Abstract
The ecology of coast redwood (Sequoia sempervirens) is examined with an emphasis on the climatic, physiographic, edaphic, and hydrologic factors responsible for its present distribution. Adaptations to fire and flood-interrupted environments, as well as resistance to depredations by mammals, invertebrates, and fungal pathogens, are recognized as important adjuvants to its regeneration and dominance throughout its range. A review of paleo-sequoian distributions for coast redwood, dawn redwood and giant sequoia is also provided.
Detailed appendices include a botanical comparison of the three redwood species, and an annotated register of the tallest coast redwoods. The etymology of Sequoia sempervirens is also discussed.
Acknowledgments
I would like to express my gratitude to Drs. Michael Kutilek, Wanna Pitts and Howard Shellhammer for their review of this manuscript, to the late Dr. H. Thomas Harvey for his guidance and inspiration during the production of “On the edge: nature’s last stand for coast redwoods,” and to Verl Clausen of the Sempervirens Fund for his resolute confidence in my ability to write and narrate the documentary video.
I would also like to extend a special note of appreciation to Hjordis Madsen and the staff of the Inter-Library Loan office at San Jose State University, whose indefatigable efforts to retrieve many of the documents cited herein were truly blessed events.
Introduction
Discovery
The superlative redwood groves of the central California coast had been known to Native Americans for nearly 11,000 years before the arrival of Don Gaspar de Portola in 1769 (Greenlee, 1983). Although redwoods may have been chanced upon by former explorations of the coast (Dewitt, 1985; Dolezal, 1974; Hewes, 1981; Weaver, 1975), the chronicle of Portola’s expedition overland provides the first written description of the trees (Carranco, 1982). On October 10, 1769, a Franciscan missionary, Fray Juan Crespi, described the Spanish transit north of the Pajaro River, “…over plains and low hills, well forested with very high trees of a red color, not known to us. They have a very different leaf from cedars, and although the wood resembles cedar somewhat in color, it is very different and has not the same odor; moreover, the wood of the trees that we found is very brittle. In this region, there is great abundance of these trees and because none of the expedition recognizes them, they are named redwood from their color (Appendix I).”
The first botanist to examine the coast redwood, or “palo colorado (Jepson, 1910),” attended Alejandro Malaspina’s landfall on Monterey Bay in September of 1791 (Thomas, 1961). The expedition fielded two botanists, Louis Nee and Thaddeus Haenke, but only Haenke collected ashore where Malaspina had seen “the red pine, a tree much taller than the rest (Eastwood, 1939).” It is curious that neither scholar attempted to name the dominant conifer, since Nee later described the coast live oak (Quercus agrifolia) and valley oak (Q. lobata) from specimens brought to him by ship’s officers. Apparently, one redwood is known to still be growing in Spain from seed collected during the expedition (Baker, 1965).
Whereas Haenke might be considered the “botanical discoverer” of coast redwoods (Jepson, 1910), a more appropriate candidate would be Archibald Menzies, a Scottish botanist and surgeon with the Vancouver Expedition of 1790-1795 (Baker, 1965; Jepson, 1923; Roy, 1966). In 1794, Menzies obtained a sample of the tree near Santa Cruz which was to become the basis for its botanical description by Aylmer Bourke Lambert in 1824 (Jepson, 1910; Hastings, 1928). m ere is some doubt, however, that Menzies personally collected the type specimen, since Vancouver was known to deny him landfalls (Eastwood, 1939) and Menzies’ “journal does not show that he himself was ever at Santa Cruz though other members of the expedition were (Shirley, 1937).”
Delayed Description
Upon Vancouver’s return to London in 1795, Menzies’ herbarium collection was cached in the British Museum (Natural History) for more than a quarter century before coming to the attention of Lambert. Inasmuch as Menzies only collected in California during the late fall and early winter (Eastwood, 1939), the unsatisfactory condition of his redwood specimen may have been responsible for the delay in its description. Indeed, Eastwood (1939) discovered many of Menzies’ California specimens still unnamed.
But Lambert acknowledged while “having only a single imperfect specimen of this species for examination, it is not without some hesitation, that I have referred it to Taxodium… leaving to future observations to determine, whether or not the place I have assigned to it be correct (Jepson, 1910).” Although Lambert recognized that the redwood appeared to be intermediate between Taxodium and Cupressus, and that a colleague, “the late Mr. Salisbury considered it as forming a new genus, and had applied to it the name Condylocarpus (Jepson, 1910),” he chose Taxodium because of the specimen’s close resemblance to the bald cypress (Taxodium distichum). But the evergreen nature of the redwood was very much unlike the deciduous cypress, so he christened the species Taxodium sempervirens to differentiate between the two habits (Jepson, 1923). The propriety of “sempervirens,” meaning “ever-living (Hewes, 1981)” or “always-alive (Coombes, 1987),” was incidental to the tree’s extraordinary longevity (Baker, 1965).
