UNDERSTANDING THE ROLE OF THE STREAM-GROUNDWATER INTERFACE IN
DISSOLVED ORGANIC CARBON CYCLING IN A THIRD-ORDER,
LOWLAND RIVER NETWORK
By
Sydney S. Ruhala

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Geological Sciences – Master of Science
2017

ABSTRACT

UNDERSTANDING THE ROLE OF THE STREAM-GROUNDWATER INTERFACE IN
DISSOLVED ORGANIC CARBON CYCLING IN A THIRD-ORDER,
LOWLAND RIVER NETWORK
By
Sydney S. Ruhala
Global carbon models currently underrepresent the contribution of carbon dioxide (CO2)
emissions from stream channels to the atmosphere since streams have traditionally been viewed
as passive drainage systems for terrestrial landscapes. However, recent studies have shown that
rivers are a significant source of CO2 (a greenhouse gas) to the atmosphere, indicating that much
of the organic carbon entering rivers (predominantly as dissolved organic carbon (DOC))
experiences significant in-stream biogeochemical degradation. The zone beneath and alongside
of the stream where stream water-groundwater interactions occur (i.e., the hyporheic zone, HZ)
may be an important location within river networks for DOC processing (e.g., degradation,
consumption), since it is a hotspot for microbial activity and biogeochemical reactions. This
thesis assesses the role of the HZ (e.g., source vs. sink of carbon) in DOC cycling in a thirdorder, lowland watershed in Michigan, USA and represents one of the first watershed-scale HZ
sampling efforts. Our findings indicate that the HZ is a location for DOC processing in stream
networks and that the processing signal is observable at the watershed scale. The HZ in this
small, groundwater-fed stream network likely acts as a sink, consuming DOC via aerobic,
microbial respiration. In addition, our analysis suggests that the HZ may act as a larger sink for
DOC in headwater streams than in higher stream orders in this river network.

Copyright by
SYDNEY S. RUHALA
2017

ACKNOWLEDGMENTS

This project was partially funded by a Geological Society of America Graduate Student
Research Grant as well as by the NSF Long-term Ecological Research Program (DEB 1027253)
at the Kellogg Biological Station and by Michigan State University AgBioResearch. I would like
to thank the Department of Earth and Environmental Sciences for their support as well as my coadvisors, Dr. David T. Long and Dr. Jay P. Zarnetske and my committee member, Dr. Stephen
K. Hamilton, for their advice and guidance.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ vii
LIST OF FIGURES ..................................................................................................................... viii
KEY TO ABBREVIATIONS ....................................................................................................... xii
CHAPTER 1: INTRODUCTION ....................................................................................................1
CHAPTER 2: A PRELIMINARY STUDY OF TRENDS IN DISSOLVED ORGANIC
CARBON QUANTITY AND QUALITY ACROSS THE STREAM-GROUNDWATER
INTERFACE OF A THIRD-ORDER, LOWLAND RIVER NETWORK .....................................6
2.1 Introduction ..................................................................................................................6
2.2 Background ..................................................................................................................8
2.2.1 Field Site Description ..................................................................................8
2.2.2 Hydrology ....................................................................................................9
2.3 Methods .......................................................................................................................12
2.3.1 Sample Collection ......................................................................................12
2.3.2 Analytical Methods ....................................................................................15
2.3.3 Optical Indices ...........................................................................................16
2.3.4 Vertical Porewater Profiles .......................................................................17
2.4 Results .........................................................................................................................18
2.4.1 Results of the Vertical Porewater Profiles.................................................18
2.4.2 DOC Variance and Mixing Zone Model ....................................................21
2.5 Discussion....................................................................................................................26
2.6 Conclusions .................................................................................................................30
CHAPTER 3: DETERMINGING HOW DISSOLVED ORGANIC CARBON IS
TRANSFORMED AT THE STREAM-GROUNDWATER INTERFACE ACROSS A THIRDORDER, LOWLAND RIVER NETWORK ..................................................................................31
3.1 Introduction ................................................................................................................31
3.2 Background ................................................................................................................35
3.2.1 Field Site Description ................................................................................35
3.2.2 Hydrology ..................................................................................................36
3.3 Methods .......................................................................................................................38
3.3.1 Sample Collection ......................................................................................38
3.3.2 Analytical Methods ....................................................................................42
3.3.3 Optical Indices ...........................................................................................42
3.3.4 Vertical Porewater Profiles .......................................................................43
3.4 Results .........................................................................................................................44
3.4.1 Composite Vertical Porewater Profiles ....................................................44
3.4.2 Rates of Change in DOC Quantity and Quality with Depth ......................47
3.4.3 DOC Metabolism .......................................................................................52
v

3.5 Discussion....................................................................................................................55
3.6 Conclusions .................................................................................................................62
CHAPTER 4: USING IN-SITU OPTICAL SENSORS TO STUDY DISSOLVED ORGANIC
CARBON DYNAMICS OF STREAMS AND WATERSHEDS: A REVIEW ............................63
4.1 Abstract .......................................................................................................................63
4.2 Introduction ................................................................................................................64
4.3 Spectroscopy as a Diagnostic Tool for DOC............................................................67
4.3.1 Absorbance Spectroscopy ..........................................................................67
4.3.2 Fluorescence Spectroscopy ........................................................................68
4.4 Technology and Limitations .....................................................................................71
4.4.1 In-Situ Optical Instruments ........................................................................71
4.4.2 Effects of Temperature on Sensors ............................................................75
4.4.3 Effects of Turbidity on Sensors ..................................................................75
4.5 Examples of Utility in Riverine DOC Studies .........................................................79
4.5.1 Revealing Seasonal, Event, and Diurnal Variations in DOC Quantity and
Quality .......................................................................................................79
4.5.2 Revealing Sources and Pathways of DOC to the Stream Channel ............84
4.5.3 Annual DOC Flux ......................................................................................87
4.6 Conclusions and Future Considerations ..................................................................90
4.7 Acknowledgments ......................................................................................................91
CHAPTER 5: SYNTHESIS & IMPLICATIONS .........................................................................92
APPENDICES ...............................................................................................................................95
APPENDIX A 2015 Synoptic Hyporheic Zone Sampling Data ........................................96
APPENDIX B 2016 Synoptic Hyporheic Zone Sampling Data ......................................123
REFERENCES ............................................................................................................................164

vi

LIST OF TABLES

Table 1: Summary of optical DOC indices ....................................................................................17
Table 2: Summary of the rate constants for DOC concentration and SUVA254 ............................48
Table 3: Summary of the known strengths and limitations of in-situ optical instruments ............78
Table A1: Supplemental information for the 2015 sampling sites ................................................97
Table A2: Synoptic sampling data for 2015 ................................................................................115
Table B1: Supplemental information for the 2016 sampling sites ..............................................124

vii

LIST OF FIGURES

Figure 1: Conceptual model of the hyporheic zone .........................................................................5
Figure 2: Topographic map of Augusta Creek watershed .............................................................10
Figure 3: 2015 USGS hydrographs for Augusta Creek .................................................................11
Figure 4: Map of Augusta Creek sampling sites in 2015...............................................................14
Figure 5: Sampling scheme ............................................................................................................15
Figure 6: Example of a vertical porewater profile .........................................................................17
Figure 7: Vertical porewater profiles of DOC concentration and optical indices .........................20
Figure 8: Model used to identify mixing zone depths ...................................................................22
Figure 9: DOC concentration and optical indices within the mixing zone ....................................24
Figure 10: 2016 USGS hydrographs for Augusta Creek ...............................................................37
Figure 11: Map of Augusta Creek sampling sites in 2016.............................................................40
Figure 12: Subplots of labeled sampling sites across Augusta Creek ...........................................41
Figure 13: Composite vertical porewater profiles .........................................................................46
Figure 14: Using a first-order rate law to determine rate constants ...............................................47
Figure 15: Changes in rate constants and VHG across the stream network ..................................51
Figure 16: Example of a metabolism plot ......................................................................................52
Figure 17: Metabolism plots for Augusta Creek............................................................................54
Figure 18: Plots of HZ function .....................................................................................................60
Figure 19: A simplified Jablonski diagram ....................................................................................69
Figure 20: Plots of DOC concentration versus time using a variation of sampling frequencies ...72
Figure 21: In-situ optical instruments ............................................................................................73
Figure 22: Plot of discharge and FDOM concentration versus time..............................................77
viii

Figure 23: Diurnal trends for CDOM and absorbance...................................................................84
Figure 24: Hysteresis trends...........................................................................................................87
Figure 25: DOC flux versus discharge...........................................................................................89
Figure A1: Map of Augusta Creek watershed with land use .........................................................98
Figure A2: Site A Vertical Porewater Profiles ..............................................................................99
Figure A3: Site B Vertical Porewater Profiles .............................................................................100
Figure A4: Site C Vertical Porewater Profiles .............................................................................101
Figure A5: Site D Vertical Porewater Profiles ............................................................................102
Figure A6: Site F Vertical Porewater Profiles .............................................................................103
Figure A7: Site G Vertical Porewater Profiles ............................................................................104
Figure A8: Site H Vertical Porewater Profiles ............................................................................105
Figure A9: Site I Vertical Porewater Profiles ..............................................................................106
Figure A10: Site J Vertical Porewater Profiles ............................................................................107
Figure A11: Site K Vertical Porewater Profiles ..........................................................................108
Figure A12: Site L Vertical Porewater Profiles ...........................................................................109
Figure A13: Site M Vertical Porewater Profiles ..........................................................................110
Figure A14: Site O Vertical Porewater Profiles ..........................................................................111
Figure A15: Site P Vertical Porewater Profiles ...........................................................................112
Figure A16: Site Q Vertical Porewater Profiles ..........................................................................113
Figure A17: Site R Vertical Porewater Profiles ...........................................................................114
Figure B1: Site A1 Vertical Porewater Profiles ...........................................................................125
Figure B2: Site A2 Vertical Porewater Profiles ...........................................................................126
Figure B3: Site A3 Vertical Porewater Profiles ...........................................................................127
Figure B4: Site B1 Vertical Porewater Profiles ...........................................................................128
ix

Figure B5: Site B2 Vertical Porewater Profiles ...........................................................................129
Figure B6: Site C1 Vertical Porewater Profiles ...........................................................................130
Figure B7: Site C2 Vertical Porewater Profiles ...........................................................................131
Figure B8: Site C3 Vertical Porewater Profiles ...........................................................................132
Figure B9: Site C4 Vertical Porewater Profiles ...........................................................................133
Figure B10: Site E1 Vertical Porewater Profiles .........................................................................134
Figure B11: Site E2 Vertical Porewater Profiles .........................................................................135
Figure B12: Site E3 Vertical Porewater Profiles .........................................................................136
Figure B13: Site F1 Vertical Porewater Profiles .........................................................................137
Figure B14: Site F2 Vertical Porewater Profiles .........................................................................138
Figure B15: Site F3 Vertical Porewater Profiles .........................................................................139
Figure B16: Site F4 Vertical Porewater Profiles .........................................................................140
Figure B17: Site G1 Vertical Porewater Profiles .........................................................................141
Figure B18: Site G2 Vertical Porewater Profiles .........................................................................142
Figure B19: Site H1 Vertical Porewater Profiles .........................................................................143
Figure B20: Site H2 Vertical Porewater Profiles .........................................................................144
Figure B21: Site H3 Vertical Porewater Profiles .........................................................................145
Figure B22: Site I1 Vertical Porewater Profiles ..........................................................................146
Figure B23: Site I2 Vertical Porewater Profiles ..........................................................................147
Figure B24: Site J1 Vertical Porewater Profiles ..........................................................................148
Figure B25: Site K1 Vertical Porewater Profiles .........................................................................149
Figure B26: Site K2 Vertical Porewater Profiles .........................................................................150
Figure B27: Site L1 Vertical Porewater Profiles .........................................................................151

x

Figure B28: Site L2 Vertical Porewater Profiles .........................................................................152
Figure B29: Site L3 Vertical Porewater Profiles .........................................................................153
Figure B30: Site M1 Vertical Porewater Profiles ........................................................................154
Figure B31: Site M2 Vertical Porewater Profiles ........................................................................155
Figure B32: Site O1 Vertical Porewater Profiles .........................................................................156
Figure B33: Site O2 Vertical Porewater Profiles .........................................................................157
Figure B34: Site R1 Vertical Porewater Profiles .........................................................................158
Figure B35: Site R2 Vertical Porewater Profiles .........................................................................159
Figure B36: Site R3 Vertical Porewater Profiles .........................................................................160
Figure B37: Site R4 Vertical Porewater Profiles .........................................................................161
Figure B38: Site S1 Vertical Porewater Profiles .........................................................................162
Figure B39: Site S2 Vertical Porewater Profiles .........................................................................163

xi

KEY TO ABBREVIATIONS

C – Carbon
CO2 – Carbon dioxide
DOM – Dissolved Organic Matter
POC – Particulate Organic Carbon
DOC – Dissolved Organic Carbon
NPOC – Non-Purgeable Organic Carbon
Cl- – Chloride
NO3- – Nitrate
SO42- – Sulfate
CaCO3 – Calcium carbonate
HCO3- – Bicarbonate
SW-GW – Stream Water-Groundwater
HZ – Hyporheic Zone
KBS – Kellogg Biological Station
USGS – United States Geological Survey
°C – Degrees Celsius
DO – Dissolved Oxygen
HDPE – High Density Polyethylene
US EPA – United States Environmental Protection Agency
SUVA254 – Specific Ultraviolet Absorbance at 254 nm
UV-VIS – Ultraviolet-Visible

xii

λ – Wavelength
SD – Standard Deviation
VHG – Vertical Head Gradient
CDOM – Chromophoric fraction of DOM
FDOM – Fluorescent DOM
QSE – Quinine Sulfate Equivalents
LED – Light Emitting Diode
FI – Fluorescence Index

xiii

CHAPTER 1: INTRODUCTION

Understanding the global carbon (C) cycle, including its dominant sources and sinks and
associated fluxes, is of critical importance to mitigating future climate change. Previous models
have underrepresented the emissions of carbon dioxide (CO2) from stream channels to the
atmosphere since streams have traditionally been viewed as passive drainage systems for
terrestrial landscapes (Boulton et al., 1998; Battin et al., 2009; Boano et al., 2014). However,
recent studies revealed that streams exhibit strong, net heterotrophy (i.e., community respiration
considerably exceeds gross primary production), globally generating 1.8 ± 0.25 Pg CO2 each
year - a rate 6 times greater than lakes and reservoirs combined (Battin et al. 2008; Butman and
Raymond, 2011; Raymond et al., 2013). This indicates that rivers* are a significant source of
CO2 (a greenhouse gas) to the atmosphere and that much of the organic C entering rivers is, in
fact, undergoing significant in-stream biogeochemical degradation (Battin et al., 2008). Thus,
further study of C cycling in rivers is warranted, so that rivers can be more accurately
represented in global C models.
Carbon in natural waters can exist is several different forms: 1) as particulate organic
carbon (POC) such as woody debris and leaf litter, 2) as dissolved organic carbon (DOC) and 3)
as dissolved inorganic carbon including CO2 and bicarbonate (HCO3-; Cory et al., 2011). Here,
we focus on DOC since it is often the major form of organic C in streams and is the most
susceptible to respiratory conversion to CO2 and subsequent emission (e.g., Aiken, 2014). DOC
is operationally defined as any organic C that is able to pass through a filter of a specified pore
size (e.g., typically 0.7 µm, 0.45 µm or smaller), and can include some colloidal matter as well as
dissolved compounds. The composition of DOC is complex, consisting of a mixture of water*

For the purposes of this thesis, the terms “river(s)” and “stream(s)” are used interchangeably.

1

soluble, organic molecules with varying chemical structures (e.g., molecular weight and
aromaticity), which is largely dependent on its source material. For example, microbially- and
algal-derived DOC that is produced within the stream (i.e., autochthonous DOC) tends to have a
lower molecular weight and be less aromatic than terrestrially-derived DOC from vascular plant
sources (i.e., allochthonous DOC; McKnight et al., 2001; McDonald et al., 2004; Fellman et al.,
2010). Some general processes that produce DOC include organic matter leaching and
decomposition, chemical degradation, and exudates from primary production (e.g., Thurman,
1985; Cory et al., 2011; Jollymore et al., 2012).
DOC plays several critical roles in stream C cycling (Cory et al., 2011). For instance,
DOC produces CO2 when remineralized, fuels stream metabolism, and regulates additional
biogeochemical cycles (e.g., nitrogen transformations, trace metal complexation; Aiken, 2014).
As a result, stream DOC dynamics influence C cycling on local to global scales (Battin et al.,
2009). Despite its importance, the nature of in-stream DOC processing, including rates, pathways
and transformations, remains inadequately quantified (e.g., Sobczak and Findlay, 2002). Thus,
the mechanisms of stream DOC processing need to be investigated including, 1) where DOC
processing occurs within rivers, 2) the dominant biogeochemical transformation processes, and
3) how those processes affect the fate of DOC transported downstream (e.g., uptake via
microbial processes, remineralization). This knowledge will help improve the current
understanding of stream C cycling as well as help predict how streams and ultimately the global
C cycle will respond to changes in loading of DOC. For example, increases in stream DOC
transport have been occurring in parts of North America and northern Europe, for reasons that
remain unclear (Monteith et al., 2007; Laudon et al., 2011).

2

The zone beneath and alongside of the stream where stream water-groundwater (SWGW) interactions occur (i.e., the hyporheic zone, HZ) may be an important location within river
networks for DOC processing. The HZ is a hotspot for microbial activity and biogeochemical
reactions. (Figure 1; Storey et al., 1999; Baker et al., 1999; Fischer et al., 2005; Battin et al.,
2009; Zarnetske et al., 2011). This increased activity is the result of the mixing of two
chemically distinct waters (i.e., SW and GW), which creates a transitional zone in the subsurface
sediments, promoting both biological and chemical diversity (e.g., microbial communities,
metabolic rates, chemical reactions; Valett et al., 1996; Hedin et al., 1998; Findlay and Sobczak,
2000; Nogaro et al., 2013; Boano et al., 2014). In addition, residence times increase in the HZ
due to slow porewater transport, creating a longer exposure time of stream-borne constituents
(e.g., C, nutrients) to microbial communities, which facilitates reaction/metabolism (Battin et al.,
2008; Harvey et al., 2013; Boano et al., 2014). Due to the combination of high biological
diversity, strong chemical gradients, and increased residence times, it is expected that the HZ is a
significant processor of stream DOC.
Previous small-scale field (e.g., reach scale) and mesocosm studies have observed
decreasing DOC concentrations along hyporheic flowpaths, suggesting that the HZ is effective at
processing DOC and may act as an important sink for DOC in stream networks (Findlay et al.,
1993; Findlay and Sobczak, 1996; Schindler and Krabbenhoft, 1998; Sobczak and Findlay, 2002;
Zarnetske et al., 2011). Although consistent patterns in DOC quantity, in terms of concentrations,
have been documented along individual hyporheic flowpaths, few studies have thoroughly
characterized changes in DOC quality (i.e., molecular characteristics, composition) within the
HZ (Findlay and Sobczak, 1996). DOC quality is a critical variable to include in stream C studies
since the quality of DOC impacts its bioavailability and reactivity (i.e., lability), which

3

ultimately affects downstream ecosystems and water quality (Fellman et al., 2010; Cory et al.,
2011). Furthermore, DOC quality can change independently of DOC concentration and only a
relatively small fraction of total DOC can drive microbial metabolism (Lutz et al., 2012), thus
changes in DOC quantity alone do not fully reflect hyporheic DOC dynamics.
Hence, the overarching objectives of this thesis are to: 1) evaluate DOC quantity and
quality within the HZ of a third-order, lowland river network to determine the role (i.e., source
vs. sink) of the HZ in stream DOC cycling at the watershed scale (Ch. 2 and 3) and 2) evaluate a
series of analytical methods used to quantify stream DOC quality in a rapid and cost-effective
way, including new, in-situ DOC sensor technologies (Ch. 4). Specifically, Chapter 2 is a
preliminary study that focuses on whether the HZ is an important location for stream DOC
processing and if a processing signal is observable at the watershed scale. Chapter 3 focuses on
how the HZ processes stream DOC and if the dominant biogeochemical transformation
mechanism(s) is consistent throughout the stream network. Lastly, Chapter 4 provides a review
of new in-situ optical methods for measuring DOC quality in stream networks, which is a critical
variable when studying in-stream C cycling (Ruhala and Zarnetske, 2017).

4

Figure 1: Conceptual model of the hyporheic zone. This diagram depicts the extent of a
representative HZ. The red dashed lines (--) indicate hyporheic flow and the black arrows ()
indicate groundwater flow. Image modified from Tonina and Buffington (2009).

5

CHAPTER 2: A PRELIMINARY STUDY OF TRENDS IN DISSOLVED ORGANIC
CARBON QUANTITY AND QUALITY ACROSS THE STREAM-GROUNDWATER
INTERFACE OF A THIRD-ORDER, LOWLAND RIVER NETWORK

2.1 Introduction
The zone beneath and alongside of the stream where stream water-groundwater (SWGW) interactions occur (i.e., the hyporheic zone, HZ) may be an important location within river
networks for dissolved organic carbon (DOC) processing (e.g., degradation, consumption) since
it is a hotspot for microbial activity and biogeochemical reactions (Storey et al., 1999; Baker et
al., 1999; Fischer et al., 2005; Battin et al., 2009; Zarnetske et al., 2011). Previous small-scale
field (e.g., reach scale) and mesocosm studies have observed decreasing DOC concentrations
along hyporheic flowpaths, suggesting that the HZ is effective at consuming DOC and therefore
may be an important sink for DOC in stream networks (Findlay et al., 1993; Findlay and
Sobczak, 1996; Schindler and Krabbenhoft, 1998; Sobczak and Findlay, 2002; Zarnetske et al.,
2011). However, hyporheic processes are also expected to impact stream DOC cycling on larger
scales (e.g., the watershed scale; Boano et al., 2014). Yet, few studies have assessed hyporheic
DOC cycling across entire watersheds, and those that have, relied heavily on mathematical
models more than empirical field data (e.g., Wondzell, 2011; Boano et al., 2014; Kiel and
Cardenas, 2014; Gomez-Velez and Harvey, 2014).
Here, we uniquely couple trends in hyporheic DOC quantity and quality from a
watershed-scale field study to assess the role of the HZ in DOC cycling within stream networks.
DOC quality is included as a variable since the quality of DOC impacts its bioavailability and
reactivity (i.e., lability), which ultimately affects downstream ecosystems and water quality
(Fellman et al., 2010; Cory et al., 2011) and may vary independently of DOC concentration

6

(Lutz et al., 2012). This approach will allow us to identify if the HZ is an important location for
DOC processing and how it controls the fate of DOC transported downstream.
The following central research questions are addressed in this study: 1) is the HZ a
significant location for processing (e.g., degradation, consumption) of stream DOC? and if so, 2)
can a processing signal be observed at larger scales (i.e., across the entire watershed and/or
within stream orders). We hypothesize that the HZ is a significant location for processing of
stream DOC. If our hypothesis is true, then we predict:
1) Decreasing DOC concentrations will be observed along hyporheic flowpaths throughout
the watershed, similar to the results found in previous small-scale field and mesocosm
studies.
2) The high rates of microbial activity in the HZ will alter overall DOC quality in the HZ, as
indicated by the presence of less aromatic and lower molecular weight DOC when
compared to stream water.
3) The variance in DOC quantity and quality will be greatest in the HZ when compared to
stream water and groundwater conditions due to the high DOC processing rates in the
HZ.
To test our predictions, we use field observations of DOC and hydrologic conditions collected
across the third-order, mixed land use watershed of Augusta Creek in southwestern Michigan,
USA (42°21'12"N, 85°21'14"W), which is adjacent to the Kellogg Biological Station (KBS), a
Michigan State University field research station.

