ArticlePDF Available

Smoldering fire in long-unburned longleaf pine forests [electronic resource] : linking fuels with fire effects /

Authors:
  • Tall Timbers Research Station

Abstract and Figures

ABSTRACT: Historically, frequent fires maintained low fuel loads in many temperate coniferous forests. Widespread 20th century fire exclusion altered community structure and species composition while facilitating the accumulation of fuel. Organic matter accumulations on the forest floor have been poorly studied and have major implications for the restoration of historically fire-maintained ecosystems. A major restoration strategy for historically fire-maintained ecosystems in general, and Pinus palustris-dominated ecosystems in the southeastern USA in particular, is to reintroduce fire. The results of reintroducing fire into longleaf pine ecosystems have been appalling; post-ignition smoldering fires generate noxious smoke and kill remnant large pines. The potential mechanisms of fire-induced pine mortality are long-duration heating to vascular tissues in the stems and roots, or a combination of factors increasing tree stress. In landscape-scale experimental fires in long-unburned pine forests, four replicated treatments based on lower forest floor (duff) moisture content were applied: dry burns were ignited at 55% duff moisture content (DMC); moist burns at 85% DMC; wet burns at 115% DMC; and there was a control treatment that was not burned. Pine mortality was delayed up to 18 months following burning, and was highest (P < 0.05) in the dry burns. The likelihood of mortality was predicted by duff consumption and canopy scorch; only duff consumption consistently predicted mortality across all burning treatments. Isolation of the effects of stem and root heating were tested in an experiment using small-scale fires designed to heat stem, root, stem and root, or neither tissues. Depth of lethal temperatures in the mineral soil across all treatments was highly related to post-fire radial growth and root nonstructural carbohydrates. Mineral soil heating, not duff or basal bark temperatures, decreased latewood growth in the year following fire (P= 0.07). Coarse root (2-5 mm diameter) carbohydrates were also linked to duration of lethal heating of the mineral soil (P < 0.01; R² = 0.59). Given the conclusions from both small- and large-scale experiments and a simulation model-based comparison, current knowledge and research needs are reviewed. Future research should focus on determining spatial patterns of forest floor moisture and combustion dynamics. Research linking forest floor fuels to fire behavior and subsequent fire effects will increase ability to manage and restore long-unburned forests. Text (Electronic thesis) in PDF format. System requirements: World Wide Web browser and PDF reader. Mode of access: World Wide Web. Title from title page of source document. Document formatted into pages; contains 123 pages. Thesis (Ph.D.)--University of Florida, 2005. Includes vita. Includes bibliographical references.
Content may be subject to copyright.
SMOLDERING FIRE IN LONG-UNBURNED LONGLEAF PINE FORESTS:
LINKING FUELS WITH FIRE EFFECTS
By
J. MORGAN VARNER, III
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2005
To Julian M. Varner, Jr. (1941-2004).
iii
ACKNOWLEDGMENTS
Many people and organizations assisted me with this work, primary among them
the Chair of my committee, Jack Putz. His inquisitiveness, energy, enthusiasm, and belief
in me and my work inspired me through trying life events. Doria Gordon acquired
funding and had faith in my abilities to balance logistical issues that tend to accompany
fires in remote locations. Alan Long, Tim Martin, and Bob Mitchell helped me to better
understand fire and the trees they affect. My friends and colleagues Kevin Hiers, Neil
Pederson, and Ben Poulter stimulated and challenged my thinking and provided
encouragement.
Research involving the use of fire requires incredible logistical support. Pete
Jerkins, Kevin Hiers, James Furman, Kevin Mock, and many others provided assistance
with fires at Eglin Air Force Base. Fires at the Carl Swisher-Katharine Ordway Preserve
were accomplished with the help and energy of Steve Coates. Field support was provided
by many at Eglin Air Force Base Jackson Guard. Assistance with data collection and
study design was provided by Diana Olson, David Wright, Roger Ottmar, Bob Vihnanek,
and many others at the USDA Pacific Northwest Research Station’s Pacific Wildland
Fire Sciences Laboratory. Joe O’Brien helped with small-scale experimental fire
temperature measurements and advice on study design and analysis. Data collection
throughout this project was assisted by Elise Owens and her quick wit. Over the course of
iv
this project, Rich Fonda, Dale Wade, and John Kush and many others provided advice
and helpful criticisms.
This research was supported by funds provided to the USDA Forest Service,
Southern Research Station, Asheville, North Carolina, from the Joint Fire Science
Program of the Departments of Agriculture (Forest Service) and Interior (Bureau of Land
Management) with supplementary support from the University of Florida School of
Natural Resources and Environment, The Nature Conservancy, and Eglin Air Force Base
Jackson Guard. I appreciate the time, patience, and support over the last year from the
Department of Forestry and Watershed Management at Humboldt State University.
Over the course of this project, I experienced personal tragedies amidst newfound
bliss. Without the support of my family and friends, this project would have never been
completed. In addition to surveying burned forests, excavating roots, and reviewing
manuscript drafts, Rachel Seman-Varner helped keep me focused and joyful. Slaton
Wheeler assisted with field research through lightning storms and humidity. Special
thanks are also due to Geoff Parks, Adam Watts, Tova Spector, Eddie Watkins, Claudia
Romero, Alex Varner, Bob Dylan, Maynard Hiss, and many others in McCarty, Carr, and
Bartram Halls.
v
TABLE OF CONTENTS
page
ACKNOWLEDGMENTS ................................................................................................. iii
LIST OF TABLES............................................................................................................ vii
LIST OF FIGURES ........................................................................................................... ix
ABSTRACT....................................................................................................................... xi
CHAPTER
1 INTRODUCTION ........................................................................................................1
2 RESTORING FIRE TO LONG-UNBURNED Pinus palustris ECOSYSTEMS:
NOVEL EFFECTS AND CONSEQUENCES.............................................................4
Introduction...................................................................................................................4
Effects of Fire Exclusion on Longleaf Pine Ecosystems.......................................6
Overstory responses to fire suppression.........................................................6
Understory responses to fire suppression.......................................................7
Midstory responses to fire suppression ..........................................................7
Forest floor characteristics after fire suppression...........................................8
Responses to Fire Reintroduction: Restoration Case Studies.......................................8
Flomaton Natural Area ..........................................................................................8
Eglin Air Force Base ...........................................................................................10
Other Examples ...................................................................................................10
Causes of Pine Mortality after Fire Re-Introduction ..................................................11
Smoldering Duff Fires and Southeastern USA Restoration .......................................15
3 TREE MORTALITY RESULTING FROM REINTRODUCING FIRE TO
LONG-UNBURNED LONGLEAF PINE FORESTS: THE ROLE OF DUFF
MOISTURE................................................................................................................22
Introduction.................................................................................................................22
Methods ......................................................................................................................24
Study Area ...........................................................................................................24
Field Data Collection...........................................................................................25
Data Analysis.......................................................................................................26
Results.........................................................................................................................27
vi
Discussion...................................................................................................................29
Mortality and Fire Damage to Stem, Canopy, and Roots....................................31
Challenges to Restoration Using Fire..................................................................33
4 EFFECTS OF BASAL BARK, DUFF, AND SOIL TEMPERATURES ON
POST-FIRE TREE STRESS AND GROWTH ..........................................................44
Introduction.................................................................................................................44
Methods ......................................................................................................................46
Study Site.............................................................................................................46
Tree Measurements .............................................................................................48
Fuel Sampling......................................................................................................49
Fire Measurements ..............................................................................................50
Data Analysis.......................................................................................................51
Results.........................................................................................................................52
Fire Behavior .......................................................................................................52
Radial Growth and Mortality...............................................................................53
Root Nonstructural Carbohydrates ......................................................................53
Discussion...................................................................................................................54
5 MODELED VERSUS OBSERVED FIRE EFFECTS IN LONG-UNBURNED
Pinus palustris ECOSYSTEMS .................................................................................68
Introduction.................................................................................................................68
Methods ......................................................................................................................70
Study Sites ...........................................................................................................70
Simulation Modeling ...........................................................................................73
Results.........................................................................................................................73
Discussion...................................................................................................................74
6 LINKING FOREST FLOOR FUELS WITH FIRE BEHAVIOR AND EFFECTS:
A REVIEW.................................................................................................................82
Introduction.................................................................................................................82
Forest Floor Accumulation Patterns ...........................................................................84
Forest Floor Composition ...........................................................................................86
Controls of Duff Moisture Content.............................................................................90
Conclusions.................................................................................................................91
LIST OF REFERENCES...................................................................................................95
BIOGRAPHICAL SKETCH ...........................................................................................111
vii
LIST OF TABLES
Table page
2-1 Reports of excessive overstory longleaf pine mortality following re-introduction
of fire into pine ecosystems after decades of fire suppression. ................................17
3-1 Fire weather and fuel moisture values for each 10 ha experimental prescribed
burn in long-unburned longleaf pine forests. ...........................................................36
3-2 Effects of 10 ha prescribed burns in long-unburned longleaf pine forests...............37
3-3 Univariate linear regression results for correlates of individual pine mortality
resulting from prescribed restoration burns in long-unburned forests. ....................38
3-4 Step-wise multiple regression results for predictors of block-level longleaf pine
mortality caused by prescribed restoration burns in long-unburned pinelands........39
3-5 Modeled nested logistic regression of post-fire individual tree longleaf pine
mortality. ..................................................................................................................40
4-1 Fire weather observations and time-of-ignition fuel moistures from 80
experimental single-tree burns... ..............................................................................59
4-2 Durations of lethal heating (i.e., temperatures >60ºC) to basal bark, duff, and
mineral soil during individual tree burns. ................................................................60
4-3 Effects of individual tree experimental burning treatments on smoldering time,
mean floor consumption, longleaf pine stem radial growth, and root
nonstructural carbohydrates. ....................................................................................61
4-4 Regression results for the analysis of fuel and soil moisture effects on basal
bark, duff, and mineral soil temperatures in smoldering fires in a long-unburned
longleaf pine stand. ..................................................................................................62
4-5 Effects of basal bark, duff, and mineral soil temperatures on longleaf pine stem
radial growth and root nonstructural carbohydrates following individual tree
fires...........................................................................................................................63
5-1 Model inputs for simulations of fire effects (FOFEM 5.2.1) across three
different moisture regimes in fire excluded longleaf pine stands. ...........................79
viii
5-2 Comparison of results from simulated (FOFEM 5.2.1) and actual fires for fire-
excluded longleaf pine forests..................................................................................80
6-1 Carbon fractions from basal fuel accumulations in a long-unburned longleaf pine
forest floor. ...............................................................................................................92
ix
LIST OF FIGURES
Figure page
1-1 Many traditional models in fire ecology research have neglected fuel and fire
characteristics and instead employed simplistic and often misleading pre-fire
environment + fire = post-fire environment models. .................................................3
2-1 A frequently-burned longleaf pine ecosystem reference condition at the
Katharine Ordway Preserve-Swisher Memorial Sanctuary in northern Florida.. ....18
2-2 Typical long-unburned (37 years since fire) longleaf pine forest at the Katharine
Ordway Preserve-Swisher Memorial Sanctuary, Florida.........................................19
2-3 Forest floor development in a long-unburned (ca. 40 years since fire) longleaf
pine forest at Eglin Air Force Base, Florida.............................................................20
2-4 Restoration fires in long-unburned longleaf pine forests damage canopy, stem,
and root tissues often leading to excessive tree mortality........................................21
3-1 Study site locations of experimental prescribed fires to examine correlates of
mortality resulting from reintroduction of fire into fire-suppressed longleaf pine
(Pinus palustris) forests. ..........................................................................................41
3-2 Effects of burning treatment on cumulative overstory longleaf pine mortality .......42
3-3 Relationship between duff consumption and cumulative post-fire mortality of
overstory longleaf pines at Eglin Air Force Base, Florida.......................................43
4-1 The Katharine Ordway Preserve-Swisher Memorial Sanctuary in Putnam
County, Florida.........................................................................................................64
4-2 Experimental fire treatments to individual pines in a long-unburned Pinus
palustris forest in northern Florida, USA.................................................................65
4-3 Locations of thermocouples during experimental fires surrounding the base of an
individual long-unburned Pinus palustris tree. ........................................................66
4-4 Relationship between the change (2003-2004) in coarse root non-structural
carbohydrates in mature Pinus palustris and duration of heating >60°C at 5 cm
below the surface of the mineral soil in experimental individual tree fires. ............67
x
5-1 Locations of study sites in northern Florida, USA...................................................81
6-1 Basal forest floor depths and composition in a long-unburned longleaf pine
forest.........................................................................................................................93
6-2 Basal smoldering near a mature longleaf pine at the Katharine Ordway Preserve-
Swisher Memorial Sanctuary. ..................................................................................94
xi
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
SMOLDERING FIRE IN LONG-UNBURNED LONGLEAF PINE FORESTS:
LINKING FUELS WITH FIRE EFFECTS
By
J. Morgan Varner, III
December 2005
Chair: Francis E. Putz
Major Department: School of Natural Resources and Environment
Historically, frequent fires maintained low fuel loads in many temperate coniferous
forests. Widespread 20
th
century fire exclusion altered community structure and species
composition while facilitating the accumulation of fuel. Organic matter accumulations on
the forest floor have been poorly studied and have major implications for the restoration
of historically fire-maintained ecosystems.
A major restoration strategy for historically fire-maintained ecosystems in general,
and Pinus palustris-dominated ecosystems in the southeastern USA in particular, is to
reintroduce fire. The results of reintroducing fire into longleaf pine ecosystems have been
appalling; post-ignition smoldering fires generate noxious smoke and kill remnant large
pines. The potential mechanisms of fire-induced pine mortality are long-duration heating
to vascular tissues in the stems and roots, or a combination of factors increasing tree
stress.
xii
In landscape-scale experimental fires in long-unburned pine forests, four replicated
treatments based on lower forest floor (duff) moisture content were applied: dry burns
were ignited at 55% duff moisture content (DMC); moist burns at 85% DMC; wet burns
at 115% DMC; and there was a control treatment that was not burned. Pine mortality
was delayed up to 18 months following burning, and was highest (P < 0.05) in the dry
burns. The likelihood of mortality was predicted by duff consumption and canopy scorch;
only duff consumption consistently predicted mortality across all burning treatments.
Isolation of the effects of stem and root heating were tested in an experiment using
small-scale fires designed to heat stem, root, stem and root, or neither tissues. Depth of
lethal temperatures in the mineral soil across all treatments was highly related to post-fire
radial growth and root nonstructural carbohydrates. Mineral soil heating, not duff or basal
bark temperatures, decreased latewood growth in the year following fire (P= 0.07).
Coarse root (2-5 mm diameter) carbohydrates were also linked to duration of lethal
heating of the mineral soil (P< 0.01; R
2
= 0.59).
Given the conclusions from both small- and large-scale experiments and a
simulation model-based comparison, current knowledge and research needs are reviewed.
Future research should focus on determining spatial patterns of forest floor moisture and
combustion dynamics. Research linking forest floor fuels to fire behavior and subsequent
fire effects will increase ability to manage and restore long-unburned forests.
1
CHAPTER 1
INTRODUCTION
In fire ecology research, limited understanding of mechanistic links among pre-fire
conditions, fire characteristics, and subsequent ecosystem effects results in a poor
understanding of fire effects (Johnson and Miyanishi 1995, 2001; Dickinson and Johnson
2001, 2004). Knowing the linkages among pre-fire fuel characteristics (types and
amounts), fire characteristics (types, temperatures, and durations), and post-fire
ecological outcomes (e.g., tree damage or death) will provide an increased understanding
of the ecology of fire and its effects on species, communities, and ecosystems (Ryan and
Frandsen 1991, Swezy and Agee 1991, Ryan 2000, Johnson and Miyanishi 2001,
Dickinson and Johnson 2001).
Development of a mechanistic fire ecology will require several lines of inquiry for
southern pine ecosystems. First, a review is needed of fuel characteristics, fire behavior
and damage, and resulting post-fire tree mortality in southern pine-dominated
ecosystems. To date, most work in coniferous ecosystems has focused on pre-fire and
subsequent post-fire communities, with little emphasis on the fuel and fire characteristics
that influence the structure and composition of the latter (Figure 1-1). Specifically, lack
of research on the role of forest floor fuels (organic soil horizons) is troubling; especially
considering their role in smoldering fire, smoke generation, and tree mortality in
coniferous ecosystems worldwide (Flinn and Wein 1977, Ryan and Frandsen 1991,
Swezy and Agee 1991, Hungerford et al. 1995, Miyanishi 2001).
2
Following this logic, descriptive work is needed on forest floor fuels in
fire-excluded coniferous ecosystems. Forest floor fuel depth, distribution, structure and
composition are topics that deserve study, with particular reference to their roles in
smoldering fire and fire-induced tree mortality. The forest floor fuel complex has been
little studied (but see Brown 1966, Brown 1970; Miyanishi 2001; Stephens et al. 2004;
Hille and Stephens 2005). Given the diversity of its physical and chemical characteristics,
the forest floor fuel complex warrants experimental study on fire behavior and
determinants of combustion. With a better empirical understanding of the forest floor
fuel complex and its combustion characteristics, we can then incorporate pre-fire forest
floor fuel loading into theoretical models of fire-caused tree mortality (Dickinson and
Johnson 2001, Miyanishi 2001).
This study evaluated the effects of varying burning treatments on overstory tree
mortality in relation to forest floor combustion. In a large field experiment,
long-unburned longleaf pine stands were burned at different duff moisture contents. After
these fires, post-fire damage and tree mortality were monitored for 3 years to capture
delayed mortality. To explore the mechanisms for the observed patterns of tree mortality
in stand-level fires, individual pines were burned so as to damage stem, roots, both stem
and roots, or neither tissues. Temperatures on the basal bark, within the duff, and at 5, 10,
and 20 cm depths in the mineral soil were monitored to determine heating effects on tree
stress, growth, and mortality. These experiments, in concert with field and laboratory
smoldering experimentation, strengthen our understanding of the mechanistic linkages
among forest fuels and fire behavior with tree damage and mortality.
3
Figure 1-1. Many traditional models in fire ecology research have neglected fuel and fire
characteristics and instead employed simplistic and often misleading pre-fire
environment + fire = post-fire environment models. The objective of this
study is to build a more rigorous approach to understanding fire effects and
the mechanisms of fire ecology, particularly in relation to longleaf pine
dominated ecosystems of the southeastern USA.
4
CHAPTER 2
RESTORING FIRE TO LONG-UNBURNED PINUS PALUSTRIS ECOSYSTEMS:
NOVEL EFFECTS AND CONSEQUENCES
1
Introduction
Southeastern USA pine forests and savannas dominated by Pinus palustris
(longleaf pine) and a biologically diverse understory covered an estimated 37 million ha
before European settlement (Frost 1993). During ensuing centuries, southeastern
forestlands have been logged, farmed, subdivided, and planted with faster-growing
southern pines (Croker 1987). Remnant areas not converted have been degraded by
several decades of fire suppression (Croker 1987, Frost 1993). In these areas, such
landscape changes caused a 97% decline in the area of longleaf pine ecosystems, making
them among the most imperiled ecosystems in the United States (Noss et al. 1995).
Of the remnant area of longleaf pine ecosystems, only about half is frequently
burned (Outcalt 2000), leading to substantial alterations in ecosystem structure and
composition. Presettlement fire regimes were typified by short fire-return intervals
(ranging from 1 to 5 years), and by low-intensity surface fires ignited by lightning and
late Holocene Native Americans (Christensen 1981). Fire suppression transforms these
once open savanna-woodland ecosystems into closed canopy forests, with reduced
understory plant species richness, and heavy accumulations of surface fuels (Heyward
1939, Engstrom et al. 1984, Mushinsky 1985, Ware et al. 1993, Gilliam and Platt 1999,
1
A version of this chapter was published as Varner, J.M., D.R. Gordon, F.E. Putz, and J. K. Hiers. 2005.
Restoring fire to long-unburned Pinus palustris ecosystems: Novel fire effects and consequences.
Restoration Ecology 13:536-544.
5
Kush and Meldahl 2000, Kush et al. 2000, Varner et al. 2000, Provencher et al. 2001b).
Overstory density, tree species richness, and basal area increase in response to fire
suppression (Ware et al. 1993, Gilliam and Platt 1999, Varner et al. 2000), while
understory species richness and cover decrease (Gilliam and Platt 1999, Kush et al. 2000,
Varner et al. 2000). Whereas organic matter on the forest floor was scarce in
presettlement ecosystems, in the absence of frequent fires there are substantial
accumulations of surficial organic horizons, particularly around the bases of large pines
(Heyward and Barnette 1936, Brockway and Lewis 1997, Varner et al. 2000, Kush et al.
2004).
To reverse or reduce the further decline of southeastern longleaf pine ecosystems,
many fire-excluded stands with remnant mature pine overstory have been targets for
ecological restoration (Hermann 1993, Landers et al. 1995, Wade et al. 1998, Provencher
et al. 2001b). In long-unburned pinelands, the objectives of restoration (Wade et al. 1998,
Varner et al. 2000, Provencher et al. 2001b) are typically:
Maintain the remnant pine overstory
Reduce the hardwood midstory
Enhance or re-establish native plants and animals
Reduce accumulated fuels
Reduce native and non-native invasive species populations.
Efforts at restoring community structure and composition have generally included
the complementary actions of altering species composition by removing invasive species,
reducing stand density, and reducing fuel loads. In highly altered systems, reintroduction
of understory species is increasingly common (Bissett 1996, Cox et al. 2004, Jenkins et
al. 2004). The most common approach to restoration of long-unburned southern pine
communities has been the reinitiation of historical fire regimes with prescribed fires.
6
The objectives for this review were:
Describe the effects of fire exclusion on southern pine ecosystems
Review the outcomes of fire reintroduction and restoration
Review the hypothesized causes of restoration fire mortality of overstory
pines in the Southeast and in analogous ecosystems worldwide
Present a fuels-based perspective for setting restoration priorities that
minimizes catastrophic overstory mortality.
Effects of Fire Exclusion on Longleaf Pine Ecosystems
Overstory responses to fire suppression
With fire exclusion, Southeastern pinelands have experienced structural and
compositional shifts from open savanna-woodlands to closed canopy forests.
Frequently-burned savanna structure is typified by a spatially variable but mostly open
canopy, with stand densities of 130 to 250 trees/ha
> 10 cm dbh, and basal areas of 12 -20
m
2
ha
-1
(Wahlenberg 1946, Platt et al. 1988, Boyer 1990, Palik and Pederson 1996,
Varner et al. 2003a; Figure 2-1). Throughout its range, longleaf pine is mono-dominant or
occurs with scattered fire-resistant oaks (primarily Quercus geminata, Q. incana, Q.
laevis, Q. margaretta, and Q. marilandica) and hickories (Carya tomentosa and C.
pallida; Peet and Allard 1993, Varner et al. 2003b). With the cessation of fire-induced
mortality, the cover and density of shrubs and trees increase in the midstory and canopy
(Gilliam et al. 1993, Brockway and Lewis 1997, Gilliam and Platt 1999, Kush et al. 2000,
Varner et al. 2000, Provencher et al. 2001a, 2001b; Figure 2-2). The species that benefit
from fire suppression include many fire-susceptible species (e.g., Q. hemisphaerica, Q.
nigra, Acer rubrum, Liquidambar styraciflua, Magnolia grandiflora, and Nyssa
sylvatica) that alter stand structure by increasing tree densities, leaf areas, and basal area.
Stand composition is degraded as canopy species richness increases.
7
Understory responses to fire suppression
Without fire in longleaf pine ecosystems, understory communities undergo radical
shifts in cover and richness. Frequently-burned pineland understory communities are
among the most species-rich outside of the tropics (Peet and Allard 1993, Kirkman et al.
2001, Provencher et al. 2003). Typical burned understories contain 20 to 30 species m
-2
,
with dominance by bunch grasses (Aristida stricta, Schizachyrium scoparium, and
Andropogon spp.), asters, legumes, and other forbs including several rare and endemic
plant species (Hardin and White 1989, Peet and Allard 1993). Without fire, increased
overstory and midstory canopy cover (and leaf litter deposition) reduce sunlight reaching
the forest floor; leading to the loss of light-demanding understory grasses, forbs, and pine
seedlings (Provencher et al. 2001a, 2001b, Waters et al. 2004). After several decades of
fire suppression, herbaceous species richness is often <2 species m
-2
; pine seedlings are
lacking, and the understory becomes dominated by woody species (Varner et al. 2000,
Kush et al. 2004).
Midstory responses to fire suppression
A marked change in fire-excluded pinelands is the advent of a woody midstory.
Most frequently-burned pinelands (particularly on sites with high net primary
productivity) lack a well-developed midstory stratum (Peet and Allard 1993, Landers et
al. 1995). The few native shrub and tree species present in frequently-burned pinelands
include oak and hickory sprouts, Ilex glabra (gallberry), Vaccinium spp., Serenoa repens
(saw palmetto), and isolated patches or “domes” of Quercus geminata (Guerin 1988, Peet
and Allard 1993). Without fire, hardwoods and shrubs ascend into the midstory, where
they increase cover and stem density dramatically (Provencher et al. 2001b).
8
Forest floor characteristics after fire suppression
Frequently-burned pinelands have very little organic matter on the forest floor,
except some litter (Oi horizon); but this condition is altered radically by fire exclusion.
Without frequent surface fires, leaf litter, sloughed bark, fallen branches, and other
organic necromass accumulate and decompose into fermentation (Oe) and humus (Oa)
horizons absent in frequently-burned communities (Figure 2-3 Heyward 1939, Switzer et
al. 1979). Roots and mycorrhizal hyphae exploit these “duff” horizons, especially near
the bases of large pines, where duff can accumulate to depths of 25 cm or more (Varner
et al. 2000, Gordon and Varner 2002, Kush et al. 2004). Litter accumulation and duff
formation further block light from reaching the forest floor (Waters et al. 2004) and may
play a significant role in driving changes in nutrient cycling (Wilson et al. 2002).
