The Black Point well (BkP) farther offshore consistently maintained greater salinity than surface water during the study. The higher salinity may indicate that the depth and distance of this well is beyond the influence of the Biscayne Aquifer. The BkP well and the other wells contained only marine groundwater during the course of the study. Although consistently marine, the Petrel Point well was 1 to 2 ppt less saline than other bay or offshore wells. The lowered salinity may be the result of mixing seawater with the brackish lens beneath Elliott Key and subsequent eastward flow due to tidal pumping, similar to that described at Key Largo (Reich and others, 2002).

One of the factors controlling groundwater flow to the bay is the geologic framework of the region. Knowledge about variability through the Biscayne Aquifer was accomplished by drilling that produced rock cores and allowed observations to be made on the geologic materials that compose the shallow subsurface of BNP. Lithologic core logs are shown in Appendix C. The cores, together with the well-known geology of the mainland (Fish and Stewart, 1991) and previous studies of the shelf geology (Perkins, 1977; Shinn and others, 1989; Lidz and others, 1997), provide the basis for a schematic cross section illustrating the various rock types and sediments beneath the seafloor (Figure 6). The cores show that along the transect from NW to SE, Biscayne Bay is underlain by the uppermost marine stratigraphic units (Q3 - Q5; Quaternary units described by Perkins, 1977) of the Miami Limestone. These units are separated by exposure horizons, surfaces that were weathered during low stands of sea level during the midto- late Pleistocene. In this part of the bay, the limestone is typically overlain by less than 6 in. of modern carbonate sediment (Wanless, 1967). A facies change occurs at Elliott Key to more reefal limestone as the Miami Limestone grades laterally into the Key Largo Limestone. The Billy's Point core did not encounter reefal limestone, which indicates the transition is laterally abrupt here, perhaps only a few tens of meters from this well to the Key Largo Limestone exposed on Elliott Key. The Key Largo Limestone is veneered with modern sediments east of Elliott Key and is increasingly buried by modern sediment east of Hawk Channel. Assuming this area of the reef tract is similar to the shelf margin off central Key Largo (Lidz and others, 1997) modern sediment in the vicinity of Alina's Reef may be 12-18 ft thick and 20-30 ft thick at Pacific Reef.

Taken together with the strontium-isotope analyses (Figure 5), the salinity of groundwater wells in BNP indicates that there may be very limited flow from the Biscayne Aquifer along the extreme western shore of Biscayne Bay near Black Point. There is no evidence from the ^87/86Sr measurements that the Floridan Aquifer is significantly contributing water to BNP. The ratio of Sr isotopes with atomic weight 87 to 86 (^87/86Sr) has been steadily increasing in seawater for the past 40 million years (Howarth and McArthur, 1997), during the time when the carbonate rocks of the Floridan and Biscayne Aquifers were being deposited. Carbonate aquifers, in turn, often transfer their strontium isotopic values to pore water, because there is much more strontium in the rock matrix than in the pore fluid. Strontium isotope values from the Floridan Aquifer are distinctly less (older) than those of modern seawater (Schmerge, 2001). The Biscayne Aquifer rocks are so recent in origin (geologically speaking) that they may appear only slightly older than modern seawater. Mixing of a few percent Floridan Aquifer with surface water would be evident because isotopic compositions are markedly different. Porter and Porter (2002) suggested that a conductivity record from Alina's Reef was evidence of polluted groundwater beneath the reefs. We did not observe low- salinity water at the reef or abnormally elevated chemical constituents that might indicate a source of land-based pollution. It is possible that some other processes were affecting the conductivity reported by Porter and Porter (C.D. Langevin and J. Wang, personal communication, 2004).

The concentrations of nutrients found in marine groundwater are not excessive. Similar concentrations are found to the south off the Florida Keys (Shinn and others, 1994) and beneath Florida Bay (Reich and Shinn, 2003). High concentrations of nutrients in brackish water near shore appear to be more closely related to surface water than to groundwater flow. Although the Biscayne Aquifer samples from the onshore well are elevated in nutrients, there are insufficient concentrations in groundwater beneath the bay to implicate a significant contribution from onshore groundwater. For nitrate and nitrite, the surface water at BPI and BkP is consistently enriched relative to groundwater. This observation, together with surface-water analyses conducted by Meeder and Boyer (2001) and Brand (2002), indicates that nitrate and nitrite levels in the near shore wells are more the result of local denitrification than direct flow from Biscayne Aquifer water. Nutrients determined in these wells appear to be within the range of groundwater values reported by D'Elia and others (1981) for groundwater influx to the reefs in Discovery Bay, Jamaica.

