GCRA  OVERVIEW  FAQ  NEWS  ARTICLES  PHOTOS  REEF ISSUES  RESTORATION  PAPERS  LINKS 

ASSOCIATION OF MARINE LABS OF THE CARIBBEAN

JUNE 4-6 2007, ST. THOMAS

Turks and Caicos Islands 2006 coral reef assessment: Large-scale environmental and ecological interactions and their management implications

Thomas J. Goreau¹, Tatum Fisher², Fernando Perez², Kathy Lockhart², & Andrew Lewin³

¹Global Coral Reef Alliance, 37 Pleasant St., Cambridge, MA 02139 USA; goreau@bestweb.net

²Department of Environment and Coastal Resources, Turks and Caicos Islands

³ LGL Limited Environmental Research Associates, King City, ONT, Canada

Abstract: Coral reef health is critical for Turks and Caicos Islands (TCI) fisheries, tourism, and shore protection from global warming and sea level rise. TCI has some of the best remaining reef in the Caribbean region, but corals are declining from episodic damage and progressively increasing environmental stress. Traditional assessment methods provide relatively little information because they cover too little area to accurately characterize complex reef ecosystems, miss large-scale patterns, and high intrinsic spatial variability of reefs inhibits the ability of data from small areas for identifying changes or causes. Spatially extensive surveys provide more information, over larger areas, in less time, than intensive ones, revealing large-scale gradients that intensive methods are inherently unable to identify. Extensive surveys were carried out during early June 2006 by diving, snorkeling, towing, from shore, and from the air, covering North Providenciales from the northern west coast to eastern north coast, from southwest South Caicos to southeast East Caicos, and all around Grand Turk, including Gibbs Cay, and Round Cay. Since algae cover exceeds coral cover, and algae species are more sharply spatially zoned than corals, they convey more environmental information useful for management, so a special focus of the survey was identifying algae zonation across environmental gradients, rarely considered in reef assessment. 26 ecological and environmental parameters were assessed at 47 coral reef sites across the Turks and Caicos Islands (TCI). Live coral cover averaged 10-20%, maximum around 40%. Dead coral exceeded live coral at all sites. Several past mortality events from disease and high temperature were identified. There was little hurricane damage in most locations: old dead elkhorn corals were largely intact, but large areas of staghorn and finger coral rubble were seen in lagoons. Dead offshore elkhorn reefs, which protect the shoreline from erosion, are slowly crumbling from boring organisms. The intensity and frequency of hurricanes, along with sea level rise, will increase with global warming, greatly increasing future beach erosion. Stress from sedimentation was seen down-current from the Cruise Ship Port dredging site, and inshore where sediments are re-suspended by waves. Unexpectedly frequent, severe, but localized, coral disease was seen at many locations, often appearing in intense patches. Non-parametric statistical analysis of the data revealed that many diseases correlate with certain species of algae, which might be reservoirs for pathogens. High algae abundance was seen at almost all sites, with zoned gradients pointing towards nutrient sources. Most dive sites had little algae in shallows, but algae increased dramatically at the drop off edge into deeper water to the thermocline. In many shallow reefs algae indicated land-based nutrient sources from contamination by outflow of creeks, marinas, salinas, fish processing plants, hatcheries, or groundwater seepage. Surprisingly, very high levels of algae were also seen at many sites with no land-based nutrient sources. Spatial zonation of algae suggest nutrient sources from persistent localized upwelling of cold, deep, nutrient-rich water offshore. Green chlorophyll-rich surface water was noted at most sites affected by land-based sources as well as offshore upwelling, suggesting widespread natural nutrient stress backgrounds. Transects were chosen parallel and perpendicular to major environmental gradients and to span their ranges. Correlations between all pairs of ecological and environmental variables were assessed by non-parametric statistics because few variables are normally distributed. The matrix of all correlations and significance was calculated, providing improved insight into interactions between species and their relations to environmental gradients. Previously known interactions related to reef deterioration were confirmed, but many unexpected correlations linked to coral diseases and algae species were also identified, that require further research. Extensive methods are far more cost effective than intensive ones and provide a new paradigm that should be much more widely used for management purposes. The TCI data are being used to develop national water quality management and restoration strategies.

Key words: large-scale coral reef ecological assessment, environmental gradient analysis, non-parametric statistics, coastal zone management, algae zonation, coral community assemblages, diseases, bleaching

INTRODUCTION

Turks and Caicos Islands: The Turks and Caicos Islands (TCI) lie at the extreme southeast end of the Bahama Islands chain, but form a separate country (Fig. 1.). The islands are on the edges or two shallow banks (Fig. 2). Most Turks and Caicos Islanders have always lived from the sea, whether from salt, fisheries, or tourism (Sadler 1997). TCI has developed the strongest coastal zone management in the Caribbean in order to save their marine resources. There is widespread awareness public awareness and support, since the current tourism and financial service economies are vulnerable to the uncontrollable whims of political forces and events outside TCI, and the people could once again be forced to live exclusively from the sea as they did in the past. TCI has some the strongest coral reef protection policies in the world in terms of the proportion of reefs with protected status. Although the Department of Environment and Coastal Resources is small and under-funded, the level of enforcement of environmental laws is among the most effective anywhere: TCI is the only country in the world that has adopted coral reef specific water quality standards, and which ensures that all developments recycle all their wastewater on their own property as plant irrigation to prevent land-based sources of pollution that could damage reefs and fisheries.

Nevertheless, as the reefs continue to decline, there is little information on the current status of the reefs, the changes they are undergoing, or the stresses that cause them, which could be used to guide coastal resource management.  Planning for the future will require learning from the past, correctly assessing the causes of current trends, and anticipating the future pro-actively. The purpose of this study was to assess current reef health trends in the light of past observations, and use them as a guide to prepare a strategy that protects and enhances TCI’s future fisheries, tourism, and shore protection needs. This study was carried out at the request of the Department of Environment and Coastal Resources, approved by the National Parks Advisory Committee, and funded by the Community Conservation Projects Programme.

Intensive versus extensive assessment: The normal methods of coral reef assessment, such as those used in the 1999 AGRRA surveys, use counts of corals (or fish) and estimates of total algae abundance along short line transects or in small quadrat frames. We refer to these as “intensive” methods in this paper. Intensive methods give quantitative information, but are extremely laborious, take a lot of time, and cover only a very tiny part of the reef being looked at. Experience with these methods shows coral reefs are so variable a habitat that totally different results can be obtained by placing the lines or quadrats just a few meters away. Vast numbers of such measurements must be made before the average measured value converges on the true mean for the reef, which is impractical due to time and funding constraints. That is why after such methods were first tried in Jamaica in the 1960s, they were abandoned as being not worth the time they took for so little information. Intensive method data give the impression of high precision, but can in fact be wildly unrepresentative of larger areas and so are often very inaccurate at characterizing the habitat. The high intrinsic variability of the measurements and the extremely small areas measured intrinsically prevent seeing larger scale ecological gradients, yet these larger scales are usually the dominant feature of reefs ecosystems. Large amounts of time and effort can be spent using intensive methods, often without seeing the forest for the trees. No forest ecologist would dare to use them to make the sorts of claims that reef ecologists routinely do, because they would lose all credibility with their colleagues.

Four brief examples of the ways intensive methods can lead to misleading or inaccurate results are given from the first author’s personal experience. They illustrate some limitations with intensive methods even in the hands of excellent researchers.

1) In Aitutaki, Cook Islands, T. Goreau and R. Hayes surveyed coral bleaching on reefs during 1994, simultaneously with groups doing line transects on the same dives. Due to the time needed to run the transect lines, they were limited to shallow water, and laid lines along the 5 meter contour. While they ran a 100 meter transect at this depth, the TG and RH would rapidly survey a vastly larger area of reef, from 40 m to the surface. They noticed that the areas shallower than 10 m had very little coral cover, typically around 20% or less, apparently the result of old typhoon wave damage and Acanthaster planci predation, but in deeper waters than those seen by the other group live coral cover was around 80% and much more diverse, with larger and different species than those documented in the shallow line transects. 

2) In Tutuila, American Samoa, TG and RH shortly afterwards looked at a reef with a wide platform, on which coral cover ranged from zero at the shoreline to about 80% on the outer lip of the platform, so they described the mean coral cover of the site as around 40%, plus or minus 40%. This number was greeted with outrage and derision by certain local experts, who said that they had run a line transect in the same bay and measured it to be 18.73%, so that the extensive estimate must be wrong.

3) In Port Antonio, Jamaica, TG dived with a student working on the relationship of Long Black Spined Sea Urchins (Diadema antillarum) and algae. While the student ran a single 50 meter transect in a nearshore seagrass bed counting sea urchins and algae, TG swam for several miles and looked at around 8 completely different reef habitats from shallow to deep water. On returning he asked the student what his findings were, and was told that the more sea urchins were packed together the less algae there were. However a look at all the sites he had not seen would have led to exactly the opposite conclusion. The student’s site was the only site that had either high sea urchins or high algae abundance, and it was the most sewage polluted.

4) In the Seychelles, Indian Ocean, TG worked with the permanent transects established by the Marine Park Authority in the national Marine Protected Areas. Running each of the 100 m shallow transect lines required several dives to record the needed information, and there were problems with species identification. In a single dive TG could take a permanent digital video record of every single coral along the transect in about 10 minutes, and could then run several more similar transects at in both deeper and at shallow levels and other areas on one tank, giving many more useful information in far less time. Because the same transects were filmed before, during, and after the 1998 bleaching event that killed most corals in the Seychelles, it was possible to quantitatively compare the survival and mortality rates of all of the corals species filmed from their time series images, using much larger areas than intensive methods allowed.

A diametrically different philosophy of environmental assessment to the intensive, small-scale counting methods normally used is to make extensive, large-scale studies of ecological gradients. In the same time it takes to count everything along a single short line or small square, a trained diver can swim over the entire reef and see large-scale spatial patterns that could never be identified by intensive methods, even with many years of work. Rather than trying to count everything, observers using extensive methods look for changes in the abundances of major groups of organisms along long distances perpendicular to and parallel to the major environmental gradients, such as distance from shore, depth, wave energy, population centers, nutrient sources, etc. and covering as much ground and different habitats as possible. Special notice is paid to the average abundances of major groups along the gradients and of the presence of unusual numbers of any organisms at specific sites. These methods are those used in the original 1950s studies by T. F. Goreau first describing the ecological zonation of coral reefs, and in drawing up marine protected area plans to protect reefs in Ocho Rios, Montego Bay, and Negril, in Jamaica, St. John in the Virgin Islands, and Buccoo Reef in Tobago.

Because the purpose of extensive surveys is to cover as much area as possible, detailed counts are not made as in intensive surveys. Instead highly experienced reef ecologists mentally average the abundances they see along the gradient and team members frequently compare their observations and notes right after each dive. If possible the entire transect is digitally filmed or photographed. Coverage of major groups of organisms, such as live coral cover or different types of algae, is estimated to the nearest 10%. While the intensive method gives the appearance of greater precision, in fact differences of less than 10% cover are almost never really significant, and significant differences in abundance can easily be estimated visually to the nearest 10% by experienced observers. Comparison shows that experienced observers agree on their coverage estimates almost always, and only rarely will two experienced observers disagree by 10% because they were looking at a somewhat different field of view even when swimming nearby. By discussing their findings at the end of each set of observations, the team has always found it possible to agree on a satisfactory average value that is consistent with the observations of all members of the team.

