From: Carew, J. L. (ed.). 1997. Proceedings oftheEighth Symposium on the Geology ofthe Bahamas and Other Carbonate Regions. Bahamian Field Station. San Salvador. Pp. 47-58.
HYDROGEOLOGY OF THE COCKBURN TOWN AQUIFER, SAN SALVADOR ISLAND BAHAMAS. AND TH.E CHANGE IN WATER QUALITY RESULTING FROM THE ,_ DEVELOPMENT OF A RESORT COMMUNITY
Jason S. Erdman and Marcus M. Key, Jr. P.O. Box 1773 Department of Geology Dickinson College Carlisle, PA 17013-2896
R. Laurence Davis Department of Biology and Environmental Science University of New Haven 300 Orange Ave: West Haven, CT 065 I 6
ABSTRACT 17 January 1995. The construction of the Club Med According to many historians, resort led to a 382% increase in groundwater Columbus landed in the New World in 1492, at pumping. There was also a 51% increase in the what is now San Salvador Island in the salinity of the water delivered from the Bahamas. Spurred by this event, five hundred wellfield. At 9 g/1, this salinity renders the years later, a large resort community opened water unusable. Thus, the combination of an on the island near Cockburn Town, the historical quirk that encouraged resort government seat. Water s~pply to both the development, and the island's complex karst town and the resort comes from a I 05 head hydrogeology, has resulted in the degradation wellfield, located nearby. We examined the of the aquifer and the water supply for both hydrogeology and water quality at the well field Cockburn Town and the resort. · to determine what impact the opening of the Ground water elevations did fluctuate resort, with its attendant increase in water with the tides, but their amplitude did not demand, had on the aquifer. dampen with distance from the ocean. The aquifer is in the Cockburn Town Conductivity measurements also showed a Member of the Pleistocene Grotto Beach complex spatial pattern of fluctuations through Formation. The we!lfield is located between the tidal cycle. Although relationships are _the ocean and two inland, hypersaline lakes on weak, it is possible that this complex pattern a platform with an elevation of approximately reflects the influence of dissolucional conduits 3 m above currenr sea level. The ocean is that connect the two lakes to the ocean and roughly 0. 7 km west of the wellfield, Flamingo which probably pass through the aquifer. Pond is roughly 1.1 km east-northeast, and Because of this, and the thinness of ·the Little Lake is roughly 1.6 km southeast. There freshwater lens, the quality of the water is a thin freshwater lens at the wellfield site, supplied to Cockburn Town. prior to the resort and groundwater elevations are tidally development was already marginal. influenced. We obtained 119 weeks of data on. INTRODUCTION pumping rates, rainfall, and salinity, covering the period 9 October 1992 to 20 January 1995 San Salvador Island, Bahamas (formerly from the Bahamian Water and Sewerage known as Watlings Island) is located Corporation and the Bahamian Department of approximately 620 km east-southeast. of Fo:t Meteorology. We also measured groundwater Lauderdale, Florida (Figure l ). The 1sland IS elevations and conductivity in 26 · of ·rhe 1 I km wide (E-W) and 19 km long (N-S) field's l 05 wells over a 25 hour period on 16- (Figure 2). Inland hypersaline lakes occupy
47 .···· ····· ..... ~ . 'Co..
;_ THE \ BAHAMA ' ISLANDS
0 100mi. I I I 0 100km Balhymetric contours in fathoms
'• (cay sat \._ \ Sank ·· .._ ·········100 .... .) _,.·-··· ... . ··-····· ..
