ATOLL RESEARCH BULLETIN NO. 466 ORIGIN OF THE PELICAN CAYS PONDS, BELIZE BY IAN G. MACINTYRE, WILLIAM F. PRECHT, AND RICHARD B. ARONSON ISSUED BY NATIONAL MUSEUM OF NATURAL HISTORY SMITHSONIAN INSTITUTION WASHINGTON, D.C., U.S.A. MARCH 2000 City, 17:N m Area Honduras Carbonate Shoals (<2m depth) Figure I . Index map showing the location of the Pelican Cays in the Belizean Barrier Reef Complex. Modified from a Landsat TM image acquired 18 September 1987. ORIGIN OF THE PELICAN CAYS PONDS. BELIZE IAN G. MACINTYRE,' WILLIAM F. PRECHT? and RICHARD B. ARONSON~ ABSTRACT Probing with interlocking steel rods and short cores indicates that the small ponds characteristic of the Pelican Cays are formed by differential coral accumulations on a polygonal karst pattern eroded into the underlying Pleistocene limestone. Rapidly accunlulating Acropora cervicornis-dominated communities have been responsible for exaggerating the karst relief, forming steep-sided ridge patterns that commonly result in small restricted ponds. A shallowing- upward facies pattern is documented within the ridges and consists of an Acropora cervicornis facies grading into a Porites divaricata facies, which finally gives way to mangrove peat. INTRODUCTION The most striking topographic feature of the Pelican Cays (Fig. I), in the south-central lagoon of the Belizean Barrier Reef, is the complex network of coral ridges, both submerged and exposed, some with mangrove cover. This unusual honeycomb topographic pattern, once colonized by the red mangroves Rhizophora mangle, forms the characteristic enclosed or partly enclosed ponds of this area of the barrier reef complex (Fig. 2). Similar ridge patterns have been reported in shallow lagoon areas in the Maldives by Purdy and Bertram (1993), who related them to the karst topography of the underlying limestone. The purpose of this study was to investigate the role of subsurface control in the formation of the Pelican Cays network of reef ridges, the extent to which Holocene differential reef accumulation contributes to the relief of these ridges, and the relationship of these ridge patterns to the origin of enclosed ponds. 'Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, DC 20560-0125. 'PBS & J Engineering Science, 2001 Northwest 107Ih Avenue, Miami, FL 33172. )Dauphin Island Sea Lab, 101 Bienville Boulevard, Dauphin Island, AL 36528. Figurc 2. Acrial view of the Pelican Cays looking SSE: Fisherman's Cay (left), Manatee Cay (right), and Cat Cay (background). Note the charactcristic enclosed and partially enclosed ponds in these islands and the network of shallow ridges in Pond C in Manatee Cay. ( Photo by T. Rath) METHODS The research focused on the well-developed network of ridges in Pond C of Manatee Cay (Fig. 2). A 0.95-cm-diameter steel probe, in 3.05-m extensions, was used to establish the contact with the underlying rock substrate (Fig. 3). This contact was easy to detect because of the relative ease of penetrating the overlying open branching coral framework consisting predominantly of Acropora cervicornis (Shinn et al., 1979; Westphall, 1986; Aronson and Precht, 1997). In addition, three short cores were collected by pushing a 7.6-cm-diameter aluminum tube into the crests of reef ridges. Core logs were expanded to correct for compaction that occurred during coring. Radiocarbon dates were determined by standard techniques by Beta Analytic Inc. (Miami, Florida). These dates are reported as radiocarbon years before A.D. 1950, conventionally termed "before present" (B.P.), using a Libby half-life of 5,568 yrs and a modern standard based on 95% of the activity of the National Bureau of Standards' oxalic acid. No corrections were made for the DeVries effect, reservoir effect, or natural isotopic fractionation. RESULTS Our probing and coring activities were limited to the complex ridge system around the opening to the large central pond in Manatee Cay, Pond C (Fig. 4). One transect of probe sites, including a core site, was established across the coral ridge that runs across the entrance to this Figure 3. Probing open coral framework with connecting steel rods to locate depth of solid rock substrate. pond, and a similar transect was set up across the northern inner ridge that cuts across the pond. In addition, a probe and core site, was also located on the crest of the southern inner ridge, an a 3-m probe section was pushed into the center of the southern pond and did not