Management Options for Reducing Wildfire Risk and
Maximizing Carbon Storage under Future Climate Changes,
Ignition Patterns, and Forest Treatments1
Final Report
November, 2012
SNPLMA (Southern Nevada Public Lands Management Act) project (P049)
PI: Robert M. Schellera
Co-PIs: Peter Weisbergb, Jian Yangb, Alison E. Stantonc
Agency Collaborator: Carl Skinnerd
Project Researcher: E. Louise Loudermilka, Thomas E. Diltsb
Final Report developed by E.L. Loudermilk and A. E. Stanton
a
Portland State University
Environmental Science and Management Department
Portland State University
PO Box 751
Portland, OR 97207 USA
b
University of Reno Nevada
Department of Natural Resources and Environmental Science
University of Nevada, Reno
1664 N. Virginia St., Mail Stop 186
Reno, Nevada 89557 USA
c
Alison E. Stanton, Consulting
3170 Highway 50 Suite #7
South Lake Tahoe, CA 96150
d
USDA Forest Service
USDA Forest Service
Pacific Southwest Research Station
3644 Avtech Parkway
Redding, CA 96002 USA
1
Citation: Loudermilk, E. Louise, Stanton, Alison E., Scheller, Robert M., Weisberg, Peter J., Yang, Jian, Dilts,
Thomas E., Skinner, Carl. 2012. Management Options for Reducing Wildfire Risk and Maximizing Carbon Storage
under Future Climate Changes, Ignition Patterns, and Forest Treatments. Final Report for Southern Nevada Public
Lands Management Act project P049, Pacific Southwest Research Station, Tahoe Center for Environmental Studies,
Incline Village, NV.
Table of Contents
List of Tables ................................................................................................................................................. 4
List of Figures ................................................................................................................................................ 5
Key Findings ................................................................................................................................................ 10
1.0 Introduction .......................................................................................................................................... 11
1.1 Carbon dynamics in a changing climate............................................................................................ 11
1.2 Wildfire dynamics in a changing climate .......................................................................................... 12
1.3 Forest fuels management ................................................................................................................. 12
1.4 Study Objectives ............................................................................................................................... 13
2.0 Methods ................................................................................................................................................ 14
2.1 Study area ......................................................................................................................................... 14
2.2 Model development, parameterization, and calibration ................................................................. 15
2.2.1 Landscape and forest composition ............................................................................................ 15
2.2.2 Climate downscaling .................................................................................................................. 15
2.2.3 Modeling carbon dynamics with the Century Extension ........................................................... 16
2.2.5 Wildfire ignition density modeling............................................................................................. 17
2.2.6 Modeling fire behavior and effects with the Dynamic Fire Extension....................................... 19
2.3 Development and implementation of fuel treatments .................................................................... 21
2.3.1 Local fuel treatment data and treatment areas ........................................................................ 21
2.2.2 Assigning fuel types and simulating treatment ......................................................................... 22
2.2.3 Fuel treatment prescriptions ..................................................................................................... 23
2.3.4 Fuel treatment scenarios ........................................................................................................... 24
2.4 Landscape simulations and analysis ................................................................................................. 25
3.0 Results ................................................................................................................................................... 26
3.1 Impacts of climate change ................................................................................................................ 26
3.1.1 Wildfire dynamics ...................................................................................................................... 26
3.1.2 Carbon dynamics........................................................................................................................ 28
3.1.3 Wildfire effects on carbon dynamics ......................................................................................... 29
3.1.4 Species response to changing climate and fire regime.............................................................. 29
3.2 Fuel treatment effects under contemporary climate ....................................................................... 30
3.2.1 Wildfire....................................................................................................................................... 30
3.2.2 Forest carbon and growth.......................................................................................................... 31
2
3.2.3 Species dynamics ....................................................................................................................... 32
3.3 Predicted wildfire ignition density .................................................................................................... 32
3.3.1 Modeling results ........................................................................................................................ 32
3.3.2 Effects of increased wildfire ignitions ........................................................................................ 34
3.4 Fuel treatment effects under a changing climate ............................................................................. 35
3.4.1 Wildfire....................................................................................................................................... 35
3.4.2 Forest carbon and growth.......................................................................................................... 35
3.4.3 Species dynamics ....................................................................................................................... 36
4.0 Discussion.............................................................................................................................................. 36
4.1 Climate change effects on growth rates and wildfire activity .......................................................... 37
4.2 Climate change effects on carbon sequestration potential and net carbon emissions ................... 38
4.3 Climate change effects on individual species ................................................................................... 39
4.3 Fuel treatment effectiveness under contemporary climate ............................................................. 40
4.4 Fuel treatments in a changing climate .............................................................................................. 42
5.0 Further considerations .......................................................................................................................... 42
6.0 Conclusions ........................................................................................................................................... 43
7.0 Acknowledgements............................................................................................................................... 44
7.0 References ............................................................................................................................................ 45
Figures ......................................................................................................................................................... 51
Appendix A .................................................................................................................................................. 97
3
List of Tables
Table 3-1. Simulated fire rotation periods (FRP) and mean and standard deviation of fire sizes at the LTB
for all fuel treatment scenarios, and across five replicate 100 year simulations.
Table 3-2. Individual species response to climate change, represented here as an approximate positive
or negative % difference (to nearest 5%) of mean aboveground live biomass between both A2 or
B1 climate and base climate at year 2110.
Table3-3. Simulated fire rotation periods (FRP) and mean and standard deviation of fire sizes at the LTB
for all fuel treatment scenarios, and across five replicate 100 year simulations.
Table 3-4. Simulated fire rotation periods (FRP) and mean and standard deviation of fire size and annual
area burned at the LTB for three climate scenarios, with and without increased ignitions across
five replicate 100 year simulations.
Table 3-5. Simulated fire rotation periods (FRP) and mean and standard deviation of fire size and annual
area burned at the LTB for three climate scenarios, with and without increased ignitions and fuel
treatment scenarios across five replicate 100 year simulations.
4
List of Figures
Figure 2-1. Map of the study area for this research; the Lake Tahoe Basin, CA and NV, USA.
Figure 2-2. LANDIS-II climate inputs and processes modeled for the forested landscape of the Lake
Tahoe Basin, CA, NV. Climate data were implemented into the Century extension and Dynamic
Fire extension using seasonal, monthly, and daily inputs, respectively of current and projected
climate (years 2010-2110). Processes in the Century extension include above and belowground
components (e.g., plant root, shoot, and leaf biomass, soil organic matter). FRP: Fire Rotation
Period
Figure 2-3. Initial landscape composition by vegetative community in the Lake Tahoe Basin (LTB) (after
Ottmar et al. 2009).
Figure 2-4. Study area of the Lake Tahoe Basin with reported human- and lightning-caused fires greater
than 0.1 ha in size between 1986 and 2009.
Figure 2-5. Maps showing road density, lightning density, elevation, average January minimum
temperature, average July maximum temperature, and annual water deficit in the Lake Tahoe
Basin, overlaid with reported human-caused (red dots) and lightning-caused (blue triangles)
fires.
Figure 2-6. Fire size distribution simulated using LANDIS-II (100 years, using base climate) and empirical
data from the study area.
Figure 2-7 Treatment areas within the LTB: the defensible space, defense zone, and extended wildland
urban interface.
Figure 2-8. Size class distribution of all fuel treatment units within each treatment area in the LTB (from
LTBMU Management unit data).
Figure 2-9. Treatment stand map representing randomly located polygons of approximately the same
size distribution as existing fuel treatments in the LTB (from LTBMU Management unit data).
Figure 2-10. Fraction biomass removed in the light thinning prescription, by species and age cohorts.
Figure 2-11. Fraction biomass removed in the moderate thinning prescription, by species and age
cohorts.
Figure 2-12. Fraction removed in the mid-seral thinning prescription, by species and age cohorts.
Figure 2-13. Continued Intensity scenario under a) 15 and b) 30 year rotation period and the Transition
to Forest Health Initiative’ scenario under c) 15 and d) 30 year rotation period.
Figure 3-1. Mean annual temperature (°C) and total annual precipitation (cm) simulated at the LTB using
current climate conditions and two GCMs of high (A2) and low (B1) emissions across 100 years.
This represents landscape level mean and standard errors across five replicates.
5
Figure 3-2. Simulated mean landscape carbon estimates of a) live C (bole, leaves, roots), b) total C (live C
+ detrital C + SOC), c) detrital C, d) SOC (3 soil pools: slow, fast, passive) simulated over 100
years (five replicates) using contemporary (base) climate and two emissions scenarios (A2, B1).
Figure 3-3. Simulated mean landscape carbon estimates of a) Aboveground Net Primary Productivity, b)
Net Ecosystem Production, and c) heterotrophic respiration, simulated over 100 years (five
replicates) and using contemporary (base) climate and two emissions scenarios (A2, B1). Note
difference in scale of y-axis between graphs.
Figure 3-4. Results from second simulation approach, where temperature and precipitation were
independently assessed using the A2 and base climate. Increasing temperatures (as opposed to
reduced precipitation) determined impacts on forest C (a, live C) and productivity (b, ANPP). A2
Temperature only: A2 temperature combined with base climate precipitation; A2 Precipitation
only: A2 precipitation combined with base climate temperature.
Figure 3-5. Comparison of simulated live carbon with and without fire using the base and A2 climate.
Figure 3-6. Simulated mean area burned (ha) across the four fuel treatment management areas within
the LTB, representing mean and standard deviations across five replicate model simulations.
Figure 3-7. Simulated mean fire severity (index from 1:5) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations.
Figure 3-8. Simulated mean area burned (ha) across the four fuel treatment management areas within
the LTB, representing mean and standard deviations across five replicate model simulations.
Figure 3-9. Simulated mean fire severity (index from 1:5) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations.
Figure 3-10. Simulated mean total carbon (live C + soil organic C + detrital C, g C m-2) and Aboveground
Net Primary Productivity (ANPP) across the LTB, representing mean and standard deviations
across five replicate model simulations.
Figure 3-11. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across
five replicate model simulations.
Figure 3-12. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across
five replicate model simulations.
6
Figure 3-13. Simulated forest growth, represented as ANPP (Aboveground Net Primary Productivity),
across the four fuel treatment management areas within the LTB, representing mean and
standard deviations across five replicate model simulations.
Figure 3-14. Change in mean aboveground live biomass (g m-2) of the six most abundant tree species at
the LTB for simulations (a) without fuel treatments and (b) with fuel treatments (continuous 15
year rotation period).
Figure 3-15. Mean aboveground live biomass (g per m-2) of white fir (Abies concolor) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across
five replicate model simulations.
Figure 3-16. Mean aboveground live biomass (g per m-2) of Jeffrey pine (Pinus jeffreyi) across the four
fuel treatment management areas within the LTB, representing mean and standard deviations
across five replicate model simulations.
Figure 3-17. Mean aboveground live biomass (g per m-2) of sugar pine (Pinus lambertiana) across the
four fuel treatment management areas within the LTB, representing mean and standard
deviations across five replicate model simulations.
Figure 3-18. Mean age of Jeffrey pine (Pinus jeffreyi) across the four fuel treatment management areas
within the LTB, representing mean and standard deviations across five replicate model
simulations.
Figure 3-19. Mean age of sugar pine (Pinus lambertiana) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations.
Figure 3-20. Mean age of white fir (Abies concolor) across the four fuel treatment management areas
within the LTB, representing mean and standard deviations across five replicate model
simulations.
Figure 3-21. Mean age of incense cedar (Calocedrus decurrens) across the four fuel treatment
management areas within the LTB, representing mean and standard deviations across five
replicate model simulations.
Figure 3-22. Max age distribution of incense cedar across the LTB. Left to right: Year 2010 (year 0 of
model), Year 2110 (year 100) without fuel treatments, Year 2110 (year 100) with continuous fuel
treatments applied on a 15 year rotation interval.
Figure 3-23. The top 6 most important predictor variables and their marginal effects on fire occurrence
density for (a) lightning-caused and (b) human-caused fires.
Figure 3-24. Landscape level mean fire occurrence density predicted by the top 30 best Poisson point
process models under the climate scenario of A2 and B1 for the lightning-caused (a and b) and
human-caused (c and d) fires. The unit of fire occurrence density is # of fires per 100 sq. km per
7
decade. Each line represents one model with red, green, and blue color indicating the ranking
order (R > G > B) of each model’s weight of evidence. The black line represents the modeling
average results.
Figure 3-25. Predicted landscape-level mean total (human-caused + lightning-caused) fire occurrence
densities (fires per 100 sq. km per decade) and their relative changes (above panel) under the
climate A2 and B1 scenarios, as well as time series of predicted proportion of lightning-caused in
the total fire occurrences (bottom panel).
Figure 3-26. Predictive maps of total ignition density (# of fires per 100 sq. km per decade) at (a) year
2000 and (b) year 2100 under the SRES A2 scenario.
Figure 3-27. Simulated mean area burned (ha) across the LTB, representing mean and standard
deviations across five replicate model simulations, using three climate scenarios with current
fire ignition estimates (left) and with increased ignitions (right).
Figure 3-28. Simulated mean total Carbon (live C + soil organic C + detrital C, g C m-2) and Net Ecosystem
Production (NEP) across the LTB, representing mean and standard deviations across five
replicate model simulations, using three climate scenarios with current fire ignition estimates
(top row) and with increased ignitions (bottom row).
Figure 3-29. Simulated mean aboveground live biomass (g m-2) of four representative tree species found
at the LTB, representing mean and standard deviations across five replicate model simulations,
using three climate scenarios with current fire ignition estimates and with increased ignitions.
Top, left to right: white fir, Jeffrey pine, Bottom, left to right: sugar pine, red fir.
Figure 3-30. Simulated mean aboveground live biomass (g m-2) of re-sprouting shrubs, representing
mean and standard deviations across five replicate model simulations, using three climate
scenarios with current fire ignitions and with increased ignitions.
Figure 3-31. Simulated mean area burned (ha) across the LTB, representing mean and standard
deviations across five replicate model simulations, using three climate scenarios with current
fire ignition estimates and with increased ignitions.
Figure 3-32. Simulated mean total carbon (live C + soil organic C + detrital C, g C m-2) across the LTB,
representing mean and standard deviations across five replicate model simulations, under base
and A2 climate, with current fire ignitions and with increased ignitions, and applying continuous
fuel treatments on a 15 year rotation period.
Figure 3-33. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across
five replicate model simulations. Base climate: No fuel treatments using base climate, Base: FT
15 RP: Continuous fuel treatments applied on a 15 year rotation period using base climate, A2
climate w/incr. ign: No fuel treatments using A2 climate with projected increase in fire ignitions,
8
A2: FT 15 RP w/incr. ign: Continuous fuel treatments applied on a 15 year rotation period using
A2 climate with projected increase in fire ignitions.
Figure 3-34. Total carbon (live C + soil organic C + detrital C, g C m-2) distribution across the LTB. Left to
right: Year 2110 (year 100 of model) for base (contemporary) climate without fuel treatments,
A2 climate without fuel treatments, Year 2110 (year 100) with continuous fuel treatments
applied on a 15 year rotation interval.
Figure 3-35. Simulated mean aboveground live biomass (g m-2) of four representative tree species found
at the LTB, representing mean and standard deviations across five replicate model simulations,
for base and A2climate, with and without increased ignitions and fuel treatments. Top, left to
right: white fir, Jeffrey pine, Bottom, left to right: sugar pine, red fir.
Figure 3-36. Simulated mean aboveground live biomass (g m-2) of re-sprouting shrubs, representing
mean and standard deviations across five replicate model simulations, using three climate
scenarios with contemporary fire ignitions and with increased ignitions, and with and without
fuel treatments (15 yr. rotation period).
9
Key Findings
Interactions among climate, forest growth, carbon (C) sequestration, and wildfire activity in the Lake
Tahoe Basin (LTB)
Our study illustrated the potential for continued forest growth and sequestration of above and
below ground C across the LTB, which remained a C sink (Net Ecosystem Productivity > 0),
despite any potential shifts in climate in the coming decades. This was a landscape legacy effect
from the Comstock Era logging in the 1880’s and the resulting regeneration and growth of the
forest into the next century.
Effects from future changes in climate at the LTB included reduced establishment ability of the
subalpine and upper montane tree species; stimulated growth of particular conifers, aspen and
re-sprouting shrub species; and enhanced wildfire activity. Changes in the wildfire regime had
the strongest impact on forest response.
An increase in wildfire activity (area burned) in a changing climate caused higher mortality rates
across the LTB and lower C sequestration potential by year 2110.
Altered mortality patterns, caused by increased wildfire activity under A2 climate, had a
stronger effect on overall productivity and C storage potential than climate change effects (e.g.,
growth, establishment) on individual species.
In a changing climate, increased wildfire activity was caused by a reduction in fine fuel moisture
across a longer growing season.
Wildfire activity was further enhanced under future-climate scenarios, especially towards the
end of the 21st century, by projected increases in natural and anthropogenic ignition sources.
Interactions with simulated fuel treatments
Continuous fuel treatments reduced area burned and fire severity across the LTB, regardless of
climate or fuel treatment scenario applied.
Projected increases in ignitions under the A2 climate dramatically reduced the ability of the
forest to sequester C, but continuous fuel treatments moderated the reduction.
The forest continued to sequester C in all fuel treatment scenarios, although at a lower rate than
without fuel treatments.
Continuous fuel treatments strongly suppressed target species (e.g., white fir and incense cedar)
in managed areas and improved the regeneration environment for more shade-sensitive species
(e.g., Jeffrey pine, sugar pine).
As wildfire ignitions increase, mainly towards the end of the 21st century, mitigation feedbacks
from fuel treatments may become more difficult but be possibly more essential, as more area
treated would intersect with more wildfires. This in turn would cause earlier net C gain from
fuel treatment application, than fuel treatment simulations using contemporary climate.
10
1.0 Introduction
1.1 Carbon dynamics in a changing climate
Understanding the effects that climate change may have on forest carbon (C) budgets requires
knowledge of climate effects on forest growth, long-term forest succession, and important past and
future disturbances, such as wildfires (Westerling et al. 2006, Millar et al. 2007, Xu et al. 2009). Climate
change affects forests by altering the physiological responses of trees and generates subsequent
changes in forest succession. In addition, forest structure and composition responds to climate-induced
changes to natural disturbances regimes (e.g., frequency and duration of wildfires) (Dale et al. 2001).
Climate change effects are often region specific, however, whereby plant species and the driving
processes (e.g., forest succession, nutrient cycling, disturbances) may behave idiosyncratically (Scheller
and Mladenoff 2008). The coupled effects from past disturbances (e.g., historic logging), large wildfires,
and future changes in climate call for a modeling approach at the regional scale that incorporates
sufficient interaction, contingency, and site specificity.
Changes in regional weather and climatic patterns will influence the physiological response of trees. In
California, a 2-5°C increase in mean annual temperature over the next century is expected to affect the
spatial distribution and phenology of extant tree species (Millar et al. 2006). Shifts in precipitation
(magnitude and seasonal distribution) will be variable between regions and dependent on local weather
patterns and topography. In the western U.S., increasing temperatures and decreasing precipitation
have been identified as a significant cause of increasing tree mortality (van Mantgem et al. 2009).
Climate projections for California however, suggest little change in annual total precipitation, although
temperatures are expected to continue increasing (Cayan et al. 2008). This emphasizes the importance
of understanding how projected alterations in temperature or precipitation, independently or in
tandem, affect various ecosystem processes. Increased temperature may reduce long-term soil
moisture availability, with subsequent effects on decomposition rates and N availability, ultimately
reducing forest productivity (Kirschbaum 1995, Lenihan et al. 2003). In the Lake Tahoe Basin (study area
of this research, CA, NV), mean annual temperatures are expected to increase by 1-5°C over the next
century, while projected precipitation is highly variable (Coats et al. 2010). Throughout the Sierra
Nevada, changing climate conditions are expected to reduce snow pack with more precipitation falling
as rain. Shorter winters may cause earlier snow melt and prolonged droughty summers (Battles et al.
2008).
