Soil Organic Matter and Pore Water Diffusion

For the overview of decomposition and the water-column recycling loop, see Death and Decomposition.

Why buried soil changes everything about a planted tank

A layer of organic soil beneath a sand cap transforms the nutrient economy of the whole system. Instead of leaning on fish food, water-column fertilizers, or the slow accumulation of detritus, a soil substrate acts as a slow-release reservoir that feeds rooted macrophytes directly through their roots for months to years. The nutrients never pass through the water column at all — they diffuse out of decomposing soil into the water trapped between the grains, and roots intercept them before they reach the surface. This is the mechanism that lets a Walstad-style planted tank run without external dosing.

The water held inside the substrate is also chemically a world apart from the water above it. It builds up dissolved CO₂ at concentrations ten to fifty times higher than the open water, giving rooted plants a private carbon supply that bypasses the pH-driven carbon limitation algae have to live with. It runs its own miniature nitrogen cycle, complete with coupled nitrification-denitrification that can strip nitrogen out of the system entirely as N₂ gas. Understanding the soil and its pore water is most of the reason planted tanks behave so differently from bare-bottom setups.

Soil organic matter: the slow-release engine

Some scenarios include a buried organic substrate — the approach Diana Walstad popularized for low-tech planted aquaria. The model treats this substrate as two pools that behave very differently from the suspended and settled detritus described in the parent page.

Two soil fractions

The labile fraction is the freshly processed material — the manure and compost end of the spectrum, with relatively short carbon chains that microbes can get at. It turns over slowly compared with anything in the water column: a fraction of a percent per day, against roughly twelve percent a day for suspended detritus — but it is steady, releasing meaningful nutrient fluxes over the first weeks and months after setup.

The refractory fraction is the condensed, humified material — the humic substances and lignin polymers of aged peat or composted bark. It breaks down roughly twenty times slower again, which is exactly what makes it a multi-year reservoir: it is the slow burn that keeps a Walstad tank fed long after the labile fraction is spent.

Both fractions carry carbon, nitrogen, and phosphorus, and as each one mineralizes its stoichiometry is preserved — nitrogen and phosphorus are released in step with the carbon, so the elemental ratio of the soil pool holds steady over time.

Where the products go: pore water

Unlike water-column decomposition, which dumps its products straight into the open water, soil mineralization releases everything into pore water — the interstitial water trapped between soil particles. Ammonium and phosphate accumulate there from ongoing mineralization (and from the initial dissolved load described below). Nitrate builds up too, produced by nitrification within the thin oxygenated skin of the sediment, and it becomes the substrate for sediment denitrification (the coupled nitrification-denitrification pathway; Nielsen 1992). Just as in the open water, a transient pulse of nitrite appears in the pore water whenever the ammonia-oxidizers there run ahead of the nitrite-oxidizers — the in-substrate echo of the new-tank nitrite spike, settling back toward zero in a mature bed.

Two more pore-water quantities matter. Dissolved oxygen in the substrate is its own thing entirely, distinct from the well-oxygenated water above: it is supplied by slow diffusion down through the sand cap and by oxygen leaking from plant roots, and drained by mineralization and re-oxidation reactions, so in any real planted substrate it settles far below the bulk value. And a short-lived dissolved-organic intermediate forms when labile soil matter is first hydrolysed, before it fully mineralizes — this is the substrate the anaerobic microbes living in the pore zone actually feed on (more on that below).

DIC — the dissolved CO₂ released by mineralization — collects in pore water too, where rooted plants can draw it directly through their internal air channels (aerenchyma; see Macrophytes — Pore Water CO2 Access) in addition to its diffusing slowly upward.

