BULLETIN OF MARINE SCIENCE. 56(1): 48-57. 1995

SURFACE CURRENTS IN SOUND, BAHAMAS AND ADJACENT AREAS WITH REFERENCE TO POTENTIAL LARVAL TRANSPORT

Patrick L. Colin

ABSTRACT Surface currents in Exuma Sound (ES) were examined using satellite-tracked drifters. Of four drifters released on a line across central ES, the two launched on the middle of the transect moved to the northern part of the Sound while those launched nearest the ,~dgesdid not. Of three drifters released near the deep opening of ES, one moved to northern ES and two launched in advance of a cold front moved into the Atlantic. Surface circulation in ES appears active with a general northwesterly flow, eddies and jets, mean current speeds of 11.2-19.8 em sec-I, and drogue drift times of 6-35 days before grounding. Drogues exiting ES into the north Atlantic moved at times to the south and east, contrary to conventional presentations of the Antilles Current and on several occasions approached shallow water areas of . Current patterns indicate the potential for complete pelagic development of larvae of various benthic invertebrates and fishes within ES.

The Bahama Archipelago consists of shallow-water carbonate banks (the Great Bahama Bank and others) with islands, deep channels and deep water basins. Two such basins are the Tongue of the Ocean (TOTO) and Exuma Sound (ES); both in excess of 1,000 m deep, with steep sides and largely enclosed by islands and shallow water banks (Fig. 1). General presentations of current patterns near the Bahamas Archipelago have the Florida Current, flowing northward through the Straits of Florida, and the Antilles Current, running in a northwesterly direction east of the Bahamas, joining to form the Gulf Stream (Wust, 1924; Wust, 1964; Iselin, 1936). Most previous studies in this area have focused on the currents of the Straits of Florida (see Leaman and Molinari, 1987 for references) and Northwest Providence Channel (Lee, 1977; Leaman and Molinari, ]987; Richardson and Finlen, ]967). Less has been published regarding the Antilles Current (Amos et aI., 1971; Ingham, 1975; Gunn and Watts, 1982; Rosenfeld et aI., 1989; Lee et a1. 1990; Leaman and Harris, 1990) and virtually no information is available regarding currents in the central Bahamas. It can be hypothesized that due to the enclosed and restricted margins of ES (and TOTO) this basin may have relatively self-contained circulation of surface waters with limited exchange with adjacent oceanic areas. Such restricted circu- lation might produce a situation where numerous fishes and other organisms with planktonic larval stages undergo their entire pelagic development within ES. Biological evidence supporting this hypothesis is a shallow-water reef fish, Gobiosoma atronasum Bohlke and Robins ]968, which is restricted to ES (Colin, 1975) despite probably having a planktonic larva (larvae of congeners are plank- tonic). Endemism is unusual among Bahamian reef fishes (Smith-Vaniz and Bohlke, 1991) and a restricted, possibly self-contained surface circulation in ES may be largely responsible for this goby's limited distribution. Similarly, more widely distributed species of reef fishes with planktonic larvae may have popu- lations resident in IESwhich are essentially self-recruiting. Despite this intriguing biological information, there is no published current data to support or refute this hypothesis. In an effort to determine whether the surface

48 COLIN: EXUMA SOUND CURRENTS 49

78°W ...... '•....

Figure I. The location of Exuma Sound (ES) in the central Bahamas and other features mentioned in the text. circulation of ES is sufficiently restricted to result in self-recruitment (in ES as a whole) among organisms having planktonic larvae, a series of current-following drogues were released in early 1989 and tracked within and near ES. The results of these drogue releases are presented here with an interpretation of the results relative to pelagic larvae of some shallow water organisms. Exuma Sound is approximately 160 km in its longest dimension (tending north- west-southeast) by 55 km in maximum width (Fig. 1). It is bounded on the east of Cat and Eleuthera Islands, on the north and west by the Exuma island chain and the great Bahama Bank and on the southeast by Long Island. As such it is a semi-enclosed body of water with two connections to the open Atlantic, a deep water (reaching over 2,000 m depth) connection 50 km in width (between Long and Cat Islands) and a shallow sill (generally about 15-30 m depth) 27 km wide between Eleuthera and Little San Salvador. Excluding these openings, the perim- eter of ES is approximately 480 km and is bordered either by islands or carbonate bank. There is free access to water exchange (no intervening islands inhibiting flow) between ES and large areas of the carbonate banks along about 55% of this perimeter.

