I have spent much of the last couple of years taking an indirect look at what happens in reef tanks with regard to various chemical constituents (Shimek, 2001, 2002a, b, c), primarily the toxic heavy metals, referred to in the hobby as “trace elements.” Some of these materials are biologically necessary, but none has been ever shown to have any benefits at concentrations above those found in natural sea water. Indeed, most of them have been shown to be both acutely and chronically toxic at even slightly higher concentrations (see, for example, the discussion in Shimek, 2002d and these references: Alutoin, et al., 2001; Breitburg, et al. 1999; Goh, and Chou, 1992; Heyward, 1988; Negri, and Heyward, 2001; Reichelt-Brushett and Harrison, 1999). Those studies of documented toxicity notwithstanding, I have found that many of these chemicals have exceptionally high concentrations in the liquid medium of reef aquaria, as well as within the food we add to the systems.

There are numerous documentations of presumed cases of metals toxicity in our systems, mostly due to copper contamination. Additionally, there are likely many other cases of “hidden” metals poisoning. Most of these events, from the point of view of the aquarist, likely occur haphazardly and sporadically enough that the actual cause of mortality has either been overlooked or attributed to some other cause. I think that most cases of “acclimation death” are due to insufficient acclimation to the heavy metals concentrations found in our tanks. Most marine organisms can detoxify these materials to one degree or another, but it takes time for the metabolic pathways which can do this (mostly the production of metallothioneins, the proteins used to bind and detoxify metals) to become active. This activation may take from several hours to several days.

Many, if not most, coral reef organisms are from shallow water and, as such, most of them are subject to significant and rapid changes in salinity. Salinity changes of several parts per thousand are not uncommon with tidal shifts and as a result of heavy rains. For example, a few hours of heavy tropical rains can often drop the salinity of a coral reef lagoon from 36 ppt to 30 ppt down to a depth of twenty feet or more. While this stresses the animals, most of them will not die. Likewise, after several days of hot, dry conditions, lagoonal animals often have to contend with salinities of 39 ppt or more. If this occurs during a spring tide period where the tidal exchange is minimal, the salinity may remain elevated for weeks. When the tidal exchanges do become more extreme, such hot saline water flows out over the reef and animals that a few minutes earlier might have had much lower salinity in waters surrounding them. Again this stresses the animals, but it generally doesn’t result in the immediate mortality often referred to in the reef aquarium hobby as “salinity shock.” Frankly, for animals in good health otherwise, transient salinity shock simply is very unlikely to cause any long-term stress. On the other hand, sudden exposure to high concentrations of heavy metals, even if they might be tolerable after acclimation, will be lethal.

I have documented the metal concentrations from samples of aquarium tank water taken from 23 tanks across the United States (Shimek 2002a, b) and are summarized in Table 1, below. These data are unambiguous and conclusive, showing that in these tanks, many of the toxic heavy metals concentrations exceeded demonstrated lethal levels for many corals and other marine organisms (Shimek, 2002d)

There has been much discussion of these results with various proposals being put forth as to why the tanks should not show the effects of toxic metals, even though the metals concentrations were apparently high enough to cause such effects. Some of the explanations for ameliorating effects were:

All of the ameliorating effects listed may be true to some degree. For example, the metals may not be toxic to some adult organisms even though they are toxic to larvae and immature organism. Toxicity is often demonstrated with larvae or immature organisms as they are known to be more sensitive. Perhaps in some cases, the adults kept in reef aquaria are not as affected as are these other stages. For example, heavy metal toxicity, particularly to adult organisms, is often a long-term process occurring over several years and would not likely be immediately noticeable in many cases, except that “previously healthy” animals die of unknown causes after four or five years in an aquarium system. Those animals that have some method to detoxify the chemicals typically build up a significant body burden of the metals, which, in turn, results in long-term effects such as abnormal behavior, neurological damage, sterility and eventually early death. Additionally, any materials in solution, be they bound to a humic acid, an iron hydroxide, or some other chemical, are ingested by the organisms during their normal feeding (Chong and Wang, 2000; Sundelin et al., 2001). Digestive processes have the potential to release any bound metals, with net effect of transferring them to the animals. Additionally, many invertebrate organisms including corals, actively absorb organic materials across their outer epithelial surfaces. Such absorption will result in ingestion of these metals bound to the organic molecules. Finally, particulate material in the water is actively consumed by many coral reef animals; most notably for aquarists, many of the corals, tridacnid clams, and other suspension feeders. Finally, the presence of organic material may increase trace metal uptake significantly over what occurs in water with less organic materials in it (Guo, et al. 2001; Vasconcelos, et al., 2001; Wang and Dei. 2001). In effect, the amelioration of metals concentrations by organic binding or precipitation likely simply replaces acute toxic effects with long-term chronic ones.

