Figure 1.  Do you think your coral reef aquarium has water anywhere near as pure as what this turtle was swimming through?  Think again; the average aquarium water has toxic concentrations of many trace metals.

We have all set up our coral reef aquaria with the goal, either specifically expressed or implicitly understood, of providing a good home for the animals we keep.  Probably all aquarists, myself included, think that the coral reef environment is a beautiful environment, generally free from the environmental degradation we see around us in our daily lives.  It follows that one of the strongest tenets of the reef aquarium hobby is that for success to occur, the aquarist has to use the best materials available and take care not to introduce sources of pollution into their systems. 

There is much “folk wisdom” concerning what aquarists must and must not do.  Some of the things that they must do are to use pure water, good salt mixes, and avoid the possibility of contamination with any potential toxic pollutants.  Unlike much of the dogma of this hobby, this example of aquarium lore is undoubtedly true, and all of us know of some examples where this folk wisdom has been violated, either purposefully or accidentally, with disastrous results.  I think every aquarist who has been keeping aquaria for more than a couple of years has had a disaster or two that can be directly traced to some sort of pollution event.  Consequently, we all take a great deal of care to avoid such events. 

Even with the best of care and intentions, however, there are a lot of “inexplicable events” or “strange deaths” in coral reef aquaria.  These odd events include such occurrences as animals that have thrived for many years suddenly, and without apparent cause, starting a downward spiral in their health leading, eventually, to their death.  Other similar events include the wholesale decline of older tanks, much to the frustration and sorrow of the aquarist who often strives mightily to keep those systems healthy.  In many cases, this type of decline involves the deaths of animals that have been kept for upwards of five years, and to which the aquarist has become quite emotionally attached.  Such events are both sad, and perplexing.

These unexplained mortalities do not only happen in older tanks.  Many deaths of newly purchased, or newly collected, animals also may fall into this category.  Although transport stress and handling can be seriously damaging to many animals, enough animals get through the distribution network to provide what should be a steady supply of animals in good health.  Nevertheless, many aquarists report that newly introduced animals which appear to be healthy, seem to sicken and die with a couple of weeks, often much faster, in home aquaria. 

I think that the cause of many such unexplained mortalities lies hidden in our tanks, and is even abetted or accentuated by well-meaning aquarists.  My hypothesis is that a primary cause of many, if not all, of these unexplained mortalities is heavy metal poisoning.  I believe this metal poisoning is specifically due to the excessively high concentrations of some of the trace elements that are present in tank water.

Methodology:

Unfortunately, there is no simple set of guide books to peruse or examine that will tell an aquarist whether or not mortality of any reef aquarium organisms has occurred from a specific cause.  Consequently, the assessment of these mortalities as being caused by toxic trace elements must be done by assessing circumstantial evidence.

If my explanation of lethal trace element concentrations is correct, I must show three things.

First:  The potentially toxic trace element concentrations in at least some aquarium systems must be shown to be much higher than are found in nature.  This was done by examining the trace element concentrations found in tank water as reported in the tank water study results in the Reefkeeping Articles in March and April.

Second:  These elevated trace element concentrations have to equal or exceed some minimum threshold of experimental mortality.  In other words, there has to be some sort of empirical or experimental evidence showing that the concentrations that are found in reef aquaria are greater than some lethal threshold.  This condition is much more difficult to assess.  It seems that toxicologists are simply not interested in assessing survival of organisms in aquaria.  The data that are available come from a variety of studies, all found and reported in the peer-reviewed scientific literature (see the “References Examined” section).   Because of the disparate nature of these reports, no two studies examined the same suite of trace elements, and generally the tests were done in different ways, and on different target organisms.  Consequently, the story of the effects of toxic trace elements is a patchwork quilt, but no less convincing for that, as I hope to show.

