The Florida Everglades has lost most of its natural flow due to drainage from agricultural and urban uses. This change is detrimental to ecosystem structure and function because hydrology maintains the ridge and slough topographic relief that contribute to structuring its food webs. The Everglades is moving towards a more homogeneous flat state with the increase of sawgrass-dominated habitats. Sloughs are being lost, but they are vital areas for connections between aquatic habitat patches, nutrient transport, and dry season concentration of prey creating high-quality food patches for apex predators. To assist with the monitoring efforts of the Comprehensive Everglades Restoration Project (CERP), I am conducting an analysis of wetland habitat patterns to improve calculations of fish biomass (m2) at spatial different scales. I am using data from a 14-year monitoring program that has sampled fish and macroinvertebrates from 147 sites selected by a generalized random-tessellation stratified (GRTS) design to provide a spatially balanced representation of the Everglades ecosystem. I am developing vegetation classification maps for each of the primary sampling units (PSU) to define sampleable habitats that can be used in conjunction with fish density data to calculate fish biomass at any scale within the study area.
Results/Conclusions
Using a supervised classification algorithm, we successfully categorized PSU’s into vegetation habitats. Based on our accuracy matrix, the average map had 93.3% accurate classification. After categorizing the sites into separate habitat classes, we were then able to reclassify the classes into sampleable and non-sampleable habitats. This allowed us to isolate the area of sampleable habit that will be used with the Horvitz-Thompson estimator to take fish biomass sampled at the 1m2 scale and interpolate the data to the whole sample unit. This work creates a framework to estimate the effects of the CERP monitoring program and allows us to compare these effects over the 14 years it has been actively monitored for fish and aquatic macroinvertebrates. These results also help to establish an understanding of the relationship between surrounding habitats and local fish communities, and to asses fish assemblage status at multiple scales.