People go about the process of starting a saltwater aquarium in a variety of ways. One of the common ways is

• See an attractive aquarium at a fish store or a friend's house.
• Learn what equipment you need for an aquarium.
• Buy the equipment, and the animals that you want to keep (hopefully not on the same day).
• As the aquarium develops, start learning more about the inhabitants, such as their feeding behavior, and their roles on the reef in nature.

This is exactly backwards. It results in unnecessary deaths and substandard conditions for a lot of what manages to survive.

The coral reefs in the Solomon Islands and elsewhere in the Indo-Pacific are so rich and successful, in short, because coral reef ecosystems evolved there. A better approach to planning an aquarium is to think of the task as one of modelling a natural coral reef ecosystem. With this in mind, we can proceed as follows.

• Learn about ecosystems in general. Many properties are common between different ecosystems, for example the mechanisms of food webs, the relationship between area and diversity, and the relationship between diversity and stability.
• Learn as much as possible about the particular ecosystem that we wish to model: the coral reef. In particular, we need to learn how each property that is common to all ecosystems is manifested in this particular ecosystem. This should ideally involve field study, but most of us have to settle for book research.
• Decide how to model each aspect of the natural ecosystem. In this step we decide what things we will try to include in the model directly, and which things will be abstrated to a greater degree. This is a fancy way of saying that you need to decide what kind of tank you want.
• Begin establishing the physical environment. This includes the physical body of water, the water chemistry, temperature, lighting, water motion, and so on.
• Introduce the basic structuring elements that define the ecosystem, through a series of block injections. This means adding live rock to the tank, gradually.
• Add the higher animals, beginning with herbivores, then low level carnivores, and finally, if ever, mid level or higher carnivores.

That's a summary of the approach, just to let you know where I'm headed. You're not expected to be ready to run off and set up a tank by this method, just yet.

Modelling, scaling, diversity, and stability

In a modelling effort of this type, we consider each aspect of the natural ecosystem, and decide how were are going to represent it in the model. There are some things that we can include in the model exactly, such as salinity and temperature, some things that we can include only to a limited degree, such as waves, and some things that we must completely simulate, such as the effects of the adjacent open ocean, which provides nutrient flushing and planktonic input. However we treat them, we need to consider each aspect of the natural ecosystem, and include it, or some approximation of it, in the model.

How we decide to model each aspect of the reef community depends on the financial and time resources available, the aesthetic considerations (spouse acceptance factor), and what we are interested in. At one extreme, we may be interested in a few species of ornamental fish, and not care much about the rest of the ecosystem. This thinking is what results in the typical fish-only tank. Setting up a tank for maintaining large fish is perhaps fine from a modelling perspective, but we still should model the other parts of the natural ecosystem. The traditional fish-only tank doesn't do this.

If there is one problem that will haunt us more than any other, it is the problem of scale. We are trying to take an ecosystem that exists over millions of gallons and hundreds of acres and compress it into a small glass box of tens of gallons to a few hundred gallons. You often hear that a larger aquarium is more stable than a smaller one. Usually this is heard in reference to chemical stability of the water, but it is also true of the stability of the biological community.

The stability of a natural ecosystem is related to the diversity of the ecosystem. This point has been debated in the scientific literature, for one can imagine ecosystems with a large number of species that are unstable. However, what seems important to achieve stability is that there be a high degree of functional redundancy in the ecosystem, at all levels of the food web. That is, there should be many species that perform the same biological role with respect to the rest of the food web [Adey, p. 281, van Voris et al 1980]. What this means for us is that to achieve a high degree of stability in our aquaria, we must establish a high level of diversity at all levels in the food web.

In nature, species diversity is largely proportional to area. In fact, diversity increases with the square root of the area. If you know characteristic things about a particular type of ecosystem, you can estimate the species diversity of the ecosystem if you know its size. Walter Adey compared the St. Croix reef and a microcosm model of it (the model was a few thousand gallons in size). He found that the expected number of stony coral species in the model was 1.3. [Adey 1991, p. 278]. Since few of us are content with a reef tank with only one species of coral, this puts what we're trying to do in perspective.

Why do larger areas support more diversity? There are several reasons. Larger areas have a larger variety of habitats. There is more area for species to carry out negative or positive interactions. Perhaps most interestingly, in any area, local populations of many species are always on the edge of extinction because of random or periodic unfavorable events. Large areas support more diversity because it's unlikely for extinction to occur in many local populations at the same time. [Adey 1991, p. 269].

In an aquarium that's going to please most of us, we need to do something to push diversity higher than it would be if it were a completely natural ecosystem. We can do several things. First, there is what we do already: human management. We place corals next to one other in ways that we believe will minimize aggression, and if we do see aggression that we don't want to see, we will move things around. Another example of this is when we add or remove "reef janitors" as a way of regulating the standing algae population.

