What is a reef

We start with the elements upon which there is general agreement. First, The primary disagreement concerns the concept of framework in reef building. The reef-forming organisms generally possess massive skeletons that are the building blocks of the reef. Many workers hold that in a "true reef" these are organized into an in-place and interlocking meshwork that provides the rigidity of the reef. Conversely, any feature lacking this interconnecting framework is not a reef (see Fagerstrom 1987 for an excellent summary).

The problem is that when ancient reefs are examined, many of them are not the organized structures that are central to this model. They have recognizable organisms, but these are loosely packed and often "floating" in a detrital matrix. In these cases it is assumed that either:

This pattern has occurred with sufficient regularity to raise questions about how (and why) modern and ancient reefs might be fundamentally different from one another. Reef organisms have certainly evolved through time; perhaps this has changed the ways in which reefs have functioned. Sea-water chemistry and the shape of ocean basins have likewise changed - many ancient reefs existed in broad, shallow seaways and intracontinental seas, in contrast to the more-exposed and steep platform margins of today. If all these factors have conspired to fundamentally change reefs over the span of geologic time, then modern reefs become very poor models for their ancient counterparts. We will show that this is not the case. In an analogy by Ginsburg, reefs are likened to Shakespearean plays in which an ever-changing cast acts out a timeless plot. The following discussion builds on this theme.

Reef classification

From a geological perspective, only biological construction and topographic relief are directly observable in the rock record. Accordingly, terms have been proposed for biologically produced deposits that, in outcrop, are moundlike (bioherms) or else occur as discontinuous sheets or lenses parallel to the beds above and below . Dunham 1970 proposed a genetic terminology to distinguish between reefs similar to those constructed by modern corals (ecologic reefs) and those that are identifiable only as "reef-like" structures in the rock record (stratigraphic reefs). The former are presumed to be dominated by in-place and interlocking organisms, as in Wilson's 1975 "Organic Framework Reefs." The latter might be "true reefs," but they could also be the product of differential compaction along a bed or the stacking of several biostromes into a feature that only appears to have had relief at the time of deposition.

To circumvent some of these interpretive problems, Heckel 1974 proposed the more-generic and descriptive term "buildup." His "encrusted skeletal buildups" are texturally similar to modern coral reefs and are implied to be analogous. All other deposits are relegated to some other category that implies less internal organization or rigidity (i.e. "loose skeletal buildup" --> detritus near to, but not within a reef; "sorted, abraded skeletal buildup" --> a hydromechanical deposit of skeletal debris, probably located some distance from the reef).

Inherent in all of these classifications is the presumption that the interiors of modern, ecologically defined reefs are dominated by corals growing on the backs of other corals. By extension, any ancient buildup worthy of being called a "reef" must possess this interlocking and rigid interior.

With the development of lightweight drilling systems, the scientific community gained access to the interior of modern reefs. With each new set of cores, it has become increasingly apparent that the interior of modern coral reefs are dominated not by in-place and interlocking framework, but rather by variable assemblages of both whole and broken corals mixed in with sediment and debris derived from their breakdown. The relative abundance of recognizable coral recovered from cores of modern reefs is less than 20%. The other material in Caribbean reefs is sediment and open voids. These data are drawn from reefs occurring in several different oceanographic settings, and show that recognizable coral comprises less than half of the reef mass in the groups sampled. If we consider that only a portion of that recovered coral is in place, then the importance of in-place and interlocking framework is put into a much different perspective.

This is not to say that interlocking framework never occurs nor that it is impossible for a single reef to be constructed largely of in-place material. However, most of what anyone would agree are true "reefs" as they swim over them are not organized collections of in-place and interlocking corals. In extreme cases, they are disorganized "garbage piles", only indirectly controlled by the original pattern of corals. Most often, they lie somewhere in-between along a continuum. Their rigidity is only partly derived from interlocking framework; it is equally the result of broken bits falling into stable positions, as well as the secondary encrustation and cementation that bind the entire mess together.

The simplest way to integrate these emerging concepts into existing models would be to expand the limits of "true reefs" within the classifications of Heckel or Wilson. This approach could lead to considerable confusion, however, given traditional views about the importance of in-place framework in reef development. This necessitates the distasteful exercise of constructing yet another reef classification.

