effects of prescribed forest fire on water quality and

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

https://tigerprints.clemson.edu?utm_source=tigerprints.clemson.edu%2Fall_theses%2F2763&utm_medium=PDF&utm_campaign=PDFCoverPages

https://tigerprints.clemson.edu/all_theses?utm_source=tigerprints.clemson.edu%2Fall_theses%2F2763&utm_medium=PDF&utm_campaign=PDFCoverPages

https://tigerprints.clemson.edu/theses?utm_source=tigerprints.clemson.edu%2Fall_theses%2F2763&utm_medium=PDF&utm_campaign=PDFCoverPages

https://tigerprints.clemson.edu/all_theses?utm_source=tigerprints.clemson.edu%2Fall_theses%2F2763&utm_medium=PDF&utm_campaign=PDFCoverPages

https://tigerprints.clemson.edu/all_theses/2763?utm_source=tigerprints.clemson.edu%2Fall_theses%2F2763&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

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.




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




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






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




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.



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



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.


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


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




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




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




26


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.


ecosystems. Science Update SRS-054, (August), 80. Retrieved from

http://www.srs.fs.usda.gov/pubs/su/su_srs054.pdf




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.


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.




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




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






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




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.



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



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.


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


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




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




Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011b). Wildfire




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., 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





ecosystems. Science Update SRS-054, (August), 80. Retrieved from

http://www.srs.fs.usda.gov/pubs/su/su_srs054.pdf




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.


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'


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.


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


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




Smith, H. G., Sheridan, G. J., Lane, P. N. J., Nyman, P., & Haydon, S. (2011). Wildfire








ecosystems. Science Update SRS-054, (August), 80.

effects of prescribed forest fire on water quality and

Documents