Diet Preferences of Juvenile Steelhead (Oncorhynchus Mykiss): A Comparison Between Three Hood Canal Rivers

Item

Title

Eng Diet Preferences of Juvenile Steelhead (Oncorhynchus Mykiss): A Comparison Between Three Hood Canal Rivers

Date

2015

Creator

Eng Davis, Sarah R

Subject

Eng Environmental Studies

extracted text

DIET PREFERENCES OF JUVENILE STEELHEAD
(ONCORHYNCHUS MYKISS):
A COMPARISON BETWEEN THREE HOOD CANAL RIVERS

by
Sarah R. Davis

A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
June 2015

© 2015 by Sarah R. Davis. All rights reserved.

This Thesis for the Master of Environmental Studies Degree
by
Sarah R. Davis

has been approved for
The Evergreen State College
by

________________________
Erin Martin, Ph. D.
Member of the Faculty

________________________
Date

ABSTRACT
Diet preferences of juvenile steelhead (Oncorhynchus mykiss):
A comparison between three Hood Canal rivers
Sarah R. Davis
Hatchery programs have been established in Washington State for decades to supplement
declining wild steelhead (Oncorhynchus mykiss) populations. However, these hatchery
programs have been implemented with a limited understanding of how the introduction of
large numbers of hatchery raised steelhead in Puget Sound river systems impact wild
steelhead populations. This information can inform current steelhead supplementation
programs throughout the Pacific Northwest by providing the composition and quantity of
important prey items and will allow for a better understanding of juvenile steelhead
needs. This study examined differences between wild and hatchery steelhead in three
Hood Canal rivers of various sizes. Results showed that Ephemeroptera aquatic larvae are
the dominant items in the drift and the dominant prey items in wild and steelhead
juvenile diets. At the river scale, wild and hatchery steelhead diets were found to differ
significantly in the middle sized river (A=0.1617, p<0.001). At the reach scale, wild and
hatchery diets were found to differ significantly in the lower reaches of the medium
(A=0.0370, p<0.001) and large sized rivers (A=0.1230, p<.001), but not in the small
river. This could possibly be due to the lack of resources in the small river, where the
abundance of items in the drift was lowest, and as such, fish had less selection, thereby
eliminating differences between wild and hatchery diet. This is consistent with the
observation that wild and hatchery diets were found to have a wider variety of prey items
in the smallest size river. Furthermore, wild and hatchery fish rejected fewer items

relative to the larger rivers, with wild fish consistently consuming all items available
relative to drift. Overall, wild fish consumed a greater diversity of prey relative to
hatchery fish. These results show that wild and hatchery steelhead have been found to
consume different items in some rivers and reaches and that river size seems to influence
the number of items available. These diet differences could possibly be due to rearing
environment differences or competition between wild and hatchery steelhead.

Table of Contents

List of Figures ………………………………………………………………………..……v
List of Tables.....................................................................................................................vii
Acknowledgements..........................................................................................................viii
CHAPTER I. Introduction……………………………………………….………………..1
CHAPTER II: Literature Review……………………………………………………….....5
CHAPTER II: Manuscript.................................................................................................32
Introduction………………………………………………………………………32
Methods…………………………………………………………………………..34
Hood Canal Steelhead Project………………………………………...…34
Study Area……………………………………………………………….38
Macroinvertebrate Drift Collection………………………………………41
Steelhead Stomach Sampling…………………………………………….42
Site Characteristics……………………………………………………….43
Data Analysis……………………………………………………….....…43
Results …………………………………………………………….……………...48
Drift Composition……………………………………………………..…48
Diet Composition……………………………………………………..….50
Diet Overlap…………………………………………………………..….55
Electivity Index…………………………………………………..……....62
Discussion………………………………………………….…………………..…68
CHAPTER IV: Conclusion………………………………………………….…………...76
References………………………………………………………………….…………….80
Appendix……………………………………………..…………………………………..88
iv

List of Figures
Figure 1 Diagrams of incomplete and complete metamorphosis (Lanham 1964)……....23
Figure 2. A map showing the locations of the supplemented (S) and control (C) streams
in Hood Canal. The Hamma Hamma River supplementation was terminated in 2007
(Berejikian et al. study plan draft)………………………………………………….…....35
Figure 3 Percent abundance of drift items collected in the Dewatto River (A),
Duckabush River (B) and South Fork Skokomish River (C)…………………………....50
Figure 4 Percent abundance of items collected from Dewatto River (A) Duckabush River
(B) and South Fork Skokomish River (C) wild and hatchery juvenile steelhead stomach
contents………………………………………………………………………….......…...53
Figure 5 Percent abundance of items collected from Dewatto River (A) Duckabush River
(B) and South Fork Skokomish River (C) hatchery juvenile steelhead stomach
contents…………………………………………………………………………..……....54
Figure 6 Percent abundance of items collected from Dewatto River (A) Duckabush
River (B) and South Fork Skokomish River (C) wild juvenile steelhead stomach
contents……………………………………………………………………………...…55
Figure 7 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap
between wild and hatchery origin steelhead juveniles in the Dewatto River (A),
Duckabush River (B) and South Fork Skokomish River (C) using the Bray-Curtis
similarity index. The proximity of the symbols indicates degree of similarity. Solid
triangles indicate hatchery steelhead, hollow triangles represent wild steelhead……..…58
Figure 8 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap
between river reaches in the Dewatto River (A), Duckabush River (B) and South Fork
Skokomish River (C) using the Bray-Curtis similarity index. The proximity of the
symbols indicates degree of similarity. Solid circles indicate the upper reaches, hollow
circles indicate the middle reaches and hollow squares indicate the lower reaches (Note:
The Dewatto River only contains upper and lower reaches, data was not collected in the
middle reach)….………………………………………………………………….…….59
Figure 9 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap
between wild and hatchery steelhead juveniles in the lower reaches of the Dewatto River
(A), Duckabush River (B) and South Fork Skokomish River (C) using the Bray-Curtis
similarity index. The proximity of the symbols indicates degree of similarity. Solid
triangles indicate hatchery steelhead, hollow triangles represent wild steelhead ……...60
Figure 10 Percent abundance of items collected from lower Duckabush River wild (A)
and hatchery (B) juvenile steelhead stomach contents………………………………….61
Figure 11 Percent abundance of items collected from lower South Fork Skokomish
River wild (A) and hatchery (B) juvenile steelhead stomach contents…………...……..62
v

Figure 12 Vanderplog and Scavia Electivity Index of prey selectivity for wild and
hatchery steelhead in the Dewatto River. Includes all items found in both the drift and
steelhead stomach contents. Grey bars represent wild steelhead, black bars represent
hatchery steelhead………………………………………………………………………..65
Figure 13 Vanderplog and Scavia Electivity Index of selectivity for wild and hatchery
steelhead in the Duckabush River. Includes all items found in both the drift and steelhead
stomach contents. Grey bars represent wild steelhead, black bars represent hatchery
steelhead……………………………………………………………………………...…..66
Figure 14 Vanderplog and Scavia Electivity Index of selectivity for wild and hatchery
steelhead in the South Fork Skokomish River. Includes all items found in both the drift
and steelhead stomach contents. Grey bars represent wild steelhead, black bars represent
hatchery steelhead……………………………………………….……………………….67

vi

List of Tables
Table 1 The physical attributes of the Dewatto, Duckabush and South Fork Skokomish
Rivers…………………………………………………………………………………….38
Table 2 The number of hatchery and wild fish included in the NMS/MRPP analysis of
the Dewatto, Duckabush and South Fork Skokomish River. Types of prey items indicates
the number of distinct prey categories included in the analysis………………...…….…45
Table 3 The number of hatchery and wild fish included in the Electivity Index analysis of
the Dewatto, Duckabush and South Fork Skokomish River. Types of prey items indicates
the number of distinct prey categories included in the analysis…………………..……..47
Table 4. The number of wild and hatchery fish included in diet analysis on the Dewatto
River, Duckabush River and South Fork Skokomish River and the number of different
item types found in stomach contents. Fish are broken down by reach, upper (U), middle
(M) and lower (L) and the drift totals are combined for all river reaches………..……...51

vii

Acknowledgements

Erin Martin, Ph.D.
Katy Doctor
Julie Davis & Mark Dixon
Keith Davis & Uma Sankar
Joshua Davis, Jonathon Davis & Danielle Heckert
Ben, Chal & Lois Martin
Laura Milleville & Chelsea Waddell
2013-2015 MES Cohort

viii

Introduction
Puget Sound steelhead (Oncorhynchus mykiss) play important cultural, economic
and ecological roles in Washington State. Declines in Puget Sound steelhead populations
in recent decades led to the development of hatchery programs in the 1960’s and 1970’s
designed to bring back historic steelhead populations (WDFW 2014a). The Puget Sound
wild steelhead population has declined 97% since the 1800’s. In 1895, steelhead
populations ranged from 330,000-820,000 fish, while the current population is at an
average of 22,000 fish (WFC (b)). These hatchery programs have been implemented with
limited understanding of how the introduction of large numbers of hatchery raised
steelhead in Puget Sound river systems impacts wild steelhead populations. Furthermore,
despite these hatchery programs, steelhead populations have continued to decline. It is
imperative that the interactions between wild and hatchery steelhead are better
understood so it can more clearly be determined if current hatchery practices are helping
or harming wild steelhead populations.
Through better understanding of wild and hatchery steelhead feeding behaviors,
hatchery programs can be developed in order to ensure that hatchery fish are not
negatively impacting wild populations. Diet analysis provides the composition and
quantity of important prey items, allowing for a better understanding of juvenile
steelhead needs (Wright 2010). An understanding of juvenile steelhead diets while in the
freshwater ecosystem can inform current steelhead supplementation plans throughout the
Pacific Northwest, where large numbers of steelhead populations also continue to
decline. Measures can be taken to allow for hatchery programs to more closely mimic

1

the conditions experienced by wild juveniles through a better understanding of the
feeding habits and the factors that influence these feeding behaviors.
A difference in diet and prey preference could have several implications for
hatchery practices and management. A difference in diet and preference may reflect
hatchery practices that are rearing fish that have diets not found in wild fish. This could
indicate that hatchery fish are not consuming the appropriate prey and are possibly not
receiving the nutrients needed to allow for their survival to maturity. However, diet
differences may also indicate competition between wild and hatchery steelhead. If larger
and more aggressive hatchery fish (Abbot 1985; Hill 2006; Keeley & McPhail 1998 &
McMichael et al. 1999) are consuming the preferred items, wild fish may be forced to
specialize their diets, consuming the less desirable items that don’t require confrontations
with hatchery fish. Conversely, if the wild and hatchery juvenile populations are found to
have similar diets it may indicate that hatchery supplementation programs are rearing fish
that exhibit the same natural feeding patterns of wild fish.
Differences in diet could also have implications in regards to trophic dynamics.
Steelhead juveniles feed almost exclusively on macroinvertebrates found in the drift. If
hatchery fish are consuming macroinvertebrates that are not often food items for wild
fish, the food chain could be drastically altered. This is due to the role of
macroinvertebrates in stream ecosystems. Macroinvertebrates influence nutrient cycling,
primary production and decomposition (Wallace & Weber 1996). If the balance of the
ecosystem is geared toward what wild steelhead consume, introducing hatchery steelhead
that consume different macroinvertebrate species may impact these essential ecosystem
functions.

2

Understanding juvenile steelhead diets can also influence habitat restoration
planning. Food webs are not often considered when planning habitat restoration projects.
By taking into consideration steelhead feeding habitats and including numerous habitat
types that allow for multiple feeding options, there is an increased chance that one of
these habitats will become favorable to steelhead when environmental and ecological
conditions change (Bellmore et al. 2013). This habitat restoration strategy will better
enable future steelhead populations to thrive in the face of continued anthropogenic
impacts including global climate change. Through the understanding of juvenile
steelhead diets and feeding behaviors, restoration teams can provide an environment that
will provide juvenile steelhead with the resources necessary to ensure proper growth and
survival of these populations. Combining this with the identification of wild and hatchery
steelhead diet preferences can provide the optimal combination of hatchery rearing and
habitat restoration practices that will allow for the most beneficial conditions to restore
wild steelhead population levels in Washington and throughout the Pacific Northwest.
In order to determine what the actual interactions are between wild and hatchery
steelhead I will be focusing specifically on juvenile wild and hatchery steelhead in three
Hood Canal rivers: the Dewatto River, Duckabush River and South Fork Skokomish
River. I will be attempting to answer the following research question: Do wild and
hatchery steelhead juveniles exhibit different diet preferences and do these preferences
vary within rivers and between rivers? The middle, upper and lower reaches of these
rivers is also of great importance to my study because it provides a more complete picture
of each river system and how habitat and prey dynamics may impact juvenile steelhead
diet and preference. These are three of six rivers included in a larger study, the Hood

3

Canal Steelhead Project (HCSP), in which the effects of hatchery steelhead
supplementation are being examined. The three supplemented rivers included in this
study contain hatchery fish that were collected from these rivers, after spawning occurred
naturally, making them genetically wild. These eggs were then raised in a conservation
hatchery setting where food and water temperatures were closely regulated.
Following this introduction, a comprehensive literature review will provide an indepth look into steelhead life history, causes for steelhead population decline, the
establishment of Washington State fish hatcheries and the known impacts of hatchery
steelhead on wild populations. The second portion of the literature review will focus
more specifically on juvenile steelhead diets and feeding behaviors. It will also provide a
detailed look into past and present research conducted in the area of juvenile steelhead
diet analysis. Next contains the findings of this study, presented in the form of a
manuscript. The manuscript contains study background, methods used and data analysis
conducted, as well as a discussion of the results and findings of the research questions.
The final chapter focuses on the key findings of this research and the implications of
these findings on steelhead management plans in Washington State and the Pacific
Northwest.

4

CHAPTER II: Literature Review
Puget Sound steelhead populations have declined 97% since the 1800’s resulting
in the development of hatchery programs in Washington State in an attempt to boost wild
steelhead populations. However, steelhead populations have continued to decline despite
these programs (WFC (b)). These hatchery programs have been implemented with a
limited understanding of the interactions between wild and hatchery steelhead and how
hatchery steelhead impact wild populations. It is imperative that these interactions are
better understood to determine if current hatchery programs are truly helping to boost
wild populations.
Understanding juvenile steelhead diets while in freshwater systems is an
important element to consider when developing hatchery supplementation plans with the
goal of improving wild steelhead populations. Steps can be taken to allow hatchery
programs to more closely align with the conditions experienced by steelhead juveniles in
the wild. For this to be possible, the feeding habits of wild and hatchery steelhead must
be understood. This literature review will provide the background information needed to
understand the complexities of wild and hatchery steelhead interactions and more
specifically, what factors have been found to influence wild and hatchery steelhead
feeding preferences and diet.
Steelhead Life History
Steelhead are anadromous salmonids. Adults return from the ocean to lay their
eggs in freshwater rivers and streams where the eggs hatch and spend 1-2 years rearing in
freshwater. Smolts out-migrate to the saltwater environment where they spend 1-3 years

5

before returning to freshwater to spawn (Sheppard 1972; Quinn 2005). Native steelhead
populations extend from the Bering Sea to southern California. Steelhead have been
introduced to freshwater lakes and streams in the United States, Canada, Europe, South
America, Africa and Asia (Gilbert & Williams 2002; Krueger & May 1991). Within the
United States steelhead have been introduced to the Appalachian Mountains (Krueger &
May 1991) and in the late 1800’s steelhead were introduced to the Great Lakes (Seelbach
1993).
Embryos
The length of time it takes for steelhead egg development is temperature
dependent. On average the eggs need 50 days of 10°C water. For every degree below
10°C, the eggs will need an additional day of incubation and for every degree above 10°C
the eggs need one less day of incubation (Sheppard 1972). Water temperatures vary
extensively between rivers resulting in a wide range of steelhead incubation periods.
Eggs that are laid in cooler streams will require much longer incubation periods than eggs
in warmer water temperatures. For example, steelhead eggs laid in a stream with an
average temperature of 5°C, will take 68 days to hatch, while in contrast, eggs in a stream
with 11°C water will only require 28 days of incubation. This acceleration of hatching in
warmer waters is due to the fact that higher temperatures increase metabolic rate and so
increase the rate of development. Salmonid embryos, including steelhead, survive best in
water temperatures between 5-11°C. Any temperatures below 2°C or above 14°C are
lethal (Quinn 2005). Eggs are typically incubating from March to May, depending on
when spawning occurred (Sheppard 1972).

