Fuels Reduction and Group Selection Silviculture
in Sierran Mixed-Conifer Forest:
Short-term Effects on Environmental Attributes
Seth W. Bigelow, Colin Dillingham, and Lauren Payne
December 3, 2012
1
U. S. Forest Service, Pacific Southwest Research Station, 1731 Research Park Drive, Davis CA
95618, seth@swbigelow.net.
2
U.S. Forest Service, Plumas National Forest, 159 Lawrence St. Quincy, CA 95971,
cdillingham@fs.fed.us.
3
U.S. Forest Service, Vegetation Management Services Enterprise Unit, lpayne@fs.fed.us.
1
Abstract
The Herger-Feinstein Quincy Library Group Forest Recovery Act of 1999 was in the
vanguard of large national forest management projects in the Sierra Nevada aimed at protecting
forests from wildfire. Detailed monitoring was done because of the newness of the management
activities and the scale at which they were carried out, and this report documents the effects of
the portfolio of forest activities carried out as a result of the act. These treatments include
thinning live trees in small and medium diameter classes; group selection; prescribed underburning; and hand piling and burning. Response variables included tree density, basal area, and
canopy cover; the amount of standing and down dead wood; cover of shrubs and herbs; and
simulated fire behavior. This report documents pre-treatment and 1-year post-treatment condition
of 107 treatment units, and compliance with targets defined by forest planners. Treatments led to
significant decreases in canopy cover and basal area; mean canopy cover decreased from 51% to
35% after treatment, and mean basal area decreased from159 ft2/acre to 111 ft2/acre. Large (≥30
inches dbh) tree density, 3.4 per acre prior to treatment, did not change significantly with
treatment. Large down woody debris loads were smaller than in similar forests elsewhere: pretreatment 2.7 tons/acre,post-treatment, 1.6 tons/acre). Snags were also initially sparse, becoming
more so after treatment (snags of dbh ≥ 15 inches; pre-treatment 3.2 per acre, post-treatment
1.6per acre). Herbaceous understory plant cover increased significantly after treatment, but shrub
cover did not change. Areas identified as potential habitat for the California spotted owl declined
in most measures of habitat quality after treatment.Fire behavior simulation modeling showed
that treatments were largely effective in moderating predicted fire behavior under moderate and
2
severe weather scenarios. We concluded that treatments generally met the primary goalof
improving forest health and resilience to wildfire, and that they had mixed success in meeting
secondary objectives related to maintenance of habitat quality. We present suggestions on how
such decreases in habitat quality may be avoided or mitigated, and how canopy cover targets
may be converted to traditional basal area targets based on quadratic mean diameter.
Keywords: Group selection, fuels-reduction thinning, monitoring, Sierra Nevada Mixed-Conifer,
canopy cover, snag, down wood, understory plant
3
4
Contents
Introduction ..................................................................................................................................... 5
Logging and Fire in the Plumas and Lassen National Forests, 1850-2000 ................................ 5
The Quincy Library Group ......................................................................................................... 7
Monitoring Program: Targets and Guidelines .......................................................................... 10
Study Site and Methods ................................................................................................................ 12
Canopy Cover and Basal Area .................................................................................................. 15
Snags ......................................................................................................................................... 16
Down Wood .............................................................................................................................. 17
Understory plants ...................................................................................................................... 18
Owl Habitat ............................................................................................................................... 19
Fire Behavior Modeling ............................................................................................................ 20
Results ........................................................................................................................................... 21
Canopy Cover and Basal Area .................................................................................................. 21
Snags ......................................................................................................................................... 25
Down Wood .............................................................................................................................. 27
Understory Plants ...................................................................................................................... 28
Owl Habitat ............................................................................................................................... 29
Fire Behavior Modeling ............................................................................................................ 31
Discussion ..................................................................................................................................... 34
The untreated forest .................................................................................................................. 34
Treatment effects on the landscape ........................................................................................... 37
Management Implications ......................................................................................................... 41
Conclusions ............................................................................................................................... 46
Acknowledgments......................................................................................................................... 47
Literature Cited ............................................................................................................................. 48
4
5
Introduction
The national forests at the junction of the Sierra Nevada and Cascade mountain ranges in
northern California have undergone many phases of management over the century since their
establishment. In the past decade they have been the focus of an unprecedented experiment in
community forestry and national forest management. The present report documents changes in
forest structure resulting from this new management.
Logging and Fire in the Plumas and Lassen National Forests, 1850-2000
The forests currently administered as Plumas and Lassen national forests have been
logged commercially from 1850 to the present, supporting a succession of industries such as
precious-metals mining and housing construction (Young, 2003).Plumas and Lassen were
established as national forest reserves in 1905 and as national forests in 1907. Records of timber
production begin in 1909, but an earlier timber inventory confirms that substantial timber
harvesting had taken place prior to establishment of the national forest reserves (Leiberg, 1902).
Major logging enterprises (i.e., processing > 40 million board feet [bf; equivalent to a block of
finished lumber 12 by 12 by 1 inches] of lumber in any one year) have included Sierra Flume
and Lumber Co. (1876-1878), Red River Lumber Co. in (1917-1939), Spanish Peak Lumber
Company (1916-1939), Collins Pine (1941- present) and Sierra Pacific Industries (ca. 1949 to
present). After establishment of the national forests, timber production gradually increased over
many decades and from the 1960s to the 1980s both forests were producing timber commonly
exceeding an astounding 200 million bf/year (fig. 1).A large tree such as a ponderosa pine 30
inches in breast-height diameter (dbh) might yield 3,000 bf. Production dropped sharply in the
1990s due to increased legal challenges, new regulations to protect the California spotted owl,
5
6
depletion of the supply of large trees, and decreased staffing of the national forest management
offices (Bernard and Young, 1996).
Cut volume (million bf)
300
250
Lassen
Plumas
200
150
100
50
0
1900
1920
1940
1960
1980
2000
2020
Year
Figure 1. Annual volume of timber cut (millions of board feet) on Plumas and
Lassen National Forests, 1909-2011. No data were available for Plumas National
Forest for 1966-1976. Conversion to cubic feet may change ranking between
forests for a given year.
Fire and fire management have marked these forests as profoundly as logging. Prior to
the 1850‟s, the fire regime consisted predominantly of low-severity fire, with a fire return
interval varying from 8-22 years (Beaty and Taylor, 2001; Moody et al., 2006). Fire suppression
became an important element of management policy in Plumas and Lassen forests beginning
with their establishment as national forests (Stephens and Ruth, 2005).Fire suppression has had
unanticipated consequences including the buildup of dead wood and organic matter, and the
expansion of white fir, a species that is vulnerable to fire when it is young, far beyond its original
range. The combination of logging and fire suppression has led to forests that are denser and
6
7
have higher canopy cover than mixed conifer forests under the pre-European settlement fire
regime (Collins et al., 2011). This in turn has led to increased regeneration of white fir and other
shade-tolerant species(Bigelow et al., 2009; Bigelow et al., 2011). High density, with droughtsensitive species, deep canopies, buildup of fuels, and warmer temperatures have combined to
make these forests vulnerable to very large fires, and there have been several fires of tens of
thousands of acres burning at high intensity on the Plumas and Lassen national forests in the past
15 years. This trend of increasing size and severity of fires has been observed throughout the
Sierra Nevada over the last quarter of the 20th century (Miller et al., 2008).
The Quincy Library Group
In response to the diminished economy associated with the decline of logging on the
national forests, and increased threat of wildfires, a group of elected officials, timber industry
representatives, and environmentalists formed the Quincy Library Group. The group is named
for their meeting place in Quincy, the small town that is the Plumas County seat (Fig. 2). The
group devised a plan to provide timber for mills, improve environmental quality, protect spotted
owl habitat and wilderness, decrease erosion, and reduce fire risk (Yost, 1994; Bernard, 2010).
Elements of the 1993 Community Stability plan were adopted by the Forest Service as the Forest
Health Pilot plan, but the Quincy Library Group was dissatisfied with the pace of
implementation and took the plan directly to Congress(Bernard and Young, 1996), where it was
signed into lawas The Herger-Feinstein Quincy Library Group Forest Recovery Act(1998)and
later renewed[Consolidated Appropriations Act, 2008 (H.R.2764)]. The vision established in the
plan was to “promote an all-age, multistory, fire resilient forest approximating pre-settlement
conditions”(Yost, 1994), and the plan prescribed specific activities and targets for doing so. One
activity was to thin small-diameter trees in a system of shaded fuel breaks (Agee et al., 2000)
7
8
referred to locally as defensible fuel profile zones (DFPZs; Weatherspoon and Skinner, 1996).
Defensible fuel profile zones are designed to interrupt crown fire and provide a safe place for fire
crews to act against large, high intensity wildfires. Another activity was implementation of group
selection (GS), a management system in which small (e.g., 1-2 acres) stands of trees are
harvested. The Pilot Project was turned into management policy(USDA Forest Service, 1999); a
Sierra-wide management plan issued in 2001 conflicted with the Pilot Project but was ultimately
revised to allow the Pilot Project to proceed as originally envisioned (USDA Forest Service,
2004).
8
9
Figure 2. Map of sampling sites and areas treated as part of the Herger Feinstein
Quincy Library Group project in Plumas, Lassen, and Tahoe (Sierraville ranger
district) National Forests in northern California.
9
10
Monitoring Program: Targets and Guidelines
The Herger-Feinstein Quincy Library Group Forest Recovery Act contained a provision
for reporting of “forest health improvements….and other natural resources-related benefits”
resulting from the implementation of the Pilot Project. Questions of particular interest defined by
forest managers in collaboration with Quincy Library Group members follow:

Did silvicultural treatments result in the desired levels of canopy cover?

Were large trees protected?

Did treatments produce the desired abundance and distribution of snags and logs?

Was the amount of early seral stage vegetation enhanced as a result of treatments?

Was habitat for the California spotted owl protected?

Did treatments reduce fire hazard?
