Interlacing roles of bottom‐up, top‐down, endogenous, and anthropogenic factors in population oscillations

The relative importance of bottom‐up, top‐down, and endogenous factors in population oscillations has been vigorously debated, yet few longer‐term studies of population dynamics have considered their potential nonadditive interactions. This study tests the hypothesis that oscillations can arise from the combined effects of bottom‐up, top‐down, and endogenous factors, with some of these factors being sensitive to anthropogenic modifications of habitats. Populations of Neodiprion abietis , a univoltine sawfly whose oscillations have strong first‐ and second‐order components, were monitored over six to eight generations in a natural and an adjacent managed (precommercially thinned) balsam fir ( Abies balsamea ) stand at each of three sites in Newfoundland, Canada. Endogenous and exogenous factors affecting N. abietis were evaluated to identify the factors contributing to first‐ and second‐order density dependence and uncover any differences in dynamics attributable to forest management. In both natural and managed stands, N. abietis populations remained at high densities for 2–4 yr before collapsing. At the beginning of outbreaks, immigration, high fecundity, female‐biased sex ratios, and high survival in later larval instars and during the cocoon stage allowed N. abietis populations to increase rapidly. After this increasing phase, top‐down (mortality by the baculovirus Neodiprion abietis nucleopolyhedrovirus [ Neab NPV ] and, under some circumstances, parasitism by Mesopolobus verditer ) and bottom‐up (host‐plant) exogenous factors combined with endogenous factors associated with recruitment to initiate the localized collapse of N. abietis populations. Because first‐order density dependence was attributable to Neab NPV (and sometimes M. verditer ) and second‐order density dependence was attributable to host‐plant effects and endogenous factors associated with recruitment, these factors are inferred to cause the periodic behavior of this species. Host‐plant effects, but not top‐down mortality factors, were more pronounced in natural than thinned stands, suggesting that recent increases in N. abietis outbreak severity and range are best explained by a management‐induced reduction of host‐plant effects. By considering nonadditive interactions between trophic forces, this study reveals the intricate relationships between the factors responsible for population oscillations that would otherwise remain hidden.