effects of prescribed forest fire on water quality and
TRANSCRIPT
Clemson UniversityTigerPrints
All Theses Theses
12-2017
Effects of Prescribed Forest Fire on Water Qualityand Aquatic Biota in the Southeastern UnitedStatesWenbo ZhangClemson University
Follow this and additional works at: https://tigerprints.clemson.edu/all_theses
This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorizedadministrator of TigerPrints. For more information, please contact [email protected].
Recommended CitationZhang, Wenbo, "Effects of Prescribed Forest Fire on Water Quality and Aquatic Biota in the Southeastern United States" (2017). AllTheses. 2763.https://tigerprints.clemson.edu/all_theses/2763
EFFECTS OF PRESCRIBED FOREST FIRE ON WATER QUALITY AND AQUATIC
BIOTA IN THE SOUTHEASTERN UNITED STATES
A Thesis
Presented to
the Graduate School of
Clemson University
In Partial Fulfillment
of the Requirements for the Degree
Master of Science
Wildlife and Fisheries Biology
by
Wenbo Zhang
December 2017
Accepted by:
Alex T. Chow, Committee Chair
Juang-Horng Chong
Marzieh Motallebi
ii
ABSTRACT
Prescribed burning has long been in the history of pine dominated forests for
thousands of years before European settlement in the Southern Unites States, and continues
to be an important management tool today. This style of management can have various
environmental effects on vegetation, soil, wildlife, air, and water quality. Among these
potential environmental effects, impact of fire on water quality have received very limited
research attention. Increasing reports of negative changes in water quality has been
reported following intense wildfire events in the recent years. However limited reports
about prescribed fire impact on water quality and aquatic ecosystems have been mixed with
examples of little to no effect, or effect to near wildfire levels in systems that undergo
frequent fire managements.
This study uses two experiments to address the lack of understanding of prescribed
forest fire influence on water quality and aquatic biota in the Southeastern pine dominated
open savanna ecosystems. In the first experiment, through a controlled experiment testing
the impact of different forest detritus on water qualities, we found that water that exported
through prescribe burned forest is less acidic, less nutritious, and have less concentrated
dissolved carbon than the unmanaged forest, which correlated to improvements in water
quality. Our second experiment, through a field survey looking and the effect of frequent
prescribed burning on aquatic macroinvertebrate communities found no significant long-
term impact of low intensity burn on the aquatic ecosystems, although we did find
iii
measurable changes in chemical and hydrological parameters through this management
method.
These results further confirm the benefits of prescribed fire management in the
Southern United States, and releases certain concerns raised by similar studies on different
ecosystems. It also validates that prescribed fire impacts are highly site-specific, fire
history, watershed condition, and management goals must be carefully considered to
determine the potential effectiveness and consequences of using this method.
Key words:
Prescribed fire, water quality, macroinvertebrates, forested watershed, savanna, Coastal
Plains, Dissolved Organic Carbon, nutrients
iv
ACKNOWLEDGEMENTS
I would like to thank my committee members, Drs. Alex Chow, Juang-Horng
Chong, Marzieh Motallebi, and my family for their unconditional support during my study.
Brian Williams, Ryan Marsh, Rob O’Neal, Christopher Olivares for all the help and fun in
the field. Dr. Adam Coates and Mary-Francis Rogers for collection and sorting of the
incubation samples. Habib Uzun, Christopher Olivares, Cagri Utku Erdem at our
collaborator Dr. Tanju Karanfil’s laboratory for the water quality analysis. Leah Gregory,
Jeff Vernon, for the laboratory assistance at Baruch Institute. Alreda Grate, Skip Van
Bloom, and the whole “family” at the Baruch Institute for an unforgettable journey through
these years.
Next, I want to thank USDA NIFA Grant (SCN-2013-02784) for funding the
research. USDA Santee Experimental Forest station, Carl C. Trettin for allowing us to
conduct the watershed experiment, Charles (Andy) Harrison, Devendra M. Amatya,
Juanita J. Lockwood for all the help through sample collections, sensor installation, and
monitoring on chemical and hydrological parameters of the experimental sites.
v
TABLE OF CONTENTS
Page
TITLE PAGE ................................................................................................................... i
ABSTRACT ..................................................................................................................... ii
ACKNOWLEDGMENTS .............................................................................................. iv
LIST OF FIGURES ........................................................................................................ vi
LIST OF TABLES ........................................................................................................ viii
CHAPTER
I. INTRODUCTION ......................................................................................... 1
References ................................................................................................ 3
II. FIELD INCUBATION OF DIFFERENT FOREST DETRITUS MATERIAL
IMPACTS ON WATER QUALITY IN A ONE YEAR PERIOD ......... 5
Introduction .............................................................................................. 5
Materials and Methods ............................................................................. 6
Results ...................................................................................................... 9
Discussion .............................................................................................. 14
Conclusions ............................................................................................ 20
References .............................................................................................. 22
III. IMMEDIATE AND LONGTERM EFFECT OF PRESCRIBED BURNING
ON WATER QUALITY AND AQUATIC BIOTA IN A PAIRED FIRST
ORDER WATERSHEDS ...................................................................... 41
Introduction ............................................................................................ 41
Materials and Methods ........................................................................... 44
Results .................................................................................................... 47
Discussion .............................................................................................. 52
Conclusions ............................................................................................ 62
References .............................................................................................. 64
IV. CONCLUSIONS.......................................................................................... 80
References .............................................................................................. 82
vi
LIST OF FIGURES
Figure Page
2.1 Set-up of litter incubation trays in the open field located at Hobcaw Barony,
Georgetown, South Carolina, USA ....................................................... 27
2.2 Aerial photo of incubation trays in the open field at Hobcaw Barony,
Georgetown, South Carolina. ................................................................. 28
2.3 Sum of daily radiation (a) average daily air temperature (b) at the incubation
site from 1/1/2016 to 1/13/2017. ............................................................ 29
2.4 Average cumulative rain volume collected per tray compared between
treatments (left axis) and sum of daily precipitation (right axis) received at
the incubation site from 1/1/2016 to 1/13/2017. .................................... 30
2.5 Average total rain volume received per tray compared between each treatment
and control. ............................................................................................ 31
2.6 Box chart for DOC concentrations between treatments (a). Ultraviolet
absorbance at 254nm over the whole incubation period (b). ................. 32
2.7 Cumulative DOC export per tray compared between each treatment and
control .................................................................................................... 33
2.8 Lab pH measurements of leachates from each treatment over the incubating
period ..................................................................................................... 34
2.9 Corrected electrical conductivity of leachate at room temperature between
treatments ............................................................................................... 35
2.10 Total mass loss between treatments ............................................................. 36
2.11 The ratio of UV absorption at 250 to 365 nm (E2:E3 ratio) of each treatment
group over the incubating period ........................................................... 37
2.12 Orthophosphate concentration compared between treatments (a). Dissolved
nitrogen concentration compared between treatments (b). .................... 38
2.13 Fuel loading (total mass) for unmanaged and managed sites (a), and
composition of litter materials (%) placed in trays (b) (Coates 2017). .. 39
vii
LIST OF FIGURES (CONTINUED)
2.14 Fluorescence excitation-emission matrices (EEMs) proportions of the leachate
from different treatments. ...................................................................... 40
3.1 Map of the Santee Experimental watersheds located in the Francis Marion
National Forest, Cordesville, South Carolina, USA (WS77 managed,
WS80 unmanaged control). ................................................................... 69
3.2 Discharge rates at the downstream watershed gauging stations 2005 ......... 70
3.3 Weather patterns of the study site during sampling year, air temperature every
15 minutes (a), and daily total precipitation (b). .................................... 71
3.4 DOC concentrations of the watersheds. ....................................................... 72
3.5 Orthophosphate concentration of the watersheds. ....................................... 73
3.6 Nonmetric multidimensional scaling ordinations for macroinvertebrate
communities of the two watersheds (WS77 treatment, WS80 control) at
different times across the whole sampling year. .................................... 74
3.7 Nonmetric multidimensional scaling ordinations for macroinvertebrate
communities for the whole post fire period (left), and whole sampling year
(right). .................................................................................................... 75
3.8 Shannon-Wiener diversity index (top) and Shannon’s equitability (bottom) of
the two watersheds compared based on six classes of macroinvertebrates.
................................................................................................................ 76
3.9 Total macroinvertebrate abundance per sample comparing the two watersheds
for the whole sampling year. .................................................................. 77
3.10 Average water depth of the sampling site at each time of invertebrate
sampling, unconnected dots indicate dried up periods. ......................... 78
3.11 Observed algal bloom one month after prescribed burn, before flow was
restored to the watersheds. ..................................................................... 79
viii
LIST OF TABLES
Table Page
3.1 Chronology of forest management practices and natural disturbances on both
the managed (Watershed 77) and unmanaged watersheds (Watershed 80)
of the Santee Experimental Forest, Cordesville, South Carolina (Amatya
& Trettin, 2007; Coates, 2017). ............................................................. 44
3.2 Comparison of chemical properties of collected water samples and on-site
measurement between the watersheds ................................................... 49
3.3 Pairwise statistical test for macroinvertebrate functional groups and taxa
abundance between the watersheds using all sampled data with negative
binomial models ..................................................................................... 52
1
CHAPTER I
INTRODUCTION
Frequent fires ignited with lightning and Native Americans maintained the longleaf
pine (Pinus palustris) dominated open savannas for thousands of years before European
settlement, it is regarded as one of the most abundant and floristically diverse ecosystems
in North America (Kush & Varner, 2009; Peet & Allard, 1993; David H. Van Lear, Carroll,
Kapeluck, & Johnson, 2005; Walker, 1993). The southern pine plantations of United States
make up 16% of the world’s timber volume (ITTO report 2016), with over 12 million
hectares of plantations dedicated for silviculture by 2000, it is the foundation of
southeastern US forest products economy (Fox, Jokela, & Allen, 2007). Prescribed burning
is an important management tool of the Southern silviculture with over 2.6 million hectares
of forest prescribe burned in 2011. Not only is it economically desirable, it is often the only
practical choice of management (Waldrop & Goodrick, 2012).
Fire have many direct and indirect effects on the environment including vegetation,
wildlife, soil, air, and water. Although acknowledged, most studies about burning impact
on water quality and aquatic ecology have only focused on intense wildfire events. Limited
reports about low to moderate intensity prescribed burning have been mixed with either
minimal short-lived response, or substantial effects similar to wildfire if managed for a
long enough period (Arkle & Pilliod, 2010; Bêche, Stephens, & Resh, 2005; Britton, 1991a,
1991b; L. E. Brown et al., 2015; Douglas, Setterfield, McGuinness, & Lake, 2015). Due to
the complexity of interactions including vegetation, soil, topography, hydrology, climate,
and management history, effect of prescribed burning on water is highly site specific and
2
unclear for the Southern coastal plains. Past studies mainly used two ways to examine
effects of fire on water quality. Field sampling in accordance with the burn, and controlled
lab extraction of burned materials with DI water. Field samples are hard to control, requires
a lot of replications, and measurements are limited to the specific sites of measurement.
Lab extraction experiments provide precise measurements and generates parameters such
as hydrophobicity of the burned materials that are hard to obtain in the field, but are not as
relatable to real conditions and can only cover a brief period. Our first experiment (chapter
2) tries to find a middle ground between these two commonly used approaches, and
compare the impacts of forest ground litter from burned and unburned forest to water
quality in both short-term and chronically. To further test the effect of frequent prescribed
burning on water and aquatic ecosystems in the southern pine forests, we also conducted a
field study of two forested first-order watersheds (chapter 3). To do so we compared the
water chemistry and stream benthic macroinvertebrate communities of two first-order
watersheds of similar historical condition, only that one watershed was consistently
managed by prescribed fire and mechanical thinning for over 50 years, while the other
watershed does not go through any forestry management.
Using both chemical and biological indicators, we try to develop an interdisciplinary
understanding of frequent prescribed fire management on the stream quality of the
Southern coastal plains. In chapter 4 we compiled the results of these two studies and tried
to answer if prescribed fire management is making positive or negative impact on the water
quality, and whether we should be concerned about the potential impacts of this forest
management technique in ecosystems like it.
3
REFERENCES
Arkle, R. S., & Pilliod, D. S. (2010). Prescribed fires as ecological surrogates for
wildfires: A stream and riparian perspective. Forest Ecology and Management,
259(5), 893–903. https://doi.org/10.1016/j.foreco.2009.11.029
Bêche, L. A., Stephens, S. L., & Resh, V. H. (2005). Effects of prescribed fire on a Sierra
Nevada (California, USA) stream and its riparian zone. Forest Ecology and
Management, 218(1–3), 37–59. https://doi.org/10.1016/j.foreco.2005.06.010
Britton, D. L. (1991a). Fire and the chemistry of a South African mountain stream.
Hydrobiologia, 218(3), 177–192. https://doi.org/10.1007/BF00038834
Britton, D. L. (1991b). THE BENTHIC MACROINVERTEBRATE FAUNA OF A
SOUTH AFRICAN MOUNTAIN STREAM AND ITS RESPONSE TO FIRE.
Southern African Journal of Aquatic Sciences, 17(1–2), 51–64.
https://doi.org/10.1080/10183469.1991.9631313
Brown, L. E., Holden, J., Palmer, S. M., Johnston, K., Sorain, J. R., & Grayson, R.
(2015). Effects of fire on the hydrology, biogeochemistry, and ecology of peatland
river systems. Freshwater Science, 34(4), 1406–1425.
https://doi.org/10.1086/683426.
Douglas, M. M., Setterfield, S. A., McGuinness, K., & Lake, P. S. (2015). The impact of
fire on riparian vegetation in Australia’s tropical savanna. Freshwater Science,
34(4), 1351–1365. https://doi.org/10.1086/684074
Fox, T. R., Jokela, E. J., & Allen, H. L. (2007). The Development of Pine Plantation
Silviculture in the Southern United States. Journal of Forestry, 105(November),
337–347.