The rationale for monitoring these elements of forest health are that canopy cover, or the
proportion of the forest floor covered by the vertical projection of the tree crowns (Jennings et
al., 1999), is associated with wildlife habitat quality (Mayer and Laudenslayer, 1988),
regeneration (Moghaddas et al., 2008), and fire hazard (Agee and Skinner, 2005). Large trees
(usually considered to be those with diameter at breast height ≥30 inches) provide habitat for
many wildlife species, store enormous amounts of carbon, and are resistant to fire. Snags provide
habitat for cavity-nesting birds, small mammals, insects, lichens, reptiles, amphibians, and fungi
(Bull, 2002; Bunnell et al., 2002; Innes et al., 2008), yet can be dangerous to fire-fighters and
other forest managers. Down wood provides habitat for animals from amphibians to
mesocarnivores, sequesters mineral nutrients, and drives restoration of plant species diversity by
creating heterogeneity in burn intensity when fire occurs (Wayman and North, 2007).Understory
10
11
plants such as shrubs, forbs, and grasses are the largest source of plant species diversity in mixed
conifer forests, and they provide forage to early-seral associates such as mule deer, towhees,
hummingbirds, and black bear. Monitoring fire hazard, though physics-based models, is key to
understanding the success or failure of the treatments. The concept is to simulate fire behavior
under conditions of high wind and heat, to predict whether a running crown fire would be
initiated and/or sustained if ignition of the understory were to occur (Albini, 1976).
The California spotted owl (Strix occidentalis occidentalis) was the only species singled
out in the monitoring program, in part because its close relative the northern spotted owl (Strix
occidentalis caurina) has a strong habitat preference for old-growth forest features such as snags,
large trees, and coarse woody debris(Buchanan et al., 1995) and has been the subject of intense
controversy over logging practices in the Pacific Northwest. Like the northern subspecies,
California spotted owls select nests and roosts in habitat with multiple vegetation strata (LaHaye
et al., 1997), large trees (e.g., ≥100 cm dbh; Moen and Gutiérrez, 1997)and high canopy cover
(Bias and Gutierrez, 1992; Bond et al., 2004; Blakesley et al., 2005).Optimal foraging habitat is
more controversial; an early study indicated that the owls selected against foraging habitat with ≤
40% canopy cover (Call et al., 1992), but recent studies have found owls preferentially foraging
in open areas of little cover (Bond et al., 2009; Williams et al., 2011). Given the increasing body
of knowledge regarding California spotted owl population habitat preferences it may be possible
to interpret changes in forest structure due to management with regard to owl habitat preferences.
A monitoring program was established at the outset of the project. In the present report
we describe short-term effects of treatments on common indicators of forest health by comparing
data collected prior to treatment with data collected up to one year after treatment. We did not
attempt to determine the effects of particular prescriptions on forest elements; our goal was
11
12
instead to describe the effects of the management regime as a whole. As the first national forests
in the Sierra Nevada to undergo large-scale fire-oriented silviculture treatments, these forests
have received intensive monitoring, in part as test of whether forest conditions can be maintained
or improved by fuel-reduction treatments in conjunction with a regime of gap-creation, i.e.,
group selection. The data collected prior to treatment provide an unusually detailed picture of the
condition of the forests at the onset of the 21th century, and such representations can have
enduring value as a reference for managers and researchers (e.g., Leiberg, 1902). Taken together,
the elements monitored provide a picture of the effects of contemporary, fuels-oriented forest
management on the forest environment.
Study Site and Methods
The study took place in Plumas and Lassen National Forests, and the Sierraville ranger
district of Tahoe National Forest (fig. 2). The region is characterized by a Mediterranean climate
with wet winters and dry summers. Long-term annual precipitation for the area is 37.4 (11.7)
inches (mean and standard deviation), and average precipitation during the June to September
summer is 2 inches (1896-2010; data from Chester, Canyon Dam, Greenville, and Quincy
weather stations via the Western Regional Climate Center; http://www.wrcc.dri.edu). The forests
span the main crest of the Sierra Nevada range; because moisture-laden winds come
predominantly from the west, the western slopes tend to have higher precipitation. The long-term
minimum winter (December-February) average temperature is 22.9 (3.3) °F, and average
maximum summer temperature (June through September) is 82.8 (1.9) °F.
Forest types in the study were classified as Sierran mixed conifer, white pine, ponderosa
pine, and aspen types in the California Wildlife Habitat Relationship system (Mayer and
12
13
Laudenslayer, 1988). The 107 units for which one-year post-treatment follow-up data were
available comprised 46 Sierran mixed conifer, 24 white fir forest types, 24 east side pine, and 13
ponderosa pine. Most sample units were within the DFPZ network, but many of the group
selection stands were not (fig.2).Treatments included group selection, mechanical thinning,
underburning, mastication, and/or hand-thinning and pile burning. As of mid-2011, 170 plots had
been established, 107 of which had been treated and re-measured (Table 1). Under-burning was a
scheduled follow-up treatment to reduce surface fuels in many areas, and the first post-treatment
measurement was intended to occur after under-burning under the original monitoring design.
Because implementation of many burns was delayed, the monitoring plan was modified to allow
the first post-treatment stand assessment to occur before under-burning was done.
Table 1—Allocation of 107 units in Plumas, Lassen, and Tahoe (Sierraville ranger
district) national forests to matrix(units in defensible fuel profile zones [DFPZ] or
aspen restoration areas), group selection (GS), and/orprescribed understory
burning treatments
DFPZ
Matrix treatment
GS
Unburned
Burned
Y
Hand Thin/Pile
N
0
7
Y
Mastication
N
3
1
Y
Mechanical Thin
N
36
31
Y
Underburn only
N
0
1
Y
Mechanical Thin
Y
15
3
Y
Underburn only
Y
0
1
N
None
Y
7
0
N
Aspen restoration
N
2
0
63
44
Totals
13
14
Each plot was made up of three ¼ acre rectangular subplots that were located randomly
within a sample unit. The ¼ acre subplots had dimensions of 165 by 66 ft and contained nested
subplots: one 1/8th acre plot (165 × 33 ft), one 1/16th acre plot (165 ×16.5 ft), five 1/250 acre
plots (13.2 ×13.2 ft), and five 1/1000 acre plots (6.6 × 6.6 ft). Permanent tags were attached to
down logs and live and dead trees. Plot boundaries were ≥ 20 ft from roads. Pre-treatment
surveys were conducted between 2001 and 2009. Post-treatment surveys were generally done
within one to three years after mechanical treatment. Data on live and dead trees were entered
and permanently archived in the USFS‟s Common Stand Exam database. The Forest Vegetation
Simulator was used to summarize data on stand basal area, and output was archived in a
dedicated database (TSSM; available upon request, cdillingham@fs.fed.us). The dedicated
database also contained details on prescriptions, dates of treatments, and treatment objectives.
Many of the attributes monitored had specific objectives or targets, and in such cases we
tested for compliance of treatments with targets. Typical canopy cover targets for dense westside
stands are 40% whereas targets for eastside stands are usually lower. Because attaining canopy
cover targets is relatively new as a management metric and many field personnel are more
accustomed to marking stands to achieve residual basal area targets, we used our data to
determine the relationship between plot basal area and canopy cover. Sierra Nevada management
plans prohibit removal of large trees under most circumstances, and management objectives for
most projects include retaining all large trees. Average snag densities should vary from 3 to 8 per
acre depending on location within the landscape. Goals for large down wood are to retain 10-15
tons/acre averaged over the treatment unit in west side vegetation types, and three large down
logs per acre in east side vegetation types (USDA Forest Service, 2004).
14
15
Canopy Cover and Basal Area
Vertically projected canopy cover was measured in the ¼ acre plots by sighting the
canopy at 16.5 ft intervals along three lines ineach¼ acre plot with a vertical sighting tube
(canopy densitometer; Geographic Resource Solutions, Arcata CA). Canopy cover was
expressed as the number of sighting points with foliage vertically overhead divided by the total
number of points (90 points per ¾ acre plot) where sightings were taken. Differences in pretreatment cover among ecological zones were analyzed with pairwise t-tests. Effects of
treatments on canopy cover were analyzed by subtracting pre-treatment canopy coverfrom posttreatment canopy cover for each unit, then applying one-way analysis of variance by ecological
zone (east side, west side, or transition). A change from pre-treatment to post-treatment cover
was deemed significant if the estimate of the difference was significantly different than zero
based on the t distribution with less than 5% probability of committing a type 1error (i.e., α =
0.05). Analyses were done with linear models (lm command)in the R statistical environment (R
Development Core Team, 2008).
Basal area was calculated from censuses of trees. All trees ≥ 5 inches dbh were measured
in the 1/16th acre subplots, all trees ≥ 16 inches dbh were measured in 1/8thacre subplots, and all
trees ≥ 30 inches dbh were measured in the ¼ acre plots. Tree densities were scaled up to acre
basis using appropriate expansion factors for each dbh class. Stand basal area and quadratic
mean diameter were analyzed incorporated trees ≥5 inches dbh. Both were analyzed similar to
canopy cover, with an analysis of variance on pre-to-post treatment differences using ecological
zone as a factor. Stand basal area (difference in pre- and post-basal area) was also analyzed with
two-way analysis of variance, with dbh class (6–11.9 inches, 12–23.9 inches, 24–29.9 inches,
and ≥30 inches) and ecological zone (east side, west side, or transition) as factors.
15
16
Canopy cover targets were assigned to most projects prior to treatment Management
guides provided suggested canopy cover targets that were specific to ecological zones, but
managers were encouraged to define their own targets for their projects (USDA Forest Service,
2004). Most targets were given as a single value (e.g., “maintain 40% minimum canopy cover”),
but some were given as ranges. In order to develop a uniform system, we defined successful
attainment as having post-treatment canopy cover within +/- 7 percent canopy cover of target
canopy cover (e.g., 33%-47% range for 40% canopy cover target). Only stands with canopy
cover targets assigned prior to treatment (88 of 107) were included in the analysis. We calculated
the proportion of stands with canopy cover above, within, and below the target range both preand post-treatment. Many stands had canopy cover within or below the target range prior to
treatment, and we statistically compared the proportion of stands within the target range pre- and
post-treatment by ecological zone. McNemar‟s chi-squared test(implemented in R,
mcnemar.test) was used because pre- and post-measurements were done on the same stands,
resulting in paired, non-independent data (Agresti, 1990).
Due to the interest expressed by managers in being able to translate canopy cover targets
to basal area, we made linear regressions of basal area against canopy cover. The relationship
between crown diameter and tree diameter may change as trees grow in size (Gill et al., 2000),
and to allow for this possibility data were segregated intothree quadratic mean diameter (QMD)
ranges: ≥ 6 – 11.9 inches, 12 – 17.9 inches, and 18 – 24 inches. Analyses were done on both pretreatment data (170 plots) and post-treatment data (107 plots).
