Producer–Scrounger Game
The Producer–Scrounger Game is an evolutionary game-theoretic model in behavioral ecology in which group-foraging animals choose between actively searching for food patches (producing) and exploiting patches discovered by conspecifics (scrounging) under negative frequency-dependent selection.
The Producer–Scrounger (PS) Game is a mathematical model in evolutionary game theory, behavioral ecology, and optimal foraging theory that analyzes the strategic conflict and equilibrium dynamics within social foraging groups where individuals adopt one of two mutually exclusive behavioral tactics:
- Producers: Individuals that invest time and metabolic energy in actively searching the environment to locate previously undiscovered food patches (for instance, probing the soil, peeling bark, digging through snow, or diving for prey schools).
- Scroungers: Individuals that refrain from independent patch searching and instead monitor group-mates with elevated head-up vigilance, joining or kleptoparasitizing food patches discovered by producers (compare Information Center Hypothesis and Emery's Rule).
First formulated in 1981 by British behavioral ecologists Christopher J. Barnard and Richard M. Sibly in Animal Behaviour ("Producers and scroungers: a general model and its application to captive flocks of house sparrows"), and developed mathematically by W. L. Vickery, Luc-Alain Giraldeau, Thomas J. Valone, and John M. McNamara (1991) and Luc-Alain Giraldeau & Thomas Caraco (2000) in their monograph Social Foraging Theory, the Producer–Scrounger Game is the social-foraging counterpart to John Maynard Smith's Hawk–Dove game and Fretwell and Lucas's Ideal Free Distribution. Because scroungers do well when producers are common (many patches are being discovered and shared) but starve when scroungers are too common (no one is finding new food), the PS game is governed by strong negative frequency-dependent selection, converging toward a mixed Evolutionarily Stable Strategy (ESS) across social animals in the biosphere.
Mathematical Formulation: The Vickery–Giraldeau Rate-Maximizing Model (1991)
Finder's Share (a), Group Size (G), and the ESS Proportion of Producers (p*)
In 1991, William L. Vickery, Luc-Alain Giraldeau, J. S. Templeton, D. L. Kramer, and C. A. Chapman formalized the fundamental rate-maximizing Producer–Scrounger Game for a foraging flock of group size G exploiting discrete, depletable food patches that each contain F total food items (or energy units).
Suppose that within the group of G foragers, a proportion p play the Producer tactic (so GP = pG individuals are actively searching for patches at personal discovery rate λ patches per unit time) and a proportion 1 − p play the Scrounger tactic (so GS = [1 − p]G individuals scan for producers that have just struck food). Assume the two tactics are incompatible (a bird looking down with its bill buried in sand searching for seeds cannot simultaneously scan the horizon with its head up to spot distant flock-mates eating).
Whenever a producer discovers a patch of total value F, it consumes an exclusive finder's share (Ffinder, where a = Ffinder / F is the finder's fractional advantage, 0 < a ≤ 1) before the GS scroungers arrive and divide the remaining food (F − Ffinder = [1 − a]F) equally among the 1 producer + GS scroungers (n = 1 + GS individuals at the patch).
Under these assumptions, the expected intake rate of a Producer (WP) and the expected intake rate of a Scrounger (WS) as a function of producer frequency p are:
WP(p) = λ F · [ a + (1 − a) / (1 + (1 − p)G) ]
WS(p) = λ F · [ (pG) · (1 − a) / (1 + (1 − p)G) ]
Because WS(p) increases monotonically with p (equals 0 when p = 0) while WP(p) is positive at p = 0, setting WP(p*) = WS(p*) yields the unique mixed Evolutionarily Stable Strategy (ESS) proportion of producers (p*):
p* = a / (1 − a) + 1 / G
| Ecological / Patch Parameter | Mathematical Symbol | Direction of Change | Predicted Shift in ESS Producer Frequency (p*) | Predicted Shift in ESS Scrounger Frequency (1 − p*) | Empirical Verification |
|---|---|---|---|---|---|
| Finder's Share (Exclusive Fraction) | a = Ffinder / F | Increases (a ↑: rapid consumption or small, quickly eaten patches) | Increases (p* ↑) (Pure Producer p* = 1 when a ≥ [G−1]/G) | Decreases (1 − p* ↓) (Less food left over for joiners) | Spice finches (Lonchura punctulata, Giraldeau et al., 1990) |