encounter a hard base. As can be seen in Table 1, the two probe transects indicate that the ridges are established on minor relief (1 m and 0.7 m) at the edge of a significant drop in elevation of the hard rock substrate (8.6 m and 5.7 m). This is well illustrated in Fig. 5, which shows a transect across the ridge that extends across the mouth of Pond C. Although the hard substrate was not located in the southern pond probe, similar relief was encountered between this site and the site on the southern inner ridge. The accumulation of Holocene sections (Table 1) indicates that reef growth on the ridges is approximately twice that found in the interior of ponds. Core 1 was collected on the crest of the ridge at the mouth of Pond C and consisted of two core intervals. The first interval reached a depth of 1.5 m and had a core recovery of 80%. The second core interval was from 1.5 m to 4.1 m and had a recovery of 72%. As the core log (Fig. 6 ) indicates, most of the coral recovered was Acropora cervicornis, with smaller amounts of Agaricia renuifolia. Poritesfurcata, and Millepora spp. There was a marked increase in the amount of A. tenul;folia and Millepora spp. in the top section of the core. This transition in reef facies is related to the colonization of A. tenuifolia following the 1986-1992 mass mortality ofA. cervicnrnis throughout this area of the lagoon (Aronson and Precht, 1997). A light grey mud matrix was found throughout both core intervals. Manatee Cay 1 o Figure 4. Map showing probe and core locations in the central area of Pond C, Manatee Cay. 10 20 30 40 HORIZONTAL DISTANCE (METERS) Figure 5. Probe transect across ridge at mouth of Pond C, Manatee Cay. The hard rock substrate shows a slight elevation in relief before a significant increase in depth. Note the exaggeration of this rim relief by the differential accumulation of the overlying section. Table 1. Probe Holes in Pond C , Manatee Cay (depths in meters) Location I Water depth Probe depth Depth of Pleistocene (below MSL) Pond Entrance Ridge Base outer slope Ridge crest Base inner slope Northern Inner Ridge Base northern slope Ridge crest Base southern slope Southern Inner Ridge Ridge crest Southern Inner Pond Center of pond Core 1 Ridae at mouth of pond 0-1.5m Mostly freshly preselved Acropora cervicornis with traces of Agaricia fenuifolia and Porites divaricata. Concentration of Millepora spp, and Agancia lenuifoiia in Halimeda sand in top 0.5m. Aii in a iight grey mud matrix. 'PC 70 t60 BPI 1.5-4.1"- Dominantly Acropora cervicornis with traces of Agaricia fenuifolia, Millepora spp. and Porifes furcala. Haiirneda sand at 1.9-2.2m. All in a light grey mud matrix. LEGEND Acropora ccervicorni. Branching Poriles sl Agaricia tenuifolia Millepora spp. .... .... . . Sand - - - - Mud Figure 6. Graphic summary of data from the two core intervals of Core 1. Recovery was only 21% for Core 2, which was located on the crest of the northern inner ridge. This core (Fig. 7) showed a sharp transition in reef facies at a depth of 2.14 m. The upper section consisted of scattered Porites divaricata, with only a trace of A. cervicornis fragments in a brown organic-rich mud matrix; in contrast, the lower section is dominantly A. cervicornis, with some Agaricia sp. and Porites spp. in a light grey mud matrix. A similar transition of reef facies was also noted in Core 3 (Fig. 8), which was collected on the crest of the southern inner ridge. The upper 0.5 m consisted predominantly of P. divaricata with only traces of A. cervicornis, over a dominantly A. cervicornis lower interval. Core 2 Noflhern Inner Ridge Wdter DeDth 1.0m 0-Z.14m Scattered Porites divancata in brown organic-rich mud. A few fragments of Acropora cervicornis I''c 510 ?80 yrs BPI 2.14-3.8m Mostly Acropora cewicornis with some Porlfes forcafa and a trace of Aoaricia so. Large Porites porites at base. All in a light grey mud matrix. LEGEND I Y acropora cewicornis I 1 @ Branching Porifes spp. I Figure 7. Graphic summary of data from Core 2, Core 3 -1 O.5.08m Acmpoc cenramr r some Agaiicia sp p- pp and Porites sp. fragments in mud grey mud matrix. Southem lnner Ridge Wdter depth 0 8 m 0 z I C a W 0 !I 2 Figure 8. Graphic summary of data from Core 3. Coral samples from both core holes and the base of peat sections exposed in undercut edges of ponds (Fig. 9) were radiocarbon dated (Table 2). Dates ranged from 780 * 60 yrs B.P. to 70 * 60 yrs B.P. Figure 9. Undercut exposure in