Professional Documents
Culture Documents
Biomass, Production and Woody Detritus in An Old Coast Redwood (Sequoia Sempervirens) Forest
Biomass, Production and Woody Detritus in An Old Coast Redwood (Sequoia Sempervirens) Forest
DOI 10.1007/s11258-005-2322-8
-1
Key words: Decay, Natural disturbance, NPP, Old growth, Pacific Northwest, Stand dynamics, Woody debris
Abstract
We examined aboveground biomass dynamics, aboveground net primary production (ANPP), and woody detritus
input in an old Sequoia sempervirens stand over a three-decade period. Our estimates of aboveground biomass
ranged from 3300 to 5800 Mg ha1. Stem biomass estimates ranged from 3000 to 5200 Mg ha1. Stem biomass
declined 7% over the study interval. Biomass dynamics were patchy, with marked declines in recent tree-fall
patches <0.05 ha in size. Larger tree-fall patches approaching 0.2 ha in size were observed outside the study plot.
Our estimates of ANPP ranged from 6 to 14 Mg ha1yr1. Estimates of 7 to 10 Mg ha1yr1 were considered to
be relatively accurate. Thus, our estimates based on long-term data corroborated the findings of earlier short-term
studies. ANPP of old, pure stands of Sequoia was not above average for temperate forests. Even though
production was potentially high on a per stem basis, it was moderate at the stand level. We obtained values of 797
m3 ha1 and 262 Mg ha1 for coarse woody detritus volume and mass, respectively. Fine woody detritus volume
and mass were estimated at 16 m3 ha1 and 5 Mg ha1, respectively. Standing dead trees (or snags) comprised
7% of the total coarse detritus volume and 8% of the total mass. Coarse detritus input averaged 5.7 to 6.9 Mg
ha1yr1. Assuming steady-state input and pool of coarse detritus, we obtained a decay rate constant of 0.022 to
0.026. The old-growth stand of Sequoia studied had extremely high biomass, but ANPP was moderate and the
amount of woody detritus was not exceptionally large. Biomass accretion and loss were not rapid in this stand
partly because of the slow population dynamics and low canopy turnover rate of Sequoia at the old-growth stage.
Nomenclature: Hickman (1993).
Biomass loss and the dynamics of dead wood in Natural disturbances in the study forest include fire,
Sequoia forests have not been fully characterized flooding and tree fall. Fire events are thought to cause
(Sawyer et al. 2000b). Bingham and Sawyer (1988) cumulative damage to Sequoia trees that may result in
noted large volume (957 m3 ha1) and mass (200 Mg death (Finney and Martin 1989). The presettlement
ha1) of coarse woody detritus in an old upland return interval for fire in Humboldt County and sur-
Sequoia forest. Other estimates for Sequoia forests rounding areas is estimated at <10 to 500 yr (Veirs
range from 10 to 280 Mg ha1 (Sawyer et al. 2000b). 1982; Stuart 1987). Veirs (1982) notes that mesic sites
However, rates of detritus input and decay have not near the coast experience light fires every 250 to 500
been widely reported. Sawyer et al. (2000b) speculate yr. Sites intermediate on the mesic-xeric continuum
that the decay rate of woody detritus is extremely slow have a fire return interval of 100 to 200 yr. Xeric
in Sequoia forests. interior sites have a return interval of about 50 yr. The
A three-decade study of the dynamics of compo- study forest is likely to have a fire return interval
sition and structure in an old Sequoia forest revealed <100 yr (Stuart 1987). However, fire suppression
slow canopy turnover (Busing and Fujimori 2002). policies implemented in the 1900s have greatly re-
Here we provide complementary information on duced fire occurrence for at least the last 50 years
fundamental ecosystem processes. Our objectives are (Stone and Vasey 1968; Veirs 1982). By contrast,
to 1) estimate aboveground wood volume and bio- flooding has occurred in the study stand in recent
mass, 2) examine the trajectory and spatiotemporal decades (Sawyer et al. 2000a). The return interval for
dynamics of biomass, 3) estimate ANPP using several floods covering most of the study stand is <100 yr.
methods, and 4) estimate woody detritus volume, Flooding replenishes nutrients (Westman and Whit-
mass, input rate and decay rate. Implications of our taker 1975), deposits sediments, and influences
findings for the dynamics of old Sequoia ecosystems understory tree survival (Stone and Vasey 1968).