Third:  There must be a reasonable and biologically meaningful explanation of the selective pattern of mortality.  This part of the story is largely incomplete.  Either many of the organisms that interest aquarists do not interest the scientists examining these problems, or the reef aquarium organisms are not particularly amenable to study for any number or reasons, or the problem of trace metal toxicity is not one that excites many coral reef oriented scientists.  Additionally, both the number of researchers examining the marine environment and the amount of funding for all aspects of marine ecology is quite limited.  The potential responses of many organisms to toxic materials, although of potential interest, may be quite low on the priority list of any given granting agency, as that agency may find other types of questions and problems more worthy of funding.  For whatever reasons, this part of the explanation must be somewhat tentative.  Nonetheless, many good data are available for comparison and workable and plausible explanations can be made for most of the effects that are seen.

Results:

In examining the peer-reviewed literature for data on trace elements in coral reef ecosystems, in particular, and polluted systems, in general, I found that there was a particular suite of elements that researchers had concentrated upon (Table 1).  These were elements that were known to pose particular human health problems.  That they may cause problems with marine organisms is really incidental to the initial focus of the studies.  Economic concerns dictated these studies, and until very recently economic concerns predicated by human health issues alone drove the studies of the effects of toxic trace elements.  Recently, in the last 15 years or so, the adverse environmental effects of abnormally high concentrations of many of these trace materials has been studied, and in some cases the environmental effects may be paramount.  Nonetheless, the major factor driving these studies, and particularly the choice of toxic materials to be studied, remains the potential for human health effects or the effect on human activities, such as fishing or recreation.  The analytical tests determining the concentrations of these elements are expensive to run, and it is difficult to justify them to granting agencies based only on the presumed environmental effects of the element.  The major side effect of such concerns, however, is that any suite of elements examined in detail is a relatively small subset of all possible elements.  Such a narrowing of focus always runs the risk of overlooking some important effects caused by some other element.  Until somebody stumbles on these effects, we have no way of predicting their magnitude or action.

It is worth noting that in doing a literature search on studies of trace elements examined in the ecological or environmental literature, and reading well over one hundred papers, going back 25 years or more, in absolutely no case were these elements considered as anything but pollutants.  This bears repeating:

Trace elements in heightened concentrations are considered to be poisons, nothing more, nothing less, by every researcher examining them.

The literature on the effects of these chemicals, specifically, Cadmium, Copper, Chromium, Lead, Mercury, Nickel, Vanadium, and Zinc, is rather depressing, frankly.  It deals with nothing more than mortality factors or how the organisms deal, perhaps more importantly, how they cannot deal, with enhanced levels of the chemicals. 

There was not a single study showing any benefit to increased concentrations of any of the above group of elements.

The concentrations of the particular trace elements found in the aquaria of the Tank Water Study (see the links above) varied significantly (Table 2).  Several important, and highly toxic, elements (Cadmium, Chromium, Lead, and Mercury) were not detected in any of the aquaria, even though in some cases they are common in the foods added to aquaria.  In data on tank exports, for a project in progress, all of these metals are found in detectable skimmate and skimming may be a major avenue for their removal from aquaria.  It is worth remembering, however, for some of those non-detectable elements, that the test detection limits were well above the concentrations found in natural seawater.  Nonetheless, several other toxic trace elements were found, and had concentrations between 12 and 540 times that of natural oceanic seawater.

As there was no single study that treated all of these elements simultaneously and in like manners, I have presented a synopsis of the studies for each element. 

Please note: This is not a scientific journal, and as it is not, to increase readability, I will not reference each datum in the manner done in a scientific paper.  However, all the references listing the specific data used in the synopses or elsewhere are listed in the “References Examined” section at the end of this article.

Types of Evidence:

In examining the evidence about the toxicity of trace metals, it is important to be aware of the types of tests, and some of the terminology.  Toxicity is generally determined by putting some test animal subjects under controlled experimental conditions of temperature and salinity.  Generally, the temperatures and salinities in studies I examined were kept at reef conditions; so the temperatures were between about 81 º F and 84 º F, and the salinities were about 36 ppt.  Temperature and salinity conditions are not neutral factors in these tests.  Animals kept at lower temperatures take longer to die if poisoned, simply because their metabolic rate is lower.  Abnormal salinity, either higher or lower than normal, stresses the animals, and often results in much more rapid mortality.