Responses to Fire Reintroduction: Restoration Case Studies
Flomaton Natural Area
The Flomaton Natural Area is a 27 ha remnant old-growth longleaf pine stand in
Escambia County, Alabama (31°01' N, 87°15' W). Fire had been suppressed in the stand
for 45 years until 1993, when a small trash fire ignited a 3-ha stand isolated by a dirt
road. The wildfire was allowed to burn out on its own with no observed canopy scorch
and limited stem char (all trees < 1 m char height). For several days after the fire,
smoldering continued in the deep duff that had accumulated around the large remnant
pines. Smoke from these fires was problematic for local residents particularly because
emissions from smoldering fires are much more hazardous to human health than
relatively benign flaming-phase fire emissions (McMahon et al. 1980, McMahon 1983).
Additionally, the danger of re-ignition remained high as long as smoldering continued.
During the first 2 years after the fire, heavy mortality was observed in the overstory
9
longleaf pines (Kush et al. 2004). Mortality was highest among large pines: 91% of the
trees >35 cm dbh died. Survival was higher among small (10 to 20 cm dbh), longleaf,
slash (P. elliottii var. elliottii), and loblolly pines (P. taeda). Most of the small trees of
fire-susceptible hardwood species (primarily Liquidambar styraciflua, Prunus serotina,
and Acer rubrum) that invaded during the fire-free period also survived the fire (Kush et
al. 2004).
In response to the loss of a high proportion of the old growth pines during the 1993
fire, an aggressive ecological restoration program was initiated on the adjacent 24 ha
unburned site. The restoration process began with the mechanized harvesting of all
hardwood stems (primarily Quercus spp.) > 10 cm dbh with a Morbark three-wheeled
feller-buncher (Morbark Inc., Winn, Michigan). Beginning in 1994, prescribed fires were
re-introduced at a FRI of 1 to 3 years (Varner et al. 2000, Kush et al. 2004). All fires
were ignited when duff moisture content was high (typically within 2 to 4 days after large
rain events) and were lit to minimize fire residence time and fireline intensity. Canopy
scorch was low (<20% of trees) in all fires. Even though the Oe and Oa horizons in the
duff were moist when the fires were ignited, smoldering was initiated in deep duff
accumulations near tree stems. Occurrences of smoldering continued to be detected for
several days post-ignition, requiring repeated extinguishing with backpack, ATV, and
tractor-mounted water sprayers. As a result of these efforts to control fire intensity and to
extinguish duff smoldering when detected, mortality of pines in the 4 years after the fire
was reduced to an annual average of 4.2% (Varner et al. 2000), still much higher than
typical longleaf pine mortality (Boyer 1979, Palik and Pederson 1996), but most death
occurred in trees < 20 cm dbh. The fires killed several pines 50 to 80 cm dbh, but losses
10
of these old pines did not exceed 2 trees ha
-1
year
-1
(Varner et al. 2000, J.S. Kush, Auburn
University, unpublished data).
Eglin Air Force Base
Eglin Air Force Base is a 188,000 ha military reservation in Okaloosa, Walton, and
Santa Rosa Counties in the Panhandle of Florida (30°38' N, 86°24' W). Among the many
natural plant communities at Eglin, longleaf pine communities cover approximately
130,000 ha. Many of Eglin’s pinelands have experienced prolonged fire-free periods
(McWhite et al. 1999, Hiers et al. 2003), leading to ecosystem conditions similar to those
observed at Flomaton.
Reintroduction of fire has been the major method for restoration of longleaf pine
ecosystems in Eglin (McWhite et al. 1999), but results have been mixed. As a result of
fire re-introduction at Eglin, some stands suffered 75 to 100% overstory pine mortality;
whereas in other stands pine mortality was 10% or less (McWhite et al. 1999, Gordon
and Varner 2002). Aside from the need to understand the mechanisms of variation in this
phenomenon, these novel fire effects in such fire-dependent forests are alarming. The
huge scale of the restoration efforts at Eglin Hiers et al. 2003) preclude the individual tree
treatments used at Flomaton (Kush et al. 2004). This situation is relevant to many
fire-excluded areas in the Southeast, as natural resource managers must operationally
manage landscapes, rather than individual trees.
Other examples
Throughout the pinelands of the Southeast, managers have experienced problems
with excessive tree mortality resulting from reintroduction of fire (Table 2-1; Gordon and
Varner 2002). As observed at Flomaton and Eglin, pine mortality after reintroduction of
fire is usually concentrated in the largest diameter classes with greatest pre-fire duff
11
accumulations (Varner et al. 2000). Resulting pine mortality combined with vigorous
resprouting of competing hardwoods, prolonged fire dangers, and smoke emissions
plague restoration of stands throughout the southeastern US. These unintended outcomes
are major deterrents to additional restoration burning region-wide.
Causes of Pine Mortality after Fire Re-Introduction
Hypothesized mechanisms for mortality of large trees after reintroduction of fire
involve the direct effects of fire, such as root damage (Ryan and Frandsen 1991, Swezy
and Agee 1991); vascular tissue damage (Martin 1963, Ryan 2000); leaf scorch (Ryan
2000, Menges and Deyrup 2001); or canopy damage (Menges and Deyrup 2001; Figure
2-4). Increased insect and pathogen attack of fire-stressed trees has also been suggested
as an indirect cause of post-fire mortality in these communities (Ostrosina et al. 1997,
Ostrosina et al. 1999, Menges and Deyrup 2001).
Where fires have been reintroduced, tree death is reportedly correlated with damage
to canopy foliage and branch meristems (Herman 1954, van Wagner 1973, Wade and
Johansen 1986, Menges and Deyrup 2001, McHugh et al. 2003). Foliage scorch is
considered less stressful than foliage consumption, which is generally associated with
damaged branch cambia (Wade and Johansen 1986). Foliage consumption has been
correlated with fire-caused mortality of slash pine in the Southeast (Johansen and Wade
1987, Menges and Deyrup 2001). Nevertheless, pine mortality after reintroduction of fire
has been observed without canopy damage (Varner et al. 2000, Kush et al. 2004).
Regardless, canopy damage is one of many stressors to a tree, exacerbating stem or root
damage, and ultimately contributing to excessive pine mortality rates after reintroduction
fires.
12
Post-fire tree decline and mortality can also result from fire-caused root damage
(Wade and Johansen 1986, Swezy and Agee 1991, Busse et al. 2000). Lateral roots of
longleaf pines are concentrated in the top 30 cm of mineral soil (Heyward 1933,
Wahlenberg 1946) and in long-unburned longleaf pine forests, numerous branch roots
grow up into duff horizons (Gordon and Varner 2002). In frequently-burned pinelands,
soil heating and resulting root mortality are negligible (e.g., Heyward 1938). With fire
suppression and duff accumulation, in contrast, pine roots in duff and in the surface
mineral soil can be heated, damaged, or consumed in long-duration smoldering fires
where temperatures can exceed lethal values for hours (Flinn and Wein 1977, Wade and
Johansen 1986). Smoldering fires spread three orders of magnitude slower than surface
fires and are typically concentrated in the lower duff (Oa horizon) beneath a thermal
blanket of overlying Oe material (Hungerford et al. 1995). Although localized and small,
the smoldering front transmits lethal heat loads (hours > 60°C) to 10 to 20 cm deep in the
mineral soil (J.M. Varner, unpublished data). A similar mechanism of duff root heating
was proposed as a cause of tree death and decline in ponderosa pine stands (P.
ponderosa; Swezy and Agee 1991, Busse et al. 2000). Given the potential physiological
impairment posed by large-scale root heating and consumption, mechanisms involving
root damage deserve further study.
Basal cambial damage is another proposed mechanism of tree mortality after fire
reintroduction. Basal damage in tree stems can occur during surface fires and during
residual smoldering of duff. During surface fires, combustion of litter causes large
amounts of heat to be released close to tree stems, leading to stem char (Wade and
Johansen 1986, Dickinson and Johnson 2001). Bark, especially the thick accumulations
13
on long-unburned trees, usually insulates the cambium sufficiently against heat damage
(Spalt and Reifsnyder 1962, Fahnestock and Hare 1964, Hare 1965, Reifsnyder et al.
1967, Vines 1968, Dickinson and Johnson 2001). In contrast, long-duration heating
during smoldering of duff around tree bases can raise temperatures to lethal levels and
cause cambial death and tree mortality (Dixon et al. 1984, Ryan et al. 1988, Ryan and
Rheinhardt 1988, Dunn and Lorio 1992, Ryan 2000, Dickinson and Johnson 2001). Duff
smoldering often continues for hours or days after ignition (Covington and Sackett 1984,
Hungerford et al. 1995), long enough to kill the cambium under even thick layers of bark.
Cambial damage, even when it does not entirely encircle the stem, is correlated with
fire-caused tree mortality in other conifers (e.g., Ryan et al. 1988, Ryan and Rheinhardt
1988, Ryan 2000). Given the long-duration heating observed in reintroduction fires and
the potential damage to whole-tree physiology, basal cambial damage appears to be an
important mechanism of overstory pine mortality when fires are reintroduced.
Indirect effects of fire re-introduction are reflected in tree physiological stress
that, in turn, renders pines susceptible to pests or pathogens. Overall tree stress may be
indicated by changes in carbon balance, as indicated by stem or root tissue carbohydrate
levels, by reduced resin exudation pressure, or by reduced radial growth (Kozlowski et al.
1991). Past work on southeastern (Davidson and Hayes 1999) and western USA conifers
(Covington et al. 1997, Ryan 2000, McHugh et al. 2003, Wallin et al. 2003, Wallin et al.
2004) shows that increased physiological stress renders trees more susceptible to pest and
pathogen attack. Re-introducing fire to long-unburned slash pine stands in south Florida
led to sharp increases in both Ips and Platypus spp. beetles and subsequent overstory
mortality (Menges and Deyrup 2001). It follows that if restoration burning in longleaf
14
pinelands increases tree stress, then growth and defenses would decline and pest and
pathogen attacks would increase. In many restoration treatments (burning and thinning,
thinning alone, and burning alone), however, resulting physiological condition varies, as
does susceptibility to decline and disease. Resin exudation pressure, a correlate of a tree’s
ability to defend itself from bark beetle attack (Raffa and Berryman 1983, Dunn and
Lorio 1992), increases after fire re-introduction in ponderosa pine ecosystems. Tree
physiological condition and growth also increase after thinning, raking, and burning in
long-unburned ponderosa pine forests (Feeney et al. 1998, Stone et al. 1999, Wallin et al.
2004). However, reduced radial growth has been correlated with restoration burning in
other ponderosa pine forests (Busse et al. 2000). To what degree restoration treatments in
southern pine stands are effective in maintaining, improving, or reducing tree
physiological conditions deserves further study, but arguably only within a mechanistic
framework that links physiological response to specific tree damages and characteristics
of the fuels and fire that caused the damage (i.e., heat damage from smoldering duff fire
to stem vascular tissues that causes physiological impairment and reduced defense
capability).
Evidence supports a mechanistic link between tree mortality and stem and root
damage after fire reintroductions. Thus an understanding of smoldering combustion is
needed to understand the mechanism behind tree mortality in long-unburned southern
pine forests. Smoldering differs from flaming combustion by being controlled mostly by
oxygen availability (as opposed to fuel availability), by lower temperatures (< 500
°
C
versus higher temperatures in flaming combustion), and by longer residence times
(Hungerford et al. 1995, Miyanishi 2001). Smoldering elevates temperatures in duff, in
15
the underlying mineral soil horizons, in roots located within these horizons, and in nearby
tree stems (Wade and Johansen 1986, Ryan and Frandsen 1991, Swezy and Agee 1991,
Hungerford et al. 1995, Schimmel and Granstrom 1996, Haase and Sackett 1998,
Dickinson and Johnson 2001, Miyanishi 2001).
Smoldering Duff Fires and Southeastern USA Restoration
Determining the correlates and mechanisms of tree mortality following fire
reintroduction should be a high priority for southeastern restoration efforts. Given that
50% of all remnant longleaf pinelands are unburned (Outcalt 2000), successful
restoration burning could double the area of functioning longleaf pinelands. Landscape
scale fire suppression has similarly affected other southern pinelands (dominated by P.
taeda, P. elliottii var. elliottii, P. elliottii var. densa, and P. echinata; Noss et al. 1995). A
better understanding of restoration burning has the potential to restore the ecological
integrity of these important communities. Without a more rigorous understanding of the
effects of restoration, continued reintroduction of fire will inevitably lead to more
catastrophic overstory mortality and hasten the decline in southeastern pine-dominated
ecosystems (Landers et al. 1995, South and Buckner 2003).
Smoldering duff and tree decline and mortality is a familiar phenomenon in
ecosystems maintained by frequent fires outside of the southeastern US where, in
response to fire suppression, deep organic horizons accumulate around large conifers,
creating a potential for mortality when fire is re-introduced (Ryan and Frandsen 1991,
Swezy and Agee 1991, Haase and Sackett 1998, Stephens and Finney 2002, McHugh and
Kolb 2003). It is likely that as native ecosystems continue to be degraded by fire
suppression and restoration efforts ensue, we will experience other novel disturbances
that will challenge future conservation and restoration.
16
It is ironic that southeastern USA pinelands are imperiled by fire suppression but
the reintroduction of fire often results in the death of a large portion of the residual pines.
Clearly, if fire is to be a useful tool for restoring the remnant stands from which it has
been excluded for decades, the fire-induced mortality problem needs to be solved. As
described, consumption of novel fuels in fire excluded stands play a major role in
contributing to fire-induced pine mortality. Reducing these novel fuelbeds, characterized
by well-developed forest floor horizons, should be a primary restoration objective for
managers attempting to reintroduce fire into excluded stands. Multiple fires over many
years may be necessary for the gradual elimination of these novel fuels prior to meeting
ancillary restoration objectives such as midstory reduction or understory restoration. At
small scales, extinguishing duff fires can save many of the large old trees for which these
ecosystems are valued, but such efforts are expensive and thus unlikely to be viable over
large areas. Nevertheless, understanding the patterns and processes of duff fire-induced
mortality represents an important step towards restoring and maintaining southeastern
pine ecosystems as viable components in our conservation landscape.
17
Table 2-1. Reports of excessive overstory longleaf pine mortality following re-
introduction of fire into pine ecosystems after decades of fire suppression.
Federal Agencies
USDA Forest Service
Ocala National Forest, FL
a
Talladega National Forest, AL
b
US Fish and Wildlife Service
Mountain Longleaf National Wildlife Refuge, AL
b
Department of Defense
Eglin Air Force Base, FL
c
Fort Gordon, GA
b
Fort Jackson, SC
b
State Agencies & Institutions
Austin Cary Forest, FL
b
Autauga Demonstration Forest, AL
b, h
Florida Division of Forestry
b, d
Florida Fish and Wildlife Conservation Commission
c
Florida Park Service
e
Georgia Department of Natural Resources
c
North Carolina Division of Parks & Recreation
c
University of Florida
f
NGO Land Management Agencies
The Nature Conservancy
Alabama Chapter
c
Florida Chapter
f
Georgia Chapter
c
Louisiana Chapter
g
Forest Industry
International Paper, Cantonment, FL
c
A- Harold G. Shenk, personal communication, September 2001
B- Personal observations
C- Varner and Kush 2004
D- Jim Meeker, personal communication, April 2001
E- Erik Johnson, personal communication, November 2003
F- Walt Thomson, personal communication, March 2003
G- Nelwin McInnis, personal communication, November 2000
H- John McGuire, unpublished data
18
Figure 2-1. A frequently-burned longleaf pine ecosystem reference condition at the
Katharine Ordway Preserve-Swisher Memorial Sanctuary in northern Florida.
Pristine pinelands are increasingly rare in current landscapes of the
southeastern USA.
19
Figure 2-2. Typical long-unburned (37 years since fire) longleaf pine forest at the
Katharine Ordway Preserve-Swisher Memorial Sanctuary, Florida. Many
pinelands throughout the southeastern USA have undergone decades of fire
suppression, leading to increases in the midstory, organic forest floor soil
horizons, and decreases in plant and animal species richness.
20
Figure 2-3. Forest floor development in a long-unburned (ca. 40 years since fire) longleaf
pine forest at Eglin Air Force Base, Florida. In frequently-burned pinelands,
only a thin Oi horizon forms; Oe and Oa horizons are signs of prolonged fire
suppression. In many long-unburned pinelands organic soil accumulations
surrounding large pines can exceed 25 cm in depth.
Oi
Oe
Oa
Mineral soil
21
Figure 2-4. Restoration fires in long-unburned longleaf pine forests damage canopy,
stem, and root tissues often leading to excessive tree mortality. Flaming and
smoldering fire can cause direct damage to canopy, stem, and root tissues.
Pine mortality has been linked to smoldering combustion of duff near trees,
perhaps caused by damage to root and/or stem tissues, or from indirect effects
due to increased physiological stress.
22
CHAPTER 3
TREE MORTALITY RESULTING FROM REINTRODUCING FIRE TO
LONG-UNBURNED LONGLEAF PINE FORESTS: THE ROLE OF DUFF
MOISTURE
Introduction
The frequency, intensity, and severity of fires have been drastically altered in many
contemporary landscapes (Agee 1993, Minnich et al. 1995, Ottmar et al. 1998, Outcalt
2000, Busse et al. 2000, Barton 2002, Wright and Agee 2004). In ecosystems historically
maintained by frequent fires, fire exclusion leads to increased tree and shrub density,
changed species composition, altered nutrient cycles, and fuel accumulation (Heyward
1939, Cooper 1960, van Wagtendonk 1985, Ware et al. 1993, Covington and Moore
1994, Minnich et al. 1995, Gilliam and Platt 1999, Everett et al. 2000, Varner et al. 2000,
Keane et al. 2002, Wright and Agee 2004, Chapter 2). When fires do occur after a
prolonged period of suppression, these legacies in fuel structure and composition result in
altered fire behavior and effects (Johnson and Miyanishi 1995). Although reinitiation of
historic fire regimes is a major component of most approaches to restoring
fire-suppressed forests, woodlands, and savannas worldwide, burning these previously
fire-suppressed ecosystems often fails to achieve the desired results (Fulé et al. 2004,
Chapter 2).
Among the impediments to restoration of fire regimes after long periods of fire
suppression is the excessive tree mortality resulting from reintroduction of fire (Wade et
al. 1997, Varner et al. 2000, Stephens and Finney 2002, McHugh and Kolb 2003, Fulé et
al. 2004, Chapter 2). In ecosystems maintained by frequent low-intensity fires, the
23
likelihood of fire-induced mortality generally decreases with increasing tree diameter
(e.g., Ryan and Reinhardt 1988, Peterson and Ryan 1985, Wade and Johansen 1986,
Ryan et al. 1988). In restoration fires, in contrast, tree mortality follows varying patterns,
sometimes increasing with increasing tree diameter (Kush et al. 2004) and occasionally
with bimodal peaks of mortality in small and large diameter trees (Swezy and Agee 1991,
Varner et al. 2000, McHugh and Kolb 2003). The mechanism responsible for increased
mortality caused by restoration fires is unclear, with different investigators emphasizing
damage to canopy (Wyant et al. 1986, Ryan and Reinhardt 1988, Menges and Deyrup
2001), stem vascular tissue (Ryan and Frandsen 1991, McHugh and Kolb 2003), and root
tissues (Swezy and Agee 1991, McHugh and Kolb 2003), as well as indirect effects of
these damages on tree stress and defense against pathogens (Ostrosina et al. 1997,
Ostrosina et al. 1999, Menges and Deyrup 2001, Feeney et al. 1998, Wallin et al. 2003,
McHugh et al. 2003). Determining causes of fire damage should help managers predict
fire-caused mortality and inform burn prescriptions to avoid or reduce post-fire tree
mortality.
Restoration of longleaf pine (Pinus palustris Mill.) ecosystems in the southeastern
USA after long periods of fire suppression is a major conservation and management goal
(Hermann 1993, Landers et al. 1995, Wade et al. 1997, Johnson and Gjerstad 1998,
Chapter 2). Reference longleaf pinelands have park-like stand structure, are often
mono
-
dominant, and have a fire return interval of 1 to 5 years (Christensen 1981, Robbins
and Myers 1992). Since European settlement, 97 % of longleaf pinelands have been lost
(Frost 1993) and only half of all remnant pinelands are burned regularly (Outcalt 2000).
Overstory mortality in restoration fires can be as high as 75 to 95%, causing radical shifts
24
in ecosystem structure and composition (Varner et al. 2000, Chapter 2) and defeating
restoration goals. This conflict between restoration objectives and outcomes has
confounded pineland restoration efforts region-wide (Chapter 2).
In 2001, a large-scale restoration experiment was initiated to examine the correlates
of mortality in long-unburned longleaf pine forests. Motivated by the reported
correlations of conifer mortality with duff consumption (Swezy and Agee 1991, Ryan
and Frandsen 1991), varying duff volumetric moisture contents were used as treatments.
We hypothesized that overstory pine mortality would increase with duff (Oe+Oa)
consumption and decrease with day-of-burn duff moisture content. Additionally, damage
to canopy, stem, and roots were examined at the stand and individual tree levels as they
related to tree mortality. The effects of tree, fuel, and burn characteristics were tested as
predictors of overstory pine mortality in restoration fires.
Methods
Study Area
Experimental burns were conducted in four blocks in long-unburned (35 to 45
years since fire) longleaf pine forests at Eglin Air Force Base on the Florida Panhandle,
USA (N 30° 38’, W 86° 24’; Figure 3-1). All blocks had heavy downed woody and duff
fuel loading (ranging from 3.14 to 7.63 kg m
-2
), remnant longleaf pine overstory (45-200
trees > 10 cm DBH ha
-1
), and the altered midstory and canopy tree species composition
typical of long-unburned pine forests on the southeastern Coastal Plain (Heyward 1939,
Peet and Allard 1993, Gilliam and Platt 1999, Kush and Meldahl 2000). All sites are
within the Southern Pine Hills District of the Coastal Plain Physiographic Province with
deep, well-drained sandy soils (Brown et al. 1990). Soils of the study sites were all typic
Quartzipsamments of the Lakeland series with mean depth to water table exceeding 200
25
cm (Overing 1995). The climate of the area is subtropical, characterized by warm, humid
summers and mild winters, with mean temperatures of 19.7°C and mean annual
precipitation of 1580 mm, most of which falls from June to September (Overing et al.
1995). Elevations of the study sites are 52 to 85 m asl and all sites have typical sandhill
topography with minimal effects of slope and aspect (Myers 1990).
Stands at each of four sites were randomly assigned to one of four burning
treatments based on collected day-of-burn volumetric duff moisture content (vdmc;
percent of dry weight): dry (60% vdmc); moist (90% vdmc); wet (120% vdmc); and an
unburned control. Moisture contents of forest floor (Oi; Oe, and Oa horizons) and woody
fuels were calculated based on day-of-burn collections (Table 3-1.) For each
experimental fire, fire weather and fire behavior were recorded periodically (Table 3-1).
All prescribed burns were ignited during the late dormant season (February to April). To
minimize variation in fire behavior, all fires were ignited using narrow strip head fires or
spot-grid ignition (10 to 30 m between strips or spots; Wade and Lunsford 1989), with
ignition managed to minimize variation in flame lengths and rate of spread.
Field Data Collection
Characteristics of ground fuels and vegetation were measured in each 10 ha stand
prior to burning. Forest floor depth was measured by horizon (litter [Oi horizon] and duff
[Oe and Oa horizons]) at the base of each of 50 randomly selected pines, 15 to 60 cm dbh
(dbh =stem diameter at 1.37 m), per plot. Forest floor depth was measured at each tree
using eight 20 cm pins buried flush with the litter surface and offset from the stem
approximately 10 cm at cardinal and ordinal directions. Total woody fuel loading was
estimated for each plot using Brown’s (1974) planar intercept method. For all overstory
26
pines (> 15 cm dbh), their dbh, total height, crown height, and distance and direction to
plot centers was recorded.
Initial post-burn measurements on all plot trees were made 3 to 4 weeks following
the experimental fires. Stem char heights were estimated with a height pole at four
cardinal directions. Scorch height and maximum height of needle consumption were
estimated using a clinometer. Percent of canopy volume scorched was estimated by the
same two trained observers on all trees. Post-fire reductions in forest floor depth at
individual trees were measured as the average difference between the pre- and post-fire
exposure of duff pins. Basal damage and evidence of pathogens were noted for all plot
trees. Following initial post-burn surveys, tree mortality and any signs of decline or
disease 6, 12, 18, and 24 months post-burn were noted to capture the temporal patterns of
mortality.
Data Analysis
The study was designed as a randomized block design with four treatments and
four replicates per treatment. The four treatments were based on day-of-burn volumetric
duff moisture content (wet, moist, dry and a no-burn control, as described above),
collected on the morning of each burn (Table 3-1).
At the burn block level, analysis of variance was used to detect effects of treatment
duff moisture (wet, moist, dry, control) on pine mortality. If differences were detected, a
post-hoc Tukey-Kramer HSD test was used to detect differences between treatments. The
same design was used to test for effects of duff moisture treatments (wet, moist, dry) on
forest floor reduction (cm and %), stem char height, and canopy damage (scorch %,
scorch height, and % canopy consumption). When necessary, non-normal data were
transformed to meet the assumptions of parametric analyses. Fuel environment
27
parameters tested in the multiple regressions were: relative humidity (%); air
temperature; moisture contents of litter (Oi), fermentation (Oe), and humus horizons
(Oa); and, moisture contents of 10-, 100-, and 1000-hour fuels. The role of crown scorch
(plot means of scorch height and percent of canopy volume scorched), height of stem
char, and duff consumption on pine mortality were also evaluated.
Given the substantial within-plot variation in fire effects and post-fire tree
mortality, individual tree post-fire mortality was modeled using the nested logistic
regression approach employed in a similar study by Stephens and Finney (2002).
