Journal of Vegetation Science 18: 879-890, 2007
© IAVS; Opulus Press Uppsala.
- Fire disturbance and forest structure in old-growth mixed conifer forests -
879
Fire disturbance and forest structure in old-growth mixed conifer
forests in the northern Sierra Nevada, California
Beaty, R. Matthew1* & Taylor, Alan H.2
1CSIRO Sustainable Ecosystems, GPO Box 284, Canberra, ACT 2601, Australia;
Department, Pennsylvania State University, University Park, PA 16802, USA; Email [email protected];
*Corresponding author; E-mail [email protected]; Tel. +62 262421590; Fax +61 262421728
2Geography
Abstract
Question: This study evaluates how fire regimes influence
stand structure and dynamics in old-growth mixed conifer
forests across a range of environmental settings.
Location: A 2000-ha area of mixed conifer forest on the
west shore of Lake Tahoe in the northern Sierra Nevada,
California.
Methods: We quantified the age, size, and spatial structure
of trees in 12 mixed conifer stands distributed across major
topographic gradients. Fire history was reconstructed in each
stand using fire scar dendrochronology. The influence of fire
on stand structure was assessed by comparing the fire history
with the age, size, and spatial structure of trees in a stand.
Results: There was significant variation in species composition
among stands, but not in the size, age and spatial patterning of
trees. Stands had multiple size and age classes with clusters
of similar aged trees occurring at scales of 113 - 254 m2. The
frequency and severity of fires was also similar, and stands
burned with low to moderate severity in the dormant season on
average every 9 - 17 years. Most fires were not synchronized
among stands except in very dry years. No fires have burned
since ca. 1880.
Conclusions: Fire and forest structure interact to perpetuate
similar stand characteristics across a range of environmental
settings. Fire occurrence is controlled primarily by spatial
variation in fuel mosaics (e.g. patterns of abundance, fuel
moisture, forest structure), but regional drought synchronizes
fire in some years. Fire exclusion over the last 120 years has
caused compositional and structural shifts in these mixed
conifer forests.
Keywords: Drought; Fire, Fire return interval; Forest structure;
Lake Tahoe; Spatial autocorrelation; Stand dynamics.
Nomenclature: Hickman (1993).
Abbreviations: FRI = Fire return interval; GCW = General
Creek Watershed; TRMI = Topographic Relative Moisture
Index.
Introduction
The structure and dynamics of western conifer forests
in North America are linked to recurring fire (Whitlock
et al. 2003; Skinner & Chang 1996), and variation in
fire regime characteristics are thought to promote forest
structural and compositional diversity at local, landscape,
and even regional scales (Romme & Knight 1981; Agee
1993; Bekker & Taylor 2001). The importance of fire as a
control on the structure and dynamics of western forests,
however, has changed since widespread EuroAmerican
settlement in the mid-19th Century, especially in the most
fire-prone forests. For example, in the mixed conifer
forests of California, suppression of wildfires in combination with logging and livestock grazing have dramatically altered forest composition and structure (Vankat
& Major 1978; Weatherspoon et al. 1992; Taylor 2000,
2004). These ubiquitous forest changes have limited our
understanding of how spatial and temporal variability in
fire regimes shaped mixed conifer forest structure and
influence their dynamics (Skinner & Chang 1996; Beaty
& Taylor 2001; Stephens & Collins 2004).
Prior to widespread EuroAmerican settlement, fire
was the most common and widespread disturbance agent
influencing the structural development of mixed conifer
forests in the Sierra Nevada (Kilgore 1973; Skinner
& Chang 1996). Frequent low intensity surface fires
– e.g. median fire return interval (MFI) = 3 - 20 a – are
thought to create a fine-grained mosaic of multi-aged
stands (Kilgore & Taylor 1979; Swetnam 1993; Skinner & Chang 1996). This structure is thought to have
developed as a result of self-organizing processes where
burn patterns and forest structure interact to maintain
the fine-grained mosaic over time (e.g. Bonnicksen &
Stone 1981, 1982). Thus, fire regimes and forest structure
are thought to be controlled by local, spatial processes,
particularly the time it takes for fuel to accumulate so a
burned patch can burn again. Yet, recent work in mixed
conifer forests suggests that interactions between fire
and forest structure may be more complex because fire
regimes and fire effects are influenced by landscape and
880
Beaty, R.M. & Taylor, A.H.
regional controls such as topography (e.g. Taylor & Skinner 1998, 2003; Taylor 2000; Beaty & Taylor 2001) or
climate variability (e.g. Swetnam 1993; Miller & Urban
1999a, 2000a, 2000b). Thus, variability in the structure
of mixed conifer forests may reflect the influence of
landscape and regional factors rather than smaller scale
self-organizing processes.
Yet, the relative importance of local vs. landscape
or regional scale influences has not been evaluated in
mixed conifer forests, and most retrospective studies
have not been designed to address this question. Most
work on mixed conifer forest structure composite data
collected over wide areas, but composite data on forest
structure do not provide an understanding of how fire
and the temporal and spatial arrangement of age-classes
develop in a particular stand. Moreover, few studies
have coupled an analysis of forest age structure, including spatial pattern, with fire disturbance history in the
same stand. In mixed conifer forests, our understanding
of the relationship between fire and forest structure is
based primarily on data (1) from giant sequoia – mixed
conifer forests, an unusual variant of limited extent that
occurs mainly in the southern Sierra Nevada, (2) from a
narrow range of environmental settings, and (3) without
records of fire disturbance and forest structure from the
same sites (e.g. Bonnicksen & Stone 1981, 1982). Thus,
it is unclear, if more typical mixed conifer forests have
similar fire-generated structures and dynamics.
This paper examines how fire regimes influence stand
structure and dynamics in an old-growth mixed conifer
forest in the northern Sierra Nevada on the west shore of
Fig. 1. Location of study area and sample plots (n = 12), Lake
Tahoe Basin, California.
Lake Tahoe. The following questions are addressed:
1. How variable are fire regime parameters and stand
structural characteristics in mixed conifer forests
across a range of environmental settings?