A warmer climate will likely alter the net C balance of forested landscapes. Northern hemisphere C
uptake estimates vary widely, ranging from 0.3-0.9 Pg of C per year in the 1990’s (Goodale et al. 2002).
This C uptake is attributed largely to northern forests (Goodale et al. 2002, Houghton 2003b) and is likely
due to fire suppression, re-growth from extensive harvesting over the past two centuries, and increased
nitrogen (N) deposition and atmospheric C fertilization (Schäfer et al. 2003). The magnitude and
duration of continued C uptake is uncertain, however, especially at the local scale at which forest
management occurs, and may decline if drier conditions reduce growth rates (Goodale et al. 2002,
Houghton 2003b, Albani et al. 2006). The ability of forests to sequester carbon and the potential
management approaches to reduce C emissions or maintain C stocks have received particular attention
11
(Galik and Jackson 2009). Difficulty in quantifying the C balance have been a major obstacle in
determining how individual forests fit into C offset programs (Richards and Andersson 2001).
1.2 Wildfire dynamics in a changing climate
The influence of a changing climate on disturbance regimes is of particular interest, especially with
regard to wildfires (McKenzie et al. 2004). Wildfire regimes in the western U.S. are directly affected by
climate change: Increasing temperatures advance the timing of snowmelt and vegetation green up, dry
fuels earlier, and extend wildfire seasons. In the western U.S., warming temperature and earlier spring
have resulted in a fourfold increase of major forest wildfires and sixfold increase of area burned since
1986, compared to the period from 1970 to 1986 (Westerling et al. 2006). This increase in area burned is
substantially controlled by a changing climate, despite increases in fuel accumulation (Littell et al. 2009).
Wildfire frequencies are also expected to increase causing a higher flux of greenhouse gas emissions
(Campbell et al. 2007), and affecting forest C pools in significant, but unknown ways (Westerling et al.
2006). The altered fire regime can cause degradation of ecosystem functions, create hazards for people,
and increase fire suppression costs (Stephens and Ruth 2005; Syphard et al. 2011).
In order to efficiently manage fuels and to reduce fire risks, it is important to understand how fire
regime would respond to the ongoing climate change in the near future. Many studies have found the
spatial distribution of fire occurrence, both human- and lightning-caused, is not completely random, and
its spatial patterns can greatly define spatial variability of fire risk at large scales (Bar Massada et al.
2011; Finney et al. 2011). Although there is a strong linkage between climate and fire occurrence, other
spatial controls such as topography, vegetation, ignition sources can also exert strong influences that
may confound our prediction of fire response to climate change (Liu et al. 2012; Parisien et al. 2012).
Comprehensive studies that examine how various controls affect fire occurrence are often conducted at
regional or global scales with coarse spatial resolutions (Krawchuk et al. 2009; Wotton et al. 2010).
However, effective fuel treatment design and its placement planning are often conducted at landscape
scales with fine spatial resolutions (Finney 2008). A spatially explicit analysis of fire occurrence and its
response to climate change at fine scales is warranted for assisting fire and fuel management at
landscape scales.
1.3 Forest fuels management
Fuel-reduction treatments (i.e., forest thinning) are used extensively throughout the western U.S. (and
worldwide) to reduce hazardous surface and ladders fuels and restructure the forest (Agee and Skinner
2005). The forests of the Sierra Nevada are of particular concern, where fuel loads and the density of
small trees have exceeded known historic conditions due to a century of fire suppression (Parsons and
DeBenedetti 1979). The resulting catastrophic wildfires have devastated forests (Westerling et al. 2006,
Littell et al. 2009) and nearby human settlements. Applying fuel treatments has become an essential
management tool for reducing wildfire intensity and severity in this region (Agee and Skinner 2005,
Schwilk et al. 2009, Syphard et al. 2011), although the trade-offs with labor costs and ecosystem carbon
(C) balance, as well as impacts on wildlife habitat are of concern (Calkin and Gebert 2006, Pilliod et al.
2006, Campbell et al. 2012).
12
In the near future, forest managers may be faced with the prospect of increased regulation of C
emissions, and may be forced to balance the use of fuel treatments for reducing fire risk against the
potential reduction of forest carbon (Hurteau et al. 2008). Furthermore, properly balancing the spatial
arrangement of management activities in order to achieve multiple objectives on the landscape (e.g.,
(Daugherty and Fried 2007, Rhodes and Baker 2008, Schmidt et al. 2008)) requires more information
about the inherent trade-offs among the objectives and improved awareness of the opportunities for
optimizing management at the landscape scale (Syphard et al. 2011). For example, fuels management
(e.g., mechanical thinning and prescribed burning) will require consideration of the immediate loss of C
from live and detrital matter with reduced fire emissions (Hurteau et al. 2008, Scheller et al. 2011c).
Previous studies that explicitly address C dynamics have typically addressed only aboveground C stocks
(e.g., (Hurteau and North 2009), although surface and soil C are also important long-term C stocks (Zinke
and Stangenberger 2000) and fluctuate in response to changes in live and detrital inputs (Scheller et al.
2011c).
Predicting the effectiveness of a suite of fuel treatments for reducing fire risk and altering ecosystem C
dynamics requires an assessment at the landscape level, where spatial arrangement of fuel treatments
and potential intersection with wildfires can be addressed (Syphard et al. 2011). The strategic
placement of fuel treatments is important for reducing landscape level wildfire spread and intensity
(Schmidt et al. 2008). Furthermore, insight into where treatments may be most effective (e.g., areas of
high ignition potential) and where treatment-wildfire intersection may be more likely, may be more
important than the amount of area treated. The re-application timeline or rotation period is also of
interest, where more intensive treatments (e.g., mechanical vs. hand thinning) may have a longer
effective period for reducing fire risk. Maintaining fuel treatments through time may restructure the
landscape, creating a more fire-resistant forest, and maintain live C stocks and reduce C emissions from
wildfire in the long term (Hurteau and North 2009, North and Hurteau 2011).
1.4 Study Objectives
Our objectives for this project were, 1) to evaluate the emergent responses of multiple interacting
processes, namely climate change and wildfire regime, on total forest carbon and succession dynamics,
and 2) to evaluate the long-term effects of fuel treatments in mitigating wildfires and sequestering
forest carbon (C), in a contemporary and climate change context, within the regional landscape of the
Lake Tahoe Basin, CA and NV. We used two future C emissions scenarios as expressed within the
Geophysical Fluid Dynamics Laboratory General Circulation Model, in combination with a landscapescale model of forest succession (Scheller et al. 2007), stochastic wildfire (Sturtevant et al. 2009), and C
dynamics (Scheller et al. 2011c) to examine the potential effects of projected climate change on: 1)
forest growth rates, 2) individual tree species response, 3) C sequestration potential and net C emissions
to the atmosphere, and 4) wildfire activity, including changes in future ignition patterns. The
independent effects of temperature and precipitation within and between emissions scenarios, as well
as fire-climate interactions, were assessed. We also examined the relative influences of different spatial
controls and resulting spatial patterns for both lightning- and human-caused fire occurrences.
Furthermore, we discuss C sequestration potential with respect to the legacy influence of past
disturbances. To address interactions with fuel treatments, forest thinning prescriptions (Syphard et al.
13
2011) were simulated to understand long-term effects of fuel treatments on wildfires, above and
belowground C dynamics, and species and community structure across the climate regimes. A multiple
fuel treatment scenario design was used to examine the interactive effects of treatment application in
terms of spatial arrangement and location, rotation period, and prescription type. We discuss the
implications of fuel treatments in a contemporary climate and climate change context.
2.0 Methods
2.1 Study area
Our study area comprised approximately 85,000 ha of forested land in the Lake Tahoe Basin (LTB)
(Figure 2- 1). The climate is Mediterranean with a summer drought period. The basin-like topography
and elevation range (ca. 1897 to 3320 m) control local temperature and precipitation patterns. The
majority of precipitation falls as snow between October and May and snowpack persists into the
summer, dependent on elevation. The western portion of the Basin has a higher water balance and
greater productivity than the northern and eastern portions of the Basin. Mean annual precipitation on
the west shore at Tahoe City, CA is 80 cm and mean annual snowfall is 483 cm. Mean annual
precipitation on the east shore at Glenbrook, NV is 46 cm and mean annual snowfall is 236 cm. At Tahoe
City, average January high temperature is 6 °C and the low is -6 °C. Summers are mild with an average
high temperature of 26 °C in August and a low of 6 °C (Western Regional Climate Center 2012). Soils are
classified as shallow Entisols or Inceptisols and the more developed soils are Alfisols. The substrate is
mainly granite with ancient volcanic bedrock lining the north shore (Rogers 1974).
Tree species distribution in the LTB is controlled by elevation and precipitation (Barbour 2002). The
lower montane zone in the west Basin is primarily a mixed conifer forest consisting of up to six codominant species including white and red fir (Abies concolor, A. magnifica), incense cedar (Calocedrus
decurrens), and Jeffrey, sugar, and lodgeppole pine (Pinus jeffreyi, P. lamertiana, P. contorta). The east
side montane zone is dominated by Jeffrey pine, red fir, and/or white fir. The subalpine zone consists of
whitebark pine (P. ablicaulis), western white pine (P. monticola), and mountain hemlock (Tsuga
mertensiana).
Almost 70% of the lower montane zone in the LTB was clearcut during the Comstock logging era
beginning around 1870 and continuing through the turn of the century. Timber harvested from the LTB
was used for shoring up mineshafts, creating timber flumes, and extending rail lines to nearby Virginia
City, NV. The timber harvest and subsequent fire suppression throughout the last century has lead to a
shift in the age and size distribution of the forest from a characteristic old-growth canopy, with an open
mid-story, to a more dense forest of younger age-cohorts (<120 years old) and more closed mid-story
(Barbour et al. 2002, Nagel and Taylor 2005). This shift has allowed more surface and ladder forest fuels
to accumulate and has increased fire risk (Beaty and Taylor 2008). In addition, shade tolerant trees like
white fir and incense cedar have increased disproportionately over fire-adapted species like Jeffrey and
sugar pine (Nagel and Taylor 2005).
14
2.2 Model development, parameterization, and calibration
To address the disturbance feedbacks of fuel treatments and wildfires on coarse-scale forest and C
dynamics (Figure 2-2), we used the Landscape Disturbance and Succession model, LANDIS-II (v.6.0). The
LANDIS-II model has been used extensively for understanding ecosystem C dynamics (Scheller et al.
2011b, Scheller et al. 2011c) and feedbacks associated with wildfire (Sturtevant et al. 2009) and fuel
treatment effectiveness (Syphard et al. 2011). LANDIS-II offers the flexibility to integrate various
ecosystem processes and disturbances that interact across large spatial extents and long time periods,
ideal for projecting forest succession and responses to human and natural disturbance.
2.2.1 Landscape and forest composition
LANDIS-II simulates the life history characteristics of individual species of trees and shrubs, each
represented as age-cohorts. Individual trees are not modeled. To characterize initial forest
communities, we utilized a database and a map of age-cohorts of trees and shrubs developed for the
LTB based on the Fuel Characteristic Classification System (FCCS,
http://www.fs.fed.us/psw/partnerships/tahoescience/fccs.shtml) and the existing vegetation map
(CALVEG) from the GIS Clearinghouse of the Pacific Southwest Region
(http://www.fs.fed.us/r5/rsl/clearinghouse/aa-ref-tmu.shtml) (Ottmar and Safford 2011). We used data
on the 10 most abundant tree species found within the LTB, each represented in our model by unique
life history characteristics including longevity, age of maturity, shade tolerance, fire tolerance, and
sexual and vegetative reproductive capabilities (Table 2-1). For simplicity, the shrubs were grouped into
four functional groups: 1) non-nitrogen fixing re-sprouters, 2) non-nitrogen fixing obligate seeders, 3)
nitrogen fixing re-sprouters, and 4) nitrogen fixing obligate seeders. These shrub functional groups are
often found together within sites and consist of various species, including Arctostaphylos spp., Artemisia
spp., Ceanothus spp., Chrysolepis spp., Purshia spp. , Quercus spp. Ribes spp., and Symphoricarpos spp.
The resulting forest community map (Figure 2-3) was coupled with Forest Inventory Analysis data from
the LTB and nearby Sierra Nevada forests to provide estimates of species composition and age
distribution by forest type (e.g., mixed conifer), relevant at the coarse scale being modeled and similar
to (Syphard et al. 2011) . This map was refined (older cohorts were removed) to account for the largest
and most significant wildfires from years 2002 to 2010, where canopy tree mortality rates were
considerable (up to 100%, (Safford et al. 2009).
2.2.2 Climate downscaling
Basin wide climate projections were processed using existing downscaled Geophysical Fluid Dynamics
Laboratory General Circulation Models (GCM) of global CO2 emissions scenarios of climate from years
2010 from 2100, specific to the LTB (Coats et al. 2010). Under the high emissions scenario (A2), carbon
emissions continue at their current pace and [CO2] reaches 800+ ppm by 2100. Under the low emissions
scenarios (B1), carbon emissions are controlled and [CO2] reaches only 550 ppm by 2100. These
downscaled climate values (precipitation and temperature) were further processed using PRISM data
(Parameter-elevation Regressions on Independent Slopes Model, http://www.prism.oregonstate.edu/).
The PRISM data were averaged across each fire region and ecoregion within each GCM cell and used to
aggregate the projected climate data down to the regional scale.
15
2.2.3 Modeling carbon dynamics with the Century Extension
Ecosystem C dynamics were modeled using the LANDIS-II Century Succession extension (Scheller et al.
2011c), based on the original CENTURY soil model (Parton 1983). This extension (hereon called
‘Century’) integrates aboveground processes of successional dynamics with C and nitrogen cycling as
well as soil decomposition and accumulation. Century parameters were developed for three ecosystem
levels: tree species, tree functional groups, and ecoregions (see below). Further parameter descriptions
and examples of calibration procedures for Century are found elsewhere (Scheller et al. 2011b, Scheller
et al. 2011c).
Successional dynamics, including tree establishment, growth, competition, and decomposition are based
on species-specific parameters that represent species sensitivity to temperature and drought in addition
to estimates of % lignin and C:N ratios in various plant parts (Table 2-2). The parameters are drawn from
the original CENTURY model (http://www.nrel.colostate.edu/projects/century/manual4/man96.html) or
from the literature, where available.
Tree species and shrub functional groups were further classified into functional types, where
temperature, Leaf Area Index (LAI), and water limitations were parameterized for monthly growth
patterns, specifically in response to the dry summers and below-freezing winter temperatures found at
the LTB (Table 2-3). Because the vegetation within the LTB is well-adapted to summer drought and
benefits mainly from the spring snowmelt and early autumn rainfall, the functional groups were
parameterized to reduce water limitations during the dry months and allow for productivity through the
summer. The CENTURY default LAI values were used for conifers and hardwoods, while shrubs were
adjusted for lower growth capacity (i.e., Net Primary Productivity) compared to trees.
The LTB landscape was represented as five ecoregions, distinguished by unique soil characteristics and
climate variables (Table 2-4). Initial values for C and N in various soil organic pools were determined for
each ecoregion (Table 2-5). The dataset was developed from national soil databases (National Resources
Conservation Service, Soil Survey Geographic database). Within each ecoregion, Century calculates
species-specific establishment probability based on species sensitivity to temperature, soil moisture,
length of dry period, and January temperatures. Detailed explanations of each Century v3 extension
parameter are found in the user's guide (www.landis-ii.org).
Six target model outputs were chosen to calibrate and validate Century parameters based on available
literature and expert opinion. These include aboveground live biomass, soil organic carbon (SOC), soil
inorganic nitrogen (mineral N), aboveground net primary productivity (ANPP), Net Primary Production
(NPP), and Net Ecosystem Production (NEP).
To validate starting conditions (year 2010), aboveground live biomass and C and N pools were
simulated. There was strong agreement between the mean simulated aboveground live biomass across
the LTB (10,006 g m-2 ) and the measured values (10,787 g m-2 ) developed for the LTB from a separate
remote sensing dataset (Dobrowski et al. 2005).Initial landscape SOC was calibrated with regional SOC
estimates (3,950 (SD=532) g C m-2, (Zinke et al. 1998)). In Century, there are three SOC pools: fast, slow,
and passive. Decay rates of the three SOC were calibrated to result in a stable rate of decay of SOC
16
throughout the simulations; changes in the decay rate of SOM2 (the slow pool) had the largest impact
on decomposition. For comparison, a range of decay rates for SOM2 (0.01, 0.02, 0.03, 0.04) were
explored. A decay rate of 0.02 for SOM2 produced accurate initial conditions and a realistic SOC
accumulation rate. Simulated mineral N values (mean (SD) across 100 years (4.3 (0.6) g N m-2) were
comparable to target mineral N estimates (5-10 g N m-2, (Miller et al. 2010, Karam et al. in press)).
ANPP was calibrated by comparing simulated mean annual ANPP values against measured values of
annual ANPP in a nearby 50-year old Ponderosa Pine plantation (Campbell et al. 2009). Very few
Ponderosa pines occur in the LTB, but this species is closely related taxonomically to Jeffrey pine
(Critchfield and Little Jr. 1966) which is widely distributed. A single reference site consisting of one
young (5 year old) age cohort of Jeffrey pine was used to initialize the simulation and five replicates
were run to year 45 (50-year old stand). Simulations resulted in mean annual ANPP of 464 (SD=55) g C
m-2 yr-1, similar to the reference site (435 (SD=69) g C m-2 yr-1) and comparable to general estimates for
temperate biomes (456 g C m-2 yr-1) (Chapin et al. 2002).
The simulated monthly fluctuations in NPP and NEP were compared to an Ameriflux site (Blodgett
Forest, CA, http://ameriflux.ornl.gov/) within the same reference site and under the same conditions as
the simulation for ANPP. From the Ameriflux data, maximum NPP and minimum NEP (summer months)
ranged from about 100 to 200 and 75 to 150 g C m-2 mo-1, respectively, while our simulations ranged
from about 100 to 150 and 50 to 125 g C m-2 mo-1, respectively.
Projected monthly climate inputs used for the Century extension (unique by ecoregion) included
average minimum and maximum temperature (C°), standard deviation of mean temperature, total
precipitation (cm), and standard deviation of precipitation. Within the Century extension, these factors
determine monthly tree growth and decomposition, establishment capability, and competition between
cohorts. The climate variables were updated every five years (e.g., 2020-2024) for each emissions
scenario. For consistency, the first five year interval of B1 climate data was used throughout the base
climate, where only current climate was simulated.
2.2.5 Wildfire ignition density modeling
Historical fire occurrence data were acquired from the Pacific Southwest Region 5 USDA USFS GIS shop
(http://www.fs.usda.gov/detail/r5/landmanagement/gis/?cid=STELPRDB5327833). The spatial point
features of fire occurrence data contained information on fire start location, ignition cause, date of
occurrence, and fire size. Although the database contained fire occurrence in the LTB reported as early
as 1949, we chose year 1986 as a cut-off date because this is the year when many agencies began
reporting fire start location as points (latitude/longitude) rather than as within an area (township, range
section). After removing a small proportion (5%) of erroneous records due to duplicates and inaccurate
locations, we obtained 1340 fire occurrence records in the LTB reported between 1986 and 2009.
Among those fires, only 142 fires had a fire size greater than 0.10 ha (0.25 acres), a commonly used
threshold for excluding small fires from statistical analysis (Figure 2-4). Such small fires contribute little
in total area burned and fire risk assessment but may disproportionally affect the analysis of spatial
patterns of fire occurrence (Miranda et al. 2012). Of the 142 fires used in our analysis, approximately
76% were human-caused and 24% were lightning-caused.
17
In addition to effects from climate change on direct tree response and on fire weather, we performed a
separate analysis of potential climate change effects on projected fire ignitions. These projected
ignitions were implemented into the model detailed in the next section (2.2.6)
Wildfire occurs as a function of ignition agents and its suitable biophysical conditions, including
vegetation, terrain, and climate (Krawchuk et al. 2009). Because most fires in the LTB have been humancaused and the spatial distribution of roadways and population determines human access to forests, we
included distance to nearest road, road density, and population density as predictor variables to
examine effects of human-caused ignition agents (Figure 2-5). Both road network and human population
data were obtained from 2000 US Topologically Integrated Geographic Encoding and Referencing
system TIGER/Line files (US Census 2000). We also obtained a GIS database for lightning strike density
from the National Lightning Detection Network to model the effects of lightning strike distribution on
spatial patterns of fire occurrence.
