Long-term records of storm deposition and coral reef community response from a late Pleistocene reef on Barbados

Transcription

1 Proceedings 9 th International Coral Reef Symposium, Bali, Indonesia October 2000 Long-term records of storm deposition and coral reef community response from a late Pleistocene reef on Barbados C. T. Perry 1 ABSTRACT Shallow water Pleistocene coral reef facies (dominated by Acropora palmata rubble), exposed along the NE coast of Barbados, comprise multiple storm depositional units. Individual units are characterised by an increase in the degree of substrate encrustation and a succession from thin, sciaphilic encruster dominated assemblages at the base of each unit to progressively thicker, photophilic encruster dominated assemblages at the top. In contrast to many modern post-hurricane or disease impacted reef systems there is, however, no evidence of long-term shifts in coral community composition following physical disturbance. Upper surfaces of each storm package are colonised by juvenile Agaricia agaricites (a common opportunist coloniser). The average sizes of juvenile colonies vary between horizons, but are consistent with growth over timescales of <10 years. These corals are overlain on each occasion by a new deposit of A. palmata rubble, the presence of which infers that A. palmata communities were able to regenerate over successive physical disturbance cycles. Keywords Hurricanes, Acropora palmata, Storm rubble, Calcareous encrusters, Reef recovery Introduction Coral fragmentation and the modification of community structure and reef topography are well documented consequences following hurricane impacts on coral reefs (Stoddart 1974, Woodley et al. 1981, Hubbard et al. 1981, Rogers et al. 1982, Hubbard 1992, Scoffin 1993). Particularly susceptible to physical disturbance are branched corals such as Acropora cervicornis and A. palmata which, in the Caribbean, characterise high energy, shallow water environments (Woodley et al. 1981). Monitoring of storm deposited fragments of branched corals on modern reefs has illustrated both high survivorship potential of fragments (e.g. Belize: 46% - Highsmith et al. 1980, St. Croix: 35-66% - Rogers et al. 1982) and rapid rates of initial regrowth (up to 18 cm/yr/ - Rogers et al. 1982). It is thus inferred that post-hurricane impacted reef communities (especially those previously dominated by fast-growing corals) should have the potential to return rapidly to prestorm impact conditions (Rogers 1993). This suggestion is supported by the observed dominance of Caribbean shallow water reef environments by Acroporiids throughout the Pleistocene and Holocene (Macintyre 1988, Jackson 1992). Post impact recovery in modern reef environments, however, often shows evidence of being suppressed, with communities failing to recover from severe physical stress (Woodley et al. 1981). This may consequently complicate our understanding of what constitutes natural rates and styles of reef recovery. In many cases, slow recovery is attributed to the additional stresses imposed through anthropogenic (e.g. over-fishing, sedimentation) and disease (e.g. White Band Disease) related impacts (Gladfelter 1982, Bythell et al. 1993). This may result in subsequent changes in community composition and structure. Commonly this is evident in the form of reduced live coral cover, changes in coral species abundance and increased macroalgal cover (the phase shift concept; Done 1992). These shifts are exemplified by the well-documented example from Jamaica (Goreau 1992, Hughes 1994,) where major community changes (primarily loss of branched coral species) and delayed recovery of branched corals have occurred following hurricane disturbance (Hughes 1994, Aronson and Precht 2000). Community trends over repetitive physical disturbance cycles are even more poorly constrained due to the short time periods (30-40 years) since detailed reef monitoring was initiated. Well-exposed Pleistocene reef systems, however, provide a potential resource for examining histories of reef growth (and disturbance) over timescales beyond those available in modern environments. They also provide a record of reef growth and framework accumulation in environments free from anthropogenic influence. Potential criteria for recognising storm depositional packages in the rock record have previously been identified based on examination of A. palmata storm rubble at Discovery Bay, north Jamaica, 3 years after Hurricane Allen in 1980 (Scoffin and Hendry 1984). A rubble layer, up to 2 m thick, was deposited on the forereef in a more or less instantaneous depositional event. Both the upper surface of the rubble pile and the top of an apparently underlying depositional unit were colonised by small (up to 10 cm diameter) juvenile A. agaricites and P. astreoides coral colonies. Excavated rubble revealed a characteristic sequence of colonisation by calcareous encrusters, from sparse, sciaphilic (= shade loving) encrusters (primarily serpulids, bryozoans and foraminifera) at the base of the rubble pile, to progressively thicker crusts dominated by photophilic (= light loving) forms (primarily coralline algae) towards the top (Scoffin and Hendry 1984; Fig. 1). This was interpreted to reflect differences in the depth of rubble burial and thus different light/turbulence regimes. 1 Department of Environmental & Geographical Sciences, Manchester Metropolitan University, John Dalton Building, Chester Street, Manchester M1 5GD, UK. c.t.perry@mmu.ac.uk