A second technique to increase diversity is through the use of refugia. Refugia were introduced by Harris in the context of forest preservation [Harris 1984]. A refugium in our context is a body of water that is attached to the main ecosystem model, but is to some degree physically distant from it, in a way that minimizes negative species interactions such as overgrazing or overpredation. My own system includes a 20 gallon refugium as a tank connected to my main system, but with no fish in it. There is an abundance of amphipods in the refugium, whereas the amphipod population in the main tank is fairly low because of the two gobies in that tank.

The physical environment

I'm going to give an abbreviated discussion of the physical environment, because this is an area that is rather familiar to most hobbyists, and where my thoughts differ from the norm, they are rather controversial. Controversial subjects can take a whole meeting time on their own.

It's my contention that the physical environment that is best for coral reef ecosystem models is the same physical environment that exists where coral reefs do best in the wild. At first this seems rather obvious, but most people are reluctant to implement what this implies.

Seawater is a highly complex soup, but our lives are made simpler, for better or worse, by the fact that we either have access to natural seawater and we use it, or we don't, so we use one of the commercial salt mixes. I use Instant Ocean for reasons of cost and consistent good results, but I don't want to get religious about salt.

I believe that coral reef communities will do best in salinities that match natural salinities on coral reefs. At tropical temperatures, that corresponds to a specific gravity of 1.025-1.026. Salinity does vary slightly over small areas of a single reef. Reef animals are able to pump ions in and out of their cells as a way to maintain the ideal internal ionic balance. This process takes considerable energy. The animals have evolved in seas of 1.025-1.026 specific gravity. There is evidence that marine animals become severely stressed when held in water of lower salinity.

Average temperature varies throughout the region where coral reefs exist, but where coral diversity is highest, in the Indo-Pacific between the Solomon Islands and Indonesia, temperature rarely drops below 80 degrees (26.6 C), and surface temperatures above 90 degrees (32.2 C) are common. If you look at a map showing ocean surface temperatures and another map showing coral diversity, measured by number of genera present, the areas of increasing temperature correspond almost exactly with the areas of increasing diversity. We could spend another Sunday arguing why this is the case, but in the meantime, I prefer to run my tank at temperatures matching wild conditions that are good for corals, not wild conditions where corals are barely able to grow because of the cool temperatures.

The issue of mineral balance in our aquaria is important. Coral reefs are all about calcification, so it follows that replenishing calcium ions and alkalinity is an essential part of a functional coral reef model. I use an aragonite/CO2 reactor to maintain my tank, but other methods work well too. The important message is to learn how to do it, and do it! Certain trace minerals are important for the ecosystem, though learning which ones really make a difference and which ones are hype is difficult. The scientific literature is not rich on this particular question. My preference would be to maintain natural seawater levels of everything, but the lack of affordable test equipment makes this effectively impossible.

The reason we have to supplement calcium and alkalinity is as a way to make up for the lack of an adjacent ocean. Thus, my aragonite reactor is one way in which my model represents the vast ocean, at a high level of abstraction. Other contributions of the ocean are represented by other hardware in the system (for example, the skimmer).

Water motion over a reef sculpts the reef itself, through its effect on growing organisms. In my opinion, simulating adequate water motion is one of the two greatest challenges to marine aquarists. We could have a whole talk on water movement and how to do it well, but suffice it to say for now that I'm not happy with switched powerheads.

Lighting is another cricial aspect of the physical environment, and one that has recently been receiving the attention of expert aquarists such as Dana Riddle and Richard Harker. Adey cites evidence that shallow marine ecosystems increase their photosynthesis right up to the level of full tropical sunlight. Yet Harker shows how dim our metal halide lights are compared to natural sunlight. The substrate of a 90 gallon tank lit by a 400 watt bulb 60 centimeters above it would receive less than 20 percent of the light it would receive at the same depth on the reef [Harker 1997]. It's not necessary to have full tropical sunlight over an aquarium to support coral growth, but from the point of view of a whole-ecosystem model, this limited light is a deficiency in the model.

There's lots of anecdotal evidence of hobbyists' successfully keeping reef animals, corals included, in water that is either of weaker salinity or lower temperature than is typical on the richest natural reefs. In response to this, I offer the suggestion that there's a difference between being able to survive in those conditions for the lifetime of most of your tanks, and being able to grow for hundreds of years and reproduce and form a beautiful ecosystem. We just don't know how long we can keep coral reef communities in unnatural physical conditions. We do know what conditions stand the test of time in nature, so why not use those conditions at home?

I want to leave plenty of time for questions, so I'll stop here for tonight. If I do another of these talks, I'll go into more detail about community structure, the nature of food webs, and what that means for establishing aquaria.

This ends the prepared talk. The text below is from the Question and Answer Session.