the edge of Pond C, Manatee Cay, showing mangrove peat overlying coral framework. Table 2. Radiocarbon Dates Location I Material - dated Depth below MSL Date From base of mangrove undercut, east side of Pond A, Cat Cay Care from ridge at entrance to Pond C, Manatee Cay Core from northern inner ridge, Pond C, Manatee Cay From base of mangrove undercut, east side of Pond A, Cat Cay From base of mangrove undercut, south side of Pond C DISCUSSION The Pelican Cays are characterized by an unusual network of reef ridges that are both submerged and emergent, some with mangrove overgrowths. This network of ridges is responsible for the formation of the ponds in this area, which are the habitat of a great diversity of marine life. The distinctive feature of the network pattern is that the ridges commonly intersect at right angles. This pattern is identical to the polygonal karst pattern that Williams (1972) documented on the exposed surface of Miocene limestones in Papua, New Guinea (Fig. 10). Indeed, Purdy and Bertram (1993) used this figure to explain the "peculiar honeycomb pattern" (p. 40) that they observed in shallow lagoon areas in the Maldives. Purdy and Bertram were convinced that preferential reef colonization on polygonal karst relief formed on limestone surfaces during Pleistocene subaerial exposure was responsible for the "honeycomb shoals" (p. 41) in the Maldives. Purdy (1974a, 1974b) also hypothesized that elevated karst Pleistocene relief was responsible for the location and initiation of many of the Holocene lagoon reefs in Belize. I I TOPOGRAPHIC RIDGES ) INTERMIRENT STREAM CHANNEL Figure 10. Plan view of polygonal karst pattern on the surface of Miocene limestones, Papua, New Guinea (Purdy and Bertram, 1993; modified from Williams, 1972). Such honeycomb shoals are common features of shallow lagoon areas and have also been well documented in the Cocos (Keeling) Islands (Searle, 1994). Searle noted that the "central southeastern part of the lagoon is occupied by steep-sided 'blue holes,' some over 15 tn deep" (p. 9, and 100 m wide (Fig. 11). These, Searle suggested, are related to "multi-generational dolins" (p. 5 ) , although he emphasized that both differential accretion and erosional relief are responsible for atoll lagoon morphology. F ig~~re 11. Aerial view of southern area of Cocos (Keeling) Islands lagoon. Note honeycomb ridge pattern ( Searle, 1994). A core hole drilled in the Pelican Cays by the University of Miami in 1984 recovered Pleistocene limestone at a depth of about 20 m (R. N. Ginsburg, personal communication, 1994) This depth coincides with the depth at which our probe encountered hard rock on the outside of Pond C. It is therefore not unreasonable to assume that our probe was recording the Pleistocene limestone surface substrate when we hit hard rock at the base of our probes. These probes confirm the hypothesis that the Pleistocene karst relief is controlling the honeycomb patterns of reef-ridge growth in the Pelican Cays, which commonly results in the formation of ponded areas. The short cores suggest that Acropora cervicornis colonized the areas of slightly elevated relief on the Pleistocene limestone surface when it was flooded by the rising seas of the Holocene Transgression. Differential growth of this fast-growing branching coral community, reported to be accumulating in this area at a rate of up to 8 m/1,000 yrs by Westphall (1986), then formed steep-sided ridges, which on catching up with sea level in some areas, were overgrown by mangrove communities. Although the ponded network pattern of reef ridges is related to polygonal karst relief on a Pleistocene limestone substrate, their relief is mainly the result of differential reef accun~ulation (Fig. 5). The origin and growth of these Holocene lagoon reefs is therefore related to a combination of karst control (Purdy, 1974a; 1974b) and differential reef growth (Halley et al., 1977). This honeycomb pattern is superimposed on a larger rhombohedra1 configuration of the shelf atolls, including the one on which the Pelican cays are located (Fig. 1). The parallelism of these rhomboidal atolls suggests that pre-Holocene reef accumulation occurred along the edges of fault blocks (Purdy, 1974a, 1974b; Precht, 1997). The cores from inside Pond C show a distinct transition from a predominantly Acropovrr cervicorrzis facies to a predominar~tly Porites divaricairc facies, dating to probably about 500 yrs