Extensive methods are much more informative if time and funding is limited, cover many times larger areas, and are therefore ideally suited to developing countries who must maximize information at minimum cost. It can reveal large-scale patterns that are not likely to emerge from much more expensive intensive assessment methods. Extensive surveys integrate the effects of patches that prevent intensive surveys from seeing the larger patterns, not seeing the forest for the trees. While intensive methods appear to be very precise, they are often inaccurate representations of the larger systems they are embedded in. In contrast, extensive methods are less precise but more accurate on larger scales, and most importantly, they integrate over patchiness of many scales. Nevertheless, extensive assessment has fallen out of fashion since 1970 because of the (incorrect) appearance of quantitative accuracy that the currently popular intensive methods appear to offer. Users of intensive methodologies regard their surveys as “quantitative” and “scientific”, and discount extensive observations as “qualitative” or “merely anecdotal”. There is a need to provide an objective and quantitative framework for analyzing extensive observations that allows the unique power of large-scale gradient observations to be expressed in a statistically significant way.  The use of non-parametric statistical analysis provides a robust way to do so. Usually parametric statistics are used to test a priori hypotheses, but can completely miss patterns that are in the data but are not specifically tested for. In this paper we are not testing prior hypotheses, we are simply looking how every pair of ecological or environmental variables assessed are related to each other over all the sites. Only then do we form a posteriori hypotheses to explain the linkages between variables found in the data. Thus this method is more consistent with mining data from natural history observations to generate testable hypotheses, than the conventional “hypothesis driven” science, which presumes what expected relationship is being tested.

Conventional parametric statistical analysis provides a mathematical framework for comparing data as long as it can be assumed that each variable being counted can be described by only two parameters, the mean value and the standard deviation. This requires assuming that the variability of all values around their mean is described by the “normal” bell-shaped Gaussian distribution curve. If the data does not in fact match this assumption, parametric methods can lead to wildly incorrect conclusions about the associations between groups. Ecological and environmental data are very rarely “normal”. The abundances of many organisms are often highly skewed, so that they are either extremely rare or extremely abundant. In such cases, the mean value and the standard deviation are fairly meaningless descriptions, and provide completely inaccurate impressions of the typical abundance of the population. Non-parametric statistics provides a way to draw strong statistical inferences about the relationships of variable parameters that do not fit “normal” distributions. The value of each variable is ranked qualitatively along a gradient. These can be average bottom cover estimates to the nearest 10% or qualitative categories, such as “very high”, “high”, “medium”, “low”, or “very low” as long as they are used in a consistent way between sites by the same observers. This allows very rapid and efficient data to be collected over large areas that are impossible with intensive methods, and for statistically sound relationships between them to be derived.

Non-parametric methods have only infrequently been used for extensive assessment of coral reef health. By comparing the Spearman rank order correlation statistics of many ecological and environmental parameters at many sites each pair of variables can be evaluated as to whether they are positively or negatively correlated and the degree of statistical significance evaluated. The long term changes of corals and algae at sites all around Jamaica were assessed with regard to climatic and anthropogenic stresses, and found that areas with the highest algae overgrowth of reefs were near the largest land-based sources of nutrients (Goreau, 1992). Goreau and Hayes (1995) examined reefs over a large area of the South Pacific, and found that several parameters were strongly negatively correlated with live coral cover, and that most of these parameters strongly positively correlated with the density of human activities on land and in the sea, but none of them correlated with bleaching, except for high temperatures alone. They concluded that many human activities were damaging corals, but bleaching was related only to high temperature stress, probably caused by global warming.  

Previous work: Although TCI has long been regarded one of the Caribbean’s top dive destinations, there is surprisingly little material available about the condition of the reefs apart from descriptions of the major dive sites (Gascoine & Lott 1991, Rosenberg 2001;,Harrigan, 1992). An assessment of TCI natural resources, including some reef sites, was done in the 1980s by Project Raleigh, and T. Topalian discussed the condition of TCI reefs in 1992, but copies of these reports could not be found for this study. Sullivan et al. (1994) identified coral species in areas of just 15-30 square meters at 5 locations on shelf edge reefs near South Caicos. A description of reefs at sites of Western Grand Turk and near the Bight, Providenciales was included in a DECR report by Goreau (1995). TCI was found to have some of the best reefs in the Caribbean, with live coral cover up to 70%, lots of groupers and big fish, no algae in shallow water except in front of large hotels, Lobophora algae only on deep walls, and conch and lobster stocks increasing. Coral reef specific water quality standards were recommended.

Although the Center for Field Studies has operated a teaching and research program in South Caicos for many years, little has been published on coral reef assessment except for one paper on coral species in reefs near Long Cay, South Caicos (Steiner 1999). G. Gaudian and P. Medley, working with DECR, published a paper on diver damage and diver carrying capacity in Grand Turk (1995) and a chapter summarizing the condition of TCI and Bahamas reefs in a book (2000), which described TCI reefs as largely “pristine”. A study of water quality has been done in the Princess Alexandra Park, Providenciales (Perez 2000a, 2000b). A report on reef assessment at major dive sites has been compiled by DECR (T. Fisher 2004).

The most comprehensive previous reef survey in the Turks and Caicos was carried out in 1999 (Riegl et al. 2003, Hoshino et al. 2003). That study examined 28 locations, most of them close to or similar to those examined in this work, and so providing a useful base for comparison. They found an average live coral cover of only 18% (range 8-28%). Because they regarded TCI reefs as “pristine”, they suggested that the reefs must be healthy but that coral cover was low for some unknown natural reason, such as hurricanes. They noted large amounts of standing dead coral in many shallow areas, which is inconsistent with hurricane damage, but did not suggest when or why these corals had died. They noted that algae cover was generally high, and higher on windward than leeward reefs, but offered no explanation of this unexpected pattern.

METHODS

Field and analytical methodology: In this study, 26 different ecological and environmental parameters were assessed at each of 47 different locations throughout the Turks and Caicos. Maps showing all of the locations examined are in Fig. 3-8. All sites were assessed by a team of four (T. Goreau of the Global Coral Reef Alliance, and T. Fisher, F. Perez, and K. Lockhart of the TCI Department of Environment and Coastal Resources). At each site long swims (often several kilometers at a stretch) were taken covering the widest range of conditions at each site, usually as onshore-offshore transects or from land-based sources of nutrients along the shore both in the upcurrent and downcurrent directions from key possible sources of stress. When diving, an effort was made to cover the largest depth range possible. Deep diving and shallow snorkeling sites were picked to represent continuous spatial gradients. Long distances were traveled by being towed behind a boat on a rope, both parallel to the shore and along onshore-offshore gradients to look at larger spatial patterns using the tow surveying method developed in 1969 (Goreau et al. 1972), later popularly re-named ‘Manta Tow”. At each location one member of the team took photographs including wide angle photographs to show the general conditions of the site, and close-ups of the most abundant or the most unusual corals, algae, and invertebrate organisms at each site, making a special effort to document coral reef diseases or bleaching.  Corals were identified to species and algae generally to genus level. Because algae were dominant at almost all sites, a special focus was on the zonation patterns found in the most abundant algae species. It was not possible to take photographs during tows, as both hands were needed to hang onto the rope, and at a few sites the camera did not work due to battery problems. DECR assessments in the past had focused on major dive sites (T. Fisher 2004), fisheries stock assessment (largely for conch and lobster, K. Lockhart, personal communication), and at environmental engineering problem response sites (largely pollution or dredging related, F. Perez, personal communication), Since the purpose of this assessment was to look at large scale ecological patterns, this gave the DECR staff a first chance to look at places that they had never dived at before. In many ways they found the range of conditions at many sites surprisingly different from those they had experienced previously. Following the dive assessments, meetings were held with the oldest divers on Providenciales and Grand Turk to discuss their observations on historical changes.

After each dive, snorkel, or tow, the team drafted group notes comparing their observations. Notes indicated the typical coverage of the site, noted any spatial gradients, and any unusual organisms seen. At the end of the field studies, the notes and the photographs were used by the team to discuss each site, and to rank the chosen ecological and environmental parameters in a consistent way at all sites. The parameters chosen in this study were all readily visually estimated. Quantitatively measured parameters could also be included in this analytical scheme, but the team had no quantitative measuring instruments, except a boat mounted GPS for part of it. Therefore the parameters selected included the major groups of organisms and bottom types, environmental parameters such as depth, distance from shore, wave energy and exposure, turbidity, water color, and distance from land-based sources of nutrients such as cess pits, septic tanks, other waste water treatment systems, garbage dumps, channels from harbors or lagoons, and the abundance of key indicator algae species.

While there are several Caribbean corals that are characteristic of very shallow or deep sites, most Caribbean corals are found in a wide range of depths, and the major zonation was historically caused by the former near complete domination of Elkhorn (Acropora palmata) and Staghorn (Acropora cervicornis) in the shallow wave breaking zone, with other coral species dominating in the shallow lagoon behind or on the deeper reef in front (Goreau 1956, 1959, Goreau and Goreau 1973). This zonation was very common in TCI in the 1970s but is now only very weakly present due to the very small amount of Acropora remaining. Since algae dominated most reefs in 2006, and since algae species are very sharply zoned along ecological gradients (Goreau 1992), algae species were used as the primary indicators of zonation rather than coral species. Indicator algae species that were picked for ranking were macrophyte (larger algae not needing a microscope for identification) species that were very abundant at several places, and that could be quickly ranked at all sites. Several very distinctly different algae communities were found at different sites, some associated with land based sources of nutrients, others with offshore exposure, and others lying in between. In some cases particular species of algae were found to be dominant at a single site, and rarely or never found at other sites. These very locally abundant but uncommonly found species were noted, but were not used as indicators because the number of sites where they were found was too rare to be statistically significant.

All parameters were ranked on a scale from 1 to 5. The parameters chosen for comparative assessment at all sites were:

1.     Live coral cover. The best indicator of reef health

2.     Bare cover. Clean hard rock and dead coral not covered by algae, suitable for new coral settlement

3.     Padina. Algae typical of shallow low nutrient areas

4.     Lobophora. Algae typical of deeper and of nutrient rich areas

5.     Sypopodium. Algae typical of moderately high wave energy

6.     Sargassum. Algae often dominant in high wave energy areas

7.     Turbinaria. Algae often found in high wave energy areas

8.     Dictyota. Very common algae that spreads at low to moderate nutrients

9.     Crustose coralline red calcareous algae. Algae typical of very low nutrients

10.    Branching red calcareous algae. Algae typical of low to moderate nutrients

11.    Laurencia. Algae typical of low to moderate nutrients

12.    Microdictyon. Algae typical of high nutrients

13.    Cladophoropsis. Algae typical of high nutrients

14.    Calcareous green calcareous algae. Algae typical of low to moderate nutrients, major beach sand source

15.    Cyanobacteria. Algae typical of very high nutrients, especially phosphorus

16.    Bleaching. Coral alive but pale and severely stressed, usually by high temperature

17.    White Plague. The disease that spreads fastest, affects most corals

18.    Black Band. Disease affecting massive corals

19.    Gorgonian disease. Disease affecting sea fans and sea whips

20.    Exposure. Degree of wave exposure, depends on geographic orientation of shoreline

21.    Depth

22.    Development. Amount of population and infrastructure on nearest shoreline

23.    Turbidity. Low visibility due to suspended sediments from dredging or wave action

24.    Green water. Low visibility due to high levels of phytoplankton

25.    Point source. Proximity to major sources of nutrients from septic tanks, canals draining salinas or creeks, marina channels, or fish processing plants

26.    Seepage. Ground water flow from inland salinas, lagoons, cesspits, septic tanks, and garbage dumps.

RESULTS

The non-parametric statistical analysis of the data, showing the Spearman correlation coefficients between each pair of variables from the rank order listings at all sites, and whether these variables are positively or negatively correlated with each other, as well as the degree of statistical significance, is shown in Table 1. Detailed spreadsheets of the field data, the field notes, the field photographs, and the 1995 Goreau report will be available as appendices on the web posted at www.globalcoral.org

An overview of the general field observations with regard to coral abundance, coral health, algae abundance and distribution, water quality with regard to water color, sediment, and pollution, and shore erosion is provided below. In addition to the underwater observations, observations of the reefs were made from the air. These are discussed below in a separate section. This section is followed by an analysis of the results of the patterns that emerged from the data analysis, many of which were not previously noticed..

