Fire

Red Hot Cycles of Renewal

#goodfire

Seeing is believing. I was lucky enough to be invited to a cultural fire Tribal training event in October of 2019, the Klamath Training Exchange (Klamath TREX).

According the the Mid Klamath Watershed Council, historically, the western Klamath Mountains experienced fires every 3 to 10 years. Fire suppression over the last 100 years and the prohibition of traditional Tribal burning has resulted in a huge fire deficit in our region. The use of prescribed fire may be the only viable long-term method for protecting our communities. Fire needs to be restored to the landscape for multiple other reasons as well: including for cultural resources, wildlife habitat, and general ecological functionality.

More on this informative field trip to the Klamath River on this page.

What follows is my 2010 book chapter on fire ecology in California, updated for 2022.

Landscape Patterns

I arrived in Aliso Campground at sunset, in a canyon in Los Padres National “Forest,” more of a shrubland with some oaks, and sat at a picnic table admiring the subtropical diversity of plants in the vicinity. Coast live oaks nestled in the canyon along with sycamores, Blue oaks, scrub oaks, Poison oaks, manzanitas, junipers, Bay trees, cottonwoods and a wide assortment of aromatic shrubs. Western bluebirds called softly among the live oaks, Oak titmice and Scrub jays bounded around camp, and California thrashers sang from the chaparral slopes above. I had the campground to myself in the summer heat. My intent was to study Cuyama Valley below me (the name taken from that of a Chumash Indian band), a broad remote river flat stretching west through the Coast Range from near Carrizo Plain to Santa Maria on the coast, a forgotten area of California that is one of my windows into the past. The imprints of fire are obvious here, the landscape showing multiple episodes of burning and vegetation renewal -- a visible pattern that I can read like the pages of a book (that is still being written).

Adjacent to the National Forest, large cattle spreads with names like Rancho Rinconada occupied the valley floor. Despite the grazing, native grasses can still be found commonly; coastal sage scrub, oak savanna, chaparral, and riparian glades still remain undisturbed by subdivision; and most intriguingly, the whole area has signs of recent burns of various ages, giving the plant communities a more pristine aspect than “protected” places. Fire still is a common occurrence in this valley.

In fact, one was burning as I arrived at camp -- way up on the chaparral ridgeline. I could see at dusk a swath of brown smoke blown eastward. The sun set into the smoke, becoming a bright ruby disc. The next morning I woke up to Poorwills calling at dawn in the chaparral above the live oaks, and soon I hiked up Aliso Canyon along a trail crossing the scrubby slopes. The blaze continued to burn on the mountain slope in the far distance; on the lower foothills it had burned the grass under an open group of oaks, leaving the trees unscathed. In the higher hills, however, the brush was completely ablaze, leaving only gray ash and black shrub skeletons. Some Big cone spruces (Pseudotsuga macrocarpa) were also burning like giant candles.

I examined the gray-green chaparral along the trail under a clear sun-filled blue sky: Ceanothus, Silktassel (Garrya), Sourberry (Rhus trilobata), California buckwheat (Eriogonum fasciculatum), Black sage (Salvia mellifera), Deerweed (Lotus scoparius), and some scattered Mountain mahogany (Cercocarpus ledifolius) and Bush poppy (Dendromecon rigida) grew with abundant Chamise (Adenostoma fasciculatum). Dry, robust, yellow Giant needlegrass bunches (Achnatherum coronatum) dotted open areas between dense shrubs. Whipple yuccas (Yucca whipllei) grew more abundantly than I had ever seen, covering many of the steep slopes with a carpet of blue-green spear-rosettes. They showed signs of old burns -- some blackened stems. Perhaps the yuccas increase in density with regular burning. The older burned scrub was more open, with many bunchgrasses, than the long-unburned patches. A Canyon wren called its long descending crescendo, and an Anna’s hummingbird fed on a scarlet-flowered penstemon. Wrentits laughed at me hidden, in the sage.

Arriving back in the afternoon, I found my once empty campground strewn with human bodies lying haphazardly on picnic tables and stretched across the ground in the shade of oaks. An exhausted fire crew had arrived from fire-fighting up on the ridge, and men lay resting, too tired to move in their smoked-yellow jackets. They were trying to extinguish the “Spanish Fire” that had burned 5,000 acres, named for an old adobe near its start. I never found out the cause, but it was late July and the monsoon season had been giving Los Angeles to the south repeated thunderstorms and flooding -- was it lightning? A few months later, a lightning bolt ignited another fire up on the ridge, bringing suppression crews back again to Los Padres National Forest.

I packed up and headed out to explore the old fire patterns on the valley itself. I sketched the vegetation mosaic of open grassland on the flats and terraces above the Cuyama River, and on parts of the sloping bajadas at the bases of the mountains. Old burns were evident, dominated by grass, mostly exotics such as wild oats (Avena spp.), and red brome (Bromus madritensis spp. rubens). But I also found several acres of native bunchgrass: pure stands of feathergrass (Stipa cernua) or mixes with bluegrass (Poa secunda) on the flats and rolling low hills. One section along the road had been just burned off and was blackened and charred, probably from a cgiarette tossed from a car passing by. The black bases of the Stipa bunchgrasses remained forming a dense pattern of dark spots on the bare dirt.

Patches of unburned cane cholla (Opuntia parryi) and prickly pear (O. phaeacantha) darkened the grassland in widely scattered groups, Feathergrasses sometimes poking up through their protective spiny arms.

California buckwheat scrub grew commonly on certain low hills around the valley edge, and in gullies and ravines interfingering with grassland. This scrub dominated the bajadas, and on old burns was less dense with shrubs more scattered. California sagebrush (Artemisia californica), Whipple yucca, and goldenbush (Ericameria sp.) grew in large patches on the valley edge slopes, mixing with patches of open grassland perhaps in a burn mosaic.

Live oaks lined the ravine riparian bottoms, and mountain bases had widespread Blue oak savanna. Chaparral of Chamise, manzanita, and Black sage grew in localized patches on lower hillsides with oaks and buckwheat scrub, but the really extensive chaparral areas were higher on the mountain ridges.

As a whole, the patterning of fire could be read on this land and its importance undoubted. I was sure the Cuyama Chumash had once played a large role in structuring this pattern before missionization. The later changing attitude towards fire upset the balance in California. Fire is one of the most crucial processes shaping the state, and every year lives and property are lost tragically when fuel, weather, and accident converge. Since the book was published in 2010, the megafire problem has grown to catastrophic proportions. I look out my window every summer living east of the Sierra Nevada and every summer now the sky is full of brown smoke and haze from wildfires burning out of control in western California.

Fire is a force that will not simply go away, as very many native vegetation communities show adaptations to cultural fire and renewal out of the ashes.

Burn mosaic

Ecologist P. V. Wells (1962) carried out a careful comparison of the vegetation mosaic around San Luis Obispo, just north of Cuyama Valley, to determine if the seemingly random jumble of grassland, scrubland, and oak woodland patches on the hills and valleys were related to geological substrates -- what grew on sandstone, basalt, clay, or volcanic ash? He found that certain plant species, such as specific manzanitas and Sargent cypress (Cupressus sargentii), were often limited to just one or two substrates, but that vegetation types were not related to the soils. They were apparently caused by the fire history of the area and by chance dispersal of seeds and nuts. Fire could interact with substrate, however, and he theorized that frequent fire in oak woodland could turn it into grassland on deep rich soils, but would turn into coastal sage scrub or chaparral on thin rocky soils.

At Gaviota State Park, on the coast south of Santa Maria, another study looked at repeat aerial photographs to see how the vegetation mosaic shifted through time. From 1947 to 1989 on unburned and ungrazed plots almost a third of grassland converted to coastal sage scrub, almost a third of coastal sage scrub converted to oak woodland or chaparral, 5 % of chaparral converted to oak woodland, and 10% of oak woodland converted back to grassland, possibly due to lack of recruitment of trees on drier sites. Dense Coast live oak woodland was more common on shady north slopes, and open oak savanna more on gentle drier topography. On plots that had fires in 1944, 1955, and 1986, grassland increased, coastal sage scrub decreased, and chaparral and oak woodland remained the same in extent. Areas outside the park that were grazed showed slower conversion rates, and in some areas coastal sage scrub was prevented from invading grassland. Fire strongly favored grassland over sage scrub, and slowed or halted the invasion of oaks into shrublands. The frequent fires also favored coastal sage scrub over chaparral, although on thin rocky soils chaparral appeared to remain unchanged indefinitely as an edaphic climax (soil climax). Not only soil, slope, aspect, fire, and grazing, but subtle biotic interactions between plants themselves could determine a “dynamic shifting mosaic” of vegetation patches on the landscape, and in some cases these are perhaps cyclic: the shift from grassland to coastal sage scrub to oak woodland, then back to grassland was observed in a few areas. And of course climatic fluctuations cannot be left out, which could over time change the transition rates between vegetation types (Callaway and Davis 1993).

Looking at the vegetation patches on the landscape I began to understand just how complex its history may be, a story of disturbances such as fire combined with cycles of succession and renewal of communites; and I think we are learning so far just a hint about this mystery of changing Nature.

I drove out east towards lonely Carrizo Plain to continue camping and studying signs on the

landscape. The fire on the mountain range burned in pockets of flames, and smoke whisped upwards, fingering into the sky and filling the whole region with an ashy haze.

The Process of Fire

Ecologists speak of the “fire regime” as the role fire plays in a natural community, its fire history. Included are measures such as the fire frequency and size, the intensity of the blaze and its pattern across the landscape. Determining these factors are the amount of dead vegetation that can as fuel, the moisture content of the vegetation, its vertical and horizontal structure, the continuity of fuels, and floristic make-up. Topography acts to mold a burn: heat goes upwards and fire accelerates upslope; south-facing slopes have higher temperatures that may increase fire intensity, yet moister north-facing slopes may have more fuel build-up. Weather, of course, also has a great influence on fire behavior -- higher air temperatures allow fire to ignite more easily, low humidity cause fires to burn more inrensely, and winds “feed” fire more oxygen, and direct the flames to spread (Reiner 2007).

My colleague Miriam Morrill is an expert on how weather, topography, and many other observable factors infuelce fire starts and spread on the landscape. I have learned so much from her about fire on the land.

See Miriam Morrill Pyroskectchology.

Marsh Fires in Old California

We do not know what the historic fire regimes were in California’s various plant communities, although guesses can be made using historical accounts and current studies of fire ecology.

Quite a few native habitats were apparently adapted to burning by wildfire and Indian-managed fire before Europeans and Americans came with differing philosophies about how to interact with the landscape.

In 1860 an observer wrote in Scenes of Wonder and Curiousity in California:

“An apparently interminable sea of tules extends nearly one hundred and fifty miles south, up the valley of the San Joaquin; and when these are on fire, as they not infrequently are, during the fall and early winter months, the broad sheet of licking and leaping flame, and the vast volumes of smoke that rise, and eddy, and surge, hither and thither, present a scene of fearful grandeur at night, that is suggestive of some earthly pandemonium” (Bohn 1969).

Biologist and native Ohlone Chuck Striplin of the San Francisco Bay Institute stressed to me as I consulted with him on a reconstruction painting of bay marshes, how an updated version of aboriginal life needed to be shown: the Indians burned the grasslands, they burned the marshes, “they burned everything,” he said (Pers. comm. October 10, 2004). I got the distinct impression that to him no more beautiful scene could be painted than an orange blaze engulfing a tidal marsh edge centuries ago.

Burning was used by the tribes in rabbit drives, to flush deer in hunting, to increase the number and quality of basketry plants, to clear the ground under oaks and pines, to encourage new growth of grasses and forbs as wild crops, to keep wildfires from damaging villages by controling local growth, and to increase ecotones between vegetation community patches that favored game animals and successional plant resources (Anderson 1993, Lewis 1993, Bean and Lawton 1993).

Marshes and wet meadows in the past may have been kept in a sort of dynamic equilibrium by alternating years of rainy flooding and dry drought. Fires favored herbaceous plants over woody ones, and so repeated marsh fires in the past probably got rid of a lot of trees. The only exception to this rule was willow (Salix), which will increase after a fire. Fires also stimulated plant productivity as minerals were released into the soil. Burns increased marsh plant growth rate, the size of the plants, and their seed production, especially three to five years after a fire; after that productivity diminished as old dead matter builds up again (Young 1986).

