Presets
Playback
Time
Seasons repeat every 365 days; speed up to see decades.
Plants, flowers and pollinators
Growth, pollination benefit, winter die-back and pollinator survival.
Rabbits and foxes
Feeding efficiency and conversion of food into offspring.
Starting densities
Changing these restarts the meadow on 1 March.
Layers
Model checks
The science
A meadow is a web of interactions. Flowers feed pollinators and pollinators raise the seed set of flowers; rabbits graze the plants; foxes hunt the rabbits; and the seasons switch the whole system on in spring and off in winter. This lab integrates four coupled equations (per hectare, time in days) with a 4th-order Runge-Kutta solver.
The equations
dP/dt = r g(t) P (1 - P) [1 + mu B/(B + b0)] - a P H / (1 + a h P) - s w(t) P + iP plants, P = cover as a fraction of carrying capacity
dB/dt = B [ eb N/(N + n0) - mb - cb B ] + iB pollinators, nectar N = P x bloom(t)
dH/dt = H [ kappa a P / (1 + a h P) - mh ] - f H C / (1 + f h2 H) + iH rabbits
dC/dt = C [ e2 f H / (1 + f h2 H) - mc ] + iC foxes
- Logistic plant growth with a season. g(t) = max(0, sin(2 pi (t - 80)/365))^0.9 switches growth on in spring and fades it in autumn; w(t) = winter die-back; the small constant iP stands for roots and seed bank.
- Pollination (mutualism). Seed set is multiplied by 1 + mu B/(B + b0), a saturating benefit. Pollinators in turn live on nectar N = P x bloom(t), where bloom(t) = max(0, sin(2 pi (t - 95)/365))^2 peaks in early summer, with a saturating (Holling II) gain and density-dependent loss cb B.
- Holling type II feeding. Grazing and predation saturate because handling takes time (h, h2): a P H/(1 + a h P) for plants, f H C/(1 + f h2 H) for rabbits. This is the Rosenzweig-MacArthur form of the Lotka-Volterra model; with linear responses, no seasons and no logistic limit it reduces to the classic Lotka-Volterra pair, which conserves V = delta x - gamma ln x + beta y - alpha ln y.
- Open meadow. The small immigration terms iB, iH, iC stand for animals wandering in from the neighbourhood; without them a deterministic model can drive a population to numerically tiny but never truly zero values.
What the checks verify
The classic Lotka-Volterra invariant tests the integrator on a model whose exact conserved quantity is known. The plant-rabbit pair has an analytic fixed point P* = mh / (a (kappa - mh h)); starting there the solver must stay there. Halving the step must shrink the error of the full model by about 16, the signature of 4th-order accuracy.
What is simplified
Parameters are illustrative round numbers chosen to give believable ranges (tens of rabbits and a fraction of a fox per hectare, a flowering peak in early summer), not fitted to a particular site. Real meadows have dozens of plant and insect species, predator switching, disease (myxomatosis), weather and mowing. The animals you see are drawn to match the equation densities: they wander, flee and chase for the picture, but the numbers come only from the differential equations. Nothing here is a forecast for a real habitat.
What to try
- Press No pollinators and compare the plant cover and the seed boost with the balanced meadow: the mutualism is worth a few percent of cover.
- No foxes: rabbits boom, the plants are grazed to a stub every summer and the whole system swings harder.
- Harsh winters shrink pollinators to almost nothing; watch how late in spring they recover behind the flowers.
- Switch Seasons off and open the rabbit-fox phase graph: the trajectory spirals toward a point instead of tracing yearly loops.
- Raise the fox attack rate: does the meadow get more rabbits or fewer? (Watch the delay: foxes follow rabbits.)