Wood ants and a geometrid defoliator of birch: predation outweighs beneficial effects through the host plant.

Soil amelioration by a wood-ant species and its consequences for the larval performance of autumnal moths feeding on mountain birch were studied at various distances from the nest mound. Soil nitrate and ammonium nitrogen did not show any clear relationship with distance. However, trees growing in the mound had over 20% more foliar nitrogen than more distant trees. When moth larvae were experimentally protected from predation, their survival rate and final weight tended to decrease with increasing distance. In a laboratory experiment with detached leaves, the relative growth rate of larvae was roughly 30% higher on leaves from trees located on the mound. Differences in larval performance refute the Plant Stress Hypothesis proposed by T.C.R. White and support P.W. Price's Plant Vigor Hypothesis. Predation by ants was examined along the same gradient in trees with and without a glue band that excluded ants from the canopy. Reduction in the daily survival rate of larvae attributable to ant predation was about 35% in trees growing in the mound and around 5% at a distance of 20 m. Other things being equal, about 25 times more larvae entering the penultimate instar would achieve the pupal stage outside the wood-ant territory than in the vicinity of the mound. While both the fertilizing and predatory influence of wood ants is clear, the domain of predation is much larger than the area where trees and their herbivores can exploit enhanced nutrient levels in and around ant mounds. The existence of undamaged green islands around ant mounds in otherwise totally defoliated mountain-birch forests cannot be explained by soil amelioration by wood ants but rather by their predatory activity.