are discussed. Furthermore, we compare and contrast Establishment of Sequoia seedlings is common after
the fundamental ecosystem dynamics of our study major floods (Sawyer et al. 2000a). Tree-fall gaps are
forest to those of other forests worldwide. also an important form of disturbance in the study
forest. Although the return interval of gap disturbance
is at least 250 yr, individual gaps can exceed 0.1-ha in
Study forest area (Busing and Fujimori 2002). They currently
cover about 20% of the total area within the study
Sequoia sempervirens occurs within 50 km of the Pa- forest.
cific Ocean from southern Oregon to central
California. Particularly massive stands of Sequoia oc-
cur in the Central Redwood Forest, which ranges from Methods
southern Humboldt County to San Francisco Bay
(Sawyer et al. 2000b). An old stand of very large trees Biomass and production
on alluvial flats within this biogeographic section was
selected for study. It lies near Bull Creek in Humboldt In 1972, trees ‡10 cm diameter at 1.7 m above ground
Redwoods State Park (4021¢ N, 1230¢ W). Elevation were measured with a diameter tape and mapped in a
of the study site is 70 to 80 m above sea level. Climate, 1.44-ha plot (Fujimori 1977). Allometric equations for
geology and soils of the site are described by Fujimori tree height and stem volume were developed for the
(1977). Moderate temperatures and high moisture in- site. Equations for tree height as a function of diam-
put are characteristic of the forest. Precipitation is high eter were developed from a subsample of live trees
except in summer. Fog drip compensates for low measured using an Abney level. An equation for
summer precipitation (Byers 1953). Sequoia stem volume as a function of diameter and
Sequoia comprises more than 99% of the total live height was developed from a sample of 8 recently
basal area in the study stand (Busing and Fujimori fallen trees (see Appendix). In 2001, we measured
2002). Overstory tree height often exceeds 90 m diameter on all live trees within the plot and noted tree
(Fujimori 1977). Other tree species in the forest are ingrowth and mortality by species since 1972.
much shorter in stature. They include Umbellularia Using the 1972 stem map, diameter growth of
californica, Lithocarpus densiflorus, Taxus brevifolia, individual live trees over the 29-yr period was
and Corylus cornuta. determined. Mean annual diameter increments of
179
each species were calculated. To account for effects plying the stem NPP by 0.25 and by 0.30 (Westman
of tree size and canopy position on growth, Sequoia and Whittaker 1975). A range of twig and leaf NPP
diameter increments were calculated for each of the values was estimated as the combined stem and
three canopy strata recognized by Fujimori (1977). branch NPP multiplied by 0.3 and by 0.5.
These included a lower stratum (diameter <30 cm), a Separate biomass and production estimates were
middle stratum (diameter 30–114 cm), and an upper made using two different specific gravity values for
stratum (diameter >114 cm). Sequoia tree heights in Sequoia. Westman and Whittaker’s (1975) estimates
2001 were estimated from diameter using one of the were based on an intermediate value of 0.33. How-
stratum-specific allometric equations (Fujimori ever, Green et al. (1999) note that 0.38 is appropriate
1977). for old-growth Sequoia. Some of our dead wood
Estimation of tree ANPP followed the general for- samples (see below) also indicated that 0.38 best
mula summarized by Newbould (1967), Ogawa represented the study forest. So, for comparative
(1977) and Long (1982) as the total annual growth of: purposes we used 0.33, but we also used 0.38 and we
1) live trees, 2) trees or tree parts lost to mortality, and considered this to be a more accurate value for the
3) tree parts consumed. Tree parts shed or consumed study forest. Other species comprised a very small
were ignored, however. Stand volume increment since proportion of the stand biomass and NPP. Total
1972 was initially calculated with ingrowth and aboveground values for these species were estimated
mortality components. Whenever possible, the mea- directly from parabolic volume and specific gravities
sured diameter increment for each tree was used. If a of each species.
measured Sequoia increment was uncertain (because
of measurement location or tree identity), the mean
increment for the corresponding canopy stratum (see Woody detritus
above) was applied. Trees that died over the 29-yr
interval were assigned a diameter increment equal to Woody detritus volume in 2001 was estimated with
half the corresponding mean increment. Diameter plot measurements for standing dead trees (or snags)
increments of ingrowth trees were calculated using an and line transects for logs. Volume of standing dead
initial diameter of 10 cm. Height increments were trees (>1.7 m tall) within the 1.44-ha plot was cal-
calculated from diameter measurements using a stra- culated from measurements of diameter and height.
