Techniques for assessing BMP effectiveness in the absence of historical baseline data

Phosphorus (P) levels in Everglades Agricultural Area (EAA) surface runoff must be reduced by at least 25%. To achieve this basin regulatory target, EAA growers must implement effective farm drainage P-reduction best management practices (BMPs). This study was conducted to address concerns regarding the absence of farm-level P drainage databases and uncertainty surrounding methods to assess P-reductions under untested BMP designs. Ten farms representative of EAA production systems were instrumented to monitor unit area P load (UAL) discharges from August 1992 through April 1996. Phosphorus discharge trends during baseline and subsequent BMP operations were compared using three different analytical methods. Each method employs some measure of hydrologic adjustment to UAL data in order to address the strikingly different rainfall distributions (which directly influence farm drainage requirements) that occurred during baseline and BMP monitoring periods. The first method involves the calculation of total UAL to rainfall (UAL:R) ratios for each period and these "normalized" UALs are then compared. The second method involves the re-expression of daily rainfall and UAL data into separate baseline and BMP cumulative databases. Plotting cumulative UAL against cumulative rainfall provides daily resolution of rainfall effects on drainage discharge and allows for cumulative distribution slope comparisons between monitoring periods. The third method assesses farm discharge trends over time using a model developed by state regulators to annually evaluate basin P-reduction compliance. The model calculates a 12-month rainfall-adjusted UAL (AUAL) based on the comparison of current load (and rainfall distribution) to an earlier 10-year period of record database. With method 1, BMP UAL:R were 3 to 33% lower than for baseline, reflecting UAL reductions under modified water management (BMPs) responses to rainfall. With method 2, BMP cumulative distribution slope values for six farms were 6 to 35% smaller than baseline, reflecting long-term UAL reductions under BMPs. Omitting disproportionate basin-wide flooding effects during a single storm event, eight farms recorded reductions. For these eight farms, AUALs calculated with method 3 declined by 15 to 83% over a three-year period. Across all three analyses, two farms suffered increasing UAL trends after major crop rotation changes without concurrent hydraulic BMP implementation. Despite high rainfall variability and limited baseline data, these three analysis methods verify fami-level P reductions under BMPs for sugarcane, vegetable, rice, and mixed-crop systems.