Although the basics of biological filtration (specifically nitrification) are understood by all good fishkeepers, it is perhaps surprising that the efficiency with which this process occurs can vary greatly depending on the physical, chemical, and biological environment. By understanding how different factors can influence filter performance, it is often easier to determine why a particular filter might not be working as expected. In some cases the problem can be resolved, and the filter then returned to full working order, whereas in others the environmental factor may be unchangeable, and the set-up of the aquarium must be adapted, or additional equipment installed.

The following provides a brief review of those environmental parameters that have the most influence on the ability of a filter to convert ammonia into nitrate.


Just like fish, the metabolic rate of nitrifying bacteria is dependant upon temperature, and so perhaps more than any other factor, temperature governs the capacity of a filter to remove ammonia. As the temperature of the water increases, so does the ability of nitrifying bacteria to remove ammonia. Fortunate really, as increased temperatures also mean greater ammonia production by the fish.

Indeed, various experiments have been conducted into the effects of temperature on ammonia removal, over different temperature ranges (from as low as 1.67oC, to as high as 35oC). Such investigations not only confirm the effect of temperature on ammonia removal, but also show us that some nitrification can occur at winter temperatures -further evidence to suggest that leaving filters running during the winter is a good idea.

From various experiments conducted on temperature effects, we can gain an idea of the 'optimum' temperature for ammonia removal, as well as the ideal range of temperatures over which it is most effective. Exact figures are difficult to quote, as many factors may influence the ideal temperature under a particular set of conditions. However, temperatures of around 28 - 35oC seem to give the best results (if oxygen levels are maintained), although filters will work efficiently from well below 1oC up to around 35oC. Some research suggests that temperatures of 35oC or more may be life-threatening to filter bacteria under certain conditions, so it is wise to avoid extremes (overheating in the summer could pose a problem). However, rapid temperature fluctuations and lack of oxygen due to overheating are more likely to be the cause of temperature-related filter inhibition.


Nitrifying bacteria are primarily aerobic, and require oxygen in order to convert ammonia into nitrite, and eventually nitrate. A fall in oxygen levels limits the ability of the filter to remove ammonia and the nitrification process may fail. Typically, nitriteremoving bacteria will usually suffer first when oxygen levels fall, as they are the final stage of the process and thus have to obtain what oxygen is left after heterotrophic and ammoniaremoving bacteria have had their share. Also, ammonia- removing bacteria produce substances (hydroxylamine) that can inhibit nitrite removers under low oxygen levels. This explains why nitrite levels can fluctuate in some ponds in summer.

In terms of the amount of oxygen needed for nitrification, for every 1 mg of ammonia that is converted to nitrate, 4.57mg of oxygen are required. A more instantly understandable approach is to talk about the concentration of oxygen that is needed, in terms of mg/l (parts per million). Again, research has thus far not been able to provide a definite minimum, as the amount needed will vary from situation to situation. It is generally agreed that the oxygen concentration in the water leaving a filter (therefore after nitrification), should be 2mg/l or above. This ensures that the final stages of the filter have received sufficient oxygen for nitrite to be converted to nitrate. In terms of the oxygen concentration of the water going into the filter, prior to nitrification, you are looking at around 6-8mg/l, depending on the condition and design of the filter itself, and other prevailing environmental factors.

Unless you are employing a trickle filter, which should provide its own oxygen supply when working correctly, the supply of oxygen to filter bacteria is dependent upon flow rate, the oxygen content of the water, and the amount of oxygen used in the filter. By ensuring that filters operate at their maximum flow rate, and kept reasonably clean (to reduce oxygen consumption by the breakdown of solid waste), you will be ensuring that maximum oxygen as possible is available to the biological filter. To increase the supply further, you will need to increase the oxygen content of the water. This is done in the usual way - air-pumps, venturis, waterfalls, etc. In large external pond filters it can be an excellent idea to add airstones to the filter itself, to ensure a sufficient supply of oxygen, and also to ensure that water exiting the filter is not deoxygenated.

Tetra Filters used as trickle filters
can improve ammonia removal

pH and Alkalinity

Nitrifying bacteria prefer alkaline conditions, with optimum pH levels ranging from around 7.5- 8.5. As with fish, fluctuating pH levels are generally more of a problem than the specific pH itself, and can inhibit ammonia removal until adaptation is complete. The range over which nitrification can occur efficiently is approximately 6.0-9.0, although given time to adapt, it is likely to occur at slightly lower pHs, albeit at a much reduced rate.