Kush, J., & Varner, J. M. (2009). Restoring Fire to the Longleaf Pine Forest. Fire Science
Brief, (30), 1–6.
Peet, R., & Allard, D. (1993). Longleaf pine vegetation of the southern Atlantic and
eastern Gulf Coast regions: a preliminary classification. Proceedings of the Tall
Timbers Fire Ecology Conference, (18), 45–81. https://doi.org/10.1086/284833
Van Lear, D. H., Carroll, W. D., Kapeluck, P. R., & Johnson, R. (2005). History and
restoration of the longleaf pine-grassland ecosystem: Implications for species at risk.
Forest Ecology and Management, 211(1–2), 150–165.
https://doi.org/10.1016/j.foreco.2005.02.014
Waldrop, T. A., & Goodrick, S. L. (2012). Introduction to prescribed fires in Southern
ecosystems. Science Update SRS-054, (August), 80.
Walker, J. (1993). Rare vascular plant taxa associated with the longleaf pine ecosystems:
Patterns in taxonomy and ecology. In Proceedings of the Tall Timbers Fire Ecology
4
Conference (Vol. 18, pp. 105–126).
ITTO 2017. Annual report 2016. International Tropical Timber Organization, Yokohama,
Japan ISBN 978-4-86507-038-5
5
CHAPTER II
FIELD INCUBATION OF DIFFERENT FOREST DETRITUS MATERIAL
IMPACTS ON WATER QUALITY IN A ONE YEAR PERIOD
INTRODUCTION
Prescribed burning, widely used fuel reduction technique especially in southeastern
United States, is one of the essential forest management practice to reduce the susceptibility
of forests to wildfire by altering the thickness and composition of forest detritus and
understory vegetation. Few, if any, other treatments have been developed that can compete
with prescribed fire for its combination of economy and effectiveness. In 2011 alone, over
2.6 million hectares were burned by prescription for forestry purposes in the 13 southern
states (Waldrop & Goodrick, 2012).
The O Horizon of forest soils, comprised of the litter and duff, serves as a critical
source of organic materials (Binkley & Fisher, 2013). This litter and duff, referred to in
this publication as forest ground litter, contributes to dissolved organic matter in forested
watersheds and subsequently impacts water quality (Bladon, Emelko, Silins, & Stone,
2014). Wildfire causes considerable removal or the O Horizon, increases variability of
stream DOC, increases suspended sediment and particulate organic carbon (POC)
downstream, and increases soil pH (Meixner & Wohlgemuth, 2004; Smith, Sheridan, Lane,
Nyman, & Haydon, 2011a; Stephens, Meixner, Poth, McGurk, & Payne, 2004).
6
Low intensity prescribed burn on the other hand, can have little to no effect on
water quality parameters, or as some reports similar effect to those of wildfire events if
management consistently (Arkle & Pilliod, 2010; Bêche et al., 2005; Britton, 1991a, 1991b;
L. E. Brown et al., 2015; Douglas et al., 2015). As prescribed fire is supposed to be applied
with high moisture content in the duff layer to ensure an organic layer remain after burn
(Waldrop & Goodrick, 2012), if conducted properly in the lower Coastal Plains like our
site, it should not necessarily increase soil erosion.
MATERIALS AND METHODS
Study site and prescribed fires
The source ground litter materials for this incubation was collected at the Tom
Yawkey Wildlife Center in Georgetown, South Carolina, USA. Part of the forest on this
site has been managed with prescribed fire since 1978 and the predominant overstory tree
species are longleaf pine (Pinus palustris Miller), loblolly pine (Pinus taeda L.), turkey
oak (Quercus laevis Walter), and sweetgum (Liquidambar styraciflua L.) (Coates 2017).
With a mild subtropical climate, the annual average precipitation (1981-2010) was around
55 inches, and air temperature around 18 °C in the area (SCSCO, 2010). At the time of this
study in 2016, managed part of the forest has been burned 16-20 times since 1978, while
the unmanaged part of the forest has not received any form of applications since this
property was gifted to South Carolina Department of Natural Resources (SCDNR) in 1976.
To test the effect of prescribed burning, ground litter materials were collected at three units
7
from the unmanaged site of the forest, and three units from the annual prescribed burned
forest before and after a dormant seasonal prescribed fire in 2015. To test the effect of
different burn practices, ground litter materials were also collected from 3 units burned
during the growing season of 2015. Collection of ground litter materials was conducted
immediately after containment of the fires. All burns were head-fires and mean flame
lengths in each of the annual burns were mostly 0.3-1 m, and fire temperatures were
recorded using in-situ thermocouples. The averaged peak burning temperatures were
between 200-315 °C (Coates et al., 2017).
Material collection and experimental set-up
Destructive sampling of ground litter was conducted with a 1 m x 1 m sampling
frame, along a transect every 50 meters at each unit of the forest we conducted the study
on. Mixture of burned (ash and charred), fine (live vegetation and woody [only 1, 10, 100
hrs. ignited]) and detrital (litter and duff) materials were collected. In 2015, unmanaged
and pre-burned materials were collected during dormant season of the forest, and seasonal
burn materials were collected in 48 h after fire extinction of the dormant and growing
season burns with Brown’s Planar Intercept Method (J. K. Brown, 1974) as modified by
Stottlemyer (2004) (Coates et al., 2017). Collected materials were dried in oven at 70 ˚C
for 48 hours, and kept in desiccators until start of the incubation experiment on Jan 11,
2016. One kilogram of collected litter material from each sample was placed in custom-
made open top aluminum trays (2ft x 2ft x 1 ft) (Figure 2.1), and three trays were assigned
to each treatment group (unmanaged, managed pre-burn, growing seasonal burned, and
dormant seasonal burned). In addition to the sample trays, three blank controls for rain
8
water (empty trays) were also placed on an open field at Hobcaw Barony, a privately-
owned research preserve located on the coast near Georgetown, South Carolina. A weather
station (Campbell Scientific CR800) is located next to the trays which recorded
precipitation, air temperature, air pressure, radiation, humidity and wind speed every 15
minutes (Figure 2.2).
Leachate water quality
This incubation experiment lasted one year, during precipitation events, rainwater
saturates the litter material in trays and drains through a tubular opening at tray bottom,
leachate was collected in 34L marked glass carboys connected underneath. Total volume
of water collected by each tray was recorded for every rain event. When enough water (>2
L) was collected in carboys, we would transfer the samples into 1L amber glass bottles and
keep them refrigerated at 4˚C until analysis. Depending on the frequency of precipitation
and measurement difficulties, we measured different amounts of water quality parameters
on selected samples. Basic water quality parameters including pH, electoral conductivity
(EC), total suspended solid (TSS) and volatile suspended solid (VSS), dissolved organic
carbon (DOC), dissolved nitrogen (DN), and orthophosphate concentrations.
Spectroscopic properties including specific ultraviolet absorbance at 254nm (UVA254),
ratio of UV absorption at 250 to 365 nm (E2:E3 ratio), fluorescence excitation-emission
matrices (EEMs) (categorized into regions I: tyrosine-like, II: tryptophan-like, III: fulvic
acid-like, IV: soluble microbial byproduct-like and V: humic acid-like [Chen et al. 2003]).
Total weight loss for each tray was also calculated at the end of experiment as a
measurement of how much total carbon were lost during the incubation.
9
Multiple incidences of clogging occurred during the incubation in every treatment,
such an incident is exceptionally likely to occur during heavy precipitation events. Clogged
trays usually relate to increases the concentration of most measured parameters such as
DOC and nutrients, thus introducing temporary noise to that set of samples within the
treatment group. However, given the high frequency of such an incidence and differences
in consistency between treatments, we decided to keep all measured values for analysis.
RESULTS
Weather patterns
Both the litter collection site and the incubation field are located under a humid
subtropical climate, which is dominated by hot and humid air during summer time and mild
winters. Daily average air temperature followed a delayed pattern with radiation (Figure
2.3a) and only dropped below freezing point twice during the incubation year (Figure 2.3b).
Average relative humidity stayed high for most of the year, especially during July to
October where the daily average relative humidity remained higher than 70% for almost
three full months. Rain events occur frequently in every season under this climate, only
two relatively dry periods were recorded, during the month of May and after Hurricane
Matthew which landed on October 8th.
A total of 38 rain events were recorded during the one year incubation period, among which
we collected 27 rounds of samples from (Figure 2.4). Treatment trays with litter in them
typically have less water leached out to the collection carboys comparing to blank control
10
trays. Due to differences in the physical properties of litter mass, or differences in ability
to hold water, there is also a small difference of total volume received between treatments
(Figure 2.5). Even though this difference is statistically significant (p-value < 0.05), the
magnitude is relatively small (around 2% between unmanaged and managed means).
Fuel mass
Documented by Coates 2017, total fuel mass loading was significantly higher in
the long term unmanaged site (4.48 ± 0.11 kg/m2) (p<0.05) compared to continuously
managed sites (1.73 ± 0.51 kg/m2). The main litter components, total fine mass loading
(live + woody materials) was similar for both unmanaged and managed sites regardless of
fires seasons. However, detritus materials (3.68 ± 0.47 kg/m2) in the unmanaged site was
significantly higher (p<0.05) than consistently burned forest pre-burn (0.89 ± 0.10 kg/m2).
Significant consumption of fuel was recorded as fuel measurements right after seasonal
burns (0.20 ± 0.05 kg/m2 after dormant season burn, and 0.52 ± 0.11 kg/m2 after growing
season burn) were significantly lower than the pre-burn level. As indicated these results
confirm that prescribed burning consumes detritus materials particularly by changing their
thickness and depth (Coates et al., 2017). Therefore, combination of fuel mass is changed
per unit weight of the materials collected from burned sites. On the other side, live
vegetations were not detected in the samples after dormant and growing season burn by
virtue consumption of the fire.
Chemistry
11
Among our measured samples, specific ultraviolet absorbance at 254nm (UVA254)
is linearly correlated to DOC measurements (R2 = 0.718, n = 30), due to the high frequency
of sampling, UVA254 was used as a surrogate measurement method for DOC
concentration in this experiment for faster processing. Our measurements show that
UVA254 is significantly higher in the unmanaged control than any of the leachate from
managed site litters (Figure 2.6), indicating a more concentrated DOC discharge coming
out from the unmanaged forest ground litter, while no significant differences between all
managed treatments were observed. This difference between managed and unmanaged
leachate is highly significant at the beginning of the experiment, later in the incubation
period it gradually became less apparent, and about 8 months into the experiment there
were no longer any statistically significant differences between all litter trays. Multiplying
the UVA254 converted DOC concentration with rain volume of each tray we estimated the
total DOC discharge from each tray as shown in Figure 2.7. Although different burn regime
did not seem to produce any significant differences in the quantity of carbon export,
unmanaged control site litter were able to export twice the amount of dissolved carbon
comparing to the managed sites with the same unit mass of litter material.
In terms of basic water quality parameters, pH of unmanaged litter tray leachate
was consistently lower than all other treatments (Figure 2.8), while there were not enough
differences between all the managed site trays or the blank control. Electrical conductivity
did not show any consistent differences between treatments or control (Figure 2.9).
TSS/VSS was measured in the first three set of samples, but expressed high levels
of inconsistency within each treatment. Turbidity was measured in substitution for later
12
samples, but proven to be inaccurate due to the influence of color on measurements. At the
end of the experiment, unmanaged site trays had significantly higher weight loss than other
treatments (figure 2.10). As estimated in Figure 2.7, difference or carbon export in the
dissolved format was about 6g, however we have over 100g difference in total mass loss
between unmanaged and managed treatment trays, suggesting more litter was lost in
particulate form in the unmanaged group than others.
E2:E3 ratio showed a decreasing trend overall (Figure 2.11). In the first rain event,
E2:E3 is significantly different between all treatments, leachate from the preburn litter and
litter from the unmanaged site is exceptionally higher than the two burned treatments,
indicating the burned litter have higher molecular weight than the treatments without any
recent burn (Callaham, Scott, O’Brien, & Stanturf, 2012; Revchuk & Suffet, 2014). This
difference is quickly reduced to none significant since the second rain event. Throughout
the one-year incubation period, this difference in molecular weight is reversed after 4 to 5
months (about 15 rain events). Litter collected from forest that undergone growing season
burn had lowest molecular weight and dormant season burned litter had second lowest,
while pre-burn and unmanaged control litters didn’t have any consistently differences.
EEMs: A gradual but substantial reduction in region I, II, and IV (tyrosine-like compound,
tryptophan-like compound, and soluble microbial byproduct-like compound) proportions
over time, while humic acid-like and fluvic acid-like proportions increased during the
incubation period (Figure 2.14). This pattern is observed in all samples but particularly
noticeable in the litters from managed treatments, a much less significant change is
observed in unmanaged control litter.
13
Nutrients
Our setup is on an annually mowed open field far from the forest and tree lines, but
still experienced small levels of wildlife interference. For example, we have observed many
incidences of small reptiles like Carolina anole (Anolis carolinensis) and various types of
invertebrate such as wasps and midges utilizing the structure of our incubation trays. Most
of the wildlife do no post any significant influence on our measurement results, however
abundant bird feces were observed in a few selected trays at a few points during the
incubation. Even though we removed as much of it as we could at the time of discovery, it
still caused small but noticeable spikes in nutrient concentration in a few measurements.
We did not exclude these measurements from the results as they did not post a large enough
influence on the overall pattern, but it is worth noting for the few measurements that we
observed one treatment suddenly showed abnormal jumps in nutrient concentration when
comparing to other treatments.