Probability of pine mortality was modeled as a function of pre-burn fuel characteristics
(dbh, pre-fire duff depth, crown height, and total tree height), duff moisture treatment,
and fire effects (% duff consumed, duff depth reduction, char height, % crown scorch,
and scorch height). When variables were highly correlated (e.g., crown scorch height and
% crown scorch), the less correlated variable was removed from subsequent model
iterations. Akaike’s Information Criteria (AIC) were utilized to rank competing
regression models.
Results
Duff consumption, stem char, and canopy damage varied among the 12
experimental fires (Table 3-2). Across all twelve burns, mean duff consumption at the
bases of the pines ranged from 2 to 63% (0.2-6.2 cm). Duff consumption in the dry burns
was six times greater than in wet burns and three times greater than in moist burns (Table
3-2). Mean height of stem char was 2.24 m across all burns; individual plot means ranged
from 0.9 to 5.1 m with no apparent treatment effect (Table 3-2). Mean percent of pine
canopy scorched and scorch height ranged from 4.3 to 71.7 % and 3.5 to 16.7 m,
respectively, across all burned plots and was independent of treatment (Table 3-2). Due
28
to the low intensity of all the prescribed burns, only 3 trees (< 0.7% of all burned trees)
suffered any needle loss during the fires (crown “consumption” in contrast to crown
“scorch,” where needles are heated, not combusted) during the fires; needle consumption
did not exceed 10% of the total canopy volume of any tree and was therefore not included
in the analyses.
Overstory longleaf pine mortality for all plots during the first two years after the
fires ranged from 0 to 42%, with only three plots exceeding 10% mortality (Table 3-2).
Variation in mortality rates was greatest among the four dry burns, where overstory pine
mortality ranged from 8 to 42%. Soon after the fires, all pines in the burned blocks
produced new foliage and none showed obvious signs of decline; mortality generally
lagged 12 to 18 months after fires. All of the pines that died showed signs of bark beetle
attack (primarily Dendroctonus terebrans and, Platypus spp.), as did several surviving
pines. Overstory pine mortality during the first two years after fires differed among
treatments; dry burns averaged 20.5% mortality, moist burns averaged 3.0 %, and both
wet and control burns suffered < 1% mortality over the two years post-burn (Figure 2; F=
10.56, df = 12, p < 0.001).
Univariate linear regressions using plots as replicates revealed that pine mortality
was related to various parameters describing the fuel environment and fire effects.
Among fuel environment variables, ambient air temperature was positively related to tree
mortality (p = 0.04), while 100-hour and both Oe and Oa horizon fuel moisture were all
negatively related to mortality (p < 0.05; Table 3-3). The best fuel environment
predictors were Oa and 100-hour fuel moisture, respectively explaining 50 and 47% of
variation in pine mortality. Among fire effects variables, consumption of forest floor
29
(litter plus duff), duff consumption, and stem char were positively related (p< 0.05) to
overstory pine mortality (Table 3-3). Forest floor reduction alone explained 71% of the
variation in overstory pine mortality.
At the whole plot level, fuel environment and fire effects variables were related to
pine mortality (Table 3-4). Among fuel variables, lower duff (Oa) moisture content was
negatively related to probability of tree mortality. Among fire effects variables, as bole
char and duff consumption increased, so did the tree death. When all fuel environment
and fire effects variables were included in a variety of single and multiple regression
models (Table 3-4), the model with the highest R
2
(0.57) and lowest AIC (5.1) included
only duff consumption (p<0.01):
log pine mortality (%) = -2.93 + 0.15 (% duff consumption).
The results of the logistic regression of individual tree mortality supported the
importance of fire effects parameters demonstrated in the plot-based analyses. Combining
trees from all of the treatments in a nested logistic regression model revealed that the
likelihood of a tree dying was related to a combination of canopy scorch and duff
consumption (Table 3-5; R
2
= 0.34, p < 0.001). This analysis revealed that while duff
consumption was an important factor across all treatments, canopy scorch was significant
only in the dry burning treatment. No fuel environment parameters were incorporated
into this model because fuel environment variables were measured at the plot level, not at
individual plot trees, which may explain the low R
2
value.
Discussion
In this study, maximum fuel reduction was linked strongly to the greatest loss of
overstory pines (Figure 3). Restoration of long-unburned southeastern pine forests is
complicated by the often opposing goals of reducing fuels while retaining mature
30
overstory pines. Further, although most models of fire-caused mortality assume that as
tree size, bark thickness, and canopy height increase, so does resistance to fire damage
(e.g., Martin 1963, Ryan et al. 1988, Dickinson and Johnson 2001), fire-caused longleaf
pine mortality rates in this study increased with tree size. This pattern appears related to
the deeper accumulations of forest floor fuels around the bases of large trees. Given that
work on other conifers has also found post-fire reductions in survival (Stephens and
Finney 2002, McHugh and Kolb 2003) and growth (Busse et al. 2000, Elliott et al. 2002)
with fire-caused reductions in forest floor and duff, these results may be more generally
relevant.
Fire-induced overstory pine mortality lagged 12 to 18 months after the burns,
consistent with reports in other conifers (e.g., Wyant et al. 1986, Stephens and Finney
2002) and observations by longleaf pine resource managers (Chapter 2). This lag may
have been due to the fact that the large, old trees had several predisposing mortality
factors (i.e., reduced physiological condition, high root volumes in drought-prone duff,
greater duff fuel accumulation, high stand density resulting from fire exclusion) in
addition to the direct fire injuries to roots, cambia, and canopy meristems. Kelsey and
Joseph (2003) suggest that the fire-caused root damage may lead to water stress and
subsequent declines in tree defenses. Delayed mortality may have also been caused by
reduced allocations to defense leading to attack by beetles or susceptibility to fungal
pathogens, both common after fires (Ostrosina et al. 1997, Ostrosina et al. 1999, Menges
and Deyrup 2001, Kelsey and Joseph 2003) and detected in our study area. This work,
however, suggests more proximate causes of the lags in tree mortality. Future work is
31
needed on the physiological mechanisms of decline and mortality in response to fire
damage (see below).
Mortality and Fire Damage to Stem, Canopy, and Roots
Duff consumption due to smoldering combustion was the only fire effect
consistently related to pine mortality across scales and treatments (Tables 3-3 to 3-5).
Lower duff (Oa) percent fuel moisture explained 78% of variation in duff consumption (p
< 0.0001), consistent with results from other coniferous forests (Frandsen 1987, Brown et
al. 1991, Ottmar et al. 1991, Sandberg et al. 2001). Duff is consumed by smoldering
combustion, causing long-duration heating to stems and organic and mineral soil (Swezy
and Agee 1991, Hungerford et al. 1995). The long duration (often many hours post-
ignition) of moderately elevated soil temperatures can damage or kill roots (Swezy and
Agee 1991) and stem vascular tissues (Ryan and Frandsen 1991).
Stem char has been correlated with decreased growth and increased tree mortality
in studies of various conifers (e.g., van Wagner 1973, Dixon et al. 1984, Peterson and
Arbaugh 1986, Wade and Johansen 1986) and char height was positively related to
longleaf pine mortality in our study (Table 3-3). The exclusion of small (dbh < 15 cm)
fire-susceptible trees in this study may also have reduced the overall effect of char height
on tree mortality. It is important to note, however, that measurements of char do not
capture the results of long-duration heating during the smoldering phase of combustion.
McHugh and Kolb (2003) supplemented char height measurements with char severity
(=stem damage) and found strong correlations with post-fire mortality. Given the high
correlation between char height, canopy damage, and fire behavior, its ease of
measurement and persistence, and its potential role in tree mortality, char height can be a
32
valuable post-fire measurement for understanding tree mortality (see review in Fowler
and Seig 2004).
Canopy scorch was a significant predictor of longleaf pine mortality in the nested
logistic regression and has been associated with post-fire mortality in many other conifers
(Peterson and Arbaugh 1986, Wyant et al. 1986, Ryan et al. 1988, Menges and Deyrup
2001, Stephens and Finney 2002). The logistic regression models nested by treatment
(dry, moist, wet burns) revealed that the role of scorch was confined to the dry treatment
(Table 3-5), a potential explanation for the contradictory results found in restoration
burns in which scorch was prevalent and a good predictor of mortality (Menges and
Deyrup 2001) or absent altogether and therefore not related to mortality (Kush et al.
2004). This finding is critical in that it highlights the challenges of making sense of fire
effects with an incomplete understanding of fire behavior. Understanding the dynamic
role of canopy scorch in tree mortality will require a more mechanistic approach that
isolates scorch from other types of fire damage.
The results of this study reinforce the need to better understand the mechanisms by
which fires kill trees (Dickinson and Johnson 2001, Kelsey and Joseph 2003, Fowler and
Seig 2004). Future work on fire-caused mortality should isolate the different types of fire
damage (e.g., isolating fire damage to canopy tissues from damage to below-ground
tissues; Carter et al. 2004) and consider their interactions (e.g., heat from smoldering duff
damages vascular tissues in surficial roots and the bases of tree stems rendering trees
susceptible to beetle attack). Such studies should combine whole tree physiology models
with considerations of fire damage and subsequent increases in tree stress (Manion 1991,
Dunn and Lorio 1992, Ryan 2000, Kelsey and Joseph 2003, Wallin et al. 2003).
33
Challenges to Restoration Using Fire
Given the clear role of duff moisture content in causing tree mortality, efforts to
understand moisture dynamics of duff should be a high research and management
priority. Duff moisture content is controlled by several factors acting at different scales.
At large scales, duff moisture is controlled by recent precipitation amounts and temporal
distributions (Ferguson et al. 2002). Within stands, localized duff moisture can vary with
forest floor composition, beneath and between tree crowns, and at small scales in basal
duff accumulations (Gordon and Varner 2002, Miyanishi and Johnson 2002). Duff fuel
particles vary in their physical and chemical composition (e.g., suberin-rich bark in
contrast to extractive and cellulose-dominated pine needle litter), both of which affect
ignition and burning characteristics at small scales.
Burning prescriptions for long-unburned forests will continue to require
consideration of traditional synoptic weather variables but should be enhanced to
acknowledge the important role of forest floor fuel complexes. Prescriptions should
incorporate more robust estimates of time and duration of localized precipitation and
drying, duff moisture (Ferguson et al. 2002), and the subsequent effects of these factors
on fire behavior, particularly residence time (Frandsen 1987, Frandsen 1991, Miyanishi
2001, Miyanishi and Johnson 2002). Given the diversity in forest floor composition and
structure in long-unburned longleaf pine forests, the drying, wetting, and burning
behavior of this fuel complex warrant further study.
Prescriptions for fire reintroduction in long-unburned forests should be tailored to
local operational scales. For example, it may be operationally feasible to use a
combination of wetting (either with backpack or tractor-mounted units) and/or raking
away forest floor fuels to minimize potential for long-duration heat damage to high-value
34
trees in small stands. Pre-fire raking has been used successfully in P. ponderosa burning
(Feeney et al. 1998, Wallin et al. 2002) but less often in longleaf pine forests. In addition
to the labor intensity of raking, a few longleaf pine forest managers have reported what
they believe to be raking-induced mortality (Varner and Kush 2004). Extinguishing
smoldering duff fires with tractor, ATV, or backpack-mounted water or foam has been
successfully employed in small (< 20 ha) restoration fires (Kush et al. 2004).
Unfortunately, none of these labor-intensive treatments are appropriate for large-scale
restoration burning (Chapter 2). Managers of large forests throughout the southeastern
USA are faced with burning large areas with limited staff and increasing limitations on
smoke production (Hiers et al. 2003). For large areas, utilizing duff moisture and
optimizing burn resources to priority stands must be the rule. In these situations,
acceptable thresholds of tree mortality must be balanced with duff and woody fuel
consumption and placed in the context of landscape objectives.
In restoration fires, managers and forest and fire scientists must recognize the
increasingly important role of forest floor fuels and the unintended damage they can
cause. Traditional fire effects such as canopy scorch and stem char are readily observed
and modeled. While more subtle, the relationships between duff consumption and
overstory mortality deserve increased attention and additional small and large-scale
experimentation. The reason for this oversight may be that this phenomenon was
formerly rare but getting more common as fires are being reintroduced after long periods
of suppression (e.g., Ryan and Frandsen 1991, Swezy and Agee 1991, Haase and Sackett
1995, Stephens and Finney 2002, McHugh and Kolb 2003). While duff smoldering is
problematic, land managers are left with the conundrum of having to reduce duff fuels
35
while maintaining a living overstory. Incorporating fuel moisture thresholds in burn
prescriptions that include measures of the forest floor fuel complex may help managers
avoid widespread overstory tree damage and mortality and successfully reduce fuels
when they reintroduce fires after long periods of exclusion.
36
Table 3-1. Fire weather and fuel moisture values for each 10 ha experimental prescribed burn in long-unburned longleaf pine forests at
Eglin Air Force Base in northern Florida, USA.
Site Treatment
1
Burn Date Rel. Hum. Air Temp. Wind Speed Fuel Moisture Content (% dry wt.)
2
(%) (°C) (m sec
-1
) Oi Oe Oa 10-hr 100-hr 1000-hr
3
RT01 Wet Feb 18/01 26 17 0.45 18 69 102 26 54 83
RT02 Wet Mar 05/02 24 12 0.89 27 70 125 43 61 --
RCX Wet Mar 14/02 64 20 1.34 22 76 133 42 45 77
RCM Wet Mar 04/02 21 8 0.89 35 91 137 52 66 86
Wet Treatment Means 34 a
4
14 a 0.89 a 26 a 76 a 124 a 41 a 57 a 82 a
RT01 Moist Mar 27/01 24 16 0.45 14 41 117 20 44 119
RT02 Moist Feb 22/02 43 17 1.34 17 43 79 25 45 --
RCX Moist Mar 08/02 35 24 1.79 16 45 112 22 28 70
RCM Moist Feb 22/02 37 18 0.45 14 27 102 18 39 64
Moist Treatment Means 35 a 19 ab 1.01 a 15 b 39 b 103 a 21 b 39 b 84 a
RT01 Dry Apr 26/01 43 32 0.45 16 18 40 18 15 92
RT02 Dry Mar 24/02 39 23 0.89 12 27 87 13 22 --
RCX Dry Apr 07/02 60 19 1.34 11 46 64 9 13 33
RCM Dry Apr 24/02 53 28 0.89 7 11 59 14 18 47
Dry Treatment Means 49 a 26 b 0.89 a 12 b 26 b 62 b 14 b 17 c 57 a
1
Prescribed fire treatments were based on day-of-burn volumetric duff moisture content.
2
Fuel moisture content for each fuel category was calculated based on day-of-burn collections of 5 samples per fuel category.
3
1000-hour fuel moisture contents are based on an average of both sound and rotten downed fuels (R. Ottmar, unpublished data).
4
Different letters following treatment means denote significant differences among column means determined with ANOVA followed
by post-hoc Tukey’s HSD with α = 0.05.
37
Table 3-2. Effects of 10 ha prescribed burns in long-unburned longleaf pine forests at Eglin Air Force Base in northern Florida, USA.
Site Treatment
1
Forest Floor Duff Stem Char Canopy Scorch Pine Mortality
2
cm loss % loss cm loss % loss m m height % SC %
RT01 Wet 1.3 (1.0)
3
24 (8) 0.2 (0.3) 2 (4) 1.0 (0.6) -- -- 0
RT02 Wet 3.7 (2.3) 33 (15) 0.5 (1.3) 5 (15) 1.4 (0.8) 4.5 (5.6) 5.8 (11.3) 0
RCX Wet 3.2 (2.3) 33 (13) 0.3 (1.3) 4 (13) 0.9 (1.7) 4.1 (4.2) 4.3 ( 5.2) 0
RCM Wet 7.9 (4.9) 49 (14) 1.2 (2.9) 9 (17) 2.1 (1.4) 10.3 (6.1) 23.6 (26.7) 2
Wet Treatment Means 4.0 35 a
4
0.6 5 a 1.4 a 6.3 11.2 a 0.5 a
RT01 Moist 3.4 (3.2) 28 (17) 0.9 (1.8) 9 (17) 1.2 (1.0) -- -- 0
RT02 Moist 3.3 (2.8) 31 (19) 1.0 (2.2) 11 (20) 1.0 (0.8) 3.5 (4.3) 8.8 (15.4) 4
RCX Moist 7.8 (5.0) 52 (16) 2.3 (2.7) 22 (20) 3.1 (1.5) 10.2 (4.7) 35.9 (32.7) 6
RCM Moist 7.4 (4.7) 56 (16) 1.4 (3.0) 16 (23) 5.1 (2.8) 16.7 (3.4) 71.7 (29.7) 2
Moist Treatment Means 5.5 42 a 1.4 14.5 a 2.6 a 10.1 38.8 a 3.0 a
RT01 Dry 10.3 (4.7) 73 (26) 6.2 (4.0) 63 (34) 1.2 (1.0) -- -- 22
RT02 Dry 6.9 (3.2) 63 (14) 1.6 (2.0) 25 (23) 2.8 (1.6) 8.4 (6.3) 34.3 (36.6) 10
RCX Dry 6.7 (3.6) 72 (11) 2.8 (2.6) 49 (18) 3.1 (1.4) 11.9 (3.1) 58.4 (31.4) 8
RCM Dry 9.6 (5.0) 73 (17) 4.6 (3.9) 49 (27) 4.1 (2.2) 15.3 (3.7) 70.4 (33.7) 42
Dry Treatment Means 8.4 70 a 3.8 46.5 b 2.8 a 11.9 54.4 a 20.5 b
1
Prescribed fire treatments were based on day-of-burn volumetric duff moisture content.
2
Percent tree mortality is cumulative mortality of all pines > 15 cm dbh. Mortality values for 2001 burns include surveys 36 months
post-burn; mortality values for all 2002 burns include surveys 24 months post-burn.
3
Values in columns are plot means (n=50 trees per plot), with standard deviations noted parenthetically.
4
Different letters following treatment means denote significant differences among column means determined with ANOVA followed
by post-hoc Tukey’s HSD with α = 0.05.
38
Table 3-3. Univariate linear regression results for correlates of individual pine mortality (n=450 trees) resulting from prescribed
restoration burns in long-unburned forests at Eglin Air Force Base in northern Florida, USA.
Dependent Variable Independent Variable R
2
P
Day-of-Burn Variables
log Pine Mortality (%) Ambient Air Temperature (C) 0.36 0.04
log Pine Mortality (%) Relative Humidity (%) 0.11 0.30
log Pine Mortality (%) 10-hour Fuel Moisture (%) 0.26 0.09
log Pine Mortality (%) 100-hour Fuel Moisture (%) 0.47 0.01
log Pine Mortality (%) 1000-hour Fuel Moisture (%) 0.31 0.12
log Pine Mortality (%) Oi (litter horizon) Fuel Moisture (%) 0.17 0.19
log Pine Mortality (%) Oe (fermentation horizon) Fuel Moisture (%) 0.36 0.04
log Pine Mortality (%) Oa (humus horizon) Fuel Moisture (%) 0.50 0.01
Fire Effects Variables
log Pine Mortality (%) Forest Floor (litter+duff) Reduction (%) 0.71 <0.001
log Pine Mortality (%) Duff Reduction (%) 0.57 0.04
log Pine Mortality (%) Canopy Scorch Volume (%) 0.43 0.06
log Pine Mortality (%) Stem Char Height (m) 0.38 0.03
39
Table 3-4. Step-wise multiple regression results for predictors of block-level longleaf pine mortality caused by prescribed restoration
burns in long-unburned pinelands at Eglin Air Force Base.
Dependent Variable Parameters R
2
P AIC
1
ß
0
ß
1
ß
2
Fuel Environment Variables
log Pine Mortality (%) = 6.26 - 0.07 (Oa% moisture) .45 0.04 19.8
Fire Effects Variables
log Pine Mortality (%) = -1.01 + 0.23 (char height) +0.03 (% duff reduction) .82 < 0.001 15.9
Combined Model (Fuel Environment + Fire Effects)
log Pine Mortality (%) = -2.93 + 0.15 (% duff reduction) .57 < 0.01 5.1
1
AIC (Akaike’s Information Criteria) values rank competing regression models; models with lower AIC values approximate modeled
responses better than models with higher AIC values.
40
Table 3-5. Modeled nested logistic regression of post-fire individual tree longleaf pine mortality at Eglin Air Force Base in northern
Florida, USA.
Fire Effects Nested within Treatments Parameter Estimate χ
2
P R
2
Tree Status (alive/dead)
Intercept 3.16 21.29 <0.001
Dry Treatment % Canopy Scorch -0.03 11.65 <0.001
Moist Treatment % Canopy Scorch -0.01 0.32 0.57
Wet Treatment % Canopy Scorch -0.03 1.39 0.24
Dry Treatment % Duff Consumption -0.04 15.44 <0.001
Moist Treatment % Duff Consumption -0.03 7.29 <0.01
Wet Treatment % Duff Consumption -0.07 5.03 <0.01
Whole Model 96.35 <0.001 0.34
41
Figure 3-1. Study site locations of experimental prescribed fires to examine correlates of
mortality resulting from reintroduction of fire into fire-suppressed (ca. 35 – 45
years since fire) longleaf pine (Pinus palustris) forests.
42
Figure 3-2. Effects of burning treatment on cumulative overstory longleaf pine mortality
at Eglin Air Force Base, Florida. Treatments were based on day-of-burn duff
moisture contents, where wet = 115% duff moisture content (DMC), moist =
85% DMC, dry = 55% DMC, and an unburned control. All treatments were
replicated 4 times over the two years of the study.
43
Figure 3-3. Relationship between duff consumption and cumulative post-fire mortality of
overstory longleaf pines at Eglin Air Force Base, Florida.
R
2
= 0.73
p < 0.0001
44
CHAPTER 4
EFFECTS OF BASAL BARK, DUFF, AND SOIL TEMPERATURES ON POST-FIRE
TREE STRESS AND GROWTH
Introduction
Although frequent fires are necessary for the maintenance of many
conifer-dominated ecosystems, when fires are reintroduced after long periods of
exclusion they can stress, reduce the growth of, and kill even large trees. Increasing our
understanding of fire induced conifer mortality is important given on-going efforts to
reintroduce fire to areas that have suffered long periods of fire exclusion (Swezy and
Agee 1991, Covington et al. 1997, Wade et al. 1997, Haase and Sackett 1998, Stephens
and Finney 2002, McHugh and Kolb 2003, Agee 2003, Varner et al. 2005). Fires
reintroduced following fire exclusion can damage root, stem, and canopy meristems and
lead to stresses that reduce the abilities of trees to defend themselves against pathogens or
weather periodic climatic stress (Feeney et al. 1998, Kolb et al. 1998, McHugh et al.
2003). A major deficiency in our understanding of post-fire tree mortality concerns the
linkages between fuels and fire behavior.
Understanding the causes of fire-induced mortality of conifers characteristically
resistant to fire damage is critical to the management of fire, fuels, and forest health.
While several of the proposed mechanisms of fire-induced mortality (i.e., damage to
stem, canopy, and root meristems and second-order effects related to pathogen
susceptibility) have received attention, the link between consumption of forest floor fuels
and tree mortality has increasing support (Swezy and Agee 1991, Stephens and Finney
45
2002, Chapter 3). For example, fire damage to surface and duff-bound roots (i.e., roots in
the mor humus) was proposed as a cause of mortality in P. ponderosa (Swezy and Agee
1991). But fire damage to stem vascular tissue also can also cause P. ponderosa death
(Ryan and Frandsen 1991, Ryan 2000). In the case of P. elliottii, canopy scorch has been
linked to post-fire mortality (Dixon et al. 1984, Wade and Johansen 1987, Menges and
Deyrup 2001). Combinations of fire-caused damage affect tree survival, and several
investigators report the strongest correlates of post-fire mortality are those that include
damages to multiple tissues (Wyant et al. 1986, Ryan and Reinhardt 1988, Ryan et al.
1988, Saveland and Neuenschwander 1990, Haase and Sackett 1998, Ryan 2000).
Distinguishing the actual causes of mortality will help managers modify their actions to
avoid or minimize post-fire tree mortality. Second-order fire effects such as bark beetle
infestations (Dixon et al. 1984, Menges and Deyrup 2001, McHugh et al. 2003) and
fungal infections (Ostrosina et al. 1997, Ostrosina et al. 1999) are cited as causes of
post-fire decline and mortality, but are likely linked to compromises in tree defenses
caused by fire-caused damages to stem, canopy, or root tissues. The inability to establish
a relationship among the responses of trees, the characteristics of the disturbing fire (i.e.,
temperatures and duration of lethal heating), and the fuels that fed them is a major
shortfall in our understanding of the ecology of fire as a disturbance and inhibits
restoration of fire-excluded ecosystems.
The objectives of this study were to link fuels, specific types of damage to
individual trees, and fire behavior to the subsequent growth, carbohydrate storage, and
mortality of mature Pinus palustris in replicated small-scale burns. Specifically, the
following hypotheses were tested:
46
Changes in first-year post-fire earlywood growth in response to burning are
minimal given that earlywood reflects the previous year’s growing conditions in
longleaf pine (Meldahl et al. 1999).
Given that latewood in longleaf pine responds to current year stresses (Meldahl et
al. 1999), decreases in latewood growth are greatest with greater duration of lethal
(>60°C) heating of basal bark, and roots in duff and mineral soil.
Changes in fine root non-structural carbohydrates in response to fire are minimal
because fine roots are either killed and replaced and/or are sinks for carbohydrates
from more coarse sources (stem and coarse roots; Guo et al. 2004).
Coarse root non-structural carbohydrate concentrations following burns decrease
with increasing duration of lethal (>60°C) heating to basal bark and roots in duff
and mineral soil.
Tree mortality increases with greater durations of duff smoldering and with more
severe fire damage.
This work was designed to inform regional restoration efforts and to contribute to our
understanding of the larger issue of post-fire overstory tree decline and mortality
common to many North American coniferous ecosystems.