Lambert’s binomial survived intact for only 23 years before yielding to the adoption of Sequoia sempervirens by Stephen Endlicher in 1847 (Chaney, 1951). In his manuscript, Synopsis Coniferarum, Endlicher redescribed the coast redwood with good reason to segregate it from Taxodium (Hartesveldt et al., 1975). But Endlicher’s publication failed to intimate the origin of the word “sequoia (Hastings, 1928);” and his untimely death in 1849 left the etymology of the new genus regrettably shrouded in mystery.
Origin of Sequoia Name
Punctilious scrutiny of Endlicher’s papers by the eminent American botanist, Asa Gray, could not betray the origin of Sequoia (Hartesveldt et al., 1975), which has most often been associated with the remarkable Cherokee Indian scholar, Sequoyah, or Sikwayi (Farquhar, 1947). Though lacking formal education (Shirley, 1937), Sequoyah devised an 85-character syllabary for the Cherokee Nation in 1821 to facilitate reading and writing in their native language (Hartesveldt et al., 1975). His ignoble death in San Fernando, Mexico in 1843 (Sargent, 1947) may have prompted Endlicher, a known philologist, to honor Sequoyah’s passing.
But “no one has ever found mention in his writings of Sequoyah’s name (Hartesveldt et al., 1975)” or the Cherokee syllabary. Rather, Gray believed that the stem of the word was a derivation from the Latin “sequi” or “sequor,” meaning “sequence (Hastings, 1928)” or “following (Harvey, 1978),” and alluded to the fact that redwoods were remnants or followers of numerous fossil ancestors. Another proposal of an obscure origin in one of California’s Indian dialects remains unsubstantiated (Hartesveldt et al., 1975), and in Cherokee, “sequoyah” means “opossum (Dolezal, 1974).” Yet the labors of the gifted Sequoyah are certainly deserving of such lasting tribute, regardless of Endlicher’s intention. Hartesveldt et al. (1975) concur that “perplexity and doubt notwithstanding, let it so remain.”
Species Associated with Sequoia
Sequoia sempervirens belongs to the redwood (Jepson, 1910) or deciduous cypress family (Hewes, 1981), Taxodiaceae, representing 15 species and ten largely monotypic genera. Included are two species closely related to the coast redwood: the giant sequoia (Sequoiadendron giganteum) and the dawn redwood (Metasequoia glyptostroboides) (Stebbins, 1948).
The giant sequoia, largest of all living things with more than 630,000 board feet in the General Sherman Tree (Harvey et al., 1981), is widely dispersed among 75 groves and 35,607 acres along the western Sierra Nevada in California (Hartesveldt et al., 1975). The range of the smaller and uniquely deciduous dawn redwood is limited to the central China provinces of Sichuan and Hubei where they were discovered by Professor T. Kan in 1941 (Hu, 1948).
While the coast redwood may be regarded as the giant sequoia’s nearest living relative geographically, only recently segregated by genus (Buchholz, 1939), the dawn and coast redwoods are more closely related (Harvey, 1978). Sequoia sempervirens may even be descendant “from hybrids between an early Tertiary or Mesozoic species of Metasequoia and some probably extinct type of taxodiaceous plant (Stebbins, 1948).” Further comparison between the three species may be found on pages 90-92 (Appendix II) .
Mesozoic Origins
The redwoods are considered to be relicts of a Mesozoic group of conifers that were once richly developed and coherent, and widespread throughout the northern hemisphere (Florin, 1963). Progenitors of Taxodiaceae might have evolved as early as 200 million years ago (Engbeck, 1976), when the cycadeoids, ginkgos, and other conifers dominated the Mesozoic forests of the dinosaurs (Simpson, 1983).
The proliferation of the sequoia line was favored by a remarkably stable climate that was generally warmer, humid and more equable than at present (Florin, 1963). This was especially true during the Cretaceous period (Appendix III), 135 to 65 million years ago, when the average temperature of high-latitude North America was 59° to 77° F warmer with perhaps 25 percent more annual precipitation (Rigby, 1987). Mild conditions persisted even in northern extremes where Florin (1963) found no evidence of Mesozoic glaciation in the arctic islands regarded by Berry (1920) as the center of sequoian evolution.
Although Engbeck (1976) preferred a more southern origin in west-central North America, the paleontological record is far from conclusive. And the oldest known sequoias are found in southern Manchuria from late Jurassic deposits (Chaney, 1951). However, by the end of the Mesozoic era, representatives of each extant genera had become established throughout the northern continents (Fig. 1).
Widest Distribution
Sequoias achieved their widest and northernmost limits of distribution during the Paleocene and Eocene epochs, 65 to 38 million years ago (Florin, 1963), when the general cooling trend in Tertiary climates was often interrupted by warmer periods (Tidwell, 1975). Ancient relatives of the giant sequoia occupied parts of central and eastern North America, Greenland, Spitzbergen and Europe; those of the coast redwood became established in central and western North America, Greenland, Spitzbergen, Europe and Japan (Florin, 1963). Early Metasequoia forests were even more widespread, colonizing eastern Asia, Japan, Greenland, Spitzbergen, northern Siberia, and Ellesmere Island.
