In this final article of my three part series on oxygen dynamics in reef aquaria, I will:

• Present data gained by measuring the function and long-term results of established aquaria of various sizes and configurations.

• Show the short-term effects on those same systems of various devices or manipulations.

• Include data on the oxygen dynamics of shipping bags and numerous other systems measured at different locations and times.

• Discuss the various components that are the principle players in oxygen dynamics and their relative importance in terms of the "average" reef aquarium.

Together with the previous installments of the series, I have begun to understand some of the average oxygen values in various containers of seawater and the dynamics of the oxygen state in closed systems. My goal is that individual aquarists can use this data set to analyze their own systems by comparison and make any appropriate or needed alterations to ensure the health of their captive systems, or their shipping practices, through maintaining adequate water oxygen levels.

Methods

The methods for this experiment are similar to those presented in my previous article on the subject (Borneman 2005a,b). When monitoring reef aquariums, the oxygen probe was positioned in the same place for each measurement, with the requirement that water flowing past the probe met the minimum level required to obtain accurate readings. Prior to removing the probe, another reading was taken by manually moving the probe back and forth in the water to ensure that readings were consistent.

Tanks Utilized

Tank 1: "Clownfish Tank"

Size: standard 10-gallon glass aquarium
Water flow: two Aquaclear 201™ powerheads (Rolf C. Hagen, Inc.)
Lighting: two 65-watt power compacts (one blue, one white)
Substrates: 6 cm aragonite sand, approximately 15 pounds of live rock
Major animals present: five Trochus snails, ten Nassarius snails, two juvenile Amphiprion percula, one Entacmaea quadricolor
Corals: Isaurus sp., Protopalythoa sp., Psammacora sp., Tubastraea sp., Zoanthus sp., Pavona sp., Pocillopora sp., Acropora sp.
Filtration: none (test 1); CPR BakPak™ skimmer (CPR Aquatic, Inc.) (test 2)

Tank 2: "Puffer Tank"

Size: standard 75-gallon glass aquarium
Water flow: two Maxi-Jet 1200™ powerheads (Aquarium Systems, Inc.)
Lighting: two 55-watt power compacts (one blue, one white) (CurrentUSA)
Substrates: 8 cm aragonite sand, approximately 80 pounds of live rock
Major animals present: one Diodon holocanthus, one Salarias fasciatus, one Ctenochaetus strigosus, one Entacmaea quadricolor
Corals: Psuedopterogorgia sp., Scolymia sp., Protopalythoa sp., Zoanthus sp., Anthelia sp., Isaurus sp., Capnella sp.
Filtration: CPR BakPak™ skimmer (CPR Aquatic, Inc.)(test 1); ATS unit (Inland Aquatics) and Remora HOB™ skimmer (Aqua C, Inc.)(test 2)

Tank 3: "Main Reef Tank"

Size: multi-tank system, approximately 600 gallons total
Water flow: two 6060 and two 6000 Turbelle™ Stream pumps (Tunze USA), two
Rio Seio M1-1500™ pumps (TAAM, Inc.), 40-gallon surge tank fed by Ampmaster 2100 (Dolphin Aquarium and Pet Products, Inc.)
Lighting: day cycle: one 1000-watt metal halide, two 400-watt metal halides (Sunlight Supply, Inc.); night cycle: one 250-watt metal halide (Sunlight Supply, Inc.), two 65-watt power compacts (white) (Lights of America, Inc.), three 1-watt LED's (blue)
Substrates: 5-15 cm aragonite sand, approximately 300-400 pounds live rock, live rock rubble
Major animals present: 500 snails (Turbo sp., Trochus sp.,), 26 fish, three Entacmaea quadricolor, Aiptasia spp., two tridacnid clams
Corals: Over 80 species and several hundred colonies
Filtration: MR-2 Beckett injection skimmer (My Reef Creations), activated carbon filter (Ocean Clear 320), ozone generator (OZX-B300T, Enaly Trade Co., Ltd.)