The tests are not open-ended; they typically have a definite duration.  Generally, the tests are made to be run rapidly, under specific conditions, and durations are specified in the methods.  In these toxicity tests, the “endpoint” of the test is predefined, and it varies with the test.  This endpoint may simply be the fraction of the test organisms surviving or acting normally at the end of the test.  In different tests, the endpoint may be defined as the concentration of the trace metal necessary to produce 50% mortality (or change in behavior) during the test period.  These latter endpoints are referred to as LC_50, an acronym for:  Lethal Concentration, 50%, or, for example, IC_50, an acronym for Inhibitory Concentration, 50%, for a behavioral inhibition.

Such endpoints are common in what are termed “acute” tests, because the effect is felt over a very short period, or acutely.  It is generally considered that acute tests give values for toxicity that are much greater than for “chronic” or long-term exposure.  For example, a given trace metal may cause acute damage or death at a concentration of one part per billion.  This is presumed to indicate that the trace metal is probably causing long-term damage at concentrations as low as one tenth of a part per billion.  In other words, the concentration values given in acute tests are typically far higher than the concentrations normally causing long–term damage at low concentrations.

Finally, as a comparison, I have included some interesting data about reef aquarium water.  The first of these data sets comes from the average concentration of the metal in question in artificial sea water mixes taken from the 1999 study of Atkinson and Bingman. These data are available in detail by following the link above or by examining my April Reefkeeping article.  The second set contains data from the concentration of some of the metals found in the water within a pile of arsenic slag in a marine ecosystem which is part of an EPA superfund site.  Unfortunately, these data are not part of the public record, so I cannot specifically cite the study.  However, I did the work and used the same lab for this study as I used in the tank water study.  You will have to trust me for the veracity of these data. 

Trace Element Synopses:

Arsenic:

This metal was found in only one aquarium in the Tank Water Study, and that tank had an arsenic concentration of 0.02 ppm (parts per million) or 20 ppb (parts per billion).  Laboratory experiments in estuarine waters (freshwater runoff may have naturally high arsenic concentrations) show mortality effects in a wide array of estuarine infauna at concentrations of 4 to 8 ppb.  I was unable to find any data on the effects of arsenic on coral reef animals.  Presumably, the low concentrations normally found around reefs have led researchers to ignore arsenic in their tests.  There were no data on arsenic in artificial salt mixes.  In the water inside the arsenic slag pile, the concentration varied between 5ppb and about 2000 ppb, most of the values being below 70 ppb.  The values for the one aquarium were within the range of concentrations found within the marine arsenic slag pile.

Copper:

Copper is an essential nutrient element which is highly toxic to many animals and algae at concentrations only slightly above those found in NSW.  In aquaria, the copper concentration in The Tank Water Study averaged = 24.4 " 5.3 ppb; with the maximum value of 38 ppb. Its effects on corals are profound and occur at concentrations below those typically found in aquaria.   In a test running only four hours, copper inhibited fertilization in several corals:  the IC₅₀ (inhibiting fertilization concentration) was 17.4 ppb for Acropora millepora, and 14.5 ppb for Goniastrea aspera.  In another test, a copper concentration of 20 ppb reduced fertilization to about 45% of normal for Goniastrea aspera.  At a seawater concentration of 100 ppb, copper reduces fertilization to about 51% of normal in Favites chinensis.  In a test running for 96 hours, 50 percent of the larvae of Pocillopora damicornis died in water with a copper concentration of 63 ppb.  Fifty percent of adult colonies of Montipora verrucosa were killed in a 96 hour test at 48 ppb.  In Porites lutea colonies exposed to 30 ppb copper and/or lowered salinity (at two-thirds of normal), the zooxanthellae were affected and had reduced primary production.  Not surprisingly, copper affects sea anemones as well:  at 10 ppb the carbonic anhydrase activity is reduced to about half of normal for Condylactis gigantea and Stichodactyla helianthus.  Carbonic Anhydrase (CA) activity can be used to provide a direct measure of metal induced stress.  A reduction in CA activity results from stress.  Snails are also affected, 50 percent of Nassarius festivus died in an acute test with a copper concentration of 360 ppb, and in long-term experiments, feeding was reduced with concentrations of 50 ppb.  In numerous other bioassays of copper effects, the generalized effect endpoint or EC₅₀ occurs at concentrations between seven and 250 ppb.