Although plate tectonic theory was not discussed by Chaney or Florin, Chaney (1948) attributed the success of Metasequoia in extreme latitudes to its deciduous habit which enabled it to endure prolonged winters without sunshine in a dormant condition. From the late Mesozoic until Miocene time, Metasequoia was also the most abundant and widely distributed genus of Taxodiaceae in North America (Chaney, 1951).
Incomplete Fossil Record
It should be kept in mind that such broad paleodistributions are biased by an incomplete fossil record spanning millions of years and do not imply that the northern hemisphere was simultaneously covered by temperate sequoian forests through 40 of latitude (Cain, 1944). The sequoias were only part of an exceptional diversity of species and taxa characteristic to subalpine and mixed conifer forests of the period, and were regularly admixed in communities far richer than any surviving today (Raven and Axelrod, 1978). Examination of the communities associated with each ancestral sequoia indicates that they inhabited environments similar to conditions occupied by their modern counterparts (Hartesveldt et al., 1975), with latitudinal distributions that fluctuated with oscillations in global climate.
For example, Tertiary Sequoia forests were spread mainly over a belt bounded by latitudes 34 and 58, yet northern outposts existed for a time in western Spitzbergen during the Paleocene or Eocene at latitude 79 (Florin, 1963). Whereas some sequoian species inhabited ranges that overlapped (Chaney, 1979), others remained as mutually exclusive as the absence of contemporary fossils of Sequoia and Sequoiadendron have shown in western North America (Mason, 1947; Raven and Axelrod, 1978). However, Florin (1963) noted that the reconstruction of past Sequoiadendron distributions has been much less satisfactory than those of Sequoia because of “difficulties involved in the identification of fossil remains (Florin, 1963).”
Species Recognized
The number of species recognized by paleontologists has also been subject to considerable debate and revision. By 1910, more than 40 species of sequoia had been described from fragmentary remains, but there was by no means agreement as to the validity of the species (Jepson, 1910). And Seward (1919) determined that many impressions of vegetative shoots and cones described as Sequoia from the Jurassic and early Cretaceous did not bear close scrutiny. The twelve species of fossil redwood reported by Shirley (1937), Hewes (1981) and others most likely refers to Jepson’s (1910) account of the number recognized by Schimpfer in 1903.
The collection of the first living specimens of Metasequoia in 1944, only three years after Shigeru Miki described the genus from Pliocene fossils in Japan, led to an important contribution in the revision of paleo-sequoian nomenclature by Chaney in 1951 (Florin, 1963). Chaney (1951) reassigned Tertiary fossils of several North American taxa, including 11 species of Sequoia, into new combinations of Metasequoia occidentalis and Sequoia affinis. Both were recognized as ancestral to their extant species to the extent of being very nearly conspecific. The immediate ancestor of the giant sequoia, Sequoiadendron chaneyi, was later described by Axelrod (1956) from Mio-Pliocene floras in Nevada.
Impact of Cooler, Drier Climate
In contrast to the conditions which encouraged periods of northern expansion, a progressively cooler and drier climate throughout the remainder of the Tertiary precipitated a gradual retreat of the sequoias to the relict areas they now occupy (Engbeck, 1976). As early as the late Oligocene, 26 million years ago, Sequoiadendron-like conifers had vanished from the floras of eastern North America, Greenland, Spitzbergen and Europe (Fig. 2), while Metasequoia had retreated to latitudes bounded by 40° and 49° in western North America, and 47° and 58° in western Siberia (Florin, 1963). Although the Oligocene range of Metasequoia was expanded across central Asia, from the Pacific Ocean to the Aral Sea, it was to perish altogether from North America by the end of the Miocene, 20 million years later. And by the close of the Tertiary, 2.5 million years ago, Metasequoia had become extinct in central Asia, Siberia, Kamchatka and Japan (Fig. 3).
Florin (1963) reconstructed similar patterns of retreat from high latitudes for Sequoia, accompanied by a lowering in its altitudinal distribution in the mountains. The northernmost limits of its range appear to have been latitude 60 during the Oligocene (western Siberia), 56 during the Miocene (Denmark), and 51 during the Pliocene (the Netherlands). There was also an Oligocene expansion of its area across central Asia, comparable to that of Metasequoia, which prefaced its southernmost record of distribution in southwestern China during the Pliocene at latitude 26 .
Yet the Eurasian migrations of Sequoia terminated at the end of the Pliocene, when it disappeared completely from Europe, Japan and Asia. In North America, its survival was aided by the equability of the climate along the Oregon and California coast (Raven and Axelrod, 1978) where the abundant rainfall and prolonged maintenance of narrow fluctuations in seasonal temperature provided refuge from an increasingly hostile continental interior (Li, 1953).