Figure !. Map of the Bahamas and .mrrotmding rc;:imt (after Car<'W and Mylroie. 1995a). the swales between the lithified Pleistocene hope that this information might help the and Holocene dune ripges that dominate the people at Club Med and the Bahamian Water topography of the island. and Sewerage Corporation to better understand According to many historians, San the nature of their water supply and perhaps, Salvador was the site of first landfall in the to develop management plans that would allow "new" world in October 1492 (Gerace, 1987). its continued use in the future. Five hundred years later, in 1992, Club Med's Columbus Isle resort opened. The resort San Salvador Geology created a large increase in the demand for freshwater on the island. The source of water San Salvador Island is located on. a for both the resort and Cockburn Town, the small carbon:1te platform in_ the southeastern island's main town and government seat, is a Bahamas (Figure I). Mullins et al. { 1992) wellfield located 1.5 km north of Cockburn suggest that these small platforms are the Town and 0.5 km _southeast of Club Med products of platform-margin collapse and (Figure 2). The goal of our study was to rt'treat spurrt>d by tectonics along the North determine the impact that Club Med's American and Caribbean plate boundary. The opening had on the water quality in the southt'astern part of the Bahamas consists of wellfield and to examine ·what role tht> many of these small platforms, most of which is.land·s karst hydrogeology might have played :tn.• exposed as single isl::lnds such as San in the changes that took place. In the end, we Salvador (Carew and Mylroie, 1995a,b; Curran
48 Wnl~e Cay
Green G.a.ut In D C4~to CAY CAY[;> c7fJ "'\... Graham's Harbour \~ut Cay . '\1 Man Head Cay ATLANTIC Barker's North Po1nt -i Rice Bay OCEAN Point ··-.. ..:=----- Hanna Bay
Bone fish 2 Bay
Crab 1 Ca
Fernandez Bay
ATLANTIC OCEAN Long Bay
Bay t N
High .-otto Cay 112 r {J ;::::::=1•ct crt Owl's ~ uoe u:~a Hole Sandy ~ Gulf Hook Q I l Government Dock low Cay
Figure 2. Map of San Salvador Island, Bahamas. #l indicates the location of the Cockburn Town wellfield. #2 indicates the location of Club Med's Columbus Isle resort (after Curran. 1989 ).
49 and White, 1995). in hydrostatic equilibrium on the denser Drilling data from the Bahamian saltwater below. The shape of the lens is Platform suggest that over 5 kilometers of determined by the difference in the density of carbonate sediment have accumulated on pre the fresh and saline waters, the amount of Triassic crystalline bedrock (Sheridan et al., recharge, and the permeability of the aquifer. 1988). The upper units on San Salvador Island If the aquifer is extremely permeable, the consist of Pleistocene and Holocene limestones freshwater will spread out on the surface of including eolian calcarenite, beachrock, fossil the saltwater and form a thin, brackish lens. coral reefrock, and minor paleosol and subtidal However, if the aquifer lacks sufficient lateral facies (Curran, 1985). Lithified eolian dune permeability, water will move more quickly ridges, 10 to 20 m high, dominate the downward than it can move horizontally, topography of the island, with shallow resulting in a mound of freshwater that brackish to hypersaline lakes occupying the develops above the base elevation of sea level. depressions between (Davis and Johnson, The depth of the freshwater/saltwater 1989). The shoreline is characterized by sand interface below sea level will theoretically be beaches, commonly containing Holocene approximately 40 times the height of the beachrock, located between headlands freshwater table above sea level. Field composed of older, eroded eolianites (Curran, investigations and groundwater sampling on 1985). San Salvador, with its karstified limestones of Elevations on San Salvador reach 10m varying permeability, have revealed that this in the area east of Club Med and southward relationship does not always hold true (Davis toward Cockburn Town (Klein et al., 1958). and Johnson, 1989). This area contains several shallow sinkholes In the interior of the island, fresh formed by the dissolution and collapse of the groundwater is found in discontinuous, surface limestone (Klein et al., 1958). isolated, lenses that lie