Stability and Failure

Every aquarium that holds together does so for the same reason: production and consumption are balanced at every level of the food web. Algae produce biomass and oxygen; grazers consume both. Bacteria break down dead matter and release nutrients; producers take those nutrients back up. When any link in that chain gets too strong or too weak relative to the others, the system starts drifting toward failure — and the warning signs are usually readable long before the collapse becomes irreversible, if you know where to look.

This page is about what stability looks like, what buffers protect it, and the handful of shapes that failure actually takes in the simulator. Everything described here is enforced by the model — there is no "stability score" baked in, just a network of physical and biological processes that succeed or fail to keep each other in check.

What a stable tank looks like

A stable ecosystem is one where every species persists and no quantity grows without bound. Populations may oscillate — grazers eat down the algae, algae recover, grazers recover — but those oscillations stay within a bounded range. Concretely you are looking for:

Dissolved oxygen above about 2 mg/L even at the morning low. Below that, hypoxia stress kicks in for most animals; below about 0.5 mg/L is mass mortality territory.
pH between roughly 6.5 and 9.0. Outside that band stress mortality ramps up for nearly every species in the catalog.
Nutrients cycling, not accumulating. Ammonium, nitrate, and phosphate should move — rising as organic matter decomposes, falling as producers take them up — rather than trending steadily in one direction.
All species still present at the end of the run. Once a population drops below its extinction threshold, growth shuts off and the species can only decline. Even if conditions later improve, it won't come back.
Detritus not piling up. Dead organic matter should decompose at roughly the rate it is produced. A steadily climbing detritus pool means the recycling side of the loop has fallen behind.

A tank that meets all five of these is doing the same thing a healthy pond does: turning the same atoms through producers, consumers, and decomposers over and over, with the major pools cycling instead of running away.

What buffers a tank against instability

Some of the most important features of a stable aquarium are not the steady-state values themselves but the buffers that resist disturbance. The simulator includes several:

Chemical buffering. Alkalinity is the water's capacity to absorb acid or base without changing pH. A tank at 2.5 mmol/L (moderately buffered freshwater) can soak up a lot of overnight CO2 before pH meaningfully drops. Very soft water below 1 mmol/L can swing by more than a full pH unit between dusk and dawn — same biology, far more dangerous chemistry.

Gas exchange with the atmosphere. In an open-top tank, oxygen and CO2 exchange across the surface, and the atmosphere acts as a nearly infinite reservoir. This single feature is what makes most aquariums forgiving — even a high-respiration tank gets oxygen back from the air overnight. A sealed jar has only the headspace to draw on, and total system oxygen is conserved rather than replenished, so the buffer is dramatically smaller.

Density-dependent feedbacks. At high population densities, consumers run into self-limiting mortality — crowding stress on Daphnia, cannibalism on copepods, reproduction suppression on cherry shrimp under poor water quality. These feedbacks prevent populations from overshooting so far that they destroy their own food supply. Every consumer in the model has at least one such mechanism; see Consumers for the per-species details.

Dormant pools. Daphnia, copepods, and ostracods all carry a dormant fraction — ephippia and resting cysts that sit metabolically inert in the substrate. The dormant pool is invisible to predators and exempt from every active-pool mortality kernel: copper, hypoxia, ammonia, nitrite, temperature, pH, hydrogen sulfide. A Cu pulse that kills 99% of the active population can leave the dormant pool at 97% of its pre-event mass, ready to re-hatch when conditions recover. This is what lets a tank with a healthy egg bank survive shocks that would permanently extinguish a single-pool population.

Biofilm maturity. As surfaces in a tank are colonized by bacteria and algae, an EPS (extracellular polymeric substance) scaffold builds up over months. A mature biofilm provides structural refugia that shelter nitrifiers from grazing and light, hide bacteria from their bacterivores, and protect surface-attached prey from being scraped off by snails and shrimp. A young tank with biofilm maturity near zero offers very little of this protection, which is one of the underlying reasons new aquariums are notoriously unstable — the structural habitat hasn't developed yet, even when the species are all present on paper.

