Islands on Land: What Island Biogeography Tells Us About Forest Fragments and City Parks
Island biogeography began with real islands. MacArthur and Wilson argued that species number on an island is a balance. Immigration adds species. Extinction removes them. Island size and distance from a mainland source set the rates. Larger islands hold more species. Isolated islands hold fewer.
That logic now travels inland. A forest patch in farmland behaves like an island. So does a city park in concrete. The “sea” is matrix habitat that many forest species cannot cross. The “mainland” is the nearest large reserve or unbroken woodland.
Area still matters. Bigger fragments usually support more species. They also support larger populations. Those populations go extinct more slowly. Small parks lose area-sensitive birds, wide-ranging mammals, and plants that need interior microclimates. Edge habitat then dominates. Light, wind, heat, and weedy species pour in from the boundary.
Isolation matters next. A grove ten kilometres from the nearest forest receives fewer colonists than a grove one kilometre away. Roads, walls, and hostile lawns raise that effective distance. Flying insects and birds may still arrive. Many amphibians, understory plants, and soil organisms do not. The result is a nested pattern. Small, remote patches keep only a subset of the species found in large, well-connected woods.
The matrix is not empty ocean. Farmland, gardens, and street trees can help or block movement. A permeable matrix raises immigration. An impermeable matrix turns every park into a true island. Management therefore includes the land between patches, not only the patch itself.
Relaxation is the delayed cost of isolation. A newly cut fragment may still hold a full species list for years. Populations then shrink. Specialists disappear first. That extinction debt means surveys right after fragmentation underestimate long-term loss. Urban parks created from remnant forest often show the same lag.
Not every prediction copies the classic model. Habitat quality can outweigh raw area. A small, wet, well-managed park may beat a large, degraded one. Frequent human disturbance adds a pressure islands in the sea rarely face. Dogs, lights, mowing, and invasive plants change extinction rates independently of size.
Corridors and stepping-stones are the practical reply. Hedgerows, riparian strips, and linked parkways raise immigration. They do not replace large core habitat. They can slow the slide toward a few generalist species. Target species still decide the design. A butterfly corridor is not a leopard corridor.
Sampling and scale need care. Species–area curves from parks mix true island effects with habitat diversity. A large park often contains more vegetation types. That alone raises species counts. Researchers therefore separate area from habitat heterogeneity. They also record matrix type, age of isolation, and management intensity.
For cities the lesson is double. First, keep the largest possible intact cores. Second, connect what remains. Street-tree networks, unmowed margins, and night-dark corridors help. Planting only ornamental exotics does not. Native structure, dead wood, and leaf litter restore the resources many forest species need after they arrive.
Island biogeography does not say fragments are worthless. It says they are leaky islands. Area, isolation, matrix, and time jointly set how many species persist. Forest remnants and urban parks follow that rule more closely than a glance at a green map suggests.