Ideal Free Distribution

The Ideal Free Distribution (IFD) is a foundational theoretical model in behavioral ecology, optimal foraging theory, and habitat selection that predicts how mobile consumers or breeding individuals distribute themselves across discrete resource patches of differing intrinsic quality. Formulated in 1969 (published in Acta Biotheoretica in 1970) by American ornithologist Stephen D. Fretwell and statistician Henry L. Lucas Jr.—building on earlier habitat-selection concepts by Svärdson (1949), Morisita (1952), and Orians (1969)—the model demonstrates that when intraspecific competition reduces per-capita resource intake within crowded habitats, animals will sort across patches until the realized fitness or intake rate of every individual is equal across all occupied patches, and no individual can improve its payoff by switching patches.

In evolutionary game theory (John Maynard Smith, 1982), the Ideal Free Distribution represents an Evolutionarily Stable Strategy (ESS) and a spatial Nash equilibrium shaped by natural selection. Bridging foraging decisions with carrying capacity, metapopulation dynamics, resource partitioning, and ecological release, the IFD serves as the null model against which territorial dominance (Ideal Despotic Distribution), perceptual constraints, and ecological traps are evaluated.

Why "Ideal" and Why "Free"?

Fretwell and Lucas (1969) named the model after two explicit idealized assumptions regarding animal cognition, movement, and social structure:

1. "Ideal" (Omniscient Patch Assessment): Every individual in the population possesses complete, accurate information regarding the intrinsic resource supply rate and current competitor density of every available patch in the ecosystem, and selects whichever patch maximizes its net energy intake rate or reproductive fitness.
2. "Free" (Unconstrained Movement and Equal Competitors): Every individual is free to enter or exit any habitat patch at zero travel cost—without being excluded by territorial aggression, dominance hierarchies, or preemption—and all individuals possess equal competitive ability.
3. Negative Density Dependence Within Patches: Within any patch i, the realized suitability or per-capita intake rate W_i(n_i) decreases monotonically as the number of conspecific competitors n_i occupying that patch increases (dW_i / dn_i < 0).

1. Continuous Input Model and Parker's Input-Matching Rule

In a continuous-input system (such as stream drift feeding in juvenile Atlantic salmon, nectar production in flower patches, or prey items arriving via conveyor inlets in laboratory aquaria), resources enter Patch i at a constant continuous supply rate R_i (items or joules per minute) and are divided equally among the n_i identical competitors currently present in that patch:

W_i(n_i) = R_i / n_i

Suppose a total population of N = n₁ + n₂ + ... + n_k individuals distributes itself across k patches. At the Ideal Free Nash equilibrium, all occupied patches must yield an identical per-capita intake rate (W₁ = W₂ = ... = W_k = W*):

R₁ / n₁* = R₂ / n₂* = ... = R_k / n_k* = ( ∑ R_j ) / N

Rearranging this equality yields Parker's (1974, 1978) Input-Matching Rule (also known as the Habitat Matching Rule):

n_i* / N = R_i / ( ∑_j=1^k R_j )

Under the Input-Matching Rule, the proportion of total foragers settling in Patch i equals the proportion of total landscape resources supplied by Patch i. If Patch 1 delivers food at 10 items min⁻¹ and Patch 2 delivers food at 2 items min⁻¹ (a 5:1 resource ratio), then 5/6 (83.3%) of the foragers aggregate in Patch 1 and 1/6 (16.7%) in Patch 2, equalizing individual intake across both patches.

2. Standing-Crop Interference Model (Sutherland's m-Coefficient)

In benthic or terrestrial habitats (such as shorebirds probing mudflats for polychaetes or grazing ungulates), foragers search across a standing crop of prey of density Q_i while suffering mutual interference (kleptoparasitism, fighting, or prey depression) as competitor density n_i increases. Hassell and Varley (1969) and William J. Sutherland (1983) modeled per-capita intake under interference using the power function:

W_i(n_i) = Q_i · n_i^−m

where m (m > 0) is the Hassell-Varley interference coefficient. Setting W₁(n₁*) = W₂(n₂*) at equilibrium yields the generalized consumer ratio:

n₁* / n₂* = ( Q₁ / Q₂ )^{1 / m}

When m = 1.0, the interference model reproduces the 1:1 linear Input-Matching Rule; when 0 < m < 1.0 (weak interference), consumer density in the richer patch exceeds the resource ratio (overmatching); and when m > 1.0 (strong interference), consumers spread more evenly across patches than the resource ratio (undermatching).