Okay, I guess I'll talk some about water motion first. It's been shown countless times in thescientific literature that, other things being equal, photosynthesis is more or less proportional tothe degree of water movement. So creating adequate water movement is crucial.

Figuring out what is "adequate" is pretty hard. We can measure water flow in the ocean, andconclude that any natural water movement that corresponds with good coral reef growth wouldbe good water motion to have in a reef tank.

But is it really necessary to simulate natural water movement, in all its complexity? Perhaps asimpler water motion will have nearly the same effect on the community.

I don't have definitive answers to that question. I don't think anyone does. For that reason I thinkit's important to push the boundaries of aquarium water flow as we know it. But for those whojust want to get something functional, and don't want to take on a research project, what can wedo?

The section in The Reef Aquarium on water flow is actually pretty good in this regard. It coversthe major things that people do, and even mentions a lot of things that hobbyists don't do.

The most popular way to move water around in a reef tank is via switched powerheads, or a"wavemaker" as the industry insists on calling it. Dana Riddle did a nice two part article inAquarium Frontiers this year. It was on water movement. He showed that only the strongestpowerheads are capable of creating the level of water velocity that is present on a reef, and then,only right at the powerhead nozzle.

I'm talking about a high energy reef. Using a strong powerhead, it's difficult to make the currentvaried enough that it has a natural level of strength, but also doesn't strip corals from theirskeletons. There are other problems with powerheads, mostly obvious. I find them boring, so let'smove on.

Before we go on, I should mention that there are new products on the market. These aremotorized devices that rotate a powerhead, or accept a connection from a large system pump, androtate that. I think these devices would be a big improvement over stationary switchedpowerheads, though I've heard that the quality of construction in the existing implementations ofthese devices isn't great.

After powerheads, the next most common form of water mover is the self starting siphon, orCarlson Surge Device as it's often called. The bowl of your toilet is an example of a CSD.Thesecan create strong, directional flow, and can have a positive effect on a system. But they havetheir problems too. They're noisy and often messy, and loud. But the biggest problem in myopinion is that the output is stationary, like a stationary powerhead. A CSD is a lot like a big powerhead on a timer. Another method of water movement is the dump bucket. These areusually mostly rectangular containers on a pivot point. One side is slanted, and the pivot is placessuch that when the bucket is empty, it's stable in the upright position, but as the bucket fills, itbecomes unbalanced, and dumps into the tank. The introduction of water into the tank from adump bucket is more sudden than from a CSD, and because of this it produces a much morepleasing effect, in my opinion. The sudden dump of water excites real waves in the tank, andsoft critters like gorgonians will sway back and forth just like in the ocean. But dump bucketshave lots of problems and challenges too. Like the CSD, they introduce a lot of bubbles into the water, they're messy, and can be loud. Since dump buckets have moving parts (the bearings), there can be maintenance issues. I think this problem is overrated. The early dump buckets hadmaintenance problems, but some modern designs I've seen, using plastic and glass ball bearingunits, are completely reliable over 4 year periods and more.

Beyond those choices, you really get into the exotic and experimental devices, such as paddlesand pistons, vacuum devices and so on.

What would you consider to be the optimal water movement device?

Wind

* cap grins

At what temperature do you run your aquariums?

Which reef ecosystem in the world is the hardiest for the home aquarist to model?

Consider coral reefs at the edges of where coral grows. For example, Hawaii and southernFlorida.

I run my tank at a minimum of 81 degrees, and it gets as hot as it gets. I use a fan to try to keep itunder 88.

<DBW> The hardest would be the ocean vents, on the plate faults. Would be very interesting,very expensive, and almost impossible to do ;-(

What daily min and max do your temps run?

One could suppose that the corals that exist in Hawaii and Florida are the hardiest corals, sincethey are the only ones that can tolerate the substandard conditions.

What do you do for water movement?

Hawaii has the problem of extreme wave action in addition to the problem of being on theoutskirts of the tropics.

My temp runs 81-81 in the winter, and 81-86 or so in the summer. It'll go above 90 if I turn thecooling fan off.

So a chiller is a waste of money?

<Mark> Cap a lot of people place importance on pulsing the current, on the weekend what I sawwas extremely strong and consistent current with the tide changes. There was no change in thedirection of the water movement until the change in tide. The tide was also much stronger in onedirection than the other.

I wish I'd had time to collect actual data about the temperatures in real reefs. I believe 86 is nothot at all by the standards of the Great Barrier Reef and surrounding areas. It gets up to 100 there.

But isnt that just surface temp, what about the temp at the depths the corals are typically at?

I'm not going to say that a chiller is a waste of money, but it's not true that you have to maintainsome particular temperature, +- one degree, and everything's going to die if it gets above 80degrees.