Shown in (Figure 2.12a), orthophosphate concentrations experienced a dramatic
decreasing trend over the incubation period, about six months into the experiment its
concentration dropped lower than 1mg/L for all treatments. This parameter is closely
related to the amount and frequency of precipitation. As there might be temporary spikes
of orthophosphate following long dry periods. Between treatments, pre-burn litter stands
out consistently having the lowest amount of phosphorus, while other treatments stayed
relatively similar with each other.
14
Dissolved nitrogen was measured less often than orthophosphate (Figure 2.12b),
while there wasn’t any noticeable trend over the course of time, unmanaged site litter
consistently leached out much more dissolved nitrogen than the other treatments. Like the
pattern observed in phosphorus concentrations, Pre-burn litter also has the lowest amount
of dissolved nitrogen.
DISCUSSION
Summarized by Wang et al 2012 in a meta-analysis about soil parameters in
response to fire, most of the 75 studies on wildfires induced significant reductions in soil
DOC, especially in the short-term after fire (<3 months, more than 50% reduction).
However, this effect is highly variable with an average of almost no influence on soil
organic matter among the limited 10 studies about prescribed fire (Q. Wang, Zhong, &
Wang, 2012). For studies about fire impact on stream properties, Multiple studies of
wildfire events reported POC and DOC concentrations either increase or vary significantly
more comparing to pre-fire conditions in downstream waters, where as prescribed fire have
been reported to have little effect on DOC concentration or discharge (Battle & Golladay,
2003; Minshall, Brock, Andrews, & Robinson, 2001; Revchuk & Suffet, 2014) In our
experiment, given the same unit mass of ground litter, unmanaged site litter leached out
over two-folds more concentrated DOC solution comparing to fire managed site litters.
Being in a controlled open field, without any deposition of fresh litter or influence
of soil, this difference in DOC concentration is gradually reduced to insignificant within 9
15
months of field incubation (Figure 2.6). There are a lot of physical and structural
differences between these litter masses that might be the driving cause for this observation.
Differences in litter composition between these treated and untreated forests and
differences in debris size and structural shape makes them have different capability of
withholding moisture and decomposition rate. The rate of POC loss is related to the shape
and size of the debris, also decomposition stages. Loss of forest floor (O horizon) after
wildfire may be near complete, with reductions of > 90% reported (Baird, Zabowski, &
Everett, 1999; Murphy, Johnson, Miller, Walker, & Blank, 2006), and leaves behind
mainly fine ash materials. In our case of low intensity fire, a lot of the litter left are lightly
charred vegetation parts, which are not necessarily more volatile than normal forest detritus
material. Unmanaged site ground litter on the other hand, due to lengthy periods of
decomposition in the thick forest fuel bed, have a much large proportion of fine detrital
material (Figure 2.13), which got flushed out during rain even more than the burned ash.
According to the total mass loss at the end of our incubation period (Figure 2.10), combined
with observation of the litter composition (Figure 2.13), we can confidently say that the
organic carbon from unmanaged site litter are in fact more volatile and more hydrophilic
comparing to the litter from the managed sites. Lab leaching experiment by Revchuk and
Suffet 2014 have observed that recently burned ash had 10 times the DOC leaching
potential compared to weathered 2-year-old ash. Similar story was also observed in our
one year field incubation study (Figure 2.6), however the unmanaged forest ground
material had even greater reduction in DOC leaching potential than the burned ash.
16
For basic water quality parameters, unmanaged site litter again showed significant
differences when comparing to the managed sites. Wildfire and prescribed fire studies have
been recorded to lead increases in soil pH, because of organic acids denaturation under
heat. As expected we observed a lower pH record from the unmanaged site leachate than
all other treatments. However significant increases is only expected at high temperature
burns (> 450-500℃), where coincidentally as the complete combustion of fuel release base
and leads to an enhancement of base saturation (Arocena & Opio, 2003; Certini, 2005).
Recorded by Coates 2016, burn temperature at our sampling site rarely even reached 300℃,
but as our collected material are isolated from soil in the experimental trays, we were able
to capture the small differences otherwise would be hard to observe in actual streams (e.g.
next experiment in chapter 3 we observed an opposite pH pattern in the paired watershed
system [managed < unmanaged], possibly due to the influence of small differences in
alkalinity and hardness). It has been reported that burn events can also cause release of
inorganic ions and ephemerally increase soil EC (Hernández, García, & Reinhardt, 1997),
however in our experiment, while excluding the influence of soil, forest detritus did not
express any statistically significant differences in conductivity between ash or unburned
detritus forest ground litter at any period of incubation (Figure 2.9).
TSS and VSS measurements are important indicators of sediment run-off after
wildfire events. For our specific incubation study because no soil was included in the trays,
we are not specifically interested in the aspect of inorganic suspended solids. As confirmed
with our TSS/VSS measurements in the first three collected samples, more than 95% of
the TSS are volatile, in other words, mainly POC. So later in the incubation we measured
17
turbidity of the leachate with a colorimeter as a faster surrogate measurement of POC at
various stages of the incubation. Although we were not able to finish any statistically
representative measurements in the end due to interference of color, observations during
sampling followed a general trend that correlates to the composition of litter materials. The
unmanaged forest ground litter had the largest proportions of detrital content other than
woody debris and live vegetation (Coates, 2017), and consistently had more fine debris
leaked out of the aluminum screen we used to bag the litter. This observation agrees with
our final mass measurement, as unmanaged litter lost considerably more mass than other
treatments. Even if we subtract the dissolved cumulative carbon export from the total mass
loss, we still see a much higher amount of litter lost from the unmanaged control trays
when comparing to the managed treatment trays, while the managed treatments did not
show any significant differences between each other. This is interesting because we
generally expect to see a much higher export of POC following a fire event (Smith,
Sheridan, Lane, Nyman, & Haydon, 2011b), but our experiment showed no sign or such a
pattern and on the contrary, had more carbon exporting out of the unmanaged forest ground
litter.
Molecular weight of the DOMs was estimated by the E2:E3 ratio (Peuravuori &
Pihlaja, 1997; J. J. Wang, Dahlgren, & Chow, 2015). Differences in detritus composition
between treatments can have substantial influence on this parameter. Majidzade et al.
conducted a controlled lab burning on similar litter materials collected for this experiment
(Majidzade, Wang, & Chow, 2015). E2:E3 was observed to shift differently before and after
burn depending on the type of litter material. Coates 2017 conducted Pyrolysis–gas
18
chromatography–mass spectrometry on the same initial litter material used in this
experiment, and suggested that frequent prescribed fire in longleaf pine forests may alter
the concentration of some individual chemicals, but does little to affect the overall chemical
integrity of forest detritus (Coates, 2017). Yet as our incubation went onward, molecular
weight differences started to become clearer and clearer, most likely due to the differences
in decomposition rate between treatments. Because this parameter is not of primary
concern in this chapter, and more detailed analysis of molecular composition on the litter
after incubation is still being conducted, we will not draw any conclusions with the limited
information we have now.
Two nutrient parameters were measured in this study, orthophosphate and
dissolved nitrogen. Primary production in near-neutral freshwater is often limited by
phosphorus, whereas nitrogen is commonly the limiting macronutrient in temperate coastal
seas (Blomqvist, Gunnars, & Elmgren, 2004) In terms of terrestrial vegetation, Southern
pine dominated forest can be deficient in either P, or both N and P depending on soil group,
stand age, and past fertilizer applications (Dickens et al. 2004). Fire can cause a short-term
release of macronutrients stored in the foliage but will probably be consumed quickly by
live vegetation following low intensity prescribed burns. Our study isolates the litter from
live vegetation and only consider the nutrient release rate of different litter types. Results
of the one year incubation period show that ground litter from prescribed fire managed
forests before the burn released the least amount of limiting nutrient among all treatments.
Unmanaged forest ground litter not only kept up with the nutrient release rate of the post-
burn ashes, it released much more dissolved nitrogen than the managed site litters. This
19
means that the unmanaged site vegetation is not as limited by nitrogen content as the
managed forests, and prescribed burning can cause measurable nutrient reduction and
eventually lead to additional nutrient deficiencies in the long run.
Protein degradation and increased proportions of humic and fuvic acid compounds
were observed in all treatment and controls over time. Burned litter had significantly higher
proportions of protein like (Region I and II) and microbial by-product like (Region IV)
compounds at the beginning of our experiment. Also documented in the lab leaching
experiment with wildfire ash by Revchuk and Suffet 2014, proteins were only observed in
recently burned ash and not the 2-year weathered ash or unburned control. This indicates
that prescribed fires have a similar impact on forest ground litter with wildfire and causes
immediate release of protein and microbial by-product like compounds. Cause for this
observed pattern is likely through the direct heat-induced mortality of microbial
communities (Hart, DeLuca, Newman, MacKenzie, & Boyle, 2005). As soil heating of
lower than 100 ℃ can cause lysing of microbial cells (DeBano and Klopatek 1988;
Covington and DeBano, 1990) and releases the protein and microbial by-products
documented in this study. The excess protein and microbial by-products are quickly
degraded and consumed by new microbes under the natural open environment, as the
regions quickly disappeared in the EEM graphs (supplementary information), and the
proportions of EEM regions became relatively similar when comparing between treatments
and controls by the end of the experiment.
20
CONCLUSION
Given the scale and long history of human introduced low-intensity fire in the
southern US, there are still a lot of potential impacts this management can have that we yet
understand. This chapter focuses on the basic water chemistry parameters in a semi-
controlled environment for a short and intermediate period, and compared differences
between fire management practices to help us understand the precise impacts low-intensity
fire can have on the southeastern pine dominated forested watersheds.
From the results of this one-year field incubation study, frequent prescribed fire
management have similar effect on chemical properties of the ground litter leachate,
regardless of the frequency or season of fire application. It has been summarized by
previous researches at this forest that low-intensity prescribed fire effectively reduces the
thickness of the forest O-horizon and decomposition rate that happens within, without
qualitative impact on the chemical composition of the ground detritus (Coates, 2017).
Regarding general water quality of the leachates in this controlled incubation system,
managed forest litter leachate is less acidic, less nutritious, and have considerably lower
total carbon output than the unmanaged control leachate. Increased runoff and erosion has
been commonly observed and regarded as one of the biggest threat to water quality
following wildfire and some prescribed fire events (Certini, 2005). However, with the same
dry mass of detritus materials, managed sites litters were not as volatile as we expected,
and significantly less litter mass was lost during the incubation period comparing to the
unmanaged control litter.
21
Quantitative differences of general water chemistry of the leachate between
treatment groups and controls were significantly reduced during the later part of incubation
when compared to the beginning. Certain parameters of the leachate such as pH and
molecular weight of the DOC may still be different after one year of field incubation, but
for most water quality parameters concerning the general aquatic biota, prescribed burning
in the southern pine forest does not seem to have a long-term direct impact on their habitat
quality, and some of the short-term effects such as shifts in DOC and nutrient
concentrations are reduced and recoverable within one-year period.
These results indicate that forest management does not need to be overly concerned
about the differences fire management frequency and season of application can have on the
water chemistry of the watershed. Managing large forested watershed will indeed impact
the downstream water quality both quantitatively and qualitatively, but most of the impacts
are recoverable within several months after application. Aquatic ecosystem and habitat
quality should not be affected chemically by low-intensity fire, further discussion on this
topic is continued in the next chapter.
22
REFERENCES
Amatya, D., & Trettin, C. (2007). Development of Watershed Hydrologic Research at
Santee Experimental Forest, Coastal South Carolina. Usda, (Figure 1), 18–22.
Arkle, R. S., & Pilliod, D. S. (2010). Prescribed fires as ecological surrogates for
wildfires: A stream and riparian perspective. Forest Ecology and Management,
259(5), 893–903. https://doi.org/10.1016/j.foreco.2009.11.029
Arocena, J. M., & Opio, C. (2003). Prescribed fire-induced changes in properties of sub-
boreal forest soils. Geoderma, 113(1–2), 1–16. https://doi.org/10.1016/S0016-
7061(02)00312-9
Baird, M., Zabowski, D., & Everett, R. L. (1999). Wildfire effects on carbon and nitrogen
in inland coniferous forests. Plant and Soil, 209(2), 233–243.
https://doi.org/10.1023/A:1004602408717
Baker, M. B. J. (1988). Hydrologic and water quality effects of fire. Effects of Fire
Management of Southwestern Natural Resources, 31–42. Retrieved from AZTNC
Science Files
Battle, J., & Golladay, S. W. (2003). Prescribed Fire’s Impact on Water Quality of
Depressional Wetlands in Southwestern Georgia. The American Midland Naturalist,
150(1), 15–25. https://doi.org/10.1674/0003-
0031(2003)150[0015:PFIOWQ]2.0.CO;2
Bêche, L. A., Stephens, S. L., & Resh, V. H. (2005). Effects of prescribed fire on a Sierra
Nevada (California, USA) stream and its riparian zone. Forest Ecology and
Management, 218(1–3), 37–59. https://doi.org/10.1016/j.foreco.2005.06.010
Binkley, D., & Fisher, R. (2013). Ecology and management of forest soils. Ecology and
management of forest soils. https://doi.org/10.1007/s11104-010-0391-5
Bladon, K. D., Emelko, M. B., Silins, U., & Stone, M. (2014). Wildfire and the Future of
Water Supply. Environmental Science & Technology, 48(16), 8936–8943.
https://doi.org/10.1021/es500130g
Blomqvist, S., Gunnars, A., & Elmgren, R. (2004). Why the limiting nutrient differs
between temperate coastal seas and freshwater lakes: A matter of salt. Limnology
and Oceanography, 49(6), 2236–2241. https://doi.org/10.4319/lo.2004.49.6.2236
Britton, D. L. (1991a). Fire and the chemistry of a South African mountain stream.
