Faulkner_H2013.pdf

Media

Part of Influence of Aquaculture on Winter Sea Duck Distribution and Abundance in South Puget Sound

extracted text

Influence of Aquaculture on Winter Sea Duck
Distribution and Abundance
in South Puget Sound

by
Hannah Faulkner

A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
June 2013


 

i
 

© 2013 by Hannah Faulkner. All rights reserved.


 

ii
 

This Thesis for the Master of Environmental Studies Degree
by
Hannah Faulkner

has been approved for
The Evergreen State College
by

Dr. Dina Roberts
Member of the Faculty

Date


 

iii
 

ABSTRACT
Influence of Aquaculture on Winter
Sea Duck Distribution and Abundance
in South Puget Sound
by Hannah Faulkner
Shellfish aquaculture is a valuable and expanding industry in Washington State, in
particular in South Puget Sound. Concurrently, long-term monitoring efforts throughout
Puget Sound reveal varying levels of decline in a significant number of over-wintering
sea duck species. However, reasons for these declines are unknown and the need for
winter habitat assessments throughout Puget Sound is evident. The overlapping
distributions of aquaculture industry and marine bird use in nearshore environments
identify a high probability of interaction. This study identified and evaluated associations
of four sea duck species/groups, Bufflehead, Scoter, Goldeneye and Merganser, in
relation to a changing aquaculture landscape. Findings illustrate that shellfish aquaculture
in South Puget Sound is both expanding and intensifying; expanding almost 3 study sites
annually by medium and large acreage operations and growing at an annual rate of 127
acres. Our results suggest that sea ducks exhibit species or group-specific responses to
aquaculture. Evaluating the location and intensity of aquaculture operations in the South
Puget Sound, Bufflehead and Scoter species abundances were positively associated with
industry to different degrees. Only Bufflehead, however, maintained significant positive
associations over time. Alternatively, Goldeneye and Merganser species abundances
demonstrated negative associations with shellfish aquaculture, however responses varied
by intensity of culture operations. The influence of shellfish aquaculture on winter sea
duck populations is clear, however variability by species demonstrate that while industry
may coexist or benefit some, can prove deleterious for others. This study highlights the
complexity in defining spatially and temporally dynamic sea duck-aquaculture relations.
We recommend continued research to better understand species-specific habitat use and
availability in relation to aquaculture development and activity of winter sea duck
populations in Puget Sound.


 

TABLE OF CONTENTS

List of Figures
List of Tables
Acknowledgements
Chapter 1 Literature Review
Introduction
Puget Sound
Sea Ducks in Puget Sound
Monitoring Efforts and Trends
Habitat Use and Requirements
Shellfish Aquaculture
History and Status in Washington
Method and Operations
Ecosystem Effects
Policy and Regulations
Sea Ducks and Aquaculture
Sea Duck Responses to Aquaculture
Conclusions
Chapter 2 Manuscript
Introduction
Methods
Study Area
Study Organisms
Data Sets
Spatial Analysis
Statistical Analysis
Results
Shellfish Aquaculture
Bufflehead
Scoter Species
Goldeneye Species
Merganser Species
Discussion
Sea Duck-Aquaculture Relationships
Future Considerations
Conclusions
Chapter 3 General Conclusions


 

v
vi
vii
1
1
2
4
5
7
10
10
11
14
19
22
23
26
28
28
32
32
33
35
38
42
49
49
49
50
51
51
58
58
63
66
68

iv
 

LIST OF FIGURES
Figure 1. Simplified flow chart depicting method of operations to collage, merge,
summarize and export data
47
Figure 2. Map depicting 192 study site polygons generation in GIS along South Puget
Sound shoreline
48
Figure 3. Growth of shellfish aquaculture in South Puget Sound by total acres cultivated
and number of study sites under cultivation from 1994-2012
53
Figure 4. Growth in total acres cultivated by species cultured across South Puget Sound
study sites from 1994-2012
53
Figure 5. Sample means plot of Bufflehead relative abundances at each acreage size class
from 1994 to 2012
54
Figure 6. Sample means plot of Scoter species relative abundances across South Puget
Sound study sites from 1994 to 2012
55
Figure 7. A comparison of sample means by acreage size class for each species/species
group with corresponding Tukey HSD comparisons
56
Figure 8. Sample means plot of Merganser species relative abundances across South
Puget Sound study sites from 1994 to 2012
57
Figure 9. Raw means analysis of average count and relative abundance of Goldeneye
species across South Puget Sound
xv
Figure 10. Raw means analysis of average count and relative abundance of Scoter species
across South Puget Sound
xv
Figure 11. Raw means analysis of average count and relative abundance of Merganser
species across South Puget Sound
xvi
Figure 12. Raw means analysis of average count and relative abundance of Bufflehead
across South Puget Sound
xvi
Figure 13. Sample means plot of Bufflehead relative abundances across South Puget
Sound study sites from 1994 to 2012
xvii
Figure 14. Sample means plot of Scoter relative abundances at each acreage size class
from 1994 to 2012
xvii
Figure 15. Sample means plot of Goldeneye species relative abundances across South
Puget Sound study sites from 1994 to 2012
xviii
Figure 16. Sample means plot of Goldeneye species relative abundances at each acreage
size class from 1994 to 2012
xviii
Figure 17. Sample means plot of Merganser species relative abundance at each acreage
size class from 1994 to 2012
xix