6

Dissolved oxygen (DO) is the second most important factor in determining the
time required between fertilization and hatching and can actually have a slowing effect in
steelhead development in warm water temperatures. This is due to the fact that DO
concentrations decrease with warmer temperatures because of the water’s limited
capacity to hold oxygen at higher temperatures. However, metabolic rate of embryo
development increases with these warmer temperatures requiring a higher exchange of
oxygen that is not readily available, resulting in delayed hatching. A decrease in DO
concentrations at 10°C can delay the hatching of steelhead by 35-40 days, depending on
the decrease in DO concentration (Quinn 2005). Therefore, water temperature is the main
factor in determining the length of time required between egg fertilization and hatching,
but DO levels can also add extra time to embryo development when water temperatures
are high.
Alevins and Fry
After hatching, steelhead briefly enter the alevin stage, which occurs while still
within the gravel of the redd. Alevins still have a yolk sac attached to their bodies to
provide nutrients until they become large enough to capture sufficient amounts of food.
Immediately after hatching alevins bury themselves deeper into the gravel, then gradually
move up through the stream’s substrate. Fry emerge from the gravel completely once the
yolk sac has been absorbed, emerging as fry. The length of the alevin life stage is related
to the same water temperature and DO factors that influence embryonic development
(Quinn 2005).

7

Once the steelhead fry emerge they feed on microscopic organisms floating by in
the current. As juvenile steelhead grow larger in size, they move to deeper parts of the
stream to establish feeding territories where there are larger rocks and riffles. Steelhead
juveniles will stay in freshwater for an average of one to four years before they begin
smoltification (Busby et al. 1996).
Smolts
During smoltification juvenile steelhead undergo physiological changes that
prepare them for their entrance into salty ocean waters. Steelhead smolts can out migrate
to the ocean any time during the year but the majority migrates from April to June, with
the peak occurring in mid-April. However, some steelhead never migrate to the ocean
and become resident steelhead, also referred to as rainbow trout (Sheppard 1972).
Marine Adults
Once steelhead smolts reach the ocean, they begin a rapid growth process,
reaching 5-30 lbs after two to three years (Sheppard 1972), though they can continue to
grow larger once they return to the ocean after spawning. This rapid growth is due to the
large amount of food available in the marine environment (Sheppard 1972). While in the
ocean, steelhead feed on zooplankton, krill, squid, amphipods and schooling fish such as
herring and sand lance (Sheppard 1972 & Quinn 2005). The distance that steelhead will
travel while in the ocean varies greatly by population. Steelhead in the Pacific Northwest
have been found to travel west to Kamchatka Peninsula in Russia. However, not all
steelhead will travel that great a distance with some southern Oregon and northern

8

California populations only spending one summer at sea (Busby et al. 1996 & Quinn
2005).
Spawning Adults
There are two steelhead runs, winter run and summer run. The winter run
steelhead begin to enter their natal stream in October and November, with the highest
numbers occurring in January through March. Winter steelhead spawning occurs from
late March to early May (Sheppard 1972). The summer steelhead, often referred to as
“stream maturing” (Quinn 2005), begin to enter their native streams in late spring and
summer, with highest densities generally reached in the late summer months of August
through September. Summer run steelhead enter the rivers as sexually immature adults
and remain in freshwater until they reach sexual maturity and spawn the following spring
(Sheppard 1972).
Steelhead will spawn in main river channels and smaller side streams (Sheppard
1972). Females dig redds in stream bottoms, where eggs are deposited and
simultaneously fertilized by males (Quinn 2005), with about 95% of the deposited eggs
becoming fertilized (Sheppard 1972). Once the eggs have been fertilized, the females will
cover the redd with gravel to protect the eggs until the eggs hatch. Female steelhead may
have multiple redds in a single spawning season with each successive redd containing
fewer and fewer eggs. The average female steelhead (~ 700m in length) will lay around
5,000 eggs. Steelhead are iteroparous, which means they can survive spawning and repeat
the migration to the ocean and return to freshwater to spawn more than once (Quinn
2005). Steelhead have been documented to make the journey from the ocean to spawn in

9

freshwater as many as ten times (Anderson 2014), meaning that depending on when
smolting occurs, steelhead may live to be 15 years old.
Steelhead Population Decline
Throughout the coastal and interior waters of the western United States, steelhead
populations have been in decline in recent decades. Currently there are eleven steelhead
populations protected under the Endangered Species Act (ESA) including: the Columbia,
Snake and Willamette rivers, Central California Valley, the California coast, and the
Puget Sound in Washington State (NOAA 2014). On May 11, 2007, Puget Sound
steelhead was listed as a threatened species. (Dept. of Commerce 2007), meaning that this
distinct steelhead population is “likely to become endangered in the foreseeable future
throughout all significant portions of its range” (Dept. of Commerce 2007). Population
growth rates continue to decline 3-10% every year putting Puget Sound steelhead at a
high risk of going extinct in the next 100 years (NOAA 2011). This population decline is
due to overfishing, habitat loss in both freshwater and estuary environments,
development of hydropower, poor ocean conditions and hatchery practices (NOAA
2011).
Overfishing has contributed to the decline in Puget Sound steelhead and has
greatly impacted the summer run populations. Since the mid-1900’s commercial fishing
of steelhead has been limited, but sport fishing of steelhead has grown in popularity
(Sheppard 1972). Management of the steelhead sport fishing industry may have
contributed to the drastic decline, and even elimination, of summer run steelhead. After
the introduction of hatchery steelhead, open fishing for steelhead shifted to earlier in the

10

season to accommodate for the returning hatchery fish, which were genetically selected
for early return in order to reduce inbreeding between wild and hatchery populations.
This focus on the early run hatchery fish also allowed for the simultaneous catching of
wild summer run steelhead due to the fact that previous state law allowed the capture of
wild steelhead. However, current Washington State law prohibits keeping wild steelhead,
year round. Only catch and release fishing of wild steelhead is permitted (DFW306243,
2012). As a result of these early regulations, the summer run populations have declined
greatly and resulted in the shifting of steelhead life histories to favor later winter runs
(McMillan 2006).
Habitat loss has also been a major factor in steelhead population decline. There
are many factors that contribute to the loss of steelhead habitat including increased fine
sediment loads due to land use practices, changes in stream temperatures and light levels
and decreased levels of large woody debris (Collins 1976, Hicks et al. 1991 & Suttle et
al. 2004).
Human caused activities, such as road building, forestry practices, livestock
grazing and mining practices (Collins 1976 & Hicks et al. 1991) have increased fine
sediment (silt and clay particles that are <0.0625mm (Woo et al. 1986)) storage
throughout the steelhead’s range (Suttle et al. 2004) resulting in the loss of suitable
habitat. An investigation of juvenile steelhead in a California stream concluded that as
fine-sediment loads increased, the growth of juveniles decreased. This reduced growth is
attributed to the impact of large amounts of fine sediment on the macroinvertebrate
populations which juvenile steelhead primarily rely on for food. An increase in fine
sediment amounts is found to shift these macroinvertebrate communities to be dominated
11

by burrowing taxa that can better escape the impacts of fine sediments. However,
steelhead juveniles only consume organisms that are available in the drift, or main water
column, making these burrowing macroinvertebrates an unusable food source. This lack
of food availability results in decreased growth of steelhead juveniles (Suttle et al. 2004).
Human impacts that cause an increase in fine sediment loads, even at low concentrations
(Suttle et al. 2004) ultimately create habitat conditions that are unsuitable for steelhead
populations, resulting in the loss of habitat available to these populations.
The removal of riparian vegetation along streams and rivers changes the light and
temperature of these waterways, which in turn can impact primary and secondary
production in the stream, as well as timing of the emergence and survival of juvenile
salmonids, including steelhead (Hicks et al. 1991). Increases in water temperatures
beyond those preferable for steelhead can inhibit adults from returning upstream to
spawn, increase risk of disease outbreaks and alter the metabolism of fish, reducing their
efficiency in converting food into energy. An increase in light and temperature can lead
to an increase in primary and secondary production which may allow for an increase in
food for steelhead juveniles. However, this increase in food production is offset by the
detrimental impacts associated with increased stream temperatures (Hicks et al. 1991).
An alteration in stream temperatures can also negatively impact steelhead in
winter months. A lack of riparian vegetation along streams in the winter reduces
insulation potential and can result in the formation of ice and even a “freeze up” in higher
elevation areas. Once air temperatures warm and the ice breaks up, it can scour stream
bottoms, disrupting the recently created redds (Hicks et al. 1991).

12

Through logging practices, as well as other land-use changes, the removal of
riparian vegetation and surrounding trees leads to a decrease in the amount of available
large woody debris (LWD). Large woody debris includes entire trees that fall into or
near a stream, as well as large branches, tree crowns and root balls that enter stream
channels (Sickle & Gregory 1990). LWD has also been removed from streams for
navigation and to reduce property damage during floods. Historically, LWD was
removed from rivers because it was initially believed to inhibit fish migration (Hicks et
al. 1990). However, LWD is now recognized to play an important role in steelhead and
salmonid fish habitat by creating dynamic areas of water movement, with stretches of
faster moving water and deep pools where steelhead can rest and hide from predators
(Roni & Quinn, 2001).
Of all the anthropogenic challenges that threaten steelhead habitat, dams have had
the greatest impact (Collins 1976). Dams create barriers for adult steelhead returning to
spawn and for out-migrating juveniles. Dams also disrupt river flows by creating large
reservoirs in areas of the river in which water used to flow freely, as well as cause
increases in water temperature. Predator and prey dynamics are also altered, food
availability is disrupted and disease rates increase with the presence of dams (Collins
1976).
Dams greatly impact juvenile steelhead, and all other salmonid populations. As
the water passes over the spillway, or through the turbines, the juvenile fish flow with it.
Juveniles that pass through the turbines may be injured or killed by the movement of the
turbine blades themselves or by the high water velocity, turbulence and larger pressure
changes associated with the large volumes of water passing through the turbines (Collins
13

1976 & Petrosky & Schaller 2010). Fish that become injured or disoriented after their
passage over the spillway or through the turbines are highly susceptible to predation from
predators waiting below the dams (Collins 1976 & Petrosky & Schaller 2010). In the
1980’s and 1990’s, turbine screens were installed in order to avoid this dam related
mortality. These screens directed out migrating juveniles to bypass systems or to
collection systems in which juveniles were collected and then transported in tanker trucks
around the dams (Collins 1976). However, these practices still can have detrimental
impacts on juvenile steelhead including high levels of stress and exposure to pathogens
while in holding areas and transport vehicles (Petrosky & Schaller 2010).
In addition to the direct dangers juvenile steelhead face during their seaward
migration, dams also negatively impact habitat and environmental conditions. For
example, a study conducted on a dammed portion of the Columbia River found that
juvenile fish were delayed anywhere from three days to a month before out migrating to
estuaries, making them more susceptible to predation and disease (Collins 1976). Due to
the increased surface area associated with the impounded waters that dams create, an
increase in water temperatures also occurs, which can reach lethal levels during the
summer months. Furthermore, thermal stratification occurs in these large water bodies, in
which warmer waters are located in the top of the water column, while cooler waters sink
to the bottom. This stratification is more detrimental to juvenile steelhead rather than
adults. Due to their small size and inability to swim into deeper waters, juveniles are only
able to occupy these warmer, top areas of the water column, also making them
susceptible to disease and death (Collins 1976).

14

Changes in river habitat are not the only challenges that steelhead face. Poor
ocean conditions are also believed to be a contributing factor in the decline if steelhead
populations. Lower steelhead survival rates have been found to be associated with
warmer ocean waters and reduced spring upwelling (Petrosky & Schaller 2010). Hatchery
practices are believed to also be contributing to steelhead population declines and will be
discussed further below.
Washington State Fish Hatcheries
The first hatchery was established in Washington State in 1895 on the Kalama
River in order to mitigate for large areas of altered habitat. Since then, Washington State
has established a large network of 146 hatcheries that are focused on salmonids and trout,
including steelhead. Eighty-three of these hatcheries are operated by the state, fifty-one
are tribal hatcheries and twelve hatcheries are federally managed (WDFW 2014a).
Hatcheries are now a large part of the state’s economy, providing salmonids and trout for
commercial and recreation fisheries. Currently 88% of steelhead caught in commercial
and recreational fisheries are of hatchery origin (WDFW 2014a). After the listing of
many salmon and steelhead populations under the Endangered Species Act in 1997 and
1998, Washington State hatcheries began supplementation programs in order to boost
wild fish population numbers (WDFW 2014a). The most recent available records
available through the Washington Department of Fish and Wildlife state that in 2013, 5.3
million hatchery steelhead were released in Washington State, with 1.2 million of
released into Puget Sound (WDFW 2014a). Release of hatchery steelhead and salmonids
is still on-going, with the release of millions of fish a year (WDFW 2014a).

15

However, these hatchery practices have been implemented with little
understanding of the impacts of hatchery fish on wild populations (Arkai et al. 2007). A
growing number of scientists and conservation organizations are concerned that these
large releases of hatchery steelhead can have significant impacts on wild steelhead
populations including: negative genetic interactions (Arkai et al. 2007; Kostow 2009 &
Mackey et al. 2001), declining steelhead survivability (Chicolte 2003 & Smith & Li
1983) and significant size differences between hatchery and wild steelhead juveniles
(Abbot 1985; Berejikian et al. 1996; Hill 2006; Keeley &McPhail 1998; Kostow 2009 &
McMichael 1999).
Many organizations have become concerned about the impacts of hatchery fish on
wild steelhead populations, including The Wild Fish Conservancy (WFC). WFC is a nonprofit conservation organization established in Duvall, WA in 1989 which states:
“through science, education and advocacy, WFC promotes technically and socially
responsible habitat, hatchery and harvest management to better sustain the region’s wildfish heritage.” In March of 2014, the WFC filed a lawsuit against the Washington State
Department of Fish and Wildlife (WDFW). The WFC claimed that WDFW was violating
the Endangered Species Act by imperiling wild steelhead, salmon and bull trout recovery
(WDFW 2014b) by operating hatchery programs without hatchery genetic management
plans (HGMP) approved by the National Oceanic and Atmospheric Administration
(NOAA). NOAA approval is required before hatchery programs can be implemented
(WFC 2014).
In April 2014, WDFW and WFC reached an agreement. The only hatchery
released steelhead in Puget Sound will be 180,000 fish into the Skykomish River in
16

Snohomish County in 2014 and 2015 to support the steelhead recreation fishery (WDFW
2014b; WFC 2014 & Yuasa 2014). The remaining steelhead will be released for sport
fishing in Washington lakes that have no connection to Puget Sound (Yuasa 2014 &
WDFW 2014b). The agreement also stated that WDFW will not release winter run
steelhead into Puget Sound rivers until the National Marine Fisheries Service has
reviewed and approved each state hatchery’s HGMP (WDFW 2014b). In addition, a 12
year research study will be conducted on the Skagit River, in which no winter run
hatchery steelhead will be released. This study will allow for the evaluation and possible
establishment of hatchery programs in the Skagit Watershed using wild hatchery stock
(WDFW 2014b). This lawsuit is a clear indication that people are concerned about
hatchery steelhead affecting wild runs, making research focused on these interactions
essential.
Impacts of Hatchery Reared Steelhead on Wild Fish
Current research into hatchery and wild steelhead interactions has focused mainly
on the genetic implications of cross-breeding between these two populations (Arkai et al.
2007; Kostow 2009 & Mackey et al. 2001). Previous research has also focused on
studying the size difference between wild and hatchery smolts at the time of hatchery
release and the dominance and aggression associated with these size differences (Abbot
1985; Berejikian et al. 1996; Hill 2006; Keeley &McPhail 1998; Kostow 2009 &
McMichael 1999). Another emerging need for steelhead research is a closer examination
of steelhead juvenile diets and the differences between wild and hatchery prey items and
feeding behaviors. These areas of steelhead research focused on wild and hatchery
interactions can allow for improved hatchery practices and can inform restoration and
17

conservation efforts to reduce negative wild and hatchery steelhead smolt relationships
and to aid in the removal of Puget Sound steelhead from the ESA’s Threatened Species
List.
Genetics
Research focused on steelhead genetics is important because genetic overlap
between wild and hatchery populations is a major concern. It is believed that hatchery
fish can reduce the fitness and survivability of wild fish when these populations interact
during spawning. Steelhead raised in a traditional hatchery setting, meaning hatchery fish
that are breed with other hatchery raised fish, have showed lower fitness than wild fish
(Arkai et al. 2007). Although there is no clear definition of fitness, as used here, it is
referring to the ability of a species or population to survive and reproduce in the
environment in which it inhabits (Orr 2009). There are three possible explanations for
this observed fitness decline in hatchery fish. The first is the accumulation of deleterious
mutations. Research has shown that in hatcheries, the survival rate of hatchery fish from
the egg to smolt stage is 85-95%, while the survival rate of wild steelhead during the
same time frame has a survival rate of only 1-5% out in the wild (Arkai et al. 2008).
However, in the wild, those fish that had genetic abnormalities would have never
survived to adulthood. Therefore, when hatchery fish are released into the wild
environment, those genetic abnormalities that may have been beneficial or had no effect
while in captivity are not conducive to the environment experienced outside the hatchery
(Akai et al. 2008 & Solberg et al. 2013). This can lead to a decrease in the ability of
hatchery fish to survive and reproduce.