Coral cover:  Live coral cover was surprisingly low at all sites. All sites had more dead coral than live coral, and in many the live coral was extremely low. Most reef sites had live coral cover between 10 and 20%. The highest values for live coral cover seen in TCI reefs were around 40% at the healthiest shelf edge wall and slope reefs (the best healthy coral reefs should have approaching 100% live coral cover). A large fraction of the dead coral appeared to have died within the last 5-10 years based on the preservation of the dead coral surface and the height difference between live and dead portions of the partially dead colonies. It is thought likely that much of this partial mortality was due to bleaching and coral disease. Dives in many of the best locations on Grand Turk in 1995 had found live coral cover of up to 70% (Goreau 1995), implying that nearly half the corals had died at the best sites in 11 years. In 1995 a coral mortality event about 5 years old was clearly visible, probably due to the severe bleaching events in 1990 and 1987, which had been the hottest years in TCI up till that time. Satellite derived monthly average sea surface temperatures for TCI from 1982 through 2003 (Fig. 9, from Goreau & Hayes submitted), shows that record high temperatures were reached in 1998, and almost every year since then has equaled or exceeded the temperatures that caused bleaching in 1987 and 2000. The dead portions of large corals seen in this study, which appeared to have died 5-10 years previously based on their height differences, are most likely to have died after bleaching in 1998. In some shallow locations, for example southeast of Grand Turk, almost all large head corals were dead on top but alive on the sides, a characteristic typical of partial bleaching mortality. An opposing pattern is often caused by White Band Disease, which usually proceeds from the bottom up. Only current bleaching, not likely past bleaching mortality, was tabulated. Bleaching was distinguished from disease, which can often be confused with it, by the presence of live, but pale, tissue over the skeleton, and the characteristic species-specific gradients of color seen across bleached corals. The bleaching seen was residual slow recovery from bleaching in 2005 (Fig. 10). This recovery appeared to be much slower in some sites than others, but without knowing the intensity of bleaching at its peak the previous year at each site, it is hard to be sure. The high, but patchy, older coral mortality after the 1998 bleaching event observed in this study was apparently not recognized in the 1999 survey by Riegl et al. A healthy patch of staghorn coral that had been found at the southern tip of Grand Turk in 1995 was completely gone 11 years later. It is clear that there has been a progressive decline in live coral cover over the past 15-20 years at all sites. A large amount of even older coral mortality was clearly visible at many sites. Huge dead standing elkhorn coral reefs were seen at many locations in northwestern Providenciales, northern Grand Turk, and the islands southeast of Grand Turk. Some locations had a few younger live elkhorn corals, but with very little overall recovery compared to the amount of old dead elkhorn at the same sites. Large amounts of dead staghorn rubble were found at some locations on northern Grand Turk. Staghorn and elkhorn coral were once very abundant, but are now rare everywhere. This has serious environmental implications because the open branching of these corals provides the best habitat and hiding places for most reef fish, the best snorkeling areas, and because they grow in shallow water and are the best corals at breaking waves, for absorbing wave energy, and reducing beach erosion at the shoreline. Discussions with the earliest divers, who founded the first dive operations in Grand Turk 1980s, revealed that these areas were already dead when they had started diving. The first dive operations in Providenciales were set up in the 1970s, but the divers concentrated on fishing and did not pay attention to corals. It is quite certain that these huge coral formations died during the Caribbean-wide epidemic of White Band Disease that killed almost all Elkhorn and Staghorn throughout the Caribbean around 1979 (Gladfelter 1982). Unfortunately there are no known older observations and photographs that would help establish the date, although interviews with the oldest fishermen should provide this information. The low coral cover was surprising, because TCI reefs are among the best in the Caribbean in terms of coral cover and because they have much lower sources of land-based sources of pollution than almost any other islands, due to the low population and semi-arid climate, which causes very low runoff. TCI reefs are widely regarded as “pristine” but it is clear that the corals are dying much faster than new coral colonies can colonize the area, as seen by the large amounts of bare substrate at some locations that is ideal for coral settlement. Recruitment of new coral larvae from planktonic sources is probably very low because there are few areas with good sources of coral larvae upcurrent from TCI due to its exposed open Atlantic location. The major sources of stress to the reefs, while linked to development, dredging, and land based sources of pollution at several shallow reef sites, appears to be largely due to global warming and new diseases, and oceanographic factors, none of which can be controlled by local measures.

Coral diseases:  Epidemic levels of coral diseases were found at most locations. The most abundant was White Plague (Fig. 11), which was seen at almost all sites. At most locations only isolated cases were found, but at many locations patches of reef in which a large portion of corals had White Plague, separated by intervening areas with few or no cases, were found. In some of these patches most of the coral species were affected. While massive corals were generally most affected, most coral species were found to have some colonies being killed by White Plague, including massive, branching, and plate corals. We counted only cases in which White Plague was actively killing corals at the time of the observation, indicated by large areas of fresh white skeleton that had only lost its tissue in the last few days, before algae could overgrow it. But we did also see many corals with older dead portions that looked as though they had died from white plague previously, as noted from the typical bottom-up mortality pattern, but that the disease had stopped before killing the entire colony (often with totally dead corals nearby). White Plague, while it will completely kill smaller corals, often stops before it completely kills larger corals, leaving dead patches of characteristic shape, usually on the sides. White Plague may be seasonal if, as seen elsewhere, White Plague spreads most rapidly in the warmer months and stops when water gets cooler. Without catching diseases in the active stage, it is very difficult to assign older mortality, so we did not include older partial mortality in our White Plague estimates. White Plague was so widespread, and so locally abundant at many places, that it is likely to be the major killer of TCI corals at this time. Black Band disease (Fig. 12) was also found to be locally very abundant, but at a smaller number of places than White Plague. This disease kills corals at a few centimeters a month (while White Plague does so at several centimeters a day) and affects a smaller number of species, primarily the large head corals. Most sites with a lot of Black Band were very remote from human pollution sources, in fact it was almost never seen at places where land-based pollution is a possible stress factor. However in many cases the white recently dead area behind the black band was very broad, up to ten times or more wider than normal, making it appear as if it were combined with White Plague. The same was true of White Band Disease, which was generally fairly uncommon because of the rarity of the affected elkhorn and staghorn corals. In contrast to White Plague and Black Band, Yellow Band, Dark Spot, White Pox, and White Spot Diseases were present but far less common, usually isolated, and rarely aggregated into intense patches. Gorgonian Disease (Fig. 13) was seen at high levels in sea fans and sea whips, but only in a small number of places, and most locations showed no signs of either. In some places large amounts of dead sea whip remains were found piled up on the bottom. Coralline Algae Disease was seen in isolated places wherever encrusting coralline red algae were abundant, and one case of possible Coralline Lethal Orange Disease was seen. Given that these observations were made in June, before the water warms up to the usual seasonal maximum, it is possible that the disease outbreaks noted were in their early stages, and so it will be very important to continue observations through the rest of the warm season and the following cool and warm seasons.

Algae: Algae were the dominant cover on hard bottom at almost all sites (Fig. 14), but the species dominance varied greatly from site to site (for details see field notes and data logs posted at www.globalcoral.org). High presence of algae indicates the nutrient supply feeding them promotes faster growth than algae eating fish and other organisms can consume. This finding was extremely surprising because land-based nutrient sources in TCI are relatively small and localized, and the fish populations are some of the healthiest in the Caribbean. Algae are the best indicators of environmental conditions, with different species being most abundant under different conditions of nutrients, light, and wave energy. While a few algae species were very broadly distributed, such as Dictyota and the calcareous sand-producing green Halimeda, most algae were clearly zoned, being most abundant in certain habitats and rare in others. Algae covered most of the hard rock surface and dead corals almost everywhere, preventing the possibility of new coral settlement. Clean white limestone rock surfaces that had been noted previously in shallow Grand Turk reef sites (Goreau 1995) were largely covered with macro-algae or micro-algae turf at most locations in this survey. The exceptional locations with very low algae cover were found at sites south and east of South Caicos, certain sites on Providenciales and southwest Grand Turk, and patches on shallow fore reefs with very high wave energy exposure. There were distinct regions where both shallow and deep reefs had high (or both had low) algae. Algae indicative of very high nutrients were focused where land-based point sources of pollution and groundwater seepage enter the sea. The factors affecting the distribution of the different major algae groups are discussed in further sections, especially the statistical data analysis, and in the appendix of detailed site descriptions. One striking feature in the algae, as previously noted (Goreau 1995) was the strong increase in Lobophora variegata and calcareous green algae with depth going down the drop off. This suggests that the nutrients at these sites are coming from upwelling and mixing of deep cold North Atlantic waters from below the thermocline, which are much higher in nutrients than warm shallow water. At two locations off northwest and southwest Grand Turk we dived down to the thermocline. At the northern location the thermocline was shallower, 90 feet, and at the southern location it was deeper, 120 feet. These depths corresponded to the maximum abundance of Lobophora with depth at each site. Upwelling of deep nutrients must therefore be considered an important source of nutrients for deep and offshore reefs, even those remote from land-based sources of nutrients.

Pollution: Since there is very little industry or agriculture in TCI, the major source of pollution is sewage. Unless sewage is treated to tertiary level to remove the nitrogen and phosphorus, these elements, which are the major limiting nutrients for algae, will remain in residual waters after treatment. If these waters are not recycled on land, the nutrients in them will be released to the ocean by direct discharge, or build up in the water table and underground seepage into the sea. Almost all major developments in TCI have secondary sewage plants with complete wastewater recycling on their own property and low discharges to the coastal zone. This is accomplished via using wastewater irrigation of trees, shrubs, and lawns on site, and TCI is the only place we know in the world to mandate this from all developers. But most older homes have cesspits or septic tanks. Algae indicative of very high nutrients were found near all major sources of nutrient pollution previously known to DECR, such as the entrance to enclosed harbors and marinas, the outflow channels of the most polluted salinas, conch and lobster processing plants, the conch farm, areas of septic and garbage disposal, as well as in previously unidentified areas where nutrient-rich groundwater soaks into the sea from salinas and septic seepage. Algae levels were much higher in 2006 at all the shallow sites that had previously been looked at in 1995, indicating an increase in land-based sources of nutrients following increasing population and tourism. However no algae buildups were found at the outlets of bodies of water, such as North Creek, which receives almost no sewage nutrients. These patterns suggest strongly that nutrient sources, not lack of algae-eating organisms, are the cause of the high nutrients in many shallow coastal waters near populated areas. However large amounts of algae were also found at many locations remote from all obvious land-based sources of nutrients, including very exposed sites where there is little fishing. The significance of those sites is discussed in the next section.