Topography influences both fire ignition and fire spread processes by directly constraining human
accessibility and lightning distribution, affecting local climate, providing fire breaks, and indirectly
affecting fuel moisture, vegetation distribution, and relative humidity (Syphard et al. 2008). Our
topographical variables included elevation, slope, heat load index (McCune and Keon 2002),
topographical position index (TPI), and vector ruggedness measure (VRM). VRM is a topographical
roughness measure that effectively captures variability in slope and aspect into a single variable (Hobson
1972, Sappington et al. 2007). Ruggedness values in the output raster can range from 0 (no terrain
variation) to 1 (complete terrain variation).
We used average July maximum temperature, average January minimum temperature, mean July and
January precipitation to represent the effect of climate on vegetation productivity, rate of fuel
accumulation, and fuel moisture content (Whelan 1995). In addition, a water-balance metric, annual
climatic water deficit (Lutz et al. 2010) was computed to represent water stress to the plants. Water
deficit was calculated as the monthly sum of the difference between reference evapotranspiration and
precipitation. Reference evapotranspiration was calculated using the Penman–Monteith model (Allen et
al. 1998), which uses temperature, radiation, precipitation and wind speed data. It has been shown that
water deficit can be well correlated with plant distribution (Lutz et al. 2010) and wildfire probability
(Parisien et al. 2012).
We used spatial point process (SPP) modeling to quantify effects of various predictor variables on spatial
distribution of fire occurrence. In particular, we fitted our fire occurrence data to the Poisson point
process. A spatial point process (e.g., Poisson, Cox, and Strauss process) is any stochastic mechanism
that generates the spatial point data x. The point process models fitted to the data are often formulated
in terms of their Papangelou conditional density λ(u; x), which may be loosely interpreted as the
conditional probability of having an event at a point u (u∈W) given that the rest of the point process
coincides with x (Baddeley and Turner 2000). For the Poisson point process, the conditional density
function is the same as the density function λ(u; x) =λ(u) because spatial locations are independent of
one another and so their interactions are not considered. The density function of a Poisson point
process is often specified through a log-linear regression model as follows:
18
λ(u) = exp(θ0 +θ1*V1 + …+θn*Vn)
where λ(u) is density at point u, which may be interpreted as the number of events that occurred per
spatio-temporal unit. The V1…Vn are spatial covariates, and θ is the parameter vector (θ0, θ1, …, θn) to be
estimated for the spatial covariates. The density λ(u) will depend on θ to reflect “spatial trend” (a
change in density across the region of observation) or dependence on a covariate. The parameter vector
θ was estimated via a maximum likelihood (Baddeley and Turner 2000) implemented in the ppm
function of the ‘Spatstat’ package in the statistical computing software R. Details about the use of SPP in
fire occurrence modeling are described in (Yang et al. 2007) and (Liu et al. 2012).
We modeled the human- and lightning-caused fires separately using the SPP modeling technique.
Instead of finding a single best choice of model, we employed a model averaging approach that accounts
for model selection uncertainty in order to obtain robust estimates of parameter (θ) and model
predictions (Johnson and Omland 2004). We included the spatial covariates that were revealed to affect
the fire occurrence patterns into the trend term of the log-linear regression model. These spatial
covariates were then transformed, and we used a polynomial function (up to a power of two) to capture
curvilinear effects. We then used all combinations of the predictor variables and their transformations
to construct a wide set of alternative models. AIC scores and Akaike weights were used to modelaverage parameter values and develop robust predictions, for the set of models within 4 AIC units of the
model with lowest AIC (Johnson and Omland 2004). The sum of the Akaike weights over all of the
models in which the parameter of interest appears was then used as a measure of the relative
importance of each predictor variable. This measure was further normalized so that the sum of all
predictor variables’ relative importance was a mathematical unity.
We used the historical climate data (PRISM) and the future climate data predicted by the Canadian
CCCma model under both the A2 and B1 SRES carbon emission scenarios to derive current and future
human- and lightning-caused fire occurrence density raster maps. We then summed the two maps to
obtain total fire occurrence density maps under current and future years. Fire occurrence density maps
were generated at a 100 meter resolution with a unit of fires per 100 km2 per decade.
2.2.6 Modeling fire behavior and effects with the Dynamic Fire Extension
The Dynamic Fire and Fuels extension (hereon called ‘Dynamic Fire’) simulated fire behavior and fire
effects based on parameterized fire regimes. Fire behavior (rate of spread and direction) is a function of
fuel type, weather, topography, and ignition rate (Sturtevant et al. 2009). To determine mortality,
Dynamic Fire uses equations that estimate crown fraction burned (CFB) as an indicator of potential fire
severity (Sturtevant et al., 2009). Actual fire severity (i.e., cohorts killed vs. those that survived)
depended on the tree species present and their relative susceptibility to fire. Susceptibility was
estimated using a combination of rate of spread, fine foliar moisture content (FFMC), and fuel-type
specific parameters. Simulated fire severity was an integer index ranging from 1 to 5, with 1 being the
least severe and 5 being the most severe. Assumed fire behavior ranges from surface fire (classes 1 and
2), torching (class 3), and intermittent crown fire (class 4), through running crown fire (class 5). If a postsimulation fire severity was 5, then all cohorts were killed. For lower severity classes, mortality was
dependent upon the age of the cohorts present and the species fire tolerance, where youngest cohorts
19
are most vulnerable. Post-fire succession depended upon species dispersal into the landscape,
depending on their capacity for dispersal and shade tolerance.
The LTB was divided into three fire regions, representing distinct fire regime characteristics that
influenced fire ignition and spread on the landscape. These regions were created by an interpolation
procedure, involving elevation, climate, and ignition density estimations. The three fire regions
represented, 1) South Lake Tahoe urban areas and some lower elevation areas around the lake shore
(9,603 ha), 2) low-to-mid elevation forested area (28,777 ha), and 3) high elevation forested areas
(31,194 ha). Fires that start in one region may spread to an adjacent region (Table 2-6).
The daily input weather variables used for the Dynamic Fire extension (unique by fire region) included
daily wind speed velocity (km/h), wind direction (degrees), fine fuel moisture code (unitless), buildup
index (unitless), fire weather class, and burn season (i.e., spring, summer, fall) (Sturtevant et al. 2009).
Daily wind direction was obtained from a local weather station (SNOTEL site, Squaw Valley, CA,
http://www.wcc.nrcs.usda.gov/nwcc/site?sitenum=784&state=ca) and used for all climate simulations.
Fine fuel moisture code and buildup index are derived from the Canadian Forest Fire Weather Index
System (Van 1987), with details found in Sturtevant et al. (2009). Daily projected climate values,
including minimum and maximum temperature (C°), total precipitation (cm), and wind speed (Coats et
al. 2010) were used to develop seasonal fire weather variables. The climate variables were updated
every five years (e.g., 2020-2024) for each emissions scenario. For consistency, the first five year
interval of B1 climate data was used throughout the base climate scenario, where only current climate
was simulated.
Although fires are modeled as stochastic events, their frequency and size were parameterized and
calibrated based on the contemporary fire regime and daily fire weather (Sturtevant et al. 2009). Fire
size distribution and fire rotation period were calibrated using local fire occurrence data acquired from
the Pacific Southwest Region 5 USDA USFS and daily fire weather data (described above). The vast
majority of recorded fires in that dataset were small, infrequent and not representative of current forest
conditions and fire regime. To best represent the contemporary fire regime (i.e., without effects from
climate change), only 12 years of data (1995-2007) were used that included the only four fires that were
larger in extent than 100 ha, including the Angora fire (1250 ha) in 2007, the Gondola fire (272 ha) in
2002, the Showers fire (119 ha) in 2002, and the Royal fire (109 ha) in 2003. Fire ignitions were set to
one per year for each fire region to simulate the low number of fires > 0.01 ha found at the LTB. The
simulated fire size distribution was calibrated to the historic record (Figure 2-6).
The resulting fire rotation period (FRP) over five 100 year simulations was 360 (SD=12) years. The FRP is
the length of time required to burn the entire area of interest, in this case the entire LTB. This estimate
was reasonable compared to the recorded estimate of 476 years, given the short term reference dataset
as well as the recent rise in wildfire activity (Safford et al. 2009) that will likely continue into the future
(Westerling et al. 2006). Furthermore, wildfire activity was markedly higher prior to 1880 than these
calibrated values (Beaty and Taylor 2008), and a 360 FRP is therefore a conservative estimate of future
wildfire regime, assuming no climate change.
20
The projected ignition densities were implemented into the landscape model (for A2 and B1 climate
only) represented as the number of fires per year within each fire region. Using the projected ignition
density maps, we used a zonal statistics approach (ArcGIS, ESRI Inc.) to calculate the mean number of
ignitions within each fire region and time step for both emissions scenarios. As these ignitions represent
fires greater than 0.10 ha, they were calibrated to represent the actual number of fires that may occur
within each fire region. This allowed for the model to respond to changes in ignitions through time.
These values were input to the Dynamic Fire table that acted like a lookup table on number of fires,
based on fire region and time step. Due to the exponential nature of the projected ignitions (see Figure
3-26), these values were updated at years 2010 (year 0 of model), 2040, 2060, 2070, 2080, 2090 for B1
and A2 climate. All other Dynamic Fire parameters were identical between time steps and within fire
regions.
2.3 Development and implementation of fuel treatments
To determine our fuel treatment prescriptions and scenarios we used an expert-knowledge approach
similar to Syphard et al. (2011). We invited LTB agency personnel at the federal, state, and local level to
two workshops and requested information on the suite of forest treatments in use at the LTB as well as
fuel treatment information and treatment efficacy. From these workshops we developed fuel treatment
strategies that represent the current and anticipated management activity at the stand to landscape
level. Currently, forest treatment prescriptions in the LTB utilize hand and mechanical treatments
including: chainsaw thinning, cut-to-length harvest with log-forwarding, whole tree yarding, slash piling,
pile burning, mastication, and chipping. Limited prescribed burning is conducted due to the proximity to
populated areas, the importance of tourism, and environmental concerns. LANDIS-ll does not currently
have the capacity to model the different types of thinning or the common slash treatments of pile
burning and mastication due to the coarse-scale nature of the model. Instead, it simulates the effects of
fuels management activities on stand structure by reducing the aboveground live biomass of the agecohorts present on a site, as well as subsequent effects on fire behavior.
2.3.1 Local fuel treatment data and treatment areas
Fuel treatments were simulated within three designated treatment areas within the LTB (Marlow et al.
2007): the defensible space, defense zone, and extended wildland urban interface (WUI) ( Figure 2-7).
These represent the actual management areas in use in the LTB, generally defined by their proximity to
urban areas, structures, or roadways. The defensible space (10, 768 ha, 16% of total forested area) was
the buffer (30 m) around structures in the developed urban core that had the highest priority in
treatment intensity and application through time. Only the forested portion of this treatment area was
simulated; true urban areas (structures, parking lots, roads, etc.) were treated as non-active sites. The
defense zone (8,245 ha, 12 % of total forested area) was defined as a 0.40 km (0.25 mile) buffer from
the edge of the defensible space, representing an area close but not in direct contact with the urban
center of the LTB. The extended WUI (20,473 ha, 30% of total forested area), or threat zone, was
defined as a 2.01 km (1.25 mile) buffer from the defensible space, including highway routes into the
basin. The remaining portion (43%) was designated as a non-treatment area, and generally included the
higher elevations where fire risk was lower, access was difficult, or the area was far from urban
structures and high anthropogenic activity sites.
21
Treatment stands were developed to represent a homogeneous area that would be completely treated
if chosen based on selection criteria and treatment interval. A comprehensive, spatial, project-level fuel
treatment dataset was obtained from the U.S. Forest Service Lake Tahoe Basin Management Unit that
included data from multiple jurisdictions in the LTB for the years 2005 through 2011. Histograms were
generated that depicted the size class distribution of all fuel treatment units within each treatment area
(Figure 2-8). To estimate the density of stands based upon the largest size class in the histogram we
divided the area of the treatment area by the average treatment size in the largest size class of the
histogram. We generated random points using ArcGIS software (ESRI 2011) based upon the density
estimated for that size class. Thiessen polygons were generated in ArcGIS using Dirichlet tesselation
(Dirichlet 1850). Polygons were then randomly selected proportional to the total area from the largest
size class in the histogram. This process was repeated for the next smallest size class of the histogram
using the remaining area of the treatment area. The resulting polygons for each size class and each
treatment area were merged together to create the final stand map. The resulting treatment stand map
represents randomly located polygons of approximately the same size distribution as existing fuel
treatments (Figure 2-9). Areas of non-forest such as meadow, impervious surface, bare ground, and
water were not included in the stand maps.
2.2.2 Assigning fuel types and simulating treatment
The Dynamic Leaf Biomass Fuel System extension (Scheller et al. 2011b) was used to assign a fuel type
for each cell (1 ha) based on the species and age-cohort composition and wood and leaf biomass. A fuel
type drives fire behavior and represents fuel bed and ladder fuels conditions with unique spread
parameters, ignition probabilities, and crown base heights (Sturtevant et al. 2009). For example, sites
with young trees and shrubs (e.g., ladder fuels) will have higher rates of spread and taller flame heights
than old growth sites even though older trees have higher biomass.
We used fuel types similar to those already created for the southern Sierra Nevada (Syphard et al. 2011)
that were based on characteristic species assemblages and age ranges that together exemplify relatively
uniform fire behavior and rates of spread. Fuel types were defined by one shrub type and four basic
forest types: mixed conifer, pine/white fir, red fir, and lodgepole/hemlock. Each type was further
divided into age groups: young, mid-aged, and old. We also created two fuel types to represent how fuel
treatments modify fire behavior and lower fire severity. We added a “fire hazard” stand selection
method (found in v. 2.1.1 of the Base Harvest extension) where we "ranked" the 15 fuel types based on
the maximum rate of spread associated with each fuel type and then averaged it across a stand (if larger
than 1 ha). Further criteria were included to implement restrictions and qualifications related to stand
composition and structure. Stands were not treated if quaking aspen (a LTB conservation species) was
present in >30% portion of the stand area. Furthermore, stands were not treated if they had been
treated or burned within the previous 10 years of the simulation. If a stand was chosen to be treated,
the entire stand was treated.
Fuel treatments were implemented into the LANDIS-II model by using the Leaf Biomass Harvest
extension (v. 2.0.1) to simulate removal of aboveground live leaf and woody biomass of designated
species age-cohorts within selected stands on the landscape (Scheller et al. 2011a). Stands were treated
22
based on their associated fuel type ranking. Essentially, a stand with a higher fuel type ranking (higher
rate of spread, more ladder fuels) had higher fire risk. Stands were then treated in order of fire risk
(highest first) until the predetermined target area to treat was achieved. Target area was based on
rotation period (Gustafson et al. 2000) and described below (see ‘fuel treatment scenarios’).
The Dynamic Fire extension simulated post-treatment effects on fire behavior and subsequent fire
effects. Immediately following treatment, fuel types were reassigned based on the fuel treatment
prescription (see below) and assumed treatment efficacy (10-15 years, depending on prescription type).
After the designated efficacy period, site level fuel types were recalculated. This approach has been
successfully used in other fuel treatment (Scheller et al. 2011a, Syphard et al. 2011) and forest
harvesting studies (Scheller et al. 2011c).
2.2.3 Fuel treatment prescriptions
Fuel treatments were designed to represent the basic prescriptions in use in the LTB including hand and
mechanical thinning of understory and midstory trees up to specified diameter limits. Simulated
treatments targeted six of the ten tree species and all shrub functional groups for thinning including
white fir, red fir, Jeffrey pine, lodgepole pine, incense cedar, and sugar pine. Aspen is considered a
‘conservation species’ in the LTB and whitebark pine, western white pine, and mountain hemlock are
restricted to elevations where fuel treatments would not realistically occur. The six tree species
targeted for thinning were categorized into three removal groups based on management restoration
and conservation goals. Greater proportions of the more shade-tolerant species (white fir and incense
cedar: group 1) that have hugely proliferated under fire suppression were removed preferentially
compared to Jeffrey pine, red fir, and lodgepole pine (group 2). Sugar pine (group 3) was grouped
separately to minimize removal as much as possible because it is a management goal to promote its
distribution and growth in the LTB (Maloney et al. 2011). All shrub functional groups were treated to
emulate mortality from thinning operations and subsequent regeneration (i.e., re-sprouting).
Light thinning
The light thinning prescription (Syphard et al. 2011) was designed to represent the most ubiquitous
treatment currently in use in the LTB: hand-thinning from below of understory and mid-story trees up to
14” (35.6 cm) in diameter, followed by post-treatment activity (e.g., prescribed burning, pile burning).
To simulate realistic thinning operations, variable thinning was distributed across age-cohorts (up to 14”
in diameter), removing more of the younger cohorts (Figure 2-10). From a fire hazard stand point, this
represents reducing ladder fuels and associated fuel loads. This treatment prescription was defined as
having a resulting canopy base height of 4 m, and was effective for 10 years. The light thinning
prescription used the fire hazard stand selection method to choose stands for treatment.
Moderate thinning
The moderate thinning prescription (Syphard et al. 2011) was designed to represent a more intense
thinning application (through mechanical means), where thinning from below of understory and midstory trees up to 30” (76.2 cm) in diameter were removed, followed by post-treatment activity (e.g.,
prescribed burning, pile burning). This treatment prescription was defined as having a resulting canopy
base height of 6 m, and because of the more extensive biomass removal, was effective for 15 years. The
23
same species groups were used for mechanical thinning, but more biomass was removed across agecohorts (up to 30” diameter) (Figure 2-11). For comparison, moderate thinning removed about 20%
more biomass than light thinning.
As mechanical treatments cannot be performed on slopes > 30%, the moderate and light thinning
prescriptions were proportionally distributed within each management area based on the amount of
area with slope ≤ 30%. The moderate thinning prescription was restricted to 52%, 25%, and 16% area
within the defensible space, defense zone, and extended WUI, respectively. The moderate thinning
prescription used the fire hazard stand selection method to choose stands for treatment.
Mid-seral thinning
The mid-seral thinning prescription was developed with an overall goal of promoting more old-growth
characteristics across the landscape (Brown et al. 2004), and was designed as a prospective prescription
that may be employed gradually after the initial round of treatments. The prescription was a
modification of the light thinning prescription (i.e., 14” limit) that targets thinning of mid-seral stage
trees (Figure 2-11), targeting the six species found in the three removal groups. To continue with
restoration efforts and reduction of ladder fuels, younger cohorts of removal group 1 (white fir, incense
cedar) were thinned identically as the light thinning approach. The mid-seral thinning prescription used
a modified form of the fire hazard stand selection method to choose stands based on their mid-seral
stand structure and composition. Fuel types were ranked so both fire hazard and canopy structure
(mid-seral dominance) was taken into account when selecting stands for treatment.
2.3.4 Fuel treatment scenarios
Fuel treatment scenarios were designed to assess how fuel treatment rotation period, prescription type,
and spatial arrangement affect the fire regime, forest composition and structure, as well as carbon
stores and fluxes across the landscape. In all scenarios, the defensible space was treated continuously
on a 15 year rotation period, as a main priority in the basin is to reduce wildfire risk nearby human
communities and infrastructure. In addition, the three treatment areas (defensible space, defense zone,
extended WUI) were fully treated during the first 15 years under all scenarios to emulate the initial
treatment period currently being implemented at the LTB under guidance of the Lake Tahoe MultiJurisdictional Fuel Treatment and Wildfire Prevention Strategy (Marlow et al. 2007). That strategy
outlines an aggressive plan to treat 68,000 acres, or 25% of the total forested land in the LTB during the
period from 2008 to 2016. This includes 49,000 ac of first entry and 19,000 ac of maintenance of
treatments that were initiated in 2001.
