How Biological Filtration Works in a Koi Pond

1. Where the Nitrogen Load Comes From

Biological filtration begins with feed. Koi food contains protein, and protein contains nitrogen. Some dietary nitrogen is retained in new tissue as the fish grows, but a significant fraction eventually returns to the water through normal metabolism and waste. Fish excrete much of their nitrogenous waste across the gills as ammonia, while faeces, uneaten food, mucus and other organic matter can release additional ammonia as they are decomposed by microorganisms.

This is why feeding rate is so important when sizing and evaluating a biofilter. Pond volume matters for dilution and hydraulic design, but it does not directly create ammonia. Two ponds of the same volume can impose very different biological loads if one contains a few lightly fed koi and the other contains a dense collection being fed heavily for growth. In practical system design, the quantity and quality of feed entering the pond are often better indicators of biological demand than litres of water alone.

Fresh solids can become a biological load later

Mechanical filtration and biological filtration are linked. A piece of faecal material is initially a particle that can potentially be removed. If it is trapped and left submerged, bacteria continue breaking it down. Part of its nitrogen can then enter the dissolved waste stream. This means efficient solids removal does more than improve clarity: it can reduce the amount of organic material competing with nitrifiers and reduce the downstream burden on the biofilter.

This is one reason high-performance koi systems aim to export solids early. The goal is not to ask the biofilter to process every form of waste. Physical waste should be physically removed where practical; the biofilter is then allowed to concentrate on dissolved nitrogen under cleaner, oxygen-rich conditions.

2. Ammonia Is Not One Single Chemical Form

In pond testing, ammonia is commonly discussed as total ammonia nitrogen, or TAN. TAN represents the combined nitrogen present as un-ionised ammonia (NH3) and the ammonium ion (NH4+). The two forms exist in an equilibrium that shifts with water chemistry.

For koi keepers, the distinction is important because un-ionised NH3 is the more toxic form. As pH and temperature rise, a greater fraction of TAN is present as NH3. The same laboratory or test-kit TAN reading can therefore represent different toxic risk under different conditions. A TAN result should never be interpreted without considering pH and temperature.

This does not mean low pH is a good ammonia-control strategy. A falling pH can suppress nitrification and eventually destabilise the biofilter. The objective is a stable pond in which ammonia production is matched by biological conversion, not a pond in which ammonia is merely shifted temporarily toward the less toxic ammonium form.

3. Nitrification: The Central Biological Process

The core function of a conventional koi biofilter is nitrification. Nitrification is an aerobic microbial process in which reduced nitrogen is oxidised toward nitrate. In simple pond language, it occurs in two linked stages: ammonia oxidation and nitrite oxidation.

Stage one: ammonia oxidation

Ammonia-oxidising microorganisms use ammonia as an energy source and convert it through intermediate reactions toward nitrite. This first stage reduces the ammonia concentration but creates another compound that is also dangerous to fish. A filter that can oxidise ammonia but cannot keep up with the second stage will show low or falling ammonia while nitrite rises.

Stage two: nitrite oxidation

Nitrite-oxidising microorganisms then use nitrite and convert it toward nitrate. In a mature, balanced system, the two stages operate continuously enough that neither ammonia nor nitrite is allowed to build up substantially between passes through the system.

The familiar textbook shorthand often assigns the first stage to Nitrosomonas and the second to Nitrobacter. Those names are useful for teaching the concept, but modern molecular studies show that real freshwater biofilters can contain much more complex nitrifying communities. Nitrospira can be important nitrite oxidisers, ammonia-oxidising archaea may contribute substantially, and some Nitrospira are capable of complete ammonia oxidation, known as comammox. For the koi keeper, the practical message is that a mature biofilter should be treated as a complex living ecosystem rather than as a culture of only two named bacterial genera.