2. How did past fires influence the spatial and temporal
patterns of tree regeneration?
To help answer these questions, we quantified the fire
history and the age, size, and spatial pattern of trees in
twelve old-growth mixed conifer stands located across
a range of environmental settings.
Methods
Study area
Fire regimes and forest structure were studied in a
2000 ha area of old-growth mixed conifer forest in the
General Creek Watershed (GCW) on the west shore of
Lake Tahoe, California (Fig. 1). General Creek drains
a formerly glaciated valley and has diverse topography. Elevations range from 1850 to 3000 m and the
topographic settings range from flat valley bottoms to
moderately steep (ca. 20°) side slopes. Climate is characterized by warm dry summers and cold wet winters:
most (80%) precipitation falls as snow during the winter.
Mean monthly temperatures at Tahoe City (15 km north)
range from –2° in January to 16° C in August, and mean
annual precipitation is 784 mm. Mixed conifer stands in
the GCW may be co-dominated by any of six species
including Calocedrus decurrens, Pinus lambertiana,
Pinus jeffreyi, Abies magnifica, Pinus monticola or Abies
concolor (Barbour & Minnich 2000). Montane chaparral,
dominated by Arctostaphylos patula, Ceanothus spp. and
Quercus spp. (dwarf oaks), may occupy sites that have
experienced severe fire or are too poor to support trees
(Wilken 1967; Bolsinger 1989; Nagel & Taylor 2005).
Forests in the Lake Tahoe basin have been influenced
by people for a long time. The Washoe, who migrated
annually from the Great Basin each summer, used the
lake and surrounding forests for hunting, and they used
fire locally (e.g. meadows and forests near Lake Tahoe)
to promote certain plants for food and fiber (Lindström
2000). EuroAmericans first settled in the Lake Tahoe
basin in large numbers in the 1860s when silver was
discovered in nearby Virginia City, Nevada. Beginning
in the 1870s, nearly 70% of the Lake Tahoe watershed
was logged to provide fuel wood for silver mining (Elliot-Fisk et al. 1996). Most forest in the basin is second
growth and is about 120 years old (Lindström 2000;
Taylor 2004). The GCW became part of the California
State Park System in 1965, and there has been an active
prescribed burning program in the park since the 1980s.
None of the stands chosen for this study were logged or
- Fire disturbance and forest structure in old-growth mixed conifer forests -
Fire history
The history of fire (i.e. dates, return interval, extent,
season) in each of the 12 plots was reconstructed from fire
scars preserved in the annual growth rings of live and dead
pines (n = 34) collected in or adjacent to (≤ 100 m) each
plot. Partial cross-sections were removed with a chainsaw
(Arno & Sneck 1977) and sample locations were identified with a GPS and plotted on a 1:24 000 topographic
map. Fire dates in the cross-sections were identified by
first sanding wood samples to a high polish and then
cross-dating (Stokes & Smiley 1968) the tree rings in
each sample. The calendar year of each tree-ring with a
fire scar lesion in it was then recorded as the fire date.
The season each fire burned was inferred from the
position of the fire scar within the annual growth ring
(e.g. Baisan & Swetnam 1990). Scar positions were
classified as either growing season (i.e., occurring in
early- or late-wood) or dormant season (i.e., occurring
at the ring boundary). In the northern Sierra Nevada, the
growing season is between May and August, and dormant
season fires represent fires that burn in late summer or
fall after trees stop growing for the year (e.g. Caprio &
Swetnam 1995).
Spatial and temporal variation in the frequency and
extent of fire based on the fire history data was evaluated
using standard statistical (i.e., S-Plus) and specialized
fire history software (i.e., FHX2, Grissino-Mayer 2001).
Fire return interval (FRI) statistics were computed for
both composite and point fire chronologies at each site.
GCW-1
GCW-2
GCW-3
GCW-4
GCW-5
GCW-6
GCW-7
GCW-8
GCW-9
GCW-10
GCW-11
GCW-12
2039
2109
2078
1975
2027
1948
2018
2100
2043
2000
1999
2073
6
21
32
1
17
2
14
21
6
1
2
32
166
148
310
172
166
320
114
147
180
350
10
350
SE
SE
NW
SE
SE
NW
E
SE
SE
N
NW
N
14
24
22
38
32
45
31
27
35
43
54
26
Plot size
(ha)
TRMI
Aspect
(direction)
Plot
Aspect
(degrees)
Old-growth mixed conifer stands were selected for
sampling based on several factors, including: (1) presence of large diameter (>1.0 m diameter at breast height
[DBH]) stems ≥ 250 a old); (2) homogeneity of stand
conditions; (3) sampled stands included the range of
variation in old-growth mixed conifer forest structure and
composition in the watershed; (4) a range of topographic
settings; and (5) the presence of fire scarred trees in or
near each stand. Stands meeting these criteria (n = 12)
were selected for sampling (Fig. 1; Table 1).
Forests were sampled in 0.5 ha (50 m × 100 m) plots,
except for two plots which were larger (0.7 ha, 70 m ×
100 m) because of low stand density. The location (UTM
coordinates with a GPS), elevation, slope aspect, slope
pitch, slope configuration, and topographic position of
each plot were recorded. The last four variables were used
to calculate each plot’s Topographic Relative Moisture
Index (TRMI; Parker 1982) a measure of site moisture
conditions that varies from 0 (xeric) to 60 (mesic).
Slope pitch
(degrees)
Field sampling and sample processing
Table 1. Characteristics of old growth mixed conifer forests
in 12 plots in the Lake Tahoe Basin, California. Values for the
Topographic Relative Moisture Index (TRMI; Parker 1982)
range from 0 (xeric) to 60 (mesic).
Elevation (m)
burned with prescribed fire.
