Evaluating the role of the dingo as a trophic regulator in Australian ecosystems

Austral Ecology (2007) 32, 492 501 doi:10.1111/j.14429993.2007.01721.x Evaluating the role of the dingo as a trophic regulator in Australian ecosystems A. S. GLEN, 1 * C. R. DICKMAN, 2 M. E. SOULÉ 3 AND B. G. MACKEY 4 1 Department of Environment and Conservation and Invasive Animals CRC, Dwellingup Research Centre, Banksiadale Rd, Dwellingup,WA 6213, Australia (Email: al.glen@dec.wa.gov.au), 2 Institute of Wildlife Research, School of Biological Sciences A08, University of Sydney, New South Wales, 4 School of Resources, Environment and Society, College of Science,The Australian National University, Canberra, ACT, Australia; and 3 Professor Emeritus, University of California, Santa Cruz, USA Abstract The importance of strongly interactive predators has been demonstrated in many ecosystems, and the maintenance or restoration of species interactions is a major priority in the global conservation of biodiversity. By limiting populations of prey and/or competitors, apex predators can increase the diversity of systems, often exerting influences that cascade through several trophic levels. In Australia, emerging evidence points increasingly towards the dingo (Canis lupus dingo) as a strongly interactive species that has profound effects on ecosystem function. Through predatory and competitive effects, dingoes can alter the abundance and function of mesopredators including the introduced red fox (Vulpes vulpes) and feral cat (Felis catus), and herbivores including the European rabbit (Oryctolagus cuniculus). These effects often benefit populations of native prey, and diversity and biomass of vegetation, but may not occur under all circumstances. For example, the social structure of dingoes is of great importance; a pack subject to minimal human interference regulates its own numbers, and such packs appear to have fewer undesirable impacts such as predation on livestock. Despite abundant observational evidence that the dingo is a strong interactor, there have been few attempts to test its ecological role experimentally. Given the wellrecognized importance of species interactions to ecosystem function, it is imperative that such experiments be carried out. To do this, we propose three broad questions: (i) do dingoes limit the abundance of other predators or prey? (ii) do dingoes affect the ecological relationships of other predators or prey (e.g. by altering their spatial or temporal activity patterns)? and (iii) does the removal or reintroduction of dingoes entrain ecological cascades? Finally, we discuss the design of appropriate experiments, using principles that may also be applied to investigate species interactions on other continents. Research might seek to clarify not only the impacts of dingoes at all trophic levels, but also the mechanisms by which these impacts occur. Key words: dingo, interaction, keystone, mesopredator release, trophic cascade. INTRODUCTION The importance of toporder predators in maintaining ecosystem function has been demonstrated in many marine and terrestrial systems (e.g. Paine 1966, 1969; Estes & Palmisano 1974; Soulé et al. 1988; McLaren & Peterson 1994; Henke & Bryant 1999). By limiting populations of their prey and/or subordinate competitors, toporder predators can modulate the diversity of a system, and may ultimately increase plant biomass via a series of trophic links (Paine 1980; Pace et al. 1999; Polis et al. 2000; Schmitz et al. 2000; Ripple & Beschta 2003). Species that perform this function are known as keystone predators (Paine 1966, 1969). The identification and conservation of such species, and *Corresponding author. Accepted for publication October 2006. maintenance of their interactions, is of vital importance in arresting the global loss of biodiversity (Soulé et al. 2003, 2005). Where toporder predators are completely or virtually removed from a system, the effects on species richness and abundance at lower trophic levels can be profound. Previously subordinate predators may increase unchecked, potentially decimating prey populations (Soulé et al. 1988; Crooks & Soulé 1999). Herbivores can become overabundant, leading to overgrazing on plant populations (McLaren & Peterson 1994). Competitive relationships between prey species may be altered (Paine 1966). In some cases, these effects lead ultimately to communitylevel trophic cascades (Polis et al. 2000) in which plant biomass is redistributed throughout the system. In Australia, a wealth of observational evidence, backed by a small but growing body of experimental work, suggests that removal of the continent s largest

THE DINGO AS A TROPHIC REGULATOR 493 terrestrial predator can lead to such impacts. The dingo (Canis lupus dingo) is a large canid, which represents an early form of domesticated wolf, and was introduced to Australia by Asian seafarers some 3500 4000 years ago (Gollan 1984; Corbett 1995). The arrival of the dingo is likely to have contributed to the extinction of the thylacine (Thylacinus cynocephalus) and the Tasmanian devil (Sarcophilus harrisii) on the Australian mainland (Corbett 1995). However, changes in human hunting technology around the same time may also have been involved (Johnson & Wroe 2003). Since the advent of the pastoral industry, the dingo has been viewed as an agricultural pest, and human persecution has seen its virtual extirpation from much of southeastern Australia (Rolls 1969; Glen & Short 2000; Fleming et al. 2001). In many areas where the dingo still persists, it is reviled and subject to continued human harassment such as poisoning, shooting and trapping (Fleming et al. 2001). Given the dingo s role as the top predator in Australian terrestrial ecosystems, what effects does its removal or persecution have upon ecosystem function? Three relatively recent arrivals in Australia, the red fox (Vulpes vulpes), feral cat (Felis catus) and European rabbit (Oryctolagus cuniculus), have become major threats to biodiversity and agriculture (Rolls 1969; Saunders et al. 1995; Williams et al. 1995; Dickman 1996). All three of these species interact with the dingo and with each other (Pech et al. 1992; Corbett 1995; Holden 1999; Holden & Mutze 2002; Glen & Dickman 2005). Rabbits, in turn, are linked to plant biomass, vegetation structure and diversity through direct grazing effects (Auld 1992; Williams et al. 1995; Lord 2000). Thus, persecution of the dingo may have impacts that cascade through the trophic levels from predator to mesopredator to herbivore and ultimately to primary producers. In many respects, the situation of the dingo mirrors that of the grey wolf (Canis lupus ssp.) in North America (e.g. Bradley et al. 2005), in that its management is governed by the conflicting goals of protecting livestock while maintaining topdown ecosystem regulation and ecosystem integrity. Here we investigate evidence for trophic regulation of ecosystems by the dingo, suggest some questions for future research, and discuss the design of experiments to test these. We see these evidencebased steps as critical for providing informed management