scispace - formally typeset
Search or ask a question
Journal ArticleDOI

A soil burn severity index for understanding soil-fire relations in tropical forests.

01 Dec 2008-AMBIO: A Journal of the Human Environment (The Royal Swedish Academy of Sciences)-Vol. 37, Iss: 7, pp 563-568
TL;DR: The soil burn severity index provides a set of indicators that reflect the range of conditions present after a fire, and is viewed as a tool for understanding the effects of fires on the forest floor, with the realization that as new information is gained, the index may be modified as warranted.
Abstract: Methods for evaluating the impact of fires within tropical forests are needed as fires become more frequent and human populations and demands on forests increase. Short- and long-term fire effects on soils are determined by the prefire, fire, and postfire environments. We placed these components within a fire-disturbance continuum to guide our literature synthesis and develop an integrated soil burn severity index. The soil burn severity index provides a set of indicators that reflect the range of conditions present after a fire. The index consists of seven levels, an unburned level and six other levels that describe a range of postfire soil conditions. We view this index as a tool for understanding the effects of fires on the forest floor, with the realization that as new information is gained, the index may be modified as warranted.

Summary (2 min read)

INTRODUCTION

  • Tropical forests vary considerably in the rates of fuel accumulation, the amount of biomass available for burning, the frequency and intensity of fires, and fire effects.
  • Within dry tropical forests, fire frequency and diversity of fire types occur because of the variety of biophysical environments and seasonal fluctuations between wet and dry cycles in many areas (1).
  • When temperatures and durations are extreme, even soil particles can be fused.
  • A full range of possible severity outcomes was included, allowing users to select, combine, or identify severity outcomes appropriate for their needs or application.

METHODS

  • Physical, chemical, and biological responses of soils are influenced by the prefire, fire, and postfire environments (13) (Fig. 1).
  • Fire severity, which occurs during the fire event, describes the direct effects from the fire-combustion process (8).
  • Descriptors include the forest floor conditions using metrics such as amount of new and old litter cover, amount of mineral soil exposed, and mineral soil color (12).
  • The authors then assembled severity definitions, with emphasis on tropical studies, and identified the minimum number of levels that would characterize the full range of severity outcomes available from the current fire literature.
  • The authors placed the soil burn severity index within the continuum to determine if the levels complemented characteristics related to the prefire, fire, and postfire environments.

Soil Burn Severity Index

  • The authors identified seven burn severity levels, including an unburned level, within the postfire environment using two indicator variables (litter abundance and mineral soil color) (Table 1).
  • For characterizing burned sites, the authors selected litter (e.g., dead, partially decomposed material such as grass, leaves, or needles) (indicator characteristic 1) because it has the potential to be present across all places that support vegetation.
  • A unique set of circumstances is required to create orange-colored mineral soils, such as slowly smoldering deep humus layers or wood.

Placing the Soil Burn Severity Index within the Context of the Fire Continuum

  • Examples of the prefire environment include land cover, physical setting, and soils (Fig. 1).
  • Vegetation differences between tropical dry and moist forests can alter how a fire burns.
  • All these factors influence the soil’s physical, chemical, and biological characteristics when a fire occurs (21).
  • If not much litter is present ( 1%), and there is a plurality of gray- or orange-colored soil, temperatures could have exceeded 2908C, indicating water repellency is destroyed (8) (levels 5 and 6).
  • Therefore, factors such as mineral soil exposed (levels 3 through 6), char color (identified in levels 2 through 6), and depth of char in relation to bud location could be useful burn severity indicators for postfire vegetation response (Table 1).

DISCUSSION

  • The authors objective was to develop a simple soil burn severity index applicable to tropical systems and useful for assessing fire effects by integrating and synthesizing information related to soil burn severity from many places.
  • Rather, the authors believe through their literature synthesis that researchers will group, select, or in some cases split levels into finer groups.
  • While developing the soil burn severity index, some questions arose concerning its applicability.
  • The objective is to quantify the most abundant characteristic.
  • Other questions concern spatial and temporal aspects.

CONCLUSION

  • Methods for assessing the effects of fire and predicting ecosystem responses to wildfire are ongoing needs in natural resource management.
  • Persistence of tree related patterns in soil nutrients following slash-and-burn disturbance in the tropics.