Hydrobiologia, 218(3), 177–192. https://doi.org/10.1007/BF00038834
Britton, D. L. (1991b). THE BENTHIC MACROINVERTEBRATE FAUNA OF A
SOUTH AFRICAN MOUNTAIN STREAM AND ITS RESPONSE TO FIRE.
Southern African Journal of Aquatic Sciences, 17(1–2), 51–64.
https://doi.org/10.1080/10183469.1991.9631313
23
Brown, J. K. (1974). Handbook for Inventorying Downed Woody Material. USDA For.
Serv. Gen. Tech. Rep., 24. https://doi.org/10.1017/CBO9781107415324.004
Brown, L. E., Holden, J., Palmer, S. M., Johnston, K., Sorain, J. R., & Grayson, R.
(2015). Effects of fire on the hydrology, biogeochemistry, and ecology of peatland
river systems. Freshwater Science, 34(4), 1406–1425.
https://doi.org/10.1086/683426.
Brown, L. E., Johnston, K., Palmer, S. M., Aspray, K. L., & Holden, J. (2013). River
ecosystem response to prescribed vegetation burning on Blanket Peatland. PloS One,
8(11), e81023. https://doi.org/10.1371/journal.pone.0081023
Callaham, M. A., Scott, D. A., O’Brien, J. J., & Stanturf, J. A. (2012). Cumulative
Effects of Fuel Management on the Soils of Eastern U . S . Watersheds. Cumulative
Watershed Effects of Fuel Management in the Eastern United States, 202–228.
Retrieved from http://treesearch.fs.fed.us/pubs/41099
Certini, G. (2005). Effects of Fire on Properties of Forest Soils : A Review. Oecologia,
143(1), 1–10. https://doi.org/10.2307/20062214
Coates, T. A. (2017). Forest Management in Coastal Pine Forests : An Investigation of
Prescribed Fire Behavior , Detrital Chemical Composition , and Potential Water
Quality Impacts. Clemson University.
Dai, Z., Trettin, C. C., & Amatya, D. M. (2013). Effects of Climate Variability on Forest
Hydrology and Carbon Sequestration on the Santee Experimental Forest in Coastal
South Carolina, 1–44. Retrieved from
http://www.dnr.sc.gov/ccworkshops/pdf/gtr_srs72.pdf
Debano, L. F. (2000). The role of fire and soil heating on water repellancy in wildland
environments: a review. Journal of Hydrology, 231–232, 195–206.
https://doi.org/10.1016/S0022-1694(00)00194-3
DeBano, L. F., & Klopatek, J. M. (1988). Phosphorus Dynamics of Pinyon-Juniper Soils
following Simulated Burning. Soil Science Society of America Journal, 52(1), 271–
277. https://doi.org/10.2136/sssaj1988.03615995005200010048x
Dickens, E.D.; Barnett, J.P.; Hubbard, W.G.; Jokela, E.J., eds. 2004. Slash pine: still
growing and growing! Proceedings of the slash pine symposium. Gen. Tech. Rep.
SRS-76. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern
Research Station. 148 p.
Douglas, M. M., Setterfield, S. A., McGuinness, K., & Lake, P. S. (2015). The impact of
fire on riparian vegetation in Australia’s tropical savanna. Freshwater Science,
34(4), 1351–1365. https://doi.org/10.1086/684074
Elliott, K. J., L. Hendrick, R., Major, A. E., Vose, J. M., & Swank, W. T. (1999).
Vegetation dynamics after a prescribed fire in the southern Appalachians. Forest
24
Ecology and Management, 114(2–3), 199–213. https://doi.org/10.1016/S0378-
1127(98)00351-X
Fox, T. R., Jokela, E. J., & Allen, H. L. (2007). The Development of Pine Plantation
Silviculture in the Southern United States. Journal of Forestry, 105(November),
337–347. Retrieved from
http://www.ingentaconnect.com/content/saf/jof/2007/00000105/00000007/art00005
Hall, S. J., & Lombardozzi, D. (2008). Short-term Effects of Wildfire on Montane Stream
Ecosystems in the Southern Rocky Mountains: One and Two Years Post-burn.
Western North American Naturalist, 68(4), 453–462. https://doi.org/10.3398/1527-
0904-68.4.453
Hart, S. C., DeLuca, T. H., Newman, G. S., MacKenzie, M. D., & Boyle, S. I. (2005).
Post-fire vegetative dynamics as drivers of microbial community structure and
function in forest soils. Forest Ecology and Management, 220(1–3), 166–184.
https://doi.org/10.1016/j.foreco.2005.08.012
Hernández, T., García, C., & Reinhardt, I. (1997). Short-term effect of wildfire on the
chemical, biochemical and microbiological properties of Mediterranean pine forest
soils. Biology and Fertility of Soils, 25(2), 109–116.
https://doi.org/10.1007/s003740050289
Kush, J., & Varner, J. M. (2009). Restoring Fire to the Longleaf Pine Forest. Fire Science
Brief, (30), 1–6. Retrieved from http://www.firescience.gov/projects/briefs/01B-3-1-
01_FSBrief30.pdf
Lammert, M., & Allan, J. (1999). ENVIRONMENTAL AUDITING: Assessing Biotic
Integrity of Streams: Effects of Scale in Measuring the Influence of Land Use/Cover
and Habitat Structure on Fish and Macroinvertebrates. Environmental Management,
23(2), 257–270. https://doi.org/10.1007/s002679900184
Majidzade, H., Wang, J., & Chow, A. T. (2015). Organic Matter and Disinfection By-
Product Precursor in Forested Watersheds – Part II . A Controlled Field Study. In In
Recent Advances in Disinfection By-Products (pp. 271–292).
Meixner, T., & Wohlgemuth, P. (2004). Wildfire Impacts on Water Quality. Southwest
Hydrology, (October), 24–25.
Minshall, G. W., Brock, J. T., Andrews, D. A., & Robinson, C. T. (2001). Water quality,
substratum and biotic responses of five central Idaho (USA) streams during the first
year following the Mortar Creek fire. International Journal of Wildland Fire, 10(2),
185–199. https://doi.org/10.1071/WF01017
Murphy, J. D., Johnson, D. W., Miller, W. W., Walker, R. F., & Blank, R. R. (2006).
PRESCRIBED FIRE EFFECTS ON FOREST FLOOR AND SOIL NUTRIENTS
IN A SIERRA NEVADA FOREST. Soil Science, 171(3), 181–199.
https://doi.org/10.1097/01.ss.0000193886.35336.d8
25
Peet, R., & Allard, D. (1993). Longleaf pine vegetation of the southern Atlantic and
eastern Gulf Coast regions: a preliminary classification. Proceedings of the Tall
Timbers Fire Ecology Conference, (18), 45–81. https://doi.org/10.1086/284833
Peuravuori, J., & Pihlaja, K. (1997). Molecular size distribution and spectroscopic
properties of aquatic humic substances. Analytica Chimica Acta, 337(2), 133–149.
https://doi.org/10.1016/S0003-2670(96)00412-6
Revchuk, A. D., & Suffet, I. H. (Mel). (2014). Effect of Wildfires on Physicochemical
Changes of Watershed Dissolved Organic Matter. Water Environment Research,
86(4), 372–381. https://doi.org/10.2175/106143013X13736496909671
Richter, D. D., Ralston, C. W., & Harms, W. R. (1982). Prescribed Fire: Effects on Water
Quality and Forest Nutrient Cycling. Science, 215(4533), 661–663. Retrieved from
http://www.jstor.org.libproxy.clemson.edu/stable/1688204
Shu-ren, Y. (2003). Effects of fire disturbance on forest hydrology. Journal of Forestry
Research, 14(4), 331–334. https://doi.org/10.1007/BF02857863
Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011a). Wildfire
effects on water quality in forest catchments: A review with implications for water
supply. Journal of Hydrology, 396(1–2), 170–192.
https://doi.org/10.1016/j.jhydrol.2010.10.043
Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011b). Wildfire
effects on water quality in forest catchments: A review with implications for water
supply. Journal of Hydrology, 396(1–2), 170–192.
https://doi.org/10.1016/j.jhydrol.2010.10.043
Stephens, S. L., Meixner, T., Poth, M., McGurk, B., & Payne, D. (2004). Prescribed fire,
soils, and stream water chemistry in a watershed in the Lake Tahoe Basin,
California. International Journal of Wildland Fire, 13(1), 27–35.
https://doi.org/10.1071/WF03002
T.D., S., Webb, J. . W., & Crossley, D. A. J. (1981). Communtiy Structure and Nutrient
Content of Canopy Arthropods in Clearcut and Uncut Forest Ecosystems Author ( s
): T . D . Schowalter , J . Warren Webb , D . A . Crossley , Jr . Published by :
Ecological Society of America Stable URL : http://www.jstor.o. Ecology, 62(4),
1010–1019.
Ursic, S. J. (1970). Hydrologic Effects of Prescribed Burning and Deadening Upland
Hardwoods in Northern Mississippi. Research Papers. Southern Forest Experiment
Station, (SO-54), 15. Retrieved from
http://www.srs.fs.usda.gov/pubs/rp/rp_so054.pdf
Van Lear, D. H., Carroll, W. D., Kapeluck, P. R., & Johnson, R. (2005). History and
restoration of the longleaf pine-grassland ecosystem: Implications for species at risk.
Forest Ecology and Management, 211(1–2), 150–165.
26
https://doi.org/10.1016/j.foreco.2005.02.014
Van Lear, D. H., Douglass, J. E., Cox, S. K., & Augspurger, M. K. (1985). Sediment and
Nutrient Export in Runoff from Burned and Harvested Pine Watersheds in the South
Carolina Piedmont1. Journal of Environment Quality, 14(2), 169.
https://doi.org/10.2134/jeq1985.00472425001400020003x
Van Mantgem, P. J., Schwartz, M., & Keifer, M. (2001). Monitoring fire effects for
managed burns and wildfires: Coming to terms with pseudoreplication. Natural
Areas Journal, 21(3), 266–273.
Waldrop, T. A., & Goodrick, S. L. (2012). Introduction to prescribed fires in Southern
ecosystems. Science Update SRS-054, (August), 80. Retrieved from
http://www.srs.fs.usda.gov/pubs/su/su_srs054.pdf
Walker, J. (1993). Rare vascular plant taxa associated with the longleaf pine ecosystems:
Patterns in taxonomy and ecology. In Proceedings of the Tall Timbers Fire Ecology
Conference (Vol. 18, pp. 105–126).
Wang, J. J., Dahlgren, R. A., & Chow, A. T. (2015). Controlled Burning of Forest
Detritus Altering Spectroscopic Characteristics and Chlorine Reactivity of Dissolved
Organic Matter: Effects of Temperature and Oxygen Availability. Environmental
Science and Technology, 49(24), 14019–14027.
https://doi.org/10.1021/acs.est.5b03961
Wang, Q., Zhong, M., & Wang, S. (2012). A meta-analysis on the response of microbial
biomass, dissolved organic matter, respiration, and N mineralization in mineral soil
to fire in forest ecosystems. Forest Ecology and Management, 271, 91–97.
https://doi.org/10.1016/j.foreco.2012.02.006
White, J. C., House, A., Punchard, N., Hannah, D. M., Wilding, N. A., & Wood, P. J.
(2016). Macroinvertebrate community responses to hydrological controls and
groundwater abstraction effects across intermittent and perennial headwater streams.
Science of the Total Environment, (2016).
https://doi.org/10.1016/j.scitotenv.2017.06.081
Stottlemyer AD (2004) Fuel characterization of the Chauga Ridges Region of the
southern Appalachian Mountains. M.S. Thesis. Clemson University, Clemson,
South Carolina.
27
Figure 2.1. Set-up of litter incubation trays in the open field located at Hobcaw Barony,
Georgetown, South Carolina, USA
28
Figure 2.2. Aerial photo of incubation trays in the open field at Hobcaw Barony,
Georgetown, South Carolina. (15 trays used for incubation on the left, arrow on the right
is pointing at the Campbell weather station.)
29
(a)
(b)
Figure 2.3. (a) Sum of daily radiation (b) average daily air temperature at the incubation
site from 1/1/2016 to 1/13/2017.
30
40
50
60
70
80
90
100
110
120
130
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
Rad
iati
on (
J/cm
^2
)
-5
0
5
10
15
20
25
30
35
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
Dai
ly a
ver
age
Tem
per
ature
(°C
)
30
Figure 2.4. Left axis (dotted lines): Average cumulative rain volume collected by each
tray, compared between treatments (measurements on 3/20/2016, 8/18/2016, and
10/8/2016 was estimated with precipitation record).
Right axis (grey bars): Daily precipitation received at the incubation site from 1/1/2016 to
1/13/2017
1/1/2016 4/1/2016 7/1/2016 10/1/2016 1/1/2017
0
100
200
300
400
500
600
0
50
100
150
200
Dai
ly p
reci
pit
atio
n (
mm
)
Cu
mu
lati
ve
rain
vo
lum
e (L
) Dormant season burn
Growing season burn
Preburn
Unmanaged
Blank
31
Figure 2.5. Average total rain volume received by each tray, compared between
treatments, error bars are ± SE of three replicates in each treatment.
Labels meanings: D = Dormant season burnt litter; G = Growing season burnt litter; P =
Preburn ground litter in managed site; U = Litter from unmanaged control site that has
not been burned for over 40 years. Different treatments soak up different amount of rain
water during rain events, resulting in a significant difference in total rain volume
recorded.
460
470
480
490
500
510
520
530
540
D G P U B
To
tal
rain
vo
lum
e (L
)
32
(a)
(b)
Figure 2.6. (a): Box chart for DOC concentrations between each treatment, values
calculated based on ultraviolet absorbance at 254nm (UVA254) measurements. (b):
UVA254 over the whole incubation period. UVA254 are all measured under 1 and then
multiplied with corresponding dilution factor for this result.