In the inner Gulf of Thailand, considered to be a polluted region, the copper concentration varies between 0.33 to 14.16 ppb, while in Manila Bay it varies from about 0.02 to 2.9 ppb.  In the examination of artificial salt mixes, the average concentration was 152 " 260 ppb, which turns out to be significantly greater than the copper concentration, 25 ppb to 77 ppb, found on the inside of the arsenic slag pile that contained a significant amount of copper.

Iron:

This material is not a toxic material, but is important in a study of toxic trace elements because iron oxides often bind and remove toxic trace elements from seawater solutions.  Iron was not detected in any of the tanks in the Tank Water Study.  This is possibly because it was bound to some of the trace elements, or possibly because it is rapidly taken up by bacteria and algae.  In artificial seawater mixes, it was found with an average concentration of 67 "146 ppb.  By comparison, in the arsenic slag pile, the iron concentrations in pore water were 50 ppb to 100 ppb.

Nickel:

Aquarium nickel concentrations in the Tank Water Study averaged 24 " 6 ppb, with a maximum concentration of 39 ppb. In various short-term tests, mostly on temperate animals, the EC₅₀ = 1.2 ppm to 320 ppm.  However, nickel concentrations of 25 ppm killed adult Condylactis gigantea and Stichodactyla helianthus within 24 hours; and 40 ppb caused carbonic anhydrase activity to be reduced to about one fifth normal in Condylactis gigantea and one tenth normal in Stichodactyla helianthus.  In laboratory treatments of various temperate Atlantic estuarine animals, the toxic effects started at around 4 ppb.  The average nickel concentration in artificial seawater is 114 " 14 ppb.  Nickel was not detected in the arsenic slag.

Tin: 

The aquarium average concentration of tin in the Tank Water Study was 95 ± 0.01 ppb.  This average was absurdly high; 200,725 times the concentration of tin in NSW.  I was not able to find any data on the effects of the tin ion on marine animals.  Studies of tin effects seem only to be concerned with the highly toxic antifouling compound tributyltin and studies of tin proper are very rare.  Tin was not tested for in the water from the arsenic slag.

Vanadium:

In aquaria from the Tank Water Study, the average vanadium concentration was 22.6 " 4.7 ppb; with the maximum value being 37.0 ppb.  At concentrations of 20 ppb, vanadium reduces carbonic anhydrase to about half normal levels in Condylactis gigantea and Stichodactyla helianthus.  In artificial seawater, the average concentration was 168 " 20 ppb.  Vanadium was not tested in the water from the arsenic slag.

Zinc:

Zinc concentrations in aquarium water averaged 212 " 21 ppb and the maximum value was 260 ppb.  The LC₅₀ for Nassarius festivus is 1760 ppb and feeding is reduced at 200 ppb.  Zinc concentrations of 100 ppb to 1 ppm cause lowered growth rates in zooxanthellae, while a concentration of 100 ppb reduces fertilization to about 7% of normal in Favites chinensis.  In laboratory treatments of estuarine organisms, effects become noticeable at 1.2 ppb to 8 ppb.  In artificial seawater, the average zinc concentration is 37 " 17 ppb.  Zinc was undetectable in the water in the arsenic slag, even though it was present in relatively large amounts in the solid slag.

Figure 2.  Condylactis gigantea, shown here with symbiotic juvenile yellow head  wrasse, suffers toxic effects from copper, nickel, and vanadium when they are present in the concentrations found in the average aquarium.

Discussion:

For the purposes of this article, I am assuming that the aquaria sampled in the Tank Water Study and reported on in the February through April issues of Reefkeeping Magazine are representative of reef tanks.  There is no particular reason to suspect that they are atypical, but neither is there any compelling reason to suspect that they might not be.  This is the only study of tank water chemical concentrations simultaneously examining tanks from across the United States.  Consequently, when I refer to “an average tank,” it should be understood that I mean an average tank from that study, but that I assume that such an average tank is representative of all reef tanks.