of the dingo in future. Evidence for impacts of dingoes Large carnivores can limit the density of smaller ones by several means, both direct and indirect. Direct effects include interspecific aggression or predation (e.g. Palomares & Caro 1999), physical exclusion from contested resources (e.g. Johnson & Franklin 1994) or kleptoparasitism (e.g. Gorman et al. 1998). Indirect effects may result from exploitation competition where prey or other resources are limiting. There is evidence to suggest that dingoes affect feral cats and foxes by all these means. Of course, predators also affect herbivore populations through direct predation (e.g. Pech et al. 1992; Banks 2000), and there is evidence that a number of native and introduced herbivores, including rabbits, are limited by predation from dingoes. In general, solitary dingoes kill small to mediumsized prey, whereas cooperative hunting by a stable pack is required to subdue larger animals (Corbett 1995). As discussed below, stable packs may occur only where there is minimal human disturbance. Direct effects The most direct and obvious means by which dingoes might suppress smaller predators is by killing them, and indeed this has been documented in several studies. For example, Marsack and Campbell (1990) directly observed dingoes preying on foxes, and found fox remains in 6.1% of 49 digestive tracts and 2.4% of 82 dingo scats. Cat remains were also found in one gut sample and one scat. Consumption of cats by dingoes has also been reported by Newsome et al. (1983), LundieJenkins et al. (1993), Thomson (1992a), Corbett (1995) and Paltridge (2002). In addition, Pettigrew (1993) reported that an adult cat fitted with a radio collar was killed by a dingo. Similar interactions have been reported in the UK between domestic dogs (Canis lupus familiaris) and red foxes. Juvenile foxes were preyed upon by dogs, while adults were harassed, chased and occasionally killed.these interactions limit the distribution of foxes in Bristol (Harris 1981). As well as exerting predatory effects, dingoes probably exclude smaller predators from resourcerich patches by scent marking. Odours produced by urine, faeces and glandular secretions are used widely by carnivores to communicate ownership of territories and food resources (such as rabbit warrens), social dominance and reproductive status (Ewer 1973; Macdonald 1979, 1987; Thomson 1992b; O Neill 2002). Such odours can also deter subordinate species, which may be vulnerable to aggression or predation (e.g. Dickman & Doncaster 1984). Observations in semiarid areas of Australia suggest that foxes are excluded from the warrens of greater bilbies (Macrotis lagotis) and nest mounds of malleefowl (Leipoa ocellata) by the presence of dingo faeces (O Neill 2002). By denying foxes access to such resources, dingoes might limit the abundance and distribution of their smaller counterparts. Perhaps more importantly, this phenomenon could reduce the impact of foxes on these endangered prey species, even if fox abundance is not reduced.

494 A. S. GLEN ET AL. Dominant predators may also exclude subordinate ones on a larger spatial scale. For example, in North America, red foxes have been observed to occupy the interstices between the home ranges of coyotes (Canis latrans) (Harrison et al. 1989), and coyotes appear similarly to avoid areas frequently used by wolves (Arjo & Pletscher 1999). In addition to killing or interfering with subordinate predators, dingoes also exert predatory effects on herbivores. Rabbits are a staple prey in many areas, and native herbivores such as kangaroos and wallabies are also commonly consumed (Robertshaw & Harden 1985, 1986; Thomson 1992a; Corbett 1995). Dingoes may be capable of limiting or regulating populations of herbivores, particularly in the presence of alternative prey. Corbett (1995) reported that rabbit populations can increase following dingo control, and suppression of rabbits by dingoes is thought to have beneficial effects on vegetation communities (Newsome 2001). In addition, Caughley et al. (1980) and Pople et al. (2000) reported that populations of red kangaroos (Macropus rufus) are more abundant in the absence than in the presence of dingoes, although differences in habitat and livestock grazing practices may contribute to this pattern (Caughley et al. 1980; Newsome et al. 2001). Similarly, Newsome (1990) observed that eastern (Macropus giganteus) and western grey kangaroos (M. fuliginosus), euros (M. robustus), feral goats (Capra hircus) and feral pigs (Sus scrofa) all occurred where dingoes were excluded by fencing, but were absent from adjacent areas where dingoes still occurred. The predatory effects of dingoes may, however, be environmentally modulated (Newsome et al. 1989), so that little or no effect of dingoes is observable when resources such as food and water are abundant (e.g. Eldridge et al. 2002) or where there is dense cover (vegetation). As well as affecting the abundance of herbivores, predators can influence the foraging behaviour of their prey. For example, Ripple and Larsen (2000) hypothesized that grey wolves alter the habitat use of elk (Cervus elaphus), thereby protecting some areas from grazing and allowing recruitment of aspen (Populus tremuloides) (see also Ripple & Beschta 2003). Dingoes, by excluding herbivores such as rabbits or red kangaroos from certain areas, may be capable of exerting similar effects. Such changes in the behaviour or distribution of a prey species may benefit other native herbivore species. Robertshaw and Harden (1986) produced evidence for the influence of dingo predation upon the timing of reproduction in a swamp wallaby population; dingo predation induced the wallabies to reproduce yearround rather than seasonally or episodically, reducing the potential for predatorswamping and potentially exposing the wallabies to a greater sizerange of predators. This could be an example of a top predator trapping a prey population in a (multipredator) predation sump. Indirect effects In addition to predation, competitive aggression or exclusion, dingoes can affect cats and foxes through competition for limiting resources. Competition is likely to be particularly strong during times of low food availability such as drought (Corbett 1995) and in regions with little vegetation cover. Substantial dietary overlap occurs between the three species, and dingoes can monopolize large carcasses which would otherwise serve as a source of carrion for cats and foxes (Corbett 1995; Mitchell & Banks 2005). Paradoxically, in some situations, dingoes may benefit cats and foxes, which scavenge from carcasses killed by their larger rivals (e.g. Paltridge et al. 1997). Similar relationships have been reported by Macdonald (1987), Dickman (1992), Creel (2001) and Switalski (2003), where subordinate competitors suffer predation or kleptoparasitism from larger predators, but may also scavenge carcasses killed by the dominant species. If dingoes do indeed limit populations of cats and foxes, we would expect to see inverse relationships between these species in terms of abundance and distribution. Although