D G P U
0
20
40
60
80
100
120
DO
C c
once
ntr
atio
n (
mg
/L)
0
1
2
3
4
5
6
7
8
9
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
UV
ab
sorb
ance
at
25
4nm
(ab
s/cm
) Dormant meanGrowing meanPreburn meanUnmanaged mean
33
Figure 2.7. Cumulative DOC export from each treatment for each tray, values calculated
based on UV254 measurements
0
2000
4000
6000
8000
10000
12000
14000
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
DO
C l
oad
(m
g)
Dormant Growing
Preburn Unmanaged
34
Figure 2.8. Lab pH measurements of leachates from each treatment over the incubating
period, error bars represent ± SE.
4
4.5
5
5.5
6
6.5
7
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
pH
Dormant mean Growing mean Preburn mean
Unmanaged mean Blank mean
35
Figure 2.9. Corrected electrical conductivity of leachate at room temperature between
treatments, error bars represent ± SE.
4
24
44
64
84
104
124
144
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
Co
rrec
ted
EC
at
25
℃ (
µS
/cm
)Dormant mean
Growing mean
Preburn mean
Unmanaged mean
Blank mean
36
Figure 2.10. Total mass loss between treatments, error bars represent ± SE.
0
50
100
150
200
250
300T
ota
l m
ass
loss
(g)
Dormant Burn Growing Burn Preburn Unmanaged
37
Figure 2.11. The ratio of UV absorption at 250 to 365 nm (E2:E3 ratio) of each treatment
group over the incubating period, error bars represent ± SE.
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
E2:E
3ra
tio
Dormant mean
Growing mean
Preburn mean
38
(a)
(b)
Figure 2.12 (a): Orthophosphate concentration between treatments, all samples were
measured within 72 hours of collection (b): Dissolved nitrogen concentration between
treatments. Arrows indicate observation of noticeable disturbance from wildlife feces,
error bars represent ± SE.
0
1
2
3
4
5
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
Ort
ho
ph
osp
hat
e co
nce
ntr
atio
n(m
g/L
) Dormant mean
Growing mean
Preburn mean
Unmanaged mean
Blank mean
0
1
1
2
2
3
3
4
1/1/2016 3/1/2016 4/30/2016 6/29/2016 8/28/2016 10/27/2016 12/26/2016
Dis
solv
ed N
itro
gen
(m
g/L
)
Dormant burn Growing burn
Preburn Unmanaged
Blank
39
(a)
(b)
Figure 2.13 Fuel loading (total mass) for unmanaged and managed sites (a), and
combination of materials (%) placed in trays (b) based on real site data. 1 kg mixture (same
combination [%] was used for each plots) of materials were placed three each individually
in aluminum trays. Error bars show standard deviation between three replicates (n=3)
(Coates, 2017).
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Unmanaged Managed
preburnt
After dormant
season burn
After growing
season burn
Fuel
mas
s, k
g/m
2
Fine (live + woody) mass
Detrital (litter + duff) mass
0
20
40
60
80
100
120
Unmanaged Managed
preburnt
After dormant
season burn
After growing
season burn
Fuel
mas
s co
mposi
tion, %
Live Woody Detrital (litter + duff)
40
Figure 2.14. Fluorescence excitation-emission matrices (EEMs) proportions of the
leachate from different treatments (Regions I: tyrosine-like, II: tryptophan-like, III: fulvic
acid-like, IV: soluble microbial byproduct-like and V: humic acid-like). Top charts
indicate the beginning of the experiment, bottom ones indicate the end.
41
CHAPTER III
IMMEDIATE AND LONGTERM EFFECT OF PRESCRIBED BURNING ON
WATER QUALITY AND AQUATIC BIOTA IN A PAIRED FIRST ORDER
WATERSHEDS
INTRODUCTION
Historically the southeastern longleaf pine ecosystems in the US dominated the
whole southeastern landscape along the Gulf and Atlantic coastal plains, from eastern
Texas to southern Virginia and inland to the Piedmont, Ridge and Valley, and Mountain
Provinces of Alabama and Georgia. Use of fire by Native Americans and lighting strikes
maintained this habitat as an open savanna with grassy forest floor that supported a great
diversity of wildlife (Kush & Varner, 2009). This ecosystem was extensively harvested
and replaced after the European settlement due to long-term fire suppression and the
replacement with fast growing species such as loblolly pine. Today, longleaf pine-
dominated open landscape is considered endangered. With more than two million hectares
of forest prescribe burned in 2011, prescribed fire is one of the most commonly applied
methods of management in the southern pine forests (Waldrop & Goodrick, 2012). The
management goals of prescribed burn are to reduce risk of wildfire, promote timber
production, and benefit a variety of domestic and wild animals that have adapted to an open
savanna habitat.
Fire is one of the most important natural disturbances influencing the heterogeneity
and diversity of terrestrial landscapes and aquatic systems. It can directly and indirectly
42
affect aquatic and riparian communities at spatial scales ranging from microhabitats to
entire watersheds, and temporal scales ranging from days to decades (Bêche et al, 2005).
Several researches have indicated that low to moderate intensity prescribed burning have
little to no direct impact on the hydrology, water quality, or aquatic biotic communities
when comparing to the effect of wildfires (Arkle & Pilliod, 2010; Debano, 2000; Richter
et al., 1982). Despite its pervasive employment, we know very little about the impacts of
prescribed fire on aquatic systems in the Coastal Plain pine forests in the Southeastern
United States. The applicability of research conducted in other areas and different
ecosystems (Bêche et al., 2005; Britton, 1991a, 1991b; Elliott et al., 1999; Hall &
Lombardozzi, 2008) to fire management and the potential responses of stream and riparian
communities to fire in the abundant pine dominated open forest is not clear.
Majority of the coastal plain pine forests in the southeastern US are located direct
upland of coastal wetlands under humid and mild subtropical climate. Water bodies here
usually contain high concentrations of dissolved organic carbon (DOC) and support unique
types of lifeforms. DOC usually serves a supportive role to wildlife as it reduces the
bioavailability of many toxins such as heavy metals. Level of DOC as high as our study
site, however, can have more effect on biota than simply binding material to toxins (e.g.
providing shade “sunscreen” effect, food to support diverse microbes, potential toxicity at
high enough concentrations, and additional difficulties to water treatment facilities).
Macroinvertebrates are ubiquitous in aquatic ecosystems and often exhibit great
taxonomic and trophic varieties. They form relatively immobile stream communities, can
be easily collected in large numbers, react both acutely and chronically to environmental
43
changes, and occupy all stream habitats and display a wide range of functional feeding
preferences. Its communities are used commonly as an indicator for water quality, also as
a parameter for habitat quality of wildlife and fish studies (Lammert & Allan, 1999). Little
agreement exists, however, about how to describe macroinvertebrate assemblages for
biological assessment (Rosenberg and Resh 1993). Several indexes have been used widely
such as the invertebrate community index (ICI) (Ohio EPA 1988), biotic index (BI), and
the adapted field biotic index (FBI) which are based on pollution tolerances (Hilsenhoff
1987). These indexes are great at providing quick and general predictions of the condition
of a stream, unfortunately given the unique environmental conditions of our site, we do not
have enough sensitive macroinvertebrate taxa to utilize these commonly used indexes.
Under this scenario, we will be focusing on comparing the whole community structures of
different sampling periods, as well as general diversity and abundance of
macroinvertebrate populations.
The primary goal of this study is to evaluate the response of aquatic invertebrate
communities to prescribed fire management in a pair of first-order watersheds in the Santee
Experimental Forest, South Carolina, USA. Managed watershed (WS77), with a drainage
area of 155 ha, has been treated with prescribed burn and logging in 3-4-year intervals.
While the paired first-order watershed (WS80), with a drainage area of 160 ha, was
established as a control watershed at 1968, where no management practices will be
implemented (Table 3.1). The objectives of this study are to determine: (1) effects of
frequent fire management on plant communities, water chemistry, and physical water
44
quality; (2) effects of fire induced habitat changes on benthic macroinvertebrate
communities.
Table 3.1. Chronology of forest management practices and natural disturbances on both
the managed (Watershed 77) and unmanaged watersheds (Watershed 80) of the Santee
Experimental Forest, Cordesville, South Carolina (Amatya & Trettin, 2007; Coates, 2017).
Year (s) Description of treatments/disturbances
1963 Watershed 77 established as a managed, treatment watershed
1968 Watershed 80 established as a control/unmanaged watershed
1977-1981 Watershed 77 was prescribed burned at various times over 5 years
1989 Hurricane Hugo damaged 80% of forest (Sept.)
1990 Watershed 77 was salvage-harvested
2001 Mastication of understory vegetation on Watershed 77 (Feb.-Nov.)
2003 Watershed 77 prescribed burned on May 10
2006 Watershed 77 whole-tree thinning of understory in early July
2007 Watershed 77 prescribed burned on June 7
2009 Watershed 77 prescribed burned on April 21
2013 Watershed 77 prescribed burned on March 5
2016 Watershed 77 prescribed burned with aerial ignition on April 19
MATERIALS AND METHODS
We conducted the study on a paired first-order watershed system that includes a
155-ha treatment watershed (WS77) that have consistently been managed with prescribed
fire for over 40 years, and an adjacent 160-ha control watershed (WS80) that have not been
burned for 60 years in the Santee Experimental Forest. This research team coordinated a
45
prescribed burn of WS 77 in April 2017 with South Carolina Department of Natural
Resources and USDA Forest Service.
WS77 and WS80 each has a single ephemeral water channel that leads to a gauging
station where grabbed water samples were collected and in-situ sensors were installed.
Continuous monitoring of the water qualities of these watersheds was complemented with
biweekly or monthly grab samples collected with ISCO-4200 flow-proportional samplers
at the outflow of the gauging station since the mid-2000s by the USDA Forest Service. in
situ Carbolyser optical sensors (S-CAN, Vienna, Austria) also were installed and recorded
within five-minute intervals. For the ISCO and grab samples, samples were filtered with a
0.45-micron Supor membrane and analyzed with a TOC-L analyzer (Shimadzu, Kyoto,
Japan) for DOC and dissolved nitrogen (DN) concentrations.
Stream benthic macroinvertebrate samples were taken at the midpoint of the
ephemeral channels (Figure 3.1) with dip-netting, or sometimes referred to as sweep
netting, using an EPA standard D-frame dip net (EPA field operation manual 2015: EPA-
841-R-14-007). Samples were taken within one month before the prescribed burn event as
our baseline data, and post-fire samples were taken biweekly when water was present in
the channels. Three samples were collected along a 100-m transect of each channel at each
sampling date. Each sample consisted of organic matters collected from 1-square-meter
area of the streambed (usually after 3-5 sweeps). During the dry season, the channels might
contain only scattered puddles or were completely dried. In this case, samples were
collected from the puddles (with the dimensions of the puddles recorded), and no sample
was collected if the channels had completely dried up. All collected materials were stored
46
in two layers of gallon-sized Ziploc bags and kept at 4˚C until sorting. All samples were
sorted within 2 weeks of collection, and all macroinvertebrates were stored in 95% ethanol
until identification to the family/genus level using the keys in Merritt et al. 2008.
Grabbed water samples were analyzed for DOC and DN at Clemson University
Rich’s Environmental laboratory, orthophosphate concentration was measured at Clemson
Baruch Institute within 72 hours using HACH DR900 colorimeter. Other water quality
parameters including Ammonia, Nitrate, Calcium, and dissolved oxygen concentrations
were measured periodically by the USDA Santee Experimental Forest station.
Statistical analysis
A variety of comparisons were done before and after burn event, but more
importantly between the two watersheds to study the effect of long-term forest
management on the aquatic ecosystem. Dissolved oxygen and pH results were compared
using binomial models to test if the two watersheds are different from each other. Other
chemical concentrations are compared with standard means of whole year’s data.
Parameters that are likely to react acutely to fire event, dissolved organic carbon and
phosphate concentrations, are plotted in relationship to time during the incubation period.
For macroinvertebrate communities, non-metric multidimensional scaling (NMDS)
plots were used to visualize the differences in community composition and dispersion
within each watershed. Shannon-Weiner diversity index and Shannon equitability were
calculated based on family level precision and compared between two watersheds for each
sample period. On top of diversity, total abundance and species richness were assessed to
determine whether the two watersheds have similar habitat quality. Populations of
47
macroinvertebrates under specific taxonomic groups were compared to look for more
specific differences between the watersheds.
RESULTS
Hydrology
The paired watersheds usually show similar hydrological patterns and go through
multiple dry periods yearly when most of the water channels dries up and no discharge
come out of the gauging station (Figure 3.2). Prescribed fire managements need to be
performed during a relatively dry period for more effective management results. The
application in 2016 was performed under such a situation where downstream discharge
was already low at the end of dormant season. Shortly after fire, due to lack of precipitation
and high transpiration rate of the forest, discharge stopped in both watersheds completely,
and upstream surface water channels dried up shortly after. The first post-fire rain event
occurred on April 22 from the treatment watershed but was too small to restart the flow at
downstream gauging station. The first discharge occurred on June 7th after two subsequent
tropical storm events.
The two watersheds had very similar discharge rates before the prescribed burn.
After the dry period, WS77 had much higher and longer discharge comparing to WS80.
this pattern is later reversed; by October WS80 started to have higher discharge than WS77.
Weather
The Santee Experimental forest is in the Francis Marian National Forest reserve
near Charleston, South Carolina, USA. Located less than 50 miles from our litter
incubation study (Chapter 2), this site shares a similar weather pattern. Air temperature
48
dropped below freezing 6 times during the sampling year, but returned to above freezing
later the same day (Figure 3.3a). Ice was never observed at the sampling locations and the
lowest measured water temperature was 8.4 ˚C. Category-1 Hurricane Matthew passed
through the area, accompanied by heavy rainfall and strong gust wind, on 8 October 2016
(Figure 3.3b). A small number of trees were uprooted during the hurricane, and most of the
vegetations suffered various levels of damage such as broken branches. This event should
have little impact on our measurements in the short term, except for possible interference
of fallen biomass with the surface water channels. However, they might have influence on
the amount of litter fall between these two watersheds, as the watersheds have different
tolerance level to hurricane stress.
