The Deep sand bed

NB. If you are a new comer to reefing. To get a better understanding of the basic principles at hand and the terms used in this section I would highly recommend that you read these two sections first. Basic Chemistry and  Filtration

As Discussed in the previous section on Filtration. DSB's / Plenums and Live Rock are probably the most commonly found method of filtration within the reef keeping world today. In this section however I'm going to go into a little more detail on the DSB. How and why this filter medium can work so well in a closed system, Its variations,  and also the do's and don'ts of their use.

Firstly the DSB or 'Deep Sand Bed' in comparison to other filtration methods such as Under Gravel Filters.

For many years the UGF or 'under gravel filter' has been the mainstay of the marine aquarium industry and not surprisingly it still has its die-hard followers even now. Without getting into any controversial arguments though, let me just say at this point, that I whole heartedly agree that in some tanks and under certain conditions the UGF in all its guises has and does work, So I'm not arguing as to its effectiveness at general filtration.. For reef keeping though, the UGF does have some drawbacks. Firstly in that 'by definition' a UGF requires a media of decent granular size to allow the easy passage of water through it to keep it healthy and biologically effective. This will usually consist of a grain size no smaller than about 2-3mm in size (typically crushed coral sand or gravel). One of the big drawbacks with this type of grain size is that each grain has an ideally large surface area for the colonisation of  problem algae, hence the deep green tinge common in the sand of UGF run systems. The other drawback to large granular size, is that it all too easily allows quite large particulate matter/ detritus to fall between the grains which over time builds up and starts clogging the bed unless the keeper regularly Hoovers the gravel/sand via the use of a gravel vacuum. Effectively the UGF is part biological filter - part mechanical filter.

If we were to leave the UGF for a period of time we would commonly see some changes occur if the system also comprised of LR. Firstly, via migration from the LR, we would see a gradual increase in the biodiversity within the surface layers of the sand bed in the form of various beasties such as Copepods etc and a general increase in the population of burrowing worms found moving in and around the gaps between the sand grains feeding on the various waste products produced by our fish. (this is commonly the case if we move a piece of LR in a UGF system. We will quite often find a multitude of critters under where the rock was). These critters are feeding on the detritus that has collected there, which we were unable to reach via the vacuum. 'Great' we might think. So if I just leave my UGF alone it will turn into a DSB all by itself. Well sorry but in this case that's a no no. Remember that large granular size I mentioned earlier?. Well this is where things start to fall down over time with the gradual build-up of trapped waste, the bed effectively starts to suffocate and turn bad with all that rotting waste which is in real danger of getting to the stage where Hydrogen Sulphide (H2S) is released. This in turn will kill off all those little critters that were doing such a wonderful job cleaning up, not to mention there is a very high probability of killing most of our livestock as well since this gas is extremely toxic.. Now the reason that this is unavoidable really comes down to critter strength. yep you heard right, they are just not strong enough to shift those massive grains around sufficiently to enable them to reach all the waste that has collected. There will be a multitude of areas where waste collects that just cannot be got at, which in turn fester and go bad, giving rise to the deteriorating conditions. 'Its a bit like taking two rooms and filling them up with people. 1 room full of children, and the other full of Sumo wrestlers. Now obviously due to their smaller size, you are going to find more people in  the room full of children than the room full of Sumo wrestlers, but at the same time it will be easier for you to push your way through the children due to their smaller size and mass in comparison to yours, than it would be to get through the room full of Sumo wrestlers who you just cant budge'.

Research has shown, that regardless of the type of media used, be it land or oceanic in origin. Many sand dwelling creatures simply cannot and will not, tolerate granular sizes larger or smaller than what they are designed or accustomed to, even down to a difference of 0.005mm in some species. Consequently they do not thrive or reproduce properly causing a drop in population or diversity  and decrease in the health of the bed in question. Another reason that the critters have trouble with larger grain sizes, is that over time the individual grains 'through lack of movement' become slightly glued together with the build up of a secretion called  'glycocalyx' which is produced by the bacteria present in the bed. This gluing effect can be extremely strong, so strong in fact that it completely prohibits water flow through certain areas causing the aforementioned dead zones. It used to be believed that this bonding was due to the precipitation of Calcium carbonate within the sand bed, although this is now believed in most cases to be incorrect. Now the other, 'and probably most important' drawback with the UGF, is that it only fulfils 'part' of our required filtration process. By this I mean the cycle we are all so used to hearing, the nitrogen cycle. Unfortunately the UGF is only effective down to the Nitrate 'or NO3' stage of the biological breakdown of waste. This is largely due to the fast water flow through the sand bed and its high oxygen content which restricts the conditions required for the growth of Anaerobic bacteria which are essential for the last stage of filtration with the breakdown of N03 to harmless nitrogen gas N2. And another other form, Dinitrogen or N2O

