THE DELIVERY, SEASONAL STORAGE, AND TROPHIC TR NSFER OF
MARINE-DERIVED NUTRIENTS WITHIN
A STR AM-RIPARIAN ECOTONE

by
Leah M. Vanden Busch
B.Sc. Un1versity of Wisconsin, 2008

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
IN
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES
(ENVIRONMENTAL SCIENCE)

UNIVERSITY OF NORTHERN BRITISH COLUMIBA
September 2015
© Leah M. Vanden Busch, 2015

Abstract

Abstract
The objectives of this study were to evaluate nutrient delivery, storage, and trophic
transfer between the Horsefly River spawning channel and its riparian zone during
the 2011 sockeye salmon run and subsequent 2012 spring. The hyporheic zone was
investigated as a pathway for nutrient exchange, transformation, and storage by
analyzing water movement and ammonium (NH4 ) concentration. Using stable
isotopic analysis, the assimilation and storage of marine-derived nutrients (MDN)
were evaluated through multiple trophic levels. Results identified lateral hyporheic
flow as a dominant pathway for MDN delivery to deep-rooted vegetation,
specifically willow trees <7-m from the streambank. Amphipods and chironomids
appeared to assimilate MDN in the fall and maintained high spring biomass,
indicating an important nutrient linkage in the salmon feedback loop. Finally, a
temporal reversal in NH4+ concentration across the stream-riparian interface and a
seasonal shift in nutrient fluxes revealed reciprocal subsidies that highlight the
interconnectedness of these adjacent systems.

..

ll

Table o- Contents

I ll

Table of Contents

..

Abstract................................................................................................................................... t t

.. .

Table of Contents ................................................................................................................ lll

List of Tables .......................................................................................................................... v

..

List of igu.res ...................................................................................................................... vtz

AcknowledgeDlent ...............................................................................................................%

1 Ecotones, Salmon, and Marine-Derived Nutrien ts: A Literature Review ............. l
1.1 Zone Ecology ............................................................................................................ 1
1.1.1 Ecotones ............................................................................................................. 1
1.1.2 Riparian Zone Ecology .................................................................................... 2
1.1.3 Hyporheic Zone Ecology ................................................................................3
1.1.4 Interstitial Zone Overlap ................................................................................ .4
1.2 Pacific Salmon.......................................................................................................... 5
1.2.1. Marine-Derived Nutrients ............................................................................. ?
1.2.2 Horsefly River in the Fraser River Watershed ............................................. 8
1.3 Nitrogen .............................................................................................................. 10
1.3.1 The Nitrogen Cycle ........................................................................................ 10
1.3.2 Marine-Derived N itrogen (b 15 N ) ............................................................. 11
1.4 Carbon..................................................................................................................... 12
1.4.1. The Carbon Cycle .......................................................................................... 12
1.4.2. Marine-Derived Carbon (b 13C).................................................................... 13
1.5 Stream-Riparian Ecosystem Response ............................................................... 14
1.5.1 Aquatic ............................................................................................................ 14
1.5.2 Terrestrial. ....................................................................................................... 16
1.6 Research Objectives and Hypotheses ................................................................ 19
1.6.1 Delivery and Storage of Ammonium in the Hyporheic Zone ................ 20
1.6.2 Storage and Transport of MDN in a the Stream-Riparian Ecosystem . .21
2 Nitrogen Cycling in the Hyporheic Zone...............................•....................................23
2.1 Hyporheic Zone as Vector for Nutrient Transport ........................................... 23
2.1.1 Defining the Hyporheic Zone ...................................................................... 23
2.1.2 Physical Characteristics of Hyporheic Flow .............................................. 26
2.1.3. Mod eling Hyporheic Flow .......................................................................... 30
2.1.4 Hypotheses ..................................................................................................... 31
2.2 Methods .................................................................................................................. 31
2.2.1 Study Site ........................................................................................................ 31
2.2.2 Study Design .................................................................................................. 37
2.2.3 Site Ch aracteristics and Aqu ifer Variables ................................................. 42
2.2.4 Groundwater Modeling ................................................................................ 44
2.2.5 Water Chemistry ............................................................................................ 47

Table of Contents

lV

2.2.6 Statistical Analysis......................................................................................... 48
2.3 Results ..................................................................................................................... 51
2.3.1 Precipitation and Discharge......................................................................... 51
2.3.2 Aquifer Variables........................................................................................... 53
2.3.3 Geographic Information Systems (GIS) .......................................... ,.......... 56
2.3.4 Ammonium .................................................................................................... 63
2.4 Discussion ............................................................................................................... 71
2.4.1. Defining the Hyporheic Zone of the Horsefly River Channel.. ............ .71
2.4.2. Sources and Uptake of Ammonium in Surface Water ............................ 74
2.4.3. Sources and Uptake of Ammonium in the Hyporheic Zone .................. 79
2.4.4. Limitations and Lessons.............................................................................. 86
2.5. Conclusions .......................................................................................................... 89
3 Marine-Derived Nutrients in the Stream-Riparian Ecosystem ............................... 91
3.1 Stream-Riparian Food Webs ................................................................................ 91
3.1.1 Aquatic Food Webs....................................................................................... 94
3.1.2 Terrestrial Food Webs................................................................................... 97
3.1.2. Hypotheses.................................................................................................... 99
3.2 Methods................................................................................................................ 101
3.2.1 Field Sampling and Laboratory Analysis................................................. 101
3.2.2 Statistical Analysis ....................................................................................... 105
3.3 Results................................................................................................................... 112
3.3.1 Temporal Trends of MDN .......................................................................... 112
3.3.2 Macroinvertebrate Assimilation of MDN ................................................ 116
3.3.3 Macroinvertebrate Diversity and Abundance ......................................... 116
3.3.4 Spatial Trends of MDN in the Riparian Zone .......................................... 121
3.3.5 Tree Cores..................................................................................................... 125
3.4 Discussion............................................................................................................. 127
3.4.1. Macroinvertebrate Community of the Horsefly River .......................... 127
3.4.2 Nutrient Uptake and Assimilation in the Riparian Zone ...................... 144
3.5 Conclusions.......................................................................................................... 151
4 Conclusions and Management Implications ............................................................ 154
4.1 Conclusions.......................................................................................................... 154
4.2 Management Implications .................................................................................. 158

Bibliography....................................................................................................................... 160

List of Tables

v

List of Tables
Table 2.1. Hydraulic conductivity (K) measurements from each study piezometer on
Apri11, 2012. The soil material type, which is associated with standard K values, is
adapted from Freeze & herry (1979)................................................................................ 54
Table 2.2. The change in ammonium (NH 4 ) concentration at each sampling location
between Fall 2011 and Spring 2012 sampling periods. "Up," "Mid," and "Down"
represent surface water samples......................................................................................... 66
Table 2.3. Sockeye salmon (0. nerka ) spawning escapement at the Horsefly River
spawning channel and mean N H 4 concentration for surface water samples during
Active pawn I and 11 ........................................................................................................... 67
Table 2.4. Variance inflation factors (VIF) used to describe multicollinearity for
predictor variables used in multiple regression analyses ...............................................68
Table 2.5. Model selection results of spatial variables for all hyporheic (group 1) and
riparian hyporheic (group 2) models. The effect size is written in parenthesis after
each predictor variable. The most parsimonious linear and multiple regressions are
ranked by second-order Akaike's Information Criterion (AICc) scores. SB =
streambank, TMB = top of meander bend, df = degrees of freedom, wi = Akaike
weights.................................................................................................................................... 70
Table 2.6. Model selection results of temporal variables for all hyporheic (group 1)
and riparian hyporheic (group 2) regression models. The effect size is written in
parenthesis after each predictor variable. The most parsimonious linear and multiple
regressions are ranked by second-order Akaike's Information Criterion (AICc)
scores. df = degrees of freedom, wi = Akaike weights.................................................... 71
Table 2.7. Examples of streams w ith increased NH 4+concentration during salmon
runs, as compared to the present study. Most values were estimated from published
graphs..................................................................................................................................... 75
Table 3.1. Statistically significant differences observed in organisms between the fall
and spring sampling periods............................................................................................ 116
Table 3.2. Taxonomic identification and enumeration of macroinvertebrates in the
Horsefly River spawning channel during the 2011 salmon spawn and the
subsequent spring (labeled as Post-winter). The scientific classification of major
taxonomic groups are identified in parentheses. Functional feeding group (FFG)
classifications are as follows: CF (collector-filterers), CG (collector-gatherers), P
(predators), SC (scrapers), and SH (shredders) . ............................................................ 118
Table 3.3. Shannon Index and Effective Number of Species (ENS) diversity values,
along with Simpson's Index and Gini coefficient evenness values for the Horsefl y
River spawning channel and the Horsefly River........................................................... 121
Table 3.4. Variables that display statistical significance in relation to terrestrial
isotopic signatures ............................................................................................................. 122
Table 3.5. P-values displaying significant differences in b13C at all vegetation sites
compared to site T1-0, which was located directly on the streambank. Isotopic

( '/

si 11 tu r .

T

v rag

b tvv

n

h

m lin

n

........................................... 12

nt f l t
r n ( ) n ni tr
t l i n t th l--1 r. 0 Ri r ~

nin

ring 2 12..................................................................... 14h

List of Figures

Vll

List of Figures
Figure 1.1. Diagram depicting the location of the hyporheic zone as an interstitial
area between stream water and groundwater. Local geomorphic features, such as
the meander bend pictured here, can form localized flow features within stream and
banks. (From Alley et al. 2002.) ............................................................................................. 4
Figure 1.2. Schematic diagram showing the hydrological ecosystems and ecotones
that can interact with the hyporheic zone and their relative location to the hyporheic
zone (adapted from Boulton et al. 1998)............................................................................. .5
Figure 1.3. Sockeye salmon populations in the Horsefly River from 1960 to 2013.
Note: 2013 data are preliminary and could change. Source: Dept. of Fisheries and
Oceans Canada ..................................................................................................................... 10
Figure 1.4. MDN transfer, assimilation, and trophic transport in a stream-riparian
ecosystem is complex in nature and influential in several interacting ecosystems ... .19
Figure 2.1. Map of British Columbia displaying the town of Horsefly, the location
where this study was conducted ........................................................................................ 32
Figure 2.2. Aerial view of the Horsefly River spawning channel in Horsefly, BC. The
Horsefly River is depicted with flow direction arrows ................................................... 34
Figure 2.3. The study site in June 2011, showing meander bend in spawning channel
and adjacent riparian zone. Piezometers F and G are located upstream (right side in
this photo) of the riparian forest and are located in the un-forested section ............... 35
Figure 2.4. Map of the experimental study site on the Horsefly River spawning
channel. Contours display 0.1-m changes in land elevation (NAD83)......................... 38
Figure 2.5. Diagram of piezometer marking levels of elevation used in the study.
Measurements of piezometer holes are displayed in the inset diagram. The variable
length of well screen (L) is described in the Hvorslov equation ................................... .39
Figure 2.6. Sampling periods timed to characterize the temporal nutrient delivery
and retention during the 2011 sockeye salmon run and the subsequent spring ........ 42
Figure 2.7. Workflow chart showing steps for hyporheic water flow analysis using
Spatial Analysis tools in ArcGIS ......................................................................................... 47
Figure 2.8. Characterization of stream conditions at the Horsefly River spawning
channel in the fall of 2011. The top graph displays precipitation and the bottom
graph displays Horsefly River discharge at two measuring sites................................. 52
Figure 2.9. Characterization of stream conditions at the Horsefly River spawning
channel in the spring of 2012. Precipitation is displayed on the top graph and the
bottom graph displays discharge measurements from two sites on the Horsefly
River....................................................................................................................................... 53
Figure 2.10. Description of three soil profiles from the riparian zone. The diagram
depicts relative heights (mASL) and distances from the streambank for each soil pit.
Soil structure and color are described from the Munsell Soil Color Book. .................. 55
Figure 2.11. Vertical hydraulic gradient in the fall (a) and spring (b) at instream

List 91_ Figures

Vlll

piezometers C and D. Fall data (a) were only collected on days for which data are
displayed ................................................................................................................................ 56
Figure 2.12. Hydraulic head (mASL) at the Horsefly River spawning channel during
each month of the study. Note that each map has a unique scale ............................ 57-59
Figure 2.13. Surface water hydrograph between Apri11 and May 8, 2012. Grey
rectangles encompass the dates used for movies depicting hydraulic head
fluctuations . .......................................................................................................................... 60
Figure 2.14. Hyporheic water flow direction and relative speed, represented by
arrow direction and size, as modeled by spatial analysis of hydraulic head contours
in ArcGIS during each month of the study. Land surface elevation is depicted by the
contour lines (0.1-m intervals) ...................................................................................... 61-63
Figure 2.15. Box plots displaying all ammonium (NH4 ) concentration data at each
sampling location. "Up," "Mid," and "Dow n" represent surface water samples ...... 64
Figure 2.16. Mean ammonium (NH4 ) concentration, represented by symbol size, of
all sampling weeks at the hyporheic sampling locations. Withdrawal depth is
measured as meters above sea level (mASL).................................................................... 65
Figure 2.17. Mean ammonium concentration (!Jg/L N ) at water sampling locations
during each sampling period, with error bars show ing standard error. The
"Instream" group represents piezometers located in the stream (n=2); "Riparian"
represents the riparian hyporheic piezometers (n=8); "Surface" represents the
surface water sampling locations (n=3) ............................................................................. 67
Figure 2.18. A conceptual model of the nitrogen cycle at the interface of surface and
hyporheic water. (From Peterson et al. 2001. Reprinted with permission from
AAAS.).................................................................................................................................... 77
Figure 2.19. A profile of a stream and adjacent riparian zone displaying the
subsurface zones that may have higher ammonium (NH /) concentrations............... 85
Figure 3.1. A decaying sockeye salmon (0. nerka) in the Horsefly River spawning
channel. Biofilms, filamentous algae, and leaf litter can be seen on the surface
substrate................................................................................................................................. 93
Figure 3.2. A typical Lorenz curve which graphically represents the inequality in a
population. The areas above and below the Lorenz curve, marked A and B,
respectively, simplifies the Gini coefficient equation, which is A / (A+B)................. 110
Figure 3.3. Dual isotope plot (b 13C %o and b15 N r'oo ) from all organisms and materials
collected during the Fall 2011 salmon run. Isotopically enriched species are located
in the upper right comer. Mean ± 1 SEM ........................................................................ 113
Figure 3.4. Dual isotope plot (b 13C %o and b15 N %o ) from all organisms and materials
collected in Spring 2012, besides for salmon tissue, collected in Fall 201 1, which is
included as reference point. Mean ± 1 SEM .................................................................... 114
Figure 3.5. Temporal differences of vegetation and macroinvertebrate species
collected in the fall and spring sampling periods. Units of b 13C and b15 N are parts per
thousand (%o ). Mean ± 1 SEM ........................................................................................... 115
Fig~ue 3.6. Bar~ displaying the percent biomass of each FFG during each sampling
pen od. The heigh t of the bars represents weekly biomass as a percentage of the total

List of Figures

lX

biomass. The actual weight of amphipods varied from a minimum of 50.2 mg during
Pre-spawn and a maximum of 682.9 mg during Active Spawn 111. ............................ 119
Figure 3.7. Pie charts displaying the relative abundance (as a percentage) of each
FFG during each sampling period .................................................................................... 120
Figure 3.8. The number of individuals during each sampling week for commonly
identified taxa at the HFC .................................................................................................. 120
Figure 3.9. Dual isotope plot of willow (S. Iucida) leaves. Each cross represents the
average of all sampling periods at each sampling location. Units of b 13C and b 15N are
parts per thousand {o/oo). Mean ± 1 SEM .......................................................................... 123
Figure 3.10. Box plots displaying b 13C (roo) of soil horizons from three soil pits across
the study site. Black dots represent actual b13C signatures......................................... 124
Figure 3.11. Stable isotopic signatures of b 13C (roo) and b 15N (roo) at soil pit 1, which
was located closest to the streambank. ............................................................................ 125
Figure 3.12. Annual ring width indices for a willow tree (Salix sp.) located centrally
on the study site between the years 1956 and 2011. The vertical red dotted Jine
delineates the year the spawning channel was built, 1989........................................... 126
Figure 3.13. Standardized values of annual willow tree growth, measured by ring
widths, and salmon escapement in the Horsefly River spawning channel since its
establishment in 1989......................................................................................................... 127
Figure 3.14. Number of individuals counted for each functional feeding group (FFG)
during each of six sampling periods, beginning in August 2011 and ending in May
2012 . ...................................................................................................................................... 131
Figure 3.15. Boxplots displaying ammonium (NH/ ) concentration of all hyporheic
water samples. Active spawn includes data from three sampling weeks, Active
spawn I, II, and III. Outliers are denoted by red dots ................................................... 132
Figure 3.16. Boxplots displaying all b13C roo (top) and b 15N %o (bottom) signatures
from the shredder functional feeding group (FFG) during each sampling week.
Black dots denote actual values........................................................................................ 134
Figure 3.17. Boxplots displaying b13C roo (top) and b 15N roo (bottom) signatures from
two species of the collector-gatherer functional feeding group (FFG). Plots on the left
show chironomid isotopic signatures and ephemeroptera are displayed on the right.
Pre-spawn isotopic signatures are missing for these species. Black dots denote actual
values ....................................................................................................................................137
Figure 3.18. Chironomid larvae and pupae isotopic signatures collected in May 2012.
Increased b15N (%o) and decreased b13C (roo) in pupae are demonstrating nutrient
changes due to metamorphosis. Black dots denote actual isotope values ................. 138
Figure 3.19. Boxplots displaying b13C %o (top) and b15N %o (bottom) signatures of
Hirudinae (leeches) from the predator functional feeding group (FFG) during the fall
sampling weeks. Spring isotope values were not obtained due to low abundance and
mass. Black dots denote actual values ............................................................................. 142

Acknowledgements

X

First and foremost, I must express my deep appreciation and forever love to two
ladies I met at UNBC- Heidi Knudsvig and Jocelyn Joe-Strack. Without your
constant friendship, support, and encouragement, this thesis would not exist. I am
beyond words happy to have you both in my life and am looking forward to all of
our adventures yet to come!
My supervisory duo, Drs. Ellen Petticrew and John Rex, deserve my sincerest
thanks, particularly for all the time, effort, experience, and guidance you have given
to myself and this project. I have grown immeasurably as a scientist and a woman in
the last 5 years, and am incredibly grateful for the continuous support and flexibility
you have given to the direction of my project and my life.
Thanks to my committee members, Drs. Dan Moore and Mark Wipfi, for your time
and insightful contributions to the development of this thesis. A special thank you
goes to Mark for offering me a job that guided me home to Alaska. Also thanks to
my external advisor, Dr. Nikolaus Gantner, for his time and comments.
During my years in British Columbia, the time I spent at the QRRC were the best of
all. I had an unforgettable international crew of encouraging and fun-loving young
scientists with whom to share meals, camping trips, and advice. These wonderful
people inc.~ude Maud Barrel, Faye Hirshfield, Alex Koiter, Peng Wu, Sam Albers,
and Anja Orster. Thanks also to Maud, Nate Clark, Franny Enyedy, and Anna
Johnson for your help in the field. Particular thanks goes to Faye and Laszlo Enyedy
for the design and fabrication of 50 skookum piezometers. I am especially thankful
to Laszlo and Rick Holmes for welcoming us into your homes and your beautiful
community of Likely. Rick, your lifestyle and joyfulness will always be my
inspiration. I am honored to be the "younger and better looking" version of
yourself.
There are countless people who have generously provided technical assistance and
expertise for which I have a lifetime of karma to repay. Huge thanks to Steffi
LaZerte for walking me through R! Thanks to Lisa Poirier for help with inverts,
Allen Esler for nutrient analyses, Kate Hrinkevich for counting tree rings, Sally
Letsinger and Shane Doddridge for GIS help, Barry Booth from TLC, and Dan
Rinella and Aaron Christ for statistical and mixing model advice. I gratefully
acknowledge financial assistance from BC Real Estate Grant, Peace I Williston
Aquatic Research Award, and UNBC Travel Award.
The majority of my graduate school years were spent away from campus, so there
are a few places where I spent a surprisingly large amount of time that I'd like to
acknowledge. Cheers to St. Norbert College and the University of AlaskaAnchorage for keeping me in an academic atmosphere. Thanks also to my little
cabin in Girdwood and the Deshka River weir for providing a peaceful place to
write.
Th~nk you to the mo~t important people in my life, my family, for learning to stop

asktng about my thesis. I started grad school with none, but now I have four sweet

nd r m un ti
I
f n tur .
f r sh
th nk

I

I

tr 111

h mag

(fifth 11th
!) wh m I h
th r , t f 111 1if t th nk m
111 u11t f
ti 11 I
, lik . n

in, till my
II is,
s lw ysl
I

I

t

v

to

Chapter 1. Literature Review

1

Chapter 1

Ecotone , almon, and Marine-Derived Nutrients: A Literature Review
1.1 Zone Ecology
1.1.1 Ecotones
An ecotone represents the boundary or transitional zone between adjacent
ecosystems. The concept of ecotones has evolved greatly since the beginning of the
1900s when it first appeared in environmental literature as "edge habitats" or "zones
of tension" (Decamps & Naiman 1990). Originally, ecotones were viewed as static
boundaries where dominant ecological systems reached their limit and were notably
important as areas of habitat, cover, and food for wildlife species. Now, they are
widely recognized to have a much richer function (Decamps & Naiman 1990, Risser
1995, Clary & Medin 1999). Ecotones act as a connection between distinct
ecosystems, providing a delicate balance that allows energy and matter to flow in a
manner which supports and links the adjoining areas (Naiman et al. 1988, Gosz
1993, Naiman & Decamps 1997, Febria et al. 2011). These interfaces provide critical
ecosystem services including buffering pollutants, attenuating and transforming
nutrients, and providing refuge for prey and a source area for predators and pests
(Evans 1956, Krause et al. 2011). As a result of their narrow spatial extent and steep
ecological gradient, ecotones offer high species richness and unique genetic diversity
(Risser 1993). When this biodiversity coincides with a high number of endemic or
taxonomically unusual species, ecotones can be considered biodiversity hot spots,
which are often areas of priority for conservation efforts (Myers et al. 2000). Ecotone
boundaries are difficult to define, as their temporal and spatial characteristics are

Chapter 1. Literature Review

2

perpetually transitory. This indistinct framework emphasizes flow across threedimensional boundaries (Pinay et al. 1990, Bretschko 1995, Kerne 2002, Febria et al.
2011). The processes that control energy and material flux are intricately connected
to the biological activities that comprise an ecotone's structure and organization;
therefore, both qualitative and quantitative attributes need to be described in order
to understand an individual ecotone (Evans 1956).
Areas in proximity to waterbodies contain numerous aquatic-terrestrial
interfaces, including riparian zones, hyporheic zones, aquatic marginal wetlands,
and lake littoral zones. These areas are of particular interest due to their sensitivity,
and reactivity, to environmental change and therefore, provide a primary landscape
to observe and understand effects of local and global change (Naiman & Decamps
1990). They are considered indicators of change and can provide early warning signs
for analogous regions (Risser 1993). Terrestrial-aquatic ecotones guide the principles
of the river continuum concept, which does not treat rivers and streams as bounded
systems, but as systems which are integrally linked to downstream, upstream, and
adjacent environmental inputs and properties (Hynes 1974, Vannote et al. 1980,
Naiman et al1998, Bo et al. 2006).
1.1.2 Riparian Zone Ecology
The land that borders a river, the riparian zone, provides an essential
connection to maintain ecological functions in these adjacent and connected habitats
(Ward & Stanford 1995, Ward et al. 1999, Allan & Castillo 2007). Riparian zones are
some of the most diverse, dynamic, and ecologically productive habitats of all
terrestrial ecosystems (Naiman & Decamps 1997, Clary & Medin 1999). Exhibiting
high moisture gradients, contrasting vegetation species, and a mosaic of habitats

Chapter 1. Literature Review

3

along a river corridor, riparian zones often contain high flora and fauna biomass,
especially in arid regions (Clary & Medin 1999). As in all ecotones, riparian zones
perform functions that drive ecological connections, such as filtering pollutants and
trapping suspended particles (Peterjohn and Corre111984) and mediating
biogeochemical transformation (Carlyle and Hi112001).
1.1.3 Hyporheic Zone Ecology
Water in rivers and streams passes between the channel bed surface and
subsurface along discrete flow paths. This interstitial area of mixing, the hyporheic
zone (Figure 1.1), regulates the exchange of nutrients and gases between
groundwater and surface water (Triska et al. 1989, Valett et al. 1993, Fisher 1997,
Malard et al. 2002). The hyporheic zone represents a hydrologic continuum between
river water and groundwater and cannot be precisely separated (Alley et al. 2002,
Plummer et al. 2013). The hyporheic zone provides a medium in which these two
inherently different, but coupled ecosystems can interact.
Many critical functions are provided by the hyporheic zone: nutrient
transformation (Ward et al. 1998, Hancock 2002), pollution attenuation (Wagner &
Bretescko 2002), refuge during low flow events (Grimm et al. 1991, Boulton &
Stanley 1995, Robertson & Wood 2010), habitat for salmonid embryo survival (Geist
& Dauble 1998, Kondolf 2000, Robertson & Wood 2010), and post-disturbance

stream recovery (Valett et al. 1994, Gardner et al. 1999, Hancock 2002). The
hyporheic zone also acts as a transient storage area, where a portion of the water
moving downstream is temporarily retained in regions of stagnant or slow moving
water. Intense biogeochemical transformations occur in storage zones (Packman &
Bencala 2000).

4

Chapter 1. Literature Review
The hyporheic zone relies on surface water to supply crucial resources to

hypogean fauna, such as food, nutrients, and dissolved gases (Wagner & Bretescko
2002). The subsurface environm~nt has habitat limitations (permanent darkness and
space restrictions) that tvpicallv support a loY\ er biodiversity than the stream's
surface water (Gilbert 1990, Ward et al. 1998, llancock 2002). l lypogean fauna are
often rare, niche organisms that act as bio-indicators for groundwater contamination
and facilitate intense biogeochemical activity that supports the entire aquaticterrestrial interface (Boulton & Stanley 1996, llancock 2002).

water table
water table

Figure 1.1. Diagram depicting the location of the hyporheic zone as an interstitial area
between stream water and groundwater. Local geomorphic features, such as the
meander bend pictured here, can form localized flow features withm stream and banks.
(From Alley et al. 2002.)

1.1.4 Interstitial Zone Overlap

Traditionally, streams and rivers were viewed as discrete systems vvithin the
bounds of the surface water and the riverbed that encapsulated this flow. We now

Chapter 1. Literature Review

5

know that lotic environments are some of the most dynamic and complex
ecosystems on this planet. Streams are intricately connected to the land and water
that immediately surround it and, especially, to the interfacing systems that join the
land to the stream and the stream to the groundwater (Figure 1.2; Triska et al. 1993,
Boulton et al. 1998). The hyporheic zone and riparian zone have dynamic, shifting
boundaries that are controlled by surface stream direction and velocity overall.
Investigating the biological, physical, and chemical processes in these overlapping
and inherently difficult to define ecotones produces complicated questions.
Processes, rather than discrete products, tend to be the focus when interpreting
hydrologic and biogeochemical interactions within the hyporheic I riparian interface.

TERRESTRIAL lANDSCAPE

---------------,

SURFACE W ATER

I

~- ------ - --- -----,
I

: HYPORHEIC ZONE

I
I
I
I

RIPARIAN

---~~----------- I

\

\

ZONE
I

"----- ~--------- ...

I

I

I
I

\

'---------------- ~

,

G ROUNDWATER

Figure 1.2. Schematic diagram showing the hydrological ecosystems and ecotones that can
interact with the hyporheic zone and their relative location to the hyporheic zone (adapted
from Boulton et al. 1998).

1.2 Pacific Salmon
Every year, millions of Pacific salmon (Oncorhynchus spp.) migrate hundreds
of kilometers from the Pacific Ocean to their natal spawning grounds (Hobbs &
Wolfe 2008). They gain about 95-99% of their biomass while at sea and deliver

Chapter 1. Literature Review

6

substantial quantities of marine-derived nutrients (MDN) to the nutrient-poor
freshwater systems in which they spawn (Naiman et al. 2002). This annual event
provides an important source of nutrients and energy to enhance fish and aquatic
productivity, as well as terrestrial wildlife and plant species (Bilby et al. 1996,
Willson et al. 1998, Helfield & Naiman 2001, Wipfli et al. 2003, Albers & Petticrew
2012).
Decreasing salmon abundance has been well documented in California,
Idaho, Washington, and Oregon (Nehlsen et al. 1991, Brown et al. 1994). In one
century, the average number of salmon returning to these rivers has declined from
45 million to less than 2 million (Naiman et al. 2002). These declines cannot be
attributed a single cause, but instead are an outcome of many factors, including
reduced ocean productivity, overfishing, freshwater and estuarine habitat loss,
exotic species introduction, and unsustainable hatchery practices (Miller et al. 2014).
Salmon stocks north of the United States-Canada border are also experiencing
declines Qacob et al. 2010, McDaniels et al. 2010, Peterman & Dorner 2011 ).
Decreases in productivity have been rapid and consistent since the 1990s across a
large geographical range, spanning from Washington and British Columbia to
southeast and southern Alaska (Beamish et al. 2012, Peterman & Domer 2012). The
Canadian government-sanctioned Commission of Inquiry into the Decline of
Sockeye Salmon in the Fraser River, the Cohen Commission, has investigated this
issue due to poor returns and low biological productivity. British Columbia's Fraser
River basin, the second most productive sockeye salmon (0. nerka) fishery in the
world, is currently experiencing problems threatening the stability of salmon stocks
(Cohen 2012). Mean summer water temperature has increased by -1.5°C since 1950

Chapter 1. Literature Review

7

and, recently, spawning salmon have encountered record high temperatures that
have been associated with high levels of migration mortality (Martins et al. 2011 ).
Lately, there has been extreme variability in Fraser River fish stocks and dramatic
shifts in migration behavior, which indicates that ecological dynamics in interior
watershed systems are under stress (Hague et al. 2011). This unpredictabi1ity raises
concern because habitat degradation continues due to land-use activities that
negatively alter salmon habitat and prevent their passage upstream (Baldwin et al.
2003, Ficke et al. 2007, Lohse et al. 2008, Jacob et al. 2010). Reduced salmon runs will
result in a smaller MDN pool and may inhibit the ability of river systems to support
current resident and anadromous fish populations (Bilby et al. 1996, Scheuerell et al.
2005). A more complete ecological picture of freshwater salmon habitats will aid
restoration and management of both disappearing and thriving salmon streams.
1.2.1. Marine-Derived Nutrients

MDN are characterized as the nutrients released from salmon throughout
their migration and spawn, such as waste products, eggs, sperm, and decomposing
salmon carcasses (Naiman et al. 2002, Lessard & Merritt 2006). The upstream
delivery of MDN by anadromous salmon goes against ordinary nutrient flow, which
tends to favor downstream delivery (Murata 2003, Mitchell & Lamberti 2005). This
nutrient pulse provides a significant nutrient subsidy to enhance primary and
secondary productivity in oligotrophic, interior systems (Bilby et al. 1996). The postspawn die-off leaves thousands of salmon to decay on the river bed and the riverine
community metabolizes the flesh in several ways (Kline et al. 1997). Biofilms, the
primary benthic producers, quickly remove nutrients from the water column
through uptake and retention (Costerton et al. 1995, Reisinger et al. 2013). Aquatic

Chapter 1. Literature Review

8

macroinvertebrates and resident fish in the river obtain MDN through direct salmon
consumption (Wipfli et al. 1998, 1999, 2003). Terrestrial animals and birds also take
advantage of the salmon surplus by predation and scavenging of carcasses
(Woodward et al. 2005,

bermeyer et al. 2006). Partially eaten fish that have been

carried across the landscape and salmon-enriched waste products from
predators/ scavengers provide MDN to riparian vegetation (Helfield & Naiman
2006, Hocking & Reimchen 2009). Biota that do not directly consume salmon
carcasses or eggs still benefit from the nutrient pulse because MDN are accumulated
through higher trophic levels (Bilby et al. 1996). Salmon runs not only enrich the
local biotic community, they drive a positive feedback loop that increases
survivorship and reproduction to elevate their own populations (Schindler et al.
2003, Wipfli et al. 2003, Koyama et al. 2005).
1.2.2 Horsefly River in the Fraser River Watershed

British Columbia's Fraser River is utilized as a transportation corridor and
spawning grounds for five types of Pacific Salmon. Some of these fish, such as Pink
salmon (0. gorbuscha), will travel a very short distance from the mouth of the Fraser
River to reach their spawning grounds, but the hardiest of salmon, the Chinook
salmon (0. tshawytscha), will migrate over 1000 kilometers to the river's headwaters
near Mount Robson (Benke & Cushing 2005). The entirety of the Fraser River main
channel is undammed, allowing for natural fish passage along much of its' length
(Evenden 2007). This is a rather unprecedented feature for large rivers in populated
regions, and the Fraser River is ranked as the third largest undammed river in North
America. However, this river and its' salmon populations are not exempt from other
damaging environmental challenges that hinder salmon and ecosystem resiliency,

Chapter 1. Literature Review

9

including warming water (Ficke et al. 2007, Jacob et al. 2010, McDaniels et al. 2010,
Hauge et al. 2011, Martins et al. 2011), sea lice (Lepeophtheirus salmonis) (Krkosek et
al. 2007), oceanic conditions (Worm et al. 2006, Thomson & Hourston 2011), and
agricultural and urban run-off (Baldwin et al. 2003, Smith et al. 2007, Lohse et al.
2008).
Sockeye salmon are prized as a valuable commercial fishery, a cultural
necessity, and a superior food source in Canada. Therefore, much of the research on
salmon is focused on monitoring sockeye populations (Cohen 2012). The
International Union for the Conservation of Nature (IUCN) conducted an
endangerment assessment on subpopulations of sockeye salmon in the North
Pacific. This assessment categorized five of eleven Fraser River sockeye salmon
stocks as threatened: one as Critically Endangered, three as Endangered, and one as
Vulnerable. Out of the 22 subpopulations of sockeye on the Fraser River, seven were
classified as "endangered" (IUCN-SSG 2013).
Sockeye salmon return to the Fraser River in relatively predictable 4-year
cycles of abundance and decline, characterized by a dominant year with very large
returns, followed by a sub-dominant year with approximately half of the dominant
year's returns, and then two years of very low returns (Henderson & Graham 1998,
Guill et al. 2012). Since the late 1970s, salmon escapement to the Horsefly River
steadily increased, with a peak occurring in 2001 of over 1.5 million returning
sockeye (Figure 1.3). This cyclical predictability was disrupted in 2009 when what
was supposed to be a peak year had returns similar to a low sub-dominant year,
with only a little over 80,000 returning. The Quesnel subpopulation, which includes
Horsefly River, remains as "least concern" (IUCN-SSG 2013). However, when

Chapter 1. Literature Review

10

performing a three-generation nonparametric escapement calculation (from Geiger &
Zhang 2002), salmon escapement on the Horsefly River has declined 5.3% per year
between 1998-2013, which is judged as a biologically meaningful decline.

2000000

500000

c

~
~

~ 1000000
.D

E

~

500000

0

1960

1970

1990

1980

2000

2010

Year

Figure 1.3. Sockeye salmon populations in the Horsefly River from 1960 to 2013. Note: 2013
data are preliminary and could change. Source: Dept. of Fisheries and Oceans Canada.

1.3 Nitrogen
1.3.1 The Nitrogen Cycle
Nitrogen, occurring in 78% of the Earth's atmosphere, is one of the most
common elements in the natural world. However, nitrogen in its most abundant
form, gaseous N-u is not readily available for biological uptake (Chapin et al. 2011).
In river ecosystems, particularly small streams, primary producers are of greatest
significance in providing energy to consumers (Allan & Castillo 2007). Nitrogenfixing organisms, typically located freely in soil organic matter or symbioticallybonded to plant species, must first alter the state of nitrogen into the bioavailable

Chapter 1. Literature Review

11

forms of (mainly) ammonium (NH 4 ) and nitrate (N03-), also known as dissolved or
inorganic nitrogen (Vitousek et al. 1997, Ashkensas et al. 2004). The process of
nitrification utilizes nitrifying organisms and partially decomposed organic matter
to convert NH 4 ' to N03· (Richey et al. 1984, Williams 2000). Certain environmental
conditions help facilitate this conversion, including an aerobic environment, a soil
pH between 7 and 9, and an approximate soil temperature of 25°C (Chapin et al.
2011). Denitrification, on the other hand, is the reduction of N03· to gaseous forms of
N (NO, N 20, N 2 ) by denitrifying bacteria, which is typically facilitated by anaerobic
environments (Payne 1973, Betlach & Tiedje 1981, Kaushal 2008). Nitrification favors
the oxygen-rich hyporheic zone, whereas the riparian zone has high potential for
denitrification due to its anoxic environment (Casey et al. 2001, Kaushal2008). The
hyporheic I riparian interface may provide a positive feedback loop whereby
nitrogen can continuously cycle between its forms and support an even greater
diversity of organisms and habitat.
1.3.2 Marine-Derived Nitrogen (b 15 N)

There are two stable isotopes of nitrogen, b14 N and b15 N, which differ from an
ordinary nitrogen atom by the number of neutrons contained in the nuclei. b 14N is
significantly more abundant, comprising 99.634% of the available forms of nitrogen
isotopes Ounk & Svec 1958). However, the marine nitrogen pool has a high ratio of
b15N compared to freshwater ecosystems. The b15 N composition of a spawning
Pacific salmon ranges from 10 to 15%o (Welch and Parsons 1993), while other
available sources of nitrogen in rivers and streams tend to be closer to the O%o b15N
composition of atmospheric nitrogen (Kendall1998). The isotopic disparities

Chapter 1. Literature Review

12

between these nutrient pools offer a technique for measuring the assimilation of
MDN by freshwater organisms, where isotopic enrichment corresponds to increased
MDN assimilation (Kline et al. 1990, Bilby et al. 1996, Chaloner et al. 2002, Hicks et
al. 2005, Scheuereii et al. 2007). The trophic position of a consumer is indicated by an
increasingly positive enrichment value on the trophic food chain (Minigawa &
Wada 1984). This occurs because b 15N is not as readily excreted as a nitrogen waste
product (Robinson 2001) and accumulates within the organism.
During the process of nitrogen-fixation, many soil microbes discriminate
against using marine-derived nitrogen (b 15N), as this has a stronger nitrogen-oxygen
bond than terrestrial nitrogen (b 14N) (Stevenson & Cole 1999). Consequently, the
products of activities like decomposition, nitrification, and denitrification are less
enriched with b15N, leaving the original substance with higher enrichment (Templer
et al. 2007).
Studies have shown that when salmon enter streams, NH/ is the initial
product of salmon carcass decay and surface water concentration increases rapidly
(O'Keefe & Edwards 2002, Drake et al. 2006, Pinay et al. 2009, Kiernan et al. 2010).
Ammoniu m values will be u sed as a substitute for stable isotopic concentrations
when analyzing su rface and hyporheic water samples.
1.4 Carbon
1.4.1. Th e Carbon Cycle

Some 3.5 billion years ago in the Precambrian period, cyanobacteria began to
bind with sediments forming stromatolites, the most ancient records of life on Earth.
Stromatolites are the pioneers of oxygenic photosynthesis, which radically

Chapter 1. Literature Review

13

transformed the atmosphere into an oxic environment and initiated the dominance
of oxygen-requiring organisms (Des Marais 2000, Kopp et al. 2005). Organic carbon,
the other byproduct of photosynthesis, is the structural building material for all of
life, as we know it, and flows freely through the biosphere as decomposable organic
matter (Pace 2001). When plants and animals die, shed, or excrete waste, the organic
carbon material undergoes a microbial-mediated transformation to inorganic
carbon, primarily carbon dioxide (C0 2). C02 is transported through interstices, or
the gaps between matter (i.e. soil, surface water, groundwater, and air), until it is
picked-up and utilized by another living organism (Fry 2006). Decomposition in a
biotic community is incredibly efficient; a long-term study of mineral soils found
relatively minor changes in total organic C content 40-years after reforestation
(Billings & Richter 2006).
Photosynthesizing cyanobacteria and stromatolites created local oxygen-rich
environments, which facilitated the colonization of other microorganisms (Kopp et
al. 2005). These highly productive and energetic colonies of autotrophic and
heterotrophic organisms, called biofilms, remain as primary producers across almost
every ecosystem. In a salmon-bearing stream, a substantial amount of organic
carbon is delivered by the annual inland salmon migration. Microbial biofilms
mediate marine-derived nutrient assimilation from the water column into aquatic
and terrestrial food webs (Cummins 1974, Hoellein et al. 2007).
1.4.2. Marine-Derived Carbon {l)13 C)

As with marine-derived nitrogen, marine carbon isotopes are distinctly
enriched compared to freshwater or terrestrial values. Marine b 13C signatures are
enriched relative to the atmosphere, whereas freshwater b 13C signatures are derived

Chapter 1. Literature Review

14

from sources of dissolved carbon dioxide in the water (Fry 2006). The stable isotopic
ratio of 13C/ 12C allows for the identification of important primary food sources, as
these stable isotopes pass through the food chain relatively unch anged (Doucett et
al. 1996). Salmon predominately possess a marine carbon signature because their life
strategy shifts from feeding to reproduction as they enter freshwater. Animal tissues
reflect the b13C of their assimilated food within about 1roo. In isotopic terms, the
expression "You are what you eat" is quite accurate and provides the basis for
studying trophic food chains (DeNiro & Epstein 1978, Rounick & Winterboume
1986, Michener & Schell1994, Fry 2006).