2 Relationships between depth, light and turbulence and the occurrence of calcareous encrusters in Caribbean reef habitats have previously been described by Martindale (1992). Such criteria were inferred to have potential application to the recognition of ancient storm deposits, a concept tested in this study. Fig. 1 Schematic diagram showing the development of encruster communities at different depths in a coral rubble pile following storm deposition. Clasts deposited at the top of the rubble layer are characterised by high light and turbulent conditions and are colonised by thick sequences of photophilic encrusters (e.g. coralline algae). Clasts buried at depth are characterised by dark, low energy conditions and colonised by sparse sciaphilic encrusters (e.g. foraminifera, serpulids) (adapted from: Scoffin and Hendry 1984). Study Area Barbados lies to the east of the Windward Islands of the Caribbean and comprises, in large part, a series of reef terraces which formed during successive periods of sealevel highstand during the Pleistocene. These have been gradually uplifted such that exposed terraces increase in age with elevation. Analysis of fossil reef communities and sampling of coral rubble was undertaken at sites in the north-east of Barbados (Fig. 2) and focused on coastal outcrops dated at 83,000 yrs (James 1971). Detailed studies and sampling were not attempted in older (higher) terraces due to problems of coral preservation and micritization of encrusters (Martindale; pers. comm). Reefs on this north-east side of Barbados would have formed as inshore fringing reefs possibly protected by offshore barrier reefs (James 1971; Fig. 2) and would have been located in a windward (high energy) setting. Coral zonation is analogous with that occurring in Holocene reef systems in the Caribbean. This paper focuses on one section at Middle Bay on the NE coast of Barbados (Fig. 2). The Middle Bay succession comprises a rudstone dominated by cobble size fragments of A. palmata (Fig. 3), typically cm in length. This material is clast supported and is interpreted to represent storm deposited rubble (see Blanchon and Eisenhauer, in press). Small in-situ colonies of A. agaricites are present along horizons within the succession (Fig. 3). Methodology Sampling was undertaken only within the A. palmata rubble facies. This facies is widespread and well-exposed in road-cut and coastal sections around Barbados, but more importantly, represents the reef crest facies (Mesolella 1967). As such it corresponds to the near reef crest environments examined by Scoffin and Hendry (1984) in their study of modern storm deposits in Jamaica. Detailed logs of A. palmata dominated sequences were constructed utilising 0.5 x 0.5 m quadrats to enable depositional fabrics to be mapped. In the field, thickness of calcareous encruster assemblages on the upper surfaces of individual clasts were measured, as were dimensions of all corals colonising the rubble deposit. High-resolution sampling (every 5-10 cm vertically through the section) of A. palmata rubble was undertaken, with samples retained for subsequent analysis of encruster community composition. Samples were prepared as 20 μm thick thin sections to enhance identification of encrusters. Thin-sections were systematically examined to determine both the thickness and composition of calcareous encruster assemblages. Encruster identification was based upon descriptions given for coralline algae (e.g. Johnson 1961, Wray 1977) and foraminifera (e.g. Rooney 1970, Martindale 1976, 1992). Since bryozoans proved both problematic to identify in thin-section and are typically rare in highenergy shallow reef environments (Martindale 1992), and since serpulids have limited application as high-resolution palaeoecological indicators (Martindale 1992) these forms were recorded only on a presence or absence basis. Results The Middle Bay succession comprises a 4-5 m thick rubble sequence dominated by A. palmata plates. Six storm depositional units are recognised within this succession on the basis of, 1). the composition and