Tank 3: "Culture Tanks"

Size: multi-tank system, approximately 500 gallons total
Water flow: divided flow from Ampmaster 2600™ (Dolphin Aquarium and Pet Products, Inc.) with eductors (KTH Sales, Inc.), two 6060 Turbelle™ Stream pumps (Tunze USA), two Rio Seio M11500™ pumps (TAAM, Inc.), passive flow
Lighting: differing for each tank: four 65-watt T5 fluorescents, four 65-watt power compacts (white and blue; white), three 175-watt metal halides, two 400-watt metal halides, one 1000-watt metal halide
Substrates: remotely located live rock (250 pounds), live rock rubble, some tanks have fine grained aragonite sand
Major animals present: 900 snails (Trochus sp., Astrea sp., Nassarius sp., Cerith sp.); one Zebrasoma flavescens, one Ctenochaetus strigosus)
Corals: 30 species, several hundred colonies
Filtration: MR-2 Beckett injection skimmer (My Reef Creations), passive activated carbon filter, ozone generator (OZX-B300T, Enaly Trade Co., Ltd.)

Other Measurements

Several shipping bags also were used to determine oxygen levels present in bag water. These included standard plastic coral shipping bags and "breathable" bags (Evert-Fresh Corp.). Sterile seawater, untreated tank water and tank water containing a coral were measured by packing them with an atmospheric air "cap." The bags were placed inside a thin plastic dish to support them in an upright position while measurements were taken. The oxygen probe was placed inside the bag along with a Teflon stir bar and the bag was tightly sealed using parafilm and tape to prevent air exchange. The entire dish, bag and probe were then placed onto a magnetic stir plate to create enough water motion to record oxygen levels.

A variety of other point-readings and measurements were taken on other tanks. These are presented as indications of potential variation between tanks and to provide information about the specific effects of organisms or conditions on oxygen levels in seawater containers. Most of these readings were taken from display tanks at retail stores, display tanks at the IMAC 2005 conference in Chicago and from shipping bags. Other readings were recorded at a local coral wholesale facility, a ten-gallon tank with aged seawater and 27g (a large handful) of Chaetomorpha sp. algae, and a fifteen-gallon tank containing only a piece of live rock.

Results

Tank 1 (Clownfish Tank)

Figure 1 shows the oxygen dynamics of Tank 1 in operation without a skimmer. Figure 2 shows the oxygen dynamics of Tank 2 with a skimmer in operation.

Tank 2 (Puffer Tank)

Figure 3 shows the oxygen dynamics of Tank 2 in operation without a reverse daylight algal turf scrubber. Figure 4 shows the oxygen dynamics of Tank 2 with an algal turf scrubber in operation.

Tanks 1 and 2: Airstone Use at Night

Because of the pronounced increase in oxygen caused by using an airstone in a hypoxic 10-gallon coral-only tank (Borneman 2005b), I decided to determine what effect the addition of an airstone powered by a Rena 400™ air pump (Aquarium Pharmaceuticals, Inc.) would have on Tanks 1 and 2 at night. Figure 5 displays the results of that test. In Tank 1, the airstone was used for only 30 minutes. It was then moved to Tank 2 where it remained all night until 0800 the following morning.

Tank 3 (Main Reef Tank)

Figures 6 and 7 show the oxygen dynamics of Tank 3 in operation on two different days, several months apart.

Tank 4 (Culture Tank)

Figures 8 and 9 show the oxygen dynamics of Tank 4 in operation on two different days, several months apart.

Other Measurements

I filled a narrow 15-gallon tank with sterile seawater and sealed its top. After two days, I added a single piece of live rock (approximately 5 pounds) to the tank; the top was again sealed, and the tank was left unstirred and isolated from light using black paper for three days. After that time, I opened the top of the tank, waited several hours and then turned on a powerhead (Maxijet 1200™, Aquarium Systems Inc.), waited several more hours and then placed a single 18" 15-watt fluorescent bulb above the tank. Several hours later, I removed the rock, and turned off the light and powerhead. The water became cloudy after about four hours, and I continued to monitor oxygen levels over the next day. The results are shown in Figure 10. Each of the respective periods with a scaled x-axis (time) is shown in Figures 11-13.