MATERIALS AND METHODS

The satellite-tracked current-following drifters used in this study were designed at the NOAAJ Atlantic Oceanographic and Meteorological Laboratory, (Bitterman and Hansen, 1986) and consisted of "holey sock" drogue, 10 m in length and 1 m in diameter, attached by a tether to a float made of 10 em PVC pipe 2.3 m in length containing the transmitter and battery pack. The drogue was ballasted so that only 10-20 em of the PVC float, where the internal antennae was located, protruded above the surface. Tether length was adjusted so the drogue hung between 5 and 15 m depth. Drogues were launched in early 1989 with times and locations indicated in Table 1. Positions were reported up to several times per day via the Service Argos system. Day numbers refer to the day of the year (Julian day). Four drogues were released at l6-km intervals along a line NE-SW across ES on 8 January 1989 (Fig. 2a shows start positions). Some were recovered after grounding and relaunched at other locations resulting in a total of seven deployments. Reported positions were estimated to be accurate within about 300 m and in three cases where drogues grounded and were recovered, they were within 100 m of the position reported by the satellite. Up to about eight positions were reported per day and, with the exception of a few which were 50 BULLETIN OF MARINE SCIENCE. VOL. 56. NO. I. 1995

Table I. Data for satellite-tracked current-following drogues launched in or near Exuma Sound, Ba- hamas

Number Distance Drogue Launch Launch Days positions travelled Speed in em sec-l number date position adrift reported (km) Mean (min-max) 3665-1 8 Jan 1989 23.870 N; 76.041 W 6 30 59 11.3 (8.4-12.9) 3665-2 18 Jan 1989 23.740 N; 75.395 W 34 217 440 14.9 (8.6-27.4) 3666-1 8 Jan 1989 24.000 N; 75.936 W 30 234 515 19.8 (9.9-35.5) 3666-2 21 Feb 1989 23.660 N; 75.215 W 345 1,110 6,415 21.5 3666-2 (day 52-66) 14 389 32.2 (17.6-69.0) 3668-1 8 Jan 1989 24.141 N; 75.830 W 35 316 412 13.6 (2.3-25.0) 3669-1 8 Jan 1989 24.278 N; 75.725 W 25 219 242 11.2 (6.1-25.0) 3669-2 21 Feb 1989 23,085 N; 75.385 W 345 1,044 6,125 20.5 3669-2 (day 52-83) 31 826 30.8 (14.9-52.4)

obviously in error, were plotted using Jandel Sigmaplot software. The tracks shown are estimates based on the reported positions. For drogues 3666 and 3669 in the open Atlantic, positions were sometimes not received for periods up to 3--4 days, possibly due to rough water conditions consistently submerging the antennae end of the float. Track length and speed data were detennined using a digitizing tablet and Jandel Sigma Scan software. A stationary recording current meter (General Oceanics 6011 MKll) was moored at 15 m depth off Lee Stocking Island, Exumas, over a 22 m deep insular shelf, from 17 October 1988 to 10 January 1989 (Fig. 2b). The meter was located near the shelf edge to characterize the currents on the island shelf and was isolated from the influences of tidal channels.