The DSB on the other hand utilises a very fine grain size (preferably below 0.2mm but above 0.050mm) to get round these problems. There are various discussions going on as to the ideal grain size, although the main consensus is that a very fine 'silt like' texture is better for what we, and all those lovely critters are looking for. Some even go as far as a mud like texture (ref the new trend of a material called 'Miracle Mud'). Having an extremely fine texture, allows the worms and other creatures present in the sand bed the freedom to move about unhindered to reach any waste that is trapped or collected.  It is very easy to see how much difference a slight change in grain size makes, in relation to  how these animals can move about in the sand layer by performing a simple experiment: 'Take two cups, and fill one with normal refined cane sugar. And fill the other with fine castor sugar. Now push your finger into each cup and see which one is harder'. You'll notice that the castor sugar offers allot less resistance to your finger being pushed into it than the coarser cane sugar. Effectively the smaller grains offer less friction than the larger ones which is why they move easily to the side as your finger is pushed in. The other factor is that the deeper the sand bed gets, the harder it becomes to move each grain because of the higher weight of the sand above. in practical terms this means that for any given depth, a fine sand bed will have less weight and friction on the lower layers than a course sand bed of the same depth. At the same time these creatures are burrowing through the sand, it causes the sand to move around or as we term it 'get turned over'. This action prevents dead areas forming, moves water 'slowly' through the sand bed via all the motion and down the masses of worm tunnels, and stops the individual grains from clogging or bonding together due to bacterial action. The other benefit is that this motion clears the way for further bacterial growth which improves the efficiency of the bed as a whole. Smaller grain size also gives a larger surface area for bacterial colonisation per inch cubed of media than the same volume comprising of larger grains. this is easily demonstrated via the use of some simple maths.

'For the sake of mathematical simplicity, lets just say that the grains we are talking about are all cubed shaped'.

If we take 1 of these cube shaped grains that has a surface area of say 10mm/sq per face. then this gives that grain a total surface area of  10mm/sq x 6 faces = 60mm/sq

Now if we take another grain of a quarter the size of the original one and work out the total surface area, we arrive at the following figure. 2.5mm/sq x 6 faces = 15mm/sq.

Remember though that we are talking volume as well, so within the same physical volume as the original 10mm/sq grain we can now fit a total of 64 smaller grains that's 4 x 4 x 4. add together the total surface area of all of these grains and we now have a total surface area of 960mm/sq. That's '16' times the total surface area of the original 10mm/sq grain but within the same total volume.

In practical terms this gives the smaller grained DSB a massive advantage over a standard UGF of the same size and volume, not to mention the DSB's greater depth as standard. In terms of total habitable surface area that can be utilised  by bacteria we are talking along the lines of 4,942sq/ft of surface area for a 12" x 12" x 4"deep sand bed using a grain size of 0.125mm. or 59,312sq/ft for a 72" x 24" sand bed of the same depth, in total area that is equal to just under 1 1/2 acres. And that's an awful lot of surface area for all those beneficial bacteria to colonise.

'Porosity can have a large part in this obviously. Another words, its just as possible to have one large very porous grain that has just as much surface area internally and externally as several smaller grains that are not as porous. However this still leaves us with the problem of physical size, and the resultant larger gaps between grains that can trap waste. These larger pores are also likely to clog over time leading to reduced water flow and diffusion through the grain itself'.

Once the bed gets beyond about 2" deep, we start to encounter increasingly anaerobic areas of extremely low oxygen content due to O2 being used up by by the aerobic bacteria above. It is this lower 'anoxic' layer that really differentiates the DSB from the UGF where water is forced through the bed via power heads. Water flow through these lower layers happens partly due to the burrowing action of the various sand dwellers but mostly due to 'diffusion'. This is a tendency that all fluids have, whereby two amounts of water containing differing concentrations of diluted chemicals will always balance themselves out over time, via the movement of molecules from the lower concentration to a higher concentration, and vice versa. This should not be confused with osmosis which involves a semi-permeable membrane which only allows the passage of 'certain' molecules  from one side of the concentration to the other. In the case of diffusion both water, and other molecules pass in both directions to balance out differences in concentration. The following diagram is a basic representation of this process.