Two kinds of wetland fires could be recognized: surface fires and deep fires. In surface fires only above-ground or above-water vegetation burned, leaving the roots and rhizomes intact. Marsh plants like Common spikerush (Eleocharis palustris) and Common reed (Phragmites australis) were fire-adapted because of their underground rhizomes: the top of the plant was killed but resprouted from the rhizomes. Threesquare (Schoenoplectus pungens) had rhizomes buried six inches deep and so was well-protected from fire -- it regrew starting a week after a burn.

During long droughts, however, a second type of fire burned through the marsh, a “peat fire” where the substrate was so dry that the burn reached the deep root layer and accumulated organic mat. These root fires could reduce productivity, change the plant species composition, and even produce open water within the marsh. This was actually good, as open water spots created good habitat for fish, as well as spots for geese and ducks to feed and rest in. Dense climax stands of marsh vegetation were removed, allowing colonizers like Ditch-grass (Ruppia maritima) and Goosefoot (Chenopodium rubrum) to come in, valuable bird foods. The marsh was renewed.

In southern California, marshy vegetation with small spring-fed lakes were called cienagas (“hundred waters”) in the Spanish period, and were scattered about the Los Angeles and San Gabriel plains and elsewhere to Arizona, New Mexico, and Sonora, Mexico. Other cienagas were created from the seasonal overflow of rivers (Gumprecht 2001). Rushes (Juncus), sedges (Eleocharis, Carex), and grasses such as Creeping wildrye grew around these waterholes, grading into surrounding meadows. O. K. Davies at the University of Arizona studied pollen cores from several of these habitats and found a similar pattern of change following European settlement, a drastic decrease in fire frequency that allowed trees like willows and cottonwoods to invade these wetlands. Frequent burning prior to European arrival kept these marshes open and herb-dominated, with silty soil; after the fire decreased old marsh plant material built up to form a peaty soil (Davies n.d.).

Fire probably provided a necessary landscape disturbance to Western toads (Bufo boreas), renewing wetlands needed for breeding by opening up shallow pools that otherwise would be matted with old vegetation. Toads may actually be declining due to fire exclusion in the northern Rocky Mountain wetlands, according to a US Geological Survey study initiated by Blake Hossack and P. S. Corn.

So much of the restoration biology I have participated in has involved manual removal of excess vegetation from marshes, ponds, and lakeshores to benefit rare animals and fish. This means getting some dirt on the hands and cutting back aggressively-growing aquatic plants like bulrush with rice-harvesting knives on long poles or machetes. This is in my view all a necessary replacement for the marsh fires that used to burn through, or grazing herds of elk or bison, or floods -- natural cyclic processes that kept areas of open water more in balance with densely vegetated areas.

Grassfire

Interior grasslands may have burned every 1 to 5 or several years according to botanists (Anderson, Barbour, and Whitworth 1998, Reiner 2007). So regular and common was burning, mostly from Indian land management, that ecologist James Barry (2003) called these communities “pyric (fire) grasslands.”

The accounts of Spanish explorers were filled with descriptions of grasslands burned off by native people. Father Crespi’s party camped near a village on the coast just north of Santa Barbara in August 1769, and he recorded that the place was “well covered with very fine grasses that nearly everywhere had been burnt off by the heathens” (A. K. Brown translation in Johnson and Earle 1993). While near San Mateo in October the expedition found areas burnt over, and scouts reported the whole country to the north and northeast of the South Bay impassable to their stock because the pasture was burned off by the natives (Pourade 1968).

Governor of California from 1774 to 1777, Fernando Rivera y Moncapa, wrote that between San Gabriel and the Santa Clara River on the broad plains there was no fodder for the horses and mules because of the “great fires of the Gentiles, who burn the fields as soon as they gather up the seeds....” He continued to describe how burning “is universal although on some occasions it happens that it may be greater or less, according to the winds or calm” (Johnson and Earle 1993). Jose Joaquin De Arrillaga wrote a letter at Mission Santa Barbara warning the “Christian Indians, and particularly the old women” not to set fires to the pastures (ibid.).

The Kumeyaay and Cahuilla of southern California burned grasslands from summer into fall. In what sounds like a carefully planned rotational burn the Kumeyaay torched some of the coastal sage scrub and chaparral slopes every five to 10 or 15 years, but left others for every 20 to 30 years to keep some sage for use (Bean and Lawton 1993). The Wintu caught grasshoppers for food by burning patches of grassland (Lewis 1993). In his studies on fire ecology, Omer C. Stewart noted the significance of Indian burning to grassland ecology: the Pomo women of Ukiah Valley set fire to the grasslands after collecting seeds for pinole; an elder told Stewart, “The grass was burned every year. The fires were started and allowed to burn in every place. Burning was to make the weeds grow better and to keep down the brush” (Lewis 1993). Sometimes the seeds of Clarkias and Mule-ears (Wyethia) were sown into recent burns (Anderson 2007). As among native people in Australia, Californians practiced a form of “firestick farming,” using patch-burning techniques instead of the plough to stimulate edible plant growth.

Ecologist Susan Bicknell studied the phytoliths of the soils of coastal prairies along northern California shores, and said that much of this habitat may be anthropogenic from Indian burning. It is reverting to woody vegetation after suppression of fire following Euro-American settlement (Bicknell, Austin, Brigg and Godar 1992; Anderson, Barbour, and Whitworth 1998).

Ethnobotanist M. Kat Anderson in her wonderfully detailed book Tending the Wild, made the point that early California was a cultural landscape, not the wilderness of European description. Burning, digging, transplanting, clipping, and other small-scale disturbances helped shape the California we inherited (Anderson 2005).

The disturbance adaptations of Purple needlegrass (Stipa pulchra) may be related to frequent burning -- seedlings grew well after fire, especially in the second year (Keeley 1990). Other species also sprouted seedlings after burns, such as Foothill needlegrass (S. lepida), Squirreltailgrass (Elymus elymoides and E. multisetus), lupines (Lupinus), clovers (Trifolium), milkvetches (Astralagus), and lotus (Lotus).

Visit Point Pinole Regional Shoreline (Contra Costa County) by San Fransisco Bay to explore a good native grassland relict that has been regularly control-burned for decades, sometimes the same ground annually. The Purple needlegrasses, California oatgrasses, and swards of Creeping wildrye are lush and dense and attest to the power of fire to renew the land.

With fire suppression, today’s grasslands may burn only once every 20 or 30 years or longer (Reiner 2007). The changed structure of grasslands may also affect the fire regime -- the continuous cover of introduced annual grasses will carry a hotter, farther-spreading burn than the patchy bunchgrasses. Normally grassfires probably burned “cool” and the heat did not penetrate far below the soil.

I talked with biologist Leslie Backus about her research on Florida prairies. After conducting control burns on various plots, she looked at how the fires affected the little Grasshopper sparrow (Ammodramus savannarum). These birds needed grass to build their nests in, and did not favor shrubs invading their open grassland territories. On the burned ground the grasses began growing back fast, within weeks, and soon the sparrows set up their small, quarter-acre territories. Of many different burn regimes, the Grasshopper sparrow territories were densest on plots burned every 3 to 5 years. If unburned for 5 to 10 years, shrubs invaded the grasslands and Grasshopper sparrows declined in density (L. Backus personal communication 2004).

In the East Bay ridge grasslands, I found Grasshopper sparrows singing their insect-like songs only in open prairies free of Coyote brush, lightly grazed by cattle but with plenty of taller standing grass. The cattle here imitated a fire regime in reducing the fuel and eliminating shrub seedlings, and in the past herds of tule elk would have accomplished this. Grasshopper sparrows were very rare in these regional parks, and absent from more heavily grazed cattle pastures were only short grass prevailed. They need a mix of short grass and tall grass, which often does not happen in livestock pastures.

The fire responses of expanding south coastal scrub plants indicated to researchers that before 1700 the grasslands around San Diego may have burned every 10 years or less, excluding the shrubs that were intolerant of fire until recent decades of fire control (Cox 1986). This may have been the case in the East Bay hills as well.

Grass-shrub mosaic

Yet since my research in the 1990s and the book publishing in 2010, I have come to appreciate much more the grass-shrub mosaic that was likely shifting with cultural fire patch burns over thousands of years. Shrubs such as coyote brush (Baccharis pilularis) are important habitat for many bird species, rodents, bobcats, foxes, and other species, and there is increasing evidence that a patch-mosaic of open grassland and shrubland provides the maximum biodiversity. Pollen evidence indiactes that a mix of grass-shrub habitats was most likely the norm over several thousand years in the past, with ever-shifting boundaries between grass and shrub. Indigenous land managers were likely well aware of this, using cultural fire to manage this biodiversity and habitat complexity, in ways that Western science is still catching up with. My field observations of many bird species using coyote brush habitats in grassland areas follow below.

Coyote brush has been much maligned by biologists and land managers recently as an "invader" species into grassland, but I think this terminology unnecssarily invokes colonial historic terminology and oversimplifies the complex Holocene changing grassland-shrubland interface.

For example, a pollen-spore-charcoal core analysis of one site in Point Reyes National Seashore,

going back to 7,000 years ago, shows the presence of coyote brush and grasslands prior to European colonization (Anderson et al. 2013). This argues against the position that coyote brush expands "unnaturally" as a "weed" without livestock grazing. Depending on which point in time in this core you

look, Baccharis and California sagebrush (Artemisia californica) shrubs have expanded and contracted in coastal grasslands according to the pollen counts over the millenia, and Indigenous cultural burning had periods of greater and lesser intensity. Climate data also indicate periods of aridity and increasing moisture. But coyote brush scrub has always been present as a native plant community, perhaps most like what we see at Tomales Point with elk grazing. The pollen analysis found coyote brush to be a natural part of plant communities in Point Reyes National Seashore:

Charcoal influx in sediments prior to c. 4000 cal. Before Present (BP) is minimal, but small amounts of charcoal were deposited consistently in sediments between c. 4000 and 2200 cal. BP, with a higher influx after 2540 cal. BP. After c. 4000 cal. BP, pollen evidence suggests the beginning of development of the modern vegetation characteristics around the study site at Glenmire, not only with abundant conifers, alder, and dogwood but also with continuing importance of open coyote bush coastal scrub grassland. During the middle and late Holocene at Glenmire, the site was surrounded by a

mosaic of vegetation types, including mixed conifer forest with coastal scrub grassland prior to c. 4000 cal. BP. Subsequently, the site witnessed an increase in hardwoods such as alder and expansion of coastal scrub (California sagebrush and coyote brush) until c. 2200 cal. BP when tan oak, and particularly Doug fir and coaast redwood, expanded.

The pollen core samples apparently record the beginning of European Spanish mission colonization, and then rapid settlement, resulting in drastic changes in vegatation: Anderson et al. (2013) then describe significant changes occurred over a very short period of time, beginning in the late 1790s but accelerating in the mid-1800s. These included the arrival of the Spanish Franciscans, removal of the native Coast Miwok population, the subsequent settlement for ranching by Mexican nationals, and land clearance for ranching and dairying by the Americans at the height of California’s gold rush. The Glenmire record documents these changes with the (1) decreasing charcoal

input following the 1793 fire suppression proclamation, (2) decline of native forest species, (3) arrival of non-native weedy invasive species associated with livestock grazing and land disturbance, (4) introduction of non-native trees for firewood and windbreaks, (5) an increase in coprophilous fungi associated with the presence of large numbers of sheep and cattle, and (6) changes in sedimentation type and rates associated with increased erosion around the site (Anderson et al. 2013).

Johnson and Cushman (2007) found that elk had an overall positive effect on native species composition. Elk also appeared to play a critical role in maintaining open grasslands and preventing certain areas from being converted to less diverse, shrubdominated systems.

Thus the complex ecological history of native grazers such as elk, and significant cultural fire management over millenia by tribes, should always be factored in to assessments of current vegetation management and conservion of species. Tribal cultural management should be consulted.

Fire-adapted cactus

Southern California grasslands also had a common admixture of various cacti, giving a southwestern character to them. Large low-growing thickets of prickly pear (Opuntia X occidentalis, O. littoralis, and O. X vaseyi, forming various hybrids) grew on grassy plains and foothills. Chollas (O. parryi and O. prolifera) also mixed into the grass and shrubland communities, increasing in abundance around San Diego and southward. Only small relict stands of the prickly pears remain in southern California today; in the past they formed a larger part of the landscape scene. J. M. Bigelow, botanist on the Whipple expedition in 1852-54 described “immense patches as large as half an acre” east of Los Angeles in the valley of what is now Rancho Cucamonga in western San Bernardino County (quoted in Benson 1969). The cactus “swarms” often grew on canyon mouths and washes where they received floodwater outwash, as well as on drier shallow soils of plains and south-facing slopes. Grassfires apparently played a large role in the adaptations of these cacti: small cactus were killed outright by fires sweeping through the grass. Larger cacti were able to resprout from underground roots or lone pads that survived the burn. But the high frequency of grassfires caused the prickly pears to form large, sprawling, clonal plants, dense patches which eventually excluded grasses within their thickets, aided by seed-eating rodents living among the spiny pads. The cacti formed vast fire-resistant “nopaleras,” their piles of joints protecting the interior of the patch, and their edges vigorously growing outwards after each grassfire (Benson 1969, 1982).