18

The second possible contributor to the decrease in the fitness of hatchery fish is
inbreeding depression. This refers to the reduction in fitness related to mating between
relatives. Some hatchery programs may only have small breeding populations to work
with, causing inbreeding to occur. Inbreeding has been shown to decrease offspring
survival rates by 10-30% (Araki et al. 2008). However, there is an ongoing debate on
these survival rate decreases in regards to the exact genetic causes of fitness decline
(Araki et al. 2008).
The third possible reason for decreased fitness in hatchery steelhead is
domestication selection, in which traits that may be beneficial to steelhead in the hatchery
environment may be detrimental to hatchery populations once they enter the natural
environment. Selection for high growth rates are favored in conventional hatchery
settings (Weber & Fausch 2003) because many hatchery programs raise steelhead to
undergo smoltification at age-1, while the majority of wild steelhead will not advance to
the smolt stage until at least age-2, and often not until age-3 or 4 (Hill et al. 2006 &
Kostow 2009). Thus, hatchery fish are larger at an earlier age. Smolt size has been linked
with improved survival to adulthood therefore hatchery programs produce fish to undergo
early smoltification in order to increase the likelihood of fish surviving to adulthood and
to reduce the cost associated of having to raise juveniles in a hatchery setting for an
additional 2 or 3 years (McMichael et al. 1999). Although having larger smolts at an
earlier age may be appropriate for the purposes of hatchery production it can have
negative impacts on fish once they are released from the hatchery. Larger fish have
higher metabolic rates and so require more food than smaller fish. If hatchery fish are
released into environments with limited food availability or where wild steelhead and
19

other fish species are present, they may be unable to meet their metabolic needs, resulting
in decreased survival (Smith & Li 1983).
Hatchery fish are believed to have lower fitness possibly due to these genetic
effects, making them less likely to survive and reproduce when they are released from the
hatchery environment into the natural environment. Researchers have found indications
that the genetic differences between wild and hatchery fish can have a negative impact on
wild steelhead. A study of Oregon steelhead populations found that when hatchery fish
made up 50% of the spawning population productivity declined (Chicolte 2003).
Productivity refers here to the relationship between spawning adults and their ability to
successfully reproduce offspring that will themselves return to spawn as adults.
Researchers found that when the steelhead population was made up of equal numbers of
wild and hatchery fish that 63% fewer recruits were produced per spawning adult
(Chicolte 2003). This decrease in productivity indicates that hatchery and wild
interactions can lead to a decrease in reproductive success.
Hatchery practices have been developed to limit the genetic interactions between
wild and hatchery steelhead populations, but are not always 100% successful. It is
common hatchery practice to breed hatchery fish so that they spawn months before the
wild fish in order to reduce inbreeding of these two populations (Mackey et al. 2001).
Hatchery fish usually return to spawn three months before wild fish. However, some
population overlap has been observed where hatchery fish return late and mix with wild
populations or wild fish return early and mix with hatchery populations (Mackey et al.
2001). With overlap occurring between wild and hatchery fish it is possible that these
hatchery fish may be uncovering and exposing wild steelhead redds and reducing the
20

chances that the wild steelhead can be successful in their reproduction (Kostow 2009).
Although it was found that spatial interactions occurred between the two populations,
wild fish tended to spawn further up stream than hatchery fish, perhaps limiting the
amount of breeding between wild and hatchery steelhead populations (Mackey et al.
2001).
Resource Competition and Size Differences
Negative ecological effects have been found to be the greatest when wild and
hatchery steelhead share a similar environment for an extended period of time, such as
during freshwater development (Kostow 2009). “Hatchery adults and their juvenile
offspring were [found to] be using resources that could have been supporting wild
populations (Kostow 2009).” As is a usual hatchery practice, large amounts of steelhead
hatchery smolts are released all at one time, which leads to increased food competition
between wild and hatchery fish. Also a common hatchery practice is to release one year
steelhead smolts, while wild steelhead generally smolt around age-3 or 4. Growth rates
are accelerated in order to produce smolts that will be big enough to survive to adulthood
in the most cost effective timeframe (McMichael et al. 1999). By releasing hatchery
smolts so early, there are often higher numbers of residual steelhead (those fish that did
not out-migrate right away but remain in freshwater) in stream systems, adding to the
reduction of resources available to wild smolts (Kostow 2009).
Territory size tends to increase with steelhead size (Keeley & McPhail 1998) and
larger fish tend to be more dominant (Abbot 1985). More dominant fish compete with
smaller, more subordinate fish can consume food at higher rates. This is due to the fact

21

that the subordinate fish have been found to reduce their own feeding rates when a
dominant fish is present, thus allowing the dominate fish to have greater access to the
food that is available (Abbot 1985).
Larger, dominant fish are also found to be more aggressive. These aggressive fish
have more access to the central water column where prey is readily available (Keeley &
McPhail 1998). Fish that are dominant can maintain their status, become larger and
continue to consume more prey (Berejikian 1996) than wild, often smaller steelhead
smolts. Hatchery fish have been found to be dominant to wild fish in most interactions,
except when wild smolts were larger than hatchery smolts (McMichael et al. 1999). This
competition between hatchery and wild steelhead limits resource availability and due to
the aggressive behavior of hatchery fish, wild fish are displaced from preferred feeding
areas. When wild fish are displaced from their original position because of avoidance of
hatchery fish, the wild fish usually stop feeding altogether (McMichael 1999). This
negative interaction found between wild and hatchery fish can harm wild steelhead
populations by limiting the amount of food that these juveniles can consume. A lack of
food can negatively impact steelhead growth, fitness and survivability.
Juvenile Steelhead Diets and Feeding Behaviors
Juvenile Steelhead Diets
Juvenile steelhead diets mainly consist of terrestrial and aquatic
macroinvertebrates (Mistak et al. 2003; Rundio & Lindley 2008 & Simpson et al. 2009)
and smaller fish, often sub-yearling salmonids (Simpson et al. 2009). Because steelhead
diets consist largely of aquatic and terrestrial macroinvertebrates, in all life stages, a brief
22

description of the macroinvertebrate life cycle is necessary before going into the further
specifics of what types of macroinvertebrates juveniles consume.
Macroinvertebrate Life Cycle
Depending on the species, macroinvertebrates undergo one of two metamorphic
life cycles, complete metamorphosis and incomplete, or gradual, metamorphosis (Figure
1) (Lanham 1964). Stoneflies (Plecoptera) and grasshoppers are examples of
macroinvertebrates that undergo incomplete metamorphosis (Lehmkuhl 1979 & Reece et
al. 2011) During incomplete metamorphosis eggs develop into nymphs, which then
develop into adults (Reece et al. 2011). Nymphs resemble adults, sometimes sharing the
same feeding behaviors and habitats, but are smaller and do not have the ability to fly or
reproduce. During this stage nymphs undergo multiple molts in which the wings become
more developed with each molt (Lanham 1964 & Reece et al. 2011). After the molt, the
macroinvertebrate emerges as a full sized, winged, sexually mature adult (Reece et al.
2011).
Figure 1 Diagrams of incomplete and complete metamorphosis (Lanham 1964).

Incomplete Metamorphosis

Complete Metamorphosis

23

Riffle beetles (Coleroptera) and butterflies are examples of macroinvertebrates
that undergo complete metamorphosis (Lehmkuhl 1979). This cycle begins with the egg
stage and progresses through the larva and pupae stages before becoming adults
(Lehmkuhl 1979 & Reece et al. 2011). Unlike incomplete metamorphosis, the young look
nothing like the adults and serve very different functions (Lanham 1964). The main
function of young during the larval stage is to feed and grow as quickly and
“economically” as possible (Lanham 1964). The majority of the actual metamorphosis
occurs during the pupae stage in which the wings and legs develop. At this stage, the
pupae stops feeding and does not move. All the development is occurring internally,
where the larval tissues are replaced by adult tissue. Pupae are often contained within a
protective shell such as a puparium, or a cocoon or chrysalis in the case of moths and
butterflies (Lanham 1964).
Prey Consumption
As steelhead juveniles age their diet preferences change. A study focused on
recently emerged steelhead fry concluded that the two main taxa that fry feed on are
Chironomids (Johnson et al. 2013), also known as midges (Pacharsky et al. 1990) and
baetids (Johnson et al. 2013), commonly referred to as mayflies (Lehmkuhl 1974),
making up 20-42% and 14-34% of fry diets respectively (Johnson et al. 2013). Terrestrial
macroinvertebrates made up 7-18% of steelhead fry diets (Johnson et al. 2013).
One year old juveniles in freshwater mostly consume insect larvae and pupae,
adult insects and amphipods (crustaceans). At age two, steelhead eat the same number of
24

insects, fewer amphipods and begin to eat small fish larvae. When steelhead juveniles are
three years old they eat more fish larvae, consume more mollusks and consume fewer
insects. And at age four, generally the age when steelhead begin to migrate out to sea,
they feed mainly on small fish larvae and adult fish (Juncos et al. 2011).
Feeding Behavior
Juvenile steelhead hold their feeding position by swimming against the current
and moving out of their position to capture prey drifting in the water column (Keeley &
McPhail 1998). When steelhead feeding behaviors and diets were examined in main river
channels and stream side channels, fish consumed almost all the prey available when in
the main channels. However, the steelhead were found to utilize the food sources
available in both the main and side channels, even though there were higher levels of
macroinvertebrates in the main channels than the side channels. The fact that steelhead
consumed macroinvertebrates in both the main and side channels, regardless of insect
levels indicates that steelhead are flexible and will consume whatever prey is available
(Bellmore et al. 2013).
“Feeding rate is a critical factor for survival during stream-rearing and subsequent
life history stages (McCarthy et al. 2009).” Steelhead metabolism is affected by water
temperature and fish body weight. The rate of metabolism determines the level of energy
left for growth (Elliot 1993). Therefore, steelhead juveniles feed on drift in both winter
and summer months because this method utilizes the least amount of energy. By feeding
on what is already available in the drift steelhead are not required to expend energy
actively looking for food resources (McCarthy et al. 2009 & Wright 2010).

25

Wild vs. Hatchery Steelhead Smolt Diets and Feeding Behavior
Wild and hatchery steelhead diets have been closely examined in the Pacific
Northwest and in the Great Lake regions of the United States. Wild steelhead feeding
during the day has been found to be dependent on prey availability in the stream (Elliot
1973). Wild steelhead diets contain a wide range of aquatic and terrestrial insects (Rundio
& Lindley 2008). When hatchery steelhead and wild steelhead are compared, hatchery
residual fish show more surface oriented feeding than wild steelhead and consume more
Hemiptera (true bugs) and Archnida (eight-legged jointed invertebrates) than wild fish.
This likely due to the fact that hatchery fish are used to feeding on the surface in
hatcheries (Simpson et al. 2009). Wild smolts tend to use deeper water and larger
substrate than hatchery fish (Hill et al. 2006). This could cause a difference in diets
between the two populations.
In regards to the amount of prey consumed by wild and hatchery steelhead
smolts, there has been no firm conclusion. Some research has found that there was no
difference in the number of invertebrates consumed by wild and hatchery fish (Goby et
al. 2007), while other research has shown that out-migrating wild steelhead smolts
consumed more prey than hatchery smolts (Simpson et al. 2009). Evidence supporting the
finding that wild smolts consumed more prey than hatchery juveniles points to the
decline in the condition of hatchery fish after release. This decline is believed to be linked
to the inability of hatchery fish to recognize available food, less time spent foraging and
lower feeding efficiency (Simpson et al. 2009 & Weber & Fausch 2003) than wild
steelhead (Simpson et al. 2009). Stomach contents of both wild and hatchery fish

26

consisted of similar taxa, including sub-yearling salmonids, Diptera (true flies),
Tricoptera (caddisflies) and Ephemeroptera (mayflies) (Simpson et al. 2009).
Methods Used in Steelhead Diet Analysis
Resident Drift Collection and Analysis
There are numerous methods that have been used to collect and analyze the prey
that are available for steelhead smolt consumption. The most common method in which
to collect the invertebrates that are drifting through the water column is to stretch drift
nets across rivers and streams to collect anything floating downstream. Because steelhead
stay in the water column, facing upstream, this is an appropriate method in which to
collect prey items. These drift nets are often set along different reaches of the stream,
such as upper, lower and middle reaches. The nets are left in the river for one to twentyfour hours. (Elliot et al. 1973; Johnson 2007; Johnson et al. 2013; Keeley & McPhail
1998 & McCarthy et al. 2009). In order to determine the prey availability on the bottom
of the river Surber samples are collected (Bellmore et al. 2013; Elliot 1973; Johnson
2007; Johnson et al. 2013 & Rundio et al. 2008), while terrestrial samples are either
included in the drift net analysis or pans are set out on the water’s surface to collect prey
that falls into the stream (Rundio et al. 2008).
Once the drift samples are collected, they are placed in a preserving liquid until
the samples can be transported to a lab and analyzed (Keeley & McPhail 1998). Many of
the invertebrate samples are dried (Johnson 2007 & McCarthy et al. 2009) and then
identified to taxon, order and family (Johnson 2007; Keeley & McPhail 1998 &McCarthy
et al. 2009). These are then often divided by functional groups (McCarthy et al. 2009).
27

The most common way to determine prey abundance is to weigh the dried samples and
calculate prey abundance (Elliot 1973 & Johnson 2007). Another method that has been
used is to measure the length of the macroinvertebrates (Keeley & McPhail 1998).
Stomach Content Collection and Analysis
Two common methods are used to determine the prey items consumed by
steelhead. The first method is the use of gastric lavage, which forces water into the
smolt’s mouth and into the stomach, causing them to expel their stomach contents
(Rundio et al. 2008 & McCarthy et al. 2009). The advantages in using gastric lavage
include high rates of prey item removal and high rates of fish survival. Research focused
specifically on steelhead found that 90% of stomach contents were able to be removed
from fish stomachs with gastric lavage. A comparison between hatchery and wild coho
showed that gastric lavage did not have a significant impact on the condition and
survivability of hatchery or wild fish 30 days after stomach flushing (Meehan & Miller
1978). Disadvantages to the gastric lavage technique include the inadequate flushing of
larger and more ridged prey items and the decreased success in removing stomach
contents from larger fish. Larger fish have greater stomach muscle mass resulting in
more difficulty in dislodging prey from their stomachs (Meehan & Miller 1978).
The other commonly used method to collect stomach contents is to euthanize the
smolts and surgically remove their stomachs entirely (Kiffney et al. 2014). It is not
always possible to use this method with ESA listed species such as steelhead due to
permitting and laws regarding “take”. Once the stomach contents have been removed ,

28

the contents are dried and identified using the same methods used with the drift, benthic
and terrestrial samples.
Mass of Stomach Contents
Total amounts of prey in fish stomachs can be calculated using the total wet mass
of stomach contents and expressed as a percentage of body mass. A liner regression
analysis can then be conducted to determine mean percent of stomach content mass
across river reaches and months of data collection. This information is useful because it
allows for the detection of differences between the different river habitats and conditions.
If relationships are found where steelhead in some rivers have a larger proportion of their
body weight as prey than steelhead in other rivers, then further investigations can be
made into the habitat conditions of those rivers. Water temperature, canopy cover, water
velocity and water depth can then be correlated with fish stomach contents.
Another important calculation used is the calculation of the mean number of prey
items found in each stomach and determining the percent composition of those prey items
in relation to the stomach contents for each specific fish sampled. In order to compare the
stomach contents between species, or in the case of this study between hatchery and wild
steelhead, the prey taxa that make up 5% or more of the diet composition in each reach
and in each month can be included in the data analysis (Mistak et al. 2003).
Electivity Indices
In order to determine if juvenile steelhead are consuming prey items base on
preference of certain prey species electivity indices can be used. “Electivity indices
measure the utilization of food types in relation to their abundance or availability in the

29

environment (Lechowicz, 1982).” Foods that make up a larger proportion of the diet than
is available can be considered as preferred. If a food item makes up a smaller proportion
than the food available, that food item can be considered as being avoided. If the
proportions are equal between food items found in the diet and that are available, than
that food item is considered to be eaten at random (Lechowicz, 1982).
Lechowicz (1982) compared three commonly used electivity indices, Ivlev’s,
Jacobs’s and Vanderploeg and Skavia’s . The author found that Ivlev’s and Jacob’s
electivity index can only be used if looking at two prey types, making it unsuitable for
most diet analysis as most species consume numerous types of food. Due to the fact that
steelhead juvenile diets will contain more than one prey species Vanderploeg and
Skavia’s electivity index is the most appropriate.
The Vanderploeg and Scavia electivity index is determined using the following equation:

where

ri= proportion of taxon i in the diet
pi= proportion of taxon in i environment
n= number of kinds of food items
The electivity index ranges from -1 to +1, with a negative number indicating avoidance
of a prey item and a positive number indicating a preference for a specific
macroinvertebrate prey species (Vanderploeg & Scavia, 1979). These preferences can be
calculated and compared for each month of collection, each river reach and each river as
a whole and can be compared between wild and hatchery steelhead populations.

