Experimental Design
For more detailed information on experimental methods,
download the study plan.
The experimental design consists of nine plots, with three replications
of early season burn, late season burn, and control treatments in a
completely randomized design. Burning treatments were implemented by
Sequoia National Park Fire Management, with the help of outside resources,
including personnel from Sequoia National Forest, Redwood National Park,
and Yosemite National Park. Early season burns were conducted June 20
and 27, 2002, and late season burns were conducted September 28, October
17, and October 28, 2001.
Data on all of the following disciplines was taken prior to the burning
treatments. Some early post-treatment data have been collected (fuel
reduction, tree fire damage, initial tree survival etc.). The majority
of post treatment data will be collected in 2003 and 2004.
Grid and plot system
All data are referenced to a series of 36 markers installed on a 50m grid
within each of the nine plots and mapped using GPS. Variables for many of
the study disciplines are measured within or just outside of ten 50m x 20m
(0.1 hectare) subplots within each plot. Subplots were randomly located
among all possible grid points, and placement relative to the grid point in
the four cardinal directions was also random.
|
|
Study Disciplines
Vegetation
Fuels and Fire
Wildlife
Soils and Forest Floor
Entomology
Pathology
Treatment Economics
Vegetation
Trees and Saplings
Data were taken on all individual trees within 50m x 20m subplots and
saplings in half of the area of each of the subplots. Trees were
defined as individuals having a diameter at breast height (dbh) greater
than or equal to 10cm and saplings were defined as individuals with a
height greater than 1.37m and a dbh less than 10cm. All trees were
labeled with a steel tag nailed into the bark at 1.37m above the base.
Data taken on individual trees included:
- Species Diameter at breast height
- Status (alive or dead)
- Crown position (dominant, codominant, intermediate, subcanopy)
- Overall health
- Height, height to base of live crown, height to base of dead crown
- Mistletoe score
- Scorch height (prescribed fire plots)
- Percent of crown volume scorched (prescribed fire plots)
- Bole char height (prescribed fire plots)
- Percent of crown circumference scorched (prescribed fire plots)
|
USGS crew taking data on forest
thinning and tree injury one season
post-fire (Unit 5, late season treatment).
(Photo by Eric Knapp)
|
Shrubs, Tree Seedlings, and Understory Species
Shrubs were sampled on the same half of the 50m x 20m vegetation subplot
that the tree saplings were counted in. Tree seedlings and all herbaceous
species (grasses, sedges, forbs, ferns) were sampled within twenty 1 m2 sub
subplots located within each subplot.
- Shrub cover by species
- Shrub cover by height class (short (<0.1m), medium (0.1 to 1.0 m), and tall (> 1.0 m)
- Tree seedling count by species and height class
- Tree seedling cover by species
- Herbaceous cover by species
- Herbaceous species count within 1m sub subplots and 50m x 20m subplots
Light Environment and Gap/Patch Distribution
Percent canopy closure was estimated at each grid point using a spherical
densiometer. Canopy cover was similarly estimated using a densitometer, which
is a small hand-held piece of pipe with a level and mirrors that allows
determination of whether a single point (cross hairs) hits either vegetation
above, or open sky. Five measurements were taken per grid point – one standing
over the grid point, and one in each of the four cardinal directions, 10m from the grid point.
- Percent canopy cover
- Percent canopy closure
Seedling Demography
To collect information on seedling and tree demography before and after fire,
a one hectare (100m x 100m) forest demography plot was established within each
of the three late season burn plots and within one control plot. These plots were set up in a similar fashion to long-term tree
monitoring plots established as part of the Sierra
Nevada Global Change Research Program and are intended to address
questions about post-fire tree recruitment patterns. All trees within these plots
were mapped with survey equipment and data are being collected on individual
trees and tree seedlings each year according to previously established protocols.
Seed traps were placed on the ground within the plots to estimate potential tree
reproduction by species.
Fuels and Fire
Mass of down, dead, woody fuels were measured using Brown’s planar intercept method.
Two 20m transects were installed at each of the 36 grid points. The number of downed
woody stems in different size classes that crossed the transect were counted and
metric tons per hectare of these fuel types were calculated using formulas given in
Brown (1974). Duff and litter depth was measured along each transect and mass calculated
using depth to weight relationships previously determined within the plots.