the presence of dingoes does not preclude the occurrence of cats or foxes (e.g. Catling & Burt 1995; Paltridge et al. 1997), it has often been observed that foxes are scarce where dingoes are abundant, and vice versa (Thompson 1983; Jarman et al. 1987; Johnson et al. 1989; Smith & Quin 1996; McRae 2004). Newsome et al. (1997) found an inverse relationship between dingo and fox activity in Kosciuszko National Park and Nadgee Nature Reserve in southeastern Australia, potentially reflecting a relationship between the densities of the two species, or behavioural avoidance of dingoes by foxes. Similarly, Dickman (1996), using data presented by Catling and Burt (1994), showed that the abundance of cats in southern New South Wales was negatively correlated with that of dingoes. However, the most striking example of this negative relationship comes from Newsome et al. (2001). These authors measured the relative abundance of foxes and feral cats in adjacent areas on both sides of the dingo fence, which excludes dingoes from much of southeastern Australia. Indices of fox abundance were between 20.6 and 7.1 times higher in the absence than in the presence of dingoes. Cats were at similarly low densities on both sides of the fence (Newsome et al. 2001). Despite the striking examples presented above, Catling and Burt (1995) found no significant relationship between the abundances of dingoes and foxes in forested habitats of eastern New South Wales. Structural complexity of habitat can moderate competitive

THE DINGO AS A TROPHIC REGULATOR 495 interactions and intraguild predation (Petren & Case 1998; Creel 2001; Finke & Denno 2002). It is possible therefore that densely vegetated areas such as those studied by Catling and Burt (1995) more readily allow foxes to coexist with dingoes.the availability of shelter may facilitate avoidance or escape from the larger predator. The same effect of habitat structure may apply to the relationship between dingoes and feral cats (Pettigrew 1993; Dickman 1996; Edwards et al. 2002). Little is known of the relationships between dingoes and smaller native predators such as quolls (Dasyurus spp.). The western (D. geoffroii) and northern (D. hallucatus) quoll generally consume small prey (Soderquist & Serena 1994; Oakwood 1997) and are therefore unlikely to suffer strong competition with dingoes for food. The spottedtailed quoll (D. maculatus) preys mainly on mediumsized mammals (Belcher 1995, 2000; Dawson 2005; Glen & Dickman 2006), and may therefore experience some exploitation competition from the dingo. There is also evidence of interference competition. Oakwood (2000) found that competitive aggression from dingoes is a major source of mortality for the northern quoll. Belcher (1995) suggested that dingoes steal prey from spottedtailed quolls, although it is also thought that quolls may scavenge from carcasses killed by dingoes (Edgar & Belcher 1995). In addition to these direct effects, dingoes likely have an indirect positive influence on quoll populations. If dingoes suppress populations of cats or foxes, their presence may be of overall benefit to quolls; a possible example of indirect commensalism. Similarly, wolves in northern Europe are thought to have benefited arctic foxes (Alopex lagopus) by suppressing populations of red foxes (Hersteinsson et al. 1989). In the USA s Greater Yellowstone Ecosystem, wolves suppress coyotes but not red foxes, so that foxes are released from competition with coyotes (Smith et al. 1999). Through limitation of cats and foxes, dingoes probably also have indirect positive effects on populations of native marsupials and rodents, including the greater bilby (Pettigrew 1993), conilurine rodents (Smith & Quin 1996) and the endangered parma wallaby (Macropus parma) (Short & Smith 1994). Miller and Harlow (unpub. 2005) have documented a similar pattern in Grand Teton National Park in Wyoming. The lower abundance and altered habitat use of coyotes in an area frequented by wolves was associated with higher rates of capture of small mammals, presumably because coyotes consume more small prey than do wolves. Effects of dingo removal If the influence of dingoes upon ecosystems is as pervasive as suggested by the evidence above, we should expect to see substantial impacts when dingoes are removed. Observational evidence suggests that removal of dingoes may allow introduced predators to invade areas where they have previously been rare or absent. For example, Short et al. (2002) reported that foxes invaded areas of the Tanami Desert in central Australia following dingo control. Such new arrivals can have devastating effects. A single immigrant fox is believed responsible for extirpating a local population of the endangered rufous harewallaby (Lagorchestes hirsutus) (LundieJenkins et al. 1993). Densities of feral cats have also been observed to increase following the removal of dingoes (LundieJenkins et al. 1993; Pettigrew 1993; Christensen & Burrows 1995). Similarly in North America, expansion in the distribution of coyotes has followed the decline of the grey wolf (Mech 1970). Such effects may not require the complete removal of the dingo from an area. A strongly interactive species might become effectively absent if its abundance is reduced below a threshold of ecological effectiveness. The relationships within and between trophic levels might break down, just as they would if the strong interactor had been removed entirely (Soulé et al. 2005). Thus, widespread lethal control could potentially reduce dingo populations below their ecologically effective density in the region, even if populations are not entirely eradicated. Such a situation would negate any potentially beneficial effects of dingoes. Remnant dingo populations following widespread lethal control may also be spatially patchy. Social disruption of dingo packs A further danger is that control of dingo populations may lead to undesired ecological effects by destroying the social structure of dingo packs. Within a stable pack, dingoes exist under a strict social system enforced by the dominant pair. This has implications for both the abundance of dingoes and their behaviour. A dominance hierarchy exists, and breeding is restricted to the dominant pair (Corbett 1995). In this way, the birth rate remains at a fraction of its potential maximum, and the pack limits its own numbers. Stable packs also occupy and defend clearly defined territories, demarcated by scentmarks (Corbett 1995).When the stable structure of a pack is disrupted, for example by human interference, these strict behavioural controls are lost. If a dominant female is killed, subordinate females may then produce several litters in place of the usual one, so that lethal control leads paradoxically to increased dingo abundance (O Neill 2002). As well as facilitating higher abundance of dingoes, disruption of packs can alter the age structure of the population (Allen & Gonzalez 1998). Young,