49
Chemistry
Table 3.2 Comparison of chemical properties of collected water samples and on-site
measurement between the watersheds, dissolved oxygen was not statistically different
between the watersheds. pH was recorded with miscalibrated probe so values not reported,
comparison of dissolved oxygen and pH were done using binomial models.
Managed (WS77) Unmanaged (WS80)
DOC Low (Figure 3.4) High (Figure 3.4)
Orthophosphate Low (Figure 3.5) High (Figure 3.5)
Dissolved Oxygen 31.1 ± 7.18% (no difference)
pH Low High
Conductivity 46.4 ± 3.9 μS/cm 124.6 ± 14.5 μS/cm
Ammonia 7.31E-2 ± 3.3E-3 mg/L 7.95E-2 ± 1.55E-2 mg/L
Nitrate 7.9E-3 ± 6.2E-4 mg/L 1.58E-2 ± 4.25E-3 mg/L
Calcium 2.01 ± 0.15 mg/L 7.99 ± 0.43 mg/L
Turbidity 34.5 ± 0.11 FTU 13.9 ± 0.02 FTU
DOC (Figure 3.4)
Water discharged from WS80 consistently had higher DOC concentration than
WS77, except immediately after burn and during the two months following the first flush
event after burn (Figure 3.4), during which there was no noticeable difference in DOC
concentration between the two watersheds. DOC concentration varied significantly
(10mg/L – 40mg/L) depending on the hydrology of the two watersheds.
PO4 (Figure 3.5)
50
Orthophosphate concentration was consistently lower in WS77 than in WS80, but
the difference was not substantial (Figure 3.5) before and immediately after the burn event.
As flow was restored to the watersheds after the dry period, the orthophosphate
concentration in WS77 dropped much lower than those in WS80, and at some point in our
sampling period to an undetectable level with our instrument. This pattern indicates that
the ecosystems in this watershed, at least during the intermediate post-fire period, was
limited by phosphorus. Immediate post-burn nutrient release should have occurred
according to the observed algal bloom, but was not captured by our measurements.
pH
WS77 consistently had lower pH than the unmanaged watershed throughout the
sampling year. Considering multiple instruments were used for this measurement, and even
though they were calibrated regularly, there have been significant differences in measured
values between devices, thus we did not feel appropriate to correlate the measured values
to any other parameters. This result did not agree with the result of our field incubation
study, given similar litter compositions, the contradictory patterns observed from these two
studies might be a result of sediment influence which will be discussed later with our
calcium measurements.
Dissolved Oxygen
Dissolved oxygen was not measured during the invertebrate sampling.
Measurements done by the USDA Forest Service had shown a relatively low level of DO
at both sites, averaging 31.1 ± 3.6% (mean ± standard error, n=30), and there were no
significant differences in DO between the two watersheds.
51
Invertebrates
Benthic macroinvertebrate communities often show temporal differences between
the watersheds (Figures 3.6). Given the already pre-existing differences, no significant
change in pattern was observed when comparing before and after the burn event. However,
if we combine data from the whole sampling period, the overall yearly communities of
these two watersheds are not very different (Figure 3.7).
Similar with what we observed in the NMDS results, Shannon-Weiner diversity
(Figure 3.8), taxonomic richness, and total abundance (Figure 3.9) are almost identical for
these two watersheds if we combine whole year’s data. Yet the diversity index and
Shannon’s equitability (indicating evenness within site) for WS80 variated significantly
more than WS77. (figure 3.8)
52
Table 3.3 Pairwise statistical test for macroinvertebrate functional groups and taxa
abundance between the watersheds using all sampled data with negative binomial models,
defined significant if p-value < 0.05
Feeding type Difference in
abundance based on
feeding type
Taxonomic groups Difference in
abundance based on
taxonomic group
Gatherer Equal Chironomidae Equal
Oligochaeta Equal
Isopoda Higher in WS80
Amphipoda Higher in WS77
Ceratopogonidae Not enough data
Tipulidae Not enough data
Polycentropodidae Not enough data
Predator Equal Dytiscidae Not enough data
Corduliidae &
Libellulidae
Higher in WS77
Hydrophilidae Not enough data
Omnivore Equal Crayfish Equal
Undefined Equal Copepoda (< 1mm) Equal
DISCUSSION
Macroinvertebrates
Studies on the impacts of forest management on first-order watersheds were usually
conducted on pristine headwater streams, where water quality is high and supports sensitive
macroinvertebrate species such as Ephemeroptera, Plecoptera and Trichoptera (EPT). Our
sampling site is located on the southeastern US coastal plains, with water quality
parameters similar to bottomland swamps. Waterbodies here are often associated with
extremely high organic contents, receive intense and frequent precipitation during summer
and fall seasons, and are also vulnerable to hurricane influences. Soil layer under the forest
53
detritus is dominated by sandy sediment, plus the highly active plant species adapted to
this mild climate. Watersheds here can also drain at an exceptionally high rate (citation? or
historical hydrograph). High fluctuations in the water table can create a lot of hydrological
stress in the surface water of low level watersheds. our sampling sites are a pair of first-
order watershed with ephemeral water channels that reacts a little differently to this
problem. Fluctuations of hydrology, unstable pH, low dissolved oxygen, limited nutrient
input, low calcium content, and temperature are all potential sources of stress for stream
biota in such an ecosystem. Being an abundant ecosystem type of the whole coastal
southeastern US, huge amount of these lands remains privately owned and are rarely
studied out of the topics of forestry and recreational wildlife. With a low species diversity
and rare encounters of EPT species in these ecosystems, common indexes such as FBI, BI,
and ICI using macroinvertebrates to assess water quality would not apply. Our study tries
to answer, what do the aquatic communities look like, whether the aquatic biota respond
to management practices, what other parameters are influenced by prescribed fire
management, and what are the limiting factors in similar aquatic ecosystems. At
intermediate time scale, the polygons on the NMDS maps had little overlaps indicating
noticeable differences in macroinvertebrate communities of the two watersheds. The extent
of difference suggested that the small sample size in our study might not be the biggest
source of error. These temporal differences in community can be a result of many factors
such as temperature, oxygen content, food availability, chemical stress, or hydrological
differences. As the results on hydrological and chemical properties indicated, most of these
54
parameters were comparable between the two watersheds, except hydrology and food
availability.
There are three impactful hydrological parameters for these ephemeral water
channels in regards to aquatic invertebrates: antecedent wet duration, presence of
observable stream flow, and distance to adjacent perennial sources (White et al., 2016).
Considering the stream gauge as a hard barrier for the benthic macroinvertebrates, as it is
located near the surface of water, and the waterfall-like design makes it impossible for
macroinvertebrates to move against the current, both monitoring watersheds are aquatically
isolated from the downstream waterbodies. This makes the permanent pool at the
watershed gauging stations the only source of fully aquatic species to the upstream water
channels, such as amphipods, isopods, or copepods. Two of these three abundance groups
of arthropods, which were found in almost every round of sampling, differed significantly
in their abundance between the two watersheds (Table 3.3). Although studies have shown
that distance to perennial waterbodies have a significant linear correlation with species
richness, the effect of this difference does not fall out of its confidence interval until we hit
a difference of 5 kilometers (White et al., 2016). In such case, difference on this parameter
between our sampling sites are likely negligible, as the distance to gauge at WS80 is
approximately 500 meters, WS77 approximately 900 meters. No data in our results
supported a difference in recovery time between species of higher mobility (swimmers like
amphipods and copepods) and species of lower mobility (crawlers like isopods or crayfish).
Macroinvertebrates may also behave differently in these two watersheds in regards of food
availabilities. WS77 have a much more open canopy and allows more sunlight to reach the
55
water channels than WS80. With lower DOC concentrations, lights that do hit the stream
in WS77 can also penetrate through the water better than in WS80. Theoretically more
radiation usually correlates to more primary production, but possibly due to nutrient
limitations, we observed very little benthic vegetation, and very limited suspended algal
growth during the whole sampling period. So not surprisingly we did not find any grazers
in our samples. WS77 on average had higher turbidity than WS80 with in situ sensor
measurements, according to total suspended solid and volatile suspended solid
comparisons, majority of the suspended solids found in these watersheds are organic. This
indirectly indicates that WS77 have higher suspended organic matter than WS80, which
makes the system better suited for filterers species. However due to the extreme lack of
species richness, a lot of the factors that may have influence on aquatic biota were not
reflected due to the lack of specialists. Medium stress-tolerant species, such as dragonfly
(Corduliidae & Libellulidae) and caddisfly (Plectrocnemia spp. Polycentropodidae) were
encountered occasionally, but only during the periods when there was frequent fresh
precipitation. According to the historical hydrological data from 2005 to 2016 (Figure 3.2),
similar frequencies of dry periods occurred almost every year. Our study suggested that
these types of watersheds, although post high levels of uniqueness and many environmental
parameters are impacted by forest management, supports limited aquatic biodiversity.
Under the current state they are not vulnerable to low level disturbances such as prescribed
burning. The lack of interdisciplinary studies in similar habitats and lack of replication on
this experiment makes a lot of the observed patterns difficult to explain. More work on
vertebrate communities and controlled toxicology studies can help us further understand
56
the complex impacts of persistent prescribed burn management on similar aquatic
ecosystems.
Shannon-Weiner diversity and equitability index are used here as general indicators
for the overall habitat quality, due to the commonly used aquatic biotic index not applicable
to our systems. For the whole sampling year, diversity and equitability stayed similar
between the two watersheds (Figure 3.8), however the unmanaged WS80 had a lot more
short-term variations than WS77. This could be simple statistical error due to limited
number of samples taken from each site every time, but can also relate to the more shallow
and inconstant water depth at WS80 (Figure 3.10). Shallow and variable water depth means
the total quantity of water in stream changes frequently, which relays to more variations in
chemical concentrations of the aquatic environment. Unfortunately, there is no great way
to validate these assumptions other than installing more in-situ water quality sensors and
implement more frequent invertebrate sampling techniques, which we did not have access
to when we conducted this experiment.
Studies about prescribed fire impact on macroinvertebrate communities in the
peatland river systems of UK, in conjunction with finding from studies of wildfire in
Yellowstone National Park USA, have shown that as fire produces large quantities of fine
debris and increases run-off of ground litter materials, it reduced taxonomic richness and
diversity, and increased dominance of Chironomidae and Baetis spp (L. E. Brown et al.,
2015; L. E. Brown, Johnston, Palmer, Aspray, & Holden, 2013; Minshall et al., 2001).
Browns et al. concluded that the patterns they observed was because of these taxa are
tolerant to a wide range of environmental conditions and feed on fine particulate organic
57
matter, which are more abundant in rivers draining from burned than unburned catchments.
In our case, WS80 and WS77 did not show any obvious differences in quantities of silky
or sandy benthic sediment material. Depending on the season, WS77 tends to have more
fresh litter deposit (especially in the form of pine needles), while WS80 have well mixture
of half decomposed broadleaf litter, pine needles, and woody debris. Although not
quantified, silky sediment material from unmanaged WS80 seems to have a finer particle
size than WS77. These detritus differences as many other researches have reported, can
have observable influence on the macroinvertebrate communities, but with our lack of
diversity and functional groups, we did not discover any obvious pattern in our results.
This contractionary observation with Brown et al warned us that findings of similar
researches on different ecosystems may not be as globally applicable as they imply and a
more careful selection of study results is needed before we make management
recommendations.
Hydrology
Some past studies have shown low-intensity prescribed fires have little or no
influence on stream flows (D. H. Van Lear, Douglass, Cox, & Augspurger, 1985), but many
other studies also observed an increase in stream run-off (Ursic, 1970; Schindler et al.,
1980). The influence of fire on hydrology is expressed indirectly by the changes of
vegetation, ground cover, soil and environmental factors that affect water cycles. In
general, as prescribed fires intensify and consume more of the forest floor, we expect to
58
see effects on stream hydrology similar to wildfires or forest harvesting (Baker, 1988; Shu-
ren, 2003).
Being adjacent watersheds of comparable size, these two forests receive similar
levels of total precipitation. However due to differences in vegetation type and density
(vegetation diagram), they show opposite patterns of transpiration at different seasons.
Unmanaged site has a much higher proportion of deciduous broadleaf hardwood, and
higher basal area coverage than the managed pine dominated forest. Under such
circumstances, WS80 is expected to have higher transpiration rates in growing season than
WS77. Management practices every 2-3 years introduces a lot of short term noise that
further complicate the hydrology of these watersheds, and can indirectly affect the general
water quality and aquatic communities.
The First major rain event post-burn occurred in the middle of the growing season,
when transpiration rate is exceptionally high at these forests. Due to the immediate impact
of management, such as reduction in detritus thickness and understory density, and long-
term impact such as different vegetation types and density. We observed a significantly
higher water discharge from the treatment watershed than the unmanaged site for the rest
of the growing season. However, this pattern was reversed during dormant season from
October 2016 until the end of the experiment in January 2017, most likely due to pines
being more active than hardwood during this time. This pattern agrees with our historical
hydrology data, as WS77 consistently have much higher discharge than WS80 during the
growing periods. Four management practices were conducted during the years we have
59
hydrological discharge data at the bottom of these two watersheds. During these years we
didn’t observe a significant enough shift in hydrological patterns.