B.P. This facies change docuinents a shallowing-upward reef sequence similar to that reported in this area by Westphall (1986), where a "catch-up" reef cominunity is being replaced by a very shallow "keep-up" community (Neumann and Macintyre, 1985). The Poriks divaricafu community was killed off on the ridges inside the pond when water conditions became restricted following an almost complete closure of the mouth of the pond. A date of 70 * 60 yrs B.P., if valid, indicates that the present restricted conditious within Pond C are recent. Outside of ponds, Porites divwicuta is most common at depths shallower than 1 n ~ . Radiocarbon dates (Table 2) indicate that red mangrove, Rhizophom mangle, communities became established on the Pelican Cays reef ridges at approximately 600 yrs B.P., and there has been a continuum of mangrove colonization ever since. They can be seen in the process of establishing themselves on some submerged ridges today (Fig. 12). Figwe 12. Mangrovcs at the initial stagc of colonization of a shallow suhmcr-gcd rccf- ridgc crcst, Cat Cay. Note the mangrove roots pcnctratin~ a snrface covcl- of Poiites clii~i~ricc~in, Tilcrlossici tesiurliwm, and flic!,.oin sp. CONCLUSIONS The Pelican Cays, located in the south-central lagoon of the Belize Barrier Reef, consist of mangrove-covered islands with a network of circular ponds. Probing and coring studies indicate that the ponds are formed by differential sediment accumulation, with the faster accumulating reef facies growing on the positive karst relief on the underlying Pleistocene limestone. Acropora cervicornis-dominated communities, with documented accumulation rates in this area of Belize of up to 8 m/1,000 yrs, have been responsible for accentuating the positive polygonal karst relief. On catching up with sea level, these reef ridges show a transition from an A. cervicornis facies to a very shallow water Porites divaricata facies. Reef growth is terminated when these ridges become shallow enough for colonization by Rhizophora mangle. Islands are eventually formed when these ridges become capped with mangrove peat. The distribution of ponds in the islands reflects the original polygonal relief pattern of the Pleistocene surface. ACKNOWLEDGMENTS Special thanks go to Victoria E. Macintyre Saville for field assistance. We would also like to thank William T. Boykins for his valuable help with graphics and text layout, and Thad Murdock for additional assistance with graphics. Fieldwork for this project was supported by the National Museum of Natural History's Caribbean Coral Reef Ecosystem Program (CCRE Contribution No. 576). REFERENCES Aronson, R.B., and W. F. Precht 1997. Stasis, biological disturbance, and community structure of a Holocene coral reef. Paleobiology 23:326-346. Halley, R. B., E. A. Shim, J. H. Hudson, and B. Lidz 1977. Recent and relict topography of Boo Bee patch reef, Belize. Proceedings ofthe Third International Coral ReefSymposium, Miami, 2:29-35. Neumann, A.C., and I.G. Macintyre 1985. Reef response to sea level rise: Keep-up, Catch-up or Give-up. Proceedings ofthe Fiflh International Coral Reef Congress, Tahiti, 3: 105-1 10. Precht, W. F. 1997. Divergent wrcnch faulting in the Belize southern lagoon: Implications for Tertiary Caribbean plate movements and Quaternary reef distribution. American Association of Petroleum Geologists Bulletin 81:329-333. Purdy, E. G. 1974a. Reef Configurution: Cause and Ejfect. Society of Economic Paleontologists and Mineralogists, Special Publication No. 18, 9-76. 1974b. Karst-determined facies patterns in British Honduras: Holocene carbonate sedimentation model. American Association qfPetroleum Geologists 585325455. Purdy, E. G., and G. T. Bertram 1993. Carbonate Concepts,from the Maldives, Indian Ocean. American Association of Petroleum Geologists, Studies in Geology, No. 34, 56p. Searle, D. E. 1994. Late Quaternary morphology of the Cocos (Keeling) Islands. Afoll Research Bulletin No. 401, 13p. Shinn, E. A,, R. B. Halley, J. H. Hudson, B. Lidz, and D. M. Robbin 1979. Three-dimensional aspects of Belize patch reefs. American Association of Pefroleum Geologists Bulletin 63528 (Abstract). Westphall, M. J. 1986. Anatomy and history of a ringed-reef complex, Belize, Central America. Master's thesis, University of Miami, Coral Gables, Florida, 135p. Williams, P.W. 1986. Morphometric analysis of polygonal karst in New Guinea. Geological Sociep of America Bulletin 83:761-796.