Actual testing of reef aquarium water for toxic effects has not been done. I hoped to report on the results of some sea urchin larval bioassays in time for this writing, by testing various aquarium waters for toxic effects. However, I have been thwarted in those efforts by Mother Nature. Hurricanes Lila and Isidore impacted the Gulf Coast earlier this autumn, resulting in excessive amounts of fresh water runoff, which in turn caused mass spawnings of the local sea urchins around the facility of my source for sea urchins. As I need “unspawned” sea urchins for the tests, I now have to wait until a different species becomes gravid, probably sometime after the first of the year

Two means exist for rendering tank waters less toxic. The first is the precipitation or removal of the metal from the aqueous milieu. Once the material has been removed from the water, it is not toxic. It has to be in the water for organisms to either absorb it or ingest it. One way this precipitation might happen is biologically, as when corals detoxify the mildly poisonous metal, strontium, by incorporating it into their skeletons. Another way to precipitate the chemicals from the water may happen by strictly inorganic means, such as the precipitation of heavy metals by various iron hydroxides. Additionally, particulate substrates such as those found in deep sand beds, have in their depths anaerobic, sulfide-rich conditions that facilitate the precipitation of many of these poisonous materials as non-toxic sulfide minerals. Deep sand beds significantly favor the precipitation and retention of heavy metal sulfides, and are thus likely a significant factor in rendering tank waters less toxic. Precipitated materials in the deep sand beds accumulate with time, but they are not toxic as long as they remain insoluble in the sediments. If those materials should become soluble, however, they would present serious and acute toxicity problems. Such precipitates may become soluble if:

Any of these events can cause either acute toxicity, or accelerate chronic poisoning, and the events can vary significantly in duration and effect. To remove or reduce the threat of some catastrophic event happening, and to avoid long-term or chronic poisoning, it is to the aquarist’s advantage to reduce the accumulation of these materials either in the water or in the sediments.

The second way to detoxify tank water is to chemically or physically export the chemical out of the system. Several export methods exist for aquarists, but their effectiveness has hitherto not been evaluated in aquariums. In this article, I will describe some of the results from examining a few of the various ways commonly employed to export materials from aquaria.

Materials and Methods:

In this study I was, through the kind and generous help of several aquarists, able to examine the chemical constituents of four types of export materials. I examined 4 samples of skimmate, the liquid resulting from the foam that makes it through a foam fractionation device. I also examined one sample of skimmer sludge, the greenish-gray or black “mud” from the inside of a skimmer. One other person was scheduled to return some sludge, but had to drop out of the study, so unfortunately there was only one sample of sludge. This material is largely bacterial in composition, but as it is washed away during skimmer cleaning, it is a reasonable material to examine. There were three samples of Caulerpa; two samples of Caulerpa cupressoides, one of Caulerpa racemosa. The results from these algae were pooled for the analyses. Two samples of Xenia of unknown species were processed. I also processed one sample of a Sarcophyton species. However, as this latter species is uncommonly used as an export, I will report on it only as a comparison to the commonly exported soft coral, Xenia.

The samples were analyzed as in the Tank Water Study (See Shimek, 2002a, for details). Precleaned and treated jars were sent to me from the analytical laboratory. I repackaged the jars and sent them to the participants. All samples were collected, frozen and returned to me. I refroze them and then sent them to the laboratory.

Varying amounts of the materials were obtained, and in some cases there was insufficient material present to run all the tests. In these cases, the tests were prioritized. The sample results were reported in “as is” basis, based on the total wet weight of the samples. Tests for metals concentrations and nitrogenous compounds were done on all of the samples. Tests of the conventional nutrients and caloric content, however, were not done on several of the samples of skimmate and one of the Xenia samples. The samples had to be dried for these analyses and skimmate is, apparently, a fairly “weak tea.” Even a liter of skimmate contained insufficient material to perform all the tests. The two Xenia samples were of differing sizes, reflecting the growth rates in the two different aquaria that they originated in. One of the samples was simply too small for all the tests.

