Giant Sequoia Ecology: Fire and Reproduction
NPS Logo

H. Thomas Harvey



The conclusions presented here are based on the findings from this and other studies. They include specific new facts about the giant sequoia and associated organisms, as well as inferences as to the relationship of physical and biotic factors to the perpetuation of this unique species. Both facts and inferences are offered as bases for interpretation of the life of the giant sequoias and for their management. One of the prime concerns is that of the position of giant sequoia in the successional pattern of western coniferous forests.

Successional role of giant sequoias

Our studies support the widely held belief that the giant sequoia (Sequoiadendron giganteum) is a fire subclimax species of the middle elevation mixed coniferous forests of the central and southern Sierra Nevada (Hartesveldt 1964; Kilgore 1970, 1973). The particularly high seedling survival rate in burn pile soils—about 10 times that of seedlings on other manipulated substrates—along with the almost total absence of seedlings from the undisturbed forest floor, show the dependence of the giant sequoia on periodic fire for successful reproduction. Burn pile soils probably approach natural conditions where heavy fuel loads have accumulated. Although giant sequoia seeds may fall throughout the year at a rate approaching 1,000,000 seeds per hectare, their germination and subsequent seedling survival are minimal without the aid of fire. Thus fire, and apparently the hotter the better, is the prime requisite for the reproduction of sufficient seedlings so that the species may survive.

Older sequoia trees may persist long enough so that they remain as relicts in climax stands of white fir (Hartesveldt 1962; Bonnicksen 1975). Their contribution to the perpetuation of the species is probably minimal unless some major disturbance occurs to the forest floor. If this does not occur for prolonged periods of time, thick reproduction of white fir and other shade tolerant species can become established, and heavy concentrations of fuel may accumulate. As a shade intolerant species, but relatively more resistant to fire than other trees, the giant sequoia is favored by such factors as fire which set succession back to an earlier stage.

Role of animals in giant sequoia reproduction

The animal species which appear to affect the giant sequoia in either the seed or the seedling stage are relatively few in number. The early instars of the true bug Ischnorrhynchus resedae feed upon the endosperm of seeds in open cones in the tree. The small long-horned beetle (Phymatodes nitidus) and the Douglas squirrel (Tamiasciurus douglasi) mainly feed on the cones but not the seeds. In so doing they provide a continuous seed dispersal force. A squirrel may feed on an average of 200 cones per large tree per year, thus releasing about 40,000 seeds per tree. On the other hand, beetle activity may release about 120,000 seeds per tree per year. Neither species damage or consume a significant proportion of the seeds. In spite of this continual supply of seeds to the forest floor, very few are likely to survive unless some disturbance exposes mineral soil. During dry years very few seedlings appear in the undisturbed ground surface, however with exceptionally high precipitation as many as a thousand seedlings per hectare may spring up. Due to desiccation and insect and fungal attacks very few if any seedlings are apt to survive more than a year or so.

In addition to releasing seeds, some animal activity reduces a tree's production of seeds. For example, the gelechiid moth (Gelechia sp.) attacks first-year cones which results in the release of seeds that are non-germinable. Some moth infestation causes extended browning of cones and release of germinable seeds. This is a minor contribution to seed release.

Two species of moth caterpillars (Sabulodes caberata and Pero bebresarius) and a camel cricket (Pristocauthophilus pacificus) were found active during the seedling stage. In a wet year the mortality of seedlings due to such insect activity was about 25%. This was second only to desiccation as a mortality factor. It is hypothesized that only in the first few years after fire, when giant sequoia seedlings are disproportionately plentiful while other plants are in short supply, the sequoias are heavily used as insect food. After two or three years have passed, sequoia seedlings are greatly reduced in number and other plant species have increased to the point that they then serve as the major insect food.

Survival strategies of the giant sequoia

The giant sequoia has apparently evolved a set of adaptations which enable it to survive and prosper after fire sweeps through the forest. It has developed a cone which reduces seed loss while insuring seed release when eaten by animals. These two strategies involve an explosive reproduction and a repeated reproduction, respectively. The explosive reproduction follows fire, when numerous cones are opened by the heat of the fire and the ground is cleared of litter and duff so that the millions of seeds released will fall on an optimal substrate. The repeated reproduction strategy involves continual release of seeds by the action of animals and the fortuitous combination of optimal soil moisture and light, combined with disturbance of the forest floor. An example of this latter type of reproduction is found in the falling of a large sequoia at the edge of a meadow. Mineral soil is exposed to seeds with ample light and soil moisture to insure their germination and the subsequent survival of the seedlings.