 

v
 

LIST OF TABLES
Table 1. 9 species common in Puget Sound, including WDFW species code, feeding
characteristics and abundance in and dependence on marine environment
xxi
Table 2. Aquaculture rate of change in acres and count of study sites under cultivation in
South Puget Sound from 1994 to 2012
53
Table 3. Results of test slices report for significant interaction effect of acreage x year on
Bufflehead relative abundances
54
Table 4. Standard least squares REML results of fixed effects (acreage, year,
acreage*year) for Scoter species relative abundances

55

Table 5. Tukey HSD crosstab report for effect of aquaculture acreage on Scoter species
relative abundances
55
Table 6. Standard least squares REML results of fixed effects (acreage, year,
acreage*year) for Goldeneye species relative abundances

56

Table 7. Tukey HSD crosstab report for effect of aquaculture acreage on Goldeneye
species relative abundances
56
Table 8. Standard least squares REML results of fixed effects (acreage, year,
acreage*year) for Merganser species relative abundances

57

Table 9. Tukey HSD crosstab report for effect of aquaculture acreage on Merganser
species relative abundances
57
Table 10. Standard least squares REML results of fixed effects (acreage, year,
acreage*year) for Bufflehead relative abundances

xix

Table 11. Tukey HSD crosstab report for effect of aquaculture acreage on Bufflehead
relative abundance
xix
Table 12. Standard least squares REML variance component estimates measuring
Bufflehead winter relative abundance across South Puget Sound
xx
Table 13. Standard least squares REML variance component estimates measuring Scoter
species winter relative abundance across South Puget Sound
xx
Table 14. Standard least squares REML variance component estimates measuring
Goldeneye winter relative abundance across South Puget Sound
xx
Table 15. Standard least squares REML variance component estimates measuring
Merganser winter relative abundance across South Puget Sound
xx


 

vi
 

ACKNOWLEDGMENTS
I want to thank Joseph Evenson from Washington Department of Fish and Wildlife for
inspiring and supporting my efforts to evaluate wintering sea duck habitat use in Puget
Sound, and for conducting and providing relevant marine bird survey data. I am fortunate
to be part of this collaborative effort between a state agency and educational institution. I
also want to thank my reader Dr. Dina Roberts for providing support, encouragement and
excitement until my final thesis was submitted. Additional thanks should also be
provided to all the parties involved in shellfish aquaculture research efforts including:
Washington Department of Fish and Wildlife Licensing and Registration, Washington
Department of Health Commercial Shellfish Licensing, and the Pacific Shellfish Institute.
A special thanks is rewarded to Andy Duff of Washington Department of Fish and
Wildlife and Greg Stewart, adjunct faculty of The Evergreen State College, for their
support in the GIS and spatial analysis portion of my work. A final thanks goes out to my
understanding network of friends and family for supporting me through this intensive
thesis process.
I could not have completed this degree of quality were it not for the inspiration of my
undergraduate program at Huxley College of the Environment with Western Washington
University and the interdisciplinary nature and encouragement of The Evergreen State
College Master of Environmental Studies Program.