Fire-Adapted Oaks

Sharing in many grassfires were the varied oak landscapes across the state. Although wildfire no longer rages through most of California’s oak parks, there are still remote valleys and ridges in the Coast Ranges where the action of fire can be seen. Near San Luis Obispo I walked through a once closed oak woodland that had recently been charred and was now an open forest of black trunks. But despite the look of destruction, the Coast live oak is very resistent to fire, and small new green leaves were appearing on some twigs. Blackberry leaves began to unfurl along vines remaining on the barren ground. It was a scene of green specks in a charcoal world -- new life was evident.

In northern Baja California, which still retains a frequent fire regime, Coast live oaks often escape burning or are only scorched. Denuded oaks resprouted vigorously after burns and developed a new canopy. Aerial photographs from 1938 compared to similar photos from recent years showed no widespread changes in their distribution (Minnich and Vizcaino 1997). Native walnut trees and buckeyes also resprouted after a fire. In another study, 50 % of crown-scorched Interior live oaks (Q. wislizenii) basally resprouted within the first year in one fire study, and 79% were still alive two years after a fire (Haggerty 1990).

On the Tehachapi Mountains I came upon a recent accidental burn next to a highway in early October 2003. Hills of Blue oak and Gray pine had been swept by a fire which singed the lower branches of the mature oaks; two or three saplings were already resprouting on their small torched crowns. The Gray pines, with their low long-needled twigs, were completely scorched brown, apparently topkilling them. The open ground between oaks was blackened with burnt grass, and I found many Purple needlegrass bunches, some unburned, others burned down to their roots, and already growing new green blades. Most acorns were unharmed lying on the ground. Some tall silvery milkvetches (Astragalus) and low little mats of Rattlesnakeweed (Euphorbia albomarginata) had greened up on the burn and were flowering -- plants responded quickly to the nutrients released and competition reduced on these fired grounds, amazingly even at the end of the long dry season.

Four years later I came back to examine the scene, and could not distinguish the old burn from the surrounding areas. Renewal was quick.

Valley oaks were similarly well-adapted to fire and if topkilled by flames, soon crown-sprouted (Sugihara and Reed 1987b). After a control burn in Briones Regional Park, Contra Cosat County in October 1999, I hiked around the blackened ground examining the scattered oaks. Most Valley oaks, being well-sheathed in thick rugged bark, were unharmed and the low grass fire had not come close to the waving bunches of leaves well up on high spreading branches. A 2.5 foot-tall oak sapling had its crown killed, but a 5-inch long green leafy resprout popped up at its base.

What kept the oak savannas open? Botanist Willis Jepson in 1910 said savannas were frequently burned by Indians, and without fire to clear the underbrush, oak savannas might eventually close in. More recent studies of fires across the Santa Monica Mountains found that groves of dense Coast live oak woodland in protected canyons were only singed by passing fires; however, in some places the fire was hot enough to kill oaks outright and open up the woodland into “park savannas” (Sauer 1977).

The Indians regularly burned the oak groves to make acorn-gathering easier, increase green fodder for game animals such as deer, and encourage the growth of new straight shoots on hazelnut, Sourberry, Redbud, and Buckbrush shrubs to provide basketry materials. Individual plants were set on fire, or at other times patches or large areas were burned.

Keeping the oak groves “clean” could mean the use of regular burning to remove pest insects, such as the California acorn weevel (Curculio unifornis) that infested acorns. Larvae fed on the acorns, then dropped out of the trees to the leaf litter below to pupate in the soil. Burning woody debris, fallen bark, and layers of leaf litter may have kept their populations down (McCarthy 1993, Powell and

Hogue 1979).

Interviewing Nisenan people in the Sierra foothills east of Sacramento, ethnologist R. L. Beals noted:

“The land was apparently burned over with considerable regularity, primarily for the purpose of driving game. As a result there were few young trees and all informants were agreed that in the area of permanent settlement even so far up in the mountains as Placerville, the timber stand was much lighter than at present.... The Indians insist that before the practice of burning was stopped by the whites, it was often a mile or more between trees on the ridges, although the canyons and damp spots held thickets of timber” (Beals 1933).

The age of fire suppression came with the growth of Euro-American society, an age we are dealing with in. Tribal fire edperts are making headway, fortunately, in educating and collaborating with the state about the need to reverse this trend, and learn to live with prescribed fire.

Decades without fire can have effects not immediately visible. Fire both renews the oak landscapes, but also maintains a certain balance between old and young trees. In many cases the lack of regular burning may have shifted the live oak community structure from open to closed groves. At Hastings Reservation east of Monterey, where a large amount of classic natural history studies have been accomplished, Coast live oaks and evergreen shrubs have gradually increased since 1937 with no grazing or fires. As fire-intolerant shrubs invaded open grasslands, good microhabitats were created for acorns -- Coast live oak requires shade for seedling germination (Griffin 1988, Muick 1990).

UC Berkeley forest ecologist Scott Stephens is studying whether Sudden Oak Death, a recent outbreak of disease spread from introduced garden shrubs in northern California, might be the result of decreased burning and lowered immunity to stress in the oak groves (Ochert 2003).

Varved sediments from the ocean basin off Santa Barbara show silts from the Santa Clara and Santa Inez Rivers containing pollen that can give a generalized picture of vegetation. From A. D. 1450 to about 1875, oak pollen was constant at 20%; after 1875 an increase occurred to 27%, reaching a maximum of 45% by the late 20th century -- perhaps indicating an increased density of oak woodland due to fire suppression (Byrne, Edlund and Mensing 1990, Mensing 1998). Coast live oaks less than four inches in diameter were often killed by low intensity fire in one study, and density and canopy cover of live oak woodlands was highest on sites where no recent fires had burned (Mensing 1998).

But a reverse trend was seen in the Garin Hills of Alameda County, an area that has had fire suppression for 50 to 100 years but was grazed by cattle at times. Researchers found Bay trees commonly reproducing from seed sprouts and vegetative propagation, but a decline in Coast live oaks. Deer were heavily pruning the oak shoots and cattle had an “enormous appetite” for acorns. Other major consumers of acorns included Pocket gophers (Thomomys bottae), Gray squirrels (Sciurus griseus), Dusky-footed woodrats (Neotoma fuscipes), Scrub jays (Aphelocoma californica), and California quail (Callipepla californica). As noted, Coast live oak is more fire resistant and Bay is not, so in the current regime with fire suppression and active acorn predation, Bay is favored. The oak woodlands in this area are apparently remnants of a more extensive type, adapted to frequent ground fires (Safford 1995).

George Gruell’s enlightening repeat photography study of the Sierra Nevada showed an increase in density of Interior live oaks in woodlands of California black oak and Jeffrey pine along the Kern River in Tulare County. Younger trees filled many of the openings between old trees. In several places in the foothills of Mariposa County, old photos from 1860 show a savanna of scattered Interior live oaks, Ponderosas (Pinus ponderosa), Gray pines (P. sabiniana), and chaparral shrubs growing as scattered individuals or in patches. By 1993 the same spots had filled in with chaparral shrubs of manzanita (Acrtostaphylos spp.), Chamise (Adenostemma fasciculatum), ceanothus (Ceanothus spp.), and Toyon, and Interior live oaks and Gray pines increased to form a denser woodland on the hills. Similar photos from as early as 1858 in Placerville, and photos from Amador, Placer, and Tuolumne counties showed similar changes in the 600 to 3,000 foot altitudinal zone. The early scenes showed Ponderosa pines relatively sparse and scattered in open live oak-Gray pine savannas, and chaparral shrubs patchy and of small size. After a century or more a remarkable increase in vegetation occurred, with young Ponderosas more common and oak woodlands closing in (Gruell 2001).

Fire suppression may be the single most important cause of this change, Gruell said. John Muir noted the long-term influence of Indian burning and lightning fires between the Tuolumne and Merced Rivers, large swaths of grasslands burning until extinguished by rains or by shallow rocky soils acting as firebreaks.

As I hiked among the oak hills and valleys I sometimes had difficulty imagining the scene 300 years ago. What we can know is that fire played a much more important role than it is allowed to do now. I carefully examined the burns that I encountered, and recorded my observations. These charred landscapes were common in California for undoubtedly millions of years.

Spice Hills

Green-gray, silvery and russet masses of low aromatic shrubs intermixed with grasslands south of Santa Barbara County. Shrubs such as California sagebrush (Artemisia californica), sages (Salvia apiana, S. mellifera, and S. leucophylla), Silver bush lupine (Lupinus albifrons), brittlebushes (Encelia farinosa and E. californica), Bush buckwheat (Eriogonum fasciculatum), and Deerweed (Lotus scoparius) were low in stature and often went dormant and lost their leaves during the summer dry season. Evergreens such as Lemonadeberry (Rhus integrifolia), Sugar bush (R. ovata), Chaparral yucca (Yucca whipllei), and various cacti mixed in the pungent community.

In some years torrential rains in the mountains let loose massive flash floods down the dry washes and river channels, often overtopping their banks to spread out onto alluvial fans and lowlands. New habitats were created by these disturbances, such as open Matchweed (Gutierrezia californica) communities on arid stony fans. Dry washes often had alluvial scrub types such as Blue elderbarry (Sambucus mexicana), Scale-broom (Lepidospartum squamatum) and Cottonthorn (Tetradymia comosa). Early botanists described White sage (Salvia apiana) as very common on dry plains towards the foothills (McCawley 1996), again probably in a fluctuating grass-shrub boundary depending on disturbance from fire and floods. Elsewhere the San Gabriel Valley was said to be “pure llano” -- prairie, with groves of oaks (ibid.).

Patches of this low scrubland mixed with grassland on dry, rocky, steep sites, river bluffs, beach dunes, and on hillslopes mixing with chaparral. Whether it was more widespread on hills and plains in the past has been a topic of debate, as the complex interactions of wildfire, climate, grazing, and competition with introduced annual grasses are difficult to tease apart, as is usual for California’s changing habitats. The boundary between grassland and south coastal scrub has moved according to various lines of evidence. Repeat aerial photos starting in the 1940s and survey plots measured on the ground by the old California Forest and Range Experiment Station from 1929 to 1934 around Riverside were re-measured in the 1990s by Richard Minnich of U. C. Riverside and Raymond Dezzani of Boston University. They found great changes in 60 years in the coastal sage scrub, with a decline in shrub species diversity, a drop in shub cover on most plots, and a decrease of California sagebrush and Bush buckwheat on alluvial fans and volcanic substrates. Only Brittlebush (Encelia farinosa) increased.

But other researchers found the opposite in their study areas: grassland shrinking and sage scrub taking over. Using the early accounts by Pedro Fages, Fray Garces, the 1849 Mexican boundary survey parties and others, evidence indicated that grassland was widespread in San Diego County during early Euro-American occupation where brushlands are now prevalent (Dodge 1975). George Cox, then of San Diego State University, discovered a pattern of mima mounds, a grassland indicator we have looked at, that matched the supposed historic distribution of grasslands in San Diego, further supporting the hypothesis of formerly widespread grasslands (Cox 1986).

What is going on here? The fast-changing coastal sage scrub illuminates the varied processes that ecologists must address in trying to reconstruct pre-contact communities. Let us look at some of the possible causes and affects.

Has long-term climate change been a factor? Probably -- paleobotanist Daniel Axelrod theorized that south coastal scrub species expanded in their distribution during the arid Xerothermic interval 8,000 to 4,000 years ago, as semidesert and desert plants moved coastward (Axelrod 1978). In our new Global Warming era we may see a return to dominance of the most arid-adapted shrubs such as Brittlebushes.

Smaller climatic cycles also affected the sage scrub. Whole stands of California sagebrush and sage died off during multi-year droughts, although they recovered again during subsequent wet El Ninos. Bush buckwheat and Chaparral yucca, however, sometimes expanded during droughts according to the longterm observations of Don Mullally in the Los Angeles area (Mullally 1994).