7

2.2 Background
2.2.1 Field Site Description
A synoptic HZ sampling campaign was conducted over an 8-day period in August 2015
across the Augusta Creek watershed, which is located in Barry and Kalamazoo counties in
southwestern Michigan and is part of the larger Kalamazoo River watershed (Figure 2). This
watershed was selected as the field site because Augusta Creek is a historically important stream
for biogeochemical and ecological research (e.g., it was an original River Continuum Concept
site; Vannote et al. 1980), has a long-term and active United States Geological Survey (USGS)
gaging station (04105700), and is easily accessible through the roadways and university lands.
The geology of the watershed has been heavily influenced by glacial activity (i.e., ice
advancement and retreat), most recently in the late Pleistocene Epoch (Dunbar, 1962). As a
result, the watershed is underlain by deep glacial drift deposits consisting of mixtures of gravel,
sand, silt, and clay, varying in thickness from 40 to 120 m across the watershed (FTWRC, 2011).
Augusta Creek is a third-order, lowland watershed draining 98 km2 (Figure 2, Strahler,
1957). Most of the tributaries originate in groundwater-fed lakes and low-lying wetlands (Manny
and Wetzel, 1973; FTWRC, 2011). The three dominant land cover types are 1) 47% agriculture
(crops and pastures), 2) 23% upland forest (mainly deciduous), and 3) 20% wetland/marsh
complexes (FTWRC, 2011; Figure A1; Table A1). The mean annual precipitation for the
Augusta Creek watershed is 925 mm, while the total precipitation in 2015 was 1,154 mm
(FTWRC, 2011; KBS-LTER, Dataset KBS002). Soils are typically sandy to loamy in uplands,
with organic (muck) soils in the riparian wetlands, and patches of muck along the stream channel
(FTWRC, 2011). Soil infiltration rates throughout the watershed range from 12.7 to 25.4 mm/h
(FTWRC, 2011). The mean stream slope is approximately 2.03 m/km (Manny and Wetzel,

8

1973) and the average grain size diameter (d50) for the upper streambed sediments (i.e.,
shallower than 20 cm), determined from stream sediment samples collected at 16 sites across the
watershed, is 0.90 mm (Table A1). In addition, clay and marl (CaCO3-rich) lenses are found
throughout the streambed sediments.
In terms of stream chemistry, Augusta Creek is a hard water stream with a total hardness
of about 280 mg/L (Mahan and Cummins, 1974; King 1978). Alkalinity typically ranges from
160 to 210 mg/L as CaCO3, while pH ranges from 7.5 to 8.7 (Manny and Wetzel, 1973; Mahan
and Cummins, 1974; King, 1978). Calcite (CaCO3) precipitates are often found on stones on the
steam bottom. Sampling of Augusta Creek from 1997-2015 has shown that stream DOC
concentrations typically range from 2 to 12 mg/L (Hamilton, unpublished data)
2.2.2 Hydrology
Augusta Creek exhibits a discharge flow regime characterized as having minimal withinand among-year variation (Url and Hart 1992, Poff et al., 1997). Based on recent and historical
discharge measurements collected at the USGS gaging station (04105700) located on the lower
main stem above the Kalamazoo River confluence (Figure 2), Augusta Creek’s discharge
typically varies between 0.57 to 2.83 m3/s, with occasional peak flow events. This range in
discharge was observed from August 2014 – August 2015 (with ~3 peak flow events) as well as
over several years prior (Figure 3A). The synoptic HZ sampling campaign took place from
August 10 – August 18, 2015, during which the average discharge was 1.23 m3/sec, slightly
higher than the previous 5-year average of 1.05 m3/sec for the month of August (USGS, 2016,
Station 04105700). During the sampling period, discharge varied some from 0.93 to 1.93 m3/sec,
with two notable precipitation events, one on August 10 and another on August 17. However,
discharge remained relatively stable over most of the sampling period (Figure 3B).

9

Figure 2: Topographic map of Augusta Creek watershed. Inset shows its location in Barry
and Kalamazoo counties, Michigan, USA.

10

Figure 3: 2015 USGS hydrographs for Augusta Creek. Stream discharge data collected from
the Augusta Creek gaging station, #04105700 from A) May 2014 – November 2015 and B)
during the synoptic HZ sampling campaign from August 10 – August 18, 2015, highlighted in
the blue shaded box. Data from US Geological Survey National Water Information System
(nwis.waterdata.usgs.gov, accessed 4 Nov 2016).

11

2.3 Methods
2.3.1 Sample Collection
Hyporheic Zone porewaters, along with stream water and groundwater, were sampled at
16 sites across the Augusta Creek watershed (Figure 4) varying in land use/cover and stream
order (Table A1). Hyporheic porewater samples were collected using a MINIPOINT porewater
sampler, which is a custom-built system for temporary installation of nested piezometers (Duff et
al., 1998; Figure 5A). The MINIPOINT consists of six 50 cm long by 0.5 cm-diameter
piezometers arranged in a 10 cm-diameter circular array (Figure 5A). The peizometers are
adjustable allowing for high-resolution, vertical porewater profile sampling that is minimally
disruptive to the sediment column as well as the ambient chemical and biological processes
occurring in the HZ (Duff et al., 1998). For this study, the 6 piezometers were vertically
staggered at sediment depths of 2.5, 5.0, 7.5, 10, 15, and 20 cm. Three MINIPOINT samplers
(for triplicate sampling) were deployed at each site as close together as possible (Figure 5B). The
placement of MINIPOINTs was largely dependent on finding areas were the sediment type was
conducive to the piezometers, which means samples were not collected in cobbly or clayey
sediments that preclude installation of the sampler. Porewater samples were extracted using
peristaltic pumps (i.e., Cole Palmer Masterflex L/S Peristaltic Pump) set at low flow rates (i.e.,
1.5-2.5 mL min-1) to minimize or eliminate the disturbances to the natural groundwater flow field
(Harvey and Fuller, 1998). Approximately 60 mL of sample per depth was pumped into a 100mL BD syringe that was connected with the peristaltic pump and corresponding piezometer
using Tygon tubing with a 1.59-mm inner diameter and Masterflex Norprene Food Tubing with a
1.59-mm inner diameter. Samples were simultaneously collected from all 6 HZ depths. This was
repeated for each of the three MINIPOINTs deployed at a site. In addition, a single stream water

12

sample per site, 60 mL in volume, was collected approximately 25-35 cm below the stream freesurface using a 10- mL BD syringe. A single groundwater sample per site, 60 mL in volume, was
collected from a drive-point well installed to a depth of 60 cm below the stream-sediment
interface. The drive-point well was purged and allowed to equilibrate before the sample was
drawn directly into a 100-mL BD syringe. All samples (i.e., stream water, groundwater, and HZ)
were filtered in the field using glass microfiber filters (i.e., Whatman GF/F 25 mm-diameter
filters, 0.7 µm pore size) into 60-mL acid-washed HDPE amber bottles (to prevent
photodegradation) and kept on ice as required by US EPA method 415.3 (Potter and Wimsatt,
2003). The samples were then filtered through cellulose acetate filters (i.e., Sartorius Stedim, 0.2
µm pore size) within 12 hours and stored at 4°C until analyzed (see Section 2.3.2).
Dissolved oxygen (DO) was measured at each site using a fiber-optic oxygen meter (i.e.,
Pyro Science FireStingO2 with Robust Oxygen Probes, Pyro Science, Aachen, Germany). DO
readings were taken at HZ depths of 2.5, 5.0, 10, and 20 cm from each MINIPOINT as well as
for stream water and groundwater in conjunction with sample collection. In addition, temperature
was measured once at each HZ depth and in the stream and groundwater samples using the fiberoptic oxygen meter. Lastly, sediment cores of the upper 20 cm of the stream bed were collected
and used to determine the grain size distribution for each site, from which D50 values were
calculated (Table A1).

13

Figure 4: Map of Augusta Creek sampling sites in 2015. A map of the 16 sites sampled
during the 2015 synoptic HZ sampling campaign across the third-order, groundwater-fed
watershed of Augusta Creek, Michigan, USA.

14

Figure 5: Sampling scheme. A) MINIPOINT sampler design (Duff et al., 1998) used for HZ
sampling. B) Field setup used at each of the 16 sites sampled across Augusta Creek.

2.3.2 Analytical Methods
All samples were analyzed in the laboratory within 4 months of collection. DOC
concentrations (reported as Non-Purgeable Organic Carbon, NPOC) were analyzed via hightemperature combustion using a Shimadzu TOC-L Analyzer (Shimadzu Scientific Instruments,
Kyoto, Japan). In addition, a Dionex ICS-2100 ion chromatograph (Thermo Fisher Scientific,
Massachusetts, USA) was used to analyze the samples for NO3-, Cl-, and SO42- concentrations.
Ion analysis was completed using an AS19 Dionex IonPac column (2 x 250 mm) with a
potassium hydroxide (KOH) eluent generator and a 0.25 mL/min flow rate. To infer DOC
quality, optically-derived DOC variables were determined from absorbance data using a
Shimadzu dual-beam UV 1800 spectrophotometer (Shimadzu Scientific Instruments, Kyoto,
Japan). Absorbance readings were taken over the entire UV-VIS range from 220 to 800 nm using
15

semi-micro, BrandTech cuvettes with a 1-cm path length and EPure water (i.e., 18 ohm,
Barnstead EPure system) as the blank. Each cuvette was triplicate rinsed with EPure water
between samples. For calculations of the optically-derived variables see Section 2.3.3.
2.3.3 Optical Indices
The structural complexity of DOC makes it difficult to study DOC composition, which is
a determining factor of DOC quality, via direct molecular techniques (Cory et al., 2011). Instead,
several optical properties of DOC determined via ultraviolet-visible (UV-VIS) absorbance
spectroscopy can be used to assess the chemical structure and thus the quality of DOC (e.g.,
Weishaar et al., 2003; Cory et al., 2011; Jollymore et al., 2012; Creed et al. 2015). In this study,
specific ultraviolet absorbance at a wavelength (λ) of 254 nm (i.e., SUVA254) was used to
indicate DOC aromaticity. SUVA254 is obtained by measuring the sample’s absorbance at λ=254
nm and dividing by the DOC concentration of the sample (Equation 2.1; units are L/mg C/m).
SUVA254 =

Absorbance254 nm
DOC concentration

(2.1)

A λ=254 nm is used because electron structures associated with aromatic C molecules absorb
energy at this wavelength, while other structures do not (Weishaar et al., 2003). Therefore the
absorbance reading is directly related to the amount of aromatic C within the total DOC sample.
Higher SUVA254 values are associated with increased aromaticity (Table 1; Weishaar et al.,
2003).
The optical data were also used to calculate a spectral slope ratio (SR) for each sample.
SR is used as a proxy for molecular weight; SR increases with decreasing molecular weight (Table
1; Helms et al., 2008). To calculate SR, absorbance values were measured over 220 to 800 nm
and a ratio of the slopes of the 275 – 295 nm (S275-295) absorbance spectra and the 350 – 400 nm
(S350-400) absorbance spectra was calculated (Equation 2.2).
16

SR =

S275 nm − 295 nm
S350 nm − 400 nm

(2.2)

A summary of the optical indices and their relationship to DOC molecular characteristics is
provided in Table 1.

Table 1: Summary of optical DOC indices
Optical Indices

Aromaticity

Molecular Weight

SUVA254

Positive Correlation

Spectral Slope (SR)

Negative Correlation

2.3.4 Vertical Porewater Profiles
To observe trends in DOC quantity and quality, vertical porewater profiles of the optical
properties as well as the sampled chemical species were created for each site using mean and
standard deviation of the triplicate hyporheic samples (e.g., DOC concentrations measured from
a depth of 2.5 cm in MINIPOINT 1, 2, and 3), which was compared with the stream water and
groundwater values (Figure 6).

Figure 6: Example of a vertical porewater profile. The sediment interface (i.e., 0 cm) is
shown by the dashed line.
17

2.4 Results
The average DOC concentration in stream water collected from the 16 sampling sites was
9.97 mg/L, which falls within the previously reported range for DOC concentrations in Augusta
Creek of 2-12 mg/L (Hamilton, unpublished data). The range of DOC measured in stream water
was 6.17 to 16.75 mg/L with a median value of 9.73 mg/L. For groundwater, the average DOC
concentration across the watershed was lower than that of stream water at 6.62 mg/L and ranged
from 3.06 to 16 mg/L with a median value of 5.76 mg/L.
2.4.1

Results of the Vertical Porewater Profiles
Individual vertical porewater profiles of DOC concentration, DO, SUVA254, and SR for

all 16 sites were averaged to create a composite vertical porewater profile, representative of the
average DOC conditions in the HZ across Augusta Creek (Figure 7, graphs labeled “All Sites”).
Vertical profiles were also grouped by stream order and then averaged, to create vertical profiles
representative of average HZ DOC conditions within each stream order.
The vertical profiles show decreasing DOC concentrations with depth through the HZ
(i.e., 2.5 – 20 cm), for the watershed-wide mean profile as well as the mean profiles for each
stream order (Figure 7A). However, a slight increase in DOC concentration was observed
between depths 7.5 and 10 cm in the third-order vertical profile (Figure 7A), which may be due
to outliers such as Site F (Figure A6), which exhibited an increase in DOC concentration at
intermediate HZ depths. DOC concentrations at the 2.5 cm depth were lower than the DOC
concentrations observed in the stream channel for all averaged vertical profiles. In addition, all of
the averaged profiles exhibited lower DOC concentration at deeper HZ depths (e.g., 10, 15,
and/or 20 cm) than observed in the underlying groundwater.

18

Dissolved oxygen also exhibited decreasing concentration with HZ depth in the
watershed mean profile and in the mean profiles of each stream order (Figure 7B). The stream
water across Augusta Creek was oxygenated with DO levels greater than 5 mg/L during the
sampling campaign, while the groundwater measurements taken from 60 cm in depth show
oxygen depletion to 2 mg/L or less. Groundwater DO was not measured at first-order sites due to
interference of clay layers and problems with the fiber-optic oxygen meter. An increase in DO
between depths 15 and 20 cm is observed in the third-order vertical profile, and thus also
reflected in the watershed averaged vertical profile. Again, this increase may be the result of
several outliers (see Appendix A).
Regarding the optical properties, decreasing SUVA254 values were observed in the
watershed mean profile and in the mean profiles for each stream order (Figure 7C). However, the
first-order vertical profile showed an increase in SUVA254 at intermediate HZ depths, which may
also be the result of outliers, including Site L (Figure A12). Lastly, SR showed slight increases
with HZ depth in the watershed vertical profile, the first-order vertical profile, and third-order
vertical profile, while the second-order profile exhibited decreasing SR values, especially at
deeper (i.e., > 10 cm) HZ depths (Figure 7D).

19

Figure 7: Vertical porewater profiles of DOC concentration and optical indices. Vertical
profiles of A) mean DOC, B) mean DO, C) mean SUVA254, and D) mean SR of all 16 sites and
by stream order. Each plot includes standard deviation bars; n=16 for “all sites”, n=6 for “1st
order sites”, n=5 for “2nd order sites” and n=5 for “3rd order sites.”

20

2.4.2 DOC Variance and Mixing Zone Model
The averaged vertical profiles showed increased variance (shown as SD) at intermediate
HZ depths (i.e., ~7.5 to 15 cm) for DOC concentration and SUVA254 (Figure 7A, C). An
exception to this is the averaged profile for SUVA254 across third-order sites, which showed
maximum variance at 20 cm. SR also exhibited increased variance at deeper HZ depths (i.e., ~20
cm), especially for second-order sites (Figure 7D).
It is important to note that while stream water variance is small (with the exception of
DOC concentration for first-order sites), groundwater variance for both DOC quantity and
quality indicators was significantly larger than expected (e.g., see first-order sites for DOC
concentration; Figure 7A). Considering that many of the sampling sites are likely groundwater
upwelling areas (see Section 2.2.1), the trends in DOC conditions observed in Figure 7 may be
influenced by the variable chemistry of upwelling groundwater. As a way to verify that the
trends we are observing are the result of HZ processes and not due simply to variable
contributions of upwelling groundwater, a mixing zone model was created to identify the
subsurface depths that showed a mixing signal (Figure 8). This was done since strong
groundwater upwelling could condense the mixing zone (i.e., HZ) to depths shallower than 20
cm (e.g., Boano et al., 2014; Creed et al., 2015).
To identify the mixing zone within the overall vertical profile, we used a conservative
(i.e., non-reactive) chemical, specifically chloride (Cl-), in combination with temperature to
observe the depth (i.e., shallower than 60 cm) at which the groundwater signal ended and the
profile shifted to resembling a mixing signal of both groundwater and stream water (e.g., Peters
and Ratcliffe, 1998). Figure 8 presents the conceptual model used to identify the mixing zone.

21

Figure 8: Model used to identify mixing zone depths. A) Conceptual model of the SW-GW interactions in a simple stream
bedform and B) the resulting vertical porewater profile of a conservative chemical. C) The vertical profile for a conservative
chemical can deviate from the groundwater signal such that the concentration decreases or increases to the streambed surface, based
on the concentration of the chemical in the stream water. The depth at which the profile deviates from the groundwater signal is
indicative of the groundwater flux (i.e., weak or strong). D) For Site J the Cl- concentration deviates from that of groundwater at a
depth of 7.5 cm. The temperature profile is used in support of the interpretation of the Cl- profile.

22

Within a simple stream bedform (e.g., a small ripple), underlying groundwater will
upwell towards the stream-sediment interface where it will then meet downwelling stream water
at a given depth to create a zone in which SW-GW interactions occur (i.e., the HZ; Figure 8A). If
concentrations differ between stream water and groundwater, conservative chemicals like Cl- can
be used to determine the extent of the mixing zone. A vertical porewater profile of Clconcentrations will resemble the Cl- concentration of groundwater at all depths below the mixing
zone (Figure 8B). Within the SW-GW mixing zone, the Cl- concentrations will reflect the
relative proportions of stream water and groundwater (e.g., dilution, Figure 8C). Thus, the Clconcentrations indicate the extent of the mixing zone (Figure 8B and C). In addition, temperature
behaves similarly to conservative chemicals and will resemble groundwater temperature until the
mixing zone is reached where temperature will begin to vary. Therefore, temperature profiles
were used to support the interpretation of Cl- profiles.
As an example, Figure 8D shows how the mixing zone was qualitatively determined for
Site J, where the Cl- concentration indicates stream water influence to a depth of 7.5 cm.
Therefore the mixing zone for Site J exists between depths of 0 and 7.5 cm. This process was
repeated for all 16 sites (Table A1) and new vertical profiles representative of the average
mixing zone for the watershed and each stream order were recalculated and are presented in
Figure 9.

23

Figure 9: DOC concentration and optical indices within the mixing zone. Vertical profiles of
A) mean DOC, B) mean DO, C) mean SUVA254, and D) mean SR including only depths with a
mixing zone signal. Standard deviation was calculated for all depths where n >3.

24

The vertical profiles of the averaged DOC quantity and quality conditions created using
the mixing zone model only extend to a depth of 20 cm or shallower since the groundwater
samples taken at 60 cm depth are assumed to be below the mixing zone. In addition, error bars
are only included where n>3. Along the mixing zone profile for the watershed and for each
stream order, DOC concentration, DO (with the exception of the second-order profile) and
SUVA254 decreased with depth, while SR trends remained relatively stable, showing slight
decreases at deeper mixing zone depths (i.e., ~10 cm; Figure 9). These trends are consistent with
those observed in Figure 7. However, after identifying the mixing zone and using only depths
with a mixing zone signal, the vertical profiles for the watershed and each stream order showed
smaller changes in variance between stream water and mixing zone depths (i.e., more consistent
variance is observed along the profile) for DOC concentration, DO, SUVA254 and SR (Figure 9).
Additionally, in some cases, the variance observed in the stream water was larger than the
variance seen at mixing zone depths (e.g., Figure 9A, B). Lastly, it was observed that the mixing
zone becomes shallower (i.e., does not extend as far into the sediment) as stream order increases,
as exhibited by the shortened vertical profiles for all variables across second- and third-order
sites (Figure 9).

25

2.5 Discussion
The averaged vertical profiles (see Section 2.4.1, Figure 7A) as well as most (i.e., 14 out
of 16) of the individual site profiles (see Appendix A) showed decreasing DOC concentration
with depth through the HZ. This supports the first prediction that DOC concentration will
decrease with increasing HZ flowpath length and is in agreement with previous findings from
small-scale field and mesocosm studies, suggesting that the HZ acts as a sink for DOC in this
stream network (Findlay et al., 1993; Findlay and Sobczak, 1996; Schindler and Krabbenhoft,
1998; Sobczak and Findlay, 2002; Zarnetske et al., 2011). Furthermore, DOC in the HZ
decreased to concentrations lower than that of stream water or groundwater, indicating that
mixing of the two end members (i.e., SW and GW) alone cannot account for the observed
decreases in DOC concentration (e.g., Pinder and Jones, 1969). Thus, physicochemical and/or
biological reactions are most likely occurring in the HZ, consuming DOC.
One potential processing mechanism of DOC in the HZ is aerobic, microbial respiration
(i.e., consumption of DOC and release of CO2). Aerobic, microbial respiration will result in
decreased DOC concentrations with concomitant consumption of DO (e.g., Findlay and Sobczak,
1996). Likewise, aerobic, microbial respiration will alter DOC quality by preferentially respiring
the more labile fraction, resulting in lower molecular weight and less aromatic residual DOC
(e.g., McKnight et al., 2001; Fellman et al., 2010; Cory et al., 2011; Mann et al., 2012; Creed et
al., 2015; Helton et al., 2015). The DO profiles showed decreasing DO with depth in the HZ,
which supports the hypothesis of aerobic, microbial respiration. In addition, SUVA254 and hence
the aromaticity of DOC decreased with depth in the HZ (with the exception of first-order
intermediate depths, Figure 7C). SR exhibited a slight increase with depth in the HZ, indicating a
decrease in the mean molecular weight of DOC (Table 1, Weishaar et al., 2003; Helms et al,

26

2008). Both optical properties therefore suggest that the DOC in the HZ changes as a result of
microbial processing, thus supporting the second prediction.
When observing the trends in variance for DOC quantity and quality, variance increased
in the HZ, supporting the third prediction. However, the groundwater variance was larger than
expected (Figure 7) since regional groundwater should be relatively homogenized (i.e.,
chemically) due to longer flow paths and increased residence times (e.g., Vannote et al., 1980;
Boano et al., 2014). The larger variance might result from the groundwater having a more
variable chemistry than previously acknowledged or the sampling depth of 60 cm may not fully
represent the regional groundwater. In addition, since Augusta Creek is groundwater-fed,
upwelling groundwater may limit the extent of the HZ, resulting in a smaller region of SW-GW
interactions (e.g., Cardenas and Wilson, 2007; Boano et al., 2009).
Therefore as described in Section 2.4.2, a mixing zone model was used to identify the
mixing zone at each site to determine if the observed DOC trends in the averaged vertical
profiles were the result of strong groundwater upwelling or HZ processes. For most sampled
sites across Augusta Creek, the mixing zone did not extend to 20 cm (Table A1). However, once
the vertical profiles were adjusted to account for shallower mixing zones, it was observed that
the DOC quantity and quality variables showed trends consistent with the results presented in
Figure 7, supporting the hypothesis of DOC processing (likely via aerobic, microbial respiration)
in the HZ (Figure 9). However, smaller changes in variance were observed between stream
water and hyporheic depths (i.e., more consistent variance was observed along the profile) for
DOC concentration, DO, SUVA254 and SR (Figure 9) and in some cases, the variance observed in
the stream water was larger than the variance seen within the HZ (e.g., Figure 9A, B).

27

This observation of decreased variance observed at subsurface depths when using the
mixing zone model no longer supports the third prediction. However, this may the result of not
capturing all of the variation within DOC quantity and quality variables due to the decreased
sample size. It may also indicate that the biogeochemical transformation mechanisms in the HZ
result in a homogenized pool of DOC. Lastly, mixing zone profiles also showed that the extent of
the HZ decreased as stream order increased, indicating that the influence of HZ processes on
stream C cycling may change across the stream network.
Overall, these findings indicate that the HZ is a location for net removal of DOC in this
river network and that the processing signal can be observed at larger scales (i.e., across the
entire watershed and within stream orders), as demonstrated from the averaged vertical
porewater profiles (Figure 7). In addition, the optical indices used to infer DOC quality suggest
that aerobic, microbial respiration may be an important processing mechanism in the HZ.
However, there are several remaining questions:
1) What is the dominant transformation mechanism(s) consuming DOC in the HZ?
2) Is the dominant transformation mechanism consistent across the river network?
3) Is the potential for DOC transformation in the HZ consistent across the river network?
4) Does the net removal of DOC occur all year? It could be lesser in cooler seasons, or
greater during autumn leaf fall and consequent leaching of labile DOC, or during spring
peaks in epilithic algal production.
To more accurately assess the dominant transformation mechanisms (e.g., biotic vs.
abiotic processes) and how they vary across the watershed, vertical profiles from individual sites
need to be analyzed. In addition, by studying DOC dynamics within the HZ on a site-by-site
basis, we can also gain an understanding of not only how the HZ transforms DOC (e.g., the

28

dominant transformation mechanism(s), source vs. sink), but also if the potential for
transformation of DOC in the HZ changes based on location in the watershed. As discussed, the
mixing zone vertical profiles from this study (Figure 9) showed that the depth of the HZ
decreased as stream order increased. This would suggest that the potential for DOC
transformation (e.g., consumption) in the HZ changes across stream order, with the HZ having a
higher potential to process DOC in headwater streams than in larger stream orders (e.g., Vannote
et al., 1980; Boano et al., 2014; Creed et al., 2015).
Additional chemical (e.g., improved DO measurements) and hydrological data (e.g.,
vertical head gradient, improved groundwater measurements) need to be collected at each site,
and more sites need to be sampled across the watershed to address these remaining questions.
Thus, a second synoptic HZ sampling campaign was conducted in August 2016 and the results
are presented in Chapter 3.

29

2.6 Conclusions
Hyporheic Zone trends in DOC quantity and quality observed from the 2015 watershedscale synoptic sampling campaign across the Augusta Creek watershed support findings from
earlier small-scale field and mesocosm studies, indicating that the HZ is a location for processing
of DOC and likely acts as a net sink for DOC in this river network. However, in strongly
gaining, groundwater-fed, stream networks, the depth of the SW-GW mixing zone (i.e., HZ) may
be limited. Thus, it is important to be able to quantify the depth of the HZ when studying
processes at the SW-GW interface. Additionally, more detailed chemical (e.g., DO) and
hydrological data are needed to accurately identify processes driving DOC transformations in the
HZ (e.g., aerobic, microbial respiration) and how the role (e.g., source vs. sink) of the HZ
changes across river networks as well as at individual sites. Nonetheless, this study has
highlighted a potentially important location for DOC processing in streams and therefore the HZ
may need to be accounted for in C budget models.

30

CHAPTER 3: DETERMINGING HOW DISSOLVED ORGANIC CARBON IS
TRANSFORMED AT THE STREAM-GROUNDWATER INTERFACE ACROSS A THIRDORDER, LOWLAND RIVER NETWORK

3.1 Introduction
To understand the role of the hyporheic zone (HZ) in stream carbon (C) cycling, studies
of HZ processes need to be conducted at the watershed scale (e.g., Boano et al., 2014). In
Chapter 2, we used synoptic sampling of the HZ across a third-order, lowland watershed to
assess if dissolved organic carbon (DOC) processing was occurring within the HZ. The findings
indicate that the HZ is a location for DOC processing in this river network and that the
processing signal is observable at the watershed scale. Yet, questions remain about how DOC is
processed in the HZ at the watershed-scale. For example, in Chapter 2 we were able to show that
on average DOC was consumed (i.e., removed from solution) over HZ depths by using
“composite” or averaged vertical porewater profiles to study trends in DOC quantity and quality.
This, suggests that the HZ may act as a sink for DOC in this stream network (e.g., Sobczak and
Findlay, 2002). However, we could not fully identify the dominant biogeochemical
transformation mechanisms (e.g., biotic vs. abiotic processes) consuming DOC in the HZ, even
though we presented evidence that microbial respiration may be important.
This chapter focuses on identifying the dominant biotic (e.g., aerobic, microbial
respiration) or abiotic (e.g., adsorption to sediments) biogeochemical processes driving DOC
transformations in the HZ and on assessing how the role of the HZ as a processor of stream DOC
varies across a river network (e.g., varying processing mechanisms, varying potential for
transformation). This is accomplished at the network scale by utilizing hyporheic trends in DOC
quantity and quality, improved dissolved oxygen (DO) measurements and local flow patterns
(i.e., upwelling vs. downwelling) on a site-by-site basis. The central research questions are: 1)
31

how is DOC processed (i.e., consumed) in the HZ? and 2) does the potential for DOC
transformation (i.e., the processing potential) in the HZ change based on location (e.g., stream
order) in the stream network?
Previous work demonstrated that the HZ is a zone with high biological diversity, strong
chemical gradients (e.g., Valett et al., 1996; Hedin et al., 1998; Findlay and Sobczak, 2000;
Zarnetske et al., 2011; Nogaro et al., 2013; Boano et al., 2014) and increased residence times
(e.g., Battin et al., 2008; Harvey et al., 2013; Boano et al., 2014), all of which promote biotic
processes (e.g., microbial metabolism). Additionally, previous work has shown that DOC
sorption to sediments (i.e., abiotic processes) also occurs in the HZ (e.g., McDowell, 1985;
Fiebig and Lock, 1991; Findlay and Sobczak 1996). While, abiotic sorption of DOC to hyporheic
sediment does occur, it is limited by sorption sites and will equilibrate with the DOC
concentration of upwelling/downwelling waters (e.g., Day et al., 1994; Kaplan and Newbold,
2000). Thus, for continued sorption of DOC, the previously sorbed DOC must be removed via
microbial metabolism in order to regenerate sorption sites (e.g., Kaplan and Newbold, 2000).
Therefore, on short time scales, sorption of DOC to hyporheic sediments should not be a
significant or consistent sink for DOC in the HZ (e.g., Dahm, 1981; Fiebig, 1995).
Given the assertion that abiotic sorption would have limited influence on DOC
concentrations, we hypothesize that biotic processes, rather than abiotic processes, are
predominantly transforming DOC in the HZ. Specifically, we hypothesize that aerobic, microbial
respiration is the dominant transformation mechanism consuming DOC in the HZ and thus, the
HZ acts as a sink for DOC in this river network. If true, then we predict that:
1) Decreasing DOC and DO concentrations will be observed with increasing HZ depth
throughout the watershed.

32

2) SUVA254 will decrease with depth in the HZ, while SR will increase with depth in the
HZ, indicating the presence of less aromatic and lower molecular weight DOC.
In addition, we hypothesize that the potential for DOC transformation in the HZ varies across the
watershed, assuming that a framework similar to the River Continuum Concept also applies to
the stream water-groundwater (SW-GW) interface (Vannote et al., 1980). If this second
hypothesis is true, then we predict that:
1) The processing potential of the HZ (i.e., acting as a consumer/sink) of stream DOC
will become less significant with increasing stream order, exhibited by lower rates of
consumption of DOC and lower rates of change in the DOC quality indicators.
2) The processing potential of the HZ will be less significant at sites where upwelling
groundwater limits the SW-GW mixing depth, again exhibited by lower rates of
consumption of DOC and lower rates of change in DOC quality indicators.
Prediction 1 is based on changes in chemical and physical conditions along the stream channel
(e.g., Vannote et al., 1980; Battin et al., 2008). For example, changes in channel slope and
streambed morphology tend to limit surface-subsurface exchange in higher stream orders (Battin
et al., 2008). In addition, subsurface flowpaths that feed higher stream orders also tend to be
more homogenized (i.e., lower chemical variation) due to longer residence times, limiting
microbial reactivity (Battin et al., 2008; Harvey et al., 2013; Boano et al., 2014). Furthermore,
the stream also becomes less connected to adjacent wetlands and the terrestrial landscape as
stream order increases, reducing inputs of complex allochthonous (e.g., terrestrial vascular-plant
derived) DOC (Vannote et al., 1980). For prediction 2, upwelling groundwater may limit the
extent of the HZ, resulting in a smaller region of SW-GW interactions, limiting HZ processing of
DOC (e.g., Cardenas and Wilson, 2007; Boano et al., 2009).

33

To test these predictions, a second synoptic HZ sampling campaign was conducted in
2016 across Augusta Creek in southwestern Michigan, USA (42°21'12"N, 85°21'14"W). A total
of 39 sites were sampled, covering more of the watershed than in the 2015 survey discussed in
Chapter 2, and additional chemical and hydrological (e.g., vertical head gradient) measurements
were taken to improve our ability to identify changes in DOC quantity and quality and to assess
DOC transformation mechanisms.

34

3.2 Background
3.2.1 Field Site Description
A synoptic HZ sampling campaign was conducted over an 8-day period in August 2016
across the Augusta Creek watershed, which is located in Barry and Kalamazoo counties in
southwestern Michigan and is part of the larger Kalamazoo River watershed (see Chapter 2,
Figure 2). This watershed was selected as the field site because Augusta Creek is a historically
important stream for biogeochemical and ecological research (e.g., it was an original River
Continuum Concept site; Vannote et al. 1980), has a long-term and active United States
Geological Survey (USGS) gaging station (04105700), and is easily accessible through roadways
and university lands. The geology of the watershed has been heavily influenced by glacial
activity (i.e., ice advancement and retreat), most recently in the late Pleistocene Epoch (Dunbar,
1962). As a result, the watershed is underlain by deep glacial drift deposits consisting of
mixtures of gravel, sand, silt, and clay, varying in thickness from 40 to 120 m across the
watershed (FTWRC, 2011).
Augusta Creek is a third-order, lowland watershed draining 98 km2 (see Chapter 2,
Figure 2; Strahler, 1957). Most of the tributaries originate in groundwater-fed lakes and lowlying wetlands (Manny and Wetzel, 1973; FTWRC, 2011). The three dominant land cover types
are 1) 47% agriculture (crops and pastures), 2) 23% upland forest (mainly deciduous), and 3)
20% wetland/marsh complexes (FTWRC, 2011; Figure A1). The mean annual precipitation is
925 mm, while the total precipitation in 2016 was 975 mm (FTWRC, 2011; KBS-LTER, Dataset
KBS002). Soils are typically sandy to loamy in uplands, with organic (muck) soils in the riparian
wetlands, and patches of muck along the stream channel (FTWRC, 2011). Soil infiltration rates
throughout the watershed range from 12.7 to 25.4 mm/h and the mean stream slope is

35

approximately 2.03 m/km (Manny and Wetzel, 1973; FTWRC, 2011). In addition, clay and marl
(CaCO3-rich) lenses are found throughout the streambed sediments.
In terms of stream chemistry, Augusta Creek is a hard water stream with a total hardness
of about 280 mg/L (Mahan and Cummins, 1974; King 1978). Alkalinity typically ranges from
160 to 210 mg/L as CaCO3, while pH ranges from 7.5 to 8.7 (Manny and Wetzel, 1973; Mahan
and Cummins, 1974; King, 1978). Calcite (CaCO3) precipitates are often found on stones on the
stream bottom. Sampling of Augusta Creek from 1997-2015 has shown that stream DOC
concentrations typically range from 2 to 12 mg/L (Hamilton, unpublished data)
3.2.2 Hydrology
Augusta Creek exhibits a discharge flow regime characterized as having minimal withinand among-year variation (Url and Hart 1992, Poff et al., 1997). Based on recent and historical
discharge measurements collected at the USGS gaging station (04105700) located on the lower
main stem above the Kalamazoo River confluence (Figure 11), Augusta Creek discharge
typically varies between 0.57 to 2.83 m3/s, with occasional peak flow events. This range in
discharge was observed from August 2015 – August 2016 (with the exception of a peak flow
event in late August 2016) as well as over several years prior (Figure 10A). The synoptic HZ
sampling campaign took place from August 15 – August 22, 2016, during which the average
discharge was 1.35 m3/sec, higher than the previous 5-year average of 1.08 m3/sec for the month
of August (USGS, 2017, Station 04105700). During the sampling period, discharge varied some
from 1.08 to 3.65 m3/sec, with one notable precipitation event on August 15 and another smaller
precipitation on August 21 (Figure 10B).

36

Figure 10: 2016 USGS hydrographs for Augusta Creek. Stream discharge data collected
from the Augusta Creek gaging station, #04105700 from A) May 2015 – November 2016 and
B) during the synoptic HZ sampling campaign from August 15 – August 22, 2016, highlighted
in the blue shaded box. Data from US Geological Survey National Water Information System
(nwis.waterdata.usgs.gov, accessed 7 Mar 2017).

37

3.3 Methods
3.3.1 Sample Collection
Hyporheic Zone porewaters, along with stream water and groundwater, were sampled at
39 sites (Figure 11 and 3.3), varying in land use/cover and stream order (Table B1) across the
Augusta Creek watershed. Porewater samples were collected using a MINIPOINT porewater
sampler, which is a custom-built system for temporary installation of nested piezometers (Duff et
al., 1998; see Chapter 2, Figure 5A). The MINIPOINT consists of six 50-cm long by 0.5 cmdiameter piezometers arranged in a 10 cm-diameter circular array (see Chapter 2, Figure 5A).
The peizometers are adjustable allowing for high-resolution, vertical porewater profile sampling
that is minimally disruptive to the sediment column as well as the ambient chemical and
biological processes occurring in the HZ (Duff et al., 1998). For this study, the 6 piezometers
were vertically staggered at sediment depths of 2.5, 5.0, 7.5, 10, 15, and 20 cm. In exchange for
increasing the number of sampling sites from the previous year (i.e., 39 sites in 2016 versus 16
sites in 2015), only a single MINIPOINT sampler was deployed at each site. The placement of
the MINIPOINT in the stream channel was largely dependent on finding areas were the sediment
type was conducive to the piezometer, which means samples were not collected in cobbly or
clayey sediments that preclude installation of the sampler. Porewater samples were
simultaneously extracted from all 6 HZ depths using a peristaltic pump (i.e., Cole Palmer
Masterflex L/S Peristaltic Pump) set at a low flow rate (i.e., 1.5-2.5 mL min-1) to minimize or
eliminate the disturbances to the natural groundwater flow field (Harvey and Fuller, 1998).
Approximately 60 mL of sample per HZ depth was pumped into a 100-mL BD syringe that was
connected with the peristaltic pump and corresponding piezometer using Tygon tubing with a
1.59-mm inner diameter and Masterflex Norprene Food Tubing with a 1.59-mm inner diameter.

38

In addition, a single stream water sample per site, 60 mL in volume, was collected
approximately 25-35 cm below the stream free-surface using a 100-mL BD syringe, and single
groundwater sample per site, 60 mL in volume, was collected from a drive-point well installed to
a depth of 60 cm below the stream-sediment interface. Prior to groundwater sampling, the drivepoint well was purged, allowed to equilibrate and then a vertical head gradient (VHG) was
calculated. This was achieved by measuring the difference in head between the drive-point well
(HGW, units in cm) and the stream free-surface (HSW, units in cm) and then dividing by the total
depth of the well, d (i.e., 60 cm; Equation 3.1). A positive VHG indicates upwelling
groundwater, while a negative VHG indicates downwelling stream water (Table 2).
VHG =

HGW − HSW
d

(3.1)

Once the VHG was determined, the groundwater sample was drawn directly into a 100-mL BD
syringe. All samples (i.e., stream water, groundwater, and HZ) were filtered in the field, first
through a glass microfiber filter (i.e., Whatman GF/F 25 mm-diameter filters, 0.7 µm pore size)
and then through a cellulose acetate filter (i.e., Sartorius Stedim, 0.2 µm pore size) directly into
two, 30-mL acid-washed HDPE amber bottles (to prevent photodegradation) and kept on ice as
required by US EPA method 415.3 (Potter and Wimsatt, 2003). Upon return to the laboratory,
the samples were stored in the dark, at 4 °C until analyzed (see Section 3.3.2).
Dissolved oxygen (DO) was measured at each site using a fiber-optic oxygen meter (i.e.,
Pyro Science FireStingO2 with Robust Oxygen Probes, Pyro Science, Aachen, Germany). DO
readings were taken at all HZ depths using flow-through cells (i.e., Pyro Science Flow-Through
Cells, Pyro Science, Aachen Germany) that were connected with each piezometer as well as for
stream water and groundwater. In addition, temperature was measured at each HZ depth and in
the stream and groundwater samples using, the fiber-optic oxygen meter.

39

Figure 11: Map of Augusta Creek sampling sites in 2016. A map of the 39 sites sampled
during the 2016 synoptic HZ sampling campaign across the third-order, groundwater-fed
watershed of Augusta Creek, Michigan, USA.

40

Figure 12: Subplots of labeled sampling sites across Augusta Creek. Map of the 39 labeled sampling sites from the 2016 campaign.
41

3.3.2 Analytical Methods
All samples were analyzed in the laboratory within 2 months of collection. DOC
concentrations (reported as Non-Purgeable Organic Carbon, NPOC) were analyzed via hightemperature combustion using a Shimadzu TOC-L Analyzer (Shimadzu Scientific Instruments,
Kyoto, Japan). In addition, a Dionex ICS-2100 ion chromatograph (Thermo Fisher Scientific,
Massachusetts, USA) was used to analyze the samples for NO3-, Cl-, and SO42- concentrations.
Ion analysis was completed using an AS19 Dionex IonPac column (2 x 250 mm) with a
potassium hydroxide (KOH) eluent generator and a 0.25 mL min-1 flow rate. To infer DOC
quality, optically-derived DOC indicators were determined from absorbance data collected on a
Shimadzu dual-beam UV 1800 spectrophotometer (Shimadzu Scientific Instruments, Kyoto,
Japan). Absorbance readings were taken over the entire UV-VIS range from 220 to 800 nm using
semi-micro, BrandTech cuvettes with a 1-cm path length and EPure water (i.e., 18 ohm,
Barnstead EPure system) as the blank. Each cuvette was triplicate rinsed with EPure water
between samples. For calculations of the optically-derived indicators see Section 3.3.3.
3.3.3 Optical Indices
The structural complexity of DOC makes it difficult to study DOC composition, which is
a determining factor of DOC quality and reactivity (Cory et al., 2011). Instead of complex
analytical methods (e.g., mass spectrometry), several optical properties of DOC, determined via
ultraviolet-visible (UV-VIS) absorbance spectroscopy can be used to assess the chemical
structure and thus the quality of DOC (e.g., Weishaar et al., 2003; Cory et al., 2011; Jollymore et
al., 2012; Creed et al. 2015). In this study, specific ultraviolet absorbance at a wavelength (λ) of
254 nm (i.e., SUVA254) was used to indicate DOC aromaticity. SUVA254 is obtained by

42

measuring the sample’s absorbance at λ=254 nm and dividing by the DOC concentration of the
sample (Equation 3.2; units are L/mg C/m).
SUVA254 =

Absorbance254 nm
DOC concentration

(3.2)

A λ=254 nm is used because electron structures associated with aromatic C molecules absorb
energy at this wavelength, while other structures do not (Weishaar et al., 2003). Therefore the
absorbance reading is directly related to the amount of aromatic C within the total DOC sample.
Higher SUVA254 values are associated with increased aromaticity (Weishaar et al., 2003).
The optical data were also used to calculate a spectral slope ratio (SR) for each sample.
SR is used as a proxy for molecular weight; SR increases with decreasing molecular weight (Table
1; Helms et al., 2008). To calculate SR, absorbance values were measured over 220 to 800 nm
and a ratio of the slopes of the 275-295 nm (S275-295) absorbance spectra and the 350-400 nm
(S350-400) absorbance spectra was calculated (Equation 3.3).
SR =

S275 nm − 295 nm
S350 nm − 400 nm

(3.3)

3.3.4 Vertical Porewater Profiles
To observe trends in DOC quantity and quality, vertical porewater profiles of the optical
properties as well as the sampled chemical species were created for each site. To do so, the HZ
data collected from the MINIPOINT along with the stream water and groundwater data for a
given site were plotted in order of increasing depth. See Appendix B for the vertical porewater
profiles of DOC concentration, DO, SUVA254, SR, NO3-, SO42-, Cl-, and temperature for each
site.

43

3.4 Results
The average DOC concentration in stream water collected from the 39 sampling sites was
8.84 mg/L, which falls within the range for DOC concentration across Augusta Creek of 2-12
mg/L (Hamilton, Unpublished Data). The range of DOC concentration measured in stream water
was 5.31 to 13.63 mg/L with a median value of 8.91 mg/L. For groundwater, the average DOC
concentration sampled across the watershed was lower than that of stream water at 6.52 mg/L
and ranged from 2.42 to 14.15 mg/L with a median value of 5.79 mg/L. The average DOC
concentration for groundwater was similar between sampling years (i.e., 6.62 mg/L in 2015 vs.
6.52 mg/L in 2016) even though the sites were different, indicating that groundwater at 60 cm
depth has relatively consistent DOC concentrations.
3.4.1 Composite Vertical Porewater Profiles
Individual vertical porewater profiles for all 39 sites were initially averaged to create a
composite, watershed vertical porewater profile, representative of the average DOC quantity and
quality conditions in the HZ across Augusta Creek (Figure 13, graphs labeled “All Sites”).
Vertical profiles were then averaged by stream order for DOC, DO, SUVA254, and SR (Figure
13).
The vertical profiles showed decreasing DOC concentration with depth through the HZ
(i.e., 2.5 – 20 cm), in the mean watershed profile and in the mean profiles for each stream order
(Figure 13A). However, an increase in DOC concentration was observed between depths 7.5 and
10 cm in the third-order vertical profile (Figure 13A), which may be due to the effect of outliers
(e.g., sites with less consistent trends in DOC conditions] such as Site F1 (Figure B13), which
exhibited an increase in DOC concentration at intermediate HZ depths. In addition, a slight
increase in DOC concentration was observed between depths 2.5 and 5 cm in the first-order

44

vertical profile (Figure 13A). Again, this is likely due to variations in DOC concentration
profiles among the individual sites (see Appendix B).
Stream water across Augusta Creek was oxygenated with DO levels greater than 5 mg/L
during the sampling campaign. Dissolved oxygen concentrations decreased with depth through
the HZ, in the mean watershed profile and in the mean profile for each stream order, typically
becoming depleted (i.e., 2 mg/L DO or less) at deeper HZ depths (Figure 13B). Groundwater
measurements taken from 60 cm depth showed lower DO concentrations than stream water, but
higher than the DO concentrations observed at the bottom of the HZ.
Regarding the optical properties, SUVA254 decreased with depth in the HZ in the mean
watershed profile as well as in the mean first-order and third-order profiles (Figure 13C).
However, the second-order vertical profile showed an increase in SUVA254 at intermediate HZ
depths (i.e., ~7.5 – 10 cm), which may also be the result of outliers, including Sites F1 and F2
(Figure B13, B14). Further, SR increased with depth in the HZ in the mean watershed vertical
profile and in the mean profiles for each stream order (Figure 13D).
Increased variance in the optical indices of DOC quality were observed in the HZ when
compared to stream water and groundwater (Figure 13C, D), especially for SUVA254 (i.e., DOC
aromaticity). In addition, increased variance in DOC concentration was observed in the HZ when
compared to stream water and groundwater, although the difference is less notable (Figure 13A).

45

Figure 13: Composite vertical porewater profiles. Vertical profiles of A) mean DOC, B)
mean DO, C) mean SUVA254, and D) mean SR of all 39 sites and by stream order. Each plot
includes standard deviation bars; n=39 for “all sites”, n=16 for “1st order sites”, n=14 for “2nd
order sites” and n=9 for “3rd order sites.”

46

3.4.2 Rates of Change in DOC Quantity and Quality with Depth
Although the trends in DOC quantity and quality are consistent between the averaged
vertical porewater profiles for the watershed and each stream order (see Section 3.4.1, Figure
13), the vertical profiles for individual sites show more variation (see Appendix B). As a result,
further analyses were conducted on a site-by-site basis. In order to quantify the changes observed
in DOC quantity and quality along the vertical porewater profiles for each site, a rate constant (k,
units of cm-1) for DOC concentration (i.e., kDOC) as well as SUVA254 (i.e., k254) was calculated
over HZ depths (i.e., 2.5 cm to 20 cm), assuming a first-order reaction. The integrated rate law
for the first-order reaction is presented in Equation 3.4.
[DOC]x = [DOC]0 e−kx

(3.4)

To solve for kDOC, the natural log of DOC concentration was plotted against depth (i.e., x) and a
linear regression model was applied (Figure 14). The slope of the linear regression represents
the negative of kDOC. Similarly, for SUVA254, a k254 was calculated for each site.
The rate constants for DOC concentration and SUVA254 for all sites are presented in Table 2

Figure 14: Using a first-order rate law to determine rate constants. An example plot of
LN(DOC) versus distance below the streambed (cm), including a linear regression model. The
slope of the linear regression represents the negative of kDOC. In this example, kDOC = 0.081 cm-1.
47

Table 2: Summary of the rate constants for DOC concentration and SUVA254
Site

Stream Order

VHG (+, gaining)

A1
A2
A3
B1
B2
C1
C2
C3
C4
E1
E2
E3
F1
F2
F3
F4
G1
G2
H1
H2
H3
I1
I2
J1
K1
K2
L1
L2
L3
M1
M2
O1
O2
R1
R2
R3
R4
S1
S2

1
1
1
3
3
2
2
2
2
2
2
2
3
2
2
2
1
1
2
2
2
3
3
2
1
1
1
1
1
1
1
3
3
1
1
1
1
3
3

0.145

0.066
0.03
0.745
0.047
-0.655
-0.041
0.381
-0.447
-0.122
-0.117

0.058
0.038

-1.885
0.092
0.051
-0.544
-0.193
0.152
0.258
0.667

48

kDOC (cm-1)

0.058
-0.008
-0.006
0.086
0.073
0.097
0.059
0.107
0.039
0.142
0.081
0.073
-0.021
0
-0.005
0.013
0.049
0.065
0.035
0.053
0.075
0.101
0.064
0.066
0.06
0.039
0.077
0.013
-0.008
0.078
0.016
0.076
0.069
0.008
0.028
0.058
0.062
0.025
0.117

k254 (cm-1)

0.017
-0.013
-0.041
0.094
0.018
0.056
0.026
0.052
0.015
0.039
0.016
0.067
-0.04
-0.054
-0.032
-0.05
0.007
-0.001
-0.015
0.097
0.092
0.035
0.146
0.036
0.048
0.039
0.063
-0.018
0.01
-0.046
0.05
0.027
0.026
0.002
0.057
0.057
0.065
0.013
0.033

The majority of sites had positive kDOC values (i.e., 33 out of the 39 sites, see Table 2),
indicative of DOC uptake (i.e., consumption or loss). The median kDOC for the watershed was
0.059 cm-1, while the median for first-order sites was 0.044 cm-1, for second-order sites was
0.063 cm-1, and for third-order sites was 0.073 cm-1 (Figure 15A). Based on the median values,
the magnitude of the rate constant for DOC concentration increased with increasing stream order
(i.e., became more positive in higher stream orders, Figure 15A). To test for significance in kDOC
between stream orders, an ANOVA was performed using the statistical program, R (version
3.3.2) with an -value of 0.05. The null hypothesis is that the results are not significantly
different between groups. The F-value was 2.215 and p = 0.124, thus p > 0.05 so the differences
in kDOC between stream orders are not statistically significant.
Additionally, the majority of sites had positive k254 values (i.e., 29 out of the 39 sites, see
Table 2), indicative of decreasing SUVA254 with depth in the HZ. The median k254 for the
watershed was 0.026 cm-1, while the median for first-order sites was 0.014 cm-1, for second-order
sites was 0.031 cm-1, and for third-order sites was 0.027 cm-1 (Figure 15B). Based on the median
values, the magnitude of the rate constant for SUVA254 was the largest for second- and thirdorder sites, with the second-order sites exhibiting a slightly higher median k254 (Figure 15B). An
ANOVA was again used to test for significance in k254 between stream orders, using an -value
of 0.05. The F-value was 0.609 and p = 0.55, thus p > 0.05, so the differences in k254 between
stream orders are not statistically significant.
The rate constants for DOC concentration and SUVA254 were also compared between
upwelling and downwelling sites (Table 2). Figure 15C shows how VHG varies across the
stream network. The median VHG for the watershed was 0.047, while the median VHG for firstorder sites was -0.193, for second-order sites was -0.041, and for third-order sites was 0.058.

49

Thus, first-order sites exhibited the largest flux of water moving into the HZ. The median kDOC
(rate constant) for upwelling sites across Augusta Creek was 0.069 cm-1 and for downwelling
sites was 0.043 cm-1. For SUVA254, the median k254 for upwelling sites was 0.027 cm-1, and for
downwelling sites was -0.012 cm-1. As before, an ANOVA was used to test for significance
between rate constants at sites with positive VHG (i.e., upwelling) versus sites with negative
VHG (i.e., downwelling). For kDOC the F-value was 1.724 and p = 0.205, thus p > 0.05, so the
differences in kDOC between upwelling and downwelling sites are not statistically significant.
While, for k254 the F-value was 5.943 and the p = 0.025, thus p < 0.05, indicating that the
differences in k254 between upwelling and downwelling sites are statistically significant.

50

Figure 15: Changes in rate constants and VHG across the stream network. A) Boxplots
of kDOC for each stream order, B) boxplots of k254 for each stream order, and C) boxplots of
VHG for each stream order. The red circle (●) represents the mean.
51

3.4.3 DOC Metabolism
It was hypothesized in Section 3.1 that the dominant mechanism consuming DOC in the
HZ across Augusta Creek is aerobic, microbial respiration. To investigate this hypothesis,
“metabolism plots” were used, based on previous methods from Findlay and Sobczak (1996) and
Battin (1999). In these metabolism plots, stream DO concentration minus hyporheic DO
concentration is plotted against stream DOC concentration minus hyporheic DOC concentration
for a given site. If aerobic, microbial respiration is the only transformation mechanism
consuming DOC over that range in depth, then the point should fall on the 1:1 molar line (Figure
16), assuming that for every 1 mole of C consumed, 1 mole of O2 is consumed (Findlay and
Sobczak, 1996). This would account for minimal O2 consumption since the ratio is not corrected
for diffusion of O2 (Findlay and Sobczak, 1996).

Figure 16: Example of a metabolism plot. Plot of DO consumption versus DOC consumption
between two given depths. If aerobic respiration accounts for all DOC loss then points should
fall on the 1:1 molar line (–), assuming that 1 mole of O2 is consumed for every mole of C
consumed.
52

Points that fall to the left of the 1:1 molar line indicate that over the given range of depth, more
DO is removed than what can be explained by DOC consumption (via aerobic, microbial
respiration) alone, while points that fall to the right of the 1:1 molar line indicate that over the
given range of depth, more DOC is removed than what can be explained by DO consumption
(via aerobic, microbial respiration) alone (Figure 16). Metabolism plots were created for each
range in sampling depth from stream water to 20 cm (e.g., stream – 2.5 cm, 2.5 – 5 cm, and so
forth). All 39 sites were plotted on the graphs to allow for assessment of the dominant
biogeochemical transformation mechanism in the HZ on a site-by-site basis (Figure 17).
Overall, the hyporheic observations clustered predominantly to the left of the 1:1 molar
line in the “stream – 2.5 cm” plot and then shifted towards the right side of the 1:1 molar line as
HZ depth increased, indicating that more DOC was removed at intermediate HZ depths than
what can be explained by DO consumption alone (Figure 17). Additionally, at these
intermediate/deeper HZ depths (i.e., 7.5 to 20 cm) the observations showed little change in DO
concentration with depth (i.e., clustered along the horizontal dashed line), but still exhibited
changes in DOC, indicating that DOC might be providing electrons to other terminal electron
acceptors such as nitrate (NO3).

53

Figure 17: Metabolism plots for Augusta Creek. Metabolism plots including all 39 sampled
sites are presented between each measurement depth from stream water to 20 cm. However,
site F1 has been removed from the 7.5 cm – 10 cm plot as well as the 10 cm – 15 cm plot as it
is an outlier (i.e., shows significant increases in DOC over those depths). If aerobic, microbial
respiration accounts for all DOC loss then the points should fall on the 1:1 molar line (–),
assuming that 1 mole of O2 is consumed for every mole of C consumed.

54

3.5 Discussion
The averaged vertical porewater profiles (see Section 3.4.1, Figure 13) for Augusta Creek
in 2016 are consistent with the averaged vertical porewater profiles from 2015 (Chapter 2, Figure
2.6), showing decreasing DOC concentrations with depth through the HZ, and is accompanied by
decreasing DO concentrations. Both DOC and DO in the HZ decreased to concentrations lower
than that of either the overlying stream water or the underlying groundwater, indicating that
mixing of the two end members (i.e., SW and GW) alone cannot account for all of the observed
decreases in DOC and DO (e.g., Pinder and Jones, 1969). Thus, some combination of biological
(e.g., microbial respiration) and/or physicochemical (e.g., sorption to hyporheic sediments) DOC
removal reactions must be occurring in the HZ. The profiles also showed decreasing SUVA254
and increasing SR with depth in the HZ, indicating that DOC aromaticity and molecular weight
decreased through the HZ (Weishaar et al., 2003; Helms et al., 2008). These optical trends are
suggestive of microbially-driven DOC transformations (e.g., McKnight et al., 2001; Fellman et
al., 2010; Cory et al., 2011; Mann et al., 2012; Creed et al., 2015; Helton et al., 2015). Whereas
we can conclude from these averaged vertical profiles that DOC was consumed in the HZ at
these sites throughout the watershed, we cannot fully determine the dominant biogeochemical
transformation mechanisms of DOC in the HZ. To do so, sampling sites from across the entire
watershed need to be assessed individually, and a broader suite of biogeochemical measurements
may be necessary.
Thus, rate constants (k, units of cm-1) for DOC concentration (i.e., kDOC) as well as
SUVA254 (i.e., k254) over HZ depths (i.e., 2.5 cm to 20 cm) were calculated for all 39 sites (Table
2, Equation 3.4). The rate constants are informative because, at a minimum, they describe the
direction of reaction. For example, a positive rate constant would indicate that DOC

55

concentration is decreasing (i.e., showing consumption of DOC) with HZ depth, while a negative
rate constant would indicate increasing DOC concentration (e.g., production of DOC) with HZ
depth (Equation 3.4). In addition, rate constants describe the relative magnitude of change (i.e.,
processing potential or removal potential). If the rate constant is large and positive for a site, then
that site would likely be a strong sink for DOC. Therefore, sites with large (positive or negative)
rate constants, in terms of magnitude, indicate the potential for high rates of change (i.e.,
transformation) in DOC quantity and quality at those locations in the river network. Based on the
results presented in Table 2, the majority of sites (see Section 3.4.2) had positive rate constants
for DOC concentration and SUVA254, indicating that across the watershed, DOC was being
removed in the HZ at most sampled locations and that the residual DOC showed decreasing
aromaticity with depth through the HZ (e.g., Weishaar et al., 2003). These result are consistent
with the averaged vertical profiles and suggest that on a site-by-site basis, aerobic, microbial
respiration may be the dominant transformation mechanism for DOC across the watershed (e.g.,
McKnight et al., 2001; Cory et al., 2011; Mann et al., 2012; Creed et al., 2015; Helton et al.,
2015). This would support findings from previous small-scale field and mesocosm studies
(Findlay and Sobczak, 1996; Schindler and Krabbenhoft, 1998; Sobczak and Findlay, 2002;
Zarnetske et al., 2011).
To determine whether or not aerobic, microbial respiration is controlling the observed
trends in DOC quantity, metabolism plots were used (Figure 17; Findlay and Sobczak, 1996;
Battin, 1999). The plots showed that at all HZ depths the hyporheic observations clustered
around the 1:1 molar line, indicating that 1) aerobic, microbial respiration is predominantly
controlling the trends in DOC quantity observed through the HZ and 2) that this dominant
transformation mechanism is consistent across the river network. These results support the first

56

hypothesis that aerobic, microbial respiration is the dominant transformation mechanism of DOC
in this river network. However, the deviation from the 1:1 molar line, exhibited by individual
observations suggests that additional physicochemical processes (e.g., dilution, sorption to
hyporheic sediments, intersecting lateral flowpaths) and/or biological reactions (e.g., anaerobic,
microbial respiration using alternative electron acceptors such as NO3, SO4) are occurring in the
HZ at some of the sites. For instance, the observations clustered predominantly to the left of the
1:1 molar line in the “stream – 2.5 cm” plot and then shifted towards the right side of the 1:1
molar line as HZ depth increased (Figure 17). It is likely that within the shallow HZ, there is an
additional source of organic C in the form of particulate organic carbon (POC) such as leaves
and buried logs (e.g., Stelzer et al., 2015). The microbes may also be using the POC for
respiration, which would account for the extra, “missing” oxygen (Findlay and Sobczak, 1996).
However, the abundance of POC often decreases with depth in a streambed (e.g., Kaplan and
Newbold, 2000; Findlay and Sobczak, 2000). Given this for Augusta Creek, it would result in the
shift towards the right of the 1:1 molar line at deeper HZ depths.
At a depth of ~7.5 cm and deeper in the HZ, many of the hyporheic observations also
began to show little to no change in DO concentrations (i.e., they clustered along the horizontal
dashed line) with depth, while still residing predominantly to the right of the 1:1 molar line
(Figure 17). This suggests that at deeper HZ depths, DOC is still consumed despite little to no
change in DO. Since sorption of DOC to hyporheic sediments tends to be in equilibrium with the
DOC concentration of upwelling/downwelling waters in the HZ (e.g., Day et al., 1994; Kaplan
and Newbold, 2000), the consumption of DOC at deeper HZ depths is likely due to a shift in
electron acceptors used in microbial metabolism. Nitrate (NO3-) or sulfate (SO42-) may be used
for anaerobic, microbial respiration at deeper HZ depths as the system becomes anoxic (e.g.,

57

Hedin et al, 1998). When using NO3- for anaerobic, microbial respiration, 4 moles of NO3- are
consumed for every 1 mole of C consumed. Likewise, when using SO42- for anaerobic, microbial
respiration, 2 moles of SO42- are consumed for every 1 mole of C consumed.
Furthermore, the rate constants as well as the metabolism plots can also be used to start to
address questions about how the role of the HZ as a processor of stream DOC varies across the
river network. The results indicate that the HZ acts as a sink for DOC and that aerobic, microbial
respiration is one of the dominant transformation mechanisms consuming DOC in the HZ.
However, this HZ function (i.e., removal of DOC) may vary based on location in the watershed.
Rate constants for DOC concentration and SUVA254 were compared between stream orders as
well as between sites with varying VHG (i.e., upwelling vs. downwelling, Section 3.4.2) to
observe changes in this hyporheic processing potential across the watershed. While the rate
constants for both DOC concentration and SUVA254 showed increases in magnitude with
increasing stream order (i.e., rate constants became more positive at higher stream orders, Figure
15A), the ANOVA results indicate that the observed changes with stream order are not
statistically significant (Section 3.4.2). This suggests that the potential for transformation in both
DOC quantity and quality in the HZ is relatively similar across stream orders.
Changes in rate constants for DOC concentration between upwelling and dowelling sites
were also not significantly different, based on the ANOVA results (Section 3.4.2), further
suggesting that the potential for DOC transformation in the HZ is similar across the watershed,
despite variation in local VHG among sites. However, the changes in rate constants for
SUVA254 between upwelling and downwelling sites are statistically significant based on the
ANOVA results (Section 3.4.2). This may indicate that local flow patterns (i.e., VHG) impact the
potential for transformation in DOC quality in the HZ, across the watershed, considering that

58

DOC quality can change independently of DOC concentration (Lutz et al., 2012). These
ANOVA results may have also been impacted by the lower sampling number since not all 39
sites have VHG measurements (Table 2). In addition, of the eight downwelling sites, four of
them have negative SUVA254 rate constants, which impacts the average downwelling rate
constant (Section 3.4.2, Table 2). However, of those four sites, three of them are for Site F (i.e.,
F1, F2, and F4). Site F is adjacent to a large wetland complex (Figure 12), which may have
complicated the trends in DOC quantity and quality and the associated reactions more so than at
other sampling sites less dominated by a large wetland outflow. Finally, it is also important to
consider that changes in stream discharge due to small storm events during the August 2016
sampling period did result in small stream stage changes (Figure 10) that may have impacted the
VHG at individual sites, causing statistically significant variation in rates between upwelling and
downwelling sites.
While the rate constants indicate that the potential for DOC transformation in the HZ is
similar (for the most part) across the watershed, rate constants need to be paired with the vertical
hydrological flux (i.e., the rate of water upwelling or downwelling through the HZ) as well as
stream transport rates to assess the overall role of the HZ in DOC cycling across the river
network (i.e., large sink versus small sink). This is because the amount of DOC processing in the
HZ is not only dependent on the rate constant, but also the local hydrology controlling the supply
of stream water containing DOC to the HZ in relation to the total flux down the stream channel.
For example, a site with a large, positive rate constant for DOC concentration (i.e., high potential
for DOC removal) and a large, downwelling VHG (i.e., indicative of a constant supply of DOC,
DO and nutrients) means that the HZ at that site is likely a large sink for DOC in the river
network. On the other hand, a site with a large, positive rate constant for DOC concentration, but

59

a small, downwelling VHG is likely not as big of a sink for DOC since less water is moving
through the HZ, thus hydrologically limiting microbial metabolism. Therefore, we used the
product of kDOC and VHG to represent the function of the HZ (i.e., fDOC; size of sink) for each
stream order in terms of DOC concentration. This HZ function was calculated only for sites with
positive kDOC and negative VHG values (i.e., n=6) since those sites indicate the potential for
removal of DOC and stream water flowing into the HZ. This is unlike sites with upwelling
groundwater, which should theoretically limit the depth of the HZ and limit the supply of labile
DOC to the HZ. Thus, stream orders with a large, negative fDOC indicate that the HZ in those
reaches of the river network functions as a large sink for DOC. This process was also repeated
for DOC quality, in terms of SUVA254 (i.e., n=4; Figure 18).

Figure 18: Plots of HZ function. A) fDOC values for first- and second-order streams (n=3 for
each), and B) f254 values for first- and second-order streams (n=2 for each).The open-circle
represents the mean fDOC value +/- the max/min value. fDOC values for third-order streams are
not included, due to insufficient VHG data needed to calculate fDOC.

The first-order streams exhibit more negative fDOC and f254 values when compared to the
second-order streams. There were no third-order sites with a positive rate constant as well as a
negative VHG, which suggests there were no third-order sites observed that have a HZ that is
60

effective at DOC removal from the stream. Thus, despite the smaller kDOC value, and therefore
lower processing potential for first-order streams (Figure 15A), the large, downwelling VHG in
those locations likely allows for more removal of DOC in the HZ (Figure 18). This indicates that
the HZ may function as a larger sink for DOC in the headwaters of this river network than in
higher stream orders, with little to no HZ processing of DOC in larger streams. However, due to
the small number of sampling sites that could be used for this analysis, these results should be reevaluated in the future using additional downwelling sites across the river network.

61

3.6 Conclusions
We identified the dominant biogeochemical transformation mechanism(s) for DOC in the
HZ across Augusta Creek using watershed-wide trends in hyporheic DOC quantity and quality,
dissolved oxygen (DO) measurements, and local vertical flow gradients (i.e., upwelling vs.
downwelling). The averaged, vertical porewater profiles along with the rate constants for
individual sites showed decreasing DOC concentration with depth through the HZ, along with
decreasing SUVA254. These results, as well as the metabolism plots, demonstrate that the HZ
likely acts as a sink for DOC and that aerobic, microbial respiration may be the dominant
processing mechanism consuming DOC in the HZ across the watershed. Additionally, the role of
the HZ as a processor of stream DOC is likely more significant (i.e., acts as a larger sink) in
headwater streams, where strong, downwelling VHG occur, despite the smaller, observed kDOC
and k254 values (i.e., lower processing potential). Lastly, the ANOVA results as well as the
averaged vertical porewater profiles for SUVA254, which showed large variation in SUVA254
with depth in the HZ indicate the need for further studies focused on hyporheic DOC quality. It
is critical to understand DOC quality, since it can affect the fate of DOC in the HZ as well as the
DOC and solutes transported to downstream ecosystems.

62

CHAPTER 4: USING IN-SITU OPTICAL SENSORS TO STUDY DISSOLVED ORGANIC
CARBON DYNAMICS OF STREAMS AND WATERSHEDS: A REVIEW†
4.1 Abstract
It is important to understand how dissolved organic carbon (DOC) is processed and
transported through stream networks because DOC is a master water quality variable in aquatic
ecosystems. High-frequency sampling is necessary to capture important, rapid shifts in DOC
source, concentration, and composition (i.e., quality) in streams. Until recently, this highfrequency sampling was logistically difficult or impossible. However, this type of sampling can
now be conducted using in-situ optical measurements through long-term, field-deployable
fluorometers and spectrophotometers. The optical data collected from these instruments can
quantify both DOC concentration and composition properties (e.g., specific ultra-violet
absorbance at 254 nm, spectral slope ratio, and fluorescence index). Previously, the use of these
sensors was limited to a small number of specialized users, mainly in Europe and North
America, where they were used predominantly in marine DOC studies as well as water treatment
and management infrastructure. However, recent field demonstrations across a wide range of
river systems reveals a large potential for the use of these instruments in freshwater
environments, heightening interest and demand across multiple environmental research and
management disciplines. Hence, this review provides an up-to-date synthesis on 1) the use of
spectroscopy as a diagnostic tool in stream DOC studies, 2) the instrumentation, its applications,
potential limitations and future considerations, and 3) the new watershed DOC research
directions made possible via these in-situ optical sensors.

†

Reprinted from Science of the Total Environment, 575, Ruhala, S.S and Zarnetske J.P., Using in-situ
optical sensors to study dissolved organic carbon dynamics of streams and watersheds: A review, 713723, Copyright (2017), with permission from Elsevier. http://dx.doi.org/10.1016/j.scitotenv.2016.09.113.

63

4.2 Introduction
Rivers were traditionally thought to play an insignificant role in the global carbon (C)
cycle (Battin et al., 2008). However, recent studies reveal that rivers globally receive 4 petagrams
(Pg) of terrestrial organic C and generate 1.8 ± 0.25 Pg C as carbon dioxide (CO2) each year; a
rate 6 times greater than lakes and reservoirs combined (Battin et al., 2008; Raymond et al.,
2013). This indicates that rivers are a significant source of CO2 to the atmosphere and that most
of the organic C entering rivers is undergoing significant in-stream biogeochemical processing
(Battin et al., 2008). Thus, efforts must be made to better understand C cycling in rivers, so that
rivers can be considered in future global C models. Dissolved organic carbon (DOC) is often the
dominant form of organic C in streams and plays several critical roles in aquatic ecosystems.
DOC fuels stream metabolism, regulates additional biogeochemical cycles (e.g., nitrogen
transformations), and impacts trace metal complexation and transport (Cory et al., 2011; Aiken,
2014). As a result, stream DOC dynamics influence C cycling on local to global scales (Battin et
al., 2009). Therefore, it is important to measure and identify how DOC is processed (e.g., rates,
transformations) and transported (e.g., its sources, pathways, and quantity) in riverine
environments.
DOC is composed of a complex grouping of molecules with varying chemical structures
(e.g., molecular weight and aromaticity). The chemical structure of DOC is a determining factor
of DOC quality and is largely dependent on its source material (i.e., allochthonous versus
autochthonous) (Cory et al., 2011). Several optical properties of DOC, determined via
absorbance and fluorescence spectroscopy, can be used to assess the chemical structure of DOC,
which in turn can be used to infer DOC qualities and bioavailability. Ultraviolet-visible (UVVIS) absorbance spectroscopy can be used to infer aromaticity and molecular weight (e.g., Cory

64

et al., 2011; Jollymore et al., 2012), while fluorescence spectroscopy provides insight on redox
state, reactivity, and source material (e.g., Fellman et al., 2010; Cory et al., 2011). DOC quality
is a critical parameter that must be included in DOC export studies as the quality of DOC
impacts its bioavailability and reactivity (i.e.,, lability), which ultimately affects downstream
ecosystems and water quality (Fellman et al., 2010; Cory et al., 2011).
Until recently, optical measurements of DOC were done using field water collections
(i.e., grab samples) that were later analyzed in the laboratory (e.g., Hood et al., 2006; Spencer et
al., 2009; Inamdar et al., 2011). Utilizing lab measurements of optical properties to assess DOC
quality is quick, while still maintaining high analytical precision, and relatively inexpensive
when compared to chemical analyses (Fellman et al., 2010). However, there is a lack of temporal
resolution in the data from using this method, because grab samples are often collected at low
frequencies. Recent studies have revealed that higher temporal resolution of DOC measurements
allows for better understanding of DOC flux and C budgets (e.g., Saraceno et al., 2009;
Jollymore et al., 2012; Pellerin et al., 2012; Wilson et al., 2013; Jones et al., 2014; Grayson and
Holden, 2016), and allows for further exploration of new ecological theories and scales of
processes (e.g., Pulse-Shunt Concept, Raymond et al., 2016). Thus, there is a growing demand
for high-quality and cost effective, high-frequency measurements of DOC quantity and quality in
riverine environments. A potential solution to this demand are the emerging optically-based field
sensor technologies (i.e., long-term, field-deployable fluorometers and spectrophotometers such
as the WET Labs, WETStar FDOM Fluorometer and s::can, spectro::lyser), because they have
data collection frequencies of minutes. This high-frequency sampling helps to capture changes in
both DOC quantity and quality as the result of shifting pathways and sources (e.g., during storm
events when it is difficult to manually collect grab samples) as well as changes due to in-stream

65

biogeochemical transformations at scales that otherwise would not be seen (e.g., sub-diurnal
dynamics, diurnal fluctuations, intra-seasonal trends; Spencer et al., 2007; Pellerin et al., 2012).
These in-situ optically-based field instruments have previously been used by a limited
group of specialized users, largely in Europe and North America. The focus of these earlier uses
was on DOC dynamics in marine and coastal environments (e.g., Klinkhammer et al., 1997;
Chen et al., 1999; Chen and Gardner, 2004; Kowalczuk et al., 2010; Gueguen et al., 2012;
Etheridge et al., 2014a; Etheridge et al., 2014b) as well as in water treatment and management
infrastructure (e.g., Langergraber et al., 2004; Brinkman and Hozalski, 2011; Boënne et al.,
2014; Graham et al., 2015). However, recent riverine field demonstrations reviewed in detail in
Section 4 have heightened demand for these optical sensors in many freshwater disciplines,
including watershed ecology, biogeochemistry and hydrology (e.g., Spencer et al., 2007;
Saraceno et al., 2009; Pellerin et al., 2012).
Despite this growing interest and demand, there is no comprehensive synthesis of the
state of the science on using these new technologies (i.e., in-situ optical sensors) in the field.
Consequently, potential new users face a daunting task of compiling this information from a
wide range of sources (e.g., primary literature to government reports), which inhibits the
potential adoption and applications of these new technologies. Consequently, we provide a
review on in-situ optical methods used for studying DOC dynamics in riverine environments,
specifically via long-term, field-deployable spectrometers and fluorometers. This review will
cover 1) what we can learn about DOC from the use of optical properties 2) why in-situ, highfrequency measurements are needed, 3) the instrumentation, its diagnostic uses, limitations, and
future considerations, and 4) an illustrative synthesis of future watershed DOC research
directions discovered through recent field demonstrations of these instruments.

66

4.3 Spectroscopy as a Diagnostic Tool for DOC
Below we briefly review how absorbance and fluorescence spectroscopy is used for
assessing DOC composition, and in turn DOC quality, across a wide range of natural waters
(McKnight et al., 2001; Belzile et al., 2006; Fellman et al., 2010; Cory et al., 2011; Jollymore et
al., 2012; Lee et al., 2015). Dissolved organic matter (DOM) is operationally defined as any
organic material that is able to pass through a filter of a particularly identified pore size (e.g.,
typically 0.7 µm, 0.45 µm or smaller; Cory et al., 2011; Jollymore et al., 2012). Since DOM is
typically about 50% C by mass (Cory et al., 2011), DOM is measured as DOC. DOC consists of
a complex mixture of compounds that vary in size, weight, aromaticity, and reactivity, making it
difficult to study DOC composition, and therefore its quality via direct molecular techniques.
Hence, spectroscopy methods are a relatively quick and non-invasive means to observe and infer
DOC and DOM composition and quality.
4.3.1

Absorbance Spectroscopy
UV-VIS absorbance spectroscopy is used to determine optical properties associated with

the light-absorbing or chromophoric fraction of DOM (i.e., CDOM; Fellman et al., 2010; Cory et
al., 2011). UV-VIS absorbance (i.e., absorbance from 220 to 720 nm) of CDOM has been shown
to be a strong proxy (i.e., r2 > 0.95, based on in-situ data) for DOC concentration (Waterloo et
al., 2006; Jollymore et al., 2012; Jeong et al., 2012; Avagyan et al., 2014; Jones et al., 2014) and
can also be used to infer aromaticity and molecular weight. It should be noted that UV-VIS
absorbance may not be fully representative of all quality properties since CDOM only makes up
a fraction of the total DOM pool (Cory et al., 2011).
Algorithms utilizing Beer Lambert’s Law are used to determine DOC concentration from
a given absorbance (e.g., 350 nm), as seen in Waterloo et al., 2006 (Equation 4.1).

67

DOC = 0.905 x Absorbance350 nm

(4.1)

While specific UV absorbance at 254 nm (i.e., SUVA254) is positively correlated with
DOC aromaticity (Cory et al., 2011). Here, the absorbance at 254 nm is divided by the DOC
concentration of the sample (units are L mg-1 C m-1) (Equation 4.2). A wavelength of 254 nm is
used because electron structures associated with aromatic C molecules absorb energy at this
wavelength, while other structures do not (Weishaar et al., 2003). Therefore the absorbance
reading is directly related to the amount of aromatic C within the total DOC sample.
SUVA254 =

Absorbance254 nm
DOC concentration

(4.2)

And lastly spectral slope ratio (SR) can be used to infer molecular weight. SR is a
dimensionless ratio of the slope of the 275 nm to 295 nm (S275-295) absorbance spectra divided by
the slope of the 350 nm to 400 nm (S350-400) absorbance spectra (Equation 4.3) and exhibits a
negative correlation with molecular weight.
SR =

S275 nm − 295 nm
S350 nm − 400 nm

(4.3)

4.3.2 Fluorescence Spectroscopy
In addition to chromophores, fluorophores (i.e., molecules that absorb and re-emit light)
also make up a fraction of DOM in natural waters (Fellman et al., 2010). When a fluorophore
(e.g., fluorescent DOM, FDOM) absorbs light energy, an electron is excited (i.e., transitions)
from the original ground state to a higher electronic state (e.g., S0 to S2 in Figure 19). Within
each electronic state there are numerous vibrational energy levels (e.g., 0, 1, 2 in Figure 19). The
electron will often rapidly relax to the lowest vibrational energy level of S1 prior to emission, via
internal conversion (Figure 19; Lakowicz, 2006). Then as the electron returns to ground state, the
68

energy is emitted as light or fluorescence (Figure 19). The specific excitation and emission
wavelengths at which fluorescence occurs is dependent on the concentration, chemical
composition, and chemical structure (e.g., bonds between molecules) of the compound
(Lakowicz, 2006; Fellman et al., 2010; Cory et al., 2011). Therefore, fluorescence can provide
insight on DOC quantity and quality (e.g., redox state, reactivity, and source material; Fellman et
al., 2010; Cory et al., 2011). However, it is important to note that FDOM only makes up
approximately 1% of the total DOM pool (Cory et al., 2011). Therefore similar to CDOM, a
limitation of using fluorescence spectroscopy is that only part of the “DOC pool” (rather than the
“whole pool”) is observable.

Figure 19: A simplified Jablonski diagram. A simplified Jablonski diagram showing the
process of fluorescence. Modified from Lakowicz (2006).

Despite its small fraction, FDOM is used as a proxy for total DOC concentration
(Fellman et al., 2010; Cory et al., 2011). To do so, concentration of FDOM is measured in parts
per billion (ppb) quinine sulfate equivalents (QSE), using a fluorometer, as seen in Saraceno et
69

al. (2009), where Vsig is the output voltage for the sample, VCW is the output voltage for clean
(i.e., blank) water, and SF is an instrument specific scaling factor (Equation 4.4; Saraceno et al.,
2009).
ppb QSE = (Vsig − VCW ) x SF

(4.4)

FDOM is also used to infer DOC quality properties. A fluorescence index (FI) value can
be calculated to assess potential FDOM source material. The FI uses an excitation wavelength of
370 nm and the resulting emission intensities at 470 nm and 520 nm are then divided (Equation
4.5; McKnight et al., 2001; Fellman et al., 2010).
FI =

470nm Emission Intensity
520nm Emission Intensity

(4.5)

High FI values (i.e., ~1.8) are indicative of microbial or autochthonous sources, while low FI
values (i.e., ~1.2) are indicative of terrestrial or allochthonous sources (McKnight et al., 2001).
In addition to FI, three-dimensional excitation emission matrices (i.e., EEMs) are
commonly used to assess DOC composition. EEMs are produced using multiple excitation
wavelengths and measuring the resulting emission intensities across a range of wavelengths
(Fellman et al., 2010; Cory et al., 2011). This technique is used frequently to study DOC
composition in terms of molecular groupings and functions as EEMs contain a large amount of
information on the origin and processing of DOC (Fellman et al., 2010; Cory et al., 2011).
Additional optical properties including protein-like components (i.e., tyrosine-like and
tryptophan-like components), humic-like components, the freshness index, humification index,
and redox index can also be determined using fluorescence spectroscopy to better understand
DOC composition in natural waters (e.g., see Fellman et al., 2010; Khamis et al., 2015; Bieroza
and Heathwaite, 2016).

70

4.4 Technology and Limitations
4.4.1 In-Situ Optical Instruments
Previous work showed that DOC concentration and composition are dynamic and can
change over short time scales (i.e., hours to days) (Carstea et al., 2009; Cory et al., 2011;
Inamdar et al., 2011; Neal et al., 2012; Grayson and Holden, 2012; Strohmeier et al., 2013; Jones
et al., 2014; Mast et al., 2016). Therefore, it necessary to collect long-term, high-frequency data
when studying DOC export as small and/or rapid shifts in DOC quantity and quality are missed
with less frequent sampling (Figure 20; Bowes et al., 2009; Inamdar et al., 2011; Jollymore et al.,
2012; Jones et al., 2014; Blaen et al., 2016). In-situ optical measurements via long-term, fielddeployable spectrophotometers and fluorometers enable researchers to collect this valuable high
resolution temporal data using the spectroscopy methods discussed above. This review focuses
on two, representative instruments frequently used in published field studies to date: 1)
submersible FDOM fluorometers (e.g., WET Labs, WETStar FDOM fluorometer, Figure 21A)
and 2) submersible spectrophotometers (e.g., s::can, spectro::lyser, Figure 21B). For brevity, this
review specifically discusses the WET Labs FDOM sensor and the s::can spectro::lyser as welldocumented examples of these technologies (i.e., these sensors have been and are used in many
published studies). However, there are many other manufacturers developing and producing insitu optical instruments based upon the same fundamental optical methods including, the
SeaPoint UV fluorometer (SeaPoint Sensors, Inc.), the Cyclops-7 fluorometer (Turner Designs),
the microFLU CDOM fluorometer (TriOS), the UviLux fluorometer (Chelsea Technologies
Group), the Satlantic UV spectrophotometer, and the ABB AV400 UV absorbance probe.

71

Figure 20: Plots of DOC concentration versus time using a variation of sampling
frequencies. Plots of DOC concentration versus time created from data collected at different
sampling intervals in order to highlight the importance of high-frequency sampling in DOC
dynamics and export studies (Image source: Jollymore et al., 2012).

FDOM fluorometers use a light emitting diode (LED) as a light source and most have a
fixed single excitation/emission pair; although there are less well documented and reliable
scanning wavelength fluorometers, which measure the emission signal over a range of
wavelengths (Saraceno et al., 2009; Lee et al., 2015). The WETStar FDOM fluorometer from
WET Labs has a 7 mm optical path length along which light with a fixed wavelength of 370 nm
is emitted to excite the FDOM in stream water. The resulting emission at 460 nm is measured to

72

quantify the amount of FDOM present using Equation 4.4 (Saraceno et al., 2009). This
measurement is typically taken at 1 hertz (Hz) for 30 s every 15 min, 30 min, or every hour and
stored on a data logger (e.g., CR1000 data logger, Campbell Scientific). Most FDOM
fluorometers can function at depths up to several hundred meters and across a wide temperature
range of 0 to 30 degrees Celsius ( ͦ C; Bergamaschi et al., 2009). However, a temperature
correction algorithm may need to be applied in studies with large temperature fluctuations (see
Section 4.3.2). Many FDOM sensors also have automated systems to reduce biofouling (e.g.,
bio-wipers; Bergamaschi et al., 2009). FDOM sensors are relatively inexpensive (e.g., $30006000) compared to CDOM sensors and have low power requirements (due to the use of LEDs),
typically consuming 500 milliwatts or less (Bergamaschi et al., 2009; Conmy et al., 2014).

Figure 21: In-situ optical instruments. Images and dimensions of example in-situ optical
sensors: A) WETStar FDOM Fluorometer from WET Labs and B) spectro::lyser from s::can.
(Image from A) WET Labs and B) s::can).
73

On the other hand, submersible spectrophotometers measure absorbance across the UVVIS spectrum. For example, the spectro::lyser from s::can was the first submersible
multiparameter probe that could measure absorbance of CDOM in stream water over the UVVIS range of 220 to 720 nm (Grayson and Holden, 2012). It can also measure NO3-N, total
suspended solids (TSS), and turbidity (Table 3). The spectro::lyser uses a xenon flash lamp and
256 pixel array detector to measure the resulting absorbance of CDOM at a 2.5 nm wavelength
interval over an open path length (Jollymore et al., 2012). The path length on the spectro::lyser
can be selected by user, providing users with several options to meet different water conditions.
For example, an instrument with a path length of 100 mm can accurately measure DOC
concentrations from approximately 0 to 15 mg/L (Jollymore et al., 2012; Grayson and Holden,
2016), however higher DOC concentrations (i.e., up to ~70 mg/L) can be measured using smaller
path lengths of 35 mm or 5 mm, but with considerable reduced accuracy (Koehler et al., 2009;
Grayson and Holden, 2016). Similar to FDOM sensors, these instruments also have systems to
reduce biofouling. For example, they use bio-wipers or compressed gas to clean the optical
window at a given interval of time (Etheridge et al., 2014a; Grayson and Holden, 2016).
However, biweekly to monthly maintenance is still required as biofouling is not completely
eliminated when using any antifouling system (Downing et al., 2009; Jollymore et al., 2012).
The spectro::lyser can take measurements over the entire wavelength spectrum (i.e., 220 to 720
nm) at a rate as fast as every 2 min, but typical deployments monitor every 15 min, 30 min, or
every hour, which are then stored on an internal data logger. External data loggers (e.g., CR1000
data logger, Campbell Scientific) can also be used with most of these in-situ sensors, including
the spectro::lyser, for high-frequency sampling over extended deployment periods (e.g.,
monitoring at a 15 min interval for greater than 2 weeks, which may exceed the internal

74

memory). When compared to FDOM sensors, the spectro::lyser requires a larger power supply to
power the xenon lamp, although there are numerous options for sustained power including
battery, solar, and wind, which are helpful for remote field deployment.
4.4.2 Effects of Temperature on Sensors
It is important to note that temperature and turbidity can both interfere with optical
measurements when using in-situ FDOM fluorometers and spectrophotometers. Temperature has
an insignificant effect on absorbance readings of UV-VIS sensors, however signals received by
FDOM sensors can be greatly affected due to thermal quenching (Baker, 2005; Downing et al.,
2012; Conmy et al., 2014; Lee et al., 2015). As temperature increases it will cause electrons to
return to the original ground state via radiationless decay, therefore reducing the fluorescence
signal (i.e., reducing measured (raw) FDOM values by an average of 0.8-1.5% for every 1 ͦ C
from 1 to 25 ͦ C) of FDOM (Watras et al., 2011; Downing et al., 2012; Lee et al., 2015). The
extent of thermal quenching depends on the degree of exposure to the heat source, which varies
with organic matter composition (Baker, 2005). In environments with large temperature
fluctuations (e.g., >10 ͦ C), thermal quenching can result in underestimating FDOM
concentrations. Thus users of FDOM sensors should: 1) apply temperature compensation
equations to field data (e.g., Watras et al., 2014) or 2) use in-situ spectrophotometers as an
alternative in studies where temperature variability will be a factor (Lee et al., 2015).
4.4.3 Effects of Turbidity on Sensors
Turbidity is an inherent issue for both FDOM and UV-VIS sensors as it fundamentally
interferes with light passage and detection. Suspended particles that make up turbidity can absorb
and scatter light thus altering instrument sensor detection (Downing et al., 2012). As a result,
UV-VIS sensor outputs increase linearly with increasing turbidity as the suspended particles

75

absorb more light than the CDOM would alone. If not accounted for, this turbidity effect results
in over estimating CDOM concentration. On the other hand for the FDOM sensor, the scattering
of light due to suspended particles will reduce both the excitation signal from the FDOM light
emitter and the resulting emission signal to the FDOM detector. If not accounted for, this
turbidity effect results in underestimating FDOM concentration (Downing et al., 2012). In fact,
FDOM sensors are more sensitive to turbidity than UV-VIS sensors, as FDOM signals decrease
exponentially with increasing turbidity (Lee et al., 2015). Saraceno et al. (2009) demonstrated
the effect of turbidity on FDOM measurements by including two FDOM sensors, one of which
filtered the water prior to measurement and the other that left the water unfiltered. The study
showed that at peak discharge during a storm flow event in a stream, when turbidity was highest,
the unfiltered FDOM sensors received a dampened signal, thus underestimating the amount of
FDOM in the stream by >10% during that point in time (Figure 22; Saraceno et al., 2009). Thus,
FDOM field data should be calibrated using turbidity correction equations. To do so, an in-situ
turbidimeter must be deployed to simultaneously measure turbidity (Downing et al., 2012). Once
turbidity is known, correction equations can be developed based on the specific fluorometer and
turbidimeter used as well as the matrix waters, since the extent of reduction of the FDOM signal
due to light attenuation is dependent on both the instrument and nature (composition and size) of
suspended particles (Downing et al., 2012; Khamis et al., 2015). For example equations see
Saraceno et al. (2009), Downing et al. (2012), and Lee et al. (2015). To account for the turbidity
effect on CDOM measurements, some UV-VIS instruments, including the s::can spectro::lyser,
estimate turbidity directly from its UV-VIS measurements (Jollymore et al., 2012).
Due to the potential interference of temperature and turbidity, it is recommended that
studies using in-situ optical instruments 1) use closed path FDOM sensors with a filtering system

76

(e.g., see Saraceno et al., 2009) to reduce turbidity effects, and 2) deploy probes for temperature
and turbidity in order to correct the FDOM signals received by the sensors with instrument
dependent correction functions (Downing et al., 2012; Khamis et al., 2015; Lee et al., 2015).
Similarly, turbidity measurements and corrections must be made for UV-VIS instruments, if
turbidity correction equations are not already built into the sensor system.
In addition to specific temperature and turbidity interferences, there are a number of other
more general logistical strengths and limitations to using FDOM and UV-VIS sensor systems in
the field that have been discussed in other studies (e.g., Bergamaschi et al., 2009; Downing et al.,
2012; Jollymore et al., 2012). We briefly summarize these strengths and limitations of the
technology in Table 3.

Figure 22: Plot of discharge and FDOM concentration versus time. Both unfiltered and
filtered FDOM is plotted showing the importance of using a filter on in-situ FDOM sensors as
turbidity can interfere with the sensor, reducing the FDOM signal (Image from Saraceno et al.,
2009).

77

Table 3: Summary of the known strengths and limitations of in-situ optical instruments.

78

4.5 Examples of Utility in Riverine DOC Studies
High-frequency, in-situ optical monitoring is emerging as a critical tool for understanding
C transport and processing from hillslopes to streams and from streams to downgradient
ecosystems (Waterloo et al, 2006; Saraceno et al., 2009; Jones et al., 2014; Blaen et al., 2016;
Grayson and Holden, 2016; Raymond et al., 2016). In order to use in-situ optical methods (e.g.,
FDOM fluorometers and spectrophotometers) most effectively, they should be coupled with
discrete DOC sampling (Downing et al., 2009; Saraceno et al., 2009; Pellerin et al., 2012;
Sobczak and Raymond, 2015). The discrete samples allow for direct calibration of DOC
concentrations measured from CDOM and FDOM proxies against laboratory measurements as
well as allow for quantification of optical properties that require different measurements outside
of the range of in-situ optical instruments (e.g., FI values). Automatic pump samplers (e.g.,
Teledyne ISCO automatic water sampler) are an option for remotely collecting these discrete
DOC samples. Some recent studies focused on DOC cycling in stream catchments have
leveraged these high-frequency, in-situ optical monitoring technologies to provide new insight
on 1) seasonal, event, and diurnal variations in DOC quantity and quality, 2) sources and
pathways of DOC to the stream channel and 3) annual DOC flux. These studies and insights are
summarized below to illustrate potential future uses of these in-situ optical DOC sampling
technologies.
4.5.1 Revealing Seasonal, Event, and Diurnal Variations in DOC Quantity and Quality
High-frequency sampling over extended time periods (e.g., multiple months to years)
allows for analysis of changes in DOC across numerous time scales. Several studies with
sampling periods of months to years have been able to observe seasonal variations in DOC
quantity and quality consistent with other long-term DOC studies. For example, Wilson et al.

79

(2013) examined DOC processes in the steep, humid, forested catchment of Bigelow Brook in
Massachusetts, USA. Using a Turner Designs Cyclops-7 CDOM fluorometer, Wilson et al.
(2013) sampled at intervals of 15 min for approximately 300 days from October 2009 to
November 2010. They observed that DOC concentrations were highest during the summer (JulyAugust) and fall (September-October) seasons. A summer peak in DOC concentrations is also
seen in many other regions. For example, Grayson and Holden (2012) saw the highest DOC
concentrations during summer months in the upland peat catchment of Cottage Hill Sike in
northern England and Jeong et al. (2012) observed large summer DOC concentrations in the
mountainous catchment of Haean Basin in northern South Korea. Both Grayson and Holden
(2012) and Jeong et al. (2012) were able to observe these seasonal trends using submersible
spectrophotometers from s::can (Messtechnik GmbH, Austria). The consistency between results
from DOC studies using in-situ optical methods (after site-specific calibration) and those from
studies using low-frequency grab samples, supports their ability to provide comparable measures
of DOC in riverine systems and confirms our previous understanding of DOC dynamics across
seasons. However, beyond just confirming these long-term seasonal patterns, they are revealing
new shorter timescale patterns and processes (Figure 20).
Given the high sampling frequency of studies that incorporated in-situ optical instruments,
these studies were also able to observe intra-seasonal and multiple individual event flow DOC
dynamics, which was previously difficult to near impossible (Figure 20). For instance, Wilson et
al. (2013) observed that the DOC response to storm events changed between seasons. DOC
concentration increased rapidly with the onset of a storm flow during the summer and fall. DOC
then quickly reached a peak concentration and remained elevated throughout the storm event,
followed by a rapid decrease in concentration, tracking the decrease in discharge as the storm

80

ceased. On the other hand, during the spring and winter seasons, DOC concentration showed
slower rates of increase at the onset of an event, and DOC also reached a lower maximum
concentration and declined more gradually after the event when compared to summer and fall
events (Wilson et al., 2013). These seasonally consistent patterns were possible because the
sensors captured all the storm events across all the seasons (Figure 20).
As another example of new insight, Saraceno et al. (2009) used in-situ optical methods
(i.e., WETStar FDOM fluorometer coupled with discrete grab sampling) to examine DOC
variations during individual storm events in the disturbed agricultural watershed of Willow
Slough in Sacramento, California, USA. The fluorometer took measurements at 1 h intervals for
4 weeks, during which there was one large precipitation induced storm flow event. Saraceno et
al. (2009) observed that in-stream DOC concentration increased with discharge, in this case to 4
times the concentration observed during baseflow (Figure 22). By coupling the in-situ FDOM
sensors with discrete sampling, Saraceno et al. (2009) was also able to measure DOC quality
parameters such as SUVA254, SR, and FI (although SUVA254 and SR could also be measured had
they used an in-situ UV-VIS sensor). These discrete samples paired with the FDOM data,
indicated that the DOC in the stream channel during the event was largely of terrestrial origin
illustrating that the storm DOC was primarily allochthonous and rapidly mobilized from
hillslopes to the stream.
Additionally, in-situ optical instruments have been used to study DOC dynamics during
snowmelt events. Pellerin et al. (2012) studied variations in DOC quantity and quality associated
with a snowmelt event in the Sleeper’s River watershed in Vermont, USA. Using a WETStar
FDOM fluorometer, measurements were taken at 30 min intervals for 2 months. Here, they also
observed increasing DOC concentrations with increasing snowmelt. It should be noted that there

81

are studies, which show clear decreasing or decoupled DOC concentration with discharge.
Grayson and Holden (2012) examined DOC processes in an upland peat catchment using an
s::can spectro::lyser, which took measurements every 15 to 20 min from July to December 2009.
They observed decreasing CDOM absorbance (i.e., CDOM concentration) along the rising limb
of numerous storm hydrographs and overall found discharge and CDOM absorbance to be poorly
correlated. Similarly, Koehler et al. (2009) used an s::can spectro::lyser to study DOC dynamics
in an Atlantic blanket bog in southwest Ireland. Sampling was conducted every 30 min over the
course of 2007. DOC concentration was shown to have a weak, negative correlation with
discharge, and instead a strong correlation with temperature, indicating that temperature
dynamics rather than flow dynamics are the primary control on DOC in these bog systems. These
studies, where discharge and DOC concentration were inversely related, hypothesized that the
decrease in DOC during the rising limb of the hydrograph is likely due to 1) dilution as the
hydrological response in peatland catchments is different from steeper hillslope catchments and
2) that DOC production in the peatlands may be limited by the time in between storms (as
microbes break down organic matter at times when the water table is low), and therefore the pool
of DOC available for export is depleted over the wet season (Koehler et al., 2009; Grayson and
Holden, 2012).
Being able to consistently monitor DOC responses across storm events and seasons will
allow for new ways to compare watersheds and their responses to flow events. This is crucial to
understand the annual DOC flux out of watersheds and delivered to downstream ecosystems and
water resources, a large portion of which occurs during these intermittent flow events (Jones et
al., 2014; Raymond et al., 2016). These in-situ optical sensors enable researchers to more safely

82

and consistently sample during both low and high magnitude storm events, the timing of which is
not always predictable (Figure 20).
At an even finer temporal resolution, in-situ optical sensors are revealing short-lived,
diurnal changes in DOC quantity and quality (e.g., Spencer et al., 2007; Worrall et al., 2015) as a
result of ecosystem processes (e.g., autochthonous DOC production). Spencer et al. (2007) used
a WETStar FDOM fluorometer along with two in-situ spectrophotometers (AC-9 photometer for
the visible range and a Satlantic UV spectrophotometer for the UV range) in order to assess
diurnal variations in DOC in the San Joaquin River in California, USA. Spencer et al. (2007)
observed that DOC does exhibit measurable diurnal variations. CDOM absorbance values
peaked in the early evening and bottomed out at dawn, while CDOM fluorescence and SR were
highest in the early morning (Figure 23). In addition to the CDOM peak in the early morning,
CDOM fluorescence exhibited a second peak at dusk. Through these unprecedented
measurements, Spencer et al. (2007) were able to conclude that the decreasing daytime CDOM
fluorescence is likely suggestive of photobleaching, while the imposed second CDOM peak on
the overall diurnal trend is hypothesized to be the result of zooplankton grazing. When the SR
trends were compared to CDOM absorbance and fluorescence trends, the decreasing SR values
throughout the daytime help support the hypothesis that DOC quantity and quality is controlled
on a diurnal scale by photochemical processing and in-situ biological production (possibly from
zooplankton grazing). However, Spencer et al. (2007) acknowledges that the diurnal variations in
DOC are complex and the controlling processes need to be further studied as these short-term
DOC variations may complicate DOC source studies. These new hypotheses along with intraseasonal, event, and diurnal DOC variations can now be more regularly evaluated across systems
where these high-frequency measurements can be made. This will allow for cross-site,

83

transferable understanding of how short-term processes may impact longer timescale
observations and ecosystem processes.

Figure 23: Diurnal trends for CDOM and absorbance. Absorbance peaks in the early
evening, while CDOM peaks in the early morning (Image from Spencer et al., 2007).

4.5.2 Revealing Sources and Pathways of DOC to the Stream Channel
In-situ optical methods also provide insight on shifts in sources and pathways of DOC
through time (e.g., during storm events, between seasons, etc.) by utilizing temporal data of DOC
quantity and molecular characteristics. As mentioned earlier, Saraceno et al. (2009) observed that
terrestrially-derived DOC increased in the stream channel during the storm event, by measuring
SUVA254, SR, and FI (again SUVA254 and SR could be measured using an in-situ UV-VIS sensor
to achieve a higher sampling frequency). However, the FI values observed (i.e., 1.5-1.7) were
84

higher than predicted for a disturbed agricultural watershed. Therefore, Saraceno et al. (2009)
concluded that the terrestrially-derived DOC was likely “fresh” or less degraded, specifically
coming from plant leachates in surface runoff over organic-rich agricultural soils. In addition, by
conducting high-frequency sampling of FDOM using an in-situ optical sensor, Saraceno et al.
(2009) observed a time lag between peak FDOM, discharge, and turbidity, where the FDOM
lagged ~11 to 15 h behind peak discharge and turbidity. This suggests that in their disturbed
agricultural watershed, shallow soil flowpaths remain important for several days after the
precipitation event and that the suspended sediment (i.e., turbidity) and FDOM may have
different watershed sources, because they peak at different times. Without the continuous high
sampling frequency of in-situ optical sensors this lag in variables would likely have gone
unnoticed.
A time lag between peak discharge and peak FDOM was also observed in other opticallybased studies. Wilson et al. (2013) saw a similar length time lag between the FDOM and
discharge during storm events in their forested headwater catchment as Saraceno et al. (2009),
while Pellerin et al. (2012) observed a much shorter (~60 min) time lag during snowmelt events
in Vermont, USA. The tight coupling of peak discharge and peak FDOM seen by Pellerin et al.
(2012), along with their observed increasing SUVA254 and decreasing FI values, indicated that
surficial flowpaths along organic-rich soils were dominant during the snowmelt events (similar
to Saraceno et al., 2009). Other studies also found that increasing DOC concentrations during
storm events are likely derived from surface and shallow subsurface flowpaths. For example,
Waterloo et al. (2006) used an s::can spectro::lyser to examine DOC variations in blackwater
streams of the Amazon rainforest. Measurements were made at 30-min intervals from February
to September 2002. By additionally estimating shallow and deep groundwater flow using

85

Darcy’s Equation, Waterloo et al. (2006) was able to use the spectro::lyser data to determine that
solute paths shifted from deeper groundwater regions to shallow subsurface layers in the valley
soils that are rich in organic material during the many observed storm events.
Finally, high-frequency data can also be used to observe and characterize detailed
hysteresis relationships between discharge and FDOM concentration to infer DOC sources and
processing. Pellerin et al. (2012) observed a counter-clockwise hysteresis relationship (Figure
24), which is hypothesized to be the result of a delayed contribution from surface and shallow
subsurface flowpaths and suggests that DOC export in the Sleeper’s River watershed is transport
limited rather than source limited. While many studies have also seen counter-clockwise
hysteresis relationships for discharge and FDOM concentration (e.g., Strohmeier et al. 2013),
some have observed clockwise hysteresis relationships. Jeong et al. (2012) saw a clockwise
hysteresis relationship when studying DOC dynamics in the mountainous watershed in South
Korea suggesting that DOC export in this watershed is source limited (i.e., DOC rich soil pools
are depleted along the falling limb of the hydrograph). The discrepancy between hysteresis
relationships for discharge and FDOM concentration is an important area of future research and
may be dependent on watershed characteristics (e.g., slope, land cover, antecedent conditions,
etc.). Having the ability to observe many hysteretic responses (via in-situ optical sensors) across
a gradient of storm types in one watershed or across many watersheds will provide greater
insight to what spatial and temporal watershed conditions are the primary controls on DOC
concentrations and molecular characteristics. Overall, the ability to generate more transferable
ideas and theories about DOC dynamics in watersheds and riverine systems is growing due to the
ability of optical sensors to more consistently document DOC storm event dynamics across a
huge range of watersheds (Figure 20).

86

Figure 24: Hysteresis trends. Counter-clockwise hysteresis observed during both snowmelt and
precipitation events for FDOM (Image from Pellerin et al., 2012).

4.5.3 Annual DOC Flux
Previous work on sampling frequency, especially across dynamic stream flow conditions,
found that infrequent sampling of DOC results in large underestimations of DOC fluxes (i.e., -10
to -88%), especially during moderate to low flow periods (Saraceno et al., 2009; Jollymore et al.,
2012; Pellerin et al., 2012; Wilson et al., 2013; Jones et al., 2014; Grayson and Holden, 2016).
As a result high-frequency, in-situ sampling of FDOM concentration should be used to improved
DOC flux quantification. Thus far, studies that have incorporated in-situ optical sensor methods
were able to better calculate short-term and annual DOC fluxes from watersheds (e.g., Koehler et
al., 2009; Jones et al., 2014).
Most studies using in-situ optical sensors found that a significant amount of the annual
DOC flux is exported during isolated brief storm events or, if in a more Mediterranean climate,
during the wet season. Wilson et al. (2013) found that 63% of the annual DOC flux was exported
during these brief, reoccurring hydrologic storm events, while Koehler et al. (2009) and Jeong et
87

al. (2012) found that 45% and 48% respectively, was exported during storm events. Additionally,
Waterloo et al. (2006) observed that 69 to 72% of annual export occurred during the wet season,
all of which corroborates the need to sample on intra-seasonal and diurnal time scales. Waterloo
et al. (2006) also quantified a relationship between DOC flux and stream discharge. The largest
fluxes of DOC were typically seen in the top 10% of high flows (Koehler et al., 2009), thus as
stream discharge increases, DOC flux will typically increase and in a predictable manner. For
example, given the high temporal characterization of DOC across a large range of flow
conditions (Figure 25), Waterloo et al. (2006) was able to generate two empirical models for
DOC flux as there appears to be a threshold behavior in the specific DOC flux response to a
given discharge (Figure 25). Specifically Waterloo et al. (2006) showed that DOC flux, FDOC,
when stream flow (Qstream) is less than 2.8 mm d-1 (Equation 4.6; r2 = 0.79) is,
FDOC = 0.0075 x Qstream1.5630

(4.6)

and when Qstream is greater than 2.8 mm d-1 (Equation 4.7; r2 = 0.99).
FDOC = 0.0278 x QStream − 0.0432

(4.7)

Overall, more accurate estimates of flux dynamics were possible using the high resolution in-situ
optical sensors in both cases of flux budgets and flux response to changes in flow.
Understanding these consistently observed “pulses” of DOC in response to variable flow
conditions is important to future research of ecosystem processes at the watershed scale.
Specifically, C gains and losses are important for understanding and accurately quantifying net
ecosystem productivity and net ecosystem C balance (Wilson et al., 2013). Thus, missing or
misappropriating the key portion of C in the form of DOC will result in erroneous net ecosystem
C balance models. In addition, understanding changes in sources and pathways of DOC to the

88

stream channel is key to determining DOC quality moving through riverine environments,
because DOC quality is largely dependent on C source material (Cory et al., 2011). The timing
of both DOC quantity and molecular characteristics impacts DOC bioavailability and, therefore,
timing of DOC flux can have a large effect on organisms, downstream ecosystems, and water
quality. These in-situ sensors seem to be capable of greatly improving our ability to “take the
pulse” of key ecosystem parameters, specifically, DOC moving through watersheds and
ecosystems (Roley et al., 2014). In taking the pulse of DOC in streams, we are also able to use
that information to guide the study design and monitoring of other watershed attributes that
control ecosystems and water quality, such as limiting nutrients (Blaen et al., 2016).

Figure 25: DOC flux versus discharge. DOC flux increases linearly with discharge, although
the relationship shifts slightly from low to high discharge (Image from Waterloo et al., 2006).

89

4.6 Conclusions & Future Considerations
High-frequency, in-situ optical monitoring allows for much higher resolution temporal
observations of DOC processes in riverine environments than ever previously thought possible
(i.e., dynamics operating on temporal scales of seconds to minutes, Figure 20 and 4.5). Thus,
new processes and understanding of DOC in watershed and riverine systems are being revealed.
In particular, these sensors are facilitating novel observations, conceptual understanding, and
models that address fundamental questions about watershed DOC quantity and quality through
time, the sources and pathways of DOC to the river network, and DOC mass flux. Future
research on these important DOC dynamics will benefit from the use of these technologies.
These early adopters of these in-situ optical sensor technologies have already made great
advances as discussed above.
The in-situ optical methods are quick, precise, and relatively inexpensive when compared
to collecting discrete water samples in the field and later completing chemical C analyses.
However, they still have limitations that must be addressed by users. In particular, in-situ FDOM
fluorometers and spectrophotometers should still be coupled with some discrete sampling to
include the full range of DOC optical properties as well as cross check DOC concentrations
against CDOM and FDOM proxies. In addition, temperature and turbidity correction equations
should be applied (if not already built-in within the sensor system), especially in watersheds with
large temperature fluctuations and/or high suspended sediment loads. There are emerging
methods to make these temperature and turbidity corrections as this is an active area of research.
It is worth noting again, as discussed in Section 4.2.1 and 4.2.2, FDOM and CDOM sensors only
measure a fraction of the total DOM pool so all interpretation of bulk DOM processes and
dynamics must be done in light of this limitation. Nonetheless, these in-situ optical methods can

90

provide valuable information on DOC processes in streams and watersheds that is otherwise
logistically prohibitive and therefore unavailable. As more users adopt this technology, usersupport will also increase, eliminating numerous potential concerns (e.g., interfacing with power
supplies and data loggers). In addition, these sensors could be used to standardize protocol
between research groups and environmental regulators, creating opportunity for better
networking and data sharing plans. As the climate and anthropogenic uses of land and water
continues to change, in-situ optical instruments are likely to become crucial to measuring the
pulse of riverine and watershed responses to these changes. Having this information will help to
diagnose and understand both associated short- and long-term changes in DOC quantity and
quality, C cycling in watersheds, as well as help guide the development of future watershed-scale
management strategies.
4.7 Acknowledgements
The authors would like to thank the editor and three anonymous reviewers for their
helpful suggestions. Support for this research was also provided by the NSF Long-term
Ecological Research Program (DEB 1027253) at the Kellogg Biological Station and by Michigan
State University AgBioResearch.

91

CHAPTER 5: SYNTHESIS & IMPLICATIONS
This thesis focused on understanding the role of the stream water-groundwater interface
in dissolved organic carbon (DOC) cycling in river networks. The central research questions
addressed in Chapters 2 and 3 were 1) is the hyporheic zone (HZ) a significant processor of
stream DOC at the watershed scale in a lowland river, and if so, 2) how is DOC being processed?
To assess these questions, two synoptic HZ sampling campaigns were conducted in August 2015
and 2016 across the Augusta Creek watershed in southwestern, Michigan. The results presented
in Chapter 2 indicate that the HZ is a (potentially important) location for DOC processing in
stream networks and that the processing signal can be observed across stream orders throughout
the watershed. However, it was concluded that even at the watershed-scale, analyses needed to
be conducted on a site-by-site basis, and that improved chemical and hydrological data (e.g.,
dissolved oxygen, vertical head gradient) were necessary to more accurately interpret the trends
in hyporheic DOC quantity and quality. With the improved chemical and hydrological data in
2016, we were able to assess how DOC is processed in the HZ (i.e., the dominant transformation
mechanisms). The results presented in Chapter 3 indicate that the HZ acts as a sink for DOC and
that aerobic, microbial respiration may be the dominant processing mechanism removing DOC
in the HZ.
By studying the HZ at the watershed scale, we can ask questions about the larger
ecosystem function (i.e., sink) of the HZ. Specifically, does the role of the HZ as a processor of
stream DOC change based on location in the watershed (e.g., across stream orders). The results
presented in Chapter 3 suggest that, the HZ may function as a larger sink for DOC in headwater
streams than in higher stream orders across the watershed, despite the observed, lower
processing potentials. However, the results were inconclusive in terms of the fate of DOC

92

quality, which appeared to be influenced by changes in local VHG. Thus, further studies focused
on DOC quality need to be conducted since changes in DOC quality due to hyporheic processes
may alter the fate of DOC transported downstream.
Chapter 4 provided a review of in-situ, optical instruments that can be used in stream
networks to better understand the short-term dynamics of DOC quality in a cost-effective way.
These new technologies offer the potential to dramatically improve our understanding of stream
DOC fluxes, including the variation in sources and transformations in space and time.
In conclusion, previous HZ studies of DOC processing have predominantly been smallscale field and mesocosm studies (e.g., Findlay et al., 1993; Findlay and Sobczak, 1996;
Schindler and Krabbenhoft, 1998; Sobczak and Findlay, 2002; Zarnetske et al., 2011), yet
hyporheic processes are also expected to impact stream DOC cycling on larger scales (e.g.,
watershed scale; Boano et al., 2014). To the best of our knowledge, the present work represents
one of the first watershed-scale HZ sampling efforts and in doing so, we were able to more
accurately assess the role of the HZ on DOC cycling throughout the stream network, concluding
that the HZ acts as a sink for DOC within the larger carbon (C) cycle of Augusta Creek,
Michigan, USA.
If these results are transferrable to other watersheds, hyporheic processing of stream
DOC may be a potential mechanistic explanation for the recently observed, large carbon dioxide
(CO2) efflux from streams (Butman and Raymond 2011; Raymond et al., 2013). Nonetheless,
this research indicates the need to include streams and the associated HZ processes in global C
budgets, since accurately modeling the global carbon (C) cycle is necessary for mitigating future
climate change. In addition, this thesis highlights several gaps in our understanding of how to
study HZ processes at the watershed scale that need to be addressed going forward:

93

1) When sampling at the watershed-scale, how important is local variation in hyporheic
DOC conditions (i.e., within a single site), versus the variation observed across a
stream reach or an entire stream order? This will control how we sample the HZ over
an entire watershed (e.g., choosing between triplicate, duplicate, or single porewater
sampling schemes).
2) The use of distributed, in-situ sensor technologies across watersheds is improving our
understanding of DOC quality dynamics in the open stream channel. However,
advancements in using similar distributed, in-situ sensor networks to study the
coupling of the stream and HZ are needed.
3) Lastly, to better identify the dominant transformation mechanisms of DOC in the HZ,
studies are needed that incorporate dissolved inorganic carbon (DIC) measurements
such as CO2 since it is often a product of biotic processes which remove DOC as well
as direct measurements of microbial respiration rates across entire stream networks.

94

APPENDICES

95

APPENDIX A

2015 Synoptic Hyporheic Zone Sampling Data

96

Table A1: Supplemental information for the 2015 sampling sites.
Sampling
Site
A
B
C
D
F
G
H
I
J
K
L
M
O
P
Q
R

Longitude

Latitude

-85.332
-85.3547
-85.3534
-85.3517
-85.3449
-85.319
-85.3233
-85.3539
-85.3366
-85.3525
-85.3425
-85.3107
-85.343
-85.3441
-85.3516
-85.3599

42.4197
42.4037
42.427
42.4176
42.407
42.4479
42.4468
42.3534
42.4573
42.4688
42.4628
42.4089
42.4213
42.4113
42.3906
42.3688

Stream
Order
1
3
2
2
3
1
2
3
2
1
1
1
1
2
3
3

Drainage
Area (km2)
1.92
69.69
41.66
47.69
68.17
1.77
32.47
95.39
26.30
7.78
9.56
7.38
0.80
50.02
73.18
86.15

Depths with
Mixing
Zone Signal
NA
2.5-15 cm
2.5-10 cm
2.5-7.5 cm
NA
2.5-20cm
2.5-7.5 cm
2.5-5.0 cm
2.5-5.0 cm
NA
NA
2.5-5.0 cm
2.5-5.0 cm
2.5-10 cm
2.5-5.0 cm
NA

97

D50
(mm)

Land Use
% Agriculture

0.41
0.81
0.69
0.47
0.29
0.46
0.47
0.43
0.53
2.71
0.38
0.71
0.52
2.56
1.93
1.01

41
48
48
53
48
60
44
43
41
30
35
39
77
53
47
46

Land Use
% Forest

Land Use
% Wetland
29
21
21
19
21
13
24
28
22
23
25
27
5
19
22
25

18
22
21
19
22
13
22
19
23
28
18
19
9
19
21
19

Figure A1: Map of Augusta Creek watershed with land use. Map of the Augusta Creek
watershed in Barry and Kalamazoo counties, MI, USA. The map shows the dominant land use
types across the watershed including upland forest, wetland, and agriculture.

98

Figure A2: Site A Vertical Porewater Profiles

99

Figure A3: Site B Vertical Porewater Profiles

100

Figure A4: Site C Vertical Porewater Profiles

101

Figure A5: Site D Vertical Porewater Profiles

102

Figure A6: Site F Vertical Porewater Profiles

103

Figure A7: Site G Vertical Porewater Profiles

104

Figure A8: Site H Vertical Porewater Profiles

105

Figure A9: Site I Vertical Porewater Profiles

106

Figure A10: Site J Vertical Porewater Profiles

107

Figure A11: Site K Vertical Porewater Profiles

108

Figure A12: Site L Vertical Porewater Profiles

109

Figure A13: Site M Vertical Porewater Profiles

110

Figure A14: Site O Vertical Porewater Profiles

111

Figure A15: Site P Vertical Porewater Profiles

112

Figure A16: Site Q Vertical Porewater Profiles

113

Figure A17: Site R Vertical Porewater Profiles

114

Table A2: Synoptic sampling data for 2015
Site
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
4.786
2.801
2.178
3.442
3.054
1.971
2.511
2.616
3.276
2.506
2.635
2.115
7.175
5.0791
5.2787
4.9858
8.784
3.744
2.9913
3.2275
7.4165
4.3689
4.4027
4.4532
7.1647
5.5331
5.2764

DOC (mg/L)
11.53
8.78
8.11
5.52
4.21
5.43
11.84
9.27
8.53
8.27
5.70
7.18
10.13
8.10
7.01
7.64
5.03
4.68
10.57
5.91
4.92
5.15
5.25
3.45
3.49
7.83
4.79
4.87
3.46
3.02
2.82
3.98
5.27
5.13
4.50
4.01
3.16
9.61
5.21

SUVA254 (L/mg C/m)
1.87
0.90
1.13
1.30
0.83
2.54
1.71
2.15
3.05
2.42
1.53
1.25
1.82
1.32
0.77
0.73
0.80
0.88
2.37
1.49
0.37
0.14
0.00
0.14
0.00
1.66
0.98
0.18
1.18
0.40
0.00
0.70
0.46
0.06
0.20
0.25
0.32
1.96
0.46

115

SR
0.72
1.20
0.99
1.46
2.25
0.85
2.22
1.85
2.26
2.46
2.08
0.91
0.48
0.61

1.93
0.95
1.43

0.68
0.53
0.67

-0.15
0.69
0.44

Cl- (mg/L)
5.98
10.96
8.48
9.71
12.03
11.61
5.74
7.48
8.85
8.91
11.35
9.33
6.51
8.58
11.56
12.43
11.91
12.25
5.72
7.85
6.75
6.70
6.47
6.55
6.37
9.37
7.13
8.82
7.31
6.62
6.97
7.46
7.80
6.72
6.29
6.64
7.14
10.12
7.32

NO3-N (mg/L)
2.44
3.30
2.62
3.40
2.65
2.66
2.34
2.57
2.62
3.24
2.65
2.61
2.32
2.62
2.64
2.72
2.67
2.56
2.36
3.64
2.42
4.46
5.15
4.82
6.30
2.21
2.46
2.58
0.02
5.87
6.39
3.72
4.12
4.55
4.78
6.10
6.46
2.03
6.34

SO4 (mg/L)
4.58
0.07
3.43
0.03
2.59
2.68
4.69
3.17
2.46
0.01
3.17
2.70
4.26
2.65
2.74
2.91
2.64
9.07
4.90
17.23
21.84
19.49
19.60
19.21
18.96
12.48
17.19
21.63
23.58
19.24
19.15
19.36
20.10
20.97
19.01
19.34
19.32
10.56
19.22

Table A2 (cont’d)
Site
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW

DO (mg/L)
4.6084
4.1218
4.3592
3.6418
6.838
5.9089
7.8844
5.3488
5.5127
7.1918
7.9835
6.822
10.9533
0.114
4.197
4.602
4.543
3.64
4.51
5.961
4.613
3.944
5.338
6.091
4.23
7.622
9.171

DOC (mg/L)
7.52
6.39
7.41
6.37
6.43
5.25
5.17
7.62
6.45
8.53
6.55
8.03
6.53
6.11
7.65
6.57
5.84
6.04
10.26
5.76
7.59
8.03
7.66
5.90
4.75
4.60
9.69
10.76
12.44
5.80
7.44
6.01
8.09
7.87
7.13
5.45
5.91
4.53
8.96

SUVA254 (L/mg C/m)
1.86
2.19
1.97
2.51
2.71
1.07
2.67
1.79
2.50
1.66
2.60
1.89
2.44
2.50
1.91
2.25
2.50
2.68
1.62
0.00
1.79
1.93
1.36
1.31
2.00
2.15
2.82
3.12
1.44
1.02
3.08
1.38
1.99
2.01
1.54
0.94
0.39
0.35
1.93

116

SR
0.63
0.60
0.69
0.68
0.65
0.41
0.56
0.65
0.78
0.56
0.65
0.59
1.00
0.84
0.67
0.67
0.55
0.67
0.73
0.68
0.84
0.81
1.21
0.54
1.65
0.67
0.73
0.69
0.72
1.05
-6.72
0.50
0.54
0.64

0.54

Cl- (mg/L)
9.90
10.34
12.34
9.96
10.56
9.70
10.39
11.63
10.28
10.32
10.40
9.24
10.19
9.96
10.16
11.89
10.57
9.51
10.23
10.14
19.89
30.92
84.72
137.45
161.76
167.58
16.04
20.08
52.29
131.16
88.26
127.94
29.82
13.70
54.54
110.13
160.87
157.84
10.69

NO3-N (mg/L)
2.09
2.09
1.15
2.14
2.30
1.69
1.96
1.93
1.92
1.97
2.18
2.47
2.00
1.99
1.96
1.96
2.07
2.47
2.01
2.30
2.05
2.26
2.61
2.95
2.82
2.76
2.05
2.63
2.89
3.08
3.00
2.66
2.20
1.95
0.71
2.74
1.78
1.60
1.88

SO4 (mg/L)
11.99
12.43
13.81
11.09
9.50
4.49
12.33
12.20
12.31
12.23
9.35
11.97
11.90
11.76
11.99
11.93
12.10
12.39
11.73
32.56
11.94
9.36
6.45
1.81
11.77
19.61
10.94
4.08
3.22
0.95
2.89
20.04
9.66
11.52
9.27
6.19
3.73
22.15
12.15

Table A2 (cont’d)
Site
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
4.036
5.0278
2.548
1.3197
4.5746
3.9368
1.5163
1.5378
4.7068
5.8287
2.9567
3.8107
9.3037
-0.0085
3.6075
2.149
3.1354
3.2389
4.677
4.342
3.5645
4.2422
4.5921
3.7545
4.2596
5.5695
8.7079

DOC (mg/L)
9.94
10.66
11.13
15.25
12.74
12.50
7.93
9.67
11.07
12.81
13.05
12.24
15.69
13.30
12.74
12.88
10.72
10.45
7.77
8.94
6.58
6.21
6.18
4.75
5.41
5.13
7.46
9.71
5.01
4.56
4.19
5.33
6.74
7.42
7.26
5.19
6.30
5.50
16.75
10.08

SUVA254 (L/mg C/m)
1.93
2.21
2.63
3.63
4.47
5.35
2.84
2.92
3.32
2.97
3.79
4.27
2.24
2.45
3.09
4.21
4.70
8.89
2.72
2.84
1.82
2.51
1.44
0.29
0.07
0.21
2.47
2.02
1.32
1.29
0.10
0.23
2.24
1.48
0.90
0.08
0.00
0.00
0.84
0.51

117

SR
0.65
0.80
0.88
1.15
2.09
3.02
0.76
0.82
0.99
0.98
1.57
1.73
0.75
0.83
1.26
1.33
2.40
1.05
0.80
0.99
0.63
0.78
0.66

0.91
0.83
0.42

0.59
0.60
0.39

0.50
1.18

Cl- (mg/L)
10.66
10.95
10.55
9.78
9.80
15.26
10.74
10.60
11.51
10.16
9.58
8.33
13.08
13.11
12.12
9.94
9.78
9.05
11.90
10.67
10.72
10.77
11.92
12.45
14.28
12.83
10.53
10.50
12.84
12.30
13.95
13.91
10.47
11.52
12.93
13.91
12.98
12.98
10.21
11.51

NO3-N (mg/L)
2.18
2.34
2.67
3.40
4.30
3.45
2.26
2.53
2.83
4.07
1.87
4.14
2.83
3.10
4.33
3.34
1.37
3.28
1.78
1.43
0.97
2.10
2.37
2.49
2.47
2.46
1.99
2.09
2.72
2.75
2.56
2.43
1.92
2.26
1.67
2.50
2.47
2.49
1.94
2.42

SO4 (mg/L)
9.49
6.11
3.34
0.07
0.01
0.02
8.07
3.92
2.88
0.01
0.07
0.02
4.73
1.38
0.06
3.11
0.64
12.97
1.67
19.72
12.73
14.13
36.93
37.98
39.17
9.22
9.97
8.25
8.57
41.36
50.19
12.00
11.63
29.29
45.35
28.51
55.45
12.36
45.36

Table A2 (cont’d)
Site
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
4.368
5.799
4.911
4.651
4.364
5.777
4.69
7.052
6.089
6.098
6.605
6.384
5.847
5.7679
6.3187
7.8805
9.0974
5.8565
4.9757
7.7452
8.9049
6.3363
6.1942
6.9396
8.924
7.2594
4.1425

DOC (mg/L)
9.37
5.79
3.38
4.45
3.42
3.46
7.52
8.11
4.92
3.55
2.99
5.26
8.41
6.83
4.41
3.73
3.46
3.31
9.53
3.86
4.38
3.65
2.78
4.47
4.88
4.35
5.42
2.94
2.81
2.37
3.26
3.09
3.54
2.95
3.48
1.50
3.08
3.83
8.75
3.06

SUVA254 (L/mg C/m)
2.02
2.21
4.05
0.13
0.00
0.00
1.70
1.89
0.51
1.58
0.20
0.00
2.01
1.11
0.43
0.24
0.00
0.00
2.24
0.16
1.48
1.18
0.07
0.00
0.00
0.16
1.77
0.00
0.00
0.04
0.00
0.00
0.06
0.27
0.00
0.00
0.00
0.00
1.70
0.00

118

SR
1.48
1.62
3.22
0.00
0.00
0.00
1.29
1.41
0.32
1.13
0.03
0.00
1.45
0.81
0.27
0.13
0.00
0.00
1.63

0.86

0.59

0.59

Cl- (mg/L)
8.97
6.49
4.42
3.45
3.60
3.13
6.82
6.94
4.82
4.12
3.24
3.42
8.51
5.88
4.12
3.47
3.41
3.24
10.36
4.06
10.57
9.13
7.61
10.11
7.76
9.08
11.56
7.47
7.61
7.53
7.93
7.54
9.18
7.40
7.59
7.67
8.98
9.17
12.53
8.86

NO3-N (mg/L)
2.15
2.32
1.64
1.57
2.27
2.18
2.26
2.17
1.58
1.61
2.37
2.30
2.17
1.53
2.38
2.33
2.27
1.61
1.94
2.34
3.59
4.86
2.86
8.79
1.04
1.04
0.02
1.10
1.06
1.10
1.08
1.06
1.08
1.56
1.05
1.09
1.05
1.51
1.97
2.67

SO4 (mg/L)
6.69
8.00
23.46
35.55
45.27
44.55
10.41
9.12
26.97
23.02
34.59
44.01
10.05
22.82
34.77
41.95
44.66
42.05
8.19
48.38
13.24
14.45
14.51
13.88
14.43
13.82
12.58
13.58
13.53
13.66
13.28
12.86
15.54
12.55
12.59
13.00
13.04
13.81
13.85
14.70

Table A2 (cont’d)
Site
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
J
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW

DO (mg/L)
4.5898
5.0039
5.4169
6.167
3.735
4.054
4.898
5.776
5.0065
5.9112
5.3322
5.4617
2.9024
0.0708
4.89
4.376
4.016
4.043
5.796
3.677
3.53
3.709
4.763
3.841
3.054
3.224
8.177

DOC (mg/L)
5.69
4.09
3.99
2.86
3.08
5.21
8.28
7.35
2.84
3.96
3.50
4.05
4.52
4.13
3.08
6.00
6.12
6.89
11.03
6.93
13.29
10.97
10.12
8.80
8.88
9.65
9.70
9.46
10.05
8.59
8.88
9.68
8.58
9.62
8.92
10.66
10.57
8.92
10.91

SUVA254 (L/mg C/m)
2.49
1.32
0.18
0.11
0.03
0.00
1.42
0.18
0.21
0.00
0.37
0.00
2.52
1.33
0.81
0.52
0.15
0.09
2.18
0.16
1.51
1.67
1.73
2.00
2.05
2.20
2.09
2.07
2.07
2.32
2.04
2.18
2.22
2.01
2.24
2.05
2.06
1.75
1.81

119

SR
0.69

0.45

0.65

1.36

0.79
1.12
1.00
1.03
1.02
0.99
0.89
1.04
1.10
1.20
0.95
1.04
0.92
1.11
0.99
1.00
1.01
0.96
0.95
0.99

Cl- (mg/L)
6.40
5.24
4.78
4.82
5.37
5.27
7.16
5.24
5.02
4.80
4.78
5.53
6.78
5.47
4.82
5.62
4.50
4.34
13.30
5.16
9.01
9.41
12.42
9.61
9.47
9.86
9.38
9.70
9.57
10.27
10.09
10.61
9.53
9.40
10.08
10.51
10.06
11.54
9.41

NO3-N (mg/L)
2.43
2.45
2.32
1.51
1.50
1.49
2.22
2.28
2.32
1.50
1.49
1.50
2.26
2.41
2.39
2.43
1.51
1.46
2.10
1.52
1.83
1.88
1.87
1.87
1.86
1.93
1.84
1.87
1.87
1.90
1.99
1.93
1.94
1.85
1.98
1.98
1.99
2.69
1.84

SO4 (mg/L)
5.39
6.15
13.66
21.44
21.15
23.13
3.96
15.09
13.45
21.24
22.56
23.15
4.86
4.16
6.09
6.71
21.70
21.08
9.45
22.91
4.36
4.21
4.29
4.21
4.16
3.27
4.26
4.33
4.14
4.35
1.34
4.06
4.69
4.81
2.47
2.03
2.65
0.03
4.24

Table A2 (cont’d)
Site
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
6.7621
5.7043
4.3215
5.2336
6.9817
3.877
4.213
5.8218
7.4587
6.0244
4.5548
6.9916
8.0165
7.796
5.2039
3.917
3.97
7.2669
6.6305
4.9617
6.3889
6.1701
6.812
5.57
5.5045
7.6948

DOC (mg/L)
8.59
12.57
13.67
9.55
8.98
7.20
9.79
10.97
9.93
7.07
6.18
6.22
9.26
11.35
8.44
8.95
6.43
5.57
12.52
16.96
8.58
8.39
6.96
3.75
5.01
5.86
10.00
3.62
6.31
4.26
5.91
3.26
12.64
8.94
7.48
5.58
3.03
3.69
10.43
3.75

SUVA254 (L/mg C/m)
3.63
4.79
7.70
10.82
5.88
2.04
3.98
5.20
6.55
13.60
5.88
2.01
2.92
3.81
2.37
4.77
1.41
3.40
2.18
0.87
2.60
2.04
1.34
1.49
1.78
1.33
2.14
2.27
1.36
2.51
0.69
2.58
2.53
2.10
2.05
1.09
2.71
2.28
2.20
6.16

120

SR
0.76
0.64
0.51
0.53
0.66
1.33
0.70
0.57
0.60
0.47
0.66
1.58
0.78
0.67
0.81
0.66
1.04
0.90
0.82
1.46
0.85
1.15
0.78
1.01
0.90
1.60
0.78
0.85
0.72
0.96
1.58
1.59
0.87
0.81
0.80
0.83
0.90
1.28
0.79
0.82

Cl- (mg/L)
8.97
9.01
8.81
5.45
5.94
4.66
9.71
9.31
6.38
6.53
4.44
4.30
9.31
9.69
8.55
16.80
5.64
4.59
9.12
6.77
8.83
10.48
12.24
11.30
11.22
11.02
8.61
10.49
9.89
10.97
11.32
10.95
9.80
9.70
10.49
10.99
10.70
10.36
9.17
12.36

NO3-N (mg/L)
1.66
1.67
1.75
2.51
3.23
2.41
1.69
1.64
3.13
2.41
3.26
3.12
1.72
1.62
1.91
2.14
2.50
3.23
1.69
2.48
2.20
2.52
2.92
3.04
2.90
2.88
2.23
1.04
2.76
2.93
2.83
2.79
2.60
2.68
2.90
3.03
2.97
2.81
2.17
2.77

SO4 (mg/L)
4.18
4.30
3.13
2.62
0.07
2.00
4.50
3.95
0.04
2.50
0.03
0.01
4.74
3.76
3.67
2.54
2.72
0.02
4.20
2.10
7.04
4.64
3.88
2.89
4.46
5.19
7.16
4.32
4.01
4.36
4.67
10.42
6.53
5.76
6.05
4.56
4.49
6.31
6.61
5.91

Table A2 (cont’d)
Site
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
7.2353
7.307
6.3216
6.087
6.0437
6.4516
5.5656
5.6459
6.8293
5.465
4.3251
7.3476
7.7004
4.7057
8.3209
7.159
5.3128
6.6728
7.5781
5.1153
7.4234
7.0445
7.409
5.1853
5.5822
8.1596

DOC (mg/L)
4.37
4.70
4.14
5.56
4.39
3.85
6.23
3.72
4.70
5.21
4.39
6.46
5.34
4.43
5.01
3.76
4.35
4.25
6.17
4.24
6.06
2.08
3.36
4.47
5.05
5.43
7.16
5.94
5.70
3.47
4.95
6.12
8.31
4.86
5.02
3.51
4.33
5.53
8.66
6.15

SUVA254 (L/mg C/m)
1.79
0.64
0.92
1.06
0.77
0.60
1.00
1.80
0.83
0.75
0.59
0.39
1.25
0.95
1.00
0.66
0.94
0.56
1.43
0.14
0.56
1.34
1.01
0.67
0.46
0.92
1.73
1.89
0.70
1.07
0.46
0.31
2.08
3.25
1.67
2.22
0.62
0.52
2.23
1.15

121

SR
1.26
0.48
0.70
0.80
0.86
0.43
0.85
0.93
0.92
0.84
0.69
0.38
0.88
0.74
1.02
0.74
1.20
0.48
1.24
1.33
0.83
1.61
1.36
1.35
1.73
1.12
1.11
0.95
1.04
1.00
0.76
1.02
1.05
1.16
1.22
1.23
1.71
0.98
1.39

Cl- (mg/L)
5.63
5.47
4.76
5.02
4.95
4.80
7.55
7.98
5.26
4.78
4.60
5.38
8.97
6.90
10.50
4.56
6.05
4.69
10.06
5.77
52.33
58.44
60.16
52.74
61.69
55.35
29.35
34.64
44.92
46.80
51.00
47.44
12.00
15.05
32.41
42.21
59.71
57.27
10.75
43.65

NO3-N (mg/L)
1.04
0.95
0.99
0.94
0.94
0.96
1.13
0.64
1.00
0.99
1.00
0.99
0.01
1.07
1.09
0.98
1.05
0.93
2.96
1.02
4.09
0.01
4.61
3.96
0.95
0.01
0.01
2.73
0.01
1.05
3.41
3.82
1.67
0.02
0.01
0.01
0.86
1.09
1.47
1.03

SO4 (mg/L)
25.09
28.39
29.15
28.44
29.03
26.45
19.69
18.10
27.21
27.98
26.34
27.06
15.64
19.83
15.44
25.93
25.02
25.38
12.44
49.80
24.15
23.63
24.24
24.75
23.92
21.83
18.68
22.71
24.43
27.86
30.98
29.50
12.42
13.98
22.34
23.54
32.51
33.18
11.76
28.98

Table A2 (cont’d)
Site
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
Q
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R

Minipoint
1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

1
1
1
1
1
1
2
2
2
2
2
2
3
3
3
3
3
3

Depth (cm)
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
-2.5
-5
-7.5
-10
-15
-20
SW
-60

DO (mg/L)
7.4736
5.704
3.9802
5.338
5.9583
5.7141
4.8569
5.1982
6.4908
5.7649
5.5324
5.9948
8.4709
0.0924
5.0548
5.0424
5.1888
4.7927
6.0829
5.38
5.501
6.145
7.2601
6.4179
5.2351
5.475
8.4221
0.9091

DOC (mg/L)
7.30
5.83
4.45
3.39
3.54
4.43
6.64
6.56
4.37
5.71
5.27
5.61
5.55
4.36
3.78
3.44
5.17
4.84
7.83
4.09
6.11
4.85
3.18
5.08
4.03
3.26
6.35
6.76
4.40
5.03
4.88
2.93
8.21
7.72
5.11
4.28
5.07
3.09
9.85
7.07

SUVA254 (L/mg C/m)
2.55
1.89
1.17
2.89
1.16
0.61
2.41
2.10
1.35
0.98
1.42
0.94
2.11
1.97
2.67
1.02
1.18
0.56
2.66
0.83
1.34
1.52
2.36
0.41
0.47
0.64
1.99
0.89
0.86
0.50
0.27
0.55
1.73
1.41
1.17
0.65
0.30
0.55
1.80
0.30

122

SR
0.92
0.90
1.04
0.79
0.91
1.49
1.06
1.22
1.37
1.24
1.33
1.47
0.95
0.95
1.85
1.57
1.15
1.37
0.95
1.84
0.89
1.19
1.23
1.50
1.82
1.61
0.98
1.06
1.73
1.59
1.82
1.07
1.07
0.98
1.14
1.55
1.86
1.53
0.93
2.96

Cl- (mg/L)
9.83
8.22
6.05
6.32
5.86
6.60
9.65
7.33
6.34
5.95
7.19
5.88
8.02
6.58
6.48
5.79
6.24
5.69
10.82
7.07
17.26
20.06
17.20
23.66
27.21
23.65
13.07
18.91
22.38
23.49
23.23
23.33
13.08
14.79
17.87
22.02
22.02
21.67
20.22
21.93

NO3-N (mg/L)
1.64
2.40
2.56
1.95
2.58
2.47
0.52
2.53
2.59
2.69
2.66
2.47
2.06
2.21
2.28
2.31
2.31
1.50
0.01
1.47
2.06
2.13
2.06
2.26
2.22
1.39
1.97
2.18
2.23
2.25
1.46
1.40
1.90
2.01
2.09
1.37
2.14
1.38
1.34
1.33

SO4 (mg/L)
12.05
6.68
3.49
1.88
4.92
7.82
9.75
4.35
4.07
2.92
3.09
3.53
12.38
13.31
14.36
17.16
16.72
4.97
12.82
27.73
17.27
16.21
17.26
18.89
20.39
23.05
15.46
15.52
18.54
17.74
22.49
24.51
15.13
15.60
13.25
22.08
21.04
21.60
13.24
27.52

APPENDIX B

2016 Synoptic Hyporheic Zone Sampling Data

123

Table B1: Supplemental information for the 2016 sampling sites
Sampling Site
A1
A2
A3
B1
B2
C1
C2
C3
C4
E1
E2
E3
F1
F2
F3
F4
G1
G2
H1
H2
H3
I1
I2
J1
K1
K2
L1
L2
L3
M1
M2
O1
O2
R1
R2
R3
R4
S1
S2

Longitude
42.42095
42.41977
42.40931
42.4043
42.40375
42.42623
42.42658
42.4276
42.42841
42.45768
42.45783
42.45801
42.40707
42.40812
42.40926
42.40984
42.44843
42.4483
42.44739
42.4469
42.44679
42.35322
42.35336
42.45716
42.4683
42.46873
42.46234
42.46291
42.46341
42.40838
42.40889
42.36286
42.36563
42.42099
42.4213
42.42145
42.42178
42.39225
42.39254

Latitude
-85.33289
-85.33311
-85.33774
-85.35506
-85.35481
-85.35413
-85.3537
-85.35283
-85.35222
-85.33739
-85.33812
-85.33878
-85.34488
-85.34597
-85.34513
-85.34469
-85.31959
-85.31949
-85.32814
-85.32623
-85.32371
-85.35327
-85.35385
-85.33657
-85.35236
-85.3525
-85.34238
-85.34268
-85.34198
-85.31111
-85.31058
-85.35401
-85.3548
-85.34346
-85.34322
-85.34313
-85.34277
-85.35455
-85.35516

124

Stream Order

1
1
1
3
3
2
2
2
2
2
2
2
3
2
2
2
1
1
2
2
2
3
3
2
1
1
1
1
1
1
1
3
3
1
1
1
1
3
3

Drainage Area (km2)
1.88
1.91
3.00
69.80
69.70
41.72
41.71
41.60
41.48
25.65
25.63
25.61
68.17
50.19
50.09
50.05
1.78
1.77
32.58
32.57
32.48
95.69
95.39
25.66
7.78
7.78
9.72
9.57
9.55
7.39
7.38
87.22
86.81
0.80
0.80
0.80
0.76
72.80
72.75

Figure B1: Site A1 Vertical Porewater Profiles

125

Figure B2: Site A2 Vertical Porewater Profiles

126

Figure B3: Site A3 Vertical Porewater Profiles

127

Figure B4: Site B1 Vertical Porewater Profiles

128

Figure B5: Site B2 Vertical Porewater Profiles

129

Figure B6: Site C1 Vertical Porewater Profiles

130

Figure B7: Site C2 Vertical Porewater Profiles

131

Figure B8: Site C3 Vertical Porewater Profiles

132

Figure B9: Site C4 Vertical Porewater Profiles

133

Figure B10: Site E1 Vertical Porewater Profiles

134

Figure B11: Site E2 Vertical Porewater Profiles

135

Figure B12: Site E3 Vertical Porewater Profiles

136

Figure B13: Site F1 Vertical Porewater Profiles

137

Figure B14: Site F2 Vertical Porewater Profiles

138

Figure B15: Site F3 Vertical Porewater Profiles

139

Figure B16: Site F4 Vertical Porewater Profiles

140

Figure B17: Site G1 Vertical Porewater Profiles

141

Figure B18: Site G2 Vertical Porewater Profiles

142

Figure B19: Site H1 Vertical Porewater Profiles

143

Figure B20: Site H2 Vertical Porewater Profiles

144

Figure B21: Site H3 Vertical Porewater Profiles

145

Figure B22: Site I1 Vertical Porewater Profiles

146

Figure B23: Site I2 Vertical Porewater Profiles

147

Figure B24: Site J1 Vertical Porewater Profiles

148

Figure B25: Site K1 Vertical Porewater Profiles

149

Figure B26: Site K2 Vertical Porewater Profiles

150

Figure B27: Site L1 Vertical Porewater Profiles

151

Figure B28: Site L2 Vertical Porewater Profiles

152

Figure B29: Site L3 Vertical Porewater Profiles

153

Figure B30: Site M1 Vertical Porewater Profiles

154

Figure B31: Site M2 Vertical Porewater Profiles

155

Figure B32: Site O1 Vertical Porewater Profiles

156

Figure B33: Site O2 Vertical Porewater Profiles

157

Figure B34: Site R1 Vertical Porewater Profiles

158

Figure B35: Site R2 Vertical Porewater Profiles

159

Figure B36: Site R3 Vertical Porewater Profiles

160

Figure B37: Site R4 Vertical Porewater Profiles

161

Figure B38: Site S1 Vertical Porewater Profiles

162

Figure B39: Site S2 Vertical Porewater Profiles

163

REFERENCES

164

REFERENCES

Aiken, G.R., 2014. Dissolved organic matter in aquatic systems. Comp. Water Quality and
Purification. 1, pp. 205–220, http://dx.doi.org/10.1016/B978-0-12-382182-9.00014-1.
Avagyan, A., Runkle, B.R.K., Kutzbach, L., 2014. Application of high-resolution spectral
absorbance measurements to determine dissolved organic carbon concentration in remote
areas, J. Hydrol. 517, 435–446, http://dx.doi.org/10.1016/j.jhydrol.2014.05.060.
Baker, M.A., Dahm, C.N., Valett, H.M., 1999. Acetate retention and metabolism in the
hyporheic zone of a mountain stream. Limnol. Oceanogr. 44, 1530-1539,
http://dx.doi.org/ 10.4319/lo.1999.44.6.1530.
Baker, A., 2005. Thermal fluorescence quenching properties of dissolved organic matter, Water
Research 39, 4405-4412, http://dx.doi.org/10.1016/j.watres.2005.08.023.
Battin, T.J., 1999. Hydrologic flow paths control dissolved organic carbon fluxes and
metabolism in an alpine stream hyporheic zone. Water Resour. Res. 35(10), 3159-3169,
http://dx.doi.org/10.1029/1999WR900144.
Battin, T.J., Kaplan, L.A., Findlay, S., Hopkinson, C.S., Marti, E., Packman, A.I., Newbold, J.D.,
Sabater, F., 2008. Biophysical controls on organic carbon fluxes in fluvial networks. Nat.
Geosci. 1(2), 95–100, http://dx.doi.org/10.1038/ngeo101.
Battin, T.J., Luyssaert, S., Kaplan, L.A., Aufdenkampe, A.K., Richter, A., Tranvik, L.J., 2009.
The boundless carbon cycle. Nat. Geosci. 2(9), 598–600,
http://dx.doi.org/10.1038/ngeo618.
Belzile, C., Roesler, C.S., Christensen, J.P., Shakhova, N., Semiletov, I., 2006. Fluorescence
measured using the WETStar DOM fluorometer as a proxy for dissolved matter
absorption, Estuar. Coast. Shelf Sci. 67, 441–449,
http://dx.doi.org/10.1016/j.ecss.2005.11.032.
Bergamaschi, B., Downing, B., Pellerin B., Saraceno J.F., 2009. In situ sensors for dissolved
organic matter fluorescence: bringing the lab to the field, Instrument News, Water
Resources Discipline, USGS, 125.
Bieroza, M.Z., Heathwaite, A.L., 2016. Unravelling organic matter and nutrient biogeochemistry
in groundwater-fed rivers under baseflow conditions: Uncertainty in in situ highfrequency analysis, Sci. Total Environ. 572, 1520-1533,
http://dx.doi.org/10.1016/j.scitotenv.2016.02.046.
Blaen, P.J., Khamis, K., Lloyd, C.E.M., Bradley, C., Hannah, D., Krause, S., 2016. Real-time
monitoring of nutrients and dissolved organic matter in rivers: Capturing event dynamics,
165

technological opportunities and future directions, Sci. Total Environ. 569-570, 647–660,
http://dx.doi.org/10.1016/j.scitotenv.2016.06.116.
Boano, F., Revelli, R., Ridolfi, L., 2009. Quantifying the impact of groundwater discharge on the
surface-subsurface exchange. Hydrol. Process. 23, 2108-2116, http://dx.doi.org/
10.1002/hyp.7278.
Boano, F., Harvey, J.W., Marion, A., Packman, A.I., Revelli, R., Ridolfi, L., Wӧrman, A., 2014.
Hyporheic flow and transport processes: Mechanisms, models, and biogeochemical
implications. Reviews of Geophysics 52(4), 603-679, http://dx.doi.org/
10.1002/2012RG000417.
Boënne, W., Desmet, N., Van Looy, S., Seuntjens, P., 2014. Use of online water quality
monitoring for assessing the effects of WWTP over flows in rivers, Environ. Sci.
Process. Impacts 16, 1510–1518, http://dx.doi.org/10.1039/c3em00449j.
Boulton, A.J., Findlay, S., Marmonier, P., Stanley, E.H., Valett, H.M., 1998. The functional
significance of the hyporheic zone in streams and rivers. Annu. Rev. Ecol. Syst. 29, 59–
81, http://dx.doi.org/10.1146/annurev.ecolsys.29.1.59.
Bowes, M.J., Smith, J.T., Neal, C., 2009. The value of high-resolution nutrient monitoring: A
case study of the River Frome, Dorset, UK, J. Hydrol. 378, 82–96,
http://dx.doi.org/10.1016/j.jhydrol.2009.09.015.
Brinkman, B.M., Hozalski, R.M., 2011. Temporal variation of NOM and its effects on
membrane treatment, Am. Water Work. Assoc. 103(2), 98–106.
Butman, D., Raymond, P.A., 2011. Significant efflux of carbon dioxide from streams and rivers
in the United States. Nat. Geosci. 4(12), 839–842, http://dx.doi.org/10.1038/ngeo1294.
Cardenas, M.B., Wilson, J.L., 2007. Dunes, turbulent eddies, and interfacial exchange with
permeable sediments. Water Resour. Res. 43, http://dx.doi.org/ 10.1029/2006WR005787.
Carstea, E.M., Baker, A., Pavelescu, G., Boomer, I., 2009. Continuous fluorescence assessment
of organic matter variability on the Bournbrook River, Birmingham, UK, Hydrol.
Process. 23(13), 1937–1946, http://dx.doi.org/10.1002hyp.7335.
Chen, R.F., 1999. In situ fluorescence measurements in coastal waters, Org. Geochem. 30, 397–
409, http://dx.doi.org/10.1016/S0146-6380(99)00025-X.
Chen, R.F., Gardner, G.B., 2004. High-resolution measurements of chromophoric dissolved
organic matter in the Mississippi and Atchafalaya River plume regions, Mar. Chem. 89,
103–125, http://dx.doi.org/10.1016/j.marchem.2004.02.026.