Snags
Snags were tallied along with live trees following the same system of diameter class
cutoffs. We tested for changes in mean snag density as a result of treatment according to
16
17
ecological zone. Fifteen inches is considered a minimum dbh for snag use by many wildlife
species, so we only considered snags greater than or equal to this size. Tests were done following
the same approach as for canopy cover, i.e., analysis of variance with post-treatment minus pretreatment density as the response variable and ecological zone as the experimental factor.
Snag retention targets, generally 3, 4, or 8 snags per acre, were assigned to projects prior
to treatment; the higher targets were assigned to more productive west side sites(p. 69; USDA
Forest Service, 2004).McNemar‟s test was applied to test whether post-treatment attainment
differed from pre-treatment attainment (Agresti, 1990). In 2010, the fate of snags that were
present before and absent after treatment wasrecorded as either cut and taken, fallen over and
present as down wood, or cut for firewood. These data were tested for a relationship between
snag size and snag fate with a chi-square test.
Down Wood
Down logs ≥ 10 inches in large-end diameter and ≥ 10 ft long that had their large end
within the ¼ acre subplots were measured for length and large-end diameter. Aluminum tags
were nailed to logs during the pre-treatment survey, and logs that were too decayed to retain a
nailed tag were disregarded. Log weight was calculated by multiplying estimated volume by a
coefficient for specific gravity that differed depending on whether logs were sound or decayed
(Brown, 1974). Volume was calculated from log large-end diameter, log length, and an estimate
of small-end log diameter estimated from a taper equation (Biging, 1984). Use of a taper
equation requires an estimate of whole-tree height, gained by creating species-specific linear
equations predicting height from diameter at breast height (dbh) from the live-tree survey. When
species-specific parameters were not available for the taper equation (i.e., for species other than
the major commercial conifers), parameters for white fir were used. Data were summarized in
17
18
10-19.9 inch and ≥20 inch large-end diameter classes. Effects of treatment on wood loads were
tested by the same statistical approach as for canopy cover, i.e., analysis of variance on postminus pre-treatment down wood loads using ecological zone as a factor.
Suggested targets for large log retention were provided as weights (tons/acre)for west
side stands and densities (logs/acre) for east side stands (USDA Forest Service, 2004). To have a
consistent metric, we converted log density targets to weights based on the assumption that one
log weighs 1000 lb. Attainment of log retention targets was defined as having log weight greater
than or equal to the target. To allow for the possibility that our sampling scheme had
underestimated log weights due to lack of measurement of highly decayed wood, we adjusted
target weights. We used an independent dataset from the same area (HFQLG soils monitoring
group) to estimate the fraction of total down wood weight contributed by sound logs (0.6 =
sound log weight/total log weight), and multiplied minimum target weights by that fraction.
Change in target attainment from pre- to post-treatment was analyzed with McNemar‟s test
(Agresti, 1990)
Understory plants
One of the main interests of managers in understory plants is as forage for wildlife. The
great majority of understory plant species in mixed conifer forest are consumed by wildlife in
some way, but there are non-palatable, less-preferred, or otherwise noxious species in the study
area such as bull thistle, chinquapin, cheat grass and bracken fern. We classified each species as
either palatable or non-palatable in consultation with wildlife experts, and analyses were only
done on palatable species. Shrub cover was assessed in five 1/250 acre (13.2 × 13.2 ft) subplots
in each ¼ acre subplot, and forb and grass or grass-like (graminoid) plant cover was assessed in
five 1/1000 acre (6.6 × 6.6 ft) plots. Total cover of any species that made up ≥10% of each plot
18
19
was estimated visually and recorded in 10% cover increments. Average cover per species in each
sampling unit was calculated by dividing the sum of cover by 15 (i.e., the number of subplots per
plot). To obtain mean cover for each life form (shrub, graminoid, or forb), cover was summed
across all species within a life form.
An exploratory analysis was done of the relationship between overstory and understory
cover by life form. Overstory plant density may limit understory community development at high
densities, so we checked for this effect by plotting understory cover (by plant lifeform) against
tree canopy cover. The relationship was evaluated with linear regression (lm command in R).
The difference between pre- and post-treatment cover of each life form was analyzed by
ecological zone with analysis of variance following the method used for canopy cover.Cover of
the three most abundant species in each life form, pre- and post-treatment, is presented in a table.
Owl Habitat
US Forest Service wildlife biologists examined pre-treatment stand data to determine
whether stands were potential habitat for the California spotted owl. In making their decisions
the biologists used personal knowledge of the treatment area, and numeric criteria based on
experience and/or published literature, e.g. QMD ≥ 12 inches, presence of large trees, canopy
cover≥40%, and/or basal area≥180 ft2/acre. Forty-seven of the 107 stands with post-treatment
data were classified as potential owl habitat prior to treatment. A two-tailed, paired t-test was
used to determine whether post-treatment measures differed from pre-treatment ones at α = 0.05
level of significance (Ott, 1989). The measures comprised large log weight, stand basal area,
large tree density, snag density, canopy cover, and QMD. No attempt was made to determine
19
20
effects of different treatments (e.g., thinning vs. group selection), or to distinguish ecological
zone effects.
California wildlife habitat relationship (CWHR; Mayer and Laudenslayer, 1988) codes
expressing canopy cover range, dominant tree size class, and forest type were assigned to each
plot before and after treatment. Classifications were assigned based on canopy cover measured
with vertical sighting tube, quadratic mean diameter, and tree species composition. No statistical
analysis is done but matrices showing CWHR codes before and after treatment (as a transition
matrix) are displayed.
Fire Behavior Modeling
Fire behavior was simulated under moderate and severe weather conditions with the fire
and fuels extension to the forest vegetation simulator (Rebain et al., 2011). The fire and fuels
extension uses physically based equations to estimate fire intensity from wind speed, air
temperature, plot slope, and fuel amounts, surface area, and moisture (Rothermel, 1972). Fire
intensity has both vertical (flame length) and horizontal (rate of spread) components. The
estimated fire intensity is combined with information on tree density, species composition, size
distribution, and crown depth to predict how a fire would behave if ignited in that stand. Fires are
classified according to whether torching (crowns of large trees being ignited by surface
fire)and/or crowning (fire passing from crown to crown of large trees) are predicted to occur. A
surface fire has neither torching nor crowning, a passive crown fire has torching but not
crowning, a conditional crown fire has crowning but not torching, and an active crown fire has
both torching and crowning.
20
21
Weather and fuel moisture conditions for the simulations were obtained from nine remote
access weather stations in the vicinity of the study plots. Wind speed, air temperature, and live
and dead fuel moistures were calculated for 90th and 97th percentile conditions for the period
1998-2008; each plot was assigned weather conditions from the nearest station. Surface fuels
(those between the soil surface and 6 feet above ground) were estimated visually in the 1/250
acre plots with the photo series method. An algorithm in the fire and fuels extension used these
measurements, along with information on shrubs and herbs (see earlier sections), and stand type
to select among a limited number of fuel models to characterize fuel bed depth, and surface fuel
loads and surface-to-volume ratios within each surface fuel size class. When the algorithm
appeared to mischaracterize surface fuels, expert opinion was used to select the most appropriate
model.
Results
Canopy Cover and Basal Area
Mean pretreatment canopy cover in west side forests, 54%, was 6 percent canopy cover
units higher than in eastside forests (Table 2), but this difference was not statistically significant
(p = 0.12). In response to treatment, canopy cover decreased on average 17% in east side forest,
18% in transition, and 15% in west side forest; all decreases in canopy cover were highly
statistically significant. Prior to treatment, canopy cover in most of the 88 stands with specified
targets was above or within the target treatment range, but a few were below it (Table 2). The
number of units above the target range decreased greatly after treatment (from 63 to 13), with
approximately equal numbers of units ending up in within and below canopy cover target ranges.
The pilot-project-wide success rate for 88 units was 47% (36–57%, 95% confidence interval).
21
22
Table 2—Tree canopy cover (mean and standard deviation) and attainment of
target canopy cover pre- and post-treatment by ecological zone.
Zone
Canopy Cover
n
Posta
Pre
Target Attainment
nb
Pre
over
in
per cent
Post
under
over
inc
under
per cent
East side
56
48 (14)
31(11) ***
52
71
14
15
10
42**
48
Trans.
5
56 (16)
39(10) **
5
80
20
0
40
60
0
West side
46
54 (15)
39(15) ***
31
68
26
6
19
48
32
All
10
51(15)
35(15)***
88
71
18
11
15
45***
40
7
a
significance values for pre vs. post comparison:*p< 0.05, **p< 0.01, ***p<0.001;
b
number of units with assigned canopy cover targets
c
significance values (defined as ina) for whether proportion of units in attainment changed
between pre- and post-treatment
Treatments led to significant decreases in basal area (Fig. 3) in the 6–11.9 inch and 12–
23.9 inch dbh classes (p< 0.001).The 6–11.9 inch dbh size class decreased by 29 ft2/acre on
average, and the 12–23.9 inch dbh size class decreased by 31 ft2/acre (Table 3). Neither the 24–
29.9 inch nor the ≥30 inch (i.e., large tree) dbh size classes changed significantly (p> 0.4). Mean
basal area of ≥ 30 inch dbh trees was 12.8 (22.6 standard deviation) ft2/acre prior to treatment
and 14.5 (SD 24.0) ft2/acre after. As a frame of reference for basal area, a 30 inch dbh tree has a
basal area of 5 ft2. On average, large trees contributed 7% of basal area of all trees ≥ 6 inches
dbh. The small, non-significant increase in mean basal area may have been due either to growth
of existing large trees or recruitment of new ones; on average, 4.5 years elapsed between the
pretreatment and 1-yr post-treatment measurements. Quadratic mean diameter (QMD) increased
22
23
significantly in all ecological zones, as is expected from treatments that predominantly remove
smaller trees.
120
(226)
(225)
EastPre
side
Post
pre
post
Transition
West side
100
80
***
60
***
***
40
-2
9.
9"
>=
30
"
24
-2
3.
9"
-1
1.
9"
6
-2
9.
9"
>=
30
"
12
-1
1.
9"
6
-2
9.
9"
>=
30
"
24
-2
3.
9"
12
-1
1.
9"
6
***
***
12
***
0
24
20
-2
3.
9"
Basal area (ft2/acre)
140
Size (DBH) Class
Figure 3. Mean basal area of stands before and up to one year after treatment, by
ecological zone and diameter size class. Vertical lines indicate standard
deviations. Bars with three asterixes differ from preceding bar at p<0.001.