The following scenario descriptions pertain to treatments simulated in the defense zone and extended
WUI after the initial 15 year treatment period. The fire hazard stand selection method was used for all
scenarios.
Fuel treatment scenarios of varying prescription types were implemented on a 15 and 30 year rotation
period. A ‘Continued Intensity’ scenario was designed to apply fuel treatments on a designated rotation
period continuously through time. After the initial treatment period, this scenario was evaluated on a
15 (Figure 2-12a) and a 30 year rotation period (Figure 2-12b). Only the light and moderate thinning
prescriptions were used for this scenario, with light thinning being predominantly used on gentle slopes
24
and within the lower elevation areas (e.g., defensible space). A ‘Transition to Forest Health Initiative’
scenario was designed to transition from the light and moderate thinning prescriptions (after the initial
treatment period) to the mid-seral thinning prescription by year 50, using a 15 (Figure 2-12c) and 30
year (Figure 2-12d) rotation period. This scenario was created to represent a prospective approach to
forest thinning that both maintains low fire hazard conditions and promotes old-growth structure.
Finally, a third ‘Long Term Urban Core’ scenario was designed to exclude fuel treatments in the defense
zone and extended WUI beyond the initial treatment period, where treatment applications may become
impeded by a lack of funding. Here, fuel treatments were continued on a 15 year rotation period in the
defensible space only.
In addition to these scenarios, sensitivity tests were simulated to assess varying treatment effective
period (0, 5, 10, 15 years) and effects of post-thinning tree and debris removal (0, 50, 100% removal) in
terms of changes in area burned and Total C, respectively.
2.4 Landscape simulations and analysis
Two approaches were used to assess how climate change might affect long-term forest change. The
first approach consisted of simulating each climate scenario (A2, B1, base climate) separately for 100
years (2010-2110). The landscape was simulated at a 1 ha spatial resolution, and each scenario was
replicated five times to account for the stochastic variation (due to climate, wildfire, and regeneration)
among replicates. We evaluated changes in forest productivity (ANPP), landscape C flux (NEP), and
above and below ground carbon (Live C, Detrital C, SOC, Total C) allocation over time as well as species
response to climate change. Live C included above and below ground live C (bole, leaves, roots).
Detrital C included dead wood (bole, roots), leaf litter, and decomposing C in the surface structural and
metabolic components. SOC included the fast, slow, and passive pools and soil structural and metabolic
components within 1 m soil depth. Total C included all C stored in the system (Live C + Detrital C + SOC).
The second approach explored how projected changes in temperature and precipitation independently
influenced C pools and carbon fluxes. For this analysis, temperature or precipitation values from the A2
climate scenarios were replaced with the values from the base climate scenario. All other parameters,
extensions, and input files were identical to the first approach. Two additional climate scenarios were
created for this experimental approach: a) “A2 temperature only”: A2 temperature with base
precipitation, and b) “A2 precipitation only”: A2 precipitation with base temperature. Furthermore,
simulations without fire were run to isolate the effects of fire (or increased fire activity from climate
change) from climate change effects alone.
For simulations with fuel treatments, there were five difference scenarios. The continued intensity and
transition to Forest Health Initiative scenarios were implemented at a 15 and 30 year rotation period.
The long term urban core scenario only simulated fuel treatments in the defensible space on a 15 year
rotation period for the duration of the simulation. See details on development above.
25
Scenarios of climate change and fuel treatments were run in an iterative process:
Climate change scenarios (Base, B1, A2) were simulated without effects from fuel
treatments or changes in fire ignition patterns.
All fuel treatment scenarios were simulated using the base climate.
Continuous fuel treatments (15 year rotation period) were simulated using the A2 and
B1 climate, without changes in fire ignitions.
A2 and B1 climate were simulated with effects from changes in fire ignition patterns
without effects from fuel treatments.
Continuous fuel treatments (15 year rotation period) were simulated using the A2 and
B1 climate, with changes in fire ignitions.
For all scenario combinations (climate change, fuel treatments, with and without changing ignitions),
five replicates were simulated over 100 years (year 2010 to 2110).
3.0 Results
We begin by assessing the impacts of climate change on wildfire (section 3.1.1) and C dynamics (3.1.2)
independently. We then examine how wildfire effects C dynamics (3.1.3) and how species respond to
predicted changes in climate and wildfire regime (3.1.4). In section 3.2, we focus on how fuel treatments
impact these same elements of wildfire activity, landscape C, and species dynamics under contemporary
climate conditions (i.e., Base climate). Next, we present results on the projected effects of climate
change on wildfire ignition densities (3.3.1) within the LTB and how these changes in ignition patterns
affect model response (3.3.2). Finally, we incorporate predicted increases in ignitions into both climate
change scenarios and examine how fuel treatments modify wildfire activity, C sequestration potential,
and tree species response.
3.1 Impacts of climate change
Both down-scaled climate scenarios showed a clear trend of increasing temperature in the LTB (Figure 31), although differences between emission scenarios were not clear until mid-century. Mean
temperature at the LTB at year 2110 was 5.2° C, 7.3° C, and 9.8° C for the base, B1, and A2 climate
respectively. The projected climate change scenarios (A2, B1) simulated ~22% less precipitation overall
compared to the base climate scenario (Figure 3-1).There was no overall difference in precipitation
between A2 and B1. The down-scaled A2 and B1 climate followed similar temperature and precipitation
trends as the original projected climate data from the LTB (Coats et al. 2010). In the following sections
we assess how these changes in climate (temperature and precipitation) impact wildfires (due to
changes in fire weather, fire season length, and fuel characteristics), landscape C, and species dynamics.
3.1.1 Wildfire dynamics
Simulated wildfire size increased and FRP decreased with increasing C emissions (Table 3-1). The A2
climate resulted in a significantly larger mean and maximum area burned (across 100 years) than base
climate and B1 climate (p< 0.01), but B1 climate was not different than base climate (p> 0.29). More
variability in fire sizes was simulated with higher emissions as well (for A2 vs. base climate only, p=0.02).
26
Although more area burned with the A2 climate, mean area burned remained below 1% of the entire
LTB for all climate scenarios. Even the largest fire (under A2 climate) was 2,600 ha, <4% of the forested
area of the LTB. There was no difference in number of fires between climate scenarios, as these were
calibrated values.
Table 3-1. Simulated fire rotation periods (FRP) and mean and standard deviation of fire sizes and mean
annual area burned at the LTB for all fuel treatment scenarios, and across five replicate 100 year
simulations for the three climate scenarios.
Climate Scenario
Base Climate
B1 Climate
A2 Climate
Fire Rotation
Period (yrs.)
360
340
248
Mean
Fire Size (ha)
70 (110)
70 (128)
89 (165)
Max
Fire Size (ha)
848 (212)
1140 (247)
1784 (1150)
Mean Annual
Area Burned (ha)
192 (91)
206 (105)
279 (165)
Changes in temperature and precipitation ultimately shaped fire weather in the LTB through seasonal
fluctuations of the Fine Fuel Moisture Code (FFMC), the Build-Up-Index (BUI), and a longer fire season
(Sturtevant et al. 2009). A 3% increase from base to A2 climate in FFMC (i.e., ignition potential of fine
fuels) by year 2110, was entirely driven by increased temperature. Changes in precipitation had no
effect on FFMC. In contrast, a 28% increase in BUI (i.e., amount of fuel available for combustion) under
the A2 climate by 2100 was dominantly driven by increased temperature. Precipitation effects on BUI
were more evident when temperature was lower. The A2 precipitation-only scenario simulated the
lowest temperature and precipitation of all scenarios. The variability in precipitation in the A2 climate
scenario created a more variable BUI within the fire season as compared to base climate and reset the
BUI throughout the season with intermittent ‘bursts’ of rainfall. Therefore, despite the lower annual
precipitation of the A2 precipitation-only scenario, the increased seasonal variability in precipitation
lowered the overall availability of combustible fuels (lower average BUI). These interacting effects of
low ignition potential (FFMC) and low available combustible fuel (BUI) over the season resulted in the
lowest overall area burned and fire severity of all scenarios. The effect of rising temperatures
dampened the effect of variable precipitation on BUI, where no differences in average BUI were found
for either A2 climate or A2 temperature-only. In other words, the increased flammability of fine fuels
ultimately outweighed fuel availability. Furthermore, as more area burned throughout the A2 climate,
fewer fuels were available simply because more were consumed by fire (as compared to base climate).
This effect of precipitation on BUI was ultimately minimal compared to the temperature effects on fine
fuel moisture and the overarching impact of rising temperatures drove changes in the system as a
whole.
Changes in precipitation (increase or decrease) will most likely not have a significant effect on fire
weather because most of the rainfall at the LTB either occurs outside the summer season or as snowfall.
The summers (fire season) are already dry and changes in temperature that may cause earlier snowmelt
(and extend the dry season) may have more impact than changes in precipitation throughout the year.
The effect of precipitation was however, inconsequential in comparison to the effects of increasing
27
temperature in determining both fuel flammability and availability with a changing climate. Higher
annual temperatures extended the fire season by 25% by the year 2110.
3.1.2 Carbon dynamics
For our first approach (A2 vs. B1 vs. base climate), our simulations resulted in continued forest growth
and carbon storage into the next century, regardless of changes in climate (Figure 3-2, 3-3). As the
forest continued to age for the next 100 years, simulated live and dead C pools more than doubled in
mass (Figure 3-2) under contemporary climate conditions. With increased temperatures of the A2
climate projections, higher mortality and lower forest productivity (by 20%, Figure 3-3) reduced overall
simulated C sequestration potential by 15% by year 2110 (Figure 3-2, Total C, A2 vs. Base Climate).
NEP consistently remained highest for the base climate scenario, but was variable through time. NEP for
A2 was half that of base climate at year 2110, but the system as a whole remained a C sink throughout
all simulations (NEP > 0). As such, reduced productivity (ANPP), higher C flux from the system (e.g.,
wildfires, heterotrophic respiration), and a reduced rate of carbon storage (in all C pools) from higher
temperatures was most evident for the A2 scenario, where temperatures clearly diverged during the
latter half of the century (yr > 2060). Although temperature differences were seen for the B1 climate
scenario (2.5° C higher than base climate by 2110), overall forest growth (ANPP) and heterotrophic
respiration were minimally affected by this increase (Figure 3-3). There was little to no difference
between all live and dead C pools simulated by B1 and base climate (Figure 3-2). As such, we focused on
comparisons of C dynamics between A2 and base climate scenarios.
Live C (bole, leaves, roots) accounted for ~65% of the total C on the LTB landscape, and followed the
same trend in accumulation as total C, with 17% reduced C storage potential by year 2110 under A2.
Detrital C accounted for ~ 5% of the total C, with a 14% lower value by 2110. Year to year variability in
the detrital pool was due to direct influence of intra-annual fluctuations of fire, age-related tree
mortality, and leaf senescence. SOC accounted for ~35% of the total C. Differences between simulations
of SOC across emissions scenarios were less apparent (compared to e.g., live C) and were most apparent
later in the century (2075-2080). SOC was 6% lower for A2 climate (compared to base climate) by 2110.
Our second approach, where temperature and precipitation effects were independently examined,
demonstrated that increased temperatures (A2 climate) had the largest influence on forest growth, NEP,
and C storage potential in all C pools (Figure 3-4). Changes in precipitation had little effect among
scenarios and may result from temperature feedbacks on long-term soil moisture. Precipitation
variability did however drive annual fluctuations of ANPP, respiration (affecting NEP), and the detrital C
pool, and annual wildfires (see “Climate change and wildfire dynamics”). Total C and live C (Figure 3-4a)
demonstrated the clearest trends; increasing temperature dominantly influenced C, with little effect
from lower precipitation. Interestingly, lower precipitation and base climate temperatures (A2
precipitation only) illustrated higher total C sequestration during the last few decades than for base
climate (same temperature, more precipitation). The indirect influence on wildfires determined these
discrepancies. ANPP (Figure 3-4b) was dominantly driven by temperature, although year-to-year
fluctuations (within simulations) were driven by a response to precipitation (i.e., A2 climate). Despite
28
the year-to-year variability, the overall trend between A2 Temp only and A2 climate was virtually
identical (Figure 3-4). NEP, SOC, and detrital C followed a similar response (data not shown).
3.1.3 Wildfire effects on carbon dynamics
Simulated wildfires influenced carbon sequestration and storage potential in the LTB, regardless of
climate scenario. The influence of climate change (mainly rising temperature) on fire weather and fuel
conditions increased mean annual area burned (by 43%, A2 climate) compared to contemporary
climate. Although a very small portion (<1%) of area across the LTB was burned annually, fire had a
considerable effect on forest carbon dynamics. For example, simulations with fire resulted in a 30-40%
reduction in C storage potential (all pools) than simulations without fire, regardless of climate scenario
(Figure 3-5). Over time, the continuous combustion of dead and live C by wildfire slowed the rate of C
sequestration in both above and belowground C pools. Essentially, the cumulative effects of relatively
small fires over time determined landscape-level mortality patterns and successional response. More
area burned coupled with reduced establishment ability (e.g., subalpine species, see “Climate change
impacts on species dynamics”) hindered landscape-scale ANPP and rate of forest C storage in all pools.
When fire was removed from the system there were little dissimilarities of aboveground C (live, detrital)
between scenarios before year 2080, when the effects of reduced establishment and lower
heterotrophic respiration began to take effect. It is important to note that simulating this system
without fire was entirely experimental and not representative of the system where wildfires are an
inherent and inevitable disturbance that shapes ecosystem dynamics.
3.1.4 Species response to changing climate and fire regime
Species-specific responses to climate change and the resulting increase in fire activity altered simulated
C pools and dynamics between various shade and fire tolerant species. Simulation results indicated that
the growth rate and establishment ability of individual species responded uniquely to changes in
climate. Sugar pine and white fir illustrated positive growth, and to a lesser degree, enhanced
regeneration in response to climate change (Table 3-2). Incense-cedar responded to climate change
during the 100 year simulation with 5% more biomass during the 1st half of the century but there was no
change in biomass by the year 2100 under either climate scenario (data not shown). Likewise, Jeffrey
pine growth was stimulated during the 1st half of the century (e.g., ~10% more biomass) but was
unaffected by the end of the century under B1. For A2 climate, warmer conditions after 2050 slowed
growth and reduced establishment ability and resulted in a net negative effect on biomass.
Both climate change scenarios caused large reductions in the establishment ability of the subalpine and
upper montane community (red fir, western white pine, hemlock, whitebark pine) but little effect on
growth. However, whitebark pine exhibited a slight positive growth response for the A2 climate. Not
surprisingly, enhanced fire activity under both climate scenarios also increased the establishment ability
of fire-adapted re-sprouting shrubs and aspen.
29
Table 3-2. Individual species response to climate change, represented here as an approximate positive
or negative % difference (to nearest 5%) of mean aboveground live biomass between both A2 or B1
climate and base climate at year 2110. Note that there were some instances were species responded to
climate change for a brief period, but illustrated little to no difference in biomass by year 2110. These
species are noted by * and see narrative below for details.
Species or functional group
Pinus jeffreyi*
Pinus lambertiana
Calocedrus decurrens*
Abies concolor
Abies magnifica
Pinus contorta
Pinus monticola
Tsuga mertensiana
Pinus albicaulis*
Populus tremuloides
Non N-fixing resprouting shrubs
Non N-fixing obligate seeding shrubs
N-fixing resprouting shrubs
N-fixing obligate seeding shrubs
Response to Climate Change
+/A2
B1
10%
0%
+
25%
25%
+
0%
0%
+
5%
10%
50%
25%
60%
40%
55%
40%
40%
40%
+
5%*
0%
+
100%
50%
+
95%
40%
+
25%
25%
+
100%
60%
+
5%
5%
Response type
growth, regeneration
growth, regeneration
growth, regeneration
growth, regeneration
regeneration
regeneration
regeneration
regeneration
growth, regeneration (-)
re-sprout after fire
re-sprout after fire
growth
re-sprout after fire
growth
The second approach parsed out the species-specific responses even further, by determining their
susceptibility to increased temperature and reduced precipitation (A2 temperature only, A2
precipitation only), independently. Similar to the carbon results, simulated increases in temperature was
more influential in determining species response than simulated reduction in precipitation, although
magnitude in change was not examined. The growth stimulation of the mixed conifer community and
reduction in establishment ability of the subalpine and montane trees was entirely temperature driven.
Otherwise, higher temperatures affected most species through the change in fire regime. The
differences in precipitation between climate scenarios had no effect on species response.
3.2 Fuel treatment effects under contemporary climate
Here we focus on the effects from fuel treatments on wildfires, landscape C, and species dynamics using
contemporary climate conditions (i.e., Base climate), without the influence from climate change or
associated changes in fire ignition patterns.
3.2.1 Wildfire
Treating all three management areas approximately doubled the fire rotation period (FRP) and cut fire
size in half compared to simulations without fuel treatments (Table 3-3). There were no distinct
differences between the “transition” and “continuous” fuel treatment scenarios on mean FRP or fire
30
size. Treatment rotation period (15 or 30 years) also had little impact. Interestingly, under the long-term
urban core scenario, where treatment continued in the defensible space only (after the first 15 years),
the FRP for the LTB increased from 360 to 510 years and mean fire size declined from 70 to 46 (Table 33).
Table 3-3. Simulated fire rotation periods (FRP) and mean and standard deviation of fire sizes and mean
annual area burned at the LTB for all fuel treatment scenarios, and across five replicate 100 year
simulations, using contemporary climate.
Fuel Treatment Scenario
No Fuel Treatments
Continuous - 15 year RP
Continuous - 30 year RP
Transition - 15 year RP
Transition - 30 year RP
Long Term Urban Core
Fire Rotation
Period (yrs.)
360
800
680
720
704
510
Mean
Fire Size (ha)
70 (110)
28 (50)
35 (56)
34 (58)
34 (74)
46 (74)
Max
Fire Size (ha)
848 (212)
486 (135)
446 (125)
418 (95)
533 (139)
610 (145)
Mean Annual
Area Burned (ha)
192 (91)
87 (56)
102 (51)
98 (53)
100 (47)
137 (66)
Within each of the management areas in the LTB, continuous fuel treatments reduced area burned
(Figure 3-6) and the fire severity index (Figure 3-7). The reduction in area burned and severity was
determined by a decrease in fire intensity associated with fuel treatments. While treatment rotation
period (15 or 30 years) had little apparent effect at this scale, treatment scenario had a small but
significant effect. The long-term urban core scenario reduced area burned in the defensible space (as
intended by the prescription) and this effect also extended into the defense zone, where a reduction in
area burned was observed by year 2110 compared to no fuel treatment, even though the area was only
treated once during the first 15 years (Figure 3-8). However, the long term urban core scenario did not
significantly reduce the fire severity index in any management zone (Figure 3-9) indicating that the
moderating effect on fire spread was due to spatial factors rather than residual structural changes to the
forest. There was no change in simulated number of fires between scenarios, as this was a calibrated
value.
3.2.2 Forest carbon and growth
Forest C sequestration continued into the next century, regardless of fuel treatment or non-fuel
treatment scenario. Fuel treatment simulations illustrated 15% lower total forest C than simulations
without fuel treatments. This C loss was evident at the landscape level (Figure 3-10) and within
management areas (Figure 3-11) with insignificant differences between rotation periods. Net C gain –
where C from simulations with fuel treatment exceeded C from simulations without fuel treatments –
did not occur until the end of the century. Significantly, the long term urban core scenario had a similar
reductive effect across the management areas on C storage (Figure 3-12) as did continuous treatments
on a 15 year rotation period indicating a strong legacy effect of the initial round of treatment.
ANPP (i.e., the rate of forest growth) was generally lower for simulations with fuel treatments, where
tree removal caused up to 23% less mean ANPP across management areas (Figure 3-10). By year 2080,
31
however, ANPP from simulations with fuel treatments began to exceed simulations without fuel
treatments and this continued into the next century. There was little to no difference in ANPP between
fuel treatment scenarios (Figure 3-13).