Chaetomorpha

A large handful of Chaetomorpha sp. algae was shaken dry and weighed at 27g. The algae clump was placed into a ten-gallon tank illuminated with a white 65-watt power compact bulb (Lights of America, Inc.). The tank also had a Maxijet 1200™ powerhead (Aquarium Systems, Inc.) which was turned on at the beginning of the test, and off after several hours. The results are shown in Figure 14.

Shipping Bags

Readings for breathable and standard plastic bags are shown in Figures 15 -17. In Figure 17, a medium sized Porites cylindrica colony was placed into one liter of tank water from Tank 3 and the bag was sealed around the probe. Point rates for several bagged livestock orders were tested, as well, and are shown in Table 1.

Effect of Stirring and an Airstone on Normoxic Seawater

To test the effects of stirring and an airstone on seawater at 35psu, I poured one liter of freshly made seawater into a beaker and placed it on a stirplate with a large magnetic stir bar. I turned the device so that a deep vortex was created that nearly reached the bottom of the beaker. I then turned off the stirplate and allowed the beaker to stand for 30 minutes. I then placed a ceramic airstone powered by a large Rena 400 air pump (Aquarium Pharmaceuticals, Inc.) into the beaker and allowed heavy aeration of the water for 45 minutes. The results are shown in Figure 18.

Other Tank Oxygen Levels

I measured oxygen levels at single point intervals in a number of systems including retailers, IMAC display tanks, a coral farm and Tank 3 one year ago. The levels are shown in Table 2.

Discussion

Overview

The results of this work show that the oxygen dynamics of reef tanks, in general, follow patterns similar to those found on coral reefs (Figure 19).

As on reefs, photosynthesis brings a rapid increase in oxygen levels within a few hours of "sunrise," whether the sun or aquarium lighting is used. The similarities between reefs and tanks are even more pronounced considering that coral reefs can become hypoxic at night (Borneman 2005a). The variation between reefs, and between different reef tanks, is also similar, as the results of my research presented in this article show.

Many factors can affect the oxygen levels of reef aquaria. Obviously, salinity and temperature are primary components of seawater's physical capacity to hold oxygen. The relative rates of respiration and primary production by plants and animals vary considerably between tanks. The total biomass and metabolic rates of organisms greatly affect the oxygen content of closed volumes of water. The oxygen microenvironment in various areas of the tank, such as within a coral colony or between live rocks, was not tested here. It is likely, however, that there are areas within such spaces with much less oxygen than is present at the top or middle of the open water column, similar to what is found on coral reefs. The flushing and flux of the water column into such spaces is important in maintaining oxygen levels within them, and this serves to underscore the importance of water flow within aquaria.

In terms of the factors tested in this article that affect oxygen levels, I used the presence or absence of protein skimmers, an algal turf scrubber, powerheads and circulating pumps, airstones, and light (by photosynthesis) to determine their effectiveness at elevating or maintaining oxygen levels in tanks. I also tested their effect on hypoxic and normoxic water, and tested for the effects of the overlying air by sealing tanks or containers and by measuring differences between having an atmospheric versus a pure oxygen environment over the water.

Tank 1 Discussion

The ten-gallon tank containing clownfish has been set up as an unskimmed system with what I consider to be an average stocking density of organisms for a tank of its size. I had assumed (wrongfully) that oxygen was maintained at high levels through the use of two powerheads that agitated the water's surface. However, once the lights went out and photosynthesis stopped, oxygen levels dropped quickly from a high of 78.7% of saturation to a hypoxic low of 16% of saturation. The levels were apparently low enough that each night, the clownfish would leave their anemone and adopt a position just under the water's surface directly above a powerhead. Out of concern, I then monitored the changes in oxygen levels at night using an airstone. Oxygen rose quickly and dramatically. At that point, I added a skimmer to the tank, with the result that oxygen is now maintained at much higher levels, ranging from a high of 130% of saturation to a low of 81.2% of saturation. However, it is only when the lights come on that oxygen reaches saturation or becomes supersaturated. It is notable that there appears to be a period early in the day when oxygen levels are maximal, with a depression to slightly subsaturated levels over the course of the afternoon. Also notable is a slight, but noticeable, drop in oxygen immediately after feeding. This measurement has been made repeatedly and is consistent.

