Cane toads (Bufo marinus) have invaded large areas of Australia, killing many native predators as they have done so. The metamorph stage of the life cycle -the first terrestrial phase, immediately after transformation from the tadpole -is critical for ecological impact (because these animals are small enough to be prey for many native predators) and for potential control of toad populations (because small body size renders metamorphs vulnerable to desiccation).To quantify the spatial and temporal distribution of metamorph toads, and the biotic and abiotic factors that might affect their distributions, we surveyed toad breeding sites in Australia's wet-dry tropics (Adelaide River floodplain, NT) in both the wet season and the dry season. Metamorphs were concentrated close to the water's edge during the dry season, especially at midday when desiccation rates were highest. During the wet season, metamorphs were widely dispersed through the landscape. Our surveys indicate that abiotic factors (risk of desiccation) are most favourable for metamorph toads close to the pond edge, but biotic factors (food supply, and risk of competition and cannibalism) are most favourable away from the water. Operative temperatures were spatially homogeneous and sublethal, and so are unlikely to influence metamorph distribution. Desiccation risk fluctuated on a diel cycle as well as seasonally. We predict that metamorph toads benefit from dispersing as soon as desiccation risk allows them to do so, and hence the distribution of metamorph toads will shift dynamically in response to weather-mediated changes in rates of evaporative water loss.

Interactions between invasive species and native fauna afford a unique opportunity to examine interspecific encounters as they first occur, without the complications introduced by coevolution. In northern Australia, the continuing invasion of the highly toxic cane toad Bufo marinus poses a threat to many frog-eating predators. Can predators learn to distinguish the novel toxic prey item from native prey (and thus, avoid being poisoned), or are longer-term genetically based changes to attack behaviour needed before predators can coexist with toads? To predict the short-term impact of cane toads on native predators, we need to know the proportion of individuals that will attack toads, the proportion surviving the encounter, and whether surviving predators learn to avoid toads. We quantified these traits in a dasyurid (common planigale, Planigale maculata) that inhabits tropical floodplains across northern Australia. Although 90% of naïve planigales attacked cane toads, 83% of these animals survived because they either rejected the toad unharmed, or killed and consumed the prey snoutfirst (thereby avoiding the toxin-laden parotoid glands). Most planigales showed one-trial learning and subsequently refused to attack cane toads for long time periods (up to 28 days). Toad-exposed planigales also avoided native frogs for up to 9 days, thereby providing an immediate benefit to native anurans. However, the predators gradually learnt to use chemical cues to discriminate between frogs and toads. Collectively, our results suggest that generalist predators can learn to distinguish and avoid novel toxic prey very rapidly -and hence, that small dasyurid predators can rapidly adapt to the cane toad invasion. Indeed, it may be feasible to teach especially vulnerable predators to avoid cane toads before the toads invade, by deploying low-toxicity baits that stimulate taste-aversion learning.

No abstract

A mechanistic understanding of factors influencing the dispersal behavior of metamorph cane toads (Bufo marinus) has direct conservation relevance in Australia. These invasive anurans are toxic to native predators, and if we can predict their distribution across the landscape, we can also predict (and perhaps, manage) the scale of their impact. We propose that the major drivers of metamorph distribution are the risk of dehydration (restricting the young toads to moist substrates near pond margins) and biotic advantages to dispersal away from the pond (especially, less risk of cannibalism). To test this model, we investigated the influence of abiotic and biotic cues on the behavior of individual toads in the laboratory. Substrate moisture levels strongly influenced metamorph activity levels and habitat selection: dry substrates induced most metamorphs to remain near water. The only biotic cue to influence metamorph dispersal was proximity of a larger (cannibalistic) conspecific; a cannibal's presence at the pond margin caused most metamorphs to spend less time there, and as a consequence, to dehydrate more rapidly. Our results suggest that the spatial and temporal distribution of metamorph cane toads reflects a trade-off between competing risks: the danger of desiccation tends to keep young toads close to the pond margin in dry conditions, whereas the danger of cannibalism stimulates dispersal.

Abstract:Immediately after their transition from aquatic to terrestrial life, juveniles of many anuran species are restricted to the margins of natal ponds. Understanding the factors determining the duration of that pondside aggregation has direct management ramifications in the case of the invasive cane toad (Bufo marinus) in tropical Australia. Previous work suggests that dispersal confers biotic advantages (reduced risk of cannibalism, enhanced feeding opportunities) to juvenile toads, but desiccation risk constrains these small animals to the moist margins of the pond. If so, juvenile dispersal should be sensitive to fluctuating hydric conditions on a diel and seasonal cycle. We tested this prediction with field observations (monitoring of dispersal to and from the pond) and field experiments (manipulating hydric regimes). Our results support a dynamic model of juvenile distribution, with a primary role for temporal variations in desiccation risk as the primary factor driving dispersal. During the dry season, strong diel cycles in desiccation risk generate a ‘tidal’ flow of juveniles, dispersing out in the moist morning but retreating to the pond margins at midday. Dispersal rates were enhanced by artificial watering during the dry season, and by the onset of the wet season.