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

Theresa B. Jain, William A. Gould, Russell T. Graham, David S. Pilliod, Leigh B. Lentile and Grizelle Gonza
´
lez
A Soil Burn Severity Index for Understanding
Soil-fire Relations in Tropical Forests
Methods for evaluating the impact of fires within tropical
forests are needed as fires become more frequent and
human populations and demands on forests increase.
Short- and long-term fire effects on soils are determined
by the prefire, fire, and postfire environments. We placed
these components within a fire-disturbance continuum to
guide our literature synthesis and develop an integrated
soil burn severity index. The soil burn severity index
provides a set of indicators that reflect the range of
conditions present after a fire. The index consists of
seven levels, an unburned level and six other levels that
describe a range of postfire soil conditions. We view this
index as a tool for understanding the effects of fires on
the forest floor, with the realization that as new informa-
tion is gained, the index may be modified as warranted.
INTRODUCTION
Tropical forests vary considerably in the rates of fuel
accumulation, the amount of biomass available for burning,
the frequency and intensity of fires, and fire effects. Throughout
these forests, structure and composition vary widely in response
to the wide disparity of climates (e.g., temperature a nd
precipitation), rates of evapotranspiration, and physical settings
(e.g., topography, geology, elevation). Within dry tropical
forests, fire frequency and diversity of fire types occur because
of the variety of b iophysical environm ents and seasonal
fluctuations between wet and dry cycles in many areas (1).
Although conditions are generally too damp and fuel is limited
to sustain fires in moist and wet tropical forests, canopy
openings created by disturbances (e.g., fires, land clearing) give
surface fuels an opportunity to dry, thus creating conditions
that favor future fires (2). Additionally, in both dry and moist
tropical forests, postdisturbance vegetation often consists of an
abundance of vines, lianas, grasses, and forbs, which add to fuel
loads and thus increase the potential for a forest to burn with
high intensity and be more susceptible to damaging fires (3, 4).
Given that fire is a component of tropical forests, fire-related
changes to the forest floor and its contribution to soil organic
matter have strong influences on the composition and structure
of postfire forest communities (5, 6).
The impacts of fires on soil range from lightly scorching the
organics to their total consumption to heating mineral soil to a
degree that water repellency occurs. When temperatures and
durations are extreme, even soil particles can be fused. Several
studies have described these conditions for boreal and
temperate f orests, and results have been widely used in
understanding fire as a disturbance (7, 8). In the Rocky
Mountains of the US, knowledge gained by understanding the
relation between prefire forest characteristics and soil burn
severity has led to the development of fuel treatments and
techniques (e.g., prescribed burning, harvesting methods,
activity slash treatments) designed to alter fire behavior and
soil burn severity outcomes (9, 10, 11).
Although many soil burn severity studies have been
conducted throughout the world, classes used to describe fire
effects on soil are inconsistent, making it difficult to synthesize
information. To partially address this deficiency, Jain and
Graham (12) used a soil burn severity index for the cold, moist,
and dry temperate forests of the Rocky Mountains of western
North America, placing available literature within the context
of a fire-disturbance continuum presented by Jain et al. (13)
(Fig. 1). Rather than attempting to redefine severity, Jain and
Graham (12) synthesized current applications of severity into
one integrated index that is applicable for a variety of fires and
objectives and that is useful at a variety of spatial and temporal
scales. A full range of possible severity outcomes was included,
allowing users to select, combine, or identify severity outcomes
appropriate for their needs or application. We used this
approach to develop a working hypothesis of a soil burn
severity index for fires in tropical forests that scientists can
apply and test for validity and provide ways to improve it. In
addition, managers can use identified relations between the
index and values at risk to develop fire-management strategies
for dry and moist tropical forests.
METHODS
Physical, chemical, and biological responses of soils are
influenced by the prefire, fire, and postfire environments (13)
(Fig. 1). The prefire environment refers to conditions that can
influence a fire’s outcome, such as land cover or physical
setting. The fire environment includes fire-behavior character-
Figure 1. The fire disturbance continuum is a cycle that contains
three components that describe different environments influencing
fires (13). The first component, the prefire environment, includes
forest vegetation and state of the environment (moisture levels,
amount of biomass, and species composition) the year the fire
occurs and immediately prior to the fire. However, factors such as
disturbance legacy and climate affect the fuel characteristics. The
second component, the fire environment, includes the characteris-
tics during combustion. This would include weather, fire behavior,
and suppression tactics. Fire intensity (indicator used to describe
fire behavior) and fire severity (direct fire effects) concentrate on
describing the fire process. The third component, referred to as the
postfire environment, includes burn severity that describes what is
left after the fire is out and the ecological, social, and economic
responses. These responses can occur immediately after the fire to
several centuries later. In addition, other disturbances (e.g., storms
or agriculture development) play a role in the postfire environment.
The postfire environment characteristics are dependent upon the
prefire environment and fire environment. For soils, the ecological
response includes the physical, chemical, and biological properties
of the soil after the fire and is used to provide a rationale for
classifying burn severity.
Ambio Vol. 37, No. 7–8, December 2008 563Ó Royal Swedish Academy of Sciences 2008
http://www.ambio.kva.se

istics associated with the fire, such as rate of spread, flame
length, and energy produced (referred to as fire intensity in the
literature) (14, 15). Fire severity, which occurs during the fire
event, describes the direct effects from the fire-combustion
process (8). Measures of soil fire severity concentrate on factors
specifically caused by a fire such as litter consumption or
changes to the soils. The postfire environment is best described
as ‘what is left behind’ or the appearance of the forest floor
(soil) after combustion is finished, which we call soil burn
severity. The intent behind burn severity characterizations is not
to measure consumption but to describe the environment
created by the fire within the context of the prefire and fire
environments. Descriptors include the forest floor conditions
using metrics such as amount of new (postfire) and old (prefire)
litter cover, amount of mineral soil exposed, and mineral soil
color (12).
Using the fire-disturbance continuum as a guide, we
synthesized relevant literature describing fire effects in the
tropics, temperate, and boreal forests and modified the soil burn
severity index developed for temperate forests (12). Keywords
used in the literature search included: i) tropical, fire, soils, fire
severity, fire intensity, ii) litter present, absent, consumed,
partially consumed, char color, and iii) mineral soil exposed
and its state or color (unburned, black, gray, white, orange). We
then assembled severity definitions, with emphasis on tropical
studies, and identified the minimum number of levels that
would characterize the full range of severity outcomes available
from the current fire literature. Using this literature base and
the fire-disturbance continuum as a framework, we identified
levels of soil burn severity and logical breaks between the levels
applicable for tropical forest soils. We placed the soil burn
severity index within the continuum to determine if the levels
complemented characteristics related to the prefire, fire, and
postfire environments.
RESULTS
Soil Burn Severity Index
We identified seven burn severity levels, including an unburned
level, within the postfire environment using two indicator
variables (litter abundance and mineral soil color) (Table 1).
Due to the common use of an unburned category in several
studies for unburned sites within tropical forests, we added a
level 0 to the index for sites with no sign of recent burning
(Table 1). These were sites identified either prior to the fire (5,
16, 17), outside the fire perimeter (6, 18), or patches within the
fire perimeter (19). Ellingson et al. (20) conducted soil sampling
prior to an initial fire as well as in reference sites adjacent to
burned sites. For characterizing burned sites, we selected litter
(e.g., dead, partially decomposed material such as grass, leaves,
or needles) (indicator characteristic 1) because it has the
potential to be present across all places that support vegetation.
Next, we partitioned percent litter cover into three classes.
Mineral soil color (characteristic 2) can serve as an indicator of
the postfire state of physical, biological, and chemical soil
components (21). Therefore, we partitioned the broad litter-
cover classes (levels 2 through 6) according to the abundance of
black-, gray-, or orange-colored char, resulting in six soil burn
severity levels. We added a third indicator characteristic, which
may not always be present, because outside of tropical forests,
these attributes have been associated with mineral soil char
(e.g., lines of differing mineral soil char) (22).
Organic matter, on the surface and in the soil, influences
physical, biological, and chemical responses and is an important
component to include in a soil burn severity index (23, 24).
Because we did not find any studies conducted in the tropics
that identified litter cover thresholds related to an ecological
response, we used thresholds identified in literature outside of
the tropics that relate to postfire erosion potential. These
thresholds include 30% litter cover (25), 40% litter cover (26),
45% litter cover (27), and 50% litter cover (28). Based on this
literature, we selected a threshold of 40% litter cover to partition
level 1 from levels 2 and 3. Level 1 consists of litter cover from
40% through 100%, and levels 2 and 3 have litter cover from 2%
through 39%. We further partitioned levels 2 and 3 by mineral
soil color, with level 2 containing a plurality of black char and
level 3 a plurality of gray or white char.
In some tropical forest studies, the absence of litter and the
state of the mineral soil were used as an indicator of severity.
Perez (19) characterized the postfire environment as containing
no litter, root mat, nor woody debris as one indication of
severity, and Chaco
´
n and Dezzeo (24) observed litter decom-
position in litterbags in forests where charcoal was in and on
soils. We designed levels 4 through 6 to reflect these postfire
characteristics (Table 1). These levels have either a trace (1%)
or no litter remaining and were partitioned by the color of the
mineral soil. Level 4 mineral soils contain a plurality of black
char, and level 5 contains a plurality of gray or white char. A
unique set of circumstances is required to create orange-colored
mineral soils, such as slowly smoldering deep humus layers or
wood. Because this level may be important under slash piles
associated with land-clearing activities, we added level 6
(plurality of orange-colored soil) for tropical forests.
Placing the Soil Burn Severity Index within the Context of the
Fire Continuum
Prefire Environment.
Examples of the prefire environment
include lan d cover, phys ical setting, and soils (Fi g. 1).
Characteristics of these are influenced by factors such as climate
Table 1. Integrated soil burn severity contains seven levels, including an unburned state, through the full range of postfire conditions.
Indicator characteristics 1 and 2 will always be present. Indicator characteristic 3 may or may not be present.
Indicator characteristic 1 Indicator characteristic 2 Indicator characteristic 3
Levels of soil
burn severity
Unburned No evidence of recent fire 0
.40% litter cover and/or root mat Both charred litter and unburned litter could
be present
Mineral soil has a combination of unburned
and black char; litterfall since fire
1
2% through 39% litter cover
and/or root mat
A plurality of black char influences the mineral
soil appearance
Lines of gray char from logs; litterfall since fire 2
A plurality of gray and/or white char influences
the mineral soil appearance
Lines of orange under logs; litterfall since fire 3
1% litter cover or root mat A plurality of black char influences the mineral
soil appearance
Gray char and orange-colored soils occur
under logs; litterfall since fire
4
A plurality of gray and/or white char influences
the mineral soil appearance
Orange-colored soils occur under logs;
litter fall since fire
5
A plurality of orange char influences the mineral
soil appearance
Black ash line present 1 through 2 cm below
soil surface
6
564 Ambio Vol. 37, No. 7–8, December 2008Ó Royal Swedish Academy of Sciences 2008
http://www.ambio.kva.se