Did fire play a role? Most definitely. But how great a role is still debated. South coastal scrub species were mostly killed by fire, but quick to recolonize newly burned areas from seed banks in the soil (sages) or from small wind-transported seeds drifting in (California sagebrush, bush buckwheat, brittlebushes). Some shrubs resprouted, especially closer to the coast where moisture levels were higher. Fire intervals of less than 3-10 years have apparently led to grass domination because the shrubs were not able to get established before the next burn (Minnich 1998, Keeley 2002). Periods of 10 to 25 years fire-free were needed for coastal sage scrub to mature (U. S. Forest Service 2000, Minnich and Dezzani 1998). So fire suppression may have led to shrub increase over grassland in places.

Many ecologists (Zedler 1977, Zedler et al. 1983, Oberbauer 1978, and Keeley 1977) were of the opinion that fire frequencies were too low in the coastal terraces, hills, and valleys to prevent the dominance of shrub communities, and that very little grassland existed except in the Valle de San Jose near Warner Springs, San Diego County. Others, however, believed that fire frequencies from Indian burning were high in coastal and valley areas, and that even infrequent lightning strikes in lowland grassland would spread far and wide, all acting to exclude shrubland from these places (Axelrod 1975, Biswell 1956, Aschmann 1977, and Dodge 1975).

On the mesas and foothills many coastal sage shrub and chaparral species apparently invaded grasslands due to the decreased fire frequency within historic times, especially on the deeper soils of mima mounds. “The original grassland boundary was probably set by thin rocky soils that are incapable of supporting a grassland community dense enough to spread fire except at long intervals,” Cox observed (Cox 1986). Sage scrub may have kept more to hilly areas and rocky sites, while grasslands dominated the valleys and basins, regardless of fire regime (Minnich and Dezzani 1998).

But to complicate the picture, coastal scrub has lost acreage to introduced annual grassland which apparently differs from the open bunchgrassland and native wildflower fields once typical of prehistoric southern California. The dense crowded growth of European annual grasses outcompeted many coastal sage shrub seedlings for soil moisture, light, and nutrients (Eliason and Allen 1997). The native forbs that in the past pioneered burns contained less hard silica in their leaves than do annual grasses such as Red brome, and so may have decomposed more quickly, posing less of a repeat-fire danger than today’s weedpatches (Minnich and Dezzani 1998).

More recently, urban and agricultural development, mining, and flood-control projects have seriously fragmented the south coastal scrub community and removed huge chunks of this habitat, endangering such endemics as the Orange-throated whiptail (Cnemidophorus hyperythrus beldingi) that lived on floodplains and streamside terraces in coastal sage scrub and chaparral habitats from Baja to Orange County. Another cosatal sage scrub specialist, the California gnatcatcher (Polioptila californica), requires dense stands of a diverse mix of shrubs, and its habitat may be converting to more open introduced weed patches (Minnich and Dezzani 1998). The infilling of the little open patches between bunches that once characterized native grassland and sage scrub with dense exotic annuals perhaps reduced the ability of many small animals to move around and forage: the Stephen’s kangaroo rat (Dipodomys stephensi) and San Diego coast horned lizard (Phrynosoma coronatum blainvillii). The Quino checkerspot butterfly (Euphydryas editha quino) preferred to find native plantain (Plantago erecta) plants for its larvae to dine on within coastal sage srcub, and habitat loss and serious fragmentation have led to the listing of this species as federally endangered.

In Old Southern California the wildflower bunchgrassland and aromatic coastal sage scrub boundaries probably fluctuated continually depending on drought and rainfall, fire frequency from Indian burning and rare summer thunderstorms, soil type and slope aspect. Minnich and Dezanni (1998) pointed out, however, the importance of knowing the history of the landscape in an area, and how waves of invading plants will tip the balance away from natives. Simply protecting a piece of surviving coastal sage scrub from development will not necessarily save it, as it is not a stable community. A “dynamic one-way process of conversion” is taking place, they pointed out, due to the newly-arrived waves of introduced bromes and mustards that have changed the face of the state. New management options will have to be tried.

Fire on the Mountain

Quick yellow tongues crackled upward, dry branches exploded with sizzling sap, flames roared, gray ashy smoke bellowed skyward building cloud columns and carrying the smell of charcoal on hot winds. Swirls of sparks jumped the fire ahead on chaparral slopes. Scenes of wildfire have been regular in California’s brushlands for hundreds of thousands of years, as charcoal deposits in ancient sediments show, yet investigations into the secrets of fire ecology have proceeded slowly and have led sometimes to great controversy.

Pedro Fages during his 1769-70 journey up California noticed the Indians making a sort of native sugar from the fruit of “a very leafy, tufted shrub six feet high with a stem of reddish color and leaves like those of the mangrove.” The ripe fruit was gathered, its pulp separated, pressed into cakes and dried (Priestley 1937). This was the manzanita (Arctostaphylos), a shapely, colorful, and varied shrub typical of the chaparral. Spanish explorers often described “rosemary” on the hills, what may be Chamise (Adenostoma fasciculatum) with its diminutive linear leaves. Their word chamiso meant “thicket.”

William Brewer complained of traveling through chaparral more than any other vegetation type on his explorations to see the different parts of California in 1860-61. Ascending the Santa Ana Mountains he encountered very dense chaparral which tore his pants into ribbons. He followed deer trails to avoid the hard shrubs. Tracks of deer, wildcats, and coyotes were numerous. The chaparral was so dense on the higher granitic crags that it had to be chopped with a hatchet. Near San Luis Obispo the chaparral was worse, and he finally bought buckskins and smoked them to harden them against the chaparral and rattlesnakes -- the “chaps” that the cowboy used in riding through dense brushy country.

Fire-adapted shrubs such as Chamise seem to invite burning on a regular basis by developing dead branches (called “dead aerial fuels”), resinous leaves, and highly flammable growth of an overlapping canopy of dense twigs -- “It burns like a torch of fat, “said observer Francis Furtz (1923).

Early botanists began to find hints that brush burns were not simply destructive but regenerative: T. S. Brandegee commented in 1891: “The Coast Range of California...is not distinctively a forest region, but its hills and mountains are covered by a thick, almost impenetrable growth of Adenostoma fasciculata [sic], commonly called ‘greasewood’ or ‘chamise,’ manzanita, Garrya, oak, Ceanothus (California lilac), etc., which seems to be periodically destroyed by fire. Indeed it is almost impossible to find a hill or mountain covered with bushes even of great age, where some old charred ends of roots may not be seen showing that at some previous time they had suffered from fire” (Brandegee 1891).

The numerous chaparral shrubs, herbs, and grasses often have special adaptations to the process of wildfire, and indeed, the whole community shows a regular succession over time as different groups of plants fill changing niches of soil quality and shade. The spectacular herbland phase often dominates the first wet season after a fire, springing from seeds waiting in the duff layer. So-called “pyrophyte endemics,” fire followers, may be found only on burn sites. The Fire poppy (Papaver californicum) is abundant in the first year after a fire and usually diappears in the second year. Wild cucumbers (Marah spp.) are sometimes the first to sprout on the blackened ground, and can grow during the winter to drape across the skeletons of shrubs. Dazzling wildflower displays start often in March and different species appear in waves, one group growing and peaking, then another late into summer. The wildflowers peak in one to five years then decline as the shrubs take over, remaining in small numbers in openings in mature chaparral or in the seed bank of the soil (Biswell 1974).

Next come short-lived pioneer shrubs, often with nitrogen-fixing roots that help restore lost nutrients to the soil. They can be abundant on burns: Deerweed (Lotus scoparius), Chaparral whitethorn (Ceanothus leucodermis), Wavyleaf ceanothus (C. foliosus), Yerba santa (Eriodictyon spp.), Golden yarrow (Eriophyllum confertiflorum), and Rush-rose (Helianthemum scoparium). Deerweed only lives about 3 - 5 years, and Wavyleaf ceanothus to about 25 years. In other areas, Bush poppy (Dendromecon rigida) pioneers new burns, its seedlings becoming dominant, then declining as other shrubs such as Bigpod ceanothus (Ceanothus megacarpus) grow large (Sauer 1977). After seven to nine years the original chaparral overstory begins to develop.

Gradually the pre-burn shrubs regain their dominance, about half of the species from seed. Plants that are obligate seeders include many ceanothus (Ceanothus crassifolius, C. cuneatus, C. greggii, C. megacarpus, C. oliganthus, C. papillosus, C. parryi, C. thyrsiflorus), several manzanitas (Arctostaphylos canescens, A. glauca, A. manzanita, A. mariposa, A. nevadensis, A. obispoensis, A. parryana, A. pilosula, A. pungens, A. stanfordiana, and A. viscida), California juniper (Juniperus californica), and Saw-toothed goldenbush (Hazardia squarrosa). Chaparral seeders often have special requirements for germination, and will lie dormant for long periods. Chamise will also reseed itself after a fire: it produces two kinds of seeds, those that germinate readily, and those that germinate after exposure to high temperature -- only a five-minute heat of 280 to 300 degrees Fahrenheit (F) will induce germination (Hale 1977, Biswell 1974, Sampson 1944). The temperatures in burning leaf litter during a fire can reach 1,000 degrees F, but soils one and a half inches deep will reach only 300 degrees F. A passing wildfire can open up the closed canopy and produce a flush of Chamise seedlings as dense as 3,000 in one square meter.

Deerbrush (Ceanothus integerrimus) seeds can be harshly treated: they can be boiled for 20 minutes and some will still germinate.

Other chaparral species resprout from roots. Scrub oaks (such as Quercus berberidifolia), manzanitas, and Ceanothus species such as Chaparral whitethorn (C. leucodermis) and Greenbark ceanothus (C. spinosus ) develop a woody platform at ground level that can be as wide as 12 feet. After each fire, the platform sends out sprouts at its periphery, putting on further growth -- these root burls can be 250 years old or more and apparently store reserves of water and nutrients for fast growth during the summer. Chamise will also sprout from such a lignotuber. Other shrubs have deep tap roots.

Ecologists also found that certain seeds need more than heat to germinate, but depend on chemicals leached from charred wood to penetrate through their seed coats with the winter rians, stimulating sprouting. Many fire annuals have this response, such as Whispering Bells (Emmenanthe penduliflora), many Phacelia species, Yerba santa (Eriodyctyon crassifolium), and Penstemon (Penstemon spectabilis) (Sweeney 1985, Keeley and Keeley 1987).

In addition, smoke stimulates many wildflowers to grow after a burn: the native wild tobaccos (Nicotiana attenuata, N. clevelandii, N. quadrivalvis, and N. obtusifolia), used by a great many tribes throughout the state. Many groups sowed tobacco seeds directly onto recently burned areas (Lewis 1993). Pedro Fages in 1769 mentioned Indians in the Central Valley gathering ”great harvests” of wild tobacco, then wrapping the leaves in tules to cure; sometimes a hollow antelope horn was used to carry the dried tobacco (Priestley 1937). I have seen hundreds of acres of wild tobacco spring up after burns in the Great Basin. The seeds may have a permeable membrane that allows smoke to penetrate directly. Thus, slow, smoldering fires under moister conditions may bring out a whole set of fire annuals that a hot fast blaze would not (Keeley 2002, Keeley and Fotherigham 1998).

Would chaparral management practices that use mechanical thinning -- cutting and chaining -- eliminate this group of wildflowers from the dynamic community? Whole communities of dormant seeds lie in the soil awaiting a fire to wake them, and I can see that looking at a chaparral hillside at any one year may hide several layers of time-released plant communities usually not visible.

The Great Chaparral Debate

Some of the earliest "megafires" in California began in the chaparral belt east of San Diego in 2003. I know, as a biologist friend who bought one of my paintings lost it in the fire that took his house which I had visited, during a wildfire that burned across dispreded homes in the mountains. Santa Ana winds of 50 miles per hour whipped up a wildfire to flames raging 300 feet high, spreading 400 feet per minute. This was devastating.

Another friend of mine who worked for Clevelenad National Forest told me that her home in Julian escaped the rush of wind and flame hy a mere two miles. But she told me that the town was largely reduced to a pre-modern state: no electicity, meaning no well pumps worked to supply water for livestock or off-city homes; no gas stations operaring, and only candlelight at night for days.

Unfortunately this would prove to be just the beginning of the Megafire devastation to wreak California in the following decades, and an unnecessary but all-to-common-pattern as we shall see below.

Heated by these mostly chaparral conflagrations in 2003, disasters fire ecologists rushed to understand how to bet protect life and property along the wildland-residential interface.