beneath the Drainage of the area is chiefly underground consolidated carbonate dunes (Davis and because the rainfall sinks into cracks and Johnson, 1989). The interior freshwater solution cavities, or collects in sinkhole lenses, fed by infiltrating rain water, drain to depressions, and infiltrates into the subsurface the brackish and hypersaline lakes, where the (Klein et al., 1958 ). water evaporates. There are similar freshwater lenses near the coast that drain to the ocean; San Salvador Hydrogeology however, these are disrupted in many places by flow through caves and conduits. Most of the The climate on San Salvador is semi interior lakes and blueholes are connected to tropical with a rainy season lasting from May the sea by conduits, but some appear to be fed to December. Mean annual precipitation entirely by precipitation and groundwater (I 02.4 em) is exceeded by mean annual seeps. The freshwater lenses on the island tend potential evapotranspiration (142.8 em) (Foos, to be less than 20 m thick, irregular in shape, 1994 ). Almost all precipitation immediately small, and fragmented (Davis and Johnson, infiltrates into the rock so there is virtually no 1989). flowing fresh surface water on the island The ocean tides also play a role in the (Davis and Johnson, 1989). hydrology of San Salvador. Tides range from The volume and geometry of the 0.3 m to nearly 1.0 in, varying with the time of freshwater lenses beneath oceanic islands are day, lunar phases, and seasons (Davis and largely determined by the climatic and Johnson, I 989). This tidal action mixes the hydrologic characteristics of the aquifer fresh groundwater with the sea water and (Vacher, 1988). In homogenous, relatively moves this mixture through the dissolution unconsolidated and porous sediments, the conduits beneath the island's surface (Kunze downward infiltrating, less dense rainwater and Quick, 1994). Phreatic conduits introduce will displace the denser saltwater forming a sea water into the freshwater system with the lens-shaped body in accordance with the incoming tides. Also, the. conduits conn~ct Dupuit-Ghyben-Herzberg relationship. In this blue holes directly to the ocean, resulting in model (Vacher, 1988), freshwater lenses float tidal fluctuations in the blue holes comparable
50 to those in the ocean (Davis and Johnson, Club Med. The limestone aquifer is capable of 1989). a higher total yield than the sands aquifer. The Cockburn Town wellfield is in the Cockburn Town Member of the Pleistocene MATERIALS AND METHODS Grotto Beach Formation (Carew and Mylroie, J995a,b). The upper 7 m of the limestone in In January 1995, we undertook a study the Cockburn Town wellfield is composed of of the Cockburn Town wellfield and the well cemented, crossbedded oolitic and underlying aquifer to examine the impact that fossiliferous limestone with low to moderate Club Med's opening had on the water quality permeabilities (Klein et al., 1958). Below 9 m of the wellfield and to examine what role the is a hard, highly ·fossiliferous, dissolution island's karst hydrogeology might have played riddled limestone (Klein et al., 1958). in the changes that took place. The Cockburn According to Klein et al. (I 958), there Town wellfield contains 105 wells in II well are two unconfined aquifers, of different lines paralleling both sides of the airport lithologies, in the vicinity of the airport and runway (Figure 3). Each well line contains 8- Club Med. Both yield moderate quantities of 12 wells with one centrally located pump per freshwater from shallow wells. One aquifer is line. The wellfield is located: 0.7 km east of in the limestone area to the north and south of the ocean which has an average salinity of 32 the airport where the Cockburn Town g/1, 1.1 km west-southwest of Flamingo Pond wellfield is located, and the second aquifer is which has an average salinity of 63 g/1, and 1.6 in the unconsolidated, sands beneath the site of km northwest of Little Lake which has an
1·3 2·2 Airport R · z~3 unway 3-3
8-3
MN N I 75 50 25 J L! t!!t t !!l ! t!! I!I IJ I I I I I 100 0 100 :iOO Meter-S Scale 1:3600
Figure 3. Map of Cockburn Town wellfield showing all wells in wei/lines 1-10. Wells sampled during the study are labeled A-B. with A indicating well line#, and B the well#.