Carbonate buildups can be classified along a continuum from those dominated by in-place skeletons to loose piles of hydromechanically deposited sediment. Any classification of ancient reefs viewed in core or outcrop should be tied to the internal character of modern reefs. The term framework is so entrenched in the literature that suggesting its removal would be impractical. However, to reinforce the point that the elements contained within a "reef" need not be either in place or interlocking, we will use the term framework element to refer to recognizable members of the constructor guild (including encrusting algae where it dominates the fabric) that remain within the specific environment in which they lived. Based on the above discussion, the primary criteria for classification as a reef are:

As an example, we will consider the margin of Salt River submarine canyon on St. Croix in the Caribbean Sea . The reef is comprised of several genera of branching, massive and platy corals, all bound together by secondary encrusters (primarily coralline algae) and submarine cements. Many of the individual framework elements have been so reworked that their origins are difficult to infer, except in thin section. Also, the seemingly random pattern of dates between the youngest parts of the reef and its surface. reflects the constant slumping and redeposition of individual corals and large blocks throughout its accretionary history. This site has been the subject of intense study by reef scientists from at least 50 Universities and government institutions. These studys show that there is virtually no original framework that is demonstrably in place within the reef. Hubbard, et al., 1986

Similarly, a shelf-edge reef off southwestern Puerto Rico is comprised largely of toppled and encrusted A. palmata branches . The pattern of radiocarbon dates within this reef similarly implies considerable reworking of the coral colonies that comprise the structure. Despite this continual reorganization, however, individual framework elements fall into an organized pattern that reflects shoaling of the reef through time (i.e. the environment of deposition is essentially the same as that in which the organisms lived). It is likely that many, if not most, modern reefs contain significant quantities of reworked materials within their interior and that reefs dominated by the in-place and interlocking framework of traditional models are in fact quite rare.

These ideas about reefs can be integrated into a general classification of carbonate buildups . "Reefs" can be comprised of mostly in-place skeletal material (Primary Framework Reefs) or else can be dominated by reworked "framework elements" that are held together in a rigid structure only by cementation or encrustation (Secondary Framework Reefs). In the latter, the surface may be veneered by in-place organisms, but the internal fabric will likely seem chaotic, with the individual framework elements mixed or "floating" within a matrix of sediment, cement or smaller detritus. This is much more similar to "reefs" that dominate the fossil record.

The best, and perhaps only examples of primary framework reefs are the massive algal ridges that occur along high-energy margins in both the Caribbean and Pacific. Most of the Caribbean examples, however, have a nucleus of reworked A. palmata and head corals upon which the ultimate algal cap formed. Adey & Burke, 1976 The classification as primary or secondary- framework-reefs thus becomes a matter of which part of the reef you are looking at. Furthermore, the in-place elements of these features are comprised mostly of organisms normally relegated to the "binder" and not the "constructor" guild.

Most modern reefs fall along the continuum between the end-members of the reef axis in the classification figure, and are therefore secondary framework reefs. The debris that comprises a large part of the reef's internal fabric is derived from the biological breakdown of the larger framework elements by organisms seeking either food or shelter (the bioeroders). Contrary to common opinion, a significant proportion of this material may be fine-grained, as reef framework also serves as a baffle, trapping fine-grained material that is produced within the reef. The rigid structure of the reef serves to isolate this material from the high wave energy that dominates at the surface.

At the far left of figure 7.4, hydromechanical buildups are constructed by sediments that have been moved by currents from their point of origin to a site more suitable for deposition. In this regard, they behave much like sand bars and spits described in siliclastic systems; their location and size are controlled by sediment supply and local current patterns.

Some hydromechanical buildups owe their existence to the baffling action of upright organisms. As such, their occurrence is largely related to the distribution of organisms that act to disrupt water flow and trap sediment. These baffled buildups are intermediate between "reefs" and hydromechanical buildups, and reflect the mixed nature of the factors responsible (currents and organisms). Examples include the mud mounds in Florida Bay, which are built up around seagrass. Epibionts that grow on the grass blades and the upright calcareous algae that typically occupy grassy environments (i.e. Halimeda, Penecillis) contribute to the sediment budget of the features along with sediment moved in from elsewhere in the bay. The baffling action of grass blades traps sediment that is typically finer than that along the adjacent bottom. Thus, baffled buildups are often finer-grained than the surrounding benthos. Porites tolerates a high-sediment environment and is part of the biomass.