30

There has been comprehensive research in regards to steelhead life histories and
some understanding of the genetic impacts and food resource competition between wild
and hatchery steelhead. There is a basic understanding of steelhead diets but knowledge
in the area of wild and hatchery steelhead diets is still not well known. And the research
that has been conducted, does not generally focus on wild and hatchery feeding
interactions and diets in Puget Sound. This makes investigating the differences between
wild and hatchery steelhead within Washington State and Puget Sound even more
imperative. The diet interactions of wild and hatchery steelhead are investigated in the
study that follows, focusing specifically on the Hood Canal region of Puget Sound.

31

CHAPTER II: Manuscript
Introduction
Steelhead (Oncorhynchus mykiss) populations have declined significantly since
the 1800’s due to overfishing and habitat loss (WFC (b)). Due to this continuous
population decline, with Puget Sound steelhead listed as threatened under the endangered
species act in 2007 (Dept. of Commerce 2007), hatchery programs have been
implemented in Washington State in recent decades in an effort to increase wild steelhead
populations (WDFW 2014a). However, there is a limited understanding of the impacts
that the release of hatchery raised steelhead into Puget Sound rivers have on the wild
population. It is of critical concern to determine if these hatchery steelhead are helping or
harming wild steelhead populations.
Although there has been research in the areas of genetic interactions (Arkai et al.
2007; Kostow 2009 & Mackey et al. 2001) and food resource competition between wild
and hatchery steelhead (Abbot 1985; Berejikian et al. 1996, Hill 2006; Kelley &McPhail
1998; Kostow 2009 & McMichael 1999), there is still a lack of understanding in how
wild and hatchery steelhead diets differ in regards to actual types of food items
consumed. Understanding differences in juvenile diets can provide baseline information
for hatchery supplementation focused on improving wild steelhead populations. Measures
can be taken to allow hatchery programs to more closely resemble the conditions
experienced by wild juveniles in order to ensure better survival of hatchery released
steelhead once they reach freshwater.

32

Understanding steelhead diets can also inform habitat restoration managers and
allow for the implementation of restoration practices that will benefit juvenile steelhead
the most. By comprehending what prey items juvenile steelhead are consuming, efforts
can be made to incorporate habitat elements that will foster an environment where these
prey species can thrive. This will ensure the proper growth and survival rates necessary to
boost wild steelhead populations. Combining beneficial habitat restoration with the
understanding of wild and hatchery steelhead diets and interactions will allow for the
optimal combination necessary to increase wild steelhead populations.
This study focuses specifically on the diets of hatchery and wild steelhead in three
rivers located on the Hood Canal in Washington State: the Dewatto River, the Duckabush
River and the South Fork Skokomish River. These three rivers were specifically
examined due to their inclusion in an ongoing study, the Hood Canal Steelhead project, in
which the impacts of hatchery steelhead on wild steelhead are being studied. Of the six
rivers included in the HCSP the Dewatto, Duckabush and South Fork Skokomish Rivers
are the experimental rivers in which hatchery steelhead are released. In addition, these
three rivers are of varying sizes and allow for diet comparisons between small, medium
and large rivers. This provides a bigger picture of the factors that may impact the diets of
juvenile steelhead. Downstream and upstream diet differences were examined in order to
identify possible differences in diet assemblages due to the differing habitats that are
present in the upper, middle and lower reaches in each of these rivers. Macroinvertebrate
species may differ depending on which reach of the river they occupy. This in turn may
have an impact on the items on which juvenile steelhead feed.

33

Drift in each of the rivers was examined in order to determine the types of items
that are available for steelhead consumption. Preference and avoidance of item types
found in juvenile steelhead diets were also examined. Comparisons were made between
wild and hatchery steelhead juveniles at the river scale, with all reaches being combined.
Comparisons were also made between wild and hatchery steelhead in the lower reaches
of each river.
The findings of this study can be used to implement hatchery practices and habitat
restoration measures that can more effectively improve wild steelhead population
numbers, as well as ensure that the implementation of hatchery programs across
Washington State are not having negative impacts on wild populations.
Methods
Hood Canal Steelhead Project
As previously discussed, this steelhead juvenile diet study is part of a larger
research study, the Hood Canal Steelhead Project (HCSP). The HCSP is an on-going
sixteen year study that began in 2006 (NOAA 2015). The HCSP aims to augment
steelhead populations by supplementing Hood Canal Rivers with juvenile and adult
steelhead over a fixed period of time. There are three supplemented rivers (Dewatto,
Duckabush and South Fork Skokomish) and three control rivers where no
supplementation has occurred (Tahuya, Little Quilcene and Big Beef Creek), although
they are not studied here (Figure 2). This project is an expansion of a 10 year pilot study
conducted on the Hamma Hamma River (Berejikian et al. 2008). The HCSP is a large

34

Figure 2. A map showing the locations of the supplemented (S) and control (C) streams in Hood Canal.
The Hamma Hamma River supplementation was terminated in 2007 (Berejikian et al. study plan draft).

35

partnership with eleven different, federal, state, tribal and non-profit agencies (LLTK
2010).
Supplementation efforts began with the collection of eyed-eggs (the stage at
which the eyes are visible within the eggs) from naturally-occurring steelhead redds on
the supplemented rivers. This differs from traditional hatchery programs where eggs are
collected and fertilized by hand. This method allows for natural spawning and egg
fertilization, creating more genetic diversity than found in traditional hatchery programs
where fertilization is human controlled. The number of eggs collected in each river will
depend on the number of steelhead redds typically found in each of the three rivers. Once
collected, these eggs are then brought to the U.S. Fish and Wildlife Service’s Quilcene
National Fish Hatchery (Dewatto and Duckabush eggs) and Washington State
Department of Fish and Wildlife (WDFW) McKernan Hatchery (South Fork Skokomish
eggs) where they are incubated and tested for pathogens (Berejikian et al study plan
draft).
Once the eggs have hatched, the fry are transported to two rearing facilities, the
Long Live the King’s Lilliwaup Hatchery (Duckabush and Dewatto) and the WDFW
McKernan Hatchery (South Fork Skokomish). Most of these fry will be reared to age-2
smolts and then released into their natal streams. This differs from traditional hatchery
programs in which hatchery smolts are released at age-1. In this study smolts were
released at age-2 in order to produce fish with “a more natural age at smolitification”
(Berejikian et al. 2013). Some juvenile steelhead will be reared to age-4 adults and
released into their natal streams for natural spawning (Berejikian et al. study plan draft).
The desired number of embryos collected and the number of smolts and adults to be

36

released into each of the supplemented rivers are as follows: Duckabush River: 8,620
embryos, 6,667 smolts and 229 adults ; Dewatto River: 9,566 embryos, 7,400 smolts and
253 adults; South Fork Skokomish River: 44,216 embryos, 34,507 smolts and 400 adults
(Berejikian et al. study plan draft). The number of embryos, smolts and adult targets for
each river are related to river size, with the lowest numbers in the small Dewatto River
and the largest numbers in the large South Fork Skokomish River.
The HCSP uses the Before-After-Control-Impact study design to test if
supplementation impacts the productivity, life-history or genetic characteristics of the
wild steelhead populations. Four years prior to the introduction of hatchery steelhead
baseline information was collected on both the supplemented and control rivers (20062010). Supplementation will last for seven years (2011-2018). Post supplementation
monitoring will be conducted for four years following the last supplementation (20192022) on supplemented and control rivers. Pre and post monitoring includes redd surveys
to estimate the number of steelhead spawners, smolt trapping to gather information on
out-migrating juveniles to estimate steelhead productivity, life history monitoring to track
the numbers of anadromous and resident steelhead (rainbow trout), and acoustic
monitoring to gather further information on the number of out-migrating juveniles.
(USDA 2011 a).
From March to July, steelhead redd surveys are conducted on the supplemental
rivers and eggs fertilized from wild spawning adults are collected. Once these eggs are
collected they are raised in similar conditions and feeding rations and growth rates mimic
those of wild steelhead (USDA 2011a). The last egg collections were conducted in May
2014.

37

The HCSP will lead to an increased understanding of the efficacy of conservation
hatcheries to restore wild steelhead populations, potential recovery of threatened
steelhead populations and improved ecosystem functioning (USDA 2011a). The design
and implementation of the HCSP allows for smaller sub-studies to be conducted to
investigate other possible differences or similarities between wild and hatchery steelhead
outside the realm of genetics. The study of juvenile steelhead diets is one such study.
Study Area
The Dewatto, Duckabush and South Fork Skokomish rivers are the focus of this
steelhead juvenile diet study because these are the three rivers within the HCSP that are
undergoing supplementation with hatchery fish. Each of these rivers varies significantly
in elevation, hydrology, water source and canopy cover. The characteristics of Hood
Canal and these rivers (Table 1) are described below.
Table 1. The physical attributes of the Dewatto, Duckabush and South Fork Skokomish Rivers.

Physical Attribute

Dewatto

Duckabush

S.F. Skokomish

14

39.4

44.2

Annual flow rate (m s )

2.01

11.8

21

Depth (m) at invert.
collection sites

0.18 (2010)
0.14 (2011)

0.38 (2010)
0.43 (2011)

0.25 (2010)
0.29 (2011)

Annual Temperature (°C)

9.5

6.8

8.1

River length (km)
2 -1

Hood Canal is an 80 km glacial-carved fjord and is the western most waterway in
the Puget Sound Basin, located within Jefferson, Kitsap and Mason counties of western
Washington State (Berejikian et al. 2013 & HCCC). The watershed is an interactive
system that depends on the continued cycling of clean water and nutrients to maintain its
“biological character” (HCCC).

38

The Dewatto River is a first order stream (WADOE 1998) and has an elevation of
134 m at its headwaters and is 14.0 km in length, however steelhead generally occupy
only the lowest 4.8 km of the river (Berejikian et al. 2013). The Dewatto watershed is
59.6 km2 (Mason County 2011). The river is located in the southwestern area of the
Kitsap Peninsula, draining into Hood Canal at the tidal marsh and mud flats (PNPTC) of
Dewatto Bay. The stream is rain fed with a mean annual flow of 2.01 m3s-1 (Berejikian et
al. 2013), with the highest flows occurring in January and the lowest flows occurring in
September (Collings et al. 1968). During data collection the average water depth taken at
invertebrate collection sites on the Dewatto River was 0.18 m in 2010 and 0.14 m in 2011
(Doctor 2014). The mean annual water temperature is 9.5°C (Berejikian et al. 2013).
Land cover on the Dewatto includes 56% floodplain and riparian zone, 39% forest and
4% wetland (Mason County 2011).
Land use practices along the Dewatto River mainly consist of logging and timber
production and development of residential areas and parks. Forestry makes up 98% of the
land use, with residential and vacant areas making up the other 2%. Land ownership
along the river is 100% private. The Port of Dewatto manages and operates a park area
near Dewatto Bay, while the Manke Timber Company and Pope Resources own and
manage the forest lands (Mason County 2011).
The Duckabush River is a third order stream (WADOE 2011) and has an
elevation of 1724 m at its headwaters and is 39.4 km in length (Berejikian et al. 2013).
The river begins in Olympic National Park on the Olympic Peninsula and enters the
northwestern side of Hood Canal. The Duckabush watershed is 202 km 2 (USFS, 1998).
The Duckabush has a transitional hydrologic regime, meaning the flow is influenced by

39

both rainfall and snowmelt and has a mean annual flow of 11.8 m3s-1 (Berejikian et al.
2013). During data collection the average water depth taken at invertebrate collection
sites was 0.38 m in 2010 and 0.43 m in 2011 (Doctor 2014). The river has a mean annul
water temperature of 6.8 °C (Berejikian et al. 2013). The riparian zone is composed of
66% mixed forest, 25% deciduous trees and shrubs, 5% conifers and 4 % grasses (Correa
2003).
Of the land in the Duckabush River watershed 89% is contained within the
Olympic National Forest and Olympic National Park, while the remaining land use
consists of forest land held by private owners, residential properties and parks (Masello
2013). A fourth of the riparian zone, located below river mile 3 consists of urban and
commercial development, rural residences, roads and dikes (Correa 2003).
The South Fork Skokomish River has an elevation of 1646 m at its headwaters
and is 44.2 km in length (Berejikian et al. 2013). The river begins in the Olympic
National Forest on the Olympic Peninsula and enters the southern Hood Canal after being
joined with the North Fork Skokomish River. The South Fork Skokomish watershed
drains 268 km2 (USDA, 2011 (b) (c)). The South Fork Skokomish also has a transitional
rainfall and snowmelt hydrologic regime, with a mean annual flow of 21 m3s-1
(Berejikian et al. 2013). During data collection the river’s water depth taken at
invertebrate collection sites was an average of 0.25 m in 2010 and 0.29 m in 2011
(Doctor 2014). The river has a mean annual water temperature of 8.1°C (Berejikian et al.
2013). A 1997 analysis of the upper reaches of the South Fork Skokomish found that
there were areas of mature old growth forest as well large clear cut areas (USDA 2011
(b)).

40

The upper portion of the South Fork Skokomish River is managed by the National
Forest Service for commercial and pre-commercial tree thinning, as well as for recreation
use. Recreation uses within the national forest include camping, hiking, horseback riding,
mountain biking, berry picking and hunting. Local tribes also utilize the lands near the
South Fork Skokomish for berry picking, hunting and harvesting of plant materials for
tribal practices (USDA 2011 (b)).
Macroinvertebrate Drift Collection
Macroinvertebrate drift collection was conducted in the upper and lower reaches
of the Dewatto River and the upper, middle and lower reaches of the Duckabush and
South Fork Skokomish rivers. Sampling locations in each of these reaches were in riffle
habitats. Riffle habitats are defined as shallow river sections where water flows over
course sediment to create mild to moderate water turbulence. The current is less than 0.5
m deep and has a flow greater than 0.3 m/s (Woo et al. 1986).
The drift data was collected in August of 2010 and August and September of
2011. Sampling occurred during the late summer months when water flows are lowest
and is the most limiting time for juvenile steelhead food resources in these rivers.
Collecting stomach and drift samples during the low summer flow period provides a
more representative snapshot of juvenile steelhead feeding behaviors and diet
preferences.
Drift nets were placed along a transect that ran the width of the riffle. The nets
were lowered ~5cm above the stream bottom to insure that all drift in the water column
and on the surface were collected, but nothing was sampled from the benthos. The drift
nets were attached and held in place with rebar installed into the stream bottom. The three
41

replicate drift samples were conducted at the same time of day in each river and reach
ranging from 9am to 2pm. A flow meter was used to determine water volume and
velocity through the net. Photos and GPS coordinates were collected at each drift
location. Three replicates of drift sampling were conducted at each river reach location
for two-hour increments in 2010 and one-hour increments during 2011 data collection.
Once the nets had soaked, the entire contents of the net were emptied into a sieve. Large
substrate was removed and finer particles and organic matter were further separated from
the macroinvertebrates. Samples were then placed in a whirlpak with a 95% ethanol
solution.
The drift samples were shipped to AquaticBio labs in Portland, Oregon. Insects
and non-insects were included. The lengths of all macroinvertebrates were measured to
0.5 mm if the organism was less than 5 mm and to the nearest 1mm if larger than 5 mm.
Insect life stages were identified to larvae, pupae and adult. Origins were identified as
aquatic or terrestrial. Nematocera were identified to family and determined to be
terrestrial or aquatic. Brachycera adults were all assumed to be terrestrial in origin.
Chrionomidae were identified to family. Aquatic larvae and pupae were identified to
PNW standard taxonomic effect.
Steelhead Juvenile Stomach Sampling
Juvenile steelhead sampling was conducted as broadly as possible throughout the
upper and lower reaches of the Dewatto River and the upper, middle and lower reaches of
the Duckabush River and South Fork Skokomish River where drift samples were
collected in August 2010 and August and September 2011. Upper and lower reaches of
each fish collection period were marked by GPS. Only steelhead larger than 90 mm were

42

collected using barbless hook-and-line sampling. An attempt was made to sample thirty
wild and thirty hatchery fish during each sampling period within each reach.
Once captured, fish were anesthetized using MS-222. Fork length, weight, rearing
history (wild or hatchery), DNA, scales and diet samples were collected in the field using
the Hood Canal Steelhead Project summer parr and diet sampling protocol methods
established by Berejikian (2010 & 2011). Fish were held in a recovery bucket until the
sampling in each reach was completed. Diet samples were collected by gastric lavage,
using a squirt bottle to flush stomach contents into a funnel that flowed into a separate
sieve. This allowed water to pass through while solid material was maintained. Stomach
contents were placed into whirlpaks containing 95% ethanol solution. Stomach contents
were sent to AquaticBio for analysis, identified and classified in the same manner as the
drift macroinvertebrate samples described in the previous section.
Site Characteristics
Habitat sampling occurred in five random sampling sites downstream of the
riffles in which the drift samples were collected. The start point for the five samples was
randomly determined. At each of the five locations water velocity, water depth, substrate
type, temperature and specific conductivity, turbidity, wetted width, gradient, bank full
width and canopy cover were collected but will not be reported for the purposes of this
study.
Data Analysis
All drift items and stomach content items were classified by order, origin (aquatic
or terrestrial) and life stage (adult, larvae, pupae) when possible. This classification
system was used in all data analyses discussed below.