So it's reasonable to suppose that corals at the outskirts of coral habitats are hardier than coralsthat can't live there. But in the case of corals that can take extreme water motion, that's not aproblem that we have in aquaria. In the case of corals that can take cooler temperatures, that's nota problem we have either, because heaters are cheap.

Then why so much talk about maintaining a constant temp. usually 77 - 79?

I guess my point for Magda is that I don't know of any reason not to try corals from the center ofcoral diversity, which is where our corals come from anyway.

The recommendations for a constant temperature of 77-79 (or lower) is probably largelyhistorical, but it makes sense when you consider that many aquaria, particularly aquaria thatpeople ran 10 years ago, had much poorer water quality than modern aquaria. All biochemicalprocesses happen faster at higher temperatures. Thus, if your water quality is screwed up in away that allows certain organisms (read algae or parasites) to gain an unnatural competitiveadvantage, then running the tank cooler *might* cause that to happen slower.

And doesn't dissolved oxygen increase with lower salinity perhaps another reason this wasrecommended for so long...at least at lower level than you were talking about as well as thatconstancy was probably better for poorer water quality also....

The approach I would take is not to model the hardiest reef ecosystem, but rather to model thehallmark reef ecosystem, and omit animals that are known to be extremely difficult, such ascertain corals, crinoids, etc.

Mindy has an excellent point. The saturation point for oxygen is higher at lower salinities andlower temperatures. In a tank with inadequate gas exchange, oxygen level can become a problemat natural temperatures and salinity. But why address one inadequacy (gas exchange) withanother (unnatural salinity or temperature). It would be better to fix the gas exchange problem.

Incidencally, good things happen faster at higher temperatures too (growth). Some people havenoted that lower temperatures don't stop RTN, they just slow it down. Nevertheless, the solutionto RTN isn't to slow it down, it's to avoid it in the first place. People are still learning how toavoid it. Craig Bingman said that the problem with his tank was inadequate water movement.When he increased the water movement, he stopped having RTN problems.

In principle, having a more diverse ecosystem model will reduce the opportunity for somethinglike RTN to have a devastating effect. This is true of anything, not just RTN.

Being new to the idea of what are the " animals to avoid" (due to their difficult nature) couldyou give some general recommendations considering I'll be stocking the reef w/ FloridaAquacultured rock...If this generalization is insufficient - how about a text that speaks to us folkswho are beginning to understand what some of the complexities of selection are?

Some people have suggested that the recent trend toward "SPS monoculture" tanks has coincidedwith the increase in RTN cases. Perhaps such a high density of SPS corals, mostly acroporidcorals, isn't so good.

Is there a formula for water movement gallons pumped per/h in a tank?

* cap Hates formulae like that.

Chuck, that question has a long answer. In your case, you're just going to get your rock and tryto get as much as possible to survive. I think a talk on how to cycle live rock (the nuts and boltsof how to start a reef tank, rather than all this theory) would be a very good thing.

<DBW> Personally I would pack as much in as you can, very hard to have too much if you aregoing for a high energy tank. But that all depends on what part of the reef you are trying to model. If it is a lagoon or something like that, then water flow is different to the reef creast.

I should mention, if it wasn't obvious, that the basic structure of my talk came from WalterAdey's book, Dynamic Aquaria. Adey has an axe to grind about the hobbyists, and he hasbasically no idea what the state of the hobby was when he wrote the book, but aside from thatand a few other things, it's the best single source of information I know of regarding aquaticecosystem modelling. It's really an excellent contribution.

What are you using for water movement in you tank? Also what part of the reef are you modelling to put this in context?

There are lots of things about my tank that I'm not happy about. I'm using switched powerheadsbecause I haven't implemented my Next Great Water Mover. I've experimented in the past with aCarlson Surge Device, and I'm now planning to make some dump buckets as an interim improvement. I'm not clearly modelling any particular part of the reef, but what I have in mindis a shallow backreef, I suppose.

<DBW> cap: just an observation from my tank. Having two pump outlets interacting create lotsof random type surges and eddies in a tank.

Are timers an alternative to use instead of a wavemaker?

DBW, yes, it's possible to create turbulent flow using only powerheads, such as you describe.

Yes. In fact I use timers. I have two banks of powerheads. One points one way and the otherpoints the other way, basically. They switch off once an hour.

References

Adey, Walter H., and Loveland, Karen. 1991. Dynamic Aquaria: building living ecosystems. Academic Press.

Harker, Richard. Lighting for the Reef Aquarium. Aquarium Frontiers, May/June 1997, pages 29-37.

Harris, L. 1984. The Fragmented Forest. Island Biogeography Theory and the Preservation of Biotic Diversity. Univ. of Chicago Press, Chicago.

van Voris, P., O'Neill, R., Emanuel, W., and Shugart, H. 1980. Functional complexity and ecosystem stability. Ecology. 61 : 1352-1360.