Poison frog larvae are interesting because, in most cases, their survival depends on their transport from a terrestrial location to a body of water by one of their parents. Such bodies of water will be the environment in which tadpoles will develop until metamorphosis and, thus, their selection is nothing trivial. The larvae of many poison frog species feed on detritus, but in some species tadpoles are voracious carnivores that can even kill and consume conspecifics (i.e. they are cannibalistic; see their powerful serrated mouthparts in the photo below!).

Cannibalism is common among animals because eating conspecifics confers nutritional benefits that can make the difference between survival and death. However, cannibals should recognise their kin; if not, they decrease the amount of family genes passed onto the next generation, and are more likely to acquire infectious diseases, among other things. We study the cannibalistic behaviour between tadpoles of the Amazonian dyeing poison frog (Dendrobates tinctorius), whose fathers devotedly care for, and which develop in unstable environments that can suddenly dry out. This project aims to establish if cannibalism confers benefits that make tadpoles more likely to survive until metamorphosis, and have larger size and better performance when they start terrestrial life; but also to determine if cannibals pay the cost of acquiring infectious diseases more easily. Our study will provide new insights into the spread mechanisms of a disease that is largely responsible for amphibian declines worldwide, and into the potential role of parental decision-making in the regulation of this counterintuitive behaviour.