Botanicals (leaf litter)

Dried Indian almond leaves, oak leaves, alder cones and guava leaves are standard practice in blackwater and shrimp tanks. They tint the water, they grow the biofilm that shrimp graze, and they are the one input most keepers add without thinking of it as feeding the tank. This page is about what a leaf actually does once it is under water, and how the simulator handles it.

For the other external organic input, see Fish & Feeding. For the tannins themselves, see Allelopathy. For where the carbon ends up, see the Carbon Cycle.

A leaf is a slow-release parcel of carbon

Everything else in a closed tank runs on a fixed budget of atoms. Add a leaf and you have added carbon, nitrogen and phosphorus that were not there before, in a form that takes months rather than hours to become available. That slowness is the whole point. Fish food dissolves and is mineralised within days. A dried leaf sits on the substrate and releases what it has in three overlapping phases, which is the classic description of leaf breakdown in streams (Petersen & Cummins 1974):

1. Leaching. Within the first day or two, the water-soluble fraction simply dissolves out. This is sugars, amino acids, small organic acids, and the tannins that stain the water. It is 10–25% of the leaf's dry mass and it is where the colour comes from.
2. Microbial conditioning. Bacteria and fungi colonise the softened leaf. This is what turns a dry leaf into something worth eating, and it is why shrimp ignore a leaf for the first week and then swarm it.
3. Fragmentation. The conditioned leaf breaks into progressively finer particles, which are grazed, decomposed and eventually mineralised.

The nutrients follow the same order. The nitrogen and phosphorus in a leaf are locked in structural tissue, and they come out slowly.

Where a leaf goes in the model

A botanical addition is a dose, in exactly the sense that iron or potassium dosing is a dose. It introduces no new pools. The dry mass you add is partitioned at the moment it enters, across four pools the simulator already tracks:

Carbon, nitrogen and phosphorus are each split as fractions of that element's own total, with the detritus slot taking the remainder. The three elements are therefore conserved by construction rather than by calibration, and analysis_tools/diagnose_mass_balance.py reads 0.00% drift on all three for a sealed litter jar (scenarios/botanicals_mass_balance_sealed.yaml).

Once partitioned, nothing about a leaf is special. The detritus decomposes through the ordinary decomposition path, the refractory DOM attenuates light the way all refractory DOM does, which is what makes the tint a real optical effect rather than a cosmetic one, and the polyphenol decays and suppresses producers through the existing allelopathy kernels.

Why the leaf body is high in carbon relative to nitrogen

Senesced leaf litter has a carbon-to-nitrogen ratio of roughly 40 to 90 (by moles), against about 7 for living algae. A tree resorbs most of the nitrogen out of a leaf before dropping it, so what falls is mostly cellulose and lignin. That ratio matters for the tank: bacteria decomposing carbon-rich material need more nitrogen than the material supplies, so they take the difference from the water. Adding leaf litter can lower your nitrate, at least at first, and that is a real predicted effect rather than a modelling artifact. It is the same immobilisation that makes gardeners warn against digging fresh wood chips into soil.

The four botanicals

The four types differ in composition and in the dry weight of one piece. They do not differ in how fast they break down, because they share one settled-detritus pool. Oak really is slower to break down than guava in a real tank, and the simulator does not capture that.

Alder cones carry the most tannin per gram, which matches their reputation, and guava the least. Guava also has the lowest carbon-to-nitrogen ratio, so it immobilises the least nitrogen and conditions fastest, which is why it is the shrimp keeper's leaf rather than the blackwater keeper's.

Leaves are released steadily, not dumped

You tell the simulator how many leaves you add and how often. What it applies is a continuous release of that mass across the interval, not a single pulse on the day you add them.

This is deliberate, and it is closer to what a litter bed does. A catappa leaf takes two to four months to break down. A bed that is topped up every three weeks is at quasi-steady state, releasing at roughly the rate it is being replenished. Dosing the whole addition at once instead produces a four-day event that has nothing to do with leaf litter: in a 35 L test tank, ten grams applied as a pulse took dissolved oxygen from 8.1 to 0.8 mg/L within four days and back to 7.9 before the next addition. That is an artifact of the delivery.

The cost of the simplification is the transient. A real tank visibly darkens within a day of adding leaves and then fades over the following weeks. Here the tint is steady. Over a run measured in months, the equilibrium is what matters, and the equilibrium is the same either way.

What leaf litter does to a tank

Measured on a real 35 L open-topped, gravel, planted betta tank, averaged over days 100 to 200 of a 365-day run, adding Indian almond leaves against an otherwise identical control:

Three things are worth reading off that table.

A light litter bed grows the invertebrate population by half. One or two leaves every three weeks gives a continuous supply of grazeable detritus where the control tank has almost none. The snails and shrimp eat better and there are more of them, and the effect lasts: at the end of the year the litter tank still carries three times the grazer biomass of the control, which by then has largely collapsed.