881
0.5
0.5
0.7
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.7
0.5
Differences in FRI statistics and the shape of the FRI
distributions among stands were assessed with distribution free statistical tests (e.g. Kruskal-Wallis H test,
Kolmogorov-Smirnov two-sample tests; Sokal & Rohlf
1995). The number of plots burned in a given year was
used as an indirect measure of fire extent. Landscape scale
FRI statistics were calculated from CFRIs from each plot
for years with: (1) small fires (one or more plots burned),
(2) intermediate fires (five or more plots burned), and (3)
widespread fires (eight or more plots burned).
Temporal changes in fire regimes in the study area
that may be related to land use change were identified by
comparing fire frequency during the pre-settlement (1700
- 1850), settlement (1850 -1900), and fire suppression
(1900 - 2000) periods using a composite of all the fire
scar samples. A composite fire chronology was used for
the temporal analysis because they are considered to be
more sensitive than individual samples to changes in the
temporal pattern of burning related to land use change
(Dieterich 1980).
Forest structure and composition
Forest structure and composition were determined by
measuring and mapping stems within a 10 m x 10 m grid
in each plot. Stem locations in a cell were determined
by measuring the distance (x, y) of all live and standing
dead trees (≥ 5.0 cm DBH), live saplings (stems >1.4
m tall < 5.0 cm DBH), and live seedlings (0.5 - 1.4 m
tall) from the origin (0, 0) of each cell with a measuring tape. The species, diameter at breast height (DBH),
height class (i.e. suppressed, intermediate, co-dominant,
dominant) and condition (live or dead) of each tree was
882
Beaty, R.M. & Taylor, A.H.
then recorded. The location (x, y), diameter, height class,
and species of logs ≥ 5 cm rooted in each cell were also
recorded. Standing dead trees and logs were visually
classified as either pre- or post-fire suppression recruits
based on diameter, species, bark morphology, and level
of decomposition.
Variation in species composition among the twelve
plots was identified by ordinating the basal area of species in each plot with principal components analysis
(PCA; Gauch 1982). Compositional gradients were then
identified by correlating PCA axis scores and species’
basal area values using Pearson’s correlation coefficient
(Sokol & Rohlf 1995). Topographic factors underlying
the compositional gradients were also identified by correlating basal areas with topographic data collected at each
site. Finally, fire regime parameters for each plot were
correlated with PCA axis scores to identify whether fire
regime parameters varied with species composition.
Tree age structure in each plot was determined by
coring all trees to the pith at 30 cm above the soil surface
in randomly selected 10 m × 10 m cells until ≥ 20% of all
trees had been cored. Then, all trees > 40 cm DBH in each
plot were cored to ensure that all stems that established
prior to the last fire were included in the sample. This
cutoff was selected based on diameter – age relationships
identified from preliminary sampling of tree ages across
a range of size classes and was used to reduce redundant
sampling of fast growing Abies concolor that established
after fire suppression. The cores were cross-dated (Stokes
& Smiley 1968) and the year of the innermost tree ring
in each core was used as the estimate of tree age. Some
cores did not reach the pith (9.8%) because of heart-rot
or their radius exceeded the length of increment borers.
The amount of missing growth was estimated using
DBH - radius regression equations for each species which
were all highly significant (r2 = 0.80 – 0.88; P < 0.001).
The number of years added to incomplete cores was the
product of the average annual radial growth/year of the
innermost portion of the intact core and the estimated
missing distance to the pith.
In order to identify differences in age structures
among stands, we compared the age-class distributions
of each plot using Kolmogorov-Smirnov two sample tests
(Sokol & Rolf 1995) and calculated the percentage of
stands with similar and different tree age distributions.
We also calculated the mean number of stems > 100 a
old and counted the number of occupied 20-a age classes
in a given plot. This analysis also provides evidence of
past fire severity as stands experiencing high severity fire
would have fewer ages classes than those that experience
low and intermediate severity burns (Agee 1993).
To identify the influence of past fires on the spatial
and temporal patterns of tree regeneration, we identified
the type, scale, and intensity of the spatial pattern of tree
ages using statistical tests of spatial auto­correlation. For
each plot, we evaluated the spatial characteristics of the
ages of large diameter (> 40 cm DBH) trees using Moran’s I (Moran 1948, Cliff & Ord 1973). Moran’s I is a
weighted correlation coefficient that detects departures
from spatial randomness. Positive values of Moran’s I
indicate that trees with similar ages occur close together
whereas negative values indicate that trees with similar
ages are separate from each other. Values of Moran’s I
were calculated for successive 3-m distance classes up
to 24 m to identify the scale of spatial autocorrelation;
significance was identified using two tailed tests (Upton
& Fingleton 1985). The frequency of significant Moran’s
I values for each distance class was then summarized for
all plots. Finally, the ages of large trees (> 40 cm DBH)
were mapped to visually interpret and confirm the spatial
autocorrelation analysis.
Results
Forest compositional patterns
Species dominance in the 12 plots varied with environmental settings. PCA of species basal area separated
stands along two compositional gradients (Fig. 2). Abies
magnifica and Pinus monticola were positively correlated
(P < 0.01) and A. concolor, Calocedrus decurrens, and
P. jeffreyi were negatively correlated with PCA axis 1
(P < 0.05; Table 2). Axis 1 was positively correlated (P
< 0.01) with slope aspect (expressed in linear form from
south (low values) to north (high values)) and negatively
correlated (P < 0.01) with slope configuration (expressed
in linear form from convex (low values) to concave
(high values)). PCA axis 2 separated plots dominated
by P. jeffreyi (positive correlation but not significant P
> 0.05) from mixed stands with abundant C. decurrens,
P. lambert­iana and P. contorta (negative correlation P
< 0.05; Fig. 2; Table 2). PCA Axis 2 scores were not
significantly correlated with any of the environmental
variables (Table 2). Composite mean and median FRIs
were positively correlated (P < 0.05) with PCA axis 1
indicating a shortening of the period between fires in
mixed stands on south slopes (Table 2).
- Fire disturbance and forest structure in old-growth mixed conifer forests -
883
Table 2. Pearson correlation coefficients of species basal area,
topography and fire history with PCA axis scores of an ordination of species basal area in 12 plots of old-growth mixed
conifer forests in the Lake Tahoe Basin (* p < 0.05, ** p <
0.01) Slope aspect and configuration were transformed into
linear values (see Parker 1982).