The simplified pathway remains useful operationally: feed and metabolism create ammonia; the biofilm oxidises ammonia to nitrite; other nitrifying organisms oxidise nitrite to nitrate. Water tests tell you whether those linked processes are keeping pace with the load.

4. Nitrification Has a Real Oxygen Cost

Nitrification is an aerobic process. The microorganisms carrying it out require oxygen, which means the biofilter competes for the same dissolved oxygen resource needed by the koi and by other aerobic organisms in the pond. This is especially important at night, during hot weather, during heavy feeding, or whenever aeration and circulation are reduced.

From the chemistry of complete nitrification, oxidising one unit of ammonia-nitrogen to nitrate requires approximately 4.6 units of oxygen by mass before allowing for the oxygen demand of fish, heterotrophic bacteria and other processes. In a heavily fed system, the biofilter itself can therefore represent a meaningful oxygen demand.

Why low oxygen can create a misleading filter problem

A keeper may see ammonia or nitrite rising and assume the pond needs more media. Sometimes the media is adequate but the nitrifiers are being starved of oxygen. A dirty mechanical stage, clogged media, failed air pump, reduced water flow or low pond oxygen can reduce the amount of oxygen reaching the biofilm. The result is lower nitrification performance even though the nominal filter volume has not changed.

Good biological filtration therefore requires oxygen to be designed into the system, not added as an afterthought. Moving beds need enough air to keep the carriers properly mixed; shower and trickling filters need strong gas exchange and unobstructed airflow; submerged static media needs adequate dissolved oxygen and hydraulic movement. In every format, oxygen must reach the active biofilm.

5. Nitrification Also Consumes Alkalinity

As ammonia is oxidised to nitrate, acidity is produced. In practical pond terms, nitrification consumes alkalinity, much of which is provided by bicarbonate in freshwater. Stoichiometrically, full nitrification consumes roughly 7.1 mg of alkalinity expressed as CaCO3 for every 1 mg of ammonia-nitrogen converted to nitrate.

This matters because alkalinity acts as a buffer against rapid pH change. If a heavily stocked koi pond receives low-alkalinity source water and the keeper does not monitor KH or alkalinity, continuous nitrification can gradually reduce the buffer reserve. The pH may remain apparently acceptable for a time and then become increasingly unstable as the reserve is depleted.

The pH–KH–biofilter connection

Nitrifying organisms are sensitive to pH. Their activity generally slows as pH falls away from their favourable range, and very low pH can severely suppress nitrification. A low-buffer pond can therefore enter a destructive feedback loop: nitrification consumes alkalinity, pH declines, nitrification slows, ammonia rises, and the system becomes less stable precisely when biological treatment is most needed.

For this reason, pH and KH should be viewed as part of filtration management rather than as unrelated water-quality numbers. The exact target depends on the overall husbandry strategy and source water, but stability is more important than chasing a single fashionable figure. In high-load systems, regular alkalinity testing is especially valuable because a pH reading alone does not tell you how much buffering reserve remains.

6. Biofilm: Where Biological Filtration Actually Happens

Most koi biofilters rely on attached-growth biology. Microorganisms colonise the surfaces of the media and form a biofilm: a living layer containing nitrifiers, heterotrophs, extracellular material and other organisms. Water flowing past the biofilm supplies ammonia, nitrite, oxygen and inorganic carbon while carrying away metabolic products.

The biofilm is not a uniform paint-like coating. Conditions change from the outer surface toward the deeper layers. Oxygen and substrates must diffuse into it, while products diffuse out. If the film becomes excessively thick or buried under organic deposits, the deeper zones can become less efficient for the aerobic nitrification that koi keepers want.

More surface area is useful, but only when the surface is active

Filter media are often marketed by specific surface area, usually expressed as square metres of surface per cubic metre of media. Surface area matters because a larger active area can support more attached biomass. However, a catalogue surface-area number should not be treated as biological capacity by itself.