After processing and analyses, the data were returned to me for further analyses. It is important to realize that, as with the other tests in this series, the statistics that are presented are simply descriptive. I ran no statistical tests comparing the effectiveness of the various exports because there aren’t enough samples for the results to be meaningful. However, there are plenty of data to examine and discuss in the remainder of this article.

All of the descriptive statistics were calculated using spreadsheets, and are subject to operator error. If you find that I have made an error in calculations, please contact me and I will correct it.

Results and Discussion:

Baseline Conditions

In the discussion that follows, the exports will be discussed in relation to three distinctly different types of water. The first is natural sea water (NSW), with the various metals concentrations as reported in Pilson(1998). It is important for our purposes to realize that these data, as authoritative as they may be, may have little to relevancy to the waters over natural reefs. Relatively little chemical oceanographic sampling has been done over reef areas and, in many cases, what has been done was performed in a manner that makes comparisons to reef tanks dubious. For example, samples taken three meters above a reef samples “reef waters,” to be sure, but how those waters relate to what is happening closer to the reef is largely unknown. Most aquarists sample their reefs in far closer proximity than many oceanographers would dream of doing, and the few samples taken close to reefs show some very significant differences from the water a few meters above them. (see, for example: Stimson and Larned, 2000).

In all marine areas within a few inches of the substrate, the chemical composition of the water largely reflects what the nearby organisms have been doing to it. This is particularly the case with biologically “interesting” chemicals, and these, of course, are precisely the items of interest to aquarists. Reef water data taken a few meters above the reef might be, and has been, legitimately argued to tell little about the reef. Unfortunately, this limits comparisons severely, and I have taken the easy way out and just discussed NSW in the familiar general terms.

There is also the comparison with artificial sea water, and here two different comparisons can be made. The first is to what I refer to as the “average reef tank” water as determined by the Tank Water Study (Table 1 and Shimek 2002 b). These comparisons will be made first. The second comparison is to the average of the artificial sea water mixes described by Atkinson and Bingman (1999) and that comparison will be made subsequently.

Except for a few elements, the concentrations of elements in the water of an average reef tank, as defined by the Tank Water Study, bear little resemblance to the concentrations of the same elements in NSW. In fact, the relationship between reef aquarium water and NSW outside of a couple of the major ions, such as sodium and calcium, is so tenuous that one could legitimately say there is NO consistent relationship. Reef aquarium water has roughly the same amount of salt and calcium as in NSW, but the concentrations of all other ions differ from the concentrations found in natural sea water. Not only that, but they differ with inconsistent magnitudes and directions (Table 1). Nonetheless, many of the ions are substantially more abundant in reef aquarium water than in NSW, and often these are the ions with the highest potential for toxic effects, such as copper, iodine, nickel, and zinc.

The water found in an average tank, however, is not gathered from some place and added to a tank “as is.” Rather, it is the result of several continuous and simultaneous processes that occur to some initial water volume. The initial water can come from either natural sea water or artificial sea water, or a mixture of the two. Once the tank is filled with water, no matter where or how it originates, the initial tank water is manipulated by both the organisms in the reef aquarium system and the aquarist. Organisms actively alter the water in several ways.

Additionally, the aquarist manipulates the levels of these chemicals by adding foods, some chemical additives, making water changes, and by exporting various materials.

What organisms are doing to these chemical balances in any given tank is open to supposition; no numerical data about any organism’s secretion or accumulation of materials are known from aquaria, and precious few of these data are available from natural reef areas. That organisms are manipulating their chemical environment is a given. However, reef aquarium concentrations of many of the dissolved trace elements are so different from actual marine concentrations that it is impossible to even reasonably speculate about what is happening in a tank. Simply put, there are too few data available to generalize.

The manipulations that aquarists make to the chemicals found in the tanks may be estimated, given a couple of baselines, and I have attempted to do just that in previous articles in this series (See Shimek: 2002a, 2002b, and especially 2002c), but prior to this article few data have been available on the various means of exportation of materials. The results of the analyses of these materials are shown in Table 2.