496 A. S. GLEN ET AL. inexperienced, leaderless individuals will be less proficient at killing large prey such as kangaroos, and dingoes hunting alone tend to kill smaller prey (Thomson 1992c; Corbett 1995).Thus, the breakdown of pack structure may change the focus of predation from large mammals (such as kangaroos and feral herbivores) to smaller ones. In addition, relatively naïve individuals must presumably devote a high proportion of their activity to exploring their new territories and establishing a dominance hierarchy. This combination of circumstances is likely to intensify predation on livestock and small native prey, which represent easier pickings than large, mobile prey such as kangaroos. Allen and Gonzalez (1998) divided three pastoral properties into treatment areas, where dingoes were subjected to poison baiting and experimental controls, where no baiting occurred. Calf losses were consistently higher where dingoes were baited than where they were left undisturbed.their results indicate that baiting does not always reduce dingo numbers and, more importantly, that livestock losses are not necessarily related to the abundance of dingoes (Allen & Gonzalez 1998). Similarly, Fleming et al. (2006) report that losses incurred by livestock producers are independent of the density of wild dogs. O Neill (2002) suggested that immigrant dingoes yet to establish their territories are also ineffective at excluding foxes and feral cats. Future research The dingo, based on the evidence reviewed here, is a strong candidate for consideration as a strongly interactive species, and therefore, for protection (Power et al. 1996; Soulé et al. 2003, 2005). It is imperative that the dingo s ecological role be further investigated using rigorous scientific testing. We propose three broad questions that can provide the focus for future experiments: (i) do dingoes limit the abundance of other predators or prey? (ii) do dingoes affect the ecological relationships of other predators or prey (e.g. by altering their spatial or temporal activity patterns)? and (iii) does the removal or reintroduction of dingoes entrain ecological cascades? Do dingoes limit the abundance of other predators or prey? Despite its conceptual simplicity, a number of logistical considerations render this question a challenging one to test experimentally. Ideally, putative interactions should be tested with controlled, replicated experiments that manipulate the densities of predators (Dickman 1996; Robley et al. 2004; Glen & Dickman 2005). However, the necessary scale of such experiments, both in terms of area and time, makes them difficult and expensive. This is not to say that such experiments should not be undertaken; they are the only means by which a robust, convincing case for competition and other interactions may be obtained. However, the time and resources needed to carry out such experiments will be more easily justified if they are accompanied by other, smallerscale experiments. These might provide preliminary evidence to justify the expenditure of greater resources on largescale, more definitive experiments. There are two ways of approaching this question: first by comparing many areas with and without dingoes at a single point in time, and second by conducting experiments at a more limited number of sites, spanning a longer time period. Do dingoes affect the ecological relationships of other predators or prey? Regardless of whether a strongly interactive predator directly affects the abundance of other species, it may exert a strong influence on ecosystems if it alters the behaviour of other species (Schmitz et al. 1997). For example, subordinate species may be forced to alter their use of habitat, or reduce their activity at certain times of day, in order to avoid aggressive encounters with the dominant. In turn, this may alter the effects of subordinate predators on prey populations. Does the removal or reintroduction of dingoes entrain ecological cascades? Assuming that dingoes directly influence their competitors or prey, do these effects lead to greater abundance or diversity of species at lower trophic levels? If dingoes can limit populations of cats and foxes, this might lead to increased abundance or expanded distribution of some prey species. Similarly, if dingoes limit rabbits or other herbivores, this might lead to a greater diversity and/or biomass of vegetation.to illustrate this, consider a hypothetical example of a trophic cascade (Fig. 1), which might occur following the reestablishment of functional dingo packs in an arid area of Australia. In the absence of dingoes, foxes and feral cats are abundant, suppressing or extinguishing populations of small and mediumsized native mammals such as rufous harewallabies. Large native and introduced herbivores virtually free of predators are abundant, leading to overgrazing, soil erosion and compaction, thereby reducing plant biomass and structural habitat complexity. Addition of dingoes to this degraded system could be expected to have myriad effects. First, foxes and feral cats would likely be reduced in abundance, allowing endangered species like the rufous harewallaby to escape predator regulation and recover.

THE DINGO AS A TROPHIC REGULATOR 497 Dingo Fox + Cat Rufous harewallaby Large herbivores + Small vertebrates + Plant biomass / understorey complexity + Fig. 1. A hypothetical example of a trophic cascade. Establishment of functional dingo packs in this previously degraded system leads to reduced abundance of foxes and feral cats, increased abundance of endangered rufous harewallabies, reduced abundance of large native and feral herbivores, increased plant biomass, increased diversity and structural complexity of the understorey, and increased abundance and diversity of small vertebrates. Solid arrows show direct effects, broken arrows show indirect effects, and signs (+, ) indicate the overall effect on the species or category. For simplicity, only selected interactions are shown. Second, dingoes could be expected to limit the density of large herbivores, reducing grazing pressure and allowing plant biomass to increase. The resultant increase in habitat complexity would probably have further benefits for smaller prey species, augmenting the benefits of release from predation by cats and foxes. Thus, in terms of measurable outcomes, the introduction of dingoes to this hypothetical system could be expected to cause: (i) reduced densities of foxes and feral cats; (ii) increased density of rufous harewallabies; (iii) reduced densities of large herbivores; (iv) increased plant biomass; (v) increased diversity and structural complexity in the understorey; and (vi) increased abundance and diversity of small vertebrates. Design of predator manipulation experiments MacNally (1983) outlined three steps for establishing the importance of interspecific competition within a guild: (i) measure overlap of resource use between species; (ii) demonstrate that the presence of one species alters the resource use of another; and (iii) demonstrate that the interaction between species has impacts upon populations (e.g. reduced density) or individuals (e.g. reduced fecundity). These last two steps require manipulative experiments in which one or more species is removed or