Abstract
Analysis of field and laboratory distributions of S. pineti on Scots pine reveals that the aphids are aggregated in colonies on needles and the colonies are aggregated on shoots. While shoot aggregation is likely to reflect differences in habitat suitability or localized numerical increase, needle aggregation is a result of gregariousness mediated by social responses between aphids. From an analysis of field egg distribution, syrphid predators appear to oviposit preferentially on aphid-infested shoots; as a result syrphid larvae are distributed discriminately on those shoots bearing aphid colonies. Aphids in large needle aggregations have a higher probability of escape from a syrphid larva than those in small aggregations. The aphids use touch to communicate the presence of a predator, there being no evidence for an alarm pheromone. A set of analytical models is developed to test whether needle aggregation is likely to give a net advantage to the aphids over their syrphid predators. These models consistently predict an advantage to the aphids if the predator searches systematically between needles. Simulations of a computer model of random search within shoots, repeated encounters and mortality during escape suggest that needle aggregation will only confer an advantage on the aphids when the predator searches preferentially among those shoots bearing aphid colonies.