Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
J. Phycol. 34, 393–406 (1998) REVIEW CORAL REEFS: ALGAL STRUCTURED AND MEDIATED ECOSYSTEMS IN SHALLOW, TURBULENT, ALKALINE WATERS Walter H. Adey Marine Systems Laboratory, Smithsonian Institution, Washington, D.C. 20560; [email protected] high magnesium calcite far exceeding the others. Traditionally the deposition of calcium carbonate as external tests and skeleta has been thought of as energy driven, a specialized process that evolved as protection against predation or improvement in performance. More recent studies indicate that calcification often results from the high pH of tropical ocean seawater and the removal of CO2 by photosynthesis. Calcification could, in turn, be supportive of photosynthesis by locally converting seawater HCO 3 2 into scarce CO 2 (the preferred carbon source) (McConnaughey and Whelan 1996). Algae are responsible for most reef accretion of CaCO3, either directly (e.g. the green alga Halimeda, coralline red algae) or through photosynthesis-mediated deposition by the dinoflagellate Symbiodinium in stony corals or by diatoms and chlorophytes in foraminifera (Lee et al. 1989, Hallock and Peebles 1993; Fig. 1). Many algae in reefs are weakly calcified (Udotea, Penicillus, dasycladaceans, Liagora, Galaxaura, squamariaceans, Padina, etc.), and under sporadic weak wave and current situations, a ‘‘dust’’ of aragonite crystals is often found coating virtually all algal surfaces. The cyanobacteria-rich, algal turfs that dominate so many reefs (see below) can approach stromatolites in terms of photosynthesis-induced calcium carbonate build-up, and have been demonstrated recently to provide a major portion of buried reef carbonate as ‘‘micritic’’ interlayers with stony corals and coralline algal crusts (Camoin and Montaggioni 1994). In contrast, many species of macroalgae probably employ chemistries that prevent the initiation of aragonite crystals; this is particularly likely in high sun species with abundant tissue cavities (e.g. many members of the red algal order Rhodymeniales). The first reef boring, carried out from a back reef island at Funa-Futi Atoll (Sollas 1904), showed for the Pleistocene and Holocene that the biotic source of reef carbonates was, in order of volume: 1) coralline algae, 2) foraminifera, 3) Halimeda, and 4) scleractinian coral. In retrospect, for an open ocean, trade wind atoll with a well-developed algal ridge to seaward, it is reasonable that corallines would dominate the deposition of carbonate. Recently, submerged Halimeda-dominated bioherms lying in the lee of the primary reef line were documented in the northern section of the Great Barrier Reef (Orme and Salama, and Marshall and Davies in Roberts and Key index words: algae; calcification; community structure; coral reef; primary production; zooxanthellae With a million species or more (Reaka-Kudla 1996), and specific and phyletic biodiversities greater than that of any of earth’s other biomes (more than 80% being animal or protozoan), coral reefs would seem to be truly animal ecosystems, the antithesis of rain forests on land. Prior to mid-century, scientists generally assumed that the primary source of energy for coral reefs was derived from the plankton brought by ocean currents and trade wind seas. However, the largely algal photosynthetic component of coral reefs, which greatly exceeds the animal component in biomass, is mostly diminutive, often symbiotic within animal tissues, and easily overlooked. Odum and Odum (1955) demonstrated for an Eniwetok reef flat that on a biomass basis, algae dominated the reef flat; they invoked the ‘‘traditional’’ trophic pyramid of biomass with 85% algae, 13.8% herbivores, and 1.1% carnivores. However, most striking was the finding that only about 20% of that algal biomass was symbiotic (6% was macroalgae and 73% algal turfs, borers, and crusts). The reef literature has grown enormously since that time, along with our understanding of reef metabolism and community structure and function. However, the basic picture has not changed. Algae directly or indirectly construct and determine coral reefs. They also possess the potential to alter the character of a reef, given natural or human perturbation involving nutrients or grazing. Stony corals (scleractinians) have evolved a mechanism that allows extremely high rates of calcification, but that greatly limits the area covered by coral tissue to the pinnacles of a framework. Algae occupy the lateral surfaces of the framework and fill its holes with more carbonate. Increased nutrients or reduced grazing cause a ‘‘phase shift’’ of community structure, wherein framework construction by corals is greatly reduced and macroalgae come to dominate. These matters, as we have come to understand them in the last 20 years, are reviewed in brief below. Calcification. Over 50 minerals have been adapted by many phyla of living organisms for use as protective shells, internal skeletons, teeth, and sensory devices (Simkiss and Wilbur 1989). In the aquatic world, the carbonate minerals dominate this array with calcium carbonate as aragonite, calcite, and 393 394 WALTER H. ADEY MacIntyre 1988). These Halimeda bioherms have a Holocene thickness as great or greater than the windward reef and occupy an area that is many times larger. Even in a modern, coral-dominated reef, average coral coverage over many kilometers typically ranges from 10%–30% (Dullo et al. 1990), and the most predictable component is an algal turfdominated carbonate (Green et al. 1987). Reef bore holes rarely pass through solid coral for many meters, but more characteristically show 50%–80% carbonate-sand-filled cavities. Typically, corals build frameworks on a narrow band of reef with moderate wave energy, but the spaces are largely filled with an algal and foraminiferan sand. Deeper fore reefs, unless vertical walls with low accretion rates, are often dominated by coralline red algae (Dullo et al. 1990) and Halimeda, and the lee section of back reefs are the primary province of Halimeda or coralline algae. Significant biomineralization occurs under four key conditions of progressively greater organismic control: 1) enclosure by tissues (or within cells) of a volume of water in which ion concentrations can be elevated, 2) specific organic molecules to provide crystal initiation sites, 3) active transport of the requisite ions, and 4) an organic matrix to control crystalline shape or texture (Simkiss and Wilbur 1989). In Halimeda spp., only the first two processes occur, and disorientated aragonite crystals fill the interutricular spaces when CO2 is removed from those spaces by photosynthesis (Borowitzka 1987). Halimeda is probably the most important bulk calcifier in coral reef ecosystems at the generic level, although it rarely provides reef framework, but rather sedimentary infill within the reef framework or in the lagoon. Occurring in hard bottom benthic photic zones throughout the world ocean, the many genera and species of the Corallinales are important framework builders, framework cementers, and infill elements in coral reef environments (Bjork et al. 1995). Calcification in corallines is not as well understood as in Halimeda (see e.g. Lobban and Harrison 1994). However, it seems clear that corallines deposit high magnesium calcite in a definitive pattern within their cell walls via membrane pumps that control the flux of calcium and carbonate with control of crystal orientation by organic wall components. Thus, while corallines can be the deepest algae (Dullo et al. 