In this diagram we have two quantities of any given fluid (in our case water). each quantity has the same total number of molecules but each sample also contains an odd number of molecules of a different type. in sample (A) we have 15 water molecules (H2O) and 10  molecules of another type (for the sake of argument say NO3). In sample (b) we have 19, H2O and only 6 NO3 molecules. Now if we could place these two samples together without any form of physical mixing we would witness 'diffusion' whereby 2 molecules of H2O would  pass from the right hand side to the left, and 2 molecules of NO3 would pass from the left to the right. The final sample would contain an even distribution of each type of molecule throughout as in the bottom sample (C). Now seeing as I'm no expert in chemistry or physics for that matter, I'll hand the description of the mechanics behind this bit over to the good Mr Hipkiss who has an alluring ability to describe the most complex chemical and physical happenings in a simple and easy to understand manner.

Andy quotes.

"Re diffusion, yeah, molecules have kinetic energy, call it heat if you like, or even petrol in your fuel tank.  Your bike keeps on going until the petrol runs out and you stop (at which point for the molecule it is at absolute zero, -273 Celsius or 0 Kelvin).  On our nice warm planet, there is always energy input (the sun), but the tendency is always to try and loose this energy, so in deep deep space, molecules are very very slow moving and almost stop.

In terms of distribution of molecules, (that is that they seem to move from an area of high concentration to an area of low concentration so the net effect is to be evenly spread), then this is just due to chance.  These molecules are bouncing around all over the place colliding with things in their way hence bouncing off in a different direction to before.  Given enough time (remember diffusion is a slow process) then even if all the molecules started off in one area, they will end up being evenly spread.

Andy then quotes this analogy.

Perhaps a reasonable example of this, is if you've ever seen the cricket coverage on television where they show the scatter of where the balls were hit.  If you look at one specific over at the start of the match then the scatter is minimal, depending on the batsman they might all have ended up in a small sector say all over mid-off.  But then look again at the scatter at the end of the innings, then there is a graph that almost looks like a full circle (plus or minus, certain shots that are physically impossible for the human body to make).

Same with molecules, given them enough time and they too will cover the whole pitch.

Similar example, think of a squash court but you are playing with one of those crazy balls that keep on bouncing.  Now rather than just the one ball you have 100, and as soon as you've hit one ball, you get another and launch that off at the back wall.  Now you are not 100% accurate, and you never hit exactly the same spot and indeed you never hit it from exactly the same place.  After all 100 balls are hit, someone looking down on your squash court would a) laugh at you  b) notice that the court seems totally full of these balls.

Many thanks to Andy Hipkiss for his input on this subject. Andy's site can be found at


The scenario above goes a long way to explaining how a DSB works in relation to the constant breakdown of nitrogenous matter all the way through to the final stage of nitrate to nitrogen gas. The diagram below shows what is happening in terms of water flow through the bed at a molecular level and how it corresponds to our water tests.

(Many thanks to Andy Hipkiss for the excellent photo's of his DSB.)

As can be seen, our tank water contains among other things like solid waste, a given amount of ammonia and phosphate etc that is released via the breakdown of faeces and food by all organisms within our system. Now some of this solid waste settles directly onto the surface of our DSB to be directly broken down. The rest is cycled via the bacteria and algae on and inside our live rock into the other various stages of nitrite, and nitrate. this cocktail then washes around in our tank until it passes over the DSB. at this stage we have a higher concentration of certain contaminants in the water column than we do in the bed itself, so by diffusion some of this contamination is drawn down through the bed in an effort to balance out the concentrations further down the bed. This however, will never really happen in the case of NO3 though as the anaerobic areas in the lower regions of the bed contain bacteria that convert nitrate into free nitrogen gas. We therefore end up (in the case of a seasoned and active DSB) with a constant draw through the bed, where waste is broken down at the surface layers either through direct bacterial action or the assimilation of the various critters and algae in the sand which in turn release their waste. The first to go in this highly oxygenated upper layer is the accumulated Ammonia, and nitrite which is converted into nitrate. the concentration of nitrate in the lower layers at this point is being depleted due to its conversion to nitrogen gas by the anaerobic bacteria. This draws the higher concentration of nitrate down through the upper layers via diffusion till it reaches the lower layers and is converted as well. The free nitrogen gas (visible as bubbles in the sand) travels back up through the bed to diffuse into the water column in an effort to balance things out above, but this is constantly drawn up by the fact that it is escaping out through the surface of our tank and back into the atmosphere. In the end we have a constant cycle going on whereby each part in the nitrogen cycle is being pulled to the next stage by diffusion and an ultimate safe conclusion for our captive animals.

