Altruism in viscous populations revisited: Competition and altruism do not exactly cancel even in the island model

Michael C. Whitlock, J. David Van Dyken

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Background: Hamilton suggested that cooperation would evolve most readily in viscous (low dispersal) populations, because nearby individuals would have high genetic relatedness. However, Taylor (1992) found that the indirect fitness gains achieved by cooperation with relatives is exactly offset by local competition, preventing true altruism from evolving no matter the degree of relatedness or population viscosity. Question: Does local density regulation exactly cancel the indirect fitness benefits of altruism in Taylor's model? Mathematical method: Inclusive fitness theory following the derivation of Taylor (1992). Key assumptions: Relatedness depends on the effective deme size, N e, but the distribution of fitness benefits depends on its census size, N. Conclusions: For Taylor's model, the exact cancellation of the indirect fitness benefits of altruism by local competition requires the special case where Ne = N, a condition not often observed in nature. Even in Taylor's model, the benefits of altruism depend on relatedness, rather than just direct benefits. When Ne ≤ N, as is commonly observed, true altruism can evolve, and it is more likely to do so when populations are more viscous. Local competition is important, but it requires restrictive conditions to exactly negate the effects of cooperation among relatives in a viscous population.

Original languageEnglish (US)
Pages (from-to)747-752
Number of pages6
JournalEvolutionary Ecology Research
Issue number7
StatePublished - Oct 1 2011
Externally publishedYes


  • Altruism
  • Competition
  • Cooperation
  • Inclusive fitness
  • Island model
  • Viscosity

ASJC Scopus subject areas

  • Ecology, Evolution, Behavior and Systematics


Dive into the research topics of 'Altruism in viscous populations revisited: Competition and altruism do not exactly cancel even in the island model'. Together they form a unique fingerprint.

Cite this