- Metric tons/ hectare of 1 hour fuel (0-6mm)
- Metric tons/ hectare of 10 hour fuel (6-25mm)
- Metric tons/ hectare of 100 hour fuel (25-76mm)
- Metric tons/ hectare of 1000 hour fuel (>76mm) rotten
- Metric tons/ hectare of 1000 hour fuel (>76mm) sound
- Metric tons/ hectare litter
- Metric tons/ hectare duff
- Fuel height
- Litter and duff consumption (using duff pins)
- Coarse woody debris – log number/ha, avg. log length, log cover/ha, log mass/ha
|
USGS crew
collecting data on coarse woody
debris prior to the prescribed burning treatments
(Photo by Eric Knapp)
|
Fuel Moisture and Fire Behavior
Prescribed
fire in Unit 6
(Late season treatment)
(Photo by Eric Knapp) |
Fuel moisture measurements were made at approximately weekly intervals starting two to
three weeks prior to the burn, and again on the day the burn treatments were applied.
Fuels (including litter, duff, woody fuels (0-6mm, 6-25mm, 25-76mm, and 76+mm), and live
fuels from individual tree species were widely collected in different microenvironments
within the plot and separately placed into air-tight zip-lock bags or nalgene bottles.
The larger woody fuels were obtained by cutting thin 1cm wide slabs out of appropriately
sized logs with a chainsaw or hand saw. Samples were weighed wet, dried at 85 degrees C and
weighed again.
|
Weather data (ambient temperature, relative humidity, wind speed, and wind direction)
were collected prior to and during the burn at approximately hourly intervals. During
the burn, flame length and rate of spread measurements were made using ocular estimates
for both heading and backing fire fronts. Data were taken opportunistically throughout
the plot and the location of each observation was referenced by approximate distance and
angle to the nearest grid point.
- Percent moisture (1hr, 10hr, 100hr, 1000hr, and live fuels)
- Air temperature
- Relative humidity
- Wind speed
- Wind direction
- Flame length (backing and heading fires)
- Rate of spread
- Flaming duration
|
USGS crew
monitoring weather
conditions during the prescribed burn
(Photo by Eric Knapp) |
Birds and Small mammals
Birds
Diversity and abundance of birds were assessed through the use of point count censuses
at predetermined widely spaced gridpoints. Each experimental unit was visited a
minimum of three times (up to six) during the breeding season. At each point, the horizontal
distance of birds relative to the survey point count was recorded for a five minute period,
beginning at first light.
Nest productivity (number of young fledged per nest initiated) was monitored in a subset
of plots. Wildlife crews randomly assigned two replicates of each treatment (including
controls) to be searched for nests. Nest sites were later visited (after fledging or
failing) to measure some simple vegetation variables. Data were also collected for cavity
nesting birds in snags, although information on egg number and fledgling number was often
not obtainable.
|
Bird nest on
a burned stump, one
season after the prescribed fire
(USGS Photo) |
Functional response of woodpeckers and other bark-gleaning birds meant observing their
foraging patterns on trees at each site. Two one-hour sessions were conducted in each plot
six times during the field season. Data were only collected for birds clearly foraging (not
singing, cavity drilling etc.) in trees.
Woodpecker activity was surveyed on a portion of trees previously noted by the entomology
crew to be dead or in decline due to bark beetles. At every other gridpoint, up to two trees
were randomly selected from among the infected trees identified (up to 36 trees per plot).
Woodpecker abundance and activity were noted on these trees, using the same methods as proposed
for the overall plot survey.
Data included:
- Bird count by species
- Nest success (percent of successful nests, number of young fledged)
- Foraging behavior (gleaning, probing, pecking, scaling, and excavating)
- Woodpecker count on bark beetle infested trees
- Woodpecker activity in “plots” on tree boles
|
Monitoring woodpecker activity
(Photo by Eric Knapp) |
Small Mammals
Each plot was trapped using Sherman traps for an eight to ten day sampling period per field season.
Traps were placed at the nearest cover (log, vegetation etc.) to the grid point and baited with a
mixture of seed, peanut butter, and rolled oats. A piece of carrot was also placed in the trap to
provide moisture for trapped animals. Captured small mammals were identified to species, weighed,
sexed, marked with numbered ear tags, and released. Microhabitat data were taken for the trap location.
Deer Mouse captured in live trap
(Photo by Eric Knapp) |
Data included:
- Percent trap success
- Number of captures by species
- Reproductive status
- Body length
- Hind foot length
- Ear length
- Tail length
- Distance between the anus and the urethral opening
- Percent canopy cover over trap
- Vegetative cover surrounding trap
|
Soils and Forest Floor
Both litter and duff layers and underlying mineral soil were collected for evaluating C and N
content. Mineral soil was also sampled for macronutrient content. Six samples were taken
distal to the outer boundary of each 50m x 20m vegetation sampling subplot. Nitrogen mineralization
and nitrification were measured by taking paired soil cores. Half were prepared for N analysis
immediately, and half were placed in a metal sleeve and incubated for 28 days in situ, prior to N
analysis. Soil biodiversity was also evaluated by estimating activity for different enzymes within
fresh soil samples.
