Why Are Water Changes Necessary?

One issue that has come up several times throughout the genesis of the Fishsheets is the reliance of the Fishsheets on water changes as a measure of cleanliness. Many users have written to me about how unnecessary water changes are and that my spreadsheet should not require water change data as a prerequisite for generating results. Some aquarists have cited sources that praise the use of live plants in freshwater aquariums while others feel that their filters are so efficient that water changes are not needed. Some such arguments are very good and I am impressed by the results that can be achieved by using live plants or highly efficient filters. Nonetheless, this website and my spreadsheets are not interested in simply what can be done. Instead, I have dedicated my studies to what should be done for the fish (thrive or survive?). Consequently, I submit the following rational for requiring that users input their water change data:


Simple Explanation: Fish Produce Substances That Cannot Be Entirely Removed And Aquarists Should Not Strive To Be As Lazy As Possible

Let's begin with a very simple thought: fish produce waste and that waste has a complex composition (source on composition of fish waste from user "THE V" from AquariaCentral.com). The fact that fish produce waste is very easy to confirm by simply looking into any aquarium with fish and noticing that fish produce feces. Fish also produce urine, though it usually dissolves readily and thus is not easy to see (though the effects may be obvious). Moreover, an aquarium may have sources of waste that do not depend entirely on the fish; uneaten food, decaying plant matter, "sludge" in the filter, accidental contaminants like household cleaners, and the occasional dead fish all contribute to the waste load that a given aquarium system may have to endure. Since waste is a sure component of any aquarium system, aquarists are stuck with the problem of how to remove it from the system so that the fish do not have to live in a toxic soup.
Of the principle methods of processing wastes in a freshwater aquarium that do not involve direct removal from the system are the Nitrogen Cycle, chemical filtration, miscellaneous biological filtration methods, and the use of live plants. All of these methods leave the wastes in contact with the water and thus do not actually "purify" the system. For example, the Nitrogen Cycle only addresses wastes with a nitrogenous component and does not usually break the wastes down entirely as the net byproduct is still nitrate (which can be toxic at high concentrations). Fortunately, however, plants absorb nitrate and thus can complete the work of the Nitrogen Cycle, but plants will not require all the components that comprise the waste of a given aquarium system. For example, sulfides are sometimes found as a component of aquarium waste, yet plants cannot use sulfides to any great degree. The same goes for the variety of alternative biological filtration methods--it is very hard to create a system that perfectly matches the composition of the waste with the exact needs of all the living entities in the aquarium.
Of course, the inability of a confined system to replicate nature by use of biological entities does not address chemical filtration. Chemical filtration can often be separated into two broad categories: 1) adsorbing/absorbing materials and 2) chemically-altering processes. The first problem with adsorbing/absorbing materials is that they tend to be very size-specific. Activated carbon, as an illustration, will only adsorb molecules of a particular size and all the other molecules are left free to enter the water column and be internalized into the system by bacteria, etc (Wikipedia: Activated Carbon; Steve Kvech and Erika Tull). What is more, I have had experience working with the College of Southern Idaho Laboratory that suggests that bacteria can actually liberate some of these adsorbed molecules, reintroducing them to the aquarium system. As to the chemically-altering chemical filtration methods, they often convert one noxious waste substance (or something that is inert) into something that, though not noxious in itself, can be harmful in high concentrations (anything can be harmful in high concentrations).
The above resources exhausted, the one sure method of actually removing wastes from the aquarium system is water changes. Many aquarists hate to hear it, but water changes are an extremely reliable method of waste removal in that they physically remove wastes from the system and replace the contaminated water with water that is relatively pure. Also, if water changes are combined with gravel siphoning (which all of my spreadsheets assume is the case), then the amount of wastes that is removed is significantly higher.
I am not saying that fish cannot live without water changes; that is a different argument entirely. What I am saying is that it is in keeping with best practices to go the extra mile and be sure that a system is safe by doing regular, relatively frequent water changes. My personal recommendation is to do 25% water changes every week in combination with thorough gravel siphoning. No less, the value that is appropriate to your system may be higher or lower depending on the stocking density of your aquarium, filtration methods, and use of live plants (all of which are considerations that my Fishsheets are designed to handle, to a reasonable limit). If you want to examine your water change schedule in terms of individual components and not total remaining waste, then I suggest taking a look at this new calculator.


The Remaining Waste Paradox: Won't The Amount Of Waste Continue To Grow If You Only Remove A Percentage Of The Total Waste Each Time?

