No abstract

High sea surface temperatures caused global coral bleaching during [2015][2016]. During this thermal stress event, we quantified within-and among-species variability in bleaching severity for critical habitat-forming Acropora corals. The objective of this study was to understand the drivers of spatial and species-specific variation in the bleaching susceptibility of these corals, and to evaluate whether bleaching susceptibility under extreme thermal stress was consistent with that observed during less severe bleaching events. We surveyed and mapped Acropora corals at 10 sites (N = 596) around the Lizard Island group on the northern Great Barrier Reef. For each colony, bleaching severity was quantified using a new image analysis technique, and we assessed whether small-scale environmental variables (depth, microhabitat, competition intensity) and species traits (colony morphology, colony size, known symbiont clade association) explained variation in bleaching. Results showed that during severe thermal stress, bleaching of branching corals was linked to microhabitat features, and was more severe at reef edge compared with lagoonal sites. Bleaching severity worsened over a very short time-frame (∼1 week), but did not differ systematically with water depth, competition intensity, or colony size. At our study location, within-and among-species variation in bleaching severity was relatively low compared to the level of variation reported in the literature. More broadly, our results indicate that variability in bleaching susceptibility during extreme thermal stress is not consistent with that observed during previous bleaching events that have ranged in severity among globally dispersed sites, with fewer species escaping bleaching during severe thermal stress. In addition, shaded microhabitats can provide a refuge from bleaching which provides further evidence of the importance of topographic complexity for maintaining the biodiversity and ecosystem functioning of coral reefs.

The effect of a pollutant on the base of the food web can have knock-on effects for trophic structure and ecosystem functioning. In this study we assess the effect of microplastic exposure on juveniles of a planktivorous fish (Acanthochromis polyacanthus), a species that is widespread and abundant on Indo-Pacific coral reefs. Under five different plastic concentration treatments, with plastics the same size as the natural food particles (mean 2mm diameter), there was no significant effect of plastic exposure on fish growth, body condition or behaviour. The amount of plastics found in the gastro-intestinal (GI) tract was low, with a range of one to eight particles remaining in the gut of individual fish at the end of a 6-week plastic-exposure period, suggesting that these fish are able to detect and avoid ingesting microplastics in this size range. However, in a second experiment the number of plastics in the GI tract vastly increased when plastic particle size was reduced to approximately one quarter the size of the food particles, with a maximum of 2102 small (< 300μm diameter) particles present in the gut of individual fish after a 1-week plastic exposure period. Under conditions where food was replaced by plastic, there was a negative effect on the growth and body condition of the fish. These results suggest plastics could become more of a problem as they break up into smaller size classes, and that environmental changes that lead to a decrease in plankton concentrations combined with microplastic presence is likely have a greater influence on fish populations than microplastic presence alone.

The accumulation of floating anthropogenic debris in marine and coastal areas has environmental, economic, aesthetic, and human health impacts. Until now, modelling the transport of such debris has largely been restricted to the large-scales of open seas. We used oceanographic modelling to identify potential sites of debris accumulation along a rugged coastline with headlands, islands, rocky coasts and beaches. Our study site was the Great Barrier Reef World Heritage Area that has an emerging problem with debris accumulation. We found that the classical techniques of modelling the transport of floating debris models are only moderately successful due to a number of unknowns or assumptions, such as the value of the wind drift coefficient, the variability of the oceanic forcing and of the wind, the resuspension of some floating debris by waves, and the poorly known relative contribution of floating debris from urban rivers and commercial and recreational shipping. Nevertheless the model was successful in reproducing a number of observations such as the existence of hot spots of accumulation. The orientation of beaches to the prevailing wind direction affected the accumulation rate of debris. The wind drift coefficient and the exact timing of the release of the debris at sea affected little the movement of debris originating from rivers but it affected measurably that of debris originating from ships. It was thus possible to produce local hotspot maps for floating debris, especially those originating from rivers. Such modelling can be used to inform local management decisions, and it also identifies likely priority research areas to more reliably predict the trajectory and landing points of floating debris.

The ability of individuals to actively control their movements, especially during the early life stages, can significantly influence the distribution of their population. Most marine turtle species develop oceanic foraging habitats during different life stages. However, flatback turtles (Natator depressus) are endemic to Australia and are the only marine turtle species with an exclusive neritic development. To explain the lack of oceanic dispersal of this species, we predicted the dispersal of post-hatchlings in the Great Barrier Reef (GBR), Australia, using oceanographic advection-dispersal models. We included directional swimming in our models and calibrated them against the observed distribution of post-hatchling and adult turtles. We simulated the dispersal of green and loggerhead turtles since they also breed in the same region. Our study suggests that the neritic distribution of flatback post-hatchlings is favoured by the inshore distribution of nesting beaches, the local water circulation and directional swimming during their early dispersal. This combination of factors is important because, under the conditions tested, if flatback post-hatchlings were entirely passively transported, they would be advected into oceanic habitats after 40 days. Our results reinforce the importance of oceanography and directional swimming in the early life stages and their influence on the distribution of a marine turtle species.

Predicting the impact of climate change on species is often done using species distribution models, but these can be problematic in topographically diverse environments. For species relying on particular moisture gradients, such as Australian rainforest frogs, accurate predictions of moisture availability are crucial. We found that while temperature gradients can be more accurately modeled with highresolution digital elevation models, moisture availability can be inaccurately represented by climate layers. Standard distribution models are also limited in their ability to account for other factors influencing habitat suitability, such as competitor species or disease. Expert knowledge can be useful for bridging these gaps.

The need to proactively manage landscapes and species to aid their adaptation to climate change is widely acknowledged. Current approaches to prioritizing investment in species conservation generally rely on correlative models, which predict the likely fate of species under different climate change scenarios. Yet, while model statistics can be improved by refining modeling techniques, gaps remain in understanding the relationship between model performance and ecological reality. To investigate this, we compared standard correlative species distribution models to highly accurate, fine‐scale, distribution models. We critically assessed the ecological realism of each species’ model, using expert knowledge of the geography and habitat in the study area and the biology of the study species. Using interactive software and an iterative vetting with experts, we identified seven general principles that explain why the distribution modeling under‐ or overestimated habitat suitability, under both current and predicted future climates. Importantly, we found that, while temperature estimates can be dramatically improved through better climate downscaling, many models still inaccurately reflected moisture availability. Furthermore, the correlative models did not account for biotic factors, such as disease or competitor species, and were unable to account for the likely presence of micro refugia. Under‐performing current models resulted in widely divergent future projections of species’ distributions. Expert vetting identified regions that were likely to contain micro refugia, even where the fine‐scale future projections of species distributions predicted population losses. Based on the results, we identify four priority conservation actions required for more effective climate change adaptation responses. This approach to improving the ecological realism of correlative models to understand climate change impacts on species can be applied broadly to improve the evidence base underpinning management responses.

No abstract