tum-specific allometric equation (Fujimori 1977). For The volumetric shape (cylindric, parabolic, or conic)
trees with a measured height value in 1972, the and stage of wood decay (stage 1–5) of each standing
increment was added to the measured height. Para- dead tree were noted. Down dead wood was sampled
bolic volume (see Appendix) was calculated for 1) with eight 100-m transects across the study forest. To
live trees in 1972, 2) live trees in 2001, and 3) live avoid transect orientation bias, four of the transects
trees in 2001 plus trees that died over the interval. ran roughly north to south and the other four ran east
Annual parabolic volume increment was then calcu- to west. The lateral distance between parallel transects
lated with and without compensation for mortality. exceeded 100 m. Fine dead wood (<10 cm diameter at
Stem volume of Sequoia was estimated using three intercept) was tallied along transect segments in size
methods (see Appendix): 1) the allometric relation classes. The three smallest size classes followed those
developed by Fujimori (1977) for the study site, 2) a of Brown (1974). Class 1 items were <0.7 cm dia-
second allometric relation developed using stem taper meter, class 2 items were 0.7 to 2.5 cm diameter, and
data on hundreds of old-growth trees in the region class 3 items were 2.6 to 7.6 cm diameter. A fourth
(Parks 1952), and 3) the estimation ratios of Westman fine wood class included items 7.7 to 9.9 cm diameter.
and Whittaker (1975). The latter method involved Items in classes 1 and 2 were tallied on one 7-m
multiplying the parabolic volume by 0.9. Stem mass segment per 100-m transect. Items in classes 3 and 4
estimates were obtained by multiplying the stem were tallied on one 14-m segment per 100-m transect.
volume by the specific gravity. Branch biomass of Coarse dead wood ( £ 10 cm diameter at intercept)
Sequoia was estimated as 0.12 of the stem mass was measured for diameter at point of intercept along
(Westman and Whittaker 1975). each 100-m transect, identified to species, and
Stem NPP of Sequoia was calculated as the annual assessed for stage of decay (stage 15; Sollins 1982).
stem volume increment times the specific gravity. A Specific gravity of Sequoia wood by stage of decay
range of branch NPP values was obtained by multi- was estimated with wood samples measured for
180
volume and dry weight. For each stage of decay a set tions (Figure 1). Both allometric relations generated
of samples, including representative proportions of mean values within one standard deviation (of the
sapwood and heartwood, was collected from the study estimate) of the mean measured value. By contrast,
site. Most items were cylindric and volume could be the parabolic estimates were much greater, overesti-
estimated from length and diameter. Other items were mating the measured value in all cases. Fujimori’s
submersed in water and the displaced volume was allometric relation overestimated the measured stem
measured. All samples were then oven dried at 55 C volume in 83% of the cases, while the second allo-
(to allow nutrient analysis) until constant dry weight metric relation underestimated the measured valued in
was obtained (ca. 2 weeks). A subset of these samples 67% of the cases. We considered our stem biomass
was subsequently dried at 105 C for two additional estimates based on the allometric relations (and a
days and the dry weight of all samples was adjusted specific gravity of 0.38) to be more accurate than any
using the mean percentage of weight lost at the higher of the parabolic estimates. These relatively accurate
temperature. estimates ranged from 3442 to 4143 Mg ha1.
Rate of coarse woody detritus input was determined Sequoia branch biomass estimates in 1972 ranged
from tree mortality in the 1.44-ha plot over a 29-yr from 359 to 616 Mg ha1 (Table 1, Table 2). Esti-
period. Stem volume of the dead individuals was mates of 413 to 497 Mg ha1 corresponded to our
estimated using the two allometric equations stem biomass estimates of 3442 to 4143 Mg ha1
described above. The parabolic approach of Westman above. Combining stem and branch compartments we
and Whittaker (1975) was not used in the estimation obtained total Sequoia aboveground biomass esti-
of dead tree volumes. mates of 3855 to 4640 Mg ha1. We considered these
Volume of fallen woody detritus was calculated estimates to be relatively accurate. It should be noted,
using Van Wagner’s (1968) formula (see Appendix). however, that our total Sequoia biomass estimates,
Mass of coarse detritus was estimated for each stage calculated using a variety of methods, had a broader
of decay using our specific gravities. Fine woody range (3348 to 5749 Mg ha1; Table 1, Table 2). Total
detritus calculations followed those of Brown (1974); stand biomass was only slightly greater, ranging from
a representative diameter was used for each size class 3350 to 5751 Mg ha1.
