Postglacial vegetation and fire history of the southern Cascade Range, Oregon

Quaternary Research 84 (2015) 348–357

Contents lists available at ScienceDirect

Quaternary Research journal homepage: www.elsevier.com/locate/yqres

Postglacial vegetation and fire history of the southern Cascade Range, Oregon Alicia White a,⁎, Christy Briles b, Cathy Whitlock c a b c

Department of Earth Sciences, Montana State University, Bozeman, MT 59717, USA Department of Geography and Environmental Science, University of Colorado at Denver, Denver, CO 80217, USA Dept. Earth Sciences and Montana Institute on Ecosystems, MSU, Bozeman, MT 59717, USA

a r t i c l e

i n f o

Article history: Received 26 March 2015 Available online 2 November 2015 Keywords: Holocene Fire history Climate history Vegetation history Pollen Charcoal Cascade Range Oregon

a b s t r a c t The Cascade Range of southwestern Oregon contains some of North America's most diverse forests, but the ecological history of this area is poorly understood. A 7900-yr-long pollen and charcoal record was examined to better understand past changes in vegetation and fire activity in relation to large-scale climate variability. From 7900 to 3500 cal yr BP, the dominance of xerophytic species and the frequent fires are consistent with a climate that was warmer and drier than at present. The period from 3500 cal yr BP to present experienced an abundance of mesophytic taxa and reduced fire frequency, suggesting cooler and wetter conditions. The regional history of Abies indicates that it was most widespread during the late-glacial period; its range contracted during the early Holocene thermal maximum, and it steadily expanded during the middle and late Holocene. In contrast, Pseudotsuga was restricted in range during the glacial period, became abundant at low-elevation sites in the Coast and northern Cascade ranges during the early Holocene, and was more prevalent in southern mid-elevation sites as the climate became cooler and wetter in the late Holocene. The sensitivity of these species to past climate change suggests that biogeographic responses to future conditions will be highly variable in this region. Published by Elsevier Inc. on behalf of University of Washington.

Introduction Our understanding of the fire, vegetation and climate history of northern California and western Oregon has come from paleoecological records from the Coast Range (Worona and Whitlock, 1995; Long and Whitlock, 2002; Long et al., 2007), the central Cascade Range and its eastern flanks (Sea and Whitlock, 1995; Minckley et al., 2007; Long et al., 2014), and the Klamath and Siskiyou mountains (Mohr et al., 2000; Briles et al., 2005, 2008, 2011; Daniels et al., 2005; Colombaroli and Gavin, 2010; Crawford et al., 2015) (Fig. 1). These records show that vegetation and fire changes throughout the Holocene were strongly governed by large-scale variations in the climate system related to the seasonal cycle of insolation and its effects on the strength of the northeastern Pacific subtropical high-pressure system in summer (Bartlein et al., 1998). Records from the Klamath and Siskiyou mountains also show that past ecological responses were also mediated by topographic gradients and substrate-dependent levels of nutrients (Briles et al., 2011). Native people were present in the Klamath region throughout the Holocene, and population numbers are thought to have been highest in the last 1500 yr (Arnold and Walsh, 2010; Crawford et al., 2015). Their influences on vegetation and fire regimes have been

interpreted to be localized to population centers at lower elevations than Hobart Lake (Lake, 2013; Crawford et al., 2015). One poorly studied area of the Pacific Northwest is the region of the southern Cascade Range, located north of the Sierra Nevada, southeast of the Oregon Coast Range, and northeast of the Klamath region, which includes the Klamath and Siskiyou mountains (Fig 1). The southern Cascades support plant communities composed of many species that are at the northern, southern, eastern or western limits of their biogeographic ranges, and because of the rich forest diversity it is an area of high conservation concern (Whittaker, 1960; 1972; Whitlock et al., 2004; U.S. Geological Survey, 2006; Odion and Sarr, 2007; Olson et al., 2012). This paper describes an 7900-yr-long vegetation and fire history of the southern Cascade Range, based on pollen, charcoal and lithologic records from Hobart Lake (41.09935°N, 122.48170°W, 1458 m, 3.2 ha) (Fig. 1). The record and its comparison with pollen data from other sites in the region help trace the Holocene history of two important conifers, Abies (fir) and Pseudotsuga (Douglas-fir), in response to past climate variations. These insights have important implications for understanding the range of vegetation responses likely in the future as a result of projected climate change. Modern setting

⁎ Corresponding author.

http://dx.doi.org/10.1016/j.yqres.2015.09.007 0033-5894/ Published by Elsevier Inc. on behalf of University of Washington.

Seasonal climate patterns at Hobart Lake are driven by the expansion of the Aleutian Low in winter, which directs frontal storms to

A. White et al. / Quaternary Research 84 (2015) 348–357

349

Pacific Northwest (Agee, 1993). Based on tree-ring and charcoal studies from similar forests in the Klamath Mountains, fire-return intervals range from 11 to 19 yr (Taylor and Skinner, 1998; Mohr et al., 2000). Methods Two sets of 5-cm-diameter long cores were obtained from Hobart Lake with a modified Livingstone square-rod piston corer from an anchored platform in the deepest location of the lake. The first core, measuring 11.47 m long, was obtained in 2004 and the second, measuring 10.01 m in length, was recovered in 2011. A unit of impenetrable clay and silt was encountered at the base of sets of cores. Cores were extruded, packaged in plastic wrap, aluminum foil, and transported to the MSU Paleoecology Lab where they were refrigerated. In the laboratory, cores were split longitudinally, described, and photographed. One-third of the 2011 core was archived. The top two meters of the 2004 core dried significantly prior to this study, and the 2011 core was used to fill in the upper section. It was correlated with the 2004 core based on litho- and chronostratigraphy (see results). Therefore, the top 2.35 m of the 2011 core and the bottom 9.17 m of the 2004 core were used for all analyses and treated as a single record. Lithologic analyses

Fig. 1. The location of Hobart Lake and other pollen records in the region on a base map showing average annual precipitation from 1981–2010 (PRISM Climate Group, 2014). Numbers in parentheses are lake elevations in m asl. Beaver Lake and Tumalo Lake are not discussed in this paper.

