Summary Introduction Methods Geologic Setting Results Rock Analysis Water Chemistry Ground Water Contamination QC/QA Conclusions Future Studies Acknowledgments References Appendices

This study was partially initiated on the premise that a region-wide subaerial unconformity, known as the Q3 unconformity (Perkins, 1977), can prevent or retard vertical movement of fluids. This unconformity had been identified as an effective aquitard beneath the Dade County Landfill, where it retards downward migration of landfill leachates (Shinn and Corcoran, 1988). Because the unconformity has been shown to inhibit downward water movement, it therefore follows that fresh water injected into a saline aquifer beneath the Q3 layer could produce a freshwater "bubble" that would tend to migrate laterally. A laterally migrating lens or "bubble" of fresh water would eventually leak upward through the occasional solution hole or fracture and would enter the water column near coral and other marine communities. That lateral migration is possible was indicated by discovery of freshwater seepage in more than 100 (30 m) of water seaward of the coral reefs off Key Largo (Simmons and Love, 1984). In addition, fresh water was once harvested offshore in Biscayne Bay, where it bubbled up through salt water from the underlying Biscayne aquifer (Kohout 1960). Offshore leakage of fresh water was also shown to influence distribution of bottom biota in Biscayne Bay (Kohout and Kolipinski, 1967). Because of a lowered water table, this process is no longer active, but it is expected to occur during and following large rainfall events.

The Q3 subaerial unconformity occurs throughout the south Florida mainland and often caps a thin freshwater limestone layer. The unconformity and underlying limestone have been called the Ft. Thompson Formation, after the type locality on the Caloosahatchee River (Parker and Cooke, 1944; Parker et al., 1955). Perkins (1977) and Harrison et al. (1984) encountered the Q3 in core holes beneath Key Largo between 25 and 35 ft (7.6-10.7 m) below the surface. The freshwater limestone facies of the Q3 is often absent in the Keys. In unpublished exploratory core holes, R.B. Halley and the senior author encountered the Q3 beneath Florida Bay, Big Pine Key and in areas of the Everglades west of Miami. An isolated topographic high on the bottom of Florida Bay near East Key consists of 43 ft (13 m) of a Key Largo-like facies patch reef underlain by 3 ft (9 m) of freshwater limestone. A typical Q3 soilstone crust caps the freshwater limestone (pers. observation by senior author). An unconformity, probably the Q3, defines the base of the freshwater lens on Big Pine Key (Hanson, 1980; Vacher et. al., 1992).

In the present study, the Q3 was recognized in all cores beneath the Keys and in those drilled within a mile of shore. The Q3 was not consistently recognized in cores taken farther offshore, however. Moreover, changes in lithology at well depths generally associated with the Q3 were common. Because the soilstone crust capping the Q3 unit is the thickest and most widespread of the four that occur in south Florida, we suspect that it did extend out to and beyond the shelf margin in the past. Perkins (1977) interpreted its presence in a 150-ft-deep (45 m) core taken at Little Molasses Island. We therefore believe the unconformity existed but was locally removed by erosion during post-Q3 transgressions and regressions. In some cases it may have been present but not recovered in the core barrel. This is considered unlikely, however.

The absence of a recognizable Q3 unconformity and the realization that older injection wells and septic-tank drain fields do not penetrate the unconformity led us to change monitoring-well strategy during the later phase of drilling.

Field observations, especially the presence of tidal pumping, indicated that the overlying Holocene sediment is a more widespread and effective confining layer than the Q3 unconformity. Further-more, overlying Holocene sediment not only retards upward movement of fluids above the Q3, but also movement of any fluids that may leak upward from below the Q3 unconformity.

Holocene sediments and reefs on the Florida reef tract were deposited on the Pleistocene limestone of the Florida Keys during the last interglacial transgression, i.e., the past ~6,000 to 7,000 years. Enos (1977), Turmel and Swanson (1976), and drilling from this study showed that the first sediments deposited as the sea flooded the shelf were fine-grained lime muds similar to those presently being deposited in the shallows of Florida Bay. Other studies (Lidz et al., in press) show that a linear island of Pleistocene limestone existed along the outer edge of the reef tract even after most of the reef tract was flooded. This island would have served as a barrier similar to modern Key Largo island to produce quiet-water conditions, like those of Florida Bay, on its leeward side, thus allowing the accumulation of fine-grained sediments. Lime muds of Florida Bay and those in Hawk Channel have extremely low permeabilities of less than 10 md (Enos and Sawatsky, 1981). In this study, even when well sites were on lime sand, we invariably encountered lime mud before entering the underlying Pleistocene limestone. The only areas where mud was absent were over topographic highs beneath coral reefs. Most coral reefs on the Florida reef tract have accumulated on Pleistocene topographic highs (Shinn et al., 1977; Shinn et al., 1989). Mud accumulates in topographic lows, not on highs (Davis et al., 1992).

The crucial observation that showed the effectiveness of the seal created by the Holocene lime mud was tidal pumping. Wherever the drill bit penetrated Pleistocene limestone beneath Holocene sediment, water either gushed out or entered the borehole depending upon whether the tide was rising or falling. The effect of tidal pumping on well completion has been described earlier. Therefore, during the drilling of the last offshore transect (OR-2,3,4, and 5) off north Key Largo, the well screens were placed only 5 ft below the top of the Pleistocene (i.e., 1.5 m below the overlying Holocene lime-mud layer). At the well site closest to shore in this transect (OR-1 A and B), where there was no Holocene sediment, two wells were drilled, one below the Q3 unconformity and one to a depth of 10 ft (3 m). The OR transect is situated offshore from a community (Ocean Reef Club) with a large disposal well field consisting of 50 wells serving a community of ~3,500 people. These disposal wells, installed before current concerns about nutrients and other contaminants, are only 30 ft (9 m) deep and therefore do not penetrate the Q3 unconformity. Furthermore, they are cased to 9 ft (2.7 m), allowing fluids to enter the receiving limestone beginning at 9 ft below the surface. The best place to sample for effluents from these wells is therefore above the Q3 unconformity and below the Holocene lime-mud layer.