1990) (and many species occur in great abundance into the arctic and subarctic), their proportional carbonate accretion rates are slower, 20%–50% of that in stony corals, though sometimes more than in Halimeda (Chisholm et al. 1990). Under high wave conditions, where the survival of neither the articulated Halimeda nor scleractinian corals is possible and where active grazing of the slowgrowing corallines is more limited, frameworks of crustose corallines become extensive. Algal ridges ring most Pacific Atolls, where trade wind sea energies are high; they also are widespread in the Ca- ribbean in those circumstances where the more limited trade wind seas are the highest (Adey 1978). The high magnesium calcite of coralline skeleta is clearly defensive. Nevertheless, there are grazers (primarily chitons and limpets; Littler et al. 1995) that bear a mutualistic relationship to their coralline hosts by preventing the establishment of the spores of macroalgae that would overgrow the corallines. In reef zones of moderate energy, scleractinian corals, with their zooxanthellae, are the primary producers of an aragonite framework. Recent research (McConnaughey and Whelan 1996) indicates that the basal disk cells of coral polyps pump calcium into and hydrogen ions out of the spaces that underlie the polyps, resulting in conditions for organic-molecule-mediated development of aragonite crystals. In the coelenteron, which acidifies as a result of calcification, bicarbonate ions (from seawater) react with hydrogen ions to produce CO2. Blockage of the process by accumulation of CO2 in the coelenteron is prevented by rapid CO2 removal through photosynthesis by zooxanthellae. This concept is in agreement with most studies that show that the majority of reef calcification occurs during daylight. For example, recent investigation of a reef on the Pacific island of Moorea, with a moderately high yearly calcification rate (8.9 kg CaCO3·m22·year21) showed that 73% occurred during daylight (Gattuso et al. 1993). It is important to note that the degree of enhancement of calcification by zooxanthellae likely varies widely, as does the calcification rate, in various genera and species of corals. Some scleractinian corals lack zooxanthellae and still calcify. However, these species calcify only slowly and are considered nonhermatypic (i.e. nonreef building). Although the coral symbiont Symbiodinium microadriaticum has been treated as a single species, the taxonomic relationships of zooxanthellae are uncertain (Buddemeier 1994, McNally et al. 1994, Rowan and Knowlton 1995, Rowan 1998). For this reason, in the remainder of this review, I will refer to the coral symbionts as Symbiodinium spp. In summary, the primary producers of carbonate, both framework and infill, in coral reefs are algae (including cyanophytes), or algal (zooxanthellae)mediated animals. Many of the secondary or tertiary animal (or protist) calcifiers also have algal symbionts (foraminifera, giant clams, etc.). In reef-building species of corals, extraordinarily high rates of calcification are achieved through the utilization of abundant seawater HCO32 and zooxanthellae to remove CO2. The basic quantitative/area relationships between calcium carbonate deposition in coral reef communities and the primary reef components responsible for that deposition and that occurring elsewhere in the world ocean are shown in Figure 2. Community structure. Coral reefs, as we treat them in this article, are recent, less than 10,000 years old, though often sited on antecedent reef platforms that have developed throughout the Pleistocene and CORAL REEFS: ALGAL ECOSYSTEMS 395 FIG. 1. Primary mechanisms of calcification in key reef-building algae and algal-mediated scleractinia. Modified from Borowitzka (1987) (Halimeda) and McConnaughey and Whelan (1996) (scleractinia); note that McConnaughey and Whelan refer to active transport as the ‘‘ping-pong pump.’’ 396 WALTER H. ADEY later Tertiary. By definition, significant past calcification and framework build-up (at least a few meters) must have occurred to differentiate a coral reef from a coral community. Two major patterns of reef development can be recognized: 1) immature or recently matured reefs, with high spatial heterogeneity (1.5–3.0 m2/m2), often 10–20 m thick (these are common in the Caribbean and less so in the IndoPacific); and 2) reefs that reached sea level 3000– 5000 years ago and have a low spatial heterogeneity (1–2 m2/m2) (these are abundant in the Indo-Pacific and scattered in the Caribbean) (Adey 1978, MacIntyre 1988) (note that units of spatial heterogeneity refer to the number of square meters of large scale and noncryptic carbonate surface area per horizontally projected square meter). Reefs (and portions of reefs) are structured by wave energy, with coralline algae dominating at high energies (algal ridges), corals dominating at intermediate energies, and Halimeda bioherms and coralline algae again at low wave energies. Coralline red algae (different genera and morphologies, and a few specialized species of Halimeda) can also dominate on deep banks and in fore reef zones (.60 m), if slopes are moderate (Dullo et al. 1990). This discussion specifically relates to the highly constructional, highly productive reef communities that occur in less than 20-m depths. Basic patterns of community structure are shown in Table 1, based on four ‘‘type’’ reefs. Except for spatial heterogeneity, community structure at the functional group level is quite similar between Pacific and Caribbean examples. Algal turfs are almost always the dominant reef community, followed by corals or macroalgae, depending upon the reef type. However, there are several omissions to these data that need to be considered. The reef crest (between fore reef and back reef) is not included because this is the surf zone where it is difficult to work and adequate quantitative data have not been collected. Here, the crustose corallines dominate, and under very high wave energies (scattered eastern Caribbean and most open Pacific atolls), the crustose corallines, often along with vermetid molluscs and sometimes Millepora spp., frequently create algal ridges. Thus, although corals appear to occupy onequarter to nearly one- third of reef surfaces, and corallines 7%–16%, a full section from 20-m depth in the fore reef to the beginning of the lagoon would balance out these figures at about 20% each. Also, data from the large Halimeda bioherms at the back of the main reef line in the northern Great Barrier Reef (as noted above) would markedly alter the proportions of macroalgae to corals in this very large system if it were to be included. The information cited above for sediment composition in lagoons and back reef flat borings suggests that although perhaps not so extreme, Halimeda dominance of back reef margins is common beyond where flat transects are typically terminated. Finally, high island fringing reef flats (Table 1; see also Adey and Burke 1976, Gattuso et al. 1997) often lose their coral and coralline coverage in favor of macroalgae. Fore reefs, in more open water, may retain coral cover, but still show an increase in macroalgae at the expense of other assemblages. An increase in macroalgae also follows eutrophication (see below). Primary productivity. In a recent review of reef gross primary production (GPP) relative to global warming and reef exploitation, Crossland et al. (1991) concluded that tropical reef communities worldwide (6 3 105·km2) fixed 700 3 1012·g C year21 (or 3.2 g C m22·day21). This is indeed relatively