Marine Ecology
Chapter 20 Coral Reefs: limiting factors, morphology, and nutrition of corals
J.S. Levinton, 1982, Marine Ecology, Prentice-Hall Inc., New Jersey, p.394-418; 419-444.
List of Figures
20-1 INTRODUCTION, DEFINITIONS AND LIMITING FACTORS
Coral reefs are wave-resistant structures notable for their great species richness, topo-graphic complexity, and remarkable beauty. Coral reefs are common in
clear open marine water throughout the tropics. Massive reef accretion can be ascribed to the production of calcium carbonate by scleractinian corals and
crustose coralline algae; numerous other calcium carbonate-producing algae and invertebrates also contribute to reef growth. The aggregate activities of these
organisms has produced the Great Barrier Reef, a 1950-km-long ribbon of coral reefs capped by small tropical islands and stretching along much of the east
coast of Australia. This complex of reefs protects the coast of eastern Australia from the wave energy of the Pacific.
The world coral reef biota can be divided into (Fig. 20-1) Atlantic and Indo-Pacific biogeographic provinces, which probably differentiated from a
pantropical province in the mid-Miocene. The Indo-Pacific province differs from the Atlantic in
1. Its higher diversity of corals and most associated reef groups.
2. The presence of atolls or rings of islands capping submarine volcanoes, which are rare in the Atlantic province.
3. Extensive development of rich coral populations on intertidal reef flats, with poor intertidal coral development in the Atlantic province.
4. Differences in dominance of some groups.
Details on diversity differences are discussed at the end of Chapter 21.
Diving operations have provided our best source of information on the biotic composition of reefs and the biological interactions between species. The beauty
of coral reefs cannot be appreciated without diving or snorkeling to view the incredible microtopographic complexity and species diversity. The bewildering
array of species, competitive inter-actions, symbioses, predator-prey interactions, and remarkable variety of color have been compared to tropical rain
forests (Connell, 1973).
Definitions and General Limiting Factors
Definition. Coral reefs are compacted and cemented assemblages of skeletons and skeletal sediment of sedentary organisms living in warm marine waters
within water depths of strong illumination. They are constructional physiographic features of tropical seas consisting fundamentally of a rigid calcareous
framework mainly composed of the interlocked and encrusted skeletons of reef-building (hermatypic) corals and crustose coralline algae. The reef framework
controls the accumulation of sediment on, in, and around itself (Wells, 1957). The corals belong primarily to the order Scleractinia. Her-matypic corals all
have zooxanthellae-endosymbiotic algae that benefit the coral host (see later).
Temperature. High calcification rates are limited to warm waters. Conse-quently, coral reefs are restricted to tropical seas (Fig. 20-1), generally between 250
north and 250 south latitude. Well-developed coral reefs are usually not established at temper-atures much below 23 to 25°C. Some reefs, however, may
develop at temperatures as low as 1 8°C, as in the Florida Keys of the United States. Reefs at the edge of the above-mentioned latitudinal range may be
strongly affected by changes in climate. A 1968 cold event plunged air temperatures below 0°C over most of the Persian Gulf and caused water temperatures
to dip to as little as 10°C (Shinn, 1974). Almost all the inshore coral reefs soon completely died off. Some offshore reefs survived and presumably provided
pro-pagules for recolonization of the inshore reef zone.
These temperature restrictions apply only to coral reefs and not necessarily to corals per se. McCloskey (1970) studied a community associated with a
marine scleractinian coral that lived in shallow waters off South Carolina. Corals with zooxanthellae grew well at temperatures as low as 14.5°C. Some
particularly resistant coral species tolerated temperatures as low as 5°C for brief periods. The high rates of calcification ascribed to hermatypic (reef-building)
corals thus require both high temperature and the presence of zooxanthellae endosymbionts. Scleractinian corals, such as Astrangia danae, have
zoox-anthellae but live in temperate and boreal waters. Calcium carbonate accretion is cor-respondingly low.
Light. After temperature, light is probably the most important limiting factor to well-developed coral reefs because of the symbiosis between hermatypic
scleractinian corals and zooxanthellae. Derived from the dinoflagellates, zooxanthellae live within the gastrodermal tissues of scleractinian corals and are
apparently essential for rapid calci-fication. Because light intensity decreases exponentially with increasing depth, active reef building is greatly diminished
below depths of 25 m in the Indo-Pacific region (Rosen, 1975). In the Caribbean active hermatypic growth is rare below 75 m, but Montastrea and Agaricia
populations can be found. Normal calcification rates of corals can be cut in half on a cloudy day.
Although active coral growth cannot occur below 15 to 20% of surface light values, some coral species are adapted to diminished light conditions. When
brought into shallow waters of high surface illumination, corals typically living in the shade or at depth usually expel zooxanthellae and soon die (Lang, 1971).
Corals kept at the same depths in the shade of a boat show no ill effects. Some evidence suggests that zooxanthellae of her-matypic corals acclimatize to the
diminished light of an overcast day (Wethey and Porter, 1976).
Wells summarized many factors influencing the growth of corals and concluded that the diversity of hermatypic corals decreases with depth (Wells, 1957; Fig.
20-2). The depth distribution of light intensity and number of coral genera fit well contrasted to the change of temperature and oxygen with depth. So we
could believe that the change of coral diversity with depth is controlled by illumination. Other studies show that coral diversity may not simply decrease with
depth (Goreau, 1959) and may even increase with depth (Loya and Slobodkin, 1971). Coral species diversity increases with depth in the reefs of Eilat,
Israel, because of the greater environmental stability of deeper waters as opposed to the large fluctuations in temperature, desiccation, and salinity in the
shallow reef flat. Consequently, light does not exert a simple control on numbers of species of corals with depth.