Leaf litter lowers pH, and it does so by respiration. All that carbon is eventually breathed back out as carbon dioxide, and dissolved carbon dioxide is an acid. This is a genuine effect and it is the right order of magnitude. It is not the same mechanism as the acidity of the tannins themselves, which the model does not yet handle (see below).

Past a threshold, the tank cannot oxidise what you are adding. Ten grams of dry leaf per three weeks in 35 litres is more organic carbon than that tank's gas exchange can supply oxygen for, so oxygen falls to near zero and stays there. The invertebrates die, and the leaf litter stops breaking down and piles up instead. If you keep a heavily-planted tank with a closed floating canopy, which throttles gas exchange further, that ceiling arrives sooner.

There is a fourth effect worth knowing about, which is that the pH benefit fades. In a planted tank the extra nitrogen and phosphorus from the litter go into plant growth, and the growing plants take the extra carbon dioxide back out. In the run above, the one-leaf tank sat around pH 7.8 through days 75 to 100 and had drifted back to 8.3 by day 175. Leaf litter buys more in a tank whose plants are already nutrient-saturated than in one that was starving.

The useful rule from this is a loading limit, not a leaf count. Up to about 0.1 g of dry botanical per litre per month is a light bed the tank absorbs comfortably. Around 0.2 g/L/month you are trading oxygen for tint. Above that, in a tank without strong surface agitation, you are running an oxygen deficit. The setup wizard shows you this number as you choose, so you can size a dose to your tank rather than copying a leaf count from someone with a bigger one.

One more measured consequence: at the heaviest dose above, nitrate fell from 9.4 to 0.3 mg/L. That is the nitrogen immobilisation described earlier, and it is large enough to be a tool. If you are using leaves partly to hold nitrate down, this is the mechanism.

Leaves suppress cyanobacteria more than they suppress algae

Decomposing leaf litter is a long-standing pond treatment for nuisance algae, and so is barley straw, which works the same way (Ridge et al. 1999). What both release is polyphenols, and the useful thing about polyphenols is that they are selective. Cyanobacteria are the sensitive group. Green algae and diatoms are comparatively spared, which is what makes leaf litter a treatment rather than a steriliser.

The simulator carries that selectivity. Half of a producer's photosynthesis is blocked at roughly 1 mg/L of dissolved polyphenol for cyanobacteria, 2 for diatoms, 3 for planktonic green algae and 4 for surface-attached green algae, all measured as gallic acid equivalents. Attached algae are the least affected because the biofilm they sit in is itself a diffusion barrier. The thresholds come from laboratory dose-response work on the individual compounds a plant releases (Nakai et al. 2000), and the simulator reports its polyphenol pool in the same unit, so you can read the column against them directly.

A word of caution on scale. A light litter bed holds a few tenths of a milligram per litre, which is well below all four thresholds. If you are adding leaves hoping to knock back blue-green algae, the dose that would do it in this model is a heavy one, and heavy doses cost oxygen.

What this feature is not

It is not the fix for the simulator's carbon dioxide gap. The engine holds dissolved carbon dioxide at roughly 1 to 3 times atmospheric where real stocked tanks measure 7 to 12 times, so simulated pH reads high across the board, in tanks with botanicals and tanks without. Leaf litter is a legitimate missing carbon input and it moves pH in the right direction, but the underlying undersaturation is a separate calibration problem and a more important one. Do not read a pH improvement here as that problem being solved.

Organic acidity is deferred, and it is a small term at aquarium hardness. Tannic and fulvic acids are weak acids, and they do drag pH down independently of any carbon dioxide the litter produces. The carbonate solver works from total alkalinity and dissolved inorganic carbon only, so representing them needs a third acid-base system in it, and that is not done. Everything the pH does on this page is carbon dioxide.

It is worth knowing how much this costs you, because the answer is less than the blackwater reputation suggests. Aquatic humic substances carry about 10 microequivalents of acid per milligram of their carbon. The two-leaf tank above holds 26 mg/L of humic carbon, which is 0.26 milliequivalents per litre against a carbonate buffer of 2.50, so it would move pH by about 0.05. Titrating enough alkalinity away to reach pH 7.5 on organic acid alone would take roughly 205 mg/L of humic carbon. Blackwater is a soft water phenomenon: leaves acidify a tank with almost no carbonate hardness, and a tank at 7 dKH mostly shrugs them off. If you want the pH drop, lower your hardness first.

Only a tenth of the tannin counts as an active allelochemical. The rest is booked as refractory DOM, where it browns and shades the water. That split is a claim about the chemistry of a dried leaf. A brown leaf has already oxidised and polymerised most of its phenolics, and what leaches is dominated by large hydrolysable and condensed tannins and by humic complexes. The standard total-phenolics assay counts all of it as gallic acid equivalent, but only the small free low-molecular-weight fraction behaves as a free allelochemical, and that is the fraction the published inhibition experiments isolated and measured. Ten percent is an order-of-magnitude estimate, and it is on the generous side for a hard-water tank, because polyphenols lose potency as pH rises and the monomers polymerise.