In photo 1, we can see nitrogen pockets forming and rising through the layers in this play pit sand bed of MadSi  ( note that when this photo was taken this bed was fairly immature, hence the lack of worm trail ). However the later shot 2, clearly shows the beginnings of  worm activity and the dispersion of produced nitrogen gas.  Image 3 shows the top 3" of my established and fully functional Aragamax (aragonite) based DSB with a small sized (5mm) rubble layer on top of the 0.1mm fine sand below.  Image 4 shows a close up view of the top 1" with concentrated worm activity. I am of the opinion though that Play pit sand is 'not' an ideal substrate for the construction of DSB's due to its sharp and hard nature.

 1                                                                                                    2



3                                                                                             4



The increase in worm activity helps disperse these larger gas pockets as the sand grains are kept looser. Hence the general lack of 'larger' bubbles in a well populated and established DSB although you will still see occasional ones.


Its this 'total' biological filtration cycle that has other methods beaten in the case of DSB's and plenums (ill get to these later). All other forms effectively stop at the Nitrate stage due to the inherent high oxygen levels and lack of anaerobic areas, especially in the case of fluidised bed filters and trickle filters which are highly effective at this type of breakdown but go no further hence the common need for regular water changes in these types of system to reduce the amount of built up NO3. I like to term this scenario as the 'bottleneck syndrome' whereby the total numbers of nitrifying bacteria are so prevalent in comparison to the sparse populations of anaerobic bacteria present, that the filtration system effectively ends up with a bottleneck where ammonia and nitrite are processed much quicker than nitrate is turned to free nitrogen gas hence the constant and ever increasing pool of accumulated nitrate. The DSB and plenums ability to offer biological filtration in a balanced manner means it is possible under certain circumstances to run these systems without any form of water change for long periods, only being interrupted due to the reduction in available trace elements that the animals have assimilated and need to have replenished in order to properly thrive. Or to reduce the build-up of other unwanted elements or compounds that this type of filtration doesn't. In fact it is quite 'common' to achieve levels of NO3, that are so low in comparison to other filtration methods that even professional test kits have trouble obtaining a reading. This in turn makes conditions extremely favourable for sensitive creatures such as corals etc. One of the largest benefits to the DSB is the large amounts of released planktonic larvae, which stem from the reproduction of the sand bed creatures. This constant influx of live food has a very beneficial side effect when keeping high energy requiring corals such as SPS, Sun corals of the Tubastrea family or soft corals such as Dendronepthalia's. As such, it is essential that any sump based DSB is placed 'AFTER' the skimmer and not before. Otherwise a large proportion of this beneficial food will be skimmed off ..........'what a waste of free food'..

The Keys to success.


Although the actual practice of building and maintaining a DSB couldn't be simpler. In order for the DSB to function properly it requires some attention to certain rules though. which if ignored can 'and do' spell doom and gloom in the long run.

1. A substrate of fine 'blunt and soft' granular nature should be used ( preferably a silt like texture below 0.2mm but above 0.05mm). If the texture is  superfine (i.e. below 0.05mm) then it will pack down, restricting water flow, diffusion, and critter movement to the degree that filtration effectively stops and the bed starts to rot. If the correct particle size and texture is maintained in which the organisms present can move freely through, the substrate will be kept loose and turned over which will prevent these dead patches. This substrate does not necessarily have to be of oceanic origin, however it should not contain any contaminants in the form of high metal concentrations etc. An aragonite based substrate is preferable over others, as this is more natural and buffers the pH drop at lower levels within the bed as 02 concentrations reduce, however any medium will do as long as it is chemically inert and of the correct granular size / texture. some heavier material may need to be spread lightly across the surface layer to reduce the possibility of the substrate being washed away in the case of high flow areas. although this layer should only be very thin and in the form of LR rubble of small size i.e. 5-10mm diameter. This will also add to the diversity on the bed, as not all sand dwellers are burrowers and as such, need something to hide around and under.

