Fragmentation
Cut a habitat into pieces and the pieces hold less than the whole — even when nothing else changes.
The geometry of isolation
A single wood of a hundred hectares supports more species than five woods of twenty hectares each, even though the area is identical. The reason is not complicated: interior species need distance from the edge. When a habitat is cut into smaller pieces, the ratio of edge to interior rises, and specialists that depend on stable, buffered conditions — low light, high humidity, minimal disturbance — lose the core they need. The fragments that remain are almost entirely edge. Edge depth turns out to matter enormously: in broadleaved woodland, measurable edge conditions can penetrate fifty metres or more, meaning a fragment narrower than a hundred metres has no interior at all.
Isolation compounds the geometry. A species that disappears from a fragment through a run of bad winters, a disease outbreak, or simple chance cannot recolonise if the next patch of habitat is beyond its dispersal range. Large populations tolerate bad years; small ones do not. The smaller the fragment, the smaller the population it supports, and the more vulnerable that population is to the kind of random variation — a cold spring, a dry summer — that a large population absorbs without much consequence. Ecologists call this demographic stochasticity: luck, applied to small numbers, matters more.
This is the mechanism behind the species-area relationship, one of the most replicated patterns in ecology. Roughly speaking, reducing habitat area by ninety percent halves the number of species it can sustain in the long run — though the losses are not immediate. There is a period after fragmentation when species counts look almost unchanged, because many individuals are still alive that the fragment can no longer recruit. This lag — sometimes decades — is called an extinction debt: the species are present, but the account is already overdrawn.
Not all species respond equally. Generalists and pioneer species adapted to disturbed, open conditions often fare well or even increase in fragments. It is the interior specialists — woodland floor plants, area-sensitive birds, soil invertebrates with limited mobility — that vanish first and return slowest. They are also the species least likely to cross the open ground between fragments, making the matrix between patches — farmland, roads, suburbs — as important as the patches themselves. A fragment surrounded by hostile terrain is effectively more isolated than its map position suggests.
Fragmentation's effects, then, are not just about what is lost. They are about what cannot come back once it is gone.