A leaf is not a surface. Real leaf litter adds a large amount of colonisable area, and much of what shrimp eat off a leaf is the biofilm growing on it rather than the leaf. Here the leaf becomes detritus and the biofilm grows on the surfaces the scenario already declares. A litter bed that carried its own surface area would grow more biofilm than this. In a heavily planted tank with a closed floating canopy the difference is small, because a litter bed on the bottom gets almost no light and grows almost no algae in any case, but in an open tank it is real.

The litter bed does not accumulate. A leaf lands in the same settled-detritus pool as everything else that dies in the tank, and that pool turns over at roughly 8% a day. Real Indian almond leaves take two to four months to break apart. So the simulator gets the rate at which leaf carbon enters the tank right and the amount of leaf sitting on the substrate badly wrong: it releases the nutrients on schedule but never builds the visible litter bed. Over a run measured in months this does not change the chemistry, because at steady state a tank mineralises what you put in regardless of how long each leaf takes. It does mean the model understates how much grazeable material a litter bed holds, and it is why the four botanical types cannot differ in breakdown speed.

Using it

In the setup wizard, botanicals live on the Maintenance step, alongside water changes and fertiliser. Choose a type, how many pieces you add, and how often. The wizard shows the resulting load in grams per litre per month and tells you whether the tank can absorb it.

In a scenario file:

botanicals:
  type: indian_almond      # indian_almond | oak_leaf | alder_cone | guava_leaf
  pieces: 3                # how many leaves or cones per addition
  grams_each: 2.0          # optional; defaults to the type's typical piece weight
  interval_days: 21        # how often you top them up
  start_day: 0             # first addition
  end_day: null            # optional; stop adding after this day

Use grams_dry_per_addition instead of pieces if you weigh your litter.

There is no per-leaf tracking and no standing litter pool in the state vector. What you see in the output is the leaf's effect: settled detritus rising, refractory DOM accumulating and shading, dissolved organic carbon feeding bacteria, and the carbon dioxide and oxygen consequences of respiring all of it.

References

Petersen, R.C. & Cummins, K.W. (1974). Leaf processing in a woodland stream. Freshwater Biology 4, 343–368.

Ostrofsky, M.L. (1997). Relationship between chemical characteristics of autumn-shed leaves and aquatic processing rates. Journal of the North American Benthological Society 16, 750–759.

Gessner, M.O., Chauvet, E. & Dobson, M. (1999). A perspective on leaf litter breakdown in streams. Oikos 85, 377–384.

Webster, J.R. & Benfield, E.F. (1986). Vascular plant breakdown in freshwater ecosystems. Annual Review of Ecology and Systematics 17, 567–594.

Aerts, R. (1997). Climate, leaf litter chemistry and leaf litter decomposition in terrestrial ecosystems. Oikos 79, 439–449.

Chyau, C.-C., Tsai, S.-Y., Ko, P.-T. & Mau, J.-L. (2002). Antioxidant properties of solvent extracts from Terminalia catappa leaves. Food Chemistry 78, 483–488.

Salminen, J.-P. & Karonen, M. (2011). Chemical ecology of tannins and other phenolics. Functional Ecology 25, 325–338.

Maie, N., Pisani, O. & Jaffé, R. (2008). Mangrove tannins in aquatic ecosystems. Limnology and Oceanography 53, 160–171.

Bärlocher, F. & Graça, M.A.S. (2005). Total phenolics. In Methods to Study Litter Decomposition, Springer, 97–100.

Nakai, S., Inoue, Y., Hosomi, M. & Murakami, A. (2000). Myriophyllum spicatum-released allelopathic polyphenols inhibiting growth of blue-green algae Microcystis aeruginosa. Water Research 34, 3026–3032.

Laue, P., Bährs, H., Chakrabarti, S. & Steinberg, C.E.W. (2014). Natural xenobiotics to prevent cyanobacterial and algal growth in freshwater: contrasting efficacy of tannic acid, gallic acid, and gramine. Chemosphere 104, 212–220.

Ridge, I., Walters, J. & Street, M. (1999). Algal growth control by terrestrial leaf litter: a realistic tool? Hydrobiologia 395/396, 173–180.

Emilson, E.J.S. et al. (2018). Climate-driven shifts in sediment chemistry enhance methane production in northern lakes. Nature Communications 9, 1801.

Oliver, B.G., Thurman, E.M. & Malcolm, R.L. (1983). The contribution of humic substances to the acidity of colored natural waters. Geochimica et Cosmochimica Acta 47, 2031–2035.