43

Drift Analysis
Abundance percentages of drift items were calculated for each river at the river
scale, meaning that drift was combined across the upper and lower reaches in the
Dewatto River and across the upper, middle and lower reaches of the Duckabush River
and South Fork Skokomish River. A percentage was calculated for each specific drift
item. Abundance percentages were calculated using the following equation:
100
Pie charts were created for Dewatto, Duckabush and South Fork Skokomish drift to
demonstrate the abundance of each of the specific items in the drift of each river.
Stomach Content Analysis
Percent Abundance
In order to determine the abundance of each specific item found in wild and
hatchery juvenile steelhead stomach contents, percent abundance calculations were made
using the following equation:
100
Numerous percent abundance calculations were made to determine the percent abundance
of all items found in juvenile steelhead stomach contents in each of these three rivers.
Additional percent abundance calculations were made to examine diet differences
between wild and hatchery steelhead juveniles. Further percent abundance calculations
were done at the reach scale in the lower reach of the South Fork Skokomish and
Duckabush River following a significant result during NMS/MRPP analysis. Only the
lower reaches were examined due to low numbers of hatchery fish captured in the upper
and middle reaches.
44

Nonmetric Dimensional Scaling & Multi-response Permutation Procedure
Nonmetric multidimensional scaling (NMS) and multi-response permutation
procedures (MRPP) were conducted on each of the Dewatto, Duckabush and South Fork
Skokomish Rivers to analyze diet overlap between wild and hatchery steelhead in each
river, as well as to analyze diet overlap in the upper, middle and lower reaches of each of
the three rivers (Clarke 1993 & Tagliaferro et al. 2015). Across the three rivers a
combined total of 523 fish were included in the NMS and MRPP analysis, consisting of
454 wild and 69 hatchery steelhead juveniles and a total of 41 different item types were
included (Table 2).
Table 2. The number of hatchery and wild fish included in the NMS/MRPP analysis of the Dewatto,
Duckabush and South Fork Skokomish River. Types of prey items indicates the number of distinct prey
categories included in the analysis.
Wild
Fish

River

Hatchery
Fish

Total
Fish

Types of
prey items

Dewatto

149

35

184

25

Duckabush

133

22

155

29

South Fork Skokomish

172

12

184

Total

454

69

523

27
41
(different
items)

The NMS and MRPP analysis were conducted using PC-ORD 6.0. These
calculations allow for the analysis of diet overlap between wild and hatchery fish and
between river reach. Data were relativized to prevent very abundant prey items from
outweighing less abundant prey items during analysis.
For NMS analysis three 2-D plots were created for each individual river using the
Sorenson’s distance measure. The similarity between items found in stomach contents of
wild and hatchery steelhead were plotted for each river (all reaches were pooled).

45

Similarities between the prey items of each reach in each individual river were assessed
separately. In addition, similarities between wild and hatchery steelhead diets in the lower
reaches of all three rivers were assessed.
Using the Sorenson’s distance measure an MRPP analysis was conducted for each
of the Dewatto, Duckabush and South Fork Skokomish Rivers. Three MRPP analyses
were conducted for each river, one calculation to determine prey item similarities
between wild and hatchery steelhead juveniles and another calculation to determine diet
overlap among river reaches within each river. An additional MRPP was conducted for
the Dewatto, Duckabush and South Fork Skokomish River to determine prey similarities
between wild and hatchery steelhead in the lower river reaches. Only the lower river
reaches were examined due to the fact that those were the only reaches that contained
enough hatchery and wild steelhead for a comparison.
The resulting A-statistic of an MRPP represents the effect size, in this case
showing the amount of similarity between wild and hatchery steelhead diets. The larger
the A value, the more difference there is within the two groups. The significance of
difference is determined by the p value and denotes there is more difference than that
which would be expected by chance (McCune & Grace, 2002). An A value of 1 indicates
that steelhead diets are identical and hatchery diets are identical, while a value of 0
indicates that the differences within dietary groups is not more than expected by chance.
A value less than 0 indicates that there is more difference within groups (i.e. hatchery or
wild) than expected by chance (McCune & Mefford 2011).

46

Electivity Index
Vanderplog and Scavia Electivity Indices (Ei) were calculated for the Dewatto,
Duckabush and South Fork Skokomish Rivers to determine if certain items were
preferred by wild and hatchery origin juvenile steelhead. Due to an inability to determine
exactly where these items were consumed, all river reaches for each river were combined.
Items that were found in stomach contents but not the drift could not be included in this
analysis and were eliminated, as well as items that were found in the drift but not in the
stomach contents. These items were excluded because the item needs to be in both the
drift and diet in order to conduct the Ei calculations. A total of 538 juvenile steelhead
were used in this analysis, 452 wild and 86 hatchery origin fish (Table 3). Due to
different drift compositions the number of prey items included varied by river (Appendix
Table 1).
Table 3 The number of hatchery and wild fish included in the Ei analysis of the Dewatto, Duckabush and
South Fork Skokomish River. Types of items indicates the number of distinct categories included in the
analysis.
Wild
Hatchery
River
Fish
Fish
Total
Orders
Dewatto

149

35

184

32

Duckabush

133

22

155

32

South Fork Skokomish

170

29

199

39

Total

452

86

538

Electivity was determined using the following equation:

where

ri= proportion of prey item i in the diet
pi= proportion of prey item i in the environment
n= number of possible kinds of food items in each river
47

The electivity index ranges from -1 to +1 where a value below 0 indicates
negative electivity (discrimination) for a specific item type and a value above 0 indicates
positive electivity (preference) for a specific item type (Vanderploeg & Scavia, 1979).
Item categorization for Ei analysis was identical to that used in the NMS/MRPP analysis.
Categories included orders, insect or non-insect, aquatic or terrestrial origin and life stage
(adult, pupae, larvae)
The electivity values for each diet item were graphed to show the relationship
between wild and hatchery steelhead diet electivity (Mistak et al. 2003 & Tagliaferro et
al. 2015). However, due to a greater proportion of wild steelhead included in this study,
not all items could be compared between wild and hatchery fish. Furthermore, not all
stomach content items were found across all three rivers preventing the comparison of all
items between rivers.
Results
Drift Composition
The most abundant item found in the drift of all three rivers was aquatic
Ephemeroptera larvae composing 27% of the total drift collected in the Dewatto River,
42% of total drift in the Duckabush and 49% of the total drift in the South Fork
Skokomish River (Figure 3, Appendix Table 1).
In the Dewatto River aquatic Crustacea: Ostracoda and aquatic Diptera larvae
were also a significant proportion of the total drift, each comprising 19% of the total drift.
Aquatic Diptera larvae were also large contributors to the drift in the Duckabush and
48

Skokomish Rivers comprising 31% and 14% respectively. In addition aquatic
Arachnadia: Acari comprised 15% of the total drift in the South Fork Skokomish River
(Figure 3).
The Dewatto River had 33 different drift item types, three of which were not
found in the other two rivers and include aquatic Arthropoda: Arachnadia, aquatic
Crustacea: Cladocera and aquatic Mollusca: Gastrapoda. The Duckabush River had 32
different drift item types and a similar composition of that found in the Dewatto with the
exception of two additional Hemipotera types, terrestrial Hemipotera: Heteroptera larvae
and terrestrial Hemipotera: Auchenorrhyncha adults. The South Fork Skokomish River
was the most variable with 40 different drift item types. The South Fork Skokomish drift
had one more Hemioptera item than the Duckabush, terrestrial Hemipoter: Heteroptera
adults, as well as seven additional item types not found in the other two rivers. These
include Cottidae, aquatic Crustacea: Amphipoda, terrestrial Lepidoptera adults, aquatic
Mollusca: Bivalvia, aquatic Odonata adults, aquatic Plecoptera adults and terrestrial
Thysanoptera adults (Appendix Table 3).

49

Figure 3 Percent abundance of drift items collected in the Dewatto River (A), Duckabush River (B) and
South Fork Skokomish River (C).

A

B

C

Diet Composition
The number of hatchery and wild juvenile steelhead diets examined differed
between the Dewatto, Duckabush and South Fork Skokomish rivers due to different
population sizes in these three river. The Dewatto River had the fewest fish, with only
837. The Duckabush had the most fish captured with 2,022 and the South Fork
Skokomish had the second highest number of fish captured, 1,405. The number of fish
collected differed by river reach, with the majority collected in the lower reaches of all
three rivers.
50

Variability in types of items consumed differed between the Dewatto River,
Duckabush River and South Fork Skokomish River. The steelhead juvenile diets in the
Dewatto River showed a greater variety in the types of items consumed with 45 item
types, while the Duckabush had less variety with 38 types and the South Fork Skokomish
had the least variety, with only 31 (Table 4).
Table 4. The number of wild and hatchery fish included in diet analysis on the Dewatto River, Duckabush
River and South Fork Skokomish River and the number of different diet item types found in stomach
contents. Fish are broken down by reach, upper (U), middle (M) and lower (L) and the drift totals are
combined for all river reaches. (Note: No data was collected in the middle reach of the Dewatto River).
Dewatto

Duckabush

Skokomish

Wild
U
M

360
n/a

546

604

720

420

L

333

612

345

U

13

0

0

20

0

131

124

36

837

2022

1405

45

38

31

Hatchery

M
L
Total # of
juvenile
steelhead
Total # of diet
item types

n/a

Overall Juvenile Steelhead Diet
When hatchery and wild juvenile steelhead stomach contents were combined at
the river scale, the most abundant item in all three rivers was aquatic Ephemeroptera

51

larvae comprising 34% of the total diet in the Dewatto River, 54% in the Duckabush
River and 59% in the South Fork Skokomish River (Figure 4). The second most abundant
item found in stomach contents was aquatic Diptera larvae comprising 33% of total
stomach contents in the Dewatto, 23% in the Duckabush and 13 % in the South Fork
Skokomish Rivers (Figure 4, see Appendix Table 2).
Hatchery Juvenile Steelhead Diets
When hatchery juvenile steelhead were analyzed separately, aquatic
Ephemeroptera larvae were found to be the most abundant item in hatchery fish diets in
the Dewatto and Duckabush rivers, comprising 37% of stomach contents in the Dewatto
and 48% in the Duckabush. Analyses of hatchery fish in the South Fork Skokomish River
showed that 61% of hatchery steelhead diets were comprised of aquatic Ephemeroptera
adults. Aquatic Diptera larvae were the second most abundant item found in hatchery
stomach contents comprising 36% in the Dewatto, 23% in the Duckabush and 10% in the
South Fork Skokomish (Figure 5).
Wild Juvenile Steelhead Diets
When wild steelhead diets were analyzed, aquatic Ephemeroptera larvae were the
most abundant item in stomach contents, consisting of 37% of wild diets in the Dewatto
River, 48% in the Duckabush River and 60% in the South Fork Skokomish River.
Aquatic Diptera larvae were found to be the second most abundant item found in wild
steelhead stomach contents in all three rivers, comprising 36% in the Dewatto River, 24%
in the Duckabush River and 13% in the South Fork Skokomish River (Figure 6).

52

Figure 4 Percent abundance of items collected from Dewatto River (A) Duckabush River (B) and South
Fork Skokomish River (C) wild and hatchery juvenile steelhead stomach contents.

A
B

C

53

Figure 5 Percent abundance of items collected from Dewatto River (A) Duckabush River (B) and South
Fork Skokomish River (C) hatchery juvenile steelhead stomach contents.

A

B

C

54

Figure 6 Percent abundance of items collected from Dewatto River (A) Duckabush River (B) and South
Fork Skokomish River (C) wild juvenile steelhead stomach contents.

A

B

C

Diet Overlap
River Scale
NMS and MRPP analysis showed that in the Duckabush River wild steelhead and
hatchery steelhead diets differed more than would be expected by chance (A=0.1617,
p<0.001), indicating that wild fish and hatchery fish are not consuming the same items.
However in the Dewatto and South Fork Skokomish Rivers wild steelhead and hatchery
steelhead diets did not differ more than expected by chance ((A=0.0005680, p=0.427,
55

A=0.001850, p=0.09612, respectively)) indicating that wild fish and hatchery fish are
consuming similar items (Figure 7).
Reach Scale
NMS/MRPP analysis showed that in the Duckabush and South Fork Skokomish
Rivers juvenile steelhead diets (wild and hatchery pooled) differed more between river
reaches than expected by chance ((A=0.0378, p<0.001, A=0.1238, p<0.001, respectively)
indicating that juvenile steelhead are not consuming the same items in the upper, middle
and lower reaches of these two rivers. Analysis of the Dewatto River showed that
juvenile steelhead diets did not differ between the upper and lower reaches of the river
(A=0.001211, p= 0.1603) (Figure 8).
To further investigate these significant A-value findings, an additional
NMS/MRPP was conducted for the reaches of the Duckabush and South Fork Skokomish
for wild and hatchery fish. However, due to low hatchery numbers in the upper and
middle reaches of both rivers, only the lower reaches were analyzed. Results for the
lower Duckabush reach showed that wild steelhead and hatchery steelhead diets differed
more than expected by chance (A=0.02348, p<0.001), indicating that wild steelhead and
hatchery steelhead are not consuming the same items (Figure 9). Analysis of the lower
South Fork Skokomish reach showed that wild steelhead and hatchery steelhead diets
also differed more than expected by chance (A=0.009190, p=0<0.05), indicating that wild
fish and hatchery fish are not consuming similar items (Figure 9).
Percent abundance calculations for wild and hatchery fish in these two lower
reaches showed that wild steelhead juveniles in the Duckabush River consumed a greater
56

number of items than the hatchery steelhead, consuming 34 and 14 item types,
respectively. Aquatic Ephemeroptera larvae were the most abundant in both diets,
comprising 41% of wild steelhead diets and 42% of hatchery steelhead diets. The second
most abundant item in wild and hatchery diets in the Duckabush and South Fork
Skokomish were aquatic Diptera larvae, comprising 24% and 16% of their diet,
respectively. (Figure 10).
In the South Fork Skokomish River there was also a difference in the number of
prey types found in steelhead juvenile diets, with wild fish consuming 21 item types and
hatchery fish consuming 14 item types. Wild and hatchery steelhead diets consisted of
vastly different compositions of items. Wild juvenile diets consisted mainly of aquatic
larval stages of Ephemeroptera and Diptera, 29% and 24% respectively, while hatchery
fish diets consisted mainly of aquatic adult stages of Ephemeroptera and Diptera, 61%
and 10% respectively. (Figure 11).
In the Dewatto River juvenile steelhead (wild and hatchery) diets did not differ
more than expected by chance between river reaches (A=0.001211, p=0.1603) indicating
that juvenile steelhead are consuming the same items in the upper and lower reaches of
the river (Figure 8).