and drought, disturbance legacy, prefire weather, and fuel
characteristics. Land cover (veget ation) reflects the type,
distribution, and flammability of the fuel (29, 30). For example,
vegetation differences between tropical dry and moist forests
can alter how a fire burns. In general, dry forests tend to contain
deciduous vegetation with some evergreen, while moist forests
are primarily evergreen. Plant biomass is lower and growth
patterns are more variable in the dry compared to the moist
forests. Dry forests tend to have high spatial diversity
(intermingled grass, forest, and agriculture lands) compared to
moist forests. Because diversity of fuel alters how a fire burns
through a forest and what a fire leaves behind, this may result in
more of a variety of soil burn severity levels in dry compared to
moist forests, where a fire may burn more homogeneously
across the landscape. Dry forests also have a wide range of
understory vegetation that can serve as surface fuels, while the
high cano py structure of moist forests inhibits understory
vegetation development. Thus, in moist forests, surface or
ground fires, depending on the litter depth, may favor
smoldering fires (31), which tend to favor soil burn severity
levels 5 and 6.
Physical setting describes where a site is located within a
landscape. This can influence burn severity by altering fire
intensity and affecting postfire responses, such as erosion and
regrowth of vegetation. The location of a burned area within a
landscape can be an important risk value, especially with
reference to human settlements, roads, and agricultural
practices. Thus, the application of fuel treatments to favor soil
burn severity levels 1, 2, or 3 could be developed to mitigate
postfire effects such as erosion potential, as noted in studies
outside of the tropics (25–28). In the US, concerted efforts to
mitigate erosion after fire using a variety of fire-rehabilitation
treatments and decisions have been based on landscape position
and exposed mineral soil (32).
Important soil characteristics relevant to burn severity
include soil texture, structure, moisture, amount of clay
minerals, amount of organic matter (on the surface and in the
soil), and the amount of exchangeable ions. All these factors
influence the soil’s physical, chemical, and biological charac-
teristics when a fire occurs (21). For instance, on the mineral
soil surface, soil structure is influenced by the combination of
mineral soil particles and organic matter. With increasing
depth, clay minerals begin to play a more important role in
influencing soil structure, and the aggregation of individual
mineral particles enhances porosity (8). Fire can influence clay
minerals and organic components; therefore, the abundance of
these components in the soil prior to a fire can influence a burn
severity outcome. Outside of the tropics, fire-induced changes in
these components are typically identified by changes in mineral
soil color and have been related to changes to the biological
(e.g.,microbialbiomass),chemical(e.g.,organicmatter
quantity and quality), and physical (e.g., water repellency and
bulk density) conditions (8, 21, 33–35). Coupling these prefire
attributes with the burn severity index (mineral soil color and
abundance of organic matter) can lead to an improved
understanding of prefire and postfire relations common in
many tropical studies (Table 2).
Elements associated with climate and weather, such as
periods of drought, affect fuel moisture, annual biomass, and
seasonal fluctuations in precipitation. For example, in the
tropics, a fire may occur in a particular year during a distinctive
dry or wet season, depending on El Nin
˜
o–Southern Oscillation
occurrence, or after an extended drought in disturbed and
undisturbed forests (36–38). The time of the year a fire burns
gives rise to expected fuel moistures, which in turn affects fire
behavior and soil temperature and duration of the fire event.
Lower fuel moisture will contribute to higher soil temperatures
during a fire. Castan
˜
o-Meneses and Palacios-Vargas (39) used
drying period (which affects moisture content) of slash to
understand diversity and population of 16 ant species in
response to slash-and-burn treatments (Table 2). A study such
as this could compare ant population responses and activities to
soil burn severity levels 0 through 3 to levels 4 through 6. In
another study related to fuel moisture, Kauffman et al. (40)
used drying period of prefire biomass to evaluate changes in soil
nutrients (Table 2). These changes could be quantified in
relation to soil burn severity levels 0 through 4, where the
greatest change in nutrient dynamics occurs during a fire. The
time of the year or particular year a fire burns can determine
vegetative response as shown by Sampaio et al. (41), who used
drying period to explain coppice regeneration (Table 2). The
extent of each level within the soil burn severity index will vary
depending on the climate and weather in any given year a fire
occurs.
Disturbance legacy (e.g., t ype and timing of the last
disturbance) is a factor that influences the prefire environment
and is especially important with the prevalence of slash-and-
burn agriculture occurring in places such as the neotropics.
Prefire disturbances (or lack of disturbances) can determine the
configuration, amount, and moisture content of surface fuels in
tropical forests. Canopy openings created by disturbances can
cause these fuels to dry (2), as well as create more fuels from
growth of secondary forest vegetation. Disturbance legacy
includes frequency of past fires (2) or other disturbances such as
slashing (39) or landslides (5). Several studies contain distur-
bance legacy as a source of variation to understand the role of
fires in tropical forests (24, 42–45) (Table 2).
Interactions among all the aforementioned factors influence
fuel characteristics. Many of the fire effects described in the
literature were attributed to the amount of fuel and fuel
moisture prior to the burning (40, 41) (Table 2). Fuels consist of
live and dead organic materials, and their density and moisture
contents change over time and space. Vertical and horizontal
arrangement of fuels can also be important. For example,
burning coarse woody debris (material .8 cm in diameter) in
contact with or in close proximity to the soil surface can transfer
large amounts of heat to the mineral soil for long periods, thus
altering its physical, chemical, and biological properties (46–48).
Under slash piles commonly associated with land-clearing
activities, the presence of orange-colored soil (soil burn severity
level 6) indicates that woody debris may have been present,
favoring long fire duration and creating high soil temperatures.
Fire and Postfire Environment. Fire severity concentrates on
soil heating, which includes temperature and duration, as well
as depth of heat penetration and amount of material consumed
by the fire (7, 49). The relation between fire intensity and fire
severity is not one-to-one. A smoldering fire (low intensity) may
favor high soil heating. In contrast, a fast-moving, high-
intensity fire may lead to low soil heating (8). Rather, the
amount and type of fuels, as well as other factors mentioned
previously, greatly influence heat pulse into the soil and the
resultant temperature. There is typically considerable variation
in temperatures produced within a fire, both horizontally and
vertically. However, there are ranges of soil temperatures and
heat duration that have to occur to create a particular burn
severity outcome. An understanding of the role of temperature
and heat duration, combined with burn severity indicators, can
provide ways to predict the physical, chemical, and biological
effects of fire. In cases where consumption, a measure of fire
severity, is of interest, the inverse of the index would be
appropriate, provided prefire characteristics of litter cover and
soil color are noted.
In a review of literature on the effects of fire on forest soil
properties, Certini (21) suggested that physical soil character-
Ambio Vol. 37, No. 7–8, December 2008 565Ó Royal Swedish Academy of Sciences 2008
http://www.ambio.kva.se