Macrophyte storage quotas. Aquatic plants and algae don't take up nutrients in real time at the rate they grow — they store excess N and P internally when supply is good and draw on that store when supply runs dry. A planted tank can sail through a week of low dissolved nitrogen without the plants slowing down, because their internal Droop quotas carry them. This is a major reason mature planted tanks tolerate missed fertilizer doses that would crash a younger system.

Allelopathic interference. Established macrophytes leak polyphenols into the water column that suppress competing phytoplankton, while cyanobacteria can do the reverse with cyanotoxins. These chemical effects are one of the canonical drivers of alternative stable states in shallow lakes — a macrophyte-dominated clearwater state and a phytoplankton-dominated turbid state can both be self-stabilizing on the same nutrient load, with the dominant primary producer chemically suppressing the other.

Starvation feedback. When food runs low, consumers receive less energy than they need for maintenance and starvation mortality activates. This is self-correcting in the same way crowding is self-correcting: overpopulation produces food scarcity, food scarcity reduces the population, and the population recovers along with its food.

How tanks fail

Despite all of the above, ecosystems do fail, and the failures tend to take one of a handful of recognizable shapes.

The algae crash

This is the classic predator-prey catastrophe. Grazer biomass climbs, total grazing rate overshoots algae growth, and algae biomass falls. Grazers don't immediately starve — most consumers eat on a Holling Type II curve and can keep eating efficiently down to surprisingly low food concentrations — and meanwhile the algae cross their extinction threshold and stop being able to recover. With no producers left to support them, the grazer population finishes itself off through starvation over the following days to weeks. The result is a dead tank: no algae, no grazers, just detritus quietly breaking down.

Daphnia are particularly dangerous on this front because their half-saturation constant for algae is low (around 0.3 mg C/L) — they can graze algae down to almost nothing before their own intake meaningfully slows. The cleanest way to avoid the crash is to let producers establish first: introduce grazers a couple of weeks after the algae, so the system enters the predator-prey dance with a producer-heavy head start.

You can also see this one coming. Watch for grazer biomass rising while algae biomass is falling, especially with algae already below ~0.3 mg C/L. By the time you notice it the system may already be past saving, but the trajectory is usually visible.

Oxygen collapse

Oxygen in the water comes from two sources: photosynthesis during the light period, and gas exchange with the headspace or atmosphere. It is consumed continuously by respiration and decomposition. Anoxia sets in when consumption outpaces supply for long enough, and the day/night cycle is central to that calculus — photosynthesis tops the oxygen pool up during the day, respiration pulls it down all night, and if the nighttime deficit exceeds the daytime surplus, oxygen trends down across successive 24-hour cycles.

Several things tip the balance toward deficit. Too much biomass means too much respiration. Short photoperiods, self-shading from dense planktonic algae, or a thick floating plant canopy can reduce photosynthesis below what respiration demands. Refractory DOM builds up over weeks of decomposition and yellows the water, cutting light penetration further. And a pulse of dead biomass from any other crash creates a surge of oxygen demand as bacteria decompose it.

Once O2 drops below about 2 mg/L, hypoxia mortality activates — at species-specific thresholds (Daphnia at ~2.2 mg/L, snails at ~1.5 mg/L, shrimp at ~1.0 mg/L). The dying biomass becomes detritus that needs more oxygen to decompose, which lowers oxygen further, which kills more animals. This positive feedback loop can take a tank from mildly low oxygen to flatly anoxic in a matter of hours.

Open-top tanks have the atmosphere as a near-infinite oxygen reservoir and almost never get into this loop. Sealed jars have only the headspace and a finite total — and if you only look at dissolved O2, you can mistake redistribution into the headspace for actual loss. Always check total system O2 (water + headspace) when diagnosing.

pH runaway

pH in the simulator is not an independent dial — it is a derived quantity computed from dissolved inorganic carbon and alkalinity using carbonate equilibrium. It moves whenever the balance between CO2-removing processes (photosynthesis) and CO2-adding processes (respiration, decomposition, nitrification) shifts.