Salinity. Hermatypic corals seem to require open-ocean salinity. Well-developed reefs are not generally found in estuarine or excessively hypersaline
conditions. Persian Gulf reefs, however, develop in salinities of more than 40%o (Kinsman, 1964). High rains, resulting in excessive runoff, cause extensive
damage to corals close to river mouths in such tropical islands as Fiji (Squires, 1962; Goodbody, 1961). Rivers also carry a large suspended sediment load
that is also detrimental to corals. Flash floods on the north coast of Jamaica carry large amounts of freshwater and suspended sediment to back-reef lagoons,
which typically have lower coral species richness.
Turbidity and sedimentation. High turbidity and sedimentation rates strongly inhibit reef growth. Turbidity increases light attenuation and thereby decreases
photo-synthesis by zooxanthellae. Correspondingly, calcification rates are diminished as well. Settling sediment tends to foul the surfaces of coral colonies;
horizontal benches often collect sediment and usually support poor reef growth (Loya, 1976a).
The effects of turbidity and sedimentation exert regional differences on reef de-velopment. Reefs in the clear waters of the windward sides of Caribbean
islands (e.g., north coast of Jamaica) grow more rapidly than on the leeward sides (e.g., south coast of Jamaica) or adjacent to continental coasts (e.g., coast
of Venezuela) where sediment accumulates.
Corals show differential adaptations for dealing with turbid environments. Species with large polyps can more easily remove suspended matter than
small-polyp forms. Massive and slow-growing hermatypic corals, such as Platygyra, produce large amounts of mucus when sedimentation is high. Mucus is
transported along the surface of the skeleton of the colony and carries away particulate matter. In contrast, rapidly growing branching forms, such as
Acropora palmata, produce less mucus because sediment does not tend to settle out on their smaller cross-sectional areas. Richman and others (1975)
measured in situ mucus production on the reef at Eilat, Israel, and found production per coral head to be (a) 6.8 ± 1.2 mg head-1 d-1 for massive forms (e.g.,
Platygyra), (b) 2.1 ± 1.0 for hemispherical species (e.g., Montipora), and (c) 1.9 ~ 0.7 for branching forms (e.g., Stylophora). Corals most resistant to burial
have also been found to be resistant to salinity and temperature changes and are thus best adapted for estuarine and low-energy lagoon environments.
Wave energy. Because coral reefs require clear water and are constructional topographic features, they tend to be located in areas of high wave energy. Erect
branching forms, such as the Caribbean Acropora palmata, living in reef crest zones must withstand wave shock and are often greater than 2 m across (Fig.
20-3). Hermatypic coral colonies are typically strong and very dense in structure. Such storm events as cyclones and hurricanes, however, often topple coral
colonies and exert massive destruction on coral reefs. A cyclone at Heron Island, Australia, obliterated a small area of coral reef, toppling large corals. Within
several years many coral colonies were reestablished by larval set-tlement from the plankton (Woodhead, 1971). In the Caribbean hurricanes exert massive
effects on coral reef communities as well. Hurricane Carmen passed over the north coast of Jamaica in the fall of 1974. The ensuing turbulence overturned
many large coral heads and tore loose epifaunal organisms from their substrata (Sammarco, 1977). Storms con-tinually renew open substrate for colonization
and space occupation by newly settling larval forms. Storms and other major physical disturbances are major sources of change in coral reef communities,
suggesting that coral reef communities are not static and constant biomes. Storm damage can, in effect, be a mechanism of coral dispersal, for pieces of living
colonies transported to new sites may survive to cement to the bottom and establish a permanent new colony.
20-2 REEF TYPES AND DEPTH ZONATION
A complex terminology has been developed to classify coral reefs. For our purposes, we divide reefs into two types: atolls and coastal reefs. Atolls are
horseshoe- or ring-shaped arrays of islands consisting of coral reef rock capping an oceanic island of volcanic origin. Coastal reefs border coasts of islands or
continents and range in dimension from the enormous Great Barrier Reef of Australia to the small reefs capping the coasts of Eilat,
N
Israel. They may rest on previous reefs, coastal bedrock, or soft sediment. Even the distinction between atolls and coastal reefs fails to include some reefs
whose occurrence is on oceanic volcanic islands but whose morphology resembles coastal reefs. The reader interested in the complexities of reef terminology
should consult Stoddart (1969).
Atolls
Atolls mainly occur in the tropical Pacific Ocean. A few are found in the Indian Ocean and the Caribbean (Glynn, 1973). Darwin correctly postulated that
atolls could only develop their characteristic array of ring-shaped to horseshoe-shaped emergent coral islands by slow subsidence of volcanic seamounts with
continuous and vigorous reef growth toward the sea surface. This hypothesis would imply the presence of great thick-nesses of reef rock capping a volcanic
basement. This prediction is confirmed by the ca. 1400 m of reef rock capping the volcanic basement of the Enewetak Atoll (Ladd et al., 1953). The reef
dates back to the Eocene (40 to 60 million years before present).
The scheme of atoll development may be summarized as follows. As the volcanic island submerges, coral reefs develop around the fringe of the island. When
the volcanic island center plunges below the sea surface, the peripheral reef continues to grow upward, developing a ring-shaped array of islands and leaving
a lagoon in the center (Fig. 20-4a).
Figure 20-4b shows a cross section of a typical atoll. Because the circular array of islands in an atoll usually resides in a unidirectional wind pattern, the
windward side of the atoll generally develops different coral reefs than the leeward side of the atoll system. Seaward and windward reefs are the zones of
most intense wave energy. At a depth of a few meters, few live corals are to be found and coral rubble accumulates. But large hermatypic corals grow
abundantly below this zone. Below the zone of high wave energy, more delicate and foliose corals are found. On the intertidal margin of windward reefs, a
large algal ridge is formed by crustose coralline red algae. It usually rises about a meter above mean low tide level. Behind this algal ridge is a broad intertidal
or slightly subtidal reef flat. Changes in salinity and temperature are most extreme on this part of the reef. Corals are most abundant near the lagoonward side
of the algal ridge and lagoonward of this zone is a large expanse of reef flat in which massive corals as much as 3 m in diameter may occur.