Fire ecologists at this point took two opposing sides with differing views on what kind of historic fire regimes California chaparral had and how to manage fire country based on those regimes. Richard Minnich, of the University of California at Riverside, formulated a model that says the fire regime under current management is different than that in precontact times, and that wildfires are basically fuel-driven, while Jon Keeley of the United States Geological Survey at the Western Ecological Research Center in Three Rivers developed another model that states today’s wildfire patterns are similar to the past, and that wildfires are dominantly wind-driven. Each model has thorough scientific evidence to support it, contained in piles of research papers full of charts, Geographical Information System maps, and eloquent discussions, but the evidence used is sometimes completely contradictory. “Who is right?” I thought to myself as I sat back and read the papers, trying to decide how to paint a past landscape. Which model is closer to the truth? This is not known, but the debate illuminates the problems associated with the study of historical ecology and the difficulties of reconstructing the past even with mountains of evidence.

The Minnich Model says that “protectionist management” has led to coarse-scale patchiness of chaparral stands and unusually large intense fires in southern California. Before suppression the pattern of stand ages following burns was probably a complex fragmentation and fine-scale patchiness (Minnich and Chou 1997). Minnich (1989) found when looking at vegetation maps and aerial photographs of the U.S.-Mexican borderlands that Baja had numerous small fires, while San Diego County had large wildfires. Most fires in Baja are not over 5,000 acres and rarely over 10,000 acres; compare that with the 280,000-acre Cedar Fire in San Diego County.

Key to Minnich’s argument is what he observed to be the behavior of fires when they meet previous burns -- they “lay down” at the borders, especially if these previous burns are less than 20 years old. So fires may fit into one another neatly over time like a puzzle. What is produced is an ever-changing mosaic of burns, age-classes, and stand structures. This is the fire behavior that Minnich believes occurred north of the border before fire suppression in the 20th century. W. V. Mendenhall, the first supervisor of the Angeles National Forest said many settlers found that “fires were not extensive due to the fact they ran into older burns and checked themselves” (Minnich 1983, 1989). In the 1890s observers on Mt. Wilson noted fires slowly burning for two to three months. During the same time fires in the San Jacinto Mountains were described as “scattered throughout the reserve in small tracts” (Minnich 1983).

This is the pattern found in northern Baja today, as if the Mexican state is a mirror for California’s past. The Mexican government does not attempt quick suppression of wildfires, as they believe that having several medium-sized fires in one season will burn the fuel and prevent one great fire. The fires burn for months in remote areas and eventually burn themselves out. Less development in rural areas also means less property loss: in the October 2003 fires, medium-sized wildfires destroyed only 10 houses in Baja (compared to nearly 3,000 in southern California). A great patchwork of recent burns prevents the build-up of solid tall brush that can carry blazes the size of the Cedar Fire. Jose Luis Rosas, Executive Coordinator for the Baja California Civil Protection Agency, told reporters, “Well, let it burn. What’s the problem?” (Hernandez and Perry 2003).

Minnich agreed that numerous small burns preclude larger conflagrations owing to the fragmentation of fuels; as stands age their flammability increases due to accumulating biomass and dead plant matter. He said recently burned chaparral will not carry fire for about 5 years; from 6 to 20 years it may burn during extreme fire weather, and then to age 50 years will burn well under normal conditions. After 50 years dead branches accumulate and can burn very hot, creating intense fires that can easily get out of control (San Diego County Wildland Fire Task Force 2003). Thus fires under this regime, he said, are self-limiting and self-organizing. Modelling predicted that large fires in Chamise chaparral occurred in older fuels -- that fires spread faster and burned more intensely with increasing age of the vegetation, and almost no large fires occurred in stands younger than 15 to 17 years plotted on fire maps on the Angeles and San Bernardino National Forests (Philpot 1977).

The national forest managers of California inherited a mosaic landscape after 1900 that became homogenized with suppression. Then large catastrophic fires broke out in the 1920s, more in the 1950s, and then the Laguna/Boulder Fire of 1970, which burned 190,000 acres. A mere few months before the 2003 firestorms, San Diego County was estimated to have over 50% of its scrubland and forest over 50 years old -- a ticking bomb (Minnich 2003, San Diego County Wildland Fire Task Force 2003). Fires in San Diego County burned 284,000 acres in a only three huge wildfires.

Suppression led to disastrous consequnces in timing. In Baja Minnich noticed that most fires occurred in summer during days of weak winds around 10 miles per hour and 20 to 40% humidity, and were unusual in fall. But in San Diego County, fire-quenching has resulted in the build-up of fuels that created anomalous delayed fires set by people during dangerous weather times such as the Santa Ana wind season, which starts after mid September. Humidity then can be as low as four percent and winds over 30 miles per hour -- wildfires get out of control and spread hotly over huge areas. Most Los Angeles news reports of fires in the late 19th century were in July and August, not during severe fall weather (Minnich and Hong Chou 1997).

Not all chaparral researchers agree with this model, however. Jon Keeley and C. J. Fotheringham (2001b) call this the “fuel-age/mosiac model” and argue against it. They believe that large wildfires in the chaparral today are not a remnant of fire suppression, as many fires will readily burn through young stands when driven by strong-enough winds, and not stop at their boundaries for lack of fuel. In their model fire is weather-driven. They claimed that previous fire history is unimportant, although they admitted that previous burns may act as fuelbreaks during moderate weather. These are the most contentious points between the models: whether Santa Ana-driven fires will lie down at young fuel patches, and whether fuel accumulation is important at all with respect to wildfire.

In San Diego County most fires are set by people, and a few by lightning from July and August thunderstorms. In Baja fires are set by cattlemen and by lightning fires. Both the Minnich and Keeley Models agree that fire suppression has not been successful in places like San Diego County where sprawling urban growth has fingered into the wildlands and allowed human ignitions to continue. Minnich argued that the huge firefighting effort has shifted larger fires to fall when Santa Ana winds whip up unstoppable firestorms. Keeley countered that Santa Ana-driven fires have always been part of southern California, and are not an artifact of firefighting. But he alloweed that fire suppression has become increasingly effective if measured as area burned per number of fire starts (Keeley and Fotheringham 2001a) -- people are setting more fires along roads and around urban edges, for example, but they are being extinguished quickly before they can effectively reduce fuel over wider areas.

Presuppression brush fire intervals for California are speculative, and range from every 40 to 100 years based on stand age structure (Minnich 1989, Keeley et al. 1989). In Baja they occur about every 70 years, longer toward the arid interior (Minnich 1989). As to particular type of chaparral, Minnich and Hong Chou (1997) further specify Chamise chaparral to have an estimated fire return interval of 72 to 77 years, and mixed chaparral of Ceanothus, scrub oak, and Chamise every 58 years. In the dry interior desert-edge chaparral and pinyon associations, fires may have been rarer -- every 291 years.

Some of the only physical evidence for pre-European fire frequencies comes from data on charcoal deposited in the Santa Barbara Channel, widely cited as evidence for Santa Ana winds creating huge fires in the watersheds of the rivers feeding these sea-bottom sediments (Mensing et al. 1999). Cores from these ocean deposits contained charcoal in layered sediments, reflecting yearly winter influx. The charcoal appeared at a continuous “background” level, punctuated by small increases and occasional large peaks of charcoal, some of which correlate with large fires in the Los Padres National Forest mountains just east of Santa Barbara. Large peaks in charcoal occurring several times back to the 16th century were interpreted as indicating huge smoke plumes from conflagrations in these mountains being driven offshore by fire-winds in fall. But looking at the charcoal peaks against known large Santa Ana fires in recent times, some peaks occur in a year when no known fire occurred. What caused these peaks? I do not see how these cores can give such well-defined results, as many unknown factors may influence how these charcoal layers formed. Could numerous small fires give the same peaks? How can wind-deposited versus river-deposited charcoal be separated?

Pollen researcher Roger Byrne and his colleagues analyzing the same cores, admitted that wind versus water transport cannot yet be separated, and the distance that charcoal can be moved by ocean currents is poorly understood. The area which could collect charcoal is huge, and includes other areas than simply the Los Padres ranges; the Channel Islands for instance. In addition, Byrne pointed out that the peaks of charcoal recorded in pre-European time were in grey layers of the core, indicating “major fire-flood events,” apparently caused by high winter flood deposits and “turbidites.” I believe this could represent El Nino rain events during Little Ice Age extremes sweeping large amounts of terrestrial charcoal from numerous small summer fires into the ocean by extensive flooding, instead of Santa Ana wind-borne charcoal from a few large fall fires -- how can we be sure of the origin of such charcoal? Interestingly, the Byrne study reported: “Apart from the grey layers, the average influx value for the prehistoric period is somewhat lower than for the modern period....” They continued: “The implication is that total burning was greater during the period 1931 - 1970 than in the prehistoric period sampled” (Byrne et al. 1977). In other words, there may have been more fuel in the scrublands during the 20th century fire-suppression era, causing larger fires, than during the earlier phase of Indian management.

Using such evidence as charcoal deposited in ocean sediments can open up some intriguing clues to the past, but over-interpretation is a danger. I would like to see similar cores analyzed from the sea bottom off northern Baja California to find out what comparisons can be made before more conclusions are drawn from the Santa Barbara Channel studies.

The low incidence of lightning fires has been cited as evidence of low pre-cultural fire frequencies. Keeley (2002) argued that before Indian burning, California scrublands must have had a fire regime similar to today’s -- few and large fires. Minnich (1989) held to the oppsite claim, that even low lightning-ignited fire frequencies can be significant in chaparral ecology. San Diego County has more lightning strikes inland along the mountain crests than coastally, and summer thunderstorms surging up from the tropical Gulf of California account for 5 to 30 lightning fires per year. In the San Jacinto and San Bernardino Mountains 2 to 11% of lightning strikes start fires requiring suppression annually (Minnich and Hong Chou 1997). Minnich pointed out that the Sierra Juarez in Baja has twice the summer rain that San Diego County has, and so lightning fires may be more common there. Keeley et al. (1989) estimated the fire interval for lightning strikes could have been as high as every 100 years in Santa Cruz County. But perhaps this was enough lightning-caused fire to matter in landscape-scale chaparral organization. If the estimated fire return interval for chaparral is somewhere between 40 and 100 years, then even rare summer lightning strikes would be enough to burn chaparral regularly. Pinnacles National Monument in the Coast Range east of the Salinas Valley had one lightining-ignited fire in its 50-year record (Keeley and Fotheringham 2001a) -- again, it would only take one fire at this rate to burn these remote hills regularly over the span of millenia.

Other records indicate lightning fires did occur with some regularity in parts of California not subject to a summer monsoon: Santa Cruz County had 34 lightning fires from 1893 to 1979; the Gabilan Range, part of the South Coast Ranges, had 142 lightning fires from 1930 to 1979, three storms ingniting 35% of these fires (Greenlee and Langenheim 1990). Fire scars on trees in chaparral indicate a fire every 50 years or less in Pinnacles National Monument and on Junipero Serra Peak near King City, Monterey County (ibid.). Interior northern California national forests, which have extensive chaparral, have some of the highest lightning fire numbers in the state (Keeley 1977). The foothills of the Kaweah River drainage in the southern Sierra Nevada had 105 lightning fires from 1930 to 1978, averaging about two a year (and ranging from zero to 10 a year), most in July and August. Ecologist David Parsons, studying the Kaweah area foothills said, “When ignited under the proper conditions, few ignitions are needed to burn large areas of highly flammable chaparral and oak woodland” (Parsons 1981). Most lightning fires started in the middle elevation foothill zone where chaparral is common, but some do occur below this at lower grassier elevations.

Minnich introduced good points about the importance of lightning fires: lone oaks sticking up amidst chaparral can act as lightning rods that start chaparral fires. Other people have noticed old oaks in southern California with broken branches, basal fire wounds, and heart rot from lightning strikes (Vogl 1977). Most lightning fires occur in humid weather in summer storm and fog periods, but they can “store” for weeks in logs and snags waiting for dry weather to kick them up. Even in a drenching thunderstorm a lightning strike could cause a fire inside a rotten oak trunk that within hours or days could ignite surrounding grass and chaparral.

Minnich reconstructed fires occurring on Mt. Wilson in the San Gabriel Mountains in the late 1800s, before suppression: one fire persisted for months through the long dry season of summer, erratically smoldering and storing for long periods in logs and snags in woodland edges, then “running” through dry shrublands. Embers flew long distances to start secondary blazes -- “spotting”. The fires often spread in a web of meandering strips, leaving numerous islands. None was apparently larger than about 17,000 acres (Minnich 1983, 1989). This is the same behavior that he found in Baja today -- where fires have been known to last two months, but this behavior no longer occurs in southern California due to immediate control measures.