51 average salinity of 48 g/1 (Davis and Johnson, runway for a 12 hour period, from 03:00 on 17 1989; Davis unpublished data). Only well lines January 1995 to 15:00 on 17 January 1995. 1-9 (Figure 3) were included in this study due Safety considerations dictated the sampling of to the remote location of well line II, and the these two areas separately. overlap of well line 10 with the northern-most We measured depth to groundwater wells in well lines 6-8 (i.e:, wells 6- 3, 7-3, and using two separate methods. The first method 8-3). involved lowering a "kerplunker", attached to The Bahamian Water and Sewerage the end of a tape measure, into each well. The Corporation (WSC) on San Salvador records "kerplunker" was lowered down the well until water consumption· and chlorinity for both the distinct sound of the "kerplunker" Club Med and CocJ
52 (Figure 4). A t-test revealed no significant on 17 September 1993 (mean== 8.70 g/ i). We change (f>0.05) in water consumption by performed a regression analysis for well lines Cockburn Town before Club Med came on line 4 through 10 combined (R2 = 0.05) for mean (mean = 45,669 1 per week) and after Club salinity over time, and it revealed a significant Med came on line (mean c 39,530 I per week). increase (f<0.05). When examining the A t-test revealed a very significant increase individual well lines, all but well lines 5 and 7 (f - - •- ClubMed -0 g 25 .. • Cockburn Town + Club Med ...... 0 ---Best Fit (Cockburn Town + - Club Med) § 20 ....., ·-c. E ::I cen 15 0 0 0 1993 1994 1995 Time (year) Figure 4. Weekly water consumption of Cockhunz To_wn . Cluh Me_d. ~nd Cockbun~ Town plus Club Med from 9 October 1992 to 20 January /995. Line indicates hest /tt lmear regress101t. 53 16 14 12 -_J -C) 10 -~ ·-s:: ·-- 8 Cl)"' c 6 Q) :E"' 4 ----, ..:..-Mean Salinityj 2 -Best Fit 0 1993 1994 1995 Time (year) Figure 5. Mean weekly salinity of well Jines 4-10 comhined. from 9 Octoher Jl)92 to 20 January 1995. Line indicates best fit linear regression. to determine if a change in the precipitation groundwater, also revealed no signific~nt rate may have influenced the increase in correlation (£>0.05 ). Regression analysis of salinity. A minimum of 0.00 mm of weekly water consumption of Cockburn Town precipitation occurred multiple · times plus Club Med versus mean salinity {R 2 = 0. 10) throughout the study period of 7 April 1993 to reveals a very significant positive correlation 13 January 1995~ The maximum daily (f 54 groundwater elevations in each well were = 0.02), or Little Lake (R2 = 0.06) increases. related to distance from the ocean (mean= 0.7 Again, these fluctuations do not exhibit a km) (R 2 = 0.00), Flamingo Pond (mean "" 1.1 simple spatial distribution. km) (R 2 "" 0.00), or Little Lake (mean • 1.6 2 km) (R .. 0.07). The results indicated no DISCUSSION significant correlation (f>0.05) for any. These elevations do not exhibit a simple spatial Beneath small, oceanic islands, fresh pattern of increased elevation with increased groundwater usually occurs as a lens overlying distance from the discharge points. the saline groundwater which seeps in from the Over the 25 hour study period, the surrounding sea (Volker et al., 1985). Such tidally induced groundwater level amplitudes islands usually lack significant freshwater. (i.e., the fluctuations between low and high Therefore, the freshwater lens is an important tide) ranged from 27 to 67 em (mean • 51 em). source of water for human consumption. These data represent minimum amplitudes as Pumping from the freshwater zone reduces the each well was measured once per hour and not thickness of the freshwater lens and, as a continuously. We predicted that the amplitude consequence, the depth of the freshwater of tidal groundwater elevations would be saltwater interface. This results in local dampened with increasing distance from the upconing of saltwater at the pumping site ocean. A regression analysis (R2 = 0.03) (Volker et al., 1985). revealed no significant correlation (f>0.05) The intrusion of saltwater into a between the amplitude of tidal groundwater freshwater aquifer can be seen, in the elevations and distance from the ocean. These Cockburn Town wellfield, as an increase in the fluctuations do not exhibit a simple spatial salinity of the water being pumped from the distribution. aquifer, which has increased due to the Values for mean salinity in each well development of the resort. The degradation of ranged from 1.80 to 11.10 g/ 1 (mean = 6.74 the Cockburn Town aquifer is