Axis 1
Axis 2
Species composition
Abies concolor
– 0.62 *
– 0.17
Abies magnifica
0.78 **
– 0.25
Calocedrus decurrens
– 0.52 *
– 0.43
Pinus contorta
– 0.16
– 0.52 *
Pinus jeffreyi
– 0.72 **
0.44
Pinus lambertiana – 0.35
– 0.76 **
Pinus
monticola
0.68 **
– 0.26
Topography
Elevation
– 0.41
0.09
Slope aspect
0.72 **
– 0.33
Slope steepness
0.18
0.12
Slope position
0.04
– 0.11
Slope configuration
– 0.73 **
– 0.22
TRMI
0.33
– 0.24
Fire history
Number of fires
– 0.38
– 0.20
Mean point fire return interval
0.33
– 0.01
Median point fire return interval
0.28
– 0.05
Mean composite fire return interval
0.48 *
– 0.10
Median composite fire return interval
0.55 *
– 0.10
Fig. 2. Principal components analysis of species basal areas
for 12 plots of old growth mixed conifer forest, Lake Tahoe
Basin, California.
Table 3. Basal area (m2.ha–1) and density (stems/ha) of live trees > 5.0 cm DBH in 12 plots of old-growth mixed conifer forest in
the Lake Tahoe Basin, California.
Plot
Abies concolor Abies magnifica
Calocedrus
decurrens
Basal area (m2.ha–1)
Pinus
contorta
Pinus jeffreyi
Pinus lambertiana
Pinus
monticola Total
GCW-1
32.7
0.7
24.5
GCW-2
24.2
29.2
21.3
GCW-3
10.5
3.6
9.8
7.0
GCW-4
7.5
0.1
26.5
GCW-5
54.9
0.0
21.2
1.4
20.8
15.9
GCW-6
4.5
28.4
0.2
14.1
3.0
GCW-7
41.2
0.1
10.2
9.7
GCW-8
60.7
1.2
46.1
GCW-9
46.4
22.1
GCW-10
32.2
24.2
1.2
4.5
6.4
GCW-11
13.1
0.7
2.1
9.3
GCW-12
26.4
12.0
6.8
0.8
8.8
Mean
29.5
8.7
20.2
1.0
17.9
6.7
7.6
Density (stems/ha)
57.8
74.8
30.9
34.0
114.3
50.2
61.2
107.9
68.5
68.4
25.2
54.7
62.3
GCW-1
GCW-2
GCW-3
GCW-4
GCW-5
GCW-6
GCW-7
GCW-8
GCW-9
GCW-10
GCW-11
GCW-12
966
20
454
68
96
29
132
2
1072
2
64
16
120
184
4
740
2
16
366
2
824
506
430
28
113
14
50
226
90
96
38
30
14
102
54
12
38
6
16
88
2
78
4
28
36
14
4
46
1084
566
169
236
1234
352
776
458
926
996
213
380
Mean
468
50
616
96
49
17
9
25
884
Beaty, R.M. & Taylor, A.H.
Fig. 3. Average (± SE) size and age distributions for
plots (n = 12) of old-growth mixed conifer forest, Lake
Tahoe Basin, California.
Forest structure
of old trees (> 200 a old). In all plots, there was a large
pulse of A. concolor establishment that began about 120
a ago (Fig. 3). Calocedrus decurrens and Pinus jeffreyi
were present in a wide range of size and age classes,
including stems >155 cm DBH and > 400 a old. Most
P. jeffreyi and C. decurrens were > 100 a old, suggesting
continuous but low recruitment at least until recently.
Pinus lambertiana and P. monticola were present at low
densities across a range of size and age classes. Dead
trees were abundant and on average, comprised 20%
of standing trees. Of the dead stems, far more were A.
concolor (60%) than P. jeffreyi (18%), C. decurrens (2%),
or A. magnifica (2%).
Live tree density and basal area in the plots averaged
616 stems/ha and 62.3 m2.ha–1, respectively. However,
both of these stand characteristics varied among species
and plots (Table 3), except for Abies concolor. A. concolor density was at least five-fold greater than any other
species. Comparisons of the age and size distributions
of each plot to every other plot (n = 66) indicated that,
overall, trees had similar size and age structures (74%
of comparisons, Kolmorgorov-Smirnov two sample
tests, P > 0.05). However, the form of the age and size
distributions varied by species (Fig. 3). A. concolor and
A. magnifica had reverse-j size-class distributions but
uni-modal age structures. Most A. magnifica stems were
< 120 a old, while A. magnifica had a larger proportion
Table 4. Spatial autocorrelation of tree ages of stems > 40 cm DBH calculated using Moran’s I for old-growth mixed conifer forests
in the Lake Tahoe Basin, California. Data are the number of plots (n = 12) exhibiting significant (P < 0.05) spatial autocorrelation
at a given scale.
3
6
9
Positive autocorrelation
Negative autocorrelation
2
0
5
1
8
0
Distance class (radius, m)
12
15
5
0
3
0
18
21
24
3
0
3
0
2
3
- Fire disturbance and forest structure in old-growth mixed conifer forests -
885
Forest spatial patterns
The spatial pattern of large tree (> 40 cm DBH) ages
were similar in the 12 plots (Table 4). In most stands
(83%), there was a pattern of positive spatial autocorrelation (P < 0.05) at distance classes of 6-12 m; three
plots (GCW-2, 4, 10) also exhibited negative spatial autocorrelation (P < 0.05) at a distance class of 24 m. This
indicates that large trees occur in small groups of similarly
aged trees with some overlap in tree ages among groups
of trees. This pattern is confirmed on three representative
stem maps from the plots (Fig. 4).
Fire history
Fire record
A total of 34 fire scarred wood samples were collected
in the 12 plots. The average fire record length for a given
plot was 189 a (range, 101 - 277 a), and all stands last
burned in the mid to late 19th century (Table 5).