The initial dissolved pulse. Fresh substrates also carry nutrients that are already dissolved, released within hours of submersion — far faster than anything has to mineralize. Commercial potting mixes hold slow-release fertilizer granules that dissolve almost at once underwater, and garden soils carry exchangeable ammonium on their peat and clay that desorbs on contact with water. Together these can push first-day pore-water ammonium into the range of tens of milligrams of nitrogen per litre before any decomposition has happened. The model captures this simply, by starting the soil presets with non-zero pore-water concentrations rather than bolting on a separate fast process — at its timestep the dissolution is effectively instantaneous. Aged or pre-fired substrates start near zero, because their soluble fraction was leached or destroyed in processing.

These pore-water nutrients reach plant roots by direct uptake and reach the open water by Fickian diffusion up through the substrate, described in the next major section.

What controls the rate

Temperature. Soil microbial activity roughly doubles for every 10 °C of warming — the same Q10 behaviour as water-column decomposition (Brady & Weil 2008; Reddy & DeLaune 2008).

Oxygen, read locally. Mineralization is faster and more complete when oxygen is present, and the model judges this on pore-water oxygen — the concentration down in the substrate — not the bulk water above. That distinction is what makes the rate physically meaningful: a Walstad jar can sit at a healthy 5–8 mg/L in the open water all day while its pore water falls below half a milligram within days, as the soil's oxygen demand outruns the trickle diffusing in from above. Above roughly 1 mg/L of pore oxygen, mineralization runs fully aerobic; below about 0.2 mg/L it drops to a low floor where only fermenters and methanogens keep working. So the substrate correctly turns anaerobic even while the water column reads "fully oxic" — the split that gives aerobic and anaerobic microbes their separate niches.

Under those anaerobic conditions the carbon book closes differently: about half the carbon flux still leaves as CO₂ while the other half is routed to methane, and that methanogenic branch only lights once both pore nitrate and pore sulfate have been drawn down — the bottom rung of the energy ladder described later on this page.

The dissolved-organic intermediate. When labile soil matter is hydrolysed, the model sends only part of the released carbon, nitrogen, and phosphorus straight to the inorganic pore pools; the rest passes through a short-lived dissolved-organic intermediate first. That intermediate is the operative food for the anaerobic microbes living in the pore zone — the denitrifier and its neighbours — and routing through it is what gives those pore-dwellers a substrate channel of their own. Without it they would have to compete for bulk-water dissolved organics against the much larger water-column bacterial population, and lose. (Refractory matter skips the intermediate: there the rate-limiting step is simply getting an enzyme onto the mineral-protected polymer, and once the bonds break the product is already labile-grade.)

Bacterial stimulation. Heterotrophic bacteria living in and around the sediment secrete enzymes that speed soil hydrolysis. The baseline first-order (abiotic) breakdown always runs — bacteria can only accelerate it, never slow it — and the boost saturates as bacterial biomass grows, in the usual Monod fashion. The labile fraction responds more strongly than the refractory one, because its simpler molecules are easier for enzymes to cleave while humic and lignocellulosic polymers resist attack even when bacteria are abundant. This sets up a self-limiting feedback: bacteria consume the dissolved organics that mineralization releases, grow, and in turn accelerate further mineralization, until the boost saturates. The stimulation coefficients and thresholds are listed in the Parameter Reference (Burns et al. 2013; Wallenstein & Weintraub 2008).

Compared to water-column decomposition

The fundamental differences are timescale and destination:

	Suspended detritus	Settled detritus	Soil OM (labile)	Soil OM (refractory)
Timescale	days	days	months	years
Products go to	water column	water column	pore water	pore water
Releases dissolved organics?	yes	yes	no	no
Bacterial stimulation?	no	yes	yes	yes

Water-column detritus recycles nutrients back to the open water in days to weeks. Soil organic matter releases them into pore water over months to years — which is what makes it the long-term foundation of a planted setup, with the initial dissolved pulse covering the first days while mineralization sustains the supply far beyond that.