- Carbon content – mineral soil and forest floor
- Nitrogen content - mineral soil and forest floor
- Soil pH
- Soil cation exchange capacity
- Soil organic matter
- Calcium content
- Magnesium content
- Potassium content
- Sodium content
- Available N
- Soil enzyme activity (acid phosphatase, b-glucosidase, and chitinase)
|
Sarah Hamman (Colorado State University)
collecting soil samples
(Photo by Eric Knapp) |
Entomology
A full-plot survey was conducted to locate all trees either containing pitch tubes or with
visible crown symptoms characteristic of bark beetle attack. Symptomatic trees were labeled
and scored into classes. In addition to the full plot search for symptomatic trees, every
pine tree was located and examined for the presence of pitch tubes, regardless of visible
crown symptoms.
Ground macroarthropod diversity and abundance were determined using pit-fall
traps. Two pit-fall traps (diameter 15cm X 12 cm deep) were placed within the
coarse woody debris transects at
odd-numbered grid-points, for a total of 36 traps per plot. A vegetation analysis
of the 2 x 2m area around each pit-fall trap was conducted using the same methods
as for the herbaceous vegetation
protocols. All trapped material was sorted to remove debris and the insects
counted according to species or morphological types.
|
Chris
Fettig (FFS entomology discipline leader)
(on left), and Scott Ferrenberg (USGS
Ecologist)(on left) looking for evidence
of bark beetle activity post-fire.
(Photo by Eric Knapp) |
- Number of trees attacked by bark beetles
- Tree species attacked
- Bark beetle species
- Magnitude of bark beetle attack (pitch tube score)
- Number of ground macroarthropods trapped by species
|
Ips and Jeffrey pine bark beetles
on a Jeffrey Pine weakened by
tissue damage caused by fire.
(Photo by Eric Knapp) |
Pathology
An initial survey of all treatment plots was conducted in order to mark trees
having pre-existing symptoms so as not to confuse these with subsequent treatment
effects. The entire central
approximately 10 hectare sampling area of each plot was searched for trees (dbh > 10cm)
with symptoms of pathogen infection. Very few trees with visible crown symptoms
of root disease were
noted. However, evidence of Heterobasidion annosum (fir type) infection
was common, based on presence of laminated wood on downed logs and fruiting
bodies at the base of stumps and downed
logs. The full-plot search was therefore devoted to locating all potential H.
annosum root rot
centers and associated gaps, rather than trying to find trees with crown symptoms.
Gap locations were drawn on a map of the plot and numbered. Due to the presence
of numerous gaps, five of the
gaps were randomly selected for more intensive evaluation. At each of these
selected gaps, the five live fir trees in closest proximity to the gap (defining
the outline of the gap) were labeled
with a numbered steel tag. Data taken on tagged trees included notes on crown
symptoms, dbh, crown position, and signs of other distress agents (such as bark
beetle pitch tubes, exit holes, etc.).
The “gap maker” where the root disease infection may have started (usually a
fallen tree) was also
mapped but not tagged.
Root samples were collected from tagged trees in the fall. Lateral roots were excavated and root
samples taken using an increment hammer. Roots from two healthy fir trees per plot were also sampled
in the same way to act as controls. Extracted cores were sealed in a labeled sterile plastic bag
and transferred to the laboratory. Pathogen isolations from these root samples were conducted
according to standard lab techniques using media specified by the Institute of Tree Root Biology (TRB).
- Number of gaps believed caused by root pathogens per plot
- Root pathogen species
- Percent of root samples containing pathogen
Treatment Economics
Costs may differ between seasons due to the greater duration of fire activity in the early season
prescribed fire units. Treatment costs and benefits are being determined through an expert opinion
survey approach. Records of personnel needed for the prescribed burns and equipment costs will be
obtained from NPS Fire Management. Since forest products are not being removed at Sequoia, the
economic analyses will be reduced compared to the other FFS sites.
- Cost per volume fuel reduction (by season of burning)
--
WERC Home --
Who We Are --
Where We Are --
What We Do --
What's New --
Outreach --
Contact Us --
Search --
USGS Privacy Policy, Disclaimer
, Accessibility
Comments to: webmaster@werc.usgs.gov
URL: http://www.werc.usgs.gov/fire/ffs-seki/index.html
Last update: 15 January 2004