Alkalinity (carbonate hardness) is also important, as nitrification is an acid-forming process (resulting in hydrogen ion production), and because it possibly provides a carbon source essential for the growth of nitrifying bacteria. Approximately 7.14 mg of bicarbonate are required to neutralise hydrogen ions produced from the conversion of 1 mg of ammonia to nitrate. Water changes (in most areas), or the addition of EasyBalance, will prevent bicarbonate loss and subsequent pH crashes due to this effect. In terms of filter efficiency, at carbonate hardness levels below 100mg/l (around 6odH), biological filtration starts to become inhibited.

Clearly, in soft water conditions some thought needs to be given to stocking levels, as biological filtration will be impeded. It is risky to rely on the fact that ammonia is less toxic at a low pH (both ammonia and ammonium may possibly be used by the filter, and so nitrite production could still occur), and so over-sizing the filter / under-stocking the tank, or using additional means of ammonia removal may be pertinent.

Salinity and Mineral Content

As we all know, nitrification can occur at all salinities commonly found in aquariums and ponds (albeit at a reduced rate in salt water) thanks to their adaptability, and to the presence of many different species that are suited to varying environments. What is important though is the temporary inhibition that can occur if salinity is suddenly altered. It is generally accepted that, provided the change in salinity is below 5ppt (5g/I), there will be little effect on nitrification. Therapeutic doses commonly used in ponds should therefore not present a problem.

Certain minerals are also important for efficient nitrification, including calcium; magnesium; molybdenum; iron; copper; sodium. Regular water changes will replace some or all of these and adding EasyBalance will maintain levels of all of them.


Natural light has a strongly negative effect on nitrifying bacteria. Total darkness is preferable for optimum filter performance, with inhibition starting atjust 1% of daylight. Again, nitrite removers are the first to suffer. It is therefore sensible to keep open-topped pond filters covered. Aquarium filters are not subject to the same light intensities, and therefore light-inhibition is not a significant factor in tanks, even with clear box filters and the like.

Solid Waste Solid waste can not only impair the flow of water through the filter, but it also encourages the growth of heterotrophic bacteria (bacteria which obtain their food from organic debris). These bacteria compete with nitrifiers for space, resources, and oxygen, and therefore inhibit them. It is better to have a fairly clean filter, especially if the biological and mechanical sections are one in the same (as in an aquarium filter). Pond filters that are left clogged in the summer often fail to remove nitrite from the water properly.

Flow rate / retention time Effect of flow rates on performance is a complicated area, as the ideal flow (and therefore retention time) will depend on the design of the filter, water conditions, filter media used and filter maintainence. In essence, a trade-off between nutrient supply, contact time for ammonia/nitrite removal, and the need to avoid excessive removal of the biofilm (where the bacteria live). Importantly, very long retention times thickens the biofilm, to a point where the bacteria are unable to remove ammonia and nitrite. Also, solid waste will settle out on the biological media and clog it.

Biofilm development under poor flow rates / stagnation

A: Biofilm is initially thin, allowing good exchange of nutrients, oxygen,
and nitrogenous products.

B: Low flow rate allows biofilm to thicken in layers.
Nutrients and oxygen cannot reach bottom layer, and nitrification here ceases. In some cases, nitrate may be reconverted to ammonia.

  C: Eventually the 'dead' layer
      sloughs off, taking with it
      the upper active layers.

      This removes nitrifying
      ability, and may impede
      overall filter performance.

Under correct flow rates, biofilm thickness is kept low, allowing more consistent removal of ammonia and nitrite.

Most pond and aquarium filters are essentially submerged, and therefore rely on the flow of water to supply nutrients and oxygen. With such filters, it is usually best to run them at their maximum flow rate, all other things being equal. With more specialised filter, such as fiuidised beds and trickle filters, an argument could be made for increasing retention time. However, work to date indicates that in most cases, a higher flow rate gives preferable results (provided it is within set limits). Given the specialised nature of these and other more advanced filters though, it is always best to follow the manufacturer's guidelines.

In the majority of situations, the biological capacity of aquarium and pond filters is more than enough to deal with ammonia production, even under varying, and sometimes less than ideal, conditions. However, most of us, particularly those who are involved with helping out novice fishkeepers, will at some time or another come across situations where filters seem to be underperforming for no obvious reason. Having knowledge of some of the factors which could be responsible for this can in these cases provide a solution to the problem.

Further reading: Lawson, T. (1991), Fundamentals of: Aquacultural Engineering Wheaton, F. (1993), Aquacultural Engineering

Last updated January 2005