Loss-on-ignition analysis at 5-cm intervals was undertaken to quantify the organic and carbonate content of the lake sediments (Dean, 1974). For each interval, 1 cm3 of sediment was heated at 80°C for 24 h to dry the material, then placed in a furnace and heated at 550°C for 2 h and re-heated at 900°C for two additional hours. Sample weight loss was measured between each procedural step. Magnetic susceptibility was undertaken for the entire 2011 core and portions of the 2004 core to assist in correlation. Magnetic susceptibility was measured with a Bartington magnetic susceptibility core logging sensor at 0.5-cm increments. Peaks in magnetic susceptibility, expressed in CGS × 10− 6 units, record the input of ferromagnetic minerals from allochthonous material (Dearing, 1999) (Fig. 2). Pollen

the region, and the northeastern Pacific subtropical high-pressure system in summer, which creates seasonally dry conditions (Western Regional Climate Center, D. R. I, 2013). At Hobart Lake, winter and summer temperatures average to 1 and 19°C, respectively, and precipitation averages 760 mm/yr (PRISM Climate Group, 2014), of which approximately 50% is received in winter and the remaining amount falls in spring and fall (Western Regional Climate Center, D. R. I, 2013). The vegetation of the southern Cascade Range, in general, is arrayed along gradients of elevation and climate (Odion and Sarr, 2007). The Interior Valley Zone, dominated by Quercus garryana (Oregon white oak) and Quercus kelloggii (California black oak), occurs below 800 m elevation and Mixed-Conifer Zone with Pinus ponderosa (ponderosa pine), Calocedrus decurrens (incense cedar), and Abies concolor (white fir) extends from 800 to 1400 m elevation. The A. concolor Zone is located from 1400 to 1700 m elevation, and subalpine forests of Abies magnifica (Red fir) and Tsuga mertensiana (mountain hemlock) in the T. mertensiana Zone lie between 1700 and 2500 m elevation. Areas above 2500 m elevation support alpine meadows (Franklin and Dyrness, 1988). Hobart Lake is a landslide-dammed lake located in the MixedConifer Zone. The watershed supports Pseudotsuga menziesii, A. concolor, P. ponderosa, and C. decurrens, although logging in the last 150 yr has favored A. concolor and P. ponderosa. Abies procera and A. magnifica hybridize in the region creating A. magnifica var. shastensis. Abies lasiocarpa, scattered farther north, does not occur near Hobart Lake. The mixed-conifer and A. concolor forests are characterized by low- to moderate-severity fire regimes with lightning ignitions occurring more frequently in these forests than in any other forest type in the

Pollen analysis was undertaken on samples, 1-cm3 volume, extracted at 16-cm intervals. Processing followed standard methods (Bennett and Willis, 2001), including the addition of Lycopodium spore tablets of a known concentration to calculate pollen concentration (grains cm−3). Pollen residues were mounted in silicone oil on slides and tallied under a microscope at 400 and 1000× magnification. Reference slides and pollen identification keys (McAndrews et al., 1973; Faegri and Iversen, 1975; Kapp et al., 2000) were consulted to identify each pollen grain to the lowest taxonomic level possible. A minimum of 300 terrestrial pollen grains was counted for each slide. Pollen percentages were calculated as a percent of total counted terrestrial pollen. Aquatic pollen percentages were calculated as a percentage of total terrestrial and aquatic pollen. This information was plotted using C2 and Tilia software, with zone designation determined using a constrained cluster analysis (CONISS) in Tilia (Grimm, 1987). Pollen accumulation rates (grains cm−2 yr−1) were calculated by dividing pollen concentration by the sample deposition time. Pinus pollen was separated into subgenus Strobus, subgenus Pinus and undifferentiated Pinus grains. Pinus subg. Strobus was attributed to Pinus lambertiana (sugar pine) or Pinus monticola (western white pine). Pinus subg. Pinus was assigned to P. ponderosa, which occurs locally, or Pinus contorta (lodgepole pine), which grows ~20 km east of Hobart Lake. Abies pollen is attributed to Abies grandis and A. concolor based on their modern phytogeography. Alnus rubra-type pollen is referred to Alnus rhombifolia (white alder), Alnus incana subsp. tenufolia (thinleaf alder) or A. rubra (red alder), all of which grow today in southwestern Oregon and northern California and have similar pollen

350

A. White et al. / Quaternary Research 84 (2015) 348–357

Fig. 2. Lithology, radiocarbon dates, magnetic susceptibility data, and percent organic and inorganic content for the 2004 and 2011 Hobart Lake cores.

morphology (Leopold et al., 2012). Quercus pollen was separated into deciduous Q. garryana-type and evergreen Quercus vaccinifolia-type (huckleberry oak) (Jarvis et al., 1992). Pollen grains that were unidentifiable were counted as ‘Unknown’ while those that were degraded, damaged, or hidden were counted as ‘Indeterminate.’ Charcoal Macroscopic charcoal analysis was undertaken on samples of 2-cm3 volume at contiguous 1-cm intervals in the 2004 and 2011 cores. Each of the 1240 samples was placed in a solution of 10% sodium hexametaphosphate and 8.25% bleach for 24 h. Samples were gently rinsed through a 125-μm mesh sieve, and the residue was transferred to a gridded petri dish. This 125-μm mesh size was selected because previous studies have shown that particles larger than this size capture a local fire signal within ~b 20 km of the lake (Whitlock and Larsen, 2001). All charcoal fragments in each sample were tallied under a stereoscope. Statistical analysis of the charcoal data was conducted using the CHARAnalysis program to reconstruct the local fire history of Hobart Lake (Higuera et al., 2009; Higuera et al., 2010). CHARAnalysis converts charcoal count data into charcoal accumulation rates (CHAR), measured in particles cm−2 yr−1. This information was continuously resampled in 8-yr bins, which was the median sample resolution of the Hobart Lake record. The slowly varying component of the charcoal time series (socalled background charcoal or BCHAR) was separated from highfrequency deviations or charcoal peaks that indicate individual fire episodes. The background charcoal trend was established using a 600-yr Lowess smoother, which maximized the signal-to-noise index and the noise distribution goodness of fit. Charcoal peaks above the

95th percentile were designated as fire episodes. A Poisson distribution was applied to the fire-episode data, and multiple peaks within a single distribution were eliminated as they likely came from a single fire event. Therefore, a charcoal peak could represent more than one fire and we refer to peaks as fire episodes. The absence of significant charcoal peaks means that charcoal levels did not exceed the prescribed threshold for peak detection, either because fires did not occur or they were too small or low in severity to be detected. We cannot rule out that some very small or low severity fires were undetectable using macroscopic charcoal analysis. For example, the signal-to-noise ratio between 4000 and 2000 cal yr BP is below 0.5, suggesting a period of limited or small fires. Fire episodes were then summarized with a 1000-yr window to reconstruct the fire frequency. The fire return interval (FRI) curve representing years between fires was constructed by interpolating the raw FRIs to annual values and applying a 1000-yr Lowess smoother. Results Lithology The 2004 and 2011 Hobart Lake sediment cores were composed primarily of fine-detritus gyttja (Fig. 2). The bottom of the 2004 core from 11.47 to 8.30 m depth (Unit 1) contained dark brown fine-detritus gyttja interspersed with inorganic clay and coarse sand layers, ranging in thickness from 1 to 3 cm. These inorganic sediments likely reached the lake through a series of erosional events. Unit 2 (8.30 to 5.20 m depth) consisted of banded brown fine-detritus gyttja. Above 5.20 m depth (Unit 3), the gyttja was largely homogenous implying a fairly closed and productive lake environment. Unit 1 had the highest and