Sequoia forests were further isolated from the westward advance of xerothermic conditions by the onset of coastal orogenic movements during the late Pliocene (Raven and Axelrod, 1978). Although subsequent Quaternary migrations extended as far south as the Santa Ynez Mountains near Santa Barbara (Chaney and Mason, 1933; Putnam, 1942), the distribution of Sequoia by the end of the Tertiary was similar to its present range, with representatives persisting only as far inland as 35 miles southeast of Santa Rosa (Axelrod, 1976, 1977).
Present Range
The last and present stronghold of Sequoia sempervirens adjoins the Pacific along a narrow and discontinuous belt from southwestern Oregon to Monterey County in California (Fig. 4). Roughly 5 to 35 miles wide and 450 miles long (Roy, 1966), 1,971,000 acres of virgin redwood forest awaited Portola’s discovery in 1769 (USDI, 1964). However, since the construction of the first water-powered sawmill in 1834 (Carranco, 1982), the redwood lumber industry has not only harvested more than 95 percent of the timber (Kelly and Braasch, 1988), but diminished its range by 100,000 acres (Stone, 1965). In California, just 68,035 acres of old-growth redwood, 3.5 percent of the original forest, have been preserved within 21 state and two national parks (Appendix IV).
Northernmost
Of the six principal redwood groves in Oregon surveyed by the U.S. Forest Service in 1964, the northernmost is bounded by two clear-cut logging sites in the Little Redwood Creek drainage of the Chetco River (Becking, 1971). Other stands are scattered along downstream tributaries to within six miles of the Chetco River mouth (Sudworth, 1927), to the east along Wheeler Creek, and throughout the Winchuck River and Bear Creek watersheds into California (Becking, 1971).
Interior Redwoods
Ranging southward to Sonoma County, the “redwood belt (Jepson, 1910)” is continuous except for a transverse break along the Kings Peak Range and headwaters of the Mattole River (Roy, 1966). Another hiatus occurs among the hills of southern Sonoma and northern Marin Counties, where “coastal” redwood stands have been displaced eastward into Napa County (Griffin and Critchfield, 1976). Only isolated colonies lie east of Napa, achieving their farthest distribution inland some 42 miles from the sea. However, to the north near Angwin and the Pope Valley, a more “interior” locality of redwood may be found upon the eastern flank of Howell Mountain (Sudworth, 1967), 36.8 miles from the coast (Zinke, 1977). “This is a most remarkable station, for not elsewhere does the Redwood occur on the waters of the Sacramento or any other interior stream (Jepson, 1910).” Three other colonies have subsequently been found to face the Sacramento Valley drainage: at Swartz Creek, Ink Grade and .i.Aetna Springs; (Griffin and Critchfield, 1976).
Southern Balance
The southern balance of the redwood belt occurs in detached and irregular areas most prominently among the canyons of Marin County, the Oakland Hills, and Santa Cruz Mountains (Jepson, 1923). Only scattered remnants are found south of Monterey, confined to coastal arroyos where “stringers of redwood (Zinke, 1977)” seldom extend more than 175 feet from main stream channels (Borchert et al., 1988). In the Soda Springs Creek drainage (Borchert, 1990; personal communication), 1.5 miles north of Salmon Creek (Griffin and Critchfield, 1976), a small windswept clump marks the southernmost limit of redwood distribution, surrounded by coastal sage scrub a few hundred meters inshore (Zinke, 1977). The often published “claim that Salmon Creek Canyon is the southern limit of the redwoods is erroneous (Havlik and Ketcham, 1968).”
Restricted by Salt Spray
Recognized as a typical lowland species (Becking, 1982) ranging in elevation from 3200 feet (Borchert et al., 1988) to near sea-level (Jepson, 1910), redwoods do not inhabit coastal margins where the influx of marine air is excessive (Zinke, 1977). Prevailing winds and ocean fogs, high in salt spray aerosols, not only contribute unsuitable amounts of sodium and magnesium to ocean terrace soils, but directly inhibit redwood growth through foliage salt burn (Zinke, 1964).
Insufficient rainfall on the immediate coast also establishes the western redwood boundary (Baker, 1965), especially in areas exposed to desiccating offshore winds (Daniel, 1942). Redwoods grow best beyond the reach of salt spray fallout atop moist, sheltered plains and valleys opening toward the sea, along river deltas and the protected flats and benches of larger streams, and upon moderate western slopes (Sudworth, 1967) between 100 and 2500 feet in elevation (Person, 1937). As one progresses inland, the influx of marine air and the occurrence of summer fog have a significant impact upon local factors which tend to favor or mitigate against the presence of redwood forest (Zinke, 1977).
Climate
Cool, wet winters and warm summer droughts exemplify the Mediterranean-type climate of the redwood region (Bakker, 1984). Maritime fog and stratus regimes ameliorate conditions throughout most of the dry season (Byers, 1953), which Thornthwaite (1941) classified as dry subhumid to superhumid. The amount of annual precipitation increases northward along the coast from 20 to 120 inches (McMinn and Maino, 1980), primarily falling as winter rain although snow sometimes accumulates upon the highest ridgetops (Sudworth, 1967).