| Foraging Group Size | G (Flock / Herd Size) | Increases (G ↑: larger flock or school) | Decreases (p* ↓) toward asymptotic floor a / (1 − a) | Increases (1 − p* ↑) (More eyes profit from each patch) | House sparrows (Passer domesticus, Barnard & Sibly, 1981) |
| Patch Clumping / Bonanza Size | F (Total items per patch, fixed handling rate) | Increases (F ↑: larger, denser seed clumps) | Decreases (p* ↓) because finder's fraction a = Ffinder/F drops | Increases (1 − p* ↑) (Large leftover bonanza rewards joiners) | Zenaida doves (Zenaida aurita), Pigeons (Columba livia) |
| Visual Obstruction / Vegetation Cover | Habitat opacity | Increases (Tall grass or visual barriers block spotting) | Increases (p* ↑) (Scrounger detection efficiency drops) | Decreases (1 − p* ↓) (Harder to detect successful producers) | Nutmeg mannikins & White-throated sparrows (Zonotrichia) |
| Forager Energy Budget / Starvation Risk | Reserve X vs. Threshold R | Positive budget (X > R, risk-averse state) | Decreases (p* ↓) (Producing has high discovery variance σ2) | Increases (1 − p* ↑) (Scrounging lowers intake variance σ2) | Dark-eyed juncos & Spice finches |
Key Predictions and Experimental Confirmations
1. Manipulation of the Finder's Share (a) in Spice Finches (Giraldeau, Hogan, & Clinchy, 1990)
The central quantitative prediction of the Vickery–Giraldeau equation (p* = a/[1 − a] + 1/G) is that the equilibrium frequency of scroungers is inversely controlled by the finder's share a. In a series of elegant aviary experiments on flocks of nutmeg mannikins / spice finches (Lonchura punctulata) and Carib grackles (Quiscalus lugubris), Luc-Alain Giraldeau, Thomas J. Valone, and Donald L. Kramer manipulated the spatial distribution of millet seeds hidden in wells across an artificial foraging grid while holding total seed number constant:
- High Finder's Share (a → 1; Dispersed Small Patches): When 100 seeds were distributed as 50 tiny patches of 2 seeds each, a producer discovering a well swallowed both seeds before any flock-mate could run across the aviary (a ≈ 1.0). Within days, the flock abandoned scrounging and converged to nearly 100% Producers (p* ≈ 0.90–1.00).
- Low Finder's Share (a → 0.1; Clumped Bonanza Patches): When the exact same 100 seeds were clumped into 5 large wells of 20 seeds each, a producer could eat only 2 or 3 seeds (a ≈ 0.10–0.15) before flock-mates converged on the well and shared the remaining 17 seeds. The exact same birds shifted their behavior dramatically, dropping producer frequency to p* ≈ 0.25–0.35 while 65% to 75% of foraging effort switched to Scrounging!
2. Phenotype-Limited (Asymmetric Dominance) Producer–Scrounger Games
In the simple symmetric PS game, every individual in the flock receives the exact same payoff at the ESS (WP = WS) and can either alternate between producing and scrounging within a bout or specialize at frequencies p* and (1 − p*). In real vertebrate hierarchies, however, individuals differ strongly in social dominance status, body size (see Rensch's Rule), and foraging skill.
In 1998, Zoltán Barta and Luc-Alain Giraldeau formulated the Phenotype-Limited (Asymmetric) Producer–Scrounger Game incorporating aggressive kleptoparasitic displacement at discovered patches:
- When dominant individuals can agonistically usurp a discovered patch from its finder (stealing the remaining [1 − a]F share rather than dividing it equally), high-ranking dominants specialize almost exclusively as Scroungers, forcing low-ranking subordinates to act as Producers!
- Field and aviary tests in house sparrows (Passer domesticus, Liker & Barta, 2002), Harris's sparrows (Zonotrichia querula, Rohwer & Ewald, 1981), ring-tailed lemurs (Lemur catta), and chacma baboons (Papio ursinus, King et al., 2009) confirmed this "shepherd effect": dominant individuals position themselves near the center of the foraging group where they can scan and usurp patches unearthed by peripheral subordinate producers.
3. Risk-Sensitive Producer–Scrounger Dynamics
By uniting the Producer–Scrounger Game with Risk-Sensitive Foraging, Thomas Caraco and Luc-Alain Giraldeau (1991) explained why an animal's internal physiological state (hunger) alters its social tactic even when patch size F and group size G are unchanged:
- Producing is a high-variance (risk-prone) tactic: a solitary searcher may go a long time finding nothing (X = 0), or it may discover a rich patch and enjoy the unshared finder's share aF before anyone arrives.
