 

“£339
" 1

LIBRARY
2W Michigan State
University

E’—

 

 

 

 

This is to certify that the
thesis entitled

PACIFIC SALMON: A CONDUIT OF MARINE NUTRIENTS
AND ORGANIC MATTER TO STREAM ECOSYSTEMS

presented by

Brittany Syra Graham

has been accepted towards fulfillment
of the requirements for the

MS. degree in Environmental Geosciences

 

 

/ 24/1 L; l ; 505/} L.“

 

H/Major Professor’s Signature
ﬂzE/Oé

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/Dale0ue.indd-p41

PACIFIC SALMON: A CONDUIT OF MARINE NUTRIENTS AND ORGANIC
MATTER TO STREAM ECOSYSTEMS.

Brittany Syra Graham

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Geological Sciences

2006

ABSTRACT

PACIFIC SALMON: A CONDUIT OF MARINE NUTRIENTS AND ORGANIC

MATTER TO STREAM ECOSYSTEMS.

Brittany Syra Graham

The inﬂuence of marine-derived nutrients (MDN), delivered by spawning salmon, on
Alaskan streams was determined by measuring the carbon and nitrogen isotopic
composition of stream biota (ﬁshes, macroinvertebrates, bioﬁlm). Our main study site,
Fish Creek (PC), was divided into two reaches, separated by a salmon barrier, and
sampled before, during, and after spawning for two years. In Salmon Creek, we assessed
the upstream-downstream study design by comparing two sites also separated by a
barrier, but with no MDN inﬂuence. Isotopic differences in taxa from these two sites
were negligible. All lower FC consumers demonstrated MDN incorporation during and
after spawning and a “legacy” effect from previous subsidies. Moreover, primary
consumers exhibited a rapid MDN response and an increased MDN incorporation
downstream. Macroinvertebrate isotope values are an ideal tool for tracking MDN
dynamics, but spatial and temporal variability in MDN incorporation should be

considered before isotopes are utilized in ﬁsheries management.

My thesis is dedicated to my parents who have made everything possible. You have always

supported me to wander, stumble, and dream. I am forever grateful.

iii

ACKNOWLEDGEMENTS

First of all, the author would like to thank all of the P15 of the MDN research project,
including Dr. Rich Merritt (the kindest and most generous professor I have ever met), Dr.
Gary Lamberti, and Dr. Mark Wipﬂi. All of you helped develop, guide, and create a project
that was an incredible experience and honor to work on. The author would also like to thank
the incoming investigators on the MDN project, including Dr. Jen Tank and Dr. Rick
Edwards. The author also thanks Dr. Dom Chaloner who was the Elmer’s glue of the project
and a continual source of support. Dr. Steve Hamilton provided great insight and knowledge
throughout my project and at my committee meetings. Using stable isotope analysis in
ecosystem ecology involves an extensive ﬁeld collection effort and the author would like to
thank all of the individuals that helped in the ﬁeld, including Nicole Mitchell, Dom Chaloner,
Craig Herbold, John Hudson, Jake Musselwight, Osvaldo Hernandez, JoAnna Lessard,
Darren Throughbourgh, Chelsea Crenshaw, and Matt Herold. The author would like to thank
all those who helped in sample preparation, lab work and discussions involving methods,
including Dr. Hasand Gandhi, Dr. Nathaniel Ostrom, Dr. Rick Doucett, and Dr. Robin Sutka.
The long days and nights in the isotope geochemistry lab were made easier by the lovely
company of Adam Pitt, Amanda Field, and Mary Russ. Ironically, the author would like to
thank Dr. Lopushinsky who will always be a reminder of the worst aspects of academia.
Thankﬁrlly, the long months at MSU were made short with the arrival of Dr. Christina
Nielsen-Marsh. A thank you goes to the Acme Jam band that brought a group of us together
nearly every week and created many unique memories and friendships. The author would like
to sincerely thank others who helped me throughout my graduate program with many smiles

and laughs, including Osvaldo Hernandez, Jenn Wade, Dave Szymanski, Jason Price, and

iv

Nick Waterson. The author would like to sincerely thank Rina Dechert and my family; even
from a distance you have always been a source of my happiness. Finally, the author would
like to especially thank Dr. Peggy Ostrom who provided me with many excellent research
opportunities, support to pursue my research tangents, and intellectual challenges, which in

combination has made me a better student and scientist.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... .vii
LIST OF FIGURES ............................................................................ viii
INTRODUCTION ............................................................................. 1
METHODS
Study Site ................................................................................. 4
Sample Collection ...................................................................... 5
Stable Isotope Analysis .............................................................. 9
MDN Quantiﬁcation .................................................................. 9
Statistical Analysis ................................................................... 11
RESULTS
Characterizing the MDN Endmember and Barrier Effect ..................... 12
Food Web Components
Bioﬁlm ........................................................................ 13
Macroinvertebrates .......................................................... 14
Freshwater Fishes ........................................................... 17
Longitudinal and Inter-Annual Trends ............................................. 19
DISCUSSION
The MDN Endmember and Barrier Effect ....................................... 21
Food Web Components
Bioﬁlm ......................................................................... 23
Macroinvertebrates ........................................................... 25
Legacy Effect ........................................................ 29
Freshwater Fishes ............................................................ 32
Longitudinal Trends .................................................................. 36
Inter-Annual Data and Salmon Escapements ...................................... 38
CONCLUSIONS/IMPLICATIONS ........................................................ 4O
LITERATURE CITED ....................................................................... 60

vi

LIST OF TABLES

Table 1. 513 C and 515N values for different Fish Creek food web components with
standard errors and sample numbers in parentheses, in upper Fish Creek (FCU) and lower
Fish Creek (FCL) reaches. 1' Drunella values represent samples collected only from pre-

spawning sampling periods ....................................................................... 42

Table 2. % MDN incorporation values. Ranges represent the estimates calculated using
two estimates of trophic fractionation (F). The trophic fraction factors were 2.7 and 3.4
%o and 0.7 and 1.0 %o for N and C, respectively. Error, as calculated by the Phillips and
Gregg (2002) model, was always less than 0.5 %, therefore, the error terms were not
included in the table. The values in parenthesis after the FFG groups represent the

assigned trophic levels included in the % MDN incorporation estimates ................ 45

Table 3. 813 C and 515 N values from above and below a waterfall barrier in Salmon Creek.

The isotope values are reported with standard errors and sample numbers in parentheses.

All data represented were collect in August 2001 ............................................. 46

vii

LIST OF FIGURES

Figure 1. Location of study site. Samples were collected in 2000 and 2001 from a 3rd

order stream, Fish Creek in Southeast Alaska ................................................. 47

Figure 2. Study design. Fish Creek was divided by an impassible waterfall into two study
reaches. The upper (reference) reach never received any spawning salmon. Salmon can
spawn in the lower (treatment) reach during the spawning season. Five sub-reaches were
sampled in both the upper and lower reaches during each sampling period. Stream
consumers and organic matter sources were sampled pre-spawning and post-spawning in

2000 and pre-, during, and post-spawning in 2001 ............................................ 48

Figure 3. Macroinvertebrate 513 C and 515 N values collected in pre-, and post-spawning
periods in the lower reach and pre-spawning for upper reach in both 2000 and 2001.
Circles represent the different sampling periods, where all of the data points delimited
represent the macroinvertebrates sampled during that period. The 513 C and 815N of adult

salmon is represented by the MDN symbol .................................................... 49

Figure 4. Fish Creek macroinvertebrate trophic structure from the upper and lower stream
reaches. Macroinvertebrate 513 C and SUN collected in (a) lower Fish Creek 2000, (b)

lower Fish Creek 2001, (c) upper Fish Creek 2000, (d) upper Fish Creek 2001. The

dashed circles encircle the 513C and SUN of primary conSumers, including Cinygmula,

viii

Epeorus, Drunella, and Baetidae. Diamond symbols represent pre-spawning and square

symbols represent post-spawning samples with the associated standard error ............ 50

Figure 5. The legacy effect of MDN in Fish Creek of (a) SN and (b) 513 C for
macroinvertebrates collected from the lower and upper reach during pre-spawning in
2000 and 2001. Two estimates of trophic fractionation were used in our models and lead

to the ranges in the "/0 MDN incorporations (Equation 2) .................................... 51

Figure 6. Dolly Varden 6‘3 C, SUN, and fork lengths collected during pre- and post-

spawning from upper and lower Fish Creek 2000 ............................................. 53

Figure 7. 5% and 5'5N YOY Coho values collected from Lower Fish during 2001. The
data points before 7/28/01 represent pre-spawning, around 8/17/01 represent during

spawning, and after 9/26/01 represent post-spawning ........................................ 54

Figure 8. 813 C and 515 N values from Cinygmula, Epeorus, Drunella, and Chloroperlidae
collected along a longitudinal gradient in lower Fish Creek. A larger distance (km) from
the barrier represents a sub-reach further downstream. Samples were collected from the

sub-reaches in 2000 and 2001 .................................................................... 55
Figure 9. Longitudinal variation in 5'5N values for C inygmula and Epeorus sampled
during spawning in 2001. % MDN incorporation were estimated from the different sub-

reaches (Equation 2) ............................................................................... 56

ix

Figure 10. Inter-annual Fish Creek dataset. 513 C and SUN values for two scraper taxa,
C inygmula and Epeorus, collected from all sampling periods: Pre-spawning in 2000 (Pre
00), post-spawning in 2000 (Post 00), pre-spawning 2001 (Pre 01), during spawning in
2001 (During 01), and post-spawning in 2001 (Post 01). The error bars represent the

standard error ....................................................................................... 57

Figure 11. Temporal trends in MDN. SUN values for Cinygmula, Epeorus, and
Chloroperlidae collected in 2001 from lower Fish Creek. The NIH maximum error
represents the MDN-NIL;+ released by adult salmon excreting with their ﬁrst arrival to
Fish Creek. The % MDN incorporation estimates were calculated using Equation 2. The

SISN value for adult salmon is represented by the MDN symbol ........................... 58

Figure 12. Hypothesized juvenile coho diet. The nitrogen isotopic composition for
different diets, indicated by dash lines, were determined by (1) using stomach contents of
post-spawning YOY coho from lower Fish Creek and associated isotope values in a mass
balance model (MDN-inﬂuenced stream diet) and (2) the carbon and nitrogen isotopic
composition of adult salmon (MDN-derived diet). Months from hatching was determined
by estimating the time between sample collection and hatching (ca. March). MDN pulses

represent the arrival of spawning salmon in the fall of 2000 and 2001 in Fish Creek... 59

INTRODUCTION

The annual return of anadromous salmon (Oncorhynchus spp.) has been a
conspicuous feature of streams in the Paciﬁc Northwest for thousands of years. Salmon
deposit marine-derived nutrients into streams by excretion, gamete release, and
subsequent decay of their own carcasses (Krokhin 1967; Brickell and Goering 1970;
Mathisen et al. 1988). Therefore, salmon symbolize a biological link between marine,
terrestrial and freshwater ecosystems and are a conduit of organic matter and nutrients
transfer across ecosystem boundaries. A new paradigm emerges that delineates material
ﬂow from marine to freshwater ecosystems, which differs from the concept of material
ﬂow from the continents to the oceans being unidirectional (Milliman and Meade 1983).

Although assimilation of MDN by autotrophs occurs by remineralization and
uptake, heterotrophic bacteria can obtain MDN via these mechanisms and direct uptake.
Both of these groups contribute to bioﬁlm Marine-derived nutrients can be available to
macroinvertebrates and ﬁsh through trophic transfers initiating with bioﬁlm and/or direct
consumption of carcasses or carcass fragments within course particulate organic matter
(CPOM) pool (Bilby et al. 1996).

Evidence for the incorporation of MDN into freshwater and terrestrial food webs
is derived from a growing base of literature that concentrates on stable isotope analysis
(Kline et al. 1990; Kline et al. 1993; Bilby et al. 1996; Johnston et al. 1997; Garman and
Macko 1998; Ben-David et al 1997, 1998; Szepanski et al 1999; Hilderbrand et a1 1999,
Minakawa and Gara 1999). The basis for these studies is the observation that the isotOpic

composition of an organism reﬂects its food source. Anadromous salmon have a distinct

marine isotopic signature, typically greater than -20.0 %o for 813C and 11.0 %o for SUN,
which differs from both freshwater and terrestrial inputs that form the base of lotic food
webs (Kline et al. 1990; Bilby et a1. 1996; Doucett et a1 1999; Johnston et al. 1997). As
organic matter cycles through the biota there is a trophic ﬁactionation or preferential
incorporation of the heavy isotope during assimilation at each successive trophic level
(Minagawa and Wada 1984; Peterson and Fry 1987). In general, trophic-related
fractionation is about 3-4%o and 0-1%o for nitrogen and carbon, respectively (Deniro and
Epstein 1978; DeNiro and Epstein 1981; Minagawa and Wada 1984; France 1994).
Rapidly growing organisms, including insects, can quickly reﬂect the isotopic
compositions of a new diet (Fry and Arnold 1982, Ostrom et a1. 1997). Therefore, if
stream organisms are incorporating MDN, and have high tissue turnover, should
demonstrate a quick change in their isotope composition to reﬂect the new MDN diet.

