Long-term relationship between phosphorus inputs and wetland phosphorus concentrations in a northern Everglades marsh.
Assessments of long-term relationships between changes in nutrient inputs and wetland nutrient concentrations can be complicated by fluctuations in other environmental factors as well as by problems typical of long-term monitoring data. Consequently, statistical analysis of these types of data sets requires careful consideration of environmental covariates, potential biases in the monitoring design, and irregularities caused by changes in field sampling protocols. We evaluated the relationship between anthropogenic phosphorus (P) inputs and water-column total P (TP) concentrations in a northern Everglades marsh by statistically analyzing available data collected from several sampling programs over the past 20 years (1978-1997). Canal inputs of agricultural runoff contribute most of the P to the marsh and have produced a zone of enrichment within the marsh during the past few decades. Regression analyses showed that both canal and marsh TP concentrations increased during the 1980s and then decreased in the 1990s. However, the statistical relationship between canal P inputs and marsh TP, while significant, generally was weak except for marsh locations adjacent to the canal. Stronger relationships existed between marsh TP and hydrologic parameters such as marsh water depth, which is controlled by changes in weather patterns and marsh management. In particular, dry conditions during the 1980s may have contributed to observed increases in marsh P concentrations and the movement of a P 'front' further into the marsh. Higher rainfall and water depths and agricultural best management programs initiated during the 1990s have been associated with reduced P concentrations in canal waters entering the marsh. While it is anticipated that this reduction eventually will result in lower marsh TP concentrations, this effect is not yet evident, possibly due to internal loading of P from enriched marsh soils. Our findings illustrate some of the environmental factors that can complicate attempts to develop empirical relationships between P inputs and wetland P concentrations and to use such relationships to forecast changes in marsh concentrations based on past monitoring data alone.