The leeward and seaward sides of the reef experience much less dramatic wave surge; consequently, corals are more abundant in shallower water. The algal
ridge is not as well developed. The windward and seaward sides of the reef may be dissected by very large surge channels, sometimes as much as 15 m
deep. Sturdy collonies [sic] of Acropora, Pocillopora, Millepora, and Heliopora line the walls of the surge channels. Surge and wave energy are important
limiting factors in these shallowest portions of the reef zone.
The relatively quiet waters of the lagoon usually contain lagoon reefs and lagoon slope and floor environments (Wells, 1957). Although the lagoon floor is
generally between 25 and 50 m deep, occasional pinnacles rise to the surface and are capped by coral reefs similar to the seaward side of the atoll. If passes
do not exist from the outside to the lagoon, the atoll floor is normally covered with fine-grained sediments and supports a community typical of tropical
calcareous muddy environments. If channels exist for natural reasons or have been blasted for shipping, currents keep the lagoon floor free of fine-grained
sediments.
Coastal Reefs
Coastal reefs parallel shorelines and range in size from barrier reefs as much as 2000 km long and fronting a lagoon of 50 km in width to small and
discontinuous reefs plastered on the shoreline. Unlike atolls whose morphogeneses are all probably similar in origin, coastal reefs consist of a group of
structures of diverse origin. The Great Barrier Reef of Australia, for instance, can consist of geologically ancient reef rock and some areas of the reef tract can
be interpreted as being due to continual subsidence (Stoddart, 1969). The Great Barrier Reef tract is an end member in the types of bordering reefs because
it fronts a "lagoon" some 40 to 50 km wide protecting the coast of eastern Australia. At the other extreme, fringing reefs cap sand, volcanic rock, or
previously dead coral reefs. An example is the fringing reef at Eilat, Israel, studied by Loya and Slobodkin (1971). In the intertidal zone the bottom is capped
by a rather continuous table of coral reef rock. At depths of about 30 m, however, patches of coral reef cap sand bodies or preexisting outcrops of rock.
Coastal Reefs of Jamaica
As an example of reef zonation, we shall discuss the windward reef tract on the north coast of Jamaica, West Indies. From the shore, the general morphology
of the reef may be described as the following (Fig. 20-5):
1. A back-reef zone consisting of shoreline and lagoon (0 to 10 m depth).
2. A reef crest (0 to 20 m).
3. Buttress zone.
4. Staghorn coral zone.
5. A break in slope at depths of 55 to 65 m, leading to a precipitous dropoff at vertical angles of 60 to 900 to water depths of greater than 300 m (Goreau,
1959; Land, 1973; Goreau and Land, 1974).
The shoreward margins of most back-reef lagoons are usually lined with either reef rock or sediment consolidated by mangroves. The nature of the lagoon
depends greatly on the hydrography of the region and the openness of the lagoon to the sea. If the lagoon is cut off from strong currents, then the bottom
consists of fine-grained sediments. Shallower bottoms support extensive growths of the marine angiosperm Thallassia tes-tudinum and other sea grasses.
Members of the bivalve family Lucinidae are abundant in soft sediment among the roots of turtle grass and feed mainly on organic detritus falling into an
anterior mucous tube (Allen, 1958; Jackson, 1973). Lucinids here live in anoxic sediments and are resistant to physiological stress, such as high temperature
and extremes of salinity.
In slightly deeper and less anoxic sediments, species richness increases and a large number of mullusks [sic] are capable of surviving, relative to the
low-diversity fauna of shallow (less than 1 m) turtle grass flats. In turtle grass meadows greater than 1 m in depth, grazing urchins (e.g., Tripneustes
ventricosus) may control seagrass biomass. The large sea cucumber Astichopus multifidus consumes sediment. Volcano-shaped sediment mounds formed by
burrowing callianassid shrimp mottle the sandy bottom.
Also in the lagoon are patch reefs ranging from a meter to 50 to 100 m across. They are often as rich in coral species as the forereef but rarely support the
large coral heads found on the outer reef.
As we approach the rearward margin of the reef crest, large numbers of coral species seem to flourish even though wave energy and suspended sediment
conditions are less optimal than on the seaward side of the crest. We then go on to a reef flat that is either intertidal or only a few centimeters to 1 m in depth.
Here corals are usually rather small. The colonial zoanthid, Zoanthus sp., is often abundant. Hermatypic corals in the Car-ibbean, do not survive intertidal
exposure as well as in the Pacific, perhaps because of the smaller tidal excursion in the Caribbean, which may result in long periods of exposure to heat and
desiccation when winds move from onshore to offshore.
Just seaward of the reef flat, the reef crest supports large coral heads of the elkhorn coral, Acropora palmata (Fig. 20-3). Branches tend to be oriented
parallel to major unidirectional currents. At about 6 to 7 m in depth, the colonies are smaller and branches are more flattened. The deeper forereef consists of
a broad and relatively flat terracelike area at 10 to 15 m depth, regularly transected at right angles to the shoreline by large and deep surge channels that form
a series of large lobes facing seaward (spur and groove topography). These lobes constitute the buttress zone and are mainly composed of a framework
constructed by the massive hermatype, Montastrea annularis. Species richness peaks in the buttress zone.
Below the buttress zone, a broad low-relief bottom is covered with thickets of the staghorn coral, Acropora cervicornis (Fig. 20-6). Pieces of coral branches
readily break and roll about, making the thicket bottom unstable. Descending the forereef slope, we encounter a steeper slope of high scleractinian coral
species richness, the forereef es-carpment. From this escarpment a slope descends at vertical angles between 20 to 60° to a second slope break at a depth of
55 to 65 m. This slope is barren of corals, although it contains a rich biota of algae, gorgonians, and burrowing organisms that occupy the predominating soft
sediment. A flattened growth form of Montastrea annularis (Fig. 20-7) can be an important framework builder on coral pinnacles protruding from the forereef
slope, but some other species are more important frame builders in the lower half of the forereef zone, such as the foliose hermatype Agaricia spp. (Fig.
