Abstract

The hydrology of the north-central Everglades was altered substantially in the past century by canal dredging, land subsidence, ground-water pumping, and levee construction. Vast areas of seasonal and perennial wetlands were converted to uses for agriculture, light industry, and suburban development. As the catchment area for the Everglades decreased, so did the sources of water from local precipitation and runoff from surrounding uplands. Partly in response to those alterations, water-resources managers compartmentalized the remaining wetlands in the north-central Everglades into large retention basins, called Water Conservation Areas (WCAs). In spite of efforts to improve how water resources are managed, the result has been frequent periods of excessive drying out or flooding of the WCAs because the managed system does not have the same water-storage capacity as the pre-drainage Everglades. Linked to the hydrological modifications are ecological changes including large-scale invasions of cattail, loss of tree islands, and diminishing bird populations in the Everglades. Complex interactions among numerous physical, chemical, and biological factors are responsible for the long-term degradation of the ecological character of the Everglades.

Ground-water discharge and recharge estimates are plotted on top of an aerial photo of the Everglades Nutrient Removal (ENR) project. The view is to the south and shows various plant communities, open water, and remnant canals from the time when this area of the Everglades was farmed. Beyond the canals on the far right is land that still is being farmed. [larger version]

Over the past 15 years, a new set of smaller wetland basins, called Stormwater Treatment Areas (STAs), have been designed and constructed by water-resources engineers on the former wetlands adjacent to WCAs. The purpose of STAs is to remove excess nutrients from agricultural drainage water prior to its input to WCAs. STAs tend to be about one-tenth the size of a WCA, and they are located on former wetlands on the northwestern side of WCAs on sites that were managed as farmland for much of the twentieth century in an area referred to as the Everglades Agricultural Area, or EAA.

The objective of the present investigation was to quantify interactions between surface water and ground water in the Everglades Nutrient Removal Project (ENR), a prototype project for the STAs that began operation in 1994. Determining the effect of ground water on the mercury balance of the ENR treatment wetland was an important additional objective. In order to broaden the relevance of conclusions to all parts of the north-central Everglades, interactions between surface water and ground water and mercury also were investigated in Water Conservation Area 2A (WCA-2A) and, to a lesser extent, in two other WCA basins, WCA-2B and WCA-3A.

Figure 1. (Left) Gray area on map is shown enlarged below. (Below) Central Everglades and adjoining areas, south Florida, showing locations of Water Conservation Areas (WCA), Everglades Nutrient Removal (ENR) project, and Stormwater Treatment Areas (STA). Southwestern coastal areas of Everglades in Monroe County are not shown. [larger image]

An important conclusion of this study is that creation of the WCA basins, and accompanying water-resources management, have appreciably increased both recharge and discharge in the north-central Everglades compared with pre-drainage conditions. Recharge and discharge are highest near the northern and northwestern edges of the Everglades, in the relatively small basins such as ENR and the STAs that share borders with both WCA-1 and the EAA. All basins experienced greater increases in recharge relative to discharge, because of the effects that land subsidence and ground-water pumping outside the Everglades had on hydraulic gradients. The highest basin-wide estimate of recharge was measured in ENR, where recharge averaged 0.9 centimeter per day (cm/d) over a 4-year study period. For perspective, that estimate of recharge is the equivalent of 30 percent of pumped surface-water inflows and 230 percent of average daily precipitation in ENR. Ground-water discharge was 10 times smaller than recharge at ENR. The present study estimated a basin-averaged recharge for WCA-2A (0.2 cm/d) that was a factor of 4 smaller than ENR. Although preliminary, that estimate of recharge is 5 times higher than previous estimates (approximately 0.04 cm/d), probably because the newer measurements were able to quantify recharge and discharge at finer spatial and temporal scales. Recharge at WCA-2A is smaller than ENR because WCA-2A has a smaller topographic gradient (3 x 10^-5 and 2 x 10^-4 in WCA-2A and ENR, respectively), as well as a smaller ratio of perimeter length to total wetland surface area (6 x 10^-5 and 4 x 10^-4 in WCA-2A and ENR, respectively), which decreases the importance of processes outside the wetlands such as land subsidence or ground-water pumping. At the present time, recharge and discharge are thought to be higher in the WCAs compared to the pre-drainage Everglades (perhaps by a factor of 4 or 5), although that comparison is uncertain because of the difficulty of estimating pre-drainage hydrologic fluxes. The reason that recharge and discharge are thought to be higher now compared to pre-drainage conditions is that water-resources management has increased fluctuations in surface-water levels. The present study showed that the magnitude of recharge and discharge, as well as temporary reversals between recharge and discharge, are related to increased surface-water fluctuations caused by large water releases from WCA-1 into WCA-2A.