introduced. Removal experiments may be undesirable if, as in the case of the dingo, the species involved is believed important to ecosystem function. In such cases, reintroduction experiments are more desirable, both ecologically and ethically (Soulé et al. 2005). As outlined by Dickman (1996), predator removal or introduction experiments should follow a BACI (beforeafter, controlimpact) design. Following an initial monitoring period, predator numbers are manipulated at a number of sites, while the remaining sites are left undisturbed, acting as experimental controls. Replication is essential, and the power of the experiment increases with more treatment and control sites (Dickman 1996).The spatial scale of such experiments must be proportional to the home ranges of the study animals. Each study site must be large enough to permit the occurrence and detection of population growth. In addition, the study sites must be separated by sufficient distance to ensure that they are relatively independent of each other (Dickman 1996). Given the highly mobile nature of dingoes, study sites would need to be of the order of tens or hundreds of square kilometres, and separated by distances of at least 50 km (unless fenced). In order to avoid pseudoreplication, treatment and control sites should either be interspersed (if there are few study sites) or assigned randomly (if there are many sites). To avoid other confounding factors, treatment methods should also be simulated in the nontreatment areas. For example, if predators are removed by spotlighting and shooting, animals at the nonremoval sites should be

498 A. S. GLEN ET AL. subjected to spotlighting and the firing of blank ammunition (Dickman 1996). As well as an appropriate spatial scale, perturbation experiments must run over a sufficiently long time so that treatment effects may be differentiated from the stochastic variations typical of Australian landscapes, especially those due to the yeartoyear variability in rainfall and associated primary productivity (Dickman 1996; Robley et al. 2004; Davey et al. 2006). Experiments must also span sufficient time to allow for a response to the perturbation.the importance of this is well illustrated by Moseby (2002), who monitored populations of small mammals within and outside a predator exclosure in arid South Australia.Three years after the exclosure was built, there was no significant difference in small mammal captures inside and outside the fence. However, in the subsequent 2 years, a significant difference became apparent, with a fourfold increase inside the exclosure.thus, even dramatic and ecologically relevant changes in abundance may be overlooked when the duration of experiments fails to reflect the time scale of major environmental perturbations. Having established areas and a multiyear time frame, careful thought must also be given to precisely what to measure. Which taxa are most likely to show a response to the predator manipulation? How can their abundance and distribution best be measured? Can detailed demographic data (e.g. fecundity) be obtained? Can the movements and habitat use of interacting species be monitored? Clearly, monitoring efforts must focus on species that we know or suspect will be affected by predator manipulation (Dickman 1996). This includes species that may be affected directly (e.g. foxes released from competition by the removal of dingoes), or indirectly (e.g. small prey released from fox predation due to the introduction of dingoes). Monitoring must also include structural and floristic surveys of vegetation, and changes in primary productivity so that communitylevel trophic cascades (sensu Polis et al. 2000) may be detected. Because structurally complex habitat may moderate interactions among predators, initial experiments should be conducted in open areas (e.g. arid central Australia), where effects are likely to be detected most readily. Subsequent experiments should be able to build upon the findings of earlier ones to investigate more subtle interactions in complex habitat (Glen & Dickman 2005). CONCLUSIONS The evidence reviewed here suggests that the dingo is an important trophic regulator in Australian ecosystems. A growing body of evidence suggests that dingoes can limit both mesopredators and prey, often with significant flowon effects (e.g. LundieJenkins et al. 1993). Compelling as this evidence appears, it is mostly observational. Few experimental studies have sought to clarify the impacts of dingoes. Given the potential significance of dingoes for ecosystem function, it is imperative that carefully controlled, replicated experiments are conducted. These must seek to clarify not only the impacts of dingoes at all trophic levels within a system, but also the mechanisms by which these impacts occur (sensu Schoener 1974; Tilman 1987). The findings of such experiments will likely have significance for wildlife management and conservation. For example, by ceasing to control dingoes in some areas, or by reintroducing them where they have been eradicated, it may be possible to achieve benefits for biodiversity through the suppression of cats, foxes and rabbits (Johnson et al. 1989; Newsome 1990; LundieJenkins et al. 1993; Short & Smith 1994). Thus, although the initial arrival of the dingo may have had a negative impact on biodiversity by contributing to the extinction on the mainland of the thylacine and Tasmanian devil, we posit that in the absence of these native carnivores the dingo now fills an important role as a beneficial toporder predator. It is not our intention to suggest that reestablishment of dingoes across all areas will be an ecological panacea. A great many anthropogenic changes have occurred since the arrival of Europeans in Australia, not the least of which is the modification and fragmentation of habitats on a vast scale (Recher 2002). The situation is also complicated by hybridization between dingoes and domestic or feral dogs, the behavioural and ecological effects of which are largely unknown. Reintroduction of dingoes to areas greatly modified by human disturbance may not restore systems to their former state. Indeed, some endangered prey populations may be so depleted that they are no longer able to sustain predation from any source, including dingoes. However, many areas might benefit from the restoration of topdown regulation. As emphasized by Boyce and Anderson (1999), population dynamics are not driven solely by either topdown or bottomup effects, but by the interaction of the two. For this interaction to occur, however, top down interactions should be restored where feasible and desirable for biodiversity. In many cases this might require retaining undisturbed dingo populations. We contend that clarification of the dingo s role is a priority for the maintenance of ecosystem function in Australia. Similar principles apply in other parts of the world where conflict with humans has decimated populations of apex predators. The experimental approach described here will also serve to clarify the role of putative highly interactive predators on other continents. We invite comments.

