Geomorphological and environmental contexts of deep-water
coral mound diversity in the Strait of Florida
Thiago B.S. Correa
A suite of sensors from an Autonomous Underwater Vehicle (AUV) provided us
with a high-resolution dataset for five deep-water sites in the Strait of Florida. By
comparing backscatter intensity with substrate cover (based on submersible track video
groundtruthing), I determined that higher backscatter intensity corresponds with harder
substrates. This relationship, in conjunction with other techniques that are currently being
explored, will allow us to classify the abundance and distribution of epibenthic fauna on
deep-water corals in the Strait of Florida.
Although deep-water corals has been studied in the Florida-Bahamas-Hatteras
region for more than half of a century, the techniques applied (echo sounders, dredges,
sub-bottom profilers, and submersibles) were not capable of producing fine-scale spatial
maps and therefore, the distribution and morphology of coral mounds has remained
largely undocumented. In December of 2005, an AUV was deployed for the first time to
study deep-water coral mounds in the Strait of Florida, covering a total area of 130 km 2.
The AUV sensors provide 1-3m high-resolution topographic and backscatter-strength
maps, side-scan sonar imagery, sub-bottom profiles and physical parameter
measurements at five sites. In 2006, a submersible was used to ground-truth these sites,
recording video transects and sampling the ocean bottom. Facies classification maps were
created in order to estimate total benthic coverage, including classification of the
abundance and distribution of epibenthic fauna on and between individual mounds, as
well as the substrate types that these animals are associated with. Facies classifications
consist of backscatter maps groundtruthed by video transects in an ArcGIS project.
Following this, acoustic texture analysis of Side Scan Sonar (SSS) imageries will be used
to distinguish the relative abundance and distribution of live and dead corals.
Preliminary results from the facies classification maps indicate that ridges covered
by live and dead corals are represented by high intensity backscatter whereas muddy
bottom between ridges and large barren sand dune fields are characterized by low
intensity backscatter. This analysis allowed us to extrapolate coral mound coverage for
other areas that were not groundtruthed. However, backscatter strengths do not
distinguish between standing dead and live coral or rocky seabed. In order to solve this
problem, acoustic texture analyses of Side Scan Sonar (SSS) imageries are being used to
distinguish the relative abundance and distribution of live and dead corals.
These distinctions are necessary in order to understand depositional processes on
mounds, including mapping areas of likely net mound accumulation (dominated by live
coral) and likely erosion (dominated by dead coral). By examining the distribution of
epibenthic fauna and substrate types, and combining these results with other ongoing
studies of antecedent topography and current regime at our sites, we can understand the
optimal conditions for mound formation and development.