Year 0Mould over the pavement 0 cmBurial 0 mm/yrWorms 0 g/m²FPS 0

Drag to orbit, wheel or pinch to zoom. A 60 cm deep cutaway of grassland (shown at true proportions). A tile pavement lies on the ground at year 0; dark worm-cast mould piles up over it. Worms (drawn on the cut face) crawl faster in warm, moist seasons; small mounds on the turf are fresh casts. The number of worms drawn is capped, and the picture is illustrative.

Darwin's Worm Stone: How Earthworm Casts Bury a Roman Pavement

Earthworms swallow soil below ground and leave it as casts on the surface, so everything lying on the grass slowly sinks. See how fast, and what temperature, wetness and food do to the answer.

Climate of the soil

Worms and food

Clock

Live readouts

Depth of mould over the pavement (cm)

Worm biomass and its food limit (g/m²)

Worm activity through the year

Experiment: how much does food matter?

Runs 100 years from a small worm herd for five litter supplies and reports the worms the land ends up carrying and the burial rate. Prediction: biomass follows K, the burial rate follows the biomass.

Experiment: what does warming do?

Shifts the mean soil temperature by -4 to +6 K and runs 100 years.

Independent numerical checks

Controls

Drag or one finger: orbit; wheel or pinch: zoom.
Sliders act on the running simulation; Reset or a preset restarts at year 0. The clock runs the chosen number of years per second and stops at year 400.

The science

Darwin's observation. In "The Formation of Vegetable Mould through the Action of Worms" (1881) Darwin estimated about 18 tons of dry casts per acre per year on old pasture and watched stones and pavements sink under a growing layer of mould, at a few millimetres a year. This simulation rebuilds that mechanism as a small model.

Activity. Worms are active in a temperature window and a moisture window. The temperature factor rises with a Q10 rule, f_T = Q10^(T−20)/10 up to 20 °C, then falls linearly to zero at 30 °C, and is zero in frozen soil. The moisture factor is f_M = 1 − ((s − 0.6)/0.4)² for relative saturation s, so worms stall in dry or waterlogged soil. Seasonal temperature and wetness are cosines, warmest and driest at day 200. The overall activity is f = f_T f_M.

Population. Biomass B (g of fresh worm per m²) grows logistically towards a food-limited capacity K = κ × litter supply: dB/dt = r f B (1 − B/K) − m B. If activity is too low for births to beat the mortality m, the population dies out.

Casting and burial. Cast mass left on the surface is q f B σ (q = soil eaten per gram of worm per day, σ = share left on the surface). The mould thickness grows by cast mass divided by the density of the deposited mould: dh/dt = q f B σ / ρ. With constant B and f this is a straight line, which the first check verifies; the real curve bends because B and f change.

Honest simplifications: r = 2/yr, m = 0.3/yr, κ = 0.6 g worm per g litter per year, q = 0.3 g/g/day and the shape of the moisture factor are typical illustrative values, not measurements; the temperature optimum of 20 °C and the dead point at 30 °C are generic. Compaction of the mould, mixing of casts into the turf, worm species, burrow structure, predation and frost kill are not modelled. The check against Darwin is an order-of-magnitude comparison with a historical estimate (typical value), not a calibration. The worms and mounds are drawn for illustration and capped in number. Models built with Blender and Three.js.

Educational model; no affiliation.