166

Conmy, R.N., Del Castillo, C.E., Downing, B.D., Chen R.F., 2014. Experimental design and
quality assurance: In situ fluorescence instrumentation, Aquat. Org. Matter Fluoresc.
190-230.
Cory, R.M., Boyer, E.W., McKnight, D.M., 2011. Spectral methods to advance understanding of
dissolved organic carbon dynamics in forested catchments. Forest Hydrology and
Biogeochemistry: Synthesis of past Research and Future Directions. 117–135,
http://dx.doi.org/10.1007/978-94-007-1363-5.
Creed, I.F. et al., 2015. The river as a chemostat: fresh perspectives on dissolved organic matter
flowing down the river continuum. J. Fish. Aquat. Sci. 72(8), 1272–1285,
http://dx.doi.org/10.1139/cjfas-2014-0400.
Dahm, C.N., 1981. Pathways and mechanism for removal of dissolved organic carbon from leaf
and leachate in streams. Can. J. Fish. Aquat. Sci. 38, 68-76,
http://dx.doi.org/10.1139/f81-009.
Day, G.M., Hart, B.T., McKelvie, I.D., Beckett, R., 1994. Adsorption of natural organic matter
onto goethite. Colloid Surface A, 89, 1-13, http://dx.doi.org/10.1016/09277757(94)02855-9.
Downing, B.D., Pellerin, B.A., Bergamaschi, B.A., Saraceno, J.F., Kraus, T.E.C., 2012. Seeing
the light: The effects of particles, dissolved materials, and temperature on in situ
measurements of DOM fluorescence in rivers and streams, Limnol. Oceanogr. Methods
10, 767–775, http://dx.doi.org/10.4319/lom.2012.10.767.
Downing, B.D., Boss, E., Bergamaschi, B.A., Fleck, J.A., Lionberger, M.A., Ganju, N.K.,
Schoellhamer, D.H., Fujii, R., 2009. Quantifying fluxes and characterizing compositional
changes of dissolved organic matter in aquatic systems in situ using combined acoustic
and optical measurements, Limnol. Oceanogr. Methods 7, 119–131,
http://dx.doi.org/10.4319/lom.2009.7.119.
Dunbar, C.O., 1962. Historical Geology. John Wiley and Sons, Inc., New York. 2nd Ed, 500.
Duff, J.H., Murphy, F., Fuller, C.C., Triska, F.J., Harvey, J.W., Jackman, A.P., 1998. A mini
drivepoint sampler for measuring pore-water solute concentrations in the hyporheic zone
of sand-bottom streams. Limnol. Oceanogr. 43(6), 1378–1383,
http://dx.doi.org/10.4319/lo.1998.43.6.1378.
Etheridge, J.R., Birgand, F., Ii, M.R.B., Smith, B.T., 2014a. Addressing the Fouling of In Situ
Ultraviolet-Visual Spectrometers Used to Continuously Monitor Water Quality in
Brackish Tidal Marsh Waters, J. Environ. Qual. 42, 1896–1901,
http://dx.doi.org/10.2134/jeq2013.02.0049.
Etheridge, J.R., Birgand, F., Osborne, J.A., Osburn, C.L., Ii, M.R.B., Irving, J., Handsel, L.,
Dixon, J., Mackenzie, L., James, J., 2014b. Using in situ ultraviolet-visual spectroscopy
167

to measure nitrogen, carbon, phosphorus, and suspended solids concentrations at a high
frequency in a brackish tidal marsh, Limnol. Oceanogr. Methods 12, 10–22,
http://dx.doi.org/10.4319/lom.2014.12.10.
Fellman, J.B., Hood, E., Spencer, R.G.M., 2010. Fluorescence spectroscopy opens new windows
into dissolved organic matter dynamics in freshwater ecosystems: a review. Limnol.
Oceanogr. 55(6), 2452–2462, http://dx.doi.org/10.4319/lo.2010.55.6.2452.
Fiebig, D.M., Lock M.A., 1991. Immobilization of dissolved organic matter from groundwater
discharging through stream bed. Freshwater Biol. 26, 45-55,
http://dx.doi.org/10.1111/j.1365-2427.1991.tb00507.x.
Fiebig D.M., 1995. Groundwater discharge and its contribution of dissolved organic carbon to an
upland stream. Archiv Für Hydrobiologie, 134, 129-155.
Figure 22, Reprinted from Journal of Geophysical Research, 114, Saraceno, J.F., Pellerin, B.A.,
Downing, B.D., Boss, E., Bachand, P.A.M., Bergamaschi, B.A., High-Frequency in situ
optical measurements during a storm event : assessing relationships between dissolved
organic matter, sediment concentrations, and hydrologic processes, G00F09, Copyright
2009 by the American Geophysical Union with permission from John Wiley and Sons.
http://dx.doi.org/10.1029/2009JG000989
Figure 23, Reprinted from Hydrological Processes, 21, Spencer, R.G.M., Pellerin, B.A.,
Bergamaschi, B.A., Downing, B.D., Kraus, T.E.C., Smart, D.R., Dahlgren, R.A., Hernes,
P.J., Diurnal variability in riverine dissolved organic matter composition determined by
in situ optical measurement in the San Joaquin River (California, USA), 3181-3189,
Copyright 2007 with permission from John Wiley and Sons.
http://dx.doi.org/10.1002/hyp.6887.
Figure 24, Reprinted from Biogeochemistry, 108, Pellerin, B.A., Saraceno, J.F., Shanley, J.B.,
Sebestyen, S.D., Aiken, G.R., Wollheim, W.M., Bergamaschi, B.A., Taking the pulse of
snowmelt : in situ sensors reveal seasonal, event and diurnal patterns of nitrate and
dissolved organic matter variability in an upland forest stream, 183-198, Copyright US
Government 2011 with permission from Springer. http://dx.doi.org/10.1007/s10533-0119589-8.
Figure 25, Reprinted from Hydrological Processes, 20, Waterloo, M.J., Oliveira, S.M., Drucker,
D.P., Nobre, A.D., Cuartas, L.A., Hodnett, M.G., Langedijk, I., Jans, W.W.P., Tomasella,
J., Araujo, A.C., Pimentel, T.P., Munera Estrada, J.C., Export of organic carbon in runoff from an Amazonian rainforest blackwater catchment, 2581-2597, Copyright 2006
with permission from John Wiley and Sons. http://dx.doi.org/10.1002/hyp.6217.
Findlay, S., Strayer, D., Goumbala, C., Gould, K., 1993. Metabolism of streamwater dissolved
organic carbon in the shallow hyporheic zone. Limnol. Oceanogr. (38)7, 1493-1499,
http://dx.doi.org/ 10.4319/lo.1993.38.7.1493.

168

Findlay, S., Sobczak, W.V., 1996. Variability in removal of dissolved organic carbon in
hyporheic sediments. J. North Am. Benthol. Soc. 15(1), 35–41,
http://dx.doi.org/10.2307/1467431.
Findlay, S., Sobczak, W.V., 2000. Microbial communities in hyporheic sediments, in Streams
and Ground Waters, edited by J.A. Jones and P.J. Mulholland, 287-306, Academic Press,
San Diego, Calif.
Fischer, H., Kloep, F., Wilzcek, S., Pusch, M.T., 2005. A river’s liver – microbial processes
within the hyporheic zone of a large lowland river. Biogeochem. 76, 349-371,
http://dx.doi.org/ 10.1007/s10533-005-6896-y.
Four Townships Water Resources Council (FTWRC), 2011. Gull and Augusta Creeks Watershed
Management Plan: the Four Township Watershed Area. Accessed from
http://www.ftwrc.org/publications/FTWA_WMP_final.pdf.
Gomez-Velez, J.D., Harvey, J.W., 2014. A hydrogeomorphic river network model predicts
where and why hyporheic exchange is important in large basins. Geophysical Research
Letters 41(18), 6403-6412, http://dx.doi.org/ 10.1002/2014GL061099.
Graham, P.W., Baker, A., Andersen, M.S., 2015. Field measurement of fluorescent dissolved
organic material as a means of early detection of leachate plumes, Water Air Soil Pollut.
226(211), http:dx.doi.org/10.1007/s11270-015-2475-6.
Grayson, R., Holden, J., 2012. Continuous measurement of spectrophotometric absorbance in
peatland streamwater in northern England : implications for understanding fluvial carbon
fluxes, Hydrol. Process. 39, 27–39, http://dx.doi.org/10.1002/hyp.8106.
Grayson, R.P., Holden, J., 2016. Improved automation of dissolved organic carbon sampling for
organic-rich surface waters, Sci. Total Environ. 543, 44–51,
http://dx.doi.org/10.1016/j.scitotenv.2015.10.149.
Gueguen, C., McLaughlin, F.A., Carmack, E.C., Itoh, M., Narita, H., Nishino, S., 2012. The
nature of colored dissolved organic matter in the southern Canada Basin and East
Siberian Sea, Deep Sea Res. II 81-84, 102–113,
http://dx.doi.org/10.1016/j.dsr2.2011.05.004.
Hamilton, S.K., Augusta Creek DOC concentrations from 1997-2015. Unpublished Laboratory
Data. Accessed 2015.
Harvey, J.W., Fuller, C.C., 1998. Effect of enhanced manganese oxidation in the hyporheic zone
on basin-scale geochemical mass balance. Water Resour. Res. 34(4), 623-636,
http://dx.doi.org/10.1029/97WR03606.
Harvey J.W., Bӧhlke, J.K., Voytek, M.A., Scott, D., Tobias, C.R., 2013. Hyporheic zone
denitrification: Controls on effective reaction depth and contribution to whole-stream
169

mass balance. Water Resour. Res. 49(10), 6298-6316, http://dx.doi.org/
10.1002/wrcr.20492.
Hedin, L.O., von Fischer, J.C., Ostrom, N.E., Kennedy, B.P., Brown., M.G., Robertson, G.P.,
1998. Thermodynamic constraints on nitrogen transformations and other biogeochemical
processes at soil-stream interfaces. Ecology 79(2), 684-703, http://dx.doi.org/
10.2307/176963.
Helms, J.R., Stubbins, A., Ritchie, J.D., Minor, E.C., 2008. Absorption spectral slopes and slope
ratios as indicators of molecular weight, source, and photobleaching of chromophoric
dissolved organic matter. Limnol. Oceanogr. 53(3), 955–969.
http://dx.doi.org/10.4319/lo.2008.53.3.0955.
Helton, A.M., Wright, M.S., Bernhardt, E.S., Poole, G.C., Cory, R.., Stanford, J.A., 2015.
Dissolved organic carbon lability increases with water residence time in the alluvial
aquifer of a river-floodplain ecosystem. J. Geophys. Res. Biogeosciences, 120(4), 693706, http://dx.doi.org/10.1002/2014JG002832.
Hood, E., Gooseff, M.N., Johnson, S.L., 2006. Changes in the character of stream water
dissolved organic carbon during flushing in three small watersheds, Oregon, J. Geophys.
Res. Biogeosciences 111(1), 1–8, http://dx.doi.org/10.1029/2005JG000082.
Inamdar, S., Singh, S., Dutta, S., Levia, D., Mitchell, M., Scott, D., Bais, H., McHale, P., 2011.
Fluorescence characteristics and sources of dissolved organic matter for stream water
during storm events in a forested mid-Atlantic watershed, J. Geophys. Res. 116,
http://dx.doi.org/10.1029/2011JG001735.
Jeong, J.J., Bartsch, S., Fleckenstein, J.H., Matzner, E., Tenhunen, J.D., Lee, S.D., Park, S.K.,
Park, J.H., 2012. Differential storm responses of dissolved and particulate organic carbon
in a mountainous headwater stream, investigated by high-frequency, in situ optical
measurements, J. Geophys. Res. Biogeosciences 117(3), 1–13,
http://dx.doi.org/10.1029/2012JG001999.
Jollymore, A., Johnson, M.S., Hawthorne, I., 2012. Submersible UV–Vis spectroscopy for
quantifying streamwater organic carbon dynamics: implementation and challenges before
and after forest harvest in a headwater stream. Sensors 12, 3798–3813,
http://dx.doi.org/10.3390/s120403798.
Jones, T.D., Chappell, N.A., Tych, W., 2014. First dynamic model of dissolved organic carbon
derived directly from high-frequency observations through contiguous storms, Environ.
Sci. Technol. 48(22), 13289–13297, http://dx.doi.org/10.1021/es503506m.
Kaplan, L.A., Newbold, J.D., 2000. Surface and subsurface dissolved organic carbon, in Streams
and Ground Waters, edited by J. A. Jones, and P. J. Mulholland, 237–258, Academic, San
Diego, Calif..

170

KBS-LTER, Dataset KBS002. Annual Temperature and Precipitation, Data Table KBS002008.61. Accessed on 7 Mar 2017. Permission to use data granted from lead investigator,
Phil Robertson.
Khamis, K., Sorensen, J.P.R., Bradley, C., Hannah, D.M., Lapworth, D.J., Stevens, R., 2015. Insitu tryptophan-like fluorometers: assessing turbidity and temperature effects for
freshwater applications, Environ. Sci. Process. Impacts 17, 740–752,
http://dx.doi.org/10.1039/C5EM00030K.
Kiel, B.A., Cardenas, M.B., 2014. Lateral hyporheic exchange throughout the Mississippi River
network. Nat. Geo. 7, 413-417, http://dx.doi.org/10.1038/ngeo2157.
King, R.H., 1978. Natural history and ecology of the stictochironomous annulicrus (townes)
(diptera:chrionomidae), Augusta Creek, Kalamazoo County, Michigan. A Dissertation.
Proquest Web. 5 Nov. 2016.
Klinkhammer, G.P., Chin, C.S., Wilson, C., Rudnicki, M.D., German, C.R., 1997. Distributions
of dissolved manganese and fluorescent dissolved organic matter in the Columbia River
estuary and plume as determined by in situ measurement, Mar. Chem. 56, 1–14,
http://dx.doi.org/10.1016/S0304-4203(96)00079-5.
Koehler, A.K., Murphy, K., Kiely, G., Sottocornola, M., 2009. Seasonal variation of DOC
concentration and annual loss of DOC from an Atlantic blanket bog in South Western
Ireland, Biogeochemistry 95(2), 231–242, http://dx.doi.org/10.1007/s10533-009-9333-9.
Kowalczuk, P., Zablocka, M., Sagan, S., Kulinski, K., 2010. Fluorescence measured in situ as a
proxy of CDOM absorption and DOC concentration in the Baltic Sea, Oceanologia
52(3), 431–471, http://dx.doi.org/10.5697/oc.52-3.431.
Lakowicz, J.R., 2006. Principles of fluorescence spectroscopy, 3rd ed., Springer Science and
Business Media.
Langergraber, G., Gupta, J.K., Pressl, A., Hofstaedter, F., Lettl, W., Weingartner, A.,
Fleischmann, N., 2004. On-line monitoring for control of a pilot-scale sequencing batch
reactor using a submersible UV/VIS spectrometer, Water Sci. Technol. 50(10), 73–80.
Laudon, H., Berggren, M., Ågren, A., Buffam, I., Bishop, K., Grabs, T., Jansson, M., Köhler, S.,
2011. Patterns and dynamics of dissolved organic carbon (DOC) in boreal streams: The
role of processes, connectivity, and scaling. Ecosystems, 14(6), 880–893,
http://dx.doi.org/10.1007/s10021-011-9452-8.
Lee, E.J., Yoo, G.Y., Jeong, Y., Kim, K.U., Park, J.H., Oh, N.H., 2015. Comparison of UV-VIS
and FDOM sensors for in situ monitoring of stream DOC concentrations, Biogeosciences
12(10), 3109–3118, http://dx.doi.org/10.5194/bg-12-3109-2015.

171

Lutz, B.D., Bernhardt, E.S., Roberts, B.J., Cory, R.M., Mulholland, P.J., 2012. Distinguishing
dynamics of dissolved organic matter components in a forested stream using kinetic
enrichments. Limnol. Oceanogr. 57, 76-89, http://dx.doi.org/10.4319/lo.2012.57.1.0076.
Mann, P.J., Davydova, A., Zimov, N., Spencer, R.G.M, Davydov, S., Bulygina, E., Zimov, S.,
Holmes, R.M., 2012. Controls on the composition and lability of dissolved organic matter
in Siberia’s Kolyma River basin. J. Geophys. Res. Biogeosciences 117(1), 1–15,
http://dx.doi.org/10.1029/2011JG001798
Manny, B.A., Wetzel, R.G., 1973. Diurnal changes in dissolved organic and inorganic carbon
and nitrogen in a hardwater stream. Freshw. Biol. 3(1), 31–43,
http://dx.doi.org/10.1111/j.1365-2427.1973.tb00060.x.
Mahan, D.C., Cummins, K.W., 1974. A profile of Augusta Creek in Kalamazoo and Barry
Counties, Michigan. Tech. Rep. No. 3. W.K. Kellogg Biological Station, Michigan State
University.
Mast, M.A., Murphy, S.F., Clow, D.W., Penn, C.A., Sexstone, G.A., 2016. Water-quality
response to a high-elevation wild fire in the Colorado Front Range, Hydrol. Process. 30,
1811–1823, http://dx.doi.org/10.1002/hyp.10755.
McDonald, S., Bishop, A.G., Prenzler, P.D., Robards, K., 2004. Analytical chemistry of
freshwater humic substances. Analytica Chimica Acta 527(2), 105-124,
http://dx.doi.org/10.1016/j.aca.2004.10.011.
McDowell, W.H., 1985. Kinetics and mechanisms of dissolved organic carbon retention in a
headwater stream. Biogeochemistry 4, 329-352, http://dx.doi.org/10.1007/BF02187376.
McKnight, D.M., Boyer, E.W., Westerhoff, P.K., Doran, P.T., Kulbe, T., Andersen, D.T., 2001.
Spectrofluorometric characterization of dissolved organic matter for indication of
precursor organic material and aromaticity. Limnol. Oceanogr. 46(1), 38–48,
http://dx.doi.org/10.4319/lo.2001.46.1.0038.
Monteith, D.T., et al., 2007. Dissolved organic carbon trends resulting from changes in
atmospheric deposition chemistry. Nature 450(7169), 537–540,
http://dx.doi.org/10.1038/nature06316.
Neal, C. et al., 2012. High-frequency water quality time series in precipitation and stream flow:
From fragmentary signals to scientific challenge, Sci. Total Environ. 434, 3–12,
http://dx.doi.org/10.1016/j.scitotenv.2011.10.072.
Nogaro, G., Datry, T., Mermillod-Blondin, F., Foulquier, A., Montuelle, B., 2013. Influence of
hyporheic zone characteristics on the structure and activity of microbial assemblages.
Freshw. Biol. 58(12), 2567-2583, http://dx.doi.org/10.1111/fwb.12233.

172

Pellerin, B.A., Saraceno, J.F., Shanley, J.B., Sebestyen, S.D., Aiken, G.R., Wollheim, W.M.,
Bergamaschi, B.A., 2012. Taking the pulse of snowmelt : in situ sensors reveal seasonal,
event and diurnal patterns of nitrate and dissolved organic matter variability in an upland
forest stream, Biogeochemistry 108, 183–198, http://dx.doi.org/10.1007/s10533-0119589-8.
Peters, N.E., Ratcliffe, E.B., 1998. Tracing Hydrologic Pathways Using Chloride at the Panola
Mountain Research Watershed, Georgia, USA. Water, Air, and Soil Pollution 105, 263275, http://dx.doi.org/ 10.1023/A:1005082332332.
Pinder, G.F., Jones, J.F., 1969. Determination of the ground-water component of peak discharge
from the chemistry of total runoff. Water Resour. Res. 5(2), 438-445, http://dx.doi.org/
10.1029/WR005i002p00438.
Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R., Prestegaard, K.L., Richter, B.D., Sparks, R.E.,
Stromberg, J.C., 1997. The Natural Flow Regime: A paradigm for river conservation and
restoration. Bioscience 47(11), 769–784, http://dx.doi.org/10.2307/1313099.
Potter, B., Wimsatt J., 2003. USEPA Method 415.3 –Determination of TOC and SUVA at
254nm in Source Water and Drinking Water. National Exposure Research Laboratory,
Office of Research and Development, USEPA, Cincinnati, OH.
Raymond, P.A., et al., 2013. Global carbon dioxide emissions from inland waters. Nature
503(7476), 355–359, http://dx.doi.org/10.1038/nature12760.
Raymond, P.A., Saiers, J.E., Sobczak, W.V., 2016. Hydrological and biogeochemical controls on
watershed dissolved organic matter transport: pulse‐shunt concept. Ecology 97(1), 5-16,
http://dx.doi.org/10.1890/14-1684.1.
Roley, S.S., Griffiths, J.R., Levi, P.S., Patrick, C.J., Sadro, S., Zarnetske, J.P., 2014. Taking the
pulse of the ecosystem: progress in quantifying aquatic ecosystem health, In P.F. Kemp
[ed.], Eco-DAS X Symposium Proceedings, ASLO,
http://dx.doi.org/10.4319/ecodas.2014.978-0-9845591-4-5-88.
Ruhala, S.S., Zarnetske, J.P., 2017. Using in-situ optical sensors to study dissolved organic
carbon dynamics of streams and watersheds: A review. Sci. Total Environ. 575, 713-723,
http://dx.doi.org/10.1016/j.scitotenv.2016.09.113.
Saraceno, J.F., Pellerin, B.A., Downing, B.D., Boss, E., Bachand, P.A.M., Bergamaschi, B.A.,
2009. High-Frequency in situ optical measurements during a storm event : assessing
relationships between dissolved organic matter, sediment concentrations, and hydrologic
processes, J. Geophys. Res. 114, http://dx.doi.org/10.1029/2009JG000989.
Schindler, J.E., Krabbenhoft, D.P., 1998. The hyporheic zone as a source of dissolved organic
carbon and carbon gases to a temperate forested stream. Biogeochem. 43(2), 157-174,
http://dx.doi.org/10.1023/A:1006005311257.
173

Sobczak, W.V., Findlay, S., 2002. Variation in bioavailability of dissolved organic carbon
among stream hyporheic flowpaths. Ecology 83(11), 3194–3209,
http://dx.doi.org/10.2307/3071853.
Sobczak W.V., Raymond, P.A., 2015. Watershed hydrology and dissolved organic matter export
across time scales: minutes to millennium, Freshwater Science 34, 392-398,
http://dx.doi.org/10.1086/679747.
Spencer, R.G.M., Aiken, G.R., Butler, K.D., Dornblaser, M.M., Striegl, R.G., Hernes, P.J., 2009.
Utilizing chromophoric dissolved organic matter measurements to derive export and
reactivity of dissolved organic carbon exported to the Arctic Ocean: A case study of the
Yukon River, Alaska, Geophys. Res. Lett. 36(6), 1–6,
http://dx.doi.org/10.1029/2008GL036831.
Spencer, R.G.M., Pellerin, B.A., Bergamaschi, B.A., Downing, B.D., Kraus, T.E.C., Smart,
D.R., Dahlgren, R.A., Hernes, P.J., 2007. Diurnal variability in riverine dissolved organic
matter composition determined by in situ optical measurement in the San Joaquin River
(California, USA). Hydrol. Process. 21, 3181–3189, http://dx.doi.org/10.1002/hyp.6887.
Stelzer, R.S., Scott, J.T., Bartsch, L.A., 2015. Buried particulate organic carbon stimulates
denitrification and nitrate retention in stream sediments at the groundwater-surface water
interface. Freshwater Science 34, 161-171, http://dx.doi.org/10.1086/678249.
Storey, R.G., Fulthorpe, R.R., Williams, D.D., 1999. Perspectives and predictions on the
microbial ecology of the hyporheic zone. Freshwater Biology 41(1), 119-130,
http://dx.doi.org/ 10.1046/j.1365-2427.1999.00377.x.
Strahler, A.N., 1957. Quantitative analysis of watershed geomorphology. Trans. Amer. Geophys.
Un. 38, 913-920.
Strohmeier, S., Knorr, K., Reichert, M., Frei, S., Fleckenstein, J.H., Peiffer, S. Matzner, E., 2013.
Concentrations and fluxes of dissolved organic carbon in runoff from a forested
catchment: insights from high frequency measurements, Biogeosciences 10, 905–916,
http://dx.doi.org/10.5194/bg-10-905-2013.
Thurman, E. M., 1985. Amount of organic carbon in natural waters. Organic geochemistry of
natural waters. Vol. II in Developments in Biogeochemistry, 7-65.
United States Geological Survey (USGS), 2016. National Water Information System data
available on the World Wide Web (USGS Water Data for the Nation). Accessed
November 4, 2016 at https://waterdata.usgs.gov/mi/nwis/uv?site_no=04105700.
United States Geological Survey (USGS), 2017. National Water Information System data
available on the World Wide Web (USGS Water Data for the Nation). Accessed March 7,
2017 at https://waterdata.usgs.gov/mi/nwis/uv?site_no=04105700.

174

Url, S., Hart, D.D., 2012. International Association for Ecology Community Organization in
Streams: The Importance of Species Interactions. 91(2), 220–228,
http://dx.doi.org/10.1007/BF00317787.
Valett, H.M., Morrice, J.A., Dahm, C.N., 1996. Parent lithology, surface-groundwater exchange,
and nitrate retention in headwater streams. Limnol. Oceanogr. 41(2), 333-345,
http://dx.doi.org/10.4319/lo.1996.41.2.0333.
Vannote, R.L., Minshall, W.G., Cummins, K.W., Sedell, J.R., Cushing, C.E., 1980. The River
Continuum Concept. Can. J. Fish. Aquat. Sci. 37(1), 130–137,
http://dx.doi.org/10.1139/f80-017.
Waterloo, M.J., Oliveira, S.M., Drucker, D.P., Nobre, A.D., Cuartas, L.A., Hodnett, M.G.,
Langedijk, I., Jans, W.W.P., Tomasella, J., Araujo, A.C., Pimentel, T.P., Munera Estrada,
J.C., 2006. Export of organic carbon in run-off from an Amazonian rainforest blackwater
catchment, Hydrol. Process. 20, 2581–2597, http://dx.doi.org/10.1002/hyp.6217.
Watras, C.J., Hanson, P.C., Stacy, T.L., Morrison, K.M., Mather, J., Hu, Y., Milewski, P., 2011.
A temperature compensation method for CDOM fluorescence sensors in freshwater,
Limnol. Oceanogr. Methods 9, 296–301, http://dx.doi.org/10.4319/lom.2011.9.296.
Watras, C.J., Morrison, K.A., Mather, J., Milewski, P., Hanson, P.C., 2014. Correcting CDOM
fluorescence measurements for temperature effects under field conditions in freshwaters.
Limnol. Oceanogr. Methods 12(1), 23-24, http://dx.doi.org/10.4319/lom.2014.12.23.
Weishaar, J.L., Aiken, G.R., Bergamaschi, B.A., Fram, M.S., Fujii, R., Mopper, K., 2003.
Evaluation of specific ultra-violet absorbance as an indicator of the chemical composition
and reactivity of dissolved organic carbon. Environ. Sci. Technol. 37, 4702–4708,
http://dx.doi.org/10.1021/es030360x.
Wilson, H.F., Saiers, J.E., Raymond, P.A., Sobczak, W.V., 2013. Hydrologic drivers and
seasonality of dissolved organic carbon concentration, nitrogen content, bioavailability,
and export in a forested New England stream, Ecosystems 16(4), 604–616,
http://dx.doi.org/10.1007/s10021-013-9635-6.
Wondzell, S.M., 2011. The role of the hyporheic zone across stream networks. Hydrol. Process.
25, 3525-3532, http://dx.doi.org/10.1002jhyp.8119.
Worrall, F., Howden, N.J.K., Burt, T.P., 2015. Understanding the diurnal cycle in fluvial
dissolved organic carbon – The interplay of in-stream residence time, day length and
organic matter turnover, J. Hydrol. 523, 830–838,
http://dx.doi.org/10.1016/j.jhydrol.2015.01.075.
Zarnetske, J.P., Haggerty, R., Wondzell, S.M., Baker, M.A., 2011. Labile dissolved organic
carbon supply limits hyporheic denitrification. J. Geophys. Res. Biogeosciences 116(4),
http://dx.doi.org/10.1029/2011JG001730.
175