Getting from patch to patch
A population split in two behaves differently from one population. What corridors do, and what the evidence actually shows.
What changes when a patch is cut off
Divide a population in two and you immediately alter the arithmetic of survival. The separated halves can no longer exchange individuals: a bad winter that kills the local breeding adults in one patch cannot be repaired by immigration from the other, and a genetic line that dwindles below a critical number stays dwindled. Each fragment now runs its own demographic experiment, and small experiments fail more often than large ones. This is not a theoretical concern — metapopulation ecology, developed in quantitative form by Ilkka Hanski and colleagues from the 1980s onwards, showed that local extinction is a near-certain outcome for small, isolated populations given enough time. The question is whether connectivity changes that outcome.
A corridor, in the ecological sense, is any linear feature that increases the probability of an individual moving between two habitat patches. It can be a hedgerow, a river bank, a green roof sequence, a motorway verge, an unsprayed field margin. The test is functional: does movement actually increase? A strip of trees between two woods looks like a corridor on a map; whether it functions as one depends on which species you are asking about and what the strip contains.
What the evidence shows — and where it is cautious
The most-cited experimental test of corridor function is the work carried out on longleaf pine patches in South Carolina, published by Nick Haddad, Ellen Damschen and collaborators over more than two decades starting in the late 1990s. They cut replicated patches of open habitat into a matrix of pine forest, with some patches connected by strips and others not. Plants with bird-dispersed seeds moved into connected patches at higher rates; butterflies and some ground beetles showed the same pattern. The connected patches also maintained higher species richness over time. These results are real and robust, but the system was designed to be testable — the matrix was fairly hostile, the corridors were clear, and the focal species were mobile enough to potentially use them.
Field evidence from less controlled landscapes is more mixed. Some tagged animals use hedgerow corridors extensively; others cross open ground directly and ignore the strip. Corridor width matters: a narrow grass margin may channel small invertebrates but offer nothing to a badger. Corridor length matters: a 200-metre strip between two woods is not the same problem as a two-kilometre crossing. The matrix matters too — a field sown with a diverse wildflower mix is a very different barrier from bare compacted soil, and some species cross the open ground without needing a formal connector at all.
What the evidence supports is a qualified claim: well-designed corridors, matched to the movement ecology of specific target species, increase connectivity and reduce local extinction rates in fragmented systems. The claim that any green linear feature improves biodiversity in general is much harder to defend precisely. Fragmentation creates different problems for different species, and the solutions are correspondingly specific.
Reading the landscape
When you are walking and you see a hedgerow running between two woods, or a rank grass strip along a ditch connecting two meadows, you are looking at a potential corridor — but potential is the operative word. The question worth asking is: potential for whom? Dormice need continuous canopy or at least touching branches; they will not drop to the ground and cross a gap. Brown hares, by contrast, move freely across open farmland and rarely need a formal connector. Fungi move as spores and are largely indifferent to the corridors that matter to vertebrates.
The presence of indicator species at intermediate points in a strip is the closest thing to a field test available without radio-tracking equipment. If you find violet ground beetles along a ditch margin all the way from one wood to the next, something is using that strip as a through-route. If the strip holds only the generalist species common to the surrounding matrix, it is functioning as habitat — locally valuable — but not clearly as a corridor in the functional sense. The distinction is worth keeping in mind: a feature can be ecologically useful without doing the connectivity job that corridor planning assumes it does.