In Mediterranean Europe, lightning is not negligible, accounting for 2 to 10% of fires in oak and scrubland (Grove and Racham 2001). Until recently, shepherds burned scrub, grass, and savanna areas every three to five years to clear underbrush and increase pasturage for their herds. Aerial photographs of Crete from World War II suggest big fires were less common than today when this “occupational burning” was more common.

The point may be made that although lightning fires are rare in much of lowland California, they may have occurred just enough to promote a regular fire rotation in remote, rugged mountain ranges. Indian prescribed fire would have accelerated this cycle in many areas.

Keeley (2002) came back to assess the Indian burning question, and summarized the two camps regarding the influence of this practice. One group of fire ecologists believed Indians had only a minor effect on pristine vegetation, and perhaps altered the landscape only in the vicinity of villages. The other group hypothesized that Indian burning was pervasive and widespread, enough to significantly change grassland and scrubland patterns from their pre-human condition. Of course this picture is further complicated by trying to determine what a pre-human conditon was for North America, as before and during arrival of people to the continent, large herds of extinct Pleistocene bison, horses, mammoths, camels, and other big game were having their own impacts on the vegetation, perhaps similar to the heavy grazing-browsing-trampling pressures seen in Africa. I wonder if these herds had the same affect on shrublands as did frequent fire: keeping them open and structured in a mosaic pattern? Even elk grazing during Holocene times cannot be ruled out as a possible cause of open vegetation.

Keeley (2002) believed that Indian burning greatly accelerated the fire frequency in lowlands, thus thinning out chaparral and coastal scrub over sizable portions of the landscape. Rugged, large interior ranges may have seen much less anthropogenic fire. But on most areas, Keeley said that if fires occurred at a frequency of more than one each decade, seed-reproducing chaparral shrubs would be eliminated and resprouters would be thinned out into islands amid a matrix of grassland.

Minnich (1983) similarly thought Indian burning had more affect on grasslands and coastal sage scrub, and was less important in chaparral. Increased burning frequencies by native people may have lead to “type conversion” of some shrublands -- a process where frequent disturbance changes the vegetation community from one trajectory onto another trajectory. On the lower edges of chaparral slopes, frequent fires may have converted chaparral into coastal sage scrub or native bunchgrassland. I have seen Small-flowered melic grass (Melica imperfecta) as a common chaparral edge species, its bunches filling in burned scrub temporarily if fires occurred longer than 10 years apart; it might grow into a more permament grassland if fires burned more frequently.

Indian burning may have been crucial in keeping valleys and lower slopes open and grassy, decreasing shrublands and woodlands, and interacting in complex ways with elk herds that grazed the same lowlands. The fire return interval in these sections was frequent, yearly or every several years. Lightning ignitions may have been more significant in starting fires in the upland hills and mountain ridges, where the fire return interval was less, perhaps ranging from every 50 to 100 years. Indians burned selected plant patches on the mountains and for game drives. Occasionally these mountain spot-burns got away and may have flamed over large areas. Ethnologist John Harrington collected notes about the Chumash around the present Los Padres National Forest: Maria Solares, a Yokuts-Chumash woman told him, “...some of the Indians were burning chamiso one day near Mt. Pinos to hunt cottontails and jackrabbits, and the fire got beyond their control” (Timbrook et al. 1993) .

Both Keeley and Minnich agreed that today weedy introduced annual grasses and mustards have added to massive fires and type conversion today in southern California. Weedy fast-growing and fast-withering brome grasses, wild oats, and mustards may take over after a burn, aided by breaks in the chaparral created by roads and fire-fighters, and carpet the newly opened ground. This type of “flush and fade” fuel promotes more frequent fire, hinders the regrowth and recolonization of native grasses, shrubs, wildflowers, and may negate the affect of young burn patches in stopping wildfires from spreading over large areas. The situation is not helped when land managers reseed burns with such exotics as Italian ryegrass (Lolium multiflorum) in an effort to stop erosion.

The two fire regime models have almost opposite recommendations for ending massive firestorms that destroy life and property. The Keeley Model states that since large fires driven by extreme fire weather are normal, and happened regularly in the last several hundered years, fire suppression should be continued with every effort to save property. Keeley urged building buffer zones around urban areas where they meet fire-prone habitats, and recommended placing a greenbelt of golf courses and lawn-parks on the edges of new developments to protect them from wildfires and ease access for fire-fighters.

The Minnich Model states that since we are apparently living in an unnatural fire regime caused by fire suppression managment, we should try to reverse this trend and restore the landscape to a pre-contact conditon by carrying out control burns all over the chaparral areas to create a patchwork of fuel ages that contain natural fuel breaks, limiting the spread of massive fires ignited during extreme fire weather. Northern Baja California could be used as a very good reference site, he said, because of the similarities -- the chaparral does not stop at the fenceline.

Fire Forest

Travelling up into the higher ranges of California, I was not surpised to see that the emerald evergreen forests of the Sierra Nevada were no different than the many other fire-shaped habitats we have looked at.

John Muir in 1894 described these Sierran forests:

“The giant pines, and firs, and Sequoias hold their arms open to the sunlight, rising above one another on the mountain benches....The inviting openness of the Sierra woods is one of their most distinguishing characteristics. The trees of all the species stand more or less apart in groves, or in small, irregular groups, enabling one to find a way nearly everywhere, along sunny colonnades and through opening that have a smooth, park-like surface, strewn with brown needles and burs” (Muir 1894).

Around 1900 Robert Milnor explored the headwaters of the south fork of the San Joaquin River and found an Indian camp, probably around 7,000 or 8,000 feet, in a small “park” dotted with groves of “aspen and monster yellow pines, and knee-deep in rich grass.” The yellow pines were probably Ponderosa pines (Pinus ponderosa). Bent-willow frame “tepees” and a cache of acorn-filled baskets stood by the rushing creek, which was “alive with trout” (Roberts 1902).

Forester George Sudworth in the same year described the Sierran forests he had examined as rarely dense, “the single big trees, or groups of three to six, stand far apart, forming a characteristically open forest” (Sudworth 1900).

Among the 100-foot tall giant Ponderosa and Sugar pines (Pinus lambertiana), spaced singly and in groups, were open areas of bare pine needle duff, montane bunchgrasses such as Western needlegrass (Achnatherum occidentalis), shrubs such as Snowbrush (Ceanothus cordulatus) and Kit-kit-dizze (Chamaebatia foliolosa), and on the east slope Bitterbrush (Purshia tridentata). Small moist meadows also interfingered with the forest groves in low spots. The forest canopy cover was not continuous, and probably shaded only 50% of the ground (Barbour et al. 1993). Among this mix some giant Ponderosas reached 260 feet tall with 9-foot base diameters, and explorer John Fremont saw one gigantic Sugar pine with a 10-foot diameter as he crossed the Sierra (Bonnicksen 2000).

Foresters early in the 20th century noticed that:

“The virgin forest is uneven-aged, or at best even-aged in small groups, and it is patchy and broken, hence it is fairly immune from extensive, devastating crown fires....” (Show and Kotok 1924).

This was soon to change. But the primeval forest was well-adapted to fire, a different kind of fire than we are used to today.

For untold millenia low-intensity ground fires crackled through these “mixed conferous forests” at intervals of 2 to 30 years, apparently both lightning-caused and Native American-set. “...[T]hese were creeping surface fires that licked a few feet up the sides of large trees, and then only briefly,” said Thomas Bonnicksen, forest scientist at Texas A&M University (Bonnicksen 2000). George Sudworth found fire scars on the lower trunks of 50 to 75% of the trees, and concluded that “the fires of the present time are peculiarly of a surface nature, and with rare exception there is no reason to beleive that any other type of fire has occurred here” (Sudworth 1900).

Mormon militiaman Henry Bigler in August 1848 on the West Fork of the Carson River in Alpine County, California near Nevada reported:

“The mountains seem to be all on fire and the valley full of smoke.... At night we could see as it were a hundred fires in the California mountains made no doubt by Indians” (in Gruell 2001).

These Indian fires were most common in lower west-slope and east-side forests. Ron Goode, a North Fork Mono man told ethnobotanist M. Kat Anderson in 1989 that he recalled in the old days people set fires in a particular area in a 5 to 10 year rotation, from 1,500 to 6,000 feet elevation in the San Joaquin River drainage. In other areas people lit fires annually or every two years when they saw that the brush was getting to high in the forest along their travel routes. “They’d burn areas when they would see it’s in need,” said a Mono woman (Anderson 2005: 152).

Lightning from summer and fall thunderstorms set many highcountry fires that burned through forests and meadows. In Yosemite National Park alone over a period of 30 years in the late 20th century lightning started an avergae of 55 wildland fires each year. In some years tens of thousand of acres burned, in other years only a few acres; on average an estimated 16,000 acres per year burned naturally (NPS 2004). In August 1987 a single storm threw more than 1,500 lightning strikes in the Sierra Foothills above Visalia and conifer forests above Fresno and Merced, causing 325 wildfires (San Francisco Chronicle August 31, 1987).

Drier open Ponderosa forests burned often, every 2 to 14 years. Reconstructions based on fire scar studies vary widely, so these ranges are only estimates. To escape the next fire Ponderosa seedlings shot up into pole-sized saplings quickly (in tree terms) -- 14 years on good sites, to 70 years on infertile spots. Flames seared their lower branches off, giving them a fuel-free trunk (ibid.). Less fire-resistant White fir (Abies concolor) and Douglas fir (Pseudotsuga menziesii) seedlings were killed by a passing fire.

Higher mixed conifer forest may have burned about avery 10 to 20 years. Giant sequoia (Sequioadendron giganteum) grove fire intervals ranged from 3 to 60 years, most commonly every 3 to 8 years. Fires in the more arid east-side Jeffrey pine (Pinus jefrreyi) forests may have been less common, perhaps every 60 years, and on steep, moist, cool north-facing slopes dominated by White fir, the fire interval may have only been every 16 to 80 years (Minnich et al. 1995, Stephenson 1999, Skinner and Chang 1996, Barbour et al. 1993, Gruell 2001). This created the uneven-aged mosaics of open park-like stands. Forests had mixes of young saplings and huge old-growth trees, as well as old standing dead snags that made superb habitat for owls, woodpeckers, and bats. Dominance shifted through time between the major tree species. Ponderosa pines and California black oaks (Quercus kelloggii) were favored with frequent low-intensity ground-fires that opened up the forest (their seedlings needed sunlight to grow), while White fir and Incense cedar (Calocedrus decurrens) were fire-sensitive, with thin bark, and they required more shade (Minnich et al. 1995, Garrison, Otahal, and Triggs 2002, Kauffman and Martin 1987, Skinner and Chang 1996). Less frequent, more variable fires favored a mix of White fir, Incense cedar, Doug fir, and Sugar pine (Skinner and Chang 1996). In Ponderosa belts, heavy fuel built up on the forest floor only if a fire did not burn through in a 20-year stretch (Bonnicksen 2000).

Using fire scar data on trees, old photos, early forest plot measurements, and comparisons with forests that had no fire suppression such as the Sierra San Pedro Martir in Baja California, researchers tried to reconstruct what past forest landscapes were like with a pre-contact fire regime. The original forests were apparently a varied and interconnected mosaic of vegetation types in differing stages of recovery from fire and other disturbances. Variations in fire return intervals and fire severity patterns over the years caused both spatial and temporal patch diversity. Some high-intensity fires created patches of native weedy pioneer vegetation. More severe fires often happened on upper slopes. By chance, other forest patches avoided burning for years, building a late successional forest with multi-layered canopy, a higher density of trees, many old dead snags, and coarse woody debris fallen on the forest floor. These patches usually grew on lower slopes or north- and east-facing slopes (U. S. Forest Service Pacific Southwest Research Station 1998). Thus patches of older trees grew within areas of younger trees.

The old-growth patches of large trees produced more cones that Chickaree squirrels (Tamiasciurus douglasi) ate, and more prey for Goshawks (Accipiter gentilis) which favored these large trees and old forests. California spotted owls (Strix occidentalis occidentalis) also needed large tree patches. Black-backed woodpeckers (Picoides arcticus) depended on freshly-killed fire snags.