A. White et al. / Quaternary Research 84 (2015) 348–357

most variable magnetic susceptibility values (ranging from 0 to 510 CGS × 10−6) and relatively low organic content values. A tephra layer at 10.90 m depth is assigned to the Mazama ash, which erupted at 7627 ± 150 cal yr BP (Zdanowicz et al., 1999). Above the Mazama ash, the core featured several meters of light-brown fine-detritus gyttja with coarsely banded clay layers that were especially abundant from 8.30 to 9.00 m depth. Magnetic susceptibility values varied widely in this section from 0 to 412 CGS × 10− 6. In Unit 2 (8.30 to 5.00 m depth), the core contained banded light and dark brown fine-detritus gyttja with a transition to primarily dark fine-detritus gyttja above 5 m m depth. An unidentified ash or sand layer from 4.34 to 4.35 m depth was associated with a peak in magnetic susceptibility (155 CGS × 10−6). The tephra layer at 2.30 m depth in the 2004 core was identified as Glass Mountain tephra (890–940 cal yr BP; Elmira Wan, USGS Tephrochronology Lab, personal communication, 2013). Above that depth was brown medium-detritus gyttja with increasing plant remains. An increase in MS at the top of the 2004 core is attributed to erosion associated with recent forest clearance (Colombaroli and Gavin, 2010). In summary, erosional events delivered pulses of inorganic material to Hobart Lake in Units 1 and 2, and Unit 3 represents largely autochthonous sediment deposition. Chronology Plant macrofossils from 11 levels in the 2004 core and three levels from the 2011 core were submitted for AMS radiocarbon dating, and the results as well as the tephra information were used to create an age-depth model (Table 1 and Fig. 3). The Glass Mountain tephra was used as the point of correlation between cores, occurring at 2.20 m depth on the 2011 core and at 2.30 m depth on the 2004 core. Radiocarbon dates younger than the tephra in the 2011 core and older than the same tephra in the 2004 core were not included in the age-depth model. Radiocarbon dates were calibrated using CALIB 6.0 (Reimer et al., 2009; Stuiver et al., 2010), and Clam 2.1 software (http://chrono.qub. ac.uk/blaauw/clam.html) was used within R (http://www.r-project. org/) to create the age-depth curve. A smoothing spline set to 0.3 was used, in which 0 means no smoothing and 1 is a straight line. The 95% confidence intervals for the curve were calculated by repeated sampling of the individual radiocarbon and tephra date confidence intervals (Blaauw, 2010). Random points were selected within each date's confidence interval and a line was drawn connecting the points. This sampling procedure was repeated 1000 times to develop confidence intervals for the age-depth model. The sedimentation accumulation

351

rate of Hobart Lake was very high for the region, with rates of 9.2 yr cm−1 at the bottom of the core and 5 yr cm−1 at the top. Pollen and charcoal The pollen record was divided into three zones (Fig. 4): Zone HL-1 (11.34–9.54 m depth; ca. 7930–6300 cal yr BP) had high percentages of Pinus (41–76%), with 2.5–17.5% of the record composed of Pinus subg. Strobus and 0.3–4.4% of Pinus subg. Pinus. Abies (0.3–6%) and Pseudotsuga (0.3–5%) values were low, and Cupressaceae percentages (7–29%) were the highest of the record. A. rubra-type pollen was found in low percentages in this zone, representing between 0 and 3% of the record. Salix percentages ranged from 0.3 to 1.6%. Deciduoustype Quercus comprised a prominent portion of Zone HL-1 (8–17%) and Artemisia was 0–1.3%. Among aquatic taxa, only Pediastrum was present in values N 1%, though these were not plotted on the pollen diagram. PARs ranged from 2773 to 10,815 grains cm−2 yr−1. Most of the charcoal particles in the record were from charred wood and needles. Raw charcoal counts ranged from 1 to 330 in 2-cm3 volume samples. Charcoal accumulation rates were high and variable from 7900 to 4000 cal yr BP and less so from 4000 cal yr BP to the present (Fig. 4). Background charcoal levels were initially low at 1.80 particles cm−2 yr−1 at 7920 cal yr BP, and by 6000 cal yr BP, they were consistently N2.00 particles cm− 2 yr−1, reaching a peak of 3.40 cm−2 yr−1 at 6930 cal yr BP. Fire return intervals were shortest (64yr fire−1) at 7400 cal yr BP and gradually lengthened to 162 yr fire−1 at 6330 cal yr BP. Although the signal-to-noise index was high for the entire dataset at 0.79, it was especially high during Zone HL-1, lending confidence to these results. The high sedimentation rate likely allows for precise detection of individual fire events. Zone HL-2 (9.54–4.26 m depth; ca. 6300–2000 cal yr BP) pollen was dominated by Pinus (38–65%) and these percentages were slightly higher at the base of the zone. Three to 11% of the Pinus record was composed of Pinus subg. Strobus and 0.5–3.5% was Pinus subg. Pinus. Abies (4–16%) and Pseudotsuga (7–18%) percentages rose from the bottom to the top of this zone. Cupressaceae (6–21%) values dropped from higher percentages in Zone HL-1. A. rubra-type values remained at low levels (0–2.6%), deciduous-type Quercus fluctuated from 5–17%, Salix ranged from 0–2.6%, and Artemisia ranged from 0–2% throughout Zone HL-2. Pediastrum araneosum values increased (0–7%) and Pediastrum boryanum had a substantial spike in its occurrence (0.3–92%). PARs spanned from 2138–8273 grains cm−2 yr−1. The pollen zone suggests a period of mixed-conifer forest based on its similarity with modern pollen spectra from Oregon (Minckley and Whitlock, 2000). Background

Table 1 Radiocarbon and tephra information for Hobart Lake. Depth (cm)a

Core

Uncalibrated 14C age (14C yr BP)

±

Calibrated age (cal yr BP) with 2-sigma rangeb

Material dated

Lab numberc

153 192 241 235 230 63 167 244 379 475 545 736 861 935 1017 1128

Hobart 2011B Hobart 2011B Hobart 2011B Hobart 2011B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B Hobart 2004B

540 610 1080 na na 155 510 1020 1795 2345 2845 3990 4455 5490 5815 7270

25 25 25 na na 30 30 35 35 35 35 35 40 35 35 35

508–550 577–653 934–1014 890–940 890–940 166–231 505–555 900–988 1686–1821 2311–2473 2866–3068 4405–4534 4959–5148 6263–6324 6501–6679 8012–8169

Seed Twig Wood Glass Mountain Tephra Glass Mountain Tephra Peat Peat Wood Grass Twig Wood Wood Wood Wood Twig Wood

OS-98623 OS-98622 OS-98621d EW-062811-HL6.30 EW-062811-HL6.30 118785d 119600d 119601 119602 119603 119604 119605 119606 119607 119608 119609

a b c d

Depth below mud surface. Ages calibrated with CALIB 6.0. National Ocean Sciences AMS Facility, Woods Hole Oceanographic Institute. Date not used in age-depth model.