Methods
Study Site
This experiment was conducted in a long-unburned (37 years since fire) longleaf
pine stand at the Katharine Ordway Preserve-Swisher Memorial Sanctuary near Melrose
(Putnam County), Florida, USA (N 29º 40’, W 81º 74; Figure 4-1). The stand was
dominated by an overstory of longleaf pine, with a thick midstory of oaks (Quercus
laevis, Q. geminata, and Q. hemisphaerica), a patchy remnant groundcover dominated by
Aristida stricta, and a thick forest floor (depths to 15 cm) typical of long-unburned xeric
southeastern pine ecosystems (Chapter 2). Soils of the site are deep, excessively
well-drained hyperthermic, uncoated Lamellic Quartzipsamments in the Candler series
(Readle 1990). The topography is gentle, with north-facing slopes < 5% and elevations
47
averaging 36 m above msl. The climate is humid, warm temperate with long, warm, and
humid summers and short, mild winters with annual temperatures and precipitation
averaging 20º C and 1432 mm, respectively (Readle 1990).
One hundred mature (30 – 50 cm DBH) individual longleaf pines across a 2 ha area
were randomly selected as experimental units for examination of fire-induced stress and
mortality. Five treatments (20 replicates per treatment) were assigned to trees in a simple
random design. The treatments allowed comparison of three hypothesized causes of
post-fire mortality: 1) root damage (ROOT); 2) root and stem damage (ROOT+STEM);
and , 3) stem damage (STEM). Two control treatments were installed to compare
background mortality and stress: a control treatment that burned but smoldering phase
combustion was prevented (CONTROL) and a control treatment that was not burned (NO
BURN). Individual pines formed the center of 1 m radius plots in which measurements
were focused (Figure 4-2). Each tree was ring-ignited with a drip torch from a raked line
1 m from the tree base over a 34-day period beginning 25 September and concluding 4
November 2003. In the ROOT+STEM treatment, fires were allowed to consume the
forest floor without protection, thereby heating both basal bark and underlying roots
(Figure 4-2). ROOT treatments were accomplished by first moving a 5 cm radius of basal
forest floor away from the stem, then protecting the stem with insulating material
(Cleveland Laminating Corp., Cleveland, OH, USA). In the STEM treatments, basal bark
was heated by burning a 20 cm radius area surrounding each tree base; the adjacent forest
floor was protected with the insulating material as described above. CONTROL burns
were extinguished with a flapper once flaming combustion ended.
48
Tree Measurements
One measure used to estimate post-fire tree stress was change in stem radial growth
following burns (Busse et al. 2001) Radial stem growth (mm) of all 80 burned treatment
trees was measured using increment cores (2 per tree, 90º apart) extracted one year
post-burn (January-February 2005). Cores were air-dried, mounted, and sanded according
to standard dendrochronological methodology (Stokes and Smiley 1968). All cores were
measured using a binocular microscope with both earlywood and latewood measured to
the nearest 0.01 mm.
Another metric used to estimate tree stress following burns was change in
non-structural carbohydrate concentrations in fine and coarse roots (Wargo et al. 1972,
Marshall and Waring 1985, Kozlowski and Pallardy 1997). Within a randomly selected
subset of 8 trees in each treatment (n=40), total non-structural carbohydrate
concentrations (starch + sugar) in roots were sampled within 10 days and again at 4
months post-burn (hereafter, “tnc
10
” and “tnc
120
”). The tnc
10
trees were sampled within 3
to10 days post-fire to minimize transformations of carbohydrates from their pre-treatment
pools (Kozlowski and Pallardy 1997). For each root carbohydrate sample, 3 g (dry
weight) of fine (1-2 mm diameter) and 3 g of coarse (2-5 mm diameter) longleaf pine
roots were unearthed from the top 20 cm of mineral soil within a 50 cm radius
surrounding treatment trees. All roots were bagged and immediately stored on dry ice.
Immediately upon removal from chilling, the roots were rinsed and oven-dried at 100° C
for 2 h, followed by drying at 70° C to a constant mass to minimize post-harvest
carbohydrate losses (Caldwell 1989).
Root non-structural carbohydrate samples were analyzed using a modified
phenol-sulphuric acid method (Buysee and Merckx 1993). From each fine and coarse
49
diameter dry root sample, 80 mg was extracted for 12 h in 10 ml 80% ethanol then
centrifuged at 2200 rpm for 15 minutes. The resulting supernatant was removed and
placed in a 50 ml flask. The residue was centrifuged a second time in 5 ml 80% ethanol
for 5 minutes, and the supernatant was transferred to the same volumetric flask. The
residue from the ethanol extractions was transferred to a glass tube, dried, and then boiled
for 3 hours in a 5 ml 3% HCl. The filtrate was adjusted to 50 ml in a volumetric flask and
used for the starch analysis. For total sugar and starch determinations, 1 ml of a solution
containing 20 to 80 µg sugar was transferred into a glass tube and 1 ml of a 28% phenol
in 80% ethanol was added. Five ml of concentrated sulfuric acid was immediately added
directly to the liquid surface. The tube was agitated for 1 minute and allowed to stand for
15 minutes prior to measuring absorbance at 490 nm in a Shimadzu UV
spectrophotometer (Shimadzu Corporation, Kyoto, Japan).
Pines from all 4 treatments (n=80) were surveyed for signs of decline and mortality
at 4, 8, 12, and 24 months after the burns. Because fire-induced mortality is often delayed
for 18 to 24 months (Chapter 2, 3), no initial mortality was expected; annual sampling
will continue in order to capture any delayed mortality.
Fuel Sampling
To understand the effects of fuels on fire intensity and tree damage, we measured
forest floor consumption surrounding each burned tree. Eight 20 cm tall steel pins were
installed flush with the forest floor 8 cm from the stem in cardinal and ordinal directions.
Following all burns, pins were measured for duff depth reduction [(initial depth -
post-fire depth) / initial depth]. To estimate moisture contents, forest floor fuels were
collected by horizon on each burn day from trees proximal to treatment trees. Samples
were divided along horizons, with Oi removed first, then Oe material, concluding with
50
Oa fuels. Since pine cones are likely vectors of duff ignition (Fonda and Varner 2005),
six cones (3 recent, 3 decomposing) within the drip line of each neighbor tree were
collected and sealed in polyethylene bags. In the laboratory, fresh mass of all fuels was
determined and were then oven-dried at 70°C to a constant weight to determine moisture
content and biomass.
Fire Measurements
Fire temperatures were measured on a subset of all treatment trees (6 trees per
burning treatment, n=24) during fires using Type J (range 0º to 1200°C; Omega
Laboratories, Stamford, CT) thermocouples connected to a Campbell Scientific CR10X
datalogger (Campbell Scientific, Logan, UT). Temperatures were measured on the bark
surface at three points 120º apart at approximately 10 cm above the mineral soil surface
(near the forest floor-bark interface; Figure 4-3). To assess root and soil heating,
thermocouples were buried 120º apart in the lower duff (Oa horizon) ca. 10 cm from the
tree base, and directly beneath these points in the mineral soil at 5, 10, and 20 cm depths
(Figure 4-3). To minimize soil disturbance, thermocouples were inserted into the soil in
narrow openings created by slicing into the soil at a slight angle with a machete and then
inserting the thermocouple lead at the predetermined depth. Maximum and mean
temperatures were stored every 2 minutes from 15 minutes pre-ignition to 1700 h when
the fires were required to be extinguished according to prescription and burn permits
from the Florida Division of Forestry. While truncated, this artificial burn extinction was
applied equally to all treatments and also fits with regional land management practices.
Fire behavior, weather, and fuel consumption were recorded for each experimental
fire. Fire measurements included maximum flame height (cm; ocularly estimated), flame
time (sec), and residual smoldering time (sec). Fire weather (2-m wind speed, air
51
temperature, and relative humidity) was recorded periodically during all fires (Table 4-1).
Post-burn forest floor consumption was estimated by measuring the difference in pin
exposure following fires.
Data Analysis
The experiment was a completely randomized design with 4 burning treatments
(ROOT, STEM, ROOT+STEM, CONTROL) with 20 replicates in each treatment. The
NO BURN trees were only used to monitor background stand mortality and root
carbohydrate changes, but future plans are to sample their radial growth and include these
trees in the larger analyses. ANOVA was used to detect overall treatment effects, with
any pair wise differences among treatments determined using a post-hoc Tukey-Kramer
HSD test (α= 0.05). The effects of treatments were tested on forest floor consumption
(%), duration (minutes) of heating >60ºC (a good approximation of lethal temperature for
plants; Byram 1958), changes in 1-year post-burn radial growth (earlywood and latewood
increment; %), mortality 2 years post-burn (%), and short-term post-burn changes in root
non-structural carbohydrates (fine and coarse diameter; %). In addition to the ANOVA,
we used a step-wise multiple regression to examine relationships between fuel moisture
contents (Oi, Oe, Oa, cones, 0-5 cm mineral soil) as predictors of fire behavior response
variables. Step-wise regressions were also used to relate heating duration in the different
strata (bark, duff, and mineral soil depths) to tree mortality (%), 1-year radial growth (%
change in earlywood and latewood increment, 2003-2004), and changes in root
non-structural carbohydrates (fine and coarse diameters, tnc
10
– tnc
120
) post-burn
response variables. To meet assumptions of parametric analyses, any non-normal data
were transformed prior to analysis according to convention (Zar 1996). Finally, given the
correlations among strata (i.e., the burning of duff fuel provides heat to underlying
52
mineral soil horizons that subsequently heat lower horizons), regression model iterations
were evaluated for their multicollinearity (Hintze 2004).
Results
Fire Behavior
Flame heights during the experimental burns ranged 0.4 to 3.0 m (mean =1.52 m ±
0.70 m, n=80), did not vary with treatment and were within the range of large restoration
fires (Chapter 3). Fire temperatures on the instrumented trees (n=24) during experimental
burns were highest above-ground, with average maximum basal bark temperatures in
individual fires ranging up to 476º C, average maximum duff temperatures to 304º C, and
average maximum mineral soil temperatures at 5, 10, and 20 cm below the surface to
134º, 117 º, and 80º C, respectively. Lethal heating durations were longest in duff (mean
= 74 ± 168.8 min), next longest on basal bark (mean = 36 ± 73.9 min), and then
decreased with increasing mineral soil depth (27 ± 4.8, 7 ± 14.4, 1 ± 2.5 min at 5, 10, and
20 cm depths, respectively; Table 4-2). Following ignition, basal bark temperatures
increased first, then duff, then sequential depths in the mineral soil. After initial heating,
basal bark temperatures across all treatments dropped below 60º C. As intended, basal
bark temperatures were higher in ROOT+STEM and STEM treatments than in ROOT
and CONTROL treatments (P = 0.06, df=3, F= 2.91). Duff and mineral soil temperatures
at all depths, in contrast, did not differ among treatments (Table 4-3). Average forest
floor consumption at the base of treatment pines differed among treatments (P < 0.001,
df=3, F= 6.69; Table 4-3), with the ROOT+STEM treatment having the greatest fuel
consumption. Smoldering time also differed among burning treatments (P= 0.018, df=3,
F= 5.49; Table 4-3), with all smoldering treatments (ROOT, ROOT+STEM, and STEM)
differing from the CONTROL treatment.
53
Temperatures on the basal bark and within the duff, and at 5, 10, and 20 cm deep in
the mineral soil were related to several fuel moisture parameters. In a step-wise
regression with day-of-burn fuel moistures and weather observations as independent
variables, the duration (min) of basal bark temperatures > 60º C was best predicted by
lower duff (Oa horizon) moisture content (P= 0.05, R
2
= 0.16). The best step-wise fit for
duration of duff temperatures > 60ºC was a function of Oe moisture (fermentation
horizon or “upper duff”) moisture content (P=0.01, R
2
= 0.24). None of the fuel moisture
predictors (including mineral soil moisture) were related to heating duration in the
underlying mineral soil (5, 10, and 20 cm depths; Table 4-4).
Radial Growth and Mortality
In the first year following fires no trees died in any treatment, and only two trees
died in the first 24 months. Of these two dead trees, one occurred in the ROOT+STEM
treatment, the other tree was in the UNBURN control treatment and appeared to be struck
by lightning. Neither earlywood nor latewood growth differed among treatments. In a
regression with instrumented trees (n=32) as replicates, radial growth (% change from
2003 ring radius) decreased with greater heating duration > 60ºC in the top 5 cm of
mineral soil (P = 0.08, R
2
= 0.16; Table 4-5). Earlywood increment was apparently
insensitive to heating durations, but the regression marginally related latewood growth to
duration of 10-cm mineral soil temperatures > 60ºC (P=0.069; R
2
= 0.17).
Root Nonstructural Carbohydrates
As with radial growth, changes in root nonstructural carbohydrates did not vary
among the burning treatments (Table 4-3). There were differences, however, when trees
from all burning treatments were combined. While fine (1-2 mm diameter) root
carbohydrates were insensitive to lethal heating durations, coarse pine root carbohydrates
54
decreased precipitously with greater burning duration (Table 4-5). Lethal heating
duration at 5 cm depths in the mineral soil significantly reduced coarse root
carbohydrates (P <0.01), explaining 59% of the variation in post-burn changes in coarse
root carbohydrates (Figure 4-4).
Discussion
Fire effects research suffers from a poor understanding of mechanistic linkages
between fire damage and post-fire stress and decline (Ryan 2000, Dickinson and Johnson
2001, 2004). In this study, post-fire stress, as indicated by reductions in coarse root
carbohydrates, was linked to the duration of mineral soil heating caused by overlying
smoldering duff. Additionally, post-burn radial latewood growth rates during the first
year following the fires declined with greater duration of lethal heating in the upper 5 cm
of mineral soil (Table 4-5). These radial growth results suggest that damage from mineral
soil heating decreases radial growth in the year following fire. These data support a
multiple damage (i.e., root plus stem damage) cause, though weak, for post-fire tree
decline. Further, in contrast to the preceding years, the post-fire year (2004) was the
wettest year in the previous decade and terminated the longest sustained drought in
northern Florida recorded history (National Climate Data Center 2005), potentially
concealing or delaying the effects of damage.
Changes in coarse root carbohydrates found in this study support the role of
mineral soil heating as a primary cause of post-fire tree stress. Whereas stem radial
growth in longleaf pine represents an aggregate result of both the current and preceding
year’s conditions (Meldahl et al. 1999), carbohydrate supplies are metrics of current tree
stress (Wargo et al. 1972, Marshall and Waring 1985, Dunn and Lorio 1992, Kozlowski
and Pallardy 1997, Guo et al. 2004). Long duration lethal heating at 5 cm depths in the
55
mineral soil explained 59% of the variation in post-burn losses in coarse root
carbohydrates (Table 4-5), which is strong support for root heating as a cause of
overstory tree stress, and potentially the cause of widespread mortality reported following
restoration fires region-wide (Chapter 2).
As has been reported in several other studies of fire effects on North American
conifers (e.g., Ryan 2000, Stephens and Finney 2002, Agee 2003, McHugh and Kolb
2003, Chapter 2), there was little mortality within the first two years following fires. The
reason for such low mortality may be the short duration of the experimental fires or other
causes (e.g., wet climate during two years following burns). As periodic post-burn
surveys continue at this site, it will be interesting to evaluate the impacts of observed
short-term changes in carbohydrates (120 days post-burn) and radial growth (1 year) on
subsequent radial growth and mortality. To detect “delayed mortality” (Agee 2003,
Chapter 2, 3), plans are to survey the trees burned in this study for several years.
One important shortcoming of this work was the lack of differences among the
burning treatments. None of the fire damage treatments (ROOT, ROOT+STEM, STEM,
or CONTROL) differed in radial stem growth, root carbohydrate drain, or tree mortality
(Table 4-3). Durations of lethal temperatures across all treatments varied (Table 4-2), but
lethal heating durations generally followed the expected pattern: CONTROL durations of
heating > 60° C in all strata (bark, duff, 5-, 10-, and 20 cm depths) were the shortest;
STEM durations were low below-ground; ROOT durations > 60°C were lowest on the
bark; and ROOT+STEM durations were consistently longer across all strata. Future
sampling will determine if treatment differences in growth and mortality materialize, as
observed in other post-fire studies (Busse et al. 2000, Ryan 2000, Stephens and Finney
56
2002, Agee 2003, McHugh and Kolb 2003, Chapter 2). Future research should seek to
segregate heating more vigorously, perhaps with the use of torches (to focus heating) and
more elaborate heat exclusion materials (to exclude heat from non-target tissues).
Fire research is fraught with operational problems; one problem with this
experimental methodology was that the fires surrounding treatment trees were artificially
extinguished prior to their “natural” extinction. All burns were extinguished with water
according to prescription. Even though small amounts of water (less than 750 L of water
were used to extinguish the 80 experimental fires) were used to extinguish experimental
fires, these inputs may have confounded results by alleviating some of the post-fire
moisture stress (Ryan 2000). Nevertheless, the small volumes of water used were
unlikely to have greatly influenced the initial post-fire stress period, especially given the
very wet year following the fires. Failure to consider the role of the post-fire watering in
cooling and abatement of temperature stress in response to soil heating may have been a
bigger oversight. In the future, researchers should allow smoldering to burn to extinction
to avoid these problems.
Among the most striking results of this experiment was the depth and duration of
heating in the lower duff and surface mineral soil (Table 4-2). In the ROOT and ROOT
+STEM treatments, mineral soil temperatures were elevated well above ambient down to
20 cm below the mineral soil surface for long durations. This finding seems important
given that in long-unburned longleaf pine stands targeted for restoration; most fine roots
grow within basal duff and the upper few centimeters of mineral soil (Heyward 1933,
Wahlenberg 1946, Gordon and Varner 2002). Mineral soil heating, the most prominent
predictor of reductions in growth and stored carbohydrates, exceeded 60ºC in the top 5
57
cm in 58% of all burns (75% of treatments designed for root-only heating), as well as at
lower depths: 42% and 25% of all burns had temperatures > 60ºC at depths of 10- and
20-cm below the surface, respectively. Data from an extended overnight burn (4
November 2003; only temperature data prior to 1700 were included in the analyses)
reveal temperatures in the mineral soil can exceed lethal values for >19 hours (these
burns were extinguished at 0900 h), perhaps indicative of how other trees would have
burned if not extinguished and how fuels smolder in many large landscape fires. Duff and
soil heating, even with the artificially truncated durations in this study, suggest
substantial root heating and fire-caused root mortality in smoldering fires. Prolonged
heating kills small pine roots, but also damages or kills higher-order roots (Guo et al.
2004) that connect large numbers of smaller roots to the tree, cascading localized effects
into more substantial whole-tree damage. Given the observed link between the duration
of lethal heating in the soil and carbohydrate drain, it appears that prolonged smoldering
exacerbates post-fire stress and may cause tree death. Root damage may not be the sole
cause of tree decline and mortality, but the links among duration of lethal temperatures,
reduced stem radial growth, and root carbohydrate drain underscore our need to
understand fire damage and the physiological response to fire-caused injury (Ryan 2000,
Johnson and Miyanishi 2001).
The findings of this study provide a mechanism for delayed post-fire mortality
observed in longleaf pine (Chapter 2, 3) and other conifers (Stephens and Finney 2002,
Agee 2003). Coarse root carbohydrate concentrations in this study were reduced
drastically by smoldering-induced heating of mineral soil (Figure 4-4). Fine root
carbohydrate concentrations were unaffected by heating duration. This pattern
58
(carbohydrate drain in coarse roots while fine root carbohydrate pools remain unchanged)
in concert with findings from landscape fires (Chapter 3) where duff consumption was
strongly linked to tree mortality is provocative. These findings support a model of
post-fire mortality where damaged trees use available carbohydrate storage to replenish
killed or damaged fine roots while source carbohydrate reserves are depleted, weakening
the tree. If stressed further, these fire-weakened trees then could succumb to second-order
pests and diseases. There is a strong need for finer scale attempts at reconciling tree death
in response to stresses associated with smoldering fires.
Regardless of mechanism (root or basal damage), one major shortcoming in our
understanding of smoldering fires is linking the fuel (basal duff) to tree damage. Basal
duff is compositionally and structurally complex (Gordon and Varner 2002; Chapter 6),
hence the determinants of its ignition, smoldering duration, and extinguishment are
poorly understood. In this study, the best predictive model explained only 19% of the
variation in duration of duff temperatures > 60°C (Table 4-4). Given the importance of
duff smoldering and its implications for conifer mortality throughout North America,
future work should focus on better characterizing duff fuels, the variation in duff
moisture content, and determinants of duff ignition and combustion. With continued
large-scale fire exclusion in southeastern pine forests and in coniferous forests elsewhere
in North America, problems with duff smoldering are sure to continue.
59
Table 4-1. Fire weather observations and time-of-ignition fuel moistures from 80
experimental single-tree burns at the Katharine Ordway Preserve-Swisher
Memorial Sanctuary, Florida, USA. All experimental fires burned between 25
September and 4 November 2003.
Weather variable Mean ± s.d. Maximum Minimum
Air Temperature (ºC) 27.8 ± 2.37 32.0 25.3
Relative Humidity (%) 58.3 ± 6.19 67.0 51.0
Wind Speed (m sec
-1
) 0.9 ± 0.41 1.3 0.4
Oi moisture (%) 14.9 ± 3.4 19.1 9.4
Oe moisture (%) 64.5 ± 59.9 186.3 11.6
Oa moisture (%) 55.9 ± 23.3 100.4 36.2
Cone moisture (%) 21.0 ± 18.4 47.7 6.5
A horizon moisture (%) 7.1 ± 3.1 12.7 3.8
60
Table 4-2. Durations of lethal heating (i.e., temperatures >60ºC) to basal bark, duff, and mineral soil during individual tree burns at the
Katharine Ordway Preserve-Swisher Memorial Sanctuary in northern Florida, USA.
Treatment
a
Duration of temperatures >60ºC (minutes)
basal bark basal duff 5 cm soil 10 cm soil 20 cm soil
STEM (n=6) 44.3 ± 77.1 42.9 ± 47.3 7.9 ± 14.1 7.3 ± 17.9 --
ROOT (n=6) 10.1 ± 7.0 145.6 ± 283.7 41.0 ± 90.5 4.6 ± 9.9 0.7 ± 0.8
ROOT + STEM (n=6) 82.2 ± 121.8 95.7 ± 187.3 56.1 ± 122.2 12.8 ± 20.9 2.1 ± 4.8
CONTROL (n=6) 9.4 ± 2.6 10.6 ± 24.6 2.7 ± 5.9 1.4 ± 3.5 0.2 ± 0.5
Means 36.5 ± 73.9 73.7 ± 168.8 26.9 ± 74.8 6.5 ± 14.4 0.8 ± 2.5
a
Treatments were intended to isolate long-duration heating to either roots (ROOT), basal stems (STEM), both roots and stem tissues
(ROOT+STEM), or surface fuels burned with no long-term smoldering (CONTROL).
61
Table 4-3. Effects of individual tree experimental burning treatments on smoldering time, mean floor consumption, longleaf pine stem
radial growth, and root nonstructural carbohydrates in northern Florida, USA.
Burning Treatment
a
ROOT ROOT+STEM STEM CONTROL N P
Smoldering 6.9 ± 6.7 A
b
8.9 ± 8.6 A 6.8 ± 6.8 A 1.1 ± 2.2 B 8 0.018
time (minutes)
Forest floor 6.3 ± 1.8 AB 9.1 ± 3.1 A 7.0 ± 3.0 AB 4.6 ± 2.0 B 8 < 0.001
consumption (cm)
Radial growth -17 ± 19 -19 ± 36 -28 ± 33 -22 ± 19 8 0.582
(% change 2003-04)
Fine root 6.3 ± 16.6 12.6 ± 14.3 -9.3 ± 16.5 8.1 ± 26.3 8 0.246
carbohydrates
(% change 2003-04)
Coarse root -5.6 ± 27.1 -12.8 ± 46.8 -3.3 ± 44.1 9.9 ± 16.2 8 0.584
carbohydrates
(% change 2003-04)
a
Treatments were intended to isolate long-duration heating to either roots (ROOT), basal stems (STEM), both roots and stem tissues
(ROOT+STEM), or surface fuels burned with no long-term smoldering (CONTROL).
b
Values followed by a different letter indicate significant differences among treatments, determined using a post-hoc Tukey-Kramer
HSD with α= 0.05 prior to analysis.
62
Table 4-4. Regression results for the analysis of fuel and soil moisture effects on basal bark, duff, and mineral soil temperatures in
smoldering fires in a long-unburned longleaf pine stand in northern Florida, USA.
Dependent variables ß
0
ß
1
R
2
P
log (basal bark temperatures > 60 C; min) 3.86 – 0.019 (% Oa moisture) 0.16 0.053
log (duff temperatures > 60 C; min) 3.61 – 0.018 (% Oe moisture) 0.24 0.015
5-cm mineral soil temperatures > 60 C (min) n/a NS
10-cm mineral soil temperatures > 60 C (min) n/a NS
20-cm mineral soil temperatures > 60 C (min) n/a NS
63
Table 4-5. Effects of basal bark, duff, and mineral soil temperatures on longleaf pine stem radial growth and root nonstructural
carbohydrates following individual tree fires in northern Florida, USA.
Dependent variable
a
ß
0
ß
1
R
2
P
Radial growth (% change 2003-2004) -0.22 0.045 log (5-cm min > 60)
0.16 0.08
Earlywood growth (% change 2003-2004) n/a NS
Latewood growth (% change 2003-2004) 21.7 - 19.4 log (10-cm min > 60) 0.17 0.07
Fine root carbohydrates (% change 2003-2004) n/a NS
Coarse root carbohydrates (% change 2003-2004) 7.1 - 36.7 log (5-cm min > 60) 0.53 0.02
a
Results are based on step-wise linear regressions with minutes of basal heating > 60° C, duff heating > 60°, 5-cm mineral soil > 60°,
10-cm mineral soil > 60°, and 20-cm mineral soil > 60° as potential predictor variables.
Figure 4-1. The Katharine Ordway Preserve-Swisher Memorial Sanctuary in Putnam
County, Florida.