Very fine pores may be difficult for oxygenated pond water to penetrate. Media can become coated with organic matter. Flow can bypass part of a chamber. Some carriers protect biofilm in internal spaces, while others depend on exposed surfaces. Different systems shear and renew biofilm differently. What matters is effective, wetted, oxygenated surface under the actual hydraulic and organic load—not the largest theoretical number printed on a brochure.

This is also why a small amount of well-operated media can outperform a much larger volume of poorly aerated, dirty or bypassed media. Biofilter design is an interaction between biology and hydraulics.

7. Nitrifiers Are Slow-Growing Compared with Many Other Bacteria

One of the most important reasons new ponds experience ammonia and nitrite problems is that nitrifying populations do not establish instantly. Many heterotrophic bacteria that consume organic carbon can grow much faster than the organisms responsible for nitrification. A new biofilter therefore needs time to develop enough active nitrifying biomass to match the feeding load.

This is why the term 'cycled' is more meaningful than simply saying a filter has been running for a certain number of days. A biofilter is functionally mature when its microbial population can process the nitrogen load being presented to it while keeping ammonia and nitrite controlled. Time helps, but the rate of maturation depends on temperature, pH, alkalinity, oxygen, inoculation, ammonia availability, flow and other conditions.

Maturity is load-specific

A filter that has matured while supporting five small koi is not automatically mature for twenty large koi. Biological capacity adapts to available substrate. When biomass or feeding increases sharply, the existing nitrifying population may need time to expand. This can produce a mini-cycle even in a pond that has been established for years.

The same principle applies after prolonged fasting. If a pond receives very little ammonia for an extended period, the nitrifying community may adjust downward. Restarting aggressive feeding immediately can create a temporary mismatch between waste production and oxidation capacity.

8. How a New Biofilter Cycles

A typical new system does not move from sterile media to perfect nitrification in one step. The biological community develops in stages as substrates become available.

Early phase: ammonia becomes measurable

With fish present, ammonia begins entering the water. At first, the population of ammonia-oxidising organisms may be too small to process the full load. TAN therefore rises. The exact pattern depends on stocking, feeding and the starting microbial population.

Middle phase: ammonia falls while nitrite rises

As ammonia oxidisers establish, ammonia control improves. Their activity creates nitrite, which may accumulate because the nitrite-oxidising population is not yet equally developed. This is the classic second phase of cycling and is why measuring ammonia alone is insufficient.

Later phase: nitrite falls and nitrate production becomes established

As the nitrite-oxidising community expands, nitrite comes under control and nitrate becomes the main oxidised nitrogen product. At this stage the system may be described as cycled for the current load, provided the results remain stable under normal feeding.

Extension guidance for recirculating aquaculture commonly describes biofilter establishment as requiring weeks rather than days, with the exact period varying widely according to conditions. Commercial bottled cultures, seeded media from a healthy established system, and careful fishless cycling can accelerate colonisation, but none of these methods removes the need to confirm performance by testing.

A mature-looking brown biofilm is not proof by itself. The safest evidence is functional: ammonia and nitrite remain controlled while the fish are being fed at the intended rate.

9. The Main Environmental Factors That Control Biofilter Performance

A biofilter can only perform as well as the environment allows. The most important operating factors interact, so troubleshooting should consider them together rather than in isolation.

10. Temperature: Why the Same Filter Behaves Differently Through the Year

Nitrification is temperature-dependent. As water cools, microbial metabolic rates generally decline. Koi metabolism and appetite also change with temperature, so the relationship is not as simple as saying 'cold equals ammonia.' If feeding is reduced appropriately, the lower waste load may partially offset the slower biofilter.

Problems occur when the two sides become mismatched. For example, a sudden warm spell can increase koi appetite and feeding faster than the nitrifying population has adjusted. Conversely, a rapid cooling event can slow bacterial activity while substantial organic matter remains in the system. Indoor dealer facilities may experience less seasonal change, but they can impose very high continuous feeding and biomass loads instead.