high for a biome scale ecosystem. However, based on the definition of coral reefs by Crossland et al., the surface covered ranged from the photic limit on the ocean side to shore (or photic limit on the other side of an atoll) and included many deep carbonate banks. Thus, for the systems we are discussing in this article, that figure is very much a minimum. In addition to information covering this rather expanded reef area, Crossland et al. (1991) also collated global data on reef flats to provide a mean GPP of 7 6 0.6 g C·m22·day21. The value that is appropriate for coral reef ecosystems as we circumscribed them above can be estimated to lie between that for their expanded reef area and the reef flat, at about 5 g C·m22·year21. The highest recently recorded level based on flow respirometry is 23.2 g C·m22·day21 (Adey and Steneck 1985—St. Croix). Crossland et al. (1991) point out that because community respiration (i.e. internal use of photosynthate) is typically high, net primary production (NPP; that which is either exported or available for export) is only about 3% of gross. However, reef flats across the Pacific were considerably higher than 3% of gross (1.0 6 0.1 g C). Our typical coral reef, from 20-m depths in the fore reef to the beginning of the mostly sandy lagoon, on the same basis as mean GPP, can be estimated to have a mean NPP of about 0.6 g C·m22·day21. If a reef is dominated by algae, and particularly if carbonate spatial heterogeneity remains high, levels of .10 g C·m22·day21 are possible. As a rough measure of the relationship between calcification and biomass production in branching hermatypic corals, we can take the data of Guillaume (1991) for Porites colonies on a reef flat in Moorea. At a calcification rate of 13 kg·m22·year21, for the dense colonies, and a mean yearly extension of 1 cm year21, and using a value of NPP available for biomass construction of ,0.5 g C·m22 (see above), less than 10% of the annual CaCO3 extension of a branch could be provided with new tissue. A more recent compilation by Lough and Barnes (1997) for Porites spp. in a wide variety of Indo-Pacific localities, based on extension rates and coral skeleton densities, shows even higher calcification/ extension rates that cannot be attributed to colony CORAL REEFS: ALGAL ECOSYSTEMS 397 FIG. 2. Comparison of productivity and global surface area of various carbonate environments. While coral reefs occupy only about 0.02% of the area of the deep sea, their rate of carbonate production (per unit area) is about 200 times greater than deep-sea carbonates. Other neritic environments also occupy small areas of the global ocean, but collectively they produce nearly as much carbonate as the deep sea. Total carbonate accumulation is about equally divided between shallow and deep-water environments. Note that total deposition in the deep sea (primarily by coccolithophores) approximately equals that in reefs and on shallow banks. (After Milliman, 1993). infill or other extraneous sources. Allowing for minimal predation, on average, the tissue of most branching and pillar-forming corals must continue to grow upward at the tips without significantly increasing biomass. It may well be that nutrients (perhaps through local plankton capture) are sufficient to build an initial biomass during colony initiation, typically within the algal turf microclimate. As the colony ‘‘climbs’’ into the water column, this advantage is lost. These hypotheses need testing, but they are particularly well demonstrated on most actively growing Caribbean Acropora palmata thickets. In these colonies, where the calcification can be determined for individual branch tips, living tissue is very thin, and basal ‘‘die back’’ is a conspicuous and regular feature (Adey and Steneck 1985). Coral reefs, given elevated levels of nutrients in the water column, can undergo a phase shift to fleshy, macroalga-dominated communities with high gross and net primary productivities (Done 1992, Hughes 1994). In this state, production of calcified framework is low. As a result of constant carbonate bioerosion, the reef begins to degenerate to a more or less flat pavement. In the more typical active reef that has high GPP and low net production and export, algal turfs and crusts on antecedent, coral, or coralline carbonate provide 30%–40% of the surface cover and biomass of a reef, and a considerably higher proportion of the primary production (see e.g. Williams and Carpenter 1990). Components of coral reef primary production. In a review of carbon and energy flux in stony corals, Muscatine (1990) concluded that the GPP of the coral/ dinoflagellate symbiosis (as a coral colony in situ) was typically between 2.0–2.6 g C·m22·day21 (with extremes of 1.4 and 13.7 g C·m22·day21). Carpenter (1985) and Larkum (1983) had earlier determined that reef algal turfs were in about the same range, perhaps a little higher at 2.3–3.3 g C·m22·day21 (Larkum data are recalculated as GPP and range from 1–6 g C·m22·day21). Klumpp and McKinnon (1992), in their extensive work on the Great Barrier Reef, found algal turf biomass levels (using similar methods) that were about the same (;30 g [dry]·m22) as Carpenter, but their productivities were a little lower (1.1–2.2 g C·m22·day21). This could be due to a failure to use surge stirring in their chambers (Carpenter et al. 1991). If the St. Croix cover data of Adey and Steneck (1985) are combined with the quite similar Great Barrier Reef data of Klumpp and McKinnon (1992) (as an approximate mean of open water coral reefs), and the productivity values cited above are applied to obtain mean component productivities, the zooxanthellae of scleractinian corals can be seen to provide about one-third of mean reef productivity and algal turfs approximately one-half. However, the whole reef summation is about 30% lower than the figure cited by Crossland et al. (1991). The productivity values for algal turfs cited here were derived from dried and slabbed carbonate plates allowed to redevelop algal turfs on a reef from 6 months to a year. The biomass levels of turf achieved are about 30 g dry·m22, about one-fifth of that found by Odum and Odum (1955) in their direct extraction analysis (148 g dry·m22). It seems likely that insufficient incubation time was allowed by these methodologies to develop a fully mature turf community, especially one with the boring components. The productivity values cited for algal turfs are probably low. Johnson et al. (1995), in applying Klumpp and McKinnon’s results, reached the same conclusion with regard to the relationship between component and whole production, but concluded instead that Muscatine’s 2.30 1.27 20.3 7.1 31.5 27.1 Stony corals 0.9 ,1 0.2 10.3 Other 1.5 2.3 2.8 26.5 Corallines Back reef (reef flat) 35.2 50.5 38.5 32.8 Algal turf 21.5 36.3 8.1 1.9 Macroalgae 20.5 3.8 18.9 1.2 Sand Algal ridges ;90% coralline ? 23 width back reef ;80% coralline ? Reef crest a 2.25 1.27 Spatial hetero. 25.9 21.1 22.7 Stony corals 6.9 5.4 2.2 Other 0.4 10.9 5.6 Corallines Fore reef 25.9 38.4 37.2 Algal turf 28.7 6.2 1.3 Macroalgae 12.4 14.4 26.2 Sand a Reef crests can be as wide or wider than fore reefs or back reef zones where wave energy is high and the crest lies at 0.5–3 m below mean low water. Once an algal ridge has formed, it typically narrows to ,10-m width. b Klumpp and McKinnon (1992). c Adey and Steneck (1985). d Morrissey (1980). e Adey et al. (1977). Martinique (three reefs)e High island reefs GBR Magnetic Isl (one reef)d Caribbean St. Croix (three reefs)c Shelf reefs Great Barrier Reef (seven reefs) Spatial hetero. TABLE 1. Comparative community structure Great Barrier Reef/Caribbean (% cover of carbonate surface). 