A 500-year record of fire from a humid coast redwood forest
Steven P. Norman. 2007. A 500-year record of fire from a humid coast redwood forest. A report to Save the Redwoods League.
Full text at Save the Redwoods League [here]
Selected excerpts:
ABSTRACT
California’s coast redwood (Sequoia sempervirens) forests have long been associated with moderately frequent to frequent fire, particularly in the southern and interior portions of the species range. The historical importance of fire in northern coast redwood forests is generally thought to be much less because lightning ignitions are rare, and cool coastal temperatures and summer fog ameliorate the fire hazard. Support for this climate-fire gradient hypothesis has been limited because of insufficient fire history data from the northern coast redwood range. Past efforts to test this hypothesis range-wide are made difficult because of methodological differences among studies and problems with scar preservation in redwood. This research revisits the fire history of an area thought to have experienced fire only a few times per millennium in Del Norte Coast Redwoods State Park. I found that fire frequency was substantially more frequent than previously thought. Between 1700 and 1850, mean fire intervals within 0.25 to 1 ha sample areas varied from 11 to 26 years. Fire intervals did not correspond to a latitudinal, coast-interior or a topographically defined moisture gradient. Instead, patterns of fire frequency better fit a cultural burning gradient inferred from the ethnographic and historical record. Areas close to aboriginal villages and camps burned considerably more often than areas that were probably less utilized. Summer season fires, the ones most likely set by the Native Tolowa for resource needs, were 10 years shorter than the mean fire interval of autumn season fires. In the dryer eastern portion of the study area, frequent fire resulted in unimodal or bimodal pulses of Douglas fir (Pseudotsuga menziesii) establishment suggesting moderate to high fire severity. Near a Tolowa village site, a frequent fire regime before the late 1700s initiated a pulse of Douglas fir establishment that dominated the forest canopy for centuries; long after the village was abandoned, possibly due to epidemic disease. While variability in coastal fog-stratus and drought may also influence fire regimes, these relationships provide a weaker explanation than human ignition history. Variable human and climate influence on old-growth redwood fire regimes suggests that old growth redwood forests are not in equilibrium, but are dynamic due to a long history of variable human influence. Remnant old growth forests are likely to continue to evolve in response to human management. Efforts by managers to restore and sustain these remarkable forests can be enhanced by understanding how complex histories give rise to biodiversity.
INTRODUCTION
Fire scientists and forest managers of the American West have long thought that differences in fire regimes among forests are largely explained by variation in the fuel. According to this model, the periodicity of fire reflects the time needed for fuel to accumulate to levels where fire will spread. It further presumes that the fire weather necessary for fire to burn occurs on a regular basis and that ignitions are not a limiting factor. During resent decades, research has demonstrated that fire regimes in the Western United States are strongly driven by broad-scale variation in regional and hemispheric climate (Swetnam 1992, Kitzberger et al. 2006). While combustible fuel is clearly important for fire spread, this discovery in the importance of broad-scale drivers has eroded the previously held belief that local conditions drive fire regimes as much as was once thought, with broad implications for climate change. Recent research has also shown that certain ecosystems are strongly ignition-limited. For example, the historical and modern fire regimes of coastal southern California are dependent on human ignitions more than lightning (Keeley 2002). If climatic controls on fire regimes are weak in such ignition-limited systems, local ignition factors rather than top-down controls define the fire regime. …
A very different view of fire regimes characterizes social science research in coast redwood. Early twentieth century ethnographers documented the pervasive land use traditions of Native American groups across the coast redwood range. Research has documented the extent to which the livelihoods of tribes were enhanced by active fire use (Anderson 2005 Lewis 1973). In parallel, archaeological evidence has documented the presence of humans in and near the coast redwoods for several times the live span of the oldest known coast redwood (Frederickson 1982). While cultural and tribal traditions appear to have been inconsistent over the millennia, the considerable length of human presence in and near the coast redwood suggests that ignitions may not have been a limiting factor in some coastal forests for a very long time. In contrast to the pattern expected from a broad-scale moisture gradient, a cultural burning model associates areas most in use by Native Americans with frequent fire, and in northwestern California, villages were concentrated along the coast and along riparian corridors (Waterman 1920; Baumhoff 1958). Such places are least likely to burn according with a moisture gradient hypothesis. In addition, the coastal tribes of California extensively utilized interior forests for procurement of food and materials. While extensive human influence across the coast redwood landscape has been questioned (Vale 2002), the actual impact of Native Americans remains an open question. …