and a correction factor for intercept bias of nonhori- Sequoia stem biomass declined 7% over the 29-yr
zontal pieces was applied to the three smallest diam- period. Although biomass declined over the 1.44-ha
eter classes (see Appendix). A specific gravity of 0.33 area taken as a whole, some patches within the study
was used for all fine detritus mass estimates. plot increased in biomass. An examination of biomass
Some authors include dead material as a biomass change against grain size (the smallest area of obser-
component (e.g., Sprugel 1985). We retain the tradi- vation and measurement; sensu Wiens 1989) revealed
tional definition of biomass as the weight of material that patches exhibiting biomass increases were in the
in living trees (Ricklefs 1973) and refer to material 0.01 to 0.4 ha size range. All larger patches or col-
from dead trees and fallen parts as detritus. lections of patches decreased in biomass. Plotting
temporal variance in stem biomass against grain size
(e.g., Levin 1992) revealed high variance at grain
Results sizes less than 0.05 ha (Figure 2a). The log-log rela-
tionship was approximately linear (Figure 2b).
Biomass
Table 1. Comparison of biomass and production estimates for Sequoia sempervirens forests using a specific gravity of 0.33 for Sequoia.
Volume, biomass, basal area, density and height values are reported for the initial sampling date (1972 for the current study). Values in
parentheses include contributions of trees dying over the course of the 29-yr study. Means or ranges of estimates from previous studies are based
on Whittaker (1966), and Westman and Whittaker (1975). Allometric estimates 1 and 2 refer to stem volume equations based on Fujimori (1977)
and Parks (1952), respectively.
Table 1. Continued.
ANPP estimates. We consider our stem ANPP depending on the allometric relation used to estimate
estimates of 4 to 5 Mg ha1yr1 to be the most accurate stem volume (see methods). Annual input averaged
(Table 2). They are based on the allometric relations 5.7 to 6.9 Mg ha1 (Table 3). Assuming constant
and a specific gravity of 0.38. Also, they include con- input and steady state total mass, a decay rate constant
tributions of trees dying over the 29-yr study period. of 0.22 to 0.026 was obtained by dividing the annual
Ranges of Sequoia branch ANPP, estimated using the input by the total mass of coarse detritus (Jenny et al.
ratios of Westman and Whittaker (1975), were 1 to 2 Mg 1949; Olson 1963; Whittaker 1975; Sollins 1982;
ha1yr1 (Table 1, Table 2). Those for twig and leaf Onega and Eickmeier 1991). The total volume of fine
components were 1 to 5 Mg ha1yr1 (Table 1, Table 2). woody detritus on the forest floor was 16 m3 ha1
Total tree ANPP estimates ranged from 6 to 14 Mg (Table 3). Total mass in this compartment was esti-
ha1yr1 in our study (Table 1, Table 2). We consid- mated at 5 Mg ha1.
ered our estimates of 7 to 10 Mg ha1yr1 to be
comparatively accurate. They were based on the allo-
metric relations and a specific gravity of 0.38 (Table 2).
Discussion
The total volume of coarse woody detritus was 797 The slight decline (7%) in biomass during the three-
m3 ha1 (Table 3). The total mass was 262 Mg ha1. decade study supports other evidence that the stand is
Fallen wood comprised most of the coarse woody in an advanced stage of development. Rapid, stand-
detritus. The log component was 743 m3 ha1 (242 wide biomass accumulation typical of early devel-
Mg ha1). The snag component was 54 m3 ha1 (20 opment (Bormann and Likens 1979; Peet 1981) has
Mg ha1). Snag basal area and density were 12.7 m2 clearly ceased. The stand has reached the old growth
ha1 and 12 stems ha1, respectively. For comparison, stage or is in transition toward old growth (Oliver
live tree basal area and density in 2001 were 325 m2 1981; Peet and Christensen 1987). Large tree boles
ha1 and 170 stems ha1, respectively. (both live and dead), characteristic of old-growth
The input of coarse woody detritus in the study plot coniferous forest structure (Franklin et al. 1981), are
over 29 yr was 437 to 522 m3ha1 (166-199 Mg ha1), abundant. Trees greater than 1000 years of age are
183
Table 2. Biomass and production estimates for the study site using a specific gravity of 0.38 for Sequoia. Volume and biomass values are
reported for the initial sampling date (1972). Allometric estimates 1 and 2 refer to stem volume equations based on Fujimori (1977) and Parks
(1952), respectively. Values in parentheses include contributions of trees dying over the course of the 29-yr study. See Table 1 for stem volume
increments.