398 WALTER H. ADEY CORAL REEFS: ALGAL ECOSYSTEMS numbers for primary production by corals must be too low. As we have noted, there is a strong basis for an alternate approach, which would elevate the algal turf component of whole reef GPP from about 50%– 60%. This approach is also the only way to achieve the relative increase in GPP resulting from change of coral cover to algal turf cover in maturing reefs of St. Croix (Adey and Steneck 1985). Grazing, nutrients, and reef health: the role of algal communities. It has become axiomatic in reef science that actively calcifying coral and algal turf-rich reefs can quickly (months to years) be transformed, by storm disturbance, increase of nutrients, or overfishing, to a high standing crop (.1 kg·m22) macroalgal-dominated community (Adey and Burke 1976, Kinsey 1988, Hughes 1994, Gattuso et al. 1997). Whether it is possible for a reef to return to the coral/algal turf-dominated state once it has been transformed is not known, but it is widely accepted that the transformation represents serious environmental degradation (Done 1992). The ‘‘delicate balance’’ between nutrients, grazing, and reef community structure has been extensively studied in the past decade or two. A brief discussion of each of the components is necessary to explore this issue. Johnson et al. (1995) developed the argument (based on Davies Reef, Great Barrier Reef) that a significant shift from coral to algal turf cover (e.g. in a reef with living coral removed by the crown-ofthorns starfish, e.g. Acanthaster or other means) would not significantly change the primary production but would greatly change trophic structure and carbon flux. Their reasoning is that production by the zooxanthellae is held within the reef, whereas algal turfs are grazed inefficiently, with much of the turf leaving the reef as current-carried detritus. This concept may be valid if largely inedible macroalgae replace the coral (rather than algal turfs). However, a very large proportion of the primary production of stony corals is released as mucous to the water column (Muscatine 1990). Thus, coral primary production would seem to be as much subject to loss from the reef community as algal turf fragments. An excessive shift from coral to algal turf is a serious issue for a coral reef community over the long term because coral calcification produces most of the framework that is responsible for spatial heterogeneity. Primary production, trophic structure and dynamics, and species diversity are all highly dependent upon spatial heterogeneity. Adey and Steneck (1985), based on St. Croix reefs, make the argument that, as a mature reef with high coral cover and high spatial heterogeneity approaches sea level, a more abundant living coral surface is replaced by algal turf. As a result, extremely high levels of primary and net productivity are achieved for short periods (centuries) with a full and complex trophic structure, until spatial heterogeneity is lost and reef pavements are formed. By contrast, the phase shift of coral and algal turf cover to macroalgae (due to el- 399 evated nutrients, the overharvest of herbivorous fish, or the disturbance of hurricanes) rapidly changes reef community structure and trophic dynamics. Although gross and net primary productivity may greatly increase (Adey et al. 1977, Guttuso et al. 1997), the net result is the loss of the reef as a community and the export of most of the productivity to lagoons or offshore to deeper water. Algal turfs dominate most actively calcifying coral reefs, covering more than 50% of the carbonate surface not occupied by living coral tissue. Table 2 lists species of each major algal group by anatomical type for algal turf assemblages in the Caribbean (see also Price and Scott 1992). It is not widely appreciated among most reef workers that in biomass, cyanobacteria, particularly Calothrix spp. and Schizothrix spp., are about as important as rhodophytes (e.g. smaller Jania spp., Amphiroa spp., Polysiphonia spp., Ceramium spp., and Herposiphonia spp.), at 30%–50% each. Chlorophytes and phaeophytes are lesser elements of algal turf, with species of Cladophora, Derbesia, Giffordia, and Sphacelaria being common. Cyanobacterial, red, and green borers are endolithic in the carbonate substrate, and it has long been known that these can extend under living coral tissue. Some workers have recently proposed the existence of a symbiotic association involving fungal filaments for this situation (Le Campion-Alsumard et al. 1995). Algal turf assemblages in coral reef environments are net nitrogen fixers and achieve high levels of primary production when strongly lit and provided with intense oscillation (wave surge) motion to break up boundary layers and overcome self shading (Hackney et al. 1989, Williams and Carpenter 1990, Carpenter et al. 1991, Carpenter and Williams 1993). Grazing by fish (especially by parrotfish, surgeonfish, and damselfish) and by a host of invertebrates (including echinoids, snails, crabs, and small crustaceans) is crucial to the maintenance of algal turfs by preventing the macroalgal sporelings, which are a persistent minor element of turf communities, from overgrowing and shading the smaller turfs. Nevertheless, fish grazers remove only about onehalf of NPP (Polunin and Klumpp 1992), the remainder being lost to invertebrate grazers or as algal fragments lost to the lagoon as a result of inefficient grazing. Many macro- and mesoinvertebrates, particularly echinoids, crustacea, gastropods, and annelids, graze on algal turfs, and because they are constantly preyed on by larger invertebrates and fish, are rarely able to reduce turf dry biomass below 50–200 g·m2. In some cases, invertebrates are more important (Carpenter 1990), and in others, fish are the dominant grazers (McClanahan 1992). It is clear that algal community structure is greatly affected by the spectrum of grazers present on an individual reef, and, in turn, fishing pressure can significantly alter the type of grazing pressure (McClanahan 1997). Macroalgae are minor elements in the best devel- 400 WALTER H. ADEY oped coral reef communities. Nevertheless, their role and performance is interesting for two critical reasons: 1) given elevated nutrients or reduced grazing, they are capable of overgrowing both corals and algal turfs, significantly degrading community structure (Done 1992, Naim 1993, Hughes 1994, Tanner 1995); and 2) the algal defense mechanisms, both physical and chemical, that allow many macroalgae to maintain a presence, are significant to the understanding of community dynamics, and in the case of the secondary compounds, offer potentially valuable chemicals to human societies. It is important to note that recent studies indicate that not only do elevated nutrients increase macroalgae and reduce coral settlement and cover, but they also directly reduce coral calcification (Marubini and Davies 1996). This phenomenon is not fully understood at this point, but it probably relates to the dynamic metabolic interaction between the animal coral cells and the zooxanthellae. Zooxanthellae receive their CO2 in part from animal (coral cell) respiration but also from calcification. Nitrogen (and phosphorus) is normally received from coral cells, probably originally derived primarily from captured plankton. Roughly 80% of the photosynthate produced by the zooxanthellae is transferred to the host as glycerol, but