THE DINGO AS A TROPHIC REGULATOR 499 ACKNOWLEDGEMENTS This review was carried out under a Linkage Grant (LP0455163) to B. Mackey and M. Soulé from the Australian Research Council. We are grateful to L. Corbett, P. de Tores, N. Marlow, D. Algar and M. Calver for discussions that were helpful in formulating the scope of the review. Sincere thanks also to the participants in the National Dingo Trophic Regulation Workshop, October 2005 and in particular to C. Holden, who coordinated the workshop. Comments from two anonymous referees helped greatly to improve the manuscript. REFERENCES Allen L. & Gonzalez T. (1998) Baiting reduces dingo numbers, changes age structures, yet often increases calf losses. In: 11th Australasian Vertebrate Pest Conference pp. 421 428. Agriculture Western Australia, Bunbury. Arjo W. M. & Pletscher D. H. (1999) Behavioral responses of coyotes to wolf recolonization in northwestern Montana. Can. J. Zool. 77, 1919 27. Auld T. (1992) The impact of grazing on regeneration of the shrub Acacia carnei in arid Australia. Biol. Conserv. 65, 165 76. Banks P. B. (2000) Can foxes regulate rabbit populations? J. Wildl. Manage. 64, 401 6. Belcher C. A. (1995) Diet of the tiger quoll (Dasyurus maculatus) in East Gippsland, Victoria. Wildl. Res. 22, 341 57. Belcher C. A. (2000) Ecology of the tiger quoll, Dasyurus maculatus, in Southeast Australia (PhD Thesis). Deakin University, Geelong. Boyce M. S. & Anderson E. M. (1999) Evaluating the role of carnivores in the GreaterYellowstone Ecosystem. In: Carnivores in Ecosystems: The Yellowstone Experience (eds T. W. Clark, A. P. Curlee, S. C. Minta & P. M. Kareiva) pp. 265 83. Yale University Press, New Haven. Bradley E. H., Pletscher D. H., Bangs E. E. et al. (2005) Evaluating wolf translocation as a nonlethal method to reduce livestock conflicts in the northwestern United States. Conserv. Biol. 19, 1498 508. Catling P. C. & Burt R. J. (1994) Studies of the grounddwelling mammals of eucalypt forests in southeastern New South Wales: the species, their abundance and distribution. Wildl. Res. 21, 219 39. Catling P. C. & Burt R. J. (1995) Why are red foxes absent from some eucalypt forests in eastern New South Wales? Wildl. Res. 22, 535 46. Caughley G., Grigg G. C., Caughley J. & Hill G. J. E. (1980) Does dingo predation control the densities of kangaroos and emus? Aust. Wildl. Res. 7, 1 12. Christensen P. & Burrows N. (1995) Project desert dreaming: experimental reintroduction of mammals to the Gibson Desert, Western Australia. In: Reintroduction Biology of Australian and New Zealand Fauna. (ed. M. Serena) pp. 199 207. Surrey Beatty & Sons, Sydney. Corbett L. K. (1995) The Dingo in Australia and Asia. UNSW Press, Sydney. Creel S. (2001) Four factors modifying the effect of competition on carnivore population dynamics as illustrated by African wild dogs. Conserv. Biol. 15, 271 4. Crooks K. R. & Soulé M. E. (1999) Mesopredator release and avifaunal extinctions in a fragmented system. Nature 400, 563 6. Davey C., Sinclair A. R. E., Pech R. P. et al. (2006) Do exotic vertebrates structure the biota of Australia? An experimental test in New South Wales. Ecosystems 9, 992 1008. Dawson J. P. (2005) Impact of wildfire on the spottedtailed quoll Dasyurus maculatus in Kosciuszko National Park (MSc Thesis). University of New South Wales, Sydney. Dickman C. R. (1992) Commensal and mutualistic interactions among terrestrial vertebrates. Trends Ecol. Evol. 7, 194 7. Dickman C. R. (1996) Overview of the Impacts of Feral Cats on Australian Native Fauna. Australian Nature Conservation Agency, Canberra. Dickman C. R. & Doncaster C. P. (1984) Responses of small mammals to red fox (Vulpes vulpes) odour. J. Zool. (Lond.) 204, 521 31. Edgar R. & Belcher C. (1995) Spottedtailed Quoll Dasyurus maculatus. In: The Mammals of Australia (ed. R. Strahan) pp. 67 9. Reed New Holland, Chatswood. Edwards G. P., de Preu N., Crealy I. V. & Shakeshaft B. J. (2002) Habitat selection by feral cats and dingoes in a semiarid woodland environment in central Australia. Austral Ecol. 27, 26 31. Eldridge S. R., Shakeshaft B. J. & Nano T. J. (2002) The Impact of Wild Dog Control. on Cattle, Native and Introduced Herbivores and Introduced Predators in Central Australia. Final administrative report to Bureau of Rural Sciences. Northern Territory Parks and Wildlife Commission, Alice Springs. Estes J. A. & Palmisano J. F. (1974) Sea otters: their role in structuring nearshore communities. Science 185, 1058 60. Ewer R. F. (1973) The Carnivores. Cornell University Press, Ithaca. Finke D. L. & Denno R. F. (2002) Intraguild predation diminished in complexstructured vegetation: implications for prey suppression. Ecology 83, 643 52. Fleming P., Corbett L., Harden R. & Thomson P. (2001) Managing the Impacts of Dingoes and Other Wild Dogs. Bureau of Rural Sciences, Canberra. Fleming P. J. S., Allen L. R., Lapidge S. J., Robley A., Saunders G. R. & Thomson P. C. (2006) A strategic approach to mitigating the impacts of wild canids: proposed activities of the Invasive Animals Cooperative Research Centre. Aust. J. Exp. Agric. 46, 753 62. Glen A. S. & Dickman C. R. (2005) Complex interactions among mammalian carnivores in Australia, and their implications for wildlife management. Biol. Rev. 80, 387 401. Glen A. S. & Dickman C. R. (2006) Diet of the spottedtailed quoll Dasyurus maculatus in eastern Australia: effects of season, sex and size. J. Zool. (Lond.) 269, 241 8. Glen A. S. & Short J. (2000) The control of dingoes in New South Wales in the period 1883 1930 and its likely impact on their distribution and abundance. Aust. Zool. 31, 432 42. Gollan K. (1984) The Australian dingo: in the shadow of man. In: Vertebrate Zoogeography and Evolution in Australasia (eds M. Archer & G. Clayton) pp. 921 7. Hesperian Press, Perth. Gorman M. L., Mills M. G., Raath J. P. & Speakman J. R. (1998) High hunting costs make African wild dogs vulnerable to kleptoparasitism by hyaenas. Nature 391, 479 81. Harris S. (1981) An estimation of the number of foxes (Vulpes vulpes) in the city of Bristol, and some possible factors affecting their distribution. J. Appl. Ecol. 18, 455 65.