Tank 2 Discussion

The 75-gallon tank containing a large porcupine pufferfish has been maintained using a small skimmer and two powerheads, and it uses the same amount of lighting as Tank 1, a much smaller tank. The light levels of this tank are quite subdued, despite housing many apparently healthy and growing zooxanthellate species. I did not expect this tank to have oxygen near saturation values, and I was especially concerned with the oxygen state of the tank at night. Peak values were 75.4% of saturation, but surprisingly dropped only to 63.2% of saturation at night. I sought to raise the nighttime oxygen levels by incorporating an algal turf scrubber operated on a "reverse daylight" photoperiod. The effect is noticeable, with oxygen levels rising quickly once the scrubber lights are turned on. There was an anomalously low reading of 63% of saturation at 3 AM, but this may have resulted from a miscalibrated probe or an accidental bump of the calibration knob in the dark room at 3 AM. I have remeasured the tank at the same time on two other nights and have not found such a low level; they have been within the 70-80% of saturation range. It should also be noted that the recently added algae screen is very poorly developed and does not yet support a lush turf population. As such, I expect that the oxygen levels will rise substantially as these algae develop, and other systems utilizing turf scrubbers support this expectation (Adey and Loveland 1998).

In contrast to Tank 1, adding an airstone does not affect the oxygen levels of this tank in so dramatic a manner as in the smaller tank. While this is expected given the difference in water volume, even after nine hours there is not a large increase in oxygen over the values that occur without the airstone. As with Tank 1, a slight depression in oxygen occurs after feeding that persists for several hours. Also like Tank 1, the oxygen levels are highest several hours after "sunrise" with a progressive decline over the afternoon. This tank does not quite reach saturation and does not become supersaturated with photosynthesis occurring, although light clearly provides a marked increase in the water's oxygen content. I attribute the failure to reach saturating or supersaturating levels to the relatively low irradiance provided to this tank.

Tank 3 Discussion

The multi-tank system that comprises my 600-gallon reef system is brightly lit during the day, and three of the five interconnected tanks are lit on a "reverse daylight" cycle. The system is skimmed, has strong water flow and a surge tank, and many overflows that I expected to provide a high oxygen level in the water throughout the day. Having measured the tank previously, I knew that oxygen levels were supersaturated during the day and still relatively high at night. This occurs despite extensive growths of large corals and over 20 fish. The readings taken in this tank over the course of full days show oxygen levels that are very close to those of the water column over natural reefs. Like the other tanks, there appears to be a consistent pattern of highest levels occurring several hours after "sunrise" with a decline over the afternoon. In this system, it does not appear that skimming or the Tunze Stream powerheads impact oxygen levels to any great degree. A depression did occur after the powerheads were turned off for several hours, but it was not precipitous.

The cyclical oxygen levels rise and fall primarily in response to irradiance similar to what occurs on natural coral reefs. I did not measure oxygen levels at night without the reverse daylight occurring, and the corresponding expected decline during nighttime hours; this measurement should be taken to determine how much oxygen is provided through lighting the organisms in the sump, refugium and surge tanks at night. If no significant changes were observed, I would assume the majority of oxygen enters the tank through the large surface areas of multiple tanks and numerous overflows, fans blowing across the water, as well as some base level caused by the skimmer and powerheads. In this system, the strong water motion is probably important in pushing oxygenated water throughout the many microenvironments found in the complex three-dimensional reef structure.

Tank 4 Discussion

Tank 4 consists of six interconnected tanks, each with a high surface area/volume ratio. Water flow is provided by strong pumps. The system also has a powerful protein skimming system in place, as well as strong lighting. Because the system is entirely lit at night and is located in a sunroom, a substantial amount of light enters the room during the day while the systems are "at night." This irradiance may be partly responsible for the relatively flat and less cyclical oxygen levels that occur in this system, although I am unable to control for this factor without shading many glass walls and ceiling panels. I suspect, however, that the levels would still be high "at night" because of the other factors listed above. Oxygen levels in this system rarely fall to below 90% of saturation, and are frequently supersaturated, though not to the degree I expected. As existing coral fragments increase in size, I will expect to see oxygen levels increase during "day" and decrease "at night." Because of the importance of this system, I was unwilling to experiment with manipulations that might compromise the stability or health of its corals, and that might have provided some interesting data on the importance of various factors in oxygenation of the water.