 

vii
 

CHAPTER ONE
LITERATURE REVIEW
Introduction
Long-term monitoring efforts of sea duck populations reveal varying levels of
decline, similar to other marine bird species associated with the Washington Coast, and
evident in South Puget Sound populations (Fresh et al. 2011, PSAT 2007, Buchanan
2006). Mechanisms causing these declines are largely unknown as habitat use,
distribution and ecology of sea ducks has been minimally studied. This lack of
knowledge has been recognized by various organizations and monitoring and research
efforts have been developed to assist in identifying possible limiting factors. To maintain
vulnerable sea duck species, it is vital to understand potential limiting factors across all
phases of their annual cycle. Puget Sound provides significant habitat for sea ducks
during their non-breeding stages, in particular, related to foraging, resting and molting
requirements. Accordingly, it is critical to monitor and assess possible factors influencing
habitat use by sea ducks to identify drivers of population density and distribution in Puget
Sound. These marine coastal environments are also home to a significant and growing
aquaculture industry. This begs the question, could aquatic farm operations play a role in
determining sea duck populations in Puget Sound? If so, how are these effects translated
into observed sea duck population dynamics? To answer these questions, I will be
assessing sea duck distribution and density of nine commonly occurring species in Puget
Sound with documented varying levels of long-term decline. I will address principals of
species-habitat associations to understand relationships between sea duck populations and
presence of aquaculture. This review will provide an exploration of current knowledge,


 

1
 

gaps in knowledge, areas of improvement and plans for future efforts concerning sea
duck populations and the role of shellfish aquaculture in Puget Sound.

Puget Sound
Puget Sound is the second largest estuary in the United Stated, a complex
landscape that supports an abundance of terrestrial, freshwater, estuarine and marine
ecosystem species and habitats (Fresh et al. 2011). Puget Sound encompasses 2,800
square miles of inland marine waters, 2,500 miles of shoreline and is fed by 20 major
river systems (PSP 2012). Puget Sound is part of the greater Salish Sea, comprising
inland marine waters that span international boundaries from the coast of British
Columbia extending from Desolation Sound to include the Strait of Georgia and the Strait
of Juan de Fuca (PSNERP). Alternatively, the Puget Sound region extends from the
Canadian boarder, throughout Puget Sound shorelines and out the Strait of Juan de Fuca
to Neah Bay (PSNERP). The nearshore ecosystems that define these shorelines, bracing
the terrestrial-marine interface, are among the more complex system types, including
coastal riparian, intertidal and subtidal zones. The nearshore is generally defined from the
top of shoreline bluffs to the depth of offshore waters where light is no longer able to
penetrate waters to productively support plant growth. This area encompasses bluffs,
beaches, mudflats, kelp and eelgrass beds, salt marshes, gravel spits and estuaries
(PSNERP). While these nearshore environments are critical to many ecological
communities, they are also the foundation of many ecosystem goods and services
important to Washington’s human communities.


 

2
 

The Puget Sound region is a vital resource for Washington, supporting thriving
coastal ecosystems, sustaining growing human populations, and providing a foundation
for many long-standing cultural traditions. It is no surprise that Puget Sound is
recognized as a ‘national treasure’ environmentally, economically, socially and culturally
(PSP 2012). While Puget Sound’s coastlines parallel only the western-most third of the
state, this area is home to nearly two-thirds of the states population at 4.1 million people
(PSP 2012). In a recent report on the state of the sound, it was identified that 70% of all
jobs and 77% of total income in the state are derived from the Puget Sound basin (PSP
2012). As Fresh et al. (2011) describe, “the (Puget Sound) region’s location, deep
harbors, natural resources and economic and cultural links to the Pacific Rim have made
it a global trade center, an economic engine for much of the Pacific Northwest and an
important component of the national economy”. Therefore, the sustained health and
function of Puget Sound is recognized across disciplines, organizations, and nations
(joining coastal waters of Canada in northern-most Washington). Consequently, the
cooperation among and within these many facets are essential to maintaining productive
operations and results among the myriad of derived resources. One such example is the
Puget Sound Partnership (PSP), established in 2007, representing a collaboration of
individuals, agencies, and organizations founded on a common goal to “ensure that the
Puget Sound forever will be a thriving natural system, with clean marine and freshwaters,
healthy and abundance native species, natural shorelines, and place for public enjoyment
and a vibrant economy that prospers in productive harmony with a healthy Sound” (PSP
2012).