coral cells may mediate this transfer by controlling the nitrogen delivered to the zooxanthellae (Muscatine 1990). Zooxanthellae that receive very limited nutrients can photosynthesize and respire but are restricted in growth, so most of their photosynthate is expelled. Excess nitrogen in the form of nitrate appears to upset this balance, allowing zooxanthellae to grow and reducing glycerol transfer to the host (Marubini and Davies 1996). Thus, it seems likely that host metabolism and ultimately the energy available for active transport of calcium and hydrogen ions is reduced. Coral reef macroalgae can provide resistance to grazing by both physical (Littler et al. 1983, Coen and Tanner 1989) and chemical means. Perhaps in part because of the potential human pharmacopeia interest, considerable scientific effort is being directed to understanding the nature and function of such secondary compounds (though much of it is still confusing; Meyer and Paul 1992, Lumbang and Paul 1996). In contrast, there is little question that the large, green, leafy seaweed Avrainvillea longicaulis is defended from fishes by the brominated compound avrainvilleol (Hay et al. 1990). In this case, the value of the compound is extended, in that several invertebrates feed on the alga and additionally derive protection either from the presence of the compound in their own tissues or in the surrounding algae when used for cover. In other cases, it would seem that multiple mechanisms, including calcification, toughness, shape, and secondary metabolites, offer protection from herbivores (Schupp and Paul 1994). Much additional work is needed to gain a comprehensive view of the role of secondary compounds in the macroalgae of reef ecosystems. However, several general points can be made: 1) both physical characteristics (e.g. calcification) and secondary compounds can provide some protection from grazing, especially by fish; 2) no method provides complete protection and ultimately the principles of the ‘‘arms race’’ apply; and 3) defenses require energy and often nutrients, or in other ways reduce growth or reproductive potential. In the extreme oligotrophic reef environment, particularly above the ‘‘microclimate’’ of the algal turf, algal growth will always be limited to some degree by lack of nutrients. Therefore, an increase of water column nutrients is likely to offset potential restrictions to growth incurred by defense against herbivory. Both a reduction in grazing (by fishing or hurricanes) or an increase in nutrients (by hurricanes or cultural eutrophication) will increase macroalgal growth and physically reduce coral calcification and settlement (also see Tanner 1995). The chemical defenses against grazing of many macroalgae and the role of nutrients in causing a phase shift of dominance of macroalgae over corals and coralline algae have been discussed. However, we have barely begun to scratch the surface of the role of chemical cues in controlling coral reef community structure, as is seen by the recent discovery of the role of coralline algae in controlling the settlement of coral larvae (Morse et al. 1994). Summary. Coral reefs are benthic communities in high light, low nutrient, low plankton, moderate to high wave and current energy open waters, where physical factors show little annual variability. As in most shallow benthic communities not subject to severe physical limitations, competition for space is a significant factor, and yet, since oceanic plankton provides only a limited food/energy resource, the filter feeding barnacles, mussels, hydroids, and bryozoans that can occupy large areas on temperate/ boreal shores are of minimal importance. Scleractinian corals become a significant part of an ecosystem in which they would otherwise be highly restricted because of symbiotic association with dinoflagellate algae. By evolving a mechanism to utilize the abundant bicarbonate for calcification, at least the dominant hermatypic genera became highly competitive for space with the highest organismic mineralization rate yet developed in the earth’s biosphere. So efficient is this mechanism that it far outruns the concomitant growth potential of the living tissue. Limited in nitrogen and phosphorus supply, most of the zooxanthellar production is therefore released as mucous. Thus, at the community level, hermatypic corals build pedestals that continually raise their biomass to the sun, but leave behind a large and spatially heterogeneous mass of calcium carbonate; this results in an even greater area for algal growth than would otherwise be available to CORAL REEFS: ALGAL ECOSYSTEMS those algae that would occupy this community in the absence of coral construction. The ‘‘pedestals’’ can form a framework; however, numerous other calcifiers, mostly algae such as Halimeda or non-coral animals, many also with symbiotic algae, provide the infill. The carbonate-based, spatial heterogeneity, which is significantly further increased by the boring of algae and invertebrates, also provides cover for the high biodiversity characteristic of these ecosystems. Specifically, it provides cover for numerous fish and invertebrate grazers, which would otherwise be limited in abundance by lack of cover. Thus, in a stony– coral framework reef, in which three-quarters of the surface is not occupied by living coral, algal turfs, which have low biomass in most shallow, benthic environments, come to dominate much of the framework’s surface. This is equivalent to some grasslands in terrestrial environments and owes its continued existence to the grazers with which, in a sense, it holds a ‘‘symbiotic’’ relationship. The dominance of the algal turf is enhanced in this nutrient-poor environment in that nitrogen-fixing cyanobacteria are a significant component. Also, since most grazers remain under cover in the community when not feeding, much of the nutrients in the system are continually recycled. Energy flux is high, but exportable production is low, generally more or less in balance with incoming planktonic biomass. In the ecosystem described above, many macroalgae are limited by intensive herbivory to small refuges around toxic invertebrates or to sites where wave action or exposure to heavy predation limit grazing. Others survive by developing a variety of strategies that deter grazing to various degrees. Thus, unless the basic environmental parameters are upset, macroalgae are a third, minor component of the coral reef ecosystem. By contrast, given a natural or anthropogenic increase of nutrients (which both increase algal growth rate and decrease calcification) or a reduction in grazing (by fishing or storms), the macroalgal component can and does overpower the reef community. When this phase shift happens, macroalgal standing crops can become very high and productivity even higher than on the previously existing reef. However, much of the algal production is lost from the reef. As algal fragments float to the lagoon and net calcification is lost, the ‘‘reef’’ floor gradually becomes a flat pavement, in part due to boring and scraping, with a relatively low faunal diversity. An extremely turbulent, wave-breaking zone (the reef crest) exists for most open water reef systems that grow to within 5–10 m of the surface. In this zone, grazing is very limited, coral colonies are consistently broken or damaged by wave action, and crustose corallines (e.g. Porolithon, Lithophyllum and Neogoniolithon), with one-fifth to one-tenth of the calcification potential of scleractinian corals, build significant calcified structures (algal ridges). 