Exceptional years have recorded as much as 153.54 inches at Monumental Station in Del Norte County (Sprague, 1941). Severe winter storms, like the 1989 blizzard which left 14 inches of snow at Prairie Creek Redwood State Park, are rare (Wilkinson et al., 1989). Eighty percent of the total precipitation falls between November 1 and April 30 (Twight, 1973), with January normally the wettest month, and August the driest (Fowells, 1965).
Summertime droughts increase southward along the coast from two to five months in duration (Kuser, 1976), aggravating the protracted moisture requirements of redwood during its growing season from mid-March until September (Roy, 1966). Kuser (1976) found that supplemental moisture from high water tables and fog was critical during this period, contributing an equivalent of 16 to 30 inches of water to the 40 to 80 inches of annual precipitation needed for optimal redwood growth.
Summer Fogs
The redwood belt has long been causally linked to frequent summer fogs (Cooper, 1917) which “seem to be more important than the amount of precipitation in delineating the redwood type (Fowells, 1965).” From early spring until September, advection fogs intermittently blanket the coast when the moist marine air borne by prevailing north-westerly winds cools upon contact with coastal upwellings of deep-sea currents, 10 to 15° F colder than the surface (Gilliam, 1962). The winds and onshore pressure gradients respond to low pressure areas above the Central Valley, driving the fogs inland (Patton, 1956), pervading mountains, river valley gaps and canyons while sliding beneath the warmer air of the interior to form a temperature inversion (Gilliam, 1962).
The boundary between the two layers remains intact as the volume of fog and cool marine air increases over land, pushing the ceiling of warm air upward. Additional condensation occurs along this plane of contact to create stratus, or high fog. Seldom found below 300 feet, the stratus layer commonly penetrates the coast at 800 to 1200 feet above the inland valleys, capped by 1000 to 3000-foot ceilings which average around 1500 feet (Byers, 1953). Diurnal temperatures above and below the layer can differ by as much as 65° F (Borchert et al., 1988). Although the fog/stratus layer normally lasts from late evening until morning, it may also perdure for several days with the more coastal sites hesitant to dissipate in the afternoon warmth (Azevedo and Morgan, 1974). The predominance of fog/stratus penetration along the coast ebbs and flows in cycles, and achieves a maximum width of 100 miles over land and sea in August (Gilliam, 1962).
The precipitative importance of summer fog drip has received considerable attention since Cannon (1901) first compared the redwood’s “fern-like” boughs to filters “by which water may be ‘combed’ out from fog (Cannon, 1901).” Redwoods require prodigious amounts of moisture during the growing season, which Golte (1974) attributed to the low efficiency of their vascular conducting system. Transpiration rates of 500 gallons per day have been reported by Hewes (1981), whereas more drought-resistant associates, such as old-growth Douglas-fir (Pseudotsuga menziesii), transpire 140 gallons daily (Kline et al., 1976).
Azevedo and Morgan (1974) determined that fog drip affects both water balances and nutrient cycling within such coastal ecosystems. They recorded as much as 3.15 inches of fog precipitation beneath one Humboldt County redwood in 48 hours. And in the mountains east of Half Moon Bay, an astounding 58.8 inches of fog drip was collected by Oberlander (1956) under an exposed, 20-foot high tanoak (Lithocarpus densiflora) in 39 days! The extent of forest cover, position and shape of the tree, wind velocity and temporal character of the fog affect the spatial distribution of fog drip upon the ground (Azevedo and Morgan, 1974), which Parsons (1960) construed as mainly a hillcrest phenomenon where sufficient exposure, elevation and temperature combine to intercept stratus below the maximum 49 F required for drip formation (Freeman, 1971). Because dense, wet surface fogs seldom penetrate the major redwood groves (Byers, 1953), “the more extravagant claims advanced for water income from this source should be treated with reserve (Kerfoot, 1968).”
The redwood’s marked mesophytic reaction to low humidities (Daniel, 1942) led Freeman (1971) to conclude that reduced insolation and the high relative humidity of fog/stratus events were more important than fog drip in reducing summer water losses. Inasmuch as the energy expended in the evaporation of leaf surface moisture is the same for dissipating transpirational water, intercepted mists that do not reach the soil would be just as significant to the foliage as moisture gained from fog drip (White, 1958). “Moreover, some of the condensed water is actually absorbed and redistributed within the plant (White, 1958).”
Fog-laden air may even increase photosynthesis as a result of its high carbon dioxide content (Wilson, 1948; Wiant, 1964), or by contributing to the reduction in leaf water deficits which limit oO2 assimilation (Hodges, 1967). It has also been suggested that the level of diffuse radiation on the forest floor might conceivably be higher on thin foggy days than when days are clear, particularly during the late afternoon (Black, 1963). The “summer fog blanket (Cooper, 1917)” not only relieves evapotranspirational stress by altering net radiation, temperature and humidity (Marotz and Lahey, 1975), but may sharply delimit redwood distribution in areas where its influence is overcome by inland heating of the land (Zinke, 1977).