Fire return intervals
The median composite FRI for all plots was 3 a (range
1 - 20 a), and median composite FRIs for individual
plots ranged from 7.5 to 17 a (Table 5). Mean point FRIs
were longer and ranged from 10 to 43 a (Table 5). The
12 plots had similar mean and median composite FRI
(Kruskal-Wallis H-test, P > 0.05) except for GCW-11,
which had a longer mean FRI (Kruskal-Wallis H-test, P
< 0.05). The plots also had similar fire interval distributions (Kolmogorov-Smirnoff test, P > 0.05) except for
GCW-11.
Fig. 4. Spatial distribution of large tree (> 40 cm DBH) ages in
three old-growth mixed conifer forest plots in the Lake Tahoe
Basin, California. Plot dimensions are 50 m by 100 m. Values
with lighter font represent estimated tree ages.
Table 5. Fire return interval (FRI, yr) statistics and forest structural characteristics in 12 stands of old-growth mixed conifer forests,
Lake Tahoe Basin, California. There were no differences in mean composite FRI among stands (p > 0.05, Kruskal-Wallis H-Test).
Point Composite Mean density Number of
Fire Return Fire Return of stems/ha
20-a age
Intervals
Intervals
classes
Nr of First Last Nr of All Stems All Stems Site
samples fire
fire
fires
Mean Median Mean Median stems > 100 a stems > 100 a
GCW-1
GCW-2
GCW-3
GCW-4
GCW-5
GCW-6
GCW-7
GCW-8
GCW-9
GCW-10
GCW-11
GCW-12
3
2
3
3
3
3
5
2
3
2
3
3
1693
1775
1774
1616
1616
1655
1694
1765
1728
1629
1728
1616
1873
1889
1875
1893
1889
1879
1887
1887
1893
1867
1879
1841
21
13
10
14
26
23
14
13
20
13
9
7
20.3
15.9
23.2
27.1
18.7
13.9
18.3
16.3
16.7
17.0
33.7
35.4
16
16
20
25.5
16
10
18.5
13
14
13
43
39
9.1
8.8
10.1
17
10.5
9.7
11.4
9.4
8.3
14.4
16.8
15.4
8
8
8
12.5
8
10
8
10
7.5
13
17
15
157.3
153.9
171.4
224.0
144.4
206.3
113.3
131.0
174.5
142.1
181.9
191.9
99.1
106.2
121.3
152.1
108.2
160.9
95.7
127.5
104.8
122.5
127.6
137.0
13
15
13
17
15
23
6
8
9
17
15
19
9
11
9
14
12
19
3
6
6
13
11
15
Fires/age
class
2.3
1.2
1.1
1.0
2.2
1.2
4.7
2.2
3.3
1.0
0.8
0.5
886
Beaty, R.M. & Taylor, A.H.
Fig. 5. Number of plots (n = 12) burned between 1700 and 1900 in old-growth mixed conifer forest, Lake Tahoe Basin, California.
Fire season
The position of fire scars within annual growth rings
was determined for 50% of the 279 scars. Scars were
mainly (75.2%) at ring boundaries (dormant season)
but scars were present in early-wood (10.6%) and latewood (14.2%) indicating that some fires burned during
the growing season.
Fire extent
Fire frequency and extent were inversely related over
the period of record (Fig. 5). The median FRI for widespread fires was 32 a (range 13-36 a) which was longer
than the median for intermediate (median = 13.5 a, range
7-23 a) or local fires (median = 3 a, range 1-9 a).
Temporal changes
Mean composite FRIs did not vary between the preEuroAmerican (1700-1849; n = 39; MFI = 3.7 a) and
EuroAmerican settlement (1850-1900; n = 14; MFI = 3.0
a) periods (t-test, P > 0.05). However, no fires burned
in the sampled stands during the fire suppression period
(1900-2000).
Forest age structure and fire regimes
The age structure and temporal patterns of regeneration in the plots was similar. All possible comparisons
of tree age-class distributions indicate that most species
and plots had a similar age structure (mean = 74%; range
9–100%; Kolmogorov-Smirnov test, P > 0.05). Moreover, trees occupied a large numbers of 20-a age classes
and on average there was 14 (range 6-23) occupied
age-classes for all stems, and 11 (range 3-15) occupied
age-classes for stems > 100 a old (Table 5). There were
no large age-class peaks in any of the plots, or across
plots, that were coincident with fire dates.
Discussion
Variation in stand structure and fire regimes
Our first research question was how variable were fire
regimes and stand structure across a range of environmental settings? There was considerable variation in species
composition across sites related to local topographic
conditions. Species composition varied from mainly pinedominated stands on south facing slopes to A. magnifica
dominated stands on north aspects. Pinus jeffreyi, and
especially P. contorta, occupied harsh valley bottom sites
that experience cold air drainage (e.g. Parker 1986). The
species-environment associations in GCW are consistent
with those of other mixed conifer forests in the Sierra Nevada (e.g. Vankat 1982; Beaty & Taylor 2001).
Although our mixed conifer stands were compo­
sitionally variable they had similar structural characteristics (i.e., size, age, spatial pattern). While we found
significant variation in stand density and basal area,
values among stands overlapped and there are broadly
similar reported values in other California mixed conifer
forests (e.g. Bonnicksen & Stone 1981, 1982; Taylor
2000; Beaty & Taylor 2001; Taylor & Skinner 2003;
North et al. 2004). Despite this variation, the stand age
and size structure were similar among stands. All stands
had populations of trees that occupied a wide range of
size and age classes and they had abundant populations
of young A. concolor that established after the last fire
in the 19th century. A mixed species overstory of large
diameter older trees and abundant young A. concolor in
the understory and mid-canopy is characteristic of mixed
conifer forests that have experienced a century or more
of fire exclusion (e.g. Bonnicksen & Stone 1981, 1982;
Taylor 2000; Beaty & Taylor 2001; Taylor & Skinner
2003; North et al. 2004).