The information that I found in the scientific literature is primarily of two kinds.  First, there are comparisons with three sites, the Inner Gulf of Thailand, Manila Bay, and an arsenic slag pile, considered by the authors of the studies to be highly polluted.  Second, there are tests showing how rapidly organisms die, or how their behavior changes, when put into solutions containing known concentrations of the pollutant.

Site Comparisons:

Two of these sites were tropical:  Manila Bay in the Philippines, and the Inner Gulf of Thailand.  The data in these articles is from surveys and, for each general area, gave a range of values for each chemical.  Generally, at least some sites in each of these regional areas were near point sources of industrial pollution.  The other site, on the west coast of the United States, is of a well-defined area consisting of a pile (about one half mile long and a few hundred yards wide) of arsenic slag that had been deposited over several decades in a marine embayment.  Twenty samples of water were collected from within the slag, and analyzed for heavy metals.  The slag contained up to four percent arsenic, two percent of zinc and correspondingly high amounts of other heavy metals by weight.

How does the water from average aquarium compare to these polluted sites?  

With regard to arsenic (when found), copper, nickel, tin, and zinc, the average tank water must be considered as being polluted with heavy metals. 

The water from the average reef tank is clearly dangerous to the organisms put into it.  Although arsenic was found only in one tank, the concentration in that tank exceeded the threshold levels for toxic effects in some tested estuarine animals.  The average copper concentration in the water of a reef tank is more than twice that necessary to inhibit fertilization of coral gametes, and is about half of what has been reported as necessary to kill small colonies of Montipora verrucosa in four days.  At the average concentration, it is more than twice the concentration necessary to affect sea anemones.  Nickel concentrations are also within the lethal range for the sea anemones Condylactis gigantea and Stichodactyla helianthus.  I could not find any tests reporting the effects of nickel on corals.  Vanadium was likewise in the range necessary to cause damage to the aforementioned anemones.  Zinc was over 100 times more concentrated than the lethal concentration for Nassarius festivus and more than twice as concentrated as the levels that effectively stopped fertilization in some corals.

It should be remembered that these are data from a rather restricted set of trace materials, and that other trace elements, so far unstudied, are likely to give similar results.

About the biological tests: 

The different bioassays or tests run with animals or other organisms are designed to show the effects of the pollutant in question upon whole classes or categories of organisms.  The most delicate of the life history stages for any animals are the gametes or reproductive cells, and often these are tested as they will rapidly show toxic effects at concentrations below the threshold for damage to healthy adults.  However, the rationale for the use of the tests is that the short duration of the tests is balanced by the fragility of the animals.  Consequently, if a fragile animal dies at a low concentration in a short time, it is presumed that a more robust animal would be killed or seriously injured at the same dose over longer time periods or at a slightly higher dose for shorter periods.  ANY indication of mortality of any of these categories of animals at the concentrations found in our tanks indicates that the animals in these tanks are stressed.  This is particularly important since the concentration of these chemicals will tend to rise with time, as more trace materials come in with each feeding event. 

Causes:

What causes these excessively high trace metal concentrations?  Initially, the problem occurs with artificial seawater mixes that have abnormally high concentrations of these materials.  Then, most foods have high concentrations of the same chemicals, and other toxic materials, such as cadmium.  In most cases, these materials are in foods (Food and Additive Study) due to the bioaccumulation of the materials in animal or plant tissues.  Organisms tend to sequester trace elements in their tissues, primarily in special proteins called metallothioneins, as a means of detoxifying them.  Also, there is inadequate export of the materials due to any number of causes, but including such factors as poor skimming, inadequate water changes, and inadequate biomass export.  Finally, in some cases well-meaning, but ill-advised aquarists often add supplements containing unknown quantities of some trace elements.