One such example is the Mattole River basin where grassland persists though redwood forest might be expected (Bakker, 1984). Desiccating offshore winds and downdrafts peculiar to the inner face of the Kings Peak Range appear to be responsible for this anomaly, generated by large back eddies in the summertime wind pattern (Zinke, 1977). Redwoods are conspicuously absent from the basin despite the fact that its overall annual precipitation of 120 inches (Cooper, 1965) is the highest in the redwood region (Zinke, 1977). The trees, however, are not strictly circumscribed by the limits of fog/stratus penetration (McBride and Jacobs, 1977), and may persist in some relatively fog-free localities (Adams, 1969; Cooper, 1965) while failing establishment in others subject to heavy summer fogs (Baker, 1962; Cooper, 1965; Zinke, 1964).
Indeed, along the California coast, the Point Reyes Peninsula weathers the greatest number of days with fog (Sprague, 1941), yet supports only Douglas-fir and Bishop pine (Pinus muricata) forests (Evens, 1988). Since the frequency range of early morning and early season stratus within the redwood belt is substantial, fog/stratus events may not serve as a dominant factor in redwood distribution (Marotz and Lahey, 1975). At least, no causal relationship between coastal fogs and redwood has been clearly established (Black, 1963; Simmons and Vale, 1975).
Temperature
Redwoods require a temperate, maritime regime of temperatures where monthly means do not fall below 36 F, nor exceed 84° F (Kuser, 1976). Mean annual temperatures throughout the region vary between 50 and 60 F, with daily extremes rarely falling below 15 F or rising above 100 F (Roy, 1966). The range in mean annual maximum and mean annual minimum temperatures increases with elevation and distance inland (Zinke, 1977), from 10 or 15° F along the coast to 30 F for the easternmost colonies (Person, 1937). Position relative to the ameliorating influence of the sea controls the gradient of increasing temperature extremes, compounded by local topographical factors (Zinke, 1977). Climate, therefore, tends to be more severe in the leeward valleys exposed to broad ranges of temperature and low humidity (Franklin and Waring, 1980).
Frost Intolerance
The sensitivity of redwood seedlings and young foliage to persistent frost also restricts the northern portion of its range (Daniel, 1942; Kuser, 1976; MacGinitie, 1933) to the warmer ridgetops, saddles and upper slopes below 1500 feet in elevation (Becking, 1971 ), where snowfall might not exceed five inches per year (Becking, 1967).
Redwood frost intolerance may be aggravated by the unusually high water content of its tissues: as much as 70 pounds, or 8.4 gallons, per one-inch by one-foot by twenty-foot board (Kuser, 1976). In growth chamber studies of redwood seedlings, Hellmers (1964) noted a marked restriction in growth at nocturnal temperatures below 59 F which might explain why the tallest redwoods grow so near their northern limit of distribution. Optimum seedling growth was achieved under 66 F day and 59 F night conditions with little evidence of the salient thermoperiod characteristic of other conifers (Hellmers, 1966). This result is consistent with those of Kuser (1976), who found the highest site index or potential productivity of coast redwood at a mean summer temperature of 64° F.
Impact of Winds
The deleterious impact of ocean winds, high in salt spray aerosols and frequency, not only precludes redwood establishment along the proximate coast, but along interior wind gaps such as those between Bodega Bay and Petaluma, the Golden Gate, and the Salinas Valley (Zinke, 1977). The severity of exposure increases southward along the Santa Lucia Mountains where less than two percent of the annual precipitation falls from June to September (Borchert et al., 1988). Wind dessication and foliage salt burn severely stunt the growth of southern redwoods near canyon mouths and ridgetops (Becking, 1971), repeatedly killing the tops of unprotected trees (Daniel, 1942). The redwood’s considerable sensitivity to transpirational water loss may result from the inability of their stomatal guard cells to close properly (Kuser, 1976).
Wind-dwarfed redwoods may even suffer a net reduction in diameter due to the cumulative dessication of their tissues over a period of years (Haasis, 1933). Hence, the southernmost colonies persist only upon western and northern exposures that are at least moderately sheltered from ocean gales (Roy, 1966). “The conditions are, on the whole, so unfavorable that… mature trees with very long branches, broad or irregular crowns, or with a flat crown like a broad, flat hat are a feature of this country (Jepson, 1910).”
Examples of wind-intensified evapotranspirational stress can be seen throughout the redwood belt in the severe dieback of many mature Sequoia crowns reduced to naked spires (Kuser, 1976; Stone, 1965; Zinke, 1964). Commonly known as “spike-tops (Fritz, 1931),” the condition is familiar among old-growth stands (Cooper, 1965) where wind speeds exceed the tolerance of redwood to drying effects on the canopy (Zinke, 1964).