Green water: Water with a strong green color and reduced visibility, due to high levels of phytoplankton algae, was very noticeable in many places, both while diving and from the boat while traveling between sites. Some areas of green water were those already known to be impacted by land-based sources of nutrients, such as the marinas on Providenciales, the Conch Farm, areas near the fish processing plants on South Caicos, and the areas of salina outflow on Grand Turk. However green water conditions were also encountered at many sites remote from any possible land based sources of pollution, for example east of the uninhabited island of East Caicos, North West Point on Providenciales, the North East Barrier Reef of Grand Turk, and the uninhabited islands southeast of Grand Turk. In all areas where the water was distinctly green, macro-algae or micro-algae turf densely covered the bottom. The areas to the south of South Caicos, and areas of Providenciales and Grand Turk, which had distinctly blue water, also had the least algae cover. This strongly suggests that many parts of TCI reefs are strongly affected by upwelling of deep nutrient rich waters, so that the algae problem affects much larger areas than those affected by land-based sources of nutrients alone. While ephemeral green water can be caused by episodic upwelling events, the distinct spatial patterning of shallow and deep reefs into blue water areas with high coral cover (where all the tourist dive sites are located) interspersed with areas of low algae cover with green water, low coral, and high algae, suggests that the upwelling patterns are chronic and persistent, and that systematic long term algae and coral abundance patterns may be linked to upwelling at distinct sites. These preferred upwelling sites are probably result due to ocean circulation patterns influenced by submarine topography, but there is little direct oceanographic data to test this hypothesis. Almost the entire Northern Barrier Reef of Grand Turk and most of the Northern Barrier Reef of Providenciales consists of limestone hardground rock overgrown by algae, with very little coral. The limestone surface is an old erosional limestone surface, sculpted into spurs and grooves by erosion. Note that real spur and groove topography is purely of erosional origin, the spurs are smoothed by erosion and the grooves are full of rounded boulders, not sand, and there is little or no live coral on their sides. Unfortunately many coral reef researchers mistakenly continue to misuse the term “spur and groove” to refer to constructional reef, which has a completely different origin, with channels covered with live corals and a sand bottom. These should be properly called “buttress and canyon” topography (Goreau 1956). These shallow spur and groove areas are not dead reef that has been recently overgrown by algae, instead they are made of older limestone, subject to very high wave stress, and have never had constructional coral reefs growing on them. Their widespread distribution suggests that the green water and algae dominated conditions at these sites has a long and continuous historic past, and is not a recent phenomenon. This has several important management implications. First, the dive sites, which are located in the clearest waters, were not selected simply because they are just the closest good dive sites among many pristine locations, but because the earliest dive operators did not find good dive sites elsewhere. As a result, there is little additional room to expand diving to if the best areas are damaged. This makes their protection from diver and anchor damage even more important. Second, it also indicates that many parts of TCI shallow water habitats have chronic excess nutrients of natural upwelling origin, and therefore that is even more crucial to prevent land-based sources of nutrients from contaminating the nearshore areas because there is already a high background of nutrients from offshore sources. The importance of upwelling was unexpected, and had not previously been identified apart from a description of sporadic upwelling in Goreau (1995). Gaudian and Medley (2000) concluded that upwelling did not affect the TCI or the Bahamas.

Turbidity: High turbidity from suspended sediments was noted in front of beaches exposed to waves, such as northern Providenciales, and in the reefs around the areas dredged for the cruise ship terminal on Grand Turk. A large area of reef down current from the terminal appeared to have suffered greatly from excessive sedimentation caused by dredging, with many corals dead on their tops and others still alive but covered with fine sediments and in poor condition. Reefs upcurrent were far less affected. This indicates that mitigation is needed for the damages to reef habitat caused by the dredging for the port. The area affected had been one of the highest live coral cover sites seen anywhere in TCI, and constitutes the best shallow water coral cover and potential shallow diving sites on Grand Turk. A rescue operation to remove the living, but damaged corals, and transplant them to a location nearby, but upcurrent of the sediment source from dredging and prop wash, would allow them to recover and form a new reef where there is now none, as well as protect Governor’s Beach from eroding. 

Shore erosion: High rates of shore erosion, indicated by an erosion scarp or cliff on the beach, trees falling into the sea, and defensive building of walls in front of buildings, was seen at the northeastern end of Providenciales, and at the northern end of Governor’s Beach, north of the Grand Turk cruise ship terminal. Given the extremely low lying nature of most of the TCI, and the fact that global sea level rise, currently 3 millimeters per year, will rise sharply as global warming causes increased melting of glaciers and ice caps and causes shallow ocean water to expand, while it increases hurricane strength and frequency, these areas of erosion may be harbinger of much more erosion to come, and require a pro-active general strategy to protect the shorelines quite apart from their economic value as tourist beaches.

Aerial observations:

Please see Fig. 2 for locations of islands discussed.

1. There are a very large number of small patch reefs on the east Grand Turk Shelf that there was no time to look at during the field survey, although they look similar to those examined on the islands south east of Grand Turk, which forms the southern end of this zone, and those examined at the very northern barrier reef end. The water was distinctly greenish all over the shelf, and all the patch reefs are very dark brown, probably because of high brow algae cover. Several things can look dark brown from the air, including either healthy corals or masses of brown algae like Lobophora, and need to be confirmed by underwater observations. However it is pretty sure that these areas are largely algae lawns from our observations of similar appearing habitat in the islands southeast of Grand Turk, and the northern barrier reef, the two ends of the shelf. Were this live coral, the coral that would have dominated this habitat, elkhorn, is distinctively golden yellow brown, not dark brown. So it looks as though all the entire east and north shelves of GT is algae covered, as Mitch Rollins, owner of one of the oldest dive shops on Grand Turk, said when asked why there was no diving on those sides (Goreau 1995). 

2. The water on the shelf off southeastern East Caicos, and the east side of South Caicos down to the point where the coast changes from north-south to east-west, was greenish, all the reefs were dark brown with algae, and the water was opaque enough that the drop off was not clearly visible from the air. Thus it looks like these areas, which South Caicos fishermen said always have green water, have chronic upwelling conditions. On the other side of the point, the south facing shore of South Caicos, where we found the lowest algae, the water was clear and blue and the drop off was clearly visible from the air. This difference from the air is completely consistent with our field observations. Thus this differences between green algae dominated sites and blue coral sites seems to be chronic, not due to some very recent upwelling episode that we had fortuitously ran into at the time. 

3. The waters on the Caicos bank are mostly very clear, almost all clean white sand, and the almost total lack of well developed sea grasses except near inhabited islands with nutrients from salina seepage, sewage, and conch and lobster processing plants dumping guts in the water. This distrubution implies that they are nutrient limited (except perhaps near the mangrove areas of North and Middle Caicos). There seems to be a very clear gradient in seagrass lushness away from land implying the nutrients are land-derived. Seagrasses need much higher nutrients than coral reefs, and turn eutrophic (being overgrown by weedy algae) at nutrient levels about 25 to 40 times higher than those that kill coral habitats by algae overgrowth. Lapointe et al. (1994) have published  hyper-eutrophication limits for seagrasses in Florida, at which they are smothered by weedy algae, 25 micromolar Dissolved Inorganic Nitrogen and 0.4 micromolar Total Phosphorus, or 25 and 40 times higher than the value at which coral reefs become algae dominated. 

4. The waters of Chalk Sound and the south side of Providenciales were amazingly clear and lacking in seagrasses depite high light levels. The channels had very clear and blue water. This implies very low nutrient loading. Because these very shallow waters are extremely vulnerable to massive algae overgrowth from nutrients, it is important to prevent any land-based sources to this area from underground seepage of septic tanks or fish processing wastes.

5. There are morphologically separate shallow reef frameworks on the wave-breaking barrier north of the Bight, north of Grace Bay, and from Leeward Going Through Sound northeast to Water Cay. The waters off north Providenciales had greenish tints, and the drop off was not clearly visible, both north of the Bight, and in the area from Leeward Cut eastward, areas where we found high algae cover of shallow fore reefs. In these areas the shallow reef crest, where there is heavy breaking wave impacts and there is exposed old limestone hardground, appears pale from the air, due to waves scouring algae off the limestone hardground, but the outer shallow reef slope immediately in front of it is dark brown with algae. In sharp contrast to those sites, the reef in front of Grace Bay, lying right between them, where we found much less algae in both deep and shallow reefs, is pale in the whole front slope reef area, markedly lacking the dark brown algae covered areas on the shallow fore reefs on either side. The water was clearly bluer, and the drop off was visible from the air. This appears therefore to reflect a chronic and consistent difference in upwelling over a short distance. The plane also went towards North Caicos before turning. The same transition to clearer water and paler fore reef takes place again further east towards North Caicos. These sharp changes over short distances are surprising, but they seem meaningful.

6. Aerial surveys and filming from a low flying air craft over the reefs all around Provo, Grand Turk, South Caicos, West Caicos, North Caicos, Middle Caicos, East Caicos, Salt Cay, and outlying islands to document patterns of water clarity, color, and the color of the hardground fore-reef frameworks, if done on a very clear calm day, will be very valuable in identifying and mapping all areas subjected to localized upwelling. It appears that there is a very complex pattern of upwelling, probably linked to bottom orientation and topography, and we hypothesize that the thermocline depth varies considerably, making an oceanographic study of upwelling, or at least of surface and deep temperature profiles, and mapping of nutrients, very desirable.

Nonparametric statistical data analysis: overview: The 26 variables whose values were tabulated at all 47 sites were compared statistically, pair by pair at all sites to see how their rank order varied between locations. Each pair of variables was contrasted for theie rank order correlation at all sites, and the sign and significance of the Pearson non-parametric rank order correlation coefficients were tabulated (Table 1). A positive correlation coefficient means that when one variable increases, so does the other. A negative correlation coefficient means that when one variable increases the other decreases. The statistically significant correlations are classed into three groups, color coded in Table 1. Statistically significant correlations (S) have more than a 95% probability of being true, and less than a 5% probability of happening by chance. Strongly significant correlations (SS) have more than a 99% probability of being true and less than a 1% probability of happening by chance. Very strongly significant correlations (VSS) have more than a 99.9% probability of being true and less than 0.1% likelihood of happening by chance.

When the entire matrix of 325 interactions between all pairs of ecological and environmentally variables assessed in this study is examined, 19.69% of the relationships between variables are statistically significant, and 80.31% are not. 64 interactions were statistically significant (41 of them positive (12.6%) and 23 of them negative (7.1%)). The breakdown by level of significance and sign is given in Table 2.

Every single variable examined had at least one significant correlation with another variable, and some had many. The distributions of variables that have more significant interactions with other factors contain more information about the overall status of the reef than other less well linked parameters. Because most of the interactions between variables were not significant, it is easier to see the structure of interactions between variables by presenting them in two diagrams (Fig. 15 and 16), the first showing the positive interactions, and the second showing the negative interactions. Significant interactions between variables are indicated by thin lines, strongly significant interactions by medium lines, and very strongly significant interactions by thick lines. In both of these diagrams the number of significant interactions of each variable is clear, and groups of variables that are strongly interacting can be easily identified. These diagrams provide rich and often unexpected insight into how the ecological parameters are related to each other, and how they are related to environmental variables, most of which could not easily be identified otherwise. In addition these linkages provide the basis for a simple dynamic simulation model that can be used to explore the results of changing any one variable on the others. These linkages are discussed below, along with plausible mechanisms to explain them. In some cases alterative explanations could also be proposed, but these seemed the most likely ones.

Nonparametric statistical data analysis: positive linkages

A.      Very strongly significant

1.     Coral cover – Bare cover      The more bare limestone hardground there is the more space there is for young coral settlement. In a healthy reef with little algae these two would be the major bottom types, and so would be inversely related to each other. The fact that they are so strongly positively related is because in most places both are subordinate to and negatively related to algae cover, which covers bare areas for coral settlement.

2.     Coral cover–depth       The highest coral cover was found on deeper dives, from the top of the drop off down the wall. The only reefs above the drop off with high coral cover were in south South Caicos, southwest Grand Turk, and the Grace Bay dive sites on north Providenciales. Live coral cover was very low at most shallow sites, because the elkhorn and staghorn corals that had dominated shallow reef sites had died and not recovered, and because many shallow areas are so heavily covered by algae. On a healthy reef, coral cover is highest in shallow water and would be negatively correlated with depth, since corals need light to grow, but when nutrients are too high corals are replaced by algae. Therefore the positive relationship between coral cover and depth results from high stress from algae overgrowth at most shallow sites.

3.     Bleaching–Lobophora           This relationship is completely unexpected, because bleaching is largely caused by high temperature, and we have no good explanation for it, since bleaching is readily predicted from temperature alone. This correlation suggests that some factor related to Lobophora delays the recovery of corals from bleaching, but the mechanism is unknown and further research is needed. One possibility, that Lobophora harbors bacteria that prolong recovery from bleaching, needs further investigation.

4.     Sargassum-Turbinaria          These are both tough brown algae strongly attached to hard surfaces, which are highly adapted to wave stress and commonly occur together on shallow exposed rock, where they are usually dominant.

5.     Bleaching-White Plague        This is also an unexpected relationship. The reason for this may be fortuitous in that both are most common where coral cover is high. Our estimates of disease and of bleaching were based on the abundance of cases seen in each area examined, not the proportion of corals affected, so both factors would be strongly proportional to coral cover.