Carbon and nitrogen isotope values have been used in c6ncert to deﬁne the (1)
estimates of the relative importance of organic matter sources to a consumer’s diet at the
base of the food web using mass balance equations and (2) trophic structure
(McConnaughey and McRoy 1979; Harrigan et al. 1989; Johnston et a1. 1987; MacAvoy
et al. 2000; see review Phillips and Gregg 2001). Consequently, multiple isotope analysis
is a powerful tool to examine marine-derived nutrient ﬂow and incorporation in stream
food webs.

Marine-derived nutrient subsidies have important implications for ﬁsheries and
ecosystem management. Microbial and algal growth are augmented with increased
delivery of marine nutrients and organic carbon to streams (Richey et a1. 1975; Wipﬂi et

a1. 1998; Fisher Wold and Hershey 1999). This in turn can increase prey abundance,

primarily invertebrates, which are an important component in juvenile ﬁsh diets (e. g.
juvenile coho salmon). The additional secondary production in stream ecosystems should
increase the freshwater survivorship of salmon, ultimately leading to elevated salmon
production (Groot and Margolis 1991; Michael 1995; Johnston et al. 1997; Wipﬂi et al.
1999). Therefore, it has been proposed that salmon runs create marine nutrient subsidies
that set forth a natural positive feedback system of nutrient deposition that is critical to
the survival of salmon and important in sustaining the Paciﬁc Northwest’s oligotrophic
freshwater food webs (Mathisen et al. 1988; Cederholm et al.1999). Recently Finney et
al. (2001) and Bilby et al. (2001) proposed that a relationship between the abundance of
spawning salmon and the stable isotope-based estimates of incorporation of MDN in
stream and riparian biota can be used to determine an escapement level of salmon needed
to sustain salmon runs.

The importance and urgency of studying MDN subsidies has recently been
emphasized by managerial actions implemented in the Paciﬁc Northwest. With limited
quantitative guidelines, resource managers have begun to place spawned-out salmon
carcasses into streams to presumably elevate stream productivity and ultimately salmon
production (Levy 1997; Barnard 2000). In addition, slow-release briquettes (i.e.
compressed, processed, salmon byproducts) and slow-release artiﬁcial fertilizers are
being considered as potential salmon stock restoration tools (Ashley and Slaney 1997;
Sterling et al. 2000). However, before we begin to make large-scale management
decisions, the relationship between salmon subsidies and ecosystem responses need to be

better understood. In order to understand the effects of MDN on stream ecosystems we

need to understand the incorporation of MDN into the food web in both a temporal and
spatial context.

The aim of my study was to evaluate the relative importance of salmon on trophic
dynamics and energy ﬂow in Southeast Alaskan streams using stable isotope analysis
(SIA). I compared the carbon and nitrogen isotope values of stream consumers in an
upstream (reference) to downstream (treatment) reach in Fish Creek, AK during pre- and
post-spawning periods of 2000 and 2001 and during spawning in 2001. Speciﬁcally, my
objectives were four-fold: (1) characterize the marine-derived endmember, (2) evaluate
the treatment/reference study design, (3) quantify MDN incorporation in major trophic
components of stream ecosystems, (4) evaluate longitudinal variation in the isotope
values of macroinvertebrates and (5) examine inter-annual variability in MDN
incorporation. The ultimate goal of this study was to assess the relative importance of the

MDN subsidy to the Fish Creek food web in both a spatial and temporal context.

METHODS

Study Site

Stable isotope samples were collected from two watersheds, Fish Creek and Salmon
Creek in southeast Alaska during 2000 and 2001 (Fig. 1). Fish Creek is located on
Douglas Island and is the primary focus of this paper. Fish Creek has an impassible
waterfall, which prevents the upstream migration of salmon (i.e. salmon barrier). Thus,

the stream was separated into an upper reach (reference) and a lower reach (treatment)

(Fig. 2). We termed the portion of the stream above and below the barrier as the upper
and lower reach, respectively. Fish Creek experiences annual runs of Pink
(Oncorhynchus gorbuscha), Chum (0. keta), and Chinook (0. tshawytscha) salmon. Pink
and Churn salmon are the predominant species in terms of both biomass and abundance.
Peak spawning occurs from late-July to August, depending upon species and stream
discharge.

Salmon Creek provided an opportunity to evaluate the MDN study design. Both Fish
and Salmon Creek are 3rd order streams dominated by snow melt and groundwater inputs,
have watersheds of comparable sizes, 36 and 26 km’, respectively, have similar gradients,
and have a large amount of woody debris in the stream channels. Longitudinally, Salmon
Creek has several waterfall barriers, samples were collected from above and below the
second waterfall. These samples enabled us to examine the effects of a barrier without the
inﬂuence of salmon in a stream that experiences spawning salmon. These returns occur
below the ﬁrst barrier where the stream is heavily inﬂuenced by the Douglas Island Pink

and Chum (DIPAC) hatchery.

Sample Collection

Fish Creek was sampled intensively on four dates, corresponding to pre-salmon and
post-salmon spawning periods during 2000 and 2001 above and below the salmon
barrier. Additional sampling periods occurred in Fish Creek prior to spawning in July
2000 and during the salmon run in August 2001. The upper and lower reaches of Fish

Creek were further sub-divided into 5 reaches. Therefore, there were 10 sub—reaches

sampled on each sampling period. The distance between the lower sub-reaches was
approximately 2 km and the upper sub-reaches were approximately] km apart. The
distance between the sub-reach furthest downstream in the upper reach and the sub-reach
furthest upstream in the lower reach were approximately 2km apart. The sub-reach
sampled furthest downstream was approximately 1 km from the estuary and was not
affected by high tide periods.

Salmon Creek was sampled on two dates, corresponding to pre-spawning and post-
spawning periods in 2001. Three sub-reaches were sampled above and below the second
barrier in Salmon Creek and the upper and lower study reaches were approximately 3km
apart.

In order to analyze the marine-derived isotope signature, adult female Pink and Chum
salmon, were sampled using seine nets as they ﬁrst entered Fish Creek. Hatchery and
‘wild’ salmon spawn in the streams. Therefore, in addition to assessing isotopic
differences between species, we collected adult Pink and Chum female salmon from a
local salmon hatchery (DIPAC) for analysis. We also wanted to assess the changes in
carbon and nitrogen isotope ratios in returning salmon as they migrate upstream (i.e.
residence time in freshwater). Thus, we collected “fresh” adult female salmon from the
estuary, spawned out ﬁsh, recently deceased ﬁsh, and decomposing carcasses from Fish
Creek.

After collection, salmon were stored on ice, brought to the lab at the Paciﬁc
Northwest Research Station and frozen prior further processing for SIA. Kline et al. 1993
and Masterson 2000 veriﬁed that transverse sections of adult salmon are representative of

the whole organism, therefore, two cross-sections were sampled from adult salmon at the

mid-section and head region. The cross-sections were dried at 60°C, homogenized (Wig-
L-Bug® ball and capsule amalgamator, Crescent Industries), lipid-extracted, and re-
homogenized to a ﬁne powder. Lipids were removed by soxhlet extraction using an
azeotropic mixture (87:13) of chloroform and methanol for 6 hours (Gould et al. 1997).
Lipid extraction was preformed because lipids are depleted in 13 C relative to the bulk
513C value. Fish can have variable amounts of lipids and the 513 C lipid can overprint the
carbon isotope ratio of the diet (e. g., Kennicutt 11 et al. 1992; Focken and Becker 1998;
Doucett et al. 1999).

Minnow traps, baited with a perforated bag of salmon roe, were used to collect
resident freshwater ﬁshes. In 2000, Dolly Varden (Salvelinus malma) were sampled from
above and below the barrier, but were only captured in the upper reach of Fish Creek in
2001. In addition, benthic-feeding coastrange sculpin (Cottus aleuticus) and drift-feeding
YOY coho were sampled in lower Fish Creek. Neither of these ﬁsh exists above the
salmon barrier. YOY coho were sampled pre- and post-spawning in 2000 and at a
sampling-interval of approximately every 2 weeks during June to August 2001.
Coastrange sculpin were sampled only in 2001 from the lower reach of Fish Creek.
Stomach contents of all resident ﬁsh were removed immediately by gastric lavage. Fish
were stored on ice until processed. If the resident ﬁsh were approximately greater than
10cm, cross-sections were sampled and if less than 100m the whole ﬁsh was processed.
Fork length and ﬁsh species were recorded and all ﬁsh samples were freeze dried,

homogenized, lipid extracted, and re-homogenized into a ﬁne powder in preparation for

SIA.

Aquatic insects were collected using D-frame kick nets. Insects were separated
from organic matter in the ﬁeld and stored on ice until sorted to taxa and associated
functional feeding group in the lab. To ensure sufﬁcient size for isotopic analysis, most of
our samples represent composites of 10-200 individuals of each taxa from a single sub-
reach. In cases of certain rare macroinvertebrate taxa, samples for individual sub-reaches
were combined. Insects were identiﬁed to the lowest taxonomic group (primarily to
family) and associated functional feeding group (Merritt and Cummins 1992). Individual
taxa were placed in scintillation vials, and dried at 60°C. The collected taxa represented
four macroinvertebrate functional feeding groups, 1) Scrapers or the primary consumers,
2) collector-gathers, 3) shredders, and 4) predators. All macroinvertebrate samples were
lipid-extracted, acidiﬁed, and homogenized. Insect samples were acidiﬁed (1 N HCl) to
remove any 13 C enriched-carbonates that may have accumulated in stomach tracts and
then dried at 60°C.

Bioﬁlm (i.e. epilithon) samples were collected from 10-20 randomly selected
large cobbles combined from each sub-reach in Fish Creek in 2000-01 (Kline et al. 1990).
Each cobble was removed from a rifﬂe and placed in a pan, scraped on the top surface
using a nylon brush, rinsed with stream water, collected in a Whirl-pak®, and transported
back to the laboratory on ice. In the laboratory, samples were centrifuged, decanted, and
then the concentrated bioﬁlm sample was placed in a scintillation vial and dried at 60°C.
Bioﬁlm samples were acidiﬁed (1 N HCl) and ground into a ﬁne powder using a mortar
and pestle. Course particulate organic matter (CPOM) samples were collected in 2000
and 2001 in areas of retention (i.e. behind woody debris). In ﬁnal preparation for SIA,

CPOM samples were homogenized to a ﬁne powder using a coffee grinder.

Stable Isotope Analysis

Macroinvertebrate and ﬁsh samples were analyzed at the Isotope
Biogeochemistry Laboratory of Michigan State University using either a Carlo Erba NA
1500 nitrogen/carbon analyzer interfaced to a PRISM (Micromass, Manchester, England)
or a Eurovector elemental analyzer interfaced to a ISOPRIME (Micromass) stable isotope
mass spectrometer. Bioﬁlm, CPOM, and FPOM samples were processed for SIA by a
modiﬁed Dumas combustion where the samples are combusted on a slow temperature
program (1 h at 850°C and cooled from 700 to 500 0.6° C min'l) (Macko et al. 1987). The
evolved CO2 and N2 gas produced from combustion were puriﬁed cryogenically and the
isotope ratios of the resulting gas were measured on a PRISM mass spectrometer
(Micromass).

Isotope values are reported relative to an international standard and expressed as:

5‘30 or 5‘5N = [(Rsmple/Rsmndmyu x 1000

where R is the ratio of the heavy light isotope. Standards are V-PDB and N2 for carbon

and nitrogen, respectively. Reproducibility of the measurements is 0.2 %o.