Live Rock Discussion

It is apparent that "live rock" indeed has a significant metabolic rate that results in nearly hypoxic conditions in the absence of light or gas exchange at the air/water interface. The fifteen-gallon tank that held only a single small piece of live rock and had a low surface area/volume ratio did not quickly gain oxygen once water flow and exposure to the room's air occurred. Instead, the oxygen curve increased the most under illumination. While the increase is not as great as occurs in Tanks 1-4, it should be noted that there was only a small piece of rock to produce oxygen by photosynthesis and irradiance was provided only by a single 18" 15-watt fluorescent lamp.

In the previous article, this same tank was rapidly oxygenated using an airstone and powerhead after being made hypoxic with nitrogen. This reinforces the effect of airstones in small water volumes that was found in Tank 1. Perhaps most interesting was the rapid decline in oxygen that occurred when the live rock was removed. The tank water became cloudy, probably from a bacterial bloom in the water column, as the water had a smell of fermentation. The rapid drop in oxygen from bacteria or other microbial flora has obvious implications for shipping bags that are frequently cloudy after long transits in dark, stagnant containers.

Chaetomorpha Discussion

Perhaps most interesting were the results of illuminating small tanks containing a handful of macroalgae. In the previous article, the fifteen-gallon tank described above used the same clump of Chaetomorpha that was described in this article. In the fifteen-gallon tank made hypoxic through the use of nitrogen and illuminated by a single 15-watt fluorescent lamp, little oxygen was provided to the tank by the algae. In the ten-gallon tank described in this article, the irradiance was provided by a 65-watt lamp. So long as water flow was provided, the tank increased its oxygen content only minimally, and it is hard to say if the same increase would have occurred without the algae. Once the water flow was turned off, however, oxygen levels rose quickly. I believe this occurred because the water flow caused much of the oxygen produced by the algae to be lost at the air/water interface. With no other organisms present (besides, obviously, any microbes present), oxygen levels remained supersaturated in the tank over many hours; longer, in fact, than I would have expected given gas exchange at the surface. The reduced effect seen previously in the fifteen-gallon tank probably resulted from irradiance levels that were inadequate to maximally stimulate photosynthesis. This experiment shows algae's potential under sufficient irradiance and slow flow, such as the conditions found in refugia, to effectively raise oxygen levels in tank water.

Shipping Bag Discussion

From the limited results shown here, it appears that "breathable" bags are indeed gas permeable. In this experiment, oxygen levels decreased, but I did not attempt to determine corresponding increases in hypoxic water using the same bags. Furthermore, the motor of the stirplate caused the bag's water to warm up, and gas bubbles were noticed on the bag's inner surface. This may explain the decrease in oxygen levels in the breathable bag over time. In contrast, the gas impermeable plastic bags neither gained nor lost oxygen. However, when a medium sized coral was placed into a gas permeable bag, oxygen levels dropped from a supersaturated state to nearly hypoxic conditions within seven hours. Porites cylindrica, if anything, is a species that does not tend to produce copious mucus that would tend to foul the bag water, and is generally found to be a "good shipper." In fact, the water in the bag remained mostly clear despite hourly stirring with the probe inside the bag. Given the rapid and consistent decline in oxygen, it would seem that packing corals in a bag full of air may be a good way to ensure the loss of many species unless the transfer from location to location is quite short in duration.

The point data taken from numerous shipping bags containing either corals or snails clearly shows that using oxygen in shipping bags provides abundant oxygen to the water, and results in supersaturation of bag water even after extended periods of time. In fact, some bag water contained such high levels of oxygen that oxygen toxicity might be a consideration for some organisms. In contrast, shipping bags sealed only with atmospheric air began declining immediately and became hypoxic after lengths of time similar to the pure oxygen containing bags. This is expected, and follows a similar decline in tanks at night and/or sealed from water/air interface exchange.