References

Walstad, D.L. (1999). Ecology of the Planted Aquarium. Echinodorus Publishing.
Reddy, K.R. & DeLaune, R.D. (2008). Biogeochemistry of Wetlands. CRC Press.
Brady, N.C. & Weil, R.R. (2008). The Nature and Properties of Soils. Pearson.
Burns, R.G. et al. (2013). Soil enzymes in a changing environment. Soil Biology & Biochemistry, 58, 216–234.
Wallenstein, M.D. & Weintraub, M.N. (2008). Emerging tools for measuring and modelling in situ soil extracellular enzyme activity. Soil Biology & Biochemistry, 40, 2098–2106.

Pore water diffusion through the substrate

Once mineralization has loaded the pore water, those nutrients still have to travel up through the substrate to reach the open water, and that journey is slow. They pass through up to three layers stacked in series: the soil matrix itself, where solutes thread their way through a tortuous network of tightly packed particles; the sand cap above it, a second diffusion barrier with its own pore structure; and finally a thin (about half a millimetre) stagnant boundary layer of water clinging to the sediment surface, where ordinary free-water diffusion takes over (Jørgensen & Revsbech 1985). The model adds up the resistance of each layer and lets the concentration difference drive the flux — upward when pore water is richer than the column (the normal case for a mineralizing soil), downward in the rarer case where the column is richer.

Why the layers matter

Stacking the layers in series, rather than only counting the sand cap, makes a real difference. Nutrients produced deep in the soil have to cross the soil matrix before they even reach the sand, and soil is the more tortuous medium — its irregular, tightly packed organic and mineral particles force a more convoluted path than clean sand grains. For a typical Walstad setup the soil layer alone accounts for more than half the total resistance, and including it slows the pore-to-column leak to roughly half what a sand-cap-only model would predict. That extra resistance is precisely what reinforces the substrate's slow-release character and sharpens the competitive edge of rooted plants with direct access to the pore water. The diffusivities, porosities, and tortuosities behind these numbers are tabulated in the Parameter Reference.

What controls the rate in practice

The concentration gradient is the main lever. In a Walstad tank pore-water ammonium can sit at several milligrams of nitrogen per litre while the open water reads a hundredth of that — a steep gradient driving a steady upward leak. As rooted plants draw the pore ammonium down, the gradient near the roots steepens, but the total leak can fall simply because there is less pore ammonium left to move.

Substrate geometry is the dominant structural control. Because the soil and sand cap carry almost all the resistance, making either one deeper slows the flux roughly in proportion — and a deeper soil also lengthens the average path a molecule has to travel within the soil before it even reaches the cap.

CO₂ deserves a special mention. Pore CO₂ can climb to many times the level of the overlying water, which (in an open tank) sits close to equilibrium with the air. That enormous gradient, combined with CO₂'s naturally fast diffusion, drives a substantial carbon flux up from the sediment — a key part of what sustains plant growth in a Walstad setup. Temperature nudges all of this too, though diffusion responds far more weakly to warming than biological reactions do.

Diffusion runs both ways

All of these solutes — ammonium, nitrate, phosphate, CO₂ — can travel in either direction; whichever side is richer drives the flux. Upward is the rule in an active soil, but nitrate in particular can run downward when water-column nitrification produces more of it than the pore water holds.

How roots intercept what diffuses upward

When rooted macrophytes are present, their roots do more than simply drink from the pore water — the zone of soil right around them, the rhizosphere, reshapes the chemistry in ways that throttle the upward leak quite apart from direct uptake. Three things happen at once.

First, the roots are a physical barrier: root tissue lowers the effective porosity and raises the tortuosity of the sediment, mechanically slowing diffusion, and at high root density this becomes the dominant effect.

Second, the roots leak oxygen down through their aerenchyma and out into the otherwise anoxic soil, creating oxidized micro-zones. In those zones pore ammonium is nitrified to nitrate, and that nitrate is then denitrified to N₂ gas by anaerobic bacteria just millimetres away — permanently exporting nitrogen (the coupled nitrification-denitrification pathway; Reddy et al. 1989; Caffrey & Kemp 1992). Even sparse roots create these biogeochemically active pockets.