Related in their process of formation are the stromatolites that span the geologic time scale from the Paleozoic to the present. These features are developed by cyanobacteria that form mats. The mats in turn trap sediment. In the Paleozoic and Mesozoic, stromatolites formed across a wide morphologic spectrum from flat-lying or gently undulating beds to domal or columnar features of considerable relief. Occurring in a transitional zone of our classification, stromatolitic mounds are generally considered as specialized baffled buildups.

Paleozoic stromatolites dominated the tropical seascape before the evolution of herbivorous grazers. Today, stromatolites are rare and exist only in a few areas where grazers are discouraged by some outside physical factor. For example, meter-high stromatolites in the northern Bahamas are alternately buried and exhumed by carbonate sand bodies in an area of active bedform migration. Dill, et al., 1986 In deeper water, digitate corals can perform a similar function. This process has probably been important in the formation of the deep-water lithoherms described in the Straits of Florida by Neumann, et al. 1977

Off-reef buildups are special features that owe their origin to gravitational transport of sediment and debris from the reef to some downslope site of deposition. Usually formed along steep escarpments, off-reef buildups form near the base of the slope. Unlike sedimentary fans found at the distal ends of terrestrial fed submarine canyons, off-reef buildups usually occur as broad aprons along the edge of the basin. Hubbard, et al., 1974 The depth difference or distance between the reef and an off-reef buildup depends upon the magnitude of the underlying topography. As an example, sands found near the base of the northern insular slope of St. Croix contain 97% shallow-water material, mostly derived from the reefs. Hubbard, et al., 1981 Water depth is roughly 2,500 meters. In most areas, the difference is much smaller.

As in any classification, there are gray areas. For example, where do backreef sedimentary wedges or talus aprons at the base of a shallow reef fit into this scheme? While we consider them as part of the reef complex, some researchers might disagree. This is a minor semantic problem, however. Whether one considers them as a separate depositional entities or simply as a sub-facies within the larger feature, the underlying suppositions of the classification remain intact.

The type of reef formed by a particular group of organisms need not remain constant through time. For example, the rudists, large clams that dominated reefs during the Cretaceous, played very different roles in reef formation throughout their evolution. The earliest rudists sat either on or in the bottom, exerting control more like bafflers. In some instances, rudists may have been passive inhabitants of the bottom and were, as such, members of the dweller guild of Fagerstrom. Through time, rudists progressively increased both their colonial affinities and their ability to produce an integrated and cemented structure. This would have been more analogous to oyster reefs of today. At that point, they constituted more of a traditional framework element and may have even produced primary framework reefs in some instances.

Groups of organisms can produce varied buildups across either space or time. This is further complicated by the evolution of reef dwellers and builders and the increasing pitfalls of taxonomic uniformitarianism as we go back further in the fossil record. Nevertheless, by understanding the likely associations between various groups of organisms, the effects that they can have on local depositional processes, and the controls that can be exerted by underlying physical factors, we stand the best chance of constructing a realistic classification that adequately deals with both modern carbonate buildups and their fossil counterparts. If the reader comes away with nothing else, the following two ideas should be retained:

Reef types

Various criteria have been used to classify reefs; the most accepted approach is morphological grouping. The shape and location of reefs are controlled by the bottom topography upon which they formed, interactions among the resident biota, and physical processes. Darwin 1842 discussed three main types of reefs - barrier reefs, fringing reefs and atolls - still part of most classifications today. Fringing reefs occur adjacent to land with little or no separation from shore. A low input of terrigenous sediment is important, and the best-developed fringing reefs occur off shorelines where rainfall is low, there is little relief, or else the hillsides are stabilized by heavy vegetation. In recent years, clear cutting of forests and poor land management have impacted fringing reefs more than any other type. Barrier reefs are separated from the shoreline by a moderately deep (usually) body of water - the lagoon. The reef may form at the shelf edge, or it may be located more inshore, usually localized on an antecedent break in slope.

Atolls are roughly circular in plan with a central lagoon that contains no significant land mass. The central lagoon is often deep (less than 25 m), but this is not a prerequisite. If land does exist, it sits atop a part of the encircling reef and is comprised solely of carbonate material derived from the reef. As originally defined for Pacific reefs, the term implies a specific genetic origin around a volcanic island. Caribbean and Atlantic atoll-like reefs are not of this type, and tend to form around isolated highs formed by local tectonics.