0 15 30 km
65
Figure 4-2. Experimental fire treatments to individual pines in a long-unburned Pinus
palustris forest in northern Florida, USA. The treatments were, clockwise
from top left: ROOT, ROOT+STEM, STEM, and CONTROL. Treatments
were designed to isolate heating to target tissues (i.e., ROOT treatments were
designed to heat only root tissues, while STEM treatments were designed to
isolate heating on stem tissues).
66
Figure 4-3. Locations of thermocouples during experimental fires surrounding the base of
an individual long-unburned Pinus palustris tree. At each measured tree (n=8
per treatment), thermocouples were attached at three locations 120° apart in
each strata: on the basal bark, inserted into lower duff (Oa horizon), and
inserted at depths of 5-, 10-, and 20-cm in the mineral soil. Thermocouples
measured temperatures from 15 minutes prior to ignition until 1700
extinguishment.
67
Figure 4-4. Relationship between the change (2003-2004) in coarse root non-structural
carbohydrates in mature Pinus palustris and duration of heating >60°C at 5
cm below the surface of the mineral soil in experimental individual tree fires
in northern Florida.
-100.0
-80.0
-60.0
-40.0
-20.0
0.0
20.0
40.0
60.0
-0.5 0.2 0.9 1.6 2.3 3.0
log Minutes >60C at 5 cm
Coarse root carbohydrate change (%)
R
2
= 0.53
P = 0.02
68
CHAPTER 5
MODELED VERSUS OBSERVED FIRE EFFECTS IN LONG-UNBURNED PINUS
PALUSTRIS ECOSYSTEMS
Introduction
In most historically fire-maintained temperate coniferous forests, woodlands, and
savannas, frequent surface fires maintained fuel loads at low levels, often with no forest
floor development (i.e., no Oe and Oa horizon or “duff” formation) and little coarse
woody debris retention (Heyward 1939, Agee 2002, Stephens 2004). With widespread
fire exclusion in many temperate coniferous ecosystems, fuels have accumulated, leading
to uncharacteristically high fuel loads (van Wagtendonk 1985, Keane et al. 2002, Chapter
2) that smolder for long periods following ignition, heating both stems and root vascular
tissues. One result of these smoldering fires is a high mortality rate for large trees (Ryan
and Frandsen 1991, Swezy and Agee 1991, Haase and Sackett 1998, Varner et al. 2000,
Stephens and Finney 2002, Agee 2003, McHugh and Kolb 2003, Chapter 2), a
phenomenon uncharacteristic of what are normally fire-resistant conifers.
Researchers and fire managers commonly use simulation models to predict effects
of fire or fuels and silvicultural treatments (Stephens 1998, Brose and Wade 2002, Agee
2003). Because prescribed fires are often expensive, operationally demanding, and
unforgiving, these models have great utility. Among other models (e.g., BehavePlus,
Andrews and Bevins 2003; CONSUME, Ottmar et al. 1993; FARSITE, Finney 1998),
fire effects predictive models (e.g., NEXUS, Scott 1999; FOFEM, Reinhardt 2003; FMA
Plus, Carlton 2005) have great value for planning prescribed fires and managing
69
lightning-ignited fires (so-called “wildland fire use” fires), as well as for determining
post-wildfire management priorities (e.g., salvage of dying timber, installing erosion
control treatments). Fire effects models also offer promise for use in predicting outcomes
of reintroduction of fire into fire-suppressed ecosystems.
The need for reintroduction of fire is mounting in many places in North America
(e.g., van Wagtendonk 1985, Covington et al. 1997), particularly in southeastern Pinus
palustris (longleaf pine) ecosystems (Hermann 1993, Landers et al. 1995). Fire exclusion
in longleaf pine ecosystems leads to decreased plant and animal species richness
(Heyward 1939, Engstrom et al. 1984, Mushinsky 1985, Aresco and Guyer 1999, Gilliam
and Platt 1999, Kush and Meldahl 2000, Provencher et al. 2003, Chapter 2), increased
overstory tree density (Gilliam and Platt 1999, Varner et al. 2000), and heavy
accumulations of surface organic material (hereafter “forest floor”; Chapter 2). Restoring
fire to long-unburned longleaf pine stands has been surprisingly problematic; among the
negative outcomes that include excessive long-duration smoke emissions and weedy
species proliferation, large-scale overstory pine mortality looms large. Examples of this
mortality of large, old longleaf pine trees are abundant across the region (Varner and
Kush 2004, Chapter 2).
The objectives of this study were to compare the predictions of the most
frequently-used (Miller and Landres 2004) fire effects model, the First-Order Fire Effects
Model (FOFEM; Reinhardt 2003) with empirical results from large landscape fires and
small individual tree burns in long-unburned longleaf pine forests. I expected that the
currently available version of this model, developed using data from well-managed
ecosystems maintained by frequent fires, would perform poorly when applied to the often
70
more common areas where fire has been suppressed for decades (Outcalt 2000). FOFEM
5.2.1 was used since it has been used by managers and researchers to test effects of
treatments (e.g., Agee 2003, Reinhardt 2003). Specific hypotheses were:
The model estimates for mortality are poor in formerly fire-excluded stands,
because they focus on damage caused by flaming fires, with little focus on
residual smoldering (Chapter 2).
The modeled duff consumption would underestimate observed consumption
because FOFEM and other models typically have difficulty modeling the spatially
patchy smoldering combustion of duff.
Modeled soil heating underestimates observed soil heating because traditional
models have not included soil heating caused by prolonged smoldering duff
combustion.
These results of these analyses will be useful to restorationists and managers of
fire-excluded longleaf pine ecosystems and potentially help identify patterns that may
hold in other fire-excluded areas. Furthermore, this work should be useful for evaluating
the utility of current predictive models and for elucidating alterations for improving these
important decision-support tools.
Methods
Study Sites
Empirical data for tree mortality and fuel consumption were gathered from
experimental fires at both stand- and individual tree scales in northern Florida. Large
operational prescribed fires were ignited in two long-unburned sites dominated by Pinus
palustris at Eglin Air Force Base in Okaloosa County (N 30° 38’, W 86° 24’; Figure 5-1).
Both stands had undergone ca. 40-45 years of fire exclusion prior to reintroduction of
fire. The stands were dominated by a remnant canopy of longleaf pine (45-200 trees > 10
cm DBH ha
-1
), with scattered sand pine (P. clausa var. immuginata), turkey oak (Quercus
laevis), sand live oak (Q. geminata), sand post oak (Q. margaretta), and a woody
71
midstory dominated by yaupon holly (Ilex vomitoria), littlehip hawthorn (Crataegus
spathulata), and scattered laurel oaks (Q. hemisphaerica) with little herbaceous species
cover (Chapter 3). Soils at both sites are deep, excessively well-drained coated Typic
Quartzipsamments of the Lakeland series with mean depth to water table exceeding 200
cm. Slopes are gentle (0 to 5%) and elevations range between 50 and 60 m (Overing et al.
1995). The climate of the area is subtropical, characterized by warm, humid summers and
mild winters, with mean temperatures of 25° C and mean annual precipitation of 1580
mm, most of which falls from June to September (Overing et al. 1995).
Twelve 10 ha blocks in each of the two sites at Eglin were randomly assigned to
one of three burning treatments based on collected day-of-burn duff moisture content
(DMC; percent of dry weight): dry (60% DMC); moist (90% DMC); and wet (120%
DMC; Chapter 3). For each experimental burn, fire weather (wind speed, air temperature,
and relative humidity), fuel moisture, and fire behavior were recorded (Table 5-1). All
prescribed burns were ignited during the spring, from late February to April. To minimize
among stand variation in fire behavior, all fires were ignited using strip head fires or spot-
grid ignition (Wade and Lunsford 1989), with distance between strips adjusted between
10 and 30 m to minimize variation in flame lengths and rate of spread.
Detailed data on soil heating, which are difficult to collect in large operational fires,
were based on individual tree fires. Given that smoldering duff fires are small, propagate
slowly, and behave in response to small-scale controls (mineral content and bulk density;
Frandsen 1991), these small fires should have emulated the same patterns observed in
large fires. The individual tree study site was near Smith Lake in the Katharine Ordway
Preserve-Carl Swisher Memorial Sanctuary (hereafter, “Ordway”) in Putnam County (N
72
29º 40’, W 81º 74; Figure 5-1). The Smith Lake Tract was last burned 37 years prior to
the study (personal communication, T. Perry). The area is dominated by a canopy of
longleaf pine, turkey oak, sand live oak, laurel oak, and a remnant understory of southern
wiregrass (Aristida stricta) with patches of Florida rosemary (Ceratiola ericoides), all
typical of fire-suppressed north Florida sandhills. Soils of the site are deep, extensively
well-drained hyperthermic, uncoated Lamellic Quartzipsamments of the Candler series
(Readle 1990). The topography is gentle, with gentle north-facing slopes < 5% and
elevations averaging 36 m.
Individual tree fires at Ordway were ignited 1 m away from the base of each of 16
randomly selected mature (30 – 50 cm DBH) pines. Using methods described earlier
(Chapter 4), three treatments were installed that subjected pines to root, stem, or both root
and stem heating. Using collected duff moisture contents, we divided the burns into three
moisture regimes that fit within the treatment fuel moistures at Eglin (wet, moist, and
dry). In all burns, duff and mineral soil temperatures were measured using Type J
thermocouples (range 0º to 1200° C; Omega Laboratories, Stamford, CT) connected to a
Campbell Scientific CR10X datalogger (Campbell Scientific, Logan, UT). Temperatures
were measured at three locations 120º apart in the lower duff (Oa horizon) ca. 10 cm
from the tree base, and directly beneath each of these points in the mineral soil at 5, 10,
and 20 cm depths, for a total of 12 measurement points per tree. Maximum temperatures
were recorded every two minutes at all points from 15 minutes prior to ignition through
the duration of the burning day (termination was required by 1700 hours on all burn
days). This constrained fire length is consistent with typical prescribed fire policies in
Florida and the region (Wade and Lunsford 1989). As a result, however, both duff
73
consumption and soil heating (both temperatures and duration) are underestimates of
potential fire effects.
Simulation Modeling
To model fire effects, we used the First-Order Fire Effects Model (FOFEM 5.2.1;
Reinhardt 2003). FOFEM simulates soil heating depth and duration, fuel consumption
(by timelag category and forest floor horizon), and overstory tree mortality. Fuel and
vegetation data were gathered from the Eglin and Ordway plots and supplemented with
data from Eglin (Ottmar and Vihnanek 2000, Ottmar et al. 2003; Table 5-1). Fuel
moisture inputs were derived from day-of-burn collections of timelag and forest floor
fuels. For fuel model input, SAF Cover Type 70 (longleaf pine; Eyre 1980) with 15 years
since fire (15 years was the maximum period of fire exclusion allowable in FOFEM) was
used, appropriate for the xeric longleaf pine stands at both sites. Flame height inputs into
FOFEM were derived from observations in the prescribed fires at Eglin (ignition patterns
described above; weather in Table 5-1).
Results
Pine mortality in the long-unburned stands at Eglin differed from FOFEM
estimates across all fuel moisture scenarios (Table 5-2). FOFEM predicted 35 %
overstory mortality across all moisture scenarios (Table 5-2). In the large prescribed fires
at each of the three duff moisture levels, there was no overstory pine mortality during the
first year following burns but during the second year, mortality was 0.5, 3.0, and 20.5 %
in the wet, moist, and dry scenarios, respectively (Table 5-2). In FOFEM simulations
across all moisture scenarios, probability of tree mortality decreased with increasing tree
diameter (Table 5-2). In contrast, observed tree mortality varied by treatment; mortality
74
was uniform across all size classes in wet burns, concentrated in smaller trees in moist
burns, and concentrated in larger pines in dry burns (Table 5-2; Chapter 3).
Observed duff consumption and smoldering times were substantially greater than
predicted in FOFEM simulations (Table 5-2). FOFEM predicted no duff consumption in
the wet and moist burns, and only 8.1% in the dry scenario. In contrast, field burns
resulted in some duff consumption across all moisture scenarios; duff consumption
averaged 5.0% in wet burns; 14.5% in moist burns; and 46.5% in dry burns (Table 5-2;
Chapter 3). FOFEM simulations predicted brief durations of smoldering across all
moisture treatments (6.8, 17.0, and 18.3 minutes in the wet, moist, and dry scenarios,
respectively), but observed smoldering durations, even with premature extinguishment,
were greater than FOFEM predicted, with increases of 38, 137, and 883% in the wet,
moist, and dry treatments, respectively (Table 5-2; Chapter 4).
Although FOFEM did not predict any mineral soil heating above lethal
temperatures (>60°C; Table 5-2), individual tree fires resulted in substantial mineral soil
heating (Chapter 4). Across all moisture treatments, mineral soil was heated to depths of
20-cm in mineral soil (Table 5-2). The dry field burns resulted in lethal temperatures
(>60° C) at 5 cm (mean duration = 61.3 min), 10 cm (mean duration = 9.7 min), and 20
cm below the surface (mean duration = 1.7 min). Again, since field burns were
extinguished at 1700 h on the day of ignition, heating duration values were conservative.
Discussion
Among the differences between predicted and observed results of fires in long-
unburned longleaf pine ecosystems, the discrepancy in tree mortality is of great
importance. In this study, observed pine mortality differed from model predictions across
all fuel moisture regimes. Whereas FOFEM predicted equal mortality across all duff
75
moisture content treatments, observed mortality increased as fuel moisture decreased.
This increase in mortality is important, given that prescribed burning prescriptions are
written within narrow fuel moisture parameters (Wade and Lunsford 1989). Because
they use empirical data (based on frequently burnable stands) and predict only first-order
fire effects, FOFEM and other fire effects models (e.g., FMA Plus; Carlton 2003) are
generally poor predictors of mortality of large trees. This problem is not restricted to
longleaf pine ecosystems. For example, in FMA Plus (Fire Program Solutions 2003)
simulations of post-fire Sierra Nevada mixed-conifer mortality, Stephens and Moghaddas
(2005) estimated minimal mortality of large conifers (21% of all trees > 25 cm DBH) but
observed post-fire conifer mortality under similar conditions was much higher (between
64 and 100 %; Haase and Sackett 1998, Stephens and Finney 2002). These errors in
mortality prediction should be addressed in future versions of these models and others
that utilize their equations (e.g., FVS-Fire and Fuels Extension, Reinhart and Crookston
2003).
Overstory pine mortality in this study varied with duff moisture content; there was
a minimal effect of tree diameter on mortality in wet and moist burns, whereas in dry
burns mortality was highest in largest trees (Table 5-2). Other investigators have found
that mortality increases with increasing tree diameter, or follows a U-shaped probability
distribution (McHugh and Kolb 2003). Because first-order fire effects on overstory trees
are based on equations incorporating bark thickness and canopy height as the sole
predictors of mortality, they underestimate the role of basal damage caused by
smoldering duff. Given that prolonged lethal soil heating has been linked to overstory
stress and mortality (Swezy and Agee 1991, Haase and Sackett 1998, Chapter 4), duff
76
consumption and soil heating should be incorporated into these models to more
accurately estimate post-fire mortality in long-unburned forests.
The durations and temperature maxima of soil heating predicted by the FOFEM
simulations were much lower than observed in the field. The failure of FOFEM to
capture this impact is of general concern since long-duration mineral soil heating due to
duff smoldering has been observed in many mesocosm and stand scale experiments
(Frandsen 1987, 1991; Swezy and Agee 1991; Haase and Sackett 1998; Chapter 4). None
of the simulations in this study resulted in lethal heating in mineral soil whereas lethal
heating (minutes > 60° C) was recorded in field burns across all moisture treatments
(Table 5-2). Even in the wet burning treatment, duff smoldering heated surface soil (5
cm) above the lethal threshold for an average of three minutes. These observations of
long-duration lethal heating suggest that there was substantial root and mycorrhizal
damage (Flinn and Wein 1977), potentially the cause of high large tree mortality in
restoration fires. Given that duff fuels are the ultimate source of long duration soil
heating, better models of duff ignition and consumption are needed (Frandsen and Ryan
1985, Hungerford et al. 1995, Frandsen 1997). Improved understanding of duff-soil
heating should improve the predictions of fire effects models and more accurately predict
the outcomes important to restorationists and fire managers.
Duration of duff smoldering increased with decreasing duff moistures in both
simulated and observed burns (Table 5-2), a result consistent with previous duff
smoldering research (Brown et al. 1991). The fact that simulations underestimated
smoldering times is alarming, particularly given the magnitude of differences (duration of
smoldering in the predicted dry scenario = 18.3 minutes; observed = 179.8 minutes;
77
Table 5-2) and the finding that smoldering duration and consumption are strongly related
to tree mortality (Chapters 3 and 4). This non-linear increase in duff smoldering with
decreases in moisture content supports results of heavy overstory mortality in dry
prescribed fires and wildfires (Varner and Kush 2004, Chapter 3).
Given that fuel consumption is a product of the duration of combustion (Brown et
al. 1991), it is not surprising that the differences between modeled and observed duff
consumption followed the same patterns as duration of duff smoldering. FOFEM
simulations predicted no duff consumption (linked to the absence of smoldering) under
wet and moist scenarios, and only minimal (8.1%) consumption in the dry scenario
(Table 5-2). In contrast, duff consumption was observed across all moisture treatments in
prescribed burns. Observed duff consumption was greatest in dry burns (Table 5-2), with
decreases in duff moisture corresponding to large increases in duff consumption. Given
that duff consumption is a major objective of restoration burns in long-unburned
southeastern pine forests (Hiers et al. 2003), the prediction errors reduce the utility of
FOFEM for land managers. Simulation models still suffer from their reliance on solely
empirical data. Future duff consumption models based on fundamental fuel
characteristics (Sandberg et al. 2001) and process-based approaches (Miyanishi and
Johnson 2002) are important links filling this understanding gap.
Although we did not measure smoke emissions from the landscape-scale fires, the
results of FOFEM simulated emissions are worth noting. The production of smoke
emissions is linked to their phase in combustion, and is therefore sensitive to durations of
flaming and smoldering during fires (Sandberg et al. 2002). Smoldering phase
combustion is responsible for dominant fractions of particulate matter (PM
10
and PM
2.5
),
78
CH
4
, and CO. In FOFEM simulations under a dry scenario, 75.7 and 75.8 % of PM
10
and
PM
2.5
emissions were generated during the smoldering phase. If FOFEM calculations
were adjusted to fit the longer smoldering durations observed, smoldering phase
emissions (i.e., PM
10
, PM
2.5
, CH
4
and CO) should increase substantially. In the dry
scenario the duration of smoldering was nearly 10X that of the simulated burn (Table 5-
2), potentially a critical difference given the human health hazards of particulate matter
emissions (Sandberg et al. 2002) and increasing concern over greenhouse gases. Future
attempts at simulating emissions should more accurately account for duration of duff
smoldering and the emissions it generates.
This study indicates the need to validate fire effects models across the variation in
fuel loadings present in contemporary landscapes. The weaknesses of current versions of
FOFEM and others using similar algorithms (e.g., FVS-FFE and FMA Plus) help reveal
gaps in our knowledge of how to manage fire in ecosystems. Incorporation of additional
empirical data across the variation in time since fire and future adaptations of
process-based approaches will improve the utility of simulation models in the restoration
and management of fire-excluded ecosystems.
79
Table 5-1. Model inputs for simulations of fire effects (FOFEM 5.2.1) across three
different moisture regimes in fire excluded longleaf pine stands in northern
Florida, USA. Values were based on replicated operational prescribed burns at
Eglin Air Force Base, Florida used in subsequent comparisons.
Moisture Scenario
Wet Moist Dry
Weather variables
Relative Humidity (%) 34 35 49
Air Temperature (°C) 14 19 26
Wind speed (at 6.1 m; m sec
-1
) 0.89 1.01 0.89
Fire behavior observations
Flame length (m) 1.3 1.3 1.3
Fuel moisture variables
A
Oi (litter; %) 26 15 12
Oe (fermentation; %) 76 39 26
Oa (humus; %) 124 103 62
10-hour (%) 41 21 14
100-hour (%) 57 39 17
1000-hour (%)
B
82 84 57
Fuel loading (kg ha
-1
)
C
1-hour --------------- 0.40 ---------------
10-hour --------------- 0.66 ---------------
100-hour --------------- 0.84 ---------------
1000-hour --------------- 1.15 ---------------
Litter --------------- 3.03 ---------------
Duff --------------- 4.24 ---------------
Live herbaceous --------------- 0.09 ---------------
Shrub --------------- 0.35 ---------------
Foliage --------------- 0.00 ---------------
Branch --------------- 0.00 ---------------
Total loading ( kg ha
-1
) --------------- 10.76 ---------------
Soil Inputs
Soil texture
D
------------- Coarse-Silt ------------
Soil moisture (A horizon; %) 12 8 4
FOFEM season of burn -------------- Spring ---------------
A
Fuel moisture data are from collected field data (R.Ottmar, unpublished data from Eglin Air
Force Base).
B
1000-hour fuel moisture contents are based on averages of collected sound and rotten downed
fuels (R. Ottmar, unpublished data from Eglin Air Force Base).
C
Fuel loading data are a composite of collected field data (1-, 10-, 100-, and 1000-hr; litter and
duff) and data from Ottmar and Vihnanek (2000) and Ottmar and others (2003) for live
herbaceous, shrub, foliage, and branch fuels.
D
Allowable soil textures are limited in FOFEM; coarse-silt approximated local sites better than
alternative soil selections.
80
Table 5-2. Comparison of results from simulated (FOFEM 5.2.1) and actual fires for fire-
excluded longleaf pine forests at Eglin Air Force Base and the Katharine
Ordway Preserve-Swisher Memorial Sanctuary in northern Florida, USA.
Data on soil heating and duration of smoldering were collected during
individual tree fires at the Katharine Ordway Preserve-Swisher Memorial
Sanctuary and tree mortality and duff consumption data were collected in
replicated prescribed fires at Eglin.
Parameter Model predictions Observed effects
Wet Moist Dry Wet Moist Dry
Soil heating (minutes)
Duration > 60°C 5-cm 0 0 0 3.0 11.6 61.3
10-cm 0 0 0 0.8 9.2 9.7
20-cm 0 0 0 0.3 0.3 1.7
Fuel reduction
Duff (% consumed) 0.0 0.0 8.1 5.0 14.5 46.5
Duration of smoldering (minutes) 6.8 17.0 18.3 9.4 40.4 179.8
Pine mortality probability (%)
Overall 35 35 35 0.5 3.0 20.5
10 – 20 cm dbh 69 69 69 0.0 0.0 6.7
20 – 30 cm dbh 47 47 47 0.0 17.9 13.4
30 – 40 cm dbh 30 30 30 3.2 3.3 30.0
40 – 50 cm dbh 18 18 18 0.0 6.8 13.4
50 – 60 cm dbh 12 12 12 3.2 3.3 30.0
81
Figure 5-1. Locations of study sites in northern Florida, USA. Both sites had undergone
37-45 y of fire exclusion prior to prescribed and simulated fires. Large
prescribed fires were conducted at Eglin Air Force Base. Individual tree
centered fires to evaluate mineral soil heating were conducted at the Katharine
Ordway Preserve-Swisher Memorial Sanctuary.
Eglin Air Force Base Katharine Ordway Preserve-
Swisher Memorial Sanctuary
82
CHAPTER 6
LINKING FOREST FLOOR FUELS WITH FIRE BEHAVIOR AND EFFECTS: A
REVIEW
Introduction
Although foresters and ecologists have long recognized the role of fire in temperate
coniferous forests (e.g., Harper 1913, Chapman 1928, Stoddard 1931, Cooper 1960), fire
exclusion is ongoing in many temperate savanna, woodland, and forest ecosystems.
Excluding fire results in drastic alterations in ecosystems including: increased overstory
density (Cooper 1960, Covington and Moore 1994, Gilliam and Platt 1999, Varner et al.
2000, Keane et al. 2002); altered overstory, midstory, and understory species composition
(Heyward 1939, Gilliam and Platt 1999, Kush and Meldahl 2000, Varner et al. 2000,
Keane et al. 2002, Provencher et al. 2003); altered vertebrate habitat (Stoddard 1931,
Leopold et al. 1963, Engstrom et al. 1984, Mushinsky 1985, Aresco and Guyer 1999);
and, changes in the distribution and loading of surface and ground fuels (van
Wagtendonk 1985, Swezy and Agee 1991, Haase and Sackett 1998, Stephens 2004,
Varner et al. 2005). Among these changes resulting from fire exclusion, the development
of deep organic matter accumulations on the surface of the mineral soil (hereafter, “forest
floor”) and the ecological implications of this alteration have received little attention.
In ecosystems maintained by frequent low-intensity fires, recently fallen litter is
typically consumed by solid-phase combustion, reducing inputs into surface organic soil
horizons (Oe and Oa, collectively termed “duff”). As a result of successful fire
suppression over the last century, many fire-prone temperate savannas, woodlands, and
83
forests currently have well-developed duff horizons (Switzer et al. 1979, Swezy and Agee
1991, van Wagtendonk et al. 1998, Varner et al. 2000, Stephens et al. 2004, Hille and
Stephens 2005; Chapter 2). With re-introduction of fire, duff accumulations represent
novel forest fuels that, in contrast to surface flaming combustion in frequently burned
ecosystems, support smoldering combustion (Chapter 2). Smoldering duff consumes the
roots it contains and concentrates long duration heating on basal meristematic tissues and
surficial roots near the mineral soil surface (Flinn and Wein 1977, Ryan and Frandsen
1991, Swezy and Agee 1991, Hungerford et al. 1995, Schimmel and Granstrom 1996,
Haase and Sackett 1998, Dickinson and Johnson 2001). Smoldering duff fires often kill
trees, alter stand structure and species composition, and generate noxious smoke
emissions (Ryan and Frandsen 1991, Swezy and Agee 1991, Hungerford et al. 1995,
Chapter 2).