1.5 Stream-Riparian Ecosystem Response
The input of marine-derived nutrients to salmon-bearing freshwater
ecosystems can be seen throughout all trophic levels of a food web. Major
freshwater and terrestrial components of these food webs are outlined in this section
(Figure 1.4).
1.5.1 Aquatic

1.5.1.1 Biofilms
In the presence of spawning salmon, biofilms become enriched with marine
isotopes (Kline et al. 1990, Holtgrieve et al. 2010). Since they are primary producers,
utilization and trapping of MDN at this level will aid increased accumulation in
species occupying higher tropic levels (Lamberti 1996, Wipfli et al. 1998). Their
growth rate varies between streams due to specific environmental conditions, such
as temperature and timber harvest, and salmon legacy effects (Rex & Petticrew 2008,

Chapter 1. Literature Review

15

Ruegg et al. 2011, Reisinger et al. 2013). Albers & Petticrew (2012) found that biofilm
density decreased during the active-spawn period due to spawning disturbance,
whereas during the post-spawn period, biofilm abundance increased to above prespawn values and utilization of marine-derived nutrients was observed.

1.5.1.2 Macroinvertebrates
Consumption of salmon carcasses and eggs by aquatic macroinvertebrates
represents an important bottom-up uptake of MDN into the food web (Lessard &
Merritt 2006, Kiernan et al. 2010). Current research has found that aquatic
invertebrates respond both positively and negatively to the MDN pulse. Reduced
invertebrate densities and altered community structure in streams with actively
spawning salmon have been observed due to disturbances from particularly high
salmon densities and redd construction (Peterson & Foote 2000, Minakawa & Gara
2003, Moore & Schindler 2008, Honea & Gara 2009). However, other studies have
discovered increased macroinvertebrate density due to MDN (Bilby et al. 1996,
Claeson et al. 2006}. Wipfli et al. (1998) documented an increase in invertebrate
densities by 8 to 25 times when adding salmon carcasses to an artificial flume.
Lessard & Merritt (2006) discovered that only certain taxa responded positively to
MDN enrichment, particularly chironomid midges and shredders.
Macroinvertebrates encompass a diverse community of feeding groups and Jife
cycle structures. The timing of spawning season within an insect's life cycle, as well
as salmon nutrient legacy, will contribute to species and location-specific enrichment
(Honea & Gara 2009, Guyette et al. 2014).

1.5.1.3 Fish

Chapter 1. Literature Review

16

Marine stable isotope enrichment is evident in anadromous and resident
fishes (Bilby et al. 1996, Wipfli et al. 2003). Increased growth rates of cutthroat trout
(0. clarki), Dolly Varden (Salvelinus malma), and anadromous salmon (Oncorhynchus

spp.) suggest a transfer of nutrients and energy from decaying salmon to subsequent
fish populations (Wipfli et al. 2003). Salmon carcasses, eggs, and fry provide a rich
food source for fish populations (Schindler et al. 2003), and also initiate a positive
feed back effect that benefits the next generation of salmon by increasing primary
food resources and juvenile salmon growth rate and energy density (Wipfli et al.
1998, Adkison 2010, Rinella et al. 2012).
1.5.2 Terrestrial

1.5.2.1 Riparian Vegetation
Several studies have recognized that b15N and b13C are incorporated into terrestrial
vegetation adjacent to salmon spawning streams (Bilby et al. 1996, Ben-David et al.
1998, Bilby et al. 2003, Reimchen et al. 2003, Hocking & Reimchem 2009). On
Chichagof Island in southeast Alaska, the trees and shrubs surrounding salmon
streams derive -22-24% of foliar nitrogen (N) from nitrogen stable isotopes (Helfield
& Naiman 2001). MDN enrichment of vegetation has been observed to extend

throughou t a 200-m rip arian zone (Ben-David et al. 1998) and at a distance of 800-m
from the stream, p resumably due to the displacement of salmon by bears
(Hilderbrand et al. 1999a). An established, mature riparian forest is able to retain
nutrients better than a small, young forest (Milner et al. 2008, DeVries 2012). Once
MDN has been established in a population of trees, they may preserve MDN within
the ecosystem for 30 or more years following declines in salmon population

Chapter 1. Literature Review

17

(Koyama et al. 2005). Evidence exists that MDN are associated with increased
growth rates of riparian vegetation (Gende et al. 2002). In southeast Alaska, the
growth rate of Sitka spruce within 25-m of spawning streams were found to be triple
that of trees in reference sites (Helfield & Naiman 2001).
The fertilization of riparian vegetation by spawning salmon initiates a positive
feedback loop, in which vegetation growth and production benefits from the influx
of salmon nutrients, and salmon spawning and rearing habitat is improved because
of streamside vegetation (Helfield & Naiman 2001). The shading of streams by
vegetation helps to moderate in-stream temperature, which influences
developmental rates of young salmon by directing optimal timing of life history
events and controlling embryo development. Riparian root systems help to stabilize
banks and filter sediments to minimize siltation and erosion, thus providing well
oxygenated redd conditions that are ideal for salmon production (Nakano &
Murakami 2001). Terrestrial inputs of organic matter support the production of
aquatic insects, which are a basic source of food for juvenile salmon (Helfield &
Naiman 2001). Large woody debris, which accumulates in streams with riparian
vegetation, plays an essential role in providing rearing habitat for salmonids (Fausch
& Northcote 1992). Large wood also enhances hyporheic exchange (Hester &

Gooseff 2010), which may influence the riparian-hyporheic interactions that are
being investigated in this study.

1.5.2.2 Terrestrial Animals
Many studies have investigated the effect of salmon abundance in stream
ecosystems on black and brown bear diet (Hilderbrand et al. 1999b, Weaver &
Zammuto 2004, Helfield & Naiman 2006). Nitrogen stable isotope analysis from hair

Chapter 1. Literature Review

18

samples of grizzly bears in the Columbia River basin and from coastal brown bears
in Alaska have documented, respectively, 33% to more than 90% of N in their diet
from salmon-derived nutrients (Naiman et al. 2002). Grizzly bears living in areas
with easy access to salmon streams also exhibit greater productivity and population
density (Darimont et al. 2010).
Although bears are the prevailing animals associated with salmon
consumption, many other terrestrial animals prey on spawning salmon and
decaying carcasses. On the Olympic Peninsula of Washington, 14 mammals,
primarily black bears, raccoons, and otters, were found consuming salmon carcasses.
In the same study, eight species of birds, varying from winter wrens to bald eagles,
were noted scavenging decaying salmon (Cedarholm et al. 1989). American dippers

(Cinclus mexicanus) nesting along southeast Alaska streams have higher reproductive
performance associated with their diet of salmonids (Obermeyer et al. 2006). It is
feasible that all terrestrial animals feeding on salmon carcasses increase MDN
enrichment around a spawning stream by physically moving salmon across the
riparian and through excretion of waste products.

19

Chapter 1. Literature Revie·w

H porheic exchange
tr r.., , ..

•

•

Groundwater flow

H

•
l IC

/one

..
Figure 1.4. MDN transfer, assimilation, and trophic transport in a stream-riparian
ecosystem is complex in nature and influential in several interacting ecosystems.

1.6 Research Objectives and Hypotheses
This study aims to describe nutrient delivery, seasonal storage and trophic
transfer within a stream-riparian ecotone. Marine-derived nutrients transported by
sockeye salmon during the 2011 salmon run will be analyzed to determine if these
nutrients are being utilized by stream macroinvertebrates and riparian vegetation,
and to determine if MDN are delivered by and stored in the hyporheic zone.
Nutrients can be removed from the water column quickly from 1nultiple trophic
levels (Naiman et al. 2002), which is why several sources will be used to quantify
nutrient storage. These observations will be rnade before and during the salmon

Chapter 1. Literature Review

20

spawn, during and after the salmon die-off, and the following spring. The Horsefly
River spawning channel is a representative sample of other similar systems,
particularly spawning channels in interior British Columbia.
1.6.1 Delivery and Storage of Ammonium in the Hyporheic Zone
The aim of the research described in Chapter 2 was to determine if
ammonium (NH4 ), presumably from spawning salmon, was being delivered to,
transported through and/ or stored within the hyporheic zone. MDN transport to
the riparian zone through nutrient deposition by bears has been well documented
(Hilderbrand et al. 1999, Helfeld & Naiman 2006), however, the delivery of salmon
nutrients via the hyporheic zone and the potential for overwinter storage may also
be a critical pathway for MDN to transfer to riparian vegetation (O'Keefe & Edwards
2002, Pinay et al. 2009). The following hypotheses were developed to evaluate the
research aim described above.

Null Hypothesis I: NH4+concentration of surface water will not be correlated with
the number of spawning salmon in the HFC.

Null Hypothesis II: There will be no difference between NH4+concentrations of
surface and hyporheic water.

Null Hypothesis III: NH/ concentration of hyporheic water will not be correlated
with distance from the streambank.

Null Hypothesis IV: There will be no difference between the mean NH 4+
concentration of surface and hyporheic water between fall and
spnng.
These hypotheses help to inform the following questions about the system:

Question One- What is the extent of the hyporheic zone at this site?
Question Two -Is NH/ being retained in the hyporheic zone during the course of
the salmon spawn?

Chapter 1. Literature Review

21

Question Three - Is NH4 stored in the hyporheic zone over the winter?
1

Question Four- Does NH 4 • concentration change as hyporheic water moves
further from the stream?
1.6.2 Storage and Trophic Transport of MDN in a the Stream-Riparian Ecosystem
The aim of the research presented in Chapter 3 was to investigate MDN
assimilation, trophic transport, and storage in a stream-riparian food web by tracing
b13C and b15N in stream macroinvertebrates and riparian vegetation and soil during
the 2011 sockeye salmon spawn and subsequent spring. In British Columbia, salmon
spawn occurs in the fall when many plants are senescing or becoming dormant.
Despite an apparent contrast in run timing and plant growth, biological uptake of
MDN can occur quickly and be stored within riparian soil and plants (Drake et al.
2006) and stream organisms (Bilby et al. 1996 ). There are still questions, however,
especially in regard to the environmental factors that limit uptake and persistence of
MDN in individual species and in the riparian forest as a whole (Drake et al. 2011,
Devries 2012, Rinella et al. 2013).
For this study, macroinvertebrates were evaluated according to their
functional feeding group (FFG) to determine if species type, abundance, and stable
isotopic concentration changes during the MDN pulse and between seasons. Three
riparian plant species with differing life strategies were also investigated to
determine the pathway, uptake, and retention of MDN in the riparian zone adjacent
to the salmon spawning channel. The following hypotheses were developed to
evaluate the aim listed above:

Null Hypothesis V: There will be no difference between stable isotopic signatures
(b 13C and b15N) of macroinvertebrate taxa throughout salmon
spawn and between fall and spring.

Chapter 1. Literature Review

22

Null Hypothesis VI: Macroinvertebrate b 15 N signatures will not be correlated with
NH4 concentration of surface and hyporheic water.

Null Hypothesis VII: There will be no difference between stable isotopic signatures
(b 13C and b 15 N) of riparian vegetation between seasons.

Null Hypothesis VIII: Stable isotopic signatures {b 13C and b 15N) of riparian
vegetation and soil will not be correlated with distance from the
streambank.

Null Hypothesis IX: Riparian vegetation b 15N signatures will not be correlated with
NH4 concentration of riparian hyporheic water.
These hypothe es help supply information about the following research
questions:

Question One- Is the MDN signal retained by macroinvertebrates and riparian
vegetation?

Question Two - Do macroinvertebrate functional feeding groups show temporal
differences to the MDN pulse?

Question Three - Does macroinvertebrate species type and abundance change
throughout the salmon spawn and between fall and spring?

Question Four- Does rooting depth of vegetation have an effect on MDN
concentration?

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

23

Chapter 2
Temporal and patial

spect of N itrogen Cycling in the Hyporheic Zone

2.1 Hyporheic Zone as Vector for N utrient Transport
The upstream delivery of marine-derived nutrients (MDN) by salmon to their
freshwater spawning grounds is evident within the water, organisms, and materials
of the surrounding watershed (Allan & Castillo 2007). MDN signatures are found
within terrestrial vegetation bordering salmon streams, as well as within nearly any
animal that permanently or temporally occupies that area (Kline et al. 1990, Naiman
et al. 2002). For example, MDN transport from the streambed to the riparian zone
through MDN nutrient deposition by bears has been well documented (Hilderbrand
et al. 1999, Helfield & Naiman 2006). The transfer of nutrients through the hyporheic
zone has also been investigated as a critical pathway to promote this nutrient
exchange (O'Keefe & Edwards 2002, Pinay et al. 2009). The hyporheic zone acts as a
transient storage area by transporting and transforming essential nutrients. The
rooting zone of riparian vegetation also can extend into hyporheic flow paths
(Helfield & Naiman 2006) potentially explaining the MDN signal observed in this
vegetation (Bilby et al. 1996, Ben-David et al. 1998, Reimchen et al. 2002, Bilby et al.
2003).

2.1.1 Defining the Hyporheic Zone
Water in rivers and streams passes between the channel bed surface and
subsurface along discrete flow paths. Grou nd water and surface water exchange is
regulated in an interstitial mixing area known as the hyporheic zone. Rapid and
frequent hydrologic exchange occurs allowing for the renewal of nutrients and

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

24

oxygen (Hancock 2002). Often the extent is only a few meters beneath the
streambed, but the hyporheic zone can be considerably large, extending several
kilometers inland (Stanford & Ward 1988). The water table, or the boundary of
saturated / unsaturated soils, is not easily delineated because it is spatially and
temporally dynamic (Triska et al. 1989, Vervier et al. 1992, Boulton et al. 1998). The
hyporheic zone is now thought to be an essential component of stream function,
with the definition of streams expanding to not just what is above the streambed,
but integrally related to what is below (Fisher 1997, Malard et al. 2002).
The hyporheic zone can be delineated using three primary componentsbiological, geochemical, and hydrologic.

2.1.1.1 Biological
The biological method for defining the hyporheic zone is the presence of
organisms living in this region of mixed waters. Observations of stream organisms
deep within the streambed gravels were used for preliminary studies in identifying
the hyporheic zone (Coleman & Hynes 1970). The idea of organisms living below the
streambed was criticized heavily at first, but continued discoveries of
macroinvertebrates beneath the streambed confirmed its existence (Hynes 1974).
When Stanford & Ward (1988) identified riverine invertebrates 2-km from the main
channel of the Flathead River in Montana, the extent of the hyporheic zone
expanded to inclu de subsurface areas adjacent to rivers and across floodplains. The
main objectives of hyporheic biological methods are to understand species type and
habitat and to identify if species are either epigean (periodic users of the hyporheic)
or hypogean (permanent residents of the hyporheic). When defining the hyporheic
zone in this way, it is important to recognize that taxa composition and abundance

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

25

can change between seasons, meaning that one-time sampling does not necessarily
produce a representative sample (Hancock & Boulton 2009).

2.1.1.2 Geochemical
Geochemical methods define the hyporheic zone as a volume of sediment that
contains a certain proportion of surface water. This volume of water varies between
stream reaches and microhabitats, and is defined by Triska et al. (1989) as being
anywhere from greater than 10% to less than 98% channel water. For example, the
hyporheic zone may be quite expansive in unconfined floodplains or may be small
due to confining hillslopes around a river. Additionally, stream microhabitats can
control hyporheic exchange, with some areas, such as riffles with high stream
discharge, reducing the proportion of surface water infiltrating the hyporheic zone
(Tonina & Buffington 2009). Many different geochemical methods can be used to
identify hyporheic flow paths including the use of natural tracers (i.e. temperature,
pH, and electrical conductivity) or injection of conservative tracers (e.g. NaCl or
dye) (Malard et al. 2002).

2.1.1.3 Hydrologic
Hydrologic methods define the hyporheic zone by using groundwater flow
models to describe the volume and magnitude of water moving through a medium.
Physical parameters of groundwater level and soil characteristics are needed to
describe hyporheic flow (McDonald & Harbaugh 1988, Wondzell 2006 ). Water levels
are measured at a point-in-time, either occasionally using manual methods of
measurement or continuously using a water level data logger, and typically describe
hyporheic flow over a period of time (Dearden & Palumbo-Roe 2010).

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

26

For this study, geochemical methods were used to find hyporheic locations
influenced by surface water, and hydrologic variables of groundwater level,
hydraulic conductivity, and soil porosity were used to define hyporheic flow.
2.1.2 Physical Characteristics of Hyporheic Flow

2.1 .2.1. Piezometers and Wells
Piezometers and monitoring wells are used to access water beneath the
surface. They can be comprised of a variety of materials (e.g. plastic tubing, PVC,
aluminum) and span a range of lengths and widths, depending on the application.
Wells are perforated along the length of the pipe to capture the height of the water
table. Piezometers are perforated only around the lower section of the pipe, which
measures pore water pressure at the intake location. For this study, piezometers
were used and functioned as perforated pipes fixed vertically in the ground to
passively intercept groundwater (Sprecher 2000). Their function is to facilitate
groundwater level measurements, water sampling, and tracer introductions (Tonina
& Buffington 2009).

2.1.2.2 Hydraulic Head and Hydraulic Conductivity

Groundwater is always moving from high to low potential energy; therefore
the distribution of hydraulic head throughout an area indicates where groundwater
will flow. The hydraulic head at a particular location represents the total potential
energy of subsurface water and is measured in units of length above a fixed datum
Oones & Holmes 1996). The relative rate of movement is described by the vertical
hydraulic gradient (VHG, m · m·1 ) and horizontal hydraulic gradient (HHG, m · m·1 )
between two piezometers. Both measurements can be calculated with the following

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

27

equation:
VHG or HHG- till/ ~I

Eqn (2.1)

where:
till = difference in height between two hydraulic heads; and,
M =flow path length between two piezometers (vertical distance for VHG and
horizontal distance for HHG).

Hydraulic conductivity (K) describes the ease with which groundwater
moves through pore spaces in the alluvium or soil, and is a measurement of distance
traveled over time (with units L · T-1). The obtained values depend on sediment
porosity and connectivity of these pore spaces, which are functions of the type and
condition of sediments (Tonina & Buffington 2009). The simplest method to obtain
hydraulic conductivity values is through a bail or slug test (Freeze & Cherry 1979).
Water is either introduced to or removed from a piezometer and the water level is
then monitored until it returns to quasi-equilibrium conditions. Many methods have
been developed to calculate hydraulic conductivity for specific subsurface
conditions; this study will use the Hvorslev Method which calculates the time it
takes for the water level to rise to 37% of the initial maximum (Hvorslev 1951,
Campbell et al. 1990). The equation is as follows:

K =rLn(L/R) I [2LTo]

Eqn (2.2)

where:
K = hydraulic conductivity;
r = radius of well casing;
R = radius of well screen;
L = length of well screen; and,
To= time require for the water level to rise to 37% of the initial change.
Within the channel, VHG measurements are used to determine if water is

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

28

moving into the ground from the river (downwelling or groundwater recharge) or
from the ground to the surface water (upwelling or groundwater discharge). These
conditions may change temporally, such as during high water events where
groundwater is recharged due to the downwelling of excess surface water (Mouw et
al. 2009) or conditions may remain steady where upwelling water is constantly
delivering limiting nutrients to the surface water and producing localized zones of
productivity {Valett et al. 1994). By observing hyporheic and surface water data,
areas of groundwater discharge and recharge can be identified. These "charged"
zones exhibit particularly active hyporheic exchange that may alter water chemistry
and redirect groundwater flow patterns Oanzen 2008).
Typically, a patchy matrix of bidirectional water exchange exists at the reach
and segment scale and creates a variety of habitats for a diverse assemblage of
organisms. Site specific characteristics, such as soil type, water inflow, and intruding
vegetation, contribute to this variable matrix (Dent et al. 2001, Poole et al. 2008,
Tonina & Buffington 2009). Localized zones around tree roots, for example, can
cause disparate conditions that depend on specific spatial and temporal variables. In
some locations, tree roots may cause upwelling by drawing water and nutrients
from the hyporheic zone (Constantz et al. 1994). However, exactly the opposite
process commonly occurs when plant roots transpire water directly from
groundwater, resulting in a drawdown and localized depression of the water table
(Winter 2000). Changes in the local water table could be a likely cause of this
disparity.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

29

2.1.2.3 Residence Time
The hyporheic zone acts as a storage site, which traps water and solutes and
then releases them over time. Water may exchange between the stream and
hyporheic zone in short increments many times along its flow path, transferring
heat, oxygen, solutes, and biota. Increased water residence time occurs when stream
water enters deeper into the alluvium or flows across riparian or floodplain soils
before entering the stream again (Harvey & Wagner 2000, Boa no et al. 2006, Hester
& Gooseff 2010). Hyporheic exchange can regulate the biogeochemical

transformations or release of essential nutrients to a stream, such as inorganic
nitrogen. Where residence time is short, nitrogen is exported downstream at a
greater rate. Increased residence time facilitates internal denitrification which
removes inorganic nitrogen from the system (Triska et al. 1989, Wondzell &
Swanson 1996b).
Zarnetske et al. (2011) performed a hyporheic residence time experiment to
investigate the spatial and temporal physiochemical conditions that control nutrient
transformation in a drainage dominated by agriculture and forestry. They
discovered that short water residence times (<6.9 hours) were associated with
biogeochemical processes of ammonification, nitrification, and rapid utilization of
DO and DOC. Residence times beyond 6.9 hours resulted in denitrification and net
removal of nitrate from the stream. Although every hyporheic location studied will
yield different outcomes, shorter residence times prompting aerobic metabolic
process and longer residence times facilitating anaerobic metabolic processes is a
commonly agreed upon response (Triska et al. 1993, Peyrard et al. 2011, Binley et al.
2013).

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

30

2.1.3. Modeling Hyporheic Flow
The primary methods used to trace water flow paths in the hyporheic zone
include water temperature differences, tracer tests, and groundwater level loggers.
These methods can be used alone or in combination and may be incorporated into
predictive hydrologic models. The USGS program MODPA THis commonly used to
simulate particle tracking in groundwater transport and to calculate water residence
time (Pollock 1994). The mapping and spatial analysis software ArcGIS has
groundwater tools for advection-dispersion modeling of groundwater flow, particle
tracking, and chemical dispersion (ESRI 2013). For this study, ArcGIS groundwater
software was used to model hyporheic flow.
A geographic information system (GIS) groundwater model was used in this
study because of its distinct spatial and visual advantages. By utilizing technology in
ArcGIS, data management and tools for modeling are integrated into a single
predictive working environment, from which results of a groundwater flow model
can be seen directly in a presentable output. The ArcGIS groundwater analysis tool
Darcy Flow models two-dimensional flow by utilizing variables of groundwater
head elevation, porosity, saturated thickness, and transmissivity (defined partially
by hydraulic conductivity). Local flow velocity and direction are defined for each
cell in the grid, and Darcy's Law is applied to the cells adjacent to the center cell in
order to generate a flow field (Tauxe 1994). Water that flows through the ground is
governed by Darcy's Law, which states that flux rates are a direct function of
hydraulic conductivity and hydraulic gradient (Freeze & Cherry 1979). A precursor
to Darcy Flow is the Flow Direction function, which confines flow direction towards
one of eight adjacent cells Oenson & Domingue 1988).

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

31

2.1.4 H potheses

The aim of this study is to determine if the hyporheic zone acts as a delivery
corridor and storage zone for ammonium (NH4 ) that is delivered to the stream by
spawning salmon. The concentration of NH4 between stream and hyporheic water
samples will be evaluated, as well as the spatial variables that control nutrient
exchange. Over-winter storage of NH4 ' will be assessed by sampling throughout the
2011 sockeye salmon spawn and the subsequent spring. The following hypotheses
will be tested:

Null Hypothesis I:

H 4' concentration of surface water will not be correlated with
the number of spawning salmon in the HFC.

Null Hypothesis II: There will be no difference between NH 4+ concentrations of
surface and hyporheic water.

Null Hypothesis III: NH4+concen tration of hyporheic water will not be correlated
with distance from the streambank.

Null Hypothesis IV: There will be no difference between the mean NH 4+
concentration of surface and hyporheic water between fall and
spnng.
2.2 Method s

2.2.1 Stu dy Site
Fieldwork was performed in Horsefly, BC (52° 19'N, 121 o 1'W) located in the
Cariboo region of British Columbia (Figure 2.1). The 131-km long H orsefly River is
the largest tributary of Quesnel Lake in the Fraser Basin of the Quesnel watershed.
This river provides im portant spawning grounds for sockeye salmon (Oncorhynchus

nerka) and, to a lesser d egree, Ch inook salmon (0. tshawytscha). Historically, the
Horsefly River has supported large sockeye salmon p opulations; however, recent

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

32

salmon spawning cycles have produced lower than average numbers (Albers 2010).

0 Horsefly, BC

lob

-,..,

BRITISH COLUMBIA

Figure 2.1. Map of British Columbia displaying the town of Horsefly, the location where
this study was conducted.

2.2.1.1 Horsefly River Spawning Channel
The Horsefly River spawning channel (HFC) is an artificial channel managed
by the Department of Fisheries and Oceans Canada (DFO) that was designed and
constructed as part of a salmon enhancement program to improve salmon
production and facilitate fisheries management by creating an optimal spawning
environment for salmon (Figure 2.2). The channel is approximately 1.6 km in length
and 10m wide with a designed slope of 1%. Maximum salmon capacity in the
channel is 23,000 adults (Holmes 2008). The approximately 1-m deep gravel layer
lining the ch annel surface is relatively uniform in particle size and shape, and was
designed for ideal spawning conditions. Additionally, the alluvium and soil

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

33

surrounding this gravel layer may have an irregular development due to channel
engineering. Water inflow is regulated by a large pump that siphons water from a
settling pond, which is connected to the Horsefly River at the upstream end of the
channel. At the start of the salmon run, a fence is installed in the Horsefly River
directly upstream of the channel mouth to restrict upstream migration and allow a
selected number of salmon to enter the HFC. The confined salmon then spawn in the
channel. There are two salmon gates within the channel that restrict movement but
also become temporary dams once the salmon spawn-out (Figure 2.1). Carcasses
become lodged against the gates, increasing the water level and creating a reservoir
that is released mid-day when the dead salmon are sampled, marked, and released
into the main channel.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

34

Figure 2.2. Aerial view of the Horsefly River spawning channel in Horsefly, BC. The
Horsefly River is depicted with flow direction arrows.

In the summer and fall of 2011, in addition to regular DFO enhancement
activities, a 900-m2 area of the channel and adjacent riparian zone was converted
into an experimental site. The average wetted width of the experimental channel in
this reach is 12.3 m, with bankfull width being only 0.1 to 0.2 m wider due to the
incised channel. The maximum water height at the channel's riparian edge is 1.4 m.
The average of three measurements were taken prior to sampling events to define
channel width and height using methods outlined by Leopold (1994). Vegetation
composition resembles other riparian zones along the Horsefly River, consisting

Chapter 2. Nitrogen Cycling in tire Hyporheic Zone

35

primarily of deciduous trees with a diameter at breast height (DBH) of less than 0.3
m. Predominant tree species include willows (Sallx sp.) and Black cottonwood

(Populus triclzocarpa), and dominant shrub species are Red-osier dogwood (Corn us
sericea) and Twin-berry (LonZtera mvolzurat ). Oth12r prevalent vegetation includes
Mountain alder (Alnus ten wfi.>lla ), Saskatoons (Am. la1ulzz r alnifolza ), Cow parsnip

(Heradeunz maxmzum ), and Pnckh rose (Rosa aL n ularzs ). The riparian zone upstream
of the study site is composed mainly of smaller shrubs and grasses, with a dense
riparian forest forming at the beginning of the meandt:)r bend (Figure 2.3). The
reason for this difference in forest maturitv is unknown, but is likely due to tree
clearing during the con~truction of th• · channel.

Figu re 2.3. The study site in June 2011, showing meander bend in spawning channel and
adjacent riparian zone. Piezometers F and Care located upstream (right side in this
photo) of the riparian forest and are located in the un-forested section.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

36

Before the introduction of salmon, the top 30 em of the channel bed gravels
were mechanically cleaned to resuspend stored fine sediment and, subsequently,
increase the flow of dissolved oxygen (DO) within the streambed. This was
undertaken by the DFO and was accomplished using a rake mounted on a
bulldozer, followed by a managed increase in water flow to flush out suspended
solids. A temporary sandbag dam was constructed about 200m upstream from the
experimental site where mechanical cleaning ceased and the turbid water was
pumped from the channel into an isolated settling pond adjacent to the channel. This
was done in order to minimize the impact of fine sediment resuspension in the
Horsefly River preceding the salmon run. Due to the location of this downstream
settling pond, the final250 m of the channel is not cleaned. The experimental site is
located within this section where the streambed has not been disturbed (Figure 2.2).
Although the HFC was created to optimize salmon spawning habitat, the
artificial construction and environmental manipulation of the channel results in
underlying differences when compared to natural systems on the Horsefly River
and the broader Pacific West Coast. Notably, the water in the channel is pumped
from a settling pond. Although water in the settling pond originates in the Horsefly
River, chemical and biological attributes of this water will be altered from being held
in an environment that is more stagnant than the free-flowing river. Some of these
changes may include reduced DO, increased dissolved nutrient concentration, and
an altered aquatic plant and animal community. Channel bed cleaning prior to the
introduction of salmon is an obvious, unnatural disturbance, which likely results in
a mass export of macroinvertebrates, biofilms, and dissolved nutrients. While the
cleaning promotes the flow of DO to benefit salmon eggs, it may unintentionally

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

37

alter the ecology of the spawning habitat in undesirable ways (e.g. decrease food
reserves for salmon fry). Also, the Horsefly River is characterized by annual spring
flooding events that create large floodplains throughout the region. The HFC is
protected from floods, which does create a safe haven for eggs and juvenile salmon
(Holmes 2008). However, it results in a geomorphology that is different from the
sandy substrate and wide, meandering gravel bars created by the floods.
2.2.2 Study D esign
The experimental site was chosen at a meander bend 50 m from the mouth of
the channel (Figure 2.4). Twenty-five piezometers were installed in a grid,
comprising five instream piezometers and 20 piezometers located on the riparian
zone adjacent to the channel. Each piezometer was spaced 1 to 2m apart, with
variability due to physical barriers, such as trees and boulders. Piezometers were
constructed of l-in (2.54-cm) inner diameter electrical metallic tubing (EMT) and
were 10ft (3.048 m) in length. Forty-four, lA-inch diameter holes were drilled near
the bottom of the pipe to allow for passive hyporheic water inflow. A pointed, steel
tip was welded to the bottom in order to drive them into the ground (Figure 2.5).
Riparian piezometers were all installed to approximately the same depth, 2.4 m
below the ground surface. Instream piezometers were installed approximately 1 to
1.5 m below the streambed .

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

38

N

A
E

•
D

0

0

Soil_a 0

0

•

c

0

K

0

Transect 2

Soil

i

G

0

10 m

•

•

0

20 m

-R

A

F

Transect 2

J

0

0

0

I

Transect 2_0 m

0

0

v

w

OTransect 1_0 m
y

Transect 1

0

10 m

-

•
•

Soil_ 2

0

Transect 1_20 m

Steep upslope

)-

0 Hyporheic water sampling points

e Add'J tracer test sampling points
0

e

Vegetation samples
Soil Pits

0

5

10 Meters

Figure 2.4. Map of the experimental study site on the Horsefl) Rher spa\\ ning channel.
Contours display 0.1-m changes in land elevation (NAD83).

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
II·

39

~t'r

.~· 1 >-" I

LnruJ

I ahon (mASL)

•
•

•
•
•

•

Hydraulic head
L-0 ~6 m

•

•
~-m

-in

.. ,

{ =
Ele~at1on head

Figure 2.5. Diagram of piezometer marking levels of elevation used in the study.
Measurements of piezometer holes are displayed in the inset diagram. The variable
length of well screen (L) is described in the Hvorslov equation.

2.2.2.1 Rhodamine Tracer Test

In order to select a subset of hyporheic sampling locations for this study, a

tracer experiment using a slug injection of a fluorometric dye tracer, RhodamineWT, was executed. Subsequent samples were collected to track the breakthrough
curves of fluorescence concentration versus time throughout the study site. This
fluorescent dye was chosen for its excellent detectability, use in low concentrations,
and low potential for sorption (Harvey & Wagner 2000). Using methods adapted

40

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

from Kilpatrick & Cobb (1985), the volume of Rhodamine-WT required for the tracer
injection was determined by the following equation:
Vs = (3.79x10-5 ) * (QL/v) * C p

Eqn (2.3)

where:
C =peak concentration at the sampling site, in micrograms per liter;
L P length of the measurement reach, in meters;
Q =stream discharge, in cubic meters per second;
Vs =volume of Rhodamine-WT 20-percent dye, in miiiiliters; and,
v =mean stream velocity, in meters per second.
The peak concentration (C p ) at the sampling site was l~J.g/L, foiiowing Kilpatrick &
Cobb (1985). A location 120-m upstream from the study site was chosen as the
injection site (Figure 2.2). The length of the reach (L) was a uniform glide, with no
obvious pools or back channel storage areas, to allow for uniform dispersion across
the channel (Turner Designs 1995). To achieve the desired concentration of
Rhodamine-WT (V s ), three serial dilutions were performed using stream water to
obtain a 100 11g/L working solution.
On July 15, 2011, water samples were extracted from 24 piezometers to record
background fluorescence prior to the tracer test. (One dry piezometer located in the
northwest corner of the site was eliminated from the test.) The following day, 100 L
of Rhodamine-WT working solution, comprised of five 20-L buckets, was released as
a bulk slug into the th alweg at the injection site. To enable rapid water collection
post-release and capture a representation of the entire study area, a subset of nine
piezometers was selected from every 3-m2 area of the study site (piezometers A, C,
E, G, J, 0, Y, V, R). Water sampling commenced once the dye plume reached the top
of the study site and continued every half-hour for two hours, at which point we

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

41

exhausted all the sterilized sample containers. A1124 piezometers were again
sampled that evening and the next morning.
Water samples were analyzed for fluorescence (~Jg/ dl) within 24 hours of
collection using a desktop fluorometer (Turner Quantech Digital Filter Fluorometer
Model FM109515, Barnstead I Thermolyne, Iowa, USA) equipped with a narrow
band filter of 540 nm and a sharp-cut filter of 585 nm.
This dye tracer test was performed under the premise that it would be
possible to sample all 25 of the piezometers at a frequency and length of time that
allowed for concurrent analysis and detection of breakthrough curves throughout
the study site. However, manual sampling with two people did not allow sufficient
intervals or quantity of data to determine the precise path of Rhodamine-WT tracer.
Consequently, this test only partially helped to define which piezometers to use for
the study. Individual trends in fluorescence and location were examined to
determine the subset of 10 study piezometers. One qualifier, as determined by the
tracer test, was that the location must have displayed an increase in fluorescence,
followed by a decrease (piezometers G, J, 0, V). The second qualifier, location, was
chosen to select piezometers that were positioned along a hypothetical path flowing
across the meander bend (piezometers F, G, I, J, K, 0), beneath the stream
(piezometers C, D), or at an increment extending into the riparian forest adjacent
from the top of the meander bend (piezometers G, I, V, W). Overall, dye attenuation
was highest at instream piezometers and riparian sites located near the upstream
bank.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

42

2.2.2.2 Sampling Scheme
Sampling occurred during the 2011 sockeye salmon run and the subsequent
spring of 2012. Seven sampling periods were timed to characterize varying delivery
and biochemical activity of MON. The fall period (August 23 to October 6) consisted
of one pre-salmon spawn sampling week, three weeks during active salmon
spawning, and a final post-spawn week once all salmon had spawned-out. There
were two sampling periods in the spring, one before and one after spring break-up
(Figure 2.6). Each sampling week consisted of 3 consecutive days in which water
samples were collected.
Aug

Oct

Week 1:
Pre-spawn

Week6
PoS't-spawn

Nov

Dec

Jan

Feb

March

April

Week 30
Before spnng break up

May

Week 34
After spnng break up

Weeks 3, 4, & 5
During spawn

Figure 2.6. Sampling periods timed to characterize the temporal nutrient delivery and
retention during the 2011 sockeye salmon run and the subsequent spring.

