Source-sink dynamics

Source-sink dynamics describe the spatial redistribution of individuals in heterogeneous landscapes, where some habitat patches (sources) produce a surplus of individuals that disperse to other patches (sinks) where local death rates exceed local birth rates. The concept was formalized by Ronald Pulliam in 1988 and has become central to metapopulation theory, metacommunity ecology, and conservation biology.

A source patch is defined demographically: births exceed deaths, and the population would persist even without immigration. A sink patch is defined complementarily: deaths exceed births, and the population persists only because immigration from sources offsets the local deficit. The distinction is not based on habitat quality alone but on the demographic balance within the patch. A high-quality habitat may function as a sink if it is occupied by a competitively inferior species; a marginal habitat may function as a source if it is occupied by a species well-adapted to its conditions.

Source-sink dynamics produce counterintuitive conservation consequences. Protecting sink populations without protecting their source populations is futile: the sink will collapse when the source is degraded. Conversely, sources may be invisible in census data because they export most of their production; a landscape that appears to support a species only in a few sink patches may actually depend on a single unrecognized source. The mapping of sources and sinks requires demographic data — birth and death rates — not merely occurrence data.

The dynamics extend beyond simple source-sink pairs. A landscape may contain multiple sources and sinks arranged in a network, with flows between them determined by dispersal distances, barriers, and landscape resistance. The source-sink structure of a landscape is a property of both the habitat patches and the species that occupy them: the same landscape may be a source-sink system for one species and a uniform habitat for another, depending on their dispersal abilities and demographic requirements.

In metacommunity theory, source-sink dynamics are one of the four paradigms (the "mass effects" paradigm), but they operate in most real metacommunities regardless of which paradigm dominates. Even in species-sorting systems, where patches are environmentally differentiated, dispersal from favorable to unfavorable patches creates source-sink structure. The question is not whether source-sink dynamics occur but whether they are strong enough to override local competitive exclusion.