401 Question (Rowan): One aspect of coralline red algae puzzles me. Many studies strongly imply that they induce the settlement of coral (and other) larvae that, once settled, will certainly compete with the algae for resources. This doesn’t seem to be in an alga’s best interest, to be a ‘‘settling plate’’ for its competitors. Do the algae have no other choice? Then again, maybe the corals will eventually provide substrate for the coralline algae either directly or indirectly, by promoting reef growth in general. Could it be that coralline algae actually ‘‘farm’’ future real estate? Answer: It has been recognized for some 35 years that crustose corallines have an intercalary meristem, like a cambium, that produces a usually colorless epithallium upward and the main tissue of the alga, perithallium, downward. At least since my Ph.D. thesis, in the early 1960s, it was known that the epithallium constantly sloughs off as an antifouling device, and more recently, Keats et al. (1997), reviewing the issue, demonstrated two modes of shedding, reiterating the antifouling property. Also, since my early publications, it has been recognized that some genera delay release of the epithallial cells producing a specialized photosynthetic zone of lesser calcification. Steneck (e.g. 1985) has suggested that these genera require grazing of this thickened epithallium to prevent fouling by other organisms, though Keats et al. question that grazing is really necessary in this case. Thus, crustose corallines constantly shed their surfaces and tend to be free of fouling. However, surface damage, often by grazing through the meristem, damages or kills spots that are then settled on by other algae and invertebrates. Under some conditions, a ‘‘brainlike’’ irregular surface of thick coralline crust forms, in which the ‘‘infolds’’ are continuously occupied by an algal turf community. This is probably not advantageous in any way to the corallines, but rather demonstrates the imperfection of all evolved defenses. As you have noted, it has also been known for some time that some invertebrates key into coralline bottoms for larval settlement. If the corallines are constantly releasing necrotic tissue, a plethora of organic compounds are certainly released, allowing ample opportunity for species to evolve that can recognize this bottom. But why key onto a bottom that you cannot land on (or at least successfully establish upon)? And finally, my anecdotal, or at least hypothetical, answer. Coralline surfaces are largely free of other organisms, but there are damaged ‘‘niches’’—mostly created by grazers—that could be occupied. It is probably worth it at the population level for the settling invertebrate larva to go for the small chance of an open hole. Having found this ‘‘niche,’’ it is virtually free of space competition because corallines are slow growers and the settlement of other larvae and algae is limited. Does this benefit the cor- Isthmia sp. Navicula sp. Nitzchia sp. Synedra sp. Numerous unidentified spp. Spores and sporelings Pseudotetraspora antillarum Chlorophyta (green algae) Anacystis dimiduata Additional cocooid forms Bacillariophyceae (diatoms) [Cyanobacteria] Cyanophyceae (blue-green algae) Unicellular Chaetomorpha* minima Valonia Thalassionema sp. Tabellaria sp. Anabaena oscillarioides Calothrix crustacea Entophysalis sp. Microcoleus lyngbyacea Nostoc spumigena Oscillatoria submembranacea Schizothrix calcicola Schizothrix mexicana Simple Branching Acetabularia1 Avrainvillea1 Bryopsis pennata Bryopsis sp. Chamaedoris penicilum Cladophora laetevirens Cladophora montagneana Cladophora vagabunda Cladophoropsis* macromeres Cladophoropsis membranacea Derbesia marina Ernodesmus spp. Neomeris spp.1 Penicillus spp.1 Pseudobryopsis spp. Rhizoclonium spp. Struvea spp. Udotea spp.1 Licmophora sp. Scytonema hofmannii Filamentous Enteromorpha chaetomorphoides Enteromorpha sp. Protoderma marinum Multiseriate, polysiphonous, or weakly corticated Complexly corticated Pseudoparenchymatous Caulerpa spp. Codium Dictyosphaeria Halimeda1 Ulva Parenchymatous TABLE 2. Genera and species consistently present in numerous surveys of Caribbean and western tropical Atlantic coral reefs. Species listed in regular type are common, persistent components of the coral reef algal turf assemblage. Genera in bold contain no persistent turf components and are treated as macroalgae. Crustose coralline genera are indicated in parenthesis. Asterisks designate genera that contain both the listed turf component species and other species that do not persist within the assemblage. Plus signs designate calcareous genera. After Adey and Hackney (1989). 402 WALTER H. ADEY Spores and sporelings Rhodophyta (red algae) Unicellular Spores and sporelings Phaeophyceae (brown algae) TABLE 2. Continued. Erythrotrichia carnea Simple Branching Acrochaetium spp. Antithamnion antillarum Antithamnion sp. Asterocytis ramosa Bangiopsis subsimplex Callithamnion halliae Callithamnion sp. Crouania attenuata Dohrniella antillarum Erythrocladia subintegra Goniotrichum alsidii Grallatoria reptans Griffithsia barbata Griffithsia globulifera Griffithsia schousboei Gymnothamnion elegans Halodictyon mirabile Pleonosporium caribaem Spermothamnion investiens Bachelotia antillarum Ectocarpus elachistaeformis Giffordia rallsiae Filamentous furcigera fusca tribuloides spp. Bangia atropurpurea Ceramium brevizonatum Ceramium byssoideum Ceramium comptum Ceramium corniculatum Ceramium cruciatum Ceramium fastigiatum Ceramium flaccidum Ceramium leptozonum Ceramium leutzelburgii Ceramium nitens Dasya* rigidula Dasya sp. Dasyopsis antillarum Falkenbergia hillebrandii Herposiphonia pectenveneris Herposiphonia secunda Herposiphonia tenella Herposiphonia sp. Heterosiphonia wurdemanni (Hydrolithon1) Hypoglossum (Lithophyllum1) (Lithothamnium1) Lophosiphonia cristata Martensia (Neogoniolithon 1) Peysonellia spp.1 Polysiphonia atlantica Polysiphonia binneyi Polysiphonia denudata Polysiphonia exilis Polysiphonia ferulacea Polysiphonia flaccidisima Sphacelaria Sphacelaria Sphacelaria Sphacelaria Acanthophora spp. Amphiroa1* fragillisima Amphiroa spp. Asparagopsis Bryothamnion Centrocerus clavulatum Champia parvula Chondria* collinsiana Coelothrix irregularis Digenia simplex Eucheuma spp. Galaxaura spp.1 Gelidiella* trinitatensis Gelidiella sanctarum Gelidiopsis intricata Gelidium* pusillum Gracilaria spp. Halymenia spp. Hypnea spp. Jania1* adherens Jania capillacea Jania pumila Jania rubens Laurencia * caraibica Liagora spp.1 Lomentaria spp. Meristotheca spp. Pterocladia spp. Wurdemannia miniata Complexly corticated Pseudoparenchymatous Multiseriate, polysiphonous, or weakly corticated Parenchymatous Colpomenia spp. Dictyopteris spp. Dictyota spp. Dilophus spp. Lobophora spp. Padina spp. Sargassum spp. Turbinaria spp. CORAL REEFS: ALGAL ECOSYSTEMS 403 404 WALTER H. ADEY allines in any way? Probably, in most cases, no. Perhaps in algal ridges and cup reefs, it could be argued that the frequent coinhabitors, vermetid gastropods, provide structural strength to the ridge, providing for community longevity (and greater spore production by the corallines). Perhaps coralline surfaces with vermetids, barnacles, or spirorbids are harder to graze and in the end more productive of spores because most conceptacles are not being removed by grazers. Now we are out on the hypothetical limb. Dunlap and Shick: Given that carbonate deposits store nearly 3 3 104 times more inorganic carbon than is in the atmosphere, that vast carbonate deposits have accumulated since the