RESULTS The tracks of the four drogues launched on 8 January 1989 are shown sequen- tially from launch to grounding in Figures 2 and 3, with each taking an extremely different path. Data on distance covered and speeds of drogues are shown in Table 1. Maximal speeds were based on positions approximately 24 h apart. Maximum instantaneous speed.s are not known. Drogue 3665-1 (Fig. 2a, b) was launched closest to the Exuma chain, 10 km offshore. Over 6 days it moved inshore finally grounding near a tidal channel between the bank and ES. Drogue 3666-1 made two cyclonic circuits of increasing diameter over 10 days after launch while generally moving southeast (Fig. 2a). It then moved parallel to the Exumas (Fig. 3a) only 2 to 6 km off the chain for 4 days, reaching the north central portion of ES. It then made two and a half cyclonic circuits roughly 15 km in diameter over 11 days, and finally grounded in the north central Exumas 30 days after launch (Fig. 3b). Drogue 3668-1 moved northwest for about 8 days after launch to the northern part of ES (Fig. 2a). It then moved southwest, coming close to the Exumas, then turned northwest paralleling the northern Exuma chain and ES. It traced two anticyclonic circuits (Fig. 3a) then made a large cyclonic circuit, moving parallel to the edge of ES north of Eleuthera. It finally moved southwest acros.s ES and grounded off the northern Exumas 35 days after launch (Fig. 3b). Drogue 3669-1 was launched furthest to the northeast, close to the bank west of Cat Island, and initially traced an anticyclonic arc of about 5 km diameter (Fig. 2a). It then reversed direction, moving in a larger cyclonic arc of about 180°, then travelled southwest, across ES, to arrive about 6 km off Great Exuma IsLand(Fig. 3a). It then moved southeast, parallel to Great Exuma, for 4 days and then reversed course and ran along the dropoff of Great Exuma for 2 days, finally grounded near Georgetown, Great Exuma 25 days after launch (Fig. 3a). COLIN: EXUMA SOUND CURRENTS 51

Km ...... • o 20 40 60 '.

;...-.....

••••••••••••••• 1 ...... -.•...~ ....

",., .....

8 Drogue 3665-1 8-14 Jan. 1989

(16 .• )

Great Bahama Bank

Figure 2. A. Tracks of four drogues in Exuma Sound, 8-18 January 1989 (Julian day 8-18). Circles indicate positions received closest to noon each day with the number of the Julian day shown at intervals. St is the starting point of the track. Dashed box contains track of drogue 3665-1, shown below. B. Enlargement of track of drogue 3665-1 off the Exuma chain 8-14 January 1989. All positions received are shown with Julian day/time (GMT) for each indicated. The location of the current meter station off Lee Stocking Island is indicated. St is the starting point of the track. C. Insert shows current patterns at current meter off Lee Stocking Island, 17 October 1988-12 January 1989. Length of arms indicate relative percentages of current in that direction with total percentages indicated in parentheses. Current percentages over 10 em-I are shown by triangular arms. 52 BULLETIN OF MARINE SCIENCE. VOL. 56. NO.1, 1995

Km A"· : . 0 20 40 60 I,:',~''--'"'''' d: 1')22 •....•.• . ·',.Island ATLANTIC ~\~...,-' " " . OCEAN \:, 'a.i --..-':t:=:... lilUc San <1\ 19 .• S_d...-

Tracks of drogues ..' -', Long 18 Jal1uary (day 18) "

Km B o 20 40 60

ATLANTIC OCEAN ' •.....•. :::.::.:;;:.. lilt •• San ", Solwoclo:w

32/.5065-2

Track of drogues 1 February (day 32) ..f..~' ....Long. l~land ta -.... I",

22 February (day 54) ~ ...... •.. -.... , '-. 1989

Figure 3. A. Continuation of tracks of three drogues and initial track of one drogue in Exuma Sound, 18 January-l February ]989 (Julian day 18-32). Circles indicate positions received closest to noon each day with the number of the Julian day shown at intervals. St is the starting point of the track. B. Continuation of track of three drogues in Exuma Sound, 1-22 February 1989 (Julian day 32-54) Circles indicate positions received closest to noon each day with the number of the Julian day shown at intervals. COLIN: EXUMA SOUND CURRENTS 53