Deer and elk roamed about searching for the latest succulent regrowth on burns -- their populations rose for 10 to 20 years on a fire-patch, then declined as the animals moved on when the vegetation grew older (Bonnicksen 2000). West slope Mule deer preferred to nibble on the newer growth of such shrubs as Deerbrush (Ceanothus integerrimus), while on the east side they took the new leader twigs of Bitterbrush (Purshia tridentata) -- both became old and less palatable without the regenerating affect of fire. In the 1950s people began to notice that deer populations were lessened as dense young conifers displaced shrubs (Gruell 2001).

The turn of the century saw heated discussions about fire in the forests, as economics drove management. Foresters knew about the old fire regime, but did not like it. An early 20th-century federal forester disdainfully described “ancient notions of ‘Piute forestry’ whose deep fire-scars remain upon so many of our giant landmark pines and sequoias: ‘burn off the rubbish, the dead limbs and stubs, the thick undergrowth and chaparral; clear the way for more forest, incidently get more grass, besides, all the tree-beetles which destroy so much standing timber” (U. S. Forest Service Cooperation 1920). He disparaged this “unscientific” old way of lighting “light surface fires” aimed at “producing a smooth forest floor” (ibid.). The new crop of forest managers had the key in their hands, but they arrogantly threw it away.

“The Forest Service is solidly opposed to every sort of ‘light burning’ beacuse they have seen it in practice many times, under all sorts of conditions; so are the foresters of all civilized nations.... The underlying principles of all scientific forestry, however, are these: Save the young growth as well as the mature trees; protect the soil; encourage reproduction; fill up all possible gaps in the forest cover -- do not make more by surface fires -- fight all fires to a finish” (ibid.).

In 1905 complete fire suppression became U. S. Forest Service policy (Barbour et al. 1993). This increased density of trees everywhere, and for the foresters the profits seemed to increase as well, for a time. White fir densities tripled in many areas and invaded open areas where they had been less common before. Leaf litter levels built up and “ladder fuels” developed -- old twiggy matter along tree trunks -- that allowed fires to climb from the ground into the treetops (Minnich et al. 1995, Stephenson 1999).

Logging began with the Gold Rush in 1848 to supply wood for mining operations, and mills were opened beginning in 1855. First the high-grade Sugar and Ponderosa pines were cut, further tipping the balance towards the White firs which were logged later (Beesley 1996).

A century later the same scenes often became blocked with a dense growth of chaparral, oaks, and young second-growth conifers. Higher up a similar closing-in of the pines, Incense cedars, and Doug firs occurred (Gruell 2001).

Surrounded by saplings that sucked up water the old trees became increasingly stressed, making them more susceptible to insect attack (such as bark beetles) during droughts. A record level of beetle infestation was reported in 2004 in Ponderosas in the Sierra and southern California mountains (and indeed throughout the West), due to the lack of fires that normally cleansed the forests and kept beetle numbers lower. U.C. Berkeley forest ecologist Scott Stephens compared the forests on either side of the California-Mexico border in the San Bernardino Mountains and Sierra San Pedro Martir. On the California side, fire suppression resulted in around 300 trees per acre; on the Baja side, where ground fires continue to burn, only 60 per acre were found. There, the trees remained healthy even during the drought due to less competition for water. In contrast the San Bernardino National Forest tree mortality was the highest seen in a century, according to Stephens -- “It’s absolutely stunning” (Ochert 2003).

Giant sequoias nearly ceased reproducing, as they needed fire to release seeds from the cones and expose ashy mineral soils and open gaps in the forest to germinate in. The small cones hung on the trees for years waiting for the heat of a fire passing below to open the cone scales, releasing a seed rain as great as 8 million per acre. The best results came when the fire was hot and flashy, burning a pile of downed branches or a fallen log -- clusters of Sequoia seedlings could then be found on these spots (Stephenson 1999, Barbour et al. 1993, Bonnicksen 2000). John Muir watched a fire in a Sequoia grove along the Kaweah River in fall 1875, coming upslope from the lower chaparral belt like an “ungovernable flood.” But as it reached the old forest the flames calmed, “creeping and spreading beneath the trees.” It left most old thick orange-barked giants alone, merely leaving black scorch-marks. Occasional branches on the ground flared up into bonfires, and in other spots the flames managed to climb up young tree trunks and ignite the leaves with a roar. Sometimes fire crawled high up a Giant sequoia and lit its top into a firey lamp that burned for days (Muir 1901). But the forest itself was unharmed.

As fire intervals lengthened (now to 100 to 300 years), fuels built up, and fire intensities increased, allowing more stand-destroying crown fires. Before fire suppression surface fuels of pine needles, fallen branches, and small shrubs amounted to fewer than 10 tons per acre; today the same figure is 40 to 50 tons per acre -- low intensity ground fires have not been allowed to “clean up” this debris (Ochert 2003). Severe hot crown fires now ravage the Sierra. I remember seeing the Fountain Fire in 1992 as I drove over Highway 120, a gigantic cloud of smoke billowing upwards like a nuclear explosion in the clear blue sky, generated by wood instead of radioactive atoms. Fire ecologists pointed out that today’s fire regime of fast-moving catastrophic large forest fires are associated with vegetation characteristsics that may be outside the historic natural range of variability: the landscape pattern of homogeneous forests that spring up (or are planted, more usually) were probably very rare before the 20th century (Skinner and Chang 1996).

In addition, fire suppression does not seem to be working. In 2002 fire fighters contained 99% of U. S. wildfires before they got out of control, yet that was the second worst season on record, with 7.2 million acres flamed to charcoal. 7.3 million acres had burned in 2000 (Ochert 2003). Through the western U. S. in the last 20 years a new phenomenon erupted: “megafires,” wildfires well over 40,000 and even to 100,000 acres. Fast, intense, incredibly high flames shooting into the air, so energetic and devastating that firefighters often could not get close enough to fight them. Since the year 2000 Colorado and Utah have seen the largest wildfires in their recorded histories. The incredible heat creates a “hydrophobic” soil which blocks out water penetration, creating conditions that these ecosystems had not seen before (Forecast Earth, The Weather Channel, July 28, 2007).

University of California at Berkeley forestry professor Harold Biswell in the 1940s had suggested using small controlled burns to remove the built-up fuels in forests to restore them to health and lessen the danger of conflagrations. His economically-minded colleagues thought this talk was insane at the time, and the dean of the School of Forestry tried to have him fired, unsuccessfully (Ochert 2003).

Only by the 1960s were government agencies reconsidering allowing some wildfire to occur to help forest health; before this fire suppression was absolute. The National Park Service took the lead, admitting that fire suppression “led to a disruption of ecological processes” (NPS 2004). Yosemite National Park, for example, began mechanically thinning (with chainsaws) the brush and small trees that are clogging the mixed conifer forest in areas along roads and near communities, placing the debris in brush piles, and burning them in the cold season. Prescribed fires were later set to restore the forests. Lightning fires were carefully allowed to burn in certain parts of the high country. All this work is ongoing (www.nps.gov/archive/yose/fire). Using fire alone as a tool, forest researchers beleive that the mixed conifer forest can be restored perhaps within 200 years.

climate and Fire

Observers are discovering evidence that other processes complicate the picture of how fire shaped the landscape: climate clearly enters into the discussion as ecologists study forest burning regimes, such as increased extreme wind events.

Fire scars within the tree ring samples of Sequoais gave a picture of how fires fluctuated with climate change in the past. Wet rainy periods had less fires, while dry years burned more. The old giants revealed a history of fires back 2,000 years, their thick bark blackened but protected with each surface fire. From A. D. 500 to 800 the Sequoia groves showed a lower fire frequency (13 to 29 fires per century), then an increased frequency from A. D. 800 to 1300 (27 to 46 fires per century). This latter period marked an extended drought punctuated by a very wet period (Kinney 1996). These extreme droughts broke out in the mid-1200s that lasted several decades, part of the Medieval Warm Period. A decrease in fire occurred after this, matching the reign of the moister Little Ice Age, although the 1500s had several more extreme droughts shown by restricted tree ring growth (Swetnam and Baisan 1989). Some very wet years had no fires in a grove. But after about 1860 fires suddenly greatly fell in frequency, a sign of the decrease in Native American burning and associated fire suppression. During low fire-frequency periods more fuels accumulated, producing more widespread and intense burns; this in turn tended to homogenize the vegetation pattern. Fuels and vegetation became more patchy during high fire-frequency phases. This matched Richard Minnich’s findings for chaparral. In the Sequoia groves, researchers realized that through time fire frequencies and sizes constantly changed -- there existed no steady-state. Change was paramount in the forest ecosystem in what ecologists call a “nonequilibrium” condition (Swetnam 1993). Because vegetation lagged behind climate change, the condition of the forest at any given point in time may be a legacy of the preceding decades or centuries. Fire in the forest, as elsewhere in California, has been called a “keystone process,” a part of nature that deeply effects every living thing (Stephenson 1999).

Yet to me this indicates not a trend upwards or downwards, but a cyclic and somewhat irregular fluctation in climate and fire regimes. We need to learn more about this cyclic climate behavior, and not assume an apocalypse.

Yosemite Valley

“A fire-glow in the distance, and then the wavy line of burning grass, gave notice that Indians were in the valley clearing the ground, the more readily to obtain their winter supple of acorns and wild sweet potato root -- ‘huckhau.’”

--H. W. Baxley in Yosemite Valley, 1865

In Yosemite Valley, climate-driven vegetation and fire changes are difficult to separate out from human-caused changes, but both processes are indicated. Pollen, leaf fossils, and charcoal from Woski Pond in the valley give evidence of the scene. During the core’s early record from 1,550 to 650 years ago the valley had a forest of Ponderosa pine, Doug fir, Incense cedar, and some White fir. The subapline conifers Lodgepole pine (Pinus contorta) and Mountain hemlock (Tsuga mertensiana) were also found, trees which do not grow in the valley today. Meadows blanketed nearby areas. Charcoal was common, indicating regular fires. Then at A. D. 1300, the start of the drier Medieval Warm Period, the pollen shifted to a more open Ponderosa-Incense cedar-California black oak mix with increased shrubs. The shift was accompanied by a temporary jump in charcoal levels, indicating a large fire disturbance swept through the area, reducing the conifers. After this the wetter Little Ice Age climate took over and moisture-loving Douglas fir and Incense cedar increased. Charcoal levels decreased (Anderson and Carpenter 1991, Woolfenden 1996).

Climate drove the fire regime and forest diversity, but people also played a role. Archaeological remains suggest Native Americans, perhaps the Central Sierra Miwok, moved permanently into Yosemite Valley at about A. D. 1200 and they may have begun a burning regime. Before this, people visited the valley seasonally to hunt and gather, but after this an intensive acorn economy developed (ibid.). Nineteenth century observers such as Joaquin Miller described elderly Miwok women in the spring looking about “for the little dry spots of headland and sunny valley, and as fast as dry spots appeared, they would be burned,” which he noted kept the forest “fruitful” and open (Lewis 1993).

In visiting Yosemite Valley in June 1860, explorer William Brewer described it as green and grassy plain dotted with trees, the Merced River winding through (Brewer 1949), a product of Indian fire management and the current climate of the time.

But with total fire suppression the meadows and open grass flats became choked with shrubs and young conifers despite the climatic conditions. Pines and Incense cedars displaced meadows, shrubs, and oaks. The gaps closed in gradually and the canopy merged (Anderson and Carpenter 1991, Gruell 2001). In Yosemite National Park resurveying of meadows from 1864 to 1943 showed that invading conifers reduced meadowlands from 745 acres to 327 in about 80 years, most probably from the lack of periodic Indian burning (Ernst 1949).

Tuolumne Meadows

Other clues in the mountains also show signs of climate change and altered fire regimes both acting to alter the face of the land. Meadow invasion by conifers has been described by ecologists as a complex and dynamic interaction of fire, short-term and long-term climate change, and vegetative response (Kinney 1996). Geographer Thomas Vale and writer Geraldine Vale shot repeat photographs at previously visited sites in the Yosemite high country, and found that open places like Tuolumne and Dana Meadows had been greatly invaded by Lodgepole pines in the last century. In addition the Lodgepole forest itself had increased in density in many photo pairs (Vale and Vale 1994).