MDN Quantiﬁcation

To determine the inﬂuence of salmon on the stream ecosystem we quantify the
inﬂuence of MDN on stream biota using mass balance equations. The following general
equation was used to derive the equation used to quantify the degree of MDN

incorporation,

5mm = fr51 + f252 (1)

1=f|+f2

where 8m is the Sconsume, including the discrimination between the consumer and its diet,
f1 and f2 are the fraction of food source 1 and 2, and 51 and 82 are the isotopic signature of
food source 1 & 2. As a result of our study design, we were able to compare an organism
in the reference reach (i.e. upper reach) to the same organism in the treatment reach (i.e.
lower reach) and we assumed the diets of the organisms between the two reaches were
similar, except for the contribution of MDN to the lower reach food web (cf. Johnston et
al. 1997; Chaloner et al. 2002). As a result, the following equation was used to quantify

the degree of MDN incorporation:

% MDN incorporation = [(ESLowcr - Supper)/((8MDN + (TL x F)) — Supper” x 100 (2)

where SLOW is the mean SUN or 813 C value of the organism in the lower reach, Supper is
the 8'5 N or 513 C value of the organism in the upper reach, SMDN is the isotopic value for

an adult salmon, TL is the proposed trophic level of the organism, and F is the degree of

fractionation. Two estimates of trophic fractionation were used in our modeling efforts,

10

2.7 and 3.4 %o for nitrogen and 0.7 and 1 %o for carbon that are commonly reported in the
literature (e.g. DeNiro and Epstein 1978, 1980, Tieszen 1983, Hobson 1992). This
resulted in two estimates of % MDN incorporation. Consequently, we consider that the
amount of MDN incorporated by consumers to fall within the range bounded by these
estimates and report our estimate of % MDN as such. Macroinvertebrate scraper,
collector-gather, and shredder functional groups were assigned to trophic level one.
Freshwater ﬁshes and the macroinvertebrate predators assigned to trophic level two.
Mixing models are highly sensitive to the isotope values incorporated into the model.
Therefore, we evaluated uncertainty in our % MDN calculations, including standard
errors. We modiﬁed the approach Phillips and Gregg (2001) to include our

reference/treatment study design. The model is as follows,

02mm: = 1/(5MDN-50pper)2[ CZSLower + fZMDNCZSMDN + (1' fMDN)2025trpper] (3)

where 025MDN, GZSLowcr, and 0’25Uppe, represent variances of the mean isotopic values for

adult salmon, and organisms in the lower reach, and the same organism in the upper

reach, respectively and fMDN represent the % MDN incorporation calculated from

equation 2. SLOW and Supper have been corrected for the associated trophic fractionation

and trophic level.

Statistical Analysis

11

This study consisted of randomly sampling stream organisms for isotope analysis
under different conditions over time. Therefore, a repeated measures AN OVA with time
as the repeated measure and a random factor. SAS version 8.2 (2001) was used for the
repeated measures AN OVA analysis. In other cases, comparisons required a students t-
test or one way ANOVA (e. g. comparison among both wild and hatchery, pink and chum

salmon). These statistics were performed with STATISTICA (release 6.0).

RESULTS

Characterizing the MDN Endmember and Barrier Effect

The ﬁrst objective of our study was to characterize the carbon and nitrogen
isotope values of the marine endmember. To do so, we compared unspawned hatchery
(DIPAC) and ‘wild’ pink and chum salmon collected from spawning periods during 2000
and 2001 (Table 1). There was no difference between salmon species or between wild
and hatchery individuals (ANOVA, P < 0.01). Therefore, we used the average carbon and
nitrogen isotope values of all adult salmon do derive the MDN endmember: -19.9 :t 0.4
and 11.4 i 0.4 %o, respectively. The second objective was to assess whether the
freshwater residence time or spawning migration of adult salmon would affect their
carbon and nitrogen isotopic composition (Table 1). There was no difference in carbon or
nitrogen isotopic compositions between adult salmon collected from the saltwater and

dead decaying salmon collected from Fish Creek in 2001 (ANOVA, P < 0.01).

12

An important strategy to evaluate the inﬂuence of MDN on the stream biota was
to compare organisms in an upstream reach above an impassible waterfall to organisms
of the same taxa from the lower reach. We ﬁrst evaluated the effects of a barrier on the
carbon and nitrogen isotopes of stream consumers. In order to do so, organisms of the
same taxa were collected from two reaches in Salmon Creek that were not inﬂuenced by
salmon, but separated by a barrier during August 2001. The macroinvertebrates, Baetidae
and Epeorus, from above and below the barrier did not signiﬁcantly differ in SUN values
collected (Table 3, ANOVA, p >0.01). The isotope values of the single Chloroperlidae
spp. collected in the upper was very similar to one collected from the lower reach
(difference of 0.3%o or less). Statistically, the Cinygmula SUN values in the upper reach
are enriched in l5N relative to the lower reach of Salmon Creek (ANOVA, p >0.01).
However, given the average isotope values (0.5 +/- 0.3 and 1.1 +/- 0.1) and the
reproducibility of the measurement 0.2, it is unlikely that the difference in these values is
truly signiﬁcant. Carbon isotope ratios for all taxa sampled were not signiﬁcantly

different between the two reaches (Table 3, ANOVA, p<0.01).

Food Web Components

Bioﬁlm and CPOM

There were no signiﬁcant differences in bioﬁlm isotope values between the upper

and lower reaches during any sampling periods in 2000 (ANOVA, P > 0.01). Sampling

of CPOM was limited. Only two Samples were obtained from upper and lower F C in

13

2001 after spawning. The 613 C values of the samples collected in the lower reach differed
by only 0.3 %o from those from the upper reach. The 515 N values of the same CPOM

samples from the lower reach were approximately 4 %o enriched in 15N relative to the

CPOM samples collected from upper Fish Creek (Table 1).

Macroinvertebrates

With the exception of Paraleptophlebiids, relative to pre-spawning, 813 C and BISN
values of macroinvertebrates in the lower reach shifted towards the MDN signature after
spawning occurred in Fish Creek (Table 1, Fig. 3). The one exception, Paraleptophlebia,
exhibited greater carbon and nitrogen isotope values in pre- relative post-spawning
periods in 2001. The Paraleptophlebia dataset is limited. Because the dataset consists of
one sample for each period sampled, it prevents us from drawing conclusions from this
apparently contradictory trend (Table 1). Macroinvertebrates in the upper (reference)
reach did not exhibit higher carbon or nitrogen isotope values after spawning relative to
pre-spawning periods (Table 1, Fig 4c and 4d). The observation that the carbon and
nitrogen isotope values of consumers stayed constant in the upper reach while those of
the same consumers increased in the lower reach after spawning is consistent with MDN
incorporation. In pre-spawning periods, there are several cases where carbon and nitrogen
isotopic variability of the macroinvertebrate predator, Chloroperlidae spp., is high
relative to that of other macroinvertebrates. The carbon and nitrogen isotope values of
Chloroperlidae spp. also did not differ signiﬁcantly between the upper and lower reaches

during the pre-spawning period in 2001 (ANOVA, p>0.05) (Table 1). After spawning in

14

2001, Chloroperlidids differed signiﬁcantly for only nitrogen (AN OVA, p<0.01) and not
carbon isotope values (ANOVA, p>0.5). In 2001, these macroinvertebrate predators did
not signiﬁcantly differ in either carbon or nitrogen isotope ratios between pre-spawning
and post-spawning periods in lower Fish Creek (ANOVA, p>0.05) (Table 1).

In addition to the evidence reﬂecting MDN incorporation in primary consumers,
several other salient features appeared in our macroinvertebrate data set.
Macroinvertebrates isotope values varied with respect to size of individuals and
functional feeding groups. Also, MDN incorporation varied among macroinvertebrates.
This was associated with apparent changes in trophic position relative to prespawning
periods. We examined small and large sized Drunella from upper and lower Fish Creek
sampled in August 2001 (Table 1). Large Drunella were more enriched in 15N relative to
the smaller individuals (t-test, p < 0.01). Carbon isotope values of large and small
individuals were not signiﬁcantly different (t-test, p > 0.1).

Macroinvertebrates within the scraper group, C inygmula, Drunella, and Epeorus
differed from each other in their carbon and nitrogen isotopes values regardless of
sampling period in both years (Fig. 4, Fig. 5, ANOVA, p<0.01). Overall, the average
513 C value of Epeorus was lower than that of the other two scraper taxa regardless of
location or sampling period (Table 1, ANOVA, p<0.01). Drunella samples had the
highest 815N values during pre-spawning periods compared to the other two scraper taxa
collected in 2000 and 2001 (AN OVA, p<0.01). After spawning in the lower reach,
Cinygmula were the most enriched in '3 C and 15N relative to the other two scraper taxa

during 2000 and 2001.

15

Qualitatively, differences in the % MDN incorporation appeared to occur between
the scraper taxa and among different functional feeding groups (Table 2). Post-spawning
data show that Cinygmula incorporated more MDN nitrogen than did the other two
scraper taxa in both 2000-2001 (56-69 % and 40—5 7 %, respectively). As for MDN
carbon incorporation, C inygmula exhibited the greater % MDN carbon incorporation (71-
77 %) than the other two scraper taxa (40-52 %) in 2000 and 2001. As indicated by
carbon and nitrogen isotope values, the macroinvertebrate trophic structure (the relative
position of an organism within the food web), in upper Fish Creek remained relatively
constant over the season and between years (Fig. 4c and 4d). In contrast, the trophic
structure of the macroinvertebrate food web changed from the pre-spawning and post-
spawning periods in lower Fish Creek, which was a result of the different degrees of
MDN incorporation by functional feeding groups (Fig. 4a and 4b).

Macroinvertebrate carbon and nitrogen isotope values were enriched in 13 C and
15 N in the lower reach relative to the upper reach, regardless of season or year
(rmAN OVA, <0.005) (Table 1). Evidence for the presence of MDN in lower Fish Creek,
in the absence of spawning salmon, is exhibited by the degree of MDN incorporation in
the stream biota during the pre-spawning sampling period (Fig. 5a and 5b). The “legacy
effect” was more apparent in 2000 relative to 2001. Estimates of % MDN nitrogen
incorporation for macroinvertebrate consumers in the lower reach prior to spawning were
lower than those based on carbon incorporation for both years (Fig. 5a and 5b). After
salmon spawned, the average % MDN carbon and nitrogen incorporation for all

macroinvertebrates collected was 44 and 50 % for 2000 and 2001. As for carbon, the

16

average % MDN incorporation for macroinvertebrates was 59 and 48 % for 2000 and

2001.

Freshwater Fishes

Resident Dolly Varden from the upper reach were signiﬁcantly depleted in 13 C
and 15N relative to Dolly Varden collected in the lower reach during 2000 (t-test, p<0.01;
Alowemppe, reaches 5'5N s 6 %o, Alowe,-uppe, reaches 513 C s 5 %o) (Table 1). In the lower reach in
2000, carbon isotopes values for Dolly Varden appeared higher after spawning than prior
to spawning, but this difference was not statistically different (t-test, p>0.05). During the
same time period, however, the nitrogen isotope values were statistically different
between pre- and post-spawning (t-test, p<0.01). In 2001, no Dolly Varden samples were
caught in minnow traps in either pre- or post-spawning periods.

Estimates of % MDN incorporation were high before spawning; 50-56 % and 49-
55 % for carbon and nitrogen, respectively. The high degree of MDN incorporation in
Dolly Varden prior to the salmon subsidy could reﬂect the ability of DV to access other
MDN resources during migration to and from the estuary (e. g. estuarine invertebrates).
After spawning, the °/o MDN incorporation for Dolly Varden was 80-91 % and 59-67 %
for carbon and nitrogen, respectively. However, because the individuals below the barrier
could be accessing MDN from other sources than the salmon subsidy, pre- and post-
spawning individuals from the lower reach were used in equation 2 to estimate the %
MDN incorporation instead of comparing the upper and lower reaches. In 2000, Dolly

Varden incorporated an average of 62 % MDN carbon and 26 % MDN nitrogen after

17

spawning. In the upper reach, Dolly Varden carbon and nitrogen isotopes were very
similar between sampling periods (5.8 to 6.3 %o and —26.0 to -27.1 %o, respectively).

Dolly Varden 515 N values were not correlated with fork length either prior to or
during spawning in the upper reach, but there was a correlation between BISN and fork
length in lower Fish Creek (Fig. 6). Interestingly, the carbon and nitrogen ratios of the
post-spawning Dolly Varden sampled in 2000 from the lower reach are positively
correlated. Also, samples from this period exhibit the most enriched isotope values,
suggesting that even though these ﬁsh are likely to be migratory there is some evidence
for MDN incorporation (Fig. 6).

Coastrange sculpin only exist in lower Fish Creek. Because we did not have
individuals from the upper reach, sculpin from pre-spawning periods were used in the
models to estimate %MDN. All coastrange sculpin were collected in 2001. Carbon and
nitrogen isotope values were signiﬁcantly different in the post-spawning compared to
pre-spawning (Table 1, t-test, p<0.01). Freshwater sculpin incorporated an average of 83-
100 % MDN carbon and 16-19 % MDN nitrogen. Unlike other ﬁsh analyzed in this
study, the average 613 C value of post-spawning sculpin (11.8: 0.9 %o, n = 10) was higher
than that of the bulk isotopic signature of adult salmon (-1 8.3 i 1.2 %o, n = 10) (Table 1).
Sculpin had a large degree of variability in 513 C value from both the pre- and post-
spawning periods. The 5‘5 N values were not positively correlated with fork length.

All YOY coho samples were collected in 2001 from Fish Creek. Carbon and
nitrogen isotope values of these YOY coho were higher after spawning relative to pre-
spawning (t-test, p<0.05, Fig. 7). The average % MDN incorporation for YOY coho was

34-57 % for carbon and 22-28 % for nitrogen. YOY coho 513 C and 815N values were

18

positively correlated and are a function of the time of sampling. Coho salmon parr
showed similar patterns to the YOY coho, including no correlation between SISN and
fork length, but a positive correlation between carbon and nitrogen isotope ratios that is a

function of the time of sampling.

Longitudinal and Inter-A nnual Trends

C inygmula, Epeorus, Drunella, and Chloroperlidae were collected from a
sufﬁcient number of sub-reaches in lower Fish Creek to evaluate longitudinal trends in
isotope data. In several cases, these taxa exhibited a relationship of increasing 813 C and
515 N values with distance downstream from the salmon barrier during and post-spawning
in both years (Fig. 8). The trend was evident for SUN values of Chloroperlidae and
C inygmula in 2000 and 2001. It was also observed for 813C values of Epeorus in 2000
and 2001 and for 513 C data for Drunella in 2001. The longitudinal trend did not occur
prior to spawning or during any sampling period in the upper reach. The slopes of the
relationship between distance downstream from the barrier and a) the carbon isotope
values of Epeorus and Drunella in 2001 and b) the nitrogen isotope values of C inygmula
and Chloroperlidae in 2000 and 2001 were similar for during and post-spawning periods,
respectively (Fig. 8). During the salmon run in 2001, differences in 813C and SUN
isotope values of Cinygmula collected 1.2 kilometers apart (endpoints of our sampling
transect) were 1.6 and 3.8 %o , respectively. The associated % MDN incorporation

estimates ranged from 16 to 36 % (Fig. 9).

l9

Estimates of the salmon escapement between 2000 and 2001 did not differ (D. T.
Chanoler, unpublished data) However, the biomass delivered to Fish Creek was slightly
greater in 2001, due to a greater contribution by chum, but historically both years
experienced low returns (D. T. Chaloner, unpublished data). The “during” spawning
period in 2001 occurred approximately two weeks after the arrival of the ﬁrst salmon
migrating upstream. As a result, only live salmon (with no sign of decay) existed in Fish
Creek during this sampling period (B. S. Graham personal observation). Isotope values
of macroinvertebrates collected during the salmon run exhibited elevated carbon and
nitrogen values relative to pro-spawning values from lower Fish Creek (Table 1, Fig. 10).
The scraper, predator, and collector-gather functional feeding groups incorporated 25-52
and 24-38 % MDN carbon and nitrogen, respectively, relative to the pre-spawning period
(Table 2). The shredder functional feeding group (i.e. utilize CPOM as a food resource)
incorporated less MDN during the salmon run at 14-15 and 19-20 % for carbon and
nitrogen, respectively. After spawning, estimates of % MDN incorporations were
different between 2000 and 2001 for the scraper and predator taxa (Table 2). Cingymula
and Epeorus exhibited similar 813 C values between years and greater 515N values in 2001

relative to 2000 (Fig. 10).

DISCUSSION

A comparison of 513 C and 5'5 N values of stream biota were made between the
treatment and reference reach of Fish Creek. Sub-reaches from each of these locations
were sampled prior to and after spawning in 2000 and 2001, and during spawning in

2001. Our Fish Creek data set allows us to examine differences (1) between reference and

20

treatment reaches, (2) among sub-reaches, (3) among pre-spawning and post-spawning
periods in 2000, and pre-, during, and post-spawning periods in 2001. The results from
this study illustrate that Fish Creek consumers incorporated MDN, but the degree of
incorporation was inﬂuenced by the spatial and temporal variation in MDN, metabolic
turnover rate of the organisms, trophic time lag, the age of the organism, and migratory

behavior.

The MDN Endmember and Barrier Eﬂect

The upstream-downstream study design applied in Salmon and Fish Creek is one
approach to evaluate MDN subsidies on stream ecosystems (Kline et al. 1993; Bilby et a1.
1996; Johnston et a1. 1997, D. T. Chanoler et al. 2002). Although this approach is
commonly used to evaluate the importance of MDN in stream ecosystems, little has been
done to evaluate the issue of pseudoreplication (Hurlbert 1984). In other words, without
any effect of MDN, do upper and lower reaches have similar food web structures?
Salmon Creek provided an opportunity to examine this issue as there were two barriers
(i.e. waterfalls) leaving two study reaches without any MDN inﬂuence.
Macroinvertebrates collected from such two reaches in Salmon Creek showed no
difference in their carbon isotope values (Table 3). C inygmula showed minor differences
in its nitrogen isotope values between the two reaches but this was deemed insigniﬁcant
considering the standard deviation associated with the replicates and error associated with

the measurement (Table 3). Overall, in the small discrete watersheds of Salmon Creek,

21

stream macroinvertebrates showed a high degree of isotopically similarity above and
below a barrier.

The carbon and nitrogen isotopic composition of adult paciﬁc salmon did not
differ (1) between pink and churn salmon, (2) between wild and hatchery individuals, and
(3) among fresh, spawned out, and decomposing salmon. The most likely interpretation
for these data is that pink and chum salmon, both wild and hatchery individuals, have
similar feeding strategies or subsist at the same trophic position. The data also shows that
salmon entering streams retain their marine isotopic signature throughout their freshwater
migration upstream and during their subsequent decay. Kline et al. (1993) ﬁrst reported
differences in 813 C and 815 N between fresh and spawned-out Paciﬁc salmon and showed
that spawned-out individuals had higher 613C and 8'5 N values relative to fresh
individuals. Doucett et al. (1999) suggested 513 C values increase as a result of energy
reserves, primarily lipids, being lost during spawning migration of Atlantic salmon
(Salmo salar), but SUN values seem unaffected by the upstream migration. The
discrepancy between our carbon isotope data and those of Kline and Doucett results
because our samples were lipid extracted. The removal of lipids produces an increase in
the carbon isotope values of the organism because lipids are at least 3 %o more depleted
in 13 C relative to the average bulk tissue 813 C value (Tiezen et al. 1983). We conﬁrm this
in our study by comparing the isotope values of lipid extracted tissue to the non-lipid
extracted counterparts of three macroinvertebrates: Drunella, Chloroperlidae, Cinygmula
and Zapada. Considering the average difference for all taxa, the lipid-extracted aliquots
were depleted in 13 C by 1.2 i 0.4 %o relative to the non-lipid extracted sub-samples. No

differences in BUN values were found between the lipid-extracted and non-lipid extracted

22

sub-samples (515N values differed by 0.3 :i: 0.2 %o between the two groups) (Fantle et al.
1999). High 513 C values observed by Kline et al. (1993) in decomposed salmon are
consistent with the loss of 13 C depleted lipids. In conclusion, our results conﬁrm lipid
content will affect the 513 C values of adult salmon, but after lipid extraction adult salmon

have similar isotopic composition throughout their migration upstream.

Food Web Components

Bioﬁlm

Autotrophic uptake of mineralized MDN could be an initial and crucial link
between the MDN subsidy and upper trophic levels in lotic ecosystems. However, both
CPOM and bioﬁlm did not show a clear response to the MDN subsidy and their trophic
relationship to macroinvertebrates was not clearly discerned based on nitrogen isotope
values (Table 1). Several studies have examined 813C and SUN values of bioﬁlm with
respect to MDN subsidies (Kline et al. 1990, Bilby et al. 1996, Johnston et a1. 1997, and
Chaloner et al. 2002). Whereas Kline et al. (1990) and Bilby et al. (1996) showed MDN
incorporation by the bioﬁlm community, this trend was not demonstrated in Chaloner et
al. (2002). Chaloner et al. (2002) discuss a variety of mechanisms that can result in
isotopic variation and overprinting of the MDN signature of the bioﬁlm community
including differences in algal growth, nutrient pool size (Finlay et al. 1998), and,
community structure. They suggest that in addition to MDN, the 513C values of bioﬁlm
could respond to algal growth rates and their inﬂuence on 13C discrimination during

photosynthesis (Laws et al. 1995). They also point out that isotopic discrimination (i.e.

23

fractionation) during nitrogen cycling (e. g. nitriﬁcation) or the pool size of the dissolved
inorganic nitrogen reservoir can inﬂuence the SISN of algae (Fogel and Cifuentes 1993).
The latter seems particularly relevant because like Bilby et al, (1996), large variation in
the size of the dissolved inorganic nitrogen pools as a consequence of the MDN subsidy
and its subsequent decline have been observed in our study and elsewhere. (Mathisen et
a1. 1988, Bilby et a1. 1996, Mitchell, 2003). Bioﬁlm isotopic values could be an artifact
of the relative proportion of heterotrophs to autotrophs (Hamilton et al. 1992, Chaloner et
al. 2002). Whereas, both autotrophs and heterotrophs take up dissolved inorganic or
organic nitrogen, only heterotrophs are likely to assimilate nitrogen from marine-derived
particulate matter (Fenchel et al. 1998). Finally, studies which analyzed ﬁlamentous
algae (Kline et al. 1990, Johnston et al. 1997) exhibited less carbon and nitrogen isotopic
variation than those that analyzed bioﬁlm randomly sampled from rock surfaces (Bilby et
al. 1996, Chaloner et al. 2002). Therefore, in the context of nutrient cycling at the base of
the stream food web, MDN input is only one among many variables affecting the isotopic
composition of bioﬁlm.

Similar to the results of Chaloner et al. (2002) in Margaret Creek, AK, the isotope
values of our bioﬁlm in pre-spawning periods in upper ﬁsh creek were frequently higher
than those of associate macroinvertebrates from the same location and time. Furthermore,
515N values of bioﬁlm in pre-spawning period were higher than those during spawning in
upper and lower Fish Creek. These observations are consistent with our conclusion above
regarding multiple factors controlling the isotope values of bioﬁlm. An important line of

future inquiry would be to better characterize the isotopic compositions of the bioﬁlm

24

(e.g. Hamilton et al. (1992) method) and therefore, provide a more comprehensible link

between the bioﬁlm community and primary consumers.

Macroinvertebrates

In stream food webs, macroinvertebrates represent an essential link between the
lower trophic levels and ﬁshes and are therefore, crucial to understanding organic matter
and nutrient cycling in streams. The increase in isotope values between pre- to post-
spawning periods for macroinvertebrates isotope values generally support the established
concept of MDN incorporation in stream ecosystems (Fig. 3). Temporal trends in carbon
and nitrogen isotope values of Chloroperlidae spp. suggest that they experienced MDN
incorporation, however differences among sampling periods in both years were not
statistically signiﬁcant. This is, in part, due to a) high variation about the mean isotope
values during pre-spawning periods and b) low sample numbers (Table 1, Fig. 4).

The notably high isotopic variability among Chloroperlids could be a
consequence of (1) variation in feeding habits with depth in substrate, (2) variation in
nitrogen isotope values at the base of the food web, and/or (3) trophic position.
Chloroperlidae spp. 8'5N values reﬂect the average 15 N trophic level of all their prey
(weighted by dietary contribution) plus approximately 3 %o (DeNiro and Epstein, 1979).
Differences in prey and associated average 515 N of the diet might occur if the habitat of
Chloroperlidae spp varies within the stream. For example, Chloroperlids located deeper
within the substrate might have access to different prey than those that live closer to the

substrate-water column interface. To illustrate that such variation in diet could inﬂuence

25

the isotope values of Chloroperlids consider the following example. If a chloroperlid at
the surface of the stream bottom consumed 50 % baetidae (BISN = 1.5 %o) and 50 %
paraleptophelibidae (515N = 8.3) then the 5‘5N value of Chloroperlidae should be
approximately 7.9 %o, including a trophic fractionation. Whereas, if a cholorperlidae‘
located deeper in the substrate consumed 10 % baetidae and 90 °/o paraleptophelibidae
then the 515N should be approximately 7.6 %o.

