Abstract
In western North America, subalpine forests experience fires that vary greatly in terms of severity and extent. However, beyond observational and dendroecological records, little is known about past fire severity and magnitude. This is because metrics used to identify fire and ecological impacts in the deep past (i.e., sedimentary charcoal and pollen data) are coarse tools for examining fine-scale environmental responses. Yet large fires should result in changes in pollen abundance and composition, which in turn can be used to identify event types, or severity of past events. We compare pollen spectra changes from a subalpine forest following identified fire events, to pollen spectra from periods between fires. From the pollen data, two types of fire events may be inferred. Fire events that affect understory plant composition (low to mid-severity) result in increases in canopy pollen. Fire events that consume both understory and canopy plants (high severity) result in decreases in canopy pollen relative to understory pollen and likely reflect differences in the recovery rates of trees compared to shrubs, grasses, and forbs. Inferred fire type showed no relationship to the size of charcoal peaks, suggesting that the quantity of charcoal observed in a sample does not provide information on fire severity. These results reveal the potential of examining fire regimes through time by combining sedimentary charcoal and pollen data, extending our interpretations of fire regimes deeper into the past.
Highlights
In western North America, subalpine forests experience fires that vary greatly in terms of severity and extent
Sedimentary records of fire occurrence have expanded our understanding of fire regimes beyond the observable record and the scope of physical evidence identified by dendroecology
Recent analyses of fire histories based on sedimentary charcoal have expanded the interpretation of these data to include some metrics of fire biomass consumption (Duffin 2008, Marlon et al 2008, Higuera et al 2009, Marlon et al 2009)
Summary
UCI 63879 NPD-LWH08 1A-1B; 53.5 cm charcoal UCI 58941 NPD-LWH08 1A-2B; 59.5 cm charcoal UCI 63880 NPD-LWH08 1A-1B; 69.5 cm charcoal UCI 63881 NPD-LWH08 1A-2B; 91.5 cm charcoal UCI 58940 NPD-LWH08 1A-2B; 105.5 cm charcoal UCI 63882 NPD-LWH08 1A-2B; 172.5 cm charcoal UCI 58938 NPD-LWH08 1A-3B; 195 cm charcoal UCI 63883 NPD-LWH08 1A-3B; 240.5 cm charcoal UGAMS 4602 NPD-LWH08 1A-3B; 333.5 cm bulk a Median age based on intcal04.14c Stuvier and Reimer. Peak charcoal (total charcoal ÷ background charcoal) was identified as a fire event only when it exceeded the ninetyfifth percentile of the noise distribution of charcoal counts within 1000 years, based on a Gaussian mixture model (Higuera et al 2009). The unequal sampling strategy between the contiguously sampled fire record (charcoal) and intermittent sampling of the environment (pollen) results in a random sampling of postfire and non-fire vegetation. Pollen data were binned into three categories: canopy, understory, and other (Figure 1; Table 2). The ratio of canopy to the understory was calculated for each sample to characterize vegetation composition of post-fire and non-fire samples. Pollen data were considered post-fire if they came from either the same or the following sample of an identified charcoal peak.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