20-7). An abrupt slope break, the "dropoff," occurs at a depth of about 55 to 65 m, below which the deep forereef descends almost vertically.
Below about 75 m, corals are no longer sufficiently abundant to be principal frame-work builders. The vertical wall dropping off to greater than 1000 m,
however, is actively growing outward. An important framework builder on these walls is the sclerosponge Ceratoporella nicholsoni, a potentially large and
massive calcifying sponge thought to be related to the extinct Stromatoporoidea (Hartman and Goreau, 1970). They are only present in shallow-reef
environments within caves and help define a distinct coral reef cave community, along with articulate brachiopods (Jackson et al., 1971). At these depths,
sclerosponges are large and become major framework builders, helping to trap sediment moving downslope on the reef. They are probably of major
importance in maintaining the dropoff as a vertical wall. At depths of 200 to 300 m, organisms like stalked crinoids have been found in great abundance on
sandy bottoms.
Indo-Pacific Reef Zonation
Coastal reefs in the Indo-Pacific share many features with those of the Caribbean. But-tresses project seaward and form a spur and groove topography. As
in the Caribbean, Acropora is dominant in wave-swept areas; coral colony branches tend to align and point into the surf. Behind the reef front, a broad algal
rim develops, as on the seaward side of atolls. The rim consists of step pavements of coralline algae, which grow seaward over the spur and groove channels,
and bears the brunt of the main attack of waves hitting the coral reef surface. The back reef consists of channels, islands, and relatively quiet lagoons filled
with soft sediment. For a detailed account of zonation on the Great Barrier Reef, the reader should consult Manton and Stephenson (1935), and Maxwell
(1968).
General features of Indo-Pacific zonation have been summarized by Rosen (1971, 1975), based on exposure to water movement. Three components are
important (Fig. 20-8): (a) exposure of the coast, (b) distance from the reef edge exposed to waves, and (c) water depth. A given coral assemblage is
therefore determined by its position with regard to these three factors. In order of decreasing exposure, the zones are
1. The algal ridge.
2. Pocillopora zone of turbulence.
3. Acropora exposed zone.
4. Faviid-Musiid zone.
5. Porites zone.
With the exception of the algal ridge, which is always at the surface, all zones are found in a sequence corresponding to decreasing water movement. Given
the three important components, the wave-resistant Pocillopora assemblage is to be found in shallow, wave-exposed coasts, nearest the oceanward edge of
the reef. Porites zone residents will occur in deeper, more protected environments away from the reef edge. In practice, all zones are not found in any one
place (Rosen, 1975).
20-3 REEF TOPOGRAPHY: ACCRETION AND EROSION
Genesis of Reef Morphology
The control of reef morphology is a subject in great dispute. We can generally agree that the very thick accumulations of coral reef rock on atolls can only be
explained by long-term subsidence and concomitant upward reef growth. However, the complex of terraces and dropoffs found, for example, in Caribbean
reefs may be explained by two alternative processes or their combined effects: (a) the constructional growth of reef tracts or (b) erosional processes that
occurred during Pleistocene lowerings of sea levels (Purdy, 1974). At low sea level stands an erosional terrain may then have subsequently controlled reef
growth when sea level rose as the glaciers retreated. This second process would explain the presence of steep slope dropoffs on oceanic margins of reefs, the
presence of barrier reef terraces, and shelf lagoons. The occurrence of blue holes-large and deep depressions in the sea bottom indicating former
development of karst or cave topography when sea level was at a lower stand-supports Purdy's hypothesis of extensive erosion at lower Pleistocene sea
levels.
Although extensive erosion is likely, massive reef accretion has occurred during postglacial rises in sea level. Because the Pleistocene was a period of several
glacial advances and retreats, there were several corresponding rises and falls in sea level. The ability of reef growth to track rises in sea level depends on
antecedent topography, the rate of sea-level rise, and wave energy (Adey, 1978). When a rise in sea level inundates a carbonate platform, initial conditions of
suspended sediment may be unfavorable for reef growth. The lag time in reef initiation may be too great for sufficient reef growth before water depth has
increased to the degree that light is limiting. Windward coasts, however, may favor the growth of species of Acropora and coralline red algae; the former are
favored in moderate wave energy and explain most rapid reef accretion.
Adey (1978) summarizes the evidence supporting the hypothesis that many reefs are capable of growing with rising sea level during glacial retreats. Reef
growth has been vigorous in both the Atlantic and Pacific; rates of 9 to 15 m upward accretion per year have been recorded in the Caribbean.
Curves of reef growth developed from C-14 dating of coral skeletons in both the Atlantic and Pacific often show strong concordance with sea-level rise
curves. The role of such organisms as sclerosponges in deep water suggests that significant lateral accretion is also possible and that near-vertical walls can be
accretionary features. Coral reef workers have generally thought of Caribbean reefs as relatively depauperate and slow growing. Although the greater species
richness found in the Pacific is not in doubt, it is becoming clear that vigorous Caribbean reef growth is the rule and that extensive reef tracts are common
(e.g., northern side of Little Bahama Bank on the Nicaraguan rise).
Bioerosion
Although dramatic reef accretion ultimately stems from the contributions of calcium carbonate-producing organisms, numerous species of animals and plants
destroy the skeletal output and may even inhibit reef accretion. Urchins and grazing fishes feed on epibionts and concomitantly scrape off bits of calcium
carbonate. The majority of sand--sized particles on reefs probably come from grazing activities.
The other major source of bioerosion comes from the activities of endolithic or-ganisms, those creatures that bore into hard substratum. A wide variety of
bivalve mol-lusks, sipunculids, and polychaetes plays a major role in boring and subsequently weak-ening coral skeletons. Sponges of the family Clionidae are
probably the most significant source of bioerosion on coral reefs. They work at reef rock at the base of the intertidal zone and are the cause of overhanging
benches of intertidal rock commonly observed in the Caribbean (Neumann, 1966). Clionids are commonly found at the base of hermatypic coral skeletons
and may weaken the base to the point of breakage (Goreau and Hartman, 1963). Although chemical dissolution is the mechanism by which the calcium
carbonate matrix is weakened, a poorly known type of contractile activity by sponge cells causes a continuous removal of small chips of calcium carbonate.