The spatial patterning of trees was similar among
stands despite variation in composition and environmen-
- Fire disturbance and forest structure in old-growth mixed conifer forests tal setting. Trees of similar age were clustered at small
spatial scales (i.e. 6-12 m distance classes) and there
was overlap in tree ages between clusters suggesting
overlap in recruitment at the patch edges. There are no
similar analyses of the spatial structure of tree ages for
other mixed conifer forests in Sierra Nevada. However,
large-diameter Sequoiadendron giganteum trees occur
in small widely spaced clusters. S. giganteum forests
are a unique variant of the California mixed conifer
forest type. No age data are reported for giant sequoia
clusters, however, so it is uncertain if they are even aged
or not. More recently, North et al. (2004) described the
spatial structure of large diameter (> 76 cm DBH) trees
in mixed conifer forests as being random; different than
what we found in GCW or in giant sequoia forests (i.e.
Bonnicksen & Stone 1982). Like the S. giganteum-mixed
conifer study, North et al. (2004) analysed only spatial
structure of tree sizes rather than age. Overall, our work
supports the view that mixed conifer forest dynamics can
be characterized as a shifting mosaic (e.g. Bonnicksen
& Stone 1982) but only with regard to structure and
not species composition Thus, the shifting mosaic is a
structural and not a compositional mosaic that is related
to the interaction between forest structure and the fire
mosaic (Miller & Urban 1999a).
Although species composition and environmental
conditions varied among stands, fire regime characteristics were quite similar. There was no difference
in the median fire return intervals (median FRI = 10 to
13 a), patterns of fire severity, or season of fire among
stands, and most fires burned only a few stands in a
fire year. The median FRI in GCW is within the range
identified for other mixed conifer forests in California
(e.g. Skinner & Chang 1996) and in northern Mexico
(Stephens et al. 2003). Several studies have identified
significant variability in fire regime parameters that
parallel species composition and environmental gradients (e.g. Taylor 2000, Beaty & Taylor 2001; Bekker &
Taylor 2001). For example, in the southern Cascades
fire return intervals decreased along a north to south
slope aspect gradient (Beaty & Taylor 2001). These
differences appear to be related to fuel bed characteristics (short needle Abies vs. long needle Pinus) and fire
season length which influence fire spread (Agee 1993).
There was some evidence of this relationship in GCW.
Fire return intervals shortened with increased dominance
of long-needle pines (P. jeffreyi) on south facing slopes
(PCA). Overall, the environmental and species composition gradients sampled in GCW were not long enough
to promote development of distinct fire regimes among
stands (i.e. Taylor 2000; Bekker & Taylor 2001).
Although stands burned with similar frequency,
they burned in different years. This suggests that the
similarity in fire regimes and forest structure among
887
stands is not the result of periodic widespread fires but
instead results from small burns that are influenced by
the self limiting process of post-fire fuel accumulation.
However, in some years fires were synchronized across
the landscape, especially during drought. Between 1700
and 1880, on average, five-fold more sites burned in
GCW in the 10 driest years than in the 10 wettest years
(PDSI GP-46; Cook et al. 2004).
Fire influences on spatial and temporal patterns of
tree regeneration
The patterns described above inform the second
question addressed in this study: how do patterns of fire
effects influence the spatial and temporal patterns of
tree regeneration? Prior to EuroAmerican settlement,
frequent surface fires influenced old-growth stand
structure by killing mainly seedlings, saplings, and
small diameter trees while leaving most thick-barked,
large diameter trees intact. This pattern of burning
promoted open stand conditions in a wide range of
size- and age-classes. This suggests that tree regeneration was temporally intermittent with spatial overlap
among regeneration patches, a pattern that has been
alluded to in other mixed conifer forests (e.g. Taylor
& Skinner 2003; Norman 2002; North et al. 2005).
The spatial autocorrelation of tree ages suggests that
successful recruitment into the canopy often occurs in
small patches (113-254 m2). These patch sizes are similar
to sapling patch sizes in mixed conifer forests in northern Mexico that have not experienced fire suppression
(Stephens & Fry 2005). Thus, the multi-age structure
of these forests is an aggregation of small overlapping
even-aged patches.
The temporal and spatial pattern of tree establishment
may be related to variation in both local patterns of fire
severity and the length of fire free periods. The spatial
autocorrelation of tree ages suggests that openings, such
as those caused by death of a canopy tree by torching, are
locations where trees establish. Subsequent survival of
established trees in a burn patch may then be dependent
on the length of time before the patch burns again (Brown
& Wu 2006). Once the trees mature their relatively thick
bark make them more resistant to fire than thin-barked
seedlings or saplings (Agee 1993). Infilling would be inhibited by fire soon after seedlings established. Given the
patchy nature of fires and the longevity (e.g. > 350 a) of
trees, periods with less fire and higher tree establishment
may result in structural attributes that persist for decades
or even centuries (Savage et al. 1996). Finally, there was
little evidence that widespread fires were more severe
than smaller burns. For example, there were no strong
peaks in the age-class distribution of stands associated
with widespread fire years.
888
Beaty, R.M. & Taylor, A.H.
Conclusions
References
Fire and forest structure interact in mixed conifer forests to create and perpetuate similar stand characteristics
across a range of environmental settings.
This suggests that, at the stand scale, fire occurrence
is controlled primarily by spatial variation in fuel mosaics
(e.g. patterns of abundance, fuel moisture, forest structure). Regional climate variation (i.e. drought) appears
to synchronize fires among stands in some years and this
may also influence stand structure. Yet, there were no
strong peaks in the age-class distribution of stands that
corresponded with years of more widespread burning.
Moreover, tree establishment in mixed conifer forests
may be more synchronized by variation in precipitation
and other climatic conditions associated with El Niño
than with surface fires (North et al. 2005). Thus, the
composition, structure and dynamics of mixed conifer
forests are influenced by both local processes that shape
the nature of patch development and large scale climate
variation. However, the dominant influence on forest
composition and structure over the last 120 years has
been fire exclusion, and this may have reduced the resilience of these forests to future fire by increasing live
and dead fuels, their susceptibility to drought and insect
induced mortality (Guarin & Taylor 2005) and the risk
of high severity fire.