3.2.3 Species dynamics
Under a no fuel treatment scenario, white fir was strongly dominant with more than two and half to six
times more biomass than any other species (Figure 3-14a). White fir was also the only species with an
increasing biomass trajectory other than sugar pine. At the landscape level, continuous fuel treatments
released red fir, lodgepole pine, and sugar pine which almost doubled in biomass compared to
simulations without fuel treatments (Figure 3-14b). White fir also continued to increase with fuel
treatments, but Jeffrey pine became co-dominant by mid-century and forest composition by the end of
the century was more characteristic of a mixed conifer system. Changes in species biomass were
especially evident within management areas. White fir was strongly suppressed in the defensible space
and defense zone and there was no difference in biomass between rotation periods (15 vs. 30 year)
within these zones. (Figure 3-15). In the extended WUI the 30 year rotation period scenario resulted in
40% higher mean landscape white fir biomass than the 15 year rotation in year 2100, indicating that
more frequent treatment is needed to suppress this fast growing species. In contrast, Jeffrey pine
biomass with fuel treatment (15 or 30 year rotation) surpassed the no treatment scenario by midcentury and by 2100 was about 40% greater (Figure 3-16). Sugar pine showed a similar but less
pronounced response than Jeffrey pine (Figure 3-17).
Fuel treatments primarily removed younger tree cohorts and influenced the resulting species age
structure. By the end of the century, the mean age of Jeffrey pine was greater compared to simulations
without fuel treatments across all management zones (Figure 3-18). There was minimal effect on sugar
pine age structure from fuel treatments in any management zone (Figure 3-19). In contrast, the mean
age of white fir (Figure 3-20) and incense cedar (Figure 3-21) was lower with fuel treatments than
without across all management zones since these two species were targeted for priority removal. The 30
year rotation period resulted in a higher mean age of these two species by year 2100 in select
management zones compared to the 15 year rotation. The 30-year rotation period allowed more time
between treatments for regeneration and forest growth in the defense zone and extended WUI. The
maximum age distribution of all species increased on the landscape as a result of aging, but fuel
treatment allowed shade intolerant species, particularly Jeffrey pine, to increase dramatically (e.g.,
Figure 3-22).
3.3 Predicted wildfire ignition density
Here we present results on the projected effects of climate change on wildfire ignition densities within
the LTB. Then we present how these changes in ignition patterns affect model response.
3.3.1 Modeling results
The relative importance of spatial controls of lightning- and human-caused fire occurrence in LTB varied
greatly. The top six most important predictor variables of lightning-caused fires, in decreasing order, are
elevation, lightning density, January minimum temperature, January precipitation, topographic position
index, and annual water deficit (Figure 3-23a). Three of the top six variables were climatic variables. In
32
contrast, variables describing ignition agents (road density, population density, and distance to road)
and topography (heat load index and elevation) contributed five of the top six variables for humancaused fires (Figure 3-23b). Only one climatic variable (annual water deficit) was identified as a top
variable
Lightning- and human-caused fires also responded very differently to important spatial controls. For
lightning-caused fires, elevation and TPI exerted strong curvilinear effects where fire occurrence density
initially increased but gradually decreased with increasing elevation and TPI. The other four top variables
had linear or only weakly curvilinear effects. Fire occurrence density responded positively to lightning
density, January minimum temperature, and annual water deficit; and it responded negatively to
January precipitation. Among the top six most important predictor variables for the human-caused fires,
only road density exerted monotonic positive effects. Fire occurrence density responded negatively to
heat load index, population density, and elevation. It initially decreased but gradually increased with
increasing distance from roads. Human-caused fire occurrence density showed a unimodal response to
annual water deficit, with the greatest occurrence of human caused fires at intermediate water deficit
values.
Mean fire occurrence density at the LTB landscape level predicted by the top 30 best SPP models varied
greatly, especially at the end of 21st century (Figure 3-24). For the lightning-caused fires, the predicted
density at year 2100 under the SRES A2 scenario could reach as high as 16 fires per 100 sq. km per
decade (about 900% of the current level) or as low as 80% of the current level (Figure 3-24a). These top
30 models had delta AIC score less than 2, indicating their relative weights of evidence were very similar.
However, the range of their predictions was very large, suggesting there would be a high uncertainty if
using a model selection approach to choose one single best model for predicting future conditions.
The modeling average approach produced a more robust lightning-caused fire occurrence density at
year 2100 under A2 scenario as 7.2 fires per 100 sq. km per decade, about 400% of the current level. The
range of predictions for human-caused fires at year 2100, comparing to its current level, was also large:
varying from 400% to 80%. However, such uncertainty was much less than that in predicting future
lightning-caused fires. The mean human-caused fire occurrence density at year 2100 under A2 scenario
was 10.1 fires per 100 sq. km per decade, about 180% of the current level (Figure 3-24c). Both the range
and mean predicted fire occurrence density under B1 scenario were lower than that under A2 scenario
in modeling lightning-caused (Figure 3-24b) and human-caused (Figure 3-24d) fires.
The predicted total fire occurrence density was then calculated as the sum of the predicted modelaveraged human- and lightning-caused fires. The mean landscape total ignition density could increase to
230% and 170% of the current level under the A2 and B1 scenario respectively. Furthermore, our
modeling results showed that the relative increase of lightning-caused fires was much greater than for
human-caused fires. The proportion of lightning-caused fires in the total fires was predicted to increase
from 24% to 42% under the A2 scenario (Figure 3-25). Consequently, the spatial patterns of fire
occurrence hotspots under future climate scenarios are expected to be influenced by the topographical
variables (e.g., land form, elevation) that had stronger effects on lightning-caused fires. For example,
areas that are currently with small fire risk, such as north shore, east shore, and higher ridge lines on the
33
west shore, might become fire occurrence hotspots in the future (Figure 3-26) due to the predicted
faster increase of lightning-caused fires.
3.3.2 Effects of increased wildfire ignitions
Simulated increases in fire ignitions into the coming century had a dramatic impact on wildfire activity
compared to simulations without increased ignitions. By 2110, mean area burned per year under the B1
climate with increased ignitions was more than double base climate values, and was three and a half
times greater under the A2 climate (Figure 3-27). Consequently, the time it took to burn the entire
landscape of the LTB (fire rotation period) was less than half the base climate under the B1 climate and
less than one third the base climate under A2 with increased ignitions (Table 3-4). However, increased
ignitions did not affect mean fire size under B1 climate, and under the A2 climate mean fire size
increased only minimally from 70 to 85 (ha). This was likely a function of the intense suppression efforts
in the LTB that have kept the mean fire size very low, influencing initial model calibration.
Table 3-4. Simulated fire rotation periods (FRP) and mean and standard deviation of fire size and annual
area burned at the LTB for three climate scenarios, with and without increased ignitions across five
replicate 100 year simulations. * with increasing ignitions implemented into model
Climate Scenario
Base Climate
Fire Rotation
Period (yrs.)
360
Mean
Fire Size (ha)
70 (110)
Max
Fire Size (ha)
848 (212)
Mean Annual
Area Burned (ha)
192 (91)
A2 Climate
A2 Climate w/Incr. Ign.*
248
105
89 (165)
85 (151)
1784 (1150)
1494 (1483)
279 (165)
655 (417)
B1 Climate
B1 Climate w/Incr. Ign.*
340
154
70 (128)
65 (100)
1140 (247)
1253 (757)
206 (105)
450 (240)
Increased ignitions also created a more distinct disparity between the three climate scenarios in terms
of forest C sequestration potential and forest growth. The greater amount of fire on the landscape
reduced C sequestration potential for the last half of the century for both the B1 and A2 climate (Figure
3-28). Compared to the continued increase using base climate, total C leveled off around year 2070 and
began to decrease for B1 and A2 climate, respectively. Despite this increase in ignitions, NEP remained
above zero for all the simulations, but was closer to a C neutral state by year 2110 than simulations with
contemporary ignition values. This reduction in forest C was caused by more area burned and higher
mortality rates (more area with high fire severity) across the entire landscape and through time.
Changes in individual tree species biomass with increased ignitions were similar to C trends, e.g. leveling
off for the B1 climate and descending for A2 climate, but the timing and magnitude of the reduction
varied by species (Figure 3-29). The subalpine and montane trees were minimally impacted by increased
ignitions (data not shown). Re-sprouting shrubs had the opposite effect, where biomass increased
dramatically with more fire activity from increased ignitions (Figure 3-30).
34
3.4 Fuel treatment effects under a changing climate
In this section we bring all elements together and implement predicted increases in ignitions in both
climate change scenarios and examine how fuel treatments modify wildfire activity, C sequestration
potential, and tree species response.
3.4.1 Wildfire
Fuel treatments mitigated the effects of increased ignitions and climate change on wildfire activity.
Under both climate scenarios, continuous fuel treatments on a 15 year rotation period prevented the
FRP from decreasing beyond base climate projections and it cut mean fire size roughly in half (Table 35). Mean annual area burned was reduced with fuel treatments under the A2 , but not the B1 climate.
With increased ignitions implemented in the A2 climate, the mean area burned with fuel treatments
was still greater than base climate, but was less than half the area burned if no treatments were applied
(Figure 3-31).
Table 3-5. Simulated fire rotation periods (FRP) and mean and standard deviation of fire size and annual
area burned at the LTB for three climate scenarios, with and without increased ignitions and fuel
treatment scenarios across five replicate 100 year simulations. * with increasing ignitions implemented into model
Fuel Treatment
Scenario
No Fuel Treatments
Climate
Scenario
Base Climate
Fire Rotation
Period (yrs.)
360
Mean Fire
Size (ha)
70 (110)
Max Fire
Size (ha)
848 (212)
Mean Annual
Area Burned (ha)
192 (91)
No Fuel Treatments
A2 Climate
248
89 (165)
1784 (1150)
279 (165)
Continuous - 15 year RP
A2 Climate
657
37 (59)
402 (153)
107 (50)
No Fuel Treatments
A2 Climate
w/Incr. Ign.*
A2 Climate
w/Incr. Ign.*
105
85 (151)
1494 (1483)
655 (417)
226
39 (75)
925 (835)
307 (192)
No Fuel Treatments
B1 Climate
340
70 (128)
1140 (247)
206 (105)
No Fuel Treatments
B1 Climate
w/Incr. Ign.*
B1 Climate
w/Incr. Ign.*
154
65 (100)
1253 (757)
450 (240)
324
33 (59)
620 (602)
214 (198)
Continuous - 15 year RP
Continuous - 15 year RP
3.4.2 Forest carbon and growth
Fuel treatments slightly reduced C sequestration under the base climate scenario, but the reduction was
no longer evident by the end of the century (Figure 3-32). Likewise, fuel treatments minimized C loss
from increased ignitions in simulations using A2 climate, but the gain was achieved much earlier in the
century (year 2070). Continuous fuel treatments substantially increased simulated mean total carbon
(live C + soil organic C+ detrital C) in the A2 climate with increased ignitions compared to no treatments
across the three management areas (Figure 3-33). For instance, C storage potential in the treated
defensible space was about 30-40% greater by the end of the century than without treatments, and
comparable to the base climate scenario with fuel treatments. This effect diminished with increasing
35
distance from the urban core, with little effect outside the treatment area. The mitigation effects from
fuel treatments in the A2 climate were notable across the landscape (Figure 3-34) where greater forest C
was retained compared to no treatment.
3.4.3 Species dynamics
The response of individual species to the combination of increased ignitions and fuel treatments
reflected differences in both treatment prescriptions and in growth and regeneration capabilities (Figure
3-35). White fir was preferentially targeted in all fuel treatment prescriptions and its prevalence on the
landscape was dramatically reduced by year 2110 by continuous 15 year fuel treatments despite
increasing ignitions or the A2 climate change scenario. In contrast, sugar pine was preferentially
retained in fuel treatment prescriptions and its prevalence on the landscape was dramatically increased.
By year 2110, the mean aboveground live biomass of sugar pine was about 40% greater with fuel
treatments compared to an ignition only A2 climate scenario. Jeffrey pine followed a similar pattern,
with almost 50% great biomass by year 2100 with fuel treatments. The subalpine and montane trees
were minimally impacted by fuel treatments in a changing climate (e.g., red fir biomass was lower under
all A2 climate scenarios, regardless of increased ignitions of treatments). Fuel treatments minimized the
strong positive response of re-sprouting shrubs to increased ignitions in a changing climate, partly
because they were mostly removed by treatment and also because less area was burned, leading to
fewer opportunities for vigorous re-sprouting (Figure 3-36).
4.0 Discussion
Our research highlights how past disturbance and future climate change may have multi-faceted effects
on a specific forested landscape (LTB). Our study illustrated the potential for continued forest growth
and sequestration of above and below ground C across the Lake Tahoe Basin, despite any potential
shifts in climate into the coming decades. The forest is essentially a C sink (+NEP), regardless of changes
in climate, with higher growth rates than emissions from ecosystem respiration and wildfires. The net C
flow into the system is predicted to slow towards the end of the 21st century – at a faster rate for A2
than base climate – and it was unclear whether this trend would continue or if it could switch to a C
source (NEP < 0). As the forest matures, climate effects – both direct (e.g., loss of establishment,
enhanced growth) and indirect (e.g., increased area burned) – may become more evident especially if
climate follows the more extreme predictions (e.g., GFDL A2 climate).
The biomass accumulation over the 100 year simulation was quite large (50% increase) and was
predominately a product of regrowth following the Comstock Era logging. As many of the LTB tree
species may live up to 500 years (Dolanc et al. 2012), those trees that established following this abrupt,
yet intense harvesting period are still relatively young and have not reached their potential size and
biomass. The high rate of growth suggests that the legacy effects of intense clear cut logging and fire
suppression will continue into the next century or longer, despite the potential effects of climate
change.
These legacy effects are evident across the world. A global analysis of carbon flux from 1850-2000
(Houghton 2003a), illustrated that land use practices are the driving mechanism of global C balance, and
36
northern mid-latitude regions (e.g., USA, Europe) were transitioning to C sinks towards the end of the
20th century due to enhanced storage of C in forests. In WI, USA, only half of the forest C has recovered
since pre-EuroAmerican settlement (1850s), and C storage may continue due to further fire suppression,
forest in-growth, and forest recovery (Rhemtulla et al. 2009). In Europe, reforestation of abandoned
areas in the European Alps has continued for over 150 years (Tasser et al. 2007). This extensive forest
recovery may trump potential effects of climate change, at least into the coming decades. For example,
vast forest growth and recovery was projected in MA, USA over the next 50 years; despite climate
change and changes in land use patterns (Thompson et al. 2011), where reforestation has been
prominent after centuries of extensive logging, settlement, and agriculture.
Our research further suggests that increases in wildfire activity from climate change, especially with
projected increases in wildfire ignitions, may exert the strongest influence on forest species response
and C storage potentials. Increased wildfire activity caused by climate change substantially increased
overall tree mortality and influenced the biomass accumulation of all species. Furthermore, simulated
fuels treatments were effective at moderating the effects of wildfire under a contemporary climate
scenario and climate change intensified the moderating effects in some instances. Although fuels
treatments may reduce C storage (net C loss) in the short term, the forest of the LTB will continue to
sequester C, remain a C sink, and eventually a net gain in C would be reached. Furthermore, the
benefits of reduced fire risk and improving forest composition and structure in the long-run could
outweigh the C lost from fuel treatments in the shorter term.
4.1 Climate change effects on growth rates and wildfire activity
The higher emissions climate scenario (A2 climate) had greater effects on forest C and community
dynamics than the lower emissions scenario (B1 climate). The influence from climate change was
primarily driven by temperature increases which affected wildfire activity, forest growth rates,
heterotrophic respiration, mortality patterns, and tree establishment ability. Yearly fluctuations in
precipitation, however, influenced annual growth increment and C flux as well as within season fuel
conditions, that when coupled with higher temperatures ultimately determined long-term forest
productivity and C storage potential.
The effect of climate change on wildfire activity was the most influential factor on forest growth and C
sequestration potential. Increasing temperature from the A2 climate ultimately shaped fire weather by
increasing fine fuel flammability and availability and increasing the fire season length. Higher summer
temperatures during a longer growing season lowered fine fuel moisture content throughout (earlier
snowmelt in the spring, (Westerling et al. 2006)) and across the season and increased their ignition
potential and availability into the coming century. Wildfires in the LTB are currently of high intensity
and cause high severity (85-100% mortality, (Safford et al. 2009)), and these more extreme fire
conditions may significantly increase area burned in the future. With continued high severity wildfires,
sequestering more forest C may become more difficult. This is especially true given current stand
conditions that promote higher wildfire severity, i.e., high surface fuel loads combined with high density
of younger trees (ladder fuels) and snags (Agee and Skinner 2005). Similar effects of changing climatic
conditions on wildfire activity have been found throughout western forests, including the Sierra Nevada
37
(Littell et al. 2009), as well as in boreal systems (Girardin et al. 2010), south African systems (Archibald et
al. 2009), and projected globally (Gonzalez et al. 2010).
The effects of increased fire activity on forest productivity and C sequestration were most likely
moderate estimates, due to the conservative nature of the contemporary fire regime. The historic LTB
fire regime, reconstructed from fire scars (c. 1700-1880) had more frequent low severity fires and was
substantially different than the current (c. 1880-2000) regime (Beaty and Taylor 2008). Based on fire
return interval data reconstructed from fire scars, it has been estimated that between 850-3200 ha
burned every year in the LTB (Manley et al. 2000) prior to European settlement. Since 1906, less than
3600 ha have burned in the LTB. For our study, we calibrated our model using data from 1995-2007 that
included a low frequency of fire, the majority of which were small. With the current state of fuel
conditions (e.g., high density of younger cohorts) and altered ignition patterns (e.g., primarily humancaused, (Safford et al. 2009), increasing fire activity is highly probable notwithstanding climate change
effects. On the other hand, compared to the surrounding Sierra Nevada,, fires in the LTB are an order of
magnitude smaller in area burned due primarily to intensive suppression in the WUI but also to rugged
topography (e.g.Littell et al. 2009). In addition, the shape and topographic structure of the Basin is selfcontained, with sparse vegetation and rocky outcrops along the mountain ridges, where wildfires would
not likely spread from or into the adjacent Sierra Nevada forest.
Although there is a strong linkage between climate and fire occurrence, other spatial controls such as
topography, vegetation, and ignition sources can also exert strong influences that may confound our
prediction of fire response to climate change (Liu et al. 2012, Parisien et al. 2012). Although the mean
area burned across the LTB was <1%,.a 40% decrease in total C sequestration potential by 2110 was
observed with increased ignitions with the A2 climate compared to contemporary climate projections.
This reduction dwarfed the 15% reduction in total C storage that was observed with the A2 climate
alone, without increased ignitions. The LTB has one of the highest ignition rates in the Sierra Nevada,
but the amount of area burned has been kept very small because most are concentrated around urban
areas where response time is the shortest in the region (Manley et al. 2000). Still, if ignitions increase
and wildfire occurrence continues to rise, this system could potentially become a net C source, with
increased emissions from wildfire and reduced C storage potential (Meigs et al. 2011, North and
Hurteau 2011).
4.2 Climate change effects on carbon sequestration potential and net carbon
emissions
Our study illustrated how a varying climate can influence the rate of C sequestration into the coming
century by limiting establishment and overall productivity and by stimulating wildfire activity. The
aboveground C pools were especially sensitive to the A2 climate, where higher temperatures lowered
live and detrital C storage potential through enhanced fire activity and limited establishment of some
species. The amount of coarse wood and litter fluctuated with annual tree and shrub mortality (from
senescence and wildfire) as well as variation in precipitation.