Parameter Value
not uncommon (Fujimori 1977), indicating that at has been tree-fall gaps created by trees without severe
least a millennium has passed since the last stand- fire injury. Our analysis of temporal variance in bio-
replacing disturbance. mass by grain size, showing greater variance at grain
Although pervasive disturbances such as fire and sizes less than 0.05 ha, shows the effect of scattered
flooding affect the dynamics of riparian Sequoia for- mortality of individual large trees. Recent canopy
ests (Stone and Vasey 1968;Veirs 1982; Agee 1993; gaps in our study plot are generally less than 0.05 ha;
Sawyer et al. 2000a), the predominant form of dis- however, a survey of the study area revealed gaps
turbance in our study forest over the last three decades approaching 0.2 ha in size (Busing and Fujimori
184
Figure 1. Comparison of calculated Sequoia stem volumes against measured volumes. Means and standard deviations for each method are
shown. Volume data in the ‘measured’ category are from Sawyer et al. (2000a). They consist of very large trees (87–110 m tall; n ¼ 12) within
the region that were measured in the field. All other categories displayed are calculated volume estimates applying different formulas based on
tree diameter and height. The various formulas range from simple parabolic volume estimates (Westman and Whittaker 1975) to allometric
volume estimates developed with field data and regression analysis (Fujimori 1977).
2002). Further analysis of biomass variance by grain lometric equations we use to estimate stem volume
size reveals a log-log linear relationship. Levin (1989) give lower values. Even our estimates using a larger
states that departure from a linear relationship indi- Sequoia specific gravity and including ANPP of trees
cates ‘...operation of distinct mechanisms on different lost to mortality tend to be lower than those of
scales.’ For biomass dynamics, it appears that distinct Westman and Whittaker (1975) when allometric stem
mechanisms of this sort have not operated during the volumes are used.
course of this study. We consider our ANPP estimates of 7 to 10 Mg
A general conclusion from the analysis of biomass ha1yr1 to be the most accurate. They are based on
dynamics is that small-patch dynamics are significant. allometric stem volume and a specific gravity of 0.38.
A shifting-mosaic of small patches influenced by tree They also include contributions of trees dying over
falls is characteristic of many old-growth forests the course of the study. These ANPP values are
(Bormann and Likens 1979; Pickett and White 1985; slightly below average for old temperate forests
Busing 1998; Fujimori 2001), including those domi- (Whittaker 1966; Fujimori 2001), but similar to those
nated by conifers (McCarthy 2001). Our short glimpse reported for other old coniferous forests of the Pacific
at biomass dynamics within a stand suggests that this Slope of North America (Long 1982). Our estimates
type of small patch dynamics can play a substantial do support the idea that old Sequoia forests are not
role in old-growth Sequoia forests as well. highly productive (Whittaker 1966; Westman and
Whittaker 1975). Even though biomass exceeds that
of most other temperate forests by a factor of ten or
Production more, our ANPP estimates are not high.
The ratio of biomass to annual NPP is very large
Our ANPP estimates of 6 to 14 Mg ha1yr1 fall (ca. 400) for our study forest. Whittaker (1975)
within the range of values reported for other temper- termed this the biomass accumulation ratio. It typi-
ate forests. Our estimates tend to be less than the cally ranges from 20 to 50 in mature forests. The
Sequoia forest ANPP means from Westman and extreme biomass accumulation ratio value from our
Whittaker (1975). One possible explanation is simply study is a result of the exceptionally large biomass, as
that our study stand has slower growth. Indeed, our ANPP is moderate.
mean radial increments were slightly lower. However, The fact that our study stand is old and strongly
the primary reason for the discrepancy is that the al- dominated by Sequoia deserves consideration in
185
Table 3. Volume and mass of woody detritus at the study site. The
minimum diameter of coarse woody detritus is 10 cm. Decay classes
range from undecayed (class 1) to well decayed (class 5).