2. In relation to DSB's,the depth of the substrate should be in excess of 3" to effectively induce the anaerobic conditions essential for the breakdown of NO3 to free nitrogen gas to a really beneficial level. Various depths have been tried from 2" - 16"+ in some commercial applications, however no real benefits have been shown in 'all' cases as to whether more or less is any definite advantage. the current trend is commonly between 4 - 6" in depth for aquarium applications. Indeed, having 'too' much depth in relation to the wrong type of grain size can and will lead to compaction due to the weight of the sand layers above. Although some people use split beds of differing grain size, it is not essential (in fact its pretty pointless, as a well populated bed will see this all mixed together over time any way). as long as both areas fall within the boundaries of what our sand dwellers will accept as liveable conditions. It is just as easy if not preferable to go with a single grain size throughout the entire depth. My own personal preference is Aragamax for the main bulk of the bed, topped off with a fine layer of Aragamax grand Bahama as a rubble layer as this type includes a portion of crushed shell fragments. In relation to SSB's or (shallow sand beds) a 2" layer has also been seen to give good all round results if used in shallower tanks where the loss of display space effectively rules out the use of a full DSB with a 6"+ depth. SSB's work in exactly the same fashion as DSB's and the same rules should be applied, the only drawback is a more limited degree of nitrate reduction due to the reduced anaerobic layer and population. In many cases a combined method is used, whereby the main tank contains a 2" SSB whilst the main DSB is moved to the sump or a remote tank.

3. Critter population and diversity is the major factor in whether a DSB does its job effectively, and for any length of time. Ideally the bed should be thriving with a multitude of worms, copepods, flatworms, mini stars, round worms, loads of types of bristle worms, snails, brittle stars, protozoon's, lots of small crustaceans, and small sea cucumbers etc etc, overall well in excess of 200 species may be present in a well fed and diverse DSB. These critters all serve slightly different rolls, as sand movers, surface scavengers, or both in the case of bristle worms and stars. The higher this population, the greater the sand bed's ability to deal with waste, which in turn improves water quality throughout the whole system. Regular influxes of fresh diversity are also recommended, due to the fact that one form will gradually become dominant over another if left to its own devices in a closed system. This diversity can easily be obtained by the introduction a good dose of crud from the bottom of your local LFS live rock holding tanks on a quarterly basis or by buying one of the critter kits that are now appearing on the market. Creatures that effectively and aggressively predate on the sand bed life forms should be avoided at all costs, such as sand sifting fish like blue cheek Gobies, large hermits, and predatory sand sifting stars. These will all decimate a worm population very quickly. I am of the opinion though, that a small amount of predation can be of benefit in the aim of boosting reproduction in this population and to lessen the possibility that one species of sand dweller will overpopulate to the detriment of others, one or two small hermits can be a bonus in this department within a decent size DSB. Passive sand sifters are also welcome such as certain cucumbers like the tiger tail  and black cucumbers. These 'passive' sifters are primarily just taking in sand for the filtering out of bacterial life as well as detritus and algae's etc. This cleaning action of the upper layer also promotes the re-growth of new bacteria essential to the biological processes within the bed as well as keeping the top layer loose in order to increase chemical diffusion capacity. It should be remembered though that excessive use of these larger sifters can steal food so quickly that they hinder the population of sand dwellers by starvation alone. IMO I would not try keeping more than 1 medium sized Cucumber per 4-5sq/ft of surface area on a DSB, in an effort to ensure both cucumber and sand dwellers get an equal share of the goodies.

Below are just a few of the commonly found beasties that you might encounter with the aid of a magnifying glass. Many of these creatures will commonly only show themselves in the late hours between lights out and lights on, however they will usually loose this shyness if food is placed directly onto the bed during lit periods, at which point all manner of creatures may be witnessed feeding.


Usually found crawling in and out of the rubble on top of the sand bed. these exelent scavengers will make very short work of uneaten food and detritus


This is another type. you may find numerous different species within a single bed or around the live rock.some can get quite large.


Although people commonly panic about these creatures, there are infact many species that go unnoticed in our reefs, many of which are just scavengers and do a good job without getting out of hand.


It is quite usual for these to be pure white in colour to match the sand they live in, however there are numerous different species from microscopic 2mm specimins. Right up to the juvinile stages of larger verieties.


It is quite common to see these little chaps scurrying about all over the place. feeding on detritus etc, the young 'nuplii' form a good food supply for your corals at hatching time.


Just like the Amphipods. these little critters will also breed to good proportions. And fullfill a similar niche in the breakdown of solid waste material.


Althogh more a LR inhabitant. These small fast moving snails will commonly be found scowering any larger sand rubble for algal films on which they feed.