The remaining three launches took place near the north end of Long Island, near the oceanic entrance to ES, using the drogues recovered from the 8 January launch. Drogue 3665-2 was launched again on 18 January NW of the north end of Long Island (Fig. 3a). It initially moved southwest paralleling the edge of ES for 2 days, then reversed course and moved NE across the mouth of ES over 4 days. Despite nearly exiting ES, it then turned west, running along the south shore of Cat Island and out into central ES. It made a large anticyclonic loop over 11 days in central ES, then moved northward (Fig. 3b) becoming northeastly and finally grounded on the shelf edge off southwestern Eleuthera (Fig. 3b). Figure 2c shows the generalized current speed and direction from October 1988 to January 1989 at the stationary current meter station off Lee Stocking Island. Currents were predominently along the shelf edge (NW-SE). NW currents oc- curred more often than SE, but were generally lower speed (no more than 30 cm sec-' while SE currents were occasionally near or over 40 cm sec-I). No tidal effect was discernable at the site of the current meter (Smith, pers. comm.). The final launches were made on 21 February 1989 with 3666-2 being started inside ES (NW of Long Island) and 3669-2 on the east side of Long Island 8 Ian south of the north end (Fig. 4). These drogues were launched immediately prior to passage of a cold front through the area with strong north and northeasterly winds and both took easterly and southeasterly tracks away from ES. Both tracked past Conception Island and and 9 days after launch were 125 Ian ESE of San Salvador, almost directly east of their launch points. At this point their tracks diverged (Fig. 4c). Drogue 3666-2 moved generally northeastward over the next two months into the central Sargasso Sea (Fig. 4a). It turned back southwest over the next three months (April-June), crossing its outbound track twice, to arrive about 300 Ian northeast of the Bahamas. It then moved south, arriving near the latitude of its launch and 300 km to the east 200 days after launch (mid-August). It then went northwest and west coming within 60 km northeast of the Little Bahama Bank (LBB) in October and moved northwest along the LBB finally appearing to be- come entrained into the eastern edge of the Gulf Stream in late December (Fig. 4a). Drogue 3669-2 moved back along a track similar to its outbound track from north Long Island, passing between Conception and Long Islands 25 days after launch (Fig. 4c). The drogue then closely approached the south end of the Cat Island 2 days later and then went north into the open Atlantic. It then turned to the west, eventually approaching to within 14 km of the northeastern edge of the LBB about 65 days after launch (Fig. 4d). The drogue continued on its generally northwesterly track, but once north of the northern extreme of the LBB, it moved back to the east and south, again approaching to within 30 km of the LBB near Great Abaco at 85 days after launch. It then moved to the east and slightly south out into the central Sargasso Sea where it again reversed course and moved back to the northwest starting about 150 days after launch. It eventually arrived again in the area north of the LBB's northern extreme and finally appeared to become entrained into the eastern edge of the Gulf Stream. Both drogues 3666-2 and 3669-2 were terminated at the end of January 1990, some 345 days after launch.

DISCUSSION Exuma Sound Circulation.-ES appears to have an active, but relatively self- contained, surface circulation dominated by both eddies and jets with a general northwestward movement of surface water along its axis. Drogue tracks indicated 54 BULLETIN OF MARINE SCIENCE, VOL. 56, NO. I, 1995

o 31

o 29

o 27

Trock of drogue 3656-2 o 25 21 Feb 1989 to 31 Jon 1990 Numbers indicate day of year 01989 o .,990 23

o B Track of dro~lue 3669-2 31 21 Feb 1989 to 31 Jan 1990 Numbers indicate day of year 01989 o • '990 29

NO: 01 27 '

