Stable between‐group differences in collective behavior have been documented in a variety of social taxa. Here we evaluate the effects of such variation, often termed collective or colony‐level personality, on coral recovery in a tropical marine farmerfish system. Groups of the farmerfish Stegastes nigricans cultivate and defend gardens of palatable algae on coral reefs in the Indo‐Pacific. These gardens can promote the recruitment, growth, and survival of corals by providing a refuge from coral predation. Here we experimentally evaluate whether the collective response of farmerfish colonies is correlated across intruder feeding guilds – herbivores, corallivores and egg‐eating predators. Further, we evaluate if overall colony responsiveness or situation‐specific responsiveness (i.e. towards herbivores, corallivores, or egg‐eaters in particular) best predicts the growth of outplanted corals. Finally, we experimentally manipulated communities within S. nigricans gardens, adding either macroalgae or large colonies of coral, to assess if farmerfish behavior changes in response to the communities they occupy. Between‐group differences in collective responsiveness were repeatable across intruder guilds. Despite this consistency, responsiveness towards corallivores (porcupinefish and ornate butterflyfish) was a better predictor of outplanted coral growth than responsiveness towards herbivores or egg‐eaters. Adding large corals to farmerfish gardens increased farmerfish attacks towards intruders, pointing to possible positive feedback loops between their aggression towards intruders and the presence of corals whose growth they facilitate. These data provide evidence that among‐group behavioral variation could strongly influence the ecological properties of whole communities.

A suite of processes drive variation in coral populations in space and time, yet our understanding of how variation in coral density affects coral performance is limited. Theory predicts that reductions in density can send coral populations into a predator pit, where concentrated corallivory maintains corals at low densities. In reality, how variation in coral density alters corallivory rates is poorly resolved. Here, we experimentally quantified the effects of corallivory and coral density on growth and survival of small colonies of the staghorn coral Acropora pulchra. Our findings suggest that coral density and corallivory have strong but independent effects on coral performance. In the presence of corallivores, corals suffered high but density-independent mortality. When corallivores were excluded, however, vertical extension rates of colonies increased with increasing densities. While we found no evidence for a predator pit, our results suggest that spatio-temporal variation in corallivore and coral densities can fundamentally alter population dynamics via strong effects on juvenile corals.

Surveying coastal systems to estimate distribution and abundance of fish and benthic organisms is labor‐intensive, often resulting in spatially limited data that are difficult to scale up to an entire reef or island. We developed a method that leverages the automation of a machine learning platform, CoralNet, to efficiently and cost‐effectively allow a single observer to simultaneously generate georeferenced data on abundances of fish and benthic taxa over large areas in shallow coastal environments. Briefly, a researcher conducts a fish survey while snorkeling on the surface and towing a float equipped with a handheld GPS and a downward‐facing GoPro, passively taking ~ 10 photographs per meter of benthos. Photographs and surveys are later georeferenced and photographs are automatically annotated by CoralNet. We found that this method provides similar biomass and density values for common fishes as traditional scuba‐based fish counts on fixed transects, with the advantage of covering a larger area. Our CoralNet validation determined that while photographs automatically annotated by CoralNet are less accurate than photographs annotated by humans at the level of a single image, the automated approach provides comparable or better estimations of the percent cover of the benthic substrates at the level of a minute of survey (~ 50 m2 of reef) due to the volume of photographs that can be automatically annotated, providing greater spatial coverage of the site. This method can be used in a variety of shallow systems and is particularly advantageous when spatially explicit data or surveys of large spatial extents are necessary.