Third, root exudates feed microbes — the sugars, organic acids, and amino acids that living roots secrete stimulate heterotrophic growth in the rhizosphere, and those microbes lock dissolved nitrogen into their own biomass rather than letting it diffuse away (immobilization; Wigand et al. 1997).

The model rolls these together into a single saturating function of root density that attenuates the upward leak (and only the upward leak — downward diffusion drives solutes into the root zone, where they are wanted, so it is left unimpeded). A newly planted tank with sparse roots intercepts only a small fraction; as the root bed fills in over months, interception climbs toward a ceiling of roughly half the raw flux. Combined with whatever the roots take up directly, this is why mature planted tanks keep such low water-column nutrient levels even over a rich soil. The coefficients are given in the Parameter Reference.

Why it matters ecologically. Rhizosphere interception reinforces the swing from an algae-dominated tank to a plant-dominated one. In a freshly set-up Walstad tank the roots are sparse, so pore-water nutrients — both the initial dissolved pulse and the ongoing mineralization flux — leak freely into the water column and feed algae. As the roots establish, that leak is progressively choked off, starving the water column of pore-derived nutrients while the plants themselves keep feeding directly through their roots. This asymmetric access is a core mechanism of the Walstad method: the plants engineer conditions that favour themselves and suppress their competitors (Walstad 1999; Scheffer et al. 1993).

References

Caffrey, J.M. & Kemp, W.M. (1992). Influence of the submersed plant Potamogeton perfoliatus on nitrogen cycling in estuarine sediments. Limnol. Oceanogr., 37, 1483–1495.
Reddy, K.R., Patrick, W.H. & Lindau, C.W. (1989). Nitrification-denitrification at the plant root-sediment interface in wetlands. Limnol. Oceanogr., 34, 1004–1013.
Scheffer, M., Hosper, S.H., Meijer, M.-L., Moss, B. & Jeppesen, E. (1993). Alternative equilibria in shallow lakes. Trends Ecol. Evol., 8, 275–279.
Walstad, D.L. (1999). Ecology of the Planted Aquarium. Echinodorus Publishing.
Wigand, C., Stevenson, J.C. & Cornwell, J.C. (1997). Effects of different submersed macrophytes on sediment biogeochemistry. Aquat. Bot., 56, 233–244.

The sediment iron reservoir

A soil substrate carries a third buried pool alongside its two organic fractions: a stock of iron(III) oxide — the rust-coloured ferrihydrite and goethite that coats soil particles. Mineral soils and volcanic clays (the basis of many commercial aquasoils) carry a heavy complement, ordinary aged garden soil somewhat less, and pure peat very little.

This iron pool is not part of the organic cycle — it doesn't mineralize, doesn't release nitrogen or phosphorus by decomposition, and doesn't respond to the temperature, oxygen, and bacterial controls above. Instead it runs its own redox cycle across the oxygen gradient through the sand cap. Where the substrate is oxygenated — at the sediment surface, and in the rhizosphere of actively photosynthesising roots — dissolved iron oxidizes and precipitates onto the oxide coating, dragging dissolved phosphate down with it. Where the substrate is anoxic, iron-reducing bacteria burn the pore-water dissolved organics and reduce the oxide back to soluble iron, releasing the iron-bound phosphate into the pore water at the same time.

That phosphate release is the classic internal P-loading mechanism — the phosphate pulse that accompanies every shift from oxic to anoxic sediment, and a big part of why an aging planted tank can develop an algae problem it didn't have when it was young. How much phosphate is in play depends on the soil's mineralogy, which the model sets per preset (mineral soils bind much more iron-associated phosphate than peat does). Rooted plants run a parallel version right at their root surfaces — the rice-paddy pattern — liberating both iron and its phosphate into the pore water where the roots can immediately use them. The full redox chemistry, kinetics, and the "cryptic" cycle that keeps dissolved iron near zero in healthy water are covered in the Iron Cycle.