confirmed by g/ 1). We predicted that mean salinity in each comparing current salinity values with those well would decrease as distance from the ocean prior to major groundwater withdrawals. A and/ or the hypersaline lakes increased. A 1958 U~S. Air Force study obtained data from regression analysis for mean salinity versus test wells in this same area (Klein et al., 1958). distance from the ocean (R2 = 0.00), Flamingo At that time, salinities ranged from 0.21 to Pond (R 2 = 0.00), and Little Lake (R 2 = 0.04), 1.81 g/ 1 (Klein et al., 1958), which are much revealed no significant correlation (£>0.05). lower than the latest value of 9.85 g/ 1 (mean of It was predicted that there would be the last four weeks of data). higher groundwater salinities at the lower Standard groundwater theory (Fetter, groundwater elevations. A regression analysis 1994) suggests ground water elevation should of mean salinity versus the elevation of the increase with increasing distance from the groundwater table (R2 = 0.08) revealed no discharge points, the ocean or lakes in this significant correlation (£>0.05). These mean case; and groundwater salinity, salinity salinities do not exhibit a simple spatial fluctuation , and tidal amplitude should distribution. decrease with increasing distance. The lack of Over the 25 hour study period, the correlation between these groundwat.er measured tidally-induced salinity fluctuations parameters . and distance from the ocean, between low and high tide ranged from 0.00 to Flamingo Pone!, and Little Lake requires an 3.22 g/ 1 (mean "" 0. 70 g/ 1). These · are the explanation. There are three possible minimum salinity fluctuations as each well was explanations for the lack of the predicted measured every other hour and not correlations. These hypotheses suggest that continuously. It was prediCted that salinity there are more complex subsurface controls on fluctuations would decrease as distance from the groundwater table ·than just distance from the ocean and/ or the hypersaline lakes the discharge points. increased. The salinity £luctuations do not First, all but well lines 3 and 4 were significantly (£>0.05) decrease as distance being pumped during the field phase of the 2 from the ocean (R ,.. 0.03), Flamingo Pond (R2 study. Pumping creates cones of depression 55 around the wells, which would alter the have distinctly horizontal orientations and are elevation of the groundwater table. The second related to changes in sea level and to ancient explanation has to do with the depth of each buried soil horizons (Tarbox, 1987). The individual well. If a well is deeper than 9 m, distribution of conduits is neither then it is connected to the dissolution-riddled I ithologically ,stratigraphically ,norstructurally deep zone of the aquifer (Klein et al., 1958). controlled, suggesting a process unique to Therefore, it may exhibit greater fluctuations conduit formation in the . Bahamas a:nd than a well drilled shallower than 7 m due to geologically similar islands (Mylroie and more ocean tide influence. The effects of the Carew, 1988). As the ocean tides rise and fall, tides oil groundwater fluctuations, as measured the conduits serve as an exchange for ocean by Klein et al. (1958), were greatest in wells water, groundwater, and inland lake water, that penetrated the highly permeable deep and in the process, cause fluctuatjons in both zone, which indicates a good hydraulic the elevation and the salinity of the connection between the deep zone and the groundwater. ocean. Groundwater fluctuations in wells penetrating only the shallow parts of, the CONCLUSIONS aquifer are much smaller in comparison to those of the deep zone wells. The development of the new Club Med The third explanation is that there may resort, supported by the Bahamian government be subsurface conduits that pass beneath the as a means of increasing the economic well wellfield. On carbonate islands, conduits are being of San Salvador, has increased the rate of often caused by dissolution of exposed freshwater withdrawal from the Cockburn limestone during subaerial exposure. San Town aquifer. The salinity has increased to Salvador was repeatedly subjected to subaerial over 9 g/1, which renders the water unusable exposure during low sea level stands for human consumption, and requires a new throughout the Pleistocene Epoch (Carew and assessment of the local groundwater resources. Mylroie, 1985, 1995a,b). Dissolution of the Therefore, the combination of an historical limestones at the