Evolutionary responses to invasion: cane toad sympatric fish show enhanced avoidance learning

Abstract

The introduced cane toad (Bufo marinus) poses a major threat to biodiversity due to its lifelong toxicity. Several terrestrial native Australian vertebrates are adapting to the cane toad's presence and lab trials have demonstrated that repeated exposure to B. marinus can result in learnt avoidance behaviour. Here we investigated whether aversion learning is occurring in aquatic ecosystems by comparing cane toad naïve and sympatric populations of crimson spotted rainbow fish (Melanotaenia duboulayi). The first experiment indicated that fish from the sympatric population had pre-existing aversion to attacking cane toad tadpoles but also showed reduced attacks on native tadpoles. The second experiment revealed that fish from both naïve and sympatric populations learned to avoid cane toad tadpoles following repeated, direct exposure. Allopatric fish also developed a general aversion to tadpoles. The aversion learning abilities of both groups was examined using an experiment involving novel distasteful prey items. While both populations developed a general avoidance of edible pellets in the presence of distasteful pellets, only the sympatric population significantly reduced the number of attacks on the novel distasteful prey item. These results indicate that experience with toxic prey items over multiple generations can enhance avoidance leaning capabilities via natural selection.

Figure 1. The the mean proportion of attacks (±SE) on the vial containing tadpoles versus an empty vial for toad sympatric and allopatric fish populations.

Proportions >0.5 represent a prefernce for the vial containing the tadpoles.

Figure 2. Mean number of attacks (±SE) on cane toad tadpoles over five days of direct exposure by cane toad allopatric and sympatric fish populations.

Figure 3. The mean (±SE) number of attacks on distasteful pellets (A) and edible pellets (B) by rainbowfish from cane toad allopatric and sympatric populations over five days.

Figure 4. The mean (±SE) proportion of rainbowfish eating distasteful pellets (A) and edible pellets (B) from cane toad allopatric and sympatric populations over five days.

Figure 5. The mean proportion (± SE) of pellets eaten during the contol period when both colours were edible and during the test treatment when one colour was edible (EP) and the other was distasteful (DP).