istics when combined with fire could influence water repellency,
structure stability, bulk density, particle-size distribution,
mineralogical assemblages, and color. DeBano (50) found that
water repellency intensifies when soils experience temperatures
between 2008C and 2608C for longer than 15 min. Soil burn
severity levels 3 and 4 (particularly level 4) would tend to
indicate conditions that could favor water repellency. If not
much litter is present (1%), and there is a plurality of gray- or
orange-colored soil, temperatures could have exceeded 2908C,
indicating water repellency is destroyed (8) (levels 5 and 6).
However, the type of soils, their physical setting, and other soil
erosion indicators must also be satisfied in order to evaluate
erosion potential (51). As mentioned previously, we selected
40% litter cover as the threshold between soil burn severity
levels 1 and 2, based on postfire erosion potential. In addition,
Lewis et al. (25) found that gray-colored char on mineral soil
coupled with less than 30% litter cover could be indicators of
potential water repellency. Certini (21) noted that mineralogy
would change becaus e of fire, but only at temperatures
exceeding 5008C. Ketterings et al. (47) verified this when they
found that severe burning (.5008C) changed soil mineralogy
under slash piles, where orange-colored soil occurred (level 6).
Although changes in mineralogy only occur at high tempera-
tures, the soil burn severity index could be used in similar
studies to gain a better understanding of soil appearance and its
relation to soil physical properties.
The soil burn severity index was designed to identify
indicators that relate to the postfire state of nutrients as a
result of the fire. Litter begins to char at ;1758C, but it is not
consumed (level 1) (52). Therefore, the presence of litter,
although charred, indicates little change in nitrogen. Black char
(remaining litter after combustion) tends to occur from ;1758C
to 3008C (level 2) (52). Gray char (levels 3 and 5) begins to occur
from ;3008C to 5008C, which may lead to both nitrogen
volatilization and mineralization (53, 54). Orange soils (level 6)
occur anywhere from ;4008Cto.5008C, indicating a major
loss of nitrogen (46). In addition, these chemical changes occur
relatively rapidly and in relatively cool temperatures (;1758Cto
4008C) (levels 1 through 4). In contrast, as the organics are
volatilized, other nutrients, such as sulfur, sodium, potassium,
magnesium, and phosphorus concentrations, can increase with
increasing temperatures (47, 55) (levels 5 and 6). The abundance
of litter (level 1) and presence of black char (levels 2 and 4), gray
char (levels 3 and 5), and orange char (level 6) should provide
an indication of nutrient abundance after a fire.
Vegetation response is strongly dependent on what plant
components are still alive or viable after a fire (41, 56). Trees are
the vegetation mo st ofte n studied after a fire. How ever,
understory vegetation, such as perennial forbs, grasses, and
shrubs, can influence the postfire environment. Germination of
seeds is dependent on their presence. Seeds tend to occur in the
surface organic layers or surface mineral soil (56). Soil burn
severity at level 0 or level 1 can indicate the presence of seeds.
Depending on their size, most seeds can withstand temperatures
ranging from ;708C to 1408C and a typical duration of 0.5 to
1.5 hr (49). Because charring of litter tends to begin at ;1758C,
any material that is not charred indicates that some seed in the
material is most likely viable. Survival of perennial forbs is
strongly dependent on reproduction method and the soil depth
of stolons, caudex, rhizomes, and/or bulbs. Therefore, factors
such as mineral soil exposed (levels 3 through 6), char color
(identified in levels 2 through 6), and depth of char in relation to
bud location could be useful burn severity indicators for postfire
vegetation response (Table 1).
Arthropods, molluscs, worms, fungi, bacteria, and other
organisms living within surface and mineral soil layers play
import ant ecolog ical roles, ranging from soil creation to
nutrient cyclin g. The influence of s oil organisms on soil
structure and processes is due, in large part, to the high
diversity and abundance of soil microbes and fauna (57). Soil
microarthropods, for example, can reach densities of thousands
to millions of individuals per square meter, even in dry soils
with low organic matter (e.g., coastal sand dunes) (58). Fire
begins to affect soil organisms at temperatures from ;608Cto
1008C (49) (level 1). In a laboratory experiment, Guerrero et al.
(55) found that 91% of the microbial carbon was lost when soil
temperatu res reached 3008C (levels 2 through 4). Thus,
relatively low burn severities can influence microbial commu-
nities (59).
Table 2. Tropical forest research studies where investigators identified or discussed ecological outcomes (postfire response) after a fire. Many
studies manipulated the prefire conditions and used this to relate to postfire response. An ‘‘x’’ under measurement time indicates if
investigators obtained measurements prefire and postfire. Postfire soil characteristics column shows the description used to characterize
burn severity. Several investigators did not use postfire characteristics in their studies.
Prefire manipulations
or source of variability Postfire responses
Measurement time
Postfire soil characteristics Literature sourcePrefire Postfire
Fuel moisture Nutrients x x None given 6
Fuel moisture Ants x x None given 39
Fuel moisture Coppice regeneration x x None given 41
Fuel moisture Biomass & nutrients x x None given 40
Fuel loading Regeneration x x None given 56
Fuel loading & moisture Nutrients x x Biomass consumption 20
Fuel loading & moisture Nutrients x x None given 17
Fuel loading & moisture Nutrients x x None given 16
Disturbance legacy Vegetation x x None given 2
None given Plant mortality x x Forest floor 19
Disturbance legacy Vegetation x Regeneration strategy 43
Fuel loading Vegetation x None given 48
Fuel loading Plant mortality x Soil color 47
Fuel loading Phosphorus x x Soil color 46
Pre- & postfire Mineralogy x x None given 5
Fuel loading Soil fertility x x None given 34
Disturbance legacy Nutrients x Fuel biomass 42
Disturbance legacy Nutrients x Biomass 44
Disturbance legacy Vegetation x Biomass distribution 45
Disturbance legacy Litter decomposition x Forest floor 24
Agriculture land Soil chemistry x Residual litter 23
Agriculture land Soil chemistry x Litter & soil char 35
566 Ambio Vol. 37, No. 7–8, December 2008Ó Royal Swedish Academy of Sciences 2008
http://www.ambio.kva.se