The giant sequoia is similar to most organisms in having reached a compromise between K selection and r selection (Krebs 1972). It appears to be K selected in its population constancy, large body size, repeated reproduction, and great age. It appears to be r selected in its survivorship curve (type III, high early stage mortality), rapid development, high rm, and relatively early reproduction.

Serotinous cones have been evolved in the giant sequoia, as was first suggested by Shellhammer. They differ from those evolved by other conifers in that they function not so much as deterrents to seed predation by rodents, but rather as a type of cone in which the scales have become a food source in themselves. The closed-cone pines have evolved cones which, because of being enlarged and hard, reduce the loss of seeds due to predation (Smith 1970; Daubenmire 1974). Giant sequoias, however, in providing a source of food in the cone scales, seem to be able to take advantage of the feeding activity of Douglas squirrels and beetles to assure a continuous release of seeds. This, along with the long period of time over which closed cones with germinable seeds may persist, coupled with large cone crops (up to 40,000 per tree per year) and numerous seeds per cone (200 on the average) make the giant sequoia fit not only for continuous reproduction but also provide it with a reservoir of large numbers of seeds for propagation when a fire releases them and sweeps the forest floor clean.

The apparent co-evolution of three animal species that feed on sequoia cones reduces competition for a food source and aids the sequoia in its repetitive reproduction strategy. Although early infestations of the gelechiid moth may bring about the release of nongerminable seeds, successful infestations may induce release of germinable seeds from second-year cones. The Douglas squirrel feeds mostly on cones that are from two to five years old, while the long-horned beetle feeds on cones four years and older. These three species apparently have partitioned the food source on a basis of age of cone, with each in turn taking a greater and greater percentage of the cone crop with increasing age of the cones. Therefore, as soon as seeds become germinable there is a continuous sequence of animals feeding upon the cones that releases a constant supply of such seeds upon the ground. This persistent fall of seeds provides reproduction during the occurrence of favorable situations on the forest floor, and thus is a vital part of the repetitive reproduction strategy.

Hot fires beneath giant sequoias may heat closed cones to the extent that they die, dry out and release their seeds much in the fashion of the closed-cone pines (Daubenmire 1974). The additional adaptations of thick bark and evanescent branches enable mature trees to resist fire and serve as a continuing source of seeds for thousands of years. The tree produces cones which are serotinous and are therefore present many years after the seeds become germinable. The annual crop of cones is added to those that persist from previous years to provide an accumulation of tens of thousands of cones per tree. These numerous closed cones are then available to shed seeds if a hot fire comes through and dries them out. Our evidence that the hottest fire conditions produce the most favorable substrate for seedling survival, when coupled with serotinous cones, indicates that giant sequoias are adapted for explosive reproduction.

Sequoia management

Fire management policy

The concept of fire as a management tool in western forests was first developed by Weaver and Kallander during the 1940s (Kilgore 1976). Their principal objective was to reduce the hazards of wildfires. In California the leader in investigating prescribed fires was H.H. Biswell, whose work started in 1951 in ponderosa pine forests (Kilgore 1976). The study reported here, which started in 1964, was also aimed at reducing wildfire hazards, but in addition, emphasis was given to understanding the effect on giant sequoia reproduction. The role of fire in sequoia groves is now seen to include both fuel reduction and enhancement of giant sequoia regeneration. Thus the use of fire as a management tool in sequoia forests may serve the role of reducing wildfire hazards while at the same time producing for a relict species, the giant sequoia, the natural conditions under which it seems to have evolved and upon which it depends for its survival.

Inasmuch as the majority of giant sequoias are under the jurisdiction of the National Park Service, management policy will be mainly discussed in that context. The other two governmental agencies with management decisions over the species are the State of California and the United States Forest Service. The use of prescribed fires as a management tool in sequoia groves by these agencies began in 1975. In that year the Forest Service conducted a small test fire in the Nelder Grove near Yosemite National Park. Also in 1975 the California Department of Parks and Recreation, under H.H. Biswell's direction, used prescribed burns in the South Calaveras Grove.

The fire policy of the National Park Service has gradually shifted from fire exclusion to fire inclusion. Beginning in 1886 in Yellowstone the policy was to suppress all fires (Agee 1974). The National Parks Act of 1916 emphasized the protection of objects, including trees, rather than processes. Therefore the early emphasis was on trying to perpetuate the status quo without fully appreciating the inexorable process of succession.