There was a strong positive relationship between basal area and canopy cover,
particularly when stands were grouped by QMD. The relationship differed slightly between pretreatment and post-treatment stands (Table 4). Prior to treatment there was a clear difference in
the basal area-canopy cover relationship for each QMD range (Fig. 4). For pre-treatment stands
with QMD 6–11.9 inch (i.e., stands with many small trees), basal area was low compared to
stands with higher QMD. The pre-treatment canopy-cover multiplication factors (i.e., the number
that a target canopy-cover must be multiplied by to estimate a basal-area marking target) were
2.4, 3.3, and 3.9, respectively, for QMD in 6–11.9 inch, 12–17.9 inch, or 18–23.9 inch ranges
(Table 4). Post-treatment, the basal area-canopy cover relationship did not differ significantly
23
24
between QMD 6–11.9 inch and 12–23.9 inch ranges, although basal area in relation to canopy
cover was higher for the 18-23.9 inch QMD range.
Table 3—Basal area and quadratic mean diameter (trees≥5 inches dbh) pre- and
post-treatment by ecological zone
Zone
Basal area
Quadratic mean diameter
Posta
Pre
Pre
ft2/acre
Post
inches
East side
135(53)
91(34)***
11.9(2.0)
14.9(2.9) ***
Transition
253(107)
142(49)***
12.7(3.8)
17.7(5.8) ***
West side
177(67)
132(57)***
13.3(2.7)
16.6(4.3) ***
All
159(68)
111(51)***
12.5(2.5)
15.7(3.4)***
a*
p< 0.05, **p< 0.01, ***p<0.001; significance values for pre vs. post comparison
Table 4—Multipliers for obtaining target basal area from desired canopy cover,
derived from pre- and post-treatment stands.
QMD range
na
na
Pre
(ft2/acre)/%
inches
Post
(ft2/acre)/%
6–11.9
61
2.75 (2.51–2.99)b
15
3.05 (2.64–3.45)
12–17.9
96
3.48 (3.30–3.66)
66
2.97 (2.75–3.18)
18–23.9
11
3.97 (3.47–4.47)
21
3.61 (3.23–4.00)
a
number of stands used to derive multiplier
b
multiplier with 95% confidence interval in parentheses
24
25
18 - 23.3 inch QMD
12 - 17.9 inch QMD
6 - 11.9 inch QMD
300
2
Basal area (ft /acre)
400
200
100
0
0
20
40
60
80
100 0
20
40
60
80
100
Canopy cover (%)
Figure 4. Basal area of untreated plots (n=170) and one-year post-treatment plots
(n=107) as a function of canopy cover [grouped by quadratic mean breast-height
diameter (QMD) class].
Snags
The number of snags ≥15 inches dbh declined after treatment: snag density was 3.2
(4.4)[mean and standard deviation] before treatment and 1.6 (2.6) after treatment (Table 5).
There was a statistical decrease in snag density in all ecological zones, as well as the project as a
whole. As expected, snag distribution was patchy: prior to treatment, 53 of 107 stands lacked
snags, and after treatment, 66out of 107 lacked snags. In most cases, the standard deviation of
snag density was greater than the mean (i.e, mean-to-variance ratio was much less than one),
indicating that snags were clumped on the landscape when measured at the 3/8 to ¾ acre scale.
The number of plots in which snag retention objectives were met decreased after treatment, as
expected from the decline in snag densities. Prior to treatment, 29% of plots were in attainment,
but after treatment only 17% of plots were (different at p< 0.01;McNemar‟s test).
25
26
Table 5—Density of snags≥15 inches dbh and attainment of snag density
objectives pre- and post- treatment according to ecological zone
na
Zone
Snag Density
Pre
Target Attainment
Postb
Pre
count per acre
Post
per cent
Eastside
56
2.7(4.0)c
1.0(1.7)***
29
12**
Transition
5
6.4(2.6)
2.1(2.2)**
100
20
Westside
46
3.4(4.9)
2.3(3.4)*
24
22
Totals
107
3.2(4.4)
1.6(2.6)***
29
17**
a
number of units
p< 0.05, **p< 0.01, ***p<0.001; significance values for pre vs. post comparison
c
mean with standard deviation
b*
Of the 130 disappeared snags whose fate could be determined during the 2010 and 2011
surveys, 107 (82%) were cut and removed during the treatments, and the rest fell down:
dbh
Cut, Removed
Fallen, remain on site
Removed as firewood
5-10 in.
64
7
0
10-15 in.
30
7
0
15-20 in.
9
3
0
>20 in.
4
6
0
The relationship between dbh and cause of disappearance was highly statistically significant (χ2,
p = 0.001). Most of the smaller snags were cut and removed (90% of 5-10 inch dbh snags, 7580% of 10–20 inch snags), but only 40% of snags >20 inches dbh were cut and removed. None
were removed for firewood or consumed in a prescribed fire.
26
27
Down Wood
Pretreatment log loads, 1.8 tons/acre of 9-19.9 inch logs, and 3.3 tons/acre of ≥20 inch
logs, were low compared to suggested targets, i.e., 10-15 tons/acre for west side forests(USDA
Forest Service, 2004). Most stands (71 of 107) had loads of ≥20 inch diameter logs that were
below the target value (Table 6). Log loads varied greatly among plots; a few plots had high
loads but more than half of the plots (which were 3/4 acre) had no ≥20 inch logs, resulting in
median values that were much smaller than means. Log loads did not differ among west side,
transition, and east side forest.
Table 6—Down log (≥ 10 inches diameter) loads pre- and post-treatment by
ecological zone; Plumas, Lassen, and Tahoe (Sierraville ranger district) national
forests
na
Zone
≥20 inches
10–19.9 inches
Pre
Postb
Pre
Postb
Target
Pre
tons per acre
Postb
per cent
East side
56
1.6 (2.1)
1.1 (1.6)**
2.5 (4.1)
1.3 (2.7)**
48
25**
Transition
5
1.4 (1.4)
1.5 (1.1)
4.8 (5.6)
5.5 (7.0)
20
40
West side
46
1.5 (1.6)
1.1 (1.3)*
2.8 (4.3)
1.5 (2.5)**
15
2**
All
107
1.6 (1.8)
1.1 (1.4)***
2.7 (4.2)
1.6 (3.0)***
33
16**
a
number of units
p< 0.05, **p< 0.01, ***p<0.001; significance values for pre vs. post comparison
b*
After treatment, 88 of 107 stands were below the target range for ≥20 inch diameter logs.
Mean attainment after treatment was 16% (= 17/107) (95% confidence interval 10%–24%). The
overall trend was for decreased log loads after treatment, but the data provided limited support
for the idea that ≥20” logs are being maintained under certain conditions. Mean loads of both
small and large logs decreased, highly statistically significantly, on plots that were not burned
27
28
before the post-treatment survey (Table 6), and small log loads decreased on plots that were
burned. However, large-log loads did not change significantly in plots that were under-burned.
Understory Plants
Pre-treatment forb and shrub cover were higher on the west side than the east side (3.3%
vs. 1.8 %, and 18.7% vs. 12.5%, respectively; Table 7). There was little difference in grass cover
3.2% vs. 4.4%, west side vs. east side. There was a weak but statistically significant relationship
(p = 0.013, r2 = 0.13) between forb cover (sum of all species) and tree canopy cover for west side
plots; as tree canopy cover increased, forb cover decreased (Fig. 5). This relationship was not
significant for forbs in east side stands. There was a similar relationship (p = 0.018, r2 = 0.10)
between shrub cover and tree canopy cover for east side plots. The formula of the relationship
for west side forbs was Forb Cover = 15 – 0.17*Percent Tree Canopy Cover, and the formula for
east side shrub cover was Shrub Cover = 22 - 0.19*Percent Tree Canopy Cover. There was no
relationship between graminoid cover and tree canopy cover.
Understory Plant Cover (%)
70
Forbs
60
Graminoids
Shrubs
east side
west side
predicted west side
predicted east side
50
40
30
20
10
0
0
20
40
60
80
0
20
40
60
80
0
20
40
60
Tree Cover (%)
Figure 5.Cover of forbs, graminoids, and shrubs (pre-treatment) in relation to
overstory tree canopy cover. There was a significant relationship between forb
28
80
29
cover and tree cover on the west side (dotted line), and between shrub cover and
tree cover on the east side (solid line).
Treatments more than doubled forb cover in eastside treatment units (Table 7).
Graminoid cover increased significantly, by more than 50%, in both east side and west side units
after treatment. There was a slight decrease in shrub cover in west side units after treatment,
which was statistically significant (p = 0.011).There were no statistically significant relationships
between change in tree canopy cover and change in plant life form cover.
Table 7—Percent cover of forbs, graminoids, and shrubs (mean with standard
deviation) before and after treatment according to ecological zone (east vs.
westside) in Plumas, Lassen, and Tahoe (Sierraville ranger district) national
forests
Zone
n
Forb
Pre
Graminoid
Post
Pre
Post
Shrub
Pre
Post
percent cover
East side
56
1.8 (2.3)
4.3 (4.1)*
4.4 (6.5)
7.0 (6.6)**
12.5 (8.9)
12.2 (10.1)
Transition
5
0
0.4 (0.6)
1.0(2.2)
1.2 (2.7)
7.9(16.1)
7.9 (11.0)
West side
46
3.3 (7.0)
3.9 (5.5)
3.2 (7.1)
5.9 (6.9)*
18.7 (15.8)
16.1 (13.0)*
All
107
2.4 (4.9)
3.9 (4.7)*
3.8 (6.6)
6.3 (6.7)***
14.9 (13.0)
13.7 (11.6)
b*
p< 0.05, **p< 0.01, ***p<0.001; significance values for pre vs. post comparison
Owl Habitat
Of 47 units identified as potential owl habitat, 37 were in west side, eight were in east
side, and two were in transitional ecological zone. Mean (with standard deviation in parentheses)
pre-treatment weight of large logs was 2.08 (2.88) tons/acre (Fig. 6), and post-treatment weight
was 1.14 (1.76). The mean difference was 0.94 (2.11) tons/acre which was significantly different
than pre-treatment (paired t-test, p< 0.01). Mean density of snags pre-treatment was 3.64 (4.55)
29
30
snags/acre, and post-treatment density was 2.43 (3.14) snags/acre, with a mean difference of 1.21
(3.72) which was significantly different than pre-treatment (paired t-test, p< 0.05).Mean pretreatment stand basal area was 182.8 (62.6)ft2/acre and post-treatment basal area was 133.5
(50.8), with a mean difference of 49.3 (42.7) which was significantly different than pre-treatment
(paired t-test, p< 0.001).