Precambrian (evidence of long residency), and that the bulk of carbonate rock is the result of biological precipitation, it is clear that calcification has had a profound geologic effect in shaping our present biosphere. The questions I present relate to the role of reef calcification on global C/CO2 cycling with consequence to atmospheric CO2 (‘‘greenhouse gas’’) levels due to chemical equilibria and exchange across the airsea interface. Question (Dunlap and Shick): There is ongoing debate as to whether coral reefs contribute in the present geologic as a source or sink of atmospheric CO2 (Buddemeier 1996, Gattuso et al. 1996, Kayanne 1996). Would you comment on the scientific basis of this debate? If coral reefs function as either a source or sink of atmospheric CO2, is it likely to be significant at the global scale? Question (Dunlap and Shick): It has been estimated that the ratio of released CO2/precipitated CO2 by calcification is approximately 0.6 (accounting for the buffering capacity of seawater) (Frankignoulle et al. 1994). This suggests that community metabolism in a heavily calcifying reef would cause an evasion of CO2 to the atmosphere. However, net CO2 consumption is an apparent feature of macroalgaldominated reefs (Gattuso et al. 1997). Would you comment on how community structure could influence metabolic equilibria in favor of CO2 release or consumption? Answers: These questions probably cannot be answered definitively today, but I will at least try to place them in perspective, from my point of view. If our averages for pantropic, coral reef function, are correct, globally, reefs are slight net primary producers and large calcium carbonate sinks. The net primary production is, at least in part, lost to sediments in lagoons and to some extent lost to the deep ocean. On the calcification side, reefs are massive world-scale net sinks of carbon. The issues relate more to where the carbon comes from and what the time scales of the fluxes are. Primary production carbon is in large measure derived, in more or less short term, from CO2 (ultimately, from the atmosphere), although some almost certainly originates from alkalinity. Macroalgal- dominated reefs are strong CO2 sinks. What percentage of pantropic reefs are macroalgal dominated? Probably less than 20%, and let’s hope that modern human society does not accept that reef eutrophication is a good thing in order to provide a new CO2 carbon sink. In the larger scale, as long as most reefs retain their recent past community structure, atmospheric CO2 sinks by coral reefs would be negligible, and the biodiversity loss associated with the macroalgal phase shift enormous. Coral/algal turf ecosystems, while very large recyclers of carbon, are not likely significant sinks of CO2 carbon. The carbon in coral reef calcification is primarily derived from alkalinity (HCO32, CO35). Some of this calcification by stony corals is derived from bicarbonate. In my view, only a few rapidly calcifying genera (including acroporids) use bicarbonate as a calcification substrate. However, this is not well established or even generally accepted at this time. Bicarbonate calcification results in some CO2 release to the atmosphere, and some rapidly calcifying reefs, rich in acroporids, undoubtedly transfer CO2 carbon to the atmosphere from the bicarbonate pool. In my anecdotal view, this is not likely to be a significant transfer on a pantropic scale. However, we need better large-scale information on coral calcification and reef community structure to be certain. I tried to outline where we stand on this issue in the review. In summary, coral reefs are clearly a significant sink, on the global scale, of oceanic alkalinity. Alkalinity carbon is ultimately carbon derived from atmospheric carbon through alkalinity pumps (see e.g. Kemp and Kazmierczak 1994). However, this is probably not of significance relative to our shortterm concern (decades to centuries) for fossil carbon introduction into earth’s atmosphere. Adey, W. 1964. The Genera Phymatolithon in the Gulf of Maine. Hydrobiology 24:377–420. 1965. The Genus Clathromorphum in the Gulf of Maine. Hydrobiology 26:539–73. 1966. The Genera Lithothamnium, Leptophytum and Phymatolithon in the Gulf of Maine. Hydrobiology 28:321–68. 1978. Coral reef morphogenesis: a multidimensional model. Science 202:831–7. Adey, W., Adey, P., Burke, R. & Kaufman, L. 1997. Holocene reef systems of eastern Martinique. Atoll Res. Bull. 218:1–40. Adey, W. & Burke, R. 1976. Holocene bioherms (algal ridges and bank barrier reefs) of the eastern Caribbean. Geol. Soc. Am. Bull. 87:95–109. Adey, W. & Hackney, J. 1989. The composition and productivity of tropical marine algal turfs in laboratory and field experiments. In Adey, W. & Farrier D. [Eds.] The Biology, Ecology and Mariculture of Mithrax spinosissimus Utilizing Cultured Algal Turfs. Mariculture Institute. Smithsonian Institution, Washington, D.C., pp. 1–83. Adey, W. & Steneck, R. 1985. Highly productive eastern Caribbean reefs: synergistic effects of biological, chemical, physical and geological factors. Symp. Ser. Underseas Res., NOAA Underseas Res. Prog. 3:164–87. Bjork, M., Mohammed, S., Bjorkland, M. & Semesi, A. 1995. Coralline algae, important coral reef builders threatened by pollution. Ambio 24:502–5. CORAL REEFS: ALGAL ECOSYSTEMS Borowitzka, M. 1987. Calcification in algae: mechanisms and the role of metabolism. CRC Crit. Rev. Plant Sci. 6:1–45. Buddemeier, R. 1994. Symbiosis, calcification and environmental interactions. Bull. Inst. Ocean. Monaco 13:119–35. Buddemeier, R. M. 1996. Coral reefs and carbon dioxide. Science 271:1298–9. Camoin, G. & Montaggioni, L. 1994. High energy coralgal-stromatolite frameworks form Holocene reefs. Sedimentology 41: 655–76. Carpenter, R. 1985. Relationships between primary production and irradiance in coral reef algal communities. Limnol. Oceanogr. 30:784–93. 1990. Mass mortality of Diaderma antillarum. I. Long term effects in sea-urchin population dynamics and coral reef algal communities. Mar. Biol. 104:67–77. Carpenter, R., Hackney, J. & Adey, W. 1991. Measurements of primary productivity and nitrogenase activity of coral reef algae in a chamber incorporating oscillatory flow. Limnol. Oceanogr. 36:40–9. Carpenter, R. C. & Williams, S. L. 1993. Effects of algal turf canopy height and microscale substratum topography on profiles of flow speed in a coral forereef environment. Limnol. Oceanogr. 38:687–94. Chisholm, J., Collingwood J. & Gill, E. 1990. A novel in-situ respirometer for investigating photosynthesis and calcification in crustose coralline algae. J. Exp. Mar. Biol. Ecol. 141:15–30. Coen, L. & Tanner, C. 1989. Morphological variation and differential susceptibility to herbivory in the tropical brown alga Lobophora variegata. Mar. Ecol. Prog. Ser. 54:287–98. Crossland, C., Hatcher, B. & Smith, S. 1991. Role of coral reefs in global ocean production. Coral Reefs 10:55–64. Done, T. 1992. Phase shifts in coral reef communities and their ecological significance. Hydrobiologia 247:121–32. Dullo, W-C., Moussavian, E. & Brachert, T. 1990. The foralgal crust facies of the deeper fore reefs in the Red Sea. Geobios 23:261–81. Frankignoulle, M., Canon, C. & Gattuso, J.-P. 1994. Marine calcification as a source of carbon dioxide: positive feedback on increasing atmospheric CO2. Limnol. Oceanogr. 39:458–62. Gattuso, J., Pichon, M., Delesalle, B. & Frankignoule, M. 1993. Community metabolism and air-sea carbon dioxide fluxes in a coral reef ecosystem (Moorea, French Polynesia). Mar. Ecol. Prog. Ser. 96:259–67. Gattuso, J.