Water masses move over reefs, seagrass beds, and sandbanks – and as they do, the seawater chemistry changes.
In the Florida Keys, changes in coral reef carbonate chemistry are driven by benthic metabolism, the origin of the water mass, and the connectivity of habitats. A new study from NOAA’s Atlantic Oceanographic and Meteorological Laboratory (AOML) shows how we can use existing monitoring data to better understand the combined influence of these factors on local reef water chemistry.
Dr. Heidi Hirsh, an Assistant Scientist with the AOML Coral Program, demonstrates how integrating the source water, or “endmember”, chemistry conditions, the benthic habitat, and the flow of water between habitats can be used to predict the nearshore carbonate chemistry on a specific coral reef.
Seawater carbonate chemistry is complex and highly dependent on biological processes. Seagrass beds, algae, and coral reefs take up dissolved inorganic carbon (DIC) through photosynthesis, raising the pH of surrounding waters and reducing acidity. Conversely, respiration adds DIC to the water column, lowering the pH and making the water more acidic.
Calcification – the process through which corals and other organisms build calcium carbonate skeletons – removes carbonate ions from seawater, potentially leading to more acidic conditions. In contrast, the dissolution of their skeletons releases carbonate back into the system, which may help buffer acidity.
The carbonate chemistry on a coral reef also depends on the origin and history of the water mass as it has traveled to a reef. Waters arriving from offshore, nutrient-rich bays, or semi-enclosed basins will have unique chemical signatures.
These complex changes in carbonate chemistry are constantly happening but understanding how they vary across space and time within the Florida Keys may allow scientists at AOML to predict the impacts of increasingly acidic waters on the integrity and structure of vital coral reefs.
Benthic communities (i.e. seagrass, coral), source water (“endmember”) chemistry and the complex flow of water (hydrodynamics) between habitats all influence the local carbonate chemistry of a coral reef. Derived from: Hirsh, et al., 2025
As part of the four-year Florida Regional Ecosystems Stressors Collaborative Assessment (FRESCA), a collaboration co-led by NOAA’s Atlantic and Meteorological Laboratory (AOML) and the University of Miami, Hirsh has developed a statistical model to predict nearshore coral reef carbonate chemistry based on modeled trajectories of currents and the interconnection between relevant sourcewater and habitats.
This approach takes into account where the water came from and the influence of marine ecosystems (i.e. benthic community metabolism) on a water mass before it arrives on a reef in a specific area.
Training a model to predict chemical changes across miles of a barrier reef requires measurements of carbonate chemistry at each reef station paired with data to describe local environmental parameters like temperature, salinity, light availability, and nutrients.
Enter the South Florida Ecosystem Restoration Program (SFER).
The Florida Coral Reef stretches 350 miles from the Dry Tortugas and Florida Keys up to St. Lucie, Florida. The points on this graph represent marked sampling stations and coordinates for the SFER Cruises within the Florida Keys National Marine Sanctuary (FKNMS). Derived from: Palacio, et al., 2023
Scientists with AOML’s Ecosystem Assessment team have led a series of SFER cruises from Biscayne Bay to the Florida Keys to the West Florida Shelf – as far north as Tampa Bay – sampling and monitoring changes in seawater chemistry and other variables for the last 27 years.
By returning to designated cruise paths (i.e. transects) with marked coordinates every other month, the team has measured changes in carbonate chemistry across the Florida Keys over seasons, years and decades – a collaboration that made this modeling possible.
“We are incredibly fortunate to have such a long-standing dataset from the SFER cruises. It allows us to ask new questions and build on decades of research in the Florida Keys.” explains Hirsh, Ph.D.
To build the model, Hirsh used the discrete sampling points along ten of these cross-shelf transects in the Florida Keys, covering an estimated 250 kilometers (155 miles) of the Florida Coral Reef, and the carbonate chemistry measurements collected between 2015 and 2021 at each point.
The relevant starting chemistry (i.e. “endmember”) of the water sampled at these points and the habitats the water was exposed to along the path to the sampling station also shape the observed reef chemistry.
To effectively capture these changes, Hirsh needed to recreate the pathways of waterflow to each reef station. Collaborators from Université Catholique de Louvain (UC Louvain) – Dr. Emmanuel Hanert and Dr. Thomas Dobbelaere -simulated water particle trajectories using a hydrodynamic model: the second-generation Louvain-la-Neuve Ice-ocean Model (SLIM).
These backward simulations (ranging from 1 to 14 days) described the possible trajectories traveled by the water before it was sampled on the SFER cruises.
All pathways for a given station were summarized into a polygon area, or “flowshed,” describing the probable upstream history of the water before it arrived at each sampling station.
Each polygon, or “flowshed”, represents the upstream area associated with each inshore sampling station. The flowsheds (with increasing size) represent the 1-day, 7-day, and 14-day water mass histories (probable area that water traveled before being sampled 1, 7, or 14-days later). Left: July 2019 and right: September 2019. Derived from: Hirsh, et al., 2025
With benthic maps incorporated to account for the varying habitat to which the water was exposed, the model could effectively identify how the spatial variation in complex habitats influenced the chemistry of each water mass before arriving at the given sampling point on the reef.
“This spatial approach to understanding carbonate chemistry on coral reefs demonstrates how we can leverage existing datasets and models to make high-resolution predictions for reefs of interest. By utilizing data already at our disposal, we can be more strategic about designing future sampling strategies to fill the knowledge gaps and increase the utility of new data, ” says Hirsh, Ph.D.
A simulation
Increasing the complexity of these statistical models fuels what the model is able to tell us, effectively increasing scientists’ ability in predicting local carbonate chemistry on the reefs of the Florida Keys. Integrating more data only enhances this.
By developing and validating the model to capture these complex processes now, scientists at AOML aim to apply it to future predictions of how exacerbated ocean acidification could impact carbonate chemistry across the only barrier reef in the continental United States and third largest in the world.
Globally, ocean acidification erodes away at the complex calcium carbonate skeleton that composes coral reefs, which provide millions in storm surge protection in Florida alone.
It also decreases coral’s ability to grow and add on to its existing skeleton by reducing the concentration of the carbonate (CO₃²⁻) ions in the water column. But these impacts are not consistent globally – they vary across time and space.
A recent decade-long study led by Ana Palacio, Ph.D., a CIMAS Assistant Scientist with AOML’s Coral Program, demonstrated that inshore reefs of the Upper Keys may act as potential refugia against ocean acidification – likely due to the proximity of dense seagrass beds – while the impacts of acidification in the Lower Keys may be exacerbated.
“One takeaway of our new modeling study is the importance of identifying the appropriate upstream endmember so that we understand the magnitude of change that is taking place across these reefs,” explains Hirsh, Ph.D.
Ultimately, this study builds on the goal of the larger project – FRESCA – to investigate the spatial variability of environmental stressors across South Florida’s ecosystems – and how the combined impacts of these stressors on key ecosystems will be exacerbated or mitigated over time.