3 thickness of calcareous encruster communities preserved on A. palmata clasts, and 2). the presence of horizons characterised by the occurrence of in-situ juvenile A. agaricites colonies (Fig. 3). Individual units vary in thickness with the upper surfaces often topographically complex and difficult to trace laterally (possibly reflecting spatial variability in the thickness of rubble deposits), however the following observations are typical. Each unit (S1 to S6; Fig. 3) exhibits, to a greater or lesser degree, a gradational increase in the thickness of the encruster sequence from base to top. Within units encruster assemblages exhibit a broad trend from thin (1-2 mm thick), sciaphilic assemblages at the base, dominated by serpulids and foraminifera (e.g. Homotrema rubrum - branched and low relief forms, and Carpenteria utricularis - conical and globose forms) to thicker (up to 20 mm thick) photophilic assemblages at the top, comprising multiple crusts of coralline algae (e.g. Porolithon sp., Mesophyllum sp.). more foliaceous crusts of coralline algae such as Neogoniolithon sp. and Lithoporella sp., to the thick coralline crusts of the upper unit surfaces. In some examples, thin crusts of coralline algae occur beneath sciaphilic dominated assemblages. These presumably represent crusts that formed on dead coral substrate prior to storm-induced fragmentation and rubble deposition. In all cases, regardless of unit thickness, upper coralline algal crusts occur intergrown with A. agaricites (and more rarely P. astreoides) colonies. These intergrowths of coralline algae and opportunist coral species mark the upper surface of each unit and occur in all sections examined. Indeed, in the thinner depositional units the occurrence of thick coralline assemblages along with A. agaricites provide the most convincing evidence of an upper storm deposit surface. Mean colony size for A. agaricites in the Middle Bay section is 88.6 mm diameter (range mm), although this varies between units. However, along individual horizons colony sizes are comparable inferring that they reflect products of related recruitment events. Overlying each coralline algal/a. agaricites horizon is a new A. palmata dominated rubble deposit with similar trends with respect to encruster sequence thickness and composition. Discussion Fig. 2 Location of Middle Bay and an idealised crosssection through reef facies on the NE sides of Barbados (after James 1971). Abbreviations used in schematic reef sections: Ap A. palmata; P Porites/Thalassia zone; Ck Mixed coral head zone; T Terrigenous sands; Ls Lagoon sands. Although comparable encruster assemblages occur within each depositional unit (Fig. 3), the extent to which the scia-photophilic sequence is developed varies between units. Typically it is best developed within the thicker units (e.g. S1, S4) where more marked transitions from open, well lit to dark, cryptic conditions would have developed. Thinner units, where this environmental gradient would have been less marked (e.g. S2, S3), are characterised by a transition from thin (2-3 mm), Shallow water (A. palmata dominated) Pleistocene sequences at Middle Bay (and at other sections in NE Barbados Perry in press) record evidence of multiple storm depositional sequences. These exhibit marked vertical variations in encruster community composition and encruster sequence thickness. In all sections, discrete depositional units can be identified, each of which exhibit, to a greater or lesser extent; 1) a gradual vertical increase in encruster succession thickness; 2) a shift from thin (1-2 mm), sciaphilic crusts on plates at the base of each unit, to thick (up to 20 mm) photophilic crusts at the top of each unit; and 3) an upper surface colonised by small in-situ, opportunist corals (mainly A. agaricites). These encruster sequences are comparable with those identified in the modern post-hurricane example from north Jamaica (Scoffin and Hendry 1984). However, the extent to which this succession is developed within successive depositional units varies. The clear sciaphilic to photophilic assemblages described by Scoffin and Hendry (1984) are best developed in the thicker units (e.g. units S1, S4; Fig. 3). In thinner units (e.g. unit S4) many of the sciaphilic encrusters are absent or rare, presumably because clasts deposited at the bases of these thinner units were still receiving sufficient light for coralline algal successions to develop. The overall style of encruster development within each unit, however, demonstrates the potential to apply the modern storm encrustation model of Scoffin and Hendry (1984) to the recognition of ancient

