Fire and Vegetation History in the Eastern Cascade Mountains, Washington

Abstract

Dendrochronological techniques were used to reconstruct a 433‐yr fire history, and to characterize the historical fire regime (frequency, size, season, severity) of the Teanaway River drainage in the eastern Cascade Mountains of Washington, USA. General Land Office section corner data were used to reconstruct aspects of the late‐19th‐century structure and composition of the forests in the study area. Systematic fire‐scar surveys (∼30 000 ha; 92 sites; 257 fire‐scarred cross‐sections; 1569 individual fire scars) revealed that fire frequency was quite variable spatially; Weibull median probability intervals ranged from seven to 43 years. Fire extent also varied widely; most fires were relatively small (<1000 ha), although several large fires (>4000 ha) were detected. Mean and median fire sizes were 1795 ha and 988 ha, respectively. Large fires occurred every 27 years, on average (every 11 years, on average, between 1708 and 1889), and coincided with periods of annual and seasonal drought (Palmer Drought Severity Index and winter Southern Oscillation Index). Intervals from 1 to 37 years occurred between fires of >4000 ha. Over 80% of fires occurred late in the growing season, or after the onset of cambial dormancy. Sampling locations in dry forest types (dominated by ponderosa pine) yielded fire‐scarred cross‐sections with numerous fire scars leading us to infer that most historical fires were of low intensity, leaving the overstory structure intact. This inference is corroborated by the composition and structure of the historical forest, which was characterized by a preponderance of very large (>100‐cm diameter) ponderosa pines. Mesic forest types (dominated by grand fir and Douglas‐fir) likely exhibited a wider range of fire severities. Moderate and occasional high‐severity understory fires or crown fires did occur within the study area, as indicated by the scarcity or lack of fire‐scar evidence and the presence of relatively even‐aged forests at several mesic forest sites. Historical section corner data indicate that small amounts of these forest types occurred in the study area. Fire frequency and size declined dramatically circa 1900, coincident with the advent of commercial timber harvesting, although most fires, despite their reduced number and size, continued to burn in the late summer or fall.

Climate drivers of regionally synchronous fires in the inland Northwest (1651 - 1900)

“…Although the reconstruction of temperature that we used was for the summer half of the year (April through September), we suggest that variation in spring temperature may have been more important than summer temperature in driving fire in our study area because the reconstruction captures more of the variation in spring than summer temperature. Our regional-scale analysis of the influence of summer climate generally corroborated local-scale results from the southern part of the study area where extensive fires occurred at individual sites during dry summers Hessl et al 2004;Wright and Agee 2004), consistent with the intuitive observation that dry weather during the fire season leads to lower fuel moisture and greater flammability.…”

“…In contrast, neither ENSO or PDO acting independently was a strong driver of fire, consistent with local-scale analyses in dry forests of the inland Northwest Hessl et al 2004;Wright and Agee 2004) and with regional-scale analysis in dry forests of the Northern Rocky Mountains just to the east, where historical fires were not strongly driven by variation in PDO (Heyerdahl et al in press). The occurrence of our highly synchronous regional fire years is consistent with climate and large-scale climate patterns across western North America (Kitzberger et al 2007).…”

“…1). Fire-scarred trees were sampled systematically over large areas (262 to 30 000 ha) in grids of multi-tree plots Wright and Agee 2004) or targeted within several topographic facets (Everett et al 2000;Hessl et al 2004), except at Cariboo, where nine 1-ha sites were sampled along an 85-km transect on the Fraser River Plateau (Daniels and Watson 2003) (Table 1). An average of 248 trees was crossdated per site (range 86-667 trees).…”

“…Over the past decade, a network of crossdated surface-fire chronologies has been developed from fire scars across the inland Northwest, with 15 sites sampled over 7 degrees of latitude in the interiors of Oregon, Washington and British Columbia (Everett et al 2000;Heyerdahl et al 2001Heyerdahl et al , 2007Daniels and Watson 2003;Hessl et al 2004;Wright and Agee 2004). The climate drivers of fire have been analysed for many of these sites individually or in small groups of sites Daniels and Watson 2003;Hessl et al 2004;Wright and Agee 2004).…”

“…The climate drivers of fire have been analysed for many of these sites individually or in small groups of sites Daniels and Watson 2003;Hessl et al 2004;Wright and Agee 2004). Large and/or widespread fires occurred during significantly dry years, some of which were El Niño years.…”