These fragments may be the principal source of silt-sized particles on reefs.
The importance of bioerosion increases with depth, for hermatypic coral growth is adversely affected by diminishing light. The rate of bioerosion is sufficiently
important at depth to affect the morphology of hermatypic corals. Although massive and branching forms are common in shallow water, platy forms (e.g.,
Agaricia) predominate in deeper water. If the base of a platy coral is eroded to the point of breakage, the plate will simply fall down to the sediment with the
polypoid surface facing upward. The coral colony will therefore survive. Because the deeper reef in Jamaica has a considerable slope, however, breakage of
the base of a more massive and hemispherical coral will cause it to roll somewhat; the whole polyp surface may then be smothered in sediment. Furthermore,
the platy growth form permits several colonies to grow as a series of adjacent interlocking plates. In some cases, the base may be completely destroyed and
yet the colony skeleton will not fall into the sediment (Goreau and Hartman, 1963).
20-4 BIOLOGY OF SCLERACTINIAN CORALS
Morphology and Reproduction
Scleractinian corals are coelenterates closely related to sea anemones that secrete a skeleton of calcium carbonate (aragonite). Some corals are solitary (only
one polyp) with polyps as much as 25 to 30 cm in diameter. Most are colonial with hundreds or thousands of polyps averaging about 1 to 15 cm in diameter
(Fig. 20-9). A colony of coral polyps is a sheet of live tissue covering a dense massive skeleton of calcium carbonate, formed by gradual accretion by the
coral colony at the surface. The periphery of the polyp oral disk region is surrounded by one to several rings of tentacles with varying degrees of adaptation
for zooplankton capture and an esophaguslike connection extends from the central mouth to the interior gastrovascular cavity. Tentacles are armed with
nematocysts designed to entrap prey.
Many species of corals have been observed to be sequentially hermaphroditic. In some species, individual colonies are of one sex, but sex changes with age.
Most com-monly, fertilization is internal with extrusion of eggs into the water occurring in a minority of species. Reproductive activity may or may not be
seasonal. After fertilization, larvae develop in the gastrovascular cavity of the parent and eventually are ejected through the mouth. The swimming planula
larvae are usually elongate or spherical and externally ciliated. The larva is the means of dispersal, carrying the species into newly opened environments
Asexual budding allows the parent colony to grow and increase in size (Connell, 1973)
A coral planula spends some time swimming in the plankton; a later stage swims to the bottom, where attachment, settlement, and metamorphosis occur.
Connell (1973) studied rates of recruitment of different coral species and found that a wide range of recruitment rates occurred but that there was no obvious
explanation for these general variations in space and time. In all areas, however, commoner species had higher re-cruitment rates and mortality. There were
very large differences in recruitment and areas as little as a few hundred meters apart differed strongly in larval recruitment rates. Larval life of planulae in the
plankton may be as little as 2 days. This short dispersal phase might result in great variation in settling concentrations over relatively short distances.
Hermatypic corals are those corals that are mainly responsible for the coralline contribution to reef growth. They differ from ahermatypic corals in having
higher rates of calcification and large numbers of zooxanthallae living within the gastroderm. Aher-matypes have few or no zooxanthallae, calcify at much
slower rates, and do not produce large coral heads. The presence of zooxanthallae as symbionts in groups of calcifying organisms seems to be correlated with
high calcification rates (as in the giant clams Hippopus and Tridacna) or inferred to be with such fossil organisms as prorichtofenid brachiopods (Cowan,
1970).
Growth of Corals
Coral-growth analysis allows an understanding of the mechanics of reef growth and reef accretion. The rate of coral reef accretion must be the aggregate of
the growth rates of the individual coral heads, with cementation of grains in between the coral colonies. Vaughan's (1915) study of corals of the Dry Tortugas
reefs showed that coral heads with a massive growth (Fig. 20-10) form grew more slowly in linear dimensions than corals with a ramose (branching) growth
form (Fig. 20-3, Fig. 20-6). The staghorn coral, Acropora cervicornis, a common coral on reefs in the Caribbean, grows as much as 10 cm y-1, as measured
by branch-tip extension (Shinn, 1966). In contrast, massive hem-ispherical colonies of coral Montastrea annularis accrete at 0.25 to 0.70 cm yr-1 in height
(Vaughan, 1915). Massive colonies, however, tend to reach a greater total weight of calcium carbonate than the branching forms.
Two types of skeletal addition can be recognized in massive corals (Dustan, 1975; Fig. 20-11): the axis of upward growth, representing an increase of
skeletal size by the same polyps, and the axis of polyp addition, or horizontal axis along which polyps are added through extratentacular budding in
conjunction with new skeletal material. Meas-uring growth along both axes, along with skeletal density (using mercury displacement), a shallow massive
growth form of Montastrea annularis can be shown to have a higher calcification rate than a deeper platelike form (Dustan, 1975). This difference is mainly
reflected in depth-dependent variation in the axis of upward growth (Fig 20-12).
The common Caribbean massive hermatypic coral Montastrea annularis has dis-tinctly different growth forms (Fig. 20-7, Fig. 20-10) in shallow and
deep-water envi-ronments (Goreau and Land, 1974; Dustan, 1975). In shallow water (10 m) the species grows in massive, hemispherical colonies with the
predominant growth vector upward. In deeper waters (30 m) M. annularis assumes a platelike growth form with a predom-inately horizontal-extending
growing edge. Platelike growth form is probably favored to maximize light capture at low light intensities and to avoid rolling when the base of the skeleton is
bioeroded. The importance of M. annularis as a reef framework builder thus affects the reef profile with depth as the shallow growth morph forms large
buttress structures. Faviid coral skeletons of New Caledonia vary regularly with depth and with a progressive reduction of septa per corallite and number of
corallites per unit area of colony (Wijsman-Best, 1972). This result is thought to be a response to lowered light, decreasing the ability of zooxanthellae to aid
calcification. Corals in deeper water contain fewer zooxanthellae (Yonge, 1940; Best, 1969).