 

3
 

Sea Ducks in Puget Sound
Puget Sound waters provide habitat to over 70 species of marine bird closely
associated with marine environments throughout some or all of their life history, for
nesting, wintering and migration. The scope of my work is focused to nine commonly
occurring sea duck species of South Puget Sound (WDFW 2010), including: black scoter
(Melanitta nigra), surf scoter (M. perspicillata), white-winged scoter (M.fusca), common
goldeneye (Bucephala clangula), Barrow’s goldeneye (B. islandica), bufflehead (B.
albedo), common merganser (Mergus merganser), hooded merganser (Lophodytes
cucullatus), and red-breasted merganser (M. serrator). Sea ducks are a typical
representation of waterfowl dependent on Puget Sound resources, distinguished by other
Anatidae by marine specialized physiological traits. The majority of sea ducks utilizing
Washington coastal habitats breed in the Boreal Forests of Canada and Alaska, and
migrate south in winter to Puget Sound during non-breeding periods. Wintering periods
can be shown to constitute a significant or majority of their life cycle (WDFW 2010),
although winter habitat affiliations of sea ducks are poorly known (Esler et al. 2000)
Wintering sea duck populations in South Puget Sound are almost exclusively associated
with the nearshore environment and are dependent on marine food resources within the
intertidal and subtidal habitat (<20m in depth) (Essington et al. 2011). These nine
common species can be further subdivided by foraging guild into benthivores (eg. scoter
species, goldeneye species, bufflehead) and piscivores (eg. merganser species).
Benthivores forage largely on ground dwelling aquatic invertebrates, such as mollusks
and crustaceans, although specific variations of prey will occur among species and by
seasonal availability (SDJV 2003a-d,f, h, j-k). Piscivores forage primarily on small fish,


 

4
 

such as salmon, trout and sculpin, but at times are documented opportunistic feeders,
feeding on other available aquatic invertebrates (SDJV 2003e, g, i).
Monitoring Efforts and Trends
Marine birds have been widely used as indicators of estuarine ecosystems (Bower
2009). Monitoring efforts can provide useful indicators of changes in ecosystem
condition or health (Gaydos and Pearson 2011), by evaluating population responses to
natural and manipulated environments. Birds, and waterfowl in particular, are well
known to respond to changes in habitat such as natural or anthropogenic alterations to
available prey (Kirk et al. 2007). Prior to the 1970s, marine bird monitoring efforts across
Puget Sound were largely restricted to citizen science efforts (Bower 2009). Not until
1978 was the first comprehensive census of marine birds in northern Puget Sound
conducted as part of the Marine Ecosystems Analysis (MESA) (Essington et al. 2011),
although failing to incorporate the entirety of the Puget Sound region. In 1992, the
Washington Department of Fish and Wildlife (WDFW) began collecting summer and
winter marine bird population estimates, as part of the Puget Sound Ambient Monitoring
Program (PSAMP). Aerial transects were flown both paralleling the shorelines, and
extending into open-water throughout the entire Puget Sound and southern shore of the
Straight of Juan de Fuca (Nysewander et al. 2005). However, following 1999, summer
surveys were eliminated and only winter surveys currently remain. PSAMP still remains
the only source of continuous marine bird monitoring efforts in Puget Sound. There are
inherent difficulties in both the method and analysis of aerial bird surveys (Butler et al.
1995) and concerted effort must be taken in methodology and monitoring and
supplementary analysis. The first report comparing MESA (1978-79) and PSAMP (1991-


 

5
 

1999) surveys documented long-term declines in a large portion of Puget Sound sea duck
species with drastic declines in many. Analyses revealed significant declines in 13 of 20
species, or species groups studied (Nysewander et al. 2005). Characteristic of these sharp
declines are wintering Scoter populations (black, white-winged and surf), demonstrating
declines as great at 56% from 1978-99 and 55% from 1994-2011 (Evenson 2012). While
other species documented stable or slowly decreasing long-term trends, such as longtailed and harlequin duck. Although coupling concerns develop due to occurrence of
some species at low densities in restricted locations (WDFW 2010). These initial results
from Nysewander et al. (2005) sparked concern over declining marine bird populations,
and reinforced involved agencies to address the conservation and management of Puget
Sound sea duck populations (eg. Bower 2009, PSP 2012, SDJV 2012, WDFW 2010).
Reasons for these declines in wintering sea duck populations throughout Puget
Sound, consistent with other documented trends throughout North America (Bower
2009), remain unanswered. As a result, several knowledge gaps common across sea duck
populations have been identified. This may be attributed to the unique challenges of
studying sea ducks, such as their broad and remote distributions (SDJV 2012). Foremost,
basic biology and ecology of sea ducks is poorly known or unknown (SDJV 2012).
Additionally, comprehensive monitoring efforts lack consistency and connectivity
throughout different life stages to provide complete detection of population trends
(Anderson et al. 2009). Many scientists, conservationists and managers recognize the
need for continued research, monitoring and assessment linked throughout the entirety of
their life cycles (SDJV 2012). Specifically, a concerted effort must be placed on nonbreeding grounds as constrains in these areas may contribute to long-term declines in


