Researchers from the Australian Institute of Marine Science (AIMS) have developed an advanced remote sensing model that will significantly enhance efforts to manage and restore coral reefs.
The , published in Remote Sensing of the Environment, uses widely available data from the Sentinel-2 satellite, validated with low-tech digital photos of reef areas. Compared to current methods this more accurately maps habitat on the ocean floor, identifying the locations where young corals have the highest chance to establish and thrive.
“The model incorporates many of the key habitat attributes that coral need for recruitment and survival,” Dr Ben Radford, lead author of the study, said.
“We can then use the model to help decide which reef areas to protect or actively manage to ensure coral life cycles continue and hence help reefs rebuild following events such as coral bleaching.”
Coral bleaching events are increasing in frequency and severity worldwide, and have caused a global decline in coral cover. Coral restoration provides an immediate and local intervention that may help reefs recover, yet restoration techniques are still largely experimental, logistically challenging and expensive.
Reef restoration has the highest chance of success in locations where a hard substrate lies in a relatively narrow range of depths, does not experience too much wave energy, too much (or too little) heat or light, and is not exposed during low tides.
The type of hard substrate is also critical: one of the key factors that influences the success of coral larvae settling on a reef is the presence of crustose coralline algae (CCA). . Conversely, turf algae competes with coral larvae for recruitment sites, while areas with destabilised coral rubble or sand provide poor habitat for larvae to settle and grow.
Dr Radford said that current models used to map coral reef habitat are not designed to specifically look for optimal places for the recruitment of coral, and this can be an important consideration for designing restoration projects.
“We want to maximise the survivorship of corals because it’s so expensive to undertake restoration efforts in the first place,” Dr Radford said.
The new model uses a rugosity metric – an indicator of the complexity of the sea floor habitat – to accurately predict areas of CCA, turf algae, coral rubble and sand at scales of tens to hundreds of metres.
“Our approach uses high-resolution imagery of how the sea floor varies with depth and combines this with field data to try and predict what habitat types could be there, based on how much hard structure is likely there,” , a co-author of the study, said.
The researchers used ‘drop’ cameras (cameras capturing images while facing the sea floor) to collect detailed information on the height and form of reef features on the Scott Reef complex off the north-west coast of Western Australia.
Dr Puotinen said that a strength of the model was that it was ‘trained’ using field data collected from numerous locations on the reef and then tested against the data collected from other locations, allowing the research team to verify how accurately the model predicted habitat types under different conditions.
The model will also inform wider reef management. Co author said one of the most essential components for reef management is understanding the connectivity within and among reefs: where different coral types are located and are producing larvae, how larvae are exchanged among parts of the reef, and where they might settle and survive.
“You can’t manage corals unless you know where they’re distributed on the reef and the environments that are, or are not, suitable for long-term survival and reproduction,” Dr Gilmour said.
“Habitat models will ultimately help to infer how reefs are going to change in the future with climate change and other disturbances. We can then explore how certain interventions or management activities, like reef restoration, may or may not help their trajectory in the future.”
This research was funded by AIMS and Woodside Energy Ltd as Operator for and on behalf of the Browse Joint Venture (BJV).
The research was a collaboration between AIMS and The University of Western Australia.