Also of concern are the extremely low oxygen levels in bags that held "cloudy" water, similar to what was found in the cloudy water of the live rock tank described above. I am not sure how such events could be prevented, although the potential of gas permeable bags to ameliorate the rate of decline might be possible. In the near future, I will be receiving corals in gas permeable bags after a long overseas transit, and I will report any significant findings, if they occur, in The Coral Forum.

Stirring and Bubbling Discussion

The effect of stirring and bubbling using an airstone on normoxic water was rather unremarkable. Only extremely vigorous stirring increased oxygen levels of freshly made seawater to near-saturation levels. An airstone and strong air pump did not have much effect on normoxic water even though only one liter of water was used. When combined with the results described above, it appears that airstones under normal use have a finite capacity to increase oxygen content of seawater in tanks from 0.27 to 75 gallons in volume to a high but subsaturating value of perhaps around 90% of saturation over the course of many hours. Similarly, water movement by stirring on a stirplate or by using pumps and powerheads in tanks does not result in oxygen saturation unless the circulation is extreme or long periods of time are involved.

Point Measurements Discussion

The numerous measurements made in various tanks suggest that most "average" tanks with circulation, lights and an "average" composition of organism inhabitants, or freshly mixed seawater, maintain oxygen levels that are above hypoxia but are not saturated or supersaturated.

Conclusions and Recommendations

While this work is not comprehensive, it does indicate that some methods are better than others at maintaining or increasing oxygen levels. Based on what I have shown in this paper, the following conclusions and recommendations are made:

• Reef tanks approximate the cyclical nature of oxygen dynamics found in the field.

• Variation on daily and seasonal cycles is the rule rather than the exception on natural coral reefs, and appears to be the rule in reef aquaria, as well.

• Aquaria can and do become hypoxic at night and such a state may pose a risk to hypoxia-intolerant organisms. Cloudy water in shipping containers and tanks is a cause for concern as oxygen levels are measured to decline rapidly and to very low levels.

• Gas impermeable bags packed with an oxygen cap result in high water oxygen levels even over long periods of time. The levels are, in some cases, extremely high and may be a cause for concern in hyperoxia-sensitive species. Gas permeable bags are not permeable enough to ensure adequate oxygen levels in bags containing living specimens over normal overnight shipping durations.

• Aquaria can and do become saturated or supersaturated with oxygen during the day, and this is a result of oxygen resulting from irradiance of photosynthetic organisms. In no case was saturation or supersaturation measured without photosynthesis.

• Airstones and skimmers appear to be a very effective means of oxygenating small water volumes. Their effect on larger water volumes appears to be less. While the effect may be relative, the larger tanks and systems described here utilized powerful skimming or air pumps, and to gain an equivalent amount of oxygen as occurs in small water volumes would likely require air pumps or skimmers far larger than those commonly employed by aquarists. This includes data from a coral farm where very large commercial sized skimmers and high surface area/volume ratios failed to produce water even nearly saturated with oxygen at night with a heavy coral population.

• Powerheads and recirculating pumps do not appear to greatly increase the oxygen saturation state of seawater aquaria. Instead, they probably serve to move oxygenated waters to areas of the tank that are locally lower in oxygen resulting from respiration within the tank.

• Using algae in reverse daylight tanks appears to be an effective means of keeping oxygen levels at normoxic levels at night. This effect is pronounced even in tanks and systems that employ protein skimmers and airstones.

References

Adey WH, and K Loveland. 1998. Dynamic aquaria: building living ecosystems. Academic Press, NY. 498 pp.

Adey WH, and R Steneck. 1985. Highly productive eastern Caribbean reefs: synergistic effects of biological, chemical, physical and geological factors. In: The ecology of Coral Reefs ( M Reaka, ed.), NOAA Symposium Series on Underwater Research, Washington D.C 3: 163-187.

Borneman EH. 2005a. The need to breathe in reef tanks: is it a given right? Reefkeeping 4(5) http://www.reefkeeping.com/issues/2005-06/eb/index.php

Borneman EH. 2005b. The need to breathe, part 2: experimental tanks. Reefkeeping 4(6) http://www.reefkeeping.com/issues/2005-07/eb/index.php