In addition to changes in diet that occur between habitats, the isotopic
composition of the base of the food web would also contribute to variation. With
increasing depth in the stream proﬁle the available nitrogen pool could be inﬂuenced by
different nitrogen cycling processes than those occurring in the stream (Ostrom et al.
2002). For example, although nitriﬁcation dominates in highly aerated streams resulting
in 15N depletion in the production (i.e. N03) and 15N enriched values in the residual
substrate (NHX) if suboxic localized regions develop in the subsurface, denitriﬁcation
would dominate. In this case, the residual substrate, N03 would be enriched in 15N. Thus,
the 15N at the base of the food web would depend on (1) the DIN that supplies nitrogen to
the foodweb (N 03 or NH4+) and (2) the predominant process that inﬂuences the source.

Chloroperlids were identiﬁed to the taxonomic level of family, while most other
macroinvertebrates were separated into separate genera. Different genera within the same
family group exhibited differences in nitrogen isotope values that could indicate
differences in feeding strategies or differences in the isotOpe value at the base of the food
web. If the former is the case, some of the nitrogen isotopic variability could result from
different trophic position of the various taxa lumped into the Chloroperlidae spp.

samples.

26

Isotope values of macroinvertebrates varied with respect to size and functional
feeding groups (Table 1). Drunella carbon isotope values were similar between large and
small individuals but nitrogen isotope values were higher in larger relative to small
individuals. Drunella appears to be consuming a similar carbon resource throughout their
development. The large and small individuals were collected from the same sub-reaches
and at the same time. Therefore, the base of the food web did not change and, thus, can
not account for the observed differences in nitrogen isotope values between the two size
classes. Instead, differences in SUN values between the two size classes may reﬂect a
change in breadth of diet. Even though food acquisition mechanisms remain similar
throughout the life stages of scraper taxa, the increase in the size of mouth parts of larger
individuals enhances the possibility of accidental ingestion of early instars (Cummins
1980). Ingestion of early instar macroinvertebrates would increase the trophic position
and, consequently, the isotope value of Drunella.

Functional feeding groups (FF G), a classiﬁcation system used extensively by
stream ecologists, are deﬁned by morpho-behavioral characteristics (Merritt and
Cummins 1996). The FFG classiﬁcation, combined with stomach content analysis are
used to examine trophic interactions and structure (e. g. Harri gan et al. 1989). Stomach
content analysis provides investigators information biased toward an organism’s most
recent feeding event and the FF G method designates a general feeding strategy related to
the organism’s life history. However, the F FG method and stomach content analysis do
not always reﬂect an organism’s diet, due to differential digestion and assimilation. SLA
is advantageous because 1) it records only the assimilated portion of the diet and 2) it

gives a time-integrated view of the diet. Thus, isotope data provides some additional

27

information to our understanding of food webs based on stomach content analysis and
FFG method. The FFG approach indicates that Cinygmula and Epeorus are scrapers. The
stable isotope data (for all sampling periods) suggests that these two taxa likely have
different feeding strategies as their 513 C and SUN values deviate from one and another
regardless of location. The FFG method indicates scrapers utilize bioﬁlm as their primary
food resource. If the different scrapers were indiscriminately feeding on bioﬁlm the
isotope values would be similar. The difference in isotope values between the taxa
observed here could result if certain scraper taxa assimilate a more labile component of
the bioﬁlm, which might have a different 513 C value than the bulk bioﬁlm 513 C value
(Hamilton et al. 1992). It appears Cinygmula and Epeorus are either utilizing different
components of the bioﬁlm or are supplementing their diets with other food resources that
are isotopically distinct.

MDN incorporation varied within scraper functional feeding group prior to and
after spawning. In both years, C inygmula exhibited an average of 67 % MDN (C & N)
incorporation in the post-spawning period, the greatest for any macroinvertebrate taxa,
whereas, Drunella and Epeorus incorporated approximately 46 and 47 % MDN,
respectively. Relative to the other two scrapers, Cinygmula could be utilizing a food
resource more heavily inﬂuenced by MDN by either (1) selectively grazing a portion of
the bioﬁlm community that incorporated more MDN, (2) feeding directly on or
accidental ingestion of MDN particles, or (3) could be located in a habitat where the base
of the food web is 15N enriched. Finally, it is important to note that a common strategy in
organisms is mixotrophy and invariably this can apply to macroinvertebrates and would

lead to variability in the carbon and nitrogen isotopic data. It is clear, however, that

28

scraper taxa, in particular Cinygmula is utilizing a food resource that is highly inﬂuenced
by the MDN subsidy.

The high MDN incorporation estimates exhibited by macroinvertebrates could
reﬂect fast tissue turnover rates and a larger degree of isotopic disparity between the
MDN subsidy and the stream-derived diet. The rate that an organism’s isotope value
changes is a function of net growth efﬁciency or in other words, the degree of growth
relative to the degree of assimilation (Fry and Arnold 1982, Tiezen et a1. 1983).
Invertebrates have rapid isotopic turnover (i.e. within 30 days, Fry and Arnold 1982,
Ostrom et al. 1997), therefore, changes in diet should have been observed during our
sampling period of 5 months. In addition, the MDN subsidy, which is isotopically distinct
from the stream diet, should lead to a conspicuous shift in isotopic values. Thus, a
response to changes in diet is more likely to be detected in an organism with fast tissue
turnover rate and short-lived compared to an organism that is longer-lived with a slow
tissue turnover rate (Hesslein et al. 1983). Furthermore, long-lived stream organisms such
as Chlroperlidae, relative to other macroinvertebrates with a shorter life span would
experience multiple pulses of MDN and their tissues would represent a time—integrated

diet.

The Legacy Eﬂect

Lower Fish Creek food web demonstrated a legacy of marine-derived nutrients

retained from the previous year’s spawning salmon. Accordingly, anadromous salmon

may represent a subsidy that sustains stream productivity throughout the year. Marine-

29

derived nutrients and organic matter can be retained in the stream by physical retention in
large-woody debris (LWD) (Bilby and Likens 1980) or by subsurface retention (Edwards
2000). The degree of retention depends on many factors, including the number and
intensity of spate events, the amount of LWD, the retention and utilization of MDN in
ground water or the hyporheic zone, the degree of absorption of MDN onto particles in
the sub-surface, and biotic uptake turnover time.

Relative to upper Fish Creek, several groups of macroinvertebrates exhibited
elevated carbon and nitrogen isotope values prior to salmon spawning in both years (Fig.
5a, 5b). This suggests that the lower reach is experiencing spring carryover from previous
MDN exposure. Previous studies have shown that over-wintering ﬁsh retain MDN
signatures into the spring (MacAvoy et al. 1998, Doucett et al. 1999) and likely represent
isotope dilution between a MDN-derived diet in the fall to a strearn-derived diet
throughout the winter (Fry and Arnold 1982). However, our study illustrates that short-
lived macroinvertebrates (e. g. grazers and collector—gathers) are acquiring a MDN
signature, in absence of spawning salmon, after hatching in lower Fish Creek in the
spring. In other words, primary consumers are incorporating marine nutrients retained in
the freshwater food web.

Although many factors could affect the isotope values of food web components
throughout the year, the conspicuous annual migration of salmon seems an important
candidate for explaining the elevated values in the lower relative to upper Fish Creek.
The range in % MDN incorporation in the lower reach during pre-spawning periods for
all macroinvertebrates in 2000 and 2001 was 4-56 and 4-37 % for carbon and nitrogen,

respectively (Table 2). Interestingly, the estimates of the legacy effect determined using

30

carbon isotope values differed from those determined using nitrogen isotope values (Fig.
5). MDN incorporation estimates are derived using a comparison between an upstream
and downstream consumer (Equation 2). If factors other than MDN input inﬂuence
isotope values between consumers from upper and lower Fish Creek this would lead to
error and perhaps account for the disparity between estimates based on carbon and
nitrogen isotope values. For example, in the case of carbon, upstream consumers are
more likely to assimilate l3C-depleted terrestrial CPOM than are downstream organisms.
Thus, one could argue that carbon isotope values of consumers in upper Fish Creek might
not provide an accurate basis for determining estimates of % MDN in our mass balance
equations. However, the Salmon Creek data set demonstrated that groups of
macroinvertebrates had similar carbon isotope values above and below a barrier (neither
exposed to MDN) sampled during August 2001. An alternative explanation for
differences in %MDN incorporation between estimates of carbon and nitrogen is that
carbon has a longer residence time the Fish Creek food web than nitrogen. In addition,
nitrogen is often a limiting nutrient in terrestrial and marine systems and is utilized more
rapidly than carbon.

A legacy of MDN in streams could represent an important source of nutrients
throughout the winter months and into the spring for young-of-the-year salmon.
Consequently, salmon might not only affect the ﬂow of energy and nutrients through the
stream food web, but may also be a crucial factor in their own production and restoration
(Kline et al. 1990, Bilby et al. 1996, Garman and Macko 1998). If MDN is indeed
retained from the previous year’s subsidy, then differences in salmon escapements would

affect the % MDN legacy during the next spring. However, before management practices

31

are implemented to enhance smolt production based on the ‘positive feedback theory’ of
salmon escapements, a more rigorous and systematic examination of the relationship
between salmon escapements and the MDN legacy needs to be conducted. For example,
the relationship between salmon escapements and smolt production might not be a linear
relationship, but instead one with an asymptote representing the degree of MDN
saturation in the stream food web. The delivery of the MDN subsidy to the
macroinvertebrate food web appears to be substantial and remains throughout the year,

whereby the macroinvertebrates can utilize the long-term resource the following year.

Freshwater Fishes

The quick and evident response of the macroinvertebrate food web to the MDN
subsidy can be contrasted to that of resident freshwater ﬁsh, which may (1) have multiple
exposures to MDN and (2) have multiple marine-derived food resources in Fish Creek.
In Southeast Alaskan streams migratory ﬁsh can acquire energy from two marine
sources: salmon-derived nutrients (SDN) and estuarine-derived nutrients (EDN). Like
SDN, relative to terrestrial and freshwater nutrients, EDN is likely to have high isotope
values because it is strongly inﬂuenced by marine nitrogen. Dolly Varden from lower
Fish Creek appear to be a migratory ﬁsh, and therefore, would have access to both SDN
and EDN. Thus their isotope values would not be a simple function of SDN. Instead the
values would depend on the amount of time the ﬁsh foraged in the stream, estuary, and
marine environments, the success of foraging in these environments, the trophic level of

their prey, and their subsequent growth rates in each environment (Hansson et al. 1997).

32

Presumably, Dolly Varden enter Fish Creek when it is most physiologically beneﬁcial,
i.e. with the arrival of the MDN subsidy. It has been well documented that Dolly Varden
feed on salmon eggs during the upstream migration of spawning salmon (Fechhelm et al.
1997). Therefore, there should be a pulse of SDN in the Dolly Varden diet that coincides
with the time salmon spawn. Dolly Varden migrate out to saltwater sometime after their
second, third, or fourth year in the stream (Reed 1967). Despite these complicating
factors, Dolly Varden isotope values still appear to be inﬂuenced by the annual pulse of
SDN. There are two lines of isotopic evidence suggesting that Dolly Varden migrate into
the stream during the pulse of SDN, incorporate SDN, and then migrate out to saltwater
as they age. First, larger Dolly Varden 815N values are more enriched in 15N than smaller
individuals. Freshwater ﬁsh that migrate and experience a new diet that is more enriched
in 15N show a speciﬁc patter in their SISN values. As the ﬁsh ages, or fork length
increases, 515N values increase (Doucett et al. 1999b). Therefore, higher S’SN values for
larger Dolly Varden would indicate these ﬁsh are consuming a diet more enriched in lsN.
The diet could be richer in EDN or SDN and/or they are feeding at a higher tr0phic level
in the estuary or stream. Second, this is most likely occurring in the stream because the
greatest SUN and 513 C observed for Dolly Varden in Fish Creek occur shortly after
spawning. However, this data could be explained by either an increase in MDN or a shift
to a higher trophic level non-MDN diet. Dolly Varden represent a highly migratory ﬁsh,
which complicates any interpretation of results when evaluating the importance of SDN.
Therefore, investigating a ﬁsh species that does not have multiple sources of MDN (i.e.
non-migratory ﬁsh) would prove valuable in assessing the importance of SDN to

freshwater ﬁshes.

33

Coastrange sculpin are resident ﬁsh in Fish Creek and are not believed to be
migratory (Reed 1967). Sculpin are effective predators of salmon eggs, salmon fry, and
benthic and drifting invertebrates (Reed 1967, McLamey 1967, F oote and Brown 1998).
Sculpin can derive the majority of their dietary requirements from salmon eggs, which
are a very labile food resource with high lipid content (McLamey 1967). To demonstrate
the high lipid content of the eggs we compared the isotope value of lipid to non-lipid
extracted eggs. Two samples of coho salmon (0. Keta) eggs were separated into two
aliquots, one set was not lipid-extracted and the other was lipid-extracted. The lipid
extracted aliquot was nearly 2 %o more enriched in 13 C compared to the non-lipid
extracted sub-sample (i.e. -18.6 i 0.1 and -20.2 i 0.5, respectively), indicating a large
loss of 13 C-depleted lipids during extraction. 8N and 513 C values of sculpin exhibit an
increase from pre- to post-spawning periods and could represent a shift from a stream-
derived diet to one inﬂuenced by MDN. Large differences were exhibited between
sculpin carbon and nitrogen MDN incorporation, with MDN C % incorporation being
nearly a magnitude greater. Sculpin live for many years, thus, their muscle tissue samples
represent a time-integrated diet. Thus, it would be difﬁcult to observe an evident response
in isotope values to a seasonal MDN subsidy.

In contrast to sculpin, salmon fry are young ﬁsh that are more actively growing
and, thus, their isotope values should respond more rapidly to a change in diet (Hesslein
et al. 1993). Salmon fry will initially have a maternal SDN ﬁngerprint when they ﬁrst
hatch in streams in early spring (ca. March). At this point the YOY salmon begin feeding
on a stream-derived diet. Their diet may again be inﬂuenced by MDN when adult salmon

return to spawn in the fall (Doucett et al. 1996). This change in diet would elicit a strong

34

response in their isotope values, but it would also be inﬂuenced by the age and growth
rate of the ﬁsh. We estimated the stream-derived diet using stomach content data for
lower Fish Creek YOY coho and associated isotope values in a mass balance model
(Gould et al., 1997). Estimates of the resulting carbon and nitrogen isotopic values of the
stream-derived diet are 1.5 i 0.5 and -26.3 :I: 1.0 %o, respectively. In contrast, a salmon-
derived diet would be 11.4 %o and -19.9 %o for nitrogen and carbon isotope values,
respectively. The MDN-inﬂuenced stream diet, estimated using stomach contents of post-
spawning YOY coho from lower Fish Creek and associated isotope values in a mass
balance model are 8.5 and -23.2 %o for SN and 513C values, respectively (Fig. 12). YOY
coho isotope values increased after their ﬁrst exposure to the MDN. At the beginning of
the MDN subsidy, the SISN values of the YOY coho differ from the MDN-inﬂuenced
stream diet by about 3 %o. After the MDN subsidy, the difference between the YOY coho
and the MDN-diet is about 1 %o, at which point the values reach an asymptote (Fig. 12).
Initially after hatching, the stream-derived diet could be used for metabolic requirements,
but the majority of YOY growth could be derived from an MDN-inﬂuenced diet. F antle
et al. (1999) showed the difference between blue crab (Callinectes sapidus) BISN values
and their diet, or the trophic fractionation, was a function of the protein quality of the
diet. In other words, trophic fractionation varies with grth rate, which is a function of
the amount of dietary protein (diet quality). Organisms that are rapidly growing, on a
high quality diet, will fractionate dietary nitrogen less than slow growing organisms
feeding on a lower quality diet. The trophic fractionation exhibit between the YOY coho
and their estimated diet suggest that immediately before the MDN subsidy the trophic

fractionation was greater than after the MDN subsidy, which is in accordance with the

35

higher quality diet of the direct consumption of MDN. Freshwater ﬁshes in Fish Creek
responded to the MDN subsidy. However, future work should make a greater effort to
constrain confounding factors, including migratory behavior and multiple exposures to

MDN before assessing the importance of the SDN subsidy to these stream consumers.

Longitudinal Trends

The inﬂuence of marine-derived nutrients varied longitudinally in Fish Creek for
macroinvertebrates during and after spawning (Fig. 8). The longitudinal trend was
documented by macroinvertebrates 813 C and 515 N values increasing downstream from the
salmon barrier after spawning in 2000 and during and after spawning in 2001. To the best
of our knowledge, this is the ﬁrst time such spatial variation in MDN incorporation has
been documented in streams. The longitudinal trend could be a result of (1) tissue
turnover rates by stream consumers and/or (2) differences in marine-derived nutrient
loading with distance downstream.

Previous stable isotope based diet switching experiments indicate that
invertebrates have tissue turnover rates on the order of 30 days (e. g. Ostrom et al. 1997).
We estimated tissue turnover based on the number of days over which isotope values
increase from the stream-derived diet to the MDN-inﬂuenced diet. If a drifting
macroinvertebrate enters lower Fish Creek from the upper reach, the time it takes their
tissues to turnover and reﬂect the MDN-inﬂuenced diet provides an estimate of turnover
rate. For example, during spawning, our grazer taxa exhibited rapid MDN incorporation
of approximately of 30% after 12-14 days of the ﬁrst arrival of salmon in Fish Creek

(Fig. 11). Given enough time, macroinvertebrate tissues will approach isotopic

36

equilibrium, reﬂecting the MDN diet. However, many macroinvertebrates drift only a
few meters (Rader 1997). Accordingly, since macroinvertebrates don’t travel large
distances longitudinally their isotopic signature should reﬂect the degree of the MDN in
their immediate environment. If the degree of MDN base of the food web varies
longitudinally then macroinvertebrates will mirror this trend. Initially MDN is deposited
in streams as excretion and later as carcass deposition and decomposition. Salmon
carcasses do decrease as you move upstream. However, the carcasses are
heterogeneously dispersed in Fish Creek (Chaloner, unpublished data). Overtime the
carcasses begin to decay, and produce a particulate pool of nutrients that will be entrained
downstream. Therefore, the longitudinal trend could be a function of the accumulation of
marine-derived nutrients downstream, in dissolved and particulate form, whereby the
main mechanism for MDN retention is through the retention of remineralized MDN by
the bioﬁlm community. Carbon and nitrogen isotope values should be highest in areas
where the most MDN is deposited. Nutrient dynamics in streams are tightly coupled with
the physical movement of water (Newbold et al. 1983). An area downstream has had
more exposure to marine-derived nutrients ﬂowing by than an area further upstream.
Bioﬁlm located downstream, relative to upstream, are exposed to a higher concentration
of dissolved and particulate marine-nutrients as a consequence of downstream ﬂow of the
salmon derived nutrient pool. Therefore, a trend of increasing isotope values in stream
consumers downstream from the salmon barrier could be a function of streams being a
longitudinally-dominated system where an organism downstream has a cumulative

exposure to MDN.

37

Longitudinal isotopic variation in streams has major implications for evaluating
the inﬂuence of salmon on stream ecosystems and associated management strategies.
Depending upon the location of sampling, the estimated % MDN incorporations, using
linear mixing models, could vary considerably and the % MDN incorporation could be at
least 20 % different than estimates sampled from a downstream reach (Fig. 9).
Accordingly, spatial sampling is crucial to assess the inﬂuence of MDN on the stream

food webs.

Inter-Annual Data and Salmon Escapements

In addition to the spatial heterogeneity of MDN in streams, temporal variation in
MDN can play a large role in the response of stream consumers. Lower Fish Creek
consumers showed little difference in the % MDN incorporation during pre-spawning
periods between 2000 and 2001 (Table 1, Fig. 10). Whereas, after spawning the
macroinvertebrate carbon and nitrogen isotope values differed. In general, isotope values
were greater in the samples collected in 2001 relative to 2000 (Table 1, Fig. 10). Initially,
we examined if differences in salmon escapements could explain the observed trends.
Using several different methods, the salmon escapement levels of Fish Creek appeared
similar between 2000 and 2001 (Chaloner, unpublished data). In regards to MDN
biomass, 2000 appears to have received more biomass than 2001 due to greater numbers
of larger chum relative to pink salmon, but the difference is small (Chaloner, unpublished
data). Thus, SUN values of stream consumers are not a simple proxy for salmon

escapement levels. Instead, nitrogen isotope values can be inﬂuenced by many factors

38

including longitudinal variation in MDN accumulation, the spring carryover of MDN,
and in general, nutrient cycling in streams.

The two-year database did reveal a rapid response to the MDN subsidy by
primary consumers (Fig. 11). If the bioﬁlm community transfers MDN to the
macroinvertebrate food web via remineralization and utilization, then the scraper taxa
will incorporate MDN prior to any other functional feeding group (e. g. predators,
shredders). In Fish Creek, scrapers responded rapidly to the MDN pulse, with increases in
SBC and SN within weeks of the arrival of the ﬁrst adult salmon. At the time we
sampled during the spawning in 2001, only fresh salmon were present with no carcasses
present. Furthermore, our sampling overlapped the peak of MDN-NH; attributed to
excretion by the salmon spawning, which occurs only brieﬂy when the salmon ﬁrst arrive
in the stream (N. Mitchell, MSc. 2002). Accordingly, a portion of the bioﬁlm community
selectively utilized by macroinvertebrates could be characterized by rapid incorporation
of marine-derived dissolved inorganic nitrogen. Fish Creek data is in agreement with the
conclusions of Matheisen et al. 1988 and Kline et al. 1991, 1993, i.e. autotrophic uptake
of MDN. Bilby et al. 1996 argued primary production was limited by low light levels,
cold temperatures, short day lengths, and frequent spate events after spawning (Bilby et
al. 1996), and limited MDN incorporation via remineralization, therefore, MDN
incorporation into the stream food web occurred by direct consumption. However, our
results showing a rapid increase in isotope values associated with a peak in NH4+
suggests an important role for incorporation of dissolved inorganic nitrogen into the food

web via the bioﬁlm community.

39

Recently, Bilby et al. (2000) and Finney et al. (2001) suggested that 515N values
of different food web components could be used to develop a historical perspective of
salmon escapements. As we have demonstrated, nitrogen isotope values of
macroinvertebrates vary longitudinally in Fish Creek. Secondly, % MDN incorporation
varies temporally and would likely reﬂect the degree of retention and/or assimilation
throughout the winter. The degree of 15N-enrichment observed in one year would be a
function of the previous years salmon escapements rather than just the current year. Pre-
spawning macroinvertebrate 813C and 8'5N values were similar in 2000 and 2001 in
lower Fish Creek (Fig. 10). As a result, we can not interpret differences in % MDN
incorporation during a sampling period as a simple reﬂection of salmon biomass
delivered to Fish Creek. Moreover, we discussed that depending on which organisms
were sampled, the degree of incorporation may be function of tissue turnover rates,
longevity, and migratory behavior. For example, in Fish Creek many grazer invertebrates
responded rapidly to the MDN subsidy (Fig. 11), but predators, both Chloroperlidae and

resident ﬁshes showed pronounced response (Fig. 7).

C onclusion/Implications

I found that the stable isotope composition of primary consumers provided an
understanding on how lower trophic levels are inﬂuenced by the delivery of marine
nutrients and organic matter from spawning salmon. My research supports the results of
previous studies, in that signiﬁcant amount of MDN was incorporated into stream food

webs, by both lower and higher trophic-level consumers. However, I also found

40

signiﬁcant temporal and spatial variation, and results were not always as predicted from
previous studies, especially with regard to the bioﬁlm assemblage. Stable isotope
analyses have great utility in the study of MDN in freshwater ecosystems, but there is still
a need to better characterize the processes underlying the link between marine subsidies
and isotopic enrichment of food web components, and how this is linked to other
measures of ecosystem structure and function. This is especially critical if stable isotope

techniques are to be explicitly used in the management of Paciﬁc salmon ﬁsheries.

41

 

 

 

 

 

 

62 33. $3. «3. 35. man. $2 he"- 53 m... $2 5 82 3. A32 m._ weeaamoa
gasoem
9853b
ikcuumzcb
59c 33- 52 EN- 6: 3 52 ﬁn emcei.
ES 3". $2 an”- 3.3 m... $3 3 :3?
Eve v.2- as: sea- 62. man. 3 man. 33 3- $2 3 62. 3. E ~.m 3:83-38
53 3."- 3 on". $2 v... 5 an mean
52 3a- 5 sea. 6:. can- 53 an”- 62. 3 52 an 52. 3 See am 323335
~52.sz
6v; man- 52 e.m~- S; in. $2. Eu- GE. 3. 52 3 $3 3- 2%... «2m 3:38an
63 oi- 52 «.3. 53 2.- 53 «gm mesa
$3 33.. $2. EN- 52. 3n- 5 3a. 33 ms. 52. a... 53s 2.- A c a... messages...
Snowmm
5:. mg. 63 3a. 52. can- 5o... nun- 3:... I. 53 ad 52 N... 52 m.» weeaeomaom
52 33- 53 an"- 53 2.- 53 em wean
$3 man- 52 SN- 52 «23.. 9x3 3a. 33 ms. 53. 3 52 n... 2 :2 3 355835
oEEMAED
Emmoxow
36.2.0202:
5..an
SIC 3h- SI. 3."- are 2- 53 5 20,5
52 Ex. 52 2 8:2 .82
52 «.5. $3 can- SE 5 $3 2 weeaaraom
53 as"- 52 EN- 52 Wm $2 am 323835
ean
oousom
Leta—z 3.590
doe :8 H: F :8 H: _ an: H: _ Dom
EoeanoU
Som coca Sow ooom 83 noon
Assv 02w 3a 22w g ozee

 

 

 

 

42

 

ES 3:-
53 n.3-
53. n.2-
ES 3m-
53 mg:-
53 n.3-
53 EN-
3 c.2-
ES n.3-
63 2h- 52 n.3-
63 «2mm- 65. n.3-
€§ mimN- 53 «.2-
3 3n- Ee. c.2-
52 «:5- 52 3&-
€ @5- 5 v.3.-
€ «.3-
€ 3a-
5 @5- A : n.2-

52 n.5-

A : remn-

 

€ c.3-
E 3n.
3 n.2-
3 @3- £2 «nnu-
€ @2- 3 ven-
See «.2-

 

63 as _
aces 3:
53 ha—
55 v.2
63 ad
52 «.2
3%... a.—
3 3
$3 ﬁn
63. m... 52 we
5% e.»- Gvod ed
as: m.~ 52. E
5 an S: we
53 m... 52 we
€2.- Ema
E E.
E 3
S a... € .2
59 I. 5 mun.

 

52
Se.—

53 3

3 m...-

e 2.-

:V n:

Sam Qm

3mm

:33qu ANBQ
chBmdm-amom
chBaam-En—
233.6%
wEEsmam-amom
wEEsR—m-oi
EEK EEO
chBQO-50m
chBmam-oi
6&8 xQx
am:— consignee-a
chBaam-Eom
wESO
wEEsam-oi
vague—Lem
weEBaam-Cmom
chBmdm-oi
oeezzeooﬁé
wEEsam-Lmom
wet—5
chBQO-oi
32—53820
Eocene-ck
chBanm-EOQ
war—=0
chaQO-Em
unscrew.
2333380
maﬁaan-uom
chBst-oi
owes—oioao—Sam
w:_=>>aam-.mom
wESQ

 

43

 

 

3x3 n.3-
€«._ c.3-

 

52 mi-

52. «.3-
53. «.2-

53. «da-
vad v.3-

5? c.3-
63 E«-

62 c.2-
5..« b.««-

 

@«e am
52. 3

 

5.3 a:

53 :—
53. «.2

52 w:
A82 «.2

53 3 5.3 «.m—
€«.o am 53 ”.2

vii new 53:0
3380-; .5 ﬁrm
tuna
525% :33
via
83:0
mpg—«Eon U<n=Q
via
85:0
SEEP-H .5 55
53k
ion—Em =35.
wemgam-umom
mcmgmqm-Bm

 

44

Table 2:

 

 

 

 

 

Food Web Component 2000 2001
%MDNC %MDNN %MDNC %MDNN
Macroinvertebrates
Scrapers (l)
Cinygmula
Pre-spawning 37-39 10 34-35 12-13
During - - 49-52 36-38
Post-spawning 71-74 56-5 8 73-77 65-69
Epeorus
Pre-spawning 22-23 4 24-25 10-1 1
During - - 32-33 36-38
Post-spawning 52-54 40-42 40-42 54-57
Drunella
Pre-spawning 37-3 8 15-16 35-37 7
During - - 25-27 24-25
Post-spawning - - 41-43 50-53
Collector-Gathers (1)
Baetidae
Pre-spawning 45-48 7-8 34-36 8
During - - 26-28 35-37
Post-spawning - - 51-54 54-57
Shredders (1)
Zapada
Pre-spawning - - - -
During - - 14-15 19-20
Post-spawning NA 38-40 33-35 44-46
Predators (2)
Chloroperlidae
Pre-spawning 46-56 32-3 7 21-25 23-26
During - - 44-50 31-34
Post-spawning 54-67 36-41 55-63 43-48
Freshwater Fish 1
Y OY Coho (2)
Pre- vs. Post- - - 34-57 22-28
Sculpin (2)
Pre- vs. Post- - - 83-100 16-19
Dolly Varden (2)
Pre-spawning 50-56 49-55 - -
Post-spawning 80-91 59-67 - -
Lower pre- vs. post- 62-80 23-30

 

 

 

 

45

 

Table 3:

 

 

 

 

 

 

 

 

5‘5N 5‘3c
FOOd Web Component Above Below Above Below
Barrier Barrier Barrier Barrier

Scrapers

Cinygmula 0.5 0.3(3) 1.1 0.1(3) -29.l 0.1(3) -29.6 0.7(3)

Epeorus 0.0 0.6(3) 0.3 0.2(3) -29.7 0.2(3) -30.4 0.7(3)
Collector-gather

Baetidae 0.6 0.1(3) 0.8 0.2(3) -28.9 0.1(3) -29.4 0.0(3)
Predator

Chloroperlidae 3.3 (l) 3.0 (1) -27.5 (1) -27.5 (1)

 

46

 

 

 

 

 

96m?-

§m£< ammoﬁsom

_ bar-—

47

0:8 ~

I
Z
«can? 53:3
guano :mE
Aeouoom 8:80.“on
53m 3993‘
Emu-833v
EoEusoE ﬁrm 9 Siam
Acozoom mew—38$ I
c

53% .833

« oawa

48

ZOE

 

 

20.2 o
comom Long: 4

526-. m:_c3waw-.won_ I

526-. mEcBmaw-oi o

v

(D
(“96) N919

or

 

NF

m oeswE

49

 

 

eavoee

 

 

 

 

 

om- em- mmu ----- «m- cm. em-
D
EM
430
Aev on
m<m %mnﬁ TIT MAE
DMD 0>U U>U
Tali-mam
m<m ”HMO D 430
I
AIDA U D in Mam TﬂiU>U
96 SC a #6 A3

 

(0%) N919

or

 

«P
4 ocean

50

53,—.

 

 

- N-
uEemIQN SESEQ SEMAEU onEBam - T
- c
9
H
- _ N
\M
- N W
2-2 “Zn—2.x. «HZQEX.
2-3 H 292$ m
n n 2928 ZEN
- v
1 N-
SSEMAEU anew—Laeoﬁo 6%:on

 

Q.
H
3. “ 292 as N
- .V \I/
%
, c
nosed E964 E . . . . . 3.3 u 292 .X. - c
58% 5de I , . cch

 

mm oSwE

mmem H ZOE o\o mmém H ZOE AX. mmém H ZOE AX.

omém H ZOE o\o

       

wiomzmw £3:xe QEEMAED mSSBBM

om$m H ZOE .X. «v.9» H ZOE .X.

canon 533 n.
:83— SQQD I

 
 

omév H ZOE Ax.

  

 

   

  

EMAEU 321008.530 0538mm

 

E:

 

 

 

Om-

3
(°%) 3:19

3
(”96) 3919

52

A53 5?»: far,—

 

 

 

3 NO 3 w
. . . _ . . V
5
Saab No n «M - w
wﬁcammmémom was -05 I E + i.e- u A
833 wﬁcammméum 4 ﬂ

 

 

 

So H mm . S
.2” + 63 u a
On.
Q U m S
m T cm. A we 2 mm- cm. W
_[ _ _ w WN
(
- 2
wEGBQOémom I
MEQBQOéE 4 ﬂ
4 - NO
3 m. S: tame; - E

 

c 8sz

53

 

 

 

 

 

 

 

2- om- Fm- NN-
_ L . _ w 1
9 and N E
Q.
My Sségou»
3.0 u «t 3‘
33.33 H > 9 O O.
6 ﬂ u
03: 0N8 05 SE wmk. wk.

 

  
    

 

vnd H my.
wdoe - 566 H >

 

vm-

mm-

A:

N—

3

("96) 3:19

(0%) N919

n mama

82132030 I 33.5.5 < 3.6an a SzEMAED o

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

. - R-
.w_ow~z 4 cm nmonw_ Z
om~-xWo-uzn\nl.M.-F\m.\mv.m 9%;de
Nwd “N”— D
camux<_u» «a. .mw_&w
Am
_w.o H mm \M
3" Na wxméww n» W
95-x:vu»
9 mm- 0 .mm-
L! ﬁlo
0|! it I
. mmou~m o
r _ . m2. Q; n» om- . 5.
A55 3?... am 595 antm=>>cO 3:8»:—
o.m wN 9N in ad ad 0N ed
mwd n E
mg+xqmux
@ O ”NH— N. . 0
~a+xq_u» on
N;
“W
a Z w M\
m N. + X0 — H —— rO—

 

Sea

scam

 

w ouswi

Etc BENO Eom ooSwEO

Nd m Wm 9N vd Nd N M:

.1 _ _ O _ _ _ N

l
v

(0%) N919

I
In

 

wicmmm a! .. a.

33339.6 IT

 

a ”5mm

56

I on.

  

 

 

 

 

BEEMAEU D
”503$ l mm-
r om- Q.
m
- vm- \WM
1 mm-
l _ h
cor—om wEEEam

S “mom 5 wEEO 5 2m co 60m co 8m

N 9 CO CO v N O
(°%)Nng

 

 

S Rama

57

BmO

 

52>: Skim SRNG
_ _ O
4 \<
Stommmw \ \
\ \
Sgwkﬁb I \ \ o\°w

035238—50 O

 

22:2 883m .0:
63::me ﬁzz

 

NF

: Sam

wﬁsoua: £89595 :3...“ 23:32

 

 

on 2 2 m o
. h . _ h. c
mom—3m ZOE _om_:m ZOE
I w
........... “my?“ wombawmmng I I I I I I I I I I I I I 9
m.
I 3 x)
IIIIIIIIIIIIIIIIIIIIIIIIIIII .. .. .. I .. - - - Wm
36 wagging
I N.
O . O
I E

E Rama

59

LITERATURE CITED

Ashley, K. I. and P. A. Slaney. 1997. Accelerating recovery of stream and pond
productivity by low-level nutrient enrichment. In: Fish Habitat Rehabilitation
Procedures for the Watershed Restoration Program. BC Ministry of Environment,

Lands and Parks and Ministry of Forests, Watershed Restoration Tech. Circ. No.
9, Vancouver, BC.

Ben-David, M. T. A. Hanley, D. R. Klien, and D. M. Schell. 1997. Seasonal changes in

diets of coastal and riverine mink: the role of spawning Paciﬁc salmon. Can. J.
Zool. 75:803-811.

Ben-David, M., T. A. Hanley, and D. M. Schell. 1998. Fertilization of terrestrial

vegetation by spawning Paciﬁc salmon: The role of ﬂooding and predator
activity. Oikos. 83:47-55.

Bilby, R. E. and G. E. Likens. 1980. Importance of organic debris dams in the structure
and function of stream ecosystems. Ecology. 61:1107-1113.

Bilby, R. E., B. R. F ransen, and P. A. Bisson. 1996. Incorporation of nitrogen and carbon

from spawning coho salmon into the trophic system of small streams: evidence
from stable isotopes. Can J. Fish. Aquat. Sci. 53:164-173.

Bilby, R. E., B. R. Fransen, P. A. Bisson, and J. K. Walter. 1998. Response of juvenile

coho salmon (Oncorhynchus kisutch) and steelhead (Oncorhynchus mykiss) to

addition of salmon carcasses to two streams in southwestern Washington, U.S.A.
Can. J. Fish. Aquat. Sci. 55:1909-1918.

Bilby, R. E., B. R. Fransen, J. K. Walter, C. J. Cederholm, and W. J. Scarlett. 2001.

Preliminary evaluations of the use of nitrogen stable isotope ratios to establish
levels for Paciﬁc salmon. Fisheries. 26:6-14.

Brickell, D. C., and J .J . Goering. 1970. Chemical effects of salmon decomposition on

60

aquatic ecosystems. In Proceedings Symposium on Water Pollution Control in
Cold Climates. Edited by R. S. Murphy. US. Government Printing Ofﬁce,

Washington, DC.

Cederholm, C. J ., M. D. Kunze, T. Murota, and A. Sibatani. 1999. Paciﬁc salmon
carcasses: essential contributions of nutrients and energy for aquatic and
terrestrial ecosystems. Fisheries. 2426-15.

Chaloner, D. T., K. M. Martin, M. S. Wipﬂi, P. H. Ostrom, and G. A. Lamberti. 2002.
Marine carbon and nitrogen in southeastern Alaska stream food webs: evidence

from artiﬁcial and natural streams. Can. J. Fish. Aquat. Sci. 59:1257-1265.

Deniro, M. J. and S. Epstein. 1977. Mechanism of carbon fraction associated with lipid
synthesis. Science. 197:261-263.

DeNiro, M. J. and S. Epstein. 1978. Inﬂuence of diet on the distribution of carbon
isotopes in animals. Geochim. Cosmochim. Acta. 42:495-506.

DeNiro, M. J. and S. Epstein. 1980. Inﬂuence of diet on the distribution of nitrogen
isotopes in animals. Geochim. Cosmochim. Acta. 45:341-351.

Doucett, R. R., G. Power, D. R. Barton, R. J. Drimmie, and R. J. Cunjak. 1996.
Stable isotope analysis of nutrient pathways leading to Atlantic sahnon. Can J.
Fish. Aquat. Sci. 53:2058-2066.

Doucett, R. R., R. K. Booth, G. Power, and R. S. McKinley. 1999a. Effects of the
spawning migration on the nutritional status of anadromous Atlantic salmon

(Salmo salar): insights from stable-isotope analysis. Can. J. Fish. Aquat. Sci.
56:2172-2180.

Doucett, R. R., W. Hooper, and G. Power. 1999b. Indentiﬁcation of anadromous and
nonanadromous adult brook trout and their progeny in the Tabusintac River, New

Brunswick, by means of multiple-stable-isotope analysis. Trans. Am. Fish. Soc.
128:278-288.

Fantle, M. S., A. I. Dittel, S. M. Schwalm, C. E. Epifanio, and M. L. Fogel. 1999. A food

61

web analysis of the juvenile blue crab, Callinectes sapidus, using stable isotopes
in whole animals and individual amino acids. Oecologia. 120:416-426.

Fechhelm, R. G., J. D. Bryan, W. B. Grifﬁths, and L. R. Martin. 1997. Summer growth
patterns of nothern Dolly Varden (Salvelinus malma) smolts from the Purdhoe
Bay region of Alaska. Can. J. Fish. Aquat. Sci. 54:1103-1110.

Finlay, J. C., M. E. Power, and G. Cabana. 1999. Effects of water velocity on algal

carbon isotope ratios: Implications for river food web studies. Limnol. Oceanogr.
44:1198-1203.

Fisher Wold, A. K. and A. E. Heshey. 1999. Effects of salmon carcass decomposition on
bioﬁlm growth and wood decomposition. Can. J. Fish. Aquat. Sci. 56:767-773.

Foote, C. J. and G. S. Brown 1998. Ecological relationship between freshwater sculpins
(genus Cottus) and beach-spawning sockeye salmon (Oncorhynchus nerka) in
Iliamma Lake, Alaska. Can J. Fish. Aquat. Sci. 55:1524-1533.

France, R. L. 1994. Nitrogen isotopic composition of marine and freshwater
invertebrates. Mar. Ecol. Prog. Ser. 115:205-207.

Fry, B. and C. Arnold. 1982. Rapid ‘3 C/ 12C turnover during growth of brown shrimp
(Penaeus aztecus). Oecologia. 54:200-204.

Fry, B. 1988. Food web structure on Georges Bank from stable C, N, and S isotopic
compositions. Limnol. Oceanog. 33:1 182-1 190.

Garman, G. C. and S. A. Macko. 1998. Contribution of marine-derived organic matter to
an Atlantic coast, freshwater, tidal stream by anadromous clupeid ﬁshes. J. N.
Am. Benthol. Soc. 17:277-285.

Groot, C., and L. Margolis (eds.) 1991. Paciﬁc Salmon Life Histories. University of
British Columbia Press, Vancouver, BC. 564 pp.

Gould, P., P. Ostrom, W. Walker, and K. Pilichowski. 1997. Laysan and Black-footed

62

Albatrosses: trophic relationships and driftnet ﬁsheries associations of non-
breeding birds. In: Albatross Biology and Conservation, G. Robertson and R.
Gales (eds.). Surrey Beatty & Sons, Chipping Norton. pp 199-207.

Hamilton, S. K., W. M. Lewis, and S. J. Sippel. 1992. Energy sources for aquatic animals
in the Orinoco River ﬂoodplain: evidence from stable isotopes. Oecologia.
89:324-330.

Hansson, S., J. E. Hobbie, R. Elmgren, U. Larsson, B. Fry, and S. Johansson. 1997. The
stable nitrogen isotope ratio as a marker of food-web interactions and ﬁsh
migration. Ecology. 78:2249-2257.

Harrigan, P., J. C. Zieman, and S. A. Macko. 1989. The base of nutritional support for the
gray snapper (Lutjanus griseus): An evaluation based on a combined stomach
content and stable isotope analysis. Bulletin of Marine Science. 44:65-77.

Hesslein, R. H., K. A. Hallard, and P. Ramlal. 1993. Replacement of sulfur, carbon, and
nitrogen in tissue of growing broad Whiteﬁsh (Coregonus nasus) in response to a
change in diet traced by 5348, 813 C, and SUN. Can. J. Fish. Aquat. Sci. 50:2071-
2076.

Hilderbrand, G. V., T. A. Hanley, C. T. Robbins, and C. C. Schwartz. 1999. Role of
brown bears (Ursus arctos) in the ﬂow of marine nitrogen into a terrestrial
ecosystem. Oecologia. 121 1546-550.

Hobson, K. A. and R. G. Clark. 1992. Assessing avian diets using stable isotopes I:
turnover of 13 C in tissues. The Condor. 94: 1 81-188.

Hurlbert, SH, 1984. Pseudoreplication and the design of ecological ﬁeld experiments.
Ecological Monographs 54: 1 87-21 1 .

Johnston, N. T., J. S. Macdonald, K. J. Hall, and P. J. Tschaplinski. 1997. A preliminary
study of the role of sockeye salmon (Oncorhynchus nerka) carcasses as carbon
and nitrogen sources for benthic insects and ﬁshes in the “early Stuart” stock
spawning streams, 1050 km from the ocean. British Columbia Ministry of
Environment, Lands and Parks, Fisheries Project Report No. RD55, 24 pp.,
Victoria.

63

Kennicutt II, M. C., R. R. Bidigare, S. A. Macko, and W. K. Keeney-Kennicutt. 1992.
The stable isotopic composition of photosynthetic pigments and related
biochemicals. Chem. Geol. 101 :235-245.

Kline, T. C., J. J. Goering, et al. 1990. Recycling of elements transported upstream by

runs of Paciﬁc salmon: 1. SUN and 513C evidence in Sashin Creek, southeastern
Alaska (USA). Can J. Fish. Aquat. Sci. 47:136-144.

Kline, T. C., J. J. Goering, O. A. Mathisen and P. H. Poe. 1993. Recycling of elements
transported upstream by runs of Paciﬁc salmon: 11. 515N and 513 C evidence in the

Kvichak River watershed, Bristol Bay, southwestern Alaska. Can. J. Fish. Aquat.
Sci. 50:2350—2365.

Laws, E. A., B. N. Popp, R. R. Bidigare, M. C. Kennicutt, and S. A. Macko. 1995.
Dependence of phytoplankton carbon isotopic composition on growth-rate and
C02: Theoretical considerations and experimental results. Geochem. Cosmochim.
Ac. 59:1131-1138.

Levy, S. 1997. Paciﬁc salmon bring it all back home. Bioscience. 47:657-660.

MacAvoy, S. E., S. A. Macko, S. P. McIninch, and G. C. Garman. 2000. Marine nutrient
contributions to freshwater apex predators. Oecologia. 122:568-5 73.

Macko, S. A., M. L. Fogel, P. E. Hare, and T. C. Hoering. 1987. Isotopic fractionation of
nitrogen and carbon in the synthesis of amino acids by microorganisms. Chemical
Geology. 65:79-92.

McLamey, W. O. 1967. Intr-stream movement, feeding habits, and population of the
coastrange sculpin, Cottus aleuticus, in relation to eggs of the pink salmon,
Oncorhynchus gorbuscha, in Alaska. PhD. Dissertation. University of Michigan.

Mathisen, O. A., P. L. Parker, J. J. Goering, T. C. Kline, P. H. Poe, and R. S. Scalan.
1988. Recycling of marine elements transported into freshwater systems by

anadromous salmon. Verh. Internat. Verein. Limnol. 23:2249-2258.

McConnaughey, T. and C. P. McRoy. 1979. 13C label identiﬁes eelgrass (Zostera
marina) carbon in an Alaskan estuarine food web. Mar. Biol. 53:263-269.

64

Michael, J. H. 1995. Enhancement effects of spawning pink salmon on stream rearing

juvenile coho salmon: managing one resource to beneﬁt another. Northwest
Science. 69:228-233.

Milliman, J. D. and R. H. Meade. 1983. World-wide delivery of river sediment to the
oceans. J. Geol. 91:1-21.

Minagawa, M. and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further

evidence and the relation between 815N and animal age. Geochimica et
Cosmochimica Acta. 48:1 135-1 140.

Minakawa, N. and R. I. Gara. 1999. Ecological effects of a chum salmon (Oncorhynchus
keta) spawning run in a small stream of the Paciﬁc Northwest. J. Freshwater Biol.
14:327-335.

Mitchell, N. L. 2002. Effects of spawning salmon on nutrient dynamics and epilithon in
Southeast Alaska streams. MS Thesis, University of Notre Dame, Notre Dame,
IN.

Newbold, J. D., J. W. Elwood, R. V. O’Neill, and A. L. Sheldon. 1983. Phosphorus
dynamics in a woodland stream ecosystem: A study of nutrient spiraling.
Ecology. 64:1249-1265.

Ostrom, P. H., M. Colunga-Garcia, and S. H. Gage. 1997. Establishing pathways of
energy ﬂow for insect predators using stable isotope ratios: ﬁeld and laboratory
evidence. Oecologia 109: 108-1 13.

Peterson, B. J. and B. Fry. 1987. Stable isotopes in ecosystem studies. Ann. Rev. Ecol.
Syst. 18:293-320.

Phillips, D. L. and J. W. Gregg. 2001. Uncertainity in source partitioning using stable
isotopes. Oecologia. 127: 1 71-179.

Rader, R. B. 1997. A functional classiﬁcation of the drift: traits that inﬂuence
invertebrate availability of salmonids. Can. J. Fish. Aquat. Sci. 54:1211-1234.

65

Reed, R. 1967. Observation of ﬁshes associated with spawning salmon. Trans. Am. Fish.
Soc. 96:62-67.

Richey, J. E., M. A. Perkins, and C. R. Goldman. Effects of Kokanee salmon
(Oncorhynchus nerka) decomposition on the ecology of a subalpine stream. J.
Fish. Res. Board Can. 32:817-820.

Sterling, M. S., K. 1. Ashley, and A. B. Bautista. 2000. Slow-release fertilizer for

rehabilitating oligotrophic streams: a physical characterization. Water Qual. Res.
J. Canada. 35:73-94.

Szepanski, M. M., M. Ben-David, and V. Van Ballenberghe. 1999. Assement of
anadromous salmon resources in the diet of the Alexander Archipelago wolf using
stable isotope analysis. Oecologia. 120:327-335.

Tieszen, L. L., T. W. Boutton, K. G. Tesdhal, and N. A. Slade. 1983. Fractionation and

turnover of stable isotopes in animal tissues: implications for 813 C analyses of
diet. Oecologica. 57:32-37.

Wipﬂi, M. S., J. Hudson, and J. Caouette. 1998. Inﬂuence of salmon carcasses on stream
productivity: response of bioﬁlm and benthic macroinvertebrates in southeastern
Alaska, USA. Can. J. Fish. Aquat. Sci. 55:1503-1511.

Wipﬂi, M. S., J. P. Hudson, D. T. Chaloner, and J. P. Caouette. 1999. Inﬂuence of
salmon spawner densities on stream productivity in Southeast Alaska. Can. J.
Fish. Aquat. Sci. 56:1600—1611

66

 

M3