The main problems with Darwin's original discussion of reef types include:

The first is easily dispensed with by decoupling reef type from tectonic history (i.e. a barrier reef is a morphologic entity, separate from its tectonic regime). The latter problem requires that a limit be set on lagoonal dimensions in fringing vs. barrier reefs. Implied in Darwin's definition is the idea that the lagoon is sufficiently large to permit open circulation behind the reef. The reef serves as a "barrier" that clearly separates lagoonal processes from those of the open ocean. Based on this, we pragmatically make the split between barrier and fringing reefs at a point when the lagoon reaches 500 m in width and 5 m in depth. In a natural setting, and in the absence of significant upland clearing of vegetation, a lagoon of this magnitude can substantially isolate the reef from direct impact by terrestrial runoff. Furthermore, circulation within the lagoon is distinctly removed from that of the open ocean beyond. We can describe some of the parameters for distinguishing different reef types.

Patch reefs are smaller features, roughly equant in plan view. While they have generally reached sea level, this is not necessarily so. Usually, patch reefs occur within the lagoon behind the barrier or atoll rim. On occasion, however, they can occur on the open shelf as pinnacles. Modern examples of exposed (i.e. non-lagoonal) patch reefs occur off the north coast of St. Croix in the Caribbean Sea. Numerous small reefs, 10-20 m across, rise out of 10-15 m of water. Their fabric of broken and piled-up coral branches has led to the local name "haystacks".

Submerged shelf-edge reefs are Caribbean platform margins that presently sit in water depths greater than 10-15 meters after being flooded by rising sea level 6,000 - 10,000 years ago. Since then, they have not been able to offset the effects of ever-deepening water, and many of them have been left behind. While coral and other calcifying organisms occur along most of these margins, they are not producing carbonate at a rate sufficient for the reef to "catch up" with sea level.

Equally problematic are reefs that occur on wider shelves (more than 5 km), and fall between the criteria for either barrier or patch reefs. They are similar to patch reefs in shape, but they are usually larger, more linear, and are aligned in roughly shore-parallel. They exist near sea level and, in some instances, have emerged to form islands. The sediments behind the reef (landward) are similar to those seaward reflecting the absence of lagoonal conditions. Because they usually occur along either insular or continental shelves, they are classified as shelf reefs.

The nature of shelf reefs changes from shore to the shelf edge. More-seaward reefs are exposed to higher wave energy. Those closer to shore come more under the influence of terrestrial sedimentation. For example, on the southern coast of Puerto Rico, the inner-shelf reefs are often subjected to fine-grained sediments derived from the adjacent hillsides . As a result, they are mostly mud mounds with scattered corals. In some instances, they have been stabilized by mangroves and have built small islands. The mid-shelf reefs are subject to the effects of open-ocean circulation and more wave action. Accordingly, coral cover is higher and the benthic-community structure is more complex.

Morphology of Caribbean reefs

The morphologic terminology developed for Discovery Bay, Jamaica by Goreau and Land 1974 is often referred to as "typical" for Caribbean reefs . While variations occur from place to place, this general scheme is a reasonable place to start, and each element of this profile is discussed below.

Shallow Reef

The crest of the main reef is generally emergent in Pacific reefs at low tide, but may be below the surface in Caribbean reefs. The seaward edge of the reef crest takes the brunt of the incoming wave energy. Roberts 1989 has shown that the reef can reduce incoming wave energy by up to 97%. As waves break, water is washed across the reef crest and into the lagoon, driving lagoonal circulation. Hubbard, et al., 1981

Because of the modification of wave forces across the reef crest, the backreef is an environment of totally different physical processes, ecology and sediment characteristics. Sediments and rubble from the reef crest are dumped behind the crest, widening the backreef flat through time. The outer reefs of the Great Barrier Reef have been at sea level for nearly 6,000 years. Hence, the wide backreef flats often exhibit distinctive front-to-back zonation. By comparison, Caribbean reef flats have only recently reached sea level and are narrower. While zonation is less pronounced, there is a general transition from branching corals and the hydrozoan Millepora near the front of the crest to sand flats and Thalassia landward. The shallow back reef may have a shallow Porites reef flat immediately behind the crest and numerous small patch reefs in a sand apron. The corals are generally well adapted to the high levels of sedimentation to which they are regularly subjected. In the Caribbean, the dominant corals include Porites porites and several head corals, especially Montastrea annularis, Porites asteroides and species of Diploria.

Forereef

The forereef extends seaward and downward from the reef crest. It is the most complex of the reef zones, owing to the large depth gradient over which it occurs. In many areas, the forereef is organized into a set of en-echelon reef promontories and sand channels, termed "spur-and-groove" topography.