Ostensibly, removing a percentage of the aquarium water removes a proportional amount of waste from the aquarium system as the replacement water will not have the same impurities. This reduction in the waste level is the justification that I and many other aquarists use to justify the validity of using water changes to reduce waste levels. Nonetheless, the premise behind a partial water change presents a problem: if you remove only part of the waste, then part of the waste will still remain. This remaining waste is added to the amount of waste that the aquarium system produces until the next partial water change. Thus, the amount of waste in the tank at the next water change is actually higher than it was the first time. Consider the following scenario (completely demonstrative; not an actual case): a 10-gallon aquarium with fish that produce 100mg (milligrams) of waste each week is cleaned once a week with 25% water changes. After the first cleaning, the aquarium will contain 7.5mg of waste per gallon.
100mg / 10 gal. = 10mg/gal.; 10mg/gal. - 0.25*10mg/gal. = 7.5mg per gallon.
However, when the aquarium is ready for its second cleaning, the waste level will be 17.5mg per gallon because the fish have added 10 mg per gallon over the course of the intervening week. This means that when the aquarium is cleaned, then it will have a new waste level of 13.125mg per gallon.
7.5mg/gal. + 10mg/gal. = 17.5mg/gal.; 17.5mg/gal. - 0.25*17.5mg/gal. = 13.125mg/gal.
If the process is repeated once more, then the waste remaining after the third cleaning will be 17.34375mg per gallon.
13.125mg/gal. + 10mg/gal. = 23.125mg/gal.; 23.125mg/gal. - 0.25*23.125mg/gal. = 17.34375mg per gallon.
As can be seen, the amount of waste is actually increasing each cleaning despite that some of the waste is being removed. This presents a serious problem for people who praise the value of water changes. Based on the above trend, it would appear that completing water changes is not much better than simply letting the waste pile up without removing any (the same tank with no water changes would have a waste level of 30mg/gal.; 3weeks*10mg/gal. = 30mg/gal.). Since this appears to be the case, a lot of people do not put much stock in water changes.
The above paradox is hard to overcome unless one considers the following question: Is there a limiting level of waste such that the amount removed by each partial water change is equal to the amount produced between each partial water change? If there is an answer to this question, then it would follow that the amount of waste would rise until that level is reached but not continue to increase afterward. For the above scenario, the equation to model this crucial question would be as follows:
X = amount of waste remaining before a partial water change;
X - 0.25*X + 10mg/gal. = amount of waste remaining one week after a partial water change;
amount before water change must equal amount one week after water change;
X = X - 0.25*X +10mg/gal.; since a single X appears on both sides, they would cancel; 0 = -0.25*X + 10mg/gal.;
the negative -0.25*X could be moved to the left to make it positive; 0.25*X = 10mg/gal.; to isolate X, divide by 0.25;
X = 10/0.25mg/gal. = 40mg/gal.; Therefore, the waste level will stabilize once the waste level reaches 40mg per gallon;
Check: Must be equal to 40mg/gal. one week after a water change:
40mg/gal. - 0.25*40mg/gal. + 10mg/gal. = 40mg/gal. - 10mg/gal. + 10mg/gal. = 40mg/gal.
This equation implies that there is always some limiting amount X that the waste level will reach before it stabilizes. This being the case, doing partial water changes is better than doing nothing as doing no water changes would result in an infinite amount of waste. No less, a major concern that this simple result hints at is that it is hard to tell what the limiting value is without doing the math for each and every case (making it hard to tell just how much cleaner the aquarium will be if a particular water change schedule and percentage is used in favor to another combination). Fortunately, my spreadsheets can compute this result pretty easily, but they do so by using the amount of fish biomass to estimate the amount of waste the fish will produce. The spreadsheets use a version of the following equation:
Where r is a ratio of effectual (after accounting for the effect of plants and filtration) daily IFUs per gallon, d is the days between cleanings, p is the percentage removed, x is the number of days since the aquarium was started, and floor() is the greatest integer function (rounds down decimal values).
A couple typical runs of the above formula might look something like this (each gap in the curve represents a water change):
Based on the above graph (which covers two years worth of data for each scenario, 730 days) it can be seen that unchecked waste quickly exceeds any partial water change schedule within the first six months (180 days or less). Also, it can be seen that a larger amount removed (the 30%) does not always equate to less waste (note how much lower the 25% curve is compared to the curve at 30%). The reason for this discrepancy is the cleaning frequency: cleaning every seven days does a much better job at keeping the tank waste levels down than cleaning every month (30 days).
These kinds of calculations are handled almost instantaneously by my Fishsheets, but it still may be useful for you to check your water change schedule against a sort of standard. One simple standard is the RWI factor, a simple calculation that uses the days between cleanings and the percentage removed per cleaning to assign a relative cleanliness index to a particular aquarium system. The RWI factor omits the ratio (r in previous formulas; 10mg/gal. in the initial example) and thus is not an entirely accurate method of judging tank cleanliness (plugging your data into the Fishsheets would be much more accurate), but it will give you a ballpark estimate of how efficient your cleaning schedule is when compared to another cleaning schedule for the same tank. For example, it can be seen from the below chart that changing the water every 12 days at 20% each time results in about three times more remaining waste than changing the water every 2 days at 10% (60 / 20 = 3).
On the above chart, RWI values between 15 and 45 are considered ideal (by me). RWI values below 15 are considered to be a bit too clean and potentially harmful if the aquarist does not take care in acclimating the fish to the large environmental changes that will occur. That is, if one chooses to clean very often or in very large volumes, then one must be sure that the fish are used to having stuff going on inside their home. Many people who have low RWI numbers suggest trying to avoid things like splashing the water too much and simply being considerate toward the fish (click here for more on water changes with a low RWI factor). RWI values above 45 are considered too dirty in that a significant amount of waste will always remain in the tank if the aquarium is fully stocked (if the aquarium is not fully stocked, the filter is highly efficient (highly), or a ton live plants are used, then the RWI factor can be higher). To examine your own water change schedule in greater detail, click here.