1.5
1.0
0.5
5
4
3
2
1
0
Canopy Cover (%)
Large tree density (#/acre)
0.0
Basal area (ft /acre)
4
200
2
2.0
250
3
2
1
150
100
50
0
0
70
20
18
16
14
12
10
8
6
4
2
0
60
QMD (inches)
2.5
5
Pre-treatment
1 yr post-treatment
Snag density (#/acre)
Log weight (t / acre)
3.0
50
40
30
20
10
0
Figure 6. Large log weight, snag density, basal area, large tree density, canopy
cover, and quadratic mean diameter (QMD) pre- and post-treatment in stands
identified as potential owl habitat. Means of 47 plots with standard error bars.
Forty of the 47 stands contained at least one large tree within the plots either before or
after treatment. Mean pre-treatment large tree density was 3.6 (4.3) trees/acre, and post-treatment
density was 3.9 (4.2) trees/acre (Fig. 6). The difference was not statistically significant (paired ttest, p = 0.09). The distribution of data was highly skewed because most stands had a few large
30
31
trees but a few stands had many large trees. Pre-treatment canopy cover was 59.2 (12.5)%, and
post-treatment canopy cover was 40.4 (11.8)%. Mean canopy cover difference was 14.8 (11.6)%,
which was statistically significant (p< 0.001). Mean quadratic mean diameter increased from
13.6 (2.6) inches before thinning to 17.5 (4.0) after thinning (standard deviation in parentheses),
consistent with removal of small-diameter trees.
The California wildlife habitat relationship codes indicated many stands changed from
dense to moderate cover, or from moderate to sparse cover, after treatment (Table 8). There were
few changes in size class or forest type due to treatment. One stand changed from Sierra mixed
conifer to an aspen type due to the aspen restoration treatment.
Table 8—Pre- to post-treatment transitions in components of California Wildlife
Habitat Relationship codes in stands identified as potential spotted owl habitat
Canopy Cover
Pre
Size Class
Post
Pre
D
M
P
S
Da
1
12
6
0
M
0
8
15
0
P
0
0
3
2
Forest Type
Post
Pre
3
4
5
3b
3
4
0
4
0
37
3
Post
ASP
PPN
SMC
WFR
PPNc
0
2
0
0
SMC
1
0
24
0
WFR
0
0
1
19
a
Canopy cover codes: Dense (60-100%), Moderate (40-59%), oPen (25-39%), Sparse (10-24%)
Size class: 3 pole tree (5-10.9” QMD), 4 small tree (11.0 – 23.9” QMD), 5 med/large tree (24” or
greater QMD)
c
Forest type: ASP (aspen), PPN (ponderosa pine), SMC (Sierran mixed-conifer), WFR (white fir)
b
Fire Behavior Modeling
Fire behavior modeling indicated that prescriptions were generally effective in changing
fire behavior (Table 9).In the 97th percentile weather scenarios, the number of stands predicted to
31
32
have active or conditional crown fire if ignited decreased from 58 to 20. The number of stands
with surface fire regime almost doubled, from (30 to 59), and the number of stands with passive
crown fire types predicted increased by 50%. Trends under 90th percentile weather conditions
were similar.
Table 9—Fire type predicted from fire behavior modeling, pre- and posttreatment. Numbers represent number of stands (of 107).
90th percentile
97th percentile
Pre
Post
Pre
Post
Surface
36
77
30
59
Passive
25
21
19
28
Conditional crown
26
5
26
9
Active crown
20
4
32
11
Changes in fire type were driven by characteristics of aerial rather than surface fuels.
Estimated crown bulk density decreased significantly with treatment, and estimated canopy base
height, an important factor in conveying fire from the forest floor to the canopy, increased
significantly (Table 10).In contrast, surface flame lengths did not change significantly with
treatment (Table 11), despite significantly increased mid-flame wind spreads (estimated as a
function of canopy cover).Increased estimated surface winds may have been mitigated by surface
fuel models, many of which were reassigned after treatment. Model 10 (Timber: litter and
understory) was used for 66 (of 107) pre-treatment stands, but only 28 post-treatment stands.
Model 9 (named „Hardwood litter‟ but which also has terms for conifer litter) was used for only
30 pre-treatment stands, increasing to 58 post-treatment stands.
32
33
Table 10—Canopy metrics and fuel models used in fire behavior simulations,
before (P0) and after (P1) fuels-reduction thinning and group selection treatment
(means with standard deviation in parentheses).
Zone
n
Canopy bulk
Canopy base
density
height
P1a
P0
P0
kg/m3
East
56
side
Trans.
West
5
46
side
All
107
P1
Fuel Modela
2
P0
5
P1
P0
8
P1
P0
feet
0.137
0.053***
9.1
14.6***
(0.098)
(0.037)
(6.5)
(10.4)
0.265
0.138***
6
19.2**
(0.115)
(0.075)
(2.5)
(14.8)
0.165
0.086***
6.9
13.9***
(0.086)
(0.05)
(5.4)
(12.9)
0.155
0.071***
8.0
14.5***
(0.097)
(0.049)
(6.0)
(11.7)
9
P1
10
P0
P1
P0
P1
n
a
3
11
3
0
0
5
21
31
29
9
0
0
0
0
0
0
1
4
4
1
0
0
5
1
0
4
8
23
33
18
3
11
8
1
0
9
30
58
66
28
Fuel models: 2 is Timber (grass and understory), 5 is Brush, 8 is Closed timber litter, 9 is
Hardwood litter, 10 is Timber (litter and understory). Data represent number of plots for which a
fuel model was used.
b**
p< 0.01, ***p<0.001; significance values for pre vs. post comparison
33
34
Table 11—Mid-flame wind speed and surface flame length estimated in fire
behavior simulations under 90th and 97th percentile weather scenarios, before and
after fuels-reduction thinning and group-selection treatment. Means with
standard deviations in parentheses.
Zone
n
Mid-flame wind speed
90th
Surface flame length
97th
90th
97th
mph
East side
Transition
West side
All
a**
56
5
46
107
ft
Pre
Posta
Pre
Post
Pre
Post
Pre
Post
2.24
3.00***
2.57
3.62***
4.00
3.97
4.58
4.65
(0.76)
(0.88)
(1.01)
(1.09)
(1.32)
(2.09)
(1.52)
(2.45)
1.36
2.26**
1.36
2.69***
3.50
3.04
3.94
3.53
(0.64)
(0.68)
(0.64)
(0.77)
(1.21)
(1.35)
(1.30)
(1.41)
2.78
3.51***
2.92
4.20***
4.92
4.13
5.71
4.90
(0.81)
(1.08)
(0.79)
(1.29)
(1.28)
(1.65)
(1.46)
(1.88)
2.43
3.19
2.66
3.83
4.24
4.03
4.49
4.73
(0.85)
(1.01)
(0.96)
(1.22)
(1.41)
(1.60)
(1.34)
(1.81)
p< 0.01, ***p< 0.001; significance values for pre vs. post comparison.
Discussion
The untreated forest
Quantitative estimates of pre-treatment forest attributes were similar in most respects to
those from similar untreated forests nearby. Canopy cover in untreated stands, about 55% in
west-side stands, was consistent with other estimates of canopy cover in untreated, firesuppressed mixed conifer stands (e.g., 63-69% canopy cover in untreated stands at a productive
34
35
site in the central Sierra Nevada; Stephens and Moghaddas, 2005a). Mean pre-treatment largelog loads however, 3.2 tons/acre, was somewhat low. For comparison, the weight of dead fuels
(≥ 6 inches diameter) in untreated forest was 7.4 tons/acre at Blodgett Forest in the central Sierra
Nevada, and 23.2 tons/acre (≥ 3inches diameter) at Teakettle Experimental Forest in the southern
Sierra Nevada (Table 12).
Table 12—Estimates of coarse woody debris load from mixed-conifer forests in
California and Mexico
Location
Min.
Max.
Min.
diameter
diameter
length
inches
inches
feet
tons/acre
Sequoia NPa
3
∞
3.3
27.4
Blodgett Forestb
6
∞
3.3
7.4
Teakettle EFc
3
11.9
0
2.4
Teakettle EFc
12
23.9
6.6
1.8
Teakettle EFc
24
35.9
6.6
4.0
Teakettle EFc
36
∞
6.6
15.3
Plumas, Lassen, and Tahoe NFd
3
9
0
2.1
Plumas, Lassen, and Tahoe NFd
10
19.9
10
1.6
Plumas, Lassen, and Tahoe NFd
20
∞
10
2.7
Plumas and LassenNFg
3
9.9
0
3.8 (2.9)
Plumas and Lassen NFg
10
19.9
0
4.8 (3.2)
Plumas and Lassen NFg
20
∞
0
8.5 (5.9)
Sierra San Pedro Martire
3
∞
0
6.0
35
Load
36
Yosemite NPf
a
∞
3
0
19.6
Southern Sierra Nevada, old-growth mixed conifer forest, fire-suppressed for 120 years
(Knapp et al., 2005)
b
Central Sierra Nevada, managed mixed conifer, fire-suppressed for 90 years, (Stephens and
Moghaddas, 2005b)
c
EF denotes Experimental Forest, Southern Sierra Nevada, old-growth mixed conifer, no fire for
140 years (Innes et al., 2006)
d
Northern Sierra Nevada and southern Cascades (present study), managed mixed conifer and
white fir forests (sound wood only)
e
Northwest Mexico, old-growth Jeffrey-pine mixed conifer, natural fire regime (Stephens, 2004)
f
Central Sierra Nevada, mixed-conifer, active fire regime. Includes sound and rotten wood (B.
Collins, unpublished data)
g
Data collected by K. Menning and S. Stephens. Sound and rotten wood; sound-only values in
parentheses
Because of the century and a half-long history of intensive resource extraction from the
forest (Fig. 1), and what is generally conceded to have been a misguided fire suppression policy
(Stephens and Ruth, 2005), pre-treatment data reflect a forest that is degraded compared to the
pre-European condition. This degradation is manifested as high density and canopy cover of
small-diameter trees, few large trees, low understory plant cover and diversity, low numbers of
snags and large down logs, and volatile fire behavior under moderate to extreme weather
conditions. Pre-settlement reference conditions for down dead wood in frequent-fire dry forest
are uncertain particularly because the accumulation of large-diameter down wood would have
been limited by the low-intensity fires that formerly burned at 8-22 year intervals over much of
the area now making up Plumas and Lassen national forests (Moody et al., 2006). Still, given the
36
37
formerly much higher density of large trees on the landscape, it may be reasonably assumed that
large wood loads were considerably greater than under present conditions despite the legacy of a
century of fire suppression management. The Pilot Project has sought to restore forest structure
and function to pre-European-settlement conditions, while increasing commercial resource
extraction beyond the very low levels of the early 1990‟s (fig. 1). Monitoring of the short-term
effects of the Pilot Project implementation shows areas both of improvement and of further
degradation in forest condition. It also documents the ability of managers to achieve management
targets and suggests some areas in which current practice may be modified to enhance
management outcomes.