57

Figure 7 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap between wild and
hatchery origin steelhead juveniles in the Dewatto River (A), Duckabush River (B) and South Fork
Skokomish River (C) using the Bray-Curtis similarity index. The proximity of the symbols indicates degree
of similarity. Solid black triangles indicate hatchery steelhead, hollow triangles represent wild steelhead.

Axis 2

A

A=0.005680
p= 0.427

Axis 2

B

A=0.1617
p <0.001

Axis 2

C

A=0.001850
p= 0.09612

Axis 1

58

Figure 8 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap between river
reaches in the Dewatto River (A), Duckabush River (B) and South Fork Skokomish River (C) using the
Bray-Curtis similarity index. The proximity of the symbols indicates degree of similarity. Hollow circles
indicate the upper reaches, hollow squares indicate the middle reaches, solid circles indicates lower
reaches. (Note that the Dewatto River only contains an upper and lower reach. No data was collected in the
middle reach of the Dewatto River).

Axis 2

A

A= 0.001211
p= 0.1603

Axis 2

B

A= 0.0378
p<0.001

Axis 2

C

A= 0.1238
p<0.001

Axis 1

59

Figure 9 Nonmetric multidimensional scaling (2D) showing the degree of diet overlap between wild and
hatchery steelhead juveniles in the lower reaches of the Dewatto River (A), Duckabush River (B) and South
Fork Skokomish River (C) using the Bray-Curtis similarity index. The proximity of the symbols indicates
degree of similarity. Solid triangle represent hatchery steelhead, hollow triangles represent wild steelhead.

Axis 2

A

A= 0.002667
p = 0.4036

Axis 2

B

A=0.02348
p <0.001

Axis 2

C

A=0.009190
p <0.001

Axis 1

60

Figure 10 Percent abundance of items collected from lower Duckabush River wild (A) and hatchery (B)
juvenile steelhead stomach contents.

A

B

61

Figure 11 Percent abundance of items collected from lower South Fork Skokomish River wild (A) and
hatchery (B) juvenile steelhead stomach contents.

A

B

Electivity Index
Vanderplog and Scavia Electivity Indices (Ei) were calculated for the Dewatto,
Duckabush and South Fork Skokomish Rivers to determine if certain items were
preferred by wild and hatchery origin juvenile steelhead. In this analysis a total of 454
wild and 69 hatchery steelhead juveniles were included (see Table 2 in methods). Due to
the fact that there were significantly more wild juvenile steelhead included in this study

62

than hatchery steelhead it is difficult to conclusively state diet preference differences
between wild and hatchery juvenile steelhead.
Hatchery Juvenile Steelhead Electivity
Nonetheless, through this analysis a few patterns did develop across all three
rivers. There was a distinct difference between the preferences of hatchery fish in the
Dewatto, Duckabush and South Fork Skokomish Rivers. In the Dewatto River, hatchery
fish preferred 100% of the available items (Figure 12), while in the Duckabush (Figure
13) and South Fork Skokomish River (Figure 14) hatchery fish showed more variation in
preference of item types. For example, of the 29 available items in the Duckabush River
included in this analysis, hatchery steelhead preferred 6 of the items indicating a
preference for only 20% of the items (Figure 13). Similarly of the 27 available items in
the South Fork Skokomish River, hatchery steelhead preferred 3 of the items indicating a
preference for only 11% of the items (Figure 14).
In regards to preferential items, no patterns emerged between the hatchery
steelhead across all three rivers. There was not one item that was preferred or not
preferred consistently across the Dewatto, Duckabush and South Fork Skokomish Rivers
Figure 12-14). However, this may be due to the fact that there are significantly fewer
hatchery steelhead included in this analysis.
Wild Juvenile Steelhead Electivity
There was a distinct difference between preferences of wild steelhead in the three
rivers. In the Dewatto River, wild fish preferred 68% of the available items (Figure 12),

63

while in the Duckabush River (Figure 13) and South Fork Skokomish River (Figure 14)
wild steelhead showed more variation in preference of items. Of the 29 available items in
the Duckabush River wild steelhead preferred 6 of the available items indicating a
preference for only 10% of the items (Figure 13). Of the 27 available items in the South
Fork Skokomish River wild steelhead only preferred 8 prey items indicating a preference
for 30% of the items (Figure 14). A few patterns did emerge in the preference of items by
wild steelhead across all three rivers. Wild steelhead did not prefer aquatic
Ephemeroptera larvae, aquatic Diptera (pupae, larvae or adult) or aquatic Coleoptera
adults. However, wild steelhead in the Dewatto, Duckabush and South Fork Skokomish
River did show a preference for aquatic Ephemeroptera adults (Figure 12-14).

64

Figure 12 Vanderplog and Scavia Electivity Index of item selectivity for wild and hatchery steelhead in the
Dewatto River. Includes all items found in both the drift and steelhead stomach contents. Grey bars
represent wild steelhead, black bars represent hatchery steelhead.

65

Figure 13 Vanderplog and Scavia Electivity Index of item selectivity for wild and hatchery steelhead in the
Duckabush River. Includes all items found in both the drift and steelhead stomach contents. Grey bars
represent wild steelhead, black bars represent hatchery steelhead.

66

Figure 14 Vanderplog and Scavia Electivity Index of item selectivity for wild and hatchery steelhead in the
South Fork Skokomish River. Includes all items found in both the drift and steelhead stomach contents.
Grey bars represent wild steelhead, black bars represent hatchery steelhead.

67

Discussion
River Scale
Drift
There was greater variety of drift items in the South Fork Skokomish River
compared to the Duckabush River and Dewatto River. This difference could possibly be
due to a difference in river size. The South Fork Skokomish River is the largest of the
three rivers, with a length of 44.2 km and encompassing the largest watershed area of 268
km2. Because of its larger size, it is possible that the South Fork Skokomish River is
receiving a larger input of drift items from lower order streams, further up in the
watershed. As these smaller streams merge together and form the larger South Fork
Skokomish River, the drift items from those streams are transported into the larger river
(Wipfli & Gregovich 2002). These drift characteristics play an important role, as drift
items are the main food source for juvenile steelhead and can greatly influence the diet
composition of juvenile steelhead in these three rivers.
Diet Composition
The diet composition of the wild and hatchery steelhead juveniles in the Dewatto,
Duckabush and South Fork Skokomish is consistent with results found in other studies
focused on juvenile steelhead diets, with the most abundant items consisting of
Ephemeroptera aquatic larvae and Diptera aquatic larvae (Godby et al. 2007; Johnson
2007 & Mistak et al 2003). This indicates that the juvenile steelhead in the these three
rivers are consuming items that are characteristic of juvenile steelhead diets worldwide.
However, there was a noticeable difference in juvenile steelhead diet composition across
all three rivers. The juveniles in the Dewatto River consumed a greater variety of item
68

types (72 types) than in the Duckabush River (56 types) and the South Fork Skokomish
River (42 types). It is not entirely clear why the juvenile steelhead in the Dewatto River
are selecting a greater variety of prey items than on the Duckabush and South Fork
Skokomish Rivers. However, the size of the Dewatto River, as well as water depth, water
velocity and water temperature, in comparison to the other two rivers, could explain this
difference in juvenile fish diets.
The Dewatto River is significantly smaller, with a length of 14 km, while the
Duckabush and South Fork Skokomish Rivers have a length of 39.4 and 44.2 km,
respectively. In addition, during the months of data collection, the Dewatto River had the
lowest stream depth and slowest water velocity when compared to the Duckabush River
and South Fork Skokomish River. These differences in stream characteristics could
influence the feeding behaviors of hatchery and wild fish in the Dewatto River. Due to
the possible lower abundance of drift items available, because of low flow conditions, the
steelhead juveniles have access to fewer resources and so are forced to eat all available
item types. These may include items that the steelhead in the Duckabush River and South
Fork Skokomish River do not often eat because these rivers are not as impacted by low
flow conditions and so have more items available. This allows the steelhead in these
rivers to be more selective about what items they eat.
In each of the three rivers, there were more item types found in juvenile steelhead
stomach contents than were found in the drift. The largest discrepancy was in the
Dewatto River where only 33 item types were found in the drift, while 72 item types
were found in juvenile steelhead stomach contents. A similar pattern was found in the
other two rivers. The Duckabush River had 32 item types found in the drift and 56 item

69

types found in steelhead stomach contents. In the South Fork Skokomish River there
were 40 item types found in the drift and 42 item types found in steelhead stomach
contents. The most likely cause for this discrepancy is that terrestrial insects that entered
the river were eaten right away by the juvenile steelhead, thus not making it into the drift.
This likely explains why there was a larger difference between the number of drift and
stomach content item types in the Dewatto River. Because the Dewatto River is much
smaller than the other two rivers, it has a narrower stream channel with more
overhanging vegetation, allowing for more terrestrial inputs into the stream. If juvenile
steelhead are eating terrestrial inputs as soon as these items enter the river and the
Dewatto has larger terrestrial inputs, due to its small size, then it would be expected that
the steelhead in this river would have a larger variety of item types in their stomach
contents than is found in the drift.
In all three rivers there was a noticeable difference in diet composition between
wild and hatchery steelhead juveniles. Wild fish consumed a greater variety of item types
than hatchery fish. In the Dewatto River, wild fish consumed 41 item types, while
hatchery fish only consumed 32 types. This pattern holds true for the other two rivers,
with Duckabush wild fish consuming 38 types and hatchery steelhead consuming 18. In
the South Fork Skokomish wild fish consumed 29 types and hatchery fish consumed 13
types. It should be noted that there were far more wild steelhead analyzed in this study
than hatchery fish, so some of these patterns may be related to the size difference
between these two sample groups and may not fully reflect diet differences.
When looking at differences across all three rivers using the MRPP, the only
significant difference at the river scale was found in the Duckabush River. There are a

70

number of possibilities for these significant findings. One explanation may be that the
Duckabush had the highest abundance of fish collected. This larger sample size may have
allowed for the detection of a diet difference. In addition, in the Duckabush River, like all
the rivers in this study, the majority of fish were collected in the lower reach. The lower
reach of the Duckabush was also found to be significant between wild and hatchery diets,
making it likely that the higher steelhead abundance in the lower reach is driving the
pattern for the whole river.
Another possible explanation may be a difference in the rearing habitats
experienced by wild and hatchery raised fish. Wild fish feed on live prey right away once
they emerge as fry, but hatchery fish are pellet fed and don’t experience live food until
they are released into the wild (Simpson et al. 2009 & Weber & Fausch 2003). It has
been suggested that hatchery fish juveniles may be unable to recognize available food
sources and have been found to spend less time foraging and have lower feeding
efficiency than their wild counterparts (Elliot 1973). This earlier experience with natural
prey consumption may explain why wild steelhead are consuming such a wider array of
prey items compared to the hatchery reared fish. This may also explain why wild fish
consumed more larvae than adults in all three rivers, compared to hatchery fish who
consumed more adult life stages. Larvae are more likely to be found in the drift than adult
life stages of macroinvertebrates, who are located at or on the water surface. Because
hatchery fish are accustomed to surface feeding, they go after the items located on top of
the water, rather than items down in the water column, where it is likely more young life
stages are available.

71

It is also possible that wild fish are eating a wider variety of items than hatchery
fish due to competition. Hatchery fish are often larger and more aggressive than wild fish
(Keeley & McPhail 1998). Studies focused specifically on competitive interactions
between wild and hatchery steelhead have found that hatchery fish limit wild fish food
resource availability due to the more aggressive behavior of hatchery fish (McMichael
1999). In the case of the wild and hatchery interactions in the Dewatto, Duckabush and
South Fork Skokomish rivers, wild fish may be forced to consume different, perhaps less
desirable, items because hatchery fish are holding the optimum feeding positions and
therefore consuming the optimum, less available, prey. Although size data was collected
it was not analyzed in this specific study. Further research is necessary in order to clearly
determine if competition is occurring between wild and hatchery steelhead juvenile in
these three Hood Canal Rivers.
Reach Scale
A significant difference was found between wild and hatchery diets on the lower
reaches of the Duckabush River and South Fork Skokomish River. These differences may
be due to the fact that the number of hatchery and wild fish captured in the three reaches
of these rivers were very different. In the Duckabush River there were no hatchery fish
captured in the upper reach and most hatchery fish were captured in the lower reach.
Similarly, on the South Fork Skokomish River hatchery fish were only captured in the
lower reach. The high abundance of hatchery fish in these lower reaches may have
provided a large enough sample size to show a trend, although the number of hatchery
fish captured in all three rivers was still much lower than wild fish.

72

Another possible factor that may have led to the significant findings in the lower
reaches of these two rivers may be that the lower reaches have a larger input of items
from upstream. Similar to the idea when discussing drift (Wipfli & Gregovich 2002)
when those items from the upper reach are combined with the middle and lower reach, a
wider variety of items are present (Wipfli & Gregovich 2002). .
Electivity
Despite the high abundance of aquatic Ephmeroptera and aquatic Diptera larvae
in juvenile steelhead diets and the drift of all three rivers, these items both received
negative electivity index values, with the exception of hatchery fish in the Dewatto River
(which showed positive electivity for these two prey items). These negative electivity
values indicate that despite high abundance in the drift, as well as high abundance in
diets, steelhead juveniles selected against Ephemeroptera and Diptera aquatic larvae.
These findings are similar to that reported by Mistak et al. (2003) in a study
focusing on juvenile steelhead diets in the Pine River. The authors found a similar
situation in which items that were highly abundant in the drift and highly abundant in
steelhead stomach contents were being selected against. They concluded that the negative
electivity values were not necessarily indicating avoidance of these items but that the
items were underutilized. A similar conclusion could be reached in the case of the
Dewatto, Duckabush and South Fork Skokomish diet and drift composition. Although
high abundance of Diptera and Ephemeroptera aquatic larvae exists in the drift and diet,
in relation to all the other items consumed these two items are not being consumed in
equal proportion to their availability. However, more research is necessary in this area to

73

more clearly understand this discrepancy between high drift and diet abundance but low
electivity values of these two items.
Due to the large discrepancy in the number of wild and hatchery fish included in
this study it is hard to distinguish specific patterns in regard to electivity index (Ei)
values. However the distinctive pattern between the Ei values of items consumed by
hatchery steelhead in the Dewatto River was noticeable. Hatchery steelhead on this river
have only positive Ei values, indicating that these hatchery fish selected for every single
available item, while in the Duckabush and South Fork Skokomish River there was more
variability in the preference and avoidance of the same items. Wild fish in the Dewatto
also showed more preference than wild steelhead in the other two rivers.
There are a number of factors that may have contributed to the positive electivity
values for both hatchery and wild steelhead in the Dewatto River. One factor may be the
habitat availability in the Dewatto. Because data was collected during late summer, when
stream flows are the lowest, there may be less habitat available to the fish in this smaller
system, forcing hatchery and wild fish to be confined to a smaller area and to compete for
the same limited food resources. This competition for food sources could be occurring in
two different ways. Hatchery fish could be outcompeting wild fish (McMichael 1999),
causing the wild steelhead to consume the less desirable (more discriminated against)
items. On the other hand, it could be that wild fish are consuming the more desirable (less
discriminated against) items, forcing hatchery fish to show a preference for all items.
Further Research
In order to delve deeper into some of the results found in this study there are three
main areas in which more research is required. The most important finding that needs

74

further investigation is the relationship between age and size differences between the wild
and hatchery steelhead sampled in these rivers. By comparing the size and age
differences between wild and hatchery fish, the idea that competition may be occurring
between the two can be further substantiated.
Further investigation into the differing habitat characteristics along the entire
lengths of the Dewatto, Duckabush and South Fork Skokomish rivers is also necessary.
Looking at canopy cover, water temperature, turbidity, conductivity, substrate type and
gradient (all of which were gathered during sample collection) may provide more insight
into the drift and diet differences found in these three rivers. Using this information, it
may be possible to determine exactly why the larger South Fork Skokomish River had a
higher variety of drift items. These habitat characteristics could also be used to determine
the diet differences observed between these three rivers.
Finally, as mentioned previously, more investigation is needed in order to explain
the negative Ei values for the diet and drift items that were found to be the most abundant
in all three rivers. Gaining a deeper understanding of this discrepancy will provide a
clearer understanding of hatchery and wild juvenile steelhead diet preferences as well as
investigate the usefulness of the Ei itself in diet analysis.