352

A. White et al. / Quaternary Research 84 (2015) 348–357

Fig. 3. Age-depth relationship for Hobart Lake. The black line indicates the interpolated age at each depth. The gray bar is a 95% confidence interval surrounding the age-depth curve. Radiocarbon and tephra dates are depicted as small teal circles with vertical error bars. Clam 2.1 (http://chrono.qub.ac.uk/blaauw/clam.html) was used within R (http://www.r-project.org/) to create the age-depth curve.

charcoal levels fluctuated between 1.40 and 2.50 particles cm−2 yr−1 from ca. 6000 until 800 cal yr BP. Fire return intervals lengthened to 247 yr fire−1 at ca. 4050 cal yr BP and then shortened to 78 yr fire−1 by ca. 3300 cal yr BP. Zone HL-3 (4.26–0.00 m depth; ca. 2000-present) pollen was dominated by Pinus (25–54%), although less so than in the other zones. Pinus subg. Strobus represented 4–15% and Pinus subg. Pinus was 0–5% of the Pinus record. Abies (2–16%) levels fluctuated throughout the zone, whereas Pseudotsuga (10–32%) levels continued to increase from the previous zone. Cupressaceae (3–21%) levels also fluctuated, but remained below the higher levels of Zone HL-1. Salix percentages ranged from 0.3–6.3%. Fraxinus and Salix occurred in low values throughout the record with the only percentages of note occurring

within Zone HL-3 (0–5.7%). A. rubra-type increased to its highest levels of the record (0.33–4.33%). Deciduous-type Quercus percentages ranged from 4–17% with no temporal trend, and Artemisia levels varied between 0 and 3%. Asteraceae subf. Asteroideae maintained low percentages (0–1%) throughout Zones HL-1 and HL-2 but saw higher percentages (3.6%) in Zone HL-3 than before. P. araneosum (0–11%) and especially P. boryanum (0–47%) reached higher percentages during this time period. PARs ranged from 806 to 8680 grains cm−2 yr−1 in Zone HL-3. Based on comparison with modern pollen data, the zone marks the establishment of modern mixed-conifer forest at the site (Minckley and Whitlock, 2000). Background charcoal levels, primarily of burned wood material, dropped from around 2 to 1.31 particles cm−2 yr−1 to 0.30 cm−2 yr−1

Fig. 4. Charcoal and pollen data for selected taxa from Hobart Lake.

A. White et al. / Quaternary Research 84 (2015) 348–357

353

at present (Fig. 4). Fire return intervals ranged between 100 and 180 yr fire−1 during the majority Zone HL-3 until lengthening substantially to 298 yr fire−1 at present. The highest charcoal influx of the record occurred at 900 cal yr BP.

Discussion Vegetation, fire, and climate history at Hobart Lake The Hobart Lake pollen record began at ca. 7900 cal yr BP, and this age is a minimum estimate for the landslide that formed the lake. Humans inhabited the region throughout the Holocene but their influence on vegetation dynamics and fire regimes in this part of the Cascade Range is not known. Other studies in the Klamath region document the presence of semi-permanent settlements at lower elevations than Hobart Lake as well as the deliberate use of fire for ease of travel, acorn production, and the creation of ideal game habitat (Lake, 2013; Crawford et al., 2015). At Hobart Lake, we assume that most of the vegetation and fire changes are driven by climate. Abundant Pinus pollen prior to 6300 cal yr BP is attributed to P. ponderosa, which grows at the site, although P. contorta may also have been present (Fig. 4). Quercus pollen, largely of the deciduous type, also composed a substantial portion of the pollen record from ca. 7900 to 6300 cal yr BP, and is attributed to an expansion of Q. garryana. Its high abundance suggests drier-than-present conditions. Cupressaceae pollen reached its highest percentages at ca. 6300 cal yr BP and both C. decurrens or Juniperus occidentalis grow near the lake today in dry settings. In contrast, Abies was less abundant from ca. 7900 to 6300 cal yr BP than subsequently. The low ratio of arboreal to non-arboreal pollen (AP/NAP) from ca. 6500 to 5700 cal yr BP indicates open forest conditions (Fig. 4). From ca. 6300 to 2000 cal yr BP, Pinus and Cupressaceae levels decreased, and Abies and Pseudotsuga increased (Figs. 4 and 5). The charcoal record indicates that fires, which had been relatively frequent (FRI = ~65 yr) prior to ca. 7600 cal yr BP, became much less frequent after ca. 4000 cal yr BP (FRI = ~250–275 yr) (Figs. 4 and 5). The changes in vegetation and fire regime after ca. 4000 cal yr BP were likely driven by a trend towards cooler, effectively wetter conditions in the middle and late Holocene, as a result of declining summer insolation (Bartlein et al., 1998) (Fig. 5). Cooling after ca. 4000 cal yr BP is also evidenced by renewed glacial activity in the mountains of the Pacific Northwest and northern California (Porter and Denton, 1967; Marcott et al., 2009; Bowerman and Clark, 2011). In addition, winter sea-surface temperatures in the northeastern Pacific, inferred from alkenone data, suggest warm conditions prior to ca. 8000 cal yr BP followed by cooling winters between ca. 8000 and 2300 cal yr BP (Barron et al., 2003). The last 2300 yr of the marine record is again characterized by warm winters but with greater interannual variability (Barron et al., 2003). BCHAR levels, which reflect levels of fuel biomass, were relatively stable from 6400 to 900 cal yr BP. Fire frequency generally declined from ca. 7900 cal yr BP to present, as suggested by lengthening FRI. Within this general trend are three prominent periods of low fire frequency from ca. 6400 to 6100 cal yr BP (FRI = ~160 yr), from ca. 4900 to 3600 cal yr BP (FRI = ~250 yr), and from ca. 2800 to 2200 cal yr BP (FRI = ~ 180 yr). The shifts from short to long FRIs correspond with marked declines in Cupressaceae pollen and increases in Abies and Pseudotsuga pollen. Both genera of Cupressaceae (J. occidentalis and C. decurrens) tolerate frequent fires today, and acquire thick bark as they age, making them increasingly resistant to low-severity surface fires (Burns and Honkala, 1990). While Cupressaceae likely did well during the periods of frequent fires, the pollen data confirm that Abies and Pseudotsuga were favored during the episodes of low fire activity (Fig. 5). The presence of large charcoal peaks from ca. 7300 to 4000 cal yr BP is an evidence of occasional severe or large fires throughout the middle Holocene.