 

6
 

some species of sea duck (Anderson et al. 2009), especially as some populations spend
most of their annual cycle on non-breeding areas (Zydelis et al. 2006). Consistent and
coherent work will not only provide insight into Puget Sound’s sea duck populations, best
directing conservation and management strategies, but will also provide an indicator to
ecosystem health and vitality (PSP 2012).
Habitat Use and Requirements
Research on sea duck population dynamics often aim to characterize functional
species requirements, identify habitat availability and limitations, and explore patterns of
habitat use. This information can then be used to identify possible mechanisms
responsible for observed and predicted spatial and temporal trends in sea duck
populations. In particular, diet provides a relative value of habitat to birds by quantifying
benefits and consequences to local and regional productivity and survivorship (Anderson
et al. 2008). This is based largely on the growing body of knowledge exploring the
behavioral and functional relationships between predators and their prey (Kirk et al.
2008) by investigating how sea duck populations utilize food resources and respond to
both natural and human-induced modifications to occupied habitats and prey landscapes.
To answer these questions, research has encompassed three complementary focuses
including, foraging dynamics, movement patterns and physiological requirements.
First, explorations in foraging dynamics address behavior responses to food
characteristics and availability to infer habitat quality. Sea ducks, like other birds, have
been documented to modify foraging behavior and efforts in response to food abundance
and quality (Lewis et al. 2008). Foraging theory suggests that fitness consequences of
foraging behavior encourage animals to optimize net energy intake when faced with


 

7
 

variation in prey attributes or abundance (Kirk et al. 2008, Lewis et al. 2008). Thus,
observations in foraging behavior can then be used to infer habitat quality, including
possible food limitations (Lewis et al. 2008). Several studies by Kirk et al. (2007, 2008)
found that certain prey characteristics (i.e. weaker byssal attachments, thinner shells,
easier accessibility) of marine mussels rendered them more susceptible, and in response,
surf scoters demonstrated greater level of depletion. Second, complementary studies have
been conducted on movement patterns to uncover mechanisms underlying foraging
strategies and animal distributions (Kirk et al. 2008). Distribution theories predict that
predator densities are closely related to prey landscapes, further suggesting that as a result
of individual movements, changes in predator distribution can reflect the underlying
differences in the availability or quality of prey within a habitat (Kirk et al. 2008). In
addition to habitat quality, animal movement patterns have also been used to estimate
minimum space requirements and evaluate flexibility of individuals to habitat change
(Kirk et al. 2008 and references therein). Finally, physiological measurements reinforce
investigations into foraging dynamics and distribution patterns, by quantifying nutritional
requirements and energy balances of sea ducks to determine functional habitat needs
(Anderson and Lovvorn 2011). Anderson et al. (2009) explains, that to understand
functional dependencies on specific foraging sites requires knowledge of the relative
contributions of food to predator conditions. This helps identify critical foraging sites and
possible sources of nutritional constraints for sea ducks. Additionally, evaluating
physiological measurements in relation to spatial and temporal variability also helps
identify distinctions in endogenous versus environmental variables of bird energy statuses
(Anderson et al. 2008). Moreover, collaborative methods can be used to infer


 

8
 

comprehensive evaluation of habitat for sea ducks, as utilized in Kirk et al. (2008)
comparing differing habitat prey landscapes of two heavily used wintering grounds in
coastal British Columbia. Results demonstrated that sites of high mussel prey density
were consequently significantly depleted, thus encouraging limited site fidelity, larger
foraging range size, and less predictable movement and distribution patterns.
Alternatively, the site of stable clam availability supported higher site fidelity, small
foraging range size and more predictable movement and distribution patterns (Kirk et al.
2008).
As mentioned above, as per investigations in habitat use and requirements, sea
duck populations interact intimately with their environment. These interactions can then
be monitored and evaluated to conceptualize patterns, evaluate inconsistencies and
identify disturbances in functionally critical sea duck habitats. Although a comprehensive
assessment of habitat requirements is lacking for Puget Sound sea duck populations, one
essential component to quantifying habitat needs is identifying potential stressors and/or
limitations (Evenson 2012). Several possible mechanisms responsible for documented
populations declines (described in Section 3) in South Sound’s wintering sea ducks
include: habitat modification and degradation, human activities, disease and
contaminants, low recruitment rates, food resource depletion, predation, and larger scale
population shifts (Evenson, pers. communication). The continuing focus of this review
will concentrate on one practice of human activity, shellfish aquaculture, concurrent with
increasing focus and concern of its role in estuarine ecosystems and its growth in
nearshore environments in Puget Sound.


 

9
 

Shellfish Aquaculture
History and Status in Washington State
Abundance of shellfish in Puget Sound is a vital resource for Washington,
continuing a time-honored cultural tradition, and today providing a thriving economic
industry. The aquaculture industry is one of many resource-based business-sectors in the
State, securing the economies of many rural western Washington communities and
providing the state a significant and indispensable value (PSP 2003c). While recreational
harvesting of shellfish still occur in Puget Sound shorelines, and can provide substantial
economic value, the industry is most strongly influenced by commercial aquaculture
(PSP 2003c).
Aquaculture in U.S. West Coast waters, although occurring since the late 1800s
(Dumbauld et al. 2009), is comparatively a young practice in regards to ancient practices
in places such as China and Japan, dating as far back as 2,000 years ago (Tidewell 2012).
In recognition of Puget Sound’s local and abundant marine resource, the 1850s marked
the advent of aquaculture (Magoon and Vining 1981). Growth was minimal until the
early 1900s, when wild shellfish stocks, heavily depleted from overfishing, were replaced
by introduced species for planting, such as the Japanese (or Pacific) oyster (Crassostrea
gigas) and Manila clam (Venerupis philippinarum) (Magoon and Vining 1981, PSP
2003a). By the 1950s, to supplement growing human populations with a quality protein
source, the industrialization of aquatic farming escalated and aquaculture on the West
Coast has since grown substantially in the following decades (Dumbauld et al. 2009).
Currently, aquaculture is a significant and still expanding industry in Washington State,
as well as throughout the West Coast of the United States. Washington produces the


 

10
 

single largest amount of cultured shellfish in the nation, comprising 85% of total West
Coast sales and contributing to a $270 million industry (PSP 2012). Oysters are the
largest shellfish crop in the State, followed by clams, mussels, and geoducks (PCSGA
2011).
With such a significant presence of shellfish aquaculture in Washington, many
agencies and organizations at different levels are invested in evaluating its role as both a
stable human food resource and economic industry, and for its implications to coastal
ecosystem functions within Puget Sound. Even State government has taken an active role
to protect Washington’s shellfish resources with the implementation of the Washington
Shellfish Initiative in 2011 to promote sustainable aquaculture economies in agreement
with critically functional aquatic ecosystems (WDOE 2012). Industry members have also
taken pro-active measures to ensure that Puget Sound nearshore ecosystems continue to
support shellfish operations, such as by addressing issues of water quality (PCSGA 2011)
and ocean acidification (WABROA 2012). As a recognized water-dependent use of state
shorelines, under proper management, aquaculture “can result in long-term over shortterm benefits and can protect the resources and ecology of the shoreline” (WDOE 2012).
Method and Operations
The term aquaculture refers to the breeding, rearing and harvesting of aquatic
plants and animals for sales in a market economy and does not include the harvest of wild
stock. There are two defining characterizations of aquaculture that distinguish it from
capture fisheries, intervention and ownership (Lucas and Southgate 2012). As described
by the Food and Agriculture Organization of the United Nations (FAO): “farming implies
some form of intervention in the rearing process to enhance production…” and


 

11
 

subsequently “implies individual or corporate ownership of stock being cultivated”
(2006). Aquaculture is a resource dependent activity, operating in freshwater, marine and
brackish environments of both natural and man-made systems. The sheltered waters of
the Puget Sound region provide an essential resource for marine cultivation in nearshore
coastal environments. Aquatic farming practices in Puget Sound are largely comprised of
the cultivation of bivalve mollusks (eg. oysters, clams and mussels). Lucas and Southgate
(2012) describe that bivalves are an ideal aquaculture species because they can be reared
using relatively simple technology (see below). What is more, because bivalves are filter
feeders, they readily obtain their energy source from surrounding waters, requiring no
additional feed labor following initial placement (Lucas and Southgate 2012).
According to the Pacific Coast Shellfish Growers Association (PCSGA) (2011),
regardless of the type of shellfish, planting and harvesting techniques follow similar
progressions, proceeding through a stage of seeding, cultivation and harvest. To begin,
shellfish seed, or spat, are planted (termed ‘broadcast’) either at the intertidal level, areas
alternatively exposed or submerged in tidal waters, or alternatively in subtidal areas,
where farms are continually submerged in marine waters. Seedling can be derived from
hatchery broodstock, an intensive culture process (i.e. requiring greater labor and capital
costs), or alternatively from natural recruitment, an extensive culture process (i.e. where
stocking rate is low to moderate and capital is limited) (Lucas and Southgate 2012).
Natural recruitment is most widespread, as long as time and locality of recruitment is
known and supportive substrate is available, or provided, for settlement (Lucas and
Southgate 2012). After a period of growth ranging from 1-6 years, matured shellfish are
then harvested, either by hand or mechanically (PCSGA 2011). Although aquaculture


