Sampling and Statistical Design of the
Florida Keys Coral Reef Monitoring Project
The Coral Reef Monitoring Project (CRMP) was established by the United States Environmental Protection Agency (US EPA) to address the question, "Are the coral reefs of the Florida Keys changing?" The group faced the task of creating a project that would give precise estimates of change over the large geographical region of the Florida Keys. The sampling protocol is designed to detect change in species richness, condition, and percent cover of particular benthic organisms in the coral reef system. It is a stratified random nested sampling design (Figure 1) with annual sampling of permanently marked stations at sites selected using US EPA EMAP protocol
Figure 1. Levels of stratified random nested sampling design used in CRMP.
Water Quality Segments 5 designated by NOAA/US EPA
Based on previous research and/or experimental sampling of the reefs in the Florida Keys, the following project design criteria needed to be met:
The nested sampling scheme surveys the community for species richness (biodiversity), coral condition (disease and bleaching), and percent cover (abundance). The sampling unit is a 2x22 meter quadrat, termed a station. There are four stations at each site. Two divers survey each station for species richness and coral condition. Video sampling of three 0.5 meter wide transects, running the length of the quadrat, provides estimates of coverage for coral, soft corals, sponges and other functional groups. Detection of change is done by statistically comparing data collected at each station over time, using a paired comparison or repeated measures experimental design.
Estimation Of Percent Cover
An estimate of percent cover (projected to the surface) of corals (stony and soft), algae, sponges, substrate, and other functional groups is provided using random point counting. Ten random points are projected on every non-overlapping, framegrabbed image of each transect. The user identifies each point and enters the data via mouse or keyboard stroke. The random points for each image of a station (3 transects) are generated at the time of framegrabbing and stored in a file along with the images. This ensures that each person examining a particular image will view the same points allowing double blind counting for quality assurance and control purposes as required by the US EPA.
PointCount’99 generates a data file in ASCII format, which can be imported, into a spreadsheet or database of the users choosing. Percent cover is calculated as the proportion of points of a particular category over the sum of all points for each transect. The mean of three transects per station provides an estimate of coral cover for the station. The CRMP uses a combination of Microsoft Excel and Access for data compilation and then a series of different programs for statistical analysis (i.e. SPSS and Statistica by StatSoft).