Spur-and-groove is common in both modern and ancient reefs. The term was originally coined from Indo-Pacific examples formed by erosion of the algal rim just below the surf zone. More recently, examples have been described from the Caribbean that appear to be the result of accretion by Acropora palmata under the influence of strong wave surge. Both the coral branches and the intervening sand channels are oriented parallel to the dominant wave-approach direction. Shinn, 1963; Roberts, 1974 Hubbard, et al. 1974 proposed that the channels serve as primary conduits for sediment export from the reef. They further proposed that spur-and-groove topography will be best-developed along windward margins where a barrier exists to bankward transport, and downslope sediment movement is the only means of export

Forereef Slope and Deep Forereef

The forereef slope is the least consistent of any of the reef zones, in either its occurrence or character. At many sites, it is totally absent and the forereef drops from shallow water to oceanic depths. Near Discovery Bay, it occurs as a gently sloping sand plain, separating the shallower reef from its deeper counterpart by several tens of meters. A nearly identical feature occurs along the southern margin of Little Bahama Bank and many other Caribbean sites. Its persistent depth throughout the region (ca. 30-35 m) may imply control by an underlying terrace cut at a lower stage of sea level.

Where a forereef slope is present, the deep forereef usually occurs as a well-defined ridge near the platform margin. Otherwise, it is simply a down-dip extension of the forereef. When occurring separate from the shallower reef zones, the location of the deep forereef is probably controlled by both the break in slope and the existence of an antecedent high left by a previous reef. The character of the reef surface is often similar to the spur-and-groove topography described above, except that the scale of both the reef promontories and the intervening channels is generally larger.

The Reef Wall

Perhaps the most dramatic feature of the deep forereef is the "reef wall". At depths ranging from 50 to 85 meters around the Caribbean, the forereef slope rolls over to a vertical or, in some places, overhanging precipice . The role of active accretion at this depth is not well understood, owing to its remoteness. Along the high-energy margin of the Great Barrier Reef, coral cover is apparently limited to a very thin veneer over the antecedent Pleistocene reef front. Isdale, 1984 Along the front of the Belize barrier reef, several episodes of reef-wall accretion have probably occurred. James & Ginsburg, 1979 On the north coast of St. Croix, accumulation of reefal debris has resulted in significant (greater than 26 m) lateral accretion in water depths of up to 30 m, primarily the result of repeated slumps stacked one in front of another.

Caribbean reef zonation

Early discussions of reef zonation were based on Caribbean reefs and, therefore, reflect their species composition. A profile across a "typical" Caribbean reef shows both morphological and species zonation. The following is a generalized description of species patterns keyed to that profile. It should be kept in mind that the distribution of reef inhabitants can vary from this composite picture in response to the physical-oceanographic processes that are discussed later on in this chapter.

The Acropora palmata Zone

Along the front of moderate- to high-energy reefs, the dominance of Acropora palmata to depths of 5-10 m is primarily a response to wave energy. While, not as strong as head corals, Acropora palmata orients its branches to minimize tensional loading - the higher the current flow, the lower is the profile presented by the branches of the colony. Acropora palmata is capable of rapid growth, which elevates the colony above the traction carpet of shifting sediment that, can "sandblast" polyps sitting closer to the bed. Because Acropora palmata is extremely sensitive to sedimentation and has no physiological mechanism for sediment removal, wave-induced surge performs this important function for the coral.

The Massive Coral Zone

Usually starting at depths of 5 - 10 m, head corals increase in importance; notable among these are Montastrea Annularis, Montastrea cavernosa, Porites astreoides, Colpophyllia natans and species of Diploria. In some areas, the branching coral Acropora cervicornis can occur as a narrow belt between this and the shallower Acropora palmata zone. Both their depth of occurrence and their massive nature make head corals in this zone more resistant to periodic disturbance by storms. As a result, this zone is often the best represented in cores from Holocene reefs throughout the Caribbean region.

Platy Coral Zone

At depth, most corals are platelike , an adaptation that concentrates the photo-receptive algae contained within their tissues along the upper surface of the colony. In shallow water, this fragile shape is disadvantageous from a structural standpoint. In deeper water, however, physical breakage is low and the need to gather light dominates. The primary species at depth vary from place to place, but usually include members of the genus Agaricia and flattened colonies of Montastrea annularis.