Remove one animal

When one species comes out, the system it shaped comes out with it

A moorland without grazing sheep returns to a specific pattern of colonisers. A river that loses its otters does not simply become a river with more fish. Subtract one animal from a system and you discover, often too late, how much architecture that animal was quietly maintaining.

The clearest examples are ungulates — hoofed mammals whose grazing keeps vegetation at an early successional stage. On British uplands, red deer at high densities suppress tree regeneration across entire catchments. Where deer are removed or substantially reduced, birch and rowan move quickly into dwarf shrub heath. The result is not the open woodland these uplands once carried; it is a novel scrub community assembled from whatever seed sources happen to be within dispersal range, regulated by soil chemistry that decades of grazing have already altered. Nothing goes back exactly when a key animal is removed, any more than it does when restoration is attempted by direct intervention.

What the middle trophic levels actually do

The language of keystone species, coined by the ecologist Robert Paine following his 1966 experiments on rocky intertidal communities in Washington State, describes an animal whose influence on the system is disproportionate to its abundance. Remove the sea star Pisaster ochraceus from a tidal patch and mussels outcompete almost everything else within months; species richness collapses. The mechanism is straightforward predation, but the outcome is structural: the physical arrangement of the community changes.

Herbivores operate through different pathways. Rabbits in lowland Britain graze turf to a close sward that favours annual plants, cushion-forming perennials and invertebrates that depend on bare soil and low vegetation. When myxomatosis arrived in 1953 and removed rabbits from much of the landscape in a matter of months, rank coarse grass and scrub spread rapidly across chalk downland that had been open for centuries. Some of the rarer chalk plants could no longer compete. The invertebrates that depended on bare warm soil lost their microhabitat. Arrested succession had been maintained by teeth; when the teeth went, succession resumed.

Burrowing and rooting animals alter soil differently. Wild boar create patches of bare disturbed earth — rooting events that function as small-scale disturbances, opening the seed bank and allowing light-demanding herbs to germinate. Where boar have been excluded, these patches disappear; the seed bank accumulates dormant viable material but receives less light. The absence of one large rooting animal changes the germination ecology of hundreds of plant species simultaneously, though most of this goes unnoticed because the process is slow and diffuse.

Water is also restructured by single species. The beaver effect on hydrology is well documented: dam building raises local water tables, creates wetland margins, slows flow and changes sediment dynamics. Remove the beaver from a valley — as happened across most of Europe over several centuries — and the wetland mosaic it maintains gradually drains, the channel narrows and incises, and the floodplain dries. The vegetation transitions accordingly, from wetland specialists toward dry grassland or scrub. The soil carbon that accumulated under wet conditions begins to oxidise. One animal's absence propagates into geology.

The return problem

Reintroduction is not the inverse of removal, and the interval matters enormously. A soil altered by centuries of dry conditions does not re-wet the moment beavers return; a scrub system that closed over chalk in the 1960s does not open again when rabbit numbers recover. The vegetation community that was present before removal is often no longer available as a seed source. The physical structure — soil texture, organic matter depth, drainage paths — has moved to a new equilibrium.

This is why ecologists speak carefully about what reintroduction can achieve. An animal returned to a landscape will reshape what is currently there, not what was there before. Sea eagles returned to Britain do not find the same coastal ecology they left; they find a different configuration of prey and habitat and they interact with it accordingly. The cascade they create will differ from any historical cascade, because every other component of the system has changed in the interval.

What single-species removals and returns make visible, more than anything else, is the degree to which communities are not simply collections of organisms. They are records of what each organism in them has been doing, for a long time, to everything else.