4 Fig. 3 Log from Middle Bay section, N.E. Barbados. Six storm depositional packages (S1-S6) are identified. Coloured boxes along individual horizons denote the presence of encrusters identified in individual thin-sections. Calcareous encrusters identified: 1 - bryozoa, 2 H. rubrum (globose form), 3 serpulids, 4 H. rubrum (branched form), 5 C. utricularis (globose form), 6 C. utricularis (conical form), 7 H. rubrum (low-relief form), 8 G. plana, 9 Lithoporella sp., 10 Tenarea sp., 11 Mesophyllum sp., 12 Neogoniolithon sp., 13 Lithophyllum sp., 14 Porolithon sp. Cryptic to open surface encruster sequences are interpreted from published shallow-water (0-8 m water depth) encruster occurrence data (Scoffin and Hendry 1984, Martindale 1992). storm sequences. The use of such criteria thus have potential to be applied to the analysis of core sections through Holocene/Pleistocene sequences to assist the interpretation of depositional events. In addition to providing evidence of storm-related depositional episodes, data from these outcrops are of interest in relation to assessing long-term records of posthurricane community response. Many modern reef communities in the Caribbean have undergone major community shifts following disturbance (Gladfelter et al. 1991, Bythell et al. 1993) and are now characterised by different coral assemblages to those present in the 1970 s (Aronson and Precht 2000). Such community changes should have good potential to be preserved in the fossil

5 record. However, Pleistocene successions examined in Barbados, provide no coral faunal evidence to suggest any major post-hurricane community transitions. Indeed, the presence of repetitive A. palmata dominated sequences implies that branched coral communities were capable of regenerating over successive disturbance cycles. Whilst the timing or longevity of disturbance intervals is hard to constrain, several lines of evidence suggest there were no prolonged changes in community structure. These include, 1). the restricted diversity of opportunist coral species along upper storm rubble surfaces, 2). the small size of opportunist corals (primarily A. agaricites) along these surfaces and 3). the lack of evidence (e.g. extensive bioerosion) suggesting prolonged exposure of rubble at the sediment-water interface following deposition and prior to substrate colonisation. A. agaricites is a well documented pioneer recruit following natural disturbances (Stoddart 1974, Bak and Engel 1979, Hughes 1985) and the absence of either large specimens of this species, or a greater diversity of other pioneers (see Bak and Engel 1979) is intriguing. Common coral species and mean maximum colony diameters on the shallow fore-reef in Jamaica are, in decreasing order of abundance; A. agaricites mm (range: mm), P. astreoides mm (range: mm), Montastrea annularis mm (range: mm), Diploria strigosa 98.5 mm (range: mm) and Diploria labyrinthiformis 120 mm (1 specimen). Such colonies would have good potential to be preserved beneath new storm rubble deposits, but contrast markedly with observations from storm horizons in Barbados. Only limited data has been published on growth rates (annual increases in diameter) of A. agaricites (Table 1), although these have not followed growth from initial recruitment of the colonies. However, they can be used to provide some indication of possible lifespans for the colonies measured in the NE Barbados sections. Using a mean annual growth rate, derived from the figures in Table 1, of 17.2 mm/yr, estimated time periods for A. agaricites growth can be approximated (Table 2). It is significant that even given a potential delay in pioneer species recruitment of several years, colony sizes imply that colonies were growing for only a few years after recruitment. It is also of interest to note that the upper surfaces of A. agaricites show no evidence of an obvious cause of mortality. Some colonies are partially overgrown by coralline algae, but in others corallite structure is exquisitely preserved, implying instantaneous mortality. It is thus suggested that possible causes of juvenile A. agaricites mortality were either overgrowth and shading by rapidly recovering in-situ A. palmata colonies and/or burial by rubble following the next storm depositional event. Although the precise nature of the recovery/burial cycle is thus hard to interpret, it is clear that repetitive accumulations of mono-specific layers of A. palmata rubble repeatedly occur over marked storm horizons. This implies that Pleistocene A. palmata communities in Barbados successfully recovered from successive physical disturbance events. It is of interest to note that there is no preserved coral or other faunal Table 1 Published lateral growth rate data for Agaricia agaricities from sites around the Caribbean. Site Depth (m) Lateral growth rate (mm/yr) Jamaica 10 mean: 7 (range: 0-24) Reference Hughes and Jackson, 1985 Curaçao Bak, 1976 Curaçao Van Moorsel, 1985 St. Croix Rogers et al Curaçao/Bonaire Bak and Engel, 1979 evidence to suggest major changes in reef community composition, or long-term cessation of branching coral dominance following disturbance. This contrasts with many modern reef systems in which recovery from physical disturbance is inhibited by additional stresses imposed through anthropogenic/disease related disturbance (Curran et al. 1994, Aronson and Precht 1997). Table 2 Estimated lifespans of A. agaricites recruits from storm rubble surface in the Middle Bay section. Ages are based on an average lateral growth rate figure of 17.2 mm/yr derived from