Benthic Habitat Composition | Methods
Quantitative assessments of the benthos will be made using a modification of the photoquadrat method. At each site we will build a large composite photomosaic, providing the imagery with which to estimate benthic community composition. For each photomosaic, a 100 m2 area of reef will be photographed and all of the organisms living in this area will be mapped. All organisms will be identified to the finest level of resolution possible (genus level for hard and soft corals, macroalgae and macroinvertebrates and functional group for algal turfs and crustose coralline algae). These data will provide a snapshot view of the composition of the coral reef community, providing a novel, spatial view of the landscape that is analogous to the visual information that satellites have been providing for the study of terrestrial ecology for decades. These photomosaics will serve as a large-scale, photographic archive and rich data source for exploring coral reefs in critical and novel ways for decades to come.
Our large-scale benthic imagery is created by capturing 2000-3000 images over the 100m2 area, which can be used to create 2-dimensional orthophotos and orthoprojections, or 3-dimensional point clouds.
An orthophoto is created by identifying and matching features within the raw images and essentially stitching together or blending pixels. While the orthophoto provides a nice visualization of the reef area, it often has stitching errors associated with creating a 2D image of a complex, 3-dimensional reef structure. Orthophotos are therefore recommended for visualization, rather than scientific purposes.
Our compromise is the creation of an orthoprojection. Like an orthophoto, an orthoprojection is a 2D representation of the reef area. Orthoprojections differ from orthophotos as they are created by taking an orthographic projection of the 3D point cloud based on a defined plane. This reduces the stitching errors associated with orthophotos and eliminates the need for custom-built or expensive software required to make annotations for analytical purposes.
The most accurate representation of the reef area is a 3D reconstruction. Advances in Structure from Motion (SfM) technology allows for the creation of a 3D model of a reef area using 2D images with a high degree of overlap (~80%) by identifying point matches between images in 3D space. The result is a 3D point cloud representing the reef area. While the creation of these 3D models is readily available through commercial software, the ability to visualize and manipulate these models, as well as create annotations for scientific purposes can be extremely time intensive and require customized software. See an example of the 3D model below: