.5. 31.4352». 3:

€ . .1 . . I.

. , - 3% w.

? . a...» a $2..

.5 , fawn .. . um?!
V .2 . . .

fin“. .A
z .i ..
3“. mt?

. ‘ ‘ x
.5“ «T
x. A

a

:3 i. .
....d.uvu._t.
:3...

n; 49..”
.1" aw

2 2:15: a it

.. 5
may

A
i,
. r:

.35.. £15m p.77
1 a“ c.
:2

.f ‘. .
,... 3:”. .n.
- NJ!
.. 3,9,...
”HR. 5..

 

3.71.:
L: ‘

, . .1.

aw.
5.... 2..

49

.. . 2
him. X

. . ~

3.x..xiL.
{‘1: 1...
13.? . .z

 

a; ‘

 

 

\' , ‘ E.:m£.\.wmd.wmn.fln§,%

 

This is to certify that the

thesis entitled

A SEASONAL INVESTIGATION OF ENERGY TRANSFER
AND FOOD WEB STRUCTURE OF THE DEEPWATER
SCULPIN (MYOXOCEPl-IALUS THOMPSONI) IN
GRAND TRAVERSE BAY, LAKE MICHIGAN

presented by
Colleen Frances Masterson

has been accepted towards fulfillment
of the requirements for

M.S. degree in W81

Geosciences

 

W6 at»

Major professor

Date EIHIOI

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY
Michigan State
University

 

 

 

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
Nov 1 2 2008

n “i 11 L‘ A
U j H U E

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDue.p65-p.15

 

 

A SEASONAL INVESTIGATION OF ENERGY TRANSFER AND FOOD WEB
STRUCTURE OF THE DEEPWATER SCULPIN (MYOXOCEPHAL US THOWSONI)
IN GRAND TRAVERSE BAY, LAKE MICHIGAN
By

Colleen Frances Masterson

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Department of Geological Sciences

2001

ABSTRACT
A SEASONAL INVESTIGATION OF ENERGY TRANSFER AND FOOD WEB
STRUCTURE OF THE DEEPWATER SCULPIN (MYOXOCEPHAL US THOMPSON!)
IN GRAND TRAVERSE BAY, LAKE MICHIGAN
By

Colleen Frances Masterson

This study attempts to elucidate the diet of the deepwater sculpin in Grand
Traverse Bay, Lake Michigan using a combination of stomach content and stable isotope
techniques.

Stomach contents revealed that the diet of the deepwater sculpin was composed
primarily of the amphipod, Diporeia hoyi. Other major prey items were mysids (Mysis
relicta) and chironomid larvae. The abundance of minor prey items in the deepwater
sculpin stomachs varied seasonally, and isotope data implied that these minor prey items
may have been more important than stomach content data suggested.

The importance of long-term seasonal sampling in food web studies was clearly
supported. Seasonal variation in carbon and nitrogen isotope values was observed in the
food web members of Grand Traverse Bay, especially at lower trophic levels. An
isotopic mixing model was used to explore the relative importance of seston, sinking
POM and sediments to the deepwater sculpin. The model results suggested that seston
was an important primary organic SOurce for the benthic food web and that atmospheric
deposition may be a significant source of contaminants to the deepwater sculpin in Grand
Traverse Bay. Difficulties in the application of isotopic mixing models in food web

studies are discussed and recommendations for future research are made.

To those who have kept me strong
and encouraged me to persist.

iii

ACKNOWLEDGMENTS

I would first like to thank my committee members, Nathaniel Ostrom, Peggy
Ostrom, Don Hall and Tom Coon for their support and thoughtful comments throughout
the progress of my research. I am gratefirl to John Skubinna for collecting the stomach
content data included in this report, and equally important, for his guidance in the early
days of my grad school career at MSU. This work would not have been possible without
the support of Michigan SeaGrant and the US EPA or the hard work of the many folks
involved in the field sampling in Grand Traverse Bay.

I would like to thank my fellow graduate students for their advice and friendship
and for providing me with a great amount of amusement in the last couple of years.
Eileen McCusker and Amy Macrellis introduced me to the lab and showed me the ropes
when I was the new kid in town. Mary Russ was a good sounding. board in times of
frustration and always had something quirky to make me smile. I would like to extend a
special thank you for the remaining 2 of the ‘3 musketeers’, Tucker McNulty and Andery
Calkins. We have had some grand times together, and my invitation to you both is to
continue to build and enjoy these friendships regardless of how many miles are
separating us. I would also like to thank Chris Harvey for his thoughtful advice and for
the many emails that had me laughing out loud.

I would like to thank my boyfriend and best friend, Chris Bzdok, for the humour
and love he has shared with me before, during and after my time at MSU. I can’t imagine
finishing this work without you by my side. You deserve part of this degree for keeping

me strong, keeping me sane and keeping me smiling. I think you’re spectacular.

iv

I would like to thank my Mom and Dad for their never-ending support. I want to
thank you for leading me to believe that I could become anything I wanted to be, and for
encouraging me in every route I’ve chosen to pursue. I’d like to thank my Sister, Karen,
for lifting my spirits when I needed it most, and for being the one who always
understands me. I’m lucky to have a best friend and a sister in the same amazing woman.
I’d also like to thank my brothers, Tim and Pat for just being downright awesome. My
family have made me who I am today and I know that with them behind me, I can do
anything.

And finally, thanks to my lil’ Emy Mahina for all the love.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... vii
LIST OF FIGURES ......................................................................... viii
INTRODUCTION ........................................................................... 1
MATERIALS AND METHODS .......................................................... 6
Study Location and Sample Acquisition ......................................... 6
Stomach Content Analysis ......................................................... 7
Stable Isotope Analysis ............................................................ 8
Estimates of Trophic Fractionation ............................................... 10
Estimates of the Relative Contribution of Primary Organic
Sources (POS) to Prey Items .................................................... 11
RESULTS ..................................................................................... 13
Stomach Content Analysis ......................................................... 13
Stable Isotope Analysis ............................................................. 15
Seasonal Variation in TN and 5'3C of Deepwater Sculpin
and Average Diet .................................................................. 16
Estimates of the Relative Contribution of POS to Prey Items ................. 18
DISCUSSION ................................................................................. 24
Stomach Content Analysis ......................................................... 24
Seasonal Variation in 8'5N and 8‘3C of Deepwater Sculpin
and Average Diet .................................................................. 25
Estimates of the Relative Contribution of POS to Prey Items ................ 31
Sources of Error and Recommendations for Future Research ................ 37
SUMMARY AND CONCLUSIONS ...................................................... 40
TABLES ....................................................................................... 41
FIGURES ..................................................................................... 49
LITERATURE CITED ...................................................................... 59

vi

Table 1.

Table 2.

Table 3.

Table 4.

LIST OF TABLES

Summary of Grand Traverse Bay deepwater sculpin stomach
contents, expressed as percent fiequency of occurrence
(% F 0), percent prey number (% PN), percent prey mass

(% PM) and percent index of relative importance (% IRI)

Nitrogen and carbon isotope values of food web members
collected from Grand Traverse Bay, Lake Michigan.
Isotope values shown are averages for each month for the
years 1997 and 1998 combined. Values are reported as

averages i 1 SD. ND = no data .....................................

Trophic fractionation factors for S'SN and 613 C between
deepwater sculpin and average sculpin diet (calculated
using 6 month seasonal averages). Trophic fractionation
factors were calculated with stomach content data expressed
as percent prey number (% PN), percent prey mass (% PM)

and percent index of relative importance (% IRI) .................

Results of isotopic mixing model attempts, predicting the
relative importance of seston, sinking POM and
sediments as sources of nitrogen and carbon to D. hoyi
and M. relicta. “ns” indicates that there was no solution

to the mixing model ....................................................

Vii

42

.44

.47

.. 48

LIST OF FIGURES

Figure 1. Location of the 2 study sites GT1 and GT3 in Grand Traverse
Bay, Lake Michigan (http://www.glerl.noaa.gov) .................. 50

Figure 2. Nitrogen and carbon isotope values of food web members
of Grand Traverse Bay, Lake Michigan in 1997 and 1998.
Values shown are seasonal averages. Squares represent
the isotope values of the deepwater sculpin, circles
represent the isotope values of the deepwater sculpin prey
items and triangles represent the isotope values of the three
POS ....................................................................... 5 1

Figure 3. DIN (A) and 813C (B) distributions of deepwater sculpin
as a function of length in Grand Traverse Bay, 1997-1998 ....... 52

Figure 4. Monthly nitrogen isotope values for deepwater sculpin an
average deepwater sculpin diet. Average diet was calculated
based on stomach contents expressed as percent prey number
(% PN), percent prey mass (% PM) and percent prey
Index of relative importance (% IRI). 8'5N values for
deepwater sculpin are shown as averages :I: 1 SE .................. 53

Figure 5. Monthly carbon isotope values for deepwater sculpin and
average deepwater sculpin diet. Average diet was calculated
based on stomach contents expressed as percent prey number
(% PN), percent prey mass (% PM) and percent prey
index of relative importance (% IRI). 815N values for
deepwater sculpin are shown as averages i 1 SE .................. 54

Figure 6. Graphical representation of the mixing model (Equations 4-6)
for Diporeia hoyt' and the three POS. Average D. hoyt' diet
was calculated by correcting the 813 C and BISN values by
trophic fractionation (0.5 0/00 for 513C and 3.4 %o for
815N). D. hoyi was assumed to be 1 trophic level above the
POS. Grey values in parentheses indicate relative
proportions based on the mixing model .............................. 55

Figure 7. Monthly nitrogen isotope values of D. hayi, seston and

sinking POM in Grand Traverse Bay, Lake Michigan,
1997 ....................................................................... 56

viii

Figure 8. Graphical representation of the mixing model (Equations 4-6)
for Mysis relicta and the three POS. Average M. relicta
diet was calculated by correcting the 813 C and SISN
values by trophic fractionation (0.5 0/00 for BBC and 3.4 %o
for 81’N). M. relicta was assumed to be 1.5 trophic levels
above the POS. Grey values in parentheses indicate relative
proportions based on the mixing model .............................. 57

Figure 9. Relative importance of the three POS to deepwater sculpin
in Grand Traverse Bay, 1997-1998 according to the isotopic
mixing model (Equations 4-7). Plots shown represent results
based on stomach content data expressed as prey number
(PN), prey mass (PM) and prey index of relative importance
(IRI) ....................................................................... 58

INTRODUCTION

In this study we present a combination of stomach content and stable isotope
analyses to elucidate the food web structure of a benthic forage fish, the deepwater
sculpin (Myoxocephalus thompsom'), in Grand Traverse Bay, Lake Michigan.
Understanding the food web structure of Grand Traverse Bay at the level of forage fishes
may provide future insight into the relative importance of different exposure routes of
organic contaminants to higher level consumers of great commercial and recreational
importance in the Great Lakes.

There are three hypothesized exposure pathways by which organic contaminants
enter the food web of Great Lakes fishes: (l) atmospheric deposition transferred through
the pelagic food web; (2) atmospheric deposition transferred, via rapidly-settling particles
through the benthic food web, and; (3) transfer from historically-contaminated, in place
sediments through the benthic food web (Baker et al. 1996). The primary organic sources
(POS) associated with each of these exposure routes are: seston for route (1); sinking
particulate organic matter (POM) for route (2); and sediments for route (3). A thorough
analysis of food web structure provides estimates of the relative importance of the three
POS to upper level consumers. The combination of these data with knowledge of
contaminant levels in the POS will allow predictions to be made about the relative
importance of the 3 exposure routes of organic contaminants to upper level consumers.

Stomach content and stable isotope analyses supply distinct, yet complimentary
information in food web investigations. While stomach content data provide a direct

assessment of ingested material, there are inherent difficulties in identifying partially

digested food items, and there is a bias towards the identification of food types with slow
digestion rates (Gu et al 1996, Gould et al. 1997b, Jennings et al. 1997). Stomach
content data provide an estimate of the food items ingested, but may be misleading if
- assimilation is not considered. Assimilation may be incomplete, and certain organisms or
parts of organisms may be preferentially digested over others. Stable isotope data
strongly indicate nutritional dependence, as the isotopic composition of an organism is
directly related to the organic matter assimilated from its diet (Tieszen et al. 1983,
Rosenfeld et al. 1992, Hesslein et al. 1993, Ostrom et al. 1996). In addition, stomach
contents provide only a snapshot of recently ingested material, while the isotopic
composition of a consumer is a reflection of its food resources assimilated over a longer
period of time from weeks to months (Harrigan et al. 1989, Hobson and Welch 1992,
Gould et al. 1997a,b). The combination of stomach content and stable isotope analyses in
food web research therefore offers a unique approach to assess trophic relationships and
food web structure.

The stable isotope technique uses natural differences in the 13Cz'ZC and 15N:"’N
ratios of organisms (conventionally expressed as 513C and 8'5N 1) to identify food

sources of carbon and nitrogen which move with predictable isotopic alterations fiom one

 

1 Stable isotope ratios are reported as 813 C and 8'5N (%o), according to:
8'12 = [(mmp../Rmm>-1] " 1000
where I is the heavy isotope of element E, either carbon or nitrogen, and R is the
abundance ratio of the heavy to light isotope. Internationally recognized standards are

Vienna Pee Dee Belemnite for carbon, and atmospheric N2 for nitrogen.

trophic level to the next (Goering et al. 1990). Trophic fractionation is a term that
describes the difference in 813C values or 8‘5N values between a consumer and its food
that is due to discrimination against the heavy isotope during metabolism and excretion
(Peterson and Fry 1987). Generally, 613C values change by 0-1 0/00 per trophic level
(e.g. DeNiro and Epstein 1978, Fry and Parker 1979, Peterson and Fry 1987, Harrigan et
al. 1989, Goering et al. 1990, Keough et al. 1996). Such small increases in 613C values
with trophic level, relative to individual variations and the precision of analytical
techniques, can make the determination of trophic position based on SEC values difficult.
Carbon isotopes are more often used to identify sources of organic carbon and to trace the
flow of these carbon sources through an ecosystem (Fry and Scherr 1984, Goering et al.
1990, Gearing 1991). In contrast to 813C values, greater increases in S'SN values of
3-4 0/00 generally take place between each trophic level (e.g. DeNiro and Epstein 1981,
Minagawa and Wada 1984, Fry 1988, Goering et al. 1990, Keough et al. 1996). For this
reason, S'SN values are often used as more robust indicators than 613C values in
determining trophic positions of organisms. With reliable estimates of trophic
fractionation of both SEC and 8'5N values, it is possible to trace each step of the food
web back to the primary sources of organic matter. The use of both carbon and nitrogen
isotope ratios can be instrumental in resolving complex food web relationships that
depend on multiple sources of organic matter (Creach et al. 1997).

While many food web studies have employed the use of stable isotopes and
stomach content analyses, relatively few of these studies have investigated seasonal
variation in food web structure (Gearing et al. 1984, Goering et al. 1990, Toda and Wada

1990, Leggett 1998, Neilson et al. 1998). Factors such as the seasonal availability of

nutrients, succession of plankton communities and changes in the spatial distribution of
aquatic organisms may greatly affect temporal food web dynamics. Temporal change in
trophic interactions is still one of the least understood factors affecting the
biogeochemical material flow in aquatic ecosystems (Y oshioka et al. 1994). Therefore,
efforts to obtain seasonal stomach content and isotope data will make important
contributions to the understanding of food web dynamics.

The present study not only combines stomach content and stable isotope
techniques to investigate the food web dynamics of the deepwater sculpin, but also
probes the nature of temporal variation in food web structure. Deepwater sculpin are
abundant benthic-dwelling secondary consumers in the benthic food web of the upper
Great Lakes, including Grand Traverse Bay. Based on previous stomach content studies,
they are thought to utilise benthic food resources nearly exclusively as adults, feeding
mainly on amphipods, and to a lesser extent on mysids (Kraft and Kitchell 1986, Wojcik
et al. 1986, Selgeby 1988). They are also known to feed to a varying extent on other
seasonally important food resources, such as chironomids, fish eggs, larval fish and
terrestrial insects (Wojcik et al. 1986). Deepwater sculpin are important prey for higher
level consumers in Lake Michigan, including burbot (Lota Iota), lake trout (Salvelinus
namaycush), steelhead (Oncorhynchus mykiss), coho salmon (Oncorhynchus kisutch) and
other salmonines (W ojcik et al. 1986).

This study attempts to elucidate the diet of the deepwater sculpin in Grand
Traverse Bay, Lake Michigan, using a combination of stable isotope and stomach content
techniques, and to assess seasonal variation in the isotopic composition and diet of food

web members. By improving our understanding of nutrient transfers and trophic

interactions, food web studies in the Great Lakes may provide valuable insight into the

bioaccumulation of organic contaminants in fish.

MATERIALS AND METHODS

M Location and Sample Acquisition

All samples were collected on approximately a monthly basis from April to
September, 1997 and 1998 in the western arm of Grand Traverse Bay, Lake Michigan
(Figure 1). In 1997, samples were collected from study sites GT1 and GT3, while in
1998, the sampling effort focused only on site GT3. Site GT1 has a depth of 98 m and
site GT3 has a depth of 112 m.

Deepwater sculpin were collected with a 4.9 m otter trawl with 33 mm body mesh
and 5 mm codend mesh. Tows of 10-20 minutes were made near sites GT1 and GT3 at
depths greater than 80 m. Collected fish were sorted for stable isotope analysis or
stomach content analysis, and the weight and total length of each individual was
recorded. Fish intended for stable isotope analysis were frozen immediately after
collection, while fish intended for stomach content analysis were preserved in formalin
(10% formaldehyde solution).

Zooplankton were collected using daytime oblique tows from 80m to the surface
using a 1m diameter plankton net with 505 um mesh. Bulk zooplankton samples were
filtered through pre-combusted (500°C, 1 hr) GF/F glass fibre filters (Whatman) and
frozen for isotope analysis.

A benthic sled was employed to collect macroinvertebrates such as amphipods,
mysids, oligochaetes and chironomids (design modified fi'om Nesler 1981). The sled was
equipped with a 90 cm x 60 cm rectangular net with 750 um mesh, and was towed along

the bottom near sites GT1 and GT3 for 10-20 minutes per sample. Samples were sieved

through 595, 833, 1400 and 2000 um mesh screens, and taxa were hand sorted and frozen
for stable isotope analysis.

Seston samples were obtained by filtering 3-5 L of water from each of five depth
intervals at low pressure through pre-combusted (500°C, 1 hr) Whatman GF/F glass fiber
filters (McCusker et al. 2000). Sediments were collected using a box core or ponar grab.
Sinking POM samples were collected in biweekly intervals by multi-sequencing sediment
traps (8 in. OD, 8:1 aspect ratio; Eadie et al. 1984), which remained at the sampling site

throughout the season.

Stomach Content AnalLsLs

John Skubinna (Ph.D. candidate, Michigan State University Department of
Fisheries and Wildlife) completed the identification of all deepwater sculpin stomach
contents. Prey taxa were identified using digestion resistant hard parts (Balcer et al.
1984, Merritt and Cummins 1996). OptimasTM image analysis software was used to
measure prey length or key morphological features of each prey item obtained. Care was
taken in avoiding duplicate counts of the same individual by choosing morphological
features that are singular in the morphology of the prey (e.g. body length, head width,
etc.) when possible. The linear measurements of length or hard part size were converted
to estimates of dry weight biomass of whole individuals using length-weight or hard part
size-weight relationships (e.g. Dumont et al. 1975, Smock 1980, Sell 1982, Culver et al.
1985). Specific conversion relationships used are described in detail in the dissertation of

John Skubinna (in prep).

To assess the importance of each prey item in the deepwater sculpin diet, four
indices of prey importance were calculated: (1) percent frequency of occurrence (% PC),
the number of stomachs containing a single type of prey divided by the total number of
stomachs containing organic matter; (2) percent of prey number (% PN), the number of
items of a single prey taxon divided by the total number of all items in the sample; (3)
percent of prey mass (% PM), the mass of a single prey taxon divided by the total mass of
all material in the sample and; (4) percent prey index of relative importance (% IRI;
Pinkas et al. 1971), the IRI of the food group divided by the IRI of the total diet, where
IRI is calculated as:

IRI=%FO (% PN +%PM) (eq. 1)

Monthly values of prey importance were obtained by pooling all prey items found
in deepwater sculpin stomachs from that particular month. The seasonal mean was
calculated for each index of prey importance, with each month weighted equally to
account for unequal sample sizes and seasonal variation in prey importance. Each index
of prey importance provides a unique perspective of diet with a different bias. Percent
IRI incorporates the three traditional indices of prey importance, and is thought to
minimize the biases of each, obtaining a more representative description of diet (Pinkas et

al. 1971).

t I no] si
To provide a manageable sample, a subsample of each fish (~10 %) was obtained
in preparation for isotope analysis. The isotopic signature of a subsample of this size was

found to differ from that of whole fish by 0.07 :t 0.42 °/oo for 8N and 0.19 i 0.39 %o for

613C (n=5). The subsample was freeze dried, ground and lipid extracted for six hours in a
soxhlet apparatus containing an azeotropic mixture of chloroform and methanol
(87:13 v/v). After lipid extraction, the tissue was dried and homogenized with a Wig-L-
Bug (Crescent Industries) mechanical mill. Bulk zooplankton and macroinvertebrate
samples (bulk samples of many individuals) were prepared in the same manner as fish,
with an additional step of acidification using 10 % HCl to remove carbonates prior to
lipid extraction. Fish, zooplankton and macroinvertebrate samples were weighed (3-4 mg
for 515N values; 04-05 mg for 813C values) into 6 mm x 4 mm ultra-light weight tin
capsules (Elemental Microanalysis Ltd.), and analysed for isotopic abundances using a
Carlo Erba NA 1500 nitrogen/carbon analyzer interfaced to a Micromass Prism mass
spectrometer (Wong et al. 1992).

In preparation for stable isotope analysis, filters containing seston samples were
dried (40°C), acidified with 10 % HCl to remove carbonates, and dried again. The seston
sample was obtained by removing the surface layer of the filter containing the seston
(McCusker et al. 2000). Sediments and sinking POM were freeze dried, acidified with
10 % HCl to remove carbonates, and dried again (40°C). Seston, sediments and sinking
POM samples were placed in precombusted (500°C, 1hr) quartz tubes with excess
precombusted CuO and Cu (approximately 3 g of each). The tubes were evacuated,
sealed, and combusted at 850°C. The combustion products were separated cryogenically
on a vacuum line and the isotopic composition of the purified carbon dioxide and
nitrogen gas was determined on a Micromass Prism mass spectrometer.

Additional sediment trap samples are available for nitrogen and carbon isotope

analysis, and modifications to the laboratory protocol (acidification with more

concentrated HCl and sample shaking to more thoroughly remove carbonates) may
further improve the accuracy of the results. However, the results presented here are

thought to be a representative subset of the available data.

Estimates of T rophic Fractionation

In diet analyses, mass balance equations can be used to estimate the relative
contribution of different sources of carbon or nitrogen to a consumer, or if this is known,
to estimate trophic fractionation. Stomach content analysis provided the relative
importance of different prey items to the deepwater sculpin (based on the three indices of
prey importance), and the average isotopic composition of the deepwater sculpin diet for
each month was calculated as:

Slstculpin diet = 2fi515Ni and 613Csculpin diet = 2fi513Ci
(eq- 2)

I= i=1

where f is the fractional contribution of individual prey items to the deepwater sculpin
(based on each index of prey importance) and n is the number of different prey items.
When the isotopic composition of a prey item was not known for a particular month,
isotope values were assumed to be equal to the nearest month with available data.
Trophic fractionation factors for both SUN values (TR) and 613 C values (TFc) were then

quantified by:

TFn = 81SFNsculpin ‘ 81Slqsculpin diet and We = 513Csculpin ' 813Csculpin diet (“1.3)

10

Estimates of the Relative Contribution ofPrimaLv Organic Sources QOSZ to Prev Items
To estimate the relative contribution of each POS to individual deepwater sculpin

prey items, an isotopic mixing model of the following form (Harrigan et al. 1989) was

used:
5‘5Nmy - rs, = aa‘st, + ba‘stmmM + c5“N,,,im,..., (eq. 4)
5%,”, - Tr, = 3813me + b6‘3c,,,ki..,m + c5‘3c,,dm.m, (eq. 5)
= a + b + c (eq. 6)

where the terms a, b and 0 represent the percent contribution for nitrogen or carbon of
each of three prey items, and TFn and TFc represent the difference in 515N or 813C
values, respectively, between the consumer and its diet (trophic fiactionation factors).
This model requires that the three POS are isotopically distinct, which has been verified
in previous studies (Takahashi et al. 1990, Yoshioka et al. 1994, Zohary et al. 1994,
Ostrom et al. 1997).

Due to the lack of stomach content data from prey items of deepwater sculpin in
this study, species-specific trophic fractionation factors for the mixing model were
' unavailable. Because trophic fractionation factors have been found to vary widely among
taxa in previous studies, the mixing model was explored using both the trophic
fractionation factors calculated in this study (between deepwater sculpin and sculpin
diet), as well as trophic fi'actionation factors taken from the literature. The trophic
fractionation factor adopted fi'om the literature for 815N was 3.4 %o per trophic level, the
widely cited average of Minagawa and Wada (1984). The trophic fractionation factor
adopted from the literature for 513C values was 0.5 %0 per trophic level, the midpoint of

the commonly reported range of 0-1 %0 (Peterson and Fry 1987).

ll

Seasonal (6-month) averages of 6'5N and 813C values of deepwater sculpin prey
items and POS were used in the mixing model to develop the best estimate of the long-
term relative contributions of POS to the nutrition of the prey items. Because of the large
seasonal variations observed in the isotopic compositions of deepwater sculpin prey
items, each month was weighted equally in the seasonal mean to prevent a
disproportionate contribution from months with larger sample sizes.

The amount of nitrogen or carbon assimilated fiom each POS by the deepwater

sculpin was then determined by the relationship:

Vs=i§lfin (eq. 7)

where Vs is the percent contribution of a POS to the deepwater sculpin, f is the fractional
contribution of the prey item to the deepwater sculpin, Vp is the fractional contribution of
the POS to a prey item and n is the number of prey items. These techniques were used in
an attempt to reach our ultimate goal of determining the relative importance of the three

primary organic sources to higher trophic levels.

12

RESULTS

Stomach Content Analgsis

Stomach content data were available fiom deepwater sculpin collected in April of
1997 and 1998, May, July, August and September of 1997 and June of 1998. An inter-
annual comparison of deepwater sculpin stomach contents between April 1997 and April
1998 (John Skubinna, Department of Fisheries and Wildlife), found that the two years did
not differ significantly in the distribution of biomass among the different prey items. A
similar comparison using % IRI data found the relative proportions of prey items to differ
by less than 1.5 % between the two years. Stomach content data fiom 1997 and 1998
were therefore combined to provide a full monthly data set from April through
September.

For each month from April to September, sculpin stomach contents were
summarized according to the 4 indices of prey importance, that included one measure of
the fi'equency of occurrence of prey items in stomachs (% F O) and 3 measures of the
relative abundance of prey items in stomachs (% PN, % PM and % IRI; Table 1). For
each index of prey importance, the mean diet was calculated by taking the average of the
indices for the six months.

Stomach content data indicated that the amphipod Diporeia hoyi was consistently
the most abundant prey item in the deepwater sculpin diet, regardless of the month
sampled (Table 1). On the basis of % FO, D. hoyi were present in 86.2 — 100.0 % of
sculpin stomachs in each month and in 95.4 % of the stomachs on an annual basis.

D. hoyi comprised 69.9 and 78.8 % of the annual sculpin diet based on % PN and % PM,

13

 

 

respec
imPO

annu

chin

respectively. The index of prey relative importance (% IRI) also substantiated the
importance of D. hoyi in the deepwater sculpin diet, with a prediction of 85.8 % of the
annual diet.

Estimates of % F 0 indicated that opossum shrimp (Mysis relicta), fish eggs and
chironomid larvae were each present in over 38 % of deepwater sculpin stomachs
annually, and estimates of % PN, % PM and % IRI indicated that each of these prey
items comprised > 3 % of the annual sculpin diet. Minor prey items found in the
stomachs of deepwater sculpin included chironomid pupae and adults, unidentified
dipteran adults, annelids, microcrustacean zooplankton, aquatic and terrestrial beetles,
and crayfish. Each of these minor prey items were found in fewer than 25 % of
deepwater sculpin stomachs annually, and each comprised < 3 % (range 3.2-14.1 %) of
the annual sculpin diet as estimated by % PN, % PM and % IRI.

While seasonal variation in the relative importance of most prey items to the
deepwater sculpin over our six month sampling period was minor, notable variation was
observed in the relative importance of some of the less abundant prey items in the
deepwater sculpin stomachs. Fish eggs were most abundant in the sculpin stomachs
during April and May (% FO of 89.7 and 100.0 %, % PN of 42.1 and 17.0 %, % PM of
10.4 and 3.5 %, and % IRI of 29.0 and 10.2 % for April and May, respectively), and
declined in importance throughout the rest of the sampling season (% FO of 17.0 %,
% PN of 3.0 %, % PM of 1.1 % and % IRI of 0.5 % by September). Beetles increased in
importance in the sculpin diet later in the season, and peaked in abundance in July and
September (% PC of 4.8 and 4.0 %, % PN of 0.2 and 0.0 %, % PM, of 6.6 and 0.6 % and

% IRI of 0.1 and 0.0 % for July and September, respectively).

14

The salient features of the deepwater sculpin stomach content data were reviewed
here solely for the purpose of comparing indices of prey importance and as an
introduction for further sections of this thesis combining stomach content and stable
isotope analyses. A more extensive investigation of this stomach content analysis can be

found in the Ph.D. dissertation of John Skubinna, Department of Fisheries and Wildlife

(in prep.).

Stable Isotope Anglysis

Isotopic compositions of food web members generally differed by less than 1 %o
between 1997 and 1998 when monthly averages were compared, and yearly averages
differed by less than 0.5 %o. For this reason, data fi'om the two years were pooled to
include both sampling seasons and increase sample sizes. One exception was the isotopic
composition of seston, which showed a dramatic enrichment in 815N values in 1998, with
5‘5N values over 6 0/oo higher than those observed in 1997 (Table 2, Figure 2). Seston
values from 1997 and 1998 are reported separately in Table 2 and Figure 2. Thus, while
average isotope values of the two years combined were adopted for all other food web
members, those analyses involving the isotopic composition of seston were repeated with
1997 seston values, 1998 seston values and the mean value of the two years combined.

No significant relationships were found between length of deepwater sculpin and
8N values (R2 = 0.0002, Figure 3A) or 5% (R2 = 0.0029, Figure 38). Consequently,
all sizes and ages of deepwater sculpin were combined for the remainder of the analyses

and discussion.

15

Nitrogen isotope values generally increased with trophic level. For example,
deepwater sculpin were enriched in 15N relative to their major prey items, which were in
turn enriched in 15N relative to the three POS (Table 2, Figure 2). Trends in carbon
isotope values were not as clear. Whereas the deepwater sculpin were enriched in '3 C
relative to D. hoyi and M. relicta, they were depleted in 13C relative to all other prey
items (Table 2, Figure 2). D. hoyi and M. relicta were enriched in 13C relative to only
one of the POS (seston), while the remaining deepwater sculpin prey items were enriched

in 13C relative to all three POS (Table 2, Figure 2).

§e_asonal Variation in 51 5 N and 513C ofDeepwater Sculpin and Average Diet

A unique feature of this study was the emphasis on seasonal trends in isotopic
compositions. Seasonal variations were observed among lower trophic level food web
members, with ranges of up to 8 0/oo in 815N values and 4 %o in 613C values (Table 2)
between April and September. Much smaller ranges in 615N and 613C values were
observed for the deepwater sculpin (0.6 %0 range in 8'5N values and 0.4 %0 range in 813C
values; Table 2).

In contrast to the small range of BISN values of deepwater sculpin from April to
September, the average deepwater sculpin diet exhibited considerable variation in 8’5N
values over the sampling season, with an observed range of approximately 4 %o (Figure
4). The 8‘5N value of the average diet decreased from April through June, peaked briefly
in July, decreased to August and began to increase slightly in September (Figure 4).

While estimates of the SISN value of the average diet derived from mass balance

16

equations using three different indices of prey importance differ slightly, they all display
a general decline throughout the season (Figure 4).

Monthly estimates of trophic fractionation for 8'5N values were calculated by
subtracting the SN value of the average diet from the 6‘5N value of the deepwater
sculpin. The constant 8'5N values of the deepwater sculpin, combined with the seasonal
decline in the SISN value of their average diet, resulted in a range of apparent trophic
fractionation factors throughout the season. The trophic fractionation factor ranged fi'om
0.9 to 4.4 %0 based on % PN, from 1.9 to 5.2 %0 based on % PM and from 1.5 to 4.9 °/oo
based on % IRI. The average trophic fractionation between SISN values of the deepwater
sculpin and its diet for the six-month sampling period was calculated by taking the mean
trophic fractionation of all sampling months. The average trophic fractionation of 8‘5N
values was determined to be 3.3 %o based on % PN, 4.0 %o based on % PM and 3.9 %o
based on % IRI (Table 3).

The seasonal trends observed in the 813C values of deepwater sculpin and their
diet closely resembled the variation in 8'5N values, although the magnitude of seasonal
variation differed between the two measures. The 613C values of deepwater sculpin
displayed little seasonal variation, with an overall range in 813C values of only 0.4 %o
(Figure 5). The SEC values of the average diet showed a substantially greater amount of
seasonal variation than the deepwater sculpin, with a range of approximately 2 %o
(Figure 5). According to all three indices of prey importance, the 813C values of the
average diet declined from April through August, followed by a slight increase in

September (Figure 5).

l7

Monthly estimates of trophic fractionation for 8‘3 C values were calculated by
subtracting the 813C value of the average diet fiom the 613C value of the deepwater
sculpin. The constant 613 C values of deepwater sculpin combined with the seasonal
variation in the 513C values of the average diet again resulted in a range of apparent
trophic fractionations throughout the season. The trophic fiactionation in 8'3 C values
ranged from 0.2 %o to 2.4 %0 based on % PN, from 1.4 to 2.7 o/oo based on % PM and
fiom 0.7 to 2.7 %0 based on % IRI. The average trophic fractionation of 813C was
calculated to be 1.5 %o based on % PN, 1.9 %0 based on % PM and 2.0 %0 based on
% IRI (Table 3).

Annual estimates of trophic fractionation for both 815N and 813 C were calculated
using linear regressions between months with available data to estimate isotope values of
food web members in months where data were lacking. While seasonal variation
undoubtedly occurs during the winter months, the estimated annual trophic fractionation
factors differed from the calculated 6-month seasonal averages by less than 0.6 %o for

SISN and 0.2 %0 for 813C.

Estimates of the Relative Contribgtion QfPOS to Prey Items

An isotopic mixing model was used to determine the relative contribution of
carbon and nitrogen from each of the three POS to the deepwater sculpin prey items. The
isotopic compositions of the average D. hoyi and M. relicta diets were calculated by
correcting the actual 8N and 813C of D. hoyi and M. relicta for trophic fractionation.
Due to the lack of stomach content data from prey items of deepwater sculpin in this

study, and since trophic fractionation factors have been found to vary widely among taxa

l8

in previous studies, the mixing model was explored using trophic fractionation factors
calculated between deepwater sculpin and sculpin diet (Table 3), as well as trophic
fractionation factors taken from the literature. The adopted literature trophic
fractionation factors were 3.4 %o for 6'5N values (Minagawa and Wada 1984) and 0.5 %o
for 813 C values (midpoint of range reported by Peterson and Fry 1987). In addition,
while yearly differences in the isotope values of most food web members were negligible
and average values of the two years combined were adopted for the mixing models, the
model was repeated with 1997 seston values, 1998 seston values and the mean value of
the two years combined because of the large inter-annual discrepancy in isotopic
composition of this single food web member.

In the application of the isotopic mixing model, it was assumed that D. hoyi were
positioned one trophic level above the POS, since they are generally classified as
detritivores and are thought to rely primarily on organic particles that settle from the
water column to the sediments for their nutrition (Gardner et al. 1985, Quigley 1988,
Gauvin et a1. 1989).. It was assumed that M. relicta were positioned at a trophic level 1.5
above that of the three POS. The trophic level of 1.5 above the three POS was chosen
because bulk M. relicta samples included individuals of varying life stages. M. relicta
are known to exhibit ontogenetic shifts in feeding behaviour, fi'om herbivory as juveniles
to increasing carnivory with maturity (Grossnickle 1982, Nero and Sprules 1986). While
larger size classes tended to have higher isotope values when size classes of M. relicta
were analysed separately in this study, sample sizes were low, and the ability to make
conclusions based on size class effects on isotope values was restricted. All data were

subsequently combined in the remainder of the analyses.

19

Table 4 outlines the results of the mixing model attempts used to determine the
relative contribution of carbon and nitrogen from each of the three POS to D. hoyi and
M. relicta. All mixing model attempts using the trophic fractionation factors calculated
between deepwater sculpin and sculpin diet (from Table 3) failed to yield plausible
solutions (Table 4). Two of the mixing model attempts using literature values of trophic
fractionation resulted in plausible solutions (Table 4). Mixing model attempts using the
estimated annual isotope values of food web members calculated by linear regressions
between months with available data failed to provide any plausible results.

A graphical representation of the isotopic mixing model of D. hoyi and the three
POS (using literature values of trophic fractionation) is shown in Figure 6, in which the
area within each triangle represents isotopic compositions of the D. hoyi diet that would
yield plausible solutions. The isotopic composition of the average D. hoyi diet did not lie
within a triangle created with the isotope values of the three POS based on overall
averages for 1997 and 1998 seston or 1998 seston values, but did provide plausible
solutions based on 1997 seston values alone. The mixing model predicted that in 1997,
73.1 % of the nutrition of D. hoyi (in the form of carbon and nitrogen) was derived from
seston, 19.7 % was derived fiom sediment, and only 7.2 % was derived fiom sinking
POM (Figure 6). Because the mixing model did not provide plausible solutions when
seston isotope values were used from 1998 or fi'om the two years combined, it was not
possible to determine the relative importance of the 3 P08 to D. hoyi in 1998 using this
analysis.

To further examine the prediction of the mixing model that seston was the

primary POS that contributed nitrogen to D. hoyi in 1997, the 6'5N monthly averages of

20

D. hoyi throughout the season were plotted with the SEN values of seston and sinking
POM (Figure 7). The seasonal trend in 615N values of D. hoyi resembled that observed in
the 1997 seston nearly perfectly from April through August. The seasonal trend in the
8'5N values of D. hoyi was less similar to that observed in the sinking POM, particularly
in the month of June (Figure 7). This may be further support that D. hoyi were indeed
relying to a greater extent on seston than sinking POM in this system.

A graphical representation of the isotopic mixing model was also created for
M. relicta and the three POS (Figure 8). While the mixing model for M. relicta was
explored using the same suite of trophic fractionation factors used in the D. hoyi model
(Table 4), the graphical representation again shows only results of the model using
literature values of tr0phic fractionation. It was assumed that M. relicta were positioned
at a trophic level 1.5 above that of the three POS. The isotopic composition of the
average M. relicta diet was located within the triangle formed by the isotope values of the
three POS when overall averages for 1997 and 1998 were used for all components
(Figure 8). The isotopic mixing model predicted that 43.4 % of the nutrition of the
M relicta diet (in the form of carbon and nitrogen) was ultimately derived from seston,
3.7 % was derived fi'om sinking POM, and 52.9 % was derived from sedimentary
sources. The mixing model could not be used to determine the relative importance of the
three POS to M relicta using 1997 or 1998 seston isotope values alone.

Estimating the relative importance of the three POS as sources of carbon and
nitrogen to fish eggs and chironomid larvae, the other main prey items of the deepwater
sculpin, was more difficult. Fish eggs do not directly utilize any of the three POS. Thus,

a mixing model analysis with fish eggs and these three endmembers would be

21

inappropriate. In addition, the isotopic mixing model of chironomid larvae and the three
POS did not provide any plausible solutions using the trophic fractionation factors chosen
and trophic levels between 1 and 1.5. The mixing model was also not applied to the
minor prey items of the deepwater sculpin because seston, sinking POM and sediments
were inappropriate endmembers for trophic relationships involving some of these prey
items, and provide implausible solutions for others.

The relative contribution of carbon and nitrogen from the three POS to the
deepwater sculpin was determined based on the results of the isotopic mixing models of
D. hoyi and M relicta alone. D. hoyi and M. relicta comprised the vast majority of the
deepwater sculpin diet. Together, they comprised 74.6 % of the deepwater sculpin diet
based on % PN, 90.3 % of the diet based on % PM and 89.6 % of the diet based on
% IRI. Given that we had accounted for this large proportion of the diet, errors in the
estimates of relative importance of POS to the deepwater sculpin were expected to be
minimal. Under these circumstances, seston, sinking POM and sediments were
discovered to be the ultimate source of 69.3-71.8 %, 6.8-7.1 % and 21.1-23.9 %,
respectively, of the carbon and nitrogen of the deepwater sculpin (Figure 9). It is
important to note that the mixing model of the deepwater sculpin actually had four
endmembers contributing carbon and nitrogen using this method of calculation: 1997
seston (from the D. hoyi mixing model results), average 1997-1998 seston (from the
M relicta mixing model results), and average 1997-1998 sinking POM and sediments
(from both the D. hoyi and M relicta model results). While the overall results of the
deepwater sculpin mixing model suggested that seston was the ultimate source of 69.3-

71.8 % of the carbon and nitrogen for the deepwater sculpin (Figure 9), 1997 seston

22

actually represented 63.8-70.0 % of this overall contribution of seston, while average

1997-1998 seston represented only 1.8-5.5 % of the total.

23

DISCUSSION

Stomach Content Analysis

The four indices of prey importance explored in this study may each provide a
unique perspective of prey importance owing to inherently different biases (Pinkas et al.
1971, Gould et al. 1997a). Sampling error, particularly in the case of low sample size,
may bias % F O to a greater extent than the other indices. Percent PN may allow
numerous small prey items to overshadow the importance of fewer large prey items,
while % PM may be biased by the occurrence of a few large prey items that are not
representative of the normal diet. Percent IRI incorporates the three previously described
indices in an attempt to minimize the biases of each and obtain a more representative
description of diet (Pinkas et a1. 1971). An ideal index of prey importance would also
incorporate a measurement of nutrition, such as calories per unit volume (Pinkas et al.
1971), but such analyses were beyond the scope of this study.

In this study, the four indices of prey importance differed in their diet assessment
in each month as expected, with % IRI integrating the suggested significance of each
prey item from % F0, % PN and % PM. The importance of D. hoyi to the deepwater
sculpin diet was supported by the highest seasonal mean value of relative importance
according to all four indices (Table 1). Large prey items, such as M relicta, were deemed
to comprise a large portion of the deepwater sculpin diet according to the traditional
measure of % PM, while numerous small prey items like fish eggs were suggested to
have a higher significance according to % F0 and % PN (Table 1). The seasonal means

of % PN, for example, suggested that M relicta comprised 4.7 % of the deepwater

24

sculpin diet and fish eggs comprised 14.1 %. In contrast, the seasonal means of % PM
suggested that M relicta comprised 11.5 % of the deepwater sculpin diet while fish eggs
comprised only 3.2 %. The seasonal means of % IRI provided a balance between the
means of % PN and % PM, with M relicta comprising 3.7 % of the deepwater sculpin
diet and fish eggs comprising 6.5 %.

The most notable example of potential bias in stomach content data based on a
single index of prey importance was evident in the comparison of results for the
September deepwater sculpin diet. In this month, a single crayfish was found in a single
deepwater sculpin stomach. Because of its large biomass relative to more common prey
items, however, % PM suggested that crayfish comprised 19.1 % of the deepwater
sculpin diet in September, and 3.2 % of the annual deepwater sculpin diet. Percent IRI,
in contrast, suggested that crayfish comprised only 0.1 % of the September diet, and
0.0 % of the seasonal average. In any event, this prey item was removed from fiirther
analyses because of the lack of analogous stable isotope data. While it is important to
take into consideration the potential biases of each index of prey importance, each index
provided a unique perspective regarding the relative importance of each prey item to the

deepwater sculpin in this study.

Sggonal Variation in 615N atnd 513C of Deepwater Sculpinfland Average Diet

Seasonal variation in the SN and 513C values of food web members was
prevalent in Grand Traverse Bay. While differing amounts of seasonal variation were
observed in the isotopic composition of the many food web members, the range of S'SN

values throughout the sampling season was always much higher than that observed in

25

613C values. In fact, the seasonal variation of SN values was consistently about twice
that observed in 5'3C values, regardless of food web member examined. Although
seasonal variability may be observed in both the 815N values of inorganic nitrogen and
the 813C values of inorganic carbon, variation in 815N values at the base of the food web
may be accentuated by shifts in phytoplankton uptake of the two inorganic nitrogen pools
(N03 and NH4), which differ isotopically by more than 10 °/oo (McCusker et al. 2000).
This large amount of isotopic variation in nitrogen at the base of the food web may then
be transferred through the food web.

The 8'5N and 613C values of the deepwater sculpin in Grand Traverse Bay
exhibited very little seasonal variation compared with lower trophic level food web
members (Figures 4 and 5). Higher trophic level organisms differ in growth and
physiology from smaller, lower trophic level organisms, and have slower tissue turnover
rates (Brett et al. 1969, Kitchell and Stewart 1977, Hansen and Christoffersen 1995,
Dittel et al. 1997, Herzka and Holt 2000). Thus, the constant 8'5N and 513C values of the
deepwater sculpin relative to its diet are not surprising, as strong seasonal signals
observed at lower trophic levels are dampened out or minimized as trophic level
increases (Cabana and Rasmussen 1996, Harvey and Kitchell 2000). Several other
studies have also noted increasing isotopic variability at lower trophic levels (Bun and
Boon 1993, Yoshioka et al. 1994, Zohary et al. 1994). The isotopic signature of growing
broad Whitefish (Coregonus nasus) was found to integrate the isotopic signature of its
diet over a period greater than 1 year (Hesslein et al. 1993). Slow metabolic rates would
be expected for organisms such as deepwater sculpin that live in extremely cold water

(.<.. 4°C). Therefore, turnover times are likely greater than the growing broad Whitefish.

26

With such slow tissue turnover, the invariable 8‘5N and 513 C values of the deepwater
sculpin over the time span of this study are not surprising.

Previous studies have noted high Sl’N values of lower trophic level food web
members in the spring, followed by a decline throughout the summer months (Goering et
al. 1990, Toda and Wada 1990, Bun and Boon 1993, Yoshioka et al. 1994). The observed
seasonal decline in the 615N values of the average sculpin diet (Figure 4) was likely due
to a combination of several factors. The 815N values of the major prey items of the
deepwater sculpin may have been tracking a similar seasonal decline in 6'5N values
observed in two of the primary organic sources at the base of the food web in Grand
Traverse Bay: seston and sinking POM (Table 2). In addition, prey items with high 6'5N
values (i.e. fish eggs) were more important to the sculpin diet in the spring, while prey
items with particularly low 8'5N values (i.e. beetles) became more important to the
sculpin diet as the season progressed (Table 1; Figure 2). Although fish eggs and beetles
comprised a relatively small portion of the overall deepwater sculpin diet, their isotopic
compositions were quite unique relative to more common prey items. A combination of
system-wide spring enrichment and the availability or selection of prey items with unique
isotopic signatures was most likely controlling the seasonal decline in 615N values of the
average deepwater sculpin diet in Grand Traverse Bay.

The relatively constant 815N and 813 C values of the deepwater sculpin, combined
with the observed variation in the monthly estimates of the average isotopic composition
of the sculpin diet, resulted in a large range of trophic fractionation factors for both SISN
and 613 C values. Consequently, it was difficult to define an accurate single estimate of

trophic fractionation of 815N and 613C values for the deepwater sculpin in Grand Traverse

27

Bay. Previous studies have suggested that trophic fractionation is better represented by
seasonal averages of 513C and 6‘5N values than by instantaneous values (Neilson et al.
1998), and given the large range in fractionation factors in this study, the best estimate of
fractionation was calculated here as a 6-month seasonal average. It is recognized that
such averages are simply estimates and that because seasonal variation likely continues
during the winter months, additional months of data would further refine these trophic
fractionation factors. Our results may be biased by the lack of sampling during the winter
months, and this stresses the importance of sampling throughout the entire year.

Trophic fractionation factors calculated between deepwater sculpin and sculpin
diet (ranging from 3.3 °/oo to 4 %o) were within the commonly reported range of 3-4 %o
(e.g. DeNiro and Epstein 1981, Minagawa and Wada 1984, Fry 1988, Harrigan et al.
1989, Goering et al. 1990, Keough et al. 1996, Gorokhava and Hansson 1999, Roth and
Hobson 2000). There is a range of trophic fractionation factors in S'SN values among
taxa, and fractionation factors well below 3%o (Dittel et al. 1997, Tatrai et al. 1999,
Adams and Sterner 2000, Herzka and Holt 2000) and as high as 5 or 6 %o (Estep and
Vigg 1985, Adams and Sterner 2000) have been recorded. The trophic fractionation of
BBC values between deepwater sculpin and sculpin diet (ranging from 1.5 %o to 2 %o)
was quite high relative to other studies, which often cite fiactionation factors of 1 %o or
less (e.g. DeNiro and Epstein 1978, Fry and Parker 1979, Peterson and Fry 1987, Goering
et al. 1990, Keough et al. 1996, Dittel et al. 1997). However, a large range of trophic
fractionation in 813C values has also been reported among taxa in the literature, ranging
fiom negative trophic fractionation factors (DeNiro and Epstein 1978, Hesslein et al.

1991, Zohary et al. 1994, Kiriluk et al. 1995) to those greater than 2 %0 (DeNiro and

28

Epstein 1978, Yoshioka and Wada 1994, Gu et al. 1996, Tatrai et al. 1999, Roth and
Hobson 2000).

Our somewhat high estimates of carbon trophic fractionation relative to those in
the literature may have been a result of a bias in our stomach content analysis. Literature
values of trophic fractionation determined with methods other than stomach content
analysis (ex. laboratory feeding experiments) would not be affected by such biases. While
stomach content analyses were exhaustive, there are inherent biases in identifying only
those food items that are still distinguishable after partial digestion and those food items
with slow rates of digestion (Gu et a1 1996, Gould et al. 1997b, Jennings et al. 1997).
Small, relatively soft-bodied prey items like chironomids and fish eggs may have been
under-represented in the results of stomach content analyses, while large-bodied prey
with protective carapaces like mysids and amphipods may have been over-represented.
Because some of the minor prey, such as chironomid larvae, fish eggs and beetles, had
very unique carbon isotopic compositions relative to more common prey items (Figure
2), a shift in diet to include a larger proportion of these minor prey could dramatically
affect the isotope values of deepwater sculpin. In addition, variations in the isotopic
composition of the minor prey items may not have been taken sufficiently into account
because of the small number of minor prey samples collected. The application of more
common literature values of trophic fractionation in this study ( ~ 3 %o for 8'5N values
and ~ 1 %o for 813C values) would imply a much larger proportion of minor prey items in
the deepwater sculpin diet (Figure 2). Calculating trophic fractionation factors in this
manner, based on a combination of stomach content and stable isotope analyses

(equations 2 and 3), assumes that ingestion and assimilation are equal. The model is, in

29

part, based on stomach content analysis and is therefore not entirely free of error
associated with differential digestion. Differential digestion may actually play a
significant role in the nutrition of the deepwater sculpin. Our trophic fractionation factors
may have, therefore, been biased somewhat by the stomach content analysis, and minor
prey items may play a larger role in the nutrition of deepwater sculpin than previously
thought.

In summary, large seasonal variations in trophic fractionation were observed in
this study. Relatively high estimates of mean trophic fractionation factors may indicate
not only the importance of sampling throughout an annual cycle, but also the increased
importance of minor prey items to the deepwater sculpin diet. Trophic fractionation may
vary among organisms and environments, and although our estimates were at the upper
limit or above most commonly reported values, they were not outside the range of
previously observed results. There is clearly a strong seasonal influence in Grand
Traverse Bay, and it is easy to recognize the importance of long-term data in isotopic
food web studies upon reviewing these results.

The models described below are particularly sensitive to estimates of trophic
fiactionation. We have made a best attempt to investigate results of the models with our
calculated trophic fractionation factors as well as literature values of trophic
fi-actionation, but but admit that errors in estimates of fractionation may greatly alter the

model results.

30

Qtimates of the Relative Contribution ofPOS to Prev Items

The isotopic mixing model assumed that seasonal averages of isotopic
compositions were the best estimates of long-term assimilation for the study organisms in
Grand Traverse Bay, and that trophic relationships could be explored using these
seasonal averages. The mixing model was used to predict the relative contribution of
carbon and nitrogen from each of the three POS to the prey items of the deepwater
sculpin. This was an important step leading to the ultimate determination of the relative
importance of the three POS to the deepwater sculpin of Grand Traverse Bay.

D. hoyi use benthic habitats almost exclusively, and are thought to rely primarily
on organic particles that settle from the water column to the sediments (Gardner et al.
1985, Quigley 1988, Gauvin et al. 1989). The mixing model indicated that seston was
the most important source of carbon and nitrogen for D. hoyi in 1997. Seasonal isotope
values supported this idea, as the SN values of D. hoyi exhibited nearly identical
seasonal trends to those observed in the seston; trends that were less similar to those
observed in the 8'5N of sinking POM (Figure 7). This heavy reliance of D. hoyi on
seston throughout the season is inconsistent with previous assertions that amphipods
depend on sinking POM for their sustenance in the spring, and rely on lipids gained
during this period throughout the remainder of the year (Gardner et al. 1985, Gauvin et al.
1989)

Several factors may be responsible for the observation that D. hoyi relied
primarily on carbon and nitrogen from seston in 1997. D. hoyi in Grand Traverse Bay
may have consumed primarily sinking POM, but may have assimilated only the portion

of sinking POM that was most similar to the seston, isotopically. That is, they may have

31

been selectively feeding or selectively assirnilating the freshest and most nutritious
portion, or the smallest size fiaction, of the sinking POM. If fresh algal material in the
sinking POM was available for a large portion of the year, then D. hoyi would be
expected to continue feeding, rather than resort to their lipid stores for sustenance during
this time period. In fact, D. hoyi collected from Grand Traverse Bay were not found to
have significantly higher lipid stores in the spring than later in the sampling season
(Stapleton et al. in prep.) This is consistent with previous observations of intense
feeding by D. hoyi on flesh settling algal particles when available (Gardner 1985, Gauvin
et al. 1989).

The physical structure of Grand Traverse Bay in 1997 may also have facilitated
the direct use of seston by D. hoyi. In 1997, a subthermocline chlorophyll maximum was
present in Grand Traverse Bay (Macrellis 1999). There was, therefore, no structural
barrier to mixing between algal production and the direct utilization of this nutritious
resource by D. hoyi. In contrast, the chlorophyll maximum was above the thermocline
for most of the year in 1998 (Macrellis 1999, McCusker et al. 1999), which may have
limited the availability of recently produced algal particles to D. hoyi.

It remains unclear what POS D. hoyi were utilising in 1998. Selective feeding
may again have played a role, but it is clear that seston, sinking POM and sediments were
not suitable endmembers for the isotopic mixing model investigating the trophic
relationships of D. hoyi in 1998 (Figure 6). Selective feeding or assimilation of particular
fiactions of the POS may pose difficulties in the selection of appropriate endmembers for
an isotopic mixing model of this form. The actual POS that contributed to the base of the

food web may have been masked in the analysis because of selective feeding by D. hoyi,

32

and may be a combination of organic sources including uncommon material at times like
partially decomposed fish corpses. In addition, organisms may assimilate nitrogen and
carbon fiom different food sources 0(ikuchi and Wada 1996, Liden and Angerbjom
1999), introducing the possibility that D. hoyi did indeed rely on sinking POM as a major
source of nutrition in the form of nitrogen, but relied on seston or another nutritional
source for carbon. Such de-coupling of nitrogen and carbon resources would present
additional difficulties in interpreting mixing model results. Finally, despite efforts to
preserve sediment trap material, it may possible that the sinking POM collected by the
sediment traps in Grand Traverse Bay were altered somewhat by partial decomposition
and rendered isotopically distinct from the sinking POM that actually arrived at the lake
bottom. The sinking POM delivered to the bottom of the bay would likely have been
relatively fresh and unaltered by microbes at the time of delivery and ingestion by
D. hoyi, relative to the sediment trap samples analysed for isotopic composition.

The role of M relicta in the food webs of fi'eshwater lakes is often difficult to
characterize (Leggett 1998). M relicta are omnivorous crustaceans which exhibit diet
vertical migrations and ontogenetic shifts in feeding behaviour (Rudstam et al. 1989,
Leggett 1998, Rudstam et al. 1998, Branstrator et al. 2000). The results of the isotopic
mixing model suggested that approximately half of the carbon and nitrogen of the
M relicta diet was ultimately derived from each of seston and sedimentary sources, while
sinking POM was minor. Although the considerable role of sediment in the nutrition of
M relicta may be somewhat surprising, selective feeding on the surface detritus of the
sedimmt layer during the day by M relicta directly, or by their zooplankton prey, may

have resulted in this observation. Large numbers of calanoid copepods and mysids in

33

contact with the sediment surface have previously been observed using a video camera
mounted on an ROV in central Lake Michigan (Lee and Hall, personal communication).

The isotopic mixing model of M relicta and the three POS provided plausible
results only when average 1997 and 1998 seston isotope values were utilized, and not
when 1997 or 1998 seston values were used alone. However, the 1997 seston values
resulted in a triangle in Figure 8 that nearly provided plausible solutions for the diet of
M relicta. Based on the discussion of the D. hoyi mixing model, the 1997 seston may be
more likely to influence the Grand Traverse Bay food web because of the subthermocline
chlorophyll maximum present in that year. In contrast, the chlorophyll maximum was
above the thermocline for much of the year in 1998. The thermocline likely acted as a
barrier for mixing the seston to greater depths. M relicta may have been less influenced
by this barrier to mixing in 1998 (resulting in a plausible model including both 1997 and
1998 seston) than D. hoyi because of their diel vertical migrations fi'om the benthos to the
metalimnion in pursuit of food (Lasenby and Langford 1973). These vertical migrations
may have allowed M relicta to take advantage of seston particles in the metalimnion
directly and it may have provided access to zooplankton prey that feed on seston particles
in the epilimnion and metalimnion.

The mixing model of M relicta and the three POS may have been biased by the
difficulty of assigning of a trophic level for M relicta. M relicta may both prey on and
compete with zooplankton for food (Johannsson et al. 1994). The assumption that
M relicta was positioned 1.5 trophic levels above the POS was based on a parsimonious
estimate between a juvenile herbivore which would be positioned 1 trophic level above

the POS, and an omnivorous or carnivorous adult which may be positioned 1.5-3 trophic

34

levels above the POS. However, studies have often found mysids to rely nearly
exclusively on zooplankton for their sustenance (Lasenby and Langford 1973, Leggett
1998). The complexity of the food web at the level of bulk zooplankton alone makes it
difficult to accurately assess the diet of M relicta using this type of model. A thorough
stomach content analysis on the M relicta collected in this study may have provided
additional insight into the feeding habits of M relicta and a more accurate trophic level
could have been assigned. A previous attempt at determining trophic fractionation
through stomach content analysis of M relicta found a large range of fi'actionation
factors, and the author resorted to literature values of fractionation as the most reliable
estimates available (Leggett 1998). The estimates of trophic fiactionation in this study
may have been an additional source of error to the M relicta isotope model.

Estimating the relative importance of the three POS to prey items other than
D. hoyi and M relicta was more difficult. Many of the prey items of the deepwater
sculpin were not expected to depend on the three POS measured in this study. Fish eggs
do not directly utilize any of the three POS. Chironomids do not feed during the pupal
stage of their life cycle. In fact, the isotopic composition of chironomid pupae was quite
high compared to chironomid larvae, possible evidence of the effects of starvation and
enrichment of 615N and 513C values (Figure 2). In addition, the isotopic composition of
chironomid adults, unidentified dipteran adults and terrestrial beetles are undoubtedly
heavily influenced by terrestrial sources of carbon and nitrogen. While many of these
minor prey items had unique isotopic signatures capable of affecting the 8'5N and 813 C
values of the deepwater sculpin, they comprised a very small amount of the total prey

consumed. Assuming minor prey items were not grossly underestimated in the stomach

35

content analyses, they were unlikely to significantly bias the interpretation of the relative
importance of seston, sinking POM and sediments as sources of carbon or nitrogen to the
deepwater sculpin.

Based on the results of the isotopic mixing models of D. hoyi and M relicta, the
relative importance of each of the three POS as sources of carbon and nitrogen to the
deepwater sculpin was determined. Seston was found to be, by far, the most important
POS to the deepwater sculpin, followed by sediments and sinking POM (Figure 9).
Finding an exclusively benthic feeder like the deepwater sculpin to ultimately derive
~ 70 % of its carbon and nitrogen from seston was somewhat unexpected. The heavy
reliance on seston was the result of the fact that the majority of the deepwater sculpin diet
consisted of D. hoyi, and D. hoyi were found to rely primarily on seston. Again, the true
mechanism of nutrient transport may have been a direct route from seston through
D. hoyi to the deepwater sculpin, or an indirect route from sinking POM with selective
feeding by D. hoyi on a fraction of the sinking POM to the deepwater sculpin. It is
important to keep in mind that these results are based on M relicta and D. hoyi only, and
exclude any influence of minor prey items on the relative proportions of the three POS to
the deepwater sculpin.

The isotopic mixing models thus predicted that the majority of the carbon and
nitrogen of deepwater sculpin in Grand Traverse Bay was ultimately derived from seston.
Because seston is the POS associated with an exposure route of organic contaminants
based on atmospheric deposition, these results imply that atmospheric deposition may in
fact be a very important source of organic contaminants to the deepwater sculpin.

However, these data must be used in conjunction with contaminant concentration data to

36

accurately determine the relative importance of the three hypothesized exposure routes of

organic contaminants to the deepwater sculpin.

Sources of Error and Recommendations for Future Research

It is clear upon completion of this study that there are many challenges in food
web research based on stable isotope analysis. Temporal variation in isotope values of
food web members, particularly at lower trophic levels, made it impossible to derive an
accurate single estimate of trophic fiactionation between deepwater sculpin and sculpin
diet. Seasonal averages were used as a best estimate of trophic fractionation, but seasonal
variation continues during winter months when samples were not collected, and
additional months of data may further refine such estimates. Calculating trophic
fractionation factors by assuming that stomach contents accurately define the assimilated
diet of the deepwater sculpin also provided a source of error to our analysis, as ingestion
may not be equal to assimilation.

Difficulties also arose in assuming a trophic fractionation factor for the mixing
model used to determine the relative importance of the three POS to D. hoyi and
M relicta. Trophic fractionation factors have been shown to vary greatly among species,
diets and tissue types (ex. DeNiro and Epstein 1978, Tieszen et al. 1983, Minagawa and
Wada 1984, Estep and Vigg 1985, Hobson and Clark 1992, Gu et al. 1996, Hogberg
1997, Focken and Becker 1998, Fantle et al. 1999, Gorokhova and Hansson 1999, Adams
and Sterner 2000, Roth and Hobson 2000). The mixing models of D. hoyi and M relicta
were explored using the fractionation factors calculated with the deepwater sculpin and

sculpin diet, as well as the most commonly reported literature values. However, given

37

that fi'actionation factors have been shown to vary even between similar types of
organisms fed known diets, in the case of two species of fish (tilapia and common carp;
Focken and Becker 1998) and two species of mysids (Gorokhova and Hansson 1999) in
laboratory experiments, it is very problematic to apply fiactionation factors among taxa.
The first step in resolving such difficulties would be to perform laboratory feeding
experiments with D. hoyi and M relicta with known diets to accurately determine
species-specific trophic fractionation factors. The effect of ontogenetic feeding shifts of
M relicta on isotopic composition should also be explored in more detail in future
research, to more accurately define the trophic position of M. relicta above the POS.

The many difficulties that arose with the mixing model efforts in this study
suggest that a more complex model may need to be developed to accurately assess
trophic relationships with stable isotopes. Even with a more extensive temporal scale, a
simple time lag approach to the current mixing model is not likely to provide accurate
results, as only a portion of the tissue of a consumer would be affected by short-term
feeding behaviour. A dynamic modeling approach incorporating temporal variations in
isotopic signatures and the growth, metabolism and tissue turnover of food web members
may increase the ability to test hypotheses and build inferences about food web linkages
in aquatic ecosystems (Harvey and Kitche112000).

Given the great potential value of stable isotopes in food web studies, surprisingly
little attention has been paid to the underlying physiological and biochemical
mechanisms that account for trophic enrichment (Adams and Sterner 2000). Recent
suggestions that organisms may disproportionately assimilate carbon and nitrogen from

different food sources (Kikuchi and Wada 1996, Fantle et al. 1999) may actually prevent

38

the reliable use of mixing models in food web studies altogether. To thoroughly
understand the implications of isotopic food web studies, scientists may need delay large
ecosystem research for a time and first investigate trophic relationships through
controlled laboratory investigations of the isotopic compositions of consumers and their
diets. Studies combining such investigations with those in natural ecosystems could
provide considerable new insight, and result in a more solid foundation from which future

food web research could evolve.

39

SUMMARY AND CONCLUSIONS

This study was an attempt to elucidate the diet of the deepwater sculpin in
Grand Traverse Bay, Lake Michigan using a combination of stomach content and stable
isotope techniques. Stomach content analysis revealed that the diet of the deepwater
sculpin was composed primarily of the amphipod, D. hoyi. Seasonal variation was
observed in the abundance of minor prey items in the deepwater sculpin stomachs, and
isotope data implied that these minor prey items may have been more important than
stomach content data suggested. Four different indices of prey importance were used to
summarize the stomach content data, and each provided a unique perspective of
deepwater sculpin diet. The importance of long-term seasonal sampling was clearly
supported by our study. Seasonal variations in isotopic composition were observed in the
food web members of Grand Traverse Bay, and were particularly noteworthy at lower
trophic levels. An isotopic mixing model was applied to the data in an attempt to
determine the relative importance of the three POS to consumers. The mixing model
results revealed that seston was an important POS of the benthic food web, suggesting
that atmospheric deposition may be an significant source of contaminants to the
deepwater sculpin in Grand Traverse Bay. Difficulties in the application of isotopic
mixing models in food web studies were discussed and recommendations for future

research were made.

40

TABLES

41

 

ad ad ad ad ad od 06 E. .x.
ad ad ad ad od ad ed in. .x.
od v.0 06 ad ad od od zn .x.
ed 09 .2? ad ad ad ad 0... .x.
$.33 32822.20
ad 06 od 06 ...o 0.0 0.0 E. .x.
to Pd 0.0 ...o «.0 od 0.0 E .x.
«.o «.0 ad ad ad ad ad zn. $
m6 od .2? No 93 0.0 0.0 On. ,3
395 3282.920
Qm od «6 od oi 0.0 me E. 3
«.m ad mi m6 ad 0.9 we in. x.
ms m.«v o. 3 m. 3 m.» ad ad. Zn. .x.
mdm ode QE «.2. «do 0.0 0.8 on. .x.
mate. emancQEO
m6 ad .1 0m 9m «.9 o.m« E. .x.
«.m .2? ad md m.« m...” «.9 SE .3
v.3 9m Ev «.m md Qt ...«v 2n. 8
«.5 0.x... «.«v 0mm Qmm 0.03 Now On. .x.
33 5.“.
En 0m fin as a... ed as E. .x.
of. m. 3 far 0.9. 0.x. 0.0 «.3 E x.
Ev on. Ym md ma ed as Zn. .x.
Yam 0.3” Yaw Qmm 93 ed «.3 on. .x.
$6.3. $9.3.
9mm V...» min o.«m «.om Nam wdm E. .x.
mm» Eb 0.9. Q: «.8 98 v.«\. E 3
mam ado OR «.9. «.2 m.«m r.«v Zn. 3
v.3 0.3 v.3 v.3 odor odor «.8 On. .x.
.30: $98.5
cams. nonEoEow 539.2. 22. 95.. >32 :54 Em: com

 

.cm. «a 3:889... 9.3.0.. .0 xouc. Enema ecu 2233 $9: .65 .522. .AZn. .5 .383: .35 “:3qu .5”. 3.
85:33 .o 55:02.. 28.6.. mm 332on vac—.350 comEBm 5233 6.8.2.03 Em 3.92... 9.20 3 meEzw .F 2%...

42

22000.00 .5523. new 00000... new «.2...on 00300 83.0... 2.

2.25.. can 8:22“. 3 225.83 «5336 5.2. 286:8 9... B 88358 enhance 22:8 nausea ..

 

 

 

 

a9. 00? a0 «0? 00 m an 0320...... 209.8% .0 .3802
0.0 0.0 0.0 d 0.0 0.0 0.0 E. e\e
0.0 0.0 0.0 0.0 «.0 0.0 0.0 .2... .x.
«.0 0.0 0.0 «.0 rd 0.0 0.0 2n. 0..
0.« 04. 3 ad v.« 0.0 n... O“. .x.
: 3003
...0 0d 0.0 0.0 0.0 0.0 0.0 E. 0..
rd 5.0 0.0 0.0 0.0 0.0 0.0 E .x.
«.F 5.0 «.0 «.0 d 0.0 «.0 za 3
0.«? 0.50 5.0 a.« 0d 0.0 «.m on. .x.
€093.88 managocxm
d d d 0.0 0.0 0.0 0.0 E. .x.
0.0 0.0 0.0 0.0 0.0 0.0 Pd 2.". 0..
ad ad .2? ad Pd 0.0 «.0 2a 0..
v.3 0.3. Pd? Nu 0.0 0.0 0.5 on. .x.
«08300
d v.0 d d «.0 rd 0.0 E. .x.
d «.0 0.0 ...0 rd Pd 0.0 .2n. 0..
0.? 0.« 0.0 0.0 «._. 0.0 ad 2.. 0..
0.0« 0.u« a. P« 0.0? a.a« 0.0m ...«.. On. .x.
mu._occ<
0.0 0.0 0.0 0.0 0.0 0.0 0.0 E. 0..
0.0 0.0 0.0 0.0 0.0 0.0 0.0 .2... 0..
0.0 0.0 0.0 0.0 0.0 0.0 0.0 2n. 0..
«.0 0.? 0.0 0.0 0.0 0.0 0.0 On. .x.
353 ceaefieEEeuE:
50.2 59.553 500-2 22. 9.3.. was. =54. 1 EB. noon.

 

€2.03 .. 0.nm...

43

 

 

 

cm H N N m o v a
mom- a. H «0&- vom- «.8. A0.... H moa- n... H EN- 2. H Su- .00.0. on...
o. 3 m... H 2: o. : «em. 3 H w. 3 3 H 3. a... H 0.«. .00.0. 22...
$9.3. $9.3.
R H m w w .- m =
Q8. 3 H 98. N... H QR- m... H Eh- ...o H QR- to H mou- o... H «.8. .00.0. 0%
to o... H N... F... H m... .-.o H a... 3 H m... m... H a... m... H N. 2 .00.0. 20....
.So: $28.5
3 ..mm~m.5mtm>5
com- oz oz oz oz oom- oz .00.0. 00...
3 oz oz oz oz 2. oz .00.0. 20...
Eme_8w
Sm- cau- SN- oom- 2.- 3&- ...-- .00.0. on...
3 3 a... a... v... a... to .00.0. we.
.200. 5%
IN- ...N. new- oz SN- ...3- now. .00.0. 00... 82
m. : 3. a... oz 3: E: 2: .00.0. 20... 82
com. ~.o~- 3h- .8. mdw- Nou- o.o~- .00.0. 00... 32
o... v... on an m... 2. o... .00.0. 20... 32
- Ear-Ho. gov-om
“mo-Sow 9:5 0 mEtl
cums. .moEm.omw $39.4. 23.. 95.. >22 :5... .3805. oo>> noon.

 

Emu o: u 02 .n_.w P H 38.95

mm Broom. 2m mm:_m> 62.2.80 mom. can 3a.. «.3. m5 .8 £58 :28 .0. $385 2m :39.» mm:.m> 320m.
.5952... 9...... Sam $.05...- n:m.0 E9. 3.03.8 «.389: no... too. .0 wo:_m> 328. 59.8 new 595.2 .N man...

44

 

o m N oz F v F c

«.8. 3 H 0.8- mam- oz EN. 98. New- Ass 0%

m. z 3 a 3: 2: oz to 3: Q9 it 2%
$.25

F oz oz oz oz F oz =

New- oz oz oz oz SN- oz 30V 0%

to oz oz oz oz to oz ABC 2%
£32 5.055 35.52::

a oz oz oz oz N oz :

3.. oz oz oz oz 3.- oz .82 0%

No oz oz oz oz 3 oz Ass 2%
9.25 32:85ch

N. N oz a F N oz :

EN- 2:- oz EN- 2:. wow. oz Ass 0%

n! E oz «.3 3; 3: oz ASE 22m
mmoao 82898ch

2 m m N F m m c

v.8- 3 H mam- new. You- o. E- New- no H 08. ASE 0%

Q9 5 a ma; mfl oi mfi ofl E H «.2 Ass 2%
82m. mmEEocefo

~ oz oz oz F F oz :

mam- oz oz oz 3N- Qua- oz Ass 0%

2: oz oz oz 3: mi oz ABC 2%
«now 5E
cum—2 .ooEmEmw “.3me 32. 9.2. >92 En? oonms. om>> coo“.

 

3:03 N $33.

45

$032205 >5... 00 3.053 953800 53 Eat 005.5200 0.03 30059.95 00 002$ 320$ =m z.

 

 

 

809 ._m 00 omxmoonvzv 5:08 zoom .2 2.300 m .0 30205 09:90; coambcmocoo 90 53$ 00 003.9 390$ ..
E “N 3 NH. 2 H 9 c
IN- «o H 3N- to H N.vN- 3 H 3N- no H EN. EN. No H 3N- Ass 0%
3: to H mm. to H on. to H 0.9 no H I; 3: to H SH A85 2%
£33m .2026me
[St
N oz oz oz oz N oz c
N. N. oz oz oz oz 3 H N. .N- oz 300 can
3 oz oz oz oz no H 3 oz Acct 2%
3:03 338:0...
N N oz oz oz oz oz c
QNN- QNN- oz oz oz oz oz Ass 0%
N9 N9 oz oz oz oz oz A32 2%
.zEngmbmo mmcqmaockm
m. N N N H m N c
EN- SN. 3 H EN. EN. EN. 3 H EN. EN. Ass 0%
mo 3 no H to no no. 3 H o. : 3: A3000 2%
2.380 8.0.5.83 0:3
c022 oooEmfimm 5:93 22. 0:3... >05. Eo< .00sz ow>> 000".

 

3:03 N 0.00...

46

Table 3. Trophic fractionation factors for 615N and 813C between deepwater sculpin and
average sculpin diet (calculated using 6 month seasonal averages). Trophic fractionation
factors were calculated with stomach content data expressed as percent prey number
(% PN), percent prey mass (% PM) and percent index of relative importance (% IRI).

 

 

8‘5N 5'3c
% PN 3.3 1.5
% PM 4.0 1.9
% IRI 3.9 2.0

 

47

Table 4. Results of isotopic mixing model attempts, predicting the relative importance of
seston, sinking POM and sediments as sources of nitrogen and carbon to D. hoyi and M.

relicta. “ns” indicates that there was no solution to the mixing model.

 

A. Calculated fractionation (sculpin % PN’): TFn" = 3.3 °/oo, TFc“ = 1.5 %o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

D. hoyi M. relicta
Relative Proportion of: 978'“ 988'” 97988'” 978 988 97988
Seston ns ns ns ns ns ns
Sinking POM ns ns ns ns ns ns
Sediments ns ns ns ns ns ns
B. Calculated fractionation (sculpin % PM'): TFn = 4.0 %o, TFc = 1.9 °/oo
D. hoyi M. relicta
Relative Proportion of: 978 988 97988 978 988 97988
Seston ns ns ns ns ns ns
Sinking POM ns ns ns ns ns ns
Sediments ns ns ns ns ns ns
C. Calculated fi‘actionation ficulpin % IRI‘): TFn = 3.9 °/oo, TFc = 2.0 o/oo
D. hoyi M. relicta
Relative Proportion of} 978 988 97988 978 988 97988
Seston ns ns ns ns ns ns
Sinking POM ns ns ns ns ns ns
Sediments ns ns ns ns ns ns
D. Literature estimates of fractionation: TFn = 3.4%0, TFc = 0.5 %o
D. hoyi M. relicta
Relative Proportion of? 978 988 97988 978 988 97988
Seston 73.1 ns ns ns ns 43.4
Sinking POM 19.7 ns ns ns 'ns 3.7
Sediments 7.2 ns ns ns ns 52.9

 

 

' PN represents prey number, PM represents prey mass and IRI represents prey index of

relative importance

.. TFn denotes the trophic fiactionation for S'SN, TFc denotes the trophic fractionation

for 5%

".978 indicates mixing model results using 1997 seston isotope values, 988 indicates
mixing model results using 1998 seston isotope values and 97988 indicates mixing
model results using average seston isotope values of the two years combined

48

FIGURES

49

45°15’

45°00’

44°85’

 

Figure 1. Location of the 2 study sites GT1 and GT3 in Grand Traverse Bay,
Lake Michigan (http://www.glerl.noaa.gov).

50

.wOm 0005 05 00 00:_0> 820$ 05 200000000 00.0002 000 080:
>000 00.000 02030000 05 00 m020> 820$ 05 000000000 00.20 .5930 02030000 05 00 0020>
820$ 05 20000000. 000003 0000080 .0000000 000 03000 m0:_0> .009 000 39 0_ 00902.2

000.. £00 00080:. 00000 00 8000.00. 002, 002 00 0020> 820$ 000000 000 000202 .N 050E

 

As 00 020
0H- or. cm. 3. mm- mm. 00. mm- mm. R- mm. mm-
0 0 0 L 0 L b 0 _ v
00:00m 0
200000004 800: - 0
02000 4
2.000 O . 0
0 00.205
2.000 T N
058900 .200 0:250 4 0
so: .0 e . 0
08000000000 .m 0 r or
1 3
m0__000< O $.00:
000.000 .2 O 4 02000 - NH
0020.
0.800230 0 - 0?
00000 800 l . 3
EEOCQEO O 50.... C .EowQEoS .3

 

 

 

9.

51

6‘N (04,0)

6‘30 (W...)

15.0

14.5

14.0

13.5

13.0

12.5

-23.6

-24.0

-24.4

-24.8

-25.2

 

 

 

 

 

o 0 O 00
0 0 0 0
” 00 00%0
o O O O O
__ o 00 o
0 0 0Q)
_ 00) O 039 O O 0 80
O O % 0% (230%?) @O
— 0 Q) 00 00° 0
P 000 O 00 0069 86’
O OO O 00 o
._ 00
_ o O
[:1
"" :1
_ U c] D
D D E]
r— D DUE] Dan E] D C]
D 0000000 U
” [:10 g u U r_']
__ '3 BU C5200 EDD [20.11938
DUE] 0 DB £1
r CI] D DQD CID Cb
D D [$19 an El
__ D D E]
DD :1 [:1 C]
[:1 ED
_ [1 C] D
[L111]!llllllllllllliqlll
50 60 70 80 90 100110 120130 140150 160170 180
Iength(mm)

Figure 3. 8‘5N (A) and 613C (B) distributions of deepwater sculpin as a
function of length in Grand Traverse Bay, 1997-1998.

52

 

15

144

fi

  
 
   
  

 

Deepwater Sculpin

Average Diet

%PN
\ A

% IRI

%PM

 

 

T I

April May June July August September

month

Figure 4. Monthly nitrogen isotope values for deepwater sculpin and
average deepwater sculpin diet. Average diet was calculated based on
stomach contents expressed as percent prey number (% PN), percent
prey mass (% PM) and percent prey index of relafive importance (% IRI).

815M values for deepwater sculpin are shown as averages +l- 1 SE.

53

 

 

 

 

 

-24 - Deepwater Sculpin
1 W;
1 t
-25
3
‘1’ Average Diet
no
Foo
-26 j
l
l
-27
‘28 T r r i I
April May June July August September
month

Figure 5. Monthly carbon isotope values for deepwater sculpin and
average deepwater sculpin diet. Average diet was calculated based on

stomach contents expressed as percent prey number (% PN), percent
prey mass (% PM) and percent prey index of relative importance (% IRI).

615N values for deepwater sculpin are shown as averages +l- 1 SE.

54

00000.. 000000 000 00 00000 0000000000 900.0. 20205 0000000000 0_ 00:_0>

>90 .000 05 9600 .05. 20090 F 00 2 005330 3; :2 .0 .220 .280 «.0 0:0 020 80 800.8
0000000000 $0000 >0 00:_0> 220 000 020 05 @0000000 >0 020.0200 003 “$0 .300 .Q 00.000>< .wOm
00:0 05 000 .0200 0.000005 02 8.0 00000030 .0020 @0000. 000 00 0000000000000 .020005 .0 059".

 

   
 

@090
mu- 0.«. mm- 00- m..- 00- mu-
0.
31.0: 31.00 M
005000 8.80 52 mm,
0
K
300.: - - - z. - ., 0 o...
.200 90.0.0 . ........ 0 N9
. ...................................................... “.0 02000 . \Iy
...... .. 82 0 0.02 0 mt
. ..... ........ we
................. m E
......0 02000 82 M. 2
fl
2

 

 

 

55

.30.. 000.0002 00.0.. .>0m 00.90.... 000.0
0. .29”. 00.0.0.0 000 02000 . .300 .Q 00 00:.0> 00200. 00000.0 >...000.2 .0. 050...".

 

5000.
000002000 003004. >.:.. 00:... >05. .00<
h p . o
I N
- v
02000 - 0
.200 8.0.0.0 - 0
. :0: .Q 1 OF

 

 

N_.

 

(°°/o) N99

56

..0008 00.008 000 00 00000 00200020 0>00.0. 0000.00. 0000000900

0. 00:.0> >05 .won. 000 0>000 0.05. 0.0000 02 00 2 00E0000 003 0.000. .3. 0220 .2 8? 0.0 000 090 .2
8.0 0.8 0000000000 0.0000 >0 00:.0> 220 000 090 000 00000000 >0 0000.00.00 003 00.0 0.0000 .2 0m0.0><
.won. 00.00 05 000 0.000. 0.0%.. .2 3.... 0000003.. .0000. 00.000. 000 00 0002000900. .00.0005 .0 0.30.“.

 

   

 

 
    

     

    

 

 

 

.8 o. 020
mm- .0- 00- 00- ..N- 00- mm.
. . _ . . . . _ 0 . r _ . . . _ . . . . . . . . . h v
3.. 0.00.
005.80 M 0
........................................... W O
t 00.0. 0.000. .3. H 9
......0m0.0>< M Nm
3.. 0.0. 3.. 0.00. .. 0 )o
...... 002 0 52 w m
..........x...... . W Dr
8.80 002 H
0 2

57

E. 0:00.00

05.x.

9.00.0

 

$0.3
80000.3

.00... 00000000.. 0>00.0. .0 .800. >000 000 020. 0000.
>20 .20. .0080: >0... 00 0080.96 0.00 .0208 0000.20 00 00000 02000. 00000.00. 03000 80.0 .03. 0000003. .0000. 00.000.
0.0200. 00. 2 00.0.0000 00070002 .>0m 00.05.... 000.0 0. 0.0.000 .2030000 2 000 00.5 05 .0 00000000.. 900—0”. .0 0.00.“.

En. 00090.0

$0.00
0908

    

08.0
.200
00.0.0.0

$0.0m
8005.3

 

Zn. 0000.00

*3.
800.0

$9 N
80083

LITERATURE CITED

Adams, TS. and R.W. Sterner. 2000. The effect of dietary nitrogen content on trophic
level 15N enrichment. Limnology and Oceanography 45(3):60l-607.

Baker, J .E., N.E. Ostrom, P.H. Ostrom, B.J. Eadie, D.S. Lee and D. Hall. 1996. Trophic
transfer of atmospheric and sedimentary contaminants into Great Lakes fish:
controls on ecosystem-scale response times. Proposal to US. Environmental
Protection Agency. 31pp.

Balcer, M.D., N.L. Korda and LI. Dodson. 1984. Zooplankton of the Great Lakes.
University of Wisconsin Press. Madison, WI.

Branstrator, D.K., G. Cabana, A. Mazumder and J .B. Rasmussen. 2000. Measuring life-
history omnivory in the opossum shrimp, Mysis relicta, with stable nitrogen
isotopes. Lirnnology and Oceanography 463-467.

Brett, J .F ., J .E. Shelbom and CT. Shoop. 1969. Growth rate and body composition of
fmgerling sockeye salmon, Oncorhynchus nerka, in relation to temperature and
body size. Journal of the Fisheries Research Board of Canada 26:2363-2394.

Bunn, SE. and RI. Boon. 1993. What sources of organic carbon drive food webs in
billabongs? A study based on stable isotope analysis. Oecologia 96:85-94.

Cabana, G. and J .B. Rasmussen. 1994. Modelling food chain structure and contaminant
bioaccumulation using stable nitrogen isotopes. Nature 372:255-257.

Creach, V., M.T. Schn'cke, G. Bertru and A. Mariotti. 1997. Stable isotopes and gut
analyses to determine feeding relationships in saltmarsh macroconsumers.
Estuarine, Coastal and Shelf Science 44:599-611.

Culver, D.A., M.M. Boucherle, D.J. Bean and J .W. Fletcher. 1985. Biomass of
freshwater crustacean zooplankton from length-weight regressions. Canadian
Journal of Fisheries and Aquatic Sciences 42:1380-1390.

DeNiro, M.J. and S. Epstein. 197 8. Influence of diet on the distribution of carbon
isotopes in animals. Geochimica et Cosmochimica Acta 42:495-506.

DeNiro, M.J. and S. Epstein. 1981. Influence of diet on the distribution of nitrogen
isotopes in animals. Geochimica et Cosmochimica Acta 45:341-353.

Dittel, A.I., C.E. Epifanio, L.A. Cifuentes and UL. Kirchman. 1997. Carbon and
nitrogen sources for shrimp postlarvae fed natural diets from a tropical mangrove
system. Estuarine Coastal and Shelf Science 45(5):629-637.

59

Dumont, H.J., 1. Van de Velde and S. Dumont. 1975. The dry weight estimate of
biomass in a selection of cladocera, copepoda and rotifera from the ploankton,
periphyton and benthos of continental waters. Oecologia 19:75-97.

Eadie, B.J., R.L. Chambers, W.S. Gardner and G.L. Bell. 1984. Sediment trap studies in
Lake Michigan: resuspension and chemical fluxes in the southern basin. Journal
of Great Lakes Research 10:307-321.

Estep, M.L.F. and S. Vigg. 1985. Stable carbon and nitrogen isotope tracers of trophic
dynamics in natural populations and fisheries of the Lahontan Lake system,
Nevada. Canadian Journal of Fisheries and Aquatic Sciences 42:1712-1719.

Fantle, M.S., A.I. Dittel, S.M. Schwalm, C.E. Epifanio and ML. Fogel. 1999. A food
web analysis of the juvenile blue crab, Callinectes sapidus, using stable isotopes
in whole animals and individual amino acids. Oecologia 120(3):416-426.

Focken, U. and K. Becker. 1998. Metabolic fractionation of stable carbon isotopes:
Implications of different proximate compositions for studies of the aquatic food
webs using delta 13C data. Oecologia 115(3)337-343.

Fry, B. 1988. Food web structure on Georges Bank from stable C, N, and S isotopic
compositions. Limnology and Oceanography 33(5):]182-1190.

Fry, B. and PL. Parker. 1979. Animal diet in Texas seagrass meadows: l3C evidence for
the importance of benthic plants. Estuarine, Coastal and Marine Science 8:499-
509. .

Fry, B. and E. Sherr. 1984. 813 C measurements as indicators of carbon flow in marine
and freshwater ecosystems. Contributions Marine Science 27:13-47.

Gardner, W.S., T.F. Nalepa, W.A. Frez, E.A. Cuichocki and RF. Landrum. 1985.
Seasonal patterns in lipid content of Lake Michigan macroinvertebrates.
Canadian Journal of Fisheries and Aquatic Sciences 42:1827-1832.

Gauvin, J .M., W.S. Gardner and MA. Quigley. 1989. Effects of food removal on
nutrient release rates and lipid content of Lake Michigan Pontoporeia hoyi.
Canadian Journal of Fisheries and Aquatic Sciences 46:1125-1 130.

Gearing, J .N. 1991. The study of diet and trophic relationships through natural
abundance l3C. Chapter 13 in Carbon Isotope Techniques, Academic Press Inc.
pp.201-217.

Gearing, J .N., R]. Gearing, D.T. Rudnick, A.G. Requejo and MJ. Hutchins. 1984.

Isotopic variability of organic carbon in a phytoplankton-based, temperate
estuary. Geochimica et Cosmochimica Acta 48:1089-1098.

6O

Goering, J., V. Alexander and N. Haubenstock. 1990. Seasonal variability of stable
carbon and nitrogen isotope ratios of organisms in a North Pacific Bay.
Estuarine, Coastal and Shelf Science 30:239-260.

Gould, P., P. Ostrom and W. Walker. 1997a. Trophic relationships of albatrosses
associated with squid and large-mesh drift-net fisheries in the North Pacific
Ocean. Canadian Journal of Zoology 75:549-562.

Gould, P., P. Ostrom and W. Walker. 1997b. Food of flesh-footed shearwaters Puflinus
cameipes associated with high-seas drifinets in the central North Pacific Ocean.
EMU 97:168-173.

Gorokhova, E. and S. Hansson. 1999. An experimental study on variations in stable
carbon and nitrogen isotope fractionation during growth of Mysis mixta and
Neomysis integer. Canadian Journal of Fisheries and Aquatic Sciences
56(11):2203-2210.

Grossnickle, NE. 1982. Feeding habits of Mysis relicta — an overview. Hydrobiologia
93: 101-1 07 .

Gu, B., C.L. Schelske and M.V. Hoyer. 1996. Stable isotopes of carbon and nitrogen as
indicators of diet and trophic structure of the fish community in a shallow
hypereutrophic lake. Journal of Fish Biology 49:1233-1243.

Hansen, B. and K. Christoffersen, 1995. Specific growth rates of heterotrophie plankton
organisms in a eutrophic lake during a spring bloom. Journal of Plankton
Research l7(2):413-430.

Harrigan, P., J .C. Zieman and SA. Macko. 1989. The base of nutritional support for the
gray snapper (Lutjanus griseus): and evaluation based on a combined stomach
content and stable isotope analysis. Bulletin of Marine Science 44(1):65-7 7.

Harvey, C.J. and J .F. Kitchell. 2000. A stable isotope evaluation of the structure and
spatial heterogeneity of a Lake Superior food web. Canadian Journal of Fisheries
and Aquatic Sciences 57:1395-1403.

Herzka, S.Z. and J .G. Holt. 2000. Changes in isotopic composition of red drum
(Sciaenops ocellatus) larvae in response to dietary shifts: Potential applications to

settlement studies. Canadian Journal of Fisheries and Aquatic Sciences
57(1):]37-147.

Hesslein, R.H., M.J. Cape], D.E. Fox and K.A. Hallard. 1991. Stable isotopes of sulfur,
carbon, and nitrogen as indicators of trophic level and fish migration in the lower

Mackenzie River Basin, Canada. Canadian Journal of Fisheries and Aquatic
Sciences 48:2258—2265.

61

Hesslein, R.H., K.A. Hallard and P. Ramlal. 1993. Replacement of sulfur, carbon, and
nitrogen in tissue of growing broad Whitefish (Coregonus nasus) is response to a
change in diet traced by 834$, 813 C and SISN. Canadian Journal of Fisheries and
Aquatic Sciences 50:2071-2076.

Hobson, K.A. and R.G. Clark. 1992. Assessing avian diets using stable isotopes: 11.
Factors influencing diet-tissue fractionation. Condor 94(1):]89-197.

Hobson, K.A. and HE. Welch. 1992. Determination of trophic relationships within a
high Arctic marine food web using 813C and 8'5N analysis. Marine Ecology
Progress Series 84:9-18.

Hogberg, P. 1997. Tansley review no. 96 15N natural abundance in soil-plant systems.
New Phytologist 137(2):]79-203.

J acoby, C. 1953. Notes on the life history of the deepwater sculpin, Woxocephalus
quadricornis L,in Lake Superior. M.S. thesis. Univ. Mich., Ann Arbor, Mich.

Jennings, S., O. Renones, B. Morales-Nin, N.V.C. Polunin, J. Moranta and J. Coll. 1997.
Spatial variation in the 15N and 13C stable isotope composition of plants,
invertebrates and fishes on Mediterranean reefs: implications for the study of
trophic pathways. Marine Ecology Progress Series 146:109-116.

Johannsson, O.E., Rudstam, LG. and DC. Lasenby. 1994. Mysis relicta: assessment of
metalirnnetic feeding and implications for competition with fish in Lakes Ontario
and Michigan. Canadian Journal of Fisheries and Aquatic Sciences 51 :2591-
2602.

Keough, J .R., M.E. Sierszen and CA. Hagley. 1996. Analysis of a Lake Superior coastal
food web with stable isotope techniques. Limnology and Oceanography
41(1):]36—146.

Kikuchi, E. and E. Wada. 1996. Carbon and nitrogen stable isotope ratios of deposit-
feeding polychaetes in the Nanakita River Estuary, Japan. Hyropbiologia
321(1):69-75.

Kiriluk, R.M., M.R. Servos, D.M. Whittle, G. Cabana and J .B. Rasmussen. 1995. Using
ratios of stable nitrogen and carbon isotopes to characterize the biomagnification
of DDE, mirex and PCB in a Lake Ontario pelagic food web. Canadian Journal
of Fisheries and Aquatic Sciences 52:2660-2674.

Kitchell, J .F. and DJ. Stewart. 1977. Applications of a bioenergetics model to yellow

perch (Percaflavescens) and walleye (Stizostedion vitreum vitreum). Journal of
the Fisheries Research Board of Canada 34:1922-1935.

62

Kraft, CE. 197 7 . Comparative trophic ecology of deepwater sculpins (Myosocephalus
quadricomis) and slimy sulpins (Cottus cognatus) in Lake Michigan. M.S. thesis,
Oceanography and Limnology, Univ. Wisconsin-Madison, Madison, Wise.

Kraft, CE. and J.F. Kitchell. 1986. Partitioning of food resources by sculpins in Lake
Michigan. Environmental Biology of Fishes l6(4):309-316.

Lasenby, DC. and RR. Langford. 1973. Feeding and assimilation of Wsis relicta.
Limnology and Oceanography. 18:280-285.

Leggett, M.F. 1998. Food-web dynamics of Lake Ontario as determined by carbon and
nitrogen stable isotope analysis. Ph.D. Dissertation, University of Waterloo.
Waterloo, ON. 266 pp.

Liden, K. and A. Angerbjorn. 1999. Dietary change and stable isotopes: A model of
growth and dormancy in cave bears. Proceedings of the Royal Society Biological
Sciences Series B 266(1430):1779-1783.

Macrellis, A. 1999. Geochemical and isotope dynamics of dissolved inorganic nitrogen
in Grand Traverse Bay, Lake Michigan. Masters Thesis, Michigan State
University. East Lansing, MI. 42 pp.

McCusker, E.M., P.H. Ostrom, N.E. Ostrom, J .D. Jeremiason and J .E. Baker. 1999.
Seasonal variation in the biogeochemical cycling of seston in Grand Traverse
Bay, Lake Michigan. Organic Geochemistry 30:1543-1557.

Merritt, R.W. and K.W. Cummins, eds. 1996. An introduction to the aquatic insects of
North America, 3rd ed. Kendall/Hunt Publishing Company. Dubuque, IA.

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

Further evidence and the relation between 8‘5N and animal age. Geochimica et
Cosmochimica Acta 48:1135-1140.

Mozley, S.C. an M.H. Winnell. 1975. Macrobenthic species assemblages of southeastern
Lake Michigan. Verh. Internat. Verein. Limnol. 19:922-931.

Neilson, R., D. Hamilton, J. Wishart, C.A. Marriott, B. Boag, L.L. Handley, C.M.
Scrimgeour, J .W. McNicol and D. Robinson. 1998. Stable isotope natural
abundances of soil, plants and soil invertebrates in an upland pasture. Soil

biology and biochemistry 30(13):1773-1782.

Nero, R.W. and W.G. Sprules 1986. Predation by three glacial opportunists on natural
zooplankton communities. Canadian Journal of Zoology 64:57-64.

63

Nesler, T.P. 1981. The relative abundance of opossum shrimp, Mysis relicta, in Twin
Lakes, Colorado using a benthic trawl. Engineering and Research Center Joint
Report with Colorado Division of Wildlife. 22pp.

Ostrom, P.H., M. Colunga-Garcia and SH. Gage. 1996. Establishing pathways of
energy flow for insect predators using stable isotope ratios: field and laboratory
evidence. Provisorische Seitenzahlen 6p.

Ostrom, N.E., S.A. Macko, D. Deibel and KL Thompson. 1997. Seasonal variation in
the stable carbon and nitrogen isotope biogeochemistry of a coastal cold ocean
environment. Geochimica et Cosmochimica Acta 61(14)2929-2942.

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

Pinkas, L., M.S. Oliphant and I.L.K. Iverson. 1971. Food habits of albacore, bluef'm
tuna, and bonito in California waters. California Fish and Game Fish Bulletin No.
152.

Quigley, MA. and J .A. Robbins. 1984. Silica regeneration processes in nearshore
southern Lake Michigan. Journal of Great Lakes Research 10:3 83-392.

Roth, J .D. and K.A. Hobson. 2000. Stable carbon and nitrogen isotopic fractionation

between diet and tissue of captive red fox: Implications for dietary reconstruction.
Canadian Journal of Zoology 78(5):848-852.

Rosenfeld, J .S. and J .C. Roff. 1992. Examination of the carbon base in sourthem Ontario
streams using stable isotopes. Journal of the North American Benthological
Society 11(1):1-lO.

Rudstarrr, L.G., K. Danielsson, S. Hansson and S. Johannsson. 1989. Diet vertical
migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the
Baltic Sea. Marine Biology 101:43-52.

Rudstam, L.G., A.L. Hetherington and A.M. Mohammadian. 1998. Effect of temperature
on feeding and survival of Mysis relicta. Canadian Journal of Fisheries and
Aquatic Sciences 25(2):363-371.

Selgeby, J .H. 1988. Comparative biology of the sculpins of Lake Superior. Journal of
Great Lakes Research 14(1)L:44-51.

Sell, D.W. 1982. Size-frequency estimates of secondary production by Mjmis relicta in
Lakes Michigan and Huron. Hydrobiologia 93:69-78.

Smock, LA. 1980. Relationships between body size and biomass of aquatic insects.
Freshwater Biology 10:375-383.

Stapleton, H.M., C. Masterson, P. Ostrom, J. Skubinna and J .E. Baker. In prep.
Accumulation of atmospheric and sedimentary PCBs and toxaphene in a Lake
Michigan food web.

Takahashi, K., T. Yoshioka, E. Wada and M. Sakamoto. 1990. Temporal variations in
carbon isotope ratio of phytoplankton in a eutrophic lake. Journal of Plankton
Research 12(4):799-808.

Tatrai, I., K. Matyas, J. Korponai, G. Paulovits, P. Pomogyi and M. Presing. 1999. Stable
isotope analysis of food webs in wetland areas of Lake Balaton, Hungary. Archiv
fuer Hydrobiologie 146(1):] 17-128.

Tieszen, L.L., T.W. Boutton, K.G. Tesdahl and NA. Slade. 1983. Fractionation and
turnover of stable carbon isotopes in animal tissues: Implications for delta carbon-
13 analysis of diet. Oecologia 57(1-2):32-37.

Toda, H. and E. Wada. 1990. Use of 15N/ 1“N ratios to evaluate the food source of the
mysid, Neomysis intermedia Czerniawsky, in a eutrophic lake in Japan.
Hydrobiologia 194:85-90.

Wells, L. 1980. Food of alewives, yellow perch, spottail shiners, trout-perch, and slimy
and fourhom sculpins in southeastern Lake Michigan. Technical Papers of the
US. Fish and Wildlife Service #98 12pp.

Wojcik, J .A., M.S. Evans and DJ. Jude. 1986. Food of deepwater sculpin,
Myoxocephalus thompsom’, from southeasrtem Lake Michigan. Journal of Great
Lakes Research 12(3):225-231.

Wong, W.W., L.L. Clarke, G.A. Johnson, M. Llaurador and PD. Klein. 1992.
Comparison of two elemental-analyzer gas-isotope-ratio mass spectrometer
systems in the simultaneous measurement of 13 C/ 12C ratios of carbon content in
organic samples. Analytical Chemistry 64:354-358.

Yoshioka, T., E. Wada and H. Hayashi. 1994. A stable isotope study on seasonal food
web dynamics in a eutrophic lake. Ecology 75(3):835-846.

Zohary, T., J. Erez, M. Gophen, I. Berman-Frank and M. Stiller. 1994. Seasonality of

stable carbon isotopes within the pelagic food web of Lake Kinneret. Limnology
and Oceanography 39(5):]030-1043.

65

MICHIGAN IIIIIIIIIIIIIIIIIIIIIIII

[IllWI1121111MHUIILWJIWBIHIHI