In this paper, I describe the fire history of a humid coast redwood forest along a coast to interior gradient and across slopes. I then compare fire regimes to the known pattern of Native American land use described by early 20th century ethnographers to determine which of these two competing hypotheses best explains the historical fire regime of this coast redwood forest. …
STUDY AREA
Del Norte Coast Redwood State Park is located in central Del Norte County California and consists of old growth and second growth forest. It extends from the Pacific Ocean, south of Crescent City California, through the Mill Creek watershed to encompass the entire core portion of the coast redwood range. The Mill Creek and Rock Creek portions of the Park consist nearly entirely of second growth forest and were only recently acquired. By 1926, logging was restricted to the western edge of the Mill Creek region (Weber 1926) and it continued through 2000, shortly before the area became part of the Park (Stillwater Sciences 2002). …
The people known as the Tolowa have lived here for centuries and are thought to have been the primary sources of human ignitions before 1850. Based on a change in material culture seen in the archaeological record, the Tolowa are believed to have entered the area from Oregon around A.D. 1300, displacing an earlier people (presumably the ancestral Karuk). …
The seasonal pattern of Tolowa resource use was similar to that of neighboring tribes. Families had seasonal camps in the mountains where they went to gather acorns, harvest seeds and to hunt. For several weeks during the summer or early fall, the coastal villages were largely abandoned as family groups dispersed to these inland sites (Drucker 1937, Gould 1975). Tanoak acorn gathering was performed by women while men hunted for venison. Berries supplemented the Tolowa diet and were collected as they became available over the growing season. Salmon were procured from the Smith River and its tributaries including Mill Creek. The sea provided supplemental food and materials including fish, seals, mussels and whales. Seals were killed using sea-worthy canoes made of redwood.
The use of fire by the Tolowa is well-established in the ethnographic record. According to Drucker (1937; p. 233), “informants maintain that nearby hills were kept clear of brush by annual burning; this also improved the grass, so that deer frequented such clearings and could be shot easily.” He further relates that “Late spring, when the old fern was quite dry and the new growth just starting, is said to have been the time for burning off the hillsides to improve the hunting grounds” (p. 232). As with other neighboring tribes, tanoak acorns provided a primary food source and regular burning of tanoak groves reduced the filbert worm and filbert weevil infestation that would otherwise reduce the quality of the annual harvest (Anderson 2005). … Other reasons for burning tanoak stands may have included a desire to stimulate oak shoot and beargrass production for basketry, reduce the severity of fires, to encourage edible mushrooms, and to increase seed production in grasses (Anderson 2005). …
The ethnographic record does not tell us if the Tolowa or neighboring tribes on the North Coast deliberately burned significant portions of the coast redwood forest that had no tanoak component. To the south, the Lolangkok Sinkyone Indians were apparently responsible for the frequent fires in the pure redwood forests of the alluvial flats of Humboldt Redwoods State Park, but these areas were used as winter villages sites (Norman et al. 2003). Fires in the shaded understory may have reduced unwanted downed wood, duff and litter, or fires may have stimulated the production of specific herbs or shrubs of value that primarily grew in these humid sites, such as ferns (Lewis 1973). In other areas, fires may have only occasionally spread into redwood stands from adjacent forests that were actively burned. …
DISCUSSION
Fire regimes
The fire intervals documented from this study indicate that fire was substantially more frequent than previously thought in this portion of the humid north coast redwood forest. While areas with long fire intervals in the coast redwood forest probably do exist, it is unlikely that the difference between this research and that done previously in the area reflect site differences. …
Human modification of the redwood environment
Native Americans have lived in or near the Mill Creek watershed for thousands of years. The population of the region may have markedly increased with the arrival of the Tolowa about AD 1300 (Frederickson 1984), but temporal and spatial changes in human use across the landscape are difficult to reconstruct with certainty. Consistent with a cultural burning gradient, the fire record of individual sites corresponds well to what is expected from the ethnographic record. Sites near villages and camps (MBH, MVT, MGH and to some degree MWM) burned more frequently between 1700-1849 than more distant sites (MRC, MRR and MWT). …
… An active program of prescribed fire management in these and similar fire-prone interior forests may be the only way to restore desired compositional and structural attributes.