p o ~ 25

Track or drogue 3669-2 N D 22-30 AprIIllil8S1 end . 15-111 ~ 1989

a 23 .,,:' 76<>W

Figure 4. A. Track of drogue 3666-2 in the western North Atlantic Ocean, 21 February 1989 to 31 January 1990. Circles show positions closest to noon at 5-day intervals with Julian day indicated. B. Track of drogue 3669-2 in the western North Atlantic Ocean, 21 February 1989 to 31 J,muary 1990. Circles show positions closest to noon at 5-day intervals with Julian day indicated. C. Detail of early COLIN: EXUMA SOUND CURRENTS 55 relatively rapidly movement (over 20 cm sec-I) for periods of several days along the NW-SE axis of ES at varying distances from the edges of ES. Likewise, drogues also circled in eddies for several days, particularly in the northern portion of ES. Drogues also moved from the east to the west side of ES. The current speeds recorded at the LSI current meter station were consistent with the speeds indicated by the drogue data. The area near the deep opening of ES has the potential to move either into or out of ES, perhaps due to effects of weather. The only clear movement of a drogue from within ES to the open Atlantic was drogue 3669-2. It is doubtful whether there is regular movement of water on the surface from ES to the Atlantic through the deep opening. None of the four drogues launched on the 8 January transect, well inside ES, exited from ES. Drogues released near the mouth of ES between Cat and Long Islands had various tracks, either into or out of the Sound. Drogue 3665-2 nearly exited south of Cat Island, but turned and moved west to remain within and travel to northern ES. 3669-2 passed by the mouth just to the east on day 77-79, but did not move into the Sound. The only drogue launched within ES which moved into the Atlantic (3669-2) may have responded to the strong northerly wind conditions just after its release to produce the observed track. There appeared to be no tendancy for drogues to exit ES via the shallow sill between Eleuthera and Little San Salvador. The only drogue which approached this area was 3665-2 which moved in a large anticyclonic loop, and at one point was moving toward the sill opening. The drogue, however, then moved away from the opening and its long arcing movement towards the north and its sub- sequent grounding on the leeward shore of Eleuthera perhaps indicates some movement of water into ES across the sill. Due to the limited number of positions available, no clear tidal influence was evident in the movement of drogue 3665-1, although it was launched closest to the bank. It is possible such factors are important in affecting currents on the margins of ES. The tidal channels between islands of the Exuma chain can have currents ranging up to 1 m sec-lor more. This results in movement of water from ES up to several km onto the bank during spring flood tides. This water which is drawn onto the banks by tidal currents can become altered during its residence time on the bank by changes in temperature and salinity. This tidal exchange of oceanic and bank water occurs along the entire perimeter of ES except in areas where large islands, such as Eleuthera and Great Exuma, block broad access to the adjacent banks. Where many islands occur, such as the Exuma chain, the tidal flow is channeled into passes between the island so that the amount of water exchanged by tidal processes remains high. The Antilles Current.-Gunn and Watts (1982) reported that the Antilles Current did not always exist and that mesoscale eddies dominate the surface circulation of the area when it did occur. They reported a net northwestward transport of water during winter (January-February 1973), but a lack of similar net transport in summer (June-July 1972). During the present study drogues 3666-2 and 3669- 2 appear to track in the direction expected during the summer if the Antilles

tracks of drogues 3666-2 (solid line) and 3669-2 (dashed line) in the east central Bahamas. Circles show positions closest to noon each day with Julian day indicated at intervals. D. Detail of two approaches of drogue 3669-2 to the northern Little Bahama Bank in April and May 1989. Circles show all positions with Julian days indicated. 56 BULLETINOFMARINESCIENCE,VOL.56, NO.I, 1995

Current exists (July-October for 3669-2) (August-October for 3666-2), However, during the spring (April-June) drogue 3669 tracked almost exactly the opposite of the direction expected of the Antilles Current and covered in about 70 days the same distance that took 130 days on the return trip during the following summer and fall. Possible Relationships between Circulation and Living Organisms.-The drogue drift times of 6-35 days before grounding within ES (all of which were launched at least 10 km from shore) imply that a parcel of surface water might spend, at most, a few months resident in ES without ever coming in over shallow-water environments. This is a result of the active circulation of ES and the tidal transport on and off of the bank along the margins of ES. Given periods of a few months, the prospects of a larva living in the surface pelagic environment eventually en- countering a shallow water insular shelf are high. Once in such areas tidal currents can carry pelagic larvae several km up onto the bank in a single tidal exchange. In contrast, larvae may also be faced with the difficulty of remaining in the pelagic environment until ready to recruit. Tidally-controlled flow along the mar- gins of ES probably does not reach very deep because the tidal channels between islands are generally no more than about 6-9 m deep. If larvae are capable of maintaining themselves at a depth in the pelagic realm below that of the tidally- influenced flow, they are unlikely to be drawn onto the banks by tidal flow, but can proceed with their planktonic development with a high chance of later, by moving into shallow surface waters, having access to shallow water when ready for recruitment. The observed movement of drogues outside ES also indicates the potential of both 1) relatively long distance transport of larvae with gyres which would tend to bring larvae in over shallow environments and 2) transport producing potential self-recruitment over periods of a few days to many weeks and months. For example, drogue 3669-2 came close to its launch point, near a former Nassau , Epinephelus striatus, aggregation area, 25 days after launch, and ap- proached the Little Bahama Bank (LEE) on several occasions after :50, 62, 85 and 300 days post launch. On any of those approaches, if larvae carried along with the drogue had the potential to detect and move towards shallow water they would have had a reasonable chance of finding such environments. However, drogue 3666-2 only approached the LEE once, about 265 days after launch; there is certainly great variation in the chance of a particular water mass coming close to shallow bottom.