That tree invasion is so rampant in these meadows today could also be a sign we are entering the new warmer drier climatic phase of Global Warming. Post-1850 climatic warming may be drying out meadows, lowering the water table and allowing trees to invade. But changing fire regimes may also be acting to dry out the meadows as well: with the previous open canopy more of this moisture soaked into the ground or ran into nearby meadows, increasing their water table. When the forest canopy closed in during the last 100 years it intercepted a lot more rain and snowfall, evaporating it back into the atmosphere and witholding it from the meadows. George Gruell thought that fire was the single most important factor in the changes, and that the lack of fire heightened the effect of other factors such as climate change; even with climate change, continuing fire would be keeping the meadows tree-free (Gruell 2001).

Lightning fires probably burned through these forests more often than in lower elevations. They were frequent, but may not have spread far due to natural fuel breaks -- granite or volcanic rock outcrops and lush green meadows, for instance. Red fir (Abies magnifica) forests burned somewhere around every 8 to 26 years, occasionally only every 70 years (Gruell 2001, Millar and Woolfenden 1999, NPS 2004). I watched a lightning-caused fire burn slowly through a Red fir-Lodgepole pine forest in Yosemite’s high country. The park let the fire burn while keeping crews nearby to monitor it. I stopped to monitor the fire too -- from a distance a quiet small white plume of smoke emerged from the dark forest. Upon close inspection I saw only a low line of flickering orange tongues creeping along the fir needle floor, crackling and popping as they hit a resin-filled cone or fallen twig, breathing out blue smoke among the trunks. The fire blackened some lower bark, but did not climb the trees -- the forest looked well. I have seen many living Red fir forests with blackened trunks and a lush green carpets of newly-grown Bracken ferns (Pteridium aquilinium) under them, sprung up after a fire.

Lodgepole pines burned about every 34 years in the southern Cascades, some fires burning small areas (55 acres), while other fires spread over 2,600 acres. In the interior of North America, hot fires may have burned only every century or two in these habitats (Bonnicksen 2000). John Muir in 1894 described creeping ground fires burning for weeks in strips across the Lodgepole forest, killing the thin-barked trees. Strong winds sometimes drove the flames into the crowns, “forming one continuous belt of roaring fire that goes surging and racing onward above the bending woods, like the grass-fires of a prairie” (Muir 1894). Fire in these subalpine forests were “stand-replacing,” that is, they killed the forest outright, unlike the lower pine forest regime. Even though Lodgepoles were fire-sensitive, the gaps in the forest allowed for quick regeneration by the seeds which sprouted within a year. Lodgepole stands were often a patchwork of different ages, each associated with a specific fire event. In the Rocky Mountains Lodgepole pines need fire to melt open their resin-filled cones and allow the seeds to come out, but this seems to happen less in the Sierra. Here, the cones usually freely open. Fire-scar studies of Lodegpole pine forests on the Yellowstone Plateau in Wyoming and Montana sadly showed decreasing fire frequencies during the first major European-generated Smallpox epidemic that raged among the Blackfeet in 1781; a second reduction in fires occurred with the next epidemic in 1837, and a third with the 1870 spread of the disease -- Indian burning left a legacy in the trees (Bonnicksen 2000). Whether Sierran Lodgepoles were burned as frequently by Indians is not known.

The repeat photographs of George Gruell (2001) showed that both Red fir and Lodgepole pine forests were more open in the late 1800s, with gaps and clearings; they have since become much denser. The boundary between Red fir forest and Lodegpole pine forest must have constantly and slowly shifted in a mosaic of patches due to influences from fire and climate change. Moister climate phases and low intensity ground fires favored Red firs, while drier times and the occasional hot crown fire favored Lodgepoles as high intensity burns opened up new areas for the pines to colonize. Fire suppression in the last century has inhibited the development of patchiness in the forest (Millar and Woolfenden 1999).

the age of megafires in california

I happened to be driving westward one day in Nevada in the late summer of 2020, and as I topped a pass, I saw an unusual giant cloud formation over the Sierra Nevada mountains. Otherwise the sky was clear blue with no winds, no other clouds. I was puzzled at what this was. I stopped and parked, and without a camera I got out my sjkecthing supplies and made notes and drwaings. Back home I investigated this, and quickly discovered this was a massive fire-caused cloud from the Creek Fire on the west-slope of the Sierra. I added notes to my skecth about this amazing sight.

The ignition date of the Creek Fire was recorded as September 4, in the Big Creek area of the San Joaquin River watershed. I sketched it about a day later. By September 26 the megafire reached 292,172 acres and was only 39% contained, with 3,263 fire crew personell on it. At the time it attained the dubious title of the largest single fire in California history (larger fires were fire complexes of multiple fires in a region).

All said and done the Creek Fire burned approximately 379,895 acres, destroyed 853 structures, and damaged an additional 64 structures on the Sierra National Forest below Big Creek. The U.S. Forest Service reported: "Despite an exhaustive investigation, the cause of the fire is officially categorized as 'undetermined.' Fire investigators determined that the most probable cause was a lightning strike" (https://www.fs.usda.gov/detail/sierra/news-events/?cid=FSEPRD932048).

What happened here?

I was lucky enought to talk with a fire crew boss recently on 2021, who worked on the Creek Fire. He described the area as chock-full of very dense beetle-killed trees from dorught years. The fuel was at a very high level, and dry. When the fire ignited, there was no wind, but days later the winds picked up and he witnessed embers being carried more than two miles ahead of the flame front, igniting new spot fires. He told me he had never seen anything like this in his career. The inferno under the pyrocumulus was astounding, with lightning strikes coming out of the cloud.

To me this seemed like a drastic and dangeous rebalancing of Sierra Nevada fuels after a century of fire suppression and cultural genocide.

Instead of the useful cultural burns of old, today’s wildfires in chaparral and conifer forests are often huge conflagrations, whipped up by extreme winds, excessive fuels, and drought conditions, creating flames raging to 300 feet in height. Winds of 50 miles per hour can push the spread of fires to 400 feet per minute.

Ironically, record-breaking wildfires ripped through southern California while I was researching the topic of fire ecology in 2003 for my book -- I emailed my friend Gale, a Cleveland National Forest biologist living in Julian at the time. She and her house escaped the rush of wind and flame by a mere two miles, but she told me how the town was reduced to a pre-modern state: no electricity in most of the region meant no well pumps working, no gas stations operating, and only candlelight at night for days. Another friend who had collected some of my paintings lost them when his house built deep in the chaparral east of San Diego burned to the ground in a chaparral wildfire.

Fast-forward to 2022, and the outbreak of megafires has only gotten worse in California, with huge loss of life, property, and impacts to local communities and wildlife.

The Camp Fire in 2018 in the Sierra Nevada killed 85 people and destroyed the town of Paradise, as raging winds toppled a tree onto a transmission line which ignited a wildfire--horizontal wind-driven flames roared through a ponderosa pine forest. The fire burned 153,336 acres.

More megafires made the news: the Mendocino Complex Fire in 2018 burned 459,123 acres in the North Coast Range and killed a firefighter; the Dixie Fire in 2021 starting in the Feather River drainage bruned an incredible 963,309 acres, cost an estimated $1.15 billion to fight, destroyed 1,329 structures, and was caused by PG&E utility infrasctructre; and the Caldor Fire in late 2021 burned paridly through forest in El Dorado and Amador Counties, closing Interstate 80 and threatening the doorsteps of Lake Tahoe communities--only October snow seemed to have staved off a worst disaster.

Many ignitions are due to high-voltage transmission lines that cross wildlands, with utilities far behind on trimming trees that fall on powerlines during wind storms. This is not just a climate change problem, but a problem of society's over-dependence on utility infrastructure for energy needs, instead of building distributed generation.

Yet lightning events still happen across California, as shown by the August Complex Fire in the North Coast Range, where dry lightning strikes in August 2020 caused 38 separate fires which then merged into a monster wildfire that became the single largest wildfire in modern California history, at 1,032,648 acres.

The ongoing devastating fire seasons in California illuminate the ongoing debate among fire ecologists about how to protect life and property along the wildland interface.

The Remembered Forest

Instead of locking forests into giant log-preserves or static wilderness areas, many Indigenous people and tribes have a different way of being upon the land, a way that has helped shaped California for thousands of years. One part of this philosophy was to start small disturbance regimes in various plant communities which created openings in the forest and brush, patches of early successional states.

Traditional ecological knowledge is being put into action in Plumas National Forest, as the United Maidu Nation people made agreements with the Forest Service to manage 2,100 acres of their homeland. Their goal is not lumber, but the ground cover: basketry plants such as willows, medicinal plants, plants used for ceremonial purposes, and oaks used for food. The Maidu Stewardship Project area is in the Wolf Creek drainage between the town of Quincy and the resevoir Lake Almanor in the Feather River watershed. I searched for Mountain yellow-legged frogs along this creek in the 1990s, and found nothing but overgrazed and highly eroded floodplains with helicopters carrying out pine logs overhead. The place was heavily impacted.

“We want to show we can bring the land back just by taking care of it,” said Lorena Gorbet, spokeswoman of the Maidu (Little 2002). They pruned willow and maple twigs for weaving materials, selectively tilled the Camas lily (Camassia quamash) beds by digging in certain areas for the bulbs, and cultivated the oaks by encouraging large low branches that would produce big acorn crops. Fire was one of their most important tools. Annually they set low-intensity ground fires to encourage such basketry plants as Beargrass (Xerophyllum tenax). One of Gorbet’s cousins remembered his grandfather using a forked stick to pull the slow-burning flames away from plants he wanted to keep unburnt. As recently as the 1970s Maidu people were given citations by the Forest Service for “illegal” burning. The Maidu plan on using ground fires again, as well as removing introduced weeds, replanting native grasses, willows, and aspens, and halting erosion on the floodplain. Students work to monitor these projects.

“The Remembered Forest” is a phrase the Klamath, Modoc, and Yahooskin tribes use today as a goal set for restoration of the forests on their Klamath Reservation in southern Oregon, forests continous with Ponderosa, Lodgepole, and mixed conifer forests along northeastern California. The people here remembered when they set fire to the old pine forests to renew berry crops and make flourishing wildlife habitat, they remembered how fire sculpted the forest, lightly, clearing the ground of excess brush, forming over time patches of old large Ponderosas and small clusters of new trees in an irregular mosaic. The Klamath tribes are now restoring the complexity to the forests, reducing stand densities built up from fire suppression by removing some White fir stands (and leaving others), clearing saplings from around the huge old pines to protect them from fire, carrying out prescribed fires, and in turn re-newing their own identity and continuity with the past.

“As the true stewards of this land, the Klamath Tribes seek to reclaim our stewardship and, to the extent possible, return the forests and waters to their original splendor,” said Klamath Tribes Chairman Allen Foreman. “For we share a vision: ‘When we heal the land, we also heal people’” (Wolf 2004). [See www.klamathtribes.org]

Living with Fire

Contrast this with the current still-dominant view of fire: as more conflagrations erupted with terrible predictability in southern California on October 25, 2007, I watched Katie Couric on the CBS Evening News talk of “war” and “evil.” The latest California wildfires were “man vs. nature,” 9,000 firefighters “battling the beast,” and “launching a counterattack” with an “air war” of planes dropping water to save houses. Sixty to 100 mph “evil devil winds” (the Santa Anas) spread the numerous fires over 700 square miles, causing the largest evacuation in state history -- “another brutal day on the front.”

The great wildfire debate will undoubtedly go on among ecologists, but both sides agree that rural migration out into the once remote firebelt has cost us dearly. San Diego will decide on whether to adopt a Rural Lands Initiative that limits the spread of houses into the remote indefensible backcountry where “intermix” wildfires around scattered houses are very difficult to fight. Buffer lands may be bought and developers offered incentives not to build in highly fire-prone areas. In addition, county supervisors in San Diego discussed adopting the Minnich Model following the 2003 disaster, planning to control-burn 27,000 acres a year in a large rotation during cooler safer weather periods (Downey 2003).

My conclusions: California scrublands, oak woodlands, and forests are complex and dynamic ecosystems, so we should remain open to as much evidence and as many points of view as possible in order to try to understand them. Fire is not “evil,” but a process of cyclic renewal. Wildfire has scorched the landscape for thousands of years and certain cultures learned to live with it and use it -- when will we?

Closely observing fire effects

Nature journaler and illustrator Robin Carlson has spent years field-sketching fire impacts to landscapes, floras, and faunas in a place-based perspective that I really admire. Spending so much time on one place over days, months, and years, observing how fire can renew the land, and how the vegetation recovers, is central to how I record change in California. Observing and sketching how butterflies and other insects return to the burned area, and wildlife reappearing, shows us that fire is not inherently bad, it is a process of change that can renew the landscape. These detailed observations are important to make, and we can all learn from them. See https://coldcanyonfirerecovery.com