6.     Bleaching-Depth  This is probably due to the fact that bleaching was most common in deep water where there were more corals.

7.     White Plague-Black Band      Both Diseases were most common where there were the most corals. Both diseases have spatially clumped aggregations, implying infectious transmission from one coral to another. Transmission of diseases will be greatest where corals are densest. Black Band was much less common than White Plague, but wherever Black Band was common, White Plague was also.

8.     White Plague-Depth     This also appears to be the result of corals being more abundant in deeper water.

9.     Development-Turbidity This association appears to have two causes, 1) due to the tourism economy the most developed areas are behind beaches, which were also the locations of fishing villages, and there is higher resuspension of sediment in front of beaches than on rock shores, and 2) the high turbidity associated with dredging at the cruise ship terminal.

10.    Development-seepage          The effects of seepage are visible through the high nutrient algae growing wherever nutrients from subterranean groundwater flow into the nearshore area. In most cases these nutrients are probably derived from cesspit and septic tank leakage from developed areas and the increased input of water recharge to the ground from recycled waste water irrigation in hotel areas.

B.      Strongly significant

1.     Coral cover-Crustose coralline algae               Crustose coralline algae cover is strongly inhibited by fleshy algae over-growth, as is coral settlement, and the two occur together in the lowest nutrient habitats. In addition crustose coralline algae are often favored sites for coral settlement.

2.     Coral cover-bleaching          Bleaching is most common where there are the most corals to bleach.

3.     Coral cover-White Plague      White Plague transmission is easiest where corals are densest.

4.     Padina-Sargassum       Padina is a brown alga commonly found on shallow rocky areas with low nutrients, and has similar ecological habits to Sargassum. However this association is spatially separated from the Sargassum-Turbinaria community.

5.     Lobophora-White Plague       Lobophora has been suggested as a possible vector for the bacteria causing White Plague and other diseases (Nugues et al, 2004; Smith et al, 2006), and our finding appears to provide strong support for this hypothesis.

6.     Lobophora-Depth Lobophora becomes much more common at the edge of the drop off and steadily increases in abundance with depth, reaching a maximum at the thermocline. This suggests that most of the deep Lobophora is getting its nutrients from upwelling of deep cold nutrient rich water (Goreau,1995). Lobophora has several very distinct forms, and the shelf like morphology is the one common in deep water. In contrast another variety, the “fluffy ruffles” morphology, is found at high levels in some shallow areas (Littler & Littler 2000).

7.     Lobophora-Green Water       Lobophora in shallow water forms dense mats where the water is green. This can be due to upwelling of deep water to the surface or to green water from land-based sources of nutrient pollution.

8.     White Plague-Gorgonian disease    Gorgonian disease is generally thought to be caused by a fungal parasite while White Plague is thought to be caused by a bacterial community that has not been fully identified. Gorgonian disease was much less common than White Plague, but wherever it was found there was a lot of White Plague. This linkage suggests a common reservoir of pathogens, which is unexpected and requires further work.

9.     Development-Point Source    This relationship is no surprise, as the major point sources of pollution are from the most developed sites. The only reason that this relationship is strongly significant rather than very strongly significant is that the nutrients from point sources, once they enter the water, are strongly carried downstream along the coast by the longshore currents and have minor effects on nearby upcurrent areas. In all cases there were very strong differences in the abundance and species of algae on the upcurrent and downcurrent sides of major point source flow entrances into the sea.

10.    Point Source-Seepage  The effects of seepage are seen through the algae stimulated by underground seepage, and like point sources of sewage, these are highest in the most developed areas with the highest population density, where there is the most sewage leakage into ground waters and the greatest recharge of water to the ground in hotel areas.

11.    Bare cover-Crustose coralline red algae  Crustose coralline algae are typical of low nutrient waters, and are quickly overgrown when fleshy algae spread, as is bare cover.

C.      Significant

1.     Coral cover-Black Band        This is a result of the clumped infectious distribution of the disease, which is mostly easily transmitted to other corals nearby.

2.     Bare cover-Depth        Although deepest areas have more algae than intermediate depth zones, this is outweighed by the much higher algae dominance of very shallow waters near to land-based sources of nutrients.

3.     Padina-Development            Padina is often extremely common on rocks in front of beaches with a lot of development nearby, for example in front of Cockburn Town on Grand Turk. This abundance is highest where there are only low levels of land-based sources of nutrients from seepage. Where nutrients are high, Padina is overgrown and replaced by high-nutrient weedy algae.

4.     Lobophora-Turbinaria           Both are brown algae that are tightly attached to hard rock in shallow waters affected by upwelling. However Lobophora is not correlated with Sargassum, as is Turbinaria, so these combinations with Turbinaria take place in distinctly different habitats.

5.     Lobophora-Microdictyon       Microdictyon is brittle green alga that reaches very high levels in shallow sites with high nutrients. Where they occur together the Lobophora is always the ruffled shallow water type.

6.     Lobophora-exposure            Very high levels of Lobophora were often found at shallow windward sites. It appears that high wave turbulence on these shores is very effective in mixing deep nutrients up to the surface, as indicated by the green water at these sites.

7.     Stypopodium-Microdictyon    These two algae formed a characteristic assemblage on the northern shallow barrier reefs.

8.     Sargassum-Cladophoropsis    These two algae also formed a characteristic assemblage on the northern barrier reefs, but spatially separated from the Stypopodium-Microdictyon assemblage. Cladophoropsis, like Microdictyon is a green alga typical of high nutrient habitats, but is much tougher, and this separation may reflect tolerance of higher wave energy or specific inhibition of one by the other.

9.     Turbinaria-Microdictyon        This association is surprising because the first is very tough, while the latter is fragile, but it was quite common on the northern barrier reef of Grand Turk.

10.    Dictyota-Development  Dictyota was most common in shallow areas in front of developed beaches with moderate nutrient loading. Nearer to nutrient sources it is replaced by higher nutrient indicating algae.

11.    Crustose calcareous red algae-Branching calcareous red algae   Both are most abundant where nutrients are low, as they are quickly overgrown by faster growing weedy algae wherever there are high nutrients.

12.    Crustose red algae-White Plague    This may be a reflection of the fact that both are most abundant where corals are densest. Alternatively the algae may be a reservoir of White Plague pathogens.

13.    Crustose red algae-depth      Once again, crustose red algae and corals are most abundant in deeper waters with less algae.

14.    Branching calcareous red algae-Black Band disease          This relationship is unexpected, and may be due to a possible pathogen reservoir effect that needs further studies.

15.    Laurencia-Gorgonian disease         Another unexpected correlation, which may be due to possible pathogen reservoir effects that need further investigation.

16.    Microdictyon-Exposure Microdictyon was extremely abundant on many windward reefs despite its fragility, almost always associated with Lobophora. This may be due to higher nutrients transport to the surface at high wave energy coasts.

17.    Bleaching-Black band disease        This may simply reflect the fact that both are most common where there are the most corals.

18.    Bleaching-Exposure             Exposed coasts have little sand and sediment to resuspend, and so corals in these areas may have clearer water and higher light exposure, which increases the speed of bleaching when the temperature is hot, and prolong their recovery.

19.    Turbidity-Seepage               Both are strongly linked to development which is near beaches where sand is resuspended and cruise ports where dredging has taken place.

20.    Green water-Point source     Green water was clearly seen to emanate from land-based point sources of pollution from marinas, fish processing plants, hatcheries, and Salinas.

Nonparametric statistical data analysis: negative linkages        

A.      Very strongly significant

1.     Coral cover-Cladophoropsis           Cladophoropsis forms tough, dense green mats that completely cover the bottom and they are very effective in overgrowing and killing corals. There almost no corals in Cladophoropsis dominated areas.

2.     Seepage-Bare cover            All hardbottom in areas where seepage was identified was completely overgrown by algae and/or cyanobacteria. In most of these areas the bubbling up of groundwater was clearly visible. This water is certainly nutrient rich, although no measurements could be made.

B.      Strongly significant

1.     Coral cover-Green water      All green water sites were heavily overgrown by algae and had few live corals.

2.     Bare cover-Green water       All green water sites had nearly complete algae cover and little or no bare substrate except where there had been very recent wave impacts.

3.     Bare cover-Point source       Hard surfaces near point sources were completely algae covered. Once again high nutrients are certainly responsible, but we were not able to confirm this by direct measurements due to lack of the equipment needed.

4.     Stypopodium-Turbidity Stypopodium is almost never found in turbid water.

5.     Microdctyon-Calcareous green algae              Microdictyon appears to smother and kill calcareous green algae. This could result from an opportunistic interaction.

6.     White Plague-Cladophoropsis There were almost no corals found in Cladophoropsis dominated areas, so infectious transmission was suppressed.

7.     Cladophoropsis-Depth   Cladophoropsis was found only on shallow hard limestone surfaces and did not go deep. This may be due to high light requirements.

C.      Significant

1.     Coral cover-Sargassum        Sargassum is a tall tough alga that is very effective at overgrowing and killing corals, so there are few corals where Sargassum is dominant.

2.     Bare cover-Cladophorpsis     Cladophoropsis forms dense tough mats completely covering the bottom, with no bare spots.

3.     Padina-Cyanobacteria          This negative correlation may be because Padina is most abundant at low to moderate nutrient levels, but cyanobacteria are most common in sites with very high nutrients. Direct nutrient measurements are needed to confirm this.

4.     Padina-gorgonian disease     Padina was not common at sites where gorgonians were abundant.  

5.     Padina-Depth              Padina is most common in very shallow water, as it appears to require high light levels. It is often found on rocks exposed at low tide.

6.     Stypopdium-Development             Stypopodium was almost never seen near developed areas. It appears to be typical of low nutrients and moderate wave energy.

7.     Stypopodium-Seepage Stypopodium appears intolerant of high nutrient, sediments, and low wave energy areas where sand accumulates in front of salinas that feed underground seepage.

8.     Laurencia-Turbidity              Laurencia appears to require high light, low nutrient conditions, and is intolerant of sedimentation, while turbidity is associated with low light, development, and point sources of nutrients.

9.     Microdictyon-Cladophoropsis Both are green algae typical of high nutrients but the former is fragile and the latter is very tough, so one possible explanation is that they occur in areas of different wave energy. Another possibility is that one of these algae is able to release chemicals that inhibit the other. Such specific inhibition, or allelochemical effects, have been suggested but can only be confirmed by direct experimental competition studies.

10.    Cladophoropsis-Black Band           This may be due to the very low coral abundance where Cladophoropsis dominates. 

11.    Cyanobacteria-Black Band            This is at first sight astonishing because the Black Band Disease is marked by a community of bacteria living in a matrix made up by a cyanobacteria species that give it the dark color. However areas with high amounts of cyanobacteria are high nutrient areas with few remaining corals, since cyanobacteria are effective at overgrowing and killing corals.

12.    Gorgonian Disease-Point Source     Areas near point sources were algae dominated with few gorgonians. This implies that the pathogen causing the disease is not carried in land-based sources of pollution. Some researchers blame air borne African dust as the carrier of the fungal pathogen, but this does not fit the clearly clumped abundance of the disease.

13.    Exposure-Development Almost all development on the Turks and Caicos is near beaches and inland from them. These areas are not found on heavy wave-exposed windward coastlines, but on the protected leeward shores.

14.    Seepage-depth           Due to the very low elevations of the TCI, the hydraulic groundwater gradient heads are very low, and seepage takes place only through shallow beach sand formations.