groundwater table and at the quirk that encouraged resort development, and halocline has produced subsurface rocks with the island's complex karst hydrogeology, have high and variable permeabilities (Oberdorfer et created the need for this study. al., 1990). The result of this is thnt the The most significant findings of this Pleistocene lithofacies on San Salvador, and study are the 382% increase in water most other Bahamian islands, are honeycombed consumption from the wellfield, and the with an extensive system of caves, conduits, resulting 51 o,'lJ increase in the salinity of this and deep sinkholes (Mylroie and Carew, 1988; water. which are a direct result of neady Eckstein and Maurath, 1995). Many of these quadrupling the pumping rates from the extend from the coast all the way to ponds far we.llrield by Club Med's Columbus Isle in the interior. Both Little Lake and Flamingo Resort. The data also show that the amount of Porid obtain sea water through such conduits. water consumed by Cockburn Town residents Dissolution conduits in the aquifer can has not significantly increased during this change the shape of the ideal freshwater lens period; yet the residents of Cockburn Town by changing the effective hydraulic have experienced an enormous degradation of conductivity of the aquifer, thereby reducing what was an already borderline water supply. the resistance to groundwater movement in Summarizing the spatially complex preferred directions (Tarbox, 1987). In patterns of groundwater elevations, tidal addition to the major controlling parameters groundwater amplitude, and salinity including landmass width, hydraulic fluctuations is probably impossible. The conductivity, and groundwater recharge, the effects of pumping, varying well depths, and dynamics of the ocean also play an important the presence of subsurface conduits can role in the thickness of the transition zone effectively overprint patterns expected in between the fresh and saltwater, groundwater homogenous, isotropic aquifers. In level fluctuations, and salinity fluctuations heterogeneous, anisotropic aquifers any of (Vacher, 1988). Many dissolution features tnesc three parameters can render nearly all 56 types of patterns unrecognizable. REFERENCES CITED The increase in mean salinity in the water supply may have detrimental health Carew, J.L., and Mylroie, J.E. 1985, The effects on the population of San Salvador. Pleistocene and Holocene stratigraphy Sodium hypertension is a direct health effect of San Salvador Island, Bahamas, with of increased salt intake, and its medical effects reference to marine and terrestrial include coronary, neurological, and renal lithofacies at French Bay, in Curran, problems (lsselbacher, 1994 ). Importantly, H.A., ed., Pleistocene and Holocene sodium hypertension is twice as likely to occur Carbonate Environments on San among people of African descent (Isselbacher, Salvador Island, Bahamas: Geological 1994), and most of the population of San Society of America, Orlando, Annual Salvador belongs to that ethnic group. Meeting Field Trip Guidebook: Ft. San Salvador needs groundwater supply Lauderdale, Florida, CCFL Bahamian management. One possible approach could be Field Station, p. 11-61. the construction of recharge basins, which frequently used to recharge unconnned Carew, J.L., and Mylroie, J.E., 1995a, aquifers like those found on San Salvador. Quaternary tectonic stability of the Basins create a substantial hydraulic head that Bahamian archipelago: Evidence from can be maintained to increase infiltration rates. fossil coral reefs and flank margin Because of the large concentration of water in caves: Quaternary Science Reviews, v. a small area, a large groundwater mound forms 14, p. 145-153. beneath these basins; then, when recharge ceases, the mound spreads throughout the Carew, J.L., and Mylroie, J.E., 1995b, aquifer (Fetter, 1994). Another solution could Depositional model and stratigraphy be the installation of a pressure-ridge system. for the Quaternary geology of the In .such a system, a line of pumping wells, Bahama Islands, in Curran, H. A., and parallel to the coast, pumps saltwater back into White, B. eds., Terrestrial and Shallow the ocean, which creates a ridge in the Marine Geology of the Bahamas and potentiometric surface. As a result, wells Bermuda: Boulder, Colorado, behind the trough could be drawn down Geological Society of America Special substantially, even below sea level, with no Paper 300, p. 5-32. fear of saltwater infiltration (Fetter, 