On the limited consensus of mountain pine beetle impacts on wildfire

Abstract

Context: The mountain pine beetle (MPB; Dendroctonus ponderosae) is a native bark beetle whose outbreaks leads to widespread conifer forest mortality. Of particular concern to forest and wildfire managers is the influence of MPB outbreaks on wildfire via spatial legacies left in impacted forest stands. There is, however, limited consensus in the literature regarding how MPB outbreaks affect wildfire across western North America.

Objectives: This meta-analysis aims to (1) summarize available evidence regarding MPB-wildfire interactions, and (2) identify environmental and methodological indicators associated with various wildfire responses (i.e., amplified, neutral, or dampened) post-outbreak.

Methods: We include peer-reviewed publications focusing on MPB outbreaks and subsequent wildfire activity in forests across western Canada and the USA between 2000 and 2021. A classification scheme was used to examine attributes of each publication to assess which indicators contribute most to their associated wildfire response.

Results: We found that spatial scale, forest fuels, and weather are main drivers of variation in wildfire response post-outbreak. Metrics of forest fuels and inclusion of weather data on a stand-scale are related to amplified fire responses, whereas dampened responses correspond to landscape-scale analyses. Furthermore, red-stage stands are associated with amplified fire response, whereas other stages are associated with dampened response-supporting current conceptual models of the importance of outbreak stage on wildfire.

Conclusions: Advancing our understanding regarding drivers of wildfire responses post-MPB outbreak is key to developing accurate, and comparative research studies. These findings provide crucial information for wildfire, and forest management agencies, especially in forests newly exposed to this disturbance interaction under climate change.

Supplementary information: The online version contains supplementary material available at 10.1007/s10980-023-01720-z.

Keywords: Disturbance interaction; Fire intensity; Fire severity; Forest health; Forest insect pest.

Fig. 1

Number of publications (left y axis) and citations (right y axis) per year between 1980 and 2021, returned from Web of Science search of “Mountain pine beetle” and “*fire”. Gray bars represent the number of studies related to MPB and fire published per year, and the black line represents the number of citations on the topic of MPB and wildfire interactions per year

Fig. 2

Grouping of the 27 papers in the primary database corresponding to the main fire behavior metric evaluated within each study. P/A represents ‘presence or absence’ of wildfire, and definitions proposed by Key and Benson (2006) were used to differentiate between fire intensity and fire severity. On the topic of wildfire activity following a MPB outbreak, we found most papers analyzed fire intensity and severity, followed by ignition, P/A of fire, and area burned, respectfully

Fig. 3

a Counts of publication fire responses (n = 35) within each time-since MPB stage across all fire metrics (see text for details). Stages include green (0–1; unattacked to year of attack), red (1–2 years post-outbreak), gray (3–10 years post-outbreak), and old (10 + years post-outbreak). See text for details. b Proportional counts of publication fire response within each time-since MPB stage across all fire metrics

Fig. 4

Publication sites of data collection across western Canada, and the USA. Red (circle), yellow (square), and blue (triangle) points represent amplified, neutral, and dampened fire responses. Point locations correspond to the coordinates of publication data collection and are labeled according to publication number. Gray areas indicate pertinent ecoregions containing MPB host trees

Fig. 5

Biplot summarizing the principal component analysis (PCA) of global indicators used to analyse each publication classified into three fire response categories: red—amplified, yellow—neutral, and blue—dampened. Points in this biplot represent publications plotted in ordination space and are coordinated to the associated studies’ conclusion(s) regarding fire in response to the indicator variables analysed/measured. Labels in each quadrant represent the most significant indicators driving associated fire responses in the corresponding direction. Ellipses corresponding to each fire response represent the default confidence interval of 68%, and were added to visually separate groups of response observations

Fig. 6

Classification tree model using filtered indicators with Pearson correlation coefficient threshold of 0.8. Publications are grouped below each fire response label and are labeled according to the list of papers outlined in Fig. 4. Blue, red, and yellow coloured circles represent publications that are correctly classified per group, whereas gray coloured circles represent misclassified publications. Consideration of Relative Humidity was identified as the most important indicator in predicting fire response. For studies that include relative humidity in their analyses, the predicted fire response outcome is neutral whereas studies that do not consider relative humidity tended to be associated with an amplified or dampened response. Studies that did not consider relative humidity, but did examine red-stage MPB stand tended to exhibit amplified fire severity responses. Studies that considered neither relative humidity nor red-stage stands tended to exhibit dampened fire severity responses