Smoldering combustion has received relatively little attention in fire research (but
see Frandsen 1987, Frandsen 1991, Frandsen 1997, Miyanishi 2001), even though
smoldering fires are commonly associated with elevated fire severity (e.g., soil damage
and plant mortality; Ryan and Frandsen 1991, Swezy and Agee 1991, Schimmel and
Granstrom 1996). Downed woody debris (100-, 1000-, and 10,000-hour downed woody
material) is often linked to smoldering and localized fire severity (Ottmar and Sandberg
1985, Brown et al. 1991, Ottmar et al. 1993, Costa and Sandberg 2004). If forest floor
accumulations surrounding trees (“basal accumulations”; Sandberg et al. 2001) ignite,
although their temperatures are modest (<500°C), they are a major cause of post-fire
conifer mortality (Swezy and Agee 1991, Ryan and Frandsen 1991, Hungerford 1995,
Haase and Sackett 1998). Deep forest floor accumulations smolder for long periods
84
(Covington and Sackett 1984, Swezy and Agee 1991, Hungerford et al. 1991, Haase and
Sackett 1998, Miyanishi 2001, Chapter 4), raising soil temperatures (Haase and Sackett
1998) and damaging surficial roots (Swezy and Agee 1991) and stem vascular cambia
(Ryan and Frandsen 1991, Ryan 2000).
Here, I summarize the available information on forest floor fuels with emphasis on
their accumulation, ignition, combustion, and role in fire effects on long-unburned
ecosystems. Research interest in duff smoldering is increasing due to the mounting
backlog of areas that have undergone decades of fire exclusion and the increasing
emphasis on restoring these communities (Parsons et al. 1986, Landers et al. 1995,
Covington et al. 1997, Chapter 2). Particular subjects of interest are spatial patterns of
accumulation, physical and chemical composition of duff accumulations, determinants of
ignition and combustion, and moisture retention.
Forest Floor Accumulation Patterns
In contrast to the continuous blankets of duff in many boreal forests (Miyanishi and
Johnson 2002) and peat-dominated forested wetlands (Hungerford et al. 1995), duff in
temperate coniferous forests tends to be patchy, usually accumulated near the base of
source trees (Ryan and Frandsen 1991, Swezy and Agee 1991, Gordon and Varner 2002,
Varner et al. 2005, Hille and Stephens 2005). In two long-unburned Pinus palustris
stands in Florida, for example, mean duff depths in one study reportedly decreased
rapidly away from pines from 20.4 cm near the stem to 16.1 and 11.9 cm at 100 and 200
cm from the stem, respectively (Gordon and Varner 2002; Figure 6-1).
Given the variation in fire effects associated with duff combustion, foresters and
ecologists need a better understanding of the spatial pattern of forest floor accumulations.
Accumulated duff supports localized smoldering near the base of conifers (Ryan and
85
Frandsen 1991, Swezy and Agee 1991), with one or more isolated actively smoldering
areas (ranging from ca. 1 to 100 cm
3
) rather than a unified smoldering front (Hungerford
et al. 1995, Miyanishi 2001, Miyanishi and Johnson 2002; Figure 6-2). Basal smoldering
is facilitated by the increased depth, bulk density, reduced moisture content, and
composition of duff and its fuel particles. Prolonged duff smoldering has been linked to
the presence of thermal cover provided by overlying Oi and Oe horizons (Miyanishi and
Johnson 2002). Thicker surface forest floor insulates the underlying smoldering front in
the Oa, aiding with distillation and volatilization of adjacent fuel particles. Duff bulk
density varies spatially, which should alter burning behavior substantially (Stephens et al.
2004). Above-ground canopy cover also adds variation: duff moisture content is lower
beneath canopies due to their role in the interception of precipitation (Miyanishi and
Johnson 2002, Hille and Stephens 2005) and perhaps, inhibition of dew formation
(Miyanishi and Johnson 2002). The fact that duff moisture is lower near the stems is
interesting given the assumption that this location should receive substantial stemflow.
Another potential cause of localized smoldering at tree bases is that bark slough is
concentrated near stems whereas needle litter dominates the duff elsewhere beneath the
canopy (Gordon and Varner 2002; Figure 6-1).
A better understanding of mechanisms responsible for the observed patterns of
forest floor development in the absence of fire is needed. Duff accumulation near conifer
stems may be linked to the resistance to decomposition of the high phenolic and suberin
content of bark slough (Susott 1982, Rogers et al. 1986, van Wagtendonk et al. 1998).
High C: N ratios in Oi and duff horizons and high remnant phenolics and incorporated
mineral soil in the duff horizons inhibit decomposition (Berg et al. 1982, Lee et al. 1983,
86
Gholz et al. 1985; Table 6-1). Decomposition may also be slowed beneath coniferous
canopies, due to reduced moisture contents (Keane et al. 2002, Miyanishi and Johnson
2002, Hille and Stephens 2005). Given the variation in contemporary fire regimes across
forest types, we should be able to determine at least the correlates, if not the mechanisms
responsible for the amounts and spatial patterns of duff accumulation in coniferous
ecosystems.
Forest Floor Composition
Forest floor fuels contain particles that differ in characteristics that influence
ignition, burning, and extinction of surface and ground fires. Forest floor fuels can be
subdivided into Oi, Oe, and Oa horizons and at finer scales by their composition (e.g.,
chemistry or moisture holding capacity) and structure (e.g., depth, particle size, and bulk
density). Forest floor fuels are often treated as either “litter layers” or “litter and duff
layers;” seldom are the complex strata and components that typify coniferous forest floor
accumulations differentiated.
Litter (Oi) horizons contain recently fallen and only slightly decomposed
necromass (Pritchett 1979). Oi horizons contain particles with large surface area:volume
ratios and large air spaces (low bulk density) resulting in the Oi having the wide
fluctuations in moisture characteristic of 1-hour timelag surface fuels (Byram 1959,
Nelson 2001, Stephens et al. 2004). Oi horizons also contain fine woody debris, cones
and cone fragments, some broadleaf litter, and coarse bark slough. Freshly fallen litter
contains the highest extractive contents of the forest floor (Table 6-1) and consequently
lowers ignition temperatures and increases fire intensity. The low moisture contents, low
bulk density, high volatile content, and high surface area:volume of Oi horizons support
flaming combustion fronts of high intensity but short duration (Fonda et al. 1998, Fonda
87
2001). Considering both the potential fire intensity created by Oi fuels and the diversity
of particles found in Oi, this topic deserves additional study both from mesocosm (e.g.,
Fonda 2001) up to ecosystem scales.
With partial decomposition and an extended period without fire, an Oe
(fermentation) horizon forms underneath the Oi. Oe horizons are absent in frequently
burned coniferous forests; their presence is indicative of a prolonged fire-free interval.
The Oe is typified by decomposed but recognizable plant parts, reduced air space, stable
moisture contents, abundant fungal hyphae, and many fine roots (Pritchett 1979, Harvey
et al. 1994). The tightly-packed and abundant roots and hyphae result in the Oe being the
most fire-resistant of the horizons. Decomposition decreases necromass particle size,
structural integrity, and cellulose-to-lignin ratios (Table 6-1). Oe horizons generally have
higher moisture content than Oi or Oa horizons. The high moisture content and smaller
particle sizes results in smoldering combustion, allowing the Oe horizon to burn
independent of wind direction and at very slow rates (Byram 1959, Hungerford et al.
1995, Nelson 2001, Miyanishi 2001). In smoldering fires, Oe accumulations insulate and
often conceal underlying Oa smoldering. Of all the forest floor horizons, Oe is the most
enigmatic because it resists water loss. Research is needed to understand the linkages
between the fermentation horizon, its imbedded roots and hyphae, and its resistance to
desiccation.
With continued decomposition, an Oa (humus) horizon forms beneath the Oe and
overlying the mineral soil surface in fire-excluded longleaf pine forests. In many
temperate coniferous forests growing on acidic sandy soils, a mor humus or Lentar
ectogranic layer is created (Heyward 1939, Mader 1953, Wilde 1966). As with
88
decomposition to a fermentation horizon, particle sizes in the Oa are small litter
structures not macroscopically recognizable, cellulose-to-lignin ratios are low, and air
space is greatly reduced (van Wagtendonk et al. 1998, Stephens et al. 2004). Mineral ash
content in Oa is high due to both the admixture of mineral soil from the underlying
mineral horizon and as remnant products from decomposition (van Wagtendonk et al.
1998). The Oa horizon consists of fine organic matter and numerous plant roots (Pritchett
1979, Harvey et al. 1994). Oa horizons are dessicated by the underlying mineral soil, so
they are often drier than the overlying Oe (Nelson 2001, Varner, unpublished data). As
with Oe, the Oa behaves as a ground fuel (Byram 1959, Hungerford et al. 1995, Nelson
2001, Miyanishi 2001) and its presence is an indicator of prolonged periods of fire
exclusion. Oe and Oa horizons, collectively termed “duff,” are often treated together in
fire management and research despite the fact that the two horizons differ substantially in
water relations, ignition, and burning behavior. Given the contrast between the two
horizons, the treatment of “duff” as a homogenous stratum may preclude understanding
of fuel moisture, ignition, and combustion dynamics.
The composition of basal forest floor accumulations varies over spatial scales of
meters and less (Figure 6-2). This variation in forest floor composition is likely critical to
fire behavior, with the ignition, combustion, and smoldering of these contrasting parts
varying widely (Fonda 2001, Miyanishi 2001, Fonda and Varner 2005). Despite this
obvious diversity, commercial peat has been used frequently as a standard to test for
effects and determinants of duff ignition, combustion, and effects on soil heating
(Frandsen 1987, Hungerford et al. 1995, Miyanishi and Johnson 2002). The combustion
of forest floor fuels ranges from short-duration needle litter (Fonda et al. 1998, Fonda
89
2001) to long-duration cones (Fonda and Varner 2005) and woody material (Costa and
Sandberg 2004). Substantial work is needed to describe spatial variation in forest floor
composition and the role that composition plays in combustion.
Given that duff accumulations near many conifers contain a large proportion of
bark slough, it is important to understand their chemical and physical characteristics.
Bark of many conifers contains high temperature suberin and lignin compounds that are
difficult to ignite, but burn with great intensity (Susott 1982, Rogers et al. 1986). Basal
accumulations in long-unburned Pinus palustris stands in Florida are dominated by bark
slough and other forests have similar patterns (R. Ottmar, unpublished data). Conifer
cones are also clumped around individual trees and likely serve as localized sources of
both high intensity and long duration combustion (Fonda and Varner 2005). Hille and
Stephens (2005) provide the only published data on spatial variability (i.e., thickness,
moisture content) of forest floor fuels. Insufficient research, therefore, has focused on the
variability in fuel depth, structure, and composition and the outcomes for fire effects.
Increased attention to this relationship would enhance our understanding and prediction
of fire effects in fire-excluded forests.
While the complexity of forest floor fuels is now beginning to be appreciated,
their relative importance in fire behavior and effects has not been rigorously examined.
One need is to determine the relative importance of fuel composition and structure on
forest floor ignition and combustion. This could be accomplished in laboratory
mesocosms with contrasting fuelbeds or perhaps in an ordination framework using more
elaborate pre-fire data in field burning conditions. Developing a process-based model
incorporating fuel heterogeneity may help suggest viable future research directions.
90
Controls of Duff Moisture Content
Moisture content is the primary determinant of duff ignition and combustion
probability (Sandberg 1980; Frandsen 1987, 1991; Hungerford et al. 1995; Miyanishi
2001). Both empirical (Anderson et al. 2003) and process-based (Miyanishi 2001,
Miyanishi and Johnson 2002) models of duff moisture variation exist. More empirical
work is necessary before generalization across fire-managed ecosystems is possible. In
addition to their structural and compositional complexity, duff fuels are located above
and below dynamic adsorptive surfaces (Oi and Oe above; mineral soil below) and
contain active living roots and mycorrhizae (Harvey et al. 1994). Root and mycorrhizal
activity within Oe and Oa duff drain these fuels of moisture but may also all substantial
wetting in response to hydraulic lift of subsurface moisture (Dawson 1993, Horton and
Hart 1998, Espeleta et al. 2004). To what degree the latter is responsible for localized
increases in moisture is yet to be determined, but the influence of active root uptake of
water surely affects forest floor moisture dynamics. There is obviously substantial room
for inquiry into duff moisture dynamics.
Although substantial work has been focused on duff combustion (Sandberg 1980;
Frandsen 1987, 1991, 1997), many questions remain. In restoration burning in Pinus
ponderosa forests in southern Oregon, for example, Agee and collaborators (Swezy and
Agee 1991, Agee 2003) observed abundant duff smoldering beyond threshold moisture
contents. Similar observations have been recorded in large field studies (Chapter 3) and
in small single-tree fires (personal observation, Flamelot Hammock, March 2002).
Mechanisms for wet duff ignition may be linked to the presence of dry duff vectors such
as 10-hr woody timelag fuels and fallen cones common in long-unburned stands. Cone
fuels have very low field moisture contents (contents lower than litter in longleaf pine
91
sites, Varner, unpublished data) and burn with both high intensity (maximum flame
lengths of individual Pinus palustris cones at field moisture contents averaged 87.1 cm)
and long duration (individual cones burned for an average of 52.8 min; Fonda and Varner
2005). Cones and small timelag fuels dry much faster than underlying duff and, as they
burn, preheat and eventually ignite underlying duff. Given that we have a poor grasp on
vector moisture dynamics and the mechanisms of duff ignition, substantial research on
these topics seems warranted.
Conclusions
Forest floor fuels are an often overlooked component of fuel loading, and
accumulations around the bases of trees even more so. Given the proximity of basal
accumulations of duff to heat-sensitive tree tissues, and the potential for long-duration
smoldering fires, understanding the implications of changes in these fuels is critical to
understanding the mechanisms and severity of fire effects. The effects of fire exclusion
obviously have far-reaching effects on forests beyond changing stand structure, species
composition, and habitat availability (Keane et al. 2002, Chapter 2). Incorporation of a
greater understanding of fire exclusion on fuels, fire behavior, and fire effects will
invigorate our ability to restore and manage rapidly changing ecosystems.
92
Table 6-1. Carbon fractions from basal fuel accumulations in a long-unburned longleaf pine forest floor at Smith Lake, Ordway-
Swisher Preserve, Florida, USA.
Mean Percent of Sample
Forest Floor Component N Extractives Hemicellulose Cellulose Lignin Ash
Litter (Oi horizon) 5 28.4 13.3 b
1
31.3 b 26.3 a 0.6 b
Fermentation (Oe) 5 28.2 11.6 b 28.1 b 31.5 b 0.6 b
Humus (Oa) 5 29.6 6.3 a 21.5 a 35.4 b 7.2 a
Intact bark 5 26.0 6.7 28.7 38.0 0.5
Sloughed bark 5 30.9 6.2 28.6 33.9 0.5
Recently fallen cones 5 18.9
2
15.4 34.6 30.7 0.4
Weathered cones 5 22.7 8.3 28.3 38.6 2.0
1
Percentages followed by different letters denote significant differences (p <0.05) among forest floor components determined using
post-hoc Tukey-Kramer HSD test.
2
Italicized percentages denote significant differences (p < 0.05) determined using paired t-tests.
93
93
Figure 6-1. Basal forest floor depths and composition in a long-unburned longleaf pine forest (n=60 trees) in northern Florida, USA.
Accumulations of forest floor are deepest near the stem. Composition of forest floor material changes with distance from
the stem.
Bark
Needle litter
Leaves
Cones
Woody
Roots
200 100 30
Distance from stem (cm)
Forest floor depth (cm) ± s.d.
11.9±2.6 16.1±3.1 20.4±3.9
Figure 6-2. Basal smoldering near a mature longleaf pine at the Katharine Ordway
Preserve-Swisher Memorial Sanctuary in northern Florida, U.S.A. Smoldering
in long-unburned longleaf pine forests is typically patchy, with small,
localized smolder cavities (arrow above) common near large pines.
95
LIST OF REFERENCES
Agee, J.K. 1993. Fire ecology of Pacific Northwest forests. Island Press, Washington,
D.C.
Agee, J.K. 1998. Fire and pine ecosystems. Pages 193-218 in: D. M. Richardson, ed.
Ecology and Biogeography of Pinus. Cambridge Press, New York.
Agee, J.K. 2002. Fire as a coarse filter for snags and logs. Pages 359-368 in: W.L.
Laudenslayer et al., tech. coords. Managing dead wood in western forests. USDA
Forest Service Gen. Tech. Rep. PSW-GTR-181. Pacific Southwest Research
Station, Albany, CA.
Agee, J.K. 2003. Monitoring post-fire mortality in mixed-conifer forests of Crater Lake,
Oregon. Natural Areas Journal 23:114-120.
Aresco, M.J. and C. Guyer. 1999. Burrow abandonment by gopher tortoises in slash pine
plantations of the Conecuh National Forest. Journal of Wildlife Management
63:26-35.
Barton, A.M. 2002. Intense wildfire in southeastern Arizona: transformation of a
Madrean oak-pine forest to oak woodland. Forest Ecology and Management 165:
205-212.
Bauman, T.A., L.G. Eckhardt, R.A. Goyer, and K.D. Klepzig. 2001. Interactions of fire,
insects, and root pathogens in longleaf pine. Pages 95-96 in: J.S. Kush, ed., Proc.
3rd Longleaf Alliance Conference. Oct. 16-19, 2000, Alexandria, LA.
Berg, B., K. Hannus, T. Popoff, and O. Theander. 1982. Changes in organic chemical
components of needle litter during decomposition. Long-term decomposition in a
Scots pine forest. Canadian Journal of Botany 60: 1310-1319.
Bissett, N.J. 1996. Upland restoration challenge: direct seeding of wiregrass and
associated species. The Palmetto Sum:8-11.
Boyer, W.D. 1979. Mortality among seed trees in longleaf pine shelterwood stands.
Southern Journal of Applied Forestry 3:165-167.
Boyer, W.D. 1990. Pinus palustris Mill. (Longleaf pine). Pages 405-412 in Silvics of
North America Vol. 1. Conifers. USDA Forest Service Agriculture Handbook 654,
Washington, D.C.
96
Brockway, D.G. and C.E. Lewis. 1997. Long-term effects of dormant-season prescribed
fire on plant community diversity, structure and productivity in a longleaf pine
wiregrass ecosystem. Forest Ecology and Management 96:167-183.
Brose, P. and D.D. Wade. 2002. Potential fire behavior in pine flatwood forests following
three different fuel reduction techniques. Forest Ecology and Management 163: 71-
84.
Brown, J.K. 1966. Forest floor fuels in red and jack pine stands. USDA Forest Service
North Central Exp. Station Research Note NC-9. St. Paul, MN. 3 p.
Brown, J.K. 1970. Physical fuel properties of ponderosa pine forest floors and cheatgrass.
Research Paper INT-74. Ogden, UT: U.S. Department of Agriculture, Forest
Service Intermountain Forest and Range Experiment Station. 16 p.
Brown, J.K. 1974. Handbook for inventorying downed woody material. Gen. Tech. Rep.
INT-16. Ogden, UT: U.S. Department of Agriculture, Forest Service Intermountain
Forest and Range Experiment Station. 24 p.
Brown, J.K. and C.D. Bevins. 1986. Surface fuel loadings and predicted fire behavior for
vegetation types in the northern Rocky Mountains. Res. Note INT-358. USDA
Forest Service Intermountain Forest Research Station, Ogden, UT. 9 p.
Brown, J.K., E.D. Reinhardt, W.C. Fischer. 1991. Predicting duff and woody fuel
consumption in northern Idaho prescribed fires. Forest Science 37: 1550-1566.
Brown, R.L., E.L. Stone, and V.W. Carlisle. 1990. Soils. In Ecosystems of Florida.
Edited by R.L. Myers and J. Ewel. University of Florida Presses, Gainesville.
Busse, M.D., S.A. Simon, and G.M. Riegel. 2000. Tree-growth and understory responses
to low-severity prescribed burning in thinned Pinus ponderosa forests of central
Oregon. Forest Science 46: 258-268.
Buysse, J. and R. Merckx. 1993. An improved colorimetric method to quantify sugar
content of plant tissue. Journal of Experimental Botany 44: 1627-1629.
Byram, G.M. 1958. Some basic thermal processes controlling the effects of fire on living
vegetation. USDA Forest Service Note SE-114. Southeastern Forest Experiment
Station, New Orleans. 2 p.
Byram, G.M. 1959. Combustion of forest fuels. In: Davis, K.P., ed. Forest Fire: Control
and Use. McGraw Hill, New York.
Carlton. D. 2003. Users’ Guide to Fuels Management Analysts Plus. Fire Program
Solutions, LLC, Estacada, Oregon. 77 p.
97
Carter, D.C, J.J. Hendricks, R.J. Mitchell, and S.D. Pecot. 2004. 2001. Fine root carbon
allocation and fates in longleaf pine forests. Forest Science 50: 177-187.
Chapman, H.H. 1932. Is the longleaf type a climax? Ecology 13(4): 328-34.
Christensen, N.L. 1981. Fire regimes in southeastern ecosystems. Pages 112-136 in H. A.
Mooney, T. M. Bonnicksen, N. L. Christensen, J. E. Lotan, and W. A. Reiners, eds.
Fire regimes and ecosystem properties. USDA Forest Service General Technical
Report WO-26. Forest Service, Washington, D. C.
Chrosciewicz, Z. 1989. Prediction of forest-floor moisture content under diverse jack
pine canopy conditions. Canadian Journal of Forest Research 19: 1483-1487.
Costa, F., and D. Sandberg. 2004. Mathematical model of a smoldering log. Combustion
and Flame 139: 227-238.
Covington, W.W., P.Z. Fulé, M.M. Moore, S.C. Hart, T.E. Kolb, J.N. Mast, S.S. Sackett,
and M.R. Wagner. 1997. Restoration of ecosystem health in southwestern
ponderosa pine forests. Journal of Forestry 95:23-29.
Covington, W.W. and M.M. Moore. 1994. Southwestern ponderosa forest structure:
Changes since Euro-American settlement. Journal of Forestry 92: 39-47.
Covington, W.W., and S.S. Sackett. 1984. The effect of a prescribed burn in
southwestern ponderosa pine on organic matter and nutrients in woody debris and
forest floor. Forest Science 30:183-192.
Cox, A.C., D.R. Gordon, J.L. Slapcinsky, and G.S. Seamon. 2004. Understory restoration
in longleaf pine sandhills. Natural Areas Journal 24:4-14.
Croker, T.C. 1987. Longleaf pine: a history of man and a forest. Forestry Report R8–
FR7. USDA Forest Service, Atlanta, Georgia.
Davidson, J. and J.L. Hayes. 1999. Effects of thinning on development of SPB
infestations in old growth stands. Southern Journal of Applied Forestry 23: 193-
196.
Dawson, T.E. 1993. Hydraulic lift and water use by plants: implications for water
balance, performance and plant-plant interactions. Oecologia 95: 565-574.
Dickinson, M.B. and E.A Johnson. 2001. Fire effects on trees. Pages 477-525 in:
Johnson, E.A. and K. Miyanishi, eds. Forest Fires. Academic Press, New York.
Dickinson, M.B. and E.A. Johnson. 2004. Temperature-dependent rate models of
vascular cambium cell mortality. Canadian Journal of Forest Research 34: 546-559.
Dijkstra, F., J.J. Boon, and J.M. van Mourik. 1998. Analytical pyrolysis of a soil profile
under Scots pine. European Journal of Soil Science 49:295-304.
98
Dixon, W.N., J.A. Corneil, R.C. Wilkinson, and J.L. Foltz. 1984. Using stem char to
predict mortality and insect infestation of fire-damaged slash pines. Southern
Journal of Applied Forestry 8:85-88.
Dunn, J.P. and P.L. Lorio. 1992. Effects of bark girdling on carbohydrate supply and
resistance of loblolly pine to southern pine beetle (Dendroctonus frontalis Zimm.)
attack. Forest Ecology and Management 50:317-330.
Elliott K.J., J.M. Vose, and B.D. Clinton. 2002. Growth of eastern white pine (Pinus
strobus L.) related to forest floor consumption by prescribed fire in the southern
Appalachians. Southern Journal of Applied Forestry 26: 18-25.
Engstrom, R.T., R.L. Crawford, and W.W. Baker. 1984. Breeding bird populations in
relation to changing forest structure following fire exclusion: a 15-year study. The
Wilson Bulletin 96: 437-50.
Espeleta, J.F., J.B. West, and L.A. Donovan. 2004. Species-specific patterns of hydraulic
lift in co-occurring adult trees and grasses in a sandhill community. Oecologia 138:
341-349.
Everett, R.L, Schellhaas, R., Keenum, D., Spurbeck, D., and Ohlson, P. 2000. Fire history
of the ponderosa pine/Douglas-fir forests on the east slope of the Washington
Cascades. Forest Ecology and Management 129: 207-225.
Eyre, F.H., ed. 1980. Forest cover types of the United States and Canada. Society of
American Foresters, Washington, D.C. 148 pp.
Fahnestock, G.R. and R.C. Hare. 1964. Heating of tree trunks in surface fires. Journal of
Forestry 62:799-805.
Feeney, S.R., T.E. Kolb, W.W. Covington, and M.R. Wagner. 1998. Influence of
thinning and burning restoration treatments on presettlement ponderosa pines at the
Gus Pearson Natural Area. Canadian Journal of Forest Research 28: 1295-1306.
Ferguson, S.A. J.E. Ruthford, S.J. McKay, D. Wright, C. Wright, and R. Ottmar. 2002.
Measuring moisture dynamics to predict fire severity in longleaf pine forests.
International Journal of Wildland Fire 11: 267-279.
Finney, M.A. 1998. FARSITE: Fire area simulator- Model development and evaluation.
USDA Forest Service Research Paper RMRS- RP-4, Pacific Northwest Research
Station. 17 p.
Flinn, M.A. and R.W. Wein. 1977. Depth of underground plant organs and theoretical
survival during fire. Canadian Journal of Botany. 55: 2550-2554.
Fonda, R.W. 2001. Burning characteristics of needles from eight pine species. Forest
Science 47:390-396.
