Influence of environmental factors and spatio-temporal covariates during the initial development of a spruce budworm outbreak
- First Online:
- Cite this article as:
- Bouchard, M. & Auger, I. Landscape Ecol (2014) 29: 111. doi:10.1007/s10980-013-9966-x
- 456 Downloads
Recurrent and synchronous spruce budworm (SBW) outbreaks have important impacts in boreal and sub-boreal forest ecosystems of North America. This study examines the early phase of an outbreak that was developing across a 268,000 km2 area over a period of 9 years (2003–2011). The territory was subdivided in 225 km2 cells, and the relative influence of forest composition, elevation, forest age, average degree-days and soil drainage were examined during three development phases of the outbreak: initial epicenter location, relatively long-distance spread (cell-to-cell expansion), and expansion inside individual cells (within-cell expansion). The results indicate that elevation is the most determinant variable for initial epicenter location. Other variables that were identified as important for outbreak development by previous studies, such as forest composition and average degree-days, were not so important during this phase. However, forest composition and average degree-days were important factors during the cell-to-cell and within-cell expansion phases. Separating outbreak development in distinct phases also allowed to integrate phase-specific spatial and temporal covariates that were highly significant in the models, such as distance from previous year defoliations during the cell-to-cell expansion phase, and the proportion of defoliated stands during the preceding year for the within-cell expansion phase. Overall, this study provides limited evidence that patterns of SBW outbreak expansion could be altered by reducing host tree species abundance in the forest [mainly balsam fir (Abies balsamea) in this region]. More generally, this study suggests that the influence of environmental variables on SBW outbreak development is clearly phase-dependent, and that this landscape-level, process-based approach could be useful to forecast insect outbreak development in forest ecosystems.