In 2011, I participated in a workshop during the International Marine Conservation Congress in Victoria, Canada, where experts came together to discuss which were the most critical factors that affect the ability of coral reefs to resist and recover from climate disturbance. After much deliberation, our expert consensus produced a list with 11 top resilience indicators, based on their perceived importance, evidence base in the literature, and feasibility for managers to measure. The top indicators included:
– Presence of resistant coral species
– Historical temperature variability
– Nutrient pollution;
– Coral diversity;
– Herbivore biomass;
– Physical human impacts;
– Coral disease;
– Coral recruitment; and
– Fishing pressure
We published a paper describing the selection of these indicators and how they might be used to prioritize sites for reef management in a changing climate.
However, during this process, I got to thinking. Site based measures of resilience indicators are nice, but they only allow us to say that one survey point is potentially better than any other survey point and leaves you with no information about all of the unsurveyed reef.
Wouldn’t it be even better, then, if we could combine the field data with satellite imagery to assess what are the main factors influencing the resilience indicators that we can derive from space? Then we could use those relationships to predictively map the resilience indicators over a broader spatial scale, such as across an entire fisheries management area (qoliqoli). This process would yield maps with relative values for every portion of the qoliqoli to enable us to make better decisions about how to plan marine protected areas (MPAs) and MPA networks.
We previously took this approach to predictively map characteristics of reef fish assemblages across the Kubulau qoliqoli in Vanua Levu, Fiji, and published the results in a paper in Remote Sensing. I went back to my co-authors from Simon Fraser University and the University of Queensland and asked, “Do you think we can take this same approach for mapping indicators of reef resilience?”
Our findings have just been published in a more recent issue of Remote Sensing, accessible here. We specifically focused on mapping resilience indicators for which data products were not already available. These included:
– Stress-tolerant coral taxa, as a measure of resistance to coral bleaching;
– Coral diversity, as a measure of resistance to bleaching and potentially past recovery potential;
– Herbivorous fish biomass, as a measure of recovery potential due to the ability of herbivorous fishes to remove macroalgae from the reef, thus preventing harmful coral-algae interactions and allowing space for new corals to settle;
– Herbivorous fish functional group richness, as a measure of recovery potential as fish remove algae in different ways (e.g. excavating, scraping, browsing) and it is necessary to have the full complement of types of herbivores to effectively remove most of the macroalgae;
– Juvenile corals, as a measure of recovery potential from recent coral recruitment and survival; and
– Cover of live coral and crustose coralline algae, as a measure of recovery potential both as a indicator of current coral-algal dynamics when coupled with the amount of macroalage, as well as a proxy for the amount of substrate available for coral settlement and source of new recruits.
We specifically used high resolution satellite data (<4 m pixels) to enable production of maps of resilience indicators at a scale meaningful for customary management systems in Fiji and the rest of the Western Pacific. We did not try to predictively map indicators which other people have successfully mapped (e.g. historical temperature variability, nutrients, sedimentation, physical impacts and macroalgae), but noted that our predictive maps could be combined with these data products within a spatial planning framework.
How well did we do? We were able to reasonably map relative differences in potential susceptibility to coral bleaching based on the composition of coral communities observed in the field. We were able to do this because some corals are more tolerant to environmental stress than others and they tend to be found in different environments and micro-habitats.
We also did a good job predicting distributions in the number of functional groups of herbivorous fish. This indicator likely also has strong environmental determinants as large excavators, such as bumphead parrotfish, steephead parrotfish and bicolor parrotfish, tend to be associated with forereef slopes and reef crests. We noted, though, that the total amount of “reef cleaning” performed by each group of herbivorous fish will be influenced by how many fish are present and their size. In addition, not all herbivorous fishes are created equal. A new, separate study from Fiji found that only 4 species of fish were responsible for eating 97% of the algae set out in a feeding experiment.
We did not do such a good job in predicting distributions of juvenile corals, but this was expected given that there are many factors that influence where corals may settle and how many survive. We also did not do very well in predicting coral diversity patterns. This was somewhat surprising given how much is already known about relationships between coral diversity and depth, exposure to waves and reef habitat. However, our results were likely influenced by errors in georeferencing and incomplete sampling across all of the habitat types in Kubulau qoliqoli. These issues could easily be improved upon in future studies.
What does this mean for reef managers? In the context of MPA planning, managers now have the potential to set targets for reef resilience indicators, in addition to habitat and feature representation, when designing MPA networks within decision support software (e.g. Marxan). This represents a considerable improvement over the current practice of designing or adapting MPAs based on site-based resilience score collected from relatively few sites across the planning region.