Ecology
Source-Sink Dynamics: How Surplus Habitats Rescue Population Sinks
In a Southwestern desert canyon, a population of Chiricahua sparrows may fill 90 percent of the available "sink" habitat while producing almost none of its own recruits — every year, immigrants pouring out of a small, high-quality "source" patch keep the whole apparently-thriving population from winking out. This counterintuitive rescue is source-sink dynamics: a framework in spatial ecology, formalized by H. Ronald Pulliam in 1988, describing how populations in different habitat patches are demographically linked by dispersal.
Formally, a source is a patch where local reproduction exceeds local mortality (finite growth rate λ > 1, a net exporter of individuals), while a sink is a patch where deaths exceed births (λ < 1) and which would go locally extinct without a steady inflow of immigrants. The framework rests on the BIDE bookkeeping — Births, Immigration, Deaths, Emigration — and explains why the most crowded habitat is often not the one sustaining a species.
- FieldSpatial / population ecology
- Coined byH. Ronald Pulliam (1988)
- Core accountingBIDE: Births, Immigration, Deaths, Emigration
- Source criterionλ > 1 (net exporter, births > deaths)
- Sink criterionλ < 1 (net importer, deaths > births)
- Applies toBirds, mammals, insects, plants, fish
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What it is and where it happens
Source-sink dynamics is a model of how spatially separated subpopulations, occupying habitat patches of different quality, are coupled by the movement of individuals. It applies wherever a landscape is heterogeneous and organisms disperse between patches — fragmented forests, coral reefs, farmland hedgerows, alpine meadows, host tissues colonized by pathogens.
The key insight is that local demography and local abundance can be decoupled. A patch can be permanently full of animals yet be a demographic drain (a sink), while a small, unremarkable patch quietly generates the surplus (a source) that fills it. Pulliam's canonical illustration was two bird habitats: a productive source where fledgling output exceeds adult mortality, exporting the excess; and a sink where reproduction alone cannot replace deaths.
- Source: λ > 1; birth rate b > death rate d; net emigration.
- Sink: λ < 1; d > b; sustained only by immigration.
Because sinks can hold large standing numbers, census counts alone are dangerously misleading for identifying which habitat actually matters.
The mechanism, step by step (BIDE)
The engine is BIDE bookkeeping — the change in a local population equals Births + Immigration − Deaths − Emigration. Pulliam's (1988) two-patch model works through it:
- 1. Source produces a surplus. In the source, per-capita births β exceed deaths, so at its carrying capacity it generates more recruits than local territories can hold.
- 2. Excess emigrates. Once source breeding sites saturate, the reproductive surplus is forced out — density-dependent emigration (E).
- 3. Emigrants become immigrants (I) to the sink. They settle in the lower-quality patch.
- 4. The sink runs a chronic deficit. There, deaths outpace births every generation; the local λ < 1.
- 5. Immigration balances the deficit. A steady inflow holds the sink population at a stable, nonzero equilibrium n* even though it cannot self-replace.
Pulliam showed that at equilibrium the sink can hold the majority of the total population. Cut the dispersal link and the sink decays toward extinction at rate λ, while the source rebounds to its own carrying capacity.
Governing relations and characteristic numbers
The per-patch finite growth rate is the master variable: λ = 1 + b − d (or n(t+1)/n(t)); λ > 1 defines a source, λ < 1 a sink. Pulliam framed the source's per-capita reproductive surplus with the balance β·P₁ = P₁ − a₁·P₁ + immigration/emigration terms, but the intuitive core is a partition of individuals into breeders and floaters.
- Site-dependent regulation: the source has a fixed number of high-quality breeding sites; adults that cannot secure one become floaters that emigrate — the source of the exported surplus.
- Sink equilibrium: n*_sink = I / (1 − λ_sink), so a modest immigration rate I can maintain a large standing sink population when λ_sink is close to 1.
Empirically, the Black-throated Blue Warbler and Chiricahua/Botteri sparrows supplied early field examples; in some songbird systems sink habitat holds >50 percent of individuals. Reproductive-value and λ estimates from mark-recapture routinely span λ ≈ 1.3 in prime forest down to λ ≈ 0.6–0.8 in marginal edge habitat.
How it is studied, measured, and regulated
Classifying a patch requires estimating its intrinsic λ — what it would do with immigration switched off — which is hard because dispersal is invisible in a snapshot census. Methods include:
- Demographic parameterization: measuring habitat-specific fecundity and survival by intensive nest-monitoring and mark-recapture, then computing λ per patch (as in Nystrand et al.'s Siberian jay study, 2010).
- Dispersal tracking: band re-sightings, radio-telemetry, stable isotopes, and increasingly genetic assignment / FST analysis to detect asymmetric gene flow from source to sink.
- Contribution metrics: the Cr and Cn indices of Runge, Runge & Nichols (2006) partition each patch's self-recruitment versus export to rank true conservation value.