Kilgore (1976) reports that ". . . present National Park Service fire management policy divides all fires into management fires or wildfires. It defines management fires as those of both natural origin and prescribed burns which contribute to the attainment of the management objectives of a park through execution of predetermined prescriptions defined in detail in a portion of the approved resources management plan." This policy allows some natural fires to burn, recognizes prescribed burning as an appropriate management tool and continues wildfire suppression in developed areas. Wildfire suppression may also be necessary in undeveloped areas where fuel levels have built up to the point where natural, lightning-caused fires might become so intense as to destroy the entire forest community.

Natural fires above about 2600 m (8,000 ft) in Sequoia and Kings Canyon National Parks have been allowed to burn since 1968, while prescription burning has been employed at lower elevations (Kilgore and Briggs 1972). These policies have resulted, by 1975, in the prescription burning of about 5,400 ha (13,730 acres) outside the Natural Fire Management Zone of the two parks (Shuft 1973; Parsons pers. comm.).

According to Kilgore and Sando (1975), the use of prescribed burns can be instrumental in reducing the probability of crown fires and the high intensity of surface fires. However, in as little as 5-8 years enough litter may fall to the ground, in part induced by the prescribed burn, to bring the ground fuel level back to that preceding the fire (Kilgore 1975; Kilgore and Sando 1975). Therefore it follows that repeated prescribed fires may be needed to reduce this fuel load until a natural fire frequency may be attained (Kilgore 1975).

Several criteria should be met in order to implement an effective fire management program in giant sequoia groves. Basically an overall policy on fire must be developed by the agency with jurisdiction over each of the groves. Different agencies have different objectives and therefore will probably develop different policies which will be codified in their management program. In the National Parks the overall tone was set by the Leopold Report of 1963 (Kilgore 1976). The report called for re-creation of primitive America and noted the unnaturally dense growth of shade tolerant trees on the west side of the Sierra which has resulted from over protection from natural surface fires. Valid objectives could include restoration of open forests, reduction of fire hazards, increased giant sequoia reproduction, and perpetuation of a given vegetation mosaic. Once the objectives of fire management and control are agreed upon, then an inventory map of the understory fuel types should be prepared (van Wagtendonk 1974). Blocks or strips to be burned must be carefully chosen to take into account topography, fuels and vegetation type. If a certain vegetation mosaic is an objective, then certain blocks may have to be left unburned.

If prescription burning is to be implemented, a detailed burning plan must be developed. This should include the objectives of the burn and the approximate dates during which the burning may be conducted. A prescription appropriate to the area should be developed and refined as burning is applied to the blocks within the area in question. At the actual time of burning, fuel stick moisture, relative humidity, wind velocity, and temperature must be assessed to be assured that they fall within the prescription parameters. If they do not, consideration should be given to immediate suppression activity.

In addition to the prescription fires, natural fires may be allowed to run their course provided they are within the Natural Fire Zone of a park and meet the objectives of the Management Program. Such fires should be closely monitored, and an appropriate committee should evaluate the impact of a given fire with the options to suppress, limit or allow continuence of the fire.

According to Henrickson (1972), the giant sequoia is a good example of the need for fire to preserve a species. He further adds that the fauna dependent on it, although of secondary importance, is almost neglected. Our studies indicate that the fauna is not greatly disturbed by small fires and that only a few of its species appear to be dependent on the earliest stages in succession, and they are not restricted to it. Therefore, the fauna associated with the giant sequoia, with the exception of a few insects, could probably survive even if the giant sequoia became extinct. Indeed, Bendell (1974) suggests that most vertebrate animals in coniferous forests are broadly adapted and persist through the changes induced by fire. The sequoia ecosystems have been subject to fires apparently for millions of years, and thus it is to be expected that organisms that are a part of that ecosystem have evolved to fit the alterations brought on by fire. One inference that can be drawn, however, from Bendell and our studies is that small patchy fires favor wildlife. This is in keeping with current guidelines proposed by Briggs (pers. comm.) that small strips not exceeding 30 m (100 ft) in width be employed, and that they be burned from the top down. This type of burning would allow wildlife to leave the fire area and then return, particularly if the fire has been somewhat patchy in its effect.

Succession and management

The studies by Bonnicksen (1975) on the pattern of succession in a giant sequoia—mixed conifer forest indicate that it consists of a mosaic of even-aged stands of different species. Our studies indicated that, at the level of treatment which was given, early-stage plants, e.g. annuals and shrubs, disappear or begin to be reduced within ten years. Giant sequoias, after developing in great numbers initially, are in less than 10 years reduced to only about 5% of the original seedling population. Kilgore (1973) reported even greater mortality, with 98% of the 54,000 seedlings dying within two years after a fire-induced population had been established. In our study the distribution of surviving trees is patchy, with thickets of sequoia saplings occupying burn pile soils and few surviving elsewhere. These relatively dense stands exemplify the mosaic pattern of reproduction identified by Bonnicksen (1975).

