Remote Sensing of Oyster Reefs

By: Michael Espriella, M.S candidate, University of Florida

Oyster reefs filter pollutants, serve as habitat for hundreds of species, and control shoreline erosion among numerous other ecosystem services. Unfortunately, these resources are in decline due to various anthropogenic and environmental stressors including over-harvest, disease, and low freshwater flow events. Given the difficulty in accessing these habitats, there is very limited monitoring to assess declines and their causes.

That’s where remote sensing comes in to the conversation. Not only do remote sensing techniques have the benefit of collecting data on areas that are difficult to access, but they also allow information to be collected without potentially harming the reef as can sometimes be the case with more traditional transect sampling (Figure 1).

Figure 1: Transect sampling counting live oysters off the coast of Cedar Key, FL.

Unoccupied aircraft systems (UASs) are one of the most cost-effective techniques to accomplish this task. They can collect data virtually any day of the year, assuming appropriate conditions. Additionally, they can collect data at a much higher spatial resolution than commonly used remote sensing data sources, such as satellite imagery. This high spatial resolution will allow for more detailed analysis on the state of an individual reef.

Figure 2: Example of UAV imagery mosaic from Little Trout Creek, located north of Cedar Key, FL. Imagery courtesy of Dr. Peter Frederick’s lab at the University of Florida.

Our lab’s objective with this project is to use high-resolution imagery to generate mosaics and delineate inter-tidal habitats with a focus on oyster reefs. This will be done along Florida’s Big Bend coastline using a Geographic Object Based-Image Analysis (GEOBIA) technique. Borrowed from terrestrial remote sensing, GEOBIA is particularly useful when processing high-resolution imagery as it allows for the segmentation of pixels into meaningful objects. These objects are then classified using spectral, structural, and topographical characteristics. From there, we can assess the spatial dynamics that contribute to a successful or unsuccessful system.