Upward runaway happens when photosynthesis dominates. Algae and macrophytes pull CO2 out of the water faster than respiration can replace it, and pH climbs. In severe cases it can cross 9.0 or even 10.0 — above the comfort zone for nearly every animal. High pH also makes ammonia far more dangerous: at pH 7.0 only ~0.6% of total ammonia is in the toxic NH3 form, but at pH 9.0 that fraction jumps to roughly 36%. The same ammonia reading is two orders of magnitude more lethal at high pH.

Downward runaway is the mirror case. Heavy decomposition and active nitrification produce CO2 and hydrogen ions faster than photosynthesis can offset them, and pH falls — potentially below 6.0 in a dark or organic-heavy system. Soft water is more vulnerable in both directions.

You can spot pH runaway in the same way as oxygen collapse — by tracking the trend, not the single reading. If the daily peak pH is climbing each day, the tank is trending alkaline. If both peak and trough are dropping, it is trending acidic.

Nutrient depletion

Producers need dissolved nitrogen, phosphate, and (for diatoms) dissolved silica to grow. When any of these runs out, growth stops. In a heavily producer-loaded tank where recycling can't keep up with uptake, dissolved N can trend toward zero and the producers starve. Phosphorus is harder to deplete in absolute terms because its half-saturation constants are tiny, but it runs out quickly once depletion starts because it's needed in such small quantities.

Silica depletion is the classic story behind the early "brown algae" bloom in a new aquarium. Diatoms consume about 1 mol Si per mol N as they grow, so the molar Si:N ratio of the water determines whether diatoms can exhaust silica before nitrogen runs out. Plain tap water sits around Si:N ~4 — not enough nitrogen to support a Si-depleting bloom — but a stocked tank with fish food coming in lifts dissolved N well above tap-water levels and tips the ratio below 1. Diatoms then bloom vigorously and crash when DSi drops below about 0.05 mg/L. Dead frustules dissolve slowly (half-life ~10 days at 20 °C), but the recovery is too gradual to prevent the crash. Silica-independent green algae and periphyton move into the vacated surfaces afterward — textbook ecological succession.

Detritus and substrate going sour

Bacteria and fungi decompose dead matter back into dissolved nutrients, but this decomposition can fall behind production for a few reasons: low oxygen (anaerobic decomposition runs at about 15% of the aerobic rate), low decomposer biomass (settled detritus is partly biotic — heterotrophic bacteria and aquatic fungi have to be there in numbers), or cold temperatures (Q10 ≈ 2 cuts the rate in half per 10 °C drop). Accumulated detritus is a nutrient time bomb: when conditions eventually allow rapid decomposition again, the simultaneous release of ammonium and phosphate can drive an algal bloom and the oxygen consumed during the decomposition pulse can drive a hypoxia event.

A related and more insidious failure mode is the substrate going sour — what aquarists have historically called "old tank syndrome." When organic detritus accumulates faster than aerobic decomposition can keep up, the substrate goes anoxic and the diagenetic ladder of anaerobic metabolisms steps in: denitrifiers, then iron reducers, then sulfate reducers, then methanogens. Each guild produces a less benign end product than the one above it. Iron reducers release dissolved iron and the phosphate it was carrying, driving the internal P loading pulses that aging planted tanks are famous for. Sulfate reducers produce sulfide, which the iron-sulfide sediment buffer absorbs for a while — and then stops absorbing when it saturates, releasing free hydrogen sulfide that is acutely toxic at sub-ppm levels. Methanogens produce methane, which can collect as headspace gas in sealed scenarios or vent to the atmosphere in open ones. None of these are necessarily fatal, but they reshape the tank chemistry around them.