The ecological implications of fire exclusion in more coastal redwood forests is complex because of the historical importance of variable human ignitions in the past. The time since the last fire has only recently matched the long interval that occurred during the late 1700s and early 1800s, and 20th century logging has so altered the system that the role of fire is constrained by factors other than those of old growth reserves. Fire undoubtedly contributed to the biodiversity of this area’s landscape, however, as the fire-sensitive tree species that occur in the forest are also those that require gaps to establish (Viers 1982). According to early Soil-Vegetation maps (1952), few forests of pure redwood existed in this area before logging, as redwood was associated with Douglas fir, western hemlock, grand fir and Sitka spruce across topographic gradients. While this compositional diversity also reflects disturbance from windstorms, the frequency of fire in this forest indicates that a substantial fraction of it is likely to have resulted from fire. Prescribed fire may be needed on intermediate and interior sites to retain and restore desired structural and compositional elements.
An integrated theory of fire in redwood
The frequent fire found in this humid northern portion of the coast redwood’s range indicates that a latitudinal climate gradient model is not useful for predicting fire regimes before 1920. The mean and median fire intervals reported here are similar to those reported from sites much farther south (Table 1, Table 3). If a latitudinal gradient existed, it may be found by revisiting studies conducted to the south that may have under-represented fire frequency given their sampling methodologies. Had I collected samples above 30 cm height, I would have documented few fires on any site.
Similarly, a coast-interior climate gradient is not supported by the fire scar record. Areas closer to the coast than where I sampled are also unlikely to provide exceedingly long fire intervals, given the abundance of large fire-formed basal hollows there (Zielinski and Gellman 1999). Such cavernous basal hollows develop from frequent fire over centuries, not rare fires that allow scars to heal over (Finney 1996). Most areas closest to the coast are where Native American density was greatest.
Of interest, wildfires that burned during the suppression era appear to relate to both a natural coast-interior and latitudinal climate gradient (Oneal et al. 2006). This fact may reflect the ease with which modern fires can be suppressed in redwood given the slow rate of spread associated with cool, moist forests. In the past, during the period when old growth forests developed, fire frequency and extent were not related to suppression efforts, but to factors related to ignition and spread. The prevalence of early season fires in the historical record suggests that if fuel conditions allowed, fires could have burned from the early fire season in mid-summer through the first rains of October or November. Only moist coastal areas and sites farthest from Native American ignitions are likely to have been refugia from fire. Such sites appear to have been relatively uncommon in the landscape.
The historical fire regimes that occurred in this landscape correspond to a cultural fire gradient as defined by the ethnographic record. Short fire intervals occurred across the forest, but frequent fire regimes were restricted to the areas that were most likely influenced by Natives. Their early season fires were targeted toward specific resources uses, such as preparing the acorn grounds before harvest, and the higher frequency of early season fires is consistent with this pattern. Further, the mixed frequency of the coastal redwood sites is also consistent with a cultural burning model. The reduction of fire at the most ignition-sensitive coastal sites during the late 1700s and early 1800s can be explained by a change in human ignitions or by the effectiveness of human ignitions, given a change in climate. Over time, such variation in fire intervals and the characteristics of fire events, such as their season or intensity, can have a strong influence on floral assemblages (Bond and van Wilgen 1996). This spatial-temporal complexity of the natural climatic and cultural burning gradients likely contributed to the diversity of the redwood forest landscape (Martin and Sapsis 1992; Anderson 2006).
Archaeological and linguistic evidence suggests that like most Native peoples of the northern redwood forest, the Tolowa arrived relatively recently, around A.D.1300 (Frederickson 1984). Seven hundred years is well within the lifespan of a large minority of the redwood trees of the area’s forest. While other tribes lived in and near the area’s redwoods for millennia, migrations and natural increases in the resident population likely led to substantial variation in human ignitions over the centuries. This inconstant influence of human ignitions is analogous to variation in fire regimes caused by changes in climate (Swetnam 1992). Over time, changes in either ignition and the fire climate are likely to lead to not only diverse vegetation, but to non-equilibrium conditions. This is especially true for forests dominated by long-lived sprouting coast redwood that can retain attributes of prior environmental conditions for centuries.
