Summary
As anthropogenic activities continue to impact ecosystems around the world, the ability to monitor and manage ecosystem health and function increases in importance. Despite this need for monitoring, management projects are often limited by the funding, time and personnel required to gather field data. These costs only increase when considering large dynamic ecosystems, such as pulsed wetlands, that have spatially and temporally variable habitat conditions, thus requiring frequent field measurements across the landscape.
The ability to use sensory data can drastically reduce the resources required to monitor field conditions, but can limit the scope and resolution at which data can be measured. Using the Florida Everglades as a model system because of its extensive monitoring and instrumentation network, we compared the effectiveness of large‐scale sensory models of hydrological data with models that incorporate additional fine‐scale habitat variables to account for variations in the dry season densities of small fishes and macroinvertebrates.
Since the magnitude of aquatic faunal densities during the dry season greatly affect higher trophic groups, such as wading birds and alligators, the ability to understand the drivers on fish and crayfish densities is of high importance for Everglades restoration and management plans.
We found that models using only sensory hydrological variables supported the same hypotheses as models that incorporated additional habitat variables, but the degree of fit was reduced considerably and varied depending on species group. Species more heavily dependent on water for reproduction and survival, such as fish, tended to have better fitting models than species less dependent on water, such as crayfish. Additionally, sensory‐only models were able to detect different responses between size classes of fish to the processes that drive their concentration into drying pools of water, with large fish (>2.5 cm) densities influenced more by landscape structure and small fish (<2.5 cm) densities influenced more by hydroperiod.
Our results support the use of sensory data to identify important variables affecting populations for management and restoration efforts. However, the degree to which sensory variables are able to fit species data is dependent upon the ability of sensors to measure major ecosystem drivers and the scale at which sensors can measure environmental change. Although tradeoffs exist between coarse sensory data and finer site‐specific measurements, the benefits of sensory data will increase as the scale of ecosystem management and restoration increases.