Stormwater run-off from the 290,000 ha Everglades Agricultural Area (EAA) is directed into South Florida's Everglades wetland ecosystem. Concerns regarding run-off water quality and environmental impact led to a 1992 regulatory program which requires P levels in basin run-off be reduced by at least 25% relative to historic trends. Farmers must collectively achieve this annual basin-level target by implementing best management practices (BMPs) to reduce P levels in farm drainage waters. At the time, proposed BMP strategies were largely untested, and to what extent they might reduce farm-level P discharge trends (also poorly documented) was unknown. Given these uncertainties, objectives of this study were to: (1) document long-term drainage P trends for EAA sugarcane systems and (2) quantify BMP effects on-farm drainage P loading. In late-1992, discharge pumps at five farm sites (cropped to sugarcane, sugarcane-vegetables, and/or sugarcane-rice) were instrumented to collect water samples for P analysis during all drainage events throughout baseline (BL pre-BMP) and BMP operations. Highly variable rainfall distributions in the region strongly influence farm drainage requirements, thus, meaningful interpretations of water quality trends require hydrologic adjustment to P load data. Five rainfall-adjustment analyses were applied to the 6-year farm-level databases. Two analysis methods compared P load trends for the entire BL and BMP monitoring periods. In Method 1, unit area P load (UAL) to rainfall ratios (UAL:R) during BMP operations were 20.4-47.3% smaller across all five sites than those recorded during BL. In Method 2, slope coefficients describing cumulative UAL versus cumulative rainfall trends during BMPs were 14.9-25.0% smaller than BL slopes. The remaining three methods assessed data trends across five consecutive "water years" (WY). In Method 3, slope coefficients describing WY96-98 cumulative UAL versus rainfall distributions were 32.8% lower in magnitude relative to WY94. In Method 4, average UAL:R for the WY96-98 period were 31.0% smaller than for WY94. Basin-level P loads are calculated every WY by state water management regulators, using a hydrologic adjustment model calibrated to a historic load and rainfall database. During the first 3 years (WY96-98) of required BMP implementation, the basin recorded a 55% P load reduction, When this model was applied to the farm data (Method 5), farm P load reductions for WY96-98 averaged 59.7%. All five analytical methods confirm favorable P-reduction trends under recommended BMP strategies for EAA sugarcane-based cropping systems. (C) 2002 Elsevier Science B.V. All rights reserved.