Older, larger redwoods are more vulnerable because their ability to conduct water through the saPwood decreases with age (Waring and Schlesinger, 1985), and the time required to transfer water from the roots to the leaves is too great in larger trees to permit roots to contribute much to daily water deficits (Waring and Running, 1976). Nocuous moisture stress may also result from fire damage to the basal cambium, diminishing the active water conducting area of the sapwood (Fritz, 1931). Wiant (1964) further proposed that since the lower limbs of older redwoods might require a greater portion of the photosynthates used in respiration, stress for both moisture and carbohydrates may contribute to the spike-top death of upper limbs and branches.
Windfirm
Although redwoods do not have tap roots (Fritz, 1978), they remain remarkably windfirm under most conditions (McCollum, 1957). Their extensive, shallow root systems, four to six feet deep and up to 50 feet in lateral spread (Shirley, 1937), frequently interlock (Becking, 1982), allowing individual trees of great height and massive crown to resist windthrow, especially among groves with uniform stand density (Sturgeon, 1964). The susceptibilty to windfall, therefore, is largely focused upon redwoods growing along the margins of virgin stands, roads and logging sites (Cooper, 1965). A combination of strong wind and saturated soil is necessary for significant windfall damage, compounded by the depth of soil, condition and size of the root system, crown size and shape, trunk strength, and collision with other falling trees (Boe, 1966).
In the aftermath of the Columbus Day hurricane of 1962, Boe (1966) surveyed the timber lost on several experimental cutting sites hammered by winds gusting to 120 knots for three to five hours. Eighty-three percent of redwood casualties was attributed to uprooting, while breakage accounted for the rest. Bole rot was not considered to be a major factor since 59 percent of the broken trees showed no signs of disease. Redwoods with the smallest and largest diameters emerged the most windfirm; and losses were proportionately lowest at the shelterwood cutting site where windfall was minimized in all diameter classes.
In the protected forests of Jedediah Smith Redwoods State Park, the damage to virgin redwood was minimal, yet poignant, for “95 percent of the old-growth trees that fell were Douglas-fir. Of the relatively few redwood trees that went down, the majority fell because Douglas-fir trees fell into them (Sturgeon, 1964).” It is important to appreciate that storms of this magnitude are exceedingly rare within the redwood belt (Decker et al., 1962; Fujimori, 1972); and that from 1960 to 1962, just three winter storms accounted for nearly all of the windfall damage discovered by Boe (1966) on the experimental cutting sites.
Topography
Characteristic of the topography that dominates much of the redwood forest is the rugged, broken landscape of the Coast Range Mountains (Person, 1937), deeply etched by rivers and streams that often parallel the coastline and the San Andreas Fault (Zinke, 1977). Interior valleys accordingly trend southeast to northwest, although many individual ranges cut obliquely across the belt and terminate at the sea (Howard, 1979). The precipitous rise of peaks along the coast can be impressive, but the total relief is considered small for a mountainous country (Person, 1937) where the average summit commands an altitude of 2000 to 4000 feet (Howard, 1979). Roughly 40 to 90 miles in width, the Coast Ranges trend about 30 west of north which tends to accentuate the interior climatic aspects of river valleys that drain to the northwest (Zinke, 1977).
By reinforcing the summer influx of marine air from the prevailing northwesterly winds, redwood distribution in areas like the Eel River Valley extends further inland. When river valleys broaden to form alluvial flats and benches, redwood stands of almost unbelievable volume occur, but they constitute only a small percentage of the total (Person, 1937). The hewn topography of the Coast Range Mountains is principally forged by long, narrow watersheds with steep slopes and considerable movement of soil (Black, 1963) punctuated by numerous linear lowlands such as the Petaluma Valley and by a few irregular basins such as Clear Lake (Howard, 1979).
Geology
The predominant rock is sedimentary (Zinke, 1977), a marbled coalition of coastal and Franciscan sandstones from the late Jurassic and Cretaceous, redoubled by Tertiary marine deposits and cleaved by an eastern belt of weakly metamorphosed Franciscan material (Irwin, 1960). The Franciscan Formation differs from the coastal series by the almost complete absence of potassium feldspar in the dominant rock, graywacke (Waring and Major, 1964), and by its characteristic assemblages of greenstone, chert, slate and minor limestones (Irwin, 1960).
In addition, marine and coastal sandstones generally lack the clay that is abundant in the matrix of Franciscan sandstone (Thomas, 1961). Significant amounts of shale and conglomerate are found in all formations (Irwin, 1960), with serpentine and schist locally abundant (Person, 1937). And in the north Coast Ranges, the absence of granitic intrusions distinguishes the area geologically from the Klamath Mountains in the east (Waring and Major, 1964), and from the Santa Cruz (Thomas, 1961) and Santa Lucia Mountains in the south (Borchert et al., 1988).