1994). Alternatively, pumping of the wellfield could Curran, H.A., ed., 1985, Pleistocene and be reduced by reactivating the now inactive Holocene carbonate environments on rainfall catchment near the airport. This and San Salvador Island, Bahamas: Field the construction of new catchments could Trip Guidebook for Geological Society substantially lessen the pressure on the aquifer. of America, Annual Meeting, Orlando, Florida: San Salvador, CCFL Bahamian ACKNOWLEDGMENTS Field Station, 164 p. This study would not have been Curran, H.A., 1989, Introduction to the possible without the facilities and logistical geology of the Bahamas and San support of the staff of the Bahamian Field Salvador Island, with an overflight Station, San Salvador. We also thank Brad guide: Pleistocene and Holocene Wolf, Jamie Fisher, and Melissa Russo for Carbonate Environments on San their help during this study. In addition, we Salvador Island, Bahamas, Guidebook thank Wellington Fernander of the WSC for his for Field Trip Tl75, 12th International help, suggestions, and patience during the field Geological Congress, American work. This manuscript was improved by the Geophysical Union, p. 1-5. critical reviews of Jeffrey Niemitz, Anthony Pires, Noel Potter, Candie Wilderman, and Curran, H. A.. and White, B., 1995. Gene Yogodzinski, all of Dickinson College. Introduction: Bahamas geology, in Curran, H. A., and White, B. eds., 57 Terrestrial and Shallow Marine water-level fluctuations in water wells Geology of the Bahamas and Bermuda: on San Salvador Island, Bahamas: Boulder, Colorado, Geological Society American Association of Engineering of America Special Paper 300, p. 1-3. Geologists, v. 31, p. 75-89. Davis, R.L., and Johnson, Jr., C.R., 1989, Mullins, H.T., Breen, N., Dolan, J., Wellner, Karst hydrology of San Salvador, in R.W., Petruccione, J.L., Gaylord, M ., Curran, H.A., ed., Proceedings of the Anderson, B., Melillo, A.J., Jurgens, Third Symposium on the Geology of A.D., and Orange, D., 1992, Carbonate the Bahamas: San Salvador, Baham ian platforms along the southeast F.ield. Station, p. 118-136. Bahamas-Hispaniola collision zone: Marine Geology, v. 105, p. !69-209. Eaton, A.D., Clesceri, L.S., and Greenberg, A.E., eds., 1995, Standard Methods for Mylroie, J.E., and Carew, J.L, 1988, Solution the Examination of Water and ·· · conduits as indicators of Late Wastewater, Nineteenth Edition: Quaternary sea level pos1t10n: Washington, D.C., American Public Quaternary Science Reviews, v. 7, p. Health Association. I ,071 p. 55-64. Eckstein, Y., and Maurath, G., 1995, Oberdorfer, J.A ., Hogan, P.J., and Hydrogeology of freshwater lenses and Buddemeier, R.W., 1990, Atoll island karst features of Andros Island, hydrogeology: Flow and freshwater Bahamas: Geological Society of oc~urrence in a tidally dominated America Annual Meeting 1995, New system: Journal of Hydrology, v. 120, Orleans, Louisiana: Field Trip Guide p. 3~7-340. Book #2, p. 49-58. Sheridan, R.E., Mullins, H.T., Austin, ·Jr., Fetter, C.W., 1994, Applied Hydrogeology, J A., Ball, M.M., and Land, J.W., 1988, Third Edition: New York. Macmillan Geology and geophysics of the College Publishing Company, 691 p. Bahamas, in Sheridan, R.E., and Grow, J.A., eds., The Atlantic continental Foos, A., 1994, Water budget for Cockburn margin, U.S: Geological Society of Town, San Salvador, Bahamas: America, The Geology of North Bahamas Journal of Science, v. 2, p. America, v. 1-2, p. 329-364. 31-35. Tarbox, D.L., 1987, Groundwater occurrence Gerace, D. T., ed., 1987. Proceedings of the and development in coastal karst First San Salvador Conference: terrains of oceanic islands in lower Columbus and His World: San Salvador, latitudes: Karst Hydrology, CCFL Bahamian Field Station, 359 p. Engineering, and Environmental Applications: Florida Sinkhole lsselbacher, K.J., ed., 1994, Harrison's Research Institute, p. 287-290. Principles of Internal Medicine: New York: McGraw-Hill, Inc., 2,496 p. Vacher, H.L., 1988, Dupuit-Ghyben-Herzberg analysis of strip island lenses: Klein, H., Hoy, N.D., and Sherwood, C.B., Geological Society of America 1958, Geology and groundwater Bulletin, v. 100. p. 580-591. resources in the vicinity of the auxiliary Air Force bases, British West Volker, R.E., Marino. M.A., and Rolston, Indies: United States Department of O.E., 1985, Transition ~one width in the Interior, Geological Survey, I 50 p. groundwater on ocean atolls: Journal of Hydrologic Engineering, v. Ill, p. Kunze, A.W.G., and Quick, T.J., l994,Tidal 659-676. 58