Soils of the LTB also sequestered C, regardless of climate regime due to a higher influx from detrital to
the SOC pools than outflux due to soil respiration. The inputs of C from the live and detrital pools
38
determined the amount of C stored in soils. The transfer of C from detritus to SOC was not immediate,
with time lags associated with humification. Soil C turnover ranges from 7-65 years in Sierra Nevada
forests (Trumbore et al. 1996), suggesting the climate driven processes and effects from varying inputs
could span into the next century, beyond the timeline of this study. Higher temperatures (A2 climate)
lowered landscape level heterotrophic respiration; although a more in-depth analysis (e.g., in situ
measurements) may reveal additional insights at individual sites within the Basin. At a nearby Sierra
Nevada forest, Trumbore et al. (1996) attributed increasing temperatures as a dominant driver
influencing soil C turnover. Furthermore, the effects of temperature on decomposition may vary within
and across forest stands (Trumbore et al. 1996), but as our study suggests, may have little impact on the
landscape SOC pool of the LTB. The effects of climate on soil respiration may however, affect overall C
flux (NEP) in the future (Conant et al. 2011).
The potential impact of temperature or drought on soil respiration (stimulation or suppression) may be
overshadowed by the overall effect of changes in wildfire regime that determine the inputs (detrital C).
In this study, wildfires suppressed SOC accumulation, regardless of climate scenario, because much of
the live and detrital C that may have otherwise been incorporated into soil through humification was
consumed during fire. The slow response of SOC to changes in detrital C coupled with increased fire
activity (less inputs) interacted in determining overall SOC, as discrepancies between A2 and base
climate were minimal until the last two decades of the simulations. This contrasts with results for live C,
where discrepancies between A2 and base climate were evident for the last 50 years. Unique to this
study, SOC was not depleted due to increased soil respiration (Trumbore et al. 1996), although
feedbacks with reduced precipitation (suppressed respiration) may cause opposing feedbacks
(Schindlbacher et al. 2012).
4.3 Climate change effects on individual species
Altered mortality patterns, caused by increased wildfire activity under A2 climate, had a stronger effect
on overall productivity and C storage potential than climate change effects (e.g., growth, establishment)
on individual species. For instance, all species experienced high mortality rates because all fires were of
high severity. Differences in successional trajectories (A2 vs. base climate) were minimal in relative
biomass, but the enhanced fire regime of the A2 climate resulted in more areas for regeneration (and
therefore more young cohorts) and more shrub cover (due to re-sprouting after fire) across the basin
(data not shown). This suggests that alterations in fire regime caused by climate change may exceed any
direct forest response. Additionally, the response of fire regimes to changes in climate is ultimately
faster than the positive or negative effects of climate on individual tree species or community response,
apart from fire (Dale et al. 2001). This emphasizes the importance of understanding how subtle regional
changes in temperature and precipitation influence fire weather, fuel conditions, and changes in fire
season, which may impact systems globally (Archibald et al. 2009).
Increased wildfire activity caused by climate change substantially increased overall tree mortality and
influenced the biomass accumulation of all species. Fires were of high severity, similar to recent
wildfires, where up to 100% of overstory trees may perish (Safford et al. 2009). These simulated
wildfires also provided regeneration opportunities and other forms of propagation (e.g., re-sprouting
aspen and shrubs). Although white fir, a shade-tolerant and fire-sensitive species, experienced high
39
mortality rates when wildfires were simulated, it dominated the landscape regardless of climate
scenario and whether fires were simulated or not (complete fire exclusion). This is not surprising, as this
species is already well established across the landscape (Barbour et al. 2002), is a prolific seeder, and
establishes and grows well in post-disturbance conditions (Zald et al. 2008). Although area burned
increased with the A2 climate change scenario, the proportion area burned was still low (<1% per year)
and canopy closure was still prevalent throughout the Basin. For more shade-intolerant species, such as
Jeffrey pine and sugar pine, this canopy closure eventually suppressed their growth and establishment.
Their growth stimulation prior to this (< 2060) reflects their adaptation to lower soil moisture
conditions, compared to the other species modeled.
The subalpine community may be especially vulnerable to climate change. We found that a 2° C
increase in mean annual temperature from current conditions (5 to 7 °C) was their threshold for
suppressing establishment. This loss of recruitment by the subalpine species had a significant impact on
forest C, although biomass of the individual species minimally declined (i.e., remained level through
time), for A2 and B1 climate. Biomass did not decline from initial conditions because those trees already
present in the subalpine area continued to grow and the community persisted. The effect of climate
change on the established, adult trees will likely demonstrate a time-lag in response to rising
temperatures. A strong biomass response may not be observed until decades beyond critical climate
thresholds. This lag effect has been found in other empirical studies of subalpine communities of the
Sierra Nevada where climate changes of the 20th century were examined (Dolanc et al. 2012). This
lagged response may be staggered with decadal climatic variability (Millar et al. 2004), and other nonmodeled disturbance (e.g., insect outbreaks (Adams et al. 2009), loss of Clark’s nutcracker for whitebark
pine (Hutchins and Lanner 1982)), further delaying landscape-level response.
These positive and negative feedbacks on individual species determined the landscape level C response
to climate change. As temperatures increase, some tree species may be stimulated (although
potentially short lived), have recruitment restricted (subalpine community), and impacted by a changing
fire regime (and other natural and anthropogenic disturbances), which all act in concert to determine
long-term C feedbacks and ultimately C sequestration potential. Although the B1 climate regime
illustrated similar landscape productivity and C sequestration potential as base climate, individual
species response was more distinct. This study suggests that for many tree species, the negative
impacts from climate change (increased wildfire, recruitment restriction) may outweigh the potential
positive effects (growth stimulation). This study provides the first comprehensive projection of changes
in total ecosystem C and forest community in response to a changing climate within the LTB.
4.3 Fuel treatment effectiveness under contemporary climate
In our study, rotation period and prescription type were of less importance than spatial arrangement
and location when examining fuel treatment effectiveness, although we recognize there may be local
scale variation evident between prescription types (Schmidt et al. 2008, Safford et al. 2009). This was
mainly because treatments were applied in each time step where the interface with simulated wildfires
was more likely. We found that strategic placement and knowledge of how fuel treatments influence
resulting fire behavior were critical factors for reducing fire spread over time and that the amount of
area treated was less important. Since all prescriptions significantly reduced fire spread potential
40
compared to non-treated areas, the differences between prescription types were minimal at the scale
modeled. We concluded that if quality treatments are strategically placed over long periods, the overall
effect on mitigating wildfires at the landscape level may overshadow the underlying differences
between prescription types.
The long-term urban core scenario demonstrated the significance of strategic placement. Implementing
an initial round of treatments across the three management areas (first 15 years) and continuing only
with treatments in the urban core lead to smaller mean fire size across the entire landscape. This was
mainly due to a significant reduction in the area burned (lower fire spread rate) within the defensible
space.
From a C sequestration standpoint, the implementation of fuel treatments may result in both short and
long-term tradeoffs. Over the near future more forest C would be removed from the system than would
be released without treatment; creating a net C ‘cost’ (Mitchell et al. 2009, Campbell et al. 2012).
However, the C loss occurs in the younger cohorts targeted by the treatments, resulting in a reduction in
ladder fuels and fire spread that mitigates wildfire risk by reducing average fire size and the annual area
burned. If the fire regime (e.g., FRP) and climate remain similar to contemporary conditions, a net gain
in C could take decades with ongoing fuels treatment, but eventually reduced fire severity (lower tree
mortality) and enhanced forest re-growth would lead to a net increase in C storage at the landscape
level. Returning stand structure to a state more similar to historic, more fire resistant forest conditions
(Hurteau and North 2009, North and Hurteau 2011) is usually a higher priority management goal than
increasing C storage potential. Ultimately, there is a balance for demand of maximizing C sequestration
and demand for reducing wildfire severity and altering forest composition and structure (Mitchell et al.
2009).
Fuel treatments were especially effective in altering forest composition and species dominance patterns,
particularly between well established and competing species, such as white fir and Jeffrey pine. In
managed areas, white fir was suppressed and better regeneration conditions were created for shadesensitive trees. Fuel treatments allowed release of Jeffrey pine, sugar pine and red fir, resulting in a
more balanced mixed-conifer system, where Jeffrey pine and white fir shared dominance and other
species were better represented on the landscape. Such changes in forest composition are important for
LTB managers for addressing the imbalances created by logging and fire suppression and restoring the
forest to more historic-like conditions.
Each fuel treatment prescription (e.g., light thinning) was calibrated to represent how on-the-ground
fuel treatments influence fire behavior at the scale modeled. These calibrations instilled two model
assumptions. One assumption was that each thinning prescription was complete, including any posttreatment activity such as pile burning or controlled burning. Another assumption was that these
treatments reduced fire spread potential for 10-15 years, depending on prescription. We tested the
second assumption by varying this effective treatment period between 5, 10, and 15 years.
Interestingly, little to no differences was found in area burned between these periods due to the
continuous application and spatial configuration of fuel treatments through time.
41
4.4 Fuel treatments in a changing climate
Under a changing climate, projected changes in the wildfire regime may be the most influential factor
affecting the forests of the LTB. Climate change may cause a longer, drier growing season with higher
fire ignition and fire spread potential. Our study also projected an increase in potential wildfire ignitions
that dramatically influenced the fire regime and resulting forest growth and C sequestration potential. If
ignitions increase as projected here, coupled with these more fire-prone conditions, tree mortality may
be high and transformations in forest structure and composition may be expected (Beaty and Taylor
2008).
Fuel treatments have the potential to mitigate for this increase in fire activity and reduce C loss from the
system. As ignitions increase however, mitigation feedbacks from fuel treatments may become more
difficult as C storage potential drops below base climate projections by the end of the century.
Interestingly, when fires were more prevalent on the landscape under the A2 climate, a net gain in
forest C from implementing fuel treatments would come earlier than base climate simulations. Under
contemporary climate, a net gain in C, when C storage with fuel treatments exceeds C storage with no
fuel treatments, may not occur on the landscape until year 2100. With impacts from climate change
and increased ignitions, the net gain in C from implementing fuel treatments could come much earlier (>
2065), where wildfires intersect more treated area over time, but the total C storage potential was still
less than base climate projections. This suggests that fuel treatment effectiveness could potentially be
more effective and more essential as the climate changes.
Regardless of climate scenario or ignition pattern, fuel treatments controlled species dynamics by
removing target species (e.g., white fir), reducing fire severity, and creating a regeneration environment
for shade sensitive species. Fuel treatments selectively targeted white fir and effectively reduced
biomass regardless of climate scenario. Interestingly, increased ignitions with the A2 climate increased
the effectiveness of fuel treatment suppression of white fir. In contrast, Jeffrey pine and sugar pine
responded positively to fuel treatments with large increases in biomass, as reduced fire severity and an
open canopy created a better regeneration environment. The subalpine and montane trees were
minimally impacted by increased ignitions as fires rarely reached into the high elevation areas. They
were more impacted by reduced establishment ability in a changing climate than potential feedbacks
from fuel treatments.
5.0 Further considerations
This research included many ecosystem processes (e.g., C cycling, fire disturbance, forest succession)
that interact in unique ways across a forested landscape and in response to active management.
Another process that may be useful to explore could include the potential influence of CO2
concentrations on growth, although temperature may have a more influential negative effect (Grace et
al. 2002). Others include mortality from drought stress and feedbacks with fire, insect, disease
outbreaks (Adams et al. 2009). To address some of these concerns, this modeling project has been
extended in the SNPLMA funded project (round 12) “Drought stress and bark beetle outbreaks in the
future forest: Extending an existing model to inform climate change adaptation” (Scheller, Loudermilk,
42
Hurteau, Weisberg, Skinner). Influences from climate change coupled with these biotic and abiotic
disturbances may accelerate the transition to a C emitting or C neutral system (Kurz et al. 1995, Adams
et al. 2009). We recommend comparative phenology studies that focus on impacts of a longer growing
season (Gunderson et al. 2011, Jeong et al. 2011), as well as a longer fire season, on C pools.
6.0 Conclusions
Forested landscapes are subject to increasingly diverse and often competing demands from society. In
the Lake Tahoe Basin (LTB), managers must balance forest health objectives to restore fire-adapted
ecosystems and protect wildlife habitat with fuels management objectives to reduce the threat of
wildfire and protect communities. In the near future, these objectives may also include storing carbon
(C) or limiting C emissions. Managing the forested landscape in the LTB to meet the multiple goals of
improved forest health, reduced fire risk, and atmospheric C regulation presents new challenges,
especially in the context of changing climate regimes and altered disturbance regimes. Faced with the
prospect of increasing regulations of carbon emissions, managers may be forced to balance the use of
forest treatments for reducing fire risk against the implications for carbon sequestration. Properly
balancing the spatial arrangement of management activities in order to achieve multiple objectives on
the landscape (e.g., Daugherty and Fried 2007, Rhodes and Baker 2008, Schmidt et al. 2008) requires
more information about the inherent trade-offs among these objectives and improved awareness of the
opportunities for optimizing management at the landscape scale (Scheller et al. Biol. Cons. in review).
This research demonstrates an operational method for explicit consideration of potential trade-offs
among management objectives for reducing fire risk, improving forest resiliency to climate induced
changes in drought and wildfire regimes, and sequestering C or reducing C emissions. Our modeling
approach provided an avenue to analyze the emergent outcomes of interactions among climate,
wildfire, forest succession, and nutrient cycling between above and belowground nutrient pools. This
research illustrated the influence of multiple landscape-level processes and disturbances, including
active management that interact over long spatial and temporal scales to drive forest carbon and
species dynamics. Our research suggests that increases in fire weather, ignitions, and wildfire activity
from climate change may exert the strongest influence on forest species response and C storage
potentials and emphasizes the need to incorporate these processes when addressing questions about
climate change on forest response. The response of individual species to climate change may be unique
within forested regions, but effects on overall landscape dynamics may be more distinctively influenced
by feedbacks associated with climate-induced disturbances like wildfire.
The future response of forest ecosystem C cycling in many forested systems worldwide may depend
more on landscape legacies related to land use or major disturbances than on projected climate change
alone. This is especially true in the LTB, where a landscape legacy effect of increasing C storage through
the coming century is an apparent result of past intensive logging and subsequent fire suppression.
Despite these legacies, simulated fuels treatments were effective at moderating the effects of wildfire in
the LTB under a contemporary climate scenario and climate change intensified the moderating effects in
some instances. Although fuels treatments may reduce C storage in the short term, the benefits of
reduced fire risk and improved forest species balance provide compelling evidence that fuels treatments
43
will likely remain an important and perhaps critical component of forest management in the coming
decades.
7.0 Acknowledgements
We would like to thank all of the Lake Tahoe Basin agency personnel at the federal, state, and local level
who participated in two workshops to help us develop fuel treatment strategies that represent the
current and anticipated management activity in the region: Dave Fournier (USDA Forest Service, Lake
Tahoe Basin Management Unit), Bruce Goines (USDA Forest Service R5), Joe Guzman (Spectir), Jeff Haas
(NV Division of State Lands), Brian Hirt (CA Tahoe Conservancy), Michael Hogan (IERS), Jack Landy
(Environmental Protection Agency), Stewart McMorrow (North Tahoe Fire Protection District), Jon Palm
Consultant, Michael Papa (USDA Forest Service, Lake Tahoe Basin Management Unit), Shane Romsos
(Tahoe Regional Planning Agency), Hugh Safford (USDA Forest Service R5), Dan Shaw (CA State Parks),
Roland Shaw (Nevada Division Forestry ), Megan Sheeline (Tahoe Regional Planning Agency), Doug
Taggert (Meeks Bay Fire Protection District), Mike Vollmer (Tahoe Regional Planning Agency). We thank
the Tahoe Regional Planning Agency and the Tahoe Center for Environmental Studies for hosting the
workshops. We would especially like to thank Tiffany Van Huysen of the USDA Forest Service Pacific
Southwest Research Station for her support, administration of the grant, and participation in the agency
workshops. We obtained a valuable GIS database from Kurt Tueber (USDA Forest Service, Lake Tahoe
Basin Management Unit). Tom Dilts (University of Nevada Reno, UNR) provided GIS support, developing
input maps for the model and processing the downscaled projected climate data for model use. Sarah
Karam (UNR) provided many input model parameters associated with species life history and
physiological attributes. Robert Coates (University of California Davis) graciously provided the
downscaled projected climate data (SRES: A2, B1 emissions scenarios) for the Lake Tahoe Basin (Coates
et al. 2010).
This research was supported using funds provided by the Bureau of Land Management through the sale
of public lands as authorized by the Southern Nevada Public Land Management Act (SNPLMA), and was
funded in part through grant 10 DG-11272170-038 from the USDA Forest Service Pacific Southwest
Research Station. We thank Portland State University (PSU) and UNR for their administrative support.
The research team of the Dynamic Ecosystems and Landscapes Lab, PSU was especially important for
the success of this project. The views in this report are those of the authors and do not necessary reflect
those of the USDA Forest Service Pacific Southwest Research Station or Bureau of Land Management.
44
7.0 References
Adams, H. D., M. Guardiola-Claramonte, G. A. Barron-Gafford, J. C. Villegas, D. D. Breshears, C. B. Zou, P.
A. Troch, and T. E. Huxman. 2009. Temperature sensitivity of drought-induced tree mortality
portends increased regional die-off under global-change-type drought. Proceedings of the
National Academy of Sciences 106:7063-7066.
Agee, J. K. and C. N. Skinner. 2005. Basic principles of forest fuel reduction treatments. Forest Ecology
and Management 211:83-96.
Albani, M., D. Medvigy, G. C. Hurtt, and P. R. Moorcroft. 2006. The contributions of land-use change,
CO2 fertilization, and climate variability to the Eastern US carbon sink. Global Change Biology
12:2370-2390.
Allen, RG, Pereira, LS, Raes, D, Smith, M (1998) Crop evapotranspiration-Guidelines for computing crop
water requirements-FAO Irrigation and drainage paper 56. FAO, Rome 300, 6541.
Archibald, S., D. P. Roy, B. W. Van Wilgen, and R. J. Scholes. 2009. What limits fire? An examination of
drivers of burnt area in Southern Africa. Global Change Biology 15:613-630.
Baddeley, A. and R. Turner. 2000. Practical Maximum Pseudolikelihood for Spatial Point Patterns.
Australian & New Zealand Journal of Statistics 42:283-322.
Bar Massada, A, Syphard, AD, Hawbaker, TJ, Stewart, SI, Radeloff, VC (2011) Effects of ignition location
models on the burn patterns of simulated wildfires. Environmental Modelling & Software 26,
583-592.
Barbour, M., E. Kelley, P. Maloney, D. Rizzo, E. Royce, and J. Fites-Kaufmann. 2002. Present and past OldGrowth Forests of the Lake Tahoe Basin, Sierra Nevada, US. Journal of Vegetation Science
13:461-472.
Battles, J., T. Robards, A. Das, K. Waring, J. Gilless, G. Biging, and F. Schurr. 2008. Climate change impacts
on forest growth and tree mortality: a data-driven modeling study in the mixed-conifer forest of
the Sierra Nevada, California. Climatic Change 87:193-213.
Beaty, R. M. and A. H. Taylor. 2008. Fire history and the structure and dynamics of a mixed conifer forest
landscape in the northern Sierra Nevada, Lake Tahoe Basin, California, USA. Forest Ecology and
Management 255:707-719.
Brown, R. T., J. K. Agee, and J. F. Franklin. 2004. Forest Restoration and Fire: Principles in the Context of
Place
Calkin, D. and K. Gebert. 2006. Modeling Fuel Treatment Costs on Forest Service Lands in the Western
United States. Western Journal of Applied Forestry 21:217-221.
Campbell, J., G. Alberti, J. Martin, and B. E. Law. 2009. Carbon dynamics of a ponderosa pine plantation
following a thinning treatment in the northern Sierra Nevada. Forest Ecology and Management
257:453-463.
Campbell, J., D. Donato, D. Azuma, and B. Law. 2007. Pyrogenic carbon emission from a large wildfire in
Oregon, United States. Journal of Geophysical Research 112:1-11.
Campbell, J., M. E. Harmon, and S. R. Mitchell. 2012. Can fuel-reduction treatments really increase forest
carbon storage in the western US by reducing future fire emissions? . Frontiers in Ecology and
the Environment 10:83-90.
Cayan, D., E. Maurer, M. Dettinger, M. Tyree, and K. Hayhoe. 2008. Climate change scenarios for the
California region. Climatic Change 87:21-42.