The monitoring scheme in the present report departs from standard research designs in
that there are no controls, only measurements before and after treatments. Therefore,
measurements reflect not only the effects of management treatments, but also changes taking
place naturally over time. In general, the factors measured change slowly over time (e.g., canopy
cover; Bigelow and North, 2012), and we have assumed that the changes detected have been due
to management treatments. The monitoring scheme did not readily allow estimation of effects of
different practices (e.g., fuels-reduction-thinning vs. group selection), and there has been no
attempt to do so in the present analysis. An additional caveat is that many of the post-treatment
measurements were done on plots in which treatment was incomplete, insofar as the application
of prescribed fire for surface treatment of fuels has lagged far behind the mechanical treatment of
fuels. Post-treatment monitoring data reflect treatments as implemented rather than as planned.
Treatment effects on the landscape
Pilot Project treatments led to large decreases in canopy cover, but even stands reduced to
35% post-treatment canopy cover still remained above presumptive pre-settlement levels (cf.
37
38
Collins et al., 2011). The increase in the proportion of stands meeting canopy cover target
guidelines (from 14% to 45%) was substantial but provided room for improvement. The diameter
class analysis indicated that treatments succeeded in making the diameter-class distribution
flatter by removing many small and small-to-intermediate diameter trees while retaining
intermediate-to-large and large ones. Although making the diameter distribution flatter was not a
specific goal of the treatments, it is consistent with the Quincy Library Group vision of
approximating the pre-settlement landscape (Yost, 1994; North et al., 2007). The goal of
maintaining large-diameter trees was achieved, although density of such trees was very low
compared to the presumed pre-settlement levels. The third census, five years after treatments
have taken place, will allow an assessment of whether the larger goal of promoting recruitment
of large trees is achieved.
Snag target guidelines represent a compromise between preserving habitat values and
protecting firefighters; notably, managers were not directed to maintain current levels of snags in
project areas but simply to consider large snag retention guidelines during project planning
(USDA Forest Service, 2004, p.69).Still, the substantial decrease in proportion of stands in
attainment with goals for snag densities is a matter for concern. With only 17% of stands in
attainment of objectives the post-treatment landscape is seriously deficient in snags, a lack which
poses a problem for snag-dependent wildlife (Bull, 2002; Bunnell et al., 2002; Innes et al., 2008).
Viewed from the perspective of engineering DFPZs for firefighter safety the loss of snags is a
boon, because low-intensity fires burning through DFPZs may burn snags through at the base,
allowing them to fall over and thereby potentially threaten firefighters working in the area.
Just as with snags, an initial scarcity of large logs on the landscape was exacerbated by
treatment. In the short term, then, HFQLG treatments are degrading rather than restoring the
38
39
forest with respect to large log loads, although, again as with snags, the five-year post-treatment
census will provide a better measurement.
The life-form spectrum, or relative abundance of plant life-forms, is a reflection of
climate, topography, disturbance regime, and competition from neighboring vegetation
(Raunkiaer, 1934). Abundant shrub cover is typical of forests in Mediterranean climates, and
shrub cover, with mean 15% cover in untreated forestsin the present study, was considerably
higher than forb and graminoid cover. The proportion of understory cover contributed by shrubs
relative to graminoids typically becomes lower in drier regions such as the southwestern USA
(Abella and Covington, 2004), and on more exposed sites or at higher elevations (Gracia et al.,
2007; Coll, 2011). The lower shrub cover in east vs. west side sites in the present study was
consistent with this trend. Although forest treatments led to increased cover of graminoid and
forb lifeforms, the increases were small compared to those seen in other manipulative studies in
mixed-conifer forest. For example, thinning and burning treatments in the southern Sierra
Nevada caused a shift from a shrub-dominated to herb-dominated communities, with herb cover
increasing by a factor >10 (Wayman and North, 2007). Understory plant cover changes along
multi-decadal trajectories after treatment (Webster and Halpern, 2010), and five-year posttreatment censuses will undoubtedly be different from one-year post-treatment censuses reported
on herein. The weak or non-existent relationship between understory cover and tree canopy
cover is consistent with studies of other Mediterranean-climate forests (Gracia et al., 2007; Coll,
2011). If belowground resources, rather than light, are the factors that limit understory growth
most strongly in frequent-fire dry forests (e.g., Riegel et al., 1992) then understory growth
response to mechanical treatments may remain limited unless the stimulation of fire also occurs.
39
40
Preserving California spotted owl habitat is a major goal of management in the northern
Sierra Nevada (USDA Forest Service, 2004), yet treatments in potential habitat areas caused a
decline in many nesting habitat elements (tree canopy cover, snags, large down logs, and basal
area; Bias and Gutierrez, 1992). Large trees, a key element of California spotted owl nesting and
roosting habitat (Moen and Gutiérrez, 1997) were maintained at present density. Much has been
learned about the distinction between habitat used for nesting or roosting versus foraging in the
years since the guidelines for maintenance of spotted owl habitat were established (Verner et al.,
1992). Recent studies support the view that California spotted owls select nests and roosts in
forest patches containing large trees (e.g., ≥100 cm dbh; Moen and Gutiérrez, 1997) with high
canopy cover (Bias and Gutierrez, 1992; Blakesley et al., 2005) and many canopy strata (LaHaye
et al., 1997). In contrast, the characteristics of preferred foraging habitat are an area of continued
scientific controversy.
Several recent studies support the view that California spotted owls prefer to forage in
habitat that is much more open than their typically nesting habitat would suggest. Owls may
preferentially forage in large (>25 acre) patches dominated by pole-size trees (Williams et al.,
2011), may preferentially forage in post-high-severity-fire habitat when available (Bond et al.,
2009), and appear not to select foraging habitat based on canopy cover (Irwin et al., 2007).
Nevertheless, a telemetry study within the same area as the present report found that California
spotted owls avoided foraging in DFPZs (Gallagher, 2010). Because of the specificity of this
study to the forest and treatments of the present study, we conclude that treatments degraded owl
foraging as well nesting habitat, recognizing that there is somewhat less evidence for the former
than the latter.
40
41
The large predicted decrease in conditional and active crown fire behaviors under
moderate and extreme weather conditions suggests that this highly important function of the
treatments has been fulfilled properly. Elsewhere within the larger study area fuels treatments
have been empirically demonstrated to be effective against high-intensity fire (Moghaddas and
Craggs, 2007; Safford et al., 2012), which lends credence to the prediction that most of the
treated stands would not support a crown fire under the foreseeable range of weather conditions.
Despite these encouraging results, limitations of models prediction must be borne in mind given
that fire spread simulation models based on the Rothermel equations may substantially underpredict conditions under which the transition from surface to crown fire will occur(Cruz and
Alexander, 2010; Moghaddas et al., 2010). A further important factor is that treatments did not
decrease predicted surface fire behavior, possibly because any decreases in surface fuels (as
estimated visually in photo-series, and reflected in choice of fuel models) may have been
counterbalanced by increases in predicted understory wind velocity.
Management Implications
Increasing large dead wood loads
Large fallen logs are important resources for many wildlife species; for example, a
mesocarnivore prevalent in Plumas County 80 years ago but now locally extirpated, the
American marten(Clark et al., 1988), relies on large logs for winter resting sites. Although
anecdotal observations suggest that large down logs may increase the intensity of severe fires
(Rothermel, 1991), large logs tend to mainly combust by smoldering after the main fire front has
passed (van Wagtendonk, 2006), contributing little to the intensity of the fire. Increased potential
41
42
damage due to retaining large down logs would probably be negligible. Given the paucity of
large down logs on the Plumas-Lassen landscape, and their importance for wildlife, it seems
prudent to strive to protect all large logs during treatment operations. Such protection might take
the form of improved commitment to communication with loggers regarding the value of down
logs, specifying their retention in management contracts, flagging large logs, and/or planning
skid routes to avoid them.
Ultimately, to increase large down wood it is essential to increase the density of large
trees(Innes et al., 2006). HFQLG treatments did not diminish current densities of large trees or
the next smaller size class (24-29.9 inches). They were also effective at protecting large
treesfrom high-intensity fire, as indicated by the fire behavior modeling. Nevertheless, large tree
density is still low (mean density < 3 stems/acre) compared to all historical evidence. Further
actions might be to reduce competition around the 24-30 inch size class to enhance recruitment
to the ≥ 30 inch size class.
Spatial variability and scale in snag management
The high degree of patchiness of snag distribution shown both before and after treatment
in the present study has been observed in other dry forest ecosystems. For example, snag
densities were highly variable in a Jeffrey pine-mixed conifer forest that had little human
exploitation and a natural fire regime until recently. There, one-third of plots lacked snags and a
few plots had high snag densities (Stephens, 2004). The role of high-density snag areas for
wildlife is uncertain, but dense aggregations of burned snags are attractive to species such as the
black-backed woodpecker (Burnett et al., 2011).Forest managers may need to consider not only
how to maintain the appropriate density of snags on the landscape, but also the appropriate level
of variability.
42
43
Stephens (2004)recommended managing for snags at the scale of hundreds of acres rather
than at the per acre scale: in other words, large areas of few or no snags might be balanced by
small patches of high snag density. A prescription incorporating this concept might be worded as
“retain 200 snags per 100 acres” as opposed to “retain an average of 2 snags per acre”.
Implementation of such an approach might require designating areas of high and low snag
density within a project area, an idea perhaps implied in the Sierra Nevada Forest Plan
Amendment Record of Decision (USDA Forest Service, 2004 p. 69) which suggests “leaving
fewer snags strategically located in treatment areas within the WUI and DFPZs”. Nevertheless it
is unclear whether the intent is that low snag densities should be offset by higher densities
outside the DPFZ networks.
