Predation risk is a major ecological factor selecting for group living. It is largely ignored, however, as an evolutionary driver of social complexity and cooperative breeding, which is attributed mainly to a combination of habitat saturation and enhanced relatedness levels. Social cichlids neither suffer from habitat saturation, nor are their groups composed primarily of relatives. This demands alternative ecological explanations for the evolution of advanced social organization. To address this question, we compared the ecology of eight populations of Neolamprologus pulcher, a cichlid fish arguably representing the pinnacle of social evolution in poikilothermic vertebrates. Results show that variation in social organization and behavior of these fish is primarily explained by predation risk and related ecological factors. Remarkably, ecology affects group structure more strongly than group size, with predation inversely affecting small and large group members. High predation and shelter limitation leads to groups containing few small but many large members, which is an effect enhanced at low population densities. Apparently, enhanced safety from predators by cooperative defense and shelter construction are the primary benefits of sociality. This finding suggests that predation risk can be fundamental for the transition toward complex social organization, which is generally undervalued. social evolution | social complexity | predation risk | ecological constraints | cooperative breeding

Summary1. Many anti-predator benefits of group living are predicted to scale with prey density. Nevertheless, evidence for a general density-dependent increase of prey survival is scarce. A possible reason for this discrepancy is the reduction of costly anti-predator behaviour of prey with increasing density, which may offset density-dependent survival gains. Benefits of group living might hence accrue by saved investment into anti-predator behaviours rather than by increased survival rates. 2. Here, we experimentally presented predators in a colony of the cooperatively breeding cichlid fish Neolamprologus pulcher to study density dependence of their anti-predator defences. Predation is a driver of the formation and stability of breeding groups in this species, but potential benefits of coloniality are yet unclear. We hypothesised that increased density of breeding groups would either increase total anti-predator behaviour or allow individuals to reduce their anti-predator effort due to enhanced predator deterrence from neighbours. 3. Confirming predictions from the second hypothesis, our results show that focal groups invested less into anti-predator behaviour at higher densities, while neighbouring groups' behaviour compensated for this reduced effort. This resulted in stable levels of anti-predator behaviours over the entire range of natural densities. Thus, aggregating in colonies allows these fish to save investment in anti-predator behaviour. 4. These results suggest that the formation of both breeding groups and colonies reflects adaptive responses to high predation pressure in this species. Two different levels of sociality seem to be favoured by the same selective force. 5. Our study provides experimental evidence in nature for an important benefit of coloniality that may explain the concomitant existence of different levels of sociality in many highly social taxa.

No abstract

Cooperative breeders serve as a model to study the evolution of cooperation, where costs and benefits of helping are typically scrutinized at the level of group membership. However, cooperation is often observed in multi-level social organizations involving interactions among individuals at various levels. Here, we argue that a full understanding of the adaptive value of cooperation and the evolution of complex social organization requires identifying the effect of different levels of social organization on direct and indirect fitness components. Our long-term field data show that in the cooperatively breeding, colonial cichlid fish Neolamprologus pulcher, both large group size and high colony density significantly raised group persistence. Neither group size nor density affected survival at the individual level, but they had interactive effects on reproductive output; large group size raised productivity when local population density was low, whereas in contrast, small groups were more productive at high densities. Fitness estimates of individually marked fish revealed indirect fitness benefits associated with staying in large groups. Inclusive fitness, however, was not significantly affected by group size, because the direct fitness component was not increased in larger groups. Together, our findings highlight that the reproductive output of groups may be affected in opposite directions by different levels of sociality, and that complex forms of sociality and costly cooperation may evolve in the absence of large indirect fitness benefits and the influence of kin selection.

Tracking wild animals over long periods of time is a non-trivial challenge. This has caused a bias in the availability of individualbased long-term datasets with the majority including birds and mammals. Visual Implant Elastomer (VIE) tags are now a widely used technique that may facilitate the collection of such data for fish and amphibians. However, VIE tags might have important drawbacks. Overall, four potential issues with VIE tags have been proposed: tag loss or misidentification, limited number of individual identifiers, enhanced mortality risk, and effects on intra-specific interactions. Here, we present three experiments in which we investigated these potential problems with VIE tagging in small freshwater fish both in the laboratory and in the wild, using the cooperatively breeding Lake Tanganyika cichlid Neolamprologus pulcher. We find VIE tags to be generally suitable for work with these fish as they did not impair survival, were recognisable up to 2 years after injection, and did not generally disturb group formation. Nevertheless, we identify specific issues of VIE tagging, including colour-and position-dependent variation in tag identification rates, and indications that specific colours may influence social behaviour. Our results demonstrate the suitability of VIE tags for long-term studies on small freshwater fish, while also highlighting the need of validating this method carefully for any species and study. Significance statement Information on the survival, dispersal, and reproductive success of wild individuals across their lifespan is among the most valuable data in Behavioural Ecology. Because tracking of free-ranging individuals over extended periods of time is challenging, there exists a bias in the taxonomic distribution of such long-term datasets. Here, we investigate the suitability of visible implant elastomers (VIE) as a tracking technique to allow for the collection of such data also in small tropical freshwater fish. We show that VIE tags neither alter social behaviour in our study species, nor do they reduce survival, but they enable the tracking of wild individuals across years. We also identify colours and tag positions that are less suitable. We conclude that VIE tags can help produce long-term datasets also for small fish, provided certain precautions are met.