500 A. S. GLEN ET AL. Harrison D. J., Bissonette J. A. & Sherburne J. A. (1989) Spatial relationships between coyotes and red foxes in eastern Maine USA J. Wildl. Manage. 53, 181 5. Henke S. E. & Bryant F. C. (1999) Effects of coyote removal on the faunal community in western Texas. J. Wildl. Manage. 63, 1066 81. Hersteinsson P., Angerbjörn A., Frafjord K. & Kaikusalo A. (1989) The arctic fox in Fennoscandia and Iceland: management problems. Biol. Conserv. 49, 67 81. Holden C. (1999) The Impact of Rabbit Calicivirus Disease on Fauna of the Flinders Ranges and Implications for Threat Abatement Programs. National Parks and Wildlife South Australia, Adelaide. Holden C. & Mutze G. (2002) Impact of rabbit haemorrhagic disease on introduced predators in the Flinders Ranges, South Australia. Wildl. Res. 29, 615 26. Jarman P. J., Johnson C. N., Southwell C. J. & StuartDick R. (1987) Macropod studies at Wallaby Creek Australia I. The area and animals. Aust. Wildl. Res. 14, 1 14. Johnson C. N. & Wroe S. (2003) Causes of extinction of vertebrates during the holocene of mainland Australia: arrival of the dingo, or human impact? Holocene 13, 1009 16. Johnson K. A., Burbidge A. A. & McKenzie N. L. (1989) Australian Macropodoidea: status, causes of decline and future research and management. In: Kangaroos,Wallabies and Rat Kangaroos (eds G. Grigg, P. Jarman & I. Hume) pp. 641 57. Surrey Beatty and Sons, Chipping Norton. Johnson W. E. & Franklin W. L. (1994) Spatial resource partitioning by sympatric grey fox (Dusicyon griseus) and culpeo fox (Dusicyon culpaeus) in southern Chile. Can. J. Zool. 72, 1788 93. Lord D. (2000) The impact of rabbit grazing on the sucker recruitment of the threatened species purplewood wattle (Acacia carnei). In: Australian Rangeland Society Centenary Symposium (ed. S. Nicolson & J. Noble) pp. 70 74. Australian Rangeland Society, Broken Hill. LundieJenkins G., Corbett L. K. & Phillips C. M. (1993) Ecology of the rufous harewallaby, Lagorchestes hirsutus Gould (Marsupialia: Macropodidae), in the Tanami Desert, Northern Territory. III. Interactions with introduced mammal species. Wildl. Res. 20, 495 511. Macdonald D. W. (1979) Some observations and field experiments on the urine marking behavior of the red fox, Vulpes vulpes. Z. Tierpsychol. 51, 1 22. Macdonald D. W. (1987) Running with the Fox. UnwinHyman, London. McLaren B. E. & Peterson R. O. (1994) Wolves, moose and tree rings on Isle Royale. Science 266, 1555 8. MacNally R. C. (1983) On assessing the significance of interspecific competition to guild structure. Ecology 64, 1646 52. McRae P. D. (2004) Aspects of the ecology of the greater, bilby, Macrotis lagotis in Queensland (MSc Thesis). University of Sydney, Sydney. Marsack P. & Campbell G. (1990) Feeding behaviour and diet of dingoes in the Nullarbor region Western Australia. Aust. Wildl. Res. 17, 349 58. Mech D. L. (1970) The Wolf: The Ecology and Behaviour of an Endangered Species. Natural History Press, New York. Mitchell B. D. & Banks P. B. (2005) Do wild dogs exclude foxes? Evidence for competition from dietary and spatial overlaps. Aust. Ecol. 30, 581 91. Moseby K. (2002) The Arid Recovery Project. Annual Report 2002. Department of Environment and Heritage, Adelaide. Newsome A. (1990) The control of vertebrate pests by vertebrate predators. Trends Ecol. Evol. 5, 187 91. Newsome A. E. (2001) The biology and ecology of the dingo. In: A Symposium on the Dingo (eds C. R. Dickman & D. Lunney) pp. 20 33. Royal Zoological Society of New South Wales, Mosman. Newsome A. E., Catling P. C. & Corbett L. K. (1983) The feeding ecology of the dingo II. Dietary and numerical relationships with fluctuating prey populations in southeastern Australia. Aust. J. Ecol. 8, 345 66. Newsome A. E., Parer I. & Catling P. C. (1989) Prolonged prey suppression by carnivores predatorremoval experiments. Oecologia 78, 458 67. Newsome A. E., Pech R. P., Smyth R., Banks P. B. & Dickman C. R. (1997) Potential Impacts on Australian Native Fauna of Rabbit Calicivirus Disease. Environment Australia, Canberra. Newsome A. E., Catling P. C., Cooke B. D. & Smyth R. (2001) Two ecological universes separated by the dingo fence in semi arid Australia: interactions between landscapes, herbivory and carnivory, with and without dingoes. Rangeland J. 23, 71 98. O Neill A. (2002) Living with the Dingo. Envirobook, Annandale. Oakwood M. (1997) The ecology of the northern quoll (PhD Thesis). Australian National University, Canberra. Oakwood M. (2000) Reproduction and demography of the northern quoll, Dasyurus hallucatus, in the lowland savanna of northern Australia. Aust. J. Zool. 48, 519 39. Pace M. L., Cole J. J., Carpenter S. R. & Kitchell J. F. (1999) Trophic cascades revealed in diverse ecosystems. Trends Ecol. Evol. 14, 483 8. Paine R. T. (1966) Food web complexity and species diversity. Am. Nat. 100, 65 75. Paine R. T. (1969) A note on trophic complexity and community stability. Am. Nat. 103, 91 3. Paine R. T. (1980) Food web linkage, interaction strength and community infrastructure. J. Anim. Ecol. 49, 667 85. Palomares F. & Caro T. M. (1999) Interspecific killing among mammalian carnivores. Am. Nat. 153, 492 508. Paltridge R. (2002) The diets of cats, foxes and dingoes in relation to prey availability in the Tanami Desert, Northern Territory. Wildl. Res. 29, 389 403. Paltridge R., Gibson D. & Edwards G. (1997) Diet of the feral cat (Felis catus) in central Australia. Wildl. Res. 24, 67 