DISCUSSION
Our objective was to develop a simple soil burn severity index
applicable to tropical systems and useful for assessing fire
effects by integrating and synthesizing information related to
soil burn severity from many places. The levels of soil burn
severity describe a continuum that includes a full range of fire
outcomes, which may or may not be detrimental, depending on
values and management objectives. For example, some may
consider a severe fire (level 6) as beneficial because the species
that occur after a fire may be more resilient to disease and other
disturbances (60). Therefore, rather than place an absolute
value (e.g., low, moderate, or high) on our soil burn severity
index levels, results from scientific investigation and manage-
ment issues, which are dependent on the objective, should
dictate a particular value.
The applicability of the soil burn severity index is dependent
on its acceptance as a hypothesis that can be tested and
evaluated. As researchers add information concerning fire in
tropical systems, the index may change to reflect lessons learned
from new information. Our intention for the index is to develop
a standardized index that permits a hierarchical evaluation.
Scientists may choose not to use all seven levels. Rather, we
believe through our literature synthesis that researchers will
group, select, or in some cases split levels into finer groups. For
example, a soil scientist interested in soil fauna may concentrate
an investigation on comparing level 0 to level 1, and then split
level 1 into sublevels to validate the litter cover threshold of
40%. However, for ease in communication among different
forums and to provide context, using the index to state that the
work was concentrated on soil burn severity level 1 allows for a
common dialog to occur and the applicability of results may be
broadened. As a result, the pursuit of knowledge concerning fire
in tropical forests will be integrated and adaptable to a variety
of uses.
While developing the soil burn sever ity index, some
questions arose concerning its applicability. For example,
considering that fuel treatments are designed to favor fire
suppression, is it possible for fuel treatments to influence burn
severity outcomes? Current studies are considering the role of
the prefire environment (fuels) and its influence on soil burn
severity (11) for specific use in developing fuel treatments and
their implementation. When scientific information is related to a
soil burn severity index such as the one we present, results can
be used to guide management activities.
Another question relates to the indicator characteristics used
to identify soil burn severity levels (Table 1). What if these
characteristics occur in more than one soil burn severity level?
Although this may often occur, the objective is to quantify the
most abundant characteristic. Thus, we chose a plurality of the
char class rather than dominant or majority. Moreover, litter
cover is the distinct and first indicator that clearly separates the
levels. Another application would be to quantify in detail the
actual cover of the litter and different char classes and then
postclassify the data into the soil burn severity index. This
method may be very appropriate when testing the effectiveness
of the index when quantifying and communicating soil effects.
Other questions concern spatial and temporal aspects. For
example, what is the minimum spatial extent in order to have a
positive or negative effect on erosion or changes in invertebrate
communities? For another example, what is the time frame
required for microbes to recover after a fire as a function of soil
burn severity? There are no clear answers to these questions,
and answers are rare in the literature. Our hope is that by testing
and implementing the soil burn severity index, a variety of
studies can then be synthesized to begin to address such
questions. Another question concerns appropriate plot size to
quantify soil burn severity. There are a variety of sampling
procedures used in the literature, and plot size varies depending
on the scientific or management objective. The soil burn severity
index is designed to quantify postfire characteristics, no matter
the plot size or study design.
CONCLUSION
Methods for assessing the effects of fire and predicting
ecosystem responses to wildfire are ongoing needs in natural
resource management. These needs will only increase with
human pressure on land for agricultural, residential, commer-
cial, and recreational uses, in addition to ecosystem services
related to air, water, and biodiversity maintenance. These needs
are increasing in the tropics, especially in areas where
population pressures are high. The lessons learned in developing
and using consistent assessment methods can be useful to
predict ecosystem responses to different types of fires, thus
increasing our understanding of soil-fire relations across regions
and disciplines. Information from these applications of soil
burn severity may result in developing prescribed fire programs
and adjusting prefire conditions to favor desired outcomes and
response. The soil burn severity index is one method that can
add to our ecological understanding of the role of fire in
tropical forests. Cochrane and Schulze (2) state that the
‘extreme heterogeneity within burned forests must be explicitly
addressed in any effort to model regional fire dynamics or to
estimate the consequences of fire to forest well-being.’ We
believe the soil burn severity index is a step toward addressing
this issue.
References and Notes
1. Pyne, S.J., Andrews, P.L. and Laven, R.D. 1996. Introduction to Wildland Fire. John
Wiley & Sons, New York, 769 pp.
2. Cochrane, M.A. and Schulze, M.D. 1999. Fire as a recurrent event in tropical forests of
the eastern Amazon: effects on forest structure, biomass, and species composition.
Biotropica 31, 2–16.
3. Uhl, C. and Kauffman, J.B. 1990. Deforestation, fire susceptibility, and potential tree
responses to fire in the eastern Amazon. Ecology 71, 437–449.
4. Pinard, M.A., Putz, F.E. and Licona, J.C. 1999. Tree mortality and vine proliferation
following a wildfire in a subhumid tropical forest in eastern Bolivia. Forest Ecol. Manag.
116, 247–252.
5. Walker, L.R. and Boneta, W. 1995. Plant and soil responses to fire on a fern-covered
landslide in Puerto Rico. J. Trop. Ecol. 11, 473–479.
6. Kennard, D.K. and Gholz, H.L. 2001. Effects of high- and low-intensity fires on soil
properties and plant growth in a Bolivian dry forest. Plant Soil 234, 119–129.
7. Wells, C.G., Campbell, R.E., DeBano, L.F., Lewis, C.E., Fredriksen, R.L., Franklin,
E.C., Froelich, R.C., et al. 1979. Effects of Fire on Soil: A State of Knowledge Review.
Forest Service National Fire Effects Workshop, 10–14 April 1978. US Department of
Agriculture, Forest Service, Washington, DC, 34 pp.
8. DeBano, L.F., Neary, D.G. and Ffolliott, P.F. 1998. Fire’s Effects on Ecosystems. John
Wiley & Sons, New York, 333 pp.
9. Graham, R.T., Harvey, A.E., Jain, T.B. and Tonn, J.R. 1999. The Effects of Thinning
and Similar Stand Treatments on Fire Behavior in Western Forests. General Technical
Report PNW-GTR-463. US De partment of Agriculture, Forest Service, Pacific
Northwest Research Station, Portland, Oregon, 27 pp.
10. Peterson, D.L., Johnson, M.C., Agee, J.K., Jain, T.B., McKenzie, D. and Reinhardt,
E.D. 2005. Forest Structure and Fire Hazard in Dry Forests of the Western United States.
General Technical Report PNW-GTR-628. US Department of Agriculture, Forest
Service, Pacific Northwest Research Station, Portland, Oregon, 30 pp.
11. Jain, T.B., Graham, R.T., Sandquist, J., Butler, M., Brockus, K., Frigard, D., Cobb, D.,
Sup-Han, H., et al. 2008. Restoration of northern Rocky Mountain moist forests:
integrating fuel treatments from the site to the landscape. In: Integrated Restoration of
Forested Ecosystems to Achieve Multi-Resource Benefits: Proceedings of the 2007
National Silviculture Workshop, 7–11 May 2007. Deal, R.L. (ed). US Department of
Agriculture, Pacific Northwest Research Station, Portland, Oregon, pp. 1–26.
12. Jain, T.B. and Graham, R.T. 2007. The relation between tree burn severity and forest
structure in the Rocky Mountains. In: Restoring Fire-Adapted Forested Ecosystems:
Proceedings of the 2005 National Silviculture Workshop, 6–10 June 2005. Powers, R. (ed).
General Technical Report PSW-GTR-203. US Department of Agriculture, Forest
Service, Pacific Southwest Research Station, Albany, California, pp. 213–250.
13. Jain, T.B., Graham, R.T. and Pilliod, D.S. 2004. Tongue-tied. Wildfire July/August, 22–
26.
14. Davis, K.P. 1959. Forest Fire: Control and Use. McGraw-Hill, New York, 584 pp.
15. Rothermel, R.C. 1972. A Mathematical Model for Predicting Fire Spread in Wildland
Fuels. Research Paper INT-115. US Department of Agriculture, Forest Service,
Intermountain Forest and Range Experiment Station, Ogden, Utah, 40 pp.
16. Døckersmith, I.C., Giardina, C.P. and Sanford, R.L. Jr. 1999. Persistence of tree related
patterns in soil nutrients following slash-and-burn disturbance in the tropics. Plant Soil
209, 137–156.
17. Giardina, C.P., Sanford, R.L. Jr., Døckersmith, I.C. and Jaramillo, V.J. 2000. The
effects of slash burning on ecosystem nutrients during the land preparation phase of
shifting cultivation. Plant Soil 220, 247–260.
18. Garcı
´
a-Oliva, F., Sanford, R.L. Jr. and Kelly, E. 1999. Effect of burning of tropical
deciduous forest soil in Mexico on the microbial degradation of organic matter. Plant
Soil 206, 29–36.
Ambio Vol. 37, No. 7–8, December 2008
567Ó Royal Swedish Academy of Sciences 2008
http://www.ambio.kva.se