The general pattern of giant sequoia succession is that of patchiness. Rarely are substrate conditions uniform in a sequoia grove, even though it may have the same general climate. In addition, disturbance factors such as fire, disease and windfall generally occur in a patchy manner. The result of this nonuniform environment is a mosaic of vegetation types or successional stages. The implication of this pattern for management is that fire as a tool probably should not be applied evenly in a short period of time throughout a large area. Prescription fires should be applied in a patchy manner thus coming closest to re-establishing the primitive mixed conifer forest. The overall long-term goal should be the establishment of conditions that would allow natural processes to operate uninterrupted in the ecosystem. The intensity and extent of treatment will be an important determinant in the response of the vegetation. Thus it is important to clearly identify the objectives of any given management burn before carrying it out.

Hazard trees

The immense size of the branches of the giant sequoia may cause damage to property and injury or even death to people. Falling sequoias may do the same. The factors that may interact in causing trees and branches to fall are varied. Roots infected by fungi may weaken support, fire scars sever root connections, heavy snowfall overload crowns, winds blow down trees, streams undercut root systems, carpenter ants weaken wood by excavation of galleries, and wet soils provide minimal support. Recent studies by Piirto (pers. comm.) specifically implicate Fomes annosus in root failures in many giant sequoias.

In 1969 four giant sequoias, or portions of them, fell and resulted in the death of a woman. One tree, with a lean toward the fire scar side and carpenter ant and fungal activity at the plane of failure, fell and knocked the top section out of a neighboring tree which in turn struck and killed the woman. However prone to falling this tree may have been, there are numerous cases where leaning fire-scarred trees have not yet fallen. There are also cases where upright, apparently sound trees have fallen. It seems prudent to minimize the probability of injury and damage by removing heavy, prolonged human activity from sequoia groves, particularly from areas of hazard tree concentrations.

Summary of findings

Natural history concepts

The giant sequoia is a relict species both geologically and successionally. Only remnant populations of a once widespread species that spanned the northern hemisphere now persist on the western slope of the Sierra Nevada of California. In these remaining groves succession tends to lead toward replacement of the giant sequoia by shade tolerant species such as white fir. The ability to live for several thousand years and grow to be the tallest tree in the forest enables the giant sequoia to remain as a relict in the successional sense.

Two reproductive strategies seem apparent in the life cycle of this species. Repeated reproduction may occur as trees fall over and provide a suitable substrate and open up the canopy so that seedling sequoias may survive. Other disturbances to the forest floor also may enable seedlings to become established, e.g. deposition of sandbars in a river. The second strategy is operative when surface fires of fairly high intensity burn through sequoia groves. The serotinous cones provide an abundant source of seeds which may be released by the heat of the fire to provide for an explosive reproduction of the tree and for grove expansion.

Although in general the perimeters of giant sequoia groves appear to be relatively stable, the youngest trees exist on the edge in a few specific cases. Such trees were dated as being about 100 years old and their beginning coincided with the last fire in the areas in question. It therefore appears that expansion of groves is possible if conditions become favorable for seedling growth.

Seedfall apparently occurs throughout most of the year mainly due to the drying of cones after they have been attacked by the small long-horned beetle, Phymatodes nitidus. Contrary to Boe (1974), seeds are not released by drying of cone scales in the fall of the year. He misquotes Buchholz (1938). Buchholz was only speculating that such a thing might happen, but then went on to say that old cones, even over 20 years old, have near normal numbers of seeds in them. Our studies indicate that squirrel activity augments that of the beetle to provide the constant seedfall.

Seeds which have fallen to the ground may be exposed to sunlight and desiccation prior to germination. These factors, plus the apparent low germination percentage of seeds released from the cone, result in low viability of seeds at the ground level. This may be due in part to the mode of action of releasing seeds from the trees. The action of the beetle causes the cones to dry and cone scales to shrink. The seeds are then exposed to the air and may lose their viability while still in the tree. Seed tests at the ground showed only about 1% viability. Given 1,000,000 seeds, however, this potential would yield 10,000 seedlings per hectare per year if conditions were favorable for their establishment.

The factors which either inhibit seed germination or seedling establishment in unburned areas can only be alluded to at this time. Desiccation appears to terminate most of the seedlings in the burned areas and probably is equally effective in unburned sections. It may be even more effective, for as Stark (1968b) has shown, giant sequoia seeds will germinate in the duff and litter if adequate moisture is present. The developing seedlings however, would be more subject to desiccation inasmuch as more of the root would not be in contact with the mineral soil.