99
Fonda, R.W., L.A. Belanger, and L.L. Burley. 1998. Burning characteristics of western
conifer needles. Northwest Science 72:1-9.
Fonda, R.W. and J.M. Varner. 2005. Burning characteristics of cones from eight pine
species. Northwest Science 78: 322-333.
Fowler, J.F. and C. H. Sieg. 2004. Postfire mortality of ponderosa pine and Douglas-fir: a
review of methods to predict tree death. USDA Forest Service Gen. Tech. Rep.
RMRS-132.
Frandsen, W.H. 1987. The influence of moisture and mineral soil on the combustion
limits of smoldering forest duff. Canadian Journal of Forest Research 17: 1540-
1544.
Frandsen, W.H. 1991. Burning rate of smoldering peat. Northwest Science 65: 166-172.
Frandsen, W.H. 1997. Ignition probability of organic soils. Canadian Journal of Forest
Research 27: 1471-1477.
Frandsen, W.H. and K.C. Ryan. 1986. Soil moisture reduces heat flux and soil
temperatures under a burning fuel pile. Canadian Journal of Forest Research 16:
244-248.
Frost, C.C. 1993. Four centuries of changing landscape patterns in the longleaf pine
ecosystem. Tall Timbers Fire Ecology Conference Proceedings 18:17-43. Tall
Timbers Research Station, Tallahassee, FL.
Fulé, P.Z., A.E. Cocke, T.A. Heinlein, and W.W. Covington. 2004. Effects of an intense
prescribed forest fire: is it ecological restoration? Restoration Ecology 12:220-230.
Gholz, H.L., C.S. Perry, W.P.J. Cropper, and L.C. Hendry. 1985. Litterfall,
decomposition, and nitrogen and phosphorus dynamics in a chronosequence of
slash pine (P. elliottii) plantations. Forest Science 31: 463-78.
Gilliam, F.S. and W.J. Platt. 1999. Effects of long-term fire exclusion on tree species
composition and stand structure in an old-growth Pinus palustris (longleaf pine)
forest. Plant Ecology 140:15-26.
Gilliam, F.S., B.M. Yurish, and L.M. Goodwin. 1993. Community composition of an old-
growth longleaf pine forest: relationship to soil texture. Bulletin of the Torrey
Botanical Club 120:287-294.
Gordon, D.R. and J.M. Varner. 2002. Old growth longleaf pine restoration and
management. Final Report to Eglin Air Force Base Jackson Guard, Niceville,
Florida. 33 p. The Nature Conservancy, Altamonte Springs, Florida.
100
Guerin, D.N. 1988. Oak dome establishment and maintenance in a longleaf pine
community in Ocala National Forest, Florida. M. S. thesis. University of Florida,
Gainesville.
Guo, D.L., R.J. Mitchell, and J.J. Hendricks. 2004. Fine root branch orders respond
differentially to carbon source-sink manipulations in a longleaf pine forest.
Oecologia 140:450-457.
Haase, S.M. and S.S. Sackett. 1998. Effects of prescribed fire in giant sequoia-mixed
conifer stands in Sequoia and Kings Canyon National Parks. Tall Timbers Fire
Ecology Conference Proceedings 20: 236-243. Tall Timbers Research Station,
Tallahassee, FL.
Hardin, E.D. and D.L. White. 1989. Rare vascular plant taxa associated with wiregrass
(Aristida stricta) in the southeastern United States. Natural Areas Journal 9:234-
245.
Hare, R.C. 1965. The contribution of bark to fire resistance of southern trees. Journal of
Forestry 63:248-251.
Harper,R.M. 1913. Economic Botany of Alabama. Monograph No. 8. Geological Survey
of Alabama, Montgomery, AL.
Harvey, A.E., J.M. Geist, G.L. McDonald, M.F. Jurgensen, P.H. Cochran, D. Zabowski,
and R.T. Meurisse 1994. Biotic and abiotic processes in eastside ecosystems: the
effects of management on soil properties, processes, and productivity. Gen. Tech.
Rep. PNW-GTR-323. USDA Forest Service, Pacific Northwest Research Station,
Portland, OR.
Herman, F.R. 1954. A guide for marking fire-damaged ponderosa pine in the southwest.
USDA Forest Service Research Note RM-13. U. S. Forest Service Rocky Mountain
Research Station, Fort Collins, CO.
Hermann, S.H., ed. 1993. The longleaf pine ecosystem: ecology, restoration and
management. Proceedings of the 18th Tall Timbers Fire Ecology Conference. Tall
Timbers Research Station, Tallahassee, FL.
Heyward, F. 1933. The root system of longleaf pine on the deep sands of western Florida.
Ecology 14: 136-148.
Heyward, F. 1937. The effect of frequent fires on profile development of longleaf pine
forest soils. Journal of Forestry 35: 23-27.
Heyward, F. 1938. Soil temperatures during forest fires in the longleaf pine region.
Journal of Forestry 36: 478-491.
Heyward, F. 1939. The relation of fire to stand composition of longleaf pine forests.
Ecology 20: 287-304.
101
Heyward, F. and R.M. Barnette. 1936. Field characteristics and partial chemical analyses
of the humus layer of longleaf pine forest soils. Florida Agricultural Experiment
Station Bulletin 302. 27 pp.
Hiers, J.K., S.C. Laine, J.J. Bachant, J.H. Furman, W.W. Greene, and V. Compton 2003.
Simple spatial modeling tool for prioritizing prescribed burning activities at the
landscape scale. Conservation Biology 17:1571-1578.
Hille, M.G. and S.L. Stephens. 2005. Mixed conifer forest duff consumption during
prescribed fires: Tree crown impacts. Forest Science. In press
Hintze, J.E. 2004. NCSS, version 2004. NCSS, Kaysville, UT.
Horton, J.L., and S.C. Hart. 1998. Hydraulic lift: a potentially important ecosystem
process. Trends in Ecology and Evolution 13:232-235.
Hough, W.A. 1982. Phytomass and nutrients in the understory and forest floor of
slash/longleaf pine stands. Forest Science 28: 359-72.
Hungerford, R.D., W.H. Frandsen, and K.C. Ryan. 1995. Ignition and burning
characteristics of organic soils. Tall Timbers Fire Ecology Conference Proceedings
19: 78-91. Tall Timbers Research Station, Tallahassee, FL.
Jenkins, A.M., D.R. Gordon, and M.T. Renda. 2004. Native alternatives for non-native
turfgrasses in central Florida: germination and responses to cultural treatments.
Restoration Ecology 12:190-199.
Johansen, R.W. and D.D. Wade. 1987. Effects of crown scorch on survival and diameter
growth of slash pines. Southern Journal of Applied Forestry 11: 180-4.
Johnson, E.A. and K. Miyanishi. 1995. The need for consideration of fire behavior and
effects in prescribed burning. Restoration Ecology 3: 271-278.
Johnson, E.A. and K. Miyanishi. 2001. Strengthening fire ecology’s roots. Pages 1-6 in:
Johnson, E.A. and K. Miyanishi, eds. Forest Fires: Behavior and Ecological
Effects. Academic Press, New York.
Johnson, R. and D.H. Gjerstad. 1998. Landscape-scale restoration of the longleaf
ecosystem. Restoration and Management Notes 16: 41-45.
Keane, R.E., K.C. Ryan, T.T. Veblen, C.D. Allen, J. Logan, and B. Hawkes. 2002.
Cascading effects of fire exclusion in Rocky Mountain ecosystems: A literature
review. Gen. Tech. Rep. RMRS-GTR-91. Fort Collins, CO: U.S. Department of
Agriculture, Forest Service Rocky Mountain Research Station. 24p.
Kelsey, R.G., and G. Joseph. 2003. Ethanol in ponderosa pine as an indicator of
physiological injury from fire and its relationship to secondary beetles. Canadian
Journal of Forest Research 33: 870-884.
102
Kirkman, L.K., K.L. Coffey, R.J. Mitchell, and E.B. Moser. 2004. Ground cover recovery
patterns and life history traits: implications for restoration obstacles and
opportunities in a species-rich savanna. Journal of Ecology 92:409-421.
Kolb, T.E., K.M. Holmberg, M.R. Wagner, and J.E. Stone. 1998. Regulation of
ponderosa pine foliar physiology and insect resistance mechanisms by basal area
treatments. Tree Physiology 18: 375-381.
Kozlowski, T.T., P.J. Kramer, and S.G. Pallardy. 1991. The Physiological Ecology of
Woody Plants. Academic Press, San Diego. 657 p.
Kozlowski, T.T. and S.G. Pallardy. 1997. Physiology of woody plants. Academic Press,
San Diego, CA.
Kush, J.S. and R.S. Meldahl. 2000. Composition of a virgin longleaf pine stand in south
Alabama. Castanea 65: 56-63.
Kush, J.S., R.S. Meldahl, and C. Avery. 2004. A restoration success: longleaf pine
seedlings established in a fire-suppressed, old-growth stand. Ecological Restoration
22:6-10.
Kush, J.S., R.S. Meldahl, and W.D. Boyer. 2000. Understory plant community response
to season of burn in natural longleaf pine forests. Tall Timbers Fire Ecology
Conference Proceedings 21: 33-39. Tall Timbers Research Station, Tallahassee,
FL.
Landers, J.L., D.H. Van Lear, and W.D. Boyer. 1995. The longleaf pine forests of the
southeast: requiem or renaissance? Journal of Forestry 93:39-44.
Lee, R.S., W.L. Pritchett, and W.H. Smith. 1983. Forest floor characteristics under
longleaf-slash pine on two Spodosols. Forest Ecology and Management 5: 193-205
Leopold, A.S., S.A. Cain, C.M. Cottam, I.N. Gabrielson, and T.L. Kimball. 1963.
Wildlife management in the national parks. Transactions of the North American
Wildlife and Natural Resources Conference 28: 29-44.
Lorio, P.L., Jr., F.M. Stephen, and T.D. Paine. 1995. Environment and ontogeny modify
loblolly pine response to induced acute water deficits and bark beetle attack. Forest
Ecology and Management 73:97-110.
Mader, D.L. 1953. Physical and chemical characteristics of the major types of forest
humus found in the United States and Canada. Soil Science Society Proceedings
155-159.
Manion, P.D. 1981. Tree Disease Concepts. Prentice-Hall, New Jersey. 399 pp.
103
Marshall, J.D. and R.H. Waring. 1985. Predicting fine root production and turnover by
monitoring root starch and soil temperature. Canadian Journal of Forest Research
15:791-800.
Martin, R.E. 1963. A basic approach to fire injury of tree stems. Tall Timbers Fire
Ecology Conference Proceedings 2: 151-162. Tall Timbers Research Station,
Tallahassee, FL.
McHugh, C.W. and T.E. Kolb. 2003. Ponderosa pine mortality following fire in northern
Arizona. International Journal of Wildland Fire 12:1-16.
McHugh, C.W., T.E. Kolb, and J.L. Wilson. 2003. Bark beetle attacks on ponderosa pine
following fire in northern Arizona. Environmental Entomology 32: 510-522.
McMahon, C.K. 1983. Characteristics of forest fuels, fire and emissions. In: Proceedings
of the 76
th
Annual Meeting of the Air Pollution Control Association. Atlanta, GA.
24 p.
McMahon, C.K., D.D.Wade, and S.N. Tsouklas. 1980. Combustion characteristics and
emissions from burning organic soils. In: Proceedings of the 73
th
Annual Meeting
of the Air Pollution Control Association. Montreal, Quebec. 17 p.
McWhite, R.W., J. Furman, C.J. Petrick, and S.M. Seiber. 1999. Integrated natural
resources transitional plan, Eglin Air Force Base, 1998 – 2001. Eglin Air Force
Base, Jackson Guard, Niceville, Florida.
Meldahl, R.S., N. Pederson, J.S. Kush, and J.M. Varner. 1999. Dendrochronological
investigations of climate and competitive effects on longleaf pine growth. Pages
265-285 in: Tree Ring Analysis: Biological, Methodological, and Environmental
Aspects. R. Wimmer and R. E. Vetter, editors. CAB International, Wallingford,
UK.
Menges, E.S. and M.A. Deyrup. 2001. Postfire survival in south Florida slash pine:
Interacting effects of fire intensity, fire season, vegetation, burn size, and bark
beetles. International Journal of Wildland Fire 10: 53-63.
Miller, C. and P. Landres. 2004. Exploring information needs for wildland fire and fuels
management. General Technical Report RMRS-127. USDA Forest Service, Rocky
Mountain Research Station. Fort Collins, CO. 36 p.
Minnich, R.A., M.G. Barbour, J.H. Burk, and R.F. Fernav. 1995. Sixty years of change in
conifer forests of the San Bernardino Mountains: Reconstruction of California
mixed conifer forests prior to fire suppression. Conservation Biology 9: 902-914.
Miyanishi, K. 2001. Duff consumption. Pages 437-475 in: Johnson, E. A. and K.
Miyanishi, eds. Forest Fires. Academic Press, New York.
104
Miyanishi, K. and E.A. Johnson. 2002. Process and patterns of duff consumption in the
mixedwood boreal forest. Canadian Journal of Forest Research 32: 1285-1295.
Mushinsky, H.R. 1985. Fire and the Florida sandhill herpetofaunal community: with
special attention to the responses of Cnemidophorus sexlineatus. Herpetologica 4:
333-342.
Mutch, R.W. 1970. Wildland fires and ecosystems: a hypothesis. Ecology 51: 1046-1051.
Myers, R.L. 1990. Scrub and high pine. In Ecosystems of Florida. Edited by R.L. Myers
and J.J. Ewel. University Presses of Florida, Gainesville.
Nelson, R.M. 2001. Water relations of forest fuels. Pages 79-149 in Johnson, E.A. and K.
Miyanishi, eds. Forest Fires. Academic Press, New York.
Noss, R.F., E.T. LaRoe, and J. M. Scott. 1995. Endangered ecosystems of the United
States: A preliminary assessment of loss and degradation. USDI National
Biological Service, Biological Report 28. 58 p.
Ostrosina, W.J., D. Bannwart, and R.W. Roncadori. 1999. Root-infecting fungi
associated with a decline of longleaf pine in the southeastern United States. Plant
and Soil 27:145-150.
Ostrosina, W.J., N.J. Hess, S.J. Zarnoch, T.J. Perry, and J. P. Jones. 1997. Blue-stain
fungi associated with roots of southern pine trees attacked by the southern pine
beetle, Dendroctonus frontalis. Plant Disease 81:942-945.
Ottmar, R.D, Alvarado, E., and P.F. Hessburg. 1998. Linking recent historical and current
forest vegetation patterns to smoke and crown fire in the Interior Columbia River
Basin. In Proceedings of the 13th Conference on Fire and Forest Meteorology, R.
Weber, ed. Moran, WY: International Association of Wildland Fire.
Ottmar, R.D., M.F. Burns, J.N. Hall, and A.D. Hanson. 1993. CONSUME user’s guide.
Gen. Tech. Rep. PNW-GTR-304. USDA Forest Service, Pacific Northwest
Research Station. Portland, OR. 17 p.
Ottmar, R.D. and D.V. Sandberg. 1985. Calculating moisture content of 1000-hour
timelag fuel in western Washington and western Oregon. Res. Pap. PNW-336.
USDA Forest Service, Pacific Northwest Research Station. Portland, OR. 16 p.
Ottmar, R.D. and R.E. Vihnanek. 2000. Stereo photo series for quantifying natural fuels.
Volume VI: longleaf pine, pocosin, and marshgrass types in the Southeast United
States. PMS 835. Boise, ID: National Wildfire Coordinating Group, National
Interagency Fire Center. 56 p.
105
Ottmar, R.D., R.E. Vihnanek, and J.W. Mathey. 2003. Stereo photo series for quantifying
natural fuels. Volume VIA: sand hill, sand pine scrub, and hardwoods with white
pine types in the Southeast United States with supplemental sites for volume VI.
PMS 838. Boise, ID: National Wildfire Coordinating Group, National Interagency
Fire Center. 78 p.
Outcalt, K.W. 2000. Occurrence of fire in longleaf pine stands in the southeast United
States. Tall Timbers Fire Ecology Proceedings 21: 178-182. Tall Timbers
Research Station, Tallahassee, FL.
Overing, J.D., H.H. Weeks, J.P. Wilson, J. Sullivan, and R.D. Ford. 1995. Soil Survey of
Okaloosa County, Florida. USDA Natural Resource Conservation Service,
Washington, D.C.
Palik, B. J., and N. Pederson. 1996. Overstory mortality and canopy disturbances in
longleaf pine ecosystems. Canadian Journal of Forest Research 26:2035-2047.
Parsons, D.J., D.M. Graber, J.K. Agee, and J.W. van Wagtendonk. 1986. Natural fire
management in national parks. Environmental Management 10: 21-4.
Peet, R.K. and D.J. Allard. 1993. Longleaf pine vegetation of the southern Atlantic and
eastern Gulf Coast regions: a preliminary classification. Tall Timbers Fire Ecology
Conference Proceedings 18:45-82. Tall Timbers Research Station, Tallahassee, FL.
Peterson, D.L. and M.J. Arbaugh. 1986. Postfire survival in Douglas-fir and lodgepole
pine: comparing effects of crown and bole damage. Canadian Journal of Forest
Research 16: 1175-1179.
Peterson, D.L. and K.C. Ryan. 1986. Modeling post-fire conifer mortality for long-range
planning. Environmental Management 10: 797-806.
Philpot, C.W. 1970. Influence of mineral content on the pyrolysis of plant materials.
Forest Science 16: 461-471.
Platt, W.J., G.W. Evans, and S.L. Rathbun. 1988. The population dynamics of a long-
lived conifer (Pinus palustris). American Naturalist 131:491-525.
Pritchett, W.L. 1979. Properties and Management of Forest Soils. Wiley, New York. 500
p.
Provencher, L., B.J. Herring, D.R. Gordon, H.L. Rodgers, K.E.M. Galley, G.W. Tanner,
J.L. Hardesty, and L.A. Brennan. 2001a. Effects of hardwood reduction techniques
on longleaf pine sandhill vegetation in northwest Florida. Restoration Ecology
9:13-27.
106
Provencher, L., B.J. Herring, D.R. Gordon, H.L. Rodgers, G.W. Tanner, J.L. Hardesty,
L.A. Brennan, and A.R. Litt. 2000. Longleaf pine and oak responses to hardwood
reduction techniques in fire-suppressed sandhills in northwest Florida. Forest
Ecology and Management 148: 1-15.
Provencher, L., B.J. Herring, D.R. Gordon, H.L. Rodgers, G.W. Tanner, J.L. Hardesty,
L.A. Brennan, and A.R. Litt. 2001b. Longleaf pine and oak responses to hardwood
reduction techniques in fire-suppressed sandhills in northwest Florida. Forest
Ecology and Management 148:1-15.
Provencher, L., A.R. Litt, and D.R. Gordon. 2003. Predictors of species richness in
northwest Florida longleaf pine sandhills. Conservation Biology 17: 1660-1671.
Raffa, K.F. and A.A. Berryman. 1983. The role of host plant resistance in the
colonization behavior and ecology of bark beetles (Coleoptera: Scolytidae).
Ecological Monographs 53: 27-49.
Readle, E.L. 1990. Soil survey of Putnam County Area, Florida. USDA Soil
Conservation Service and Florida Agricultural Experiment Station, Gainesville, FL.
224 pp.
Reifsnyder, W.E., L.P. Harrington, and K.W. Spalt. 1967. Thermophysical properties of
bark of shortleaf, longleaf, and red pine. Yale University Forestry Bulletin No. 70.
41 p.
Reinhardt, E.D. 2003. Using FOFEM 5.0 to estimate tree mortality, fuel consumption,
smoke production, and soil heating from wildland fire. Sec. PS-2 in Proceedings of
Fifth Fire and Forest Meteorology Conference, Orlando, FL.
Reinhardt, E.D. and N.L. Crookston. 2003. The fire and fuels extension to the Forest
Vegetation Simulator (FVS). USDA Forest Service RMRS-GTR-116. USDA
Forest Service Rocky Mountain Research Station, Ogden, UT.
Robbins, L.E. and R.L. Myers.1992. Seasonal effects of prescribed burning in Florida: A
review. Misc. Pub. No. 8, Tall Timbers Research Station, Tallahassee, FL. 96 pp.
Rogers, J.M., R.A. Susott, and R.G. Kelsey. 1986. Chemical composition of forest fuels
affecting their thermal behavior. Canadian Journal of Forest Research 16: 721-726.
Ryan, K.C. 2000. Effects of fire injury on water relations of ponderosa pine. Tall Timbers
Fire Ecology Conference Proceedings 21: 58-66. Tall Timbers Research Station,
Tallahassee, FL.
Ryan, K.C. and W.H. Frandsen. 1991. Basal injury from smoldering fires in mature Pinus
ponderosa Laws. International Journal of Wildland Fire 1:107-118.
107
Ryan, K.C. and N.V. Noste. 1985. Evaluating prescribed fires. Pages 230-238 in: Lotan,
J.E., B.M. Kilgore, W.C. Fischer, and R.W. Mutch. Proceedings of a symposium
and workshop on wilderness fire. General Technical Report INT-182. USDA
Forest Service Intermountain Research Station, Ogden, UT.
Ryan, K.C., D.L. Peterson, and E.D. Rheinhardt. 1988. Modeling long-term fire-caused
mortality of Douglas-fir. Forest Science 34: 190-199.
Ryan, K.C. and E.D. Rheinhardt. 1988. Predicting postfire mortality of seven western
conifers. Canadian Journal of Forest Research 18: 1291-1297.
Sandberg, D.V. 1980. Duff reduction by prescribed underburning in Douglas-fir. USDA
Forest Service Research Paper PNW-RP-272. 18 p.
Sandberg, D.V., R.D. Ottmar, and G.H. Cushon. 2001. Characterizing fuels in the 21st
century. International Journal of Wildland Fire 10: 381-387.
Sandberg, D.V., R.D. Ottmar, J.L. Peterson, and J. Core. 2002. Wildland fire in
ecosystems: Effects of fire on air. USDA Forest Service General Technical Report
RMRS-GTR-42-vol. 5. Ogden, UT. 79 p.
Saveland, J.M. and L.F. Neuenschwander. 1990. A signal detection framework to
evaluate models of tree mortality following fire damage. Forest Science 36:66-76.
Schimmel, J. and A.Granstrom. 1996. Fire severity and vegetation response in the boreal
Swedish forest. Ecology 77: 1436-1450.
Scott, J.H. 1999. NEXUS: A system for assessing crown fire hazard. Fire Management
Notes 59:20-24.
South, D. B., and E.R. Buckner. 2003. The decline of southern yellow pine timberland.
Journal of Forestry 101:30-35.
Spalt, K.W. and W.E. Reifsnyder. 1962. Bark characteristics and fire resistance: A
literature survey. USDA Forest Service Occasional Paper 193 and Yale School of
Forestry. Washington, D.C. 19 p.
Stephens, S.L. 1998. Evaluation of the effects of silvicultural and fuels treatments on
potential fire behaviour in Sierra Nevada mixed-conifer forests. Forest Ecology and
Management 105: 21-35.
Stephens, S.L. 2004. Fuel loads, snag abundance, and snag recruitment in an unmanaged
Jeffrey pine-mixed conifer forest in Northwestern Mexico. Forest Ecology and
Management 199: 103-113.
Stephens, S.L. and M.A. Finney. 2002. Prescribed fire mortality of Sierra Nevada mixed
conifer tree species: effects of crown damage and forest floor combustion. Forest
Ecology and Management 162: 261-271.
108
Stephens, S.L., M.A. Finney, and H. Schantz. 2004. Bulk density and fuel loads of
ponderosa pine and white fir forest floors: impacts of leaf morphology. Northwest
Science 78: 93-100.
Stephens, S.L. and J.J Moghaddas. 2005. Silvicultural and reserve impacts on potential
fire behavior and forest conservation: 25 years of experience from Sierra Nevada
mixed conifer forests. Biological Conservation. In Press.
Stoddard,H.L., Sr. 1931. The Bobwhite Quail: Its Habits, Preservation and Increase.
Charles Scribner's Sons, New York, NY.
Stone, J.E., T.E. Kolb, and W.W. Covington. 1999. Effects of thinning on presettlement
Pinus ponderosa in northern Arizona. Restoration Ecology 7: 172-182.
Stott, P. 2000. Combustion in tropical biomass fires: a critical review. Progress in
Physical Geography 24: 355-377.
Susott, R.A. 1982. Characterization of the thermal properties of forest fuels by
combustible gas analysis. Forest Science 28: 404-420.
Swezy, D.M. and J.K. Agee. 1991. Prescribed-fire effects on fine-root and tree mortality
in old-growth ponderosa pine. Canadian Journal of Forest Research 21: 626-634.
Switzer, G.L., M.G. Shelton, and L.E. Nelson. 1979. Successional development of the
forest floor and soil surface on upland sites of the east Gulf Coastal Plain. Ecology
60: 1162-1171.
Taylor, K.L. and R.W. Fonda. 1990. Woody fuel structure and fire in subalpine fir
forests, Olympic National Park, Washington. Canadian Journal of Forest Research
20: 193-199.
Tinker, P.B. and P.H. Nye. 2000. Solute Movement in the Rhizosphere. Oxford
University Press, New York. 444 pp.
van Mantgem, P. and M. Schwartz. 2004. An experimental demonstration of stem
damage as a predictor of fire-caused mortality for ponderosa pine. Canadian
Journal of Forest Research 34: 1343-1347.
van Mantgem, P., M. Schwartz and M. Keifer. 2001. Monitoring fire effects for managed
burns and wildfires: coming to terms with pseudoreplication. Natural Areas Journal
21: 266-273.
van Wagner, C.E. 1973. Height of crown scorch in forest fires. Canadian Journal of
Forest Research 3: 373-378.