ACKNOWLEDGMENTS

This work was based at the Caribbean Marine Research Center on Lee Stocking Island. Exumas, Bahamas and I thank the staff of CMRC for their assistance. For their help in the field I would like to thank L. J. Bell, I. Hunt von Herbing, M. Tupper, R Wicklund. G. Wenz and E. Wishinski. The RI V J. W. POWELLand UNDERSEAHUNTERof CMRC were utilized for launching drogues. D. Bitterman of NOAA, AOML constructed the drogues and assisted with start up of the program. This work was supported by the National Undersea Research Program, Office of Ocean and Atmosphe];c Research, NOAA with assistance from the Perry Foundation and the Bahamas Undersea Research Foundation.

LITERATURE CITED

Amos, A. E. A. L. GorCion and E. D. Schneider. 1971. Water masses and circulation patterns in the region of the Blake ..Bahamas Outer Ridge. Deep-Sea Res. 18: 145-165. Bitterman, D. S. and D. V. Hansen. 1986. The design of a low cost tropical drifter buoy. Marine Data Systems Intemational Symposium Proceedings. Mar. Tech. Soc.: 575-581. Colin, P. L. 1975. Neon: gobies. T.EH. Publications, Neptune City. New Jersey. 304 pp. COLIN:EXUMASOUNDCURRENTS 57

Gunn, J. T. and D. R. Watts. 1982. On the currents and water masses north of the Antilles/Bahamas arc. J. Mar. Res. 40: 1-18. Ingham, M. C. 1975. Velocity and transport of the Antilles Current northeast of the Bahama Islands. Fish. Bull. U.S. 73: 626--632. Iselin, C.O'D. 1936. A study of the circulation of the Western North Atlantic. Pap. Phys. Oceanogr. Met. 4. 101 pp. Leaman, K. D. and J. E. Harris. 1990. On the average absolute transport of the deep western boundary currents east of Abaco Island, the Bahamas. J. Phys. Oceanogr. 20: 467-475. --- and R. L. Molinari. 1987. Topographic modification of the Florida Current by Little Bahama and Great Bahama Banks. J. Phys. Oceanogr. 17: 1724-1736. Lee, T. N. 1977. Coastal currents along the southern shore of Island. Bull. Mar. Sci. 27: 803-820. ---, W. Johns, F. Schott and R. Zantopp. 1990. Western boundary current structure and variability east of Abaco Island, Bahamas at 26.5°N. J. Phys. Oceanogr. 20: 446-466. Richardson, W. S. and J. R. Finlen. 1967. The transport of Northwest Providence Channel. Deep-Sea Res. 14: 361-367. Rosenfeld, L. K., R. L. Molinari and K. D. Leaman. 1989. Observed and modeled annual cycle of transport in the Straits of Florida and east of Abaco Island, the Bahamas (26.5"N). J. Geophys. Res. 94(C4): 4867-4878. Smith-Vaniz, W. F. and E. B. Bohlke. 1991. Additions to the ichthyofauna of the Bahamas Islands, with comments on endemic species. Proc. Acad. Nat. Sci. Philadelphia 143: 193-206. Wust, G. 1924. Florida- und Antillenstrom. Eine hydronamische Untersuchung. Veroff. Inst. Meeresk. Univ. Berlin, Heft 12. 48 pp. ---. 1964. Stratification and circulation in the Antillean-Caribbean basin. Columbia University Press, New York. 201 pp.

DATEACCEPTED:July 7, 1993.

ADDRESS: Research Foundation, Chuuk Atoll Research Laboratory, P.O. Box 70, Weno, Chuuk, Federated States of Micronesia 96942.