1. Milinski's (1979) Three-Spined Stickleback Aquarium Experiment

The landmark experimental test of the IFD was conducted by German behavioral ecologist Manfred Milinski (1979) using groups of six three-spined sticklebacks (Gasterosteus aculeatus) in a glass aquarium equipped with two automated pipettes at opposite ends that dropped live water fleas (Daphnia magna) at controlled continuous-input ratios. When the pipettes delivered Daphnia at a 2:1 ratio, the six sticklebacks sorted within two minutes into a mean distribution of 4.0 fish at the rich end and 2.0 fish at the poor end; when switched to a 5:1 profitability ratio, the sticklebacks redistributed to 5.0 fish at the rich end and 1.0 fish at the poor end, and rapidly reversed positions when the pipette rates were swapped mid-trial.

2. Harper's (1982) Mallard Ducks and Parker's (1970) Dung Flies

David G. C. Harper (1982) tested the IFD on a flock of 33 mallard ducks (Anas platyrhynchos) at Cambridge University by having two experimenters 20 m apart throw bread pieces at fixed ratios (1:1, 2:1, and 5:1). The flock split into sub-flocks matching the food input ratio within 90 seconds, though individual tracking revealed that dominant ducks monopolized a larger share of items than subordinates. Similarly, Geoffrey A. Parker (1970, 1974) showed that male yellow dung flies (Scathophaga stercoraria) distribute their search times across cattle dung pats of differing ages (where female arrival decays as exp(−λt)) so that every male achieves an equal fertilization rate per minute.

When field censuses of foraging animals are compared against patch profitability, researchers observe systematic deviations when the assumptions of omniscience, equal competitive ability, or free entry are relaxed:

Ideal Despotic and Ideal Preemptive Distributions

In their original 1969 paper and Fretwell's 1972 book Populations in a Seasonal Environment, Fretwell and Lucas contrasted the Ideal Free Distribution with the Ideal Despotic Distribution (IDD): when dominant resident individuals in high-quality Habitat 1 use territorial defense or aggression to depress newcomers' perceived suitability, subordinate individuals are forced to settle in undefended, lower-quality Habitat 2 at a much earlier population threshold. Under an Ideal Despotic or Ideal Preemptive Distribution (Pulliam & Danielson, 1991), average per-capita fitness remains higher in prime habitats than in marginal habitats—providing the behavioral foundation for H. Ronald Pulliam's (1988) source-sink metapopulation dynamics.

1. Fretwell, S. D., & Lucas, H. L., Jr. (1969). On territorial behavior and other factors influencing habitat distribution in birds. I. Theoretical development. Acta Biotheoretica, 19(1), 16–36. https://doi.org/10.1007/BF01601953
2. Milinski, M. (1979). An evolutionarily stable feeding strategy in sticklebacks. Zeitschrift für Tierpsychologie, 51(1), 36–40. https://doi.org/10.1111/j.1439-0310.1979.tb00669.x
3. Sutherland, W. J. (1983). Aggregation and the 'ideal free' distribution. Journal of Animal Ecology, 52(3), 821–828. https://doi.org/10.2307/4456
4. Parker, G. A., & Sutherland, W. J. (1986). Ideal free distributions when individuals differ in competitive ability: Phenotype-limited ideal free models. Animal Behaviour, 34(4), 1222–1242. https://doi.org/10.1016/S0003-3472(86)80182-880182-8)
5. Kennedy, M., & Gray, R. D. (1993). Can ecological theory predict the distribution of foraging animals? A critical analysis of experiments on the ideal free distribution. Oikos, 68(1), 158–166. https://doi.org/10.2307/3545322