75

CHAPTER IV: Conclusion
There were two key findings in this study that were the most noteworthy. The
first key finding of importance was the large variety of items in juvenile steelhead diets in
the Dewatto River, while the Duckabush and South Fork Skokomish juvenile steelhead
diets showed less variability. This finding is important because it may reveal a significant
relationship between stream size and drift composition and how these two factors
influence feeding behaviors of both hatchery and wild steelhead. Further research in this
specific area will allow for a clearer understanding of the different diet requirements of
steelhead in different sized rivers. Finding a link between stream size and diet may help
inform management practices and allow for the greater success of steelhead juveniles in
these lower order streams.
If steelhead juveniles are indeed relying on a larger proportion of the available
drift items in small order streams, measures can be taken to ensure that these stream
habitats are able to support these macroinvertebrate populations in the future. Climate
change and other anthropogenic changes, such as alteration in land use practices, can
negatively impact these ecosystems. In Washington State, climate change models are
predicting increased precipitation levels in the winter, with more falling as rain rather
than snow, and hotter and drier summers (Leung et al. 2004). This reduction in
precipitation and large increase in temperatures could have great impacts on these smaller
streams, creating even more drastic low flow conditions in the late summer months.
Growing populations in the Puget Sound area as well will likely result in land use change,
possibly resulting in reduced stream flows (Konrad & Booth 2002).

76

By anticipating these changes and incorporating them into current and future
management plans, measures can be taken to maintain and/or create habitats in which low
flow adapted macroinvertebrates can thrive and continue to be a food source for juvenile
steelhead. These could include ensuring that there are enough pool areas available for
juvenile steelhead and their food sources during low flow conditions, as well as ensuring
that the hyphoreic zone is functioning properly and allowing for groundwater intrusion to
keep stream sediments moist (Boulton et al. 1998). Many of the food sources that
steelhead are eating in the Dewatto River have life history stages that are already adapted
to flow conditions (Bunn & Arlington 2002). These macroinvertebrate species rely
heavily on pools and burrowing into moist sediments for refuge (Williams & Hayes
1977). By creating areas where these macroinvertebrate and other food sources can
survive, it will ensure that juvenile steelhead have access to necessary food sources.
The second finding of key importance was the significant diet differences
between hatchery and wild steelhead juveniles in the Duckabush River, both at the river
and reach scale. Depending on what further research reveals about the relationship
between wild and hatchery steelhead diets in the Dewatto, Duckabush and South Fork
Skokomish rivers, changes can be implemented to current hatchery program practices.
If the diet differences between wild and hatchery fish are due to rearing
environments, measures can be taken to ensure that hatchery raised steelhead are exposed
to live food items while still in the hatchery setting, as opposed to the current use of
pellets, delivered at the water surface. If the management goal is ultimately to boost
steelhead populations in Puget Sound, then steps need to be taken to ensure that once
released, these hatchery fish have the best chance of survival. Providing live prey items

77

while in the hatchery may be a way in which to ensure that these hatchery fish are
adapted to the conditions experienced by wild steelhead, thus improving rates of survival.
Previous research in this area has shown that the exposure of hatchery raised
salmonids to live prey while still in a hatchery setting are better able to forage for novel
live prey items (Brown et al. 2003 & Maynard et al. 1994). However, this only occurred
in settings where there was the addition of live prey as well as the addition of habitat
complexity (i.e. tanks with rocks, wood and temperature variations). A similar approach
could be used in the hatchery management plans in Washington State and could prove to
be extremely beneficial in improving the post-release survival of hatchery steelhead
(Brown et al. 2003 & Maynard et al. 1994), especially if Hood Canal rivers undergo
hydrologic changes in the future. Hatchery fish that are able to identify novel prey items
will be able to adapt more easily to changes in macroinvertebrate assemblages, making
these fish more resilient. Ultimately hatchery fish with previous exposure to live food
items will likely fare better and may be able to improve Puget Sound steelhead
population numbers.
By looking deeper into the causes for the significant diet differences between wild
and hatchery steelhead on the Duckabush Rivers, such as size and age differences
between wild and hatchery fish and habitat characteristics that may be influencing drift
and fish diet, a clearer understanding will be reached. It is imperative that fisheries
managers and restoration ecologists work together to produce hatchery fish that are better
adapted to natural environments and to create environments that promote healthy
macroinvertebrate populations. Hatchery managers need to create programs in which
hatchery fish are exposed to live prey items before release into streams to ensure that they

78

will be able to feed efficiently and effectively in these wild environments. This will
ultimately ensure the fitness of these animals in the wild. Restoration ecologists need to
incorporate habitat structures and elements, such as pools and a healthy hyphoreic zone,
to ensure that the food sources that juvenile steelhead rely on remain in these freshwater
ecosystems.
The implementation of these management strategies is imperative and should
begin as soon as possible. In addition, the ecological interactions between wild and
hatchery steelhead, and salmonids in general, need to be more clearly understood.
Despite the implementation of hatchery programs throughout Washington State in recent
decades, steelhead populations have continued to decline, indicating that current
management practices are not effective. A change in hatchery management plans that
puts hatchery and wild steelhead on equal footing may be the only way for steelhead
populations to bounce back.

79

References
Abbott, J.C., Dunbrack, R.L., & Orr, C.D. (1985). The interaction of size and experience
in dominance relationships of juvenile steelhead trout (Salmo gairdneri).
Behavior, 92, 241-253.
Araki, H., Ardren, W.R., Olsen, E., Cooper, B., & Blouin, M.S. (2007). Reproductive
success of captive-bred steelhead trout in the wild: Evaluation of three hatchery
programs in the Hood River. Conservation Biology, 21, 181-190.
Araki, H., Berejikian, B.A., Ford, M.J. & Blouin, M.S. (2008). Fitness of hatchery-reared
salmonids in the wild. Evolutionary Applications, 1, 342-355.
Anderson, S. (Director). (2014). Wild Reverence: The Wild Steelhead’s Last Stand
[Motion Picture]. United States: North Fork Studios.
Bellmore, J.R., Baxter, C.V., Martens, K., & Connolly, P.J. (2013). The floodplain food
web mosaic: A study of its importance to salmon and steelhead with implications
for their recovery. Ecological Applications, 23, 189-207.
Berejikian, B.A. (2010). Hood Canal Steelhead Project summer parr/diet sampling
protocol.
Berejikian, B.A. (2011). Hood Canal Steelhead Project summer parr/diet sampling
protocol.
Berejikian, B.A., Campbell, L.A. & Moore, M.E. (2013). Large-scale freshwater habitat
features influence the degree of anadromy in eight Hood Canal Oncorhynchus
mykiss populations. Canadian Journal of Fisheries and Aquatic Sciences 70, 756765.
Berejikian, B.A., Endicott, R.S., Lee-Waletermire, J. (2008). Increase in steelhead
(Onchorhynchus mykiss) redd abundance resulting from two conservation
hatchery strategies in the Hamma Hamma River, Washington. Canadian Journal
of Fisheries and Aquatic Sciences 70, 754-764.
Berejikian, B.A., Mathews, S.B., & Quinn, T.P. (1996). Effects of hatchery and wild
ancestry and rearing environments on the development of agnostic behavior in
steelhead trout (Oncorhynchus mykiss) fry. Canadian Journal of Fisheries and
Aquatic Sciences, 53, 2004-2014.
Berejikian, B.A., Vandoornik, D., Bush, R., Petrie, M., Tezak, S., Endicott, R., Tatara, C.
& Flagg, T (n.d.). Hood Canal Steelhead Supplementation Project DRAFT Study
Plan.
Boulton, A.J., Findlay, S., Marmonier, P., Stanley, E.H. & Vallet, H.M. (1998). The
80

functional significance of the hyphorheic zone in streams and rivers. Annual
Review of Ecological Systems, 29, 59-81.
Brown, C., Davidson, T. & Laland, K. (2003). Environmental enrichment and prior
experience of live prey improve foraging behavior in hatchery-reared Atlantic
salmon. Journal of Fish Biology, 63, 187-196.
Bunn, S., & Arlington, A. (2002). Basic principles and ecological consequences of
altered flow regimes for aquatic biodiversity. Environmental Management. 30,
492-507.
Busby, P.J., Wainwright, T.C., Bryant, G.J., Lierheimer, L.J., Waples, R.S., Waknitz,
F.W. & Lagomarsino, I.V. (1996). Status Review of West Coast Steelhead from
Washington, Idaho, Oregon and California (NMFS-NWFSC-27), Seattle, WA:
National Marine Fisheries Service.
Chicolte, M.W. (2003). Relationship between natural productivity and the frequency of
wild fish in mixed spawning populations of wild and hatchery steelhead
(Oncorhynchus mykiss). Canadian Journal of Fisheries and Aquatic Sciences, 60,
1057-1067.
Clarke, K.R. (1993). Non-parametric multivariate analyses of changes in community
structure. Australian Journal of Ecology, 18, 117-143.
Collings, M.R., Smith, R.W., & Higgins, G.T. (1968). The hydrology of four stream in
western Washington as related to several Pacific salmon species. United States
Department of the Interior Geological Survey.
Collins, G.B. (1976). Effects of dams on Pacific Salmon and Steelhead Trout. Marine
Fisheries Review, 39-46.
Correa, G. (2003). Salmon and steelhead limiting factors water resource inventory Area
16 Dosewallips-Skokomish Basin. Report prepared for Washington State
Conservation Commission.
Department of Commerce. (2007). Endangered and threatened species: Final listing
determination for Puget Sound steelhead. Federal Register, 72, 26722-26735.
DFW306243 (2012)
Doctor, K. (2014). [Hood Canal Steelhead Project]. Unpublished raw data.
Elliot, J.M. (1973). The food of brown and rainbow trout (Salmo trutta ans S. gairdneri)
in relation to the abundance of drifting invertebrates in a mountain stream.
Oecologia, 2, 329-347.

81

Gilbert, C.R. & Williams, J.D. (2002). National Audubon Society field guide to fishes.
Alfred A. Knopf: New York.
Godby, N.A., Rutherford, E.S. & Mason, D.M. (2007). Diet, feeding, growth, mortality
and production of juvenile steelhead in a Lake Michigan tributary. North
American Journal of Fisheries Management, 27, 578-592.
Hicks, B. J., Hall, J. D., Bisson, P. A. & Sedell, J. R. (1991). Responses of salmonids to
habitat changes. Journal of Applied Ichthyology, 29, 221-225.
Hill, M.S., Zydlewski, G.B. & Gale, W.L. (2006). Comparisons between hatchery and
wild steelhead trout (Oncorhynchus mykiss) smolts: Physiology and habitat use.
Canadian Journal of Fisheries and Aquatic Sciences, 63, 1627-1638.
Hood Canal Coordinating Council. (n.d.). Hood canal geography. Retrieved from
http://hccc.wa.gov/About+Us/HoodCanalGeography/default.aspx.
Johnson, J.H. (2007). Comparative diet of subyearling Chinook salmon (Onchorhynchus
tshawytscha) and steelhead (O. mykiss) in the Salmon River, New York. Journal
of Great Lakes Research, 33, 906-911.
Johnson, J.H., McKenna, J.E. & Douglass, K.A. (2013). Movement and feeding ecology
of recently emerged steelhead in Lake Ontario tributaries. Journal of Applies
Ichthyology, 29, 221-225.
Juncos, R., Milano, D. Macchi, P.J., Alonso, M.F., & Vigliano, P.H. (2011). Response of
rainbow trout to different food web structures in northern Patagonia: Implications
for growth, bioenergetics, and invasiveness. Transections of the American
Fisheries Society, 140, 415-428.
Keeley, E.R. & McPhail, J.D. (1998). Food abundance, intruder pressure and body size as
determinants of territory size in juvenile steelhead trout (Oncorhynchus mykiss).
Behavior, 135, 65-82.
Kiffney, P.M., Buhle, E.R., Naman, S.M. Pess, G.R. & Klett, R.S. (2014). Linking
resource availability and habitat structure to stream organisms: An experimental
and observational assessment. Ecosphere, 5, 1-27.
Konrad, C.P. & Booth, D.B. (2002). Hydrologic trends associated with urban
development
in selected streams in the Puget Sound Basin, Western Washington. (Water
Resources Investigations Report 02-4040). Tacoma, WA: U.S. Geological
Survey.
Kostow, K. (2009). Factors that contribute to the ecological risks of salmon and

82

steelhead hatchery programs and some mitigating strategies. Reviews of Fish
Biology and Fisheries, 19, 9-31.
Krueger, C.C. & May, B. (1991). Ecological and genetic effects of salmonid
introductions in North America. Canadian Journal of Fisheries and Aquatic
Sciences, 48, 66-77).
Lanham, U.N. (1964). The Insects. New York: Columbia University Press.
Lechowicz, M.J. (1982). The sampling characteristics of electivity indices. Oecologia,
52, 22-30.
Lehmkuhl, D.D. (1979). How to know the aquatic insects. Dubuque, IA: Wm. X. Brown
Company Publishers.
Leung, L.R., Qian, Y., Bian, X., Washington, W.M., Han, J., & Roads, J.O. (2004). Midcentury ensemble regional climate change scenarios for the western United States.
Climate Change, 62, 75-113.
Long Live the Kings. (2010). The Hood Canal Steelhead Project. Retrieved from
http://www .lltk.org/rebuilding-populations/hcsp/summary.
Mackey, G., McLeab, J.E. & Quinn, T.P. (2001). Comparisons of run timing, spatial
distribution and length of wild and newly established hatchery populations of
steelhead in Forks Creek, Washington. North American Journal of Fisheries
Management, 21, 717-724.
Mason County (2011). Mason County shoreline and characterization report: June 2011
draft.
Masselo, D.M. (2013). Variations in surface soil organic carbon at the Duckabush River
Delta, Washington. (Published master thesis). The Evergreen State College,
Olympia, Washington.
Maynard, D.J., Flagg, T.A. & Mahnken, V.W. (1994). A review of seminatural culture
strategies for enhancing the postrelease survival of anadromous salmonids.
American Fisheries Society Symposium, 15, 307-314.
McCarthy, S.G., Duda, J.J., Emlen, J.M., Hodgson, G.R., & Beauchamp, D.A. (2009).
Linking habitat quality with trophic performance of steelhead along forest
gradients in the South Fork Trinity River watershed, California. Transections of
the American Fisheries Society, 138, 506-521.
McCune, B. & Grace, J. (2002). Analysis of Ecological Communities. MjM Software,
Gleneden Beach, OR.