Given low initial snag densities, if any snags are to be felled during treatment activities
then consideration must be given to snag recruitment. Prescribed fire is likely to be the most
effective means for this purpose. Intentional tree mortality might be achieved by intentionally
scorching and/or torching isolated trees, or trees in small groups, with a hot fire. Trees with thick
accumulations of duff at their bases may be vulnerable to backing fires with long flame residence
times(Bova and Dickinson, 2005).
Tree canopy and understory plant cover
Canopy cover influences many forest attributes including understory fire climate (Albini
and Baughmann, 1979; Bigelow and North, 2012) and wildlife habitat. Basal area was analyzed
as a function of canopy cover in the present study because canopy cover targets were reached
with only moderate success, and we wished to provide an alternative target for field personnel
more accustomed to estimating basal area. Our data suggest that multiplying desired canopy
cover by 2.8 yields an appropriate target basal area for stands expected to have post-treatment
43
44
QMD <18”, and multiplying canopy cover by 3.5 is appropriate for stands expected to have posttreatment QMD ≥ 18”. Vaughn and Ritchie (2005) analyzed data from 37 areas of interior
ponderosa forest in northern California that had many species in common with the present study.
They found no effect of thinning or fire on the basal area-canopy cover relationship. We reanalyzed their data, making basal area a linear function of canopy cover, and determined that
multiplying canopy cover by 3.5 gave the target basal area (p< 0.001, R2 = 0.96).The close
agreement between Vaughn and Ritchie (2005) and the present study suggest that conversion of
canopy cover targets to basal area targets may be an effective approach and merits further
evaluation.
Despite the obvious importance of tree canopy cover for understory light and other
elements of microclimate (Bigelow et al., 2011), the monitoring data did not show a strong tree
canopy influence on understory vegetation. The weak correlation in untreated stands between
tree and shrub cover (east side; R2 0 =0.10) and between tree and forb cover (west side; R2 =
0.13) suggests that tree canopy cover has limited utility at best in predicting understory plant
cover. (We did not attempt to relate one-year post-treatment tree canopy cover to understory
cover, because recent harvest disturbance is expected to uncouple these factors.)We conjecture
that the explanation for the weak understory-overstory relationship is that other factors (e.g., soil
water) may limit understory growth more strongly than light(e.g., Riegel et al., 1992).This notion
is consistent with the observation that that treatments lead to increases in cover of most
understory plant lifeforms, but the increase was not proportional to the change in canopy cover.
Implementing prescribed fire
Prescribed fire is essential to a number of goals and targets of the HFQLG program. It
can be an effective agent for recruiting snags, and stimulates growth of understory plants, which
44
45
is desirable for forage production and forest health (Wayman and North, 2007). Prescribed fire
kills young white fir much more readily than young ponderosa or Jeffrey pine(van Mantgem and
Schwartz, 2003), thus reducing ladder fuels, and restoring the historic tree species composition.
Finally, prescribed fire is uniquely effective at reducing surface fuels, particularly fine fuels,
which are a major factor in elevating crown fire potential(Rebain et al., 2011). Nevertheless, the
current monitoring study points to a lack of firm institutional commitment to timely application
of prescribed fire. One third of the units with prescribed under-burns planned (54 of 107 units)
had not received them by early 2012, and the mean time elapsed since mechanical treatment of
these yet-unburned units was 7 years. Given the centrality of prescribed fire to achieving the
objectives of the HFQLG and the broader regional forest management (USDA Forest Service,
2004),much more consistent application of prescribed fire is called for.Fire that is prescribed
should be applied.
Owls and Fire Simulation Modeling
In the present monitoring study, conclusions on owls and fire are primarily derived from
mathematical or conceptual models and the strength of conclusion is directly related to strength
of the model. In the case of the California spotted owl, requirements for nesting and roosting
habitat are well understood, but important questions remain regarding foraging habitat. In the
future, monitoring questions must acknowledge the difference between nesting and foraging
habitat, perhaps making an a priori declaration of which area is potential nesting habitat versus
which is potential foraging habitat.
For wildfire, which has well-established, predictive mathematical models, there is still
some uncertainty. Empirical evidence increasingly indicates the effectiveness of fuels treatments
(Safford et al., 2012), yet there is still doubt about the ability to accurately predict conditions
45
46
under which crown fires may occur. Standard fuel models are not sensitive indicators of
understory fuel loads. Reliance on models for management runs the risk of focusing excessively
on eliciting particular model behaviors („managing the model‟) rather than ecosystem behaviors.
Conclusions
The HFQLG „Pilot Project‟ program features fuels-reduction thinning in a system of
defensible fuel profile zones, and area-regulated group selection with reserves. It has taken place
in Plumas, Lassen, and Tahoe (Sierraville ranger district) national forests from 1999 to the
present day. Monitoring of short-term effects of the Pilot Project has examined the ability of
treatments both to achieve explicit quantitative management targets, and the implicit goal of
moving the landscape closer to a pre-settlement reference structure. Canopy cover targets were
achieved 45% of the time, but snag and down wood targets were almost impossible to achieve
given that starting levels of dead wood were generally much lower than „retention‟ targets.
Treatments had only minor effect on understory plant cover hence little effect on available
forage. Treatments successfully avoided large trees (and intermediate trees also to a great extent)
thus maintaining a key element both for California spotted owl habitat, and for landscape-level
fire resistance. Ultimately, treatments were predicted to be largely effective at decreasing the
probability of crown fire, a key purpose of the project.
46
47
Acknowledgments
Jo Ann Fites established the study design and Ron O‟Hanlon and Wayne Johanssen
coordinated the early implementation of the study. We thank Jim Baldwin for statistical advice,
the Quincy Library Group for their support, Rhonda Hammer for database creation and
maintenance, Nicole Vaillant for fire modeling and guidance, and Mike Landramfor data on cut
timber volume for 1909-1966. For critiques that lead to improvements in this report we thank
Claire Gallagher, Tom Rickman, Mark Williams, and Craig Wilson. Thanks to the many people
who have collected data over the past 10 years of the treated stand structure monitoring program,
especially crew leaders Tim Taylor, Bob Jiron and Beth Stewart. We thank the wildlife biologists
who provided assessments of habitat suitability: Russell Nickerson, Tom Rickman, Cindy
Roberts, Gary Rotta, Mark Williams and Craig Wilson.
47
48
Literature Cited
Abella, S.R., Covington, W.W., 2004. Monitoring an Arizona ponderosa pine restoration:
sampling efficiency and multivariate analysis of understory vegetation. Restoration
Ecology 12, 359-367.
Agee, J.K., Bahro, B., Finney, M.A., Omi, P.N., Sapsis, D.B., Skinner, C.N., van Wagtendonk,
J.W., Weatherspoon, C.P., 2000. The use of shaded fuelbreaks in landscape fire
management. Forest Ecol. Manag. 127, 55-66.
Agee, J.K., Skinner, C.N., 2005. Basic principles of forest fuel reduction treatments. Forest Ecol.
Manag. 211, 83-96.
Agresti, A., 1990. Categorical Data Analysis. Wiley, New York.
Albini, F.A., 1976. Estimating wildfire behavior and effects. General Technical Report INTGTR-30, USDA Forest Service, Intermountain Forest and Range Experiment Station,
Ogden, Utah.
Albini, F.A., Baughmann, R.G., 1979. Estimating windspeeds for predicting wildland fire
behavior. Research Report RP-INT-221, U. S. Forest Service, Intermountain Forest and
Range Experiment Station, Ogden, UT.
Beaty, R.M., Taylor, A.H., 2001. Spatial and temporal variation of fire regimes in a mixed
conifer landscape, southern Cascades, California, USA. J. Biogeogr. 28, 955-966.
Bernard, T., 2010. Hope and Hard Times: Communities, Collaboration and Sustainability. New
Society Publishers, Gabriola Island, British Columbia, Canada.
Bernard, T., Young, J.M., 1996. The Ecology of Hope: Communities Collaborating for
Sustainability. New Society, East Haven, Connecticut.
Bias, M.A., Gutierrez, R.J., 1992. Habitat associations of California spotted owls in the central
Sierra Nevada. J. Wildlife Manage. 56, 584-595.
Bigelow, S.W., North, M., 2012. Microclimate effects of fuels-reduction and group-selection
silviculture: implications for fire behavior in Sierran mixed-conifer forests. Forest Ecol.
Manag. 264, 51-59.
Bigelow, S.W., North, M.P., Horwath, W.R., 2009. Models of resource-dependent growth for
Sierran mixed-conifer saplings. The Open Forest Sci. J. 2, 31-40.
Bigelow, S.W., North, M.P., Salk, C.F., 2011. Using light to predict fuels-reduction and groupselection effects on succession in Sierran mixed-conifer forest. Can. J. Forest Res. 41,
2051-2063.
Biging, G.S., 1984. Taper equations for second-growth mixed conifers of Northern California.
Forest Sci. 30, 1103-1117.
Blakesley, J.A., Noon, B.R., Anderson, D.R., 2005. Site occupancy, apparent survival, and
reproduction of California spotted owls in relation to forest stand characteristics. J.
Wildlife Manage. 69, 1554-1564.
Bond, M.L., Lee, D.E., Siegel, R.B., Ward, J.P., 2009. Habitat use and selection by California
spotted owls in a postfire landscape. J. Wildlife Manage. 73, 1116-1124.
Bond, M.L., Seamans, M.E., Gutiérrez, R.J., 2004. Modeling nesting habitat selection of
California spotted owls (Strix occidentalis occidentalis) in the central Sierra Nevada
using standard forest inventory metrics Forest Sci. 50, 773-780.
Bova, A.S., Dickinson, M.B., 2005. Linking surface-fire behavior, stem heating, and tissue
necrosis. Can. J. Forest Res. 35, 814-822.
48
49
Brown, J.K., 1974. Handbook for inventorying downed woody material. General Technical
Report INT-GTR-16, U.S. Forest Service, Intermountain Forest and Range Experiment
Station, Ogden, Utah.
Buchanan, J.B., Irwin, L.L., McCutchen, E.L., 1995. Within-stand nest-site selection by spotted
owls in the eastern Washington Cascades. J. Wildlife Manage. 59, 301-310.
Bull, E.L., 2002. The value of coarse woody debris to vertebrates in the Pacific Northwest. In:
Laudenslayer, W.F., Shea, P.J., Valentine, B.E., Weatherspoon, C.P., Lisle, T.E. (Eds.)
General Technical Report PSW-GTR-181, U.S. Forest Service, Pacific Southwest
Research Station, Albany, California.