The most important geologic factor affecting interactions between surface water and ground water in the north-central Everglades is the hydraulic conductivity (K) of the Surficial aquifer. Estimates of K in the top 40 feet (ft) of the aquifer at both ENR and WCA-2A are higher (by more than an order of magnitude) than previously published estimates of K for the northern Everglades (typically, reported as 5 ft/day). Finding higher than anticipated hydraulic conductivities in the upper sand and limestone layers of the Surficial aquifer has important implications. In particular, it was found that the upper sand and limestone layers with high permeability are the main parts of the aquifer with appreciable freshwater. Sampling of major-ion chemistry in ground water showed that freshwater was usually located only at shallow depths, approximately the top 40 ft of the 200-ft deep Surficial aquifer. Hydraulic and chemical results, therefore, indicate that in many areas of the north-central Everglades, interactions between surface water and ground water primarily involve the top layers (layers 2 and 3) of the Surficial aquifer, causing appreciable recharge and discharge to a depth of approximately 40 ft.

Geochemical measurements provided further information about the source of the thin layer of fresh ground water beneath the north-central Everglades. Water-stable isotopic ratios of hydrogen and oxygen showed that the source of fresh ground water was recharge of Everglades surface waters, specifically, recharge of surface waters from wetland sloughs that had been present long enough in the surface flow system to be substantially evaporated. An exception was the portion of the aquifer beneath the interior of ENR, where stable isotopes indicated that recharge occurred quickly and without appreciable evaporation. This result, along with the distinct ionic signature of the water, is consistent with an interpretation that the source of recharging water beneath ENR was precipitation onto ENR during the time period it was managed for agricultural purposes. The "light" stable isotopic composition of that water indicates that precipitation infiltrated quickly through the unsaturated zone without appreciable evaporation. Another exception is apparent in ground water near levees. The amount and type of salts in ground water in the vicinity of levees indicate that ground-water seepage beneath the levee causes deep mixing in the Surficial aquifer that results in upward movement of relict seawater from the bottom two-thirds of the aquifer (below 60 ft) to shallow ground water and to wetland surface water.

Part of the motivation for the present study was a concern that arose among the water-resources managers that designed the ENR treatment wetland. The concern was whether mercury methylation, and, thus, mercury bioavailability, might increase when agricultural soils were re-flooded and managed once again as wetlands. The present study complements the work of many other mercury investigators in the Everglades by specifically addressing the effect of interactions between surface water and ground water on mercury cycling.

Total dissolved mercury (Hg_T) was detectable in all monitoring wells in ENR and WCA-2A at an average concentration of 0.7 nanogram per liter (ng/L), which is slightly below the average concentration in surface water (1 ng/L). An important exception was shallow wells on the western side of ENR, where the average concentration was 40 percent higher (1.4 ng/L) than surface water. Higher concentrations of Hg_T in ground water on the western side of ENR was the result of recharge from ENR surface water combined with release of Hg_T from solid phases in peat to recharging water. Dissolved methylmercury (MeHg) in ground water was undetectable in all deep wells (greater than 40 ft deep) and most shallow wells (less than 0.02 ng/L compared with 0.1 ng/L in ENR surface water). Shallow wells beneath the interior of ENR were the exception, with detectable MeHg concentrations as high as 0.2 ng/L. Wells with detectable MeHg are of interest because they are the same wells classified by water-stable isotopes and major ion chemistry as "agricultural recharge water." In general, Hg_T and MeHg concentrations were not positively correlated with sulfate concentrations at either ENR or WCA-2A.

A budget was developed for ground-water fluxes of mercury at ENR, which made possible a comparison with the surface and atmospheric components of the mercury budget for ENR developed by other researchers. Recharge of Hg_T from surface water to ground water was a major pathway for transport of total dissolved mercury but not MeHg. Recharge of Hg_T accounted for a loss from ENR surface water equivalent to 10 percent of the total inputs of Hg_T to surface water. In comparison, recharge of MeHg was not detectable and accounted, therefore, for none of the losses of MeHg from surface water in ENR.

Chemical data and water-stable isotopic ratios indicate that most surface water recharged in ENR is discharged to a seepage canal on the western and northern side of ENR. Transport of recharged water through the Surficial aquifer to the seepage canal appears to take place in a matter of weeks to months, with only relatively minor mixing with deeper ground water. Measurements of Hg_T in the seepage canal suggested that Hg_T had not yet discharged to the canal at the end of the 4-year study period. Because the flow path between points of recharge in ENR and discharge in the seepage canal was short, it was concluded that mercury was retained or delayed in its transport through the aquifer by interaction with aquifer sand or limestone or fine organic sediments at the base of the seepage canal.

Acknowledgments

The authors are indebted especially to the following individuals whose help was critical to project success: Gene Shinn, Chris Reich, Don Hickey, David Krabbenhoft, Bill Orem, Jonah Jackson, Eric Nemeth, Jessica Thomas, Brent Banks and Aaron Higer of the USGS; and Larry Fink, Carl Miles, Pete Rawlik, Darren Rumbold, Wossenu Abtew, Kevin Rohrer, Jim Vincent, Doug Ellington, Pete Dauenhauer, Bruce Webb, Stan Jones, Paul McGinnes, Mark Hummel, Ben Harkenson, Sharon Niemczyk, Anne Shoffner, Jennifer Cornwell, Emily Hopkins, Heidi Bazell, Angela Chong and John Lukasiewicz of SFWMD.

Groundwater-Surface Water Interactions and Relation to Water Quality in the Everglades