More Photos will be added as they become available. If you have any to donate then please feel free to forward them to me via E mail

4. To maintain this high level of diversity it is essential to feed these animals in an effort to increase population and keep things running smoothly. Running untreated / raw water from the overflows, is the most common method as this also carries any uneaten food to the bed as well as fish waste etc. It is commonly a surprise as to how much food and waste a DSB can deal with, which can be seen easily by dropping a couple of cubes of fine textured fish food onto the sand surface and then observing how many worms etc come poking their heads out of the sand within only a couple of minutes, making a beeline for the food source. Regular feedings of this type are commonly used in the early days in an effort to boost populations quickly. This is a simple exercise whereby you place a small amount of food under a clam shell or small flat stone each day and gradually move it round the bed each time in an effort to improve distribution. Once well established with an extensive population, somewhat heavy feeding of the tank is quite possible without any adverse side effects simply because there is always something or someone who wants what's left over. just about any food will do as long as its fine enough for the fauna to consume, however larger pieces will be quite eagerly set upon by numerous bristle worms etc all at the same time. It should be noted that the term 'heavy' feeding is a very open one. Beds take time to populate and the critter population needs to adapt to the available food supply if it is to keep up with any increases. As such, care should be taken when increasing feeding levels to the main tank. at all times let your test kits dictate your actions. The term 'heavy feeding' effectively means a system that is fed to the maximum of its nutrient cycling abilities whilst not building up a pool of surplus No3 or Po4.

5. Flow rate over the DSB. There is a lot of debate as to whether beds run better with high or lower flow. But as always in this hobby, there is a compromise to be met. Whilst higher flow will increase the transfer of 'dissolved' organic nutrients through a bed, it also has a drawback in that it limits the settlement of solid particulates or food. Its this non-dissolved matter that is essential for the larger organisms that are responsible for keeping the beds structure loose and prevents it from binding over time. If you don't let 'some' solid matter settle out, then diversity will fall over time due to starvation and the bed will effectively bind together through lack of movement increasingly lowering its diffusion capability until it collapses. All critter based sand beds need both solid food 'and' dissolved organics to function effectively over the long term.. in the case of sump based DSB's The main trick  is to look at the bed and see whether some of the water born matter is settling out. The rest should be removed by the skimmer. On the subject of suspended waste. Some have the miss-assumption that detritus is bad for a sand bed. In fact it isn't in the case of true DSB's. Organic matter that has been broken down to its final non biologically active state is nothing more than silt. and silt actually 'adds' to the fine consistency of a sand bed, so have no fear. As long as your critter population is maintained to the degree its movement around and through the bed keeps the structure loose and free to diffuse effectively, the non biologically active remnants of organic matter that makes its way into the bed over time will not hinder the beds performance. Too slow a flow rate on the other hand will limit both the transfer of nutrients through the bed, and will possibly allow  too much solid matter to fall out of suspension to the degree that the bed effectively suffocates through too much settlement, mixed with lack of dissolved 02. The same rule again applies. compromise. You want a flow rate that is fast enough that it keeps water moving nicely across the bed, whilst at the same time not so fast that there is no settlement taking place.

Experiments by Al. Huettel, W. Ziebis, and S. Forster 1996 on flow induced particulate matter uptake in sediments, are frequently quoted as an argument for increasing the DSB's efficiency by way of increased flow rate across its surface. What is commonly missed in these references, is that this study specifically relates to the infusion of very fine particulates into the substrate on the leeward and pressure side of undulating substrate surfaces, and that those tests were conducted under laboratory conditions without any reference to linking critter diversity or its 'need' for solid particulate waste or the merits of such a population on a reef system as a whole. In fact there is some merit to the argument that you might not 'want' particulates 'forced' into the substrate by flow induced pressure to break down at a bacterial level only in a closed system. Solids that settle out on the surface are prime pickings for the fauna that scavenges across and through this area encouraging that life to move upwards and downwards which helps break up the surface layer on a constant basis . The whole purpose of which is to 'assimilate' part of that free nutrient mass into living body tissue and for it to be re-cycled back into the system by way of reproduction as a viable food source for our corals etc in the form of plankton or larvae before it gets a chance to break down.. The act of burying waste to break down bacterially 'before' it has had the chance to be consumed by the critter population effectively forces the bed to work harder at a biological level rather than having its input distributed via alternate pathways of nutrient cycling such as assimilation into body mass by larger fauna. Whilst this study does offer a valuable insight into the hydrodynamic function of substrates and show the need for 'some' flow in the interests of inducing chemical diffusion above and beyond passive transfer, it should be kept in context as a 'study' of particulate movement into and out of substrates relevant to bacterial breakdown. It isn't a guide to building and utilising a DSB within a closed system however, and does not factor in the importance or impact of critter population or diversity on the DSB or the hazards inherent with increasing flow rates across the surface to the degree that water born particulates don't get a 'chance' to settle out to feed that population.  These guys were interested in what was happening at the substrate / water interface, i.e. that section of water 'just' above the substrate. In a DSB however, we are interested in the 'entire' water volume above the substrate. and the only way to get waste that is 2 or more inches above the substrate to make contact with it before it passes strait past and back to the rest of the system, is to have a flow rate that allows suspended particulates to 'fall' as they travel, and this is where aquariums differ from the wild or a lab. All in all, cross referencing studies such as these is worthwhile admittedly, but they shouldn't be considered the be-all and end all of what will or wont go on in a reef tank. the variables are just too vast when put up against the differences between one system and another compared to the precise and controlled conditions of a laboratory experiment. Possibly more appropriate for this instance would be to tie this study to 'Bucket' DSB's instead, which function under different parameters (see notes at the end of this section on the differences between bucket DSB's and critter based DSB's)