- Scrounging is a low-variance (risk-averse) tactic: by pooling the discovery rates of all pG producers in the flock, a scrounger receives frequent, small, predictable shares of food, minimizing the probability of a zero-food bout.
- Consequently, when birds are on a positive energy budget (μ > R), they shift toward scrounging to minimize starvation variance; when starved onto a negative energy budget (μ < R) where a scrounger's small split share is insufficient to survive the night, they shift toward producing to gamble on capturing a full finder's share!
The "Price of Anarchy" in Social Foraging: Why Scrounging Reduces Group Productivity
A profound macroecological insight of the Producer–Scrounger Game is that individual Darwinian optimization at the ESS (p* < 1) lowers the average intake rate of the group below the cooperative optimum (p = 1)—an ecological example of the Tragedy of the Commons or Price of Anarchy:
Because scroungers do not discover any new food patches (λscrounger = 0), total food discovered per unit time by the entire flock of G birds is strictly proportional to the number of active producers (Total Group Intake = p · G · λF). Thus, a flock of 100% Producers (p = 1) achieves the highest possible per-capita intake rate (Wmax = λF). Yet because any single mutant scrounger in a 100%-producer flock can free-ride on the discoveries of G − 1 producers whenever a < (G − 1)/G, natural selection drives the flock down to the ESS producer fraction p*, reducing everyone's average food intake rate to WESS = p* · λF < λF.
Frequently Asked Questions
How does the Producer–Scrounger Game differ from the Ideal Free Distribution (IFD)?
In Fretwell and Lucas's Ideal Free Distribution, habitats or patches are already known to all competitors, and animals distribute themselves across space among patches of different baseline quality so that all individuals achieve equal intake rates. In the Producer–Scrounger Game (Barnard & Sibly, 1981), animals are foraging together within the same habitat, and the strategic choice is between investing effort to discover hidden patches (Producer) versus parasitizing the search effort of group-mates (Scrounger).
What is the "incompatibility assumption" between producing and scrounging?
In 1999 and 2002, Isabelle Coolen, Luc-Alain Giraldeau, and Mark V. Abrahams demonstrated that the classical PS game requires a sensory trade-off (the incompatibility constraint): an animal cannot search for cryptic food and scan for successful companions at 100% efficiency simultaneously. For example, granivorous birds searching deep gravel must point their heads downward (producing), whereas spotting a distant flock-mate pecking rapidly requires holding the head upward (scrounging, which also provides antipredator vigilance; see Giving-Up Density). When food patches are easy to spot while simultaneously searching (opportunistic joining), pure scrounging shrinks and foragers combine both modes.
Do producers and scroungers exist outside foraging—such as in brood care, nest building, and microbial biofilms?
Yes! The mathematical structure of the Producer–Scrounger Game applies whenever one phenotype synthesizes a diffusible or stealable public resource and another exploits it:
- Kleptoparasitic Seabirds and Dung Beetles: Frigatebirds (Fregata) and skuas (Stercorarius) scrounge fish caught by boobies and puffins (Central Place Foraging producers); "tunneling" male dung beetles (Onthophagus) produce brood balls while hornless "sneaker" males scrounge matings and burrows.
- Microbial Siderophore and Invertase Games: In bacterial (Pseudomonas aeruginosa) and yeast (Saccharomyces cerevisiae) populations, "producer" cells secrete costly extracellular iron-chelating siderophores or sucrose-cleaving invertase enzymes, while mutant "scrounger (cheater)" cells cease enzyme synthesis and absorb the liberated iron and hexose sugars under negative frequency-dependent selection.
References
- Barnard, C. J., & Sibly, R. M. (May 1981). Producers and scroungers: a general model and its application to captive flocks of house sparrows. Animal Behaviour, 29(2), 543–550. https://doi.org/10.1016/S0003-3472(81)80117-0
- Vickery, W. L., Giraldeau, L.-A., Templeton, J. J., Kramer, D. L., & Chapman, C. A. (Jun 1991). Producers, scroungers, and group foraging. The American Naturalist, 137(6), 847–863. https://doi.org/10.1086/285197
- Giraldeau, L.-A., & Caraco, T. (Dec 31, 2000). Social Foraging Theory. Princeton University Press, Princeton, NJ. https://doi.org/10.1515/9780691188348
- Coolen, I., Giraldeau, L.-A., & Lavoie, M. (May 2001). Head position as an indicator of producer and scrounger tactics in a ground-feeding bird. Animal Behaviour, 61(5), 895–903. https://doi.org/10.1006/anbe.2000.1678