Social structures in amphibians

Little is known about population-wide social structure in amphibians, which are generally considered as the least social vertebrates and assumed to lack the cognitive capabilities for higher sociality as a result of their small brains. Nevertheless amphibians show a wide array of social behaviours and especially poison frogs dispose of most, if not all, behavioural prerequisites that accompany longer lasting social structures. In the current project I want to investigate the prevalence and consequences of social structures in a captive and a semi-natural population of the territorial poison frog Allobates femoralis. I will monitor both populations with microphone arrays to continuously record calling behaviour and interactions, and conduct extensive behavioural observations in the field. The field study will employ a state-of-the-art embedded, wireless sensor network with sound source localization and acoustic caller identification. Using social network analysis I will then explore and quantify structures in the communication and social network of these populations. For the natural population I will finally correlate individual network metrics with reproductive success to investigate the influence of an individuals position in its social network on its reproductive fitness. The proposed project is exceptional in taking an integrative approach to investigate social structure and its consequences in an amphibian using state-of-the-art technology and social network analysis in both a fully controlled laboratory population and a semi-natural field population. The proposed project shall also pioneer the further development of sensor networks for studies on sociality and promote their use in behavioural research.

In this research project, we could show, for one more time, that frogs are not the "mere reflex machines" as they are still often portrayed in the wider public, but also amongst some researchers focusing on other taxa. We used the South American Poison Frog Allobates femoralis as a model species to investigate anuran social behaviour. In our experiments frogs showed to utilize diverse environmental and social cues to make graded, dynamic, and flexible decisions in their social behaviour, namely territoriality and parental care. In playback trials, simulating calling territorial intruders, the frogs not only used simple but unreliable cues like signal strength, but also complex, more reliable characteristics, such as the reverberation signature, to assess the distance of potential aggressors. However, in the context of stable territorial systems, we could show that our frogs do not use the individually distinct voices of their neighbours to differentiate between known and unknown intruders - the so called "dear-enemy effect", which has been found in some other frog species. Poison frogs, unlike the native frogs of Austria, do not lay their eggs in water but on land, in the forest leaf litter. Then they have to transport the tadpole to small water bodies, where they finish their development. In A. femoralis, the male, who has to leave his territory and faces considerable threats, for the tadpoles, but also for himself, usually performs this task. We investigated, if the males spend time during tadpole transport or subsequent homing to explore the area for new waterbodies, which they could use to deposit their tadpoles, or if they prefer to take direct paths, and to utilize known waterbodies to minimize the exposure and risks when transporting. Indeed, we could find no evidence for any search behaviour related to tadpole transport, as the frogs were moving in almost straight lines to deposition sites and back to their territory. They predominantly visited deposition sites, which they likely had used previously, even when we had removed those water bodies experimentally, which demonstrates the importance of spatial memory for these frogs. However, by now we still do not know how they had discovered those sites in the first place. Regarding selection of suitable waterbodies for their tadpoles, we could demonstrate that transporting frogs make their decisions based on various factors. They not only take into account the location of their own natal pool, but also evaluate the distance between their current territory and other pools, as well as the presence and absence of predators in those pools.