6. Algae or no algae, Light or No light.  Here again we see no definite right or wrong way dependent on the amount of space available to you. It is my preference though, to run DSB's without algae present, and in the dark. There are two main reasons for this. 1. Sand dwelling critters are more active during dark periods so a permanently dark bed will have a higher degree of constant activity than a lit bed. 2. rooting algae such as Caulerpa will over time bind the beds structure together limiting the ability of the critter population to move freely throughout the bed. Non rooting algae's such as Cheatomorphyllia whilst not directly damaging the bed will slow the flow rate across the substrate and will trap water born waste within their structure . In severe cases the algae effectively filters and collects the waste from the water to a degree it  limits settlement to the bed and starts to starve the underlying critter population. It is commonly noted in such cases that the critter population actually moves from the substrate up into the algae mass above in search for food leaving the bed under populated thereby reducing the stirring capacity that limits clumping of the substrate over the long term. My personal preference is to run these in two separate areas. A dark area for the DSB followed by a lit area with algae present to take advantage of any remaining dissolved organics like No3, after settlement of the solid matter in the DSB area.

Building the DSB.

Well surprisingly, this bit couldn't be simpler. It largely depends on whether you are upgrading an existing shallow sand bed or starting from fresh. If you are upgrading then it is advisable to only increase the bed depth by about 1" per month to enable the existing creatures time to colonise this new layer before the next is added. going strait from say 1" up to 4-5" will trap creatures in the lower layer and deprive them of oxygen before they have had a chance to dig there way back to the upper layers, this will also leave a layer of detritus and waste between the two layers which would otherwise have never made it that deep under normal circumstances. If this layer was allowed to decompose without the breaking up action of the creatures that are now in the upper layers it may cause problems later with reduced diffusion and blockages. Building a sand bed from scratch is probably the simplest exercise you will ever undertake in reefkeeping. What could be simpler than just gently dumping the sand in around your existing rock, up to the depth you desire. and then letting mother nature take her course with the help of some crud from your LFS tanks. The rest of the diversity will naturally migrate from your live rock. It is worth gently raking over the surface of the bed during the break-in period to prevent areas of compaction or clumping due to the lack of critter movement. Don't worry too much if the initial addition of new substrate causes a bit of cloudiness when added. Light mechanical filtration through floss will remove this dust from the water, and the corals themselves are well adapted to dealing with the occasional dust storm in the wild. However, if you are concerned that the settling dust may be smothering them, you can always go over them with a gentle power head to clean them off.

One important aspect to remember before dumping all this sand into your system is 'chemical imbalance'. As such, any substrate that comes in a bag or whatever, will have a differing surface composition to the water that is in your tank. All surfaces in our tanks over time develop a coating of bacteria and various precipitated chemicals that effectively seal them and attune them to the prevailing conditions so that they match their surroundings, This can commonly be seen when adding a piece of bare rock into the tank, which will usually develop various different coatings prior to any calcification or higher life forms taking root. The fact that this alien surface needs to develop this coating also goes the other way in that any substance that is on this surface will inevitably react with our water and dissolve into the water column. To this end it is essential to pre-soak our sand in some old tank water for a short period (IMO I would say 2 weeks to be sure whilst stirring every day) to minimise the possibility of affecting our water quality once added by either introducing adverse chemicals or by drawing large amounts of chemicals out of our water through precipitation which may effect the other inhabitants leaving them deprived of one form of element or another. Commonly Calcium (Ca), Alkalinity and magnesium (Mg) will all take a bit of a hit where new sand is concerned. By soaking first, this will be avoided.