109
van Wagtendonk, J.W. 1985. Fire suppression effects on fuels and succession in short-
fire-interval wilderness ecosystems. Pages 119-126 in: Lotan, J.E., B.M. Kilgore,
W.C. Fischer, and R.W. Mutch, eds.. Proceedings of a symposium and workshop
on wilderness fire. Gen. Tech. Rep. INT-182. USDA Forest Service Intermountain
Research Station, Ogden, UT.
van Wagtendonk, J.W., W.M. Sydoriak, and J.M. Benedict. 1998. Heat content variation
of Sierra Nevada conifers. International Journal of Wildland Fire 8: 147-158.
Varner, J.M., D.R. Gordon, F.E. Putz, and J.K. Hiers. 2005. Novel fire effects in
southeastern pine forests: smoldering fire and overstory pine mortality. Restoration
Ecology 13:536-544.
Varner, J.M. and J.S. Kush. 2004. Old-growth longleaf pine forests and savannas of the
southeastern USA: Status and trends. Natural Areas Journal 24:141-149.
Varner, J.M., J.S. Kush, and R.S. Meldahl. 2000. Ecological restoration of an old-growth
longleaf pine stand utilizing prescribed fire. Tall Timbers Fire Ecology Conference
Proceedings 21: 216-219. Tall Timbers Research Station, Tallahassee, FL.
Varner, J.M., J.S. Kush, and R.S. Meldahl. 2003a. Structure of old-growth longleaf pine
(Pinus palustris Mill.) forests in the mountains of Alabama. Castanea 68: 211-221.
Varner, J.M., J.S. Kush, and R.S. Meldahl. 2003b. Vegetation of frequently burned old-
growth longleaf pine ( Pinus palustris Mill.) savannas on Choccolocco Mountain,
Alabama, USA . Natural Areas Journal 23: 43-52
Vines, R.G. 1968. Heat transfer through bark, and the resistance of trees to fire.
Australian Journal of Botany 16:499-514.
Wade, D.D., G. Custer, J. Thorsen, P. Kaskey, J. Kush, B. Twomey, and D. Voltolina.
1998. Reintroduction of fire into fire-dependent ecosystems: Some southern
examples. Tall Timbers Fire Ecology Conference Proceedings 20: 94-98. Tall
Timbers Research Station, Tallahassee, FL.
Wade, D.D. and R.W. Johansen. 1986. Effects of fire on southern pine: observations and
recommendations. General Technical Report SE-41. USDA Forest Service,
Southeastern Forest Experiment Station, Asheville, NC. 14 p.
Wade, D.D. and R.W. Johansen. 1987. Relating wildland fire to defoliation and mortality
in pine. Pages 107-110 in Phillips, D.R., ed. GTR SE-42. USDA Forest Service,
Southeastern Forest Experiment Station, Asheville, NC.
Wade, D.D., and J.D. Lunsford. 1989. A guide for prescribed fire in southern forests.
USDA Forest Service Technical Publication SRS-11, Asheville, NC.
110
Wahlenberg, W.G. 1946. Longleaf Pine: Its Use, Ecology, Regeneration, Protection, and
Management. Charles Lathrop Pack Forestry Foundation and USDA Forest
Service. Washington, D.C. 429 p.
Wallin, K.F., T.E. Kolb, K.R. Skov, and M.R. Wagner. 2003. Effects of crown scorch on
ponderosa pine resistance to bark beetles in northern Arizona. Environmental
Entomology 32:652-661.
Wallin, K.F., T.E. Kolb, K.R. Skov, and M R. Wagner. 2004. Seven-year results of
thinning and burning restoration treatments on old ponderosa pines at the Gus
Pearson Natural Area. Restoration Ecology 12:239-247.
Ware, S., C.C. Frost, and P.D. Doerr. 1993. Southern mixed hardwood forest: The former
longleaf pine forest. Pp. 447-493 in: Martin, W. H., S. G. Boyce, A. C.
Echternacht, eds. Biodiversity of the Southeastern United States: Lowland
Terrestrial Communities. Wiley, New York.
Wargo, P.M., J. Parker, and D.R. Houston. 1972. Starch content in roots of defoliated
sugar maple. Forest Science 18: 203-204.
Waters, M., R.E. Will, R.J. Mitchell, and J.K. Hiers. 2004. Effects of light and litter
accumulation on understory development in longleaf pine sandhill ecosystems.
89th Annual Meeting of the Ecological Society of America, 1-6 August 2004,
Portland, Oregon. Ecological Society of America, Washington, D.C.
Wilde, S.A. 1966. A new systematic terminology of forest humus layers. Soil Science
101:403-407.
Williamson, G.B. and E.M. Black. 1981. High temperature of forest fires under pines as a
selective advantage over oaks. Nature 293: 643-644.
Wilson, C.A, R.J. Mitchell, L.R. Boring, and J.J. Hendricks. 2002. Soil nitrogen
dynamics in a fire-maintained forest ecosystem: results over a 3-year burn interval.
Soil Biology and Biochemistry 34: 679-698.
Wright, C.S. and J.K. Agee. 2004. Fire and vegetation history in the eastern Cascade
Mountains, Washington. Ecological Applications 14: 443-459.
Wyant, J.G., P.N. Omi, and R.D. Laven. 1986. Fire-induced tree mortality in a Colorado
ponderosa pine / Douglas-fir stand. Forest Science 32: 49-59.
111
BIOGRAPHICAL SKETCH
J. Morgan Varner, III was born in Columbus, Georgia on February 3, 1971. A
seventh-generation Alabamian, he grew up in several towns throughout Alabama. Before
receiving his Bachelor of Science degree in forest management from the University of
Idaho in 1997, he worked as a research assistant at the J.W. Jones Ecological Research
Center (Newton, GA). In the landscape surrounding the Jones Center, he became inspired
to understand more about open-canopied pine forests and their conservation. His master’s
work at Auburn University focused on the composition, structure, and conservation of
old-growth longleaf pine communities in northern Alabama. During this work and in
similar projects in lower Alabama, he was confronted with the effects of fire exclusion
and its implications for southern pine forests, the subject of his dissertation research. He
entered the University of Florida’s doctoral program in 2000. In August 2004 he accepted
a one-year lectureship at Humboldt State University (Arcata, CA). In July 2005, he was
named Assistant Professor of Wildland Fire Management in the Department of Forestry
and Watershed Management. He is married to Rachel and has three daughters, Zoe,
Maya, and Clara.
... The duff layer also plays an important role in fire effects, especially in temperate coniferous forests where duff tends to have patchy distributions and accumulates near the base of trees (Hille and Den Ouden 2005;Hood 2010;Varner et al. 2005;Varner 2005). The duff layer is characterized by high bulk density, abundant fungus, many fine roots and high moisture contents. ...
... The duff layer is characterized by high bulk density, abundant fungus, many fine roots and high moisture contents. Due to this, it burns relatively independent of wind direction and at very slow rates (smoldering) (Byram 1959;Miyanishi 2001;Nelson 2001;Varner 2005). That is commonly associated with elevated fire severity and undesired fire effects including soil damage, plant mortality, alteration of stand structure and species composition and noxious smoke emissions beside others (Hungerford et al. 1995;Ryan and Frandsen 1991;Varner 2005). ...
... Due to this, it burns relatively independent of wind direction and at very slow rates (smoldering) (Byram 1959;Miyanishi 2001;Nelson 2001;Varner 2005). That is commonly associated with elevated fire severity and undesired fire effects including soil damage, plant mortality, alteration of stand structure and species composition and noxious smoke emissions beside others (Hungerford et al. 1995;Ryan and Frandsen 1991;Varner 2005). Therefore, to better understand fire effects, it is necessary to know fuel load characteristics and availability because they will affect ignition and combustion, resulting in a mosaic of consumption and effects distributed across the burn unit. ...
Research
Full-text available
Land managers involved in longleaf restoration are increasingly using fire as management tool to restore and maintain this highly diverse and endangered fire dependent ecosystem. However, to achieve restoration goals it is necessary to better understand the relationships between stand conditions, fuel loadings, fire behavior, fire effects and ecosystem responses. This information would help to plan accordingly to reach short and long term management goals and avoid undesired fire effects (longleaf mortality). In this case study, detailed analysis of pre, during and post burn conditions were conducted in a mature longleaf-wiregrass stand in the Calloway Forest Preserve (NC). Before burning a fuel loading inventory was conducted using two different fuel sampling methods, planar intersect (PI) and biomass collection destructive sampling. Estimates from both methods were compared for five different fuel types. Fire conditions and fire behavior were recorded and related with fire intensities and fire effects. Fuel consumption turned out to be higher than expected in the burn plan and an equation to predict forest floor depth removal was developed for future prediction. Fire intensity was found highly variable and hotter than expected. Fire severity effects analyzed in the overstory pointed to no direct association between percentage of crown scorched and tree mortality. Longleaf seedling mortality was found to be related with fire temperature and height; especially in seedlings between 0.3 and 1.5 m height and a mortality predictive model based on these variables was developed. Finally, a review of burn objectives and accomplishments was conducted and management recommendations based on fire effects were developed.
... In longleaf pine stands, the distribution of fuels is related to the distribution and density of pine trees (Platt et al. 1988b;Robertson 2005;Varner 2005;Mitchell et al. 2009), tending to increase concomitantly with stand basal area and proximity to mature trees. Fine fuels and coarse fuels from pine trees differ in their burning characteristics and may therefore differ in their ecological effects. ...
... Although pine cones represent a potentially significant source of fuel ), the ecological effects of pine cones as a fuel source remain poorly understood (Fonda and Varner 2005). Because distribution of cones is influenced by the distribution and density of pine trees (Platt et al. 1988;Varner 2005;Mitchell et al. 2009), the impact of longleaf pine cones on fire effects is likely greater following mast seeding years when pine cones are particularly abundant. ...
Thesis
Full-text available
Fire is a globally distributed disturbance and is important for structuring and maintaining the diversity of many frequently burned ecosystems. Competitive relationships and vegetation patterns may be affected by fire‘s influence on survival, reproduction, and recruitment of plants. Each fire is unique and the effects of a given fire are not necessarily uniform throughout the area burned. Within the southeastern United States, the once-vast longleaf pine (Pinus palustris Mill.) ecosystem has been reduced to a mere fraction of its former range by overharvesting and land conversion. This fire-dependent system is also home to one of the most diverse plant assemblages known in North America. The effectiveness of restoration and management of this imperiled system may be improved by increased knowledge of fire-related germination and recruitment responses for functionally important groups of plant species. This project first investigated potential mechanisms of fire-induced germination and then sought to understand how fine-scale fire heterogeneity may influence germination of common native legume species. For the first part of the project, six legume species were used to examine the effects of exposure to various components of fire: 1) elevated temperatures (five levels with dry heat temperatures ranging from presumably non-lethal to lethal) over three durations; 2) duration (three intervals) of moist heat (i.e., steam); 3) mechanical scarification; and 4) smoke. Legume species used represented a range of seed hardness, varying from soft-seeded to extremely hard-seeded. Results indicate that response to both dry and moist heat was species-specific, but moist heat had the greatest potential to stimulate germination. The legumes studied also exhibited high heat tolerance, which may be a function of their low seed moisture contents. Smoke had no detectable effect on germination of most species, but may have inhibited Centrosema virginianum. Two studies were used for the second part of the project to examine post-fire germination of nine legume species in response to fuel variation in laboratory and field settings. For both studies, fine fuel loads (longleaf pine needles) and woody fuels (longleaf pine cones) were manipulated. Germination and mortality were assessed for six species after experimental laboratory fires. Germination of seven species was assessed for four months in the field following experimental burns of small plots. In both studies, exposure to fire reduced seed germination compared to unburned controls. Furthermore, germination of seeds that were under burned pine cones was reduced more than treatments burned with pine needles alone. Manipulation of biomass of fine fuel loads did not affect germination in either study. Together, the results of all experiments indicate the potential influence of fire on germination and mortality of legume seeds. Such impacts may, in turn, affect recruitment of these and similar species, thereby influencing post-fire vegetation patterns. Additionally, a hypothetical model is presented that describes the potential influence of pine cones on post-fire seedling recruitment.
... The forest floor is generally divided into three layers depending on the degree of decomposition, which differ from one another by their chemical and physical properties (Miyanishi 2001, Varner 2005, Banwell & Varner 2014, Berg & McClaugherty 2014. The litter layer (L) or Oi is the uppermost layer and consists of unaltered, recently deposited organic matter such as leaves, needles, twigs, moss and lichens. ...
Article
Estimation of forest floor loading is important for many forest management applications , especially those related to fire management and carbon balance. We quantified the physical properties (depth, fuel load, bulk density) and carbon stocks of endemic Dalmatian black pine (Pinus nigra J.F. Arnold subsp. dalmatica [Vis.] Franco) forest floor layers. We also examined how these properties differ with stand age and layer. Forest floor depths ranged from 1.5 cm to 11.5 cm and forest floor fuel (FFF) loads ranged from 11.9 Mg ha-1 in the young stand to 197.3 Mg ha-1 in the old stand. Forest floor carbon (FFC) stocks ranged from 6.4 Mg C ha-1 in the young stand to 85.8 Mg C ha-1 in the old stand. We developed regression equations that can be used to convert the investigated forest floor depth into load in each layer individually and across all layers. These equations, together with the organic carbon (OC) concentration determined here for individual forest floor layers, simplify quantification of carbon stocks in the forest floor. Bulk density (BD) values reported here can also be used to convert depth measurements to loads for each layer and the entire forest floor. The results presented here are suitable for rapid estimation of FFF loads and FFC stocks based solely on forest floor depth, without the need to sample and analyze large amounts of forest floor fuels. Similarly, spatial distribution in FFF loads and carbon stocks can be assessed simply by measuring forest floor depths.
... A primary assumption of this modelling framework is the treatment of the peatland surface as a matrix of fuels and air, in contrast to the packed fuel beds of Albini (1985). The presence of partially buried surface fuels such as downed log adding roughness to the surface as well as acting as a potential vector for deep smouldering (Varner 2005) is not considered here, but would complement this study, given the assumption used here of continuous moss-derived peat at and near the surface. ...
Article
Full-text available
The Peatland Smouldering and Ignition (PSI) model was developed to quantify the heat transfer from a wildfire to an organic soil or moss surface in a Sphagnum-black spruce peatland. The Canadian Fire Behaviour Prediction system was used as a basis for the relationship between wind speed and rate of spread. Convection, conduction, and radiation processes were modelled before and during the arrival of the flaming front. The net heat flux to the soil from fire varied between 1.1 and 8.6MJm-2, with moderate-intensity fires transferring more energy to the surface compared with higher-intensity fires under higher winds. Radiative heat transfer to the soil surface both before the fire's arrival and within the flaming front were the primary mechanisms of energy gain to the peatland surface. The role of convective and conductive cooling was no greater than 30% of gross energy gain. Peatland surface ignition in hummock and hollow microforms was modelled under normal and drought conditions. Hollow microforms dried out significantly during the course of a summer and increased their ignition vulnerability. Small-scale changes in slope and aspect of the peatland surface increased the amount of heat transferred by radiation by up to 30%, allowing some areas of higher soil moisture content to ignite. While no direct model validation is available, model outputs showing the preferential combustion of lichen and feathermoss, and the lack of ignition in Sphagnum in all but the most severe drought generally mimic observed ignitions patterns. The modelled peak of net energy input to the surface occurred at moderate wind speeds, suggesting that high-intensity fires do not necessarily lead to greater energy transfer and risk of smouldering combustion.
... This form of combustion typically takes place at much lower temperatures than flaming combustion (500 °C to 700 °C versus 1500 °C to 1800 °C; Rein et al. 2008). In addition to occurring to some extent in woody fuels, smoldering is chiefly the type of combustion found in duff (e.g., Varner 2005), peat, and muck, and characterizes fires found in ecosystems in which these soils or fuel types dominate during dry conditions. In desiccated mucks and peats, smoldering can occur under surprisingly high soil-moisture contents, depending less on parent material than on mineral content (Frandsen 1997, Benscoter 2011, Watts 2013. ...
Article
Full-text available
abStraCt Although fires in wetlands would seem to be rare or impossible by definition, these eco-systems do occasionally experience fire. A common feature of fires in wetlands is smol-dering combustion in organic soils, such as peat and muck. Increasing occurrence and size of these events from the Arctic to the tropics has been matched by increasing research interest, yet our understanding of smoldering lags behind that of flame-based combustion. Smoldering fires represent hazards to human health and safety locally, and global ecologi-cal concerns due to their potential for carbon release. Additionally, ecological effects of smoldering ground fires are generally perceived to be negative, particularly where their historical frequencies are thought to be low. This synthesis describes some aspects of smoldering combustion, and discusses some of the particular ecological aspects of ground fires, focusing on examples from the southeastern United States. We suggest that despite the well-recognized negative aspects of ground fires, there may exist under-recognized ecological benefits that should be further studied and weighed against known hazards posed by these events.. 2013. Smoldering combustion and ground fires: ecological effects and multi-scale significance. Fire Ecology 9(1): 124-132. doi: 10.4996/ fireecology0901124 introduCtion
... Based on the results of the dry heat treatments, the germination of R. michauxii does not appear to be linked to dry heat shock. Our application of dry heat at temperatures between 60 and 140 °C were based on maximum wildfire surface temperatures and residence times recorded in southeastern US forests (Heyward 1938;Swift and others 1993;Varner 2005). Once seeds are dispersed to the seedbank and are under a fine layer of soil or detritus, the heat-buffering capacity of soils are likely to protect seeds from maximum surface fire temperatures. ...
Article
The federally listed Michaux's sumac (Rhus michauxii Sarg. [Anacardiaceae]) is one of the most endangered shrubs in the southeastern US. The endangered status of this fire-adapted species may be primarily attributed to fire suppression. Many hard-seeded plants in fire-adapted ecosystems can survive high temperatures and germinate in response to scarification caused by high fire temperatures. Unfortunately, little is known about the germination strategy of Michaux's sumac. Using seeds collected from Virginia Army National Guard, Maneuver Training Center, Fort Picket, that were subsequently manually scarified, we discovered that imbibition of intact seeds was prevented by endocarp impermeability. We then examined if dry heat, as an analogue for wildfire, could be used to break this physical dormancy. Replicated (n = 3) dry heat treatments applied at 60, 80, 100, 120, and 140 °C (140, 176, 212, 248, and 284 °F) for 5 and 10 min durations, as well as a nontreated control, yielded no germination after 4 wk. All seeds were subsequently manually scarified. During the next 4 wk, a range of 47.8 ± 4.8 to 56.7 ± 12% germination was observed for scarified seeds of the 60 and 80 °C (at 5 and 10 min) treatment groups, not significantly different than the 61.6 ± 6.2% germination observed with scarified control seeds. Very low germination was observed when seeds were exposed to 100 °C for 5 min; all longer duration and higher treatment temperatures were lethal.
... Prescribed fire in southern Appalachian pine-hardwood forests produced flame temperatures near 800uC, while soils only reached 60uC at 5 cm (Swift et al. 1993). Similarly, in prescribed fires of a Florida longleaf pine forest, maximum soil temperatures recorded at 5, 10, and 20 cm below the surface soil were 134, 117 and 80uC, respectively (Varner 2005). In Florida and Mississippi longleaf pine forests, Heyward (1938) demonstrated that soil heat pulses above 90uC can persist for many minutes (approx. ...
Article
Full-text available
The requirement for heat shock to break physical dormancy of hard seeds is a widely known strategy for recruitment in fire maintained ecosystems. Despite extensive work by fire ecologists in the southeastern United States, heat shock response has been demonstrated for only a few eastern North American temperate species. In this study germination responses to dry and wet heat were investigated for three species with physical dormancy: Galactia regularis (Fabaceae), Lupinus perennis (Fabaceae), and Rhus copallinum (Anacardiaceae). Control, wet heat (1 min boiling), and dry heat (10 min at 60, 70, 80, 90, 100, and 110°C) treatments were independently applied in the laboratory. Maximum germination occurred at 80°C for G. regularis and L. perennis and was significantly greater than both control and boiling treatments. For R. copallinum, maximum germination occurred at 90°C for all populations investigated and was significantly greater than the control group in three of four populations. Heat shock germination may play a larger role in post fire recruitment in the flora of the eastern United States than currently recognized.
... Live fine root biomass was approximately evenly distributed between the forest floor and the top 30 cm of the mineral soil (Figure 4). These results are similar to results from other long-unburned Coastal Plain longleaf stands (Varner 2005). Fire damaged study trees had a minimum of 20 % forest floor consumption (mean = 50.2 ...
Article
Full-text available
Abstract One important legacy of fire exclusion in ecosystems dependent upon frequent fire is the development of organic soil horizons (forest floor) that can be colonized by fine roots. When fire is re-introduced, the forest floor is often consumed by fire and heavy overstory mortality, often delayed by months, results. We hypothesized that the delayed post-fire tree mortality is a manifestation of a cascade of physiological stresses initiated by root damage that can also magnify the impact of other kinds of damage. We investigated the physiological impact of forest floor consumption on longleaf pines (Pinus palustris Mill.) subjected to a wildfire in 2005 in a long-unburned (>50 years) forest by measuring forest floor consumption, whole tree water use, and leaf chlorophyll content. Ten of the 23 study trees died within three years post fire. Post-fire sap flux was unrelated to crown scorch, but was negatively correlated with forest floor consumption. A segmented linear regression revealed declines in sap flux until a threshold of 31 % forest floor consumption, after which further consumption had no additional effect on tree water use. Trees with >30 % forest floor consumption beneath their crowns were more than 20 times as likely to die as those with less consumption. Chlorophyll content in needles that flushed post fire was negatively correlated with crown scorch (R 2 = 0.60, P = 0.009) though all trees with scorch also experienced varying degrees of forest floor consumption. Our results suggest that the consumption of the forest floor with the likely concomitant loss of roots initiated a decline spiral, driven by an inability to supply sufficient water to the crown. Though we did not measure loss of stored carbohydrates in consumed roots directly, we infer that this likely effect, coupled with decreased crown photosynthetic capacity, eventually resulted in substantial overstory tree mortality.
Article
Full-text available
Article
were confronted with an upland forest that was dominated by one tree species-long­ leaf pine (Pinus palustris), They described these forests as open and grassy, with ma­ ture trees scattered across a fire-maintained landscape (Bartram, 1791), Stretching from the coastal plains of Virginia across a broad belt of the South Atlantic and Gulf Coasts into eastern longleaf his­ torically occurred on nearly 36 million hectares (90 million acres) and is thought to ha\'e been predominant on O\'er 24 mil­ lion of those hecrares (60 million acres)
Article
After fire exclusion in 1966, annual breeding bird censuses were conducted on an 8.6-ha plot of oldfield pine forest in N Florida. Only 11 of 43 bird species were encountered every year of the study. Most finches and brush nesting species no longer occur on the study area while several species associated with mesic conditions have increased in abundance. -from Authors
Article
Core samples of organic soil were collected in south-central Florida and transferred to a combustion laboratory at Macon, Georgia. The samples were burned in a manner which simulated the slow smoldering. This paper reports soil combustion characteristics and emissions of several criteria pollutants.
Article
Experimental plots were burned every year (1E), every 2yr (2E), or every 7yr (7E); the control plot (CE) has not burned for 20yr. Amphibians and reptiles of 27 species were captured. Severe cold in December 1983 may have caused a large decline in herpetofauna in 1984. Both Shannon-Weiner and Simpson's diversity indexes indicated that plot 2E had the lowest diversity each year. Greatest diversity was found on 1E or 7E. The 2yr fire periodicity produced a dense layer of grasses and herbaceous plants that was not readily occupied by sandhill herpetofauna. The most abundant reptile was 6-lined racerunner Cnemidophorus sexlineatus (33% of all captures). The highest density of racerunners was found on 1E, while lizards on 7E showed the greatest philopatric tendencies (especially in 1983, the year 7E was burned). -from Author
Article
Fire suppression has produced large forest floor fuel loads (ground fuels) in many western coniferous forests. Combustion of the forest floor can produce significant ecological effects due to potentially high fuel loads and proximity to living tissues. Forest floor consumption is estimated from depth changes after burning using species-specific data for bulk density. To quantify forest floor bulk density and fuel loads, 40 white fir and 61 ponderosa pine forest floor samples were randomly collected from Giant Forest, Sequoia National Park. Multivariate ANOVA determined there was a significant difference in both bulk density and fuel load for species and strata. Multiple regression analysis related mean bulk density (Mg ha-1 cm-1) of a given stratum to stratum depth and total depth of the forest floor. Bulk density increased non-linearly with profile depth and varied from 5.35-34.97 Mg ha -1 cm-1 (0.053-0.35 g cm-3) and 1.84-13.92 Mg ha-1 cm-1 (0.018-0.139 g cm-3) for white fir and ponderosa pine. Bulk density of the lowest stratum in white fir forest floors is approximately four times greater than ponderosa pine. Forest floor fuel load varied from 6.35-146.02 Mg ha-1 and 3.68-125.19 Mg ha -1, for white fir and ponderosa pine. Forest floor bulk density and fuel load were non-linearly related to depth for both species but negatively related to total depth for ponderosa pine. Leaf morphology and litter quality of the two species probably affected the bulk density of their forest floors.
Article
Loss of over 98% of the original extent of longleaf pine (Pinus palustris Mill.) systems has resulted in the need for development of understory restoration techniques. In natural longleaf pine systems recruitment of understory dominant species like wiregrass (Aristida stricta Michx.) likely occurs following opening of the canopy by fire or other localized disturbances. We investigated site preparation and sowing methods for reestablishing dominant understory species in heavily disturbed xeric sandhills. Wiregrass establishment was significantly higher in burned and irrigated plots than in plots that were only burned or were burned and had soil disturbance. However, without irrigation, burned and disturbed sites showed greater establishment than did either treatment alone. Overall, species richness and cover showed the same patterns when irrigation was present, but were higher in disturbed soils without fire when not irrigated. We recommend sowing native seed mix with a hayblower onto areas with light soil disturbance if irrigation is not possible. Subsequent rolling of seed into soil will increase seedling establishment. Mechanical re-establishment of native groundcover in xeric longleaf pine systems is clearly possible.