DISCUSSION

Coral disease patterns: Many long known correlations between variables, such as those related to intensity of land development nearby, are shown in the network of significant associations revealed by this data. Of special interest is the large numbers of statistically significant linkages between coral diseases and algae. These were wholly unexpected, since they could probably not been detected with intensive reef assessment methods, were not really noted in the field observations (previous sections), and emerged purely from the large scale correlations at all sites. Previous studies had suggested possible links between certain diseases and certain algae, based on small-scale quadrat observations in Curacao (Nugues et al. 2004), and from experimental laboratory manipulations (Smith et al. 2006). But the specific correlations our study found were different from those had been proposed earlier. These TCI results suggest a very specific range of algae to be tested as potential hosts for each disease. It shows little link between disease and development or pollution, which have often been pointed to as possible causes or promoters of disease.  It also shows that several of these diseases correlate with each other, suggesting they are affected by a common stress or group of pathogens.  It is clear that much would be learned about conditions promoting spread these diseases if both large and small-scale studies were combined with microbiology, and much further research along these lines is needed to test the hypotheses generated by the data analysis of this paper.  In addition the lack of strong linkages between coral diseases and indicators of human impact imply that the distribution of pathogens, not the “health” of the corals with regard to anthropogenic stress, is the dominant factor.

Algae distributions:  Little work has been done on reef algae community associations, but our results confirm that algae zonation is very well developed (Goreau, 1992), and that there are characteristic species assemblages in many well characterized habitats, so that algae zonation potentially reveals much more information about ecological patterns than coral distributions. Algae ecology has been traditionally been given short shrift, or less, in most reef assessment, which in effect amounts to throwing out the richest and most diagnostic set of data. Furthermore the data reveals a wealth of both positive and negative interactions between many algae species. These presumably result from competitive, parasitic, commensal, or exclusionary allellopathic interactions that have not been described. Much further research is warranted to test the hypotheses generated in this paper and ascribe sound mechanisms that could explain them. In addition the data reveals those algae species whose abundances are negatively correlated with coral abundance, indicating that these are the most effective at overgrowing and killing corals.

Upewlling:  The role of upwelling in promoting high nutrient backgrounds and algal dominance in areas remote from human nutrient impacts or of fishing strongly supports the primary role of nutrients in controlling algae productivity, and indicates that many “pristine” areas in the sense of having little human impact may not have healthy coral reefs. Therefore more stringent controls on land-based nutrient sources are needed wherever there are high backgrounds. The areas of upwelling appear to be systematic and are interspersed with areas of lower algae and blue water,. Much work needs to be done on hydrodynamic-topographic controls over upwelling to identify the areas affected.

CONCLUSIONS

1.                  Live coral cover is declining:  Dead coral exceeded live coral at all sites, often by very large amounts. The best reefs had about 40% live coral cover, but even there dead coral was more abundant than live coral and some of these sites had been observed to have up to 70% live coral cover 11 years before (Goreau 1995). Large areas of shallow reef framework seem to have died from White Band Disease around 1979, and have shown little signs of recovery since. Progressive mortality appears to be underway from episodes of coral bleaching, coral diseases, algae overgrowth, and in some locations near dredging, from excessive turbidity. The result is a continuing decline in biodiversity, shore protection, fisheries habitat, and ecotourism environmental services. Even though many sites still have better coral cover than most Caribbean reefs, this is no cause for complacency because progressive decline is visible everywhere.

2.                  Most TCI reefs are algae dominated: Healthy growing coral reefs are dominated by corals, and algae-dominated reefs are unable to keep up with erosion and sea level rise. An algae-dominated reef should be called precisely that, and not confused with a healthy coral reef, because it cannot serve the same shore protection, biodiversity, or fisheries habitat roles. Because most current coral reef researchers have never seen healthy reefs, there is now a common practice of describing reefs as “healthy” even when they have very little live coral. For example Vroom et al. (2006) examined Pacific reefs with very low coral and very high algae cover that were regarded as being very healthy coral reefs on grounds of their remoteness from human activity. However most of their sites had very green water, as is made clear by their photographs, 91% of which show water as green as the greenest found in TCI. Deep Pacific waters have around twice the nitrate and phosphate concentration of deep Atlantic waters, and so shallow reef areas in the Pacific are even more vulnerable to high nutrient loading, especially in areas with very shallow thermoclines subject to wind-driven upwelling. Much of the current reef literature suggests that algae-dominated conditions are simply “an alternative stable state” or “phase shift”, from which “resilient” coral reefs will bounce back by themselves, and algae dominated conditions are often blamed on fishermen, who are claimed to have eaten all the algae eating fish, or on sea urchin diseases. None of these “explanations” fit the situation in TCI. TCI has some of the least overfishing in the Caribbean, and the algae domination has little or no relationship to protection from fishing or absence of herbivorous fish. The high algae abundance preceded the epidemic die off of Diadema antillarum sea urchins in 1983 according to the oldest divers. Diadema was found to be abundant at some sites, and all of these had high algae cover, so they were clearly not controlling the algae and could not explain low algae cover at sites with no Diadema. The only plausible explanation for the high abundance of algae at most sites in TCI is an excess of nutrients. There is no evidence of algae-dominated reefs reverting to coral dominated reefs except for those few cases where all land-based sources of nutrients have been eliminated (Goreau 2003). This is possible only where land-based sources of nutrients dominate, but not in places where upwelling is the major source. Ocean fertilization, as (incorrectly) proposed by some as a carbon sink, would make the reef algae problem worse.

3.                  There is a very high incidence of coral disease:  Severe, but patchy, epidemic outbreaks of infectious coral diseases are underway, with White Plague, Black Band (apparently combined with White Plague), and Gorgonian Disease being the major ones. These diseases typically spread faster during warm seasons, and may have been in the early stages when examined. The survey data gave a clear indication that many of these diseases are correlated with the abundance of certain algae species, lending strong circumstantial support to suggestions that some algae may serve as a reservoir for coral pathogens (Nugues et al. 2004; Smith et al. 2006). Many sites with high disease had not had them when previously surveyed by DECR personnel. In contrast, Riegl et al. found low diseases at most sites in 1999, with a maximum of 13% at two sites on South Caicos, no difference between shallow and deep areas, exposed and sheltered areas, and very low disease in Grand Turk. Most disease they saw was White Plague, very little Black Band, and they reported negligible gorgonian disease. They also found low amounts of recent coral mortality. The rapid recent spread of coral diseases is a matter of great concern, and it will be important to understand the factors involved for there to be any hope of minimizing their impacts.

4.                  Coral bleaching is a major near term threat:  There have been several serious coral bleaching episodes in TCI, and partial bleaching and partial mortality from recent and past bleaching episodes were clearly visible.  Bleaching has become a chronic, nearly annual, event in the last 10 years. Only very small further rises in sea surface temperature or one unusually hot year could cause serious coral mortality any year now.

5.                  There is a high natural nutrient background:  While brief episodes of green water conditions have long been known to divers, there is clear evidence from fishermen that many sites in TCI have chronic green water conditions, and the distribution of algae suggests that there are sites that have consistent upwelling, and others where it is always much rare. Upwelling not only causes green water conditions, it is a source of high nutrients from deep cold water that cause chronic blooms of benthic algae, preventing coral settlement or recovery from coral disease epidemics. The observations strongly suggest that many, possibly most, TCI shelf edge reefs suffer from chronic high nutrient backgrounds, and that due to local circulation patterns, these upwelling events are much more frequent at some sites than others. This means that there are probably not many high quality unutilized dive sites that could be used if the current tourist dive sites are damaged from overuse.

6.                  It is crucial to control land based sources of nutrients:  Land-based sources of nutrients have clearly caused a spread of algae over many shallow reefs used for snorkeling and fishing near populated areas. Several shallow inshore areas in Grand Turk that were previously dominated by clean white limestone rock (Goreau 1995), were algae covered in 2006. Because of the high natural nutrient background caused by upwelling, TCI waters are highly vulnerable to eutrophication (nutrient-caused proliferation of weedy algae), and only very small increases in land-based sources of nutrients can trigger massive algae overgrowth of entire reefs (Goreau 1992, Goreau & Thackerr1994, Lapointe et al. 1994. Lapointe 1997, Goreau 2003, Lapointe et al. 2004). TCI cannot afford to be complacent that it is less vulnerable to eutrophication than other Caribbean islands, such as Jamaica, that have much higher populations, rainfall, and hydraulic gradients. In those islands, the thermocline is much deeper, and natural offshore backgrounds of nutrients are lower than in TCI, which sits in an area of shallow thermoclines and higher deep Atlantic nutrient levels. Control of land-based sources of nutrients is even more crucial in TCI because off the higher natural background nutrient levels.

7.                  Damage to shallow reefs and sea level rise pose major medium term erosion risks:  The dead elkhorn reef frameworks that protected much of the TCI coastline prior to 1979 is slowly being broken apart by hurricanes and bio-eroding sponges, clams, worms, algae, and fungi, and new settlement and growth is far too low to replace them. While at the moment beach erosion is confined to a few relatively limited sites, global sea level rise, now 3 millimeters a year, but due to rise rapidly in coming decades, coupled with increasing hurricane strength and frequency, is medium term threat to TCI beaches that requires a pro-active strategy to protect the beaches and low lying developed areas.

8.                  Coral reef restoration is important for long term sustainability of tourism, fisheries, and shore protection:  Long term maintenance of the ecosystem services that only healthy coral reefs can provide, in the face of accelerating threats to corals from local, regional, and global stresses, means that relying on the reefs to come back by themselves, as they did after past hurricanes and shipwrecks, will no longer work. Future management will require active strategies to restore critical reef areas by reducing stresses to reefs and actively growing corals using methods that increase their growth rate, their resistance to environmental stress, and their quality for fish and shellfish habitat.

RECOMMENDATIONS

1.      Implement a national coral reef management strategy:  Because of their long history of surviving from the sea, Turks and Caicos islanders have an exceptionally strong consensus for protecting their marine resources from damage. Yet there has been surprisingly little assessment of the reef condition, identification of the factors affecting them, or awareness of their current condition. To manage them effectively TCI needs to greatly increase the information base on which to inform effective management. A coral reef assessment and management team should be established in DECR, and the large scale assessments carried out in this study should be applied to all reef areas in TCI, including all the areas that could not be covered during this study, especially the reefs on North, Middle, and East Caicos and the southern islands. Key sites should be documented by digital film or photography, to provide a baseline to assess future change through repeated assessment.

Most countries view marine park management as basically a policing operation to control damaging activities in the water, with little effort to understand the root causes of long-term change. As a result it is impossible to point to a marine park that has more corals now than when the park was established. The solution to this is to ensure that the problems and solutions are quickly identified and that a strong public consensus is maintained that prevents the need to allocate scarce resources in patrol boats and park wardens, as so many other countries do with so little real result. The strong support of TCI islanders for coastal protection needs to be enhanced through education, so that management does not become simply a policing operation against destructive acts by an unenlightened public, which will always be far more expensive and less effective than prevention through education and informed consensus.

Effective management of reef resources must be based on large-scale assessment of reefs over long periods, not observations at a few points at one time. TCI, like most tropical countries, is using its reef resources, and the sand beaches that reefs produce and protect, to base its economic future on tourism. Since all its competitors are marketing themselves as pristine paradises, any honest assessment of the fact that all are suffering serious declines in coral and fish populations is usually viewed as a threat by the tourism public relations apparatus, and often the all too real facts about reef decline are ignored or even suppressed as unwanted bad news that might have a negative economic impact. Any country admitting their reefs are declining is immediately placed at a publicity disadvantage with regard to competitors who are lying about their reefs. Yet without honest assessment of the problems, the need for management efforts to reduce the damage and restore degraded areas cannot even begin, because the problems cannot be admitted to. TCI, with its strong dependence on the reefs and low lying coastlines can really not afford to fool itself this way, even if other countries do so. Taking a pro-active strategy to reef protection, rather than a reactive one, would make TCI a world leader in coastal zone management.

A serious coral reef management strategy, which no country in the world yet has, requires learning from the lessons of past historical changes, understanding the causes of the changes presently underway, and anticipating and proactively planning for the changes that are most likely to take place in the future. Instead coral reef management in most countries is largely confused with 1) intensive monitoring on a small spatial scale that misses the larger gradients and on a short temporal scale that misses the long-term changes, and so provides only a snapshot of selected sites that generates little guide to effective management, and 2) setting up “marine protected areas” in the hope that “resilient” reefs will just bounce back by themselves. This works only in areas that are well managed, where water quality is excellent, where there are pristine reefs up-current to provide new larval corals and fish to replace those lost, and where the major threats are local stresses such as use of inappropriate fishing gear or over-harvesting that can be controlled by adequately staffed marine protected areas.