45
Chapin, F. S., P. A. Matson, and H. A. Mooney. 2002. Terrestrial Production Processes. Principles of
Terrestrial Ecosystem Processes. Springer-Verlag New York, Inc., New York, NY.
Coats, R., J. Reuter, M. Dettinger, J. Riverson, G. Sahoo, G. Schladow, B. Wolfe, and M. Costa-Cabral.
2010. The effects of climate change on Lake Tahoe in the 21st century: meteorology, hydrology,
loading and lake response. UC Davis.
Conant, R. T., M. G. Ryan, G. I. Ågren, H. E. Birge, E. A. Davidson, P. E. Eliasson, S. E. Evans, S. D. Frey, C.
P. Giardina, F. M. Hopkins, R. Hyvönen, M. U. F. Kirschbaum, J. M. Lavallee, J. Leifeld, W. J.
Parton, J. Megan Steinweg, M. D. Wallenstein, J. Å. Martin Wetterstedt, and M. A. Bradford.
2011. Temperature and soil organic matter decomposition rates – synthesis of current
knowledge and a way forward. Global Change Biology 17:3392-3404.
Critchfield, W. B. and E. L. Little Jr. . 1966. Geographic distribution of the pines of the world.
Miscellaneous Publication 991, U.S. Department of Agriculture, Washington, DC.
Dale, V. H., L. A. Joyce, S. McNulty, R. P. Neilson, M. P. Ayres, M. D. Flannigan, P. J. Hanson, L. C. Irland,
A. E. Lugo, C. J. Peterson, D. Simberloff, F. J. Swanson, B. J. Stocks, and B. Michael Wotton. 2001.
Climate Change and Forest Disturbances. BioScience 51:723-734.
Daugherty, P. J. and J. S. Fried. 2007. Jointly Optimizing Selection of Fuel Treatments and Siting of Forest
Biomass‐Based Energy Production Facilities for Landscape‐Scale Fire Hazard Reduction. INFOR:
Information Systems and Operational Research 45:17-30.
Dobrowski, S. Z., J. A. Greenberg, and S. L. Ustin. 2005. Tahoe Basin Exisiting Vegetation Map v. 4.1.
Dolanc, C. R., J. H. Thorne, and H. D. Safford. 2012. Widespread shifts in the demographic structure of
subalpine forests in the Sierra Nevada, California, 1934 to 2007. Global Ecology and
Biogeography:no-no.
Finney, MA (2008) A computational method for optimising fuel treatment locations. International
Journal of Wildland Fire 16, 702-711.
Finney, MA, McHugh, CW, Grenfell, IC, Riley, KL, Short, KC (2011) A simulation of probabilistic wildfire
risk components for the continental United States. Stochastic Environmental Research and Risk
Assessment 25, 973-1000.
Galik, C. S. and R. B. Jackson. 2009. Risks to forest carbon offset projects in a changing climate. Forest
Ecology and Management 257:2209-2216.
Girardin, M. P., A. A. Ali, and C. Hély. 2010. Wildfires in boreal ecosystems: past, present and some
emerging trends. International Journal of Wildland Fire 19:991-995.
Gonzalez, P., R. P. Neilson, J. M. Lenihan, and R. J. Drapek. 2010. Global patterns in the vulnerability of
ecosystems to vegetation shifts due to climate change. Global Ecology & Biogeography 19:755768.
Goodale, C. L., M. J. Apps, R. A. Birdsey, C. B. Field, L. S. Heath, R. A. Houghton, J. C. Jenkins, G. H.
Kohlmaier, W. Kurz, S. Liu, G.-J. Nabuurs, S. Nilsson, and A. Z. Shvidenko. 2002. Forest Carbon
Sinks in the Northern Hemisphere. Ecological Applications 12:891-899.
Gunderson, C. A., N. T. Edwards, A. V. Walker, K. H. O'Hara, C. M. Campion, and P. J. Hanson. 2011.
Forest phenology and a warmer climate - Growing season extension in relation to climatic
provenance. Global Change Biology:1365-2486.
Gustafson, E. J., S. R. Shifley, D. J. Mladenoff, K. K. Nimerfro, and H. S. He. 2000. Spatial simulation of
forest succession and timber harvesting using LANDIS. Canadian Journal of Forest Research
30:32-43.
Houghton, R. A. 2003a. Revised estimates of the annual net flux of carbon to the atmosphere from
changes in land use and land management 1850–2000. Tellus B 55:378-390.
Houghton, R. A. 2003b. Why are estimates of the terrestrial carbon balance so different? Global Change
Biology 9:500-509.
46
Hurteau, M. and M. North. 2009. Fuel treatment effects on tree-based forest carbon storage and
emissions under modeled wildfire scenarios. Frontiers in Ecology and the Environment 7:409414.
Hurteau, M. D., G. W. Koch, and B. A. Hungate. 2008. Carbon protection and fire risk reduction: toward a
full accounting of forest carbon offsets. Frontiers in Ecology and the Environment 6:493-498.
Hutchins, H. E. and R. M. Lanner. 1982. The central role of Clark's nutcracker in the dispersal and
establishment of whitebark pine. Oecologia 55:192-201.
Jeong, S.-J., C.-H. Ho, H.-J. Gim, and M. E. Brown. 2011. Phenology shifts at start vs. end of growing
season in temperate vegetation over the Northern Hemisphere for the period 1982–2008.
Global Change Biology:no-no.
Johnson, J. B. and K. S. Omland. 2004. Model selection in ecology and evolution. Trends in Ecology
&amp; Evolution 19:101-108.
Karam, S. L., P. J. Weisberg, R. M. Scheller, D. W. Johnson, and W. W. Miller. in press. Development and
evaluation of a nutrient cycling extension for the LANDIS-II landscape simulation model.
Ecological Modelling.
Kirschbaum, M. U. F. 1995. The temperature dependence of soil organic matter decomposition, and the
effect of global warming on soil organic C storage. Soil Biology and Biochemistry 27:753-760.
Krawchuk, M., M. Moritz, M. Parisien, J. Van Dorn, and K. Hayhoe. 2009. Global Pyrogeography: the
Current and Future Distribution of Wildfire. PLoS.
Kurz, W. A., M. J. Apps, S. J. Beukema, and T. Lekstrum. 1995. 20th century carbon budget of Canadian
forests. Tellus B 47:170-177.
Lenihan, J. M., R. Drapek, D. Bachelet, and R. P. Neilson. 2003. Climate change effects on vegetation
distribution, carbon, and fire in California. Ecological Applications 13:1667-1681.
Littell, J. S., D. McKenzie, D. L. Peterson, and A. L. Westerling. 2009. Climate and wildfire area burned in
western U.S. ecoprovinces, 1916–2003. Ecological Applications 19:1003-1021.
Liu, Z., J. Yang, Y. Chang, P. J. Weisberg, and H. S. He. 2012. Spatial patterns and drivers of fire
occurrence and its future trend under climate change in a boreal forest of Northeast China.
Global Change Biology 18:2041-2056.
Lutz, JA, van Wagtendonk, JW, Franklin, JF (2010) Climatic water deficit, tree species ranges, and climate
change in Yosemite National Park. Journal of Biogeography 37, 936-950.
Maloney, P. E., D. R. Vogler, A. J. Eckert, C. E. Jensen, and D. B. Neale. 2011. Population biology of sugar
pine (Pinus lambertiana Dougl.) with reference to historical disturbances in the Lake Tahoe
Basin: Implications for restoration. Forest Ecology and Management 262:770-779.
Manley, P.N., J.A. Fites-Kaufman, M.G. Barbour, M.D. Schlesinger, and D.M. Rizzo. 2000. Biological
integrity. P. 403–600 in Lake Tahoe Basin watershed assessment, Vol. 1, Murphy, D.D., and C.M. Knopp
(eds.). Gen.Tech. Rep. PSW-GTR-175. USDA For. Serv., Pacific Southwest Res. Sta., Albany, CA.
Marlow, D., C. French, W. Graham, R. Hall, S. Howle, K. Whisennand, J. York, M. Taylor, S. Holl, and K.
Teuber. 2007. Lake Tahoe Basin: MultiJurisdictional Fuel Reduction and Wildfire Prevention
Strategy 10 Year Plan. Unpublished report on file at the USDA Forest Service, Lake Tahoe Basin
Management Unit, South Lake Tahoe, CA, South Lake Tahoe, CA.
McKenzie, D., Z. e. Gedalof, D. L. Peterson, and P. Mote. 2004. Climatic Change, Wildfire, and
Conservation. Conservation Biology 18:890-902.
Meigs, G., D. Turner, W. Ritts, Z. Yang, and B. Law. 2011. Landscape-Scale Simulation of Heterogeneous
Fire Effects on Pyrogenic Carbon Emissions, Tree Mortality, and Net Ecosystem Production.
Ecosystems 14:758-775.
47
Millar, C., R. Westfall, D. Delany, J. King, and L. Graumlich. 2004. Response of Subalpine Conifers in the
Sierra Nevada, California, U.S.A., to 20th-century Warming and Decadal Climate Variability.
Arctic, Antarctic, and Alpine Research 36:181-200.
Millar, C. I., D. Neilson, R. Batchelet, R. Drapek, and J. M. Lenihan. 2006. Climate change at multiple
scales. Pages 31-62 in H. Salwasser and M. Cloughesy, editors. Forests, carbon, and climate
change: a synthesis of science findings. Oregon Forest Resources Institute, Portland, OR.
Millar, C. I., N. L. Stephenson, and S. L. Stephens. 2007. Climate Change and Forests of the Future:
Managing in the face of Uncertainty. Ecological Applications 17:2145-2151.
Miller, W. W., D. W. Johnson, S. L. Karam, R. F. Walker, and P. J. Weisberg. 2010. A Synthesis of Sierran
Forest Biomass Management Studies and Potential Effects on Water Quality. Forests 1:131-153.
Miranda, B. R., B. R. Sturtevant, S. I. Stewart, and R. B. Hammer. 2012. Spatial and temporal drivers of
wildfire occurrence in the context of rural development in northern Wisconsin, USA.
International Journal of Wildland Fire 21:141-154.
Mitchell, S. R., M. E. Harmon, and K. E. B. O'Connell. 2009. Forest fuel reduction alters fire severity and
long-term carbon storage in three Pacific Northwest ecosystems. Ecological Applications 19:643655.
Nagel, T. A. and A. H. Taylor. 2005. Fire and Persistence of Montane Chaparral in Mixed Conifer Forest
Landscapes in the Northern Sierra Nevada, Lake Tahoe Basin, California, USA. Journal of the
Torrey Botanical Society 132:442-457.
North, M. P. and M. D. Hurteau. 2011. High-severity wildfire effects on carbon stocks and emissions in
fuels treated and untreated forest. Forest Ecology and Management 261:1115-1120.
Ottmar, R. and H. Safford. 2011. FCCS Fuelbeds for the Lake Tahoe Basin Management Unit. Final Report
P018, USDA Forest Service.
Parisien, M-A, Snetsinger, S, Greenberg, JA, Nelson, CR, Schoennagel, T, Dobrowski, SZ, Moritz, MA
(2012) Spatial variability in wildfire probability across the western United States. International
Journal of Wildland Fire 21, 313.
Parsons, D. J. and S. H. DeBenedetti. 1979. Impact of fire suppression on a mixed-conifer forest. Forest
Ecology and Management 2:21-33.
Parton, W. J., D.W. Anderson, C.V. Cole, J.W.B. Stewart. 1983. Simulation of soil organic matter
formation and mineralization in semiarid agroecosystems.in R. L. T. R.R. Lowrance, L.E.
Asmussen and R.A. Leonard, editor. Nutrient cycling in agricultural ecosystems. The University of
Georgia, College of Agriculture Experimental Stations, Athens, Georgia.
Pilliod, D. S., E. L. Bull, J. L. Hayes, and B. C. Wales. 2006. Wildlife and Invertebrate Response to Fuel
Reduction Treatments in Dry Coniferous Forests of the Western United States: A Synthesis.
Rocky Mountain Research Station, USDA Forest Service Fort Collins, CO.
Rhemtulla, J. M., D. J. Mladenoff, and M. K. Clayton. 2009. Historical forest baselines reveal potential for
continued carbon sequestration. Proceedings of the National Academy of Sciences 106:60826087.
Rhodes, J. J. and W. L. Baker. 2008. Fire probability, fuel treatment effectiveness and ecological tradeoffs
in western US public forests. The Open Forest Science Journal 1:1-7.
Richards, K. and K. Andersson. 2001. The leaky sink: persistent obstacles to a forest carbon
sequestration program based on individual projects. Climate Policy 1:41-54.
Rogers, J. H. 1974. Soil survey Tahoe Basin Area: California and Nevada. U.S. Department of Agriculture,
Soil Conservation Service, Washington, D.C.
Safford, H. D., D. A. Schmidt, and C. H. Carlson. 2009. Effects of fuel treatments on fire severity in an
area of wildland–urban interface, Angora Fire, Lake Tahoe Basin, California. Forest Ecology and
Management 258:773-787.
48
Sappington, J, Longshore, KM, Thompson, DB (2007) Quantifying landscape ruggedness for animal
habitat analysis: a case study using bighorn sheep in the Mojave Desert. The Journal of wildlife
management 71, 1419-1426.
Schäfer, K. V. R., R. Oren, D. S. Ellsworth, C. T. Lai, J. D. Herrick, A. C. Finzi, D. D. Richter, and G. G. Katul.
2003. Exposure to an enriched CO2 atmosphere alters carbon assimilation and allocation in a
pine forest ecosystem. Global Change Biology 9:1378-1400.
Scheller, R. and D. Mladenoff. 2008. Simulated effects of climate change, fragmentation, and interspecific competition on tree species migration in northern Wisconsin, USA. Climate Research
36:191-202.
Scheller, R., W. Spencer, H. Rustigian-Romsos, A. Syphard, B. Ward, and J. Strittholt. 2011a. Using
stochastic simulation to evaluate competing risks of wildfires and fuels management on an
isolated forest carnivore. Landscape Ecology 26:1491-1504.
Scheller, R., S. Van Tuyl, K. Clark, J. Hom, and I. La Puma. 2011b. Carbon Sequestration in the New Jersey
Pine Barrens Under Different Scenarios of Fire Management. Ecosystems 14:987-1004.
Scheller, R. M., J. B. Domingo, B. R. Sturtevant, J. S. Williams, A. Rudy, E. J. Gustafson, and D. J.
Mladenoff. 2007. Design, development, and application of LANDIS-II, a spatial landscape
simulation model with flexible temporal and spatial resolution. Ecological Modelling 201:409419.
Scheller, R. M., D. Hua, P. V. Bolstad, R. A. Birdsey, and D. J. Mladenoff. 2011c. The effects of forest
harvest intensity in combination with wind disturbance on carbon dynamics in Lake States Mesic
Forests. Ecological Modelling 222:144-153.
Schindlbacher, A., S. Wunderlich, W. Borken, B. Kitzler, S. Zechmeister-Boltenstern, and R. Jandl. 2012.
Soil respiration under climate change: prolonged summer drought offsets soil warming effects.
Global Change Biology 18:2270-2279.
Schmidt, D. A., C. N. Skinner, and A. H. Taylor. 2008. The influence of fuels treatment and landscape
arrangement on simulated fire behavior, Southern Cascade range, California. Forest Ecology and
Management 255:3170-3184.
Schwilk, D. W., C. N. Skinner, K. W. Outcalt, J. J. Moghaddas, S. L. Stephens, A. Youngblood, D. A. Yaussy,
T. A. Waldrop, J. McIver, E. E. Knapp, J. E. Keeley, J. D. Bailey, R. J. Harrod, C. E. Fiedler, and C. J.
Fettig. 2009. The national Fire and Fire Surrogate study: effects of fuel reduction methods on
forest vegetation structure and fuels. Ecological applications : a publication of the Ecological
Society of America 19:285-304.
Stephens, S. L., and L. W. Ruth. 2005. Federal forest-fire policy in the United States. Ecological
Applications 15:532–542.
Sturtevant, B. R., R. M. Scheller, B. R. Miranda, D. Shinneman, and A. Syphard. 2009. Simulating dynamic
and mixed-severity fire regimes: A process-based fire extension for LANDIS-II. Ecological
Modelling 220:3380-3393.
Syphard, AD, Radeloff, VC, Keuler, NS, Taylor, RS, Hawbaker, TJ, Stewart, SI, Clayton, MK (2008)
Predicting spatial patterns of fire on a southern California landscape. International Journal of
Wildland Fire 17, 602-613.
Syphard, A. D., R. M. Scheller, B. C. Ward, W. D. Spencer, and J. R. Strittholt. 2011. Simulating landscapescale effects of fuels treatments in the Sierra Nevada, California, USA. International Journal of
Wildland Fire 20:364-383.
Tasser, E., J. Walde, U. Tappeiner, A. Teutsch, and W. Noggler. 2007. Land-use changes and natural
reforestation in the Eastern Central Alps. Agriculture, Ecosystems &amp; Environment 118:115129.
49
Thompson, J. R., D. R. Foster, R. M. Scheller, and D. Kittredge. 2011. The influence of land use and
climate change on forest biomass and composition in Massachusetts, USA. Ecological
Applications 21:2425-2444.
Trumbore, S. E., O. A. Chadwick, and R. Amundson. 1996. Rapid Exchange Between Soil Carbon and
Atmospheric Carbon Dioxide Driven by Temperature Change. Science 272:393-396.
van Mantgem, P. J., N. L. Stephenson, J. C. Byrne, L. D. Daniels, J. F. Franklin, P. Z. Fulé, M. E. Harmon, A.
J. Larson, J. M. Smith, A. H. Taylor, and T. T. Veblen. 2009. Widespread Increase of Tree
Mortality Rates in the Western United States. Science 323:521-524.
Van, W. 1987. Development and structure of the Canadian forest fire weather index system. Ottawa
(Ont.).
Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam. 2006. Warming and Earlier Spring
Increase Western U.S. Forest Wildfire Activity. Science 313:940-943.
Whelan, RJ (1995) 'The ecology of fire.' Cambridge University Press, New York, NY. 346 pp.
Wotton, B, Nock, C, Flannigan, M (2010) Forest fire occurrence and climate change in Canada.
International Journal of Wildland Fire 19, 253-271.
Xu, C., G. Z. Gertner, and R. M. Scheller. 2009. Uncertainties in the response of a forest landscape to
global climatic change. Global Change Biology 15:116-131.
Yang, J., H. S. He, S. R. Shifley, and E. J. Gustafson. 2007. Spatial Patterns of Modern Period HumanCaused Fire Occurrence in the Missouri Ozark Highlands. Forest Science 53:1-15.
Zald, H. S. J., A. N. Gray, M. North, and R. A. Kern. 2008. Initial tree regeneration responses to fire and
thinning treatments in a Sierra Nevada mixed-conifer forest, USA. Forest Ecology and
Management 256:168-179.
Zinke, P. J. and A. G. Stangenberger. 2000. Elemental storage of forest soil from local to global scales.
Forest Ecology and Management 138:159-165.
Zinke, P. J., A. G. Stangenberger, W. M. Post, W. R. Emanuel, and J. S. Olson. 1998. Global Organic Soil
Carbon and Nitrogen. Oak Ridge National Laboratory Distributed Active Archive Center, Oak
Ridge, Tennessee.
50
Figures
Figure 2-1. Map of the study area for this research; the Lake Tahoe Basin, CA and NV, USA.
51
Figure 2-2. LANDIS-II climate inputs and processes modeled for the forested landscape of the Lake
Tahoe Basin, CA, NV. Climate data were implemented into the Century extension and Dynamic Fire
extension using monthly and seasonal (daily) inputs, respectively of current and projected climate (years
2010-2110). Processes in the Century extension include above and belowground components (e.g.,
plant root, shoot, and leaf biomass, soil organic matter). FRP: Fire Rotation Period
52
Figure 2-3. Initial landscape composition by vegetative community in the Lake Tahoe Basin (LTB) (after
Ottmar et al. 2009).
53
Figure 2-4. Study area of the Lake Tahoe Basin with reported human- and lightning-caused fires greater
than 0.1 ha in size between 1986 and 2009.