Temperature should therefore be considered whenever interpreting ammonia or nitrite trends. A filter that is adequate at one operating condition may have less safety margin at another.

11. Organic Waste Can Compete with Nitrification

A biological filter is not colonised only by nitrifiers. Heterotrophic bacteria that consume organic carbon also grow on media and wet system surfaces. When large quantities of dissolved and particulate organics enter the biofilter, these faster-growing organisms can occupy surface, consume oxygen and build thicker biofilms.

This is one of the strongest arguments for effective mechanical filtration upstream of the main nitrifying stage. Removing faeces and feed particles before they are broken down keeps the biological filter cleaner and helps reserve oxygen and surface for nitrification. It also reduces sludge accumulation and makes performance more predictable.

A dirty biofilter is not necessarily a more mature biofilter

Koi keepers sometimes equate dark, thick deposits with 'good bacteria.' In reality, excessive sludge can be a warning sign. Healthy biofilm is valuable; trapped decomposing solids are not automatically beneficial. The aim is to preserve attached biological activity while preventing the media from becoming a storage bed for waste.

12. Water Flow, Contact and Hydraulic Design

Water has to bring the dissolved waste to the biofilm. If flow is too low, too much of the pond may pass through the biofilter too slowly and local media zones may receive inadequate substrate or oxygen. If the hydraulic layout allows water to short-circuit, a large chamber can behave like a much smaller one because the water follows the easiest path.

At the other extreme, simply increasing pump size without considering the filter design is not a universal solution. Each biofilter format has an appropriate hydraulic loading range. Excessive flow can cause unacceptable head loss in packed media, poor settling upstream, media carry-over, or insufficient dwell characteristics for a particular unit. The correct objective is not 'slow contact' or 'fast contact' as a slogan; it is even distribution and enough total treatment capacity at the system's actual flow.

Biofilter performance should therefore be evaluated as part of a circulation loop. Bottom drains, skimmers, mechanical separation, biofiltration, pumps, pipe diameters and returns all influence the amount and quality of water that reaches the biological stage.

13. Common Biofilter Formats and How Their Biology Differs

Different filter designs create different physical environments, but all successful aerobic nitrifying biofilters must ultimately provide active surface, oxygen, water movement and access to nitrogenous substrate.

Moving-bed biofilm reactors

Moving-bed systems use buoyant plastic carriers kept in motion by aeration. The movement improves mixing and continually exposes the protected biofilm to oxygenated water. Collision and shear help control excess external biofilm while protected internal areas remain colonised. Their strengths include high aeration, good mixing and tolerance of changing load when correctly sized and operated.

Shower and trickling filters

In shower or trickling configurations, water is distributed over media while air moves freely around the water and surfaces. These systems can combine nitrification with strong gas exchange, including oxygenation and carbon dioxide removal. Their performance depends heavily on water distribution, media wetting, ventilation and adequate flow across the full media bed.

Submerged fixed media

Matting, blocks, static plastic media and other submerged fixed surfaces can support substantial biofilm. Their main management challenge is maintaining clean flow paths and preventing solids from clogging the media. Aeration and upstream mechanical separation become increasingly important as the media bed becomes denser.

Bead and pressurised media filters

Some pressurised filters perform both solids capture and biological treatment. They can support nitrification, but because captured solids remain within the bed until backwashing, maintenance frequency and oxygen conditions are central to their biological performance. A combined filter should be judged on both of its jobs rather than assuming its bio-capacity from vessel size alone.

Later articles in this filtration series examine moving beds, showers, RDF systems and other components in greater depth. The important point here is that equipment type changes the engineering environment, not the basic microbial need for oxygen, substrate, alkalinity and usable surface.

14. How Much Biological Media Does a Koi Pond Need?

There is no universal answer based only on pond volume. Media requirements depend on the amount of ammonia the system must process and on the realised nitrification rate of that media under the actual operating conditions.