-P., Frankignoulle, M., Smith, S. V., Ware, J. R. & Wollast, R. 1996. Coral reefs and carbon dioxide. Science 271: 1298. Gattuso, J.-P., Payri, C., Pichon, M., Delesalle, B. & Frankignoulle, M. 1997. Primary production, calcification and air-sea CO2 fluxes of a macroalgal-dominated coral reef community (Moorea, French Polynesia). J. Phycol. 33:729–38. Green, D., Bradbury, R. & Reichelt, R. 1987. Patterns of predictability in coral reef community structure. Coral Reefs 6:27– 34. Guillaume, M. 1991. Growth and calcium carbonate production of massive Porites (barrier reef flat of French Polynesia). Proc. ISRS Congress, Noumea 1990:87–90. Hackney, J., Carpenter, R. & Adey, W. 1989. Characteristic adaptions to grazing among algal turfs on a Caribbean coral reef. Phycologia 28:109–19. Hallock, P. & Peebles, M. 1993. Foraminifera with chlorophyte endosymbionts: habits of six species in the Florida Keys. Mar. Micropaleontol. 20:277–92. Hay, M., Duffy, J., Paul, V., Renaud, P. & Femcal, W. 1990. Specialist herbivores reduce their susceptibility to predation by feeding on the chemically defended seaweed Avrainvillea longicaulis. Limnol. Oceanogr. 35:1734–43. Hughes, T. 1994. Catastrophes, phase shifts and large scale degradation of a Caribbean coral reef. Science 265:1547–51. Johnson, C., Klumpp, D., Fied, J. & Bradbury, R. 1995. Carbon flux on coral reefs: effects of large shifts in community structure. Mar. Ecol. Prog. Ser. 126:123–43. Kayanne, H. 1996. Coral reefs and carbon dioxide. Science 271: 1299–300. 405 Keats, D., Knight, M. & Pueschel, C. 1997. Antifouling effects of epithallial shedding in three crustose coralline algae (Rhodophyta, Coralinales) on a coral reef. J. Exp. Mar. Biol. Ecol. 213:281–93. Kemp, S. & Kazmierczak, J. 1994. The role of alkalinity in the evolution of ocean chemistry, organization of living systems, and biocalcification processes. Bull. Inst. Ocean. Monaco 13: 61–117. Kinsey, D. 1988. Coral reef systems response to some natural and anthropogenic stresses. Galaxea 7:113–28. Klumpp, D. & McKinnon, A. 1992. Community structure, biomass and productivity of epilithic algal communities on the Great Barrier reef; dynamics at different spatial scales. Mar. Ecol. Prog. Ser. 86:77–89. Larkum, A. 1983. The primary productivities of plant communities on coral reefs. In Barnes, D. [Ed.] Perspectives on Coral Reefs. Australian Institute of Marine Science, Townsville, Australia, pp. 221–30. Le Campion-Alsumard, T., Golubic, S. & Priess, K. 1995. Fungi in corals; symbiosis or disease. Mar. Ecol. Prog. Ser. 117:137– 47. Lee, J., McHenry, M., Ter-Kuile, B., Eriz, J., Roettger, R., Rockwell, R., Faber, W. & Lagziel, A. 1989. Identification and distribution of endosymbiotic diatoms in larger foraminifera. Micropaleontology 35:353–66. Littler, M., Littler, D. & Taylor, P. 1995. Selective herbivore increases biomass of its prey; a chiton-coralline reef-building association. Ecology 76:1666–81. Littler, M., Taylor, P. & Littler, D. 1983. Algal resistance to herbivory on a Caribbean coral reef. Coral Reefs 2:111–8. Lobban, C. & Harrison, P. 1994. Seaweed Ecology and Physiology. Cambridge University Press, Cambridge, 366 pp. Lough, J. & Barnes, D. 1997. Several centuries of variation in skeletal extension, density and calcification in massive Porites colonies from the Great Barrier Reef: a proxy for seawater temperature and a background of variability against which to identify unnatural change. J. Exp. Mar. Biol. Ecol. 211:29–67. Lumbang, W. & Paul, V. 1996. Chemical defenses of the tropical green seaweed Neomeris annulata. J. Exp. Mar. Biol. Ecol. 201: 185–95. MacIntyre, I. 1988. Modern coral reefs of the western Atlantic: new geological perspective. Bull. Am. Assoc. Petrol. Geol. 72: 1360–9. Marubini, F. & Davies, P. 1996. Nitrate increases zooxanthellae population density and reduces skeletogenesis in corals. Mar. Biol. 127:319–28. McClanahan, T. 1992. Resource utilization, competition, and predation: a model and example from coral reef grazers. Ecol. Model. 61:195–215. 1997. Primary succession of coral-reef algae: differing patterns on fished versus unfished reefs. J. Exp. Mar. Biol. Ecol. 218:77–102. McConnaughey, T. & Whelan, J. 1996. Calcification generates protons for nutrient and bicarbonate uptake. Earth Sci. Rev. 967:1–23. McNally, K., Govind, N., Thomé, P. & Trench, R. 1994. Smallsubunit ribosomal DNA sequence analysis and a reconstruction of the inferred phylogeny among symbiotic dinoflagellates (Pyrrophya). J. Phycol. 30:316–29. Meyer, K. & Paul, V. 1992. Intraplant variation in secondary metabolite concentration in three species of Caulerpa and its effects on herbivorous fishes. Mar. Ecol. Prog. Ser. 82:249–57. Milliman, J. 1993. Production and accumulation of calcium carbonate in the ocean: budget of a nonsteady state. Global Biogeochem. Cycles 7:927–57. Morrissey, J. 1980. Community structure and zonation of macroalgae and hermatypic corals on a fringing reef flat of Magnetic Island (Queensland, Australia). Aquat. Bot. 8:91–139. Morse, D., Morse, A. N. C., Raimondi, P. & Hooker, N. 1994. Morphogen-based chemical flypaper for Agaricia humilis coral larvae. Biol. Bull. 186:172–81. Muscatine, L. 1990. The role of symbiotic algae in carbon and 406 WALTER H. ADEY energy flux in reef corals. In Dubinsky, Z. [Ed.] Ecosystems of the World. 25: Coral Reefs. Elsevier, New York, pp. 75–87. Naim, O. 1993. Seasonal responses of a fringing reef community to eutrophication, Reunion Island. Mar. Ecol. Prog. Ser. 99: 137–51. Odum, H. & Odum, E. 1955. Trophic structure and productivity of a windward coral reef community on Eniwetok Atoll. Ecol. Monogr. 35:291–320. Polunin, N. & Klumpp, D. 1992. Algal food supply and grazer demand in a very productive coral reef zone. J. Exp. Mar. Biol. Ecol. 164:1–15. Price, I. & Scott, F. 1992. The Turf Algal Flora of the Great Barrier Reef: I. Rhodophyta. James Cook University, Townsville, Australia, 266 pp. Reaka-Kudla, M. 1996. The global biodiversity of coral reefs. A comparison with rain forests. In Reaka-Kudla, M., Wilson, D. & Wilson, E. O. [Eds.] Biodiversity II, Understanding and Protecting our Natural Resources. Joseph Henry/National Academic Press, Washington, D.C., pp. 83–108. Roberts, H. & MacIntyre, I. 1988. Halimeda. Coral Reefs 6:123– 279. Rowan, R. 1998. Diversity and ecology of zooxanthellae on coral reefs. J. Phycol. 34:407–17. Rowan, R. & Knowlton, N. 1995. Intraspecific diversity and ecological zonation in coral-algal symbiosis. Proc. Nat. Acad. Sci. 92:2850–3. Schupp, P. & Paul, V. 1994. Calcium carbonate and secondary metabolites in tropical seaweeds; variable effects on herbivorous fishes. Ecology 75:1172–85. Simkiss, K. & Wilbur, K. 1989. Biomineralization. Academic Press, San Diego, 337 pp. Smith, S. V. & Kinsey, D. 1976. Calcium carbonate production, coral reef growth, and sea level change. Science 194:937–9. Sollas, W. 1904. The Atoll of Funa Futi, Borings into a Coral Reef and the Results. Report of the Coral Reef Committee, Royal Society of London, London, 428 pp. Steneck, R. 1985. Adaptations of crustose coralline algae to herbivory: patterns in space and time. In Toomey, D. & Nitecki, M. [Eds.] Contemporary Research and Applications. Springer-Verlag, Berlin, pp. 352–6. Tanner, J. 1995. Competition between scleractinian corals and macroalgae: an experimental investigation of coral growth, survival and reproduction. J. Exp. Mar. Biol. Ecol. 190:151–68. Williams, S. & Carpenter, R. 1990. Photosynthesis/photon flux density relationships among components of coral reef algal turfs. J. Phycol. 26:36–40.