Citations
More filters
Journal ArticleDOI
TL;DR: In this article, the effects of fire and burn on vegetation and soil using field methods, remote sensing and models are discussed. But instead of collapsing many diverse, complex and interacting fire effects into a single severity index, they suggest instead of directly measuring and then integrated into severity index keys specifically designed for objective severity assessment, they highlight best practices for selecting imagery, designing an index, determining timing and deciding what to measure.
Abstract: Comprehensive assessment of ecological change after fires have burned forests and rangelands is important if we are to understand, predict and measure fire effects. We highlight the challenges in effective assessment of fire and burn severity in the field and using both remote sensing and simulation models. We draw on diverse recent research for guidance on assessing fire effects on vegetation and soil using field methods, remote sensing and models. We suggest that instead of collapsing many diverse, complex and interacting fire effects into a single severity index, the effects of fire should be directly measured and then integrated into severity index keys specifically designed for objective severity assessment. Using soil burn severity measures as examples, we highlight best practices for selecting imagery, designing an index, determining timing and deciding what to measure, emphasising continuous variables measureable in the field and from remote sensing. We also urge the development of a severity field assessment database and research to further our understanding of causal mechanisms linking fire and burn severity to conditions before and during fires to support improved models linking fire behaviour and severity and for forecasting effects of future fires.