2.2.3 Site Characteristics and Aquifer Variables

Daily precipitation and air temperature measurements were recorded at an
Environment Canada hydrometeorological station located next to Gruhs Lake,
Horsefly, BC (52° 21' 56" N, 121 o 21' 19" W, el. 777 m). Horsefly River hydrometric

data was obtained from two Environment Canada discharge measurement sites on
the Horsefly River: Horsefly River above McKinley Creek (52° 17' 23" N, 121 o 3' 37"
W) and Horsefly River above Quesnel Lake (52° 26' 46" N, 121 o 25' 7" W). The
McKinley Creek station is located about 30 km upstream from the study site and the
Quesnel Lake station is approximately 20 km downstream.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

43

A land survey was conducted to map sampling locations, piezometers and
topography of the study site using an electronic theodolite survey station (Leica
Flexline TS06 Total Station, Heerbrugg, Switzerland). Over 100 survey points were
loaded into the geospatial processing program ArcMap (ArcGIS, Esri, California) to
build a contour map of the study site (Figure 2.4).
A subset of ten piezometers, chosen to represent hyporheic flow paths,
contained capacitance water level probes (Odyssey Capacitance Water Level Logger,
Christchurch, NZ) that recorded water level in the piezometers at 15-minute
intervals between August 22- October 8, 2011 and at 5-minute intervals between
March 30- May 7, 2012. Surface water staff gauge readings were recorded at midday during each sampling day.
Two pressure transducers (HOBO Water Level Logger U20-001-04,
Massachusetts, USA) were installed at the study site from March 30- May 7, 2012 to
measure surface water level during spring melt. The logger recorded absolute
pressure (kPa) and temperature (°C) every 15 minutes. HOBOware Pro software was
then used to convert pressure readings into surface water level readings by means of
barometric pressure compensation.
Hydraulic conductivity (K) was measured on April1, 2012 during baseflow
conditions using a falling-head slug test (Bouwer 1989, Wondzell & Swanson 1996a).
For this test, the time interval on all capacitance water level loggers was increased to
1-minute. Then, a known amount of water was withdrawn from each piezometer.
After 3 hours, the recorded data for the slug test were downloaded and then the
instruments were reset to 5-minute intervals to continue ongoing groundwater level
data collection. Hydraulic conductivity was calculated according to the Hvorslev

44

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
equation (Equation 2.2).
Porosity refers to the fraction of the soil's volume that is composed of pore
space which can be filled by water and/ or air (Flint & Flint 2002). In sandy
substrates, porosity has an inferred proportional relationship with hydraulic
conductivity. However, this relationship does not exist for all substrate types. For

example, clay typically has very low hydraulic conductivity, yet high porosity due
to its structure. Soil pits were dug to extract samples from hyporheic-influenced soils
and porosity measurements of three samples, located -777.8-778.0 mASL, were
performed in the laboratory. To calculate porosity, samples were first oven dried at
75°C for at least 12-h. Then water was added to a known volume of soil (VtotaJ) in a
beaker until the water covered the top surface of the soil. The volume of water
added, which can also be referred to as pore space (V void), was determined by
subtracting the water left in the beaker from the original total volume. The unit-less
measurement of porosity is the ratio of volume of pore space in soils to the volume
of material, with more porous substances closer to 1.0 or 100% saturation:
Porosity = V void/Vtotal

Eqn (2.4)

2.2.4 Groundwater Modeling

The water level inside the piezometers, presented as a height in meters above
sea level (mASL), indicates the hydraulic head of sediments surrounding the
piezometer opening. However, for hyporheic flow analysis, this study treats treats
these measurements as water table elevations and are used to interpret the lateral
flow direction, while assuming that vertical flow is negligible. Since there were no
wells installed on the study site to directly measure the water table and determine

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

45

the VHG of riparian piezometers, negligible VHG remains an underlying
assumption of hyporheic flow analysis.
Water levels inside the piezometers exhibited natural temporal fluctuations,
with some locations experiencing as much as a 1-m rise between late fall and spring
break-up. Changes in channel surface water level occurred daily during peak spawn
as a result of the build up and release of a "salmon dam" caused by carcasses
becoming lodged against gates separating sections of the spawning channe I (Figure
2.2). Hyporheic water levels changed as surface water fluctuated, which indicates
that this riparian zone is hydrologically connected to the surface water. During the
spring monitoring period, natural rises in water level were observed during
snowmelt and two large precipitation events that initiated spring break-up in 2012.
To better visualize the fluctuation in water level across the study site, ArcGIS 10.1
technology was used to create rasters that display hydraulic head at one point-intime. These rasters were then compiled into separate movies during three peak
water events in the spring of 2012: 1) A snowmelt event of 6-days between April7
and April13, in which the snowpack decreased from 23-cm to 0-cm; 2) A two-day
period between April26 and 28 after 27.4-mm of precipitation, initiating the start of
spring freshet; 3) A 24-hour period on May 1 where surface water rose considerably
after precipitation and in conjunction with the start of the spring break-up cycle.
Hyporheic flow path maps were generated to compare water movement at a
point-in-time at midnight during each month of the study. The day of the month
was chosen to reflect a steady state condition by eliminating outside influences of
water level change, i.e. salmon dam or precipitation. Two different techniques were
used to analyze and validate results of flow direction and magnitude in ArcGIS, one

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

46

with and the other without the Darcy Flow function. However, due to limited
sampling of aquifer properties, the Darcy Flow model did not produce realistic
results.

onsequently, this study will only describe the hyporheic flow analysis that

did not use the Darcy Flow function.
To generate the flow path maps (Figure 2.7), a hydraulic head raster was
created using thin plate spline interpolation. The Spline tool in ArcGIS uses an
algorithm that minimizes surface curvature and generates a smooth surface that
passes through the input points (DeMarsily 1986). From this hydraulic head raster,
two new rasters were created to represent direction and magnitude, or relative
speed, of flow. The Flow Direction function, located in the Spatial Analysis
hydrology toolbox, was executed to determine direction by locating the steepest
descent in hydraulic head from each cell (l.O-m 2 area) of the raster Genson &
Domingue 1988). By using the percent rise in slope from the hydraulic head raster, a
proxy for magnitude of flow was determined. Points from these new rasters were
symbolized as arrows that depict the flow direction and magnitude of flow.
Utilizing ArcGIS technology to generate computed flow patterns can be a
useful tool when modeling spatial environments; however, there are several sources
of potential error in the models generated by this study. Particularly, the low
number of sample data points (n=10 or 0.01 points per m 2 ) may result in errors
during the interpolation routine. Although interpolation is a valid method for
predicting values of an area when data points are limited (DeMarsily 1986 ), error is
reduced with greater sampling density (Ke1m edy et al. 2008). Some uncertainty also
exists in hydraulic head values because, in low permeable substrates, piezometers
may produce a lagged response (Hvorslev 1951). Also, this study's assumption that

47

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
VHG of riparian piezometers is negligible may not be accurate. If strong vertical

hydraulic gradients do exist in riparian piezometers, then the models will produce
an unrealistic representation of groundwater level and flow Oones & Holmes 1996).

Fill

Resample Raster

+---

..
I

Resample

+---

Flow direction Raster

- --

"
_____.

Raster to Point Shapefile ____.

Extract

/

+--- r

I

.

~to Points

•

Flow- "')

Ftll Raster

(

Dlr~

slope Raster

···----------------·
: Data Points representing :.
1

Direction and Magnitude 1

~-----------------~

Figure 2.7. Workflow chart showing steps for hyporheic water flow analysis using
Spatial Analysis tools in ArcGIS.

2.2.5 Water Chemistry
2.2.5.1 Field Sampling

Hyporheic water samples were collected from the chosen ten piezometers
around mid-day for three consecutive days during each week of the sampling
period. Water was extracted using a hand-operated vacuum pump with an attached
food-grade silicon hose and collected into separate 125-ml Erlenmeyer flasks.

Surface water samples were taken 5 m upstream, 5 m downstream, and at the
midpoint of the study site using a 60-ml syringe with Luer-Lok tip. Samples were
then filtered with 0.2-#Jm syringe filters. Approximately 12 to 15 ml of filtered water
was collected in sterile plastic centrifuge tubes and placed imtnediately on ice. The
samples were transferred to a freezer within 4 hours of collection and kept frozen

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

48

until laboratory analysis. All water extraction materials that made contact with
water were acid-washed in a 10% hydrochloric acid bath for a 2 hour minimum soak
time and triple-rinsed with deionized water (DI).

2.2.5.2 Laboratory Analysis
At the University of Northern British Columbia's Central Equipment
Laboratory, water samples were analyzed for ammonium (NH4 ) concentration
using an AutoAnalyzer (Bran+Luebbe AutoAnalyzer 3, Norderstedt, Germany). The
method used is a modification of the Berthelot reaction, using salicylate reagents and
a measurement of 660 nm (Rowland 1983). Before analysis of the samples, a
calibration was performed using standards of 0, 10, 25, 100, and 1000 ~g/L N. Two
quality control samples of 0 and 100 ~g/L N were inserted after every tenth sample.
About 10% of the samples were re-analyzed the following day for quality assurance.
Samples that exceeded 1000 ~g/L N were diluted and re-analyzed to achieve results
within the calibrated range.
2.2.6 Statistical Analysis
All statistical computations and graphs were conducted using the program R
(R Core Team 2013), as well as the R packages" AICcmodavg", "car", "ggplot2",
"gridExtra", "lattice", "mgcv", "plyr", and "scales". For each sampling week, the
three observations fro m each site were averaged to obtain mean NH/ concentration.
To confi rm statistical independence, temporal autocorrelation tests were conducted
using R's built-in autocorrelation function [ad()] to determine if NI--l/ displayed
"persistence" between sampling weeks (Chatfield 2004). Ammonium concentration
of three location al groupin gs (su rface water, instream piezometers (C and D), and

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

49

riparian piezometers) were averaged as a whole and between seasons, square-root
transformed for normality, and regressed through linear modeling to determine
statistical differences between locations and seasons. Pearson's product-moment
correlations were also performed to test the relationship between the arrival of
salmon and presence of NH 4 • For all analyses, the confidence level for statistical
significance was set tops; 0.05. Additionally, highly significant results were
identified if p s; 0.001.
Several site specific variables were chosen to be analyzed as predictors of
NH4 +concentration. Spatial variables included distance from stream bank, distance from

top of meander bend, withdrawal depth, and hydraulic conductivity. The temporal
variables used were precipitation and staff gauge reading. Before statistical analysis, all
variables were standardized to eliminate discrepancies between units of
measurement and enable direct interpretation (Schielzeth 2010). The following
calculation was used to standardize variables:
standardized parameter = {x,- x) I a

Eqn (2.5)

where:
Xi= observation i;
:X = mean of all observations; and,
a= standard deviation of all observations.
Hyporheic sites were separated into two groups for statistical analysis of
NH/: 1) all hyporheic samples (n=10) and 2) riparian hyporheic samples (n=8).
Correlation between predictor variables, known as multicollinearity, was evaluated
with variance inflation factors (VIF) using the vif() function in R. Lower VIF values
indicate less correlation, thus increasing the strength of multiple regression analysis

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

50

because the variables are more independent (O'Brien 2007). Linear and multiple
regressions tested the relationships between mean NH4 · concentration and one or
two predictor variables. The temporal variables precipitation and staff gauge reading
were tested separately from spatial variables. Multiple regression analysis included
the interaction term and was performed using variables with low or moderate
correlation (1 < VIF < 5). Mean NH 4 • concentrations were log-transformed for
normality. An information theoretic approach, specifically Akaike's 'An Information
Criterion (AIC)', was used to identify the most parsimonious models. AIC provides
a means for selecting the model that is as simple as possible, but no simpler
(Burnham & Anderson 2002). Second-order AIC (AICc), which is adjusted for small
sample size, was used and has the following equation:
AICc= (2k * (n / {n-k-1))-2ln(L)

Eqn (2.6)

where:
k = number of predictors;
n =sample size, and;
L =maximized value of the likelihood function.
For the spatial analysis, two linear regressions and three multiple regressions with
the lowest AIC values were retained as a candidate set to further rank these models
according to Akaike weights (w~). Akaike weights represent relative likelihoods, or
probabilities, for each model in the candidate set (Mazerolle 2015). For the temporal
analysis, the best linear and multiple regression were retained as a candidate set.
Lastly, the resulting effect size from regression equations was analyzed to interpret
the meaning of the best models.
In ArcGIS, exploratory statistics were analyzed for normality and volume

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

51

balance residuals were analyzed for steady and constant field flow during spatial
interpolation.
2.3 Results
2.3.1 Precipitation and Discharge
Throughout the fall sampling period, the Horsefly region experienced
relatively little precipitation, with an average of 0.5 mm I day and 2 out of 3 rain-free
days. Compared to long-term climate normals recorded by Environment Canada,
the total rainfall was low during this season. Through the month of September, 15.9
mm of rain was recorded, compared to the long-term September average of 43.9 mm
of rain. Horsefly River discharge, calculated from the McKinley Creek and Quesnel
Lake stations, remained relatively constant, but decreased slightly as the fall
progressed (Figure 2.8). The spawning channel is located approximately halfway
between these two discharge stations. Although discharge at the spawning channel
was not measured in the fall, it was likely very similar to measurements obtained
during a study conducted in 2010 because water inflow at the channel is regulated
by a siphon. Using Albers (2010)'s calculations, maximum (-3 m 3 Is) and minimum
(-1.5 m 3 ls) discharge was much lower at the channel during salmon spawn than at
the Horsefly River stations (McKinley Creek max: -12m 3 Is, min: -6 m 3 Is; Quesnel
Lake max: -19m3 Is, min: -14m 3 Is).
During the spring, water levels in the H orsefly River and spawning channel
were more variable due to snowmelt, precipitation events, and the onset of spring
break-up (Figure 2.9). The largest p recipitation event of the spring season occurred
on Apri1 26, 2012 w ith 27.4 mm of p recipitation falling in a 24-hour period. A peak in

52

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
discharge at McKinley Creek station was observed the next day. This rise in
discharge and continuing rain carried downstream to the Quesnel Station when a

peak in discharge was observed on May 3, 2012. As compared to Intensity-DurationFrequency (IDF) curves provided by Environment Canada, no rare or large-scale
rainfall events occurred during this study.

- - - -

-

7

Active

;,

spawn I

spawn II

Pre-spawn

6

es
.§.

Art •P

•£

Post-spawn

spawn Ill

~·

1e

~3

~2

0..

0
30

..

7

-McKinley

.. .. ...

'"'e
-2D
~
Cll

Quesnel

...

.•

··- ...

~

...

.. .

0

10

Aug22

Aug29

s..,os

Sap 12

Sep 19

S<lp26

Od03

Figure 2.8. Characterization of stream conditions at the Horsefly River spawning channel
in the fall of 2011. The top graph displays precipitation and the bottom graph displays
Horsefly River discharge at two measuring sites.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
1
•

~fore Break-up

---

..

E

After Break-up

...·.

-

·.·.
·. ..

eft

..,

100

~

53

McKmley
•• Qunnel

··.

(II

~

i5

50

.................................. . o••
Apr01

Apr 15

May01

May 15

Figure 2.9. Characterization of stream conditions at the Horsefly River spawning channel
in the spring of 2012. Precipitation is displayed on the top graph and the bottom graph
displays discharge measurements from two sites on the Horsefly River.

2.3.2 Aquifer Variables
The substrate and soil material in which hyporheic water flowed was mostly
comprised of "silt, sandy silts, clayey sands, and till" throughout the study site, as
defined by common hydraulic conductivity values listed in Freeze & Cherry (1979)
(Table 2.1). Piezometer 0, located deepest in the substratum and beneath the
spawning channel, displayed the highest conductivity and consisted of silty and fine

sands.
Porosity values from hyporheic-influenced soil from the C horizon ranged
from 0.30 total porosity at soil pit 1 to 0.62 and 0.65 at soil pits 2 and 3, respectively
(Figure 2.10). These samples were all extracted within a depth of 20-cm from each
other, or between -777.8 and 778.0 mASL. Soil pits 2 and 3 displayed a greater

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

54

occurrence of clay mottling, especially soil pit 3 which had a C horizon characterized
by gley, or waterlogged, soils. During sampling, hyporheic water began to infiltrate
soil pits 2 and 3, whereas the water table was not hit while digging soil pit 1.
Overall, soil pits 2 and 3 were considerably sandier than the coarse material at soil
pit 1. These porosity measurements and observations support representative
porosity values of geologic material from Freeze & Cherry (1979), which states that
the 0.30 value at soil pit 1 is comparable to course sand or gravel, and the porosity
values at soil pit 2 and 3 are indicative of clay.
VHG at the instream piezometers, C and D, indicated downwelling flow
throughout all sampling periods. There were three days in the fall (of a total of
fifteen observations) at site 0 where water was slightly upwelling, but still nearly
level with stream su rface water (Figure 2.11a). Piezometer C consistently displayed a
stronger downwelling gradient. In the spring, two sharp decreases in VHG mirror
surface water level changes during peak flow events (see Figure 2.11 b) indicating
that ground water recharge is greater during precipitation and break-up events.
Table 2.1. Hydraulic conductivity (K) measurements from each study piezometer on
April 1, 2012. The soil material type, which is associated with standard K values, is
adapted from Freeze & Cherry (1979).
Category K

c

K (cm/s)
0.000062

10 E -OS

Material
Silt. sandy silts, clayey sands, till

D

0.000631

10 E -04

Silty sands, fine sands

F
G
I

0.000111
0.000021
0.000066
0.000036
0.000026
0.000166
0.000083
0.000064

10 E -04
10 E -OS
10 E -OS
10 E -OS
10 E -OS
10 E -04
10 E -OS
10 E -OS

Silt. sandy silts, clayey sands, till
Silt, sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till
Silt. sandy silts, clayey sands, till

Location

J
K

0

v
w

55

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
Soli pit I : 779 mASL
4.8-m from streambank
Horizon A: -().23 em
- Blocky, subangular structure
- Black to very dark brown
(I OYR, value=2, chroma= I)
- Porosity = 60

Soli pit l: 778.8 mASL
17.6-m from streambank
Most evidence of m_ottling

Horizon B: -23-93 em
- Granular structure
• Sand lenses
• Dark olive brown
(2.5Y, value=], chroma=3)

Soli pit 3:778.6 mASL
4.8-m from streambank
Much sandier than other two pits

Horizon A: -Q-19 em
- Blocky, subangular structure
- Black to very dark brown
(IOYR. value=2. chroma= I)
- Porosity = 56

Horizon A:. -Q-12 em
- Blocky, subangular structure
- Brownish black color
(2.5 Y, value=3, chroma= I)
- Porosity = 60

50-cm: Evidence of
mottling appears

Horizon B: -12-64 em

Horizon B: -19-85 em

52-cm: Evidence of
mottling appean
~

/
~

40-cm: Evidence of
mottling appears

Horizon C: -93 em +
- Course material
- Granular structure
-Dark brown
(IOYR. value=3, chroma=l)
- Porosity = 30

-~ ~

Sand lens
.___ _ _ _ ___... 68-cm: Hyporhelc water
infiltration

87-cm Fine sand
~

I 00-cm: Hyporheic water
inflltration
~

Horizon C: -70 em +
-Clay; silt
- Granular structure
-Dark brown
(IOYR. value=3, chroma=3)
- Porosity = 65

Horizon C: -as em +
- Gley; clayey sands
- Greenish grey
(SG, value=S, chroma= I ).___San_d_y_m_aten
__·al_ _,
- Porosity = 62

I

Figure 2.10. Description of three soil profiles from the riparian zone. The diagram depicts relative heights (mASL) and distances
from the streambank for each soil pit, i.e. soil pit 1 and 3 were located closest to the streambank and soil pit 1 was highest in
elevation. Soil structure and color are described from the Munsell Soil Color Book.

56

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

...
p
%
~

;:

!

..

--

.·..

'

·:;
"

................
.........,..
...
...

~·:...:.·

~

\. .

\

....

vI '.-/
II

Figure 2.11. Vertical hydraulic gradient in the fall (a) and spring (b) at instream
piezometers C and D. Fall data (a) were only collected on days for which data are
displayed.

2.3.3 Geographic Information Systems (GIS)
In order to visualize the spatial and temporal fluctuations of hyporheic water
at the study site, hydraulic head and flow direction maps were created in ArcGIS
10.1 to observe reoccurring and unique patterns during each month included in the
study.
2.3.3.1 Hydraulic Head Contour Maps

Figure 2.12 shows variation in hydraulic head during each study month.
Hydraulic head is consistently higher further away from the stream, and generally
depicts a gradual, one-directional flow towards the surface stream. Note that each
map has a unique scale, as hydraulic head changed as much as 1-m between seasons
at some locations.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

o

57

Piezometers
Channel

Hydraulic Head (mASL
776.81 - 776.91
776.92 - 777.02
777.03 - 777.12
777.13 - 777.22
777.23 - 777.33
777.34 - 777.43
777.44 - 777.53
777.54 - 777.64
777.65 - 777.74
777.75- 777.84
I

00

o

I I
25

I

I
5M

Piezometers
Channel

oc

Hydraulic Head (mASL)
...____, 776.73 - 776.88
776.89- 777.03
777.04 - 777.18
777.19 - 777.33
777.34 - 777.48
777.49 - 777.63
777.64- 777.79
777.80- 777.94
777.95 - 778.09
OF

ov

778.1 0 - 778.24
I

I I
25

I

I

I
5M

Chapter 2. Nitrogen Cycling zn the Hyporheic Zone

58

Oct 2011

o

Piezometers
Channel

Hydraulic Head (mASL

77 6.80 - 77 6. 92
776.93 - 777.04
777.05 - 777. 15
777.16 - 777.27
777.28 - 777.39
777.40 - 777.5 I
777.52 - 777.62
777.63 - 777.74
777.75 - 777.86
777.87- 777.98

OF

I

I I

I

I

, N

A

00

o

Piezometers
Channel

oc

Hydraulic Head (mASL)

.._, 777.21 - 777.29
777.30 - 777.37
777.38 - 777.45
777.46 - 777.52
777.53 - 777.60
777.61 - 777.68
777.69- 777.76

OG

777.77- 777.84
777.85 - 777.91
777.92 - 777.99

OF.

I

0

I I
25

I

I
SM

_

59

Chapter 2. Nitrogen Cycling m the Hyporheic Zone
N

A

D

May 2012

o

Piezometers
C hannel

oc

I

Hydraulic Head (mASL) 1

D

776.60 - 776.83
776.84 - 777.07
777.08 - 777.30
777.31 - 777.53
777.54 - 777.77
777.78 - 778.00
778.0 I - 778.24
778.25 - 778.47
778.48 - 778.70
778.71 - 778.94
I

I

I

I

I

25

I
5M

Figure 2.12. Hydraulic head (mASL) at the Horseflv River spawning channel during each
month of the study. Note that each map has a umque scale.

Movies were created to illustrate patterns in hydraulic head fluctuations
during three peak flow events in the spring of 2012 (Figure 2.13). These movies can
be viewed online at YouTube by typing in the links below or searching with the
keywords "horsefly river spawning channel":
Snowmelt, April8-13- http: I I youtu.be I zeOw7XueWV c
Spring break-up, April26-28- http:/ /youtu.be/DObVT9CvGZk
Spring break-up, May 1 - http: I I youtu.be I zP1SkxSKNRM
Each movie represents a different time interval and magnitude of change, but the
overall pattern of water movement remains the same. Water level is highest and
responds the quickest at piezometers W & I during all events. Piezotneter K
consistently shows a depression in the water table and has a slower response time.
All locations were responsive to changes in surface water level indicdting that the

Chapter 2. Nitrogen Cycling m the Hyporheic Zone

60

hyporheic zone is well connected to the channel and there is a high rate of
groundwater recharge in the spring.

nes

-

_,

ne4

C/)

§.ne3

_,~

,._ na2

g

ftS

~

~ ne1

~

::l

C/)

nao

AprO '

Apr08

Apr 22

Apr29

May06

Figure 2.13. Surface\-\ ater h) drograph between Apnl 1 and May 8, 2012. Crey rectangles
encompass the dates used for mo\ ies depicting hvdraulic head fluctuations.

2.3.3.2 Hyporheic Flow Maps

A synthesis of geospatial and temporal information was used to characterize
subsurface flow direction and the magnitude, or relative speed, of water movement
during each study month. Hyporheic flow path maps show groundwater generally
flowing towards the stream, or from a position with higher hydraulic head to a
lower one (Figure 2.14). In the fall months, the greatest relative speed of riparian
hyporheic water is typically flowing from piezometer W to V, where the hydraulic
gradient between these points was large. In the spring, the magnitude of flow is
greatest between piezometers F, G, & I, vvhere the land surface had the highest
slope.