Agee, J.K. 1993. Fire ecology of Pacific Northwest forests.
Island Press, Washington, DC, US.
Arno, S.F. & Sneck, K.M. 1977. A method for determining fire
history in coniferous forests of the mountain west. USDA
Forest Service General Technical Report INT-GTR-42,
Washington, DC, US.
Baisan, C.H. & Swetnam, T.W. 1990. Fire history on a desert
mountain range: Rincon Mountain Wilderness, Arizona,
U.S.A. Can. J. For. Res. 20: 1559-1569.
Barbour, M.G. & Minnich, R.A. 2000. California upland forests and woodlands. In: Barbour, M.G. & Billings, W.D.
(eds.) North American terrestrial vegetation, pp. 159-200.
Cambridge University Press, Cambridge, UK.
Beaty, R.M. & Taylor, A.H. 2001. Spatial and temporal variation
of fire regimes in a mixed conifer forest landscape, Southern
Cascades, California, USA. J. Biogeogr. 28: 955-966.
Bekker, M.F. & Taylor, A.H. 2001. Gradient analysis of fire
regimes in montane forests of the southern Cascade Range,
Thousand Lakes Wilderness, California, USA. Plant Ecol.
155: 15-28.
Bolsinger, C. 1989. Shrubs of California’s chaparral, timberland, and woodland: area ownership and stand
characteristics. USDA Forest Service Resource Bulletin
PNW-RB-160, Washington, DC, US.
Bonnicksen, T.M. & Stone E.C. 1981. The Giant SequoiaMixed Conifer forest community characterized through
pattern analysis as a mosaic of aggregations. For. Ecol.
Manage. 3: 307-328.
Bonnicksen, T.M. & Stone, E.C. 1982. Reconstruction of a presettlement Giant Sequoia-Mixed Conifer forest community
using the aggregation approach. Ecology 63: 1134-1148.
Brown, P. M. & Wu, R. 2006. Climate and disturbance forcing
of episodic tree recruitment in a south-western ponderosa
pine landscape. Ecology 86: 3030-3038
Caprio, A.C. & Swetnam, T.W. 1995. Historic fire regimes
along an elevational gradient on the west slope of the
Sierra Nevada, California. In: Brown, J.K., Mutch, R.W.,
Spoon, C.W. & Wakimoto, R.H. (eds.) Symposium on fire
in wilderness and park management, pp. 173-179. USDA
Forest Service Intermountain Research Station, Missoula,
MT, US.
Cliff, A.D. & Ord, J.K. 1973. Spatial autocorrelation. Pion,
London, UK.
Cook, E.R., Woodhouse, C.A., Eakin, C.M., Meko, D.M.
& Stahle, D.W. 2004. Long-term aridity changes in the
Western United States. Science 306: 1015-1018.
Dieterich, J. 1980. The composite fire interval – a tool for
more accurate interpretation of fire history. In: Stokes,
M. & Dieterich, J. (eds.) Proceedings of the fire history
workshop, pp. 8-14. U.S. Forest Service General Technical
Report RM-GTR-81, Washington, DC, US.
Elliot-Fisk, D., Rowntree, R. A., Cahill, T. A, Goldman, C. R.,
Gruell, G., Harris, R., Liesz, D., Lindström, S., Kattelmann,
R., Machida, D., Lacey, R., Bucks, P., Sharkey, D.A. &
Ziegler, D.S. 1996. Lake Tahoe Case Study. Sierra Nevada
Acknowledgements. We would like to thank M. Johnson and
J. Swanson (Lake Tahoe Basin Management Unit) and G.
Walter, E. Mandeno, and K. Anderson (California State Parks)
for important administrative and logistic support during the
field phase of this project. L. Kirsch, J. Rubini, B. Schorr, L.
DeMais, M. Do, J. Sperry, J. Sakulich and A. Scholl assisted in
the field. A. Scholl provided thoughtful comments on an earlier
version of this manuscript. This research was partially supported
by the USDA Forest Service, Lake Tahoe Basin Management
Unit (PA-05-98-19-030), a George S. Deike Research Grant,
an AAG and dissertation enhancement grants from the AAG
Biogeography Specialty Group and from the Ruby S. Miller
Endowment for Geographic Excellence.
- Fire disturbance and forest structure in old-growth mixed conifer forests Ecosystem Project: Final Report to Congress. Addendum,
pp. 217-276. University of California Centers for Water
and Wildland Resources, Davis, CA, US.
Gauch, H.G. 1982. Multivariate analysis in community ecology.
Cambridge University Press, New York, NY, US.
Grissino-Mayer, H.D. 2001. FHX2 - software for analyzing
temporal and spatial patterns in fire regimes from tree
rings. Tree-Ring Res. 57: 115-124.
Guarin, A. & Taylor, A.H. 2005. Drought triggered tree mortality in mixed conifer forests in Yosemite National Park,
California, USA. For. Ecol. Manage. 218: 229-244.
Hickman, J.C. 1993. The Jepson manual: higher plants of Cali­
fornia. University of California Press, Berkeley, CA, US.
Kilgore, B.M. 1973. The ecological role of fire in Sierran
conifer forests. Quat. Res. 3: 496-513.
Kilgore, B.M. & Taylor, D. 1979. Fire history of a Sequoiamixed conifer forest. Ecology 60: 129-142.
Lindström, S. 2000. A contextual overview of human land
use and environmental conditions. In: Murphy, D.D. &
Knopp, C.M. (eds.) Lake Tahoe watershed assessment, pp.
23-122. USDA Forest Service General Technical Report
PSW-GTR-175, Washington, DC, US.
Miller, C. & Urban, D.L. 1999a. Interactions between forest
heterogeneity and surface fire regimes in the southern Sierra
Nevada. Can. J. For. Res. 29: 202-212.
