The ecological integrity of Florida’s Everglades is driven by water flows, depths, spatial distribution, and quality. We developed the Everglades Forecasting (EVER4CAST) application that simulates future water levels to help determine optimal outcomes for a suite of ecological models. EVER4CAST creates near-term (up to 6-months) forecasted water levels by repeatedly selecting the closest one-month historical analog to real-time conditions while adjusting for forecasted precipitation (i.e. central tendency). A routine then creates Monte Carlo simulations of near-term forecasted water levels by using historic variation around the central tendency. Simulations are categorized into 4 bins based on high/low water and high/low variability.
For 2017, a central tendency forecast and 100 simulations were generated and run through a set of Everglades species distribution models: wading birds (Great Egret, White Ibis, and Wood Stork), Burmese Python, Cape Sable Seaside Sparrow (CSSS), and aquatic fauna. For each species, simulations were then ranked and scored (1-100) according to the predicted species responses. The Burmese Python model was ranked inversely to represent the negative effect of this invasive species on native Everglades wildlife. The optimal simulation was determined by assigning equal weighting to each species; however, weights can be modified according to desired management objectives.
Results/Conclusions
A simulation belonging to the low water/low variability category was identified to have the highest overall ranking of 80.33 with high predicted suitability for CSSS (score 100), Great Egrets (score 95), White Ibises (score 89), Wood Storks (score 97), and low suitability for Burmese Pythons (score 88) and aquatic fauna (score 13). This example illustrates the positive effect that low water levels with few abrupt changes (i.e., low variability) has on 1) the CSSS by providing continuously dry conditions, 2) wading birds by concentrating aquatic fauna in shallow pools, however, limiting aquatic prey production for subsequent years, and 3) limiting suitable habitat for python movement. This approach provides water managers with daily, spatially-explicit water level forecasts to develop regional water management targets to maximize landscape-scale benefits and the flexibility to define objectives by weighting species importance.