Chapter 2. Nitrogen Cycling m the Hyporlzeic Zone

<I <f <f / J <f ~ A '\) ~--..­
<f<f<f<f<f<f8A~
<f¢¢<1¢<1 A~~

0$- <I <I <I <I <I <I ~
# #

~

'l

~

~

~ -

61

Aug 2011

oo

A

OJ

OG

01

w

Chapter 2. Nitrogen Cycling in tire Hyporheic Zone

<I <I <I <lo~ <I A

Oct 20 II

<1<1<1</<I<IA

oo

<1<1<1<1<1<1~
o~ <t1 .t~ <t1 <t1 ¢ A
<I

62

d

OJ

~

Of

--

-=::~

-W

ov
April 2012

oo

oc

63

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

'\> ~ ~ ~ \,D':>

May 2012

~'\>~~
~~~~~~

OQ

~ ~~'\>~~
\>~~~~~
~~~'\>~~
~~~~~
,...__ ....___ .__ ~

OJ
OK

w

ov
Figure 2.14. Hyporheic water flow direction and relative speed, represented by arrow
direction and size, as modeled b} spatial analysi of hydraulic head contours in ArcGIS
during each month of the study. Land surface ele\ ation is depicted by the contour lines
(0.1-m intervals).

2.3.4 Ammonium

Wide variation in NH4 concentration was observed both temporally over the
period of study and spatially within the experimental site. Values ranged from <5
~g/L Nat surface water sites to >500 ~g/L Nat hyporheic water sites near the

spawning channel (Figure 2.15). Surface water samples ranged from < 5 to 80 ~g / L
N, with mean, median, and mode values of 12.5, 8.5, and 14.0 ~tg / L N, respectively.
Ammonium concentration at hyporheic sites ranged fro1n a lc'"'-' of 25 ~tg / L Nat thL
site furthest from the streambank (site W) to a high of 606 ~tg / L N belo\-\' the

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

64

streambed (site D). The overall average value of hyporheic N H / was 160 11g/L N,
with both median and mode values of 119 11g / L N (Figure 2.16). Surface water NH/
concentration was found to be highly significantly different from hyporheic water (p
< 0.001) throughout the study.
0
0

1/)

-

0
0

~
0)

""

:::J

0
0
M

1z

0
0
N

-

...

0
0

0

C

D

F

G

J

K

0

V

W

Up

Mid Down

Location

Figure 2.15. Box plots displaying all ammonium (NH4) concentration data at each
sampling location. "Up," "Mid," and "Down" represent surface water samples.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

•

65

•

J

• Kl

n67s 1

-<

Mean Ammonium
Concentration (ugfl.)

_J

CJ)

E

5mso...
a.

•

c

•o

~

•v

•w

•

100

•

200

(0

....~

"0

~ n6.25 ...

moo~

I

I

I

I

I

I

-5

0

5

10

15

20

Distance from Streambank (m)

Figure 2.16. Mean ammonium (N~ ) concentration, represented by symbol size, of all
sampling weeks at the hyporheic sampling locations. Withdrawal depth is measured as
meters above sea level (m.ASL).

2.3.4.1 Seasonality of Ammonium
Ammonium was present in surface and hyporheic water during all sampling
periods. Linear regressions evaluating the change in NH4 + concentration between
Fall 2011 and Spring 2012 indicate that not only is NH4+ still present after the winter,
but there is a statistically significant overall increase in NH4 + concentration in surface
water (p < 0.001) and hyporheic water (p < 0.05) during the spring season (Table 2.2).
The only sites exhibiting a decrease in spring NH 4+ concentration are at instream
piezometers C and D, which were drawing hyporheic water from below the
streambed (Figure 2.17). In fact, a highly significant decrease from fall to spring

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

66

values was identified at instream piezometers (p < 0.001 ).
A tren d of decreasing NH 4 ' concentration is evident as the fa ll spawning
season progresses (Figure 2.17). There were six water sampling days (during Active
spawn I and Active spawn II) in which salmon were entering the spawning channel
and correlations between salmon number and NH4' could be analyzed (Table 2.3).
The results display a statistically significant (p < 0.05) negative correlation between
number of salmon and surface water NH4 • concentration (correlation coefficient of
-0.919). This outcome is contrary to the hypothesis that more salmon will result in an
increase of NH 4 in the water column.
Table 2.2. The change in ammonium (NH 4 ) concentration at each sampling location
between Fall 2011 and Spring 2012 sampling periods. "Up," "Mid," and "Down"
represent surface water samples.
Change in

Site

c

D
F
G
I

J
K
0

v
w
Up

spring NH 4+

p-value

Decrease
Decrease
None
Increase
Increase
Incre ase
Increase
None
Increase
None
Increase
Increase
None

Mid
Down
All hyporheic
water
Increase
All surface
water
Increase

2e-04
2.8e-06

Effect size
- 1.23
-0.651

0.53
0.694

7.3e-05
0.004
0.013
2.3e-04

1.1
0.838
0.628
0.554

0.572
0.361
0.282
0.504

4.7e-05

0.708

0.591

0.008
0.006

0.58
0.631

0.319
0.339

0.045

0.228

0.019

2.4e-04

0.613

0.2

R-squared

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

67

400-

-

lnstream

..

••••• Riparian

II· • ••••• • ..

--- Surface

0

..

••.

.. .

~

c

..~ ·-..

~ 100~

}------oI

PriHipawm

.
...

..,.·

--~---------.---------- ~ ---------~------I

I

I

Active sp;1MI I Active spawn II Active spawm Ill

I

Post-spawn

,.

-·

I

I

___ ________

Before Bruk-up After Break-up

Figure 2.17. Mean ammonium concentration (JJg/L N) at water sampling locations
during each sampling period, with error bars showing standard error. The "InslTeam"
group represents piezometers located in the stream (n=2); "Riparian" represents the
riparian hyporheic piezometers (n=8); "Surface" represents the surface water sampling
locations (n=3 ).
Table 2.3. Sockeye salmon (0. nerka) spawning escapement at the Horsefly River
spawning channel and mean NH4+ concentration for surface water samples during
Active Spawn I and II.

Date
Sept.12
Sept.13
Sept.14
Sept.20
Sept.21
Sept.22

Number of
Salmon
2131
2243
2244
2339
2341
2346

Mean NH4
(ug/L N)

27.8
24.8
23.0
14.7
19.3
19.0

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

68

2.3.4.2 Variables Affecting Hyporheic Ammonium
Predictor variables describing spatial and temporal attributes were used to
assess NH/ concentration for two groups: 1) a11 hyporheic sites (n=10) and 2)
riparian hyporheic sites (n=8). For group 1, the variables distance from streambank,

withdrawal depth, hydraulic conductivzty, precipitation, and staff gauge reading were
analyzed. Group 2 additiona11y considered the variable distance from top of meander

bend. Variance inflation factors (VIF) of group 1 and group 2 variables were a111ow
to moderate, meaning that each variable is reasonably independent of the other
variables (Table 2.4). Therefore, all variables were considered for regression analysis.
Table 2.4. Variance inflation factors (VIF) used to describe multicollinearity for predictor
variables used in multiple regression analyses

Variable

Type

distance from streambank
distance from top of meander bend
hydraulic conductivity
withdrawal depth
precipitation
staff gauge reading

Spatial
Spatial
Spatial
Spatial
Temporal
Temporal

Group I

Group 2

VIF
1.54

VIF

n/a
4.61
4.82
1.09

1.04

2.09
1.76
1.80
2.70
1.22
1.22

Linear regressions were used to analyze the statistical significance between
each variable and NH4 +concentration. For group 1, the variables distance from

streambank, withdrawal depth, and hydraulic conductivity were considered statistically
significant to predict NH4+concentration (p < 0.05). The group 2 variables that were
statistically significant (p < 0.05) included distance from streambank, distance from top of

meander bend, and precipitation.
AICc values from linear and multiple regressions of spatial variables were

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

69

analyzed to determine a candidate set of the most robust models, which induded
the two best linear regressions and three best multiple regressions (Table 2.5). The
most effective spatial variables in predicting NH 4 concentration varied between
groups, but multiple regressions were consistently stronger models than linear
regressions. For all hyporheic samples (group 1), the predictor variables distance

from streambank, withdrawal depth, and their interaction term produced the most
robust regression with an Akaike weight (wr) of 0.58, which describes the probability
of this model being the best model in the candidate set. When analyzing the riparian
hyporheic samples (group 2), the multiple regression of distance from top of meander

bend, hydraulic conductivity, and their interaction term had the lowest AICc and W;=
0.62. The strongest singular spatial variables to predict NH 4' concentration were

hydraulic conductivity for group 1 and distance from stream bank for group 2.
The effect sizes associated with the regression results aid in understanding
the real world application of these models (Table 2.5). The interpretation of group l's
best model explains that, independently, an increase in distance from streambank
(effect size= -0.20) and withdrawal depth (effect size= -0.20) tend to result in lower
NH4 +concentrations. However, when distance from streambank and withdrawal
depth increase together (effect size= 0.36), this will cause NH 4+concentration to
decrease less. The multiple regression of distance from streambank and withdrawal

depth had similar results for both groups. The interpretation of this model for group
2 explains that when controlling for withdrawal depth, a 1-m increase in distance from

streambank will cause NH/ concentration to decrease by 0.27 ~gIL N (effect size=
-0.27).

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

70

Table 2.5. Model selection results of spatial variables for all hyporheic (group 1) and
riparian hyporheic (group 2) models The effect size is written in parenthesis after each
predictor variable. The most parsimonious linear and multiple regressions are ranked by
second-order Akaike's Information Criterion (AIC,) scores. SB streambank, TMB -= top
of meander bend, df - degrees of freedom, w, Aka ike weights.
Group

I

2
2
2

2

2

Model variables (effect size)
distance from SB (-0.20) +withdrawal depth (-0.20) +Interaction term (0.36)
distance from SB (-0.18) +hydraulic conductivity (0.33)
distance from SB (-0.29) +withdrawal depth (-0. 33)
hydraulic conductivity (0.36)
withdrawal depth (-0.27)

df
5

distance from TMB (-0.32) +hydraulic conductivity (0.06) +interaction term (0.20)
distance from SB (-0.08) +withdrawal depth (-0.07) + interaction term (0.25)
distance from SB (-0.1)4)
distance from SB (·0.27) + withdrawal depth (-0.12)
distance from TMB (-0.03)

5
5
3

""
""3
3

""3

AIC<

MIC<

127.-45
129.29
130.01
133.-43
143.17

0.00
1.83
2.56
5.98
15.72

w,
0.58
0.23
0.16
0.03
0.00

100.89
103.66
10-4.71
105.38
105.54

0.00
2.78
3.83
-4.-49
4.65

0.62
0.16
0.09
0.07
0.06

When analyzing the temporal variables, AICc results demonstrated that, for
both groups, the linear regression using precipitation was the strongest model,
followed by the multiple regression of precipitation and staff gauge reading (Table 2.6).
The results state that for each incremental increase in precipitation, NH 4+
concentration will increase by 0.03 ~g / L N (group 1) or 0.05 ~g / L N (group 2).
When controlling for staff gauge reading, each incremental increase in precipitation will
cause an even greater increase in NH4+concentration (0.20 ~g / L N for group 1 and
0.27 ~g/L N for group 2). However, when the stream channel level rises, as signified
by a higher staff gauge reading, N~+ concentration will decrease by 0.10 ~g / L N
(group 1) or 0.12 ~g/L N (group 2) when controlling for precipitation.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

71

Table 2.6. Model selection results of temporal variables for all hyporheic (group 1) and
riparian hyporheic (group 2) regression models The effect size is written in parenthesis
after each predictor variable. The most parsimonious lmear and multiple regressions are
ranked by second-order Akaike's Information Criterion (AICc) scores. df degrees of
freedom, w, = Aka ike weights
MICe
0.00

4

AICC
150.71
I51.75

1.04

0.63
0.37

3
4

103.50
103.76

0.00
0.25

0.47

Group Model variables (effect size)
predphatlon (0.03)
I
precipitation (0.20) + staff gauge reading (-0.1 0)
I

df
3

pre<:ipitation (0.05)
precipitation (0.27) + staff gauge reading (-0.12)

2
2

w,

0.53

2.4 Discussion
2.4.1. Defining the Hyporheic Zone of the Horsefly River Spawning Channel
Hyporheic zones respond to changes in surface water levels and stream
discharge (Valett et al. 1994), and this concept is substantiated by our study. The
hypothesis that a hyporheic zone is present beneath the stream and below the
riparian zone was confirmed through hydraulic head contour and flow mapping, as
well as perceptible changes in hydraulic head that corresponded with surface water
level fluctuations. Hydraulic head beneath the riparian zone varied between a lower
limit of 80-cm and 164-cm deep during low flow in the fall and from an upper limit
of 10-cm to 56-cm below the surface at the peak of spring break-up. Directly beneath
the stream, the saturated thickness of the hyporheic zone extended from the

streambed to at least 155-cm deep and possibly deeper, but this is unknown as 155
em was the deepest point sampled. During specific times at particular study sites,
hydrologic and chemical characteristics implied the presence of a boundary between
saturated and unsaturated soils.
Hyporheic flow is controlled primarily by factors of sediment characteristics

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

72

(grain size, type, and connectivity), groundwater and surface water discharge rates,
and hydraulic conductivity (Valett et al. 1994, Dent et al. 2001, Tonina & Buffington
2009). The degree to which these variables exert control over hyporheic exchange
depends upon scale, time, and place. Attributes of this aquifer were evaluated
according to hydraulic head, hydraulic conductivity, and soil characteristics.

2.4.1.1. Aquzfer Properties
Negative VHG at instream piezometers (C and D) affirm that the
experimental site is a location of groundwater recharge. Slight upwelling and
neutral VHG occurred infrequently at piezometer D during low flow (Figure 2.11a),
suggesting that the upper limit of the hyporheic / groundwater boundary is around
155-cm below the streambed (or 775.943 mASL). High flow events had a discernible
effect on groundwater recharge; for example, VHG dropped from -0.37 to -1.29 m ·
m-1 at C and -0.27 to -0.84 m · m-1 at D during a rain event on April 26, 2012.
Hydraulic conductivity (K) values were relatively consistent throughout the
study site (K- 104 to l0-5 ), indicating a fairly homogeneous soil type of silt / clayey
sands at lower K and silty / fine sands at higher K (Table 2.1; Freeze & Cherry 1979).
The lowest K was observed at riparian site G and coincides with personal
observations of slow-refilling and particulate-rich water. During low water events,
piezometer G contained only enough water to fill one 12-ml sample, which suggests
that the water table was occasionally situated just above the piezometer openings
located 164 em below the land surface (or 777.292 mASL).
During soil sampling, evidence of clay mottling was observed in the B
horizon of all soil pits, which marks the perched height of the water table
(specifically between 778.2 and 778.5 mASL). Gley and blue/ green soils, which

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

73

characterize a persistent water table, were identified in the C horizon of the soil pit
located furthest frmn the streambank (soil pit 2).
Since hydraulic conductivity defines the ease of water movement in soil, it
seems to infer that a location with high porosity will also exhibit high hydraulic
conductivity. However, the resistance of water movement in soil depends on the
size and connectivity of pore spaces. As sediment particle size decreases, the
proportion between porosity and K fails because the actual correlation is between K
and pore throat radii (Soil Survey Manual 1993). Clay, being the typical example of
this, consists of ultra-fine grained minerals and displays a strong apparent cohesion
due to electrostatic forces between ions of minerals and oxygen atoms. This
structure is highly porous, and often supersaturated, but exhibits low hydraulic
conductivity because the many pore spaces are not well connected (Barnes 1995).

2.4.1.2. Hyporheic Flow Paths
Elevated hydraulic head values at locations furthest from the stream indicate
one-directional flow of groundwater toward the stream channel (Figure 2.12), which
is an expected result and typical of both surface and groundwater flow direction
(Triska et al. 1993, Chestnut & McDowell 2000). The flow maps suggest there was a
greater opportunity for hyporheic flow through the riparian meander bend in the
fall because of the variable flow directions and a small magnitude of flow toward the
stream. In the sp ring, the pattern of flow was strongly one-directional toward the
channel, ind icating that hyporheic flow at this time was likely limited (Figure 2.14).
Although water is generally flowing into the channel, hyporheic exchange between
the stream and streambank still occur on a smaller scale (Duval & Hill2007).
Evidence of this exch ange from the stream to riparian zone occurred during the

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

74

Rhodamine-WT tracer test when fluorescence of hyporheic water increased after the
slug injection. Hyporheic exchange between the stream and streambed was
confirmed from VHG measurements.
Although VHG indicated a dominantly downward flow of water into the
streambed, the flow of riparian groundwater towards the stream is known to
promote upwelling conditions (Hill et al. 1998). This discrepancy could be the result
of a localized influence associated with bedforms and also the artificial construction
of the channel (Kasahara & Wondzell 2003). The streambank at the HFC is relatively
steep compared to natural meander bends on the Horsefly River, where there is a
gradual transition between the stream and riparian zone. Perhaps there is a limited
connection between the flow of water from the riparian zone and the flow of water
into the streambed at this site. To improve on the interpretation of hyporheic flow
across the riparian and beneath the stream, piezometers installed upstream of the
meander bend may have resulted in hydraulic head values that inferred different
flow directions and magnitudes.
2.4.2. Sources and Uptake of Ammonium in Surface Water
Contrary to other experiments of natural or artificial salmon additions
(Minakawa & Gara 1999, O'Keefe & Edwards 2002, Johnston et al. 2004, Chaloner et
al. 2004, Mitchell & Lamberti 2005, Claeson et al. 2006, Janetski et al. 2009, Levi et al.
2011), NH/ concentration of surface water did not increase upon the arrival of
salmon and, thus, this study fails to reject the null hypothesis (I). Average values
decreased from 27 11g/L N before the arrival of salmon to post-spawn values of 5
/lg/L N. In fact, a strong negative correlation (r = 0.91, p < 0.05) between salmon

Chapter 2. Nitrogen Cycling in the Hyporheic Zone
abundance and

75

H 4 concentration was calculated during active salmon spawn,

which infers that the presence of salmon significantly reduced surface water NH4 +
concentration (Table 2.3). Ammonium concentration naturaJJy varies between rivers,
and some examples of systems which gain NH 4 in the presence of salmon are listed
in Table 2.7. In a Washington stream, NH4 ' level remained unchanged during
salmon spawn and this was attributed to the rapid removal of NH4 ' from solution
due to biological uptake (Bilby et aJ. 1996 ).
Table 2.7. Examples of streams with increased NH4 concentration during salmon runs,
as compared to the present study. Most values were estimated from published graphs.
NH 4+with

Authors
Present stU

Location

Species

Horsefly channel, BC

Sockeye

Minakawa & Gara 1999

Uttle Kennedy Creek, WA Chum

O'Keefe & Edwards 2002 Lynx Creek, AK

Sockeye

NH 4 + without

salmon {"giL) salmon (}lg/L)

s

27

-2SO
147
-IS
-12

-so

<S
-10

Peterson Creek, AK

Pink, Chinook, Coho

Chaloner et at. 2004

Salmon Creek, AK

Mitchell & Lamberti 2005

Fish Creek, AK

Pink, Chinook, Coho
Chum, Pink. Coho,
Chinook

40-60

Levi et al. 20 I I

Prince of Wales Island, AK

Pink, Chum

-160

2

-o

-s

Certainly ammonia was delivered to the spawning channel by salmon
(Mitchell & Lamberti 2005), but since we cannot attribute NH 4+ concentration in
water samples to salmon, a new question arose. Where is NH/ coming from?
Possibly the most influential local source was the resuspension of sediment and
bacteria during the channel cleaning that occurred between August 8-12 and 15-18,
2011, ending 5 days prior to pre-spawn sampling. Over a 10-day cleaning period, a
year's worth of accumulated streambed nutrients were released, potentially causing
a dissolved nutrient spike which mostly attenuated before sampling commenced.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

76

When N availability increases, microbes absorb, or mineralize, dissolved organic N
(DON) as energy for growth and excess N is excreted as NH4 +in a process called
ammonification. Nitrifying bacteria then absorb, or immobilize, this additional NH4 +
and nitrification becomes the predominant transformation mechanism (Figure 2.18;
Peterson et al. 2001, Chapin et al. 2011 ). This process of peak NH4 ' attenuation
followed by a decline has been observed in other salmon-bearing rivers Oohnston et
al. 2004, Claeson et al. 2006, Rinella et al. 2013). Furthermore, 2011 was a weak year
for Horsefly River sockeye returns and the spawning channel was filled to less than
10% of capacity. In contrast, the sub-dominant 2010 salmon escapement surpassed
the total 2011 escapement on only the second day of spawning channel operation,
and was eventually filled to 100% capacity. In another northern BC stream, Johnston
et al. (2004) found that NH4+concentrations exceeded N03- during a year of high
salmon abundance, but the reverse was true the following year during low salmon
abundance. At the HFC, low spawner density in conjunction with resu spension of
the previous year's dissolved nutrients likely overshadowed NH 4+delivery by 2011's
spawners.

77

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

NH 4 and N03
m export

NH 4 and N0 3 In
stream flow

1

ASSimilatory

NH4 uptake

D1rect
n.tnf1cat1011

Ass•m•latory

N03 uptake

Stream
bottom

NH4 and N03 m
seepage from catchment

Figure 2.18. A conceptual model of the mtrogen cvcle at the mterface of surface and
hyporheic "'ater. (From Peterson et al 2001. Repnnted "'ith perm1ssion from AAAS.)

InN-limiting ecosystems, actinorhizal plants, such as alder (Alnus spp.),
provide an important source of N to forested ecosystems through symbiotic nitrogen
fixation. Alder-fixed nitrogen (AFN) can be transported to the stream by advection
of water through the soil and direct leaf litterfall and, similar to MDN enrichment, is
utilized by surrounding microbes and vegetation (Helfield & Naiman 2002,
Compton et al. 2003). TheN supplied by alders is transferred to riparian soils
primarily through root secretions and leaf litter, and has been shown to increase soil

Nlf4 and N03- concentrations (Shaftel et al. 2012). Alders cotntnonly grow near
water throughout the Cariboo region, which encompasses the Horsefly watershed (J.
Holmes, pers. comm.), and Mountain alders (A. tenuifolia) were present in the
riparian forest of our study site. In 1nature forests, AFN may be a more important
contributor of soluble nutrients than MDN, perhaps, in part, because the timing of

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

78

maximum stream N from alder occurs in both the spring and the fall (Wondzell &
Swanson 1996b, Helfield & Naiman 2002, Shaftel et al. 2012).
Land-use activities in the Horsefly River watershed include mainly logging,
agriculture, mining, and urbanization (Holmes 2008, Smith & Owens 2014), which
all can deliver substantial amounts of N to the stream. Small to medium sized beef
cattle and hay production farms border much of the river upstream from the
spawning channel. Further upstream, logging has been active in the past and
present. Paved and gravel roads run adjacent to and over the spawning channel
within the 1000 person town of Horsefly. However, these anthropogenic activities
appear to be a minor source of dissolved nutrients. Surface water samples from the
Horsefly River, taken directly upstream of the spawning channel in March 2008,
displayed concentrations of dissolved inorganic N (DIN; nitrite +nitrate +
ammonia) at 99 JJg/L and soluble reactive phosphorus (SRP; orthophosphate) at 3.3
JJg/L (Holmes 2008). Concentrations at these levels are considered low (Meybeck
1982, Tank & Dodds 2003) and are an order of magnitude or more lower than
studies from N-rich agricultural streams by Hill et al. 1998 (>1000 JJg/L N) and
Bohlke et al. 2009 (1400- 14,000 JJg/L N). However, DIN and SRP values during the
summer and fall would provide a better picture of the effect of land-use activities.
All of these sources (nutrient resuspension during channel cleaning, N-

fixation by alder, and anthropogenic land-use activities) likely contributed DIN to
the stream to varying degrees, but the fate of NH 4+still remains unclear due to its
rapid immobilization from the water column. Peterson et al. (2001) compared longterm ammonium- 15 N ad ditions in 11 streams from varying biomes and found that on
average 20-30% of NH4+ was removed by nitrifying bacteria, while 70-80% was

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

79

assimilated by benthic organisms. Since discharge in the channel is relatively slow
(-2 m 3 /s), it can be assumed that the NH4 ' uptake length is short (Peterson et al.

2001, Albers 2010) and that NH4 · was utilized directly in the stream by micro- and

macro-organisms (Bilby et al. 1996, Webster et al. 2003, Levi et al. 2013). Desirable
nitrification conditions of high dissolved oxygen (DO) and plentiful DON and/ or
NH 4+can contribute to the rapid uptake of NH4 ' and low surface water NH4 +
concentrations (Chapin et al. 2011, Levi et al. 2011). However, some streams,
particularly oligotrophic streams, seem to have a fixed capacity for nitrification
(Newbold et al. 1983). AnN overload event was documented in other northern
watersheds of the Fraser River basin by Johnson et al. (2004), when years of high
salmon abundance and high NH 4' were followed by a year of low salmon and NH4 +
abundance. In theory, a profusion of instantaneously immobilized NH4 +would
remain mostly undetected in the spawning channel, until nitrifying bacteria and
benthic organisms reached their N-uptake limit, at which point NH 4+concentration
would begin to increase (Chapin et al. 2011 ).
2.4.3. Sources and Uptake of Ammonium in the Hyporheic Zone

Hyporheic water samples had significantly higher NH 4+ concentrations
compared to surface water values, both spatially and temporally, with an overall
average enrichment of 1380% in the fall and 1035% in the spring. Therefore, this
study rejects the null hypothesis (II) that there would be no difference between NH 4+
concentration in surface and hyporheic water. This observation is supported by
other studies in which hyporheic water was distinctly enriched in NH 4+compared to
surface water (Valett et al. 1990, 1994, McClain et al. 1994, Fiebig 1995, Chestnut &

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

80

McDowell 2000, Greenwald et al. 2008). Riparian and hyporheic environments are
known to be a net source of N to streams (Wondzell & Swanson 1996b, Shibata et al.
2004). Aerobic surface water is capable of rapidly removing NH4 + from solution
through nitrification, whereas anaerobic subsurface environments facilitate
denitrification (Peterson et al. 2001, Seitzinger et al. 2006). Advective transport of
water through the hyporheic zone leads to longer contact time with sediments and
associated microbes, increasing opportunities for biogeochemical transformation of
N (Triska et al. 1990). These conditions complicate our ability to track the flux of
NH4 + from salmon because the transformation of NH4 ' to other forms of DIN or
assimilation by organisms can occur both quickly and repeatedly.

2.4.3.1. Temporal Controls of NH4
A major aim of this study was to determine if hyporheic storage of salmonderived NH4+occurred over the winter. Although NH4+present in the hyporheic
zone cannot be directly attributed to salmon, temporal trends of NH 4+concentration
were observed, as well as a marked difference between instream and riparian
hyporheic locations (Figure 2.17). Overall, the riparian zone displayed a significant
increase in NH/ concentration from fall to spring. Conversely, a significant decrease
in NH4 +was observed in hyporheic water directly below the streambed (Table 2.2).
Apparently these adjacent hyporheic systems have differing sources of Nor
disparate subsurface attributes that control biogeochemical processing. These results
reject the null hypothesis (IV) that NH4 +concentration will not differ between
seasons.
The riparian hyporheic zone may beN-limited in the fall and less limited in
the spring when N becomes more available. From the regression analysis, it was

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

81

found that precipitation was a strong predictor of riparian NH4 +concentration,
indicating that NH/ increases by 0.05 J.Jg/L N with every millimeter of precipitation
{Table 2.6). Past research has suggested that spring precipitation events induce
lateral N inputs from the hill slope (Wondzell & Swanson 1996b, Shibata et al. 2004,
Wondzell & Gooseff 2013), and that snowmelt and rain can aid N percolation from
the overlying organic soil layers and leaf litter (Chafiq et al. 1999). Also, an
interesting phenomenon of freeze-thaw in the spring bursts a multitude of microbial
cells, releasing pulses of nutrients and inducing rapid accumulation of NH4 +and
N03- (Chapin et al. 2011 ). Triska et al. (1994) found riparian hyporheic locations to
have lower NH4 + in the winter, except at sites with alders. During the anoxic winter
months, alders may continue to fix N, thus contributing to the accumulation of NH/.
In the spring, the higher water table is capable of reaching the shallow rooting zone
of alders.
Peak streamflow conditions in the spring induced strong downwelling
gradients into the hyporheic zone beneath the streambed, which likely facilitated
better nitrification conditions by delivering DO and lowering NH 4+ (Wondzell &
Swanson 1996b, Zarnetske et al. 2011 ). During the salmon spawn, leaf litter, biofilms,
and streambed sediments continue to assimilate N from the surface water (Albers &
Petticrew 2012). This storage and accumulation of nutrients and organic matter may

have contributed to an increase in water residence time and decrease in DO, thereby
increasing NH4+ concentrations beneath the streambed in the fall (Hinkle et al. 2001,
Argerich et al. 2008, Heppell et al. 2014).
Flow direction maps provide evidence to a seasonal shift of nutrient
exchanges between the channel and riparian zone (Figure 2.14). In the fall, weaker

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

82

and more variable flows suggest that hyporheic water could be delivering NH4 +and
other nutrients to the riparian zone. During the spring, break-up events (e.g.
snowmelt and precipitation) promote strong subsurface flows toward the channel,
resulting in a lateral flux of nutrients to the stream. The observed increase of surface
water NH4 +concentration in the spring, alongside increased riparian hyporheic
NH/, substantiates this flow direction (Figure 2.17). Seasonally shifting nutrient
flow between adjacent ecosystems is known as reciprocal subsidies. This interaction
typically occurs when terrestrial and aquatic productivity peaks at different times of
the year (Nakano & Murakami 2001). Temporally shifting hyporheic flow at the
HFC may be evidence of another pathway for the exchange of reciprocal subsides in
stream-riparian ecosystems.
2.4.3.2. Spatial Controls of NH4

Ammonium content varied greatly between hyporheic locations, but
statistical regressions confirmed a few spatial variables that can predict NH/
concentration at this site. Distance from streambank continually ranked as a strong
predictor of NH/ in the hyporheic and riparian zones, demonstrating that NH 4+
concentration decreased as distance from the streambank increased. This result
supports the rejection of the null hypothesis (III) that the spatial variable, distance
from the streambank, does not correlate with NH 4+concentration. While controlling
for other variables, NH/ has the potential to decrease between 0.20 and 0.29 J.lg / L N
with every meter further from the streambank (Table 2.5). Although some research
has shown that anaerobic conditions conducive to denitrification result in more
NH4+with distance from streambank (Triska et al. 1993, Hinkle et al. 2001, Duval &
Hill2007), lateral distance from the stream is not always this straightforward.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

83

Hidden beneath the surface are dominant patterns of hyporheic flow and micro-sites
of accumulated organic matter, plant roots, and limited nutrients (McClain et al.
2003, Poole et al. 2004, Claret & Boulton 2009). Other studies also agree with our
results of decreasing NH4 ' with distance from the stream (Chestnut & McDowell
2000, O'Keefe & Edwards 2002). Since the hydraulic head maps indicate that
subsurface water flow is generally toward the stream, it may be inferred that
nutrients, sediment, and organic matter accumulate more readily in soils close to the
streambank. Certain biogeochemical reactions may be occuning in these soils
because of either increased N or C availability. Where N is not limiting, soil
microbes will mineralize more DON as energy for growth and excrete excess NH 4+
(Wondzell & Swanson 1996b, Chapin et al. 2011). Organically-rich subsurface
sediments can induce a microbially-mediated reaction known as the dissimilatory
reduction of nitrate to ammonium (DNRA). During this reaction, NH 4+availability
increases in the absence of oxygen (Hill et al. 1998, Lewandowski & Niitzmann
2010).
Hydraulic conductivity was the strongest singular predictor of NH/
concentration when considering all hyporheic locations (group 1), indicating that as

K increased, NH4 +also increased. This positive relationship conflicts with the
assumption that higher K typically resu lts in shorter water residence time, higher
DO, and, therefore, a nitrification environment characterized by less NH 4+ (Pinay et
al. 1995, Chestnut & McDowell2000). However, K values were only measured once
in the spring, so interpretation of this variable would have been improved by
conducting several K tests per study month and log averaging the values (U.S. Dept.
of Agriculture 1993, Genereu x et al. 2008, Hester 2008, Claret & Boulton 2009). In

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

84

addition, slug tests themselves are a source of significant uncertainty when
estimating K (Cardiff et al. 2011). Post-hoc reanalysis revealed that the significantly
positive relationship between K and NH4 occurred only during the fall sampling
period. It is possible that soils around the piezometers settled during the winter and
altered K of the sampling area. Also, when removing piezometer 0 from the
analysis, which appears to be an outlier, the relationship between K and NH/
weakens. Without piezometer 0, the linear regression of distance from streambank
becomes the most robust model (effect size = -0.02). However, if hyporheic water is
well oxygenated, then a positive relationship between K and NH4 may hold true.
Past studies have demonstrated that N0 3- concentrations can increase along a flow
path, despite increased residence times, if there is enough DO and NH4 + available
for nitrification to occur (Triska et al. 1990, Claret & Boulton 2009, Argerich et al.
2011).

The interaction of distance from streambank along with withdrawal depth
was found to be a good predictor of NH4 +in the hyporheic zone (Table 2.5). Multiple
regression results indicate that for every increase in withdrawal elevation at a given
distance from the stream, there will be more NH 4+ than if water was being drawn
from a lower depth (Figure 2.19). However, this result contradicts the theory that
deep hyporheic water is characterized by denitrifying anaerobic environments and

higher NH/ concentrations (Hill et al. 1998, Hinkle et al. 2001). By scrutinizing
Figure 2.15, it appears that statistical significance may be an artifact of the
positioning of these piezometers. When piezometer 0, located about 0.5 m deeper
than the next deepest location, is removed from multiple regression analyses, this
interaction is no longer significant. However, without piezometer 0, the only

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

85

difference between the two groups is piezometer C. A better way to explore the
effect of withdrawal depth on NH 4+concentration would have been to position
piezometers in a nested design, where water would be extracted from three depths
at approximately the same location.

RIPARIAN

STREAM

Figure 2.19. A profile of a stream and adjacent riparian zone displaying the subsurface
zones that may have higher ammonium (NH4 ) concentrations.

To better understand the fate of NH4 + at this meander bend, variables of DO
and N03- would have been advantageous. The presence or absence of oxygen
controls the mechanisms behind subsurface N cycling (Argerich et al. 2011, Peyrard
et al. 2011 ), so sufficient spatial and temporal density of DO values could have acted
as a proxy for N concentration and transformation. For similar reasons,
concentrations of N0 3- would have been desirable. Since NH4+ additions from
spawning salmon were likely rapidly converted to N03-, knowing these values could
address this assumption and, if confirmed, increased my ability to address N in
terms of salmon input Oohnston et al. 2004). Nitrate values would have also helped
discern the nutrient cycling relationship between NH4 + and N03- in the hyporheic
zone. The fate of most N0 3- is assimilation and mineralization by primary

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

86

producers, but N03• can also be assimilated and stored for an extended period of
time or denitrified to atmospheric N (Ashkenas et al. 2004, Hall et al. 2009). Nitrate
assimilation is related to seasonal and die] availability of sunlight, with greater
uptake and overall gross primary production during light hours (Mulholland et al.
2006). The relationship between values of N03 and NH4 ;, as well as DO values,

would help identify the subsurface mechanisms behind spatial and temporal N
cycling.
2.4.4. Limitations and Lessons
Subsurface environments are inherently difficult to study because of the
challenges involved with collecting data from an environment in which we have
limited access. To compensate for this complication, averages, interpolation, and
modeling are used to represent attributes of the many locations that cannot be
examined directly. The present study was based on others that utilize an array of
point measurements from which spatial variability for an entire study area is
partially defined, particularly O'Keefe & Edwards (2002) and Pinay et al. (2009).
Kennedy et al. (2008) explored the underlying question of error when averaging or
interpolating spatial distribution of subsurface attributes. Their results indicate that
a sampling density of 0.05 points per m 2 was sufficient to reduce error in reachaverage values. To produce a realistic image of spatial distribution, sampling
density of 0.08 to 0.09 points per m 2 was needed. Much of the uncertainty when
answering our study questions arises from limited spatial resolution, as the
sampling density for values of hydraulic head and conductivity were 0.011 points
per m 2 • Using Kennedy et al. (2008) as a guideline, 45 piezometers (0.05 points per
m 2 ) would be needed at our study site to obtain enough values for reach-average

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

87

estimates of subsurface properties and 72 piezometers (0.08 points per m 2 ) would be
required to map a realistic image of hyporheic flow.
Consequently, the Rhodamine-WT tracer test, which had a sampling density
of 0.028 points per m 2, only partially helped in designating the subset of study
piezometers. To improve on this design, I would have first and foremost performed
the topographic land survey and generated a contour map to identify likely
subsurface flow paths, which can be estimated by subtle changes in the surface
topography (Gordon et al. 1992, Hutchinson & Moore 2000). An in situ analysis of
hydraulic conductivity and porosity would have been helpful in estimating the time
required for the tracer to fall below detection limits. The temporal resolution is also
an important consideration and the shortage of sample containers during this
study's tracer test restricted the power of reliably detecting a breakthrough curve.
The most ideal experiment would involve data loggers capable of continuously
logging fluorescence.
The literature repeatedly reports "high variability" in describing the
heterogeneity of streambed and soil properties (Chestnut & McDowell2000, Hinkle
et al. 2001, Malard et al. 2002, Saunders et al. 2006, Zhang 2007, Claret & Boulton
2009, Chapin et al. 2011, Wondzell & Gooseff 2013), and this site is no different.
Johnson et al. (2014) points out that the residence time and removal of solutes in

transient storage zones is still being debated, largely due to the spatial and temporal
heterogeneity of hydraulic and biogeochemical processes in these storage zones.
Therefore, it is crucial for any researcher undertaking a study of the hyporheic zone
to design an experimental site with a high enough sampling density to reliably
address the study questions.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

88

The characterization of subsurface flows using hydraulic head measurements
from piezometers is not strictly correct (Hutchinson & Moore 2000). Hydraulic head
is a measure of pressure that varies with both horizontal location and vertical
position. Whereas, water table elevation, which can be evaluated using wells,
provides a more robust indicator of lateral flow direction Oones & Holmes 1996). To
improve on this design and produce a more accurate representation of subsurface
flows, a well could be installed next to each sample piezometer to quantify both
lateral and vertical hydraulic gradients.
Since the HFC was artificially-constructed, this site is an imperfect
comparison to a meander bend on the Horsefly River. In particular, riparian zone
flooding at the experimental site is uncommon, whereas flooding is an annual spring
event throughout much of the Horsefly River floodplain (Clark 2013). Flood events
cause the deposition of fine sediments, which would result in sandier substrates
along the Horsefly River. Aquifer variables would likely be characterized with
higher hydraulic conductivity and porosity when compared to the well-developed
riparian soils of the HFC. Accordingly, salmon nutrients may exhibit longer
hyporheic travel times and infiltrate to greater distances in the floodplain. However,
due to the large number of variables and interactions that influence nutrient flow
and transformation, this conclusion is merely speculation. For example, floodplains

can store considerable amounts of sediment and dissolved nutrients (Smith &