The diagenetic ladder: anaerobic biology in the substrate

In a real soil-substrate aquarium the carbon book closes through a sequence of microbial guilds, each breathing a different substitute for oxygen once oxygen itself runs out. This is the textbook diagenetic ladder (Reddy & DeLaune 2008): aerobic heterotrophs consume oxygen within millimetres of the cap, denitrifiers below them respire nitrate, then iron-reducers draw on the buried iron-oxide reservoir, then sulfate-reducers turn to pore sulfate, and finally methanogens disproportionate organic carbon directly into methane. Each rung yields less energy than the one above, so the energy yield per unit carbon drops sharply down the ladder.

The model implements these rungs as explicit organisms competing for substrate, rather than as a hardcoded ordering. The ordering instead emerges from inhibition: while nitrate is plentiful the denitrifiers dominate and the lower rungs idle; once nitrate drains, iron-reduction takes over until the oxide reservoir is drawn down; sulfate-reduction establishes only after that; and methanogens fire only when both nitrate and pore sulfate are exhausted. The practical upshot is that a sealed, organic-rich peat substrate collapses through the whole sequence quickly, with methanogenesis becoming the dominant flux within weeks — while a planted Walstad tank, fed a steady trickle of nitrate from nitrification, rarely lights the lower rungs at all.

A handful of related reactions are genuine chemistry rather than biology and stay separate: the re-oxidation of reduced iron, sulfide, and methane where they meet oxygen at the oxic–anoxic boundary, the precipitation of iron sulfide, the bubbling-off of methane, and the oxidation of methane in the water column. Only the reduction steps are biology. The iron-bound phosphate release travels with the iron-reducers, carrying the internal P-loading coupling with it, using the same per-soil mineralogy as before.

Bioturbation: burrowing animals reset the sediment clock

Everything above sets a physical ceiling on how fast the pore water and the open water can equilibrate — molecules can only travel as fast as diffusion through a stagnant porous medium allows. In an undisturbed substrate that ceiling is low: pore-water nutrients build up over weeks while the water column stays lean, and reduced products like iron sulfide and pore sulfide accumulate in the deeper layers indefinitely. Over months in a sealed organic substrate, that build-up of buried iron sulfide and hydrogen sulfide pockets is exactly what aquarists call old tank syndrome — until it breaks through the surface oxic film and crashes the system.

Burrowing animals break that ceiling. The model's first explicit bioturbator is the Malaysian trumpet snail (Melanoides tuberculata); other burrowers like tubificid worms could plug into the same machinery later. When the snails are present and have soft substrate to dig into, they drive three effects whose strength scales with how densely they have colonized the floor.

They ventilate the pore water. Burrow construction and ventilation actively pump pore water through and around the burrow walls, multiplying the effective diffusion rate of every dissolved species — a fully colonized substrate can exchange several times faster than bare sediment, consistent with field measurements of burrowing-fauna irrigation in lake sediments (Mermillod-Blondin 2011). This is the dominant effect: it carries pore ammonium up to where nitrifiers can process it, brings sulfate down, and speeds dissolved methane and sulfide out of the substrate before they can accumulate.

They re-expose buried iron sulfide. The black iron monosulfide of an aging substrate — formed where sulfate-reducers' sulfide meets pore-water iron — is physically lifted by burrowing back into the oxygenated surface film, where it re-oxidizes. Ordinarily that only happens when oxygen slowly penetrates down to the iron-sulfide layer; burrowing fauna bypass the diffusion limit and do it whenever they are active, ventilating the oxygen they need from the water above.

They resuspend settled detritus. Digging kicks fine settled organic matter back into suspension, where it decomposes faster and feeds the planktonic detritivores — a modest effect, calibrated as a small fraction of ordinary physical resuspension so it accelerates recycling without dominating sediment dynamics.

When no burrowing fauna are present — or there is no soft substrate to burrow into — all three effects simply switch off and the substrate behaves as a stagnant, diffusion-limited medium again.