163 citations


Cites background from "A soil burn severity index for unde..."

  • ...Given the numerous definitions of fire and burn severity, it is important for all users to explain how they define and assess severity (Simard 1991; Lentile et al. 2006; Jain et al. 2008)....

    [...]

Journal ArticleDOI
TL;DR: In this article, the main objective of the study was to determine whether visual signs of soil burn severity levels are related to changes in soil chemical and microbial properties immediately after fire, and the results showed that such changes are not correlated with the severity of soil burns.
Abstract: Aims Areas affected by wildfire comprise spatially complex mosaics of burned patches in which a wide range of burn severities coexist. Rapid diagnosis of the different levels of soil burn severity and their extents is essential for designing emergency post-fire rehabilitation treatments. The main objective of this study was to determine whether visual signs of soil burn severity levels are related to changes in soil chemical and microbial properties immediately after fire.

155 citations


Cites background from "A soil burn severity index for unde..."

  • ...Therefore, evaluation of this parameter is a critical step in the decisionmaking process for soil rehabilitation tasks (Jain et al. 2008; Parsons et al. 2010)....

    [...]

  • ...…organic layer) and mineral soil (changes in colour or structure) and the deposition of ash from the aboveground combustion of biomass after wildfire (Ryan and Noste 1985; Ulery and Graham 1993; Neary et al. 1999, 2005; Ketterings and Bingham 2000; Ryan 2002; Jain et al. 2008; Parsons et al. 2010)....

    [...]

  • ...Several visual indicators of soil burn severity have been defined on the basis of the immediate changes (observed in the field) in the forest floor (level of consumption of organic layer) and mineral soil (changes in colour or structure) and the deposition of ash from the aboveground combustion of biomass after wildfire (Ryan and Noste 1985; Ulery and Graham 1993; Neary et al. 1999, 2005; Ketterings and Bingham 2000; Ryan 2002; Jain et al. 2008; Parsons et al. 2010)....

    [...]

01 Jan 2012
TL;DR: In this article, the effect of forest fire on various properties of soil, which are important in maintaining healthy ecosystem, is reviewed, and the impact of fire on forest soil depends on various factors such as intensity of fire, fuel load and soil moisture.
Abstract: Forest fire is very common to all the ecosystems of the world. It affects both vegetation and soil. It is also helpful in maintaining diversity and stability of ecosystems. Effect of forest fire and prescribed fire on forest soil is very complex. It affects soil organic matter, macro and micro-nutrients, physical properties of soil like texture, colour, pH, Bulk Density as well as soil biota. The impact of fire on forest soil depends on various factors such as intensity of fire, fuel load and soil moisture. Fire is beneficial as well as harmful for the forest soil depending on its severity and fire return interval. In low intensity fires, combustion of litter and soil organic matter increase plant available nutrients, which results in rapid growth of herbaceous plants and a significant increase in plant storage of nutrients. Whereas high intensity fires can result into complete loss of soil organic matter, volatilization of N, P, S, K, death of microbes, etc. Intense forest fire results into formation of some organic compounds with hydrophobic properties, which results into high water repellent soils. Forest fire also causes long term effect on forest soil. The purpose of this paper is to review the effect of forest fire on various properties of soil, which are important in maintaining healthy ecosystem.

131 citations

Journal ArticleDOI
TL;DR: This paper is a review of the current state of the art literature dealing with the risk of post-fire mobilization of the metals into surface water resources to inform discussion on the preparation of suitable management plans and policies during and after fire events in order to maintain potable water quality in a cost-effective manner.

73 citations

Journal ArticleDOI
TL;DR: In this paper, a tropical montane forest (TMF) zonation pattern focusing on a case study from the Cordillera Central, Hispaniola where a striking discontinuity in forest composition occurs consistently at � 2000m elevation, with cloud forest below and monodominant pine forest above.
Abstract: Gradual changes in vegetationstructure and composition are expected to result from continuous environmental change with increasing elevation on mountains. Hence, the occurrence of abrupt or discrete ecotones in vegetation patterns is intriguing and may suggest key controls on community assembly in montane forests. We review tropical montane forest (TMF) zonation patterns focusing on a case study from the Cordillera Central, Hispaniola where a striking discontinuity in forest composition occurs consistently at � 2000m elevation, with cloud forest below and monodominant pine forest above. We propose that a discontinuity in climatic factors (temperature, humidity) associated with the trade-wind inversion (TWI) is the primary cause of this and other ecotones in TMFs that occur at a generally consistent elevation. Low humidity, fires and occasional frost above the TWI favor pine over cloud forest species. Fires in the high-elevation pine forest have repeatedly burned down to the ecotone boundary and extinguished in the cloud forest owing to its low flammability, reinforced by high humidity, cloud immersion and epiphytic bryophyte cover. Small-scale fire patterns along the ecotone are influenced by topography and where forest structure is impacted by hurricanes and landslides. Analogous patterns are observed worldwide in other TMFs where the TWI is important, high-elevation fires are frequent, and the flora contains frost-tolerant species (often of temperate lineage). The response of this and other TMFs to anthropogenic climate change is highly uncertain owing to potentially countervailing effects of different climatic phenomena, including warming temperatures and decreased frost; changes in the TWI, high-elevation drought or cloudiness; and increased frequency or

57 citations


Cites background from "A soil burn severity index for unde..."

  • ...The intensity of the March 2005 fire ranged from severe, stand-replacing crown fire (Class 4–6 burn severity per Jain et al. 2008); to moderate burns that scorched tree trunks and killed understory vegetation (Class 2–3); to light surface fires where understory shrubs and pine saplings survived…...

    [...]