Fig. 5. Fire and selected pollen records from Hobart Lake compared with regional environmental proxies. A. Modern and historic sea surface temperatures from ODP 1019 approximately 50 km off the California–Oregon border (Barron et al., 2003). B. Cupressaceae pollen percentages at Hobart Lake. C. The July 45°N insolation anomaly (Berger, 1978). D. Pollen percentages of Abies (dark green) and Pseudotsuga (light green) at Hobart Lake. E. CHAR (charcoal accumulation rate) in black and BCHAR (background charcoal accumulation rate) in red, with the fire return intervals (FRI) in orange at Hobart Lake. The vertical gray bars show the relationship between particular fire episodes and the other proxy data.

In the last 2000 yr, Pinus percentages at Hobart Lake were at their lowest levels of the record, Cupressaceae values were low, and Abies and Pseudotsuga continued to increase until 900 cal yr BP (Figs. 4 and 5). The increases of these genera in the forest are likely a response to increased summer insolation and warmer, more mesic conditions than before. After 900 cal yr BP, the pollen data indicate that Pseudotsuga became the forest dominant (Figs. 4 and 5). BCHAR levels dropped slightly from ca. 1100 to 700 cal yr BP and FRIs increased to ~150 yr. This shift occurred during the Medieval Climate Anomaly (ca. AD 900–1300; Steinman et al., 2012), which was a period of severe droughts in western U.S. (Graham et al., 2007). A notable fire episode at ca. 900 cal yr BP is inferred from the large charcoal peak at Hobart Lake, as well as at Bolan, Sanger, and Upper Squaw lakes (Colombaroli and Gavin, 2010; Briles et al., 2011). Apparently, this widespread conflagration occurred near the time of Glass Mountain eruption and ash-laden substrates may have reduced fuel and soil moisture, heightening the forest flammability during already dry Medieval Climate Anomaly.

354

A. White et al. / Quaternary Research 84 (2015) 348–357

After ca. 300 cal yr BP, Abies percentages increased significantly while those of Pseudotsuga declined. BCHAR also declined to very low levels, and FRIs lengthened to more than 250 yr, suggesting few fires on the landscape. The shift to Abies dominated forest in recent decades may reflect cool humid conditions of the Little Ice Age (ca. 1450– 1850 AD) (Steinman et al., 2012) and few if any fires. The Hobart Lake record contains no clear evidence of vegetation changes that might be associated with Euro-American activities, although the uppermost sediments were not recovered. History of Abies and Pseudotsuga in Oregon and Northern California A comparison of pollen records from sites west of the Cascade crest in Oregon and northern California helps clarify the history of Abies and Pseudotsuga, two of the dominant conifers in the region. These taxa were chosen because they are fairly abundant in the vegetation at most of the paleoecological sites in southwestern Oregon and northern California, and pollen stratigraphies from these sites show changes through time that are likely related to variations in climate. The history of P. menziesii has received the most attention (Tsukada et al., 1981; Worona and Whitlock, 1995; Briles et al., 2008; Gugger et al., 2010) because pollen type can be assigned to the species, and the pollen grains do not travel far from the source so low percentages of Pseudotsuga pollen can accurately identify the species presence. The abundance of Abies pollen is also fairly faithful to the local abundance of the conifer (Minckley and Whitlock, 2000). Other conifers were considered for the analysis, but interpreting their historical distribution based on pollen data proved more problematic. An analysis of Pinus, for example, was confounded by the high levels of production and wide distribution of its pollen, which far exceeded its actual distribution. Its pollen presence did not match well with the presence of the conifer. T. mertensiana pollen, in contrast, was not present at enough sites to show an interpretable pattern. In the southwestern Pacific Northwest, Abies pollen comes from Abies amabilis (Pacific silver fir), A. grandis, A. concolor, and A. magnifica (California red fir) (Fig. 6), and each species has slightly different climate requirements and fire sensitivity. A. amabilis in the southern Cascade Range occupies relatively cool wet conditions at high elevations (1000 to 1500 m elevation) (Franklin and Dyrness, 1988; Burns and Honkala, 1990; U.S. Geological Survey, 2006). Infrequent, highseverity fire regimes are characteristic of forests dominated by A. amabilis and its thin bark and shallow root system make the species especially susceptible to fire (Agee, 1993). The range of A. grandis is primarily west of the southern Cascade Range but also extends east of the Cascade Range in northern Oregon and Washington. A. grandis is largely restricted to low-elevation mesic settings, although small populations grow at drier mid-elevations around Bolan and Hobart lakes (Fig. 6) (U.S. Geological Survey, 2006). A. grandis is moderately resistant to fire, occupying areas that burn infrequently and only under the most severe conditions (Burns and Honkala, 1990). A. concolor extends farther south than that of the other three Abies species and is abundant in the Sierra Nevada and southern Cascade Range (Fig. 6). Seedlings are highly susceptible to fire, but become more resistant with age (Burns and Honkala, 1990; Agee, 1993). A. magnifica is a high-elevation species that prefers moist and cool to cold climates where temperatures rarely exceed − 30°C in the winter or 30°C in the summer (Burns and Honkala, 1990). Forests dominated by A. magnifica experience moderate-severity fire regimes. Young trees are susceptible to fires, but acquire some resistance with age as their bark thickens (Agee, 1993). A. lasiocarpa is a common species at high elevations farther north in the Pacific Northwest, but in the southern Cascade Range it reaches the limits of its distribution.