Woody detritus
(eds), Primary productivity of Japanese forests. University of wood forest: history, ecology and conservation of the coast red-
Tokyo Press, Tokyo, Japan, pp. 29–37. woods. Island Press, Washington, D.C., USA, pp. 3979.
Oliver C.D. 1981. Forest development in North America following Sollins P. 1982. Input and decay of coarse woody debris in conif-
major disturbances. Forest Ecology and Management 3: erous stands in western Oregon and Washington. Canadian
153168. Journal of Forest Research 12: 1828.
Olson J.S. 1963. Energy storage and the balance of producers and Sprugel D.G. 1985. Natural disturbance and ecosystem energetics.
decomposers in ecological systems. Ecology 44: 322331. In: Pickett S.T.A. and White P.S. (eds), The ecology of natural
Olson D.F., Roy D.F. and Walters G.A. 1990. Sequoia sempervirens disturbance and patch dynamics. Academic Press, New York,
(D. Don) Endl. In: Barnes R.M. and Honkala B.H. (eds), Silvics USA, pp. 335352.
of North America. Agriculture Handbook 654, Vol. 1, Conifers, Stokland J.N. 2001. The coarse woody debris profile: an archive of
USDA Forest Service, Washington, D.C., USA, pp. 541551. recent forest history and an important biodiversity indicator.
Onega T.L. and Eickmeier W.G. 1991. Woody detritus inputs and Ecological Bulletins 49: 7183.
decomposition kinetics in a southern temperate deciduous forest. Stone E.C. and Vasey R.B. 1968. Preservation of coast redwoods on
Bulletin of the Torrey Botanical Club 118: 5257. alluvial flats. Science 15: 157161.
Parks W.H. 1952. Redwood log characteristics: log taper and bark Stuart J.D. 1987. Fire history of an old-growth forest of Sequoia
thickness. Technical Report. California Redwood Association, sempervirens (Taxodiaceae) forest in Humboldt Redwoods State
San Francisco, California, USA. Park, California. Madrono 34: 128141.
Peet R.K and Christensen N.L. 1987. Competition and tree death. Van Wagner C.E. 1968. The line intersect method in forest fuel
BioScience 37: 586594. sampling. Forest Science 14: 2026.
Peet R.K. 1981. Changes in biomass and production during sec- Veirs S.D. 1982. Coast redwood forest: stand dynamics, succes-
ondary succession. In: West D.C., Shugart H.H. and Botkin D.B. sional status, and the role of fire. In: Means J.E. (ed.), Forest
(eds), Forest succession: concepts and application. Springer- succession and stand development research in the Pacific
Verlag, New York, USA, pp. 324338. Northwest. Oregon State University Forest Research Lab, Cor-
Pickett S.T.A. and White P.S. 1985. The ecology of natural distur- vallis, Oregon, USA, pp. 119141.
bance and patch dynamics. Academic Press, New York, USA. Westman W.E. and Whittaker R.H. 1975. The pygmy forest region
Ricklefs R.E. 1973. Ecology. Chiron Press, Newton, Massachusetts, of northern California: studies on biomass and primary produc-
USA. tivity. Journal of Ecology 63: 493–520.
Sawyer J.O., Sillett S.C., Libby W.J., Dawson T.E., Popenoe Whittaker R.H. 1966. Forest dimensions and production in the
J.H., Largent D.L., Van Pelt R., Veirs S.D., Noss R.F., Great Smoky Mountains. Ecology 47: 103121.
Thornburgh D.A. and Del Tredici P. 2000. Redwood trees, Whittaker R.H. 1975. Communities and ecosystems. Macmillan,
communities, and ecosystems: a closer look. In: Noss R.F. New York, USA.
(ed.), The redwood forest: history, ecology and conservation of Whittaker R.H. and Likens G.E. 1975. The biosphere and man. In:
the coast redwoods. Island Press, Washington, D.C., USA, pp. Lieth H. and Whittaker R.H. (eds), Primary productivity in the
81118. biosphere. Springer-Verlag, New York, USA, pp. 305328.
Sawyer J.O., Sillett S.C., Popenoe J.H., LaBanca A., Sholars T., Wiens J.A. 1989. Spatial scaling in ecology. Functional Ecology 3:
Largent D.L., Euphrat F., Noss R.F., and Van Pelt R. 2000. 385–397.
Characteristics of redwood forests. In: Noss R.F. (ed.), The red-