Canadian Forest Service Publications

Burton, P.J.; Boulanger, Y. 2018. Characterizing combined fire and insect outbreak disturbance regimes in British Columbia, Canada. Landsc. Ecol. 33: 1997-2011.

Year: 2018

Issued by: Laurentian Forestry Centre

Catalog ID: 39400

Language: English

Availability: PDF (request by e-mail)

Available from the Journal's Web site.
DOI: 10.1007/s10980-018-0710-4

† This site may require a fee

Mark record


Context Fires and insect outbreaks are important agents of forest landscape change, but the classification and distribution of these combined processes remain unstudied aspects of forest disturbance regimes.

Objectives We sought to map areas of land characterized by homogenous fire regime (HFR) attributes and by distinctive combinations of fire, bark beetles and defoliating insect outbreaks, and how their distribution might change should current climatic trends continue.

Methods We used a 41-year history of mapped fires and forest insect outbreaks to classify HFRs and combined fire and insect disturbance regimes (HDRs). Spatially constrained cluster analysis of 2524 20-km grid cells used mean annual area burned, ignition Julian date, fire size and fire frequency to delineate HFR zones. Mean annual areas burned, affected by bark beetles, and affected by defoliators were used to delineate HDR zones. Random forests classification used climate associations of HDRs to project likely changes in their distribution.

Results Eighteen HFR zones accounted for 30% of variance, compared to 27 HDR zones accounting for 59% of variance. Fire regime designation had low predictive power in explaining 23 homogenous insect outbreak regimes or the 27 HDRs. Climate change projections indicate a northward migration of current HDR zones. Conditions suitable for defoliator outbreaks are projected to increase, resulting in a projected increase in the total rate of forest disturbance.

Conclusions When describing forest disturbance regimes, it is important to consider the combined and possibly interacting agents of tree mortality, which can result in emergent properties not predictable from any single agent.