P. menziesii var. menziesii has a distribution that extends from British Columbia to northern California (Fig. 6). As with A. concolor, young Pseudotsuga are vulnerable to fires but older trees (N 100–150 yr) develop thick bark that resists ground fires (Burns and Honkala, 1990). This helps the species live as long as 800 yr in this region (Burns and Honkala, 1990). The pollen data from Hobart Lake and ten previously published paleoecological records (Fig. 6) (Sea and Whitlock, 1995; Worona and Whitlock, 1995; Mohr et al., 2000; Briles et al., 2005, 2011; Daniels et al., 2005) suggest that Abies was abundant in the central Coast Range during the pre-glacial and glacial interval, prior to ca. 16,000 cal yr BP. A regional shift to warmer and wetter conditions after ca. 16,000 cal yr BP resulted in a contraction of Abies distribution in the Coast Range, and Abies (likely A. grandis) did not regain abundance until ca. 2000 cal yr BP (Worona and Whitlock, 1995). An exceptional expansion of Abies (inferred to be A. amabilis) occurred at Indian Prairie Fen in the Cascade Range during the late-glacial period (ca. 13,000 to 10,000 cal yr BP) and then again after ca. 8000 cal yr BP (Fig. 6) (Sea and Whitlock, 1995). An increase in Abies is also noted at Hobart Lake from ca. 8000 cal yr BP onward. In the Siskiyou Mountains, data from Bolan and Sanger lakes show remarkably similar trends in Abies pollen, which may have come from A. concolor, A. grandis, A. magnifica, or some combination of the three (Briles et al., 2008). At those sites, Abies was present through the Holocene, although its abundance declined ca. 10,000 cal yr BP and increased again at ca. 6500 cal yr BP at Sanger Lake and ca. 8000 cal yr BP at Bolan Lake. The patterns suggest an upslope shift in distribution in the early Holocene, followed by a downslope expansion in the middle and late Holocene. More frequent fires in the early Holocene may also have contributed to the Abies decline at Bolan and Sanger lakes (as well as at Indian Prairie Fen), and it expanded at these sites in the late Holocene as fire activity decreased. In summary, Abies was abundant during the late-glacial period except at sites with ultramafic substrate. Its range and/or abundance contracted during the early Holocene, and it gradually became more widespread and abundant in the middle and late Holocene. These temporal changes confirm the preference of Abies for cooler climates and lower fire frequency, especially at middle elevations. The history of Pseudotsuga in the western U.S. has been discussed in several pollen-based studies (Tsukada et al., 1981; Worona and Whitlock, 1995; Briles et al., 2008; Gugger et al., 2010). Little Lake data show that Pseudotsuga was present but uncommon in the central Coast Range until the early Holocene (Fig. 6) (Worona and Whitlock, 1995). Worona and Whitlock (1995) speculated that a glacial refugium for Pseudotsuga existed to the south of Little Lake and that the species migrated northward during the late-glacial period. Pseudotsuga pollen was present at Bolan and Sanger lakes in the early late-glacial period, suggesting that possibility of low-elevation glacial refugia in the Klamath–Siskiyou region. Warm dry conditions in the early Holocene apparently allowed Pseudotsuga abundance to increase and become the dominant species in the central Coast Range by ca. 7300 cal yr BP. A similar pattern of abundance occurred in the western Cascade Range at Indian Prairie Fen (Sea and Whitlock, 1995). As temperatures cooled and fires became less frequent in the middle Holocene, Pseudotsuga abundance decreased in the Coast and Cascade ranges from their early-Holocene levels (Sea and Whitlock, 1995; Worona and Whitlock, 1995). Effectively moister conditions in these wet settings likely limited this drought-tolerant species (Minore, 1979). Pseudotsuga was never abundant at high elevations in the Cascade Range, even during the early Holocene, based on its low representation at Gold Lake Bog (Sea and Whitlock, 1995). In the southern Cascade Range, Pseudotsuga was barely present (1–2% of the pollen) when the Hobart Lake record began at 7900 cal yr BP, suggesting that the early-Holocene climate

Fig. 6. Abies and Pseudotsuga pollen percentages at paleoecological sites in western Oregon and northern California (U.S. Geological Survey, 2006). The light green dot indicates Abies pollen abundance (%) and a dark green dot indicates Pseudotsuga abundance. A black dot at a site indicates that the pollen record does not extend that far back in time and Abies and Pseudotsuga abundance is therefore unknown.

A. White et al. / Quaternary Research 84 (2015) 348–357

355

356

A. White et al. / Quaternary Research 84 (2015) 348–357

was also not ideal for the species even at middle elevations. Pseudotsuga abundance steadily increased during the late Holocene, and by ca. 900 cal yr BP, it was a dominant species. Siskiyou sites, such as Bolan and Sanger lakes, show a similar distribution pattern to that of Hobart Lake (Briles et al., 2008). Pseudotsuga was present in consistently low levels throughout the late-glacial and early-Holocene periods, implying the persistence of small populations. It was not until the late Holocene that climate and fire conditions were suitable for its expansion. In the Klamath Mountains, Pseudotsuga is and was absent from the Mumbo Lake record and barely registered in the pollen records at Cedar and Crater lakes (Mohr et al., 2000; Briles et al., 2011), all of which lie on ultramafic substrates that are unsuitable for Pseudotsuga (Alexander et al., 2006). Nonultramafic southern Klamath sites (Taylor and Campbell lakes) have Pseudotsuga records similar to those of southern Cascade and Siskiyou sites (Briles et al., 2011), in showing a late Holocene increase in abundance. Despite the broad climate space currently occupied, Pseudotsuga became most abundant under a specific set of climatic conditions. Northern low elevations, which are relatively wet, had highest levels of Pseudotsuga, during the early Holocene dry period. Southern drier sites were too dry to support Pseudotsuga in the early Holocene, and the species became abundant only with the onset of cooler wetter conditions in the middle and late Holocene. Conclusions Hobart Lake provides a high-resolution history of vegetation and fire from the west side of the southern Cascade Range, spanning the last 7900 yr. These data show changes consistent with large-scale variations in the climate system, particularly the importance of slowly varying changes in the seasonal cycle of insolation and the indirect effects of insolation on the size and strength of the northeastern Pacific subtropical high-pressure system. During the early and middle Holocene, when the climate was warm and dry, xerothermic species, such as P. ponderosa, J. occidentalis, and C. decurrens, were more abundant than at present and fires were more frequent than today. With cooler wetter conditions in the late Holocene, mesophytic taxa, such as Abies and Pseudotsuga, flourished and fire frequency declined. Fluctuations in Cupressaceae, Abies and Pseudotsuga pollen also occurred on sub-millennial time scales in response to changing fire activity (Fig. 5). Following a very large fire episode at ca. 900 cal yr BP, both Abies and Pseudotsuga decreased in abundance, in contrast to previous fire episodes in which Abies decreased after fires but Pseudotsuga increased. This may suggest that this widespread fire initiated forest structural changes that, along with the climate fluctuations, left long lasting effects across the region. The changing abundance and distribution of Abies and Pseudotsuga in southwestern Oregon and northern California reflect changes in fire activity and climate as well as site-specific differences related to elevation. Abies is not well adapted to fire and has a clear affinity for cool climates, as evidenced by its early-Holocene decline and its middle- and late-Holocene expansion. Pseudotsuga has been present in western Oregon and northern California during the glacial period, the warm and dry early Holocene, and the cool and wet late Holocene, although centers of abundance have shifted. Pseudotsuga was most abundant in the Coast Range and central Cascade Range during the early Holocene and less so during the late Holocene. In the southern Cascade Range and Siskiyou region, its abundance was limited in the early Holocene and greater in the late Holocene. These data suggest that drying of moist sites created suitable habitat for Pseudotsuga colonization when the climate was warmer and drier than at present in the early Holocene, but it was not until the late Holocene that dry sites were sufficiently moist to support the species. The paleoecologic record suggests that the forests of the southern Cascade Range have been highly sensitive to changes in climate and fire frequency, and similar changes can be anticipated moving forward.