TCI has some the highest proportion of marine protected areas of any place in the world, and while preventing fishing in certain areas has clearly helped preserve fish, lobster, and conch populations in TCI as a whole, our findings suggest it does not seem to have preserved the coral. That is because the major factors killing corals are not due to fishing, boating, and tourist diving activities, but due to new diseases, global warming, land-based sources of pollution, and possible changes in ocean currents. These factors are beyond the control of any MPA of activities within its own borders because the critical stresses come from outside. Truly effective coral reef management lies in controlling the stresses that are really killing corals, wherever they come from, and in restoring areas that have been destroyed, damaged, or degraded. This should not be construed as an argument that coral reef management is hopeless or should be scaled back, but rather that the best reef areas need to be identified and protected, and a focus be placed on identifying the causes of degradation and identifying strategies to control them at the local, regional, or global scale as needed. TCI has done about the best job of any country in the world in protecting its coastal resources despite very limited funding and personnel, but it will need to better to face the accelerating challenges that global warming and increasing development pressures will cause in coming decades. Strict control of all stresses caused by inappropriate development on land will be the first step, but more will be needed.

2.      Implement a national coral reef restoration strategy:  No country in the world now has a coral reef restoration strategy that is designed to rehabilitate degraded reefs to bring back their lost biodiversity, fisheries, tourism, and shore protection values. The international agencies, such as the World Bank and the International Coral Reef Initiative, and the US and Australian Governments that are their major source of funding, have in fact stated that none is needed, that all that countries with damaged reefs need to do is to wait, and the “resilient” reefs will come back all by themselves! This “head in the sand” strategy is clearly doomed to failure as global warming, global sea level rise, new diseases, and land-based sources of pollution continue to increase at accelerating rates. Island nations wanting to preserve their marine resources will need to invest in training their people to grow back damaged habitats for their children’s future. The technology now exists to grow reefs of any size or shape on which corals grow much faster than normal, are more resistant to environmental stresses including high temperatures, nutrients, and sediments. These provide superior habitat for fish and shellfish that support extremely dense populations, and have turned severely eroding beaches into rapid growth (Goreau & Hilbertz 2005). There is a real cost, but it is less than losing the services of the reef and having to import sand for tourist beaches, import fish to eat, and build walls around the shore to keep the buildings and roads from collapsing into the sea.

The areas of immediate concern in TCI are the beaches vulnerable to erosion that are important to the tourism economy, such as Governor’s Beach on Grand Turk and northeastern Providenciales, as well as snorkeling areas for tourists. At present there are only two good nearshore snorkeling area for tourism on Providenciales, which are small and over-used, and none on Grand Turk, where tourists are taken snorkeling on remote dead reefs. Hotels could be encouraged to restore reefs to provide snorkeling in front of their beaches and take the pressure off remote areas, reducing the time, gas costs, and hydrocarbon pollution to get tourists to sites with corals and fish, while increasing fish populations in surrounding reefs through spill-over effects. If all beachfront hotels were to do so, they could play a significant pro-active role in enhancing TCIs ecotourism resources, fisheries habitat, and shore protection. One major hotel on Providenciales has asked for assistance in doing so, since there is so little good shallow reef for snorkeling. Governor’s Beach, which could provide a prime location for beach and snorkeling activities because it is right next to the cruise ship terminal, is heavily eroded at the northern end and the reef area in front is barren of life. A shore protection reef at this site would provide a prime tourism attraction and make the beach grow. Given the large area of reef to the south of the cruise ship terminal that was badly damaged by sediments from dredging and cruise ship propeller wash, restoring this area would provide a cost effective mitigation that would be greatly appreciated by cruise ship tourists. The corals downcurrent of the dredged area are dying from sediment stress, and could be rescued by being transplanted to the upcurrent side where they could form a reef that would be a prime tourism attraction and protect the eroding beach. Because shore erosion will dramatically increase in the future as global warming causes the protective corals to be killed by bleaching, as hurricane frequency and intensity increase, and as global sea level rises, the low lying TCI will need to start planning its long term coastal defenses and carefully control development in low lying areas.

3.      Implement a national strategy of land-based nutrient management: The major thing TCI can do on its own to protect coral reefs is to prevent land-based sources of nutrients to the coastal zone. While TCI has done excellent work on developing guidelines for sewage treatment for tourism developments, this does not include tertiary treatment, which is essential to remove the nutrients from the liquids remaining after sewage treatment. The best use for this wastewater is to use it as nutrient rich water for watering ornamental plants, lawns, and golf courses, as being done by TCI to excellent effect. TCI is already the world leader in mandating on-site wastewater reuse on all developments, something all other tourist islands also need, but they are unaware of TCIs unique success in establishing a regulatory planning framework to ensure recycling and compliance. However studies are needed to determine how efficiently these nutrients are absorbed by the vegetation, since if nutrients are provided in amounts beyond the plant’s uptake capacity, the excess will soak into the groundwater. Also there are still a large number of older houses that use cesspits or septic tanks, most of which release their nutrients into groundwater, from where it flows into the coastal zone. The impacts of nutrients from groundwater seepage and outflows of sewage-contaminated salinas was very visible in terms of dense and spatially localized masses of algae and cyanobacteria typical of extremely high nutrient levels, the “end of the sewage pipe” algae. As tourism expands, and the demand for reverse osmosis water increases, the water tables under the big hotel waste water discharge areas will rise, increasing the flow into the coastal zone. Although it is assumed that all nutrients in irrigation waste water are absorbed by plants, it is possible to add nutrients in excess of the plant’s uptake capacity, in which case nutrients will then build up in ground water, from where they may flow into the coastal zone. Direct measurements of nutrients in wastewaters and ground water are needed to determine if this is a problem needing to be managed. It may become necessary to use excess waste waters to water vegetated areas inland, whose growth is clearly limited by lack of both water and nutrients due to the semi-arid climate and very poor soils. Another may be to use charcoal as a soil fertilizer, because it greatly increases the water and nutrient holding capacity of the soil, speeding plant growth and increasing the efficiency of nutrient recycling. With regard to the large number of older homes without sewage treatment, providing links to sewage treatment may be very expensive. Unless funds can be obtained for complete sewer hookup to sewage treatment facilities that recycle all the nutrients on land, the alternative may be to institute a program of composting toilets, which prevent groundwater nutrient pollution by producing a sterile organic fertilizer with a clean earth smell when managed properly.

4.      Implement national nutrient mapping capability:  Management of coastal zone water quality is essentially a matter of managing nutrient levels, especially in TCI with high background nutrients from upwelling. No nutrient management strategy will be effective unless coastal zone managers know the nutrient levels and their sources, which none in the world now do because they lack the data they need. While the strongest nutrient sources are clearly identifiable by eye through the green water and dense algae populations they cause, the actual amounts of nutrients entering the water from land-based point and non-point sources are unknown, as is the natural background from upwelling and wind and wave induced mixing of deep waters, which this study has shown is clearly of very great importance, and probably dominant in most parts of TCI except on top of the shallow Caicos Bank. Understanding where and how large the upwelling sources are is the key to understanding where land-based sources can be most effectively controlled, and assessing the effectiveness of land-based control measures. This requires real-time continuously-recording nutrient and water quality mapping capability in TCI. At present nutrient samples must be shipped abroad for analysis, which is not only extremely expensive, it introduces sampling and storage artifacts of unknown magnitude, making the measurements unreliable. Because the data come from widely spaced samples, they cannot measure large-scale gradients and can miss intervening sources or sinks of nutrients. Because they are not available in real time, they cannot be used to track sources directly to their origin. Equipment is now available that overcomes all of these obstacles, and allows real time continuous measurements of all the major nutrients (nitrate, ammonia, phosphate) and of all the related water quality parameters needed to fully interpret them (temperature, salinity, chlorophyll, oxygen, turbidity, plus acidity, oxidation-reduction potential, and hydrocarbons). With such equipment in one day one could map the nutrients all around an island, identify every source, and track them to their origin, determine seasonal changes in both land-based and offshore-based sources, and examine the effects of severe events, such as upwelling after hurricanes. We recommend that DECR seek funds to develop its own water quality mapping capability. Doing so will place it at the leading edge of coastal zone management programs worldwide. This could be done in regional cooperation with other British Overseas Territories in the Caribbean, such as Cayman, Anguilla, British Virgin Islands, and Montserrat.

5.      Implement research programs on coral disease:  Coral diseases are at the moment the major coral killer in TCI and clearly getting worse. There are few management prescriptions that can be offered, because many of the pathogens have not yet been identified, and their sources, mode of spread, and mechanism of action are largely unknown. Our data suggest that certain species of algae may be serving as reservoirs of coral disease pathogens. Much further research is needed to determine if this is the case, and to see if there are possibilities of control. The first step needed is to identify the microorganisms present in diseased coral, healthy coral, water, and on different species of algae. This can be readily done now by using Polymerase Chain Reaction (PCR) amplification to create libraries of the genetic information of all the microbes present, identifying all of those present, whether they are previously known or not. Once they are identified they may be cultured to determine which are pathogenic, and how they act and spread. We recommend working with top researchers on marine diseases and microorganisms to do the work in TCI before the problem expands. Because of the potential role of algae communities as potential disease reservoirs, more work is needed to understand the basis of algae species zonation.

6.      Develop a national strategy to address global change:  Even though TCI and the Bahamas are the most vulnerable parts of the Caribbean to global sea level rise and global warming, TCI is unable to do anything to stop global warming by itself except by providing moral leadership. Since the United Kingdom represents TCI in all international environmental meetings, TCI should encourage the UK to take TCI’s climate change vulnerability into account in seeking a comprehensive and effective solution to the problem of atmospheric greenhouse gas buildup from fossil fuels. Although the British Overseas Territories are not members of the Association of Small Island States because they are not independent, they have much in common and would benefit from working directly together with the Association of Small Island States (AOSIS) at United Nations meetings on climate change and sustainable development. Of course TCI will be in a superior moral position to make its case if it is also making serious steps to develop its own sustainable energy resources. The cost of energy is among the highest in the world, due to the fact that it is all imported and there are few economies of scale. Yet the huge solar and wind resources are essentially untapped. At sites where currents reach 4 knots or above, tidal current turbines could produce significant energy. The Global Coral Reef Alliance and the MIT First Step Coral Project, have just set up tidal, wind, and solar powered reef restoration efforts at a remote and poor fishing island in the Philippines, and TCI could easily do the same. Another possible sustainable energy option that needs to be looked into is the production of biofuels from the glasswort, Salicornia, which occupies large hypersaline areas in the larger northern Caicos Islands.

7.      Educate the public and tourists:  Public education on the importance of preserving reefs is so obvious a strategy that it is widely promoted, however in almost all cases these programs have had two major flaws that should be avoided. Firstly, they focus on the beauty of idealized reefs under perfect conditions, not the widespread reality of the threats that affect typical reefs. Secondly, they focus on symbolic management steps to control actions like tourists stepping on corals, anchor damage, ship groundings, etc. that can be locally important but are really minor coral killers compared to the real major coral killers: global climate change, land-based sources of pollution, and other factors linked to population growth and energy use. Corals are the ultimate down-stream ecosystem, being affected by all of our waste, energy, land, and ocean management practices on a global scale, and can be protected only if this is clearly conveyed not only to locals, but also to those whose actions elsewhere kill reefs here. Prevention through education of an informed and enlightened public is far cheaper than enforcement against an uninformed and uncooperative one. Because of its history, TCI residents are unusually supportive of efforts to protect and manage their marine environment. As a result, although DECR is both underfunded and understaffed, it has accomplished much more meaningful protection of reef and fisheries resources than countries spending far more, which lack the support of fishermen and do not control nutrients from sewage and golf course developments. By more widespread public education, it should be possible to build a consensus for large-scale coral reef restoration to protect the fisheries, the beaches, the tourism economy, and the very islands themselves from the greatly increased stresses they will undergo in the coming decades if global climate change is not promptly reversed. Up till now, TCI has been able to benefit from its coral reefs as a “free good”, like a goose that lays golden eggs or a cow that keeps producing milk without ever being fed. The human caused and “natural” stresses to the reefs have greatly increased in the last few decades, and will get much worse in the near future. Given the low lying nature of the islands and their extreme vulnerability to global warming, global sea level rise, and intensifying hurricanes, the only real alternative to investment in coral reef protection and restoration will be a disorderly retreat after catastrophic extreme-event disasters. Expanding the human and financial resources and the mandate of DECR is the best way forward, if strong, high-level political will, along with new funding can be generated. For this to happen, political leaders must become spokespersons for sustainable environmental management not only in the policies they apply at home, but also to international funding agencies and climate negotiations.