Owens 2014), so an influx of ammonium from salmon may be undetectable if the
floodplains are not N-limited.

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

89

2.5. Conclusions
A hydraulically responsive hyporheic zone was discovered beneath the
streambed and across a riparian meander bend at the Horsefly River spawning
channel. This zone was we11-connected, even during periods of low flow, and
responded quickly to changes in surface water level. Although NH4 +was present in
surface and hyporheic water samples throughout the study, we were unable to
correlate NH 4' concentration with the arrival of spawning salmon. This outcome
was likely the result of nutrient resuspension during channel cleaning, which
occurred prior to the first sampling event and created a profuse export of inorganic
N from the water column. Ammonium concentration in surface water samples were
significantly lower th an hyporheic water samples, with mean concentrations of 12.5
1-1g / L Nand 160 1-1g / L N, respectively. This study also found that hyporheic NH4 +
concentration decreased with increased distance from the streambank.
Overall, mean NH4 +concentration of surface and hyporheic water displayed a
significant increase from fall to spring. However, a significant decrease was
observed in the hyporheic water samples beneath the streambed. Ammonium
appears to accumulate beneath the streambed during salmon spawn and / or
throughout the summer, whereas N H4+in the riparian hyporheic zone accumulates
and stores over the w inter. This reversal of NH 4+ concentration between seasons
could indicate reciprocal subsidies and the presence of a nitrification-denitrification
coupling system on a larger time and spatial scale. During spring break-up,
hydraulic head in the rip arian to rose up to 10 em beneath the land surface,
providing water and nutrients to sh allow-rooting plants at the onset of the growing
season. This combination of snowmelt and p recipitation coincided with greater

Chapter 2. Nitrogen Cycling in the Hyporheic Zone

90

riparian NI-1 4+concentration and a diffuse export of hyporheic water into the surface
stream. These events occur around the time salmon fry emerge from the gravels,
delivering a pulse of limiting nutrients and organic matter to activate stream
microbial activity and supply energy for growth.
Anthropogenic sources of N from fertilizers and fossil fuel combustion persist
and downstream eutrophication is an ever-growing global problem (Seitzinger et al.
2006), which has the potential to harm salmon at all life stages. The primary

mechanism that permanently removes inorganic N is denitrification (Zametske et al.
2011 ). Since riparian and hyporheic zones are especially proficient at this process,

understanding the diverse interaction of conditions that facilitate denitrification is
critical to improving and protecting freshwater and marine environmental health.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

91

Chapter 3

Uptake, Storage, and Trophic Transfer of Marine-Derived Nutrients in the
Stream-Riparian Ecosystem
"When one tugs at a single thing in nature,
he finds it attached to the rest of the world." -John Muir

3.1 Stream-Riparian Food Webs
Salmon provide an isotopically-enriched marine-derived nutrient (MDN)
signature when they migrate to freshwater rivers and lakes, which can be traced
using stable isotopes of carbon and nitrogen (b 13C and b15 N) (Mathisen et al. 1988,
Kline et al. 1990, 1993, France 1994, 1995). Previous studies have estimated MDN
incorporation in freshwater and terrestrial organisms to be as much as nearly 75%
total N and over SO% total C (Bilby et al. 1996, Chaloner et al. 2002, Honea & Gara
2009). In interior British Columbia, this annual pulse of salmon nutrients occurs in
the fall (Albers & Petticrew 2012), providing a significant nutrient subsidy to
enhance stream productivity (Cedarholm et al. 1999, Gende et al. 2002, Naiman et al.
2002, Schindler et al. 2003, Janetski et al. 2009). However, certain types of organisms
may have a better advantage to benefit from MDN because of their trophic role, the
local ecology, and the timing of salmon spawn (Hicks et al. 2005, Lessard et al. 2009,
Wipfli & Baxter 2010, Rinella et al. 2012, 2013). Additionally, nutrient pulses are
carried through aquatic systems more quickly than terrestrial, so location within or
proximity to a stream may affect nutrient assimilation and retention (Ben-David et
al. 1998, Bilby et al. 2003, Nowlin et al. 2008).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

92

At the base of a stream food web, primary producers (i.e. algae, diatoms, and
microscopic phytoplankton) extract MDN from sediments and the water column
(Fischer 2003, Foreman & Covert 2003, Tiegs et al. 2011 ). Bacteria and fungi colonize
streambed substrates and salmon carcasses, and eventually entrap sediments and
other primary producers, such as algae, to form a nutritious mass of biofilms (Figure
3.1; Costerton et al. 1995, Wipfli et al. 1998, Schindler et al. 2003). Consumers,
including zooplankton and macroinvertebrates, obtain most of their energy from
primary producers and biofilms, and some may incorporate MDN through direct
consumption (Kline et al. 1997, Bilby et al. 1998, Minakawa & Gara 1999, Wipfli et al.
1999). Smaller consumers provide a rich and abundant food source for larger
consumers, such as fish and aquatic, or semi-aquatic, mammals, which also may
predate on spawning salmon or scavenge salmon carcasses (Cedarholm et al. 1989,
Wipfli et al. 1998, Woodward et al. 2005). Stream food webs, however, exhibit even
great complexity because they are imbedded within the surrounding terrestrial
landscape (Cummins 1974, Vannote et al. 1980, Ward et al. 1998, Piccolo & Wipfli
2002). Soil, leaf litter, terrestrial insects, birds, and predacious mammals may either
consume or be consumed by aquatic organisms (Nakano & Murakami 2001, Allan et
al. 2003, Foreman & Covert 2003, Allan & Castillo 2007, Christie et al. 2008).

Chapter 3. Marine-Derived Nutrient Dynamhs m tlze Stream-Riparian Ecosystem

93

Figure 3.1. A decaving sockeve salmon (0 nt rka) m the Horsefly River spawning
channel. Biofilms, filamentous algae, and leaf htter can be seen on the surface substrate.

The interconnectivity of stream-riparian food webs can be observed through
stable isotopic analysis of carbon and nitrogen, which is measured by the ratio of the
heavy to light isotope ( 13C: 12C and 1sN: 14 N) and expressed as a b-value in units of
parts per thousand ((,oo). Stable nitrogen isotopes (b 1"N) can indicate trophic position,
as consuming organisms accumulate approximately 3-4coo of the heavier isotope
with each successive trophic transfer (DeNiro & Epstein 1981, Minagawa & Wada
1984). Since stable carbon isotopes (b 13C) change relatively little with movement
through the food web, this ratio reflects an organism's diet and can be used to
identify basal energy sources, including salmon (DeNiro & Epstein 1978, Staal et al.
2007). Although not used in this study, stable sulfur isotopes (b 34S) can also be used
to track MDN assimilation in aquatic and riparian organisms. Sulfur isotopes can
provide useful insight into MDN enrichment because they are a conservative diet
tracer unaffected by confounding variations in trophic position (llesslein et al. 1991,
Barnes & Jennings 2007).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

94

The annual resource pulse of salmon nutrients can help support
macroinvertebrate populations (Minakawa et al. 2002, Lessard & Merritt 2006,
Verspoor et al. 2011), enrich resident fish populations and increase juvenile salmon
survivorship (Bilby et al. 1996, 1998, Wipfli et al. 2003, Rinella et al. 2012, Guyette et
al. 2014), increase nutrient storage in basal resources, like biofilms (Albers 2010,
Ruegg et al. 2011, Reisinger et al. 2013), and promote terrestrial plant growth
(Helfield & Naiman 2001, Koyama et al. 2005, Drake et al. 2006, Hocking &
Reimchen 2009). Evidence of MDN enrichment may even persist in streamside
vegetation and soils for years after salmon have left an area, known as a salmon
legacy effect (Naiman et al. 2002, Schindler et al. 2003, Koyama et al. 2005, Reisinger
et al. 2013).
3.1.1 Aquatic Food Webs
Macroinvertebrates, also referred to as aquatic insects or simply invertebrates
throughout this chapter, are good indicators of the function and health of a stream
ecosystem (Cummins 1974, Vannote et al. 1980). Each organism fulfills a particular
ecological role and provides a crucial intermediate link that aids the flow of energy
and nutrients between primary food sources and higher consumers (Malmqvist
2002, Woodward & Hildrew 2002). They are often categorized into the following
functional feeding groups (FFG) based on their mechanism for obtaining food and
the particle size of the food they consume: 1) shredders, 2) collector-gatherers, 3)
collector-filterers, 4) scrapers, and 5) predators (Thorp & Covich 1991, Merritt &
Cummins 1996). FFGs simplify the study of stream ecosystems by reducing a large
number of taxa into a few collective groups; however, it must be recognized that
most invertebrates feed opportunistically and cannot always be placed in a fixed

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

95

FFG (Anderson & Sedell 1979, Ward et al. 1998). Additionally, some food sources
contribute directly to all consumers, particularly biofilms which are a ubiquitous
assemblage of autotrophic and heterotrophic food sources (Allan & Castillo 2007,
Allen et al. 2009). Detritus represents another ubiquitous food source in a stream
that is composed of all non-living organic materials from terrestrial and aquatic
origins (Flecker 1984).

3.1.1.1 Shredders
Shredders are the dominant FFG in low-order, headwater streams and in
depositional microhabitats, such as pools, where the main food source is coarse
particulate organic matter (CPOM) (Vannote et al. 1980, Honea & Gara 2009). Their
functional significance is fundamental to freshwater food webs as they initiate
nutrient cycling processes by breaking down, or shredding, CPOM (Gra«;a 2001 ).
Shredders are thought to preferentially consume terrestrial leaf litter (Vannote et al.
1980, Cummins et al. 1989, Bilby et al. 1996), and studies often report the highest
abundance of shredders during fall (Hawkins & Sedell1981, Reece & Richardson
1998, Thompson 2007). However, they will supplement their diet with other
nutritious and available materials (Anderson & Sedell1979, Gra«;a 2001).
Amphipods, which were abundant in the HFC, are versatile and omnivorous
shredders that can adapt to seasonally shifting food sources, including terrestrial
detritus, algae, biofilms, macrophytes, other invertebrates, fish, salmon roe, and
carcasses (Brown & Diamond 1984, Mihuc & Mihuc 1995, MacNeil et al. 1997,
Friberg & Jacobsen 1999, Thompson 2007).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

96

3.1.1. 2 Collectors
By converting CPOM to fine particulate organic matter (FPOM), shredders
create a niche for collector-gatherers, which gather FPOM from the substrate, and
collector-filterers, which filter FPOM from the water column (Short & Maslin 1977).
As rivers grow larger and particles reduce in size, collecting organisms increasingly
become the dominant macroinvertebrate groups (Vannote et al. 1980). Their precise
diet is difficult to discern because they consume particles of everything that was
once living, whether from terrestrial, aquatic, or marine origins (Allan & Castillo
2007).

3.1.1.3 Scrapers
Scrapers, also known as grazers, are adapted to shear or scrape algae from the
surface of substrates and macrophytes (Lamberti et al. 1995). These organisms
become more prevalent in mid-sized rivers where community consumption and
respiration exceeds production via photosynthesis (Vannote et al. 1990). Due to their
mechanism of feeding, scrapers consume primary producers and typically hold the
lowest trophic position of macroinvertebrates (Anderson & Cabana 2007, Rinella et
al. 2013). However, their food source is highly abundant and scraper populations
can often coexist even though they have been known to consume over 90% of the
plant biomass (Feminella & Resh 1991, Lamberti et al. 1995).

3.1.1.4 Predators
Pred atory invertebrates derive their energy from live animal tissue and are
often size selective, choosing prey smaller than themselves to consume (Hawkins &
Sedell1981, Gra~a 2001). The density of predators is not altered according to stream

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

97

order, but instead is limited by prey availability (Vannote et al. 1990, Honea & Gara
2009).

3.1.2 Terrestrial Food Webs
In a stream-riparian ecosystem, allochthonous nutrient transfer to streams
from the surrounding terrestrial environment has been well established (Vannote et
al. 1980, Bilby & Bisson 1992, Polis & Hurd 1996, Naiman & Decamps 1997).
However, nutrient fluxes also move in the opposite direction, from the stream to
riparian, through flooding, advective hyporheic water transport, faunal delivery,
and as emergent aquatic insects (Kennedy 1950, Cederholm et al. 1989, Kielland et
al. 1997, O'Keefe & Edwards 2002, Ballinger & Lake 2006, Drake et al. 2006). The
presence and ecological importance of nutrients transported from anadromous
salmon to the riparian zone has been demonstrated through stable isotope anaysis of
riparian biota (Bilby et al. 1996, Helfield & Naiman 2001, Naiman et al. 2002, Willson
et al. 2004, Hicks et al. 2005, and see Chapter 1).
3.1.2.1 Willows

Willow trees, Salix sp., are common annual species in riparian zones, with
deep roots that can penetrate the water table (Pinay et al. 2009). They rely mostly on
groundwater during the growing season and are significantly limited by N
availability (Menezes 2009). Their roots tend to exhibit increased growth toward the
water table (Peter et al. 2012), with an average maximum rooting depth of 90 em.
Isotopic enrichment of Salix sp. is variable between studies, with some citing
enrichment due to spawning salmon (Bartz & Naiman 2005, Koshino et al. 2013) and
others documenting a lack of MDN assimilation (Hicks et al. 2005).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

98

3.1.2.2 Horsetail

Horsetail, or Equisetum sp., has a lineage dating back to the 20-m tall
Pseudobornia of the Triassic period, which also colonized streamside habitats, and
may be the oldest living genus of vascular plants (Hauke 1963, Husby 2013). They
have remained genetically isolated since the Jurassic period and retain ancient
features of rhizomatous colonial growth and silicon-dependence (Channing et al.
2011). Today, horsetails have subsurface rhizomes that extend horizontally for long
distances and can penetrate more than 1 m below the surface (Mitich 1992, Hodson
et al. 2009). Marsh et al. (2000) found that the roots and rhizomes of Equisetum sp.
concentrate in the upper soil C horizon, suggesting that this is where most of its
nutrients are acquired. Their extensive rhizome system allows them to assimilate
nutrients and sunlight from a wide-reaching area.
3.1.2.3 Dandelions

The common dandelion, Taraxacum officinale, has been a pervasive invasive
species in North America for around 1000 years and colonizes a wide range of
habitats, particularly wet, disturbed areas (Vavrek et al. 1997). They possess a
highly-regenerative tap root that penetrates deep into the soil to reach moisture and
nutrients beyond the limit of many other herbaceous species Oackson 1982). Tap
roots of 10 to 15 em in length are common, but they are capable of growing up to 1
or 2m (Stewart-Wade et al. 2002). Taraxacum sp. are commonly used in ecological
studies due to their ubiquitous distribution across temperate climates. They possess
an ability to concentrate a wide array of anthropogenic pollutants from the soil and
are often utilized to monitor metal contamination in an area (Kabata-Pendias &
Dudka 1991, Collier et al. 2010, Gworek et al. 2011, Maleci et al. 2014).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

99

3.1.2.4 Soil

Hyporheic and riparian soil is a porous medium through which water can
flow (Triska et al. 1989, Harvey & Wagner 2000). Surface water enters and moves
through these soils, delivering nutrients, dissolved organic matter (DOM), and
oxygen that have traveled downstream or have been laterally displaced from the
surrounding land (Allan & Castillo 2007). This infiltration supports a diverse
population of millions of microorganisms that live within soil, perpetually
transforming organic and inorganic compounds (Lowell et al. 2009, Febria et al.
2011 ). Numerous combinations of sources, processes, and products makes
understanding the mechanisms behind Nand C cycling challenging, but large scale
patterns exist among the variation (Foreman & Covert 2003, Billings & Richter 2006,
Borch et al. 2010). Richter et al. (2000) conducted a 40-year forest development
assessment and discovered two overwhelming conclusions about soil Nand C
budgets of anN-limiting system. First, there was a massive net transfer of N from
mineral soils to the top organic soil layer and forest biomass, meanwhile N loss from
the ecosystem was negligible. Second, total organic C content in mineral soils
remained relatively unchanged, whereas the organic soil layer accumulated
substantial amounts of C each year.

3.1.2. Hypotheses

This study aims to investigate MDN assimilation, trophic transport, and
storage in a stream-riparian food web by tracing b13C and b15 N in stream
macroinvertebrates and riparian vegetation and soil during the 2011 sockeye salmon
spawn and subsequent spring. In British Columbia, salmon spawn occurs in the fall

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

100

when many plants are senescing or becoming dormant. Despite an apparent contrast
in run timing and tree growth, biological uptake of MDN can occur quickly and be
stored within riparian soil and plants (Drake et al. 2006) and streams (Bilby et al.
1996). There are still questions, however, especially in regard to the environmental
factors that limit uptake and persistence of MDN in individual species and as the
riparian forest as a whole (Drake et al. 2011, Devries 2012, Rinella et al. 2013).
For this study, macroinvertebrates were evaluated according to their
functional feeding group to determine if species type, abundance, and stable isotopic
concentration changes during the MDN pulse and between seasons. Three riparian
plant species with differing life strategies were investigated to determine the
pathway, uptake, and retention of MDN in the riparian zone adjacent to the salmon
spawning channel. The following hypotheses were developed to evaluate the aim
listed above:

Null Hypothesis V: There will be no difference between stable isotopic signatures
(b 13C and b15 N) of rnacroinvertebrate taxa throughout salmon
spawn and between fall and spring.
Null Hypothesis VI: Macroinvertebrate b15N signatures will not be correlated with
NH4 +concentration of surface and hyporheic water.
Null Hypothesis VII: There will be no difference between stable isotopic signatures
(b 13C and b15 N) of riparian vegetation between seasons.
Null Hypothesis VIII: Stable isotopic signatures (b 13C and b15 N) of riparian
vegetation and soil will not be correlated with distance from
the streambank.
Null Hypothesis IX: Riparian vegetation b15 N signatures will not be correlated with
NH4+concentration of riparian hyporheic water.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

101

3.2 Methods
3.2.1 Field Sampling and Laboratory Analysi
Macroinvertebrates and salmon carcasses were sampled from the Horsefly
River spawning channel (HFC) in Horsefly, British Columbia (Figure 2.1).
Vegetation and soil were sampled from a riparian meander bend adjacent to the
channel. The spawning channel was artificially constructed in 1989 to enhance
salmon habitat; therefore, development of the channel bed and bank, as well as nearstream soil, is not natural. A meander bend located on the Horsefly River would
likely be composed of sandier material. Still, the HFC remains hydrologically
connected to the Horsefly River throughout the year, providing a year-round, offchannel habitat for stream and riparian organisms. The dominant microhabitat for
most of the channel was a glide, with occasional pools representing the nondominant habitat. The 625-m2 experimental site encompasses the channel and
riparian zone of a single meander bend (Figure 2.4). The stream habitat here was a
uniform glide, with a fluctuating depth of approximately 0.5 to 1.0 m between low
and high water events. The substrate consisted of a roughly 1-m deep gravel layer
with uniform particle size and shape designed for ideal spawning conditions. The
majority of the riparian zone was a forest comprised of predominantly deciduous
trees and shrubs with a diameter at breast height of less than 0.3 m. At the upstream
portion of the study site, the forest becomes a grassy area with small shrubs, which
was likely disturbed by the building of a temporary road to the channel and is
currently in an early successional stage. The sampling sites located within this
grassy area were piezmneters F and G, soil pit 1, and vegetation samples from 0 and

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

102

10 m on Transect 1 and 0 m on Transect 2. Terrestrial MDN deposits within the
study site, as well as observed bear activity, were minimal; a total of five salmon
carcasses and two piles of bear feces were recorded during the salmon spawn.
Additional details about the study site and design can be found in Section 2.2.1.

3.2.1.1 Macroinvertebrates
Benthic macroinvertebrate samples were collected during each study week in
Fall2011 (n=5) and once after break-up in May 2012 (Figure 2.6) with a Surber
sampler (500-,um mesh size). Sampling prior to spring break-up was not possible
due to ice cover. While avoiding salmon redds, three random samples were taken at
a cross-section about 5 m downstream from the study site at river right, mid-stream,
and river left by disturbing a 0.093-m 2 area of the gravel bed for 1.5 minutes to a
depth of 15 em. Precise locations were notre-sampled between weeks; therefore
recolonization of a sampling area was not measured. Sampling procedures were
designed to quantify benthic macroinvertebrates residing on and within streambed
gravels. The samples were preserved with 70% ethanol (Flecker 1984), and most
extraneous materials (i.e. substrate, algae, leaf litter) were kept in the sample to
avoid accidentally discarding invertebrates, besides for large rocks capable of
crushing the sample. Although Surber samplers are designed for use in water <30
em deep (Peckarsky 1984), the stream was shallow enough for the sampler to
function as a kicknet for this study.
In the laboratory, macroinvertebrates were identified to the lowest practical
taxonomical level (either genus or species) using An Introduction to the Aquatic Insects

of North America (Merritt & Cummins 1996). Each taxon was assigned to one of the
following functional feeding groups (FFG): collector-gatherers, collector-filters,

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

103

scrapers, shredders, or predators. The family Chironomidae, however, includes
species from two FFGs and was not identified further than family level because
accurate identification required equipment and expertise unavailable for this study.
onsequently, chironomid FFG assignment was generalized as 90% collectorgatherers and 10% predators, in accordance with Merritt & Cummins (1996).
After identification, individuals were sorted into taxa groups, dried to a
constant mass (48 hat 75° C in a drying oven), and weighed on a microbalance
(resolution of 10-7 g). The total mass of a group was used to calculate taxa biomass
(g/m 3) per sampling week (Mason et al. 1983).

3.2.1.2 Riparian Vegetation
Vegetation samples were collected during three sampling weeks: pre-spawn
(August 2011), post-spawn (October 2011), and after spring break-up (May 2012).
This sampling schedule was selected to characterize nutrient uptake during the
salmon spawn and after the winter. A total of six samples were taken along two
transects at 0 m, 10m, and 20m. Transect 1 runs perpendicular to the stream at the
top of the meander bend. Transect 2 runs parallel to the stream across the meander
bend (Figure 2.4). From here on, locations at Transect 1 are referred to as T1-0, T1-10,
and T1-20. Locations at Transect 2 are T2-0, T2-10, and T2-20. Three species were
collected at each location: Common horsetail (Equisetum arvense ), Pacific willow

(Salix Iucida), and Common dandelion (Taraxacum officinale). These species were
chosen to represent different rooting depths and root structure. After collection,
samples were transported to a freezer within 4 hours and remained frozen until
laboratory analysis.
To obtain information of annual tree growth, three tree cores were extracted

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

104

using an increment borer from a willow tree located centrally on the riparian
meander bend. Cores were analyzed by UNBC's Tree Ring Lab following standard
dendrochronological procedures. After mounting and sanding the tree cores, annual
rings were measured to 0.001 mm using a Velmex measuring system (Stokes &
Smiley 1968). Raw ring widths were averaged between cores and mean
standardized chronologies were developed using the program ARSTAN to produce
an index value (Veblen et al. 1991).

3.2.1.3 Additional Sampling
Soil samples were collected in May 2012 at three riparian locations at the
study site (Figure 2.4). Soil pits were excavated to approximately 1-m deep and soil
profiles were described by identifying horizons A, B, and C and measuring horizon
widths (Figure 2.10). A soil horizon is a layer that is distinctly distinguishable from
its adjacent layers. Certain soil characteristics, such as particle shape and color, were
recorded using a Munsell® color chart. Porosity measurements performed in the
laboratory also aided in describing the geologic material (refer to Section 2.2.3).
Samples were extracted from each horizon, with additional samples collected where
clay mottling was present, and frozen within 4 hours of collection. Soil sampling
methods generally followed those of Schoeneberger et al. (2002).
During the 2011 salmon spawn, samples of freshly dead salmon tissue and
bear feces were collected from the study site as baseline salmon nutrient data.
Samples were placed in a freezer within 4-hours of collection.

3.2.1.4 Stable Isotope Analysis
In preparation for stable isotope analysis, macroinvertebrates, vegetation, and

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

105

soil samples were dried to a constant mass in a drying oven (for 48 hours at 75° )
and samples of salmon tissue and bear feces were freeze-dried for 48 hours. Dried
samples were crushed using a mortar and pestle and weighed into tin capsules (at a
resolution of 10-7 g) at the University of Northern British Columbia's Central
Equipment Laboratory. These samples were then shipped to the University of
California-Davis Stable Isotope Facility for analysis of carbon and nitrogen stable
isotopic signatures (b 13C and b15 N) using a continuous flow-isotope ratio mass
spectrometer (dual-inlet Europa 20 / 20, PDZ Europa, Sercon Ltd., Cheshire, UK.).
Isotope enrichment (reported in parts per thousand o/oo) was determined as follows
(McConnachie & Petti crew 2006 ):

%o b 13C or b15N = [ (Rsample- Rstandard) I R standard ] * 1000

Eqn (3.1)

where R is the ratio of the heavy isotope to the light isotope. A more positive b value
is isotopically enriched, which means that the sample contains proportionally more
of the heavy stable isotope. Final values are expressed relative to internationally
accepted standards of V-PDB (Vienna PeeDee Belemnite) for carbon and Air for
nitrogen (Bilby et al. 1996). For quality assurance and control, about 10% of the
samples were randomly split into duplicates prior to submission to UC-Davis. The
mean relative percent difference (RPD) in molar C:N ratios of duplicated
invertebrate and vegetation samples was 1.9% . Mean RPD for duplicated soil

samples was 29.3%. This significant difference highlights the high variability of soil
over short spaces (Freeze & Cherry 1979). Consequently, all samples were retained
for data analysis and values from duplicate samples were averaged.
3.2.2 Statistical Analysis
Dual isotope plots were created to visualize mean b13C and b15 N signatures of

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

106

sampled organisms and materials for Fall2011 and Spring 2012 by using the
package "ggplot2" in R (Wickham 2009). The following statistical analyses were
performed in R using the packages "stats", "nlme", and / or "base".

3.2.2.1 Aquatic Samples
In order to examine temporal trends in macroinvertebrate isotopic signatures,
autocorrelation tests were conducted using R's built-in autocorrelation function
[acf()] to determine if b13C and b 15 N concentration of each species was statistically
independent between location (river right, mid-stream, and river left) and sampling
week (Chatfield 2004). Dependence was discovered between sampling locations, so
the three locations were averaged by taxa per week. All data were evaluated for a
normal distribution by observing the histogram and quantile-quantile plot (Q-Q
plot).
Only three insect orders (Amphipoda, Diptera, and Ephemeroptera) had
sufficient biomass during both seasons to perform temporal analysis on stable
isotopic signatures of b13C and b15 N. Chironomidae was the only family present in
the order Diptera, so this group will simply be referred to as chironomids from here
on. Linear regressions were conducted to analyze MDN concentration between
sampling week and season. Multiple and linear regressions of MDN were analyzed
against predictor variables of mean surface water NH 4+concentration, mean
hyporheic water NH/ concentration (from piezometers C and D), number of
salmon, insect biomass, number of individuals per order, and surface water level
that corresponded with each sampling period. Before analysis, all variables were
standardized using Equation 2.5 to eliminate discrepancies between units of
measurement and enable direct interpretation. Regression output results of p-value,

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

107

R-squared, and effect size were used to interpret the significance, goodness-of-fit,
and strength of a particular model. The confidence level for statistical significance
was set top~ 0.05. Additionally, highly significant results were identified if p ~
0.001.
3.2.2.2 Diversity and Evenness of Macroinvertebrate Communities

To better understand how the macroinvertebrate community of the HFC
compares to that of the natural river, diversity and evenness calculations were
evaluated with samples from the Horsefly River, taken by the B.C. Ministry of
Environment on Oct. 4, 2011. These indices are helpful in understanding
environmental health because a community that is highly uneven or has low
biological diversity is less resistant to environmental stress (Peet 1974, Wittebolle et
al. 2009). To assess community diversity, two well-known heterogeneity indices
were computed: Shannon Diversity Index and Simpson Index. These measurements
are based upon a combination of species richness and heterogeneity (Peet 1974). The
Shannon Index equation, which is more sensitive to community changes in rare
species, is computed as follows:

-

n·

n·

(-t * ln(-t))
. N
N
t

L

Eqn (3.2)

The Simpson Index, which is more sensitive to changes in the most abundant
species, is calculated as:

I:i ni(ni - 1)
N(N- 1)
Eqn (3.3)

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

108

where:
n; =individuals of one species; and,
N =total number of individuals.
In order to define these diversity indices in terms of number of species, the
effective number of species (ENS) was computed. This measurement describes
communities as the number of equally-common species which would produce the
same heterogeneity as the sample (Peet 1974, Jost 2006).
ENS= exp(H)

Eqn (3.4)

where:
H = Shannon Index.
Finally, a Lorenz curve was graphed to visualize population evenness of a
community. The Lorenz graph, originally developed to measure income inequality
(Lorenz 1905), displays a straight 45 degree line to represent perfect equality in a
population and the Lorenz curve, which shows population evenness, or inequality
(Figure 3.2). The Gini coefficient was calculated from the Lorenz graph, which is a
single value (ranging from zero to one) used to describe evenness (Wittebolle et al.
2009).
The sampling methods employed by the Ministry of Environment 0. McLeod,
pers. comm.) differed from this study in several ways, and, therefore, the results can
only be used as a broad comparison between the two environments. They
performed two kick samples for a 6-min duration by walking from one bank to the
other. Since our method was designed to capture invertebrates residing in the
hyporheic zone, samples were collected at three locations for 1.5 minutes each to
allow for collection 15 em beneath the sediment surface. The microhabitat they

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

109

sampled was a riffle, whereas our location was a glide. However, since these indices
are only based on the number of individuals and species, and do not factor common
species among locations, the microhabitat type is not a variable that affects
individual diversity and evenness values. The Ministry used a net with a mesh size
of 400 1Jm, as compared to ours at 500 1Jm, so our results may not include
invertebrates between 400 and 500 1Jm. Due to these differences, I would expect
samples from the Horsefly River to contain a larger total number of invertebrates
because of a longer sampling duration and smaller mesh size. I would also expect
the spawning channel samples to include a greater abundance of chironomids, as
the study design aimed to capture deeper-residing organisms.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

110

cu
u

A

~
c

""0
j

-"'
..0

~0.

0

<(,JY

c

0

·e

tt
e

&.

~~

0
'-

a..

~e

c.i'
f§'"'

v

o'

Cumulative proportion of species

cu

B

>
·a

"'

"'3
E
j

u

100%

Figure 3.2. A typical Lorenz curve which graphically represents the inequality in a
population. The areas above and below the Lorenz curve, marked A and B, respectively,
simplifies the Gini coefficient equation, which is A / (A+B).

3.2.2.3 Riparian Samples
In order to examine temporal trends in vegetation isotopic signatures,
autocorrelation tests were conducted using R's built-in autocorrelation function
[acf()] to determine if b13C and b15 N concentration of each species was statistically
independent between the three sampling periods (Chatfield 2004). Since
independence was identified, the b13C and b15 N signatures across the six sampling
locations were averaged between each species within a sampling period. All data
were evaluated for a normal distribution by observing the histogram and quantilequantile plot {Q-Q plot).
Linear least squares regressions were performed to test the hypothesis that

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

111

vegetation isotope ratios will be higher with proximity to the stream. Model output
results of p-value, R-squared, and effect size were used to interpret the significance,
goodness-of-fit, and strength of a particular model. Additional predictor variables,
including distance along transect, sample location, elevation, and rooting depth,
were also applied to linear regression models in order to analyze other factors that
may affect b13C and b15 N signatures.
Stable isotopic signatures of soil samples were also analyzed for a normal
distribution. Subsequent linear least squares regressions of b 13C and b15 N signatures
against predictor variables of depth and distance from streambank were performed.
Since the surface elevation varied at each soil pit, depth measurements were
standardized to meters above sea level (mASL).
To test the relationship between vegetation isotopic signatures and NH 4 +
concentration of hyporheic water, linear regressions were performed for the Prespawn, Post-spawn, and Spring periods. Isotopic signatures from Transect 2 were
paired with NH/ concentration from nearby piezometers. Vegetation samples from
T2-0 were tested against the average of NH4 +concentrations from sites F and G,
which were both within 2m of the vegetation samples. T2-10 isotopes were paired
with NH4 +from K, which was also 2m away. T2-20 isotopic signatures were
matched with hyporheic site 0, at a distance of 3 m from the vegetation samples.
Stable isotopic signatures and NH4 + concentration were standardized before
statistical analysis (Equation 2.5).
Statistical significance between spring vegetation and soil isotopes was tested
through Pearson's product-moment correlations and linear regressions. The
confidence level for statistical significance was set to p ~ 0.05. Additionally, highly

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

112

significant results were identified if p ~ 0.001. For each soil pit, isotopic signatures
from horizon A, B, and C were tested against vegetation isotopic signatures, with
the notion that S. Iucida isotopes will most closely correlate with horizon C and E.

arvense and T. officinale isotopic signatures may reflect horizon A or B because of
their shallower rooting depth. Each soil pit was paired with the closest vegetation
transect: Soill and T2-0, Soil 2 and Tl-20, and Soil 3 and T2-20. The relationship
between soil isotopes and hyporheic water NH4 t concentration could not be tested
because there were not enough replicate soil observations collected.

3.3 Results
3.3.1 Temporal Trends of MDN
Dual isotope plots of b 13C and b 15 N visually display the stream-riparian food
web components at this site and provide an estimate of trophic levels (Figures 3.3
and 3.4). Predacious organisms of higher trophic level are located in the upper right
corner of the graph (higher b 13C and b 15 N) and autotrophs are located in the lower
left (lower b 13C and b15 N). In the fall, the predacious subclass Hirudinea were the
most enriched with b15N compared to all macroinvertebrate species. Riparian
vegetation displays a lower (more negative) b 13C value than the invertebrate groups.
Vegetation b15N signatures vary widely between species, with E. arvense (Equisetum)
exhibiting levels of b15 N closer associated to those of invertebrates and salmon
tissue. In the spring, macroinvertebrates remain enriched in b 13C relative to riparian
vegetation.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

12

113

Macroinvertebrates

Salrnon Ttssue
8

Hiruni<Jae
Amphipods

Ephemeroptera

....z

-

.0

Charonom1ds

- - EqUtsetum

Vegetation

Salix

•

Taraxacum

Bear Feces

11'-== +

0

-35

-30

-25

Figure 3.3. Dual isotope plot (b 13C %o and b15 N %o) from all organisms and materials
colJected during the FalJ 2011 salmon run. Isotopically enriched species are located in the
upper right comer. Mean ± 1 SEM.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

114

12

Salmon Tissue
Physa

8

AmphtpodS

Ephemeroptera

...-.cz

-

ChironomtdS

-

Chtronomed pupa

Soil
-

Vegetation

EQuiSetum

-salix

Taraxacum

0

30

-35

-25

·20

Figure 3.4. Dual isotope plot (b13C %o and b15 N o/oo) from all organisms and materials
collected in Spring 2012, besides for salmon tissue, collected in Fall2011, which is
included as reference point. Mean ± 1 SEM.

The isotopic signatures of species that were collected in both the fall and
spring are displayed in Figure 3.5. The invertebrates Ephemeroptera and
Chironomids exhibited clear differences in b15 N and b13C concentrations between

seasons, with both groups displaying a significant decrease in spring b15N
signatures. T. officinale (Taraxacu m) was the only group with a higher overall b15N
signature in the spring.
Regression analyses determined significant changes in isotopic signatures
between seasons. pring b15 N signatures were significantly lower for Equisetum (p <

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

115

0.05), Chironomids (p < 0.001), and Ephemeroptera (p < 0.001). Conversely, spring
b15N signatures for Taraxacum were significantly higher (p < 0.05). Signatures of b13C
were significantly lower in spring for Ephemeroptera (p < 0.05), but higher for
Chironomids (p < 0.001); Table 3.1).

Ephemeroptera

~
.

8-

I

!,/

Am phi pod~

Equasetum

~
l.

z

-

Ill

-.o
4-

Season

Chironomids

0

+

Taraxacum

.

.

Fall

b. Spring

-..
Salix

o-

T

T

T

1

-35 0

-32.5

-30.0

-27.5

'

-25.0

r

-22 5

Figure 3.5. Temporal differences of vegetation and macroinvertebrate species collected in
the fall and spring sampling periods. Units of b13C and b15N are parts per thousand (roo).
Mean± 1 SEM.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

116

Table 3.1. Statistically significant differences observed in organisms between the fall and
spring sampling periods.

Isotope

Group

p-value

Effect size R-squared

51~

Eqursetum

0.041

-3.843

0.284

51~

Taraxacum

0.015

5.083

0.319

515N

Chironomidae

7.2e-09

-4.242

0.969

51~

Ephemeroptera

1.2e-04

-2.3n

0.959

513c

Chlronomldae

0.001

1.987

0.663

613c

Ephemeroptera

0.009

- 1.003

0.775

"' Bold indicates an increase in isotope value from fall to spring

3.3.2 Macroin vertebrate Assimilation of M D N

Macroinvertebrate groups Chironomidae and Ephemeroptera had a highly
statistically significant association between b15 N signatures and both surface water
NH/ and hyporheic water NH/. Regression analysis showed that hyporheic NH/
concentration increased with b 15 N signatures for both groups (p < 0.001).
Simultaneously, b15N concentration of these species decreased when surface NH4 +
increased (p < 0.001). There were no significant relationships detected between
invertebrate isotopic signatures and the other tested predictor variables.
3.3.3 Macroin vertebrate Diversity and Abundance

Throughout the study, 13 different taxa from four phyla (Annelida,
Arthropoda, Cnidaria, and Mollusca) were identified (Table 3.2). A cumulative total
of 3,213 individual macroinvertebrates were enumerated and grouped into one of
five functional feeding groups (FFG). Graphs of invertebrate biomass show that the
majority of the community's biomass during every sampling week were shredders

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

117

(Figure 3.6). The shredder FFG was entirely composed of the species Hyalella azteca
(referred to as amphipods for convenience). Biomass data are based on the mass of
invertebrate groups; therefore the graphs of relative abundance (number of
individuals in a group / total number of individuals) demonstrate a different ratio
(Figure 3.7}. These graphs show shredders as a dominant group, but they are
outnumbered by collector-gatherers during some weeks, particularly in the fall. The
collector-gatherers group contained taxa from the family Chironomidae, the subclass
Oligochaeta, and the orders Ephemeroptera and Tricoptera, but consisted primarily
of Chironomidae. When observing the taxon most commonly identified throughout
the study, amphipods and chironomids were discernibly the most abundant species
(Figure 3.8).
According to the Shannon Index and its related function, effective number of
