Episodic correlations in behavioural lateralisation differ between a poison frog and its mimic

We tested three primary hypotheses: 1) behaviours will be lateralised, either at the individual or population level, 2) behaviours will show complementary laterality such that different tasks will be allocated to different sides of the animal, and 3) the scale of lateralisation will differ between social and non-social behaviours, such that social behaviours will be lateralised at the population level while non-social behaviours will be lateralised only at the individual level, allowing for social behaviours to be coordinated while non-social behaviours remain unpredictable. We also conducted a further exploratory analysis to determine the presence of lateralisation-based within-individual correlations. Experiments were conducted on sympatric populations of the Ecuador poison frog, Ameerega bilinguis, and its Batesian mimic the sanguine dart frog, Allobates zaparo. Individuals repeatedly experienced four types of trials where they were exposed to a conspecific (Am. bilinguis/Al. zaparo), a heterospecific (Am. bilinguis/Al. zaparo), prey (live termites) or a simulated predator (the human observer). All trials were then analysed for lateralised eye use inspecting the stimulus (in the conspecific, heterospecific and prey trials) or for escape direction (in the simulated predator trials).