published data in Table 1. Horizon Mean colony Estimated age diameter Top S1 90mm 5.2 years Top S2 47.5mm 2.8 years Top S3 113mm 6.6 years Top S4 62.5mm 3.6 years Top S5 130mm 7.5 years Acknowledgements The author thanks Bill Martindale for advice on localities. Paul Blanchon and Noel James are thanked for comments on the manuscript, and Mairi Best and co-convenors of this mini-symposium for their hard work. Financial support for fieldwork from The Royal Society is gratefully acknowledged. References Aronson RB, Precht WF (2000) Herbivory and algal dynamics on the coral reef at Discovery Bay, Jamaica. Limnol and Oceanog 45: Aronson RB, Precht WF (1997) Stasis, biological disturbance, and community structure of a Holocene coral reef. Paleobiol 23: Bak RPM, Engel MS (1979) Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia)

6 and the importance of life history strategies in the parent coral community. Mar Biol 54: Bythell JC, Gladfelter EH, Bythell M (1993) Chronic and catastrophic natural mortality of three common Caribbean reef corals. Coral Reefs 12: Curran HA, Smith DP, Meigs LC, Pufall AE, Greer ML (1994) The health and short-term change of two coral patch reefs, Fernandez Bay, San Salvador Island, Bahamas. In: Ginsburg R.N. (ed.) Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards and History. University of Miami, Miami: Done TJ (1992) Phase shifts in coral reef communities and their ecological significance. Hydrobiol 247: Gladfelter WB (1982) White band disease in Acropora palmata: implications for the structure and growth of shallow reefs. Bull Mar Sci 32: Gladfelter EH, Bythell JC, Gladfelter WB, Lewis SK, Woodbury M (1991) Ecological studies of Buck Island National Monument St. Croix, US Virgin Islands: a quantitative assessment of selected components of the coral reef ecosystems and establishment of long-term monitoring sites. Part 1. US Dept. of Interior, National Park Services, Special Report : 177 pp Goreau TJ (1992) Bleaching and reef community change in Jamaica: Am Zool 32: Highsmith RC, Riggs AC, D Antonio C.M (1980) Survival of hurricane-generated coral fragments and a disturbance model of reef calcification/growth rates. Oecol 46: Hubbard DK (1992) Hurricane-induced sediment transport in open-shelf tropical systems an example from St. Croix, U.S. Virgin Islands. J Sed Res 62: Hubbard DK, Parsons KM, Bythell JC, Walker ND (1981) The effects of Hurricane Hugo on the reefs and associated environments of St. Croix, U.S. Virgin Islands a preliminary assessment. J Coastal Res 8: Hughes TP (1985) Life histories and population dynamics of early successional corals. Proceedings of 5th International Coral Reef Symposium, Tahiti, 1: Hughes TP (1994) Catastrophes, phase shifts and largescale degradation of a Caribbean coral reef. Science 265: Hughes TP, Jackson JBC (1985) Population dynamics and life histories of foliaceous corals. Ecol Mono 55: Jackson JBC (1992) Pleistocene perspectives of coral reef community structure. Am Zool 32: James NP (1971) Late Pleistocene reef limestones, north Barbados, West Indies. unpublished Ph.D thesis. McGill University, Montreal: 242 pp Johnson JH (1961) Limestone-building algae and algal limestones. Johnson Publishing Company, Boulder, Colorado: 297 pp Macintyre IG (1988) Modern coral reefs of western Atlantic: new geological perspectives. Am Ass Petrol Geol Bull 72: Martindale W (1976) Calcareous encrusting organisms of the Recent and Pleistocene reefs of Barbados, West Indies. Unpublished Ph.D, University of Edinburgh: 156 pp Martindale W (1992) Calcified epibionts as palaeocological tools: examples from the Recent and Pleistocene of Barbados. Coral Reefs 11: Mesolella KJ (1967) Zonation of uplifted Pleistocene coral reefs on Barbados, West Indies Science 156: Perry CT (in press) Storm-induced coral rubble deposition: Pleistocene records of natural reef disturbance and community response. Coral Reefs. Rogers CS (1993) Hurricanes and coral reefs: the intermediate disturbance hypothesis revisited. Coral Reefs 12: Rogers CS, Suchanek TH, Pecora FA (1982) Effects of Hurricanes David and Frederic (1979) on shallow Acropora palmata reef communities: St. Croix, U.S. Virgin Islands. Bull Mar Sci 32: Rooney WS (1970) A preliminary ecologic and environmental study of the sessile foraminifer Homotrema rubrum (Lamarck). Bermuda Biological Station Special Publication 6: 7-18 Scoffin TP (1993) The geological effects of hurricanes on coral reefs and the interpretation of storm deposits. Coral Reefs 12: Scoffin TP, Hendry MD (1984) Shallow-water sclerosponges on Jamaican reefs and a criterion for recognition of hurricane deposits. Nature 307: Stoddart DR (1974) Post-hurricane changes on the British Honduras Reefs: re-survey of Proceedings of 2nd International Coral Reef Symposium, Brisbane 2: Woodley JD et al.(1981) Hurricane Allen s impact on Jamaican coral reefs. Science 214: Wray JL (1977) Calcareous Algae. Elsevier Scientific, Amsterdam: 185 pp