Cannibalism in the red wood ant, Formica polyctena (Hymenoptera: formicidae).

(1) Wars between colonies of the red wood ant, Formica polyctena (Foerst.), are very common in the study area, a dune valley near The Hague, The Netherlands. In an extensive study Mabelis (1979) has put forward a hypothesis that explains the occurrence of wars in terms of protein supply in periods of insufficient prey availability. As prey density is low in early spring, most wars occur in this period. While Mabelis' evidence was mainly experimental, this field study was carried out to evaluate his hypothesis by permanent simultaneous measurements of prey/ protein intake of a colony, availability of prey in the foraging area and the protein demand. (2) The participation of the study colony in wars with three of its neighbours is described. (3) Conspecific war casualties, Bibionidae and Lepidoptera larvae amounted to 86.3% of the total prey intake during the first 80 days of the season of 1980. (4) There was an obvious, sudden switch around 6 May from conspecific to other prey. (5) Before this switch no Bibionidae or Lepidoptera larvae could be detected in the field. (6) Also prey species other than Bibionidae and Lepidoptera larvae are, for different reasons, not regarded as a significant food resource during the period of wars. (7) During spring battles there was a fluctuating but relatively high protein demand. (8) These results support Mabelis' hypothesis. (9) Proximate causes of cannibalism, other hypotheses on wood ant wars and the role of cannibalism in the regulation of wood ant and prey populations are discussed.