54
Figure 2-5. Maps showing road density, lightning density, elevation, average January minimum
temperature, average July maximum temperature, and annual water deficit in the Lake Tahoe Basin,
overlaid with reported human-caused (red dots) and lightning-caused (blue triangles) fires.
55
0.5
0.45
% Simulated
Frequency (%)
0.4
% Empirical
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
20
50
100 120 200 300 400 500 1300
Fire Size Category (ha)
Figure 2-6. Fire size distribution simulated using LANDIS-II (100 years, using base climate) and empirical
data from the study area.
56
Figure 2-7. Treatment areas within the LTB: the defensible space (16% of total area), defense zone
(12%), and extended wildland urban interface (30%), with 42% of no treatment area.
57
45
Urban
Defense
Extended
Non-WUI
40
35
Frequency
30
25
20
15
10
5
0
5
15
25
35
45 55 65
Size class (ha)
75
85
95
200
Figure 2-8. Size class distribution of all fuel treatment units within each treatment area in the LTB (from
LTBMU Management unit data).
58
Figure 2-9. Treatment stand map representing randomly located polygons of approximately the same
size distribution as existing fuel treatments in the LTB (from LTBMU Management unit data).
59
100
90
80
70
60
50
40
30
20
10
0
Target species
% biomass removed
Other mixed conifer
Sugar pine
10
30
50
70
Age-cohort
90
110
Figure 2-10. Fraction biomass removed in the light thinning prescription by species and age cohorts.
Target species: white fir, incense cedar; Other mixed conifer: Jeffrey pine, red fir, lodgepole pine; sugar
pine: sugar pine only.
% biomass removed
Target species
100
90
80
70
60
50
40
30
20
10
0
Other mixed conifer
Sugar pine
10
30
50
70
90
Age-cohort
110
Figure 2-1. Fraction biomass removed in the moderate thinning prescription by species and age cohorts.
60
% biomass removed
100
90
80
70
60
50
40
30
20
10
0
Target species
Other mixed conifer
Sugar pine
10
30
50
70
Age-cohort
90
110
Figure 2-12. Fraction biomass removed in the mid-seral prescription by species and age cohorts.
Figure 2-13. Continued Intensity’ scenario under a) 15 and b) 30 year rotation period and the Transition
to Forest Health Initiative’ scenario under c) 15 and d) 30 year rotation period.
61
Figure 3-1. Mean a) annual temperature (C°) and b) total annual precipitation (cm) simulated at the LTB
using current climate conditions and two GCMs of high (A2) and low (B1) emissions across 100 years.
This represents landscape level mean and standard errors across five replicates.
62
Figure 3-2. Simulated mean landscape C estimates of a) live C (bole, leaves, roots), b) total C (live C +
detrital C + SOC), c) detrital C, d) SOC (3 soil pools: slow, fast, passive) simulated over 100 years (five
replicates) using contemporary (base) climate and two emissions scenarios (A2, B1).
63
Figure 3-3. Simulated mean landscape C estimates of a) Aboveground Net Primary Productivity, b) Net
Ecosystem Production, and c) heterotrophic respiration, simulated over 100 years (five replicates) and
using contemporary (base) climate and two emissions scenarios (A2, B1). Note difference in scale of yaxis between graphs.
64
Figure 3-4. Results from second simulation approach, where temperature and precipitation were
independently assessed using the A2 and base climate. Increasing temperatures (as opposed to reduced
precipitation) determined impacts on forest C (a, live C) and productivity (b, ANPP). A2 Temperature
only: A2 temperature combined with base climate precipitation; A2 Precipitation only: A2 precipitation
combined with base climate temperature.
Figure 3-5. Comparison of simulated live C with and without fire using the base and A2 climate.
65
Figure 3-6. Simulated mean area burned (ha) across the four fuel treatment management areas within
the LTB, representing mean and standard deviations across five replicate model simulations. FT 15 or 30
RP: Continuous fuel treatments applied on a 15 or 30 year rotation period.
66
Figure 3-7. Simulated mean fire severity (index from 1:5) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year rotation period.
67
Figure 3-8. Simulated mean area burned (ha) across the four fuel treatment management areas within
the LTB, representing mean and standard deviations across five replicate model simulations. FT 15 RP:
Continuous fuel treatments applied on a 15 year rotation period. Long Term Urban Core: Only the
defensible space was treated after the initial treatment period.
68
Figure 3-9. Simulated mean fire severity (index from 1:5) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations. FT 15 RP: Continuous fuel treatments applied on a 15 year rotation period. Long Term
Urban Core: Only the defensible space was treated after the initial treatment period.
69
Figure 3-10. Simulated mean total carbon (live C + soil organic C + detrital C, g C m-2) and Aboveground
Net Primary Productivity (ANPP) across the LTB, representing mean and standard deviations across five
replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year
rotation period.
70
Figure 3-11. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across five
replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year
rotation period.
71
Figure 3-12. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across five
replicate model simulations. FT 15 RP: Continuous fuel treatments applied on a 15 year rotation period.
Long Term Urban Core: Only the defensible space was treated after the initial treatment period.
72
Figure 3-13. Simulated forest growth, represented as ANPP (Aboveground Net Primary Productivity),
across the four fuel treatment management areas within the LTB, representing mean and standard
deviations across five replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied
on a 15 or 30 year rotation period.
73
Figure 3-14. Change in mean aboveground live biomass (g m-2) of the six most abundant tree species at
the LTB for simulations (a) without fuel treatments and (b) with fuel treatments (continuous 15 year
rotation period).
74
Figure 3-15. Mean aboveground live biomass (g per m-2) of the white fir (Abies concolor) across the four
fuel treatment management areas within the LTB, representing mean and standard deviations across
five replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year
rotation period.
75
Figure 3-16. Mean aboveground live biomass (g per m-2) of the Jeffrey pine (Pinus jeffreyi) across the
four fuel treatment management areas within the LTB, representing mean and standard deviations
across five replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30
year rotation period.
76
Figure 3-17. Mean aboveground live biomass (g per m-2) of the sugar pine (Pinus lambertiana) across the
four fuel treatment management areas within the LTB, representing mean and standard deviations
across five replicate model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30
year rotation period.
77
Figure 3-18. Mean age of Jeffrey pine (Pinus jeffreyi) across the four fuel treatment management areas
within the LTB, representing mean and standard deviations across five replicate model simulations. FT
15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year rotation period.
78
Figure 3-19. Mean age of sugar pine (Pinus lambertiana) across the four fuel treatment management
areas within the LTB, representing mean and standard deviations across five replicate model
simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year rotation period.
79
Figure 3-20. Mean age of white fir (Abies concolor) across the four fuel treatment management areas
within the LTB, representing mean and standard deviations across five replicate model simulations. FT
15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year rotation period.
80
Figure 3-21. Mean age of incense cedar (Calocedrus decurrens) across the four fuel treatment
management areas within the LTB, representing mean and standard deviations across five replicate
model simulations. FT 15 or 30 RP: Continuous fuel treatments applied on a 15 or 30 year rotation
period.
81
Year
Base2010
Climate – yr. 100
Year 2110 – No FT
Year 2110
FT – 15RP
Legend
CaloDecu-MAX-0<VALUE>
5.31372549 - 25
Max Age
25.50588236 - 4
1 - 50
51 - 70
71- 95
96 - 140
141- 180
181 – 271+
49.94901962 - 7
70.14117648 - 95
95.64705883 - 1
140.282353 - 18
180.6666668 - 2
Figure 3-22. Max age distribution of incense cedar across the LTB. Left to right: Year 2010 (year 0 of
model), Year 2110 (year 100) without fuel treatments, Year 2110 (year 100) with continuous fuel
treatments applied on a 15 year rotation interval.
82
Figure 3-23. The top 6 most important predictor variables and their marginal effects on fire occurrence
density for (a) lightning-caused and (b) human-caused fires.
83
Figure 3-24. Landscape level mean fire occurrence density predicted by the top 30 best Poisson point
process models under the climate scenario of A2 and B1 for the lightning-caused (a and b) and humancaused (c and d) fires. The unit of fire occurrence density is # of fires per 100 sq. km per decade. Each
line represents one model with red, green, and blue color indicating the ranking order (R > G > B) of
each model’s weight of evidence. The black line represents the modeling average results.
84
Figure 3-25. Predicted landscape-level mean total (human-caused + lightning-caused) fire occurrence
densities (fires per 100 sq. km per decade) and their relative changes (above panel) under the climate A2
and B1 scenarios, as well as time series of predicted proportion of lightning-caused in the total fire
occurrences (bottom panel).
85
(a)
(b)
Figure 3-26. Predictive maps of total ignition density (# of fires per 100 sq. km per decade) at (a) year
2000 and (b) year 2100 under the SRES A2 scenario.
86
Figure 3-27. Simulated mean area burned (ha) across the LTB, representing mean and standard
deviations across five replicate model simulations, using three climate scenarios with current fire
ignition estimates (left) and with increased ignitions (right).
87
Figure 3-28. Simulated mean total carbon (live C + soil organic C + detrital C, g C m-2) and Net Ecosystem
Production (NEP) across the LTB, representing mean and standard deviations across five replicate model
simulations, using three climate scenarios with current fire ignition estimates (top row) and with
increased ignitions (bottom row).
88
Figure 3-29. Simulated mean aboveground live biomass (g m-2) of four representative tree species found
at the LTB, representing mean and standard deviations across five replicate model simulations, using
three climate scenarios with current fire ignition estimates and with increased ignitions. Top, left to
right: white fir, Jeffrey pine, Bottom, left to right: sugar pine, red fir.
89
Figure 3-30. Simulated mean aboveground live biomass (g m-2) of re-sprouting shrubs, representing
mean and standard deviations across five replicate model simulations, using three climate scenarios
with current fire ignitions and with increased ignitions.
90
Figure 3-31. Simulated mean area burned (ha) across the LTB, representing mean and standard
deviations across five replicate model simulations, using three climate scenarios with current fire
ignition estimates and with increased ignitions.
91
Figure 3-32. Simulated mean total carbon (live C + soil organic C + detrital C, g C m-2) across the LTB,
representing mean and standard deviations across five replicate model simulations, under base and A2
climate, with current fire ignitions and with increased ignitions, and applying continuous fuel treatments
on a 15 year rotation period.
92
Figure 3-33. Simulated mean total carbon (live C + soil organic C + detrital C) across the four fuel
treatment management areas within the LTB, representing mean and standard deviations across five
replicate model simulations. Base climate: No fuel treatments using base climate, Base: FT 15 RP:
Continuous fuel treatments applied on a 15 year rotation period using base climate, A2 climate w/incr.
ign: No fuel treatments using A2 climate with projected increase in fire ignitions, A2: FT 15 RP w/incr.
ign: Continuous fuel treatments applied on a 15 year rotation period using A2 climate with projected
increase in fire ignitions.
93
Base Climate
A2 Climate
A2 Climate
FT – 15RP
200 - 6,800
6,801- 15,500
15,501- 24,000
24,001 - 31,500
31,501 - 37,000
37,001 - 42,500
42,501 - 52,000
Figure 3-34. Total carbon (live C + soil organic C + detrital C, g C m-2) distribution across the LTB. Left to
right: Year 2110 (year 100 of model) for base (contemporary) climate without fuel treatments, A2
climate without fuel treatments, Year 2110 (year 100) with continuous fuel treatments applied on a 15
year rotation interval.
94
Figure 3-35. Simulated mean aboveground live biomass (g m-2) of four representative tree species found
at the LTB, representing mean and standard deviations across five replicate model simulations, for base
and A2climate, with and without increased ignitions and fuel treatments. Top, left to right: white fir,
Jeffrey pine, Bottom, left to right: sugar pine, red fir.
95
Figure 3-36. Simulated mean aboveground live biomass (g m-2) of re-sprouting shrubs, representing
mean and standard deviations across five replicate model simulations, using three climate scenarios
with contemporary fire ignitions and with increased ignitions, and with and without fuel treatments (15
yr. rotation period).
96
Appendix A
Table 2-1. Tree species and functional group attributes used in LANDIS-II modeling of the Lake Tahoe
Basin, CA, NV, USA.
Species or functional group
Longevity
(yrs.)
Age of Sexual
Maturity (yrs.)
Shade
tolerance (1-5)
Fire tolerance
(1-5)
Effective
seeding
distance (m)
Maximum
seeding
distance (m)
Vegetative
Reproduct
ion
probabilit
y
Minimu
m
resprouti
ng age
Maximu
m
resprouti
ng age
Post-fire
regenerati
on
Pinus jeffreyi
500
25
2
5
50
300
0
0
0
none
Pinus lambertiana
550
20
3
5
30
400
0
0
0
none
Calocedrus decurrens
500
30
4
5
30
2000
0
0
0
none
Abies concolor
450
35
4
3
30
500
0
0
0
none
Abies magnifica
500
40
3
4
30
500
0
0
0
none
Pinus contorta
250
7
1
2
30
300
0
0
0
none
Pinus monticola
550
18
3
4
30
800
0
0
0
none
Tsuga mertensiana
800
20
5
1
30
800
0.0005
100
800
none
Pinus albicaulis
900
30
3
2
30
5000
0.0001
100
900
none
Populus tremuloides
Non N-fixing resprouting
shrubs
Non N-fixing obligate
seeding shrubs
175
15
1
2
30
1000
0.9
1
175
resprout
80
5
2
1
30
550
0.85
5
70
resprout
80
5
2
1
30
1000
0
0
0
none
N-fixing resprouting shrubs
N-fixing obligate seeding
shrubs
80
5
1
1
30
500
0.75
5
70
resprout
80
5
1
1
30
800
0
0
0
none
97
Table.2-2. Physiological parameters for 11 tree species and 4 shrub functional groups found in the Lake Tahoe Basin, CA, NV, USA. GDD:
Growing Degree Days
Species or functional group
Functional
type
Nitrogen
tolerance
GDD
min
GDD
max
Min.
Jan.
Temp.
(C)
Max
Drought
Leaf
longevity
(yrs.)
Leaf
lignin
(%)
Fine
root
lignin
(%)
Wood
lignin
(%)
Coarse
root
lignin
(%)
Leaf
CN
ratio
Fine
root
C:N
ratio
Wood
C:N
ratio
Coarse
root
C:N
ratio
Litter
C:N
ratio
Pinus jeffreyi
1
N
555
2149
-5
0.94
6.0
0.28
0.2
0.25
0.25
48
48
250
167
100
Pinus lambertiana
1
N
815
2866
-5
0.90
2.5
0.17
0.2
0.25
0.25
53
53
278
185
100
Calocedrus decurrens
1
N
837
2938
18
0.99
4.0
0.1
0.2
0.25
0.25
48
48
500
333
100
Abies concolor
1
N
540
2670
-10
0.93
8.0
0.17
0.2
0.25
0.25
30
30
333
222
100
Abies magnifica
1
N
483
1144
-18
0.87
8.0
0.17
0.2
0.25
0.25
30
30
250
167
100
Pinus contorta
1
N
276
993
-18
0.87
3.5
0.25
0.2
0.25
0.25
48
48
500
333
100
Pinus monticola
1
N
155
1016
-18
0.82
7.0
0.31
0.2
0.25
0.25
37
37
500
333
100
Tsuga mertensiana
1
N
235
894
-18
0.80
4.5
0.24
0.2
0.25
0.25
80
80
333
222
100
Pinus albicaulis
1
N
230
950
-18
0.90
5.5
0.27
0.2
0.25
0.25
80
80
333
222
100
Populus tremuloides
2
N
600
3000
-10
0.82
1.0
0.18
0.2
0.25
0.25
62
62
333
222
100
Non N-fixing resprouting shrubs
Non N-fixing obligate seeding
shrubs
3
N
400
4000
-10
0.99
1.5
0.25
0.2
0.25
0.25
56
56
333
222
100
3
N
400
4000
-10
0.97
1.5
0.25
0.2
0.25
0.25
59
59
333
222
100
N-fixing resprouting shrubs
3
Y
400
4000
-10
0.97
1.5
0.25
0.2
0.25
0.25
20
28
333
222
50
N-fixing obligate seeding shrubs
3
Y
400
4000
-10
0.99
1.5
0.25
0.2
0.25
0.25
20
30
333
222
50
98
Table 2-3. Functional type parameters for the species list in Table 1.
Name
Functional
type index
PPDF1
mean
PPDF2
max
PPDF1
shape 1
PPDF1
shape 2
NPP
leaf
(%)
BTOLAI
KLAI
MAXLAI
PPRPTS2
PPRPTS3
Wood
decay
rate
Wood
mortality
(%/mo.)
Age
mortality
shape
Leaf
Drop
Month
Conifers
1
23.0
40.0
0.05
6.0
0.2
0.002
5000
10
0.2
0.1
1.0
0.002
10.0
9.0
Hardwoods
2
23.0
35.0
0.05
7.0
0.3
0.002
5000
20
0.2
0.1
1.0
0.002
10.0
9.0
Shrubs
3
23.0
32.0
0.05
10.0
0.3
0.002
500
5
0.2
0.1
1.0
0.002
10.0
9.0
Table 2-4. Ecoregion fixed parameters for the Lake Tahoe Basin, CA, NV, USA.
Ecoregion
Soil
depth
(cm)
% Clay
fraction
% Sand
fraction
Field
capacity
Wilting
point
Storm
flow
fraction
Base
flow
fraction
Drainage
Atmospheric
N (slope)
Atmospheric
N
(intercept)
Latitude
Decay
rate
Surface
soil
Decay
rate
SOM1
Eastside Forest & Woodland
100
0.070
0.714
0.109
0.055
0.4
0.4
0.691
0.08
0.005
39.02
0.4
1
0.02
0.0002
0.5
Upper Montane
100
0.050
0.786
0.090
0.046
0.4
0.4
0.815
0.08
0.005
39.02
0.4
1
0.02
0.0002
0.5
Subalpine
100
0.061
0.744
0.097
0.051
0.4
0.4
0.735
0.08
0.005
39.02
0.4
1
0.02
0.0002
0.5
Montane shrubland
100
0.078
0.718
0.090
0.060
0.4
0.4
0.826
0.08
0.005
39.02
0.4
1
0.02
0.0002
0.5
Riparian areas
100
0.068
0.704
0.123
0.053
0.4
0.4
0.523
0.08
0.005
39.02
0.4
1
0.02
0.0002
0.5
99
Decay
rate
SOM2
Decay
rate
SOM3
Denitrification
Table 2-5. Initial values of Carbon and Nitrogen in various soil organic pools.
Ecoregion
SOM1 C (surface)
SOM1 N (surface)
SOM1 C (soil)
SOM1 N (soil)
SOM2 C
SOM2 N
SOM3 C
SOM3 N
Mineral N
Eastside Forest & Woodland
75.0
3.0
100.0
10.0
3000.0
50.0
300.0
15.0
3.0
Upper Montane
75.0
3.0
100.0
10.0
3000.0
50.0
300.0
15.0
3.0
Subalpine
75.0
3.0
100.0
10.0
3000.0
50.0
300.0
15.0
3.0
Montane shrubland
75.0
3.0
100.0
10.0
3000.0
50.0
300.0
15.0
3.0
Riparian areas
75.0
3.0
100.0
10.0
3000.0
SOM1: Soil Organic Matter fast pool; SOM2: Soil Organic Matter fast pool; SOM3: Soil Organic Matter passive pool
50.0
300.0
15.0
3.0
Table 2-6. Fire regime parameter inputs for the Dynamic Fire extension used for the Lake Tahoe Basin,
CA, NV. FMC: Foliar moisture content.
Fire Region
Region
size
(ha)
sigma
Maximum
Size or
Duration
Spring
FMC
Low
Spring
FMC
High
Spring High
Proportion
Summer
FMC
Low
Summer
FMC
High
Summer
High
Proportion
mu
Lower-elevation
9,603
5.2
0.32
4500
135
175
0.05
85
100
0.89
Mid-elevation
Higher-elevation
28,777
5.3
0.32
4500
135
175
0.05
85
100
0.89
31,194
5.2
0.32
4500
135
175
0.05
85
100
0.89
100
101