species (ENS), the Horsefly River has a greater diversity of species than the
spawning channel (Table 3.3). However, results from the Simpson's Index and Gini
coefficient indicate that the community in the spawning channel has greater species
equality, or a similar number of individuals per species group (Table 3.3).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

118

Table 3.2. Taxonomic identification and enumeration of macroinvertebrates in the
Horsefly River spawning channel during the 2011 salmon spawn and the subsequent
spring (labeled as Post-winter). The scientific classification of major taxonomic groups
are identified in parentheses. Functional feeding group (FFG) classifications are as
follows: CF (collector-filterers), CG (collector-gatherers), P (predators), SC (scrapers), and
SH (shredders).
Af:dve

Post-spawn

135

31;5

452

72

43

169

259

-105

2

4

5

8

304
61
215

I

89

56

123

2

16

20

28

63
II

3

3

18
I

22
I

6

8

FFG Pre-spawn

Hyakno aztcca

SH

30

181

Diptera (order)
Chi ronom1 dae
Ch1ronom dae pupa
Pupa casing

CG
CG

13

CG
P

Ephemeroptera (order)
Lept.of*Jebiido!
parorepte>/)hlel»a sp.
cG

Clttellata (class)
Ol1gochaeta
Hlrud nea

Bot!odoe
untd enttfiabl e

CG

Plecoptera (order)

P

Tricoptera (order)
Hydroptifidae axyechka sp.
Hydropsymidoe
arctapsyme sp.

3

CG

10
II
14

SC

unto enttfiable

8

Hydrozoa (class)
Hydro sp.

P

Toa.l

2

IS
7

8
12

10
10

16

7
7

12

II

6

5

Bivalvia (class}
CF

79

2

CF

Sp#loeriida-e

Winter

2

sc
Promet~evn um.bh'icate»us

Post-

~n

Taxon

7

59

373

430

855

4

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

119

30

C/)

~
E 20o

Functional Feeding Group
Collector-Filterer5

CD
~

0

Collector-Gatherers

0

Predators

t-

G>

~
'E
~

Scrapers

·o -

Shredders

~

~

o-

r

Pre-spawn

I

Active
spawn

Actrve
spawn II

Actrve
Post-spawn
spawn I I

Spring

Figure 3.6. Bars displaying the percent biomas~ of each FFG during each sampling
period. The height of the bars represents weekly biomass as a percentage of the total
biomass. The actual weight of amphipods varied from a minimum of 50.2 mg during Prespawn and a maximum of 682.9 mg during Active Spawn II I.

Chapter 3. Marine-Derived Nutrient Dynamics in tile Stream-Riparian Ecosystem
Active Spawn I

Act,ve Spawn I

Pre spawn

AcllvP Sp

Figure 3.7. Pie charts displaying the relative abundance (as a percentage) of each FFG
during each sampling period.

400-

Taxon
AmphiPoda
Ch ronom1dae

-

Ephemeroptera

0

.... 200-

-

CJ)

Gastropoda

.0

E
~

z

Oligoctlaeta

00-

oI

I

Pre-spawn

Act ve
spawn I

ActiYe

spawn II

I

I

Active
spawn II

Post-spawn

Spnng

Figure 3.8. The number of individuals during each sampling week for commonh
identified taxa at the HFC.

120

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

121

Table 3.3. Shannon Index and Effective Number of Species (ENS) diversity values, along
with Simpson's Index and Cini coefficient evenness" a lues for the Horsefly River
spawning channel and the Horsefly River.
Diversity

Spawning channel
Post -spawn
Spawning channel
Fall & Spring
Horsefly River
Post-spawn

Even ness
Gini
Simpson's
Index coefficient

Shan non
Index

ENS

1.1 33

3.1

0.789

0.745

1.364

3.9

0 646

0.763

1.938

6.9

0.562

0.794

3.3.4 Spatial Trends of MDN in the Riparian Zone
S. Iucida (Salix) isotopic signatures of both b 13C and () h N displayed a highly
significant decrease as the distance from the streambank increased (p < 0.001; Table
3.4). According to linear regression analysis, with every 1-m increment further away
from the stream, b13C signatures decreased by 0.177~ ·oc and b 1 N signatures decreased
by 0.242( oo (Figure 3.9). Transect distance also displayed significance (p < 0.001 (b 13C)
and p < 0.05 (b 1'i N)). When transects Tl and T2 were tested separately there was no
significance detected between b 1'iN and T2, indicating that distance from streambank
is the primary variable when determining b1'iN concentration.
Regardless of season, E. arvense b 15 N signatures were found to be uniform
across the study site. However, statiscially significant b 13C trends from linear
regressions detected that with every 1-m increase from the stream bank values
decreased by 0.085( ·oo (p < 0.05; Table 3.4).

T. officinale isotopes were generally consistent throughout the stud) site and
no clear trends in isotope concentration were noted with distance from the stream.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

122

Rooting depth was not a significant variable in predicting isotope concentration of
vegetation species, nor were corresponding soil isotopes or hyporheic Nll4
concentration. Therefore, this study fails to reject the null hypothesis (IX) that there
is not a correlation between riparian vegetation b 1 N signatures and riparian
hyporheic NH 4 concentration
Overall, vegetation h 13C signatures were highest at site Tl-0, which was
located directly on the streambank (Table 3.5). T1-0 exhibited significantly greater bnc
signatures than sites T2-10 and T2-20 for all species. S. Iucida b 13C at T1-0 was
significantly higher than all sites (except T2-0), and E. arvense b 13C was greater than
all locations (besides T1-1 0).
Table 3.4. Variables that displa\ statistical significance in relation to terrestrial isotopic
signatures.

Isotope

Group

Variable

p-value

Effect size R-squared

615N

Salix

Distance from
Streambank

0.001

-0.242

0.457

613c

Salix

Distance from
Stream bank

I.Oe-04

-0.178

0.529

613c

Equisetum

Distance from
Stream bank

0.013

-0.085

0.369

613c

Soil

Depth

0.003

-1.834

0.5+1

Chapter 3. Marine-Derived Nutrient Dynamtcs m the Stream-Riparian Ecosystem

123

Distance from

Str&ambank{m)
15

10

5

Figure 3.9. Dual isotope plot of willow (S. Iucida) leaves. Each cross represents the
average of all sampling periods at each sampling location. Units of b 13C and b N are
parts per thousand (C ·oc ). Mean ±- 1 SEM.
Table 3.5. P-va lues disp laying significant di fferences in b 13C at all vegetation sites
compared to site Tl-0, which was located directly on the streambank. Isotopic signatures
were averaged between each sampling period.

All species

Tl-10
Tl-20

Salix

Equisetum Taraxacum

<0.05
<0.05

<0.001

TI-0

<0.05
<0.05

TI-10

<0.05

<0.05

<0.05

<0.001

TI-20

<0.05

<0.05

<0.05

<0.001

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Ripatian Ecosystem

124

There was a significant interaction between soil signatures of b 13C and depth,
with b13C increasing by 1.834o/oo for every meter deeper below the surface (Figure
3.10, Table 3.4). Isotopic signatures of soil b15 N were statistically uniform throughout

the study site regardless of depth, but a closer look at soil pit 1, which was located
closest to the streambank, reveals a positive trend with depth and both b13C and b 15 N
(Figure 3.11). There was no statistical significance detected with soil b15 N and
distance from streambank thereby supporting a failure to reject this component of
the null hypothesis (VIII).

a

..

-25-

.,.,<.>

•

-26-

•

-.e
-27-

•
I

I

A

B

Horizon

I

c

Figure 3.10. Box plots displaying b13C (%o) of soil horizons from three soil pits across the
study site. Black dots represent actual b13C signatures.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

125

•c

Soil pit 1

s o-

55-

B

...

.7-oQ

50-

4 5-

A
40

•
T

J

-27

·25

Figure 3.11. Stable isotopic signatures of b13C (roo) and b15 N (roo) at soil pit 1, which was
located closest to the streambank.

3.3.5 Tree Cores

Salix sp. tree cores were dated back to the year 1956. The spawning channel
was constructed in 1989 and ring width analysis indicates that substantial tree
growth began in 1996, with a record growth year in 1998 (Figure 3.12). A multiple
regression analysis of ring width indices against salmon escapement in the
spawning channel and total annual precipitation revealed that escapement and
precipitation on their own do not predict ring width; however the interaction
between these terms produces a significant regression (p < 0.05). When graphing
standardized ring widths and salmon abundance (Figure 3.13), a particularly
synch ronous relationship is evident between 1997-2000, when salmon population
was zero in 1997 and escalates to - 25,000 the next year. Another distinct peak in ring

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

126

width and escapement occurred in 2006, when salmon return was nearly 20,000
individuals.

Willow growth

-

Numbet' of
salmon

2

.. ....
.:. ..

.

..
.......
.
..
•.
: '
•.
... .... .
.... '
.'. ...• ·... ... ...
'.
.
.
..' .'.'
•. .
.
.
.
.'
.
.
• .
.
.
'
.
.
.
.
'
.'....
.. ... ...
..
.
.
• •
•. .
.. .
.'
...
....
.
''
. ...
..
~

1990

$

, ·,

'

-1

.'..'.
:
...

1995

2000

~

...
.'.

•
.
.... ....
..•.
..
.' ..' ..'. ...
.•. ..·.. ...
.'...
.. '
'

2005

..
••.

2010

Figure 3.12. Annual ring width indices for a willow tree (Salix sp.) located centrally on
the study site between the years 1956 and 2011. The vertical red dotted line delineates the
year the spawning channel was built, 1989.

Chapter 3. Marine-Derfl,ed Nutrient Dynamics in the Stream-Riparian Ecosystem

127

2

-gso

1970

1980

• 990

2000

2010

Year

Figure 3.13. Standardized values of annual willm"' tree growth, measured by ring widths,
and salmon escapement in the Horsefl) River spawning channel since its establishment
in 1989.

3.4 Discussion
3.4.1. Macroinvertebrate Community of the Horsefly River
The invertebrate community of the spawning channel is consistent with the

theorized structure of low order, headwater streams described in the River
Continuum Concept (Vannote et al. 1990). Shredders and collector-gatherers
constitute the dominant, year-round FFGs, which are fueled by terrestrial-inputs
provided by the bordering riparian forest (Doucett et al. 1996 ). Species richness i~
relatively low at this site, with only 13 taxa identified, although richness \Vould

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

128

likely increase slightly if some groups had been identified further, such as
invertebrates from the family Chironomidae. In comparison, the community of the
Horsefly River contained invertebrates from 41 taxa, which decreases to 34 total taxa
when combining species from the Chironomidae family and the genus Baetis.
Species diversity, according to the Shannon Index, is higher in the Horsefly River,
although the spawning channel has greater evenness, according to the Lorenz curve
and Gini coefficient, because fewer rare taxa were identified (Table 3.3).
Theoretically, the healthiest communities are those with a large number of species
and a nearly even distribution. Uneven communities are less resistant to
environmental stress because each FFG provides a critical role to whole ecosystem
function (Peet 1974, Wittebolle et al. 2009).
The richness and composition of an invertebrate community is highly
dependent on the specific site, as well as sampling techniques. Temperature,
substrate, surface flow, and microhabitat type are important variables that regulate
species distribution (Hill & Hawkins 2014, Pingram et al. 2014, Silva et al. 2014,
Dohet et al. 2015). The high abundance of amphipods (Hyalella azteca) defines this
site as being more pond-like compared to the Horsefly River, which largely reflects
the glide microhabitat of the spawning channel compared to the riffle that was
sampled in the Horsefly River (France 1992). Insect drift from the settling pond at
the top end of the channel may also have contributed to high amphipod abundance
downstream. Although amphipods are widespread shredders commonly found
throughout freshwater and marine environments (Strong 1972, Lowry &
Springthorpe 2001), they were not observed in the Horsefly River samples and are
not prevalent in the MDN literature. Shredders documented in other MDN studies

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

129

typically are from the orders Plecoptera and Tricoptera, but the only shredders at
the spawning channel were amphipods. Most of these studies are conducted in riffle
habitats (Honea & Gara 2009), thus examining MDN dynamics of a glide habitat
contributes new information to a limited area of study (see Thompson 2007).
Investigating glide and pool environments may be especially important for sockeye
salmon conservation because they supply early habitat and nourishment for eggs
and fry, as well as potential rearing habitat for juveniles (Fillatre Miller et al 2011).
In the six week period between pre-spawn and post-spawn the total number
of invertebrates increased markedly from 59 to 993 individuals, confirming the
hypothesis that population would increase upon the arrival of salmon. The largest
jump in community abundance and biomass occurred between weeks active spawn
II and active spawn III (increase of 425 individuals), by which point the channel was
at its maximum salmon capacity for this year (Figure 3.6 ). Abundance increased for
all taxonomic and FFGs during this period, including the first appearance of
collector-filtering organisms, specifically bivalves from the Sphaeriidae family and

Arctopsyche sp., a Tricoptera species (Figure 3.14). Although MDN may play a role in
increased abundance, the life history of most macroinvertebrates displays
population growth during the summer (Hynes 1970). The daily build-up and release
of the salmon dams was also most apparent between these weeks, with surface
water levels fluctuating about 1 m during a day (see Section 2.2.1.1). The dams were
composed of anything in the water column that could float downstrean1, from
FPOM to salmon carcasses and tree branches, and constituted a dense, nutrient-rich
food source for invertebrates and fishes to colonize and congregate. Total
invertebrate abundance increased slightly between active spawn III and post-spawn,

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

130

but this increase was only observed for amphipods and chironomids. In the spring,
most species declined in number except for chironomids, which retained high
abundance over the winter (Figure 3.8). Significantly more chironomid pupae were
identified in the spring, as well as over 200 pupae casings. Chironomids remain in
their pupal stage for a short period of time, indicating that chironomid morphing
and emergence began at the onset of spling break-up and/ or when channel ice
thawed (Danell & Sjoberg 1977).
The study design aimed to capture pre-spawn abundance and isotopic
signatures of macroinvertebrates as temporal control baseline conditions. However,
preparatory cleaning of the spawning channel, which occurred between August 8-12
and 15-18, ending five days prior to inve1tebrate sampling, likely altered water
chemistry (see Section 2.4.2) and may have affected pre-spawn taxa abundance. This
activity ceased approximately 200-m upstream from my study site, where a
temporary barrier inhibited the downstream flow of CPOM and a siphon pumped
water and materials into an isolated settling pond (see Figure 2.2). Although the
substrate at my study site was physically undisturbed by channel cleaning, the
removal of upstream food sources and insect drift through siphoning may have
displaced invertebrates, leading to initial low abundance. A high sediment to food
ratio may also have adversely affected invertebrate density (Pingram et al. 2014). It
is also plausible that with practice I became more effective at kick sampling during
the sampling period, reflecting a lower abundance of invertebrates in Pre-spawn
samples.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

131

3.4.1.1 Macroinvertebrate Assimilation of MDN
An increase in species richness was directly attributable to the presence of
salmon at this site; however, macroinvertebrate assimilation of MDN was variable
between species. As detailed in Chapter 2, NH4+ enrichment of stream water could
not be directly connected to the presence or abundance of salmon, a likely result of
rapid nitrification of surface water NH 4+ (see Section 2.4.2). However, NH/ values of

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

132

hyporheic water were significantly reduced during the post-spawn period,
suggesting that salmon did contribute measurable nutrients to the stream (Figure
3.15). Accordingly, b 1c;N signatures of macroinvertebrate groups Chironomidae and
Ephemeroptera increased alongside hyporheic N114 concentration, which may
signify MDN assimilation. This outcome results in a rejection of the null hypothesis
(VI) that b 1c;N signatures of these taxa will not be correlated with NI 14 concentration
of hyporheic water. As well, the statistically significant negative correlation between
Chironomidae and Ephemeroptera b 1 N signatures and surface water Nl i 4 results
in a rejection of that component of the null hypothesis (VI). The relationship
between salmon and each FFG is described in greater detail below .

•

600-

••
z 400.::::!
Ol
E
E

•
•
••
••

•
•

•

.~

c
0

E
E

<(

200-

I

•
•

I
o-

I

I

Pre-spawn

Ac11ve spawn

Post-spawn

Figure 3.15. Boxplots displaying ammonium (NH 4 ) concentration of all h'r porheic ""ater
samples. Active spawn includes data from three sampling"" eeks, Active spa"' n I, II, and
III. Outliers are denoted by red dots.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

133

Shredders:

High shredder density has been reported to coincide with leaf fall in autumn
(Hawkins & Sedell 1981, Reece & Richardson 1998), which supports this study's
finding of increasing amphipod abundance throughout the fall and lower
abundance in the spring. The concurrent timing of salmon spawn and vegetation
senescence enhances microbial activity on leaves and generates a bioenergetic
advantage for shredders by promoting rapid growth rates (Minakawa et al. 2002, Ito
2003, Honea & Gara 2009). A trend of gradually increasing mean amphipod growth
from Pre-spawn (7.97 mm) to Post-spawn (10.58 mm) was observed at this site.
However, increased growth was not statistically significant, likely due to size
variability within each sampling period (SO = 3.2-4.3 mm).
Weekly shredder b 15N signatures also increased during the fall while salmon
were actively spawning in the channel (Figure 3.16). Additions of nitrogen and
phosphorus to a stream stimulates microbial activity on leaf litter, and shredders
will preferentially feed on microbially-conditioned leaves (Ito 2003). The addition of
MDN to the stream may be indirectly reflected in increased b 15N of amphipods, as
microbes colonizing leaves can quickly assimilate salmon-derived nutrients
(Peterson et al. 2001). However, b13C continually decreases throughout salmon
spawn, which is contrary to the expected b 13C enrichment during conditioning of
detrital matter (Rounick & Winterboum 1986). This finding resembles Albers
(2010)'s results of decreased b 13C in post-spawn biofilms at the spawning channel.
Thus, amphipods may also be feeding on another food source, such as biofilms or
algae. Amphipods can exhibit plasticity when different food sources become
available and have also been found to consume fish, macroinvertebrates, and other

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

134

amphipods (MacNeil et al. 1997). Throughout the study, many amphipods were
found residing within clumps of filamentous algae attached to the substrate, which
may also be an important year-round food source.

-23-

•

~

•
-24-

•
•
•

90-

<1'1

z

8.5-

•

$

8.0r
Pre-spawn

1

Active
spawn I

r
Active
spawn II

I

Active
spawn Ill

•

I

T

Post-spaYm

Spnng

Figure 3.16. Boxplots displaying all b13C o/oo (top) and b15 N o/oo (bottom) signatures from
the shredder functional feeding group (FFG) during each sampling week. Black dots
denote actual values.

Contrary to the other invertebrates, amphipod mean isotopic signatures
remained consistent between seasons, likely because the brood sampled in the fall
was identical to the spring (Menon 1966) and possibly because their primary food
source also remained isotopically-enriched. Therefore, amphipods fail to reject the
null hypothesis (V) that b13C and b15 N will remain the same between seasons.
Terrestrial leaves are thought to be the preferred food source of amphipods
(Hawkins & Sedell 1981) and at this site isotopic ratios of fresh willow leaves are

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

135

stable between seasons. The presence of N-fixing alders at the spawning channel
likely provides a year-round contribution of N to forest soils Oohnson & indberg
1992), which may stabilize b15 N in willow. Helfield & Naiman (2002) found that at
spawning sites bordered by alder, spruce b15 N was similar to alder b 15 N and
decreased significantly at sites without alder.
Amphipods are not often referred to in MDN literature; however, in brownwater streams of Western Kamchatka, amphipods were found to be an essential
element of the salmon feedback loop. They comprised more than 88% of invertebrate
biomass, were observed feeding directly on salmon carcasses, and composed 68-88%
of juvenile salmonid diets (Thompson 2007). Our results of isotopically-enriched
amphipods in the spring support the notion that they may play an important role in
the salmon feedback loop by linking MDN from spawners to their progeny. Since
amphipods will migrate to find the highest quality food source (Cothran et al. 2014),
it is possible that they move to spawning grounds in the fall to locate conditioned
leaves and then relocate to the lakes and ponds they primarily occupy (Clifford
1991). Juvenile sockeye salmon preferentially rear in lakes (Eggers 1978), so an
amphipod migration that mimics salmon movement may indicate an amplified
response of the importance of amphipods to perpetuating the feedback loop.
Collector-Gatherers:

Chironomids were the most abundant taxa in the collector-gatherers FFG,
comprising on average 67% of total group abundance. Aquatic earthworms,
commonly called oligochaetes, were the second most abundant collector-gatherers,
fo llowed by two mayfly taxa, Paraleptophlebia sp. and Baetidae. Tricopteran, or

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

136

caddisfly, collector-gatherers made an appearance during post-spawn sampling.
Stable isotopic signatures were obtained for the taxa Chironomidae and
Ephemeroptera (mayflies). Due to the small mass of these organisms, duplicate
samples were not always possible. Therefore, taxa analysis for some samp1ing weeks
is based upon only one data point. Conclusions based upon singular data have a
high likelihood for error and must be interpreted cautiously. There were no clear
trends in isotopic composition for these invertebrates during the salmon spawn,
although pre-spawn values were not obtained (Figure 3.17). Nitrogen isotopes for
both taxa were very similar, suggesting similar trophic enrichment from a common
food source. However, fall b13C signatures were slightly enriched for
Ephemeroptera, which could be an indication of differing lipid content between the
taxa, as proteins and carbohydrates are enriched in b 13C relative to lipids (Post et al.
2007).
In the fall, chironomids had the least enriched b13C and b15 N signatures of all
the invertebrates, indicating that these chironomids are likely not predacious and
are mostly or entirely collector-gatherers (Figure 3.3). Chironomidae and
Ephemeroptera b15 N signatures in the fall were significantly enriched (p < 0.05) over
spring values, suggesting that they benefitted from assimilation of salmon nutrients
into the food web (Figure 3.5). This outcome results in a rejection of the null
hypothesis (V). Peterson et al. (2001) found that after NH4 +-15 N injections around 7080% of the NH4 +was directly assimilated by benthic organisms, with the rest
removed by nitrifying bacteria. The quick removal of NH 4+from surface water could
be a result of direct assimilation by collector-gathering organisms or indirect
assimilation through the consumption of bacteria located among the streambed

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

137

gravels (Bilby et al. 1996, Webster et al. 2003).
Ephemeraptsra

Chronomid

-24-

-25-

•

...

•

~

-26

•
•

-27-

•

10-

•
8-

•

I

• .

•

i

...

!!
6-

4I

T

Active
spawn I

Ac1ive
spawn II

T

T

Active Post-spawn
spawn Ill

•

I

I

I

Spring

Adlve

Active

spawn I

spawn II

T

J

Active Post~
spawn Ill

r

Spnng

Figure 3.17. Boxplots displaying b13C %o (top) and b15N %o (bottom) signatures from two
species of the collector-gatherer functional feeding group (FFG). Plots on the left show
chironomid isotopic signatures and ephemeroptera are displayed on the right. Pre-spawn
isotopic signatures are missing for these species. Black dots denote actual values.

A major difference between these species was the significant depletion of b13C
in the spring for Ephemeroptera and a significant b13 C enrichment for chironomids.
Both of these results reject the null hypothesis (V) that b13C signatures will remain

unchanged between seasons. This result, along with relatively enriched spring b15N
signatures of Ephemeroptera, indicates that they are consuming distinctly different
food sou rces or that their specific life histories alter tissue nutrient composition.
During times of starvation, chironomid b13C signatures become significantly more
enriched and, since chironomids hibernate during the winter, enriched spring () 13C

- - - - - - -

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

138

concurs with hibernation and subsequent starvation (Doi et al. 2007). Isotopic
signatures also change during chironomid metamorphosis, with results of enriched
b13C and decreased b15 N in larvae, which could explain significantly lower b 15 N
signatures in the spring. The slight enrichment of b15 N and depletion of b 13C in
chironomid pupae relative to chironomid larvae (Figure 3.18) demonstrates the
isotope turnover that is occurring during metamorphosis from the larval to adult
stage (Doi et al. 2007).

-24.0-24 5-

-

-25 o-

w

("')

•

-255-26.0-

-

-26.5-

•

4.8-

•

r

4.5-

4.2 -

I•

3.9 -

I•
I

I

Chirononwd
Larvae

Ch.-onomid
Pupae

Figure 3.18. Chironomid larvae and pupae isotopic signatures collected in May 2012.
Increased b15 N (%o) and decreased b13C (%o) in pupae are demonstrating nutrient changes
due to metamorphosis. Black dots denote actual isotope values.

In the spring, snowmelt, rain, and freshet induce lateral inputs from the
riparian zone and elevates stream DOC (Holmes et al. 1996, Duff & Triska 2000,

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

139

Harvey et al. 2013). Lower carbon and nitrogen isotopic signatures in the spring for
Ephemeroptera could be reflecting a greater reliance on relatively depleted
terrestrial food sources compared to enriched autumnal MDN assimilation. Spring
freshet also may have contributed to lower collector-gatherer b15N signatures, as
high water levels initiated downwelling hyporheic conditions and supplied DOC, N,
and DO creating a nitrifying environment (Chapin et al. 2011, Zametske et al. 2011,
Heppell et al. 2014). Nitrification creates b15 N-depleted N0 3· (Chapin et al. 2011 ),
which may partially be the reason for decreased invertebrate b15N.
Chironomids are a fairly common benthic invertebrate identified in MDN
studies and appear to benefit from salmon nutrients during the salmon spawn and
retain high abundance after the winter (Hicks et al. 2005, Lessard & Merritt 2006,
Lessard et al. 2009, Verspoor et al. 2011, Campbell et al. 2012). They are classified as
hyporheos, or hyporheic-dwelling organisms, and are able to avoid spawning
mortality by retreating deeper into the substratum during redd building
disturbances that causes some invertebrates to decrease in abundance (Lessard &
Merritt 2006, Honea & Gara 2009). Also, chironomids have been observed feeding on
salmon carcasses (Chaloner & Wipfli 2002), so although isotopic signatures did not
reflect salmon predation, this taxa could have altered its feeding habits to carcass
consumption after fall conditions reduced primary productivity (Lessard & Merritt
2006 ). This study supports research that identifies chironomids as important
organisms to the aquatic food web integration of MDN and their effect in providing
source nutrients to future salmon stocks by maintaining high abundance (Chaloner
et al. 2002b, Lessard & Merritt 2006).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

140

Collector-Filterers:

Collector-filterer abundance has been found to increase after additions of
coarse detritus, particularly leaf litter, as a response to increased FPOM from
shredding organisms (Richarson 1992). They also may display enriched isotopic
signatures coinciding with salmon spawn, indicating direct or indirect MDN
incorporation (Honea & Gara 2009). The appearance of collector-filterers during
active spawn at the HFC could be due to the cumulative effects of nutrient sources
from both leaf senescence and salmon spawn.

Scrapers:

Gastropods constituted the entire scraper FFG and were represented by two
species, Physa sp. and Promenetus umbilicatellus. Scraper abundance remained
relatively uniform throughout the study, with total post-spawn abundance (n=21)
nearly matching spring (n=20), showing persistence of scrapers in this stream. The
response of scrapers to spawning salmon produces a variable response between
regions, including reports of no discernible response (Lessard & Merritt 2006). Some
studies have found a lack of MDN assimilation in scrapers (Rinella et al. 2013), while
others have observed an accumulation of MDN and increased biomass in spawning
locations (Bilby et al. 1996, Honea & Gara 2009).

Scraper isotopic signatures in the spring were b15 N-enriched relative to all
other taxa (Figure 3.4). Biofilms are the primary food source of these low-trophic
order species (Anderson & Cabana 2007), which infers that either the gastropods
and I or their food source remained enriched in MDN throughout the winter season.
Adult Physa sp. likely spawn after spring break-up, so the brood sampled in the fall

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

141

wou ld be identical to the spring, which supports the assumption of b15 N
accumulation (Clampitt 1974). Additionally, winter can induce nutritional stress,
illustrated by an increasing C:N ratio of source materials, which is known to cause
b15N enrichment (Dekar et al. 2009).

Predators:

The relative abundance of predatory invertebrates was greatly influenced by
the high number of chironomids of which 10% were classified as predators (Figures
3.7, 3.14). The next most prevalent predator group from the subclass Hirudinae,
commonly known as leeches, continually increased upon the arrival of salmon and
decreased after spawning. Predatory invertebrates are limited by their available
prey (Richardson 1993, Honea & Gara 2009, Verspoor et al. 2011 ), thus, leech
abundance may have fluctuated with resident fish and juvenile salmonid
populations. Many riverine fish species will follow spawning salmon to gorge on
salmon eggs and flesh (Cederholm et al. 1999, Moore & Semmens 2008), which likely
were important nutrient sources for leeches.
The weekly isotopic signatures of leeches were represented by only one data
point, so it is difficult to conclusively discuss trends (Figure 3.19). Carbon and
nitrogen isotopic signatures generally decreased from active spawn to post-spawn,
which may correspond to a reduction in live salmon, as well as a reduced number of
resident fish following the spawners. Other fishes observed in the channel during
the fall were Redside shiners (Richardsonius balteatus), Rainbow trout (0. mykiss), and
Kokanee salmon (0. nerka).

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

142

-23 .0 -

-23 5-

-24 0-

-24 5-

11.7511.50-

..z
VI

11.2511.00-

10.75-

r

I

I

I

I

Pre-spawn

Active
spawn I

Actrve
spawn II

ActN"e

Post-spawn

spawn Ill

Figure 3.19. Boxplots displaying b13C %o (top) and b15 N %o (bottom) signatures of
Hirudinae (leeches) from the predator functional feeding group (FFG) during the fall
sampling weeks. Spring isotope values were not obtained due to low abundance and
mass. Black dots denote actual values.

3.4.1.2 Nutrient Sources in Aquatic Systems
The uptake and retention of salmon nutrients (MDN) were the focus of this
study. However, they represent only one of many resource pulses to a stream that
can enhance population productivity and growth (Richardson et al. 2010). Terrestrial
leaf litter inputs, particularly in the fall, are an impo1iant food source to
invertebrates with low trophic position (Wallace et al. 1999). Additionally, leaf
leachate provides dissolved nutrients to aquatic primary producers and has been
found to contribute 30% of total dissolved organic carbon (DOC) to a stream
(Richardson et al. 2010). Lateral flows of groundwater can deliver DOC and

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

143

dissolved inorganic nitrogen (DIN) to the stream from organic soil horizons,
especially during high precipitation, flooding, and spring break-up events
(Wondzell & Swanson 1996b, Mei et al. 2014). Nutrients are also transported from
upstream sources (Wipfli et al. 2007) and, in this case, water in the spawning channel
originates in the upstream settling pond, which is rich with macrophytes and algae.
The rationale in tracing MDN using stable isotopic analysis was that salmon
are often migrating to oligotrophic streams and, therefore, contribute detectable
amounts of N exceeding that of atmospheric N deposition (<5 kg ha· 1 year· 1), in
conjunction with the natural10-1Sroo b15 N enrichment of marine sources relative to
terrestrial sources (Welch & Parsons 1993). However, organisms and food sources
can appear enriched in stable isotopes regardless of MDN presence (e.g. due to
isotopic discrimination; Marshall et al. 2007, Staal et al. 2007), and substantial
nutrient inputs can come from sources other than salmon (Triska et al. 1993, Helfield
& Naiman 2002). When N is not the limiting nutrient in a system, the ability to trace

marine-derived nitrogen can be significantly reduced owing to confounding effects
of isotopic fractionation (Naiman et al. 2002). A nitrogen fractionation, a.k.a.
enrichment, factor of approximately 3-4%o occurs with each successive trophic
position (Minagawa & Wada 1984), so an organism that utilizes many N sources will
have a b15N value that incorporates multiple sources. In most systems, however, N
sources are limited and demand exceeds supply (Aber et al. 1989). While this
narrows the possible N options, it remains difficult to discriminate between relative
source contributions when there is more than one N-rich source (Marshall et al.
2007). Since b 13C signatures of a consumer reflects its' source with a smaller
fractionation (-oAroo), they can be helpful to determine primary nutrient sources

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

144

(DeNiro & Epstein 1978). Although this study did not aim to identify the food
sou rces of invertebrates, isotopic signatures of possible sources (e.g. aquatic
autotrophs, detritus, leaf litter, bed sediments, and resident fish) would have greatly
helped discern patterns of enrichment and the specific impact of MDN.
3.4.2 Nutrient Uptake and Assimilation in th e Riparian Zone
Riparian plants at this site possessed distinctly different b 13C and b 15 N
signatures, which lends to insights of plant function and overall nutrient availability
(Menezes 2009, Chapin et al. 2011 ). For instance, E. arvense typically were enriched in
b15N compared to other species, which could be the result of a preference for NH4 +
(Chapin et al. 2011). Variations in isotopic fractionation are mostly due toN
transformations, such as nitrification, denitrification, and volatilization (Finlay &
Kendall 2007); plants that prefer NH4 +will be isotopically enriched relative to N03absorbing plants because nitrification has a fairly large fractionation effect (Chapin
et al. 2011 ). E. arvense may also display enrichment because horsetail is especially
effective at accumulating and retaining nutrients (Husby 2013). They even act as a
nutrient pump by delivering nutrients from the soil C horizon to its shoots aboveground (Marsh et al. 2000).
All terrestrial plants obtain carbon from a well-mixed atmospheric reservoir
in the form of C02 (Ben-David et al. 1998), thus b13C variation due to the source is
insignificant and the p resence or absence of MDN should not be a factor. Variation
occurs from fractionation during C02 assimilation, which is primarily related to
water use and plant growth rate (Finlay & Kendall2007, Marshall et al. 2007). S.

Iucida consistently displayed b13 enrichment relative toT. offici'nale and E. arvense
leaves, wh ich is an expected result when comparing long-lived perennial species

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

145

and annual or herbaceous species (Marshall et al. 2007). Seasonal variation in b13C
was not detected, but there was a significant b13C enrichment in plants located
directly on the streambank (Table 3.5). Plants that experience water stress
discriminate against b13C, so stream-side vegetation with a reliable supply of water
will exhibit higher (less negative) b13C signatures (Wang et al. 2010). More available
sunlight, resulting in a greater rate of photosynthesis, may also be contributing to
b13C enrichment at this location (Leavitt & Long 1986), as the vegetation sampled

here was not shaded by the riparian forest.
The nutritional value (C:N) of all sampled vegetation species decreased as fall
progressed and were most nutritionally rich in the spring (Table 3.6). Prior to
senescence trees will reabsorb N from the leaves, so a decrease in total Nand an
increased C:N is to be expected (Drake et al. 2006). S. Iucida reabsorbed about 35% ±
10.6% of leaf N between August and September 2011, whereas E. arvense absorbed
slightly less (30% ± 12.8%) and T. officinale reabsorbed the least N (15% ± 14.1 %). The
%C inS. Iucida samples remained steady throughout seasons averaging around 46%,
while the perennial T. officinale and E. arvense species had substantial reductions in
%C from spring to fall, especially for E. arvense. This decreased nutritional value of
senescent leaves amplifies the importance of the microbial-conditioning of leaf litter
that coincides with salmon spawn for amphipod growth in the fall and winter
energy storage.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

146

Table 3.6. Stable isotopic signatures, percent of total carbon (C) and nitrogen (N), and the
C:N ratio of sampled riparian vegetation at the Horsefly River spawning channel during
Fall 2011 and Spring 2012.

Species

Date

~13c

~15N

%C

%N

C:N

Salix Iucida
Salix Iucida
Salix Iucida

Aug 2011

-28.40
-29.23
-28.38

2.37
1.87
1.23

46.76%
46.60%
46.86%

2.45%
1.60%
3.88%

19.01
29.15
12.01

Equisetum arvense
Equisetum arvense
Equisetum arvense

Aug 2011

Sept 2011
May 2012

-29.34
-29.88
-30.03

10.39
8.67
6.55

30.33%
29.81%
39.22%

1.82%
1.26%
3.49%

16.84
23.68
11.17

Taraxacum officinate Aug 2011
Taraxacum officinale Sept 2011
Taraxacum officinale May 2012

-32.53
-33.24
-31 .64

0.54
1.09
5.28

37.87%
38.00%
43.67%

2.52%
2.15%
4.32%

14.93
17.52
10.05

Sept 2011
May 2012

Salix Iucida (Pacific willow):

S. Iucida were the only plants sampled that displayed a statistically significant
spatial trend in b15 N, specifically decreasing b15 N signatures with distance from
streambank (Table 3.4). Similar results have been used to prove the influence of
spawning salmon on riparian vegetation (Bilby et al. 1996, Ben-David et al. 1998,
Helfield & Naiman 2002, H icks et al. 2005, Drake et al. 2006, Hocking & Reimchen
2009). The reduction of b15 N in inland willows suggests that less marine N reached

these locations and potentially all b15 N carne from microbial denitrification. Inland
willows had b15N signatures between -1 and O%o, which represent standard values of
atmospheric N assimilation (Kohl & Shearer 1980). At our site, the willow tree
directly on the streambank had an average b15 N enrichment of 57'oo relative to the
willows 20-m inland in the fall, and 3%o in the spring. This result confirms the

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

147

hypothesis that near-stream vegetation will retain MDN over the winter. Similar1y,
the near-stream willows exhibited an average b13C enrichment of nearly 4%o in the
fall and greater than 3r'oo in the spring. This could be due to high water and sunlight
availability (Leavitt & Long 1986, Marshal] et a1. 2007, Wang et a1. 2010), as well as
MDN enrichment.
The roots of Salix sp. protrude into the hyporheic zone and have been found
to rely on groundwater throughout their growing season (Menezes 2009). Since a
trend of decreasing b15N with distance from the streambank is only apparent in
deep-rooted willow trees, this infers that substantial quantities of MDN is
transferred to the near-stream riparian zone by advective hyporheic water. This
corresponds with results reported in Chapter 2, where NH 4+concentration in
hyporheic water also decreased with distance from the streambank and did not
appear to infiltrate inland locations during active salmon spawn. These data support
the rejection of the null hypothesis (VIII) that stable isotopic signatures of S. Iucida
will not be correlated with distance from the streambank. Nitrogen isotope
enrichment in riparian vegetation by 15N-NH 4+is substantiated by a study from
Drake et al. (2006), where they discovered that tree roots began to absorb b 15N
within 7 days and that 37% of the 15 N-NH/ tracer was contained within tree tissues
after six months.
From the tree ring analysis it appears that seven years after the construction
of the spawning channel, this particular willow tree began exhibiting increased
annual growth (Figure 3.12). Trees are capable of developing a physiological legacy
effect that allows them to efficiently assimilate nutrients when they can depend on
resource pulses (Drake et al. 2006), which seems to be occurring at this site. The

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

148

standardized values of ring width and salmon abundance also display a
ynchronized relationship during certain years, particularly between 1997-2000,
when salmon population was zero in 1997 and escalates to -25,000 the next year,
and in the year 2006 when salmon return was nearly 20,000 individuals (Figure
3.12). Recent studies have linked salmon abundance and tree ring growth (Drake &
Naiman 2007, Reimchen & Fox 2013), however high precipitation, temperature, or
other climate variables could also lead to increased growth (Garfinkel & Brubaker
1980). At this site, multiple variables plausibly acted together to increase tree
growth, as evidenced by a significant multiple regression with the interaction of
precipitation and salmon abundance. The construction of the spawning channel may
also have created a higher local water table, thus promoting tree growth. Lastly, the
strength of this analysis is questionable because only one willow tree was sampled;
however, these preliminary results provide justification to continue tree ring
research at this site.

Equisetum arvense (Common horsetail):

E. arvense showed a statistically significant b15N depletion in the spring which
could be attributed to a decrease in available N, in conjunction with greater reliance