Light starvation

Light is the energy source for every producer in the tank, and several mechanisms can drop it below the level that sustains photosynthesis. Self-shading from dense planktonic algae attenuates light exponentially with depth (Beer-Lambert) and can darken even a shallow container. A floating plant canopy at full coverage blocks more than 95% of incident light, shading out everything submerged below it. Refractory DOM accumulates over weeks of decomposition and yellows the water with humic substances, measurably reducing light availability. The three attenuators stack multiplicatively — a tank with moderate plankton, some floating cover, and mature tea-stained water can be much darker than any one factor would suggest.

Reading the signs

Quick health check

A handful of readings give you a fast sense of whether a tank is on or off the rails:

Dissolved oxygen. Above 4 mg/L is comfortable; 2–4 mg/L is concerning; below 2 mg/L is critical.
pH. Between 6.5 and 8.5 is the safe zone for most species. Above 9.0 or below 6.0 means the chemistry is drifting.
Algae. Present and holding steady. A steady descent toward nothing is a crash underway.
Grazers. Present if you stocked them, and not declining in lockstep with the algae — that pairing usually means overgrazing. A sit-and-wait predator like hydra follows its own dynamics rather than tracking the algae.
Detritus. Stable or slowly cycling. Dead organic matter that climbs and never falls means decomposition has fallen behind production.

Trend, not snapshot

The day/night cycle makes any single reading misleading. A tank at pH 8.9 at noon might be fine; the same value at midnight is a danger signal. Compare values across multiple days rather than within one:

Day-over-day oxygen minima. If the morning low is getting lower each night, the system is losing net oxygen.
Day-over-day pH peaks. Rising daily peaks mean photosynthesis is outpacing respiration and the tank is trending alkaline.
Week-over-week biomass. Short-term oscillations are normal. It is the multi-day average, not any single day, that tells you whether a population is growing, stable, or declining.

Working from the symptom to the cause

When something looks wrong, it helps to work from the whole-tank view inward. Start with the overall trends over several days — which way are oxygen, pH, and the major populations heading? Then zoom in on the oxygen budget across a full 24-hour cycle, because the decisive question is almost always whether nighttime respiration is outrunning daytime photosynthesis. From there, focus on whichever pool is actually misbehaving and ask the two questions that explain any pool: what is adding to it, and what is draining it. Finally, sanity-check that the major elements are still being conserved — nitrogen, carbon, and phosphorus should balance, allowing for the legitimate losses a tank really does have, like nitrogen venting as gas through denitrification or floating plants pulling carbon dioxide straight from the air.

Failure-mode lookup

Sign	Likely cause	What confirms it
Dissolved oxygen falling lower night after night	Net oxygen deficit	nighttime respiration is outrunning daytime photosynthesis over the full cycle
pH climbing past 9.0 on bright afternoons	Photosynthesis dominance	dense algae, strong light, and weakly buffered (soft) water
pH sliding below 6.5	Respiration and nitrification dominance	a heavy detritus load, active nitrifiers, and low alkalinity
Algae fading while grazers hold steady or rise	Overgrazing	grazer intake exceeds algae growth
Algae fading while dissolved nutrients sit near zero	Nutrient depletion	nitrogen, phosphate, or (for diatoms) silica has run out
Grazers dying back while algae stay abundant	Environmental stress	oxygen, pH, temperature, or ammonia outside the animals' tolerances
Detritus piling up steadily	Decomposition bottleneck	low oxygen, too few decomposers, or cold water
A rotten-egg smell and sulfide creeping into the water	Sulfate reducers active, iron-sulfide buffer saturating	an aged, anoxic substrate whose iron buffer is used up
Phosphate spiking just as iron appears in the water	Internal P-loading from iron reduction	an anoxic substrate releasing iron-bound phosphate
Everything declining at once	Anoxia or extreme pH	oxygen near zero, or pH far outside the safe band