Aquarists would be well advised to remember that if trace elements are necessary, they are necessary only in trace amounts, and secondly that unless materials are exported or permanently locked into some insoluble form, they remain in the aquarium and have the potential of accumulating in the water column.  Soluble trace elements are very likely poisonous in the extreme as the data for the bioassays on the metals mentioned above indicate, and they should never be added to a system. There are NO data that any trace element additions are beneficial, and for any trace element for which there are data, excess amounts are detrimental.  No adequate test kits exist for the vast majority of these materials, and few supplements list their ingredients in a trustworthy manner.  Consequently, it is prudent not to add any at all to a system.

Biological Accommodation:

Organisms can, to some extent, adjust to the excessive loads of some of the trace metals they encounter.  They may incorporate them into inert materials that are exterior to the body, such as the skeleton of corals.  Corals have long been known to incorporate materials such as excess trace elements into their skeletons, and this is a very functional way to deal with such material.  Similarly, some crustaceans incorporate materials into their integument, and these materials are shed with each molt.  Unfortunately for aquariums, as the molt breaks down, the materials are released back into the small closed water volume.  All animals incorporate excessive trace metals into specific proteins called “metallothioneins” which do nothing more than bind the metal and keep it from interfering with the organism’s metabolism. 

However, when all of these processes become saturated, the trace element concentration in the water may increase to toxic levels.  At first, this accumulation forces animals to enter a state of chronic poisoning as the trace element concentration rises above the species’ critical threshold.  Then, if the material continues to accumulate, the threshold for acute poisoning is reached, and the animal sickens and dies, often in a very short period.

Although the final effect of toxic trace metal build up in a tank is the death of one or more organisms; not all species are equivalently susceptible to heavy metal poisoning, and likewise not all individuals within a species.  This causes mortality to be “spotty” and very problematical to the aquarist.  Once death by this kind of poisoning starts to occur, it may accelerate and spread throughout a tank.  As each victim succumbs, the metals (and possibly other harmful substances) accumulated in their bodies are released, further increasing the toxic load in the tank.  Such a cascade effect could well be the cause of “old tank wipeout” or the progressive inexplicable sickening, followed by death, of animals that had once been thriving earlier in an aquarium system’s history.  The older a tank, the more accumulation of these toxic materials is likely to be a problem.  Delicate organisms often die first, and replacement animals never quite seem to thrive and they die. Then, some other organisms succumb, and eventually the downward spiral destroys the inhabitants of a tank, and the aquarist’s self-confidence.

Additionally, all organisms have a finite capability of handling stress.  As the load of chemicals increases in the system, the inhabitants have less capability of responding to other stresses such as temperature extremes, either too high or too low, or salinity fluctuations.  Such stressed organisms are ripe for mortality due to other causes such as diseases, competition, or malnutrition.

Validation of My Explanation:

I mentioned earlier, for my explanation of lethal trace element concentrations to be correct, I needed to show three things:

I have shown that the first two conditions are unequivocally true for at least some of the elements in tank water.  In the average reef tank, the concentrations of several trace metals are excessively high, and far in excess of their concentrations in NSW.  Several of these metals show lethal effects or damage to organisms at the concentrations found in reef tanks.  Finally, there is a partial explanation for the last condition; enough, I think, to validate this explanation as being true.

Solution to the Problem:

The best solution would be to use natural seawater in a system with frequent large water changes.  This option is not feasible for the vast majority of hobbyists living more than a few miles from the seashore.  For those aquarists who must use artificial seawater, it is imperative to find out the contents of the salt mixes that are available, and to use the one with the lowest concentrations of trace materials.  If possible, RO/DI water should be used, or water should be tested to learn what trace elements are in it.  Trace element supplements should not be added because of toxic levels already present for some elements in artificial seawater.  As most reef aquaria are generally kept on a starvation diet, cutting back on feeding is not an option, so trace element accumulation will still occur.  Regular and large water changes will help slow the accumulation of trace elements, and aggressive skimming may assist in this process.  Additionally, the use of filtration media such as Polyfilters™ and activated carbon, which may remove heavy metals, is indicated for probably ALL systems.  As toxic trace elements will likely accumulate in live rock and sediments, there is likely a finite life span for a reef tank.  As a guess, I would suspect that for hobbyists that have to rely on artificial sea water, it will be prudent to breakdown and re-establish a tank every four or five years, perhaps more frequently.  For those that use natural seawater, the period between breakdowns will likely be longer, but I can’t hazard a guess as to how long it might be.