The dynamics are regulated chiefly by density-dependent habitat selection: as the ideal-despotic distribution predicts, dominants monopolize source sites and subordinates are pushed into sinks. Removing source individuals (a removal experiment) reveals the source by watching sinks decline.
How it differs from its close cousins
Source-sink dynamics is often confused with related spatial-ecology ideas; the distinctions are demographic, not just semantic.
- Metapopulation (Levins) theory tracks patch occupancy (colonization vs. extinction of identical patches); source-sink instead tracks within-patch demography and quality differences. A metapopulation can have all-equivalent patches; a source-sink system cannot.
- Pseudo-sink (Watkinson & Sutherland, 1995): a patch that looks like a sink (λ < 1) only because immigration has pushed it above its own carrying capacity, triggering density dependence. Cut immigration and it persists at a lower K — a true sink would go extinct.
- Ecological trap: a genuine sink that organisms actively prefer because settlement cues have become decoupled from real fitness (often via human land-use change). Traps are worse than passive sinks because they actively bleed the source.
- Ideal free distribution assumes animals equalize fitness across patches; source-sink arises when they cannot (despotic exclusion).
Why it matters: conservation, disease, and open questions
Source-sink thinking overturned a naive rule of conservation — protect where the animals are. If a reserve is drawn around a crowded sink, the population can collapse the moment the unprotected source is lost. Correctly identifying sources is therefore central to reserve design and species recovery (e.g., spotted owls, songbirds in fragmented forest, marine reserves seeding fished waters via larval export).
- Disease and pests: the same math governs pathogen reservoirs — a source host population or environmental reservoir can sustain infection in a sink where transmission alone (R₀ < 1) could not persist; likewise agricultural sinks can mask pest sources.
- Evolution: maladaptive gene flow from a source can swamp local adaptation in the sink and constrain a species' niche and range limits — a live question in evolutionary ecology.
Open questions: How does source-sink structure behave under climate-driven range shifts (yesterday's source becoming tomorrow's sink)? How do we reliably detect sources without decades of demographic data? And when does dispersal rescue versus merely subsidize a doomed trap?
| Patch type | Local growth (λ) | Net dispersal | Fate if isolated |
|---|---|---|---|
| Source | λ > 1 (b > d) | Net exporter (emigration surplus) | Persists and grows |
| True sink | λ < 1 (d > b) | Net importer (immigration sustains it) | Deterministic extinction |
| Pseudo-sink | λ < 1 only at high density | Net importer (over carrying capacity) | Persists at lower K (Watkinson & Sutherland 1995) |
| Ecological trap | λ < 1 but preferred | Attracts settlers by false cues | Extinction; drains the source |
| Balanced / self-sustaining | λ ≈ 1 | Roughly zero net flux | Persists near equilibrium |
Frequently asked questions
What is the difference between a source and a sink habitat?
A source is a habitat patch where local births exceed local deaths, giving a finite growth rate λ > 1; it produces a surplus of individuals that emigrate. A sink has deaths exceeding births (λ < 1) and cannot sustain itself — it persists only because immigrants continually arrive from sources. Crucially, a sink can still be densely populated, so abundance alone does not tell you which is which.
Who came up with source-sink dynamics?
H. Ronald Pulliam formalized the concept in a landmark 1988 paper in The American Naturalist, 'Sources, Sinks, and Population Regulation.' He built it on the BIDE framework (Births, Immigration, Deaths, Emigration) and showed that a large fraction of a population can occupy sink habitat where reproduction cannot balance mortality. The paper has been cited well over 2,000 times.
What is the BIDE model?
BIDE is the demographic bookkeeping underlying source-sink theory: the change in a local population equals Births plus Immigration minus Deaths minus Emigration. In a source, B and E dominate (it exports individuals); in a sink, I offsets a chronic deficit where D exceeds B. Setting the BIDE terms to balance gives each patch's equilibrium population size.
What is a pseudo-sink and how is it different from a true sink?
A pseudo-sink, described by Watkinson and Sutherland in 1995, is a patch that appears to be a sink (λ < 1) only because heavy immigration has pushed its population above its own carrying capacity, triggering density-dependent mortality or reduced fecundity. Remove the immigration and it persists — at a lower carrying capacity. A true sink, by contrast, cannot support any self-sustaining population and would go extinct if isolated.
How is source-sink dynamics different from metapopulation theory?
Classic metapopulation (Levins) theory tracks the occupancy of equivalent patches through colonization and extinction, ignoring differences in patch quality and within-patch demography. Source-sink theory instead focuses on quality differences: patches have distinct λ values, and dispersal flows asymmetrically from high-quality sources to low-quality sinks. Source-sink is essentially a demographic, quality-explicit refinement of spatial population theory.
Why does source-sink dynamics matter for conservation?
Because the most crowded habitat may be a demographic sink, protecting where animals are most abundant can fail catastrophically if the true source is left unprotected and lost. Managers use demographic estimates of λ and contribution metrics to locate sources, design reserves, and evaluate marine protected areas that export larvae. The framework also warns against ecological traps — attractive habitats that actually drain source populations.