The greater concentrations of silica in near shore surface water may be an indication of interaction of groundwater with quartz sand encountered at BPI or co- mixing of groundwater and surface water. The relatively high concentrations of these nutrients found at near shore sites may, in part, reflect the groundwater contribution to the bay along its western margin. The concentrations indicate that near shore ammonium (NH₄ ⁺) may be primarily associated with runoff to the bay or with decaying organic matter.

Biscayne Bay sediments are known to contain elevated levels of some heavy metals, primarily north of BNP (Hoare, 2002). In particular, lead, silver, copper, zinc, and mercury have been identified as contaminants in some sediment samples (Corcoran 1984; Corcoran and others, 1984; Hoare, 2002). Shinn and Corcoran (1987), however, did not find significant concentrations of heavy metals in groundwater from onshore wells near Goulds Canal. Results from this study did not find excessive concentrations of these metals in bay surface water. The common heavy elements are enriched in groundwater because they have a source in the surrounding rocks and sediments and they become more soluble in lower pH (reduced) groundwater. These elements include aluminum, barium, copper, iron, lead, and zinc. There are no standards for most metals determined during this study, particularly for seawater. Although heavy metals are often enriched and more soluble in reduced groundwater, their surface-water concentrations do not appear to be excessive when compared to oceanic waters (Millero, 1996). In coastal waters, metal concentrations can be considerably greater than in the open ocean but are much less than those acceptable for drinking water. For example, copper, lead, and zinc guidelines for drinking water are 1000, 15, and 5000 ppb. In Biscayne Bay surface water, these metals are about 150, 2, and 10 ppb, respectively.

Shinn and Corcoran (1987) found traces of pesticides, plasticizers, and aliphatic hydrocarbons in samples from shallow wells (15 and 30 ft) in the Biscayne Aquifer south of the Goulds Canal. Concentrations at 30 ft were about half that of the 15-ft sample. The contaminants were not found in a well on the north side of Goulds Canal, nearer the landfill. This distribution indicates that contamination may be local, on the south side of the canal, and may be the result of surface water entering the Upper Biscayne Aquifer. During this study, wastewater compounds in groundwater (G-3613, BPI-1A, MB-1B, and AR-1B) were encountered in 3.5% of the samples and in 5.2% of the surface-water samples (BPI and Gulf Stream). Twenty-two different compounds were recognized in samples and field blanks. Nineteen of the 22 compounds were detected below the method-detection limit (MDL), indicating that while present, they are not of sufficient concentration to be measured accurately by the methods used in this study. Eight of the compounds occurred in blanks, six of those occurred below MDL. Twelve of the compounds were single occurrences. The most commonly recognized compound was DEET, occurring in nine surface-water samples, 16 groundwater samples, and four blanks. Only DEET, acetophenone, and total para-nonylphenol were encountered above the MDL. DEET and acetophenone are components of personal-care products, and total para-nonylphenol is used in detergents. Although QA/QC procedures were carefully followed, the unusual field conditions during sample and blank collections may have resulted in contamination. It is also possible that because of the extremely low detection limits for these compounds, generally in the range of 0.5 - 1 ppb, some contamination could occur during transport and analyses. No contaminants were detected consistently at any sample locations. Nor were any contaminants found to be above the MDL that did not also occur in blanks (sampling/transport/analysis contamination). A recent study statistically comparing results from 13 study units across the United States has shown that similar compounds and concentrations as found in this study (e.g., acetophenone, phenol and DEET) have also been found in field and source-solution blanks (de-ionized water samples that have not come into contact with sampling equipment) (J. Kingsbury, pers. comm., 2004).

The limestone beneath BNP is very porous and permeable and is expected to exchange water with the surface. Whereas this exchange may occur quickly in high-energy offshore settings (Tribble and others, 1992), the exchange may take as long as a few decades in similar inshore sites (Böhlke and others, 1997). In particular, the modern sediments of the middle shelf form a comparatively low-permeability layer, restricting limestone beneath from surface exchange and creating a leaky trap for groundwater rising from below. Wells at Alina's Reef should have encountered low-salinity groundwater if it were present. Our measurements do not exclude the possibility of springs acting as point sources of contaminants in BNP. But until such springs are located, sampled, and analyzed, they remain hypothetical. Based on this study, no regional groundwater contamination is evident in the BNP area sampled.

Figure 6. Geologic cross-section showing core sites and interpretations across the southeast Florida shelf. (vertical exaggeration is 1:650; for key to lithologic patterns see Appendix C) [larger image]