References
More filters
Journal ArticleDOI
TL;DR: This work is a review of the up-to-date literature dealing with changes imposed by fires on properties of forest soils, and ecological implications of these changes are described.
Abstract: Many physical, chemical, mineralogical, and biological soil properties can be affected by forest fires. The effects are chiefly a result of burn severity, which consists of peak temperatures and duration of the fire. Climate, vegetation, and topography of the burnt area control the resilience of the soil system; some fire-induced changes can even be permanent. Low to moderate severity fires, such as most of those prescribed in forest management, promote renovation of the dominant vegetation through elimination of undesired species and transient increase of pH and available nutrients. No irreversible ecosystem change occurs, but the enhancement of hydrophobicity can render the soil less able to soak up water and more prone to erosion. Severe fires, such as wildfires, generally have several negative effects on soil. They cause significant removal of organic matter, deterioration of both structure and porosity, considerable loss of nutrients through volatilisation, ash entrapment in smoke columns, leaching and erosion, and marked alteration of both quantity and specific composition of microbial and soil-dwelling invertebrate communities. However, despite common perceptions, if plants succeed in promptly recolonising the burnt area, the pre-fire level of most properties can be recovered and even enhanced. This work is a review of the up-to-date literature dealing with changes imposed by fires on properties of forest soils. Ecological implications of these changes are described.

2,268 citations


"A soil burn severity index for unde..." refers background in this paper

  • ...All these factors influence the soil’s physical, chemical, and biological characteristics when a fire occurs (21)....

    [...]

  • ...Certini, G. 2005....

    [...]

  • ...In a review of literature on the effects of fire on forest soil properties, Certini (21) suggested that physical soil character- Ambio Vol. 37, No. 7–8, December 2008 565 Royal Swedish Academy of Sciences 2008 http://www.ambio.kva.se istics when combined with fire could influence water repellency, structure stability, bulk density, particle-size distribution, mineralogical assemblages, and color....

    [...]

  • ...Certini (21) noted that mineralogy would change because of fire, but only at temperatures exceeding 5008C....

    [...]

  • ...In a review of literature on the effects of fire on forest soil properties, Certini (21) suggested that physical soil character-...

    [...]

Book
28 Oct 2017
TL;DR: In this paper, a mathematical model for predicting rate of fire spread and intensity applicable to a wide range of wildland fuels is presented from the conceptual stage through evaluation and demonstration of results to hypothetical fuel models.
Abstract: The development of a mathematical model for predicting rate of fire spread and intensity applicable to a wide range of wildland fuels is presented from the conceptual stage through evaluation and demonstration of results to hypothetical fuel models. The model was developed for and is now being used as a basis for appraising fire spread and intensity in the National FireDanger Rating System. The initial work was done using fuel arrays composed of uniform size particles. Three fuel sizes were tested over a wide range of bulk densities. These were 0.026-inch-square cut excelsior, 114-inch sticks, and 112-inch sticks. The problem of mixed fuel sizes was then resolved by weighting the various particle sizes that compose actual fuel arrays by either surface area or loading, depending upon the feature of the fire being predicted. The model is complete in the sense that no prior knowledge of a fuel's burning characteristics is required. All that is necessary are inputs describing the physical and chemical makeup of the fuel and the environmental conditions in which it is expected to burn. Inputs include fuel loading, fuel depth, fuel particle surface-area-to-volume ratio, fuel particle heat content, fuel particle moisture and mineral content, and the moisture content at which extinction can be expected. Environmental inputs are mean wind velocity and slope of terrain. For heterogeneous mixtures, the fuel properties are entered for each particle size. The model as originally conceived was for dead fuels in a uniform stratum contiguous to the ground, such as litter or grass. I t has been found to be useful, however, for fuels ranging from pine needle litter to heavy logging slash and for California brush fields. The concept of fuel models is introduced, wherein parameters of wildland fuels necessary for inputs to the model are categbrized and tabulated. These are then used to predict fire spread and intensity; this eliminates the necessity for repeatedly measuring such parameters. The conceptual approach recognizes that fuels have inherent characteristics that are repeatable.

1,907 citations

Journal ArticleDOI
TL;DR: Dry forest is widespread, usually transitional between semidesert or savanna and moist forest, and the largest proportion of dry forest ecosystems is in Africa and the world's tropical islands, where they account for 70-80% of the forested area.
Abstract: widespread, usually transitional between semidesert or savanna and moist forest. About 40% of the earth's tropical and subtropical landmass is dominated by open or closed forest. Of this, 42% is dry forest, 33% is moist forest, and only 25% is wet and rain forest (sensu Holdridge, 42; 15). We will never know the true original or potential extent of dry forest because many savannas and scrub or thorn woodlands are thought to be derived from disturbed dry forest. Walter (110), for example, considers most or all of the grassland in India to have been derived from seasonal or dry forest. Some of the processes that cause this conversion are addressed later in this review. The largest proportion of dry forest ecosystems is in Africa and the world's tropical islands, where they account for 70-80% of the forested area. In South America they

1,587 citations


"A soil burn severity index for unde..." refers background in this paper

  • ...Land cover (vegetation) reflects the type, distribution, and flammability of the fuel (29, 30)....

    [...]

Book
01 Jan 1984
TL;DR: In this paper, Wildland Fire Fundamentals and Wildland Fuels: Wildland Forests, Wildland Fires, and Forests of the United States: A Review of Fire Management.
Abstract: FIRE ENVIRONMENT. Wildland Fire Fundamentals. Fire Behavior. Wildland Fuels. Fire Weather. FIRE REGIME. Fire Ecology. Fire and Culture. Administration of Fire Regimes. FIRE MANAGEMENT. Programs for Fire Management. Fire Suppression. Prescribed Fire. GLOBAL FIRE. Global Fire. Bibliography. Credits. Index.

1,086 citations

Journal ArticleDOI
TL;DR: In this article, the authors describe the heat transfer mechanisms operating as heat moves downward in the soil along steep temperature gradients during both wildfires and prescribed fires, and the relationship between the formation of fire-related watershed condition and subsequent surface runoff and erosion from wildland ecosystems is explored.

1,018 citations


"A soil burn severity index for unde..." refers background in this paper

  • ...DeBano (50) found that water repellency intensifies when soils experience temperatures between 2008C and 2608C for longer than 15 min....

    [...]

Frequently Asked Questions (2)
Q1. What contributions have the authors mentioned in the paper "A soil burn severity index for understanding soil-fire relations in tropical forests" ?

The soil burn severity index provides a set of indicators that reflect the range of conditions present after a fire. 

Responses in plant, soil inorganic and microbial nutrient pools to experimental fire, ash and biomass addition in a woodland savanna.