Temperatures are expected to rise in this region 1.7 to 5.8°C by 2100, with more of this warming occurring during summer than winter (Cayan et al., 2006; Mote and Salathé, 2010). Vegetation models suggest that middle-elevation conifer forests of southern Oregon, as seen at Hobart Lake, will shift to a mixed-evergreen forest as a result of warmer and effectively drier conditions (Lenihan et al., 2008). The insights gained Hobart Lake record and other sites in the southern Pacific Northwest imply that species responses to these changes will likely be complex. Species poorly adapted to drought and fire, such as Abies, will likely become less abundant at middle elevations due to increased moisture stress and fire activity and confined to higher elevations and mesic habitats. Other species at their biogeographic limits in the southern Cascade Range (e.g., J. occidentalis at its western distributional limit and P. menziesii, Acer macrophyllum, and Q. garryana at their eastern limits; (Thompson et al., 1999; U.S. Geological Survey, 2006)) may undergo substantial range shifts as well. As the future unfolds, paleoecological data provide an important ecological baseline on which to anticipate future vegetation and fire conditions in the southern Cascade Range. Acknowledgments This research was supported by a USDA Forest Service Challenge Cost-share agreement grant (10-JV-11272162-044) and a Mazamas Graduate Student Research grant. We thank James Benes, Paul Bodalski, Caitlyn Florentine, Elmira Wan, Scott Starratt, Jennifer Kusler, Holly Olson, and Carl Skinner for field, laboratory and analytical support. Dave McWethy, Greg Pederson, and two anonymous reviewers provided helpful suggestions on the manuscript. References Agee, J.K., 1993. Fire Ecology of Pacific Northwest Forests. Island Press, Washington, D.C. (493 pp.). Alexander, E., Coleman, R., Keeler-Wolf, T., Harrison, S., 2006. Serpentine Geoecology of Western North America. Oxford University Press, New York, NY. Arnold, J., Walsh, M., 2010. California's Ancient Past: From the Pacific to the Range of Light. Society for American Archaeology Press, Washington, DC. Barron, J.A., Heusser, L., Herbert, T., Lyle, M., 2003. High-resolution climate evolution of coastal northern California during the past 16,000 years. Paleoceanography 18, 1020–1029. Bartlein, P.J., Anderson, K.H., Anderson, P.M., Edwards, M.E., Mock, C.J., Thompson, R.S., Webb, R.S., Whitlock, C., 1998. Paleoclimate simulations for North America over the past 21,000 years: features of the simulated climate and comparisons with paleoenvironmental data. Quaternary Science Reviews 17 (6–7), 549–585. Bennett, K.D., Willis, K.J., 2001. Pollen. In: Smol, J.P., Birks, H.J.B., Last, W.M. (Eds.), Tracking Environmental Change Using Lake Sediments. Kluwer Academic Publishers, Dordrecht. Berger, A.L., 1978. Long-term variations of caloric insolation resulting from Earth's orbital elements. Quaternary Research 9, 139–167. Blaauw, M., 2010. Methods and code for ‘'classical’ age-modelling of radiocarbon sequences. Quaternary Geochronology 5, 512–518. Bowerman, N.D., Clark, D.H., 2011. Holocene glaciation of the central Sierra Nevada, California. Quaternary Science Reviews 30 (9–10), 1067–1085. Briles, C., Whitlock, C., Bartlein, P.J., 2005. Postglacial vegetation, fire, and climate history of the Siskiyou Mountains, Oregon, USA. Quaternary Research 64 (1), 44–56. Briles, C., Whitlock, C., Bartlein, P.J., Higuera, P., 2008. Regional and local controls on postglacial vegetation and fire in the Siskiyou Mountains, northern California, USA. Palaeogeography, Palaeoclimatology, Palaeoecology 265 (1–2), 159–169. Briles, C., Whitlock, C., Skinner, C.N., Mohr, J., 2011. Holocene forest development and maintenance on different substrates in the Klamath Mountains, northern California, USA. Ecology 92 (3), 590–601. Burns, R., Honkala, B., 1990. Silvics of North America. U.S. Department of Agriculture, Forest Service, Washington, DC (675 pp.). Cayan, A. Luers, Franco, G., Croes, B., 2006. Climate change scenarios for California: an overview. California Energy Commission PIER working paper. Colombaroli, D., Gavin, D.G., 2010. Highly episodic fire and erosion regime over the past 2,000 y in the Siskiyou Mountains, Oregon. Proceedings of the National Academy of Sciences 107 (44), 18909–18914. Crawford, J.N., Mensing, S., Lake, F.K., Zimmerman, S.R., 2015. Late Holocene fire and vegetation reconstruction from the western Klamath Mountains, California, USA: a multi-disciplinary approach for examining potential human land-use impacts. The Holocene 25 (8), 1341–1357. Daniels, M.L., Anderson, R.S., Whitlock, C., 2005. Vegetation and fire history since the Late Pleistocene from the Trinity Mountains, northwestern California, USA. Holocene 15 (7), 1062–1071.