76. Pech R. P., Sinclair A. R. E., Newsome A. E. & Catling P. C. (1992) Limits to predator regulation of rabbits in Australia: evidence from predatorremoval experiments. Oecologia 89, 102 12. Petren K. & Case T. J. (1998) Habitat structure determines competition intensity and invasion success in gecko lizards. Proc. Natl. Acad. Sci. USA 95, 11739 44. Pettigrew J. D. (1993) A burst of feral cats in the Diamantina: a lesson for the management of pest species?. In: Cat Management Workshop Proceedings (eds G. Siepen & C. Owens) pp. 25 32. Queensland Department of Environment and Heritage, Brisbane. Polis G. A., Sears A. L. W., Huxel G. R., Strong D. R. & Maron J. (2000) When is a trophic cascade a trophic cascade? Trends Ecol. Evol. 15, 473 5. Pople A. R., Grigg G. C., Cairns S. C., Beard L. A. & Alexander P. (2000) Trends in the numbers of red kangaroos and emus on either side of the South Australian dingo fence: evidence for predator regulation? Wildl. Res. 27, 269 76. Power M. E., Tilman D., Estes J. A. et al. (1996) Challenges in the quest for keystones. Bioscience 46, 609 20. Recher H. F. (2002) What revolution? In: A Zoological Revolution. Using Native Fauna to Assist in Its Own Survival

THE DINGO AS A TROPHIC REGULATOR 501 (eds D. Lunney & C. R. Dickman) pp. 116 29. Royal Zoological Society of NSW, Mosman. Ripple W. J. & Beschta R. L. (2003) Wolf reintroduction, predation risk, and cottonwood recovery in Yellowstone National Park. For. Ecol. Manage. 184, 299 313. Ripple W. J. & Larsen E. J. (2000) Historic aspen recruitment, elk, and wolves in northern Yellowstone National Park, USA. Biol. Conserv. 95, 361 70. Robertshaw J. D. & Harden R. H. (1985) The ecology of the dingo in northeastern New South Wales II. Diet. Aust. Wildl. Res. 12, 39 50. Robertshaw J. D. & Harden R. H. (1986) The ecology of the dingo in northeastern New South Wales IV. Prey selection by dingoes, and its effect on the major prey species, the swamp wallaby, Wallabia bicolor (Desmarest). Aust. Wildl. Res. 13, 141 63. Robley A., Reddiex B., Arthur T., Pech R. P. & Forsyth D. (2004) Interactions Between Feral Cats, Foxes, Native Carnivores, and Rabbits in Australia. Final report for the Australian Government Department of the Environment and Heritage. Department of Sustainability and Environment, Heidelberg. Rolls E. C. (1969) They All Ran Wild. The Story of Pests on the Land in Australia. Angus and Robertson, Sydney. Saunders G., Coman B., Kinnear J. & Braysher M. (1995) Managing Vertebrate Pests: Foxes. Australian Government Publishing Service, Canberra. Schmitz O. J., Beckerman A. P. & O Brien K. M. (1997) Behaviourally mediated trophic cascades: effects of predation risk on food web interactions. Ecology 78, 1388 99. Schmitz O. J., Hambäck P. A. & Beckerman A. P. (2000) Trophic cascades in terrestrial systems: a review of the effects of carnivore removals on plants. Am. Nat. 155, 141 53. Schoener T. W. (1974) Resource partitioning in ecological communities. Science 185, 27 39. Short J. & Smith A. (1994) Mammal decline and recovery in Australia. J. Mammal. 75, 288 97. Short J., Kinnear J. E. & Robley A. (2002) Surplus killing by introduced predators in Australia: evidence for ineffective antipredator adaptations in native prey species? Biol. Conserv. 103, 283 301. Smith A. P. & Quin D. G. (1996) Patterns and causes of extinction and decline in Australian conilurine rodents. Biol. Conserv. 77, 243 67. Smith D. W., Brewster W. G. & Bangs E. E. (1999) Wolves in the greater yellowstone ecosystem: restoration of a top carnivore in a complex management environment. In: Carnivores in Ecosystems: theyellowstone Experience (eds T. W. Clark, A. P. Curlee, S. C. Minta & P. M. Kareiva) pp. 103 25. Yale University Press, New Haven. Soderquist T. R. & Serena M. (1994) Dietary niche of the western quoll. Dasyurus geoffroii, in the jarrah forest of Western Australia. Aust. Mamm. 17, 133 6. Soulé M. E., Bolger D. T., Alberts A. C. et al. (1988) Reconstructed dynamics of rapid extinctions of chaparralrequiring birds in urban habitat islands. Conserv. Biol. 2, 75 92. Soulé M. E., Estes J. A., Berger J. & Martinez Del Rio C. (2003) Ecological effectiveness: conservation goals for interactive species. Conserv. Biol. 17, 1238 50. Soulé M. E., Estes J. A., Miller B. & Honnold D. L. (2005) Strongly interacting species: conservation policy, management and ethics. Bioscience 55, 168 76. Switalski T. A. (2003) Coyote foraging ecology and vigilance in response to gray wolf reintroduction inyellowstone National Park. Can. J. Zool. 81, 985 93. Thompson M. B. (1983) Populations of the Murray River tortoise, Emydura (Chelodina): the effect of egg predation by the red fox, Vulpes vulpes. Aust. Wildl. Res. 10, 363 72. Thomson P. C. (1992a) The behavioural ecology of dingoes in northwestern Australia. II. Activity patterns, breeding season and pup rearing. Wildl. Res. 19, 519 30. Thomson P. C. (1992b) The behavioural ecology of dingoes in northwestern Australia. III. Hunting and feeding behaviour, and diet. Wildl. Res. 19, 531 41. Thomson P. C. (1992c) The behavioural ecology of dingoes in northwestern Australia. IV. Social and spatial organisation, and movements. Wildl. Res. 19, 543 63. Tilman D. (1987) The importance of the mechanisms of interspecific competition. Am. Nat. 129, 769 74. Williams K., Parer I., Coman B., Burley J. & Braysher M. (1995) Managing Vertebrate Pests: Rabbits. Australian Government Publishing Service, Canberra.