Phosphorus load reductions under best management practices for sugarcane cropping systems in the Everglades Agricultural Area

Abstract

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.