 

12
 

farms may follow these similar progressions, this 3-step outline is extremely simplified,
and variation does occur during their ocean phase as a reflection of differences in biology
and habitat of cultured species, and local factors of farming process such as labor and
costs, and market prices (Lucas and Southgate 2012).
Oysters are sessile, and require a substrate on which to attach for growth, often
preferring hard rough surface that is non-greasy and clear of silt and algae (Lucas and
Southgate 2012). Spat collectors, or culch, used to culture oysters include lime- or tarcoated wooden sticks, flexible plastic strips, oysters shells, cement tiles, tree branches
and bamboo. Similarly, marine mussels non-motile and require firm substrate on which to
attach, generally in the form of fibrous material such as rope. Alternative to oysters and
mussels, clams are a burrowing species, with developed ventral foot that allow travel
throughout particulate tidal beds, from thick mud to sand (Lucas and Southgate 2012).
Cultured bivalves can be fundamentally differentiated by operational method: within the
seabed; on or just above the seabed; or near the ocean surface. The following discussion
outlines common bivalve culture methods, synthesized from descriptions provided by
Lucas and Southgate (2012) and PCSGA (2011).
Within the seabed
Bottom-inhabiting clams are cultivated within natural muddy to sandy intertidal
seabeds requiring minimal intervention. Occasionally, culture area may require
preparation or additional input via seed fertilization and/or disturbance of seabed to
loosen substrate and remove potential predators (eg. starfish). Additionally, following
seeding, culture method often include protective netting applied to surface of intertidal
area to deter predator disturbance (Lucas and Southgate 2012).


 

13
 

On or just above the seabed
Grids of horizontal or vertical stakes (generally wood or bamboo) are used in
intertidal oyster and mussel culture, particularly in sheltered muddy coasts with good
tidal range. Advantages of this method are that installation, maintenance and harvest can
be performed at low tides. Additionally, bivalves can be cultured on racks above the
seabed in mesh boxes or baskets, trays and horizontal wooden and asbestos-cement
batons.
Surface or suspended culture
Surface methods include hanging bivalves on ropes or in culture units from rafts,
longlines and floats. Rafts are rectangular metal, wooden or bamboo frames buoyant
provided air-filled drums or floats. Ropes with attached bivalves then hang down from
rafts, typically densely packed with attached mussels. Alternatively, longlines are 50+
meter ropes supported by floats at the surface in regular intervals and help in place by
terminal anchors. Bivalves are then cultured on vertical ropes suspended from longlines.
Attached end structures varying by species, such as cylindrical or pyramid nets, and
additional roping. Finally, floats are similar to longline systems, but occur most
commonly as large single structures supporting culture system directly below.
Ecosystem Effects of Aquaculture
Puget Sound nearshore environments are among the most complex ecosystems.
Already, extensive accumulations of anthropogenic modifications to Puget Sound’s
coastlines have altered physiochemical and ecological processes that support local human
and wildlife communities (Fresh et al. 2011). It is no surprise, that with increasing tidal
area devoted to aquaculture practices, concerns have sparked questioning of the


 

14
 

ecological role of industry processes to both local environments and extended regional
systems. Shellfish aquaculture, with its growing industry, has increasingly been
recognized as a possible significant source of additional modification to these nearshore
ecosystems, chemically, biologically and physically. Implications of aquaculture
activities to ecosystem functioning have been addressed across scales – from local
alterations to benthic and water column biological communities and chemical processes,
to regional influence on mobile marine animals such as fish and seabirds. As described
by Dumbauld et al. (2009), aquaculture effects on the environment can be described, at
their most basic level, as a disturbance. Using a definition provided by Picket and White
(1985), “a disturbance is any relatively discrete event in time that disrupts ecosystem,
community, or population structure and changes resources, substrate availability, or the
physical environment” (Dumbauld et al. 2009). Although this term may invoke negative
qualities, disturbance merely describes influences to species and ecosystems, but leave
positive or negative value judgments to readers and managers (Dumbauld et al. 2009).
An outline of the most prominent environment effects when addressing the role of
aquaculture is described below. As earlier summarized by Dumbauld et al. (2009) these
effects can be understandably partitioned into 1) material process effects, 2) physical
structure effects, and 3) pulse disturbance effects.
Material process effects
Bivalve mollusks are filter feeders, meaning they filter suspended particulate
matter from the water column, ranging in size from bacterioplankton to less mobile
zooplankton and include both living and non-living material (Dumbauld et al. 2009).
Waste is then expelled in the form of uningested pseudofeces and unassimilated feces,