on associated mycorrhizal fungi. As much as 70-80% of E. arvense roots are colonized

by arbuscular mycorrhizal fungi (Hodson et al. 2009), which supply an alternative
source of N; plants may rely more heavily on this symbiotic relationship when N
availability is low, thus leading to relatively depleted b15N of E. arvense in the spring
(Chapin et al. 2011). Rinella et al. (2013) also reported lower b15 N signatures of
horsetail in the spring, as well as an apparent b15 N enrichment in the fall that did not

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

149

vary with spawner density. Horsetail is known to be especially effective at
accumulating and retaining nutrients (Marsh et al. 2000), which could account for its
persistent b15N enrichment over the other plant species and may infer an
accumulated salmon legacy effect. Carbon isotopic signatures remained the same
between seasons, resulting in a failure to reject the null hypothesis for b13C and a
rejection of the null hypothesis for b15N (VII).

Taraxacum officinale (Common dandelion):

A significant increase in b15 N signatures of T. officinale in the spring
demonstrates that herbaceous annual plants in the same area can exhibit vastly
different nutrient assimilation techniques and that large spatial variations in nutrient
concentration can occur in relatively close proximity (Chapin et al. 2011). This result
supports the rejection of the null hypothesis (VII) that b15N signatures will remain
the same between seasons. Half of the T. officinale sampling locations (3 out of 6)
exhibited a sharp b15 N increase in the spring (average of 8.21%o ± 3.51), while the
other locations remained similar to September signatures (average increase of 0.17roo

± 0.93). Although b15N of a single herbaceous species has been documented to vary
by >1%o even when grown hydroponically on a known source (Robinson 2001),
differences of this magnitude are probably due to microhabitat variation of theN

pool. Dandelions flourish best in moist loam (Stewart-Wade et al. 2002) where
conditions of high water content would limit DO and create denitrification
conditions. These factors promote b 13 depletion along with NH4 +-b 15 N enrichment
of theN pool (Marshall et al. 2007, Pinay et al. 2009). Low b 13C and high b 15 N are
present at three locations in the spring (T1-0, T2-10, and T2-20), which supports this

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

150

theory. Also, arbuscular mycorrhizal fungi is common on dandelion roots (Hodson
et al. 2009), so the b15N depleted plants could have been in locations where N was
low and their reliance on N from the fungi was increased.

Soil:

Soil isotopic signatures of b 13C followed a pattern of increasing enrichment
with depth that is commonly found across ecosystems due to microbial preference
for the lighter isotope and subsequent accumulation of the heavy isotope in older
soils (Figure 3.9; Ehleringer et al. 2000, Garten et al. 2007). Patterns of b15 N in soil
were less obvious because nutrient pooling and mechanisms of N uptake and
transformations all factor into b15 N signatures {Garten et al. 2007). For example, as
with b13C, soil microbes tend to discriminate against the heavier N isotope in
processes such as denitrification, resulting in a nitrate pool that is b15 N-enriched
(Mariotti et al. 1981). b15 N enrichment can occur at localized sites, such as
floodplains, where moisture saturation leads to low oxygen levels and higher
denitrification rates, or near plant roots because vegetation is often preferential to
the lighter N isotope (Nadelhoffer & Fry 1994, Naiman et al. 2002).
A distinct variation in soil b15N signatures was observed in the deepest
location sampled (horizon C), which was within the zone of hyporheic water

inundation. Soil pit 1, located closest to the streambank on the upstream side of the
meander bend, had a considerably high b15N value of 6.21 %o, compared to soil pit 2
(3.76%o), located furthest from the streambank, and soil pit 3 (4.09%o ), located on the

downstream side of the meander bend (Figure 3.10). Given samples were taken at
similar depths, enriched soil b15 N could indicate a persistence of MDN infiltration in

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

151

this location close to the stream's edge.
Carbon isotopic signatures across horizon C were nearly identical (-24.58%o ±

o.o2roo), but the rate of change in b13C of vertical organic matter decomposition
differed between sites. The regression coefficient between b13C abundance and the
log C concentration were used to define rate of change (Garten et al. 2007). Soil pit 1

b13 signatures changed at a rate of 0.64roo along the depth continuum, whereas soil
pit 2 and 3 displayed slower changes with b13C change rates of 0.46roo and 0.53%o,
re pectively. This greater rate of change at soil pit 1 could indicate the presence of
MDN-enriched hyporheic water inundating deep soil horizons located near to the
stream's edge.

3.5 Conclusions
The broad sample design of this study captured a few primary nutrient
pathways linking MDN to aquatic and terrestrial food webs at the Horsefly River
spawning channel. All macroinvertebrate FFGs displayed characteristics that could
be positively associated with the presence of salmon and MON, including an
increase in community biomass, species richness, and species diversity during the
spawning period. Amphipods (shredders) and chironomids (primarily collectorgatherers) were the most abundant taxa, with both groups combined comprising
between 55% of total abundance in the spring and 88% during post-spawn.
Chironomids appeared to assimilate MDN during salmon spawn, as evidenced by
enriched b15 N signatures in association with increased NH 4+concentration of
hyporheic water beneath the streambed, which may have provided energy to help
them maintain elevated biomass in the spring. Juvenile salmon are highly

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

152

dependent on chironomid larvae as a food source (Lessard & Merritt 2006) and this
study further suggests that they provide a critical link between the MDN pulse in
the fall and salmon fry emergence in the spring. Past studies have reported
shredders as a group that maintains an elevated biomass throughout salmon spawn
(Minakawa et al. 2002, Lessard & Merritt 2006, Honea & Gara 2009), but, to my
knowledge, this is the first study in North America that identifies freshwater
amphipods (in particular, H. azteca) as the only shredding invertebrate located in a
spawning area. Since amphipods are known to migrate to find the highest quality
food source (Cothran et al. 2014), it is possible that they travel to the spawning
channel in the fall to locate microbially-conditioned leaf litter, which facilitates
growth prior to winter (Ito 2003 ). Then they potentially relocate to ponds and lakes
sometime after the salmon decomposition period, while maintaining an isotopicallyenriched signature over the winter. This study suggests that amphipods, as well as
chironomids, may be particularly connected to the sockeye salmon feedback loop, as
their seasonal migration evidently delivers an abundant, MDN-enriched food source
from spawning grounds to the lakes where juvenile sockeye rear.
Patterns of isotopic enrichment were variable between riparian plant species,
which is likely the result of differing morphologies and mechanisms of N-fixation.
The shallower rooting species of E. arvense and T. officinale did not exhibit decreasing
b15 N with distance from the streambank, nor a common temporal isotope change,
suggesting a lack of MDN assimilation from the fall nutrient pulse. However, S.

Iucida displayed a clear trend of decreasing b13C and b15 N concentration with
increased distance from the stream, as well as persistent isotopic enrichment
between seasons, indicating that this deeper rooted species is assimilating MON.

Chapter 3. Marine-Derived Nutrient Dynamics in the Stream-Riparian Ecosystem

153

Since riparian deposits of salmon carcasses and bear feces were not abundant at the
experimental site, the lateral transfer of MDN through hyporheic flow appears to be
the prominent delivery pathway and is most evident in willows <7-m from the
streambank. This conclusion is further supported by enriched b 15 N found in deep,
hyporheic-influenced soil samples located near the stream's edg . Nevertheless,
riparian and hyporheic ecotones are boundaries, by definition, comprised of a
mosaic of redox gradients that vary spatially and temporally (Naiman & Decamps
1997, Boulton et al. 2010, Febria et al. 2011 ). Markedly different N pools can exist
within close proximity, thus, this pattern of near-stream MDN enrichment would be
more conclusive with greater spatial and temporal resolution.

ontinued research of

the mechanisms that facilitate nutrient transfer and storage within the salmon
feedback loop will require a greater focus into the life histories and physiology of
species that appear to provide critical linkages between salmon generations.

Chapter 4. Conclusions

154

Chapter 4

Conclusions and Man gement Implic tions
4.1 Conclusions
The results of this thesis demonstrate that marine-derived nutrients from
spawning salmon are delivered to the riparian zone of an interior British Columbia
stream via lateral hyporheic transport. Previous studies have demonstrated
hyporheic transport of MDN (O'Keefe & Edwards 2002, Pinay et al. 2009), but this
thesis expands on preceding knowledge by providing a seasonal component that
identifies the assimilation and storage of MDN through multiple trophic levels
within the stream-riparian ecosystem. The primary aims of this thesis were to: 1)
identify the hyporheic zone at this site, 2) determine if NH 4+,presumably from
spawning salmon, was retained in the hyporheic zone during salmon spawn and
after the winter, 3) identify how macroinvertebrate FFGs responded to the MDN
pulse and salmon spawn, and 4) determine if the rooting depth and distance from
the stream of riparian vegetation had an effect on MDN concentration. Chapter 2 of
this thesis defines the hyporheic zone at this location and discusses the
biogeochemical factors which influence subsurface nitrogen retention and
transformation. Chapter 3 identifies willow trees, amphipods, and chironomids as
species which particularly benefit from spawning salmon and MON. By examining
the temporal effect of MDN, this thesis provides evidence of an asynchronous
exchange of nutrient subsidies between the stream and riparian zone.
Results from Chapter 2 indicate that NH 4+concentration in riparian hyporheic
water decreased with distance from the streambank and was not delivered to inland

Chapter 4. Conclusions

155

locations during the salmon spawn. This outcome corroborates results from Chapter
3 where the reduction in b15N of S. Iucida with distance from the streambank
sugg sts that less marine N reached these inland locations and potentially all of the
b 15N that was detected in inland willows came from microbial denitrification.

Shallow-rooted plants did not display a trend of decreasing b15N with distance from
the streambank, which further confirms that the hyporheic zone is a dominant
pathway in tran ferring nutrients from the stream to riparian zone. However, the
extent of this nutrient delivery is limited at the Horsefly River spawning channel
because hyporheic water is generally flowing toward the channel. Accordingly,
hyporheic nutrient delivery in any system will be controlled by the variables which
drive hyporheic flow, particularly hydraulic head gradients and hydraulic
conductivity (Wondzell & Gooseff 2013). Although this study was unable to
correlate NH4 +concentration with the arrival of spawning salmon, the corroborating
evidence of b 15N in near-stream willows suggests that salmon provide a substantial
amount of MDN to this system.
The vertical movement of MDN from salmon carcasses and enriched
metabolic wastes deposited on the riparian zone and subsequently transferred to
terrestrial vegetation has been well established (Ben-David et al. 1998, Hilderbrand
et al. 1999, Gende et al. 2001, Klinka & Reimchen 2002, Belfield & Naiman 2006).
However, much of this research has been performed around densely-populated bear
habitat and in coastal forests where precipitation and nutrient leaching rates are
high (Reimchen et al. 2002, Drake et al. 2006). Whereas, the Horsefly River is located
in an interior sub-alpine zone, with an average annual precipitation of 517 mm
(Burford et al. 2009). Approximately 35 mm of rain was recorded during the two

Chapter 4. Conclusions

156

month sampling period in the fall of 2011. Bear activity in the experimental site was
minimal and riparian deposits of salmon carcasses (n=S) and bear feces (n=2) were
low in number. With these considerations, it appears that drier, interior climates
may be less influenced by vertical nutrient leaching and more reliant on lateral
hyporheic delivery.
Hyporheic NH 4 ' concentrations beneath the streambed were also positively
correlated with chironomid b15 N signatures, which suggests that these hyporheicdwelling macroinvertebrates can rapidly assimilate MDN. This nutrient subsidy in
the fall is apparently beneficial to chironomids because their high abundance was
retained after the winter. Previous studies have confirmed the particular ability of
chironomids to utilize MDN and avoid spawning mortality (Hicks et al. 2005,
Lessard & Merritt 2006, Lessard et al. 2009, Verspoor et al. 2011, Campbell et al.
2012), however, this study presents a dominant assimilation pathway in the form of
b15 N-NH/ that may also be facilitated by hyporheic flow.
Although shredding invertebrates commonly appear in MDN literature, the
high occurrence of amphipods at this site provides a unique perspective. Amphipod
biomass and abundance exhibits a rapid increase in the fall, presumably due to the
consumption of leaf litter that has been microbially-conditioned with the aid of
salmon nutrients, and amphipods also maintain high b15N concentration in the
spring. Analogous to the preferred rearing habitat of sockeye salmon, amphipods
are typically associated with lentic environments. These results, in conjunction with
evidence of high amp hi pod predation by juvenile salmon (Thompson 2007), allow
inference that amphipods may be particularly important to the feedback loop of
sockeye salmon.

Chapter 4. Conclusions

157

The seasonal measurements of nutrients associated with salmon in this study
reveal reciprocal subsidies between the stream and riparian zone (Nakano &
Murakami 2001). In the fall, an upstream delivery of MDN from spawning salmon
was observed to increase macroinvertebrate community biomass, species richness,
and species diversity. Simultaneously, terrestrial leaf litter fall contributed a
compounding nutrient source to shredders, which are known to initiate nutrient
cycling processes in the stream food web (Gra<;a 2001). Hyporheic pathways
delivered MDN from the stream to the riparian zone, which could then be
assimilated by near-stream vegetation. Nutrient storage is evident in the riparian
zone over the winter, but these nutrients eventually return to the stream during
spring freshet. This seasonal shift in nutrient fluxes is further demonstrated by the
reversal between increased hyporheic NH4+ concentration beneath the stream in the
fall and higher riparian hyporheic NH4 + in the spring.
Boundaries between oxic and anoxic environments create redox gradients at
the intersections of ecotones, root micro-sites, pockets of organic matter, and
hotspots of bacterial assemblages (Stanford & Ward 1993, McClain et al. 2003, Claret
& Boulton 2009, Borch et al. 2010). Redox, or electron transfer, reactions provide

energy for biological growth, and strong redox gradients enhance biogeochemical
activity and microbial processing (Borch et al. 2010, Rezanezhad et al. 2014).
Nitrification and denitrification represent a tightly coupled oxidation-reduction
(redox) reaction that facilitates rapid organic matter decomposition and an overall
loss of nitrate from the aquatic environment (Zametske et al. 2011 ). The seasonal
reversal of hyporheic NH4+ concentration beneath the stream and below the riparian
could indicate the presence of a nitrification-denitrification coupling syste1n on a

Chapter 4. Conclusions

158

larger time and spatial scale. This discovery highlights the importance of
considering the entire stream-riparian ecosystem when implementing plans for
salmon conservation and management.

4.2 Managem nt Implications
Pacific salmon populations in the United States and Canada are displaying
signs of stress, such as decreasing returns and inconsistent cycles. Continued
research of these populations in their marine and terrestrial environments is
required, as well as the implementation of conservation action plans (Nehlsen et al.
1991, Beamish et al. 2012, Peterman & Domer 2012, Sharma et al. 2013). Salmon are
tough and resilient species, and can immediately return to areas that have been
blocked by dams for decades (Anderson et al. 2014). However, their physiology is
sensitive in regards to climate change pressure, particularly warming water (Hague
et al. 2011, Martins et al. 2011). Therefore, it is critical that we understand what
makes a salmon's spawning and rearing habitat persistent and stable, so we may
recognize changes in these ecosystems and implement meaningful conservation
solutions.
Results from this study identified that MDN support the whole
macroinvertebrate community and provide energy for growth to deep-rooted
riparian vegetation bordering the stream. In turn, riparian vegetation provides
shade for spawning salmon (Naiman et al. 2002), woody debris for juvenile salmon
habitat (Fausch & Northcote 1992), and leaf litter to support the macroinvertebrate
population (Vannote et al. 1980, Ito 2003). MDN stored in the riparian zone over the
winter are delivered back to the stream in the spring, activating microbial activity

Chapter 4. Conclusions

159

and supplying energy for salmon fry (Nakano & Murakami 2001). These nutrients
are recycled throughout the stream-riparian ecosystem (Wipfli et al. 2010), thus
maintaining strong salmon returns is crucial for healthy salmon habitat. Doing this
involves a more complete understanding of the biogeochemical interactions that link
and support these ecosystems, as well as identifying the key reactions, elements, and
species that are particularly critical elements of the salmon feedback loop.
The Horsefly River spawning channel offers a unique opportunity to study
stream-riparian food web dynamics of an interior sockeye salmon system. Gates
along the channel and a water pump dictating flow allow for experimental control
and manipulation. The macroinvertebrate community is relatively simple, with
dozens, as opposed to hundreds, of specific taxa, including some FFGs that are
represented by only one species or family. The present study, in conjunction with
studies from Albers (2010) and Albers & Petticrew (2012, 2013), present a useful
baseline understanding of the biogeochemical nutrient dynamics that occur because
of salmon spawn. From this simplified framework, strong food web models can be
developed and interpreted with the intention of discovering critical linkages that
elucidate the resilience of salmon populations. Nitrate removal experiments could
also be implemented (e.g. Peter et al. 2012) and the microbial assemblages that
facilitate denitrification could potentially be identified. The results presented in this
thesis illustrate the innate connectivity of the stream-riparian ecotone, as well as the
resilient salmon feedback loops embedded within this landscape.

Biblios rapll y

160

Bibliography
Aber, J.D., K.J. Nadelhoffer, P. Studler & J.M. Melillo. 1989. Nitrogen Saturation in
Northern Forest Eco ystems, Bioscience, 39: 378-385.
Adkison, M.D. 2010. Models of the effects of marine-derived nutrients on salmon
(Oncorhynchus spp.) population dynamics. Canadian Journal of Fisheries and
Aquatic Sciences, 67: 5-15.
Albers, S.J. 2010. The Salmon Disturbance Regime: Effects on Biofilm, Sediment and
Water. Masters thesis, University of Northern British Columbia, Prince
George.
Albers, S.J. & E.L. Petticrew. 2012. Ecosy tern response to a salmon disturbance
regime: Implications for downstream nutrient fluxes in aquatic systems.
Limnology and Oceanography, 57(1 ): 113-123.
Albers, S.J. & E.L. Petticrew. 2013. Biogeomorphic impacts of migration and
disturbance: Implications of salmon spawning and decay. Geomorphology, 202:
43-50.
Allan, J.D., M.S. Wipfli, J.P. Caouette, A. Prussian & J. Rodgers. 2003. Influence of
streamside vegetation on inputs of terrestrial invertebrates to salmonid food
webs. Canadian Journal of Fisheries and Aquatic Sciences, 60: 309-320.
Allan, J.D. & M. M. Castillo. 2007. Stream ecology: structure and function of running
waters. 2nd edition. Springer, Dordrecht, The Netherlands.
Allen, A.R., B.J. Peterson, A.D. Huryn, M.B. Flinn & W.B Bowden. 2009. Foodweb
characteristics based on stable isotopic analy es of carbon and nitrogen from
streams in the central and upper Noatak River basin: Gates of the Arctic Park
and Preserve and the Noatak National Preserve, Alaska. Final Report to the
National Park Service Arctic Network, pp. 1-24.
Alley, W.M., R.W. Healy, J.W. LaBaugh & T.E. Reilly. 2002. Flow and Storage in
Groundwater Sy terns. Science, 296: 1985-1990.
Anderson, C. & G. Cabana. 2007. Estimating the trophic position of aquatic
consumers in river food webs using stable nitrogen isotopes. Journal of the
North American Benthological Society, 26(2): 273-285.
Anderson, J.H., G.R. Pess, R.W. Carmichael, M.J. Ford, T.D. Cooney, C.M. Baldwin
& M.M. McClure. 2014. Planning Pacific Salmon and Steelhead
Reintroductions imed at Long-Term Viability and Recovery. North American
Journal of Fisheries Management, 34(1): 72-93.

Bibliography

161

Anderson, N.H. & J.R. Sedell. 1979. Detritus processing by macroinvertebrates in
stream ecosy terns. Annual Review of Entomology, 24:351-377.
Argerich, A. E. Marti, F. Sabater, M. Ribot, D. von Shiller & J.L. Riera. 2008.
Combined Effects of Leaf Litter Inputs and a Flood on Nutrient Retention in a
Mediterranean Mountain Stream During Fall. Limnology and Oceanography,
53(2): 631-641.
Argerich, A. E. Marti, F. Sabater & M. Ribot. 2011. Temporal variation of
hydrological exchange and hyporheic biogeochemistry in a headwater stream
during autumn. Journal of the North American Benthological Society, 30(3): 635652.
Ashkenas, L.R., S.L. Johnson, S.V. Gregory, J.L. Tank & W.M. Wollheim. 2004. A
stable isotope tracer study of nitrogen uptake and transformation in an oldgrowth fore t stream. Ecology, 85(6): 1725-1739.
Baldwin, D.H., J.F. Sandahl, J.S. Labenia & N.L. Scholz. 2003. Sublethal effects of
copper on coho salmon: Impacts on nonoverlapping receptor pathways in the
peripheral olfactory nervous system. Environmental Toxicology and Chemistry,
22: 2266-2274.
Ballinger, A. & P.S. Lake. 2006. Energy and nutrient fluxes from rivers and streams
into terrestrial food webs. Marine and Freshwater Research, 57: 15-28.
Barnes, C. & S. Jennings. 2007. Effect of temperature, ration, body size and age on
sulfur isotope fractionation in fish. Rapid Communications in Mass Spectrometry,
21: 1461-1467.
Bartz, K.K. & R.J. Naiman. 2005. Effects of Salmon-Borne Nutrients on Riparian Soils
and Vegetation in Southwest Alaska. Ecosystems, 8: 529-545.
Beamish, R.J., C. Neville, R. Sweeting & K. Lange. 2012. A synchronous failure of
juvenile Pacific salmon and herring production in the Strait of Georgia in 2007
and the poor return of sockeye salmon to the Fraser River in 2009. Marine and
Coastal Fisheries: Dynamics, Management, and Ecosystem Science, 4:4 03-414.
Ben-David, M., T.A. Hanley & D.M. Schell. 1998. Fertilization of terrestrial
vegetation by spawning Pacific salmon: the role of flooding and predator
activity. Oikos, 83: 47-55.
Benke, A.C. & C.E. Cushing. 2005. Rivers of North America. Elsevier Academic
Press.
Betlach, M.R. & J.M. Tiedje. 1981. Kinetic Explanation for Accumulation of Nitrite,
Nitric Oxide, and Nitrous Oxide During Bacterial Denitrification. Applied and
Environmental Microbiology, 42(6): 1074-1084.

Bibliography

162

Bilby, R.E. & P.A. Bisson. 1992. Allochthonous versus autochthonous organic matter
contributions to the trophic support of fish populations in clear-cut and oldgrowth forested streams. Canadian Journal of Fisheries and Aquatic Sciences, 49:
540-551.

Bilby R.E., B.R. Fransen, & P.A. Bisson. 1996. Incorporation of nitrogen and carbon
from spawning coho salmon into the trophic system of small streams:
evidence from stable isotopes. Canadian Journal of Fisheries and Aquatic Science,
53: 164-173.

Bilby, R.E., B.R. Fransen, P.A. Bisson & J.K. Walter. 1998. Response of juvenile coho
salmon( Oncorhynchus kisutch) and steelhead (Oncorhynchus mykiss) to the
addition of salmon carcasses to two streams in southwestern Washington,
U.S.A. Canadian Journal of Fisheries and Aquatic Science, 55: 1909-1918.
Bilby, R.E., E.W. Beach, B.R. Fransen, J.K. Walter & P.A. Bisson. 2003. Transfer of
Nutrients from pawning Salmon to Riparian Vegetation in Western
Washington. Transactions of the American Fisheries Society, 132: 733-745.
Billings, S.A. & D.O. Richter. 2006. Changes in stable isotopic signatures of soil
nitrogen and carbon during 40 years of forest development. Oecologia, 148:
325-333.

Binley, A., S. Ullah, A.L. Heathwaite, C. Heppell, P. Byrne, K. Lansdown, M.
Trimmer & H. Zhang. 2013. Revealing the spatial variability of water fluxes at
the groundwater-surface water interface. Water Resources Research, 49: 39783992.

Bo, T., M. Cucco, S. Fenoglio & G. Malacarne. 2006. Colonisation patterns and
vertical movements of stream invertebrates in the interstitial zone: a case
study in the Apennines, NW Italy. Hydrobiologia, 568: 67-78.
Boano, F., C. Camporeale, R. Revelli & L. Ridolfi. 2006. Sinuosity-driven hyporheic
exchange in meandering rivers. Geophysical Research Letters, 33: 1-4.
Borch, T., R. Kretzschmar, A. Kappler, P. Van Cappellen, M. Cinder-Vogel, A.
Voegelin & K. Campbell. 2010. Biogeochemical redox processes and their
impact on contaminant dynamics. Environmental Science & Technology, 44: 1523.

Boulton, A.J. & E.H. Stanley. 1995. Hyporheic processes during flooding and drying
in a Sonoran Desert stream. II. Faunal dynamics. Archiv for Hydrobiologie, 134:
27-52.

Boulton, A.J. & E.H. Stanley. 1996. But the story gets better: subsurface invertebrates
in stream ecosystems. TREE, 11(10) 430.

Bibliography

163

Boulton, A.J., S. Findlay, P. Marmonier, E.H. Stanley, & H.M. Valett. 1998. The
functional significance of the hyporheic zone in streams and rivers. Annual
Review of Ecology, Evolution, and Systematics, 29: 59-81.
Bouwer, H. 1989. The Bouwer and Rice slug test- an update. Ground Water, 27: 304309.
Bretschko, G. 1995. River I land ecotones: scales and patterns. Hydrobiologia, 303: 8391.
Brown, A.F. & M. Diamond. 1984. The consumption of Rainbow Trout (Salmo
gairdneri Richardson) eggs by macro invertebrates in the field. Freshwater
Biology, 14: 211-215.
Brown, L.R., P.B. Moyle & R.M. Yoshiyama. 1994. Historical decline and current
status of coho salmon in California. North American Journal of Fisheries
Management,14:,237-261.
Brune, A., P. Frenzel & H. Cypionka. 2000. Life at the oxic-anoxic interface: microbial
activities and adaptations. FEMS Microbiology Reviews, 24: 691-710.
Burford, J.E., S.J. Dery & R.D. Holmes. 2009. Some aspects of the hydroclimatology
of the Quesnel River Basin, British Columbia, Canada. Hydrological Processes,
DOl: 10.1002/hyp.7253.
Burnham, K.P. & D.R. Anderson. 2002. Model Selection and Multimodel Inference: a
practical information-theoretic approach. Second edition. Springer: New
York.
Campbell, M.D., M. Sneed Starrett, J.D. Fowler & J.J. Klein. 1990. Slug tests and
hydraulic conductivity. Ground Water Management, 4: 85-99.
Campbell, E.Y., R.W. Merritt, K.W. Cummins & M.E. Benbow. 2012. Spatial and
Temporal Variability of Macroinvertebrates in Spawning and Non-Spawning
Habitats during a Salmon Run in Southeast Alaska. PloS ONE, 7(6): 1-8.
Cardiff, M., W. Barrash, M. Thoma & B. Malama. 2011. Information content of slug
tests for estimating hydraulic properties in realistic, high-conductivity aquifer
scenarios. Journal of Hydrology, 403: 66-82.
Carlyle, G.C. & A.R. Hill. 2001. Groundwater phosphate dynamics in a river riparian
zone: effects of hydrologic flowpaths, lithology and redox chemistry. Journal
of Hydrology, 247: 151-168.
Casey, R.E., M.D. Taylor & S.J. Klaine. 2001. Mechanisms o~ Nutrient Attenuation in
a Subsurface Flow Riparian Wetland. Journal of Envuonmental Quality, 30:

Bibliography

164

1732-1737.
Cedarholm, C.J., D.B. Houston, D.L. Cole & W.J. Scarlett. 1989. Fate of coho salmon
(Oncorhynchus kisutch) carcasses in spawning streams. Canadian Journal of
Fisheries and Aquatic Science, 46: 1347-1355.
Cedarholm, C.J., M.D. Kunze, T. Murota & A. Sibatani. 1999. Pacific Salmon
Carcasses: Essential Contributions of Nutrients and Energy for Aquatic and
Terrestrial Ecosystems. Fisheries, 24(10): 1-10.
Chafiq, M. J. Gibert & C. Claret. 1999. Interactions among sediments, organic matter,
and microbial activity in the hyporheic zone of an intermittent stream.
Canadian Journal of Fisheries and Aquatic Science, 56: 487-495.
Chaloner, D.T. & M.S. Wipfli. 2002. Influence of decomposing Pacific salmon
carcasses on macroinvertebrate growth and standing stock in southeastern
Alaska streams. Journal of the North American Benthologial Society, 21(3): 430442.
Chaloner, D.T., M.S. Wipfli, & J.P. Caouette. 2002a. Mass loss and macroinvertebrate
colonisation of Pacific salmon carcasses in south-eastern Alaskan streams.
Freshwater Biology, 47: 263-273.
Chaloner, D.T., K.M. Martin, M.S. Wipfli, P.H. Ostrom & G.A. Lamberti. 2002b.
Marine carbon and nitrogen in southeastern Alaska stream food webs:
evidence from artificial and natural streams. Canadian Journal of Fisheries and
Aquatic Science, 59(8): 1257-1265.
Channing, A. A. Zamuner, D. Edwards & D. Guido. 2011. Equisetum thermale sp. nov.
(Equisetales) from the Jurassic San Agustin hot spring deposit, Patagonia:
Anatomy, paleoecology, and inferred paleoecosphysiology. American Journal
of Botany, 98(4): 680-697.
Chapin, F.S., P.A. Matson &P.M. Vitousek. 2011. Principles of Terrestrial Ecosystem
Ecology, Second Edition. Springer, New York.
Chatfield, C. 2004. The analysis of time series, an introduction, sixth edition: New
York, Chapman & Hall I CRC.
Chestnut, T.J. & W.H. McDowell. 2000. C and N Dynamics in the Riparian and
Hyporheic Zones of a Tr?phic Stream,. Luqui~lo Mountains, Puerto Rico.
Journal of the North Amerzcan Benthologzcal Soczety, 19(2): 199-214.
Christie, K.S., M.D. Hocking & T.E. Reimchen. 2008. Tracing salmon nutrients in
riparian food webs: isotopic evidence in a ground-foraging passerine.
Canadian Journal of Zoology, 86: 1317-1323.

Bibliography

165

Claeson~ S._M., J.L. Li, J.E ..c~mpton & P.A. Bisson. 2006. Response of nutrients,

btofilm, and benthic Insects to salmon carcass addition. Canadian Journal of
Fisheries and Aquatzc Sczence, 63: 1230-1241.
Clampitt, P.T. 1974. easonal migratory cycle and related movements of the
freshwater pulmonate snail, Physa integra. Amerzcan Midland Naturalist, 92:
275-300.
Claret, C. & A.J. Boulton. 2009. Integrating hydraulic conductivity with
biogeochemical gradients and microbial activity along river-groundwater
exchange zones in a subtropical stream. Hydrogeology Journal, 17: 151-160.
Clark, D.E. 2013. Effect of Mining on Fine Sediment Geochemistry in the Horsefly
River Catchment, British Columbia, anada. Masters thesis, Utrecht
University, Netherlands.
Clary, W .P. & D.E. Medin. 1999. Riparian Zones- The Ultimate Ecotones? USDA
Forest Service Proceedings, RMRS-P-11. pp. 49-55.
Clifford, H.F. 1991. Aquatic Invertebrates of Alberta: An Illustrated Guide.
University of Alberta Press, Edmonton, Alberta.
Cohen, B. I. 2012. The uncertain future of Fraser River sockeye. In: Final Report,
Cohen Commission of Inquiry Into the Decline of Sockeye Salmon in the
Fraser River (Canada). Vol1, Chapter 4. DFO Overview: Organization,
Science, Policies.
Coleman, M.J. & H.B.N. Hynes. 1970. The vertical distribution of the invertebrate
fauna in the bed of a stream. Limnology and Oceanography, 15: 31-40.
Collier, M.H., B. Keane & S.H. Rugstad. 2010. Productivity differences between
dandelion (Taraxacum officinale; Asteraceae) clones from pollution impacted
versus non-impacted soils. Plant Soil, 329: 173-183.
Compton, J.E., M.R. Church, S.T. Lamed & W.E. Hogsett. 2003. Nitrogen export
from forested watersheds in the Oregon Coast Range: the role of Nrfixing red
alder. Ecosystems, 6: 773-785.
Constantz, J., C.L. Thomas & G. Zellweger. 1994. Influence of diurnal variations in
stream temperature on streamflow loss and groundwater recharge. Water
Resources Research, 30: 3253-3264.
Costerton, J., z. Lewandowski, D. Caldwell, D. Korber & H. Lappinscott. 1995.
Microbial Biofilms. Annual Review Of Microbiology, 49: 711-745.
Cothran, R.D., A.B. Stoler & R.A. Relyea. 2014. Leaves and litterbugs: how litter
quality affects amphipod life-history and sexually-selected traits. Freshwater

Bibliography

166

Science, 33(3): 812-819.
raine, J.M. 2011. Nitrogen Isotopes. In F.S. Chapin, P.A. Matson & P.M. Vitousek
(eds.) Principles of Terrestrial Ecosystem Ecology, Second Edition. Springer,
New York. pp. 283.
Cummins, K.W. 1974. Structure and function of stream ecosyst ms. BioScience, 24:
631-641.
Cummins, K.W., M.A. Wilzbach, D.M. Gates, J.B. Perry & W.B. Taliaferro. 1989.
Shredders and Riparian Vegetation. BioScience, 39(1 ): 24-30.
Danell, K. & K. Sjoberg. 1977. Seasonal emergence of chironomids in relation to
egglaying and hatching of ducks in a restored lake (northern Sweden).
Wildfowl, 28:129-135.
Darimont, C.T., H.M. Bryan, S.M. Carlson, M.D. Hocking, M. MacDuffee, P.C.
Paquet, M.H.H. Price, T.E. Reimchen, J.D. Reynolds & C.C. Wilmers. 2010.
Salmon for terrestrial protected areas. Conservation Letters, 3: 379-389.
Dearden R. & B. Palumbo-Roe. 2010. Technical Note: Hyporheic Zone Sampling
Procedures. British Geological Survey Open Report, OR/10/048. pp. 21.
Decamps, H & R.J. Naiman. 1990. Towards an Ecotone Perspective. In R.]. Naiman
and H. Decamps (ed.) Ecology and management of aquatic-terrestrial
ecotones, Man and the Biosphere Series, Unesco.
Dekar, M.P., D.O. Magoulick & G.R. Huxel. 2009. Shifts in the trophic base of
intermittent stream food webs. Hydrobiologia, 635: 263-277.
DeMarsily, G. 1986. Quantitative Hydrogeology, Academic Press.
DeNiro, M.J. & S. Epstein. 1978. Influence of diet on the distribution of carbon
isotopes in animals. Geochimica et Cosmochimica Acta, 42: 495-506.
DeNiro, M.J. & S. Epstein. 1981. Influence of diet on the distribution of nitrogen
isotopes in animals. Geochimica et Cosmochimica Acta, 45: 341-351.
Dent, C.L., N.B. Grimm & S.G. Fisher. 2001. Multiscale effects of surface-subsurface
exchange on stream water nutrient concentrations, Journal of the North
American Benthological Society, 20(2): 162-181.
Des Marais, D.J. 2000. Wh n Did Photosynthesis Emerge on Earth? Science,
289(5485): 1703-1705.
DeVries, P. 2012. Salmonid influences on rivers: A geomorphic fish tail.
Geomorphology, 157-158: 66-74.

Bibliography

167

Doi, H. E. Kikuchi. S. Takagi & S. hikano. 2007. Changes in carbon and nitrogen
stable i otopes of chironomid larvae during growth, starvation and
metamorphosis. Rapid Communications in Mass Spectrometry, 21: 997-1002.
Drake, D.C. & R.J. Naiman. 2007. Reconstruction of Pacific Salmon Abundance from
Riparian Tree-Ring Growth. Ecological Applications, 17(5): 1523-1542.
Drake, D.C., R.J. Naiman & J.S. Bechtold. 2006. Fate of Nitrogen in Riparian Forest
Soils and Trees: An 15N Tracer Study Simulating Salmon Decay. Ecology,
87(5): 1256-1266.
Drake, D.C., P.J. Sheppard & R.J. Naiman. 2011. Relationships between salmon
abundance and tree-ring 15N: three objective tests. Canadian Journal of Forestry
Research, 41: 2423-2432.
Dohet, A., D. Hlubikova, C.E. Wetzel, L. L'Hoste, J.F. lffly, L. Hoffmann & L. Ector.
2015. Influence of thermal regime and land use on benthic invertebrate
communities inhabiting headwater streams exposed to contrasted shading.
Science of the Total Environment, 505: 1112-1126.
Doucett, R.R., G. Power, D.R. Barton, R.J. Drimmie & R.A. Cunjak. 1996. Stable
isotope analysis of nutrient pathways leading to Atlantic salmon. Canadian
Journal of Fisheries and Aquatic Science, 53: 2058-2066.
Duff, J.H. & F.J. Triska. 2003. Nitrogen biogeochemistry and surface-subsurface
exchange in streams. In J.B. Jones & P.J. Mulholland (eds.). Streams and
Ground Waters. Academic Press, San Diego.
Dunne, J.A., U. Brose, R.J. Williams & N.D. Martinez. 2005. Modeling food-web
dynamics: complexity-stability implications. In A. Belgrano, U.M. Scharler,
J.A. Dunne & R.E. Ulanowicz (eds.) Aquatic Food Webs An Ecosystem
Approach, Oxford University Press, pp. 117-129.
Duval, T.P. & A.R. Hill. 2007. Influence of base flow stream bank seepage on riparian
zone nitrogen biogeochemistry. Biogeochemistry, 85: 185-199.
Eggers, D.M. 1978. Limnetic feeding behavior of juvenile sockeye salmon in Lake
Washington and predator avoidance. Limnology and Oceanography,23(6): 11141125.
Ehleringer, J.R. N. Buchmann & L. B. Flanagan. 2000. Carbon Isotope Ratios in
Belowground Carbon Cycle Processes. Ecological Applications, 10(2): 412-422.
Evans, F.C. 1956. Ecosystem as the Basic Unit in Ecology. Science, 123(3208): 11271128.

Bibliography

168

ESRI. 2013. ArcGIS De ktop: Release 10.1. Redlands, California: Environmental
Sy terns Research Institute.
Evenden, M.D. 2007. Fish versus Power: An Environmental History of the Fraser
River. ambridge: Cambridge University Press.
Fausch, K.D. & T.G. Northcote. 1992. Large Woody Debris and Salmonid Habitat in
a Small Coastal British Columbia Stream. Canadian Journal of Fisheries and
Aquatic Science, 49: 682-693.
Febria, C.M, P. Beddoes, R.R. Fulthorpe, & D.O. Williams. 2011. Bacterial
community dynamics in the hyporheic zone of an intermittent stream. The
ISME Journal, 1-11.
Feminella, J.W. & V.H. Resh. 1991. Herbivorous caddisflies, macroalgae, and
epilithic microalgae: dynamic interactions in a stream grazing system.
Oecologia, 87: 247-256.
Ficke, A.D., C.A. Myrick & L.J. Hansen. 2007. Potential impacts of global climate
change on freshwater fisheries. Reviews in Fish Biology and Fisheries, 17: 581613.
Fiebig, D.M. 1995. Groundwater discharge and its contribution of dissolved organic
carbon to an upland stream. Archiv fur Hydrobiologie, 134: 129-155.
Fillatre Miller, E.K., I.R. Bradbury & D.O. Heath. 2011. Juvenile habitat partitioning
and relative productivity in allochronically isolated sockeye salmon
(Oncorhynchus nerka). Ecology and Evolution, doi: 10.1002/ ece3.55.
Findlay, S. 2003. Bacterial Response to Variation in Dissolved Organic Matter. In
S.E.G. Findlay & R.L. Sinsabaugh (eds.) Aquatic Ecosystems: Interactivity of
Dissolved Organic Matter, Elsevier Science, USA.
Finlay, J.C. & C. Kendall. 2007. Stable isotope tracing of temporal and spatial
variability in organic matter sources to freshwater ecosystems. In R.H.
Michener & K. Lajtha (eds.). Stable Isotopes in Ecology and Environmental
Science, Second Edition. Blackwell Publishing, pp. 283-333.
Fischer, H. 2003. The Role of Biofilms in the Uptake and Transformation of
Dissolved Organic Matter. In S.E.G. Findlay & R.L. Sin abaugh (eds.) Aquatic
Ecosystems: Interactivity of Dissolved Organic Matter, Elsevier Science, USA.
Fisher, S.G. 1997. Creativity, idea generation, and the functional morphology of
streams. Journal of the North American Benthological Society, 16(2): 305-318.
Flecker, A.S. 1984. The Effects of Predation and Detritus on the Structure of a Stream
Insect Community: A Field Test. Oecologia, 64(3): 300-305.

Bibliography

169

Flint, L.E. & A.L. Flint. 2002. Porosity. In J.H. Dane & G.C. Topp (eds.) Methods of
soil analysis: Part 4. Physical methods. SSSA Book Series 5, Madison, WI.
Foreman, C.M. & J.S. overt. 2003. Linkage between Dissolved Organic Matter
Composition and Bacterial Community Structure. In S.E.G. Findlay & R.L.
Sinsabaugh (eds.) Aquatic Ecosy terns: Interactivity of Dissolved Organic
Matter, Elsevier cience, USA.
France, R.L. 1992. Biogeographical variation in size-specific fecundity of the
amphipod Hyalella azteca. Crustaceana, 62: 240-248.
France, R.L. 1994. Nitrogen isotopic composition of marine and freshwater
invertebrates. Marine Ecology Progress Series, 115: 205-207.
France, R.L. 1995. Differentiation between littoral and pelagic food webs in lakes
using stable carbon isotopes. Limnology and Oceanography, 40(7): 1310-1313.
Freeze, R.A. & J.A. Cherry. 1979. Groundwater. Prentice-Hall, Englewood Cliffs,
New Jersey.
Friberg, N. & D. Jacobsen. 1999. Variations in growth of the detrivores-shredder
Sericostoma personatum (Tricoptera). Freshwater Biology, 42: 625-635.
Fry, B. 2006. Stable isotope ecology. Springer, New York.
Gardner, K. M. 1999. The importance of surface water I groundwater interactions.
Issue paper EPA-910-R-99-013. Seattle, Washington, US EPA, 23 pp.
Garfinkel, H.L. & L.B. Brubaker. 1980. Modem climate-tree growth relationships and
climatic reconstruction in sub-arctic Alaska. Nature, 286: 872-874.
Garten, C.T., P.J. Hanson, D.E. Todd, B.B. Lu & D.J. Brice. 2007. Natural 15 N- and 13Cabundance as indicators of forest nitrogen status and soil carbon dynamics. In
R. Michener & K. Lajtha (ed.) Stable Isotopes in Ecology and Environmental
Science, Blackwell Publishing, USA.
Geiger, H.J. & Xinxian Zhang. 2002. A Simple Procedure to Evaluate Salmon
Escapement Trends that Emphasizes Biological Meaning Over Statistical
Significance. Alaska Fishery Research Bulletin, (9)2: 128-134.
Geist, D.R. & D.O. Dauble. 1998. Redd Site Selection and Spawning Habitat Use by
Fall Chinook Salmon: The Importance of Geomorphic Features in Large
Rivers. Environmental Management, 22(5): 655-669.
Gende, S.M., T.P. Quinn & M.F. Wilson. 2001. Consumption choice by bears feeding
on salmon. Oecologia, 127: 372-82.

Bibliography

170

Gende, S.M., R.T. Edwards, M.F. Willson & M.S. Wipfli. 2002. Pacific salmon in
aquatic and terrestrial ecosystems. BioScience, 52(10): 917-928.
Genereux, D.P., S. Leahy, H. Mitasova, C.D. Kennedy & D.R. Corbett. 2008. Spatial
and temporal variability of streambed hydraulic conductivity in West Bear
Creek, North Carolina, USA. Journal of Hydrology, 358: 332-353.
Gilbert J, Dole-Olivier M, Marmonier P, Vervier P. 1990. Surface water-groundwater
ecotones. In Ecology and management of aquatic-terrestrial ecotones. Edited
by R.J. Naiman & H. Decamps. UNESCO, Paris, France.
Gordon, N.D., T.A. McMahon & B.L. Finlayson. 1992. Dissecting Data with a
Statistical Scope. In Stream Hydrology: An Introduction for Ecologists. John
Wiley & ons, New York, pp. 346-402.
Gra~a, M.A.S. 2001. The Role of Invertebrates on Leaf Litter Decomposition in

Streams- a Review. International Review of Hydrobiology, 86: 383-393.
Greenwald, M.J., W.B. Bowden, M.N. Gooseff, J.P. Zametske, J.P. McNamara, J.H.
Bradford & T.R. Brosten. 2008. Hyporheic exchange and water chemistry of
two arctic tundra streams of contrasting geomorphology. Journal of
Geophysical Research, 113: 1-14, doi:10.1029/2007JG000549.
Grimm, N.B. & S.G. Fisher. 1984. Exchange between interstitial and surface water:
implications for stream metabolism and nutrient cycling. Hydrobiologia, 111:
219-228.
Grimm, N.B., S.G. Fisher, H.M. Valett & E.H. Stanley. 1991. Contribution of the
hyporheic zone to the stability of an arid lands stream. Verhandlugen der
Internationalen Vereinigungfor Theoretische und Angewandte Limnologie, 24:
1595-1599.
Gosz, J.R. 1993. Ecotone Hierarchies. Ecological Applications, 3(3): 369-376.
Guill, C., B. Drossel, W. Just & E. Carmack. 2012. A three-species model explaining
cyclic dominance of Pacific salmon. Journal of Theoretical Biology, 276: 16-21.
Guyette, M.Q., C.S. Loftin, J. Zydlewski & R. Cunjak. 2014. Carcass analogues
provide marine subsidies for macroinvertebrate and juvenile Atlantic salmon
in temporate oligotrophic streams. Freshwater Biology, 59: 392-406.
Gworek, B. A. Deckowska & M. Pierscieniak. 2011. Traffic Pollutant Indicators:
Common Dandelion (Taraxacum officinale), Scots Pine (Pinus silvestris), SmallLeaved Lime (Tilia cordata). Polish Journal of Environmental Studies, 20: 87-92.
Hall, R.O., J.L. Tank, D.J. Sobota, P.J. Mulholland, J.M. O'Brien, W.K. Dodds, J.R.

Bibliography

171

Webster, H.M. Valett, G.C. Poole, B.J. Peterson, J.L. Meyer, W.H. McDowell,
S.L. Johnson, S.K. Hamilton, N.B. Grimm, S.V. Gregory, C.N. Dahm, LW.
Cooper & others. 2009. Nitrate removal in stream ecosystems measured by
15N addition experiments: Total uptake. Limnology and Oceanography, 54(3):
653-665.
Hancock, P.J. 2002. Human impacts on the stream-groundwater exchange zone.
Environmental Management, 29: 763-781.
ague, M.J., M.R. Ferrari, J.R. Miller, D.A. Patterson, G.L. Russells, A.P. Farrell & S.
G. Hinch. 2011. Modelling the future hydroclimatology of the lower Fraser
River and its impacts on the spawning migration survival of sockeye salmon.
Global Change Biology, 17: 87-98.
Hancock, P.J. & A.J. Boulton. 2009. Sampling groundwater fauna: efficiency of rapid
assessment methods tested in bores in eastern Australia. Freshwater Biology,
54: 902-917.
Harvey, J.W. & K.E. Bencala. 1993. The Effect of Streambed Topography on SurfaceSubsurface Water Exchange in Mountain Catchments. Water Resources
Research, 29(1): 89-98.
Harvey, J.W. & B.J. Wagner. 2000. Quantifying Hydrologic Interactions between
Streams and Their Subsurface Hyporheic Zones. In Streams and Ground
Waters, Edited by Jeremy Jones & Patrick Mulholland. Academic Press.
Harvey, J.W., J.K. Bohlke, M.A. Voytek, D. Scott & C.R. Tobias. 2013. Hyporheic
zone denitrification: Controls on effective reaction depth and contribution to
whole-stream mass balance. Water Resources Research, 49: 6298-6316.
Hauke, R. L. 1963. A taxonomic monograph of the genus Equisetum subgenus
Hippochaete. Nova Hedwigia, 8: 1-123.
Hawkins, C.P. & J.R. Sedell. 1981. Longitudinal and Seasonal Changes in Functional
Organization of Macroinvertebrate Communities in Four Oregon Streams.
Ecology, 62(2): 387-397.
Helfield J.M. & R.J. Naiman. 2001. Effects of salmon-derived nitrogen on riparian
forest growth and implications for stream habitat. Ecology, 82: 2403-2409.
Helfield J.M. & R.J. Naiman. 2006. Keystone Interactions: Salmon and Bear in
Riparian Forests of Alaska. Ecosystems, 9: 167-180.
Henderson, M.A. & C.C. Graham. History and current status of Pacific salmon in
British Columbia. North Pacific Anadramous Fish Commission Bulletin, 1: 13-22.
Heppell, C., A.L. Heathwaite, A. Binley, P. Byrne, S. Ullah, K. Lansdown, P. Keenan,

Bibliography

172

M. Trimmer & H. Zhang. 2014. Interpreting spatial patterns in redox and
coupled water-nitrogen fluxes in the streambed of a gaining river reach.
Bzogeochemrstry, 117: 491-509.
Hesslein, R.H., M.J. Capel, D.E. Fox & K.A. Hallard. 1991. Stable isotopes of sulfur,
carbon, and nitrogen as indicators of trophic level and fish migration in the
lower Mackenzie River basin, Canada. Canadian Journal of Fisheries and Aquatic
Sciences, 48: 2258--2265.
Hester, E.T. 2008. Impact of Geomorphic Structures on Hyporheic Exchange,
Temperature, and Ecological Processes in Streams. Ph.D. dissertation,
University of North Carolina, Chapel Hill.
Hester, E. & M. Gooseff. 2010. Moving beyond the banks: Hyporheic restoration is
fundamental to restoring ecological services and functions of streams.
Envzronmental Science Technology, 44: 1521-1525.
Hicks, B.J., M.S. Wipfli, D.W. Lang & M.E. Lang. 2005. Marine-derived nitrogen and
carbon in freshwater-riparian food webs of the Copper River Delta,
southcentral Alaska. Oecologia, 144: 558--569.
Hilderbrand, G., T. Hanley, C. Robbins, & C. Schwartz. 1999a. Role of brown bears
(Ursus arctos) in the flow of marine nitrogen into a terrestrial ecosystem.
Oecologia, 121: 546-550.
Hilderbrand, G., S. Jenkins, C. Schwartz, T. Hanley, & C. Robins. 1999b. Effect of
seasonal differences in dietary meat intake on changes in body mass and
composition in wild and captive brown bears. Canadian Journal of Zoology, 77:
1623-1630.
Hill, A.R., C.F. Labadia & K. Sanmugadas. 1998. Hyporheic Zone Hydrology and
Nitrogen Dynamics in Relation to the Streambed Topography of aN-Rich
Stream. Biogeochemistry, 42(3): 285-310.
Hill, R.A. & C.P. Hawkins. 2014. Using modelled stream temperatures to predict
macro-spatial patterns of stream invertebrate biodiversity. Freshwater Biology,
59: 2632-2644.
Hinkle, S.R., J.H. Duff, F.J. Triska, A. Laenen, E.B. Gates, K.E. Bencalada, D.A. Wentz
& S.R. Silva. 2001. Linking hyporheic flow and nitrogen cycling near the
Willamette River- a large river in Oregon, USA. Journal of Hydrology, 244:
157-180.
Hobbs, W. & A. Wolfe. 2008. Recent paleolimnology of three lakes in the Fraser
River Basin (BC, Canada): no respons~ to the collapse of so.ckeye salmon
stocks following the Hells Gate landslides. Journal of Paleolzmnology, 40: 295308.

Bibliography

173

Hocking, M.D. & T.E. Reimchen. 2002. Salmon-derived nitrogen in terrestrial
invertebrates from coniferous forests of the Pacific Northwest. BMC Ecology,
2:4.
Hocking, M.D. & T.E. Reimchen. 2009. Salmon species, density and watershed size
predict magnitude of marine enrichment in riparian food webs. Oikos, 118:
1307-1318.
Hodson, E., F. Shahid, J. Basinger & S. Kaminsky. 2009. Fungal endorhizal associates
of Equisetum species from Western and Arctic anada. Mycological Progress,
8(1 ): 19-27.
Hoellein, T.J., J.L. Tank, E.J. Rosi-Marshall, S.A. Entrekin & G.A. Lamberti. 2007.
Controls on spatial and temporal variation of nutrient uptake in three
Michigan headwater streams. Limnology and Oceanography, 52(5): 1964-1977.
Holmes, R. 2008. Horsefly River Watershed Based Fish Sustainability Plan Watershed Profile. Technical report, Cariboo Envirotech, Ltd., Likely, British
Columbia.
Holtgrieve, G.W., D.E. Schindler, C. Gowell, C.P. Ruff & P.J. Lisi. 2010. Stream
geomorphology regulates the effects of ecosystem engineering and nutrient
enrichment by Pacific salmon on periphyton. Freshwater Biology, 55: 2598-2611.
Honea, J.M. & R.I. Gara. 2009. Macroinvertebrate community dynamics: strong
negative response to salmon redd construction and weak response to salmonderived nutrient uptake. Journal of the North American Benthological Society,
28(1): 207-219.
Husby, Chad. 2013. Biology and functional ecology of Equisetum with emphasis on
the giant horsetails. The Botanical Review, published online 15 January 2013.
Hutchinson, D.G. & R.D. Moore. 2000. Throughflow variability on a forested
hillslope underlain by compacted glacial till. Hydrological Processes, 14: 17511766.
Hvorslev, M.J. 1951. Time lag and soil permeability in groundwater observations.
Waterways Experimental Station Bulletin 36. US Army Corps of Engineers,
Vicksburg, Mississippi.
Hynes, H.B.N. 1974. Further studies on the distribution of stream animals within the
substratum. Limnology and Oceanography, 19(1 ): 92-99.
International Union for the Conservation of Nature Salmonid Specialist Group
(IUCN- SSG), 2013. Red List Assessment of Sockeye Salmon (Oncorhynchus
nerka) Retrieved January 30, 2014 from

Bibliography

174

http: I I www .stateofthesalmon.org I iucn I.
Ito, T. 2003. Indirect effect of salmon carcasses on growth of a freshwater amphipod,
Jesogammarus jesoensis (Gammaridea): An experimental study. Ecological
Research, 18: 81-89.
Jackson, B.S. 1982. The lowly dandelion deserves more respect. Canadian Geographer,
102:54-59.
Jacob, C., T. McDaniels & S. Hinch. 2010. Indigenous culture and adaptation to
climate change: ockeye salmon and the St' at'imc people. Mitigation and
Adaptation Strategies for Global Change, 15: 859-876.
Janetski, D.J., D.T. Chaloner, S.D. Tiegs & G.A. Lamberti. 2009. Pacific salmon effects
on stream ecosystems: a quantitative synthesis. Oecologia, 159: 583-595.
Janzen, K. 2008. Hyporheic flow in a mountainous riverine system. Masters thesis,
University of Saskatchewan Saskatoon.
Jenson, S.K. & J.O. Domingue. 1988. Extracting Topographic Structure from Digital
Elevation Data for Geographic Information System Analy is. Photogrammetric
Engineering and Remote Sensing, 54(11): 1593-1600.
Johnson, D.W. & S.E. Lindberg. 1992. Atmospheric deposition and forest nutrient
cycling. New York:Springer-Verlag.
Johnson, Z.C. J.J. Warwick & R. Schumer. 2014. Factors affecting hyporheic and
surface transient storage in a western U.S. river. Journal of Hydrology, 510: 325339.
Johnston, N.T., E.A. Macisaac, P.J. Tschaplinski & K.J. Hall. 2004. Effects of the
abundance of spawning sockeye salmon (Oncorhynchus nerka) on nutrients
and algal biomass in forested streams. Canadian Journal of Fisheries and Aquatic
Science, 61: 384-403.
Jones, J.B. & R.M. Holmes. 1996. Surface-subsurface interactions in stream
ecosystems. Trends in Ecology and Evolution, 11: 239-242.
Jost, L. 2006. Entropy and diversity. OIKOS, 113(2): 363-375.
Junk, G. & H. Svec. 1958. The absolute abundance of the nitrogen isotopes in the
atmosphere and compressed gas from various sources. Geochim. et Cosmochim.
Acta, 14: 234-243.
Kabata-Pendias, A. & S. Dudka. 1991. Trace metal contents of Taraxacum officinale
(dandelion) as a convenient environmental indicator. Environmental
Geochemistry and Health, 13(2): 108-113.

Bibliography

176

Kopp, R.E., J.L. Kirschvink, I.A. Hilburn & .z. Nash. 2005. The Paleoproterozoic
snowball Earth: A climate disaster triggered by the evolution of oxygenic
photosynthesis. Proceedings of the National Academy of Sciences of the United
States of America, 102(32): 11131-11136.
Koshino, Y. H. Kudo & M. Kaeriyama. 2013. Stable isotope evidence indicates the
incorporation into Japanese catchments of marine-derived nutrients
transported by spawning Pacific Salmon. Freshwater Biology, 58: 1864-1877.
Koyama, A., K. Kavanagh & A. Robinson. 2005. Marine nitrogen in central Idaho
riparian forests: evidence from stable isotopes. Canadian Journal of Fisheries
and Aquatic Science, 62: 518-526.
Krause, S., D.M. Hannah, J.P. Sadler & P.J. Wood. 2011. Ecohydrology on the edge:
interactions across the interfaces of wetland, riparian and groundwater-based
ecosystems. Ecohydrology, 4: 477-480.
Krko"ek, M., J.S. Ford, A. Morton, S. Lele, R.A. Myers & M.A. Lewis. 2007. Declining
wild salmon populations in relation to parasites from farm salmon. Science,
318: 1772-1775.
Lamberti, G.A. S.V. Gregory, L.R. Ashkena , J.L. Li, A.D Steinman & C.D. Mcintire.
1995. Influence of grazer type and abundance on plant-herbivore interactions
in streams. Hydrobiologia, 306: 179-188.
Lamberti, G.A. 1996. The role of periphyton in benthic food webs. In Algal ecology.
Edited by R.J. tevenson, M.L. Bothwell, and R.L. Lowe. Academic Press,
Menlo Park, Calif. pp. 533-573.
Lamberti, G.A., D.T. Chaloner & A.E. Hershey. 2010. Linkages among aquatic
ecosystems. Journal of the North American Benthological Society, 29(1): 245-263.
Leavitt, S.W. & A. Long 1986. Stable-Carbon Isotope Variability in Tree Foliage and
Wood. Ecology, 67: 1002-1010.
Leopold, L. 1994. A View of the River. Harvard University Press, Cambridge, MA
Lessard, J.L. & R.W. Merritt. 2006. Influence of marine-derived nutrients from
spawning salmon on aquatic insect communities in southeast Alaskan
streams. Oikos, 113: 334-343.
Levi, P.S., J.L. Tank, S.D. Tiegs, J. Ruegg, D.T. Chaloner & G.A. Lamberti. 2011. Does
timber harvest influence the dynamics of marine-derived nutrients in
Southeast Alaska streams? Canadian Journal of Fisheries and Aquatic Sciences, 68:
1316-1329.

Bibliography

177

Levi, P.S., J.L. Tank, S.D. Tiegs, D.T. haloner & G.A. Lamberti. 2013.
Biogeochemical transformation of a nutrient subsidy: salmon, streams, and
nitrification. Biogeochemistry, 113: 643-655.
Lohse, K.A., D.A. Newburn, J.T. pperman & A.M. Merenlender. 2008. Forecasting
Relative Impacts of Land Use on Anadrornous Fish Habitat to Guide
Conservation Planning. Ecological Applications, 18(2): 467-482.
Lorenz, M.O. 1905. Methods of Measuring the Concentration of Wealth. Publications
of the American Statistical Association, 9(70): 209-219.
Lowell, J.L., N. Gordon, D. Engstrom, J.A. Stanford, W.E. Holben & J.E. Gannon.
2009. Habitat Heterogeneity and Associated Microbial ommunity Structure
in a Small-Scale Floodplain Hyporheic Flow Path. Microbial Ecology, 58: 611620.
Lowry, J.K. & R.T. Springthorpe. 2001. Amphipoda: Families. Version 1, Retrieved
January 16,2015 from http:/ / www.crustacea.net/.
MacNiel C., J.T.A. Dick & R. Elwood. 1997. The trophic ecology of freshwater
Gammarus spp. (Crustacea: Amphipoda): Problems and perspectives
concerning the functional feeding group concept. Biological Reviews, 72: 349364.
Malard, F., K. Tockner, M. Dole-Olivier, & J.V. Ward. 2002. A landscape perspective
of surface-subsurface hydrological exchanges in river corridors. Freshwater
Biology, 47: 621-640.
Maleci, L. G. Buffa, M. Wahsha & C. Bini. 2014. Morphological changes induced by
heavy metals in dandelion (Taraxacum officinale Web.) growing on mine soils.
Journal of Soils and Sediments, 14(4): 731-743.
Malmqvist, B. 2002. Aquatic invertebrates in riverine landscapes. Freshwater Biology,
47: 679-694.
Mariotti,A., J.C. Germon, P. Hubert, P. Kaiser, R. Letolle, A. Tardieux & P. Tardieux.
1981. Experimental determination of nitrogen kinetic isotope fractionation:
some principles; illustration for the denitrification and nitrification processes.
Plant Soil, 62: 413-430.
Marsh, A.S., J.A. Arnone, B.T. Bormann & J.C. Gordon. 2000. The role of Equisetum in
nutrient cycling in an Alaskan shrub wetland. Journal of Ecology, 88: 999-1011.
Marshall, J.D., J.R. Brooks & K. Lajtha. 2007. Sources of variation in the stable
isotopic composition of plants. In R. Michener & K. Lajtha (ed.) Stable
Isotopes in Ecology and Environmental Science, Blackwell Publishing, USA.

Bibliography

178

Martins, E.G., S.G. Hinch, D.A. Patterson, M.J. Hague, S.J. Cookes, K.M. Miller, M.F.
LaPointe, K.K. English & A.P. Farrell. 2011. Effects of river temperature and
climate warming on stock-specific survival of adult migrating Fraser River
sockeye salmon (Oncorhynchus nerka). Global Change Biology, 17: 99-114.
Mason, W.T., P.A. Lewis & C.l. Weber. 1983. An evaluation of benthic
macroinvertebrate biomass methodology. Environmental Monitoring and
Assessment, 3: 29-44.
Mathisen, O.A., P.L. Parker, J.J. Goering, T.C. Kline, P.H. Poe & R.S. Scalan. 1988.
Recycling of marine elements transported into freshwater systems by
anadromous salmon. Verhandlungen des Internationalen Verein Limnologie, 23:
2,249-2,258.
Mazerolle, M.J. 2015. Model Selection and Multimodel Inference Based on (Q)AIC(c).
R Package,
http: I I cran.r-project.org I web I packages I AICcmodavg I AICcmodavg. pdf.
McClain, M.E., E.W. Boyer, C.L. Dent, S.E. Gergel, N.B. Grimm, P.M. Groffman, S.C.
Hart, J.W. Harvey, C.A. Johnston, E. Mayorga, W.H. McDowell & G. Pinay.
2003. Biogeochemical Hot Spots and Hot Moments at the Interface of
Terrestrial and Aquatic Ecosystems. Ecosystems, 6: 301-312.
McConnachie, J.L. & E.L. Petticrew. 2006. Tracing organic matter sources in riverine
suspended sediment: Implications for fine sediment transfers. Geomorphology,
79: 13-26.
McDaniels, T., S. Wilmot, M. Healy & S. Hinch. 2010. Vulnerability of Fraser River
sockeye salmon to climate change: A life cycle perspective using expert
judgments. Journal of Environmental Management, 91: 2771-2780.
McDonald, M.G. & A.W. Harbaugh. 1988. A modular three-dimensional finite
difference groundwater flow model, OR-83-875, U.S. Geological Survey,
Reston, VA.
Mei, Y. G.M. Hornberger, .A. Kaplan, J.D. Newbold & AK. Aufdenkampe. 2014.
The delivery of dissolved organic carbon from a forested hillslope to a
headwater stream in southeastern Pennsylvania, USA. Water Resources
Research, 50: 5774-5796.
Menezes, R.S.C. 2009. Isotopic evidence of the effects of herbivory and landscape
position on plant nitrogen sources in a riparian ecosystem. Isotopes in
Environmental and Health Studies, 45(1): 41-52.
Menon, P.S. 1966. Population ecology of Gammarus lacustris Sars in Big island Lake.
Ph.D. thesis. University of Alberta, Edmonton, Alberta.

Bibliography

179

Merritt, R.W. & K.W. Cummins. 1996. An Introduction to the Aquatic Insects of
North America, 3rd edn. Kendall Hunt, Iowa.
Meybeck, M. 1982. Carbon, nitrogen, and phosphorus transport by world rivers.

American Journal of Science, 282: 401-450.
Michener, R.H. & D.M. Schell. 1994. Stable isotope ratios as tracers in marine
aquatic food webs. In Stable Isotopes in Ecology and Environmental Science.
Edited by K. Lajtha & R.H. Michener. Blackwell, London, pp. 138-157.
Mihuc, T.B. & J.R. Mihuc. 1995. Trophic ecology of five shredders in a Rocky
Mountain Stream. Journal of Freshwater Ecology, 10: 209-216.
Miller, K.M., A. Teffer, S. Tucker, S. Li, A.D. Schulze, M. Trudel, F. Juanes, A. Tabata,
K.H. Kaukinen, N.G. Ginther, T.J. Ming, S.J. Cooke, J.M. Hipfner, D.A.
Patterson & S.G. Hinch. 2014. Infectious disease, shifting climates, and
opportunistic predators: cumulative factors potentially impacting wild
salmon declines. Evolutionary Applications, 7: 812-855.
Milner, A.M., A.L. Robertson, K.A. Monaghan, A.J. Veal & E.A. Flory. 2008.
Colonization and development of an Alaskan stream community over 28
years. Frontiers in Ecology and the Environment, 6: 413-419.
Minigawa, M. & E. Wada. 1984. tepwise enrichment of b15N along food chains:
further evidence and the relationship between b15N and animal age. Geochim.

Cosmochim. Acta, 48: 1135--1140.
Minakawa, N. & G.l. Gara. 1999. Ecological Effects of a Chum Salmon
(Oncorhynchus keta) Spawning Run in a Small Stream of the Pacific
Northwest, Journal of Freshwater Ecology, 14(3): 327-335.
Minakawa, N., G.I. Gara & J.M. Honea. 2002. Increased individual growth rate and
community biomass of stream insects associated with salmon carcasses.

Journal of the North American Benthological Society, 21 (4): 651-659.
Minakawa, N. & G.I. Gara. 2003. Effects of chum salmon redd excavation on benthic
communities in a stream in the Pacific northwest. Transactions of the American

Fisheries Society, 132: 598-604.
Mitchell, N. & G. Lamberti. 2005. Responses in dissolved nutrients and epilithon
abundance to spawning salmon in southeast Alaska streams. Limnology and

Oceanography, 50: 217-227.
Mitich, L. 1992. Intriguing World of Weeds: Horsetail Weed Technology, 6: 779-781.
Moore, J.W. & D.E. Schindler. 2008. Biotic disturbance and benthic community
dynamics in salmon-bearing streams. Journal of Animal Ecology, 77: 275-284.

Brbliography

180

Moore, J.W. & B.X. Semmens. 2008. Incorporating uncertaintyand prior information
into stable isotope mixing models. Ecology Letters, 11: 470-480.
Mouw, J.E.B., ].A. Stanford & P.B. Alaback. 2009. Influences of flooding and
hyporheic exchange on floodplain plant richness and productivity, River

Research and Applrcatwns, 25: 929-945.
Mulholland, P.J., S.A. Thomas, H.M. Valett, J.R. Webster & J. Beaulieu. 2006. Effects
of light on nitrate uptake in small forested streams: Diurnal and day-to-day
variations. Journal of the North American Benthological Society, 25: 583-595.
Mulholland, P.J. and others. 2009. Nitrate removal in stream ecosystems measured
by 15N addition experiments: Denitrification. Limnology and Oceanography, 54:
666-680.

Murota, T. 2003. The marine nutrient shadow: a global comparison of anadromous
salmon fishery and guano occurrence. In: Nutrients in salmonid ecosystems:
sustaining production and biodiversity. Edited by J.G. Stockner. American
Fisheries Society, Bethesda, MD, pp 17-31.
Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A.B. Da Fonseca & J. Kent. 2000.
Biodiversity hotspots for conservation priorities. Nature, 403: 853-858.
Nadelhoffer, K.J. & B. Fry. 1994. Nitrogen isotope studies in forest ecosystems. InK.
Lajtha & R. H. Michener. Stable isotopes in ecology and environmental
science. Oxford, Blackwell Scientific Publications, pp. 22-44.
Naiman, R. ]., & H. Decamps, editors. 1990. The ecology and management of
aquatic-terrestrial ecotones. UNESCO Paris & The Parthenon Publishing
Group, Paris.
Naiman, R.J. & H. Decamps. 1997. The Ecology of Interfaces: Riparian Zones. Annual

Review of Ecology and Systematics, 28: 621-658.
Naiman, R.J., H. Decamps, J. Pastor, & C:A. Johnston. 1998. The Potential
.
Importance of Boundaries of Fluvial Ecosystems. Journal of the North Amencan

Benthological Society, 7(4): 289-306.
Naiman, R.J., R.E. Bilby, D.E. Schindler & J.M. Helfield. 2.002 .. Pacific Salmon,
Nutrients, and the Dynamics of Freshwater and Rtpanan Ecosystems.

Ecosystems, 5: 399-417.
Nakano, S. & M. Murakami. 2001. Reciprocal Subsidies: Dy~amic Interdependence
between Terrestrial and Aquatic Food Webs, ~roceedmgs of the Natrona[

Academy of Sciences of the United States of Amenca, 98(1): 166-170.

Biblio lrtlph y

181

Nehlsen, W., ].E. Williams, & ].A. Lichatowich. 1991. Pacific Salmon at the
Crossroads: Stocks at Risk from California, Oregon, Idaho, and Washington.
Fisheries, 16(2): 4-21.
Newbold, ] .D., J.W. Elwood, M.S. Schulz, R.W. Stark & ].C. Barmeier. 1983.
Continuous ammonium enrichment of a woodland stream: uptake kinetics,
leaf decomposition, and nitrification. Freshwater Biology, 13: 193-204.
Nowlin, W.H., M.J. Vanni & L.H. Yang. 2008. Comparing resource pulses in aquatic
and terrestrial ecosystems. Ecology, 89: 647-659.
bermeyer, K.E., K.S. White & M.F. Willson. 2006. Influence of Salmon on the
Nesting Ecology of American Dippers in Southeastern Alaska. Northwest
Science, 80: 26-33.
O'Brien, R.M. 2007. A aution Regarding Rules of Thumb for Variance Inflation
Factors. Quality & Quantity, 41: 673-690.
O'Keefe, T. & R. Edwards. 2002. Evidence for hyporheic transfer and removal of
marine-derived nutrients in a Sockeye stream in southwest Alaska. American
Fisheries Society Symposium, 33: 99-107.
Pace, N.R. 2001. The universal nature of biochemistry, Proceedings of the National
Academy of Sciences of the United States of America, 98(3): 805-808.
Packman, A.I, & K.E. Bencala. 2000. Modelling methods in the study of surfacesubsurface hydrologic interactions. In Streams and Ground Waters. Edited by
].B. Jones & P.J. Mullholland, Academic Press, pp. 45-80.
Payne, W.J. 1973. Reduction of Nitrogenous Oxides by Microorganisms.
Bacteriological Reviews, 37(4): 409-452.
Peckarsky, B. 1984. Sampling the stream benthos. In A manual of methods for the
assessment of secondary productivity in freshwater, 2nd edition. Edited by J.
Downing & F. Rigler. Oxford, Blackwell Scientific Publications, IBP
Handbook 19, pp. 131-160.
Peet, R.K. 1974. The Mea urement of Species Diversity. Annual Review of Ecology and
Systematics, 5: 285-307.
Peter, S., R. Rechsteiner, M.F. Lehmann, R. Brankatschk, T. Vogt, S. Diem, B. Wehrli,
K. Tockner & E. Durisch-Kaiser. 2012. Nitrate removal in a restored riparian
groundwater system: functioning and importance of individual riparian
zones. Biogeosciences, 9: 4295-4307.
Peterjohn, W.T. & D.L. Correll. 1984. Nutrient dynamics in an agricultural
watershed: observations on the role of a riparian forest. Ecology, 65: 1466-1475.

Bibliography

182

Peterman, R.M. & B. Dorner. 2011. Fraser River sockeye production dynamics.
Cohen Commission Technical Report 10: 133, Vancouver, B.C.
www .cohencommission.ca.
Peterman, R.M. & B. Dorner. 2012. A widespread decrease in productivity of
sockeye salmon populations in western North America. Canadian Journal of
Fisheries and Aquatic Sciences, 69: 1255-1260.
Peter on, D. & C. Foote. 2000. Disturbance of small-stream habitat by spawning
ockeye salmon in Alaska. Transactions of the American Fisheries Society, 129:
924-934.
Peterson, B.J., W.M. Wollheim, P.J. Mulholland, J.R. Webster, J.L. Meyer, J.L. Tank,
E. Marti, W.B. Bowden, H.M. Valett, A.E. Hershey, W.H. McDowell, W.K.
Dodds, S.K. Hamilton, S. Gregory & D.D. Morrall. 2001. Control of nitrogen
export from watersheds by headwater streams. Science, 292: 86-90.
Peyrard, D., S. Delmotte, S. Sauvage, Ph. Namour, M. Cerino, P. Vervier & J.M.
Sanchez-Perez. 2011. Longitudinal transformation of nitrogen and carbon in
the hyporheic zone of anN-rich stream: A combined modelling and field
study. Physics and Chemistry of the Earth, 36: 599-611.
Piccolo, J.J. & M.S. Wipfli. 2002. Does red alder (Alnus rubra) in upland riparian
forests elevate macroinvertebrate and detritus export from headwater streams
to downstream habitats in southeastern Alaska? Canadian Journal of Fisheries
and Aquatic Sciences, 59: 503-513.
Pinay, G., H. Decamps, E. Chauvet & E. Fustec. 1990. Functions of Ecotones in
Fluvial Systems. In Ecology and management of aquatic-terrestrial ecotones.
Edited by R.J. Naiman & H. Decamps. UNESCO, Paris, France, pp. 141-169.
Pinay, G., T.C. O'Keefe, R.T. Edwards & R.J. Naiman. 2009. Nitrate removal in the
hyporheic zone of a salmon river in Alaska. River Research and Applications, 25:
367-375.
Pingram, M.A., K.J. Collier, D.P. Hamilton, B.J. Hicks & B.O. David. 2014. Spatial
and tempo al patterns of carbon flow in a temperate large river food web.
Hydrobiologia, 729: 107-131.
Plummer, L.N., W.E. Stanford & P.D. Glynn. 2013. Characterization and
conceptualization of groundwater flow systems. In Isotope methods for
dating old groundwater. Edited by International Atomic Energy Agency,
Vienna, Austria, pp. 5-19.
Pollock, D.W. 1994. User's Guide for MODPATH/MODPATH-PLOT, Version 3.
pen File Report 94-464. United States Geological Survey.

Bibliography

183

Polis, G.A. & S.D. Hurd. 1996. Allochthonous input across habitats, subsidized
consumers and apparent trophic cascades: examples from the ocean-land
interface. In Food Webs: Integration of Patterns and Dynamics. Edited by G.A.
Polis & K.O. Winemiller, Chapman and Hall, New York, pp. 275-285.
Poole, G.C., S.J. O'Daniel, K.L. Jones, W.W. Woessner, E.S. Bernhardt, A.M. Helton,
J.A. Stanford, B.R. Boer & T.J. Beechie. 2008. Hydrologic spiralling: The role of
multiple interactive flow paths in stream ecosystetns, River Research and
Applications, DOl: 10.10021 rra.
Post, D.M., C.A. Layman, D.A. Arrington, G. Takimoto, J. Quattrochi & C.G.
Montana. 2007. Getting to the fat of the matter: models, methods and
assumptions for dealing with lipids in stable isotope analysis. Oecologia, 152:
179-189.
R Core Team. 2013. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. URL http: I I www.Rproject.orgl.
Reece, P.F. & J.S. Richardson. 1998. Seasonal changes of benthic macroinvertebrate
communities in southwestern British Columbia. Environment Canada, FRAP
Report Number: DOE-FRAP1998-33.
Reimchen, T.E. & C.H. Fox. 2013. Fine-scale spatiotemporal influences of salmon on
growth and nitrogen signatures of Sitka Spruce tree rings. BMC Ecology, 13: 113.
Reimchen, T.E., D.O. Mathewson, M.D. Hocking, J. Moran & D. Harris. 2003.
Isotopic Evidence for Enrichment of Salmon-Derived Nutrients in Vegetation,
Soil, and Insects in Riparian Zones in Coastal British Columbia. American
Fisheries Society Symposium, 34: 59-69.
Reisinger, A.J., D.T. Chaloner, J. Ruegg, S.D. Tiegs & G.A. Lamberti. 2013. Effects of
spawning Pacific salmon on the isotopic composition of biota differ among
southeast Alaska streams. Freshwater Biology, 58: 938-950.
Rex, J. F. & E.L. Petticrew. 2008. Delivery of marine-derived nutrients to streambeds
by Pacific salmon. Nature Geoscience, 1: 840-843.
Rezanezhad, F., R.-M. Couture, R. Kovac, D. O'Connell & P. Van Cappellen. 2014.
Water table fluctuations and soil biogeochemistry: An experimental approach
using an automated soil columns stem. Journal of Hydrology, 509: 245-256.
Richardson, J.S. 1992. Seasonal food limitation of detritivores in a montane stream:
an experimental test. Ecology, 72: 873-887.

Bibliography

184

Richardson, J.S., Y. Zhang & LB. Marczak. 2010. Resource subsidies across the landfreshwater interface and responses in recipient communities. River Research
and Applications, 26: 55-66.
Richey, J.S., W.H. McDowell & G.E. Likens. 1984. Nitrogen transformations in a
small mountain stream. Hydrobiologia, 124:129-139.
Richter, D.O., D. Markewitz, P.R. Heine, V. Jin, J. Raikes, K. Tian & C.G. Wells. 2000.
Legacies of agriculture and forest regrowth in the nitrogen of old-field soils.
Forest Ecology and Management, 138: 233-248.
Rinella, D.J., M.S. Wipfli, C.A. Stricker, R.A. Heintz & M.J. Rinella. 2012. Pacific
salmon (Oncorhynchus spp.) runs and consumer fitness: growth and energy
storage in stream-dwelling salmonids increase with salmon spawner density.
Canadian Journal of Fisheries and Aquatic Science, 69: 73-84.
Rinella, D.J., M.S. Wipfli, C.M. Walker, C.A. Stricker & R.A. Heintz. 2013. Seasonal
persistence of marine-derived nutrients in south-central Alaskan salmon
streams. Ecosphere, 4(10): 122.
Risser, P.G. 1993. Ecotones. Ecological Applications, 3: 367-368.
Risser, P.G. 1995. The Status of the Science Examining Ecotones. BioScience, 45(5):
318-325.
Robertson, A.L. & P.J. Wood. 2010. Ecology of the hyporheic zone: origins, current
knowledge and future directions. Archiv for Hydrobiologie, 176(4): 279-289.
Robinson, D. 2001. b15N as an integrator of the nitrogen cycle. Trends in Ecology &
Evolution, 16: 153-162.
Rounick, J.S. & M.J. Winterboum. 1986. Stable Carbon Isotopes and Carbon Flow in
Ecosystems. BioScience, 36(3): 171-177.
Rowland, A.P. 1983. An Automated Method for the Determination of Ammonium-N
in Ecological Materials. Communications in Soil Science and Plant Analysis,
14(1): 49-63.

Ruegg, J .~ S.D. Tiegs, D.T. Chaloner, P.S. Levi, J.L. Tank & G.A. Lamberti. 2011.
Salmon subsidies alleviate nutrient limitation of benthic biofilms in southeast
Alaska streams. Canadian Journal of Fisheries and Aquatic Science, 68: 277-287.
Saunders, T.J., M.E. McClain & C.A. Llerena. 2006. The biogeochemistry of dissolved
nitrogen, phosphorus, and organic carbon along terrestrial-aquatic flowpaths
of a montane headwater catchment in the Peruvian Amazon. Hydrological
Processes, 20: 2549-2562.

Biblio, rraphy

185

Scheuerell, M.D., P.S. Levin, R.W. Zabel, J.G. Williams & B.L. Sanderson. 2005. A
new perspective on the importance of marine-derived nutrients to threatened
stocks of Pacific salmon (Oncorhynchus spp.). Canadian Journal of Fisheries and
Aquatic Science, 62: 961-964.
Schielzeth, H. 2010. Simple means to improve the interpretability of regression
coefficients. Methods in Ecology and Evolution, 1: 103-113.
Schindler, D., M. Scheuerell, J. Moore, S. Gende, T. Francis & W. Palen. 2003. Pacific
salmon and the ecology of coastal ecosystems. Frontiers in Ecology and the
Environment, 1: 31-37.
Schoeneberger, P.J., D.A. Wysocki, E.C. Benham & W.O. Broderson. 2002. Field book
for describing and sampling soils. Version 2.0. Lincoln, USA, National Soil
Survey Center, Natural Resources Conservation Service, USDA.
Seitzinger, S., J.A. Harrison, J.K. Bohlke, A.F. Bouwman, R. Lowrance, B. Pterson, C.
Tobias & G. Van Drecht. 2006. Denitrification across landscapes and
waterscapes: A synthesis. Ecological Applications, 16(6): 2064-2090.
Shaftel, R.S., R.S. King & J.A. Back. 2012. Alder cover drives nitrogen availability in
Kenai lowland headwater streams, Alaska. Biogeochemistry, 107: 135-148.
Sharma, R., L.A. Velez-Espino, A.C. Wertheimer, N. Mantua & R.C. Francis. 2013.
Relating spatial and temporal scales of climate and ocean variability to
survival of Pacific Northwest Chinook salmon (Oncorhynchus tshawytscha).
Fisheries Oceanography, 22: 14-31.
Shibata, H. 0. Sugawara, H. Toyoshima, S.M. Wondzell, F. Nakamura, T. Kasahara,
F.J. Swanson & K. Sasa. 2004. Nitrogen Dynamics in the Hyporheic Zone of a
Forested Stream during a Small Stom, Hokkaido, Japan. Biogeochemistry,
69(1 ): 83-103.
Short R.A. & P.E. Maslin. 1977. Processing of leaf litter by a stream detritivore: effect
on nutrient availability to collectors. Ecology, 58: 935-938.
Silva, D.R.O., R. Ligeiro, R.M. Hughes & M. Callisto. 2014. Visually determined
stream mesohabitats influence benthic macroinvertebrate assessments in
headwater streams. Environmental Monitoring and Assessment, 186: 5479-5488.
Smith, I.M., K.J. Hall, L.M. Lavkulich & H. Schreier. 2007. Trace Metal
Concentrations in an Intensive Agricultural Watershed in British Columbia,
Canada. Journal of American Water Resources Association, 43{6): 1455-1467.
Sprecher, S.W. 2000. Installing monitoring wells/piezometers in wetlands, WRAP
Technical Notes Collection (ERDC TN-WRAP-00-02), U.S. Army Engineer
Research and Development Center, Vicksburg, MS.

Bibliography

186

Staal, M., R. Thar, M. Kuhl, M.C.M. Van Loosdrecht, G. Wolf, J.F.C. De Brouwer &
J.W. Rijstenbil. 2007. Different carbon isotope fractionation patterns during
the development of phototrophic freshwater and marine biofilms.
Biogeosciences, 4: 613-626.
Stanford, J.A. & J.V. Ward. 1988. The hyporheic habitat of river ecosystems. Nature,
335: 64-66.
Stevenson, F.J. & M.A. Cole. 1999. The internal cycle of nitrogen in soil, chapter 6. In
Cycles of soil: carbon, nitrogen, phosphorus, sulfur, micronutrients. Wiley,
New York.
Stewart-Wade, S.M., S. Neumann, L.L. Collins & G.J. Bolland. 2002. The biology of
Canadian weeds. 117. Taraxacum officinale G.H. Weber ex Wiggers. Canadian
Journal of Plant Science, 82: 825-853.
Strong, D.R. 1972. Life History Variation Among Populations of an Amphipod
(Hyalella Azteca). Ecology,53(6): 1101-1111.
Stokes, M.A. & T.L. Smiley. 1968. An introduction to tree-ring dating. University of
Arizona Press, Tucson, Arizona, USA.
Stonedahl, S.H., J.W. Harvey, A. Worman, M. Salehin & A.l. Packman. 2010. A
multiscale model for integrating hyporheic exchange from ripples to
meanders. Water Resources Research, 46: 1-14.
Tank, J.L. & W.K. Dodds. 2003. Nutrient limitation of epilithic and epixylic biofilms
in 10 North American streams. Freshwater Biology, 48: 1031-1049.
Tauxe, J.D. 1994. Porous Medium Advection-Dispersion Modeling in a Geographic
Information System. Ph.D. dissertation, University of Texas, Austin.
Thompson, A.M. 2007. Amphipods are a strong interactor in the foodweb of a
brown-water salmon river. Masters Thesis, University of Montana.
Thomson, R.E. & R.A.S. Hourston. 2011. A matter of timing: the role of ocean
conditions in the initiation of spawning migration by late-run Fraser River
sockeye salmon (Oncorhynchus nerka). Fisheries Oceanography, 20(1): 47-65.
Thorp, J.H. & A.P. Covich. 1991. Ecology and Classification of North American
Freshwater Invertebrates. Academic Press, Toronto.
Tiegs, S.D., P.S. Levi, J. Ruegg, D.T. Chaloner, J.L. Tank & G.A. Lamberti. 2011.
Ecological Effects of Live Salmon Exceed Those of Carcasses During an
Annual Spawning Migration. Ecosystems, 14: 598-614.

Bibliography

187

Tonina, D. & J.M. Buffington. 2009. Hyporheic exchange in mountain rivers I:
Mechanics and Environmental Effects. Geography Compass, 3/3: 1063-1086.
Triska, F., V.C. Kennedy, R.J. Avanzino, G.W. Zellweger & K.E. Bencala. 1989.
Retention and transport of nutrients in a third-order stream in northwest
California: Hyporheic processes. Ecology, 70: 1893-1905.
Triska, F.J., J.H. Duff & R.J. Avanztno. 1993. The role of water exchange between a
stream channel and its hyporheic zone in nitrogen cycling at the terrestrialaquatic interface. Hydrobiologia, 251: 167-184.
Turner Designs. 1995. Preparation of Standards for Dye Studies Using Rhodamine
WT. Techmcal Notes - Fluorescence and Fluorometry, Turner Designs, California.
U.S. Dept. of Agriculture, Soil Survey Division Staff. 1993. Soil Survey Manual, U.S.
Department of Agriculture, Washington, D.C., Handbook No. 18.
Valett, H.M., S.G. Fisher & E.H. Stanley. 1990. Physical and chemical characteristics
of the hyporheic zone of a Sonoran Desert stream. Journal of the North
American Benthological Society, 9(3): 201-215.
Valett, H.M., S.G. Fisher, N.B. Grimm & P. Cahill. 1994. Upwelling, downwelling,
and algal recovery from floods in a desert stream. Ecology, 75: 548-560.
Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell & C.E. Cushing. 1980. The
River Continuum Concept. Canadian Journal of Fisheries and Aquatic Science, 37:
130-137.
Vavrek, M.C., J.B. McGraw & H.S. Yang. 1997. Within-Population Variation in
Demography of Taraxacum Officinale: Season- and Size-Dependent Survival,
Growth and Reproduction. Journal of Ecology, 85(3): 277-287.
Veblen, T.T., K.S. Hadley, M.S. Reid, & A.J. Rebertus. 1991. Methods of detecting
past spruce beetle outbreaks in Rocky Mountain subalpine forests. Canadian
Journal of Forest Research, 21: 242-254.
Verspoor, J.J. D.C. Braun, M.M. Stubbs & J.D. Reynolds. 2011. Persistent ecological
effects of a salmon-derived nutrient pulse on stream invertebrate
communities. Ecosphere, 2(2): art18. doi:10.1890/E510-00011.1.
Vervier P., J. Gibert, P. Marmonier & M.J. Dole-Olivier. 1992 A perspective on the
permeability of the surface freshwater-groundwater ecotone. Journal of the
North American Benthological Society, 11: 93-102.
Vitousek, P.M., J.D. Aber, R.W. Howarth, G.E. Likens, P.A. M~tson, D.W. Schlinder,
W.H. Schlesinger, & G.D. Tilman. 1997. Human ~lteratto~ of.the global
nitrogen cycle: sources and consequences. Ecologzcal Applzcatwns, 54: 119-140.

Bibliog1'aphy

188

Wagner, B.J. & J.W. Harvey. 1997. Experimental design for estimating parameters of
rate-limited mass transfer: Analysis of stream tracer studies. Water Resources
Research, 33(7): 1731-1741.
Wagner, F.H. & G. Bretschko. 2002. Interstitial flow through preferential flow paths
in the hyporheic zone of the Oberer Seebach, Austria. Aquatic Sciences, 64: 307316.
Wallace, J.B., M.R. Whiles, S. Eggert, T.F. Cuffney, G.H. Lugthart & K. Chung. 1995.
Long-Term Dynamics of Course Particulate Organic Matter in Three
Appalachian Mountain Streams. Journal of the North American Benthological
Society, 14: 217-232.
Wang, G., L. Zhou, M. Liu, J. Han, J. Guo, A. Faiia & F. Su. 2010. Altitudinal trends
of leaf b13C follow different patterns across a mountainous terrain in north
China characterized by a temperate semi-humid climate. Rapid
Communications in Mass Spectrometry, 24: 1557-1564.
Ward, J.V. & ].A. Stanford. 1995. The serial discontinuity concept: extending the
model to floodplain rivers. Regulated Rivers: Research and Management, 10: 159168.
Ward, J.V., G. Bretschko, M. Brunke, D. Danielopol, J. Gibert, T. Gonser & A.G.
Hildrew. 1998. The boundaries of river systems: the metazoan perspective.
Freshwater Biology, 40: 531-569.
Weaver, J.L. & R.M. Zammuto. 2004. Grizzly Bears and Chinook Salmon in the
Inland Rainforest, Wildlife Conservation Society Report, 14 pp.
Webster, J.R., P.J. Mulholland, J.L. Tank, H.M. Valett, W.K. Dodds, B.J. Peterson,
W.B. Bowden, C.N. Dahm, S. Findlay, S.V. Gregory, N.B. Grimm, S.K.
Hamilton, S.L. Johnson, E. Marti, W.H. McDowell, J.L. Meyer, D.O. Morrall,
S.A. Thomas & W.M. Wollheim. 2003. Factors affecting ammonium uptake in
streams- an inter-biome perspective. Freshwater Biology, 48: 1329-1352.
Welch, D.W., & T.R. Parsons. 1993. b13C- b15N values as indicators of trophic
position and competitive overlap for Pacific salmon (Oncorhynchus spp.).
Fisheries oceanography, 2: 11-23.
Williams, D.O., C.M. Febria, & J.C.Y Wong. 2010. Ecotonal and other properties of
the hyporheic zone. Fundamental and Applied Limnology, 147: 349-364.
Willson, M.F., S.M. Gende & B.H. Marston. 1998. Fishes and the Forest. Bioscience,
48(6): 455-462.
Willson, M.F., S.M. Gende & P.A. Bisson. 2004. Anadromous fishes as ecological

Bibliography

189

links between ocean, freshwater and land. In Food Webs and the Landscape
Scale. Edited by G.A. Polis, M.E. Power & G.R. Huxel. The University of
Chicago Press, Chicago, IL. pp. 284-300.
Winter, T.C. 2000. Interaction of Ground Water and Surface Water. In Proceedings of
the Ground-Water/Surface-Water Interactions Workshop, Environmental
Protection Agency, 15-20.
Wipfli, M., J. Hudson & J. Caouette. 1998. Influence of salmon carcasses on stream
productivity: Response of biofilm and benthic macroinvertebrates in
southeastern Alaska, USA. Canadian Journal of Fisheries and Aquatic Sciences,
55: 1503-1511.
Wipfli, M.S. J.P. Hudson, D.T. Chaloner & J.P. Caouette. 1999. Influence of salmon
spawner densities on stream productivity in Southeast Alaska. Canadian
Journal of Fisheries and Aquatic Sciences, 56: 1600-1611.
Wipfli, M.S., J.P. Hudson & J.P. Caouette. 2003. Marine Subsidies in Freshwater
Ecosystems: Salmon Carcasses Increase the Growth Rates of Stream-Resident
Salmonids. Transactions of the American Fisheries Society, 132: 371-381.
Wipfli, M.S. & C.V. Baxter. 2010. Linking Ecosystems, Food Webs, and Fish
Production: Subsidies in Salmonid Watersheds. Fisheries, 35(8): 373-387.
Wittebolle, L., M. Marzorati, L. Clement, A. Balloi, D. Daffonchio, K. Heylen, P. De
Vos, W. Verstraete & N. Boon. 2009. Initial community evenness favours
functionality under selective stress. Nature, 458: 623-626.
Wonzell, S.M. & F.J. Swanson. 1996a. Seasonal and Storm Dynamics of the
Hyporheic Zone of a 4th-Order Mountain Stream. I: Hydrologic processes.
Journal of the North American Benthological Society, 15: 3-19.
Wonzell, S.M. & F.J. Swanson. 1996b. Seasonal and Storm Dynamics of the
Hyporheic Zone of a 4th-Order Mountain Stream. II: Nitrogen cycling ..
Journal of the North American Benthological Society, 15: 20-34.
Wondzell, S.M. 2006. Effect of morphology and discharge on hyporheic exchange
flows in two small streams in the Cascade Mountains of Oregon, USA.

Hydrological Processes, 20: 267-287.
Wondzell, S.M. & M.N. Gooseff. 2013. Geomorphic Controls on Hyporheic Exchange
Across Scales: Watersheds to Particles. In Treatise on Geomorphology. Edited
by J. Shroder & E. Wohl, Academic Press, San Diego, CA, vol. 9, Fluvial
Geomorphology, pp. 203-218.
Woo, M.K. & J.M. Waddington. 1990. Effects of beaver dams on subarctic wetland
hydrology. Arctic, 43: 223-230.

Biblzography

190

Woodward, G.R. & A.G. Hildrew. 2002. Food web structure in riverine landscapes.
Freshwater Biology, 47: 777-798.
Woodward, G., R. Thompson, C.R. Townsend & A.G. Hildrew. 2005. Pattern and
process in food webs: evidence from running waters. In Aquatic Food Webs
An Ecosystem Approach, Edzted by A. Belgrano, U.M. Scharler, J.A. Dunne &
R.E. Ulanowicz. Oxford University Press, pp. 51-66.
Worm, B., E. B. Barbier, N. Beaumont, J.E. Duffy, C. Folke, B.S. Halpern, ].B.C.
Jackson, H.K. Lotze, F. Micheli, S.R. Palumbi, E. Sala, K.A. Selkoe, J.J.
Stachowicz & R. Watson. 2006. Impacts of biodiversity loss on ocean
ecosystem services. Sczence, 314: 787-790.
Zarnetske, J.P., R. Haggerty, S.M. Wondzell & M.A. Baker. 2011. Dynamics of nitrate
production and removal as a function of residence time in the hyporheic
zone. Journal ofGeophyszcal Research, 116: 1-12.
Zhang, Z. 2007. Effect of Hydrological Regimes on Groundwater Phosphorus
Transfer in a Riparian Wetland. Masters Thesis, University of Waterloo,
Canada.