Several methods have been developed to measure coral skeleton growth. The most simple technique is to measure increments of skeletal addition relative to
metal spikes driven into coral heads or to measure distances to branching tips from specified reference points in branching colonies. When introduced in
suspension, the dye Alizarin Red-5 is deposited in the coral skeleton and remains as a fixed time plane as future growth occurs (Barnes, 1972). This method
can be used in the field by introducing the dye into a polyethylene bag enclosing the coral colony (allowing for some water exchange). The coral skeleton is
later sliced with a diamond saw and the distance from the stained time horizon to the new growing surface is measured.
Growth can also be assessed in coral colonies through growth bands, which are visible in diamond-saw sections of coral skeletons and represent variations in
skeletal density. Increments correlate with cycles, such as fortnightly tides, and provide time horizons between which growth can be measured. Annual bands
corresponding to seasonal temperature change can readily be employed to measure long-term growth rates of coral colonies (Weber et al., 1975). Coral
growth was shown by this technique to be inhibited where bottom sediments were resuspended in Bermudian lagoons (Dodge et al., 1974).
Measurement of uptake of the radioisotopes Ca-45 and C-14 permits short-term studies (< 1 h) of calcification and allows controlled laboratory experiments
on the effects of temperature and salinity (Goreau, 1959; Goreau and Goreau, 1959). The Ca-45-based estimate of 20 mm y-1 growth of branch tips of the
Pacific coral Porites compressa compares favorably with more direct measurements (Buddemeier and Kinzie, 1976). In Pocillopora damicornis, calcification
is six times greater at branch tips than at the lateral regions. Using Ca-45 uptake, Goreau (1959) confirmed the previously discovered greater uptake of
calcium in the light. Corals may acclimatize to environmental temperature.
20-5 NUTRITION OF CORALS AND THE ZOOXANTHELLAE PROBLEM
Hermatypic corals participate in a remarkable symbiosis with zooxanthellae. Zooxan-thellae cultivated outside their hosts change from a typical spherical
shape maintained within the coral endoderm to biflagellate motile dinoflagellates. Zooxanthellae also occur in bivalves, other coelenterates, and in gastropods,
but it is not clear as to whether the same dinoflagellate species infects all hosts or a variety of species are host-specific (Taylor, 1971, 1974). Zooxanthellae
isolated from different hosts may have differing isozyme patterns (Schoenberg and Trench, 1976). Zooxanthellae taken from one host may or may not be
beneficial when introduced into another host (Kinzie and Chee, 1979).
For many years a controversy has raged concerning the nature and cause of this symbiosis. Because zooxanthellae are photosynthetically active, several
obvious hy-potheses have been proposed
1. They are a possible source of food.
2. They may be a source of oxygen for corals in an oxygen-depleted environment (as reefs are typically oxygen rich, we omit this factor in our discussion).
3. They aid in lipogenesis.
4. They facilitate the excretory processes of hermatypic corals.
5. Through absorption of carbon dioxide they aid or affect calcification rates in her-matypic corals.
Nutrition
The hypothesis that zooxanthellae are a possible source of food for hermatypic corals raises the question of feeding behavior and digestion. In a classic series
of experiments, Yonge (1930, 1931) concluded that corals are microcarnivores feeding on zooplankton in the overlying water. The feeding tentacles
surrounding the mouth are crowded with nematocysts-eversible structures capable of stinging or entrapping prey. Several amino acids and the peptide
glutathione consistently evoke feeding behavior in corals (Lenhoff, 1968). The puncture of prey by nematocysts releases stimulants, which initiates tentacle
movement, transfer of food, and opening of the mouth. Ciliary and muscular movement transports the food to the mouth and down the esophagus.
Mesentarial filaments extend from the mesentary and secrete digestive enzymes that digest the animal prey tissue.
Reactions of some corals to starvation also suggest their adaptations as microcar-nivores. If starved for any length of time, corals extrude their zooxanthellae
and sub-sequently die. This behavior hardly indicates that corals can subsist solely with a zoox-anthellae food source. Johannes and co-workers (1970),
however, calculated that zooplankton productivity was not adequate to sustain the energy requirements for the reef corals of Bermuda. Although
upstream~ownstream comparisons of phytoplankton and zooplankton show that reef organisms retain a large amount, the accrual is less than 20% of net
community metabolism (Glynn, 1973). As will be shown, corals differ in their dependence on zooxanthellae.
The suggestion that zooxanthellae provide a food source for reef-dwelling scler-actinian corals is supported by other organisms having zooxanthellae, where a
clear nutritional involvement has been demonstrated. In the giant clam Tridacna, large numbers of zooxanthellae are maintained in ameboid blood cells in the
blood sinuses of the mantle tissue. Although the shell's life position is with the umbo of the shell downward, the body is rotated 1800 with respect to the shell,
a unique adaptation in the Bivalvia. This rotation presents the zooxanthellae-rich mantle tissue to the light so that maximum pho-tosynthesis can occur.
Zooxanthellae are concentrated in translucent structures projecting as nodes up from the mantle tissue. Digestive cells carry zooxanthellae from this mantle
region and digest the zooxanthellae. Indigestible remains are then carried toward the excretory organs. The entire digestive system of the giant clam has been
reduced and highly specialized in order to accommodate this symbiosis (Yonge, 1936b).
Muscatine and Hand (1958) showed that C-14 was fixed by zooxanthellae and demonstrated by autoradiographic studies that it was later found widely
dispersed through-out the tissues of coelenterates. Nutrient transport therefore occurs between zooxanthellae and the body of the corals as well. Trench
(1974) investigated the tropical coral reef-dwelling coelenterate Zoanthus sociatus, a colonial anthozoan with zooxanthellae as intracellular endosymbionts. He
demonstrated a direct transfer of photosynthate from zooxanthellae to host; glycerol, glucose, and alanine were released to the anthozoan. Z. sociatus was
also capable of feeding on zooplankton. Thus the animal maintained the flexibility of autotrophy and heterotrophy. The animal did not digest the zooxanthellae
themselves. Zooplankton feeding may provide nitrogen.