A. White et al. / Quaternary Research 84 (2015) 348–357 Dean Jr., W., 1974. Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: comparison with other methods. Journal of Sedimentary Petrology 44 (1), 242–248. Dearing, J., 1999. Environmental Magnetic Susceptibility: Using the Bartington MS2 System. Chi Publishing, Kenilworth, England. Faegri, K., Iversen, J., 1975. Textbook of Pollen Analysis. Third ed. Hafner Press, New York. Franklin, J.F., Dyrness, C.T., 1988. Natural Vegetation of Oregon and WashingtonRep. Oregon State University Press, Oregon State University, Corvallis, Oregon. Graham, N.E., Hughes, M.K., Ammann, C.M., Cobb, K.M., Hoerling, M.P., Kennett, D.J., Kennett, J.P., Rein, B., Stott, L., Wigand, P.E., Xu, T., 2007. Tropical Pacific — mid-latitude teleconnections in medieval times. Climate Change 83, 241–285. Grimm, E.C., 1987. CONISS: a Fortran 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers & Geosciences 13 (1), 13–35. Gugger, P.F., Sugita, S., Cavender-Bares, J., 2010. Phylogeography of Douglas-fir based on mitochondrial and chloroplast DNA sequences: testing hypotheses from the fossil record. Molecular Ecology 19, 1877–1897. Higuera, P.E., Brubaker, L.B., Anderson, P.M., Hu, F.S., Brown, T.A., 2009. Vegetation mediated the impacts of postglacial climate change on fire regimes in the South-Central Brooks Range, Alaska. Ecological Monographs 79 (2), 201–219. Higuera, P.E., Gavin, D., Bartlein, P., Hallett, D.J., 2010. Peak detection in sediment charcoal records: impacts of alternative data analysis methods on fire-history interpretations. International Journal of Wildland Fire 19, 996–1014. Jarvis, D., Leopold, E.B., Liu, G., 1992. Distinguishing pollen of deciduous oaks, evergreen oaks and certain rosaceous spp. of S.W. Sichuan Province, China. Review of Palaeobotany and Palynology 75, 12. Kapp, R., Davis, O., King, J., 2000. Pollen and Spores. Second ed. American Association of Stratigraphic Palynologist Foundation, College Station, Texas. Lake, F.K., 2013. Historical and cultural fires, tribal management and research issue in Northern California: trails, fires and tribulations. Occasion: Interdisciplinary Studies in the Humanities 5, 22. Lenihan, J., Bachelet, D., Neilson, R., Drapek, R., 2008. Response of vegetation distribution, ecosystem productivity, and fire to climate change scenarios for California. Climatic Change 87 (1), 215–230. Leopold, E.B., Burkebek, J., Reinink-Smith, L., Jayachandar, A.P., Narváez, P., Zaborac-Reed, S., 2012. Pollen morphology of the three subgenera of Alnus. Palynology 36 (1), 131–151. Long, C., Whitlock, C., 2002. Fire and vegetation history from the Coastal Rain Forest of the Western Oregon Coast Range. Quaternary Research 58, 215–225. Long, C., Whitlock, C., Bartlein, P.J., 2007. Holocene vegetation and fire history of the Coast Range, western Oregon, USA. Holocene 17 (7), 917–926. Long, C., Power, M., Minckley, T., Hass, A., 2014. The impact of Mt Mazama tephra deposition on forest vegetation in the Central Cascades, Oregon, USA. The Holocene 24 (4), 8. Marcott, S.A., Fountain, A.G., O'Connor, J.E., Sniffen, P.J., Dethier, D.P., 2009. A latest Pleistocene and Holocene glacial history and paleoclimate reconstruction at Three Sisters and Broken Top Volcanoes, Oregon, U.S.A. Quaternary Research 71, 181–189. McAndrews, J., Berti, A., Norris, G., 1973. Key to the Quaternary Pollen and Spores of the Great Lakes Region. Royal Ontario Museum, Life Sciences, Toronto, Canada (61 pp.). Minckley, Whitlock, C., 2000. Spatial variation of modern pollen in Oregon and southern Washington, USA. Review of Palaeobotany and Palynology 112, 97–123. Minckley, Whitlock, C., Bartlein, P.J., 2007. Vegetation, fire, and climate history of the northwestern Great Basin during the last 14,000 years. Quat. Sci. Rev. 26 (17–18), 2167–2184. Minore, D., 1979. Comparative autecological characteristics of Northwestern tree species: a literature review. Pacific Northwest Forest and Range Experiment Station. U.S. Department of Agriculture, Forest Service, Portland, OR (72 pp.).

357

Mohr, J.A., Whitlock, C., Skinner, C.N., 2000. Postglacial vegetation and fire history, eastern Klamath Mountains, California, USA. Holocene 10 (5), 587–601. Mote, P., Salathé Jr., E., 2010. Future climate of the Pacific Northwest. Climatic Change 102 (1), 29–50. Odion, D., Sarr, D., 2007. Managing disturbance regimes to maintain biological diversity in forested ecosystems of the Pacific Northwest. Forest Ecology and Management 246, 57–65. Olson, D., Dellasala, D.A., Noss, R., Strittholt, J., Kass, J., Koopman, M., Allnutt, T., 2012. Climate change refugia for biodiversity in the Klamath–Siskiyou Ecoregion. Natural Area Journal 32 (1), 10. Porter, S.C., Denton, G.H., 1967. Chronology of neoglaciation in the North American Cordillera. American Journal of Science 265, 177–210. PRISM Climate Group, 2014. Oregon State University (edited). Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac, F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C.E., 2009. IntCal09 and Marine09 radiocarbon age calibration curves, 0–50,000 years cal BP. Radiocarbon 51, 1111–1150. Sea, D.S., Whitlock, C., 1995. Postglacial vegetation and climate of the Cascade Range, Central Oregon. Quaternary Research 43 (3), 370–381. Steinman, B.A., Abbott, M.B., Mann, M.E., Stansell, N.D., Finney, B.P., 2012. 1500 year quantitative reconstruction of winter precipitation in the Pacific Northwest. Proceedings of the National Academy of Sciences 109, 11619–11623. Stuiver, M., Reimer, P.J., Reimer, R.W., 2010. CALIB 6.0 (edited). Taylor, A.H., Skinner, C.N., 1998. Fire history and landscape dynamics in a latesuccessional reserve, Klamath Mountains, California, USA. Forest Ecology and Management 111 (2–3), 285–301. Thompson, R.S., Anderson, K.H., Bartlein, P., 1999. Atlas of relations between climatic parameters and distributions of important trees and shrubs in North AmericaRep. U.S. Geological Survey. Tsukada, M., Sugita, S., Hibbert, D.M., 1981. Paleoecology of the Pacific Northwest. I. Late Quaternary vegetation and climate. International Association of Theoretical and Applied Limnology, Proceedings 21, 730–737. U.S. Geological Survey, 2006. Digital Representations of Tree Species Range Maps from “Atlas of United States Trees” by Elbert L. Little, Jr. (and Other Publications) edited, USGS (http://esp.cr.usgs.gov/data/atlas/little/). Western Regional Climate Center, D. R. I, 2013. Climate of Oregon (edited). Whitlock, C., Larsen, C., 2001. Charcoal as a fire proxy. In: Smol, H.J.B.B.J.P., Last, W.M. (Eds.), Tracking Environmental Change Using Lake Sediments. Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 74–97. Whitlock, C., Skinner, C.N., Bartlein, P.J., Minckley, T., Mohr, J.A., 2004. Comparison of charcoal and tree-ring records of recent fires in the eastern Klamath Mountains, California, USA. Canadian Journal of Forest Research 34 (10), 2110–2121. Whittaker, R.H., 1960. Vegetation of the Siskiyou Mountains, Oregon and California. Ecological Monographs 30 (3), 280–338. Whittaker, R.H., 1972. Evolution and measurement of species diversity. Taxon 21 (2/3), 213–251. Worona, M.A., Whitlock, C., 1995. Late quaternary vegetation and climate history near Little Lake, central Coast Range, Oregon. GSA Bulletin 107 (7), 867–876. Zdanowicz, C.M., Zielinski, G.A., Germani, M.S., 1999. Mount Mazama eruption: calendrical age verified and atmospheric impact assessed. Geology 27 (7), 621–624.