Fire geomorphology: Interactions among climate, fire, and vegetation

In steep forested landscapes, fire plays a crucial role in the timing and magnitude of erosion. Fire is a natural disturbance to vegetation and surface erosion and is a primary driver of ecosystem change across the earth (e.g. Swetnam, 1993). In order to understand, predict, and manage for future effects of fire, it is important to understand historical relationships among fire, climate, vegetation, and erosional response. It is well established that severe summer drought is a primary driver of widespread fire on annual timescales (e.g. Westerling et al., 2006). However, fire regimes reflect both climate conditions and vegetation (fuel) characteristics, which change over decadal to millennial timescales. The natural range of variability of an ecosystem describes the nature of fires occurring over an extended period of time (Brown, 1995). But what timescale is relevant to assess fire variability in order to predict and manage for future changes? If climate is a primary driver of wildfire, then timescales that involve millennial global climate cycles are relevant.

The Middle Fork Salmon River (MFSR) flows from the southwest to northeast and drains 7440 km² of the Frank Church-River of No Return Wilderness in central Idaho (Figure 1). The lack of prior management influences (e.g. fire suppression and logging) on the landscape makes the MFSR an excellent place to study the natural role of fire without the influences of human disturbance. During the last 30 yrs, large wildfires have increased in duration and frequency throughout the western United States (Westerling et al. 2006). Fires have burned over 40% of the MFSR during the last 30 years (Figure 1). Big fire years have occurred during dry and warm summers (Figure 2).

Have large wildfires occurred in the past or are recent increases in fire frequency, duration, and size an anomaly over Holocene (1000's of years) timescales?

This study investigates the impact of Holocene climate change on spatial and temporal variations in the timing, frequency, and severity of fire. We ¹⁴C-date charcoal fragments found in discrete deposits within multiple stratigraphic profiles exposed in recently incised alluvial fans (Figure 3). One stratigraphic profile is a snapshot of a sequence of depositional events that have been preserved at one location in the larger fan complex (Figure 4). Alluvial fan deposition is autogenic and controlled by fan topography, type of depositional process, and event magnitude. Therefore, multiple profiles within each fan complex are necessary to capture a complete fan deposition history.

Within each profile, individual deposits are described according to color, texture, depth, continuity, thickness, degree of sorting, sediment size, percent gravel (% > 2mm), and charcoal abundance. Detailed descriptions of deposit characteristics were then used to identify depositional process. Fire-related deposits were determined using charcoal abundance within the deposits and/or the presence of a burn surface directly below the deposit (Figure 4). Charcoal fragments to radiocarbon date were carefully chosen within each deposit based on their in-situ presence within the deposit, lack of bioturbation (visual appearance of mixing by animal disturbance), and the angularity of the charcoal fragment indicating if the charcoal had been reworked by multiple events (rounding the fragments). Modern erosional processes within the sub-basins are useful to compare with ancient deposits within the stratigraphic profiles (Figure 4). For example, recent large fire-related debris flows occurred following moderate to severe fires. Fire history, fire frequency, and fire severity are inferred from the timing of fire, the relative abundance of charcoal, number of events per timescale, and deposit characteristics (e.g. Meyer et al., 1995; Pierce et al., 2004).

The reconstructed fire history of the MFSR spans the last 14,000 cal yr BP (Figure 5). The number of dated deposits decreases with time. This is likely because there is a decreased likelihood of older deposits being preserved and exposed (Meyer et al., 1995). Records indicate that fires are grouped both spatially (multi-basin fires) and temporally (fire frequency) during the Holocene suggesting that the timing of fire is not random and fire synchronicity and periodicity are driven by changes in climate and vegetation distributions. Temporal spacing refers to fire frequency and was determined by the number of fires burning during a given time period. Spatially grouped fires were determined by comparing the timing of fire with the spatial extent of fire throughout the larger watershed. Times when fires were burning in multiple small basins are inferred as periods of widespread fires in the MFSR.

The late Pleistocene and early Holocene (14 - 8 ka) was a period of infrequent severe fires that produced large debris flows (Figure 6). This period was followed by an increase in fire frequency between 7–5.5 ka and an absence of large debris flows in the stratigraphic record. During this time, fires were region-wide, upper and lower basin ecosystems were burning, and sheetflooding was the dominant depositonal process identified (Figure 6). From 4–2.5 ka fires and fire-related debris flows were becoming relatively more frequent. At 3 ka, fires were synchronous, occurring in multiple (upper and lower) basins. Fires were burning in multiple sub-basins, determined by individual fires with the statistically same age distributions, at 110, 900, 1500, 2070, 3780, and 6460 cal yr BP. The frequency of synchronous fires across different ecosystems increased following 3 ka.

This increase in fire frequency, severity, and synchronicity around 3 ka roughly corresponds to the arrival of lodgepole pine to high elevation forests of the Middle Fork Salmon River (Whitlock et al., 2010) (Figure 6). Lodgepole pine forests have generally dense vegetation with abundant fuel. They are adapted to survive fire by having abundant seed supplies and being able to regenerate quickly following high severity stand-replacing crown fires. A shift in fire regimes appears at approximately 3 ka inferred from an increase in the number of fire-related debris flows in all study site sub-basins throughout the MFSR (Figure 6). An increase in fire activity across a range of ecosystems during the late Holocene suggests overall higher fuel loads, combined with severe summer drought, which likely triggered severe region-wide fires. In contrast, millennial-scale dry conditions of the mid-Holocene are associated with relatively high fire frequency and low fire severity (inferred from the absence of large fire-related debris flows).

• Brown, P.M., Kaufmann, M.R., and Shepperd, W.D., 1999, Long-term, landscape patterns of past fire events in a montane ponderosa pine forest of central Colorado: Landscape Ecol., v. 14, p. 513-532.
• Cook, E., 2004, North American Summer PDSI Reconstructions, NOAA/NGDC Paleoclimatology Program: Boulder CO, World Data Center for Paleoclimatology
• Davis, L.G., Muehlenbachs, K., Schweger, C.E., and Rutter, N.W., 2002, Differential response of vegetation to postglacial climate in the Lower Salmon River Canyon, Idaho: Palaeogeography Palaeoclimatology Palaeoecology, v. 185, p. 339-354.
• Meyer, G.A., Wells, S.G., and Jull, A.J.T., 1995, Fire and alluvial chronology in Yellowstone National Park: Climatic and intrinsic controls on Holocene geomorphic processes: Geol. Soc. Am. Bull., v. 107, p. 1211-1230.
• Pierce, J.L., Meyer, G.A., and Jull, A.J.T., 2004, Fire-induced erosion and millennialscale climate change in northern ponderosa pine forests: Nature, v. 432, p. 87-90.
• Swetnam, T.W., 1993, Fire history and climate-change in giant sequoia groves: Science, v. 262, p. 885-889.
• Westerling, A.L., Hidalgo, H.G., Cayan, D.R., and Swetnam, T.W., 2006, Warming and earlier spring increase western US forest wildfire activity: Science, v. 313, p. 940-943.
• Whitlock, C., Briles, C.E., Fernandez, M.C., and Gage, J., 2010, Holocene vegetation, fire and climate history of the Sawtooth Range, central Idaho, USA: Quaternary Research, v. 75, p. 114-124.