A more defensible design process starts with peak feeding rather than litres of pond water. From the expected feed input, the designer estimates nitrogen production, selects a conservative media loading or ammonia-conversion rate appropriate to the filter technology, and adds safety margin for temperature, growth, future stocking and maintenance variation.

Why manufacturer surface-area figures are not enough

Two media products can have very different nominal specific surface areas, but this does not automatically mean the one with twice the published area supports twice the ammonia load. The accessible protected area, hydraulic conditions, biofilm thickness, oxygen transfer, organic fouling and reactor configuration all influence realised nitrification.

For the keeper, the most useful capacity figure is a validated treatment rate under conditions similar to the planned system, preferably expressed against feed load or TAN removal. If only theoretical surface area is provided, treat it as one design input rather than as proof of capacity.

Safety margin is part of good design

A filter that only just matches today's average load is fragile. Koi grow, feeding rates change, summer temperatures rise, mechanical screens become dirty, pumps lose flow and maintenance is occasionally delayed. Commercial and dealer systems are especially variable because fish may arrive in batches. Good design leaves biological headroom rather than planning to operate at the edge of measured capacity.

15. How to Tell Whether a Biofilter Is Actually Working

The filter should be judged by the water entering and leaving the fish, not by the colour of the media or the cost of the equipment. The most useful routine indicators are ammonia, nitrite, pH, alkalinity, dissolved oxygen where measurable, temperature, feed input and trend over time.

Ammonia and nitrite should be interpreted together

If ammonia rises while nitrite remains low, ammonia oxidation may be under-capacity or inhibited. If ammonia is controlled but nitrite rises, the second stage may be lagging. If both rise, the entire nitrification chain may be overloaded, oxygen-limited, chemically inhibited or disrupted by a hydraulic problem.

Nitrate confirms oxidation but does not prove perfect filtration

Rising nitrate is consistent with active nitrification, but nitrate alone does not prove that ammonia and nitrite are always safe. It is possible to have nitrate accumulation and still experience temporary ammonia or nitrite spikes during high feeding. Nitrate also has its own long-term management requirements.

Trend data is more powerful than a single test

A one-off reading tells you what the water was doing at that moment. A log showing feed amount, temperature, pH, KH, ammonia and nitrite makes it easier to see when the system begins losing margin. For valuable koi collections, a simple spreadsheet or monitoring log can reveal problems earlier than visual observation of fish behaviour.

16. What Causes an Established Biofilter to Fail or Lose Capacity?

A mature biofilter is resilient, but it is not indestructible. Most apparent 'biofilter crashes' have an underlying change in load or environment.

17. Cleaning a Biofilter Without Destroying Its Function

Biological media should not be treated like a dirty kitchen surface that must be sterilised. The goal is to remove excessive trapped solids and maintain flow while preserving the useful biofilm.

The correct cleaning method depends on the filter. Moving-bed media is normally self-agitated by air and usually requires management of the chamber rather than scrubbing each carrier. Fixed mats may need periodic rinsing. Bead filters require backwashing. Shower media may need debris removed if upstream mechanical filtration is inadequate. In every case, cleaning should restore hydraulics without unnecessarily stripping all biological growth.

Avoid simultaneous over-cleaning of every biological stage

If a system contains multiple mature biological zones, cleaning them all aggressively on the same day removes redundancy. Staggering heavy maintenance can preserve more active biomass. Similarly, allowing biological media to dry out, leaving it stagnant without oxygen for extended periods, or exposing it to strong disinfectants can damage the community.

Municipal source water may contain chlorine or chloramine, so large-scale rinsing or refilling should follow appropriate dechlorination practice. The principle is simple: clean the system, but do not sterilise the nitrifying habitat unless sterilisation is deliberately required for a biosecurity reason and the keeper is prepared to restart the biological cycle.