Bunnell, F.L., Houde, I., Johnston, B., Wind, E., 2002. How dead trees sustain live organisms in
western forests. In: Laudenslayer, W.F., Shea, P.J., Valentine, B.E., Weatherspoon, C.P.,
Lisle, T.E. (Eds.) General Technical Report PSW-GTR-181, U.S. Forest Service, Pacific
Southwest Research Station, Albany, California.
Burnett, R.D., Taillie, P., Seavy, N., 2011. Plumas-Lassen Administrative Study 2010 Post-fire
Avian Monitoring Report. PRBO Conservation Science.
Call, D.R., Gutierrez, R.J., Verner, J., 1992. Foraging habitat and home-range characteristics of
California spotted owls in the Sierra-Nevada. Condor 94, 880-888.
Clark, T.W., Anderson, E., Douglas, C., Strickland, M., 1988. Martes americana. Mammalian
Species 289, 1-8.
Coll, L., 2011. Predicting understory maximum shrubs cover using altitude and overstory basal
area in different Mediterranean forests. European Journal of Forest Research 130, 55-65.
Collins, B.M., Everett, R.G., Stephens, S.L., 2011. Impacts of fire exclusion and recent managed
fire on forest structure in old growth Sierra Nevada mixed-conifer forest. Ecosphere 2, 114.
Cruz, M.G., Alexander, M.E., 2010. Assessing crown fire potential in coniferous forests of
western North America: a critique of current approaches and recent simulation studies.
International Journal of Wildland Fire 19, 377-398.
Gallagher, C.V., 2010. Spotted Owl Home Range and Foraging Patterns Following FuelsReduction Treatments in the Northern Sierra Nevada, California. Master's Thesis,
University of California, Davis, California.
Gill, S.J., Biging, G.S., Murphy, E.C., 2000. Modeling conifer tree crown radius and estimating
canopy cover. Forest Ecol. Manag. 126, 405-416.
Gracia, M., Montané, F., Piqué, J., Retana, J., 2007. Overstory structure and topographic
gradients determining diversity and abundance of understory shrub species in temperate
forests in central Pyrenees (NE Spain). Forest Ecol. Manag. 242, 391-397.
HFQLG, 1998. The Herger-Feinstein Quincy Library Group Forest Recovery Act: Pilot Project
for Plumas, Lassen, and Tahoe National Forests to Implement Quincy Library Group
Proposal. P.L. 103-354, Section 401(j), October 1998.
Innes, J.C., North, M.P., Williamson, N., 2006. Effect of thinning and prescribed fire restoration
treatments on woody debris and snag dynamics in a Sierran old-growth, mixed-conifer
forest. Can. J. Forest Res. 36, 3183-3193.
Innes, R.J., Van Vuren, D.H., Kelt, D.A., 2008. Characteristics and use of tree houses by duskyfooted woodrats in the northern Sierra Nevada. Northwestern Naturalist 89, 109-112.
Irwin, L.L., Clark, L.A., Rock, D.C., Rock, S.L., 2007. Modeling foraging habitat of California
spotted owls. J. Wildlife Manage. 71, 1183-1191.
49
50
Jennings, S.B., Brown, N.D., Sheil, D., 1999. Assessing forest canopies and understorey
illumination: canopy closure, canopy cover and other measures. Forestry 72, 59-73.
Knapp, E.E., Keeley, J.E., Ballenger, E.A., Brennan, T.J., 2005. Fuel reduction and coarse
woody debris dynamics with early season and late season prescribed fire in a Sierra
Nevada mixed conifer forest. Forest Ecol. Manag. 208, 383-397.
LaHaye, W.S., Gutiérrez, R.J., Call, D.R., 1997. Nest-site selection and reproductive success of
California spotted owls. Wilson Bulletin 109, 42-51.
Leiberg, J.B., 1902. Forest Conditions in the northern Sierra Nevada, California. Professional
Paper No. 8, Series H, Forestry, 5, US Government Printing Office, Washington DC. my
files.
Mayer, K.E., Laudenslayer, W.F. (Eds.), 1988. A Guide to Wildlife Habitats of California.
California Department of Forestry and Fire Protection, Sacramento California.
Miller, J.D., Safford, H.D., Crimmins, M., Thode, A.E., 2008. Quantitative evidence for
increasing forest fire severity in the Sierra Nevada and southern Cascade Mountains,
California and Nevada, USA. Ecosystems 12, 16-32.
Moen, C.A., Gutiérrez, R.J., 1997. Spotted owl habitat selection in the Central Sierra Nevada.
The Journal of Wildlife Management 61, 1281-1287.
Moghaddas, J., Craggs, L., 2007. A fuel treatment reduces fire severity and increases suppression
efficiency. International Journal of Wildland Fire 16, 673-678.
Moghaddas, J.J., Collins, B.M., Menning, K., Moghaddas, E.E.Y., Stephens, S.L., 2010. Fuel
treatment effects on modeled landscape-level fire behavior in the northern Sierra Nevada.
Can. J. Forest Res. 40, 1751-1765.
Moghaddas, J.J., York, R.A., Stephens, S.L., 2008. Initial response of conifer and California
black oak seedlings following fuel reduction activities in a Sierra Nevada mixed conifer
forest. Forest Ecol. Manag. 255, 3141-3150.
Moody, T.J., Fites-Kaufman, J., Stephens, S.L., 2006. Fire history and climate influences from
forests in the northern Sierra Nevada, USA. Fire Ecology 2, 115-141.
North, M., Innes, J., Zald, H., 2007. Comparison of thinning and prescribed fire restoration
treatments to Sierran mixed-conifer historic conditions. Can. J. Forest Res. 37, 331-342.
Ott, L., 1989. An Introduction to Statistical Methods and Data Analysis. PWS-Kent, Boston.
R Development Core Team, 2008. R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria.
Raunkiaer, C., 1934. The Life Forms of Plants and Statistical Plant Geography. Clarendon Press,
Oxford.
Rebain, S., Reinhardt, E.D., Crookston, N.L., Beukema, S.J., Kurz, W.A., Greenough, J.A.,
Robinson, D.C.E., Lutes, D.C., 2011. The fire and fuels extension to the forest vegetation
simulator. Updated model documentation, U. S. Forest Service, Forest Management
Service Center, Fort Collins, CO.
Riegel, G.M., Miller, R.F., Krueger, W.C., 1992. Competition for resources between understory
vegetation and overstory Pinus ponderosa in northeastern Oregon. Ecol. Appl. 2, 71-85.
Rothermel, R.C., 1972. A mathematical model for predicting fire spread in wildland fuels.
Research Paper INT-RP-115, U.S. Forest Service, Intermountain Forest and Range
Experiment Station, Ogden, Utah.
Rothermel, R.C., 1991. Predicting behavior and size of crown fires in the northern Rocky
mountains. Res. Pap. INT-RP-438, U.S. Department of Agriculture, Forest Service,
Intermountain Research Station, Ogden, UT.
50
51
Safford, H.D., Stevens, J.T., Merriam, K., Meyer, M.D., Latimer, A.M., 2012. Fuel treatment
effectiveness in California yellow pine and mixed conifer forests. Forest Ecol. Manag.
274, 17-28.
Stephens, S.L., 2004. Fuel loads, snag abundance, and snag recruitment in an unmanaged Jeffrey
pine-mixed conifer forest in Northwestern Mexico. Forest Ecol. Manag. 199, 103-113.
Stephens, S.L., Moghaddas, J.J., 2005a. Experimental fuel treatment impacts on forest structure,
potential fire behavior, and predicted tree mortality in a California mixed conifer forest.
Forest Ecol. Manag. 215, 21-36.
Stephens, S.L., Moghaddas, J.J., 2005b. Fuel treatment effects on snags and coarse woody debris
in a Sierra Nevada mixed conifer forest. Forest Ecol. Manag. 214, 53-64.
Stephens, S.L., Ruth, L.W., 2005. Federal forest-fire policy in the United States. Ecol. Appl. 15,
532-542.
USDA Forest Service, 1999. Record of Decision: USDA Forest Service Final Environmental
Impact Statement Herger-Feinstein Quincy Library Group Forest Recovery Act.
http://www.fs.fed.us/r5/hfqlg/publications/1999_feis/Record_of_Decision.htm.
USDA Forest Service, 2004. Record of Decision: Sierra Nevada Forest Plan Amendment Final
Supplemental Environmental Impact Statement. Management Bulletin R5-MB-046, U.S.
Forest Service, Pacific Southwest Region, Vallejo, California.
http://www.fs.fed.us/r5/snfpa/final-seis/.
van Mantgem, P., Schwartz, M., 2003. Bark heat resistance of small trees in Californian mixed
conifer forests: testing some model assumptions. Forest Ecol. Manag. 178, 341-352.
van Wagtendonk, J.W., 2006. Fire as a physical process. In: Sugihara, N.G., van Wagtendonk,
J.W., Shaffer, K., E., Fites-Kaufman, J.A., Thode, A.E. (Eds.), Fire in California's
Ecosystems. University of California Press, Berkeley, pp. 38-57.
Vaughn, N., Ritchie, M.W., 2005. Estimation of crown cover in interior ponderosa pine stands:
effects of thinning and prescribed fire. Western Journal of Applied Forestry 20, 240-246.
Verner, J., McKelvey, K.S., Noon, B.R., Gutiérrez, R.J., Gould, G.I., Beck, T.W., 1992. The
California Spotted Owl: A Technical Assessment of Its Current Status. PSW-GTR-133,
Pacific Southwest Research Station, Albany, California.
Wayman, R.B., North, M., 2007. Initial response of a mixed-conifer understory plant community
to burning and thinning restoration treatments. Forest Ecol. Manag. 239, 32-44.
Weatherspoon, C.P., Skinner, C.N., 1996. Landscape-level strategies for forest fuel management.
In, Sierra Nevada Ecosystem Project: Final Report to Congress, vol. II. University of
California Centers for Water and Wildland Resources., Davis, pp. 1471-1492.
Webster, K.M., Halpern, C.B., 2010. Long-term vegetation responses to reintroduction and
repeated use of fire in mixed-conifer forests of the Sierra Nevada. Ecosphere.
Williams, P.J., Gutiérrez, R.J., Whitmore, S.A., 2011. Home range and habitat selection of
spotted owls in the central Sierra Nevada. J. Wildlife Manage. 75, 333-343.
Yost, M., 1994. Quincy Library Group Community Stability Proposal: Silviculture, Timber
Management and the Desired Future Condition (3rd draft).
Young, J., 2003. Plumas County: History of the Feather River Region. Arcadia Publishing,
Charleston, South Carolina.
51