Intense algal bloom was observed at the gauging station of WS77 before flow was
restored to the watershed (Figure 3.11), while the invertebrate sampling sites were still
dried up. As commonly observed in forests that undergo fire management, a short-term
bloom in vegetation usually occur after fire as heat volatilizes organic nutrient and turns
them into inorganic for vegetation to utilize. However, this burst of nutrient usually gets
consumed or lost in the atmosphere quickly as r-selected vegetation species take over the
forest floor. Our study is one of the first to report this pattern in the aquatic environment in
a prescribed burn scenario.
Chemical parameters
Being adjacent watersheds of comparable total area, the environmental parameters
and amount of precipitation each watershed receives are very similar in these two
watersheds. The observed DOC difference is expected to be the result of active forest
management. WS77 has been regularly managed mainly with prescribed fire every 2-3
years, the direct result of this management is reduction of detritus thickness, or as
commonly referred to in forestry terms, fuel reduction. In addition to fuel reduction, regular
fire management also prevents succession of the forest into mixed hardwood forest and
reduces understory density. Quantitatively these factors all translate into reductions in
DOC formation and our observations accurately confirms this expected pattern.
As many wildfire studies have suggested and some prescribed fire studies have
observed, fire can lead to increases in DOC concentration. Similar is observe in our study
60
for a brief period after fire, however this is a period of very low flow in the watersheds.
After major flush events in August, the difference in DOC concentration between the
watersheds became even more significant than before burn, not surprisingly because of
detrital layer reduction. Heating inevitably releases organic nutrients into inorganic forms
and available for plants to utilize, but this process reduces the total nutrient pool of the
forest floor and after the quick consumption of inorganic nutrient burst, differences in
nutrient concentrations between the watersheds became more significant than pre-burn
conditions.
As we observed in the incubation study (chapter 2), and most reports of wildfire,
heating of soil inexorably leads to the increase in pH as a result of organic acid denaturation
(Certini, 2005). Surprisingly with our field sampling at the paired watersheds, we observed
a consistent lower pH record at WS77 than WS80. Due to uncertainties about the accuracy
in measurement instruments, the scale of this difference is this pattern is still uncertain.
Hardness and alkalinity at these two watersheds might have played a vital role in the
abnormal pattern of pH. Low hardness and alkalinity levels in water means low buffer
capacity and higher chance for pH to be unstable, which is observed in both watersheds.
Brown et al. (2013) also observed decreases in pH and Calcium after prescribed burning
on blanket peatland, however did not define a cause of this observation. Lands comprising
our experimental forest have been used for agricultural and forestry purposes since the
early 1700s (Dai et al. 2013). Agricultural history of our sites before it was established as
experimental forest is a possible explanation of the differences in hardness and pH. Natural
water with low Calcium ion concentration may contain significant stress for arthropod
61
species due to their need of Calcium ion in forming exoskeletons (T.D., Webb, & Crossley,
1981). However, this is one of the chronic stress parameters that may not be easily
represented with our invertebrate dataset. Although there is a substantial difference in
calcium concentration between the two watersheds, we did not observe any significant
difference in arthropod abundance between the two watersheds. Most likely it is because
of overshadowing influences of other environmental stresses, or a lack of identification
precision, we did not find this parameter a significant enough factor to cause differences in
community assemblages at our sites.
Vegetation
Vegetation density and type play a substantial role at the litter and duff composition
of the forest floor (Majidzade et al., 2015). Due to consistent management practices for
almost half a century, the two forests show significant differences in both canopy and
understory vegetation. Documented in Coates 2017, WS77 with the historical pine
dominated forest pattern had 84% of the basal area from pine species and 16% from
multiple hardwood species, while the unmanaged WS80 being undisturbed since 1964,
went through multiple stages of succession and have 41% pine and 59% hardwood in terms
of basal area proportions. Regarding tree density, WS77 have much lower basal area (33.72
m2/ha) compared to the unmanaged WS80 (46.35 m2/ha). Understory compositions were
not quantified based on taxonomy, but WS77 being constantly disturbed are dominated by
grass species like giant cane, and ferns. WS80 having a thick detritus layer on the ground
and being at a more advanced stage of succession, does not have as many grass species,
and understory is scattered with few shrub species such as American holly and old bay.
62
Being a direct result of prescribed fire management. The differences in vegetation density
and type did not only alter the composition and accumulation of forest detrital layer, but
also shifts the watershed hydrology into different patterns especially during the growing
season of the forests.
CONCLUSIONS
Despite common conception, there is no evidence from this study that support any
measurable direct or immediate negative influence on stream macroinvertebrate population
or community composition following a low-severity prescribed fire in these high stress
first-order watersheds. However, it can have moderate influence on biotic communities
indirectly by altering vegetation composition, soil properties, nutrient availability, and
general hydrology of the habitat. Small ephemeral water channels usually carry high stress
levels to the biota inhabiting them. Species that live in these ecosystems are adapted to
most environmental stresses and use them to out-compete less stress tolerant species and
avoids many potential predators that could not live with so much environmental changes.
Most water quality parameters reacted to forest management in the long term,
inorganic nutrient parameters such as orthophosphate and DOC had the most significant
fluctuations during the short period after fire and first few precipitation events. Other
chemical parameters such as DOC, dissolved nitrogen, calcium, or electrical conductivity
are statistically different between the two watersheds, but the scales of these differences
are low or irrelevant to the aquatic biota. We did not observe any significant difference in
corresponding macroinvertebrate abundance that correlated to the chemical measurements
63
(e.g. calcium difference did not result in significant differences in arthropod population).
It is most likely due to the overshadowing influences of other environmental stresses such
as pH and hydrology, or a lack of identification precision, we did not find many of these
chemical parameters significant enough to cause differences in aquatic invertebrate
community assemblages. Although we often observed temporal differences in community
between the two watersheds, if we combine the data from the whole sampling year, both
watersheds share similar macroinvertebrate species richness, abundance, and diversity.
According to our combined NMDS graphs (Figure 3.7), few samples lies out of the
overlapped area of these two watersheds, and the centroid of the two graphs are very close
together, meaning that these two watersheds have similar species pool, and similar overall
communities. They only respond to temporal stresses such as seasonal hydrological cycles
differently as an indirect influence of forest management.
Communities are influenced by many factors, and without being able to control
every variable, our research here at these two watersheds provides further information
about what to generally expect out of the aquatic biota when a forest is frequently managed
by prescribed fire in the Southeastern Coastal Plains. Given the low severity of prescribed
fire on the Coastal Plain and relatively flat topography, increased runoff and erosion
seldom becomes a major concern. However, our results shown that, even after over one
month of dry period with almost no precipitation, the first high intensity rainfalls post-burn
can still introduce elevated DOC and nutrients into the downstream water and adversely
affect water quality.
64
REFERENCES
Amatya, D., & Trettin, C. (2007). Development of Watershed Hydrologic Research at
Santee Experimental Forest, Coastal South Carolina. Usda, (Figure 1), 18–22.
Arkle, R. S., & Pilliod, D. S. (2010). Prescribed fires as ecological surrogates for
wildfires: A stream and riparian perspective. Forest Ecology and Management,
259(5), 893–903. https://doi.org/10.1016/j.foreco.2009.11.029
Arocena, J. M., & Opio, C. (2003). Prescribed fire-induced changes in properties of sub-
boreal forest soils. Geoderma, 113(1–2), 1–16. https://doi.org/10.1016/S0016-
7061(02)00312-9
Baird, M., Zabowski, D., & Everett, R. L. (1999). Wildfire effects on carbon and nitrogen
in inland coniferous forests. Plant and Soil, 209(2), 233–243.
https://doi.org/10.1023/A:1004602408717
Baker, M. B. J. (1988). Hydrologic and water quality effects of fire. Effects of Fire
Management of Southwestern Natural Resources, 31–42. Retrieved from AZTNC
Science Files
Battle, J., & Golladay, S. W. (2003). Prescribed Fire’s Impact on Water Quality of
Depressional Wetlands in Southwestern Georgia. The American Midland Naturalist,
150(1), 15–25. https://doi.org/10.1674/0003-
0031(2003)150[0015:PFIOWQ]2.0.CO;2
Bêche, L. A., Stephens, S. L., & Resh, V. H. (2005). Effects of prescribed fire on a Sierra
Nevada (California, USA) stream and its riparian zone. Forest Ecology and
Management, 218(1–3), 37–59. https://doi.org/10.1016/j.foreco.2005.06.010
Binkley, D., & Fisher, R. (2013). Ecology and management of forest soils. Ecology and
management of forest soils. https://doi.org/10.1007/s11104-010-0391-5
Bladon, K. D., Emelko, M. B., Silins, U., & Stone, M. (2014). Wildfire and the Future of
Water Supply. Environmental Science & Technology, 48(16), 8936–8943.
https://doi.org/10.1021/es500130g
Blomqvist, S., Gunnars, A., & Elmgren, R. (2004). Why the limiting nutrient differs
between temperate coastal seas and freshwater lakes: A matter of salt. Limnology
and Oceanography, 49(6), 2236–2241. https://doi.org/10.4319/lo.2004.49.6.2236
Britton, D. L. (1991a). Fire and the chemistry of a South African mountain stream.
Hydrobiologia, 218(3), 177–192. https://doi.org/10.1007/BF00038834
Britton, D. L. (1991b). THE BENTHIC MACROINVERTEBRATE FAUNA OF A
SOUTH AFRICAN MOUNTAIN STREAM AND ITS RESPONSE TO FIRE.
Southern African Journal of Aquatic Sciences, 17(1–2), 51–64.
https://doi.org/10.1080/10183469.1991.9631313
Brown, J. K. (1974). Handbook for Inventorying Downed Woody Material. USDA For.
Serv. Gen. Tech. Rep., 24. https://doi.org/10.1017/CBO9781107415324.004
65
Brown, L. E., Holden, J., Palmer, S. M., Johnston, K., Sorain, J. R., & Grayson, R.
(2015). Effects of fire on the hydrology, biogeochemistry, and ecology of peatland
river systems. Freshwater Science, 34(4), 1406–1425.
https://doi.org/10.1086/683426.
Brown, L. E., Johnston, K., Palmer, S. M., Aspray, K. L., & Holden, J. (2013). River
ecosystem response to prescribed vegetation burning on Blanket Peatland. PloS One,
8(11), e81023. https://doi.org/10.1371/journal.pone.0081023
Callaham, M. A., Scott, D. A., O’Brien, J. J., & Stanturf, J. A. (2012). Cumulative
Effects of Fuel Management on the Soils of Eastern U . S . Watersheds. Cumulative
Watershed Effects of Fuel Management in the Eastern United States, 202–228.
Retrieved from http://treesearch.fs.fed.us/pubs/41099
Certini, G. (2005). Effects of Fire on Properties of Forest Soils : A Review. Oecologia,
143(1), 1–10. https://doi.org/10.2307/20062214
Coates, T. A. (2017). Forest Management in Coastal Pine Forests : An Investigation of
Prescribed Fire Behavior , Detrital Chemical Composition , and Potential Water
Quality Impacts. Clemson University.
Dai, Z., Trettin, C. C., & Amatya, D. M. (2013). Effects of Climate Variability on Forest
Hydrology and Carbon Sequestration on the Santee Experimental Forest in Coastal
South Carolina, 1–44. Retrieved from
http://www.dnr.sc.gov/ccworkshops/pdf/gtr_srs72.pdf
Debano, L. F. (2000). The role of fire and soil heating on water repellancy in wildland
environments: a review. Journal of Hydrology, 231–232, 195–206.
https://doi.org/10.1016/S0022-1694(00)00194-3
DeBano, L. F., & Klopatek, J. M. (1988). Phosphorus Dynamics of Pinyon-Juniper Soils
following Simulated Burning. Soil Science Society of America Journal, 52(1), 271–
277. https://doi.org/10.2136/sssaj1988.03615995005200010048x
Dickens, E.D.; Barnett, J.P.; Hubbard, W.G.; Jokela, E.J., eds. 2004. Slash pine: still
growing and growing! Proceedings of the slash pine symposium. Gen. Tech. Rep.
SRS-76. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern
Research Station. 148 p.
Douglas, M. M., Setterfield, S. A., McGuinness, K., & Lake, P. S. (2015). The impact of
fire on riparian vegetation in Australia’s tropical savanna. Freshwater Science,
34(4), 1351–1365. https://doi.org/10.1086/684074
Elliott, K. J., L. Hendrick, R., Major, A. E., Vose, J. M., & Swank, W. T. (1999).
Vegetation dynamics after a prescribed fire in the southern Appalachians. Forest
Ecology and Management, 114(2–3), 199–213. https://doi.org/10.1016/S0378-
1127(98)00351-X
Fox, T. R., Jokela, E. J., & Allen, H. L. (2007). The Development of Pine Plantation
Silviculture in the Southern United States. Journal of Forestry, 105(November),
66
337–347. Retrieved from
http://www.ingentaconnect.com/content/saf/jof/2007/00000105/00000007/art00005
Hall, S. J., & Lombardozzi, D. (2008). Short-term Effects of Wildfire on Montane Stream
Ecosystems in the Southern Rocky Mountains: One and Two Years Post-burn.
Western North American Naturalist, 68(4), 453–462. https://doi.org/10.3398/1527-
0904-68.4.453
Hart, S. C., DeLuca, T. H., Newman, G. S., MacKenzie, M. D., & Boyle, S. I. (2005).