Among-species morphological heterogeneity in skeletal morphology and polyp size suggests a spectrum of dependence on zooxanthellae as a food source
(Porter, 1976). Polyp diameter P is positively correlated with tentacle length and is a good indicator of zooplankton-capturing ability. The surface area of the
coral skeleton covered by live tissue, S, divided by the volume of the skeleton plus tissue, V, is a good index of light-capturing ability. As S/V increases in
branching and platy forms, more light can be intercepted because not all incident radiation will be captured by a single intercepting plane. A multilayered
morphology, as in the branching Caribbean coral Acropora palmata, allows S to be three times the surface area of bottom substrate that they cover (Dahl,
1973). Light can then be intercepted by more surface area of live coral tissue. Zooplankton capture favors a single continuous surface of tissue because
feeding structures cannot be saturated with light, as can the photosynthetic zooxanthellae. Branches increase the number of feeding mouths, but a single plane
of feeding polyps would effectively remove all the zooplankton.
S/V and polyp diameter are hyperbolically inversely correlated (Fig. 20-13). Thus corals with a shape well adapted to zooplankton capture have large polyps
similarly adapted to this function (Porter, 1976). Although this factor suggests a range of depend-ency of hermatypes on zooxanthellae as food, even species
with low values of (S/V)/P show a strong apparent need for such a food source. The large-polyped Montastrea cavernosa can only obtain 10 to 20% of its
daily energy requirement during the two most successful hours of its 12-hour night feeding period (Porter, 1974a, 1976).
Dissolved organic matter might be an important source of food for corals as well. Along with many other phyla, several species of coral are capable of
absorbing dissolved organic matter (for example, Stephens and Schinske, 1961). Carbon-14-labeled glucose is taken from solution by the coral Fungia and
uptake rates are independent of light intensity or presence of bacteria. Stephens guessed that the body wall is the site of uptake for organic matter. Although
the amounts of glucose necessary for the maintenance of coral metabolism seem too high, given the rate of uptake measured in Stephens' exper-iments,
uptake of organic matter-perhaps by mesentarial filaments-may be an auxiliary method of feeding. The extrusion of mesentarial filaments by a coral colony
might provide an extensive and large area for absorption of dissolved organic matter by the colony (Muscatine, 1973).
Lipogenesis
Zooxanthellae apparently are important in lipogenesis for hermatypic corals (Patton et al., 1977). Because lipid constitutes about one-third of the dry weight
of anemones and corals, it is probably a primary energy source. The hermatype Pocillopora capitata elevates lipid synthesis 300% in the light relative to
individuals maintained in darkness. Carbon-14-labeled acetate can be shown to be incorporated by zooxanthellae and used in lipogenesis (Patton et al.,
1977); therefore zooxanthellae are responsible for the efficient transfer and conversion of acetate to lipids. It seems likely that a diversity of fatty acids taken
up because of the coral's carnivorous habit is quickly oxidized to acetate in digestive cells. The acetate would then be converted to a narrow spectrum of
saturated fatty acids with the aid of the energy available from photosynthesis. Acetate produced in lipid breakdown could be recycled to the zooxanthellae for
lipid synthesis.
Polyunsaturated fatty acids are less common in corals and may either indicate lipogenesis by the animal instead of by zooxanthellae or may indicate external
sources (dietary) for the fatty acids. Acropora palmata, a resident of shallow water, contains predominantly saturated fatty acids and so probably depends
mainly on zooxanthellae for lipogenesis. Corals in deeper water have greater proportions of unsaturated fatty acids and may thus depend to a lesser extend on
algal lipogenesis (Meyers, 1979).
Excretion
Zooxanthellae may facilitate the excretory processes of hermatypic corals. In the giant clam Tridacna, Yonge (1 936b) demonstrated that clams kept in sealed
containers even-tually depleted all the phosphorus in the water in the container. Another bivalve mollusk, Spondylus, with no zooxanthellae, did not reduce
dissolved phosphorus. Phosphorus levels increased because of the excretory activities of the bivalve. Yonge and Nicholls (1931) measured phosphorus
excretion by several hermatypic coral genera and the aher-matypic coral Dendrophyllia. Although all species excreted phosphorus, the ahermatypic coral,
lacking zooxanthellae, excreted more phosphorus than the hermatypic corals. They concluded that zooxanthellae take up phosphate from the overlying water
and also remove phosphate as an excretory product from hermatypic corals. Zooxanthellae might therefore by "automatic organs of excretion" (Yonge,
1940). But the strong unidirectional currents common on coral reefs would probably remove any phosphorus excreted in the overlying water, thus preventing
poisoning or fouling of the water immediately above the coral colonies (Muscatine, 1973). The major beneficiaries in the removal of excretory products are
the zooxanthellae, who gain nutrients not otherwise available in the open water. CaCO3 precipitation can be shown to be inhibited by dissolved phosphate,
however. So zooxanthellae might enhance calcification by removal of phosphate excretions (Weber and Woodhead, 1970).
Calcification
Goreau and Goreau (1959) demonstrated that zooxanthellae play a role in the calcification of hermatypic corals by measuring light and dark uptake rates of
Ca-45 under both controlled field and laboratory conditions. Their experiments are consistent with field observations that calcification on cloudy days is about
50% of calcification on sunny days. DCMU, an inhibitor of photosynthesis, also lowers rates of calcification in her-matypic corals.
Inhibition of the enzyme carbonic anhydrase also decreases the rate of calcification. Goreau and Goreau (1959) suggested that uptake of carbon dioxide for
photosynthesis was the important factor enhancing rates of calcification in the coral. As carbon dioxide is removed, rates of calcium carbonate deposition are
increased, perhaps through effects on the carbonate, bicarbonate, carbon dioxide system, leading to deposition of calcium carbonate (see Chapter 1). The
skeleton of the coral is secreted by the ectoderm of the basal disk. Calcium carbonate is secreted onto an organic matrix, which provides nu-cleation sites for
crystals of calcium carbonate.