Miller, C. & Urban, D.L. 1999b. Forest pattern, fire and climatic
change in the Sierra Nevada. Ecosystems 2: 76-87.
Miller, C. & Urban, D.L. 2000a. Connectivity of forest fuels
and surface fire regimes. Landscape Ecol. 15: 145-154.
Miller, C. & Urban, D.L. 2000b. Modeling the effects of fire
management alternatives on Sierra Nevada mixed-conifer
forests. Ecol. App. 10: 85-94.
Moran, P. 1948. The interpretation of statistical maps. J. R.
Stat. Soc. B 10: 243-251.
Nagel, T. & Taylor, A.H. 2005. Fire and persistence on montane
chaparral in mixed conifer forest landscapes in the northern
Sierra Nevada, Lake Tahoe Basin, California. USA. J.
Torrey Bot. Soc. 132: 442-457.
Norman, S.P. 2002. Legacies of anthropogenic and climate
change in fire prone pine and mixed conifer forests of
northeastern California. Ph.D. Dissertation Geography,
Pennsylvania State University, University Park, PA, US.
North, M., Chen, J., Oakley, B., Song, B., Rudnicki, M., Gray,
A. & Innes, J. 2004. Forest stand structure and pattern of
old-growth western hemlock/Douglas-fir and mixed conifer
forests. For. Sci. 50: 299-311.
North, M., Hurteau, M., Fiegener, R. & Barbour, M. 2005.
Influence of fire and El Niño on tree recruitment varies by
species in Sierran mixed conifer. For. Sci. 51: 187-197.
Parker, A.J. 1982. The topographic relative moisture index: an
approach to soil moisture assessment in mountain terrain.
Phys. Geogr. 3:160-168.
Parker, A. J. 1986. Persistence of lodgepole pine forests in the
Sierra Nevada. Ecology 67: 1560-1567.
Romme, W.H. & Knight, D.H. 1981. Fire frequency and subalpine forest succession along a topographic gradient in
Wyoming. Ecology 62: 319-326.
889
Savage, M., Brown, P.M. & Feddema, J. 1996. The role of climate in a pine forest regeneration pulse in the southwestern
United States. Ecoscience 3: 310-318.
Skinner, C.N. & Chang, C. 1996. Fire Regimes, past and
present. Sierra Nevada Ecosystem Project: Final Report
to Congress. Vol. 2, Assessments and scientific basis for
management options, pp. 1041-1069. University of California Centers for Water and Wildland Resources, Davis,
CA, US.
Sokal, R. & Rohlf, F. 1995. Biometry: the principles and
practice of statistics in biological research. W.H. Freeman,
New York, NY, US.
Stephens, S.L. & Collins, B.M. 2004. Fire regimes of mixed
conifer forests in the north-central Sierra Nevada at multiple
spatial scales. Northwest Sci. 78: 12-23.
Stephens, S.L. & Fry, D.L. 2005. Spatial distribution of regeneration patches in an old-growth Pinus jeffreyi mixed conifer
forest in northern Mexico. J. Veg. Sci 16: 693-702.
Stokes, M.A. & Smiley, T.L. 1968. An Introduction to Tree-Ring
Dating. University of Chicago Press, Chicago, IL, US.
Stephens, S.L., Skinner, C.N. & Gill, S.J. 2003 Dendro­
chronology-based fire history of Jeffrey pine-mixed conifer
forests in the Sierra San Pedro Martir, Mexico. Can. J. For.
Res. 33: 1090-1101.
Swetnam, T.W. 1993. Fire history and climate change in giant
sequoia groves. Science 262: 885-889.
Taylor, A.H. 2000. Fire regimes and forest changes in mid and
upper montane forests of the southern Cascades, Lassen
Volcanic National Park, California, USA. J. Biogeogr.
27: 87-104.
Taylor, A.H. 2004. Identifying reference conditions for ecosystem management in late 19th century cut-over lands in
the Sierra Nevada, Lake Tahoe, Nevada. Ecol. Appl. 14:
1903-1920.
Taylor, A.H. & Beaty, R.M. 2005. Climatic influences on fire
regimes in the northern Sierra Nevada, USA J. Biogeogr.
32: 425-438.
Taylor, A.H. & Skinner, C.N. 1998. Fire history and landscape
dynamics in a late-successional reserve, Klamath Mountains, California, USA. For. Ecol. Manage. 44: 1-17.
Taylor, A.H. & Skinner, C.N. 2003. Spatial and temporal
patterns of historic fire regimes and forest structure as a
reference for restoration of fire in the Klamath Mountains.
Ecol. Appl. 13: 704-719.
Upton, G.J.G. & Fingleton, B. 1985. Spatial data analysis
by example. Vol. 1: Point pattern and quantitative data.
Wiley, London, UK.
Vankat, J.L. 1982. A gradient perspective on the vegetation
of Sequoia National Park, California. Madroño 29: 200214.
Vankat, J.L. & Major, J. 1978. Vegetation changes in Sequoia
National Park, California. J. Biogeogr. 5: 377-402.
Weatherspoon, C.P., Husari, S. & van Wagtendonk, J.W. 1992.
Fire and fuels management in relation to owl habitat in
forests of the Sierra Nevada and southern California. In:
Verner, J., McKelvey, K.S., Noon, B.R., Gutierrez, R.J.,
Gould, Jr., G.I. & Beck T.W. (eds.) California spotted owl:
890
Beaty, R.M. & Taylor, A.H.
a technical assessment of its current status, pp. 247-260.
U.S. Forest Service General Technical Report PSW-GTR133, Washington, DC, US.
Whitlock, C., Shafer, S.L. & Marlon, J. 2003 The role of
climate and vegetation change in shaping past and future
fire regimes in the northwestern US and implications for
ecosystem management. For. Ecol. Manage. 178: 5-21.
Wilken, G. 1967. History and fire record of a timberland
brushfield in the Sierra Nevada of California. Ecology
48: 302-304.
Received 11 October 2006;
Accepted 31 March 2007;
Co-ordinating Editor: E. Ezcurra.