The data presented herein and in the earlier tank water study papers, indicate that there are some very significant problems with the salt water used as a medium in reef tanks, at least in the United States.  Many of the 31 elements which were examined (see earlier articles) may be considered as both potentially toxic and commonly found in natural marine communities.  However, only a few of them were found in detectable amounts in the samples of reef tanks. That does not imply that other materials might be equally or more important as the elements I chose to discuss, but simply mean that at present, definitive data are lacking for other trace elements.  As more research is done, other toxicants will be found and their specific effects determined.

References Examined:

Alutoin, S., J. Boberg, M. Nyström, and M. Tedengren. 2001. Effects of the multiple stressors copper and reduced salinity on the metabolism of the hermatypic coral Porites lutea.  Marine Environmental Researh 52: 289-299.

Breitburg, D. L., J. G. Sanders, C. C. Gilmour, C. A. Hatfield, R. W. Osman, G. F. Riedel,  S. P. Seitzinger, and K. G. Sellner.  1999. Variability in responses to nutrients and trace elements and transmission of stressor effects through an estuarine food web.  Limnology and Oceanography. 44: 837-863.

Carey, J. 1981.  Nickel mining and refinery wastes in coral reef environs. Proceedings of the Fourth International Coral Reef Symposium, Australia.  Volume 1.  137-146.

Cheung, S. G., K. K. Tai, C. K. Leung, and Y. M. Siu. 2002.  Effects of heavy metals on the survival and feeding behaviour of the sandy shore scavenging gastropod Nassarius festivus (Powys).  Marine Pollution Bulletin.  In  Press.

DiBacco, C. and L. A. Levin.  2000.  Development and application of elemental fingerprinting to track the dispersal of marine invertebrate larvae.  Limnology and    Oceanography, 45: 871-880.

Gilbert, A. and H. E. Guzman. 2001. Bioindication potential of carbonic anhydrase activity in anemones and corals.  Marine Pollution Bulletin. 42:742-744.

Goh, B. P. L. and L. M. Chou.  1992.  Effect of low levels of zinc on zooxanthellae cells in culture.  Proceedings of the Seventh International Coral Reef Symposium, Guam.  Volume 1.  367-372.

Heyward, A. J. 1988.  Inhibitory effects of copper and zinc sulphates on fertilization in   corals.  Proceedings of the Sixth International Coral Reef Symposium, Australia. Volume 2.  299-309.

Negri, A. P. and A. J. Heyward. 2001. Inhibition of coral fertilization and larval metamorphosis by tributyltin and copper. Marine Environment Research. 51:17-  27.

Parametrix, Inc.  1995. Phase 1 Data Report and Phase 2 Sampling and Analysis Approach.  Expanded Remedial Investigation and Feasibility Study,  ASARCO Sediments Superfund Site.  Final Report.  Unpublished Report Prepared For ASARCO, Inc., Salt Lake City, Utah.  Available from EPA Region X Library, Seattle, WA.

Reichelt-Brushett, A. J. and P. L. Harrison. 1999.  The effect of copper, zinc, and cadmium on fertilization success of gametes from scleractinian reef corals.  Marine Pollution Bulletin.  38:182-187.

Rumbold, D. G. and S. C. Snedaker. 1997.  Evaluation of bioassays to monitor surface microlayer toxicity in tropical marine waters.  Archives of Environmental Contamination and Toxicology.  32: 135-140.

Scott, P. J. B. and M. Davies. 1997.  Retroactive determination of industrial contaminants in tropical marine communities.  Marine Pollution Bulletin. 34:975-980.

Velasquez, I. B., G. S. Jacinto, and F. S. Valera. 2002. The speciation of dissolved copper, cadmium and zinc in Manila Bay, Philippines.  Marine Pollution Bulletin. In  Press.