Post-fire vegetative dynamics as drivers of microbial community structure and
function in forest soils. Forest Ecology and Management, 220(1–3), 166–184.
https://doi.org/10.1016/j.foreco.2005.08.012
Hernández, T., García, C., & Reinhardt, I. (1997). Short-term effect of wildfire on the
chemical, biochemical and microbiological properties of Mediterranean pine forest
soils. Biology and Fertility of Soils, 25(2), 109–116.
https://doi.org/10.1007/s003740050289
Kush, J., & Varner, J. M. (2009). Restoring Fire to the Longleaf Pine Forest. Fire Science
Brief, (30), 1–6. Retrieved from http://www.firescience.gov/projects/briefs/01B-3-1-
01_FSBrief30.pdf
Lammert, M., & Allan, J. (1999). ENVIRONMENTAL AUDITING: Assessing Biotic
Integrity of Streams: Effects of Scale in Measuring the Influence of Land Use/Cover
and Habitat Structure on Fish and Macroinvertebrates. Environmental Management,
23(2), 257–270. https://doi.org/10.1007/s002679900184
Majidzade, H., Wang, J., & Chow, A. T. (2015). Organic Matter and Disinfection By-
Product Precursor in Forested Watersheds – Part II . A Controlled Field Study. In In
Recent Advances in Disinfection By-Products (pp. 271–292).
Meixner, T., & Wohlgemuth, P. (2004). Wildfire Impacts on Water Quality. Southwest
Hydrology, (October), 24–25.
Minshall, G. W., Brock, J. T., Andrews, D. A., & Robinson, C. T. (2001). Water quality,
substratum and biotic responses of five central Idaho (USA) streams during the first
year following the Mortar Creek fire. International Journal of Wildland Fire, 10(2),
185–199. https://doi.org/10.1071/WF01017
Murphy, J. D., Johnson, D. W., Miller, W. W., Walker, R. F., & Blank, R. R. (2006).
PRESCRIBED FIRE EFFECTS ON FOREST FLOOR AND SOIL NUTRIENTS
IN A SIERRA NEVADA FOREST. Soil Science, 171(3), 181–199.
https://doi.org/10.1097/01.ss.0000193886.35336.d8
Peet, R., & Allard, D. (1993). Longleaf pine vegetation of the southern Atlantic and
eastern Gulf Coast regions: a preliminary classification. Proceedings of the Tall
Timbers Fire Ecology Conference, (18), 45–81. https://doi.org/10.1086/284833
Peuravuori, J., & Pihlaja, K. (1997). Molecular size distribution and spectroscopic
properties of aquatic humic substances. Analytica Chimica Acta, 337(2), 133–149.
67
https://doi.org/10.1016/S0003-2670(96)00412-6
Revchuk, A. D., & Suffet, I. H. (Mel). (2014). Effect of Wildfires on Physicochemical
Changes of Watershed Dissolved Organic Matter. Water Environment Research,
86(4), 372–381. https://doi.org/10.2175/106143013X13736496909671
Richter, D. D., Ralston, C. W., & Harms, W. R. (1982). Prescribed Fire: Effects on Water
Quality and Forest Nutrient Cycling. Science, 215(4533), 661–663. Retrieved from
http://www.jstor.org.libproxy.clemson.edu/stable/1688204
Schindler, D.W., R.W. Newbury, K.G. Beaty, J. Prokopowich, T. Ruszczynski, and J.A.
Dalton. 1980. Effects of a windstorm and forest fire on chemical losses from
forested watersheds and on the quality of receiving stream. Can. J. Fish. Aquat. Sci.
37: 328-334.
Shu-ren, Y. (2003). Effects of fire disturbance on forest hydrology. Journal of Forestry
Research, 14(4), 331–334. https://doi.org/10.1007/BF02857863
Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011a). Wildfire
effects on water quality in forest catchments: A review with implications for water
supply. Journal of Hydrology, 396(1–2), 170–192.
https://doi.org/10.1016/j.jhydrol.2010.10.043
Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011b). Wildfire
effects on water quality in forest catchments: A review with implications for water
supply. Journal of Hydrology, 396(1–2), 170–192.
https://doi.org/10.1016/j.jhydrol.2010.10.043
Stephens, S. L., Meixner, T., Poth, M., McGurk, B., & Payne, D. (2004). Prescribed fire,
soils, and stream water chemistry in a watershed in the Lake Tahoe Basin,
California. International Journal of Wildland Fire, 13(1), 27–35.
https://doi.org/10.1071/WF03002
T.D., S., Webb, J. . W., & Crossley, D. A. J. (1981). Communtiy Structure and Nutrient
Content of Canopy Arthropods in Clearcut and Uncut Forest Ecosystems Author ( s
): T . D . Schowalter , J . Warren Webb , D . A . Crossley , Jr . Published by :
Ecological Society of America Stable URL : http://www.jstor.o. Ecology, 62(4),
1010–1019.
Ursic, S. J. (1970). Hydrologic Effects of Prescribed Burning and Deadening Upland
Hardwoods in Northern Mississippi. Research Papers. Southern Forest Experiment
Station, (SO-54), 15. Retrieved from
http://www.srs.fs.usda.gov/pubs/rp/rp_so054.pdf
Van Lear, D. H., Carroll, W. D., Kapeluck, P. R., & Johnson, R. (2005). History and
restoration of the longleaf pine-grassland ecosystem: Implications for species at risk.
Forest Ecology and Management, 211(1–2), 150–165.
https://doi.org/10.1016/j.foreco.2005.02.014
Van Lear, D. H., Douglass, J. E., Cox, S. K., & Augspurger, M. K. (1985). Sediment and
68
Nutrient Export in Runoff from Burned and Harvested Pine Watersheds in the South
Carolina Piedmont1. Journal of Environment Quality, 14(2), 169.
https://doi.org/10.2134/jeq1985.00472425001400020003x
Van Mantgem, P. J., Schwartz, M., & Keifer, M. (2001). Monitoring fire effects for
managed burns and wildfires: Coming to terms with pseudoreplication. Natural
Areas Journal, 21(3), 266–273.
Waldrop, T. A., & Goodrick, S. L. (2012). Introduction to prescribed fires in Southern
ecosystems. Science Update SRS-054, (August), 80. Retrieved from
http://www.srs.fs.usda.gov/pubs/su/su_srs054.pdf
Walker, J. (1993). Rare vascular plant taxa associated with the longleaf pine ecosystems:
Patterns in taxonomy and ecology. In Proceedings of the Tall Timbers Fire Ecology
Conference (Vol. 18, pp. 105–126).
Wang, J. J., Dahlgren, R. A., & Chow, A. T. (2015). Controlled Burning of Forest
Detritus Altering Spectroscopic Characteristics and Chlorine Reactivity of Dissolved
Organic Matter: Effects of Temperature and Oxygen Availability. Environmental
Science and Technology, 49(24), 14019–14027.
https://doi.org/10.1021/acs.est.5b03961
Wang, Q., Zhong, M., & Wang, S. (2012). A meta-analysis on the response of microbial
biomass, dissolved organic matter, respiration, and N mineralization in mineral soil
to fire in forest ecosystems. Forest Ecology and Management, 271, 91–97.
https://doi.org/10.1016/j.foreco.2012.02.006
White, J. C., House, A., Punchard, N., Hannah, D. M., Wilding, N. A., & Wood, P. J.
(2016). Macroinvertebrate community responses to hydrological controls and
groundwater abstraction effects across intermittent and perennial headwater streams.
Science of the Total Environment, (2016).
https://doi.org/10.1016/j.scitotenv.2017.06.081
69
Figure 3.1. Map of the Santee Experimental watersheds (WS77 managed, WS80
unmanaged control). Blue lines indicate the main water channels within the watersheds
70
Figure 3.2. Discharge rates at the downstream watershed gauging stations 2005
0
20000
40000
60000
80000
100000
120000
WS77 discharge (m^3)
WS80 discharge (m^3)
71
(a)
(b)
Figure 3.3. Weather patterns of the study site during sampling year, air temperature
every 15 minutes (a), and daily total precipitation (b).
-10
-5
0
5
10
15
20
25
30
35
40
45
3/24/2016 5/23/2016 7/22/2016 9/20/2016 11/19/2016 1/18/2017 3/19/2017
Air
tem
per
ature
℃
0
50
100
150
200
250
300
3/24/2016 5/23/2016 7/22/2016 9/20/2016 11/19/2016 1/18/2017 3/19/2017
Dai
ly r
ain (
mm
)
Hurricane
Matthew
72
Figure 3.4. DOC concentrations of the watersheds (WS77 treatment, WS80 control)
Jan Mar May Jul Sep Nov
10
15
20
25
30
35
40
45
50
DO
C (
mg
/L)
WS77
WS80
73
Figure 3.5. Orthophosphate concentration of the watersheds (WS77 treatment, WS80
control)
Jan Apr Jul Oct Jan
0.0
0.1
0.2
0.3
0.4
Ort
ho
ph
osp
hat
e (m
g/L
) WS77
WS80
74
Figure 3.6. Nonmetric multidimensional scaling ordinations for macroinvertebrate
communities of the two watersheds (WS77 treatment, WS80 control) at different times
across the whole sampling year. Time periods: (a) Pre-burn samples (March – April
2016); (b) – (g) post-burn samples from April 2016 to March 2017. Each graph
represents three sampling rounds.
75
Figure 3.7. Nonmetric multidimensional scaling ordinations for macroinvertebrate
communities for the whole post fire period (left), and whole sampling year (right).
76
Figure 3.8. Shannon-Wiener diversity index (top) and Shannon’s equitability (bottom)
of the two watersheds compared based on six classes of macroinvertebrates.
Disconnected points are periods of drought where sampling sites dried up completely.
Feb Apr Jun Aug Oct Dec Feb Apr
0.6
0.8
1.0
1.2
1.4
1.6
WS77 H'
WS80 H'
Feb Apr Jun Aug Oct Dec Feb Apr
0.4
0.5
0.6
0.7
0.8
0.9
WS
77 S
E
WS77 SE
WS80 SE
77
Figure 3.9. Total macroinvertebrate abundance per sample comparing the two
watersheds for the whole sampling year.
Feb Apr Jun Aug Oct Dec Feb Apr
0
20
40
60
80
100
120 WS77 abundance
WS80 abundance
78
Figure 3.10. Average water depth of the sampling site at each time of invertebrate
sampling, unconnected dots indicate dried up periods
Feb Apr Jun Aug Oct Dec Feb Apr
0
10
20
30
40
50W
ate
r D
epth
(cm
)
WS77 Depth
WS80 Depth
79
(a)
(b)
Figure 3.11. Observed algal bloom one month after prescribed burn, before flow was
restored to the watersheds. Burned watershed 77 (a), unmanaged watershed 80 (b).
80
CHAPTER IV
CONCLUSIONS
From our field incubation study of different forest ground litter, burning did not
necessary increase the pH of the acidic rainwater, but did prevent it from dropping to
acidity levels of the leachate from unmanaged forest. In conjunction with less nutrition
output and considerably lower DOC output by prescribed burning, it can be considered as
a way to improve water quality of the headwaters in the Southern Coastal Plain forests.
Most concerned issue with burning on water quality and aquatic ecosystems is increased
runoff and erosion (Smith et al., 2011a). Such a pattern was not observed by our litter
incubation, and on the contrary, forest ground litter can be just as mobile if not more in
forests that without management. However, by reducing the detrital layer thickness,
prescribed burning does have the potential to cause measurable increases in runoff in
areas of different topography (e.g. upland forests with significant slopes). USDA Forest
Service is aware of the concerns on erosion and runoff, and in theory needs to be more
careful to conduct such an approach in areas with considerable slopes when conducting
burns (Waldrop & Goodrick, 2012).
Field collected water samples from our second study on the paired watersheds
also did not express any significant water quality concerns, prescribed fire not only did
not reduce water quality, but in a way improved certain water quality parameters for the
aquatic ecosystem such as considerable lower dissolved nutrients. These results
combined with the tray incubation study, confirms that by regular prescribed fire
management we are not adding risk to the aquatic ecosystems in the Southern pine
81
dominated open savanna, aside from a short nutrient increase following the fire that can
potentially cause algal bloom. Such a risk can be easily mitigated by conducting the
burning during a period of frequent rain event.
The field study was not replicated due to logistic reason, which is a commonly
recognized difficulty in fire-related researches (Bêche et al., 2005; Van Mantgem,
Schwartz, & Keifer, 2001). The combined biotic and abiotic parameters suggest that
prescribed burning in the Southern Coastal Plains does not have any acute impact on the
aquatic ecosystem, but by indirectly altering the terrestrial system, it indirectly changes
the hydrological patterns and thus have a small effect on the stream water quality and
biotic communities.
Ultimately, fire history, watershed condition, and management goals must be
carefully considered for each watershed to determine the potential effectiveness and
consequences of using prescribed fire. For the Southern pine dominated open savanna on
the Coastal Plains, prescribed burning historically has been, and to our present knowledge
still is a very effective method of management at mitigating risk of wildfire, ecosystem
preservation, improving wildlife habitat, maximizing timber growth, and enhancing
aesthetics. Our work on water quality and aquatic biota suggest that on top of the
standard management procedures to ensure prescribed fire stay on low to moderate
intensities, it will be more ideal to conduct the practice during a wet season where
frequent precipitation events are expected. With these considerations in mind, fire
management should not have any negative impact on the aquatic ecosystem and
ultimately can even improve water quality downstream.
82
REFERENCES
Bêche, L. A., Stephens, S. L., & Resh, V. H. (2005). Effects of prescribed fire on a Sierra
Nevada (California, USA) stream and its riparian zone. Forest Ecology and
Management, 218(1–3), 37–59. https://doi.org/10.1016/j.foreco.2005.06.010
Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011). Wildfire
effects on water quality in forest catchments: A review with implications for water
supply. Journal of Hydrology, 396(1–2), 170–192.
https://doi.org/10.1016/j.jhydrol.2010.10.043
Van Mantgem, P. J., Schwartz, M., & Keifer, M. (2001). Monitoring fire effects for
managed burns and wildfires: Coming to terms with pseudoreplication. Natural
Areas Journal, 21(3), 266–273.
Waldrop, T. A., & Goodrick, S. L. (2012). Introduction to prescribed fires in Southern
ecosystems. Science Update SRS-054, (August), 80.