The factor of supersaturation of seawater with respect to calcium carbonate is given by the ion activity product (IAP), divided by the equilibrium constant (K)
for the reaction. Calculations by Berner (1971) show that the ratio of IAP to K in seawater is approximately 3.1. This supersaturation suggests that removal
of carbon dioxide during photosynthesis by the zooxanthellae might therefore not have an appreciable kinetic effect on the rate of deposition of calcium
carbonate. Muscatine (1973) makes this argument from an intuitive point of view by pointing out that the apical polyps of acroporid corals calcify faster than
the lateral polyps, although the former have fewer zooxanthellae (Pearse and Muscatine, 1971). Two possible alternative hypotheses are as follows.
1. Zooxanthellae might manufacture some important factor for the organic matrix for the calcium carbonate deposited by corals. It is certainly possible
because we know of cases in bivalve mollusks where different types of amino acids influence the mineral phase of CaCO3 precipitated in the molluscan shell.
According to this hypothesis, however, zooxanthellae must have evolved the ability to code for pro-teins that are subsequently secreted and transported to the
site of calcium carbonate deposition on the coral skeleton. To date, no such evidence exists.
2. Organic phosphate inhibits calcification and so the removal of phosphate by zoox-anthellae during the act of photosynthesis may increase the rate of
calcium carbonate deposition. In recent years phosphate has been shown to be an important inhibitor on the surface of crystals, particularly in the case of
aragonite, the crystal form of calcium carbonate found in hermatypic corals.
To summarize, zooxanthellae seem to confer several advantages to corals; nutrition, lipogenesis, and calcification are clearly enhanced, It is not clear whether
any of these factors are of primary importance-that is, that one particular factor was predominant in favoring selection for the mutualism between
zooxanthellae and coelenterates. Because noncalcifying forms-anemones-thrive with zooxanthellae, the roles of nutrition and lipogenesis in maintaining the
symbiosis might be emphasized. Yet the advantage to calcifying hermatypes is of obvious importance. Realistically, zooxanthellae probably confer varying
advantages, depending on morphology, habitat, and availability of alter-native nutritive sources.
SUMMARY
1. Coral reefs are wave-resistant structures dominated by scleractinian corals and calcium carbonate-secreting algae. They are tropical in occurrence and
thrive best in clear, open-ocean water at wave-exposed sites. Reef growth is a net result of calcification and upward growth keeping pace with recent global
rises in sea level. Some of the topographic features of reefs may be related to erosional processes occurring at Pleistocene low stands of sea level.
2. Scleractinian corals are coelenterates that are closely related to sea anemones and that secrete a skeleton of calcium carbonate. Reef-building (hermatypic)
corals harbor large numbers of algal endosymbionts (zooxanthellae) in the gastroderm. The value of zooxanthellae seems to based on a benefit to corals in
terms of:
(a) nutrition-zooxanthellae release photosynthate to the coral;
(b) increased calcification rate in the presence of zooxanthellae;
(c) an acceleration of lipogenesis.
Figure 20-1 Approximate limits of the tropical Indo-Pacific and Caribbean marine coral reef provinces, as compared with minimum average sea
temperatures. (After Newell, 1971)
Figure 20-2 Distribution of reef-building corals at Bikini Atoll, as compared to the depth distribution of several physical variables. (After Wells, 1957,
courtesy The Geological Society of America)
Figure 20-3 Colonies of the elkhorn coral, Acropora palmata, a dominant of the reef crest of Caribbean coral reefs. Note the preferred orientation of the
branches of the colony. (Photograph courtesy James W. Porter)
Figure 20-4 (a) The hypothetical origin of a coral atoll; (b) cross section showing major subhabitats.
Figure 20-5 (a) Map view of coral reef environments, off the north coast of Jamaica. (After Goreau and Land, 1974) (b) Cross-sectional view of depth
zonation of the coral reef at Discovery Bay, Jamaica. (After Goreau, 1959. Copyright 1959, the Ecological Society of America)
Figure 20-6 The branching coral Acropora cervicornis, in thickets, Discovery Bay, Jamaica. (Photograph courtesy James W. Porter)
Figure 20-7 The deep-water growth form of the massive reef-building coral, Montostrea annularis, in association with several species of Agaricia.
(Photograph courtesy Philip Dustan)
Figure 20-8 Schematic diagram relating major Indo-Pacific shallow-water coral asso-ciations to strength of water movement. Water movement is dissected
into three major components, arranged perpendicularly. 0 = theoretical point of zero water movement. Sequence of associations is given by numbers: 1 =
Calcareous algae; 2 = Pocillopora; 3 = Acropora; 4 = Faviid; 5 = Porites. Stippling suggests increasing effects of loss of illumination with depth. (Modified
after Rosen, 1975)
Figure 20-9 Closeup of the hermatypic coral, Monrastrea cayernosa, showing expanded polyps. (Photograph courtesy James W. Porter)
Figure 20-10 The shallow-water mound-building form of Montastrea annularis, being overtopped by the elkhorn coral, Acropora palmata. (Photograph
courtesy James W. Porter)
Figure 20-11 Montastrea annularis. Photograph of saw-cut sections of skeletons after 1-year growth period. Top: deeper water flat form; bottom: shallow
water mound-shaped form. A-B axis of upward growth used for both forms; A-B and C-D are axes of polyp addition used for round and flat forms,
respectively. (From Dustan, 1975, from Marine Biology, volume 33)
Figure 20-12 Mean colony skeletal extension rates (mm/yr) between 10 and 45 m depth on Dancing Lady Reef, north coast of Jamaica. (After Dustan,
1975, from Marine Biology, volume 33)
Figure 20-13 The S/V ratio of Caribbean reef-building coral species, as a function of polyp diameter. (After Porter, 1976)