There have been topical debates raging over whether DSB's are viable for long-term use within the marine aquarium, without going bad, for years, plus other conversations concerning the build up of toxic chemicals in the lower layers which can be released at any point causing massive damage to the tank inhabitants. After speaking to various individuals, and reading various posts scattered about on the internets reefing bulletin boards. These are IMO largely unfounded. My reasoning is as follows.

1. If the bed is functioning as it should do, with a diverse and plentiful population of sand dwellers then the possibility of the bed fouling even over long periods is very remote if not impossible. The mere fact that a well populated bed can turn its own volume of sand over every day without any trouble, negates this possibility. A healthy population will expand with food availability and decrease with lack of food. As such the two scenarios are very closely tied. Keep the bed fed with a regular influx of fresh diversity and the possibility of fouling is virtually impossible.

2. The build up of toxic chemicals is another hotly debated subject especially in light of recent revelations from the  Reefing DSB's founding father Dr. Ron Shimeck. 'Dr. Ron' as he is commonly known, has made recent reference to the 'possible' build up of heavy metals in the bed 'and other areas of our aquariums', via precipitation. And that any drastic drop in parameters such as Ph 'might' cause the release of these back into the water column with consequent damaging effects to our livestock. This is, 'I might add' an experiment still in progress which I eagerly await the outcome of, as the results might change the way we all view and approach the long-term husbandry of our charges. As for the conclusions, Dr Ron has already suggested that replacement of the media might be necessary every four to five years in an effort to rid the system of this build up, plus it may be stemmed from happening for longer periods via the use of 'Polyfilters' which are specially designed to deal with this type of impurity, ideally when preparing fresh salt mixes up prior to addition to the system. In this respect there 'might' end up being some argument in keeping DSB's in a detachable sump or one not unlike my design, ( DIY 4) So that when this type of maintenance is carried out, it will not damage the rest of the system with a sudden release of un-treated Amm, NO2, and NO3 that is being worked on in the sand layers at that time. If this truly is the case, then my only suggestion rather than starting from scratch each time, (which none of us can realistically afford), Would be to house the DSB in a separate sump which can be shut off from the rest of the system, and then replace  1/4 of the DSB every year to keep things at acceptable levels. Any pollution that is emitted from the bed will be kept away from the main tank, and can be simply changed by flushing the DSB through with some tank water prior to reattaching to the rest of the system. As for any other chemicals / gases that are part of the breakdown process. then I personally find it hard to believe that they could possibly accumulate to dangerous levels in a well maintained and well fed DSB. the mere fact that all the sand is being moved around so much, actually increases the possibly that any accumulated harmful chemicals are dispersed via diffusion before they get a chance to do any damage, to be removed via other methods such as skimming, Carbon or Polyfilters.

3. The difference between normal DSB's and BUCKET DSB's...Do 'NOT' confuse the two, they are very different beasts. A bucket DSB is a 'chemical' filter that works solely on the principle of breaking down dissolved nutrients such as Ammonia, Nitrite and Nitrate to free nitrogen gas and extremely fine organic matter that is too small to be mechanically filtered from the incoming water column. It does not rely on a critter population to handle solid matter and keep the bed loose. In the case of bucket DSB's, the bed is run remotely from the main sump system in a separate container or tank and is fed only pre-filtered water that has had the vast majority of solid particulates removed. Usually by way of a 5micron filter sock or other such fine filter material on the input side.  The main aim here is to let the bed function at a bacterial level, allowing it to work chemically without the hindrance of dealing with solid matter breakdown. The filtering effectively stops particulates entering the area which would otherwise clog the bed over time without the presence of a critter population to deal with it. It has commonly been the case that people confuse these two methods and start applying the rules of one to the other. The simple fact is that you cant have your cake and eat it all the time. Suddenly cutting off the supply of solid food to an existing critter based DSB will probably herald disastrous results as the critter population starves and dies back  leaving the bed clogged with dead or decaying matter and no way of dealing with it. whereas a Bucket DSB based system relies on diverting solid matter handling to alternate routes, such as the filter on the input, and more reliance on heavy skimming to take up the surplus. Neither method is wrong, nor right, They simply work in different ways and should be treated as such.


In closing, I would say that for anybody looking for probably the most natural and healthy way to filter a reef-tank, whilst at the same time supplying their corals with the most natural food source available, and having a fun time at night with a torch watching all the different critters that pop up to munch on that days waste, then the DSB has just about any other method of filtration beaten, hands down in my experience.

For an in-depth look at a newly developing DSB    Click Here.