REFERENCES

J.L. Campbell, T. Fisher, & F. Perez, 2006, A Management Plan for the Columbus Landfall National Park and Other Protected Areas in and Around Grand Turk, DECR

Environmental Charter agreed between the Turks and Caicos Islands and the United Kingdom, 2002, Plan for Biodiversity Management and Sustainable Development around the Turks and Caicos RAMSAR Site, p. 166-167

T. Fisher, 2004, Assessment of reefs at dive sites in Providenciales, Grand Turk, and South Caicos, DECR

B. Gascoine & C. Lott, 1991, Diving, Snorkeling, Visitor’s Guide to the Turks and Caicos Islands, Graphics Reproductions

G. Gaudian & P. A. H. Medley, 1995, Diver carrying capacity and reef management in the Turks and Caicos Islands, Bahamas Journal of Science, 3: 9-14

G. Gaudian & P. A. H. Medley, 2000, Turks and Caicos Islands, p. 587-594 in C. R. C. Sheppard (editor), Seas at the Millenium: An Environmental Evaluation, Elsevier

W. B. Gladfelter, 1982, White-Band Disease in Acropora Palmata: Implications for the structure and growth of shallow reefs, Bulletin of Marine Science, 32: 639-643

T. F. Goreau, 1956, A Study Of The Biology And Histochemistry Of Corals, 227 p., Yale University, New Haven

T. F. Goreau, J. C. Lang, E. A. Graham, & P. D. Goreau, 1972, Structure and ecology of the Saipan reefs in relation to predation by Acanthaster planci (Linnaeus), Bulletin of Marine Science, 22:113-152

T. F. Goreau TF, & N. I. Goreau, 1973, The ecology of Jamaican coral reefs. II. Geomorphology, zonation, and sedimentary phases, Bulletin of Marine Science 23:399 464

T. J. Goreau, 1992, Bleaching and reef community change in Jamaica: 1951-1991, American Zoologist, 32: 683-695

T. J. Goreau, 1995, Artificial reef development in the Turks and Caicos: Environmental assessment and proposed projects, 17 p., DECR, Ministry of Natural Resources

T. Goreau, 2003, Waste Nutrients: Impacts on coastal coral reefs and fisheries, and abatement via land recycling, 28p., United Nations Expert Meeting On Waste Management In Small Island Developing States, Havana, Cuba

T. J. Goreau & R. L. Hayes, 1995, A survey of coral reef bleaching in the South Central Pacific during 1994: Report to the International Coral Reef Initiative, 201p., Global Coral Reef Alliance, Chappaqua, New York

T. J. Goreau & R. L. Hayes, in preparation, Climate Change Impacts On Coral Reef Ecosystems: a global atlas of satellite-derived sea surface temperatures at coral reef sites, 1982-2003

T. J. Goreau & W. Hilbertz, 2005, Marine ecosystem restoration: costs and benefits for coral reefs, World Resrource Review, 17: 375-409

T. J. Goreau & K. Thacker, 1994, Coral Reefs, sewage, and water quality standards, Proceedings of Caribbean Water and Wastewater Association, 3: 98-116

S. Harrigan, 1992, Water and Light: A Diver’s Journey To A Coral Reef, Houghton Mifflin, Boston, 277p.

F. Homer & D. Shim, 2000, Status of coral reefs in the Turks and Caicos Islands, An overview for the Global Coral Reef Monitoring Network, p. 1-6

K. Hoshino, M. Brandt, C. Manfrino, B. Riegl, & S. Steiner, 2003, Assessment of the coral reefs of the Turks and Caicos Islands (Part 2: Fish communities), p. 480-499 in J. Lang (editor), Status of reefs in the Western Atlantic: Results of the initial surveys, Atlantic and Gulf Rapid Reef Assessment (AGRRA) Program

B. E. Lapointe, 1997, Nutrient thresholds for bottom-up control of macroalgal blooms on coral reefs in Jamaica and Southeast Florida, Limnology & Oceanography, 42: 1119-1131

B. E. Lapointe, D. A. Tomasko, & W. R. Matzie, 1994, Eutrophication and trophic state classification of seagrass communities in the Florida Keys, Bulletin of Marine Science, 54: 696-717

B. E. Lapointe, P. J. Barile, C. S. Yentsch, M. M. Littler, D. S. Littler, & B. Kakuk, 2004, The relative importance of nutrient enrichment and herbivory on macroalgal communities near Norman’s Pond Cay, Exumas Cays, Bahamas: a “natural” enrichment experiment, Journal of Experimental Marine Biology and Ecology, 298: 275-301

D. S. Littler & M. M. Littler, 2000, Caribbean Reef Plants, OffShore Graphics, Washington DC, 542 p.

M. M. Nugues, G. W. Smith, R. J. Van Hooidonk, M. I. Seabra, & R. P. M. Bak, 2004, Algal contact as a trigger for coral disease, Ecology Letters, 7: 919-923

M. M. Nugues, 2002, Impact of a coral disease outbreak on coral communities in St. Lucia: What and how much has been lost, Marine Ecology Progress Series, 229: 61-71

F. Perez, 2000, Rapid assessment of impacts on the coastal environment with particular emphasis on protected areas in Providenciales, Phase 1, DECR

F. Perez, 2000, Impacts on the coastal environment with particular emphasis on protected areas in Providenciales, Phase 2, DECR

B. Riegl, C. Manfrino, C. Hermoyian, M. Brandt, & K. Hoshino, 2003, p. 460-479 in J. Lang (editor), Status of reefs in the Western Atlantic: Results of the initial surveys, Atlantic and Gulf Rapid Reef Assessment (AGRRA) Program

S. Rosenberg, 2001, Diving and Snorkeling Turks and Caicos, 127p., Lonely Planet, Melbourne

M. A. Rudd, 2003, Fisheries landings and trade of the Turks and Caicos Islands, Fisheries Centre Research Reports, 11: 149-161

H. E. Sadler, 1997, Turks Islands Landfall: A history of the Turks and Caicos Islands, 299p., United Cooperative Printers, Kingston

C. Sear, M. Hulme, N. Adger, & K. Brown, 2001, The impacts of global climate change on the UK Overseas Territories, National Resource Institute, The Tyndall Centre, p. 1-51

J. E. Smith, M. Shaw, R. A. Edwards, D. Obura, A. Pantos, E. Sala, S. A. Sandin, S. Smriga, M. Hatay, & F. M. Rohwer, 2006, Indirect effects of algae on coral: algae-mediated, microbe-induced coral mortality, Ecology Letters 9: 835-845

S. C. C. Steiner, 1999, Species presence and distribution of Scleractinia (Cnidaria: Anthozoa) from South Caicos, Turks and Caicos Islands, Bulletin of Marine Science, 65: 861-871

K. Sullivan, M. Chiappone, & C. Lott, 1994, Abundance patterns of stony corals on platform margin reefs of the Caicos Bank, Bahamas Journal of Science, 1: 2-12

T. Topalian, 1992, An Assessment of the Status and Needs of Marine Conservation, Education, Policy and Resource Management in Turks and Caicos Islands, British West Indies" in the Proceedings of the 7^th International Coral Reef Symposium, Guam, 1992, Vol. 2, 1083-1090.

P. Vroom, K. N. Page, J. C. Kenyon, & R. E. Brainard, 2006, Algae-dominated reefs, American Scientist 94: 430-437

ACKNOWLEDGMENTS

We thank the entire staff of DECR for assistance with office and field work and logistics that made this study possible. We thank the dive operators for discussion on long-term change in TCI reefs, especially Art Pickering, Algrove Alexander Smith, Mitch Rollins, Everette Freites, and Michael Clarke. In particular we thank Wesley Clervaux, Brian Riggs, and the boat captains and staff.

FIGURE LEGENDS

Fig. 1 The Turks and Caicos Islands are in the open North Atlantic Gyre east of and upcurrent from the Bahamas

Fig. 2 The Turks and Caicos reefs largely occur at the edge of the drop off around the two  banks, the large and very shallow Caicos Bank, and the smaller Grand Bahama Bank, separated by deep water of the Columbus Passage.

Fig. 3 Locations assessed in Northeast Providenciales.

Fig. 4 Locations assessed in Northwest Providenciales.

Fig. 5 Locations assessed in Southeast North Caicos.

Fig. 6 Locations assessed in Northeast South Caicos

Fig. 7 Locations assessed in South South Caicos

Fig. 8 Locations assessed around Grand Turk.

Fig. 9 Monthly average TCI sea surface temperatures (in degrees C), 1982-2002 (from Hayes & Goreau submitted). The number on the horizontal axis is the number of months starting in January 1982. The low values in 1982 are an artifact of the satellite measurement and were caused by the eruption of a Mexican volcano, El Chichon, that injected a large amount of high sulfur aerosols into the upper atmosphere, causing satellite-derived sea surface temperature measurements to be contaminated with upper atmosphere haze temperature readings. Even when that year is removed from the record, there is a clear upward trend in maximum, minimum, and mean temperatures, caused by global warming. The warmest year was 1998, but almost all years since 1998 have been very warm, making bleaching a nearly annual event. 2004 and 2005, not shown, continued this pattern, with hotter conditions and more bleaching in 2006. Years of known bleaching are identified, and the monthly average temperatures causing bleaching are seen to be 29.4 degrees C and above.   

Fig. 10.  Bleached coral. The white coral on top is recently dead, the part below is alive but the tissue has lost its color. The brown line marks the dying tissue.

Fig. 11.  White Plague. The brown coral is normally colored, the white part is recently dead, not bleached. 

Fig. 12. Black Band disease. The areas on top within the black bands are dead, the parts below are alive but bleached.

Fig. 13.  Sea Fan Disease (Aspergillosis). Note that the photo was taken in 2006, the date is incorrect due to camera memory error.

Fig. 14.  Algae smothering a reef. This species, Microdictyon marinum, forms dense green mats that overgrow and kill corals.

Fig. 15          Network of statistically significant positive correlations between ecological and environmental variables emerging from the non-parametric data analysis. Thin lines indicate correlations significant at the 95% confidence level (P<0.05), medium lines indicate significance at the 99% level (P<0.01), and thick lines at the 99,9% level (P<0.001).

Fig. 16 Network of statistically significant negative correlations between ecological and environmental variables emerging from the non-parametric data analysis. Thin lines indicate correlations significant at the 95% confidence level (P<0.05), medium lines indicate significance at the 99% level (P<0.01), and thick lines at the 99,9% level (P<0.001).

TABLE 1.   Matrix of Spearman rank order correlation coefficients between all pairs of variables.  The triangular lower left half of the matrix gives the numerical value of the correlation coefficient. Numbers in black are not statistically significant (P>0.05), numbers in red are statistically significant at the P<0.05 level, numbers in blue are strongly statistically significant at the P<0.01 level, and numbers in green are very strongly statistically significant at the P<0.001 level. The triangular upper right half of the matrix gives the sign of the correlation.

TABLE 2.  Summary of the frequency of positive and negative correlations by their degree of statistical significance.