Geological Types
The geological types trend in a manner similar to the topography of the region, thereby reinforcing the edaphic and topographical controls on the distribution of flora (Zinke, 1977). Intrusions of basalt are commonly shot through the thin-bedded sandstones of the Franciscan Formation, cemented by clays and silicious elements that readily erode into good forest soils (Baker, 1962). But site quality and timber types vary with the character of the binding cement and the amount of basalt in soil deposits, both of which affect porosity.
Younger, less consolidated sandstones near the coast generally produce deeper soils with greater water-holding capacities than those older and harder and farther inland (Zinke, 1977). However, coastal terraces adjacent to the sea often have ancient surfaces covered with old, infertile soils and hardpans which restrict drainage and soil aeration, promoting depauperate vegetation of pygmy forest types such as cupressus pygmaea and Pinus contorta ssp. bolanderi in Mendocino County (Westman and Whittaker, 1975).
And where local outcrops of serpentine and peridotite give rise to shallow soils that are extremely high in magnesium (Baker, 1962) and low in potassium, calcium and phosphorous (Zinke, 1977), redwood distribution abruptly ends (Becking, 1971; Zinke, 1964). Similar barriers occur upon glades or grassland openings where high-pH, heavy clay soils are derived from rocks richer in basic elements than those of adjoining forests (Zinke, 1977).
Grassland prairies further dominate the heavy clay soils that arise from metamorphic rocks like glaucophane schist. The exclusion of redwoods from these grassland areas might even be advanced by the unfavorable influence of humus and grass-root remains upon the sod (Zinke, 1964).
Soils
A wide range of soil types occur within the redwood region (Black, 1963), marked by considerable variation in texture: from thin rocky loams on some of the steepest slopes to deep, fine sandy loams on alluvial flats and benches (Person, 1937). Among the most productive are the Mendocino (Zinke, 1964), Empire (Roy, 1966), Hugo, Larabee, Melbourne, Josephine (Black, 1963), Gamboa (Borchert et al., 1988) and Ferndale series (Waring and Major, 1964).
Conifers typically do best upon medium textured, deep soils with more than four feet to bedrock, where permeable, well-drained profiles are moderately acidic increasing with depth (Storie and Wieslander, 1952), which, for coast redwood, ranges in pH from 5.0 to 7.5, with an optimum pH of 6.5 (Zinke, 1964). These conditions are roughly analogous to the upland forest soils of the Hugo series whose gravelly to clay loams (Lenihan, 1990) support nearly 80 percent of redwood stands in central Humboldt County (Waring and Major, 1964; Zinke, 1964).
The sovereignty of the Hugo series is maintained by the youthful topography of the region (Zinke, 1964), subject to frequent earthquakes and relatively rapid uplift (Black, 1963), facilitating an equilibrium between soil erosion and weathering rates (Zinke, 1964). Both the Hugo and its more inland counterpart, the Josephine series (Storie and Wieslander, 1952), are developed on Franciscan sandstone, and normally possess 15 to 20 percent gravel with sufficient clays to qualify as loams (Black, 1963). Although Waring and Major (1964) found the Hugo to be the most variable of those studied, the series may be broadly classified as a gray-brown podzolic soil (Black, 1963; Zinke, 1964) or, on occasion, a fine, loamy, mixed mesic dystric xerochrept (Lenihan, 1990)!
Subsoil Development
Debris avalanche is a dominant hillslope erosion process in the Coast Range Mountains (Borchert et al., 1988), yet pockets of colluvium may remain stable for as long as 17,000 years to permit significant pedogenesis (Marron, 1982). Subsoil development may even be advanced by slope creep, superimposing one subsurface layer upon another to form a double B horizon, and possibly improving site quality (Zinke, 1964). Within the maritime province of northern redwood forests, upland soils weather to produce loams that are comparable to the Melbourne and Larabee series, with deeply developed horizons on subsoil clay formations that often support almost pure stands of redwood (Black, 1963). The gravelly to clay loams of the Larabee arise from weakly consolidated conglomerates, gray-brown to pale brown in color (Waring and Major, 1964), with common clay contents of 20 to 25 percent (Black, 1963). Larabee deposits may also exhibit a semi-permeable horizon of clays, three or more feet below the surface, which affects subsurface water drainage. Residual soils of the Melbourne have contrastingly darker profiles, with dark brown to brown clay loams and gleyed subsurface horizons that are chiefly derived from graywacke sandstones, shales and conglomerates (Lenihan, 1990).
A more complex assemblage of parent rock material fosters the most productive, albeit weakly developed, series for redwood in the Santa Lucia Mountains (Borchert et al., 1988). Known as the Gamboa series, these gravelly to very gravelly loams generally develop on debris slide colluviums of Franciscan shales and sandstones, abundantly admixed with granitic-metamorphic deposits of gneiss, schist, marble, quartz, diorite and tonalite. Borchert et al. (1988) concluded that the severity in slope topography was largely responsible for the weak pedogenesis of southern subsoils, since the period of pocket development is probably longer and the duration of stability shorter than in most northern colluviums.