83

McCune, B. & Mefford, M.J. (2011). PC-ORD multivariate analysis of ecological data.
Version 6. MjM Software, Gleneden Beach, Oregon.
McMichael, G.A., Pearsons, T.N. & Leider, S.A. (1999). Behavioral interactions among
hatchery-reared steelhead smolts and wild Onchorhynchus mykiss in natural
streams. North American Journal of Fisheries Management, 19, 948-956.
McMillan, B. (2006). Historic Steelhead Abundance: Washington NW Coast and Puget
Sound. Wild Fish Conservancy, Duvall, WA.
Meehan, W.R. & Miller, R.A. (1978). Stomach flushing: Effectiveness and influence on
survival and condition of juvenile salmonids. Journal of Fishers Research Board
of Canada, 35, 1359-1363.
Mistak, J.L, Hayes, D.B. & Bremigan M.T. (2003). Food habitats of coexisting
salmonines above and below Stronach Dam in the Pine River, Michigan.
Environmental Biology of Fishes, 67, 179-190.
National Oceanic and Atmospheric Administration (2011). 5 year review: Summary and
evaluation of Puget Sound Chinook Hood Canal Summer Chum Puget Sound
Steelhead. Portland, OR: National Marine Fisheries Service.
National Oceanic and Atmospheric Administration (2014). Steelhead. Retrieved from
http://www.westcoast.fisheries.noaa.gov/protected_species/salmon_steelhead/sal
mon_and_steelhead_listings/steelhead/steelhead.html.
Orr, H.A. (2009). Fitness and its role in evolutionary genetics. Nature Reviews Genetics,
10, 531-539.
Pacharsky, B.L., Fraissinet, P.R., Penton, M.A., & Conklin, D. J. (1990). Freshwater
macroinvertebrates of northeastern North America. Ithaca, NY: Cornell
University Press.
Petrosky, C.E. & Schaller, H.A. (2010). Influence of river conditions during seaward
migration and ocean conditions on survival rates of Snake River Chinook salmon
and steelhead. Ecology of Freshwater Fish, 19, 520-536.
Quinn, T.P. (2005). The behavior and ecology of Pacific salmon and trout. University of
Washington Press: Seattle, Washington.
Reece, J.B., Urry, L.A., Cain, M.L., Wasserman, S.A., Minorsky, P.V., & Jackosn, R.B.
(2011). Campbell Biology. Boston: Benjamin Cummings.
Roni, P. & Quinn, T.P. (2001). Density and size of juvenile salmonids in response to
placement of large woody debris in western Oregon and Washington streams.
Canadian Journal of Fisheries and Aquatic Sciences, 58, 282-292.
84

Rundio, D.E. & Lindley, S. T. (2008). Seasonal patterns of terrestrial and aquatic prey
abundance and use by Onchorhynchus mykiss in a California coastal basin with a
Mediterranean climate. Transections of the American Fisheries Society, 137, 467480.
Seelbach, P.W. (1993). Population biology of steelhead in stable-flow, low-gradient
tributary of Lake Michigan. Transactions of the American Fisheries Society, 122,
179-198.
Sheppard, D. (1972). The present status of the steelhead trout stocks along the Pacific
coast. H. Rosenberg (editor), A review of the oceanography and renewable
resources of the northern Gulf of Alaska, p. 519-556. IMS Report R72-23, Sea
Grant Report 73-3. Institute of Marine Science, University of Alaska, Fairbanks.
Sickle, J.V. & Gregory, S.V. (1990). Modeling inputs of large woody debris to streams
from falling trees. Canadian Journal of Forest Research, 20, 1593-1601.
Simpson, W.G., Kennedy, B.M., & Ostrand, K.G. (2009). Seasonal foraging and
piscivory by sympatric wild and hatchery-reared steelhead from an integrated
hatchery program. Environmental Biology of Fishes, 86, 473-482.
Smith, J.J. & Li, H.W. (1983). Energetic factors influencing foraging tactics of juvenile
steelhead trout, Salmo gairdneri. Developments in Environmental Biology of
Fishes, 2, 173-180.
Solberg, M.F., Zhang, Z., Nilsen, F., & Glover, A. (2013). Growth reaction norms of
domesticated, wild and hybrid Atlantic salmon families in response to differing
social and physical environments. Evolutionary Biology, 13, 3-23.
Suttle, K.B., Power, M.E., Levine, J.M. & McNeely, C. (2004). How fine sediment in
riverbeds impairs growth and survival of juvenile salmonids. Ecological
Applications, 14, 969-974.
Tagliaferro, M., Arismendi, I., Lancelotti, J. & Pascul, M. 2015. A natural experiment of
dietary overlap between introduced rainbow trout (Oncorhyncus mykiss) and
native Puyen (Galaxias maculates) in the Santa Cruz River, Patagonia.
Environmental Biology of Fishes, 98, 1311-1325.
United States Department of Agriculture, Forest Service. (2011a). Title II Project
Submission Form. Hood Canal Steelhead Project: Duckabush River wild
steelhead supplementation and monitoring component.
United States Department of Agriculture Forest Service (2011b). USDA Forest Service

85

Watershed Condition Framework FY2012 Transition watershed restoration action
plan Pacific Northwest Region Olympic National Forest, Hood Canal Ranger
District Upper South Fork Skokomish River.
United States Department of Agriculture Forest Service (2011c). USDA Forest Service
Watershed Condition Framework FY2012 Transition watershed restoration action
plan Pacific Northwest Region Olympic National Forest, Hood Canal Ranger
District Lower South Fork Skokomish River.
United State Forest Service, Hood Canal Ranger District (1998). Duckabush Watershed
Analysis.
Vanderploeg, H.A. & Scavia, D. (1979). Two electivity indices for feeding with spatial
reference to zooplankton grazing. Journal of Fisheries Research Board of
Canada, 36, 362-365.
Wallace, J.D. & Weber, J.R. (1996). The role of macroinvertebrates in stream ecosystem
function. Annual Review of Entomology, 41, 115-139.
Washington State Department of Ecology. (1998). Ambient stream biological monitoring
site Dewatto River. Retrieved from http://www.ecy.wa.gov/apps/ watersheds
/streambio/station.asp?sta=39.
Washington State Department of Ecology. (2011). Ambient stream biological monitoring
site Duckabush River. Retrieved from http://www.ecy.wa. gov/apps/watersheds/
streambio/station.asp?wria=16.
Washington State Department of Fish and Wildlife (2014a). Salmon Hatcheries
Overview.
Retrieved From http://wdfw.wa.gov/hatcheries/overview.html.
Washington State Department of Fish and Wildlife. (2014b). WDFW and Wild Fish
Conservancy settles lawsuit over 'early winter' hatchery steelhead releases [Press
release]. Retrieved from http://wdfw.wa.gov/news/apr2514b/.
Weber, E.D. & Fausch, K.D. (2003). Interactions between hatchery and wild salmonids
in streams: Differences in biology and evidence for competition. Canadian
Journal of Fisheries and Aquatic Sciences, 60, 1018-1036.
Wild Fish Conservancy. (2014). Lawsuit settlement big advance for wild steelhead
recovery [Press release]. Retrieved from http://wildfishconservancy.
org/about/press-room/ press-releases /lawsuit-settlement-big-advance-for-wildsteelhead-recovery
Wild Fish Conservancy. (n.d.(b)). Wild Steelhead. Retrieved from http://wildfish
conservancy.org.
86

Williams, D. & Hayes, H. (1977). The ecology of temporary streams. International
Review of Hydrobiology, 62, 53-61.
Wipfli, M.S. & Grerovich, D.P. (2002). Export of invertebrates and detritus from fishless
headwater streams in southeastern Alaska: Implications for downstream salmonid
production. Freshwater Biology, 47, 957-969.
Woo, H.S., Julien, P.Y. & Richardson, E.V. (1986). Washload and fine sediment load.
Journal of Hydraulic Engineering, 112, 541-545.
Wright, L.A. (2010). Winter feeding ecology of coho salmon (Oncorhynchus kisutch),
steelhead (O. mykiss), and cutthroat trout (O. clarkii) in the Skokomish River,
Washington. (Published masters thesis). The Evergreen State College, Olympia,
Washington.
Yuasa, M. (2014, April 25). Deal takes all but 180,000 hatchery steelhead out of Puget
Sound river systems. The Seattle Times. Retrieved from http://seattletimes.com.

87

Appendices
Table 1 The type and abundance of prey items found in wild and hatchery steelhead juvenile diets in the Dewatto,
Duckabush and South Fork Skokomish River. Insect prey items were classified by order and life stage. Non-insect
prey items were classified by taxon. * indicates prey items found in stomach contents but not found in drift and were
therefore excluded from the Electivity Index (Ei) calculations.

Insect Order

Life Stage

Life
Stage

Dewatto
Abundance

Duckabush
Abundance

Skokomish
Abundance

Coleoptera

Aquatic

Adult

33

8

54

Coleoptera

Aquatic

Larvae

24

2

13

Coleoptera

Terrestrial

Adult

14

28

69

Coleoptera

Terrestrial

Larvae

4

23

3

Diptera

Aquatic

Adult

40

369

144

Diptera

Aquatic

Larvae

732

1036

616

Diptera

Aquatic

Pupae

40

6

133

Diptera

Terrestrial

Adult

7

1

27

Diptera

Terrestrial

Larvae

6

2

2

Diptera

Terrestrial

Pupae

1*

0

0

Ephemeroptera

Aquatic

Adult

29

1

253

Ephemeroptera
Hemiptera:
Auchenorrhyncha
Hemiptera:
Auchenorrhyncha

Aquatic

Larvae

752

2389

2838

Terrestrial

Adult

3*

7*

14

Terrestrial

Larvae

3

2

0

Hemiptera: Heteroptera

Aquatic

Adult

1*

9*

1*

Hemiptera: Heteroptera

Terrestrial

Adult

0

0

62

Hemiptera: Heteroptera
Hemiptera:
Sternorrhyncha
Hemiptera:
Sternorrhyncha

Terrestrial

Larvae

1*

7

2

Terrestrial

Adult

6

10

28

Terrestrial

Larvae

4

8

1

Hymenoptera

Terrestrial

Adult

15

23

61

Lepidoptera*

Terrestrial

Adult

3*

6

0

Lepidoptera

Terrestrial

Larvae

18

75

10

Lepidoptera

Terrestrial

Pupae

0

2*

0

Megaloptera

Aquatic

Larvae

1

0

0

Neuroptera

Terrestrial

Larvae

11

8

0

Orthoptera

Terrestrial

Adult

3

7

1

Plecoptera

Aquatic

Adult

0

2*

3*

Plecoptera

Aquatic

Larvae

35

139

139

Pscoptera

Terrestrial

Larvae

1

7

1

88

Thysanoptera

Terrestrial

Adult

1

0

0

Thysanoptera

Terrestrial

Larvae

1*

0

0

Trichoptera

Aquatic

Adult

10*

99*

11

Trichoptera

Aquatic

Larvae

199

34

212

Trichoptera

Aquatic

Pupae

1

6

24*

Dewatto
Abundance

Non-Insect Taxon
Annelida: Oligochaeta

Aquatic

Annelida: Oligochaeta

Duckabush
Abundance

Skokomish
Abundance

1

1

0

Terrestrial

1*

0

0

Arachnida: Acari

Aquatic

70

27

42

Arthropoda: Arachnida

Terrestrial

10

45

17

Arthropoda: Chilopoda

Terrestrial

1*

3

1*

Arthropoda: Collembola
Arthropoda:
Microcoryphia

Terrestrial

3

2

0

Terrestrial

53*

5

0

Cottidae

Aquatic

2*

2*

0

Crustacea: Isopoda

Terrestrial

3

0

0

Crustacea: Ostracoda

Aquatic

3

0

0

Mollusca: Gastropoda

Aquatic

49

0

0

Mollusca: Gastropoda

Terrestrial

4*

0

0

Nemata

Aquatic

3

8

25

Nematomorpha

Terrestrial

11

5

2

2132

4293

4780

Total Number of Prey
Items

89

Table 2. The type and abundance of prey items found in wild and hatchery steelhead juvenile diets in the
lower reach of the Duckabush River.
Insect Order
Coleoptera Aquatic Adult
Coleoptera Terrestrial Adult
Coleoptera Terrestrial Larvae
Cottidae
Diptera Aquatic Adult
Diptera Aquatic Larvae
Diptera Aquatic Pupae
Diptera Terrestrial Adult
Diptera Terrestrial Larvae
Ephemeroptera Aquatic Larvae
Hemiptera: Auchenorrhyncha Terrestrial Adult
Hemiptera: Auchenorrhyncha Terrestrial Larvae
Hemiptera: Heteroptera Terrestrial Adult
Hemiptera: Heteroptera Terrestrial Larvae
Hemiptera: Sternorrhyncha Terrestrial Adult
Hemiptera: Sternorrhyncha Terrestrial Larvae
Hymenoptera Terrestrial Adult
Lepidoptera Terrestrial Adult
Lepidoptera Terrestrial Larvae
Lepidoptera Terrestrial Pupae
Neuroptera Terrestrial Larvae
Orthoptera Terrestrial Adult
Plecoptera Aquatic Adult
Plecoptera Aquatic Larvae
Psocoptera Terrestrail Larvae
Trichoptera Aquatic Adult
Trichoptera Aquatic Larvae
Trichoptera Aquatic Pupae
Non-Insects

Hatchery

Wild

3
0
0
0
38
28
3
12
0
102
0
0
0
0
0
0
2
0
1
21
0
0
0
8
0
0
6
0
Hatchery

0
12
14
1
80
441
114
20
10
748
20
20
4
6
39
17
35
3
54
1
12
21
3
53
12
5
20
21
Wild

Annelida: Oligochaeta Aquatic
Arachnida: Acari Aquatic
Arthropoda: Arachnida Terrestrial
Arthropoda: Chilopoda Terrestrial
Arthropoda: Collembola Terrestrial
Arthropoda: Microcoryphia Terrestrial
Nemata Aquatic

9
0
0
0
0
2
3

1
3
16
1
2
2
2

Nematomorpha Aquatic

0

0

238

1813

Total Number of Prey Items

90

Table 3. Items found in the Dewatto River, Duckabush River and South Fork Skokomish River. Those in
bold indicate the drift items that were only found in one of the three rivers.
Dewatto River

Duckabush River

Annelida: Oligochaeta Aquatic
Arachnida: Acari Aquatic

Annelida: Oligochaeta Aquatic
Arachnida: Acari Aquatic

Arthropoda: Arachnida Aquatic

Arthropoda: Arachnida Terrestrial

Arthropoda: Arachnida Terrestrial
Arthropoda: Collembola Terrestrial
Coleoptera Aquatic Adult

Arthropoda: Collembola Terrestrial
Coleoptera Aquatic Adult
Coleoptera Aquatic Larvae

Coleoptera Aquatic Larvae

Coleoptera Terrestrial Adult

Coleoptera Terrestrial Adult
Coleoptera Terrestrial Larvae

Coleoptera Terrestrial Larvae
Crustacea: Copepoda Aquatic

Crustacea: Cladocera Aquatic

Crustacea: Ostracoda Aquatic

Crustacea: Copepoda Aquatic
Crustacea: Isopoda Terrestrial
Crustacea: Ostracoda Aquatic
Diptera Aquatic Adult
Diptera Aquatic Larvae
Diptera Aquatic Pupae
Diptera Terrestrial Adult

Diptera Aquatic Adult
Diptera Aquatic Larvae
Diptera Aquatic Pupae
Diptera Terrestrial Adult
Diptera Terrestrial Larvae
Ephemeroptera Aquatic Adult
Ephemeroptera Aquatic Larvae

Diptera Terrestrial Larvae
Ephemeroptera Aquatic Adult

Hemiptera: Auchenorrhyncha Terrestrial Adult
Hemiptera: Auchenorrhyncha Terrestrial
Larvae

Ephemeroptera Aquatic Larvae

Hemiptera: Heteroptera Terrestrial Larvae

Hemiptera: Auchenorrhyncha Terrestrial
Larvae
Hemiptera: Sternorrhyncha Terrestrial
Adult
Hemiptera: Sternorrhyncha Terrestrial
Larvae

Hemiptera: Sternorrhyncha Terrestrial Adult
Hemiptera: Sternorrhyncha Terrestrial Larvae
Hymenoptera Terrestrial Adult

Hymenoptera Terrestrial Adult

Lepidoptera Terrestrial Larvae

Lepidoptera Terrestrial Larvae

Nemata Aquatic

Mollusca: Gastropoda Aquatic

Nematomorpha Aquatic

Nemata Aquatic

Neuroptera Terrestrial Larvae

Neuroptera Terrestrial Larvae

Orthoptera Terrestrial Adult

Orthoptera Terrestrial Adult
Plecoptera Aquatic Larvae
Psocoptera Terrestrial Larvae

Plecoptera Aquatic Larvae
Psocoptera Terrestrial Larvae
Trichoptera Aquatic Larvae

Trichoptera Aquatic Larvae
Trichoptera Aquatic Pupae

Trichoptera Aquatic Pupae

South Fork Skokomish
River
Annelida: Oligochaeta
Aquatic
Arachnida: Acari Aquatic
Arthropoda: Arachnida
Terrestrial
Arthropoda: Collembola
Terrestrial
Coleoptera Aquatic Adult
Coleoptera Aquatic Larvae
Coleoptera Terrestrial
Adult
Coleoptera Terrestrial
Larvae
Cottidae Aquatic
Crustacea: Amphipoda
Aquatic
Crustacea: Copepoda
Aquatic
Crustacea: Isopoda Aquaitc
Diptera Aquatic Adult
Diptera Aquatic Larvae
Diptera Aquatic Pupae
Diptera Terrestrial Adult
Diptera Terrestrial Larvae
Ephemeroptera Aquatic
Adult
Ephemeroptera Aquatic
Larvae
Hemiptera:
Auchenorrhyncha
Terrestrial Adult
Hemiptera:
Auchenorrhyncha
Terrestrial Larvae
Hemiptera: Heteroptera
Terrestrial Adult
Hemiptera: Heteroptera
Terrestrial Larvae
Hemiptera: Sternorrhyncha
Terrestrial Adult
Hemiptera: Sternorrhyncha
Terrestrial Larvae
Hymenoptera Terrestrial
Adult
Lepidoptera Terrestrial
Adult
Lepidoptera Terrestrial
Larvae
Mollusca: Bivalvia
Aquatic
Nemata Aquatic
Nematomorpha Aquatic
Neuroptera Terrestrial
Larvae
Odonata Aquatic Adult
Orthoptera Terrestrial
Adult
Plecoptera Aquatic Adult
Plecoptera Aquatic Larvae
Psocoptera Terrestrial
Larvae

91

Thysanoptera Terrestrial
Adult
Trichoptera Aquatic Adult
Trichoptera Aquatic Larvae
33 Total Item Types

32 Total Item Types

40 Total Item Types

92

Media

Davis_SMESthesis2015.pdf