Insect feeding on, and sun scald of seedlings would probably be less in an unburned area than in a recently burned site. Stark (1968a) reported less sun scald in seedlings when litter was present. Insect feeding and sun scald were mortality factors in burned sites, although at a much lower level of intensity than desiccation.

Both light conditions and allelopathic substances hypothetically should be more adverse to seedlings in unburned areas than in burned. Lower light intensity could induce shorter roots and thus greater chance of desiccation. Allelopathic substances might also build up in unburned soils.

The one factor which appears to release the greatest number of viable seeds quickly and to prepare the best of seedbeds for the giant sequoia is fire. And given the fire adaptations of the tree—thick fire resistant bark and high canopy—seldom is much severe damage done to the mature trees.

Prescription burns of relatively small size (10 ha) apparently have little effect on bird and mammal populations, which agrees with conclusions made in other similar studies (Kilgore 1971; Bendell 1974; Vogl 1973). The mosaic of fire intensities caused by the existing vegetation mosaic allow animals to flee to refugia during fires. The extensive use of fire probably would increase the numbers of woodpeckers, flycatchers and other insectivorous birds and ground squirrels. The greater the intensity of the fire, the earlier the stage in the successional pattern that will be developed.

The Douglas squirrel is the only vertebrate that appears to exert much of an effect on sequoia reproduction. The small size of the seeds and their wide dispersal and their loss in the litter and duff make them of minimal food value to mammals. Rodents show a very low interest in sequoia seeds, apparently preferring the larger seeds of the other conifers in the mixed conifer forest of the mid-elevations of the Sierra Nevada.

Few giant sequoia seedlings were observed to be eaten by vertebrates. Vertebrate predation accounted for less than 2% of the mortality of one to three year old seedlings. Only immediately after a fire, when other ground food sources were reduced and giant sequoia seedlings were abundant, did vertebrates ever seem to turn to them for food.

Douglas squirrels store only giant sequoia cones in a sequoia—mixed conifer forest. Although they were observed eating the seeds of other conifers, no caches were found other than of giant sequoia cones. They feed on sequoia cones throughout the year, leaving them only as other food sources come to maturity. Other foods included fungi and seeds or fruits of such nonconifers in season as hazelnuts. The preoccupation of sequoian squirrels with giant sequoia cones marks them as rather distinct behaviorally, probably as the result of early learning on the part of young squirrels. Their feeding on sequoia cones probably has a minimal effect on the giant sequoia as only about 200 cones per tree per year are eaten. However, as high as 2,000 cones may be cut from one tree during a year. No nests of the Douglas squirrel were observed other than in a cavity in a large sequoia.

Territoriality was strongly exhibited by Douglas squirrels in high population years. The giant sequoia appeared to be the stimulus and focal point of their agonistic behavior. Only one squirrel ever occupied a given large giant sequoia, loudly fending off any would-be intruders. When populations were small, possibly due to severe winters, and squirrels were widespread, their territorial behavior dropped practically to zero, and they moved between several giant sequoias and reduced their caching activities.

The insect fauna associated with the giant sequoia is highly variable. As discussed earlier a few insects affect seed production and early seedling growth. Once the tree is grown, a relatively small but fascinating fauna develops. Several fluid-feeding insects concentrate on the lower foliage of mature trees. Aphids, treehoppers and true bugs had representatives in this category, while chewing insects such as geometrids and leaf beetles were concentrated in the upper portion of the crown.

In conclusion, our findings indicate that cone predation by animals is somewhat partitioned among them so that competition is minimized between predators. First year cones are utilized by a moth, while cones 2 to 5 years old are fed upon by Douglas squirrels, and older cones (4 to 11 yrs) are the domain of a minute long-horned beetle. The squirrel exploits the cones primarily for the food value of the cone scales, thus releasing the seeds. The beetle mines throughout the cone and in so doing interrupts the vascular system which may sustain the cone in a green, closed condition for over 20 years. When the vascular system is severed the cone turns brown and the scales shrivel and release their seeds. Therefore, through the activity of the beetle and the squirrel there is a constant release of seeds. The establishment of seedlings, however, is best after a fire, which also may release seeds. Without fires intense enough to open the canopy, shade killing of young sequoias may occur and leave only sequoias as successional relicts surrounded by shade tolerant white firs. For these reasons it is essential that fire be re-introduced in its natural role if the sequoia ecosystem is to survive in anything resembling its primitive state.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 06-Mar-2007