulate pre-cane toad (ie baseline) data. It is highly unlikely that new monitoring programs will have time to provide similar data for many floodplain and lowland habitats. Data from major past and present monitoring programs within Kakadu may provide an alternative, noting that they were developed with objectives other than cane toad impacts in mind.
7.4.1 Broad scale surveys
The two major fauna surveys of the last 20 years provided information on abundances, distribution and habitat preferences of birds, mammals, reptiles and amphibians in a range of habitats similar to those identified in this report. The information from these surveys is not appropriate to use as current baseline. However, the established sites provide the opportunity for re-sampling before cane toads arrive. Not all habitat types were included in these surveys.
A proposed representative re-sampling of the Stage 3 Wildlife Survey, which is hoped to be undertaken next Dry season could possibly provide one season of pre-cane toad data.
7.4.2 Ongoing monitoring programs
The only major ongoing fauna monitoring programs in Kakadu National Park are those associated with assessing potential environmental impact downstream of ERA Ranger Mine and the Jabiluka lease area. Monitoring programs are being conducted by eriss and ERA/EWL Sciences (Energy Resources of Australia Ltd/Earth Water Life Sciences).
Aquatic macroinvertebrates are monitored at sites in the Magela Creek system (since 1988) and a number of control sites elsewhere in the Park. Sites from other areas have also been monitored regularly in the past (ie upper South Alligator River and Baroalba, Nourlangie and Gulungul Creeks). Though these studies were not designed for detecting cane toad impacts, inferences would be enhanced if cane toad invasion/distribution was monitored. Billabongs sampled in the Magela and Nourlangie Creek systems may provide information on (potentially vulnerable) freshwater snails.
Fish communities in the Magela, Nourlangie and upper East Alligator systems have been monitored annually since 1994, and data exist for fish migration patterns in Magela Creek from 1985 to 1996.
'Whole-ecosystem' monitoring by ERA/EWLS has also been conducted at sites in Swift, Magela and Nourlangie Creek systems. Zooplankton, macroinvertebrates, fish, frogs, reptiles, bushbirds, waterbirds and mammals were surveyed in 1994/95 and again in 2000/01.
7.4.3 Other surveys or monitoring programs
Other past programs may also contribute to background information, including surveys of waterbirds on the Magela and Nourlangie floodplains. It has been proposed to re-survey the original Magela floodplain sites, in order to update/add to the existing information on birds.
Information from the CSIRO Kapalga fire study from the late 1980s to the mid 1990s will provide a useful basis for detecting and assessing impacts once cane toads arrive there. Mammals, reptiles and insects were sampled originally and were re-sampled for small mammals in 1999.
It will be very difficult to obtain adequate baseline data for a cane toad impact monitoring program. While the ongoing programs will be of some use, they are not necessarily targeted at the priority species identified in this report.
8. Risk management and reduction
Given the outcomes of the assessment, some relevant issues can be discussed that may assist Park managers in developing a risk management strategy.
Parks Australia North has already been active with regards to management of cane toad issues, having initiated a cane toad identification training program and rapid response strategy to manage human assisted incursions of cane toads. Additionally, frog recording stations have been established at four sites in Kakadu (and more are planned). Baseline data have been collected for the past two Wet seasons.
Very little will be able to be done to reduce cane toad numbers in Kakadu. Particular measures may prove effective in localised areas (eg townships, caravan parks), but efforts would need to be ongoing. Management of areas damaged by feral pigs may help reduce the densities of cane toads in pig-affected areas. Chemical and biological control methods are insufficiently developed at this stage.
It is recommended that Parks Australia North manage the invasion of cane toads initially by i) ensuring that monitoring efforts are underway to assess the impacts of cane toads upon the natural and cultural values of Kakadu, and ii) investigating measures by which cane toads can be managed on a localised basis.
The preliminary risk assessment provides a starting point from which Parks Australia North can determine the monitoring requirements for fauna. In addition, it provides an overview of the potential cultural and socio-economic impacts, which could be studied in greater detail by appropriate experts.
The full report can be ordered from publications_ssd@environment.gov.au
van Dam RA, Walden DJ and Begg GW 2002. A preliminary risk assessment of cane toads in Kakadu National Park. Supervising Scientist Report 164, Supervising Scientist, Darwin.
This report is available in sections and as one whole report. You will need Adobe Acrobat Reader installed on your computer to view the PDF files.
Download report sections:
If you are unable to access the report, please contact the Supervising Scientist Division (telephone 61 (0)8 8920 1100; email enquiries_ssd@environment.gov.au) to organise a suitable alternative format.