 

15
 

which sink to the seabed as ‘biodeposits’ (Dumbauld et al. 2009). Although these
cultured bivalves are exploiting naturally occurring matter, thus not resulting in
additional nutrient loading such as with finish culture, presence of aquaculture can adjust
material processes and nutrient cycling (Dumbauld et al. 2009). As a result, these changes
to physio-chemical characteristics can alter estuarine environments through changes to
water quality, sediment properties and resources for primary producers (Dumbauld et al.
2009). The largest question in regards to water quality address particulate matter
depletion and bivalve carrying capacity, as it relates to the availability of phytoplankton
as food and susceptibility of eutrophication (Dumbauld et al. 2009 and references
therein). Concerns of sediment properties largely revolve around accumulation of
biodeposits observed under or within aquaculture operations. For example, local
concentrations of biodeposits can reduce sediment grain size and increase organic
content. In turn, this can also reduce oxygen content and alter nitrogen cycling
(Dumbauld et al. 2009) As a result of these physiochemical changes, local alterations
have been observed in species richness, composition and dominance such as the
displacement of some benthic organisms (eg. urchins), propagation of opportunistic
species (eg. marine worms), and/or changes in local infaunal species (Forrest et al. 2009
and references therein). Alternatively, nutrient and light resources to primary producers
(eg. eelgrass) may not be substantially affected by introduction of aquaculture. Although
eelgrass has shown responses to direct manipulations of nutrients and light, presence of
bivalve aquaculture has not demonstrated significant changes in West Coast studies
(Dumbauld et al. 2009 and references therein).
Physical structure effects


 

16
 

Shellfish aquaculture can influence local environments through creation of
structured habitat, acting as ecosystem engineers or foundation species (Dumbauld et al.
2009). However, it must also be recognized that these artificial structures cannot be
directly compared to natural habitats, such as bivalve reef (Dumbauld et al. 2009). It has
been documented that aquaculture, as ecosystem engineers, have both positive and
negative effects on ecological communities, providing habitat and resources for some
species, while simultaneously displacing others (Dumbauld et al. 2009). A major concern
on West Coast estuaries occupied by aquaculture structures is displacement or alteration
to eelgrass habitat, as these two systems commonly occupy and compete for, similar or
adjacent nearshore space (Tallis et al. 2009). The complex structure and invertebrate
assemblages of eelgrass communities provide valuable habitat for fish and wildlife and
are a major source of detrital carbon for estuarine foodwebs in this region (Simenstad and
Fresh, 1995). Studies of this relationship along West Coast estuaries demonstrate variable
results. One study by Ruesink and Rowell (in prep) as described by Dumbauld et al.
(2009) found that geoduck clams at a south Puget Sound aquaculture site reduced
eelgreass density by roughly 30% during summer months, although this difference
neutralized during winter months when shoot densities thinned naturally in control plots.
Alternatively, concluding this same assessment of West Coast studies, Dumbauld et al.
(2009) found that eelgrass could coexist with shellfish cultured at low densities used in
on-bottom aquaculture on soft sediments.
Off-bottom aquaculture structures have been demonstrated to introduce novel
habitat in some estuarine environments. For example, applied aquaculture structures,
such as stakes and racks, can introduce new attachment sites for growth by wild mussels


 

17
 

(Kirk et al. 2008). Structures may also provide refuge for structure-oriented feeders and
crevice-dwelling fish (Dumbauld et al. 2009). In turn, this could provide additional food
resources for large mobile predators, such as crabs, fish and birds. Most studies outlined
in Dumbauld et al. (2009) document that while aquaculture structures demonstrate
increased community level effects (abundance and diversity) of some fish and
invertebrates compared to open un-structured seabeds, values were still lesser when
compared to adjacent eelgrass