18. Power Failures and Stagnant Biofilters

Power loss is one of the most important emergency scenarios because pumps and aeration may stop at the same time. The koi continue consuming oxygen and producing waste, while the biofilter stops receiving fresh oxygenated water. In a heavily loaded submerged chamber, oxygen can fall quickly.

The danger is not only what happens during the outage but what is returned to the pond when power comes back. Water trapped in a dirty filter can become oxygen-depleted and chemically degraded. If an outage has been long enough for the filter to become stagnant or foul-smelling, blindly pumping that water straight back to the koi may be unwise.

High-value systems should therefore include emergency aeration, backup power where practical, alarms, bypass or drain options and a restart procedure. Redundancy protects both fish and the biological treatment process.

19. Medication, Salt and Other Treatments: Think About the Biofilter Too

Whenever a pond is treated, the keeper should consider two separate questions: is the treatment safe for the koi at the intended dose, and is it compatible with the biological filter? Some medications, oxidising treatments and disinfectants can suppress microbial activity or damage biofilm, especially at high concentrations or repeated exposure.

The effect depends on the compound, dose, duration, water chemistry and filter design, so blanket statements such as 'all medication kills the filter' or 'the filter is never affected' are both unreliable. Follow product-specific guidance and monitor ammonia and nitrite more closely after significant treatments. A temporary reduction in feeding can also reduce nitrogen loading while the filter is recovering, provided fish health and treatment requirements allow it.

Salt at ordinary husbandry concentrations is generally not used as a method to improve nitrification itself. It may be used in particular fish-health or nitrite-management contexts, but it does not replace a functioning biofilter and should not distract from correcting the underlying biological or hydraulic problem.

20. Nitrate Is the End of Nitrification, Not the End of Water Quality Management

Conventional aerobic biological filtration does not remove nitrogen from the system completely. It moves the nitrogen toward nitrate. Nitrate is usually much less acutely toxic than ammonia or nitrite at concentrations encountered in well-managed freshwater systems, which is why nitrification is so valuable. But nitrate can still accumulate in recirculating ponds.

Long-term nitrate control may involve water changes, plant uptake in appropriate systems, specialised denitrification, lower feed loading, removal of nitrogen-rich solids before decomposition, or combinations of these approaches. The best choice depends on the pond design and husbandry goals.

This distinction is important when advertising or evaluating a filter. A very effective nitrifying biofilter can actually produce nitrate rapidly because it is successfully oxidising ammonia and nitrite. That does not mean the unit also removes nitrate.

21. Special Considerations for Dealer, Quarantine and High-Stock Systems

Dealer ponds and quarantine tanks are much less predictable than a stable garden pond. Biomass can double in a day when a shipment arrives, fish may be stressed and producing abnormal waste, feeding may change rapidly, and systems are often run at high density. Biological filtration must therefore be designed around peak plausible loading, not only normal average occupancy.

A centralised biofilter serving many tanks can be efficient, but it also connects those tanks biologically. This has biosecurity implications because water is shared. Separate quarantine circuits reduce cross-contamination risk but require each circuit to have enough independent biological capacity. Seeded media can help establish a quarantine filter, but media moved between systems can also transfer organisms, so biosecurity should determine the procedure.

Build for recovery, not only steady state

Commercial systems benefit from spare aeration, redundant pumps, bypass lines, isolation valves, easy drain access and extra mature media capacity. The objective is not merely to achieve zero ammonia on a quiet day; it is to remain controllable when a pump is serviced, a shipment arrives, a screen blocks, a power event occurs or feeding is temporarily increased.

22. Common Biological Filtration Myths

Myth 1: Clear water means the biofilter is working

Clarity is largely a solids and algae issue. Ammonia and nitrite are dissolved and invisible. Always confirm biological performance with water testing.

Myth 2: More media always means more capacity

Extra media helps only if water, oxygen and waste reach an active surface. Poorly distributed or clogged media may contribute little despite occupying a large chamber.

Myth 3: A filter becomes fully mature after a fixed number of days

Cycling time varies. The meaningful endpoint is stable ammonia and nitrite control at the intended load, not a date on the calendar.

Myth 4: Nitrifying bacteria live only inside the filter

Nitrifiers can colonise many wet system surfaces, including pipework and pond walls. The dedicated biofilter is important because it deliberately concentrates favourable habitat and flow, but it is not the only place microbes exist.

Myth 5: Nitrosomonas and Nitrobacter are the only important nitrifiers

Those genera remain useful teaching examples, but molecular research has shown more diverse communities, including ammonia-oxidising archaea and Nitrospira, in freshwater recirculating biofilters.

Myth 6: A biofilter removes nitrate

Standard aerobic nitrification produces nitrate. Nitrate removal requires export, uptake or a separate nitrogen-removal process such as denitrification.

23. Practical Biofilter Health Checklist

• Ammonia and nitrite remain controlled at the normal peak feeding rate, not only during fasting.
• pH is stable and alkalinity/KH is monitored often enough to detect declining buffer reserve.
• The biological stage receives strongly oxygenated water and has dependable aeration or gas exchange.
• Mechanical filtration removes solids before they are allowed to accumulate heavily in the biofilter.
• Water is distributed through the full media volume without obvious bypass, channeling or stagnant zones.
• Flow rate is measured or realistically estimated rather than assumed from a pump label.
• Media is kept biologically active but is not allowed to become packed with sludge.
• Heavy cleaning is performed in a way that preserves enough mature biofilm to maintain function.
• Temperature, feeding and fish biomass are considered together when interpreting test results.
• There is biological safety margin for fish growth, seasonal feeding changes and short-term loading spikes.
• Power failure and aeration failure have an emergency response plan.
• After medications, major maintenance or stocking changes, ammonia and nitrite testing frequency is increased.
• Nitrate is managed separately rather than assumed to disappear through ordinary nitrification.

24. A Better Way to Think About Biological Capacity

The most useful mental model is to treat the biofilter as a living reactor. Its job is to process a continuous mass of nitrogen, not merely to contain a certain volume of plastic. Capacity comes from the interaction of active biomass, effective media surface, oxygen transfer, alkalinity, temperature, hydraulics and waste loading.

This explains why two apparently identical filters can perform very differently on two ponds. One may receive pre-screened, oxygen-rich water at a stable pH and sensible flow. The other may be buried in organics, poorly aerated, hydraulically bypassed and fed by a pump that delivers half its claimed flow after head loss. The equipment name is the same; the biological environment is not.

For serious koi keeping, the strongest design question is therefore not 'How many litres of media do I need?' in isolation. It is: what peak nitrogen load must be processed, how will this reactor keep enough active biofilm supplied with oxygen and alkalinity, how will solids be prevented from overwhelming it, and how will I verify performance under real feeding conditions?

Conclusion

Biological filtration is the living foundation of a stable koi pond. Koi metabolism and decomposing organic matter introduce ammonia into the water. Aerobic nitrifying communities living mainly in biofilm oxidise that ammonia toward nitrite and then nitrate. In doing so they consume oxygen and alkalinity, depend on stable water chemistry, and require usable surface and reliable flow.

Understanding those requirements changes how filtration is designed and maintained. It explains why mechanical solids removal should protect the biofilter, why pH and KH belong in filtration management, why a power failure can affect bacteria as well as fish, why new filters take time to mature, and why feed load is often more useful than pond volume when thinking about biological demand.

Most importantly, biological filtration should be judged by function. A successful biofilter is one that keeps ammonia and nitrite controlled at the pond's real operating load with adequate safety margin. Media colour, chamber size and marketing surface-area figures are secondary to that measurable outcome.

The next article in this series, FL-004, will move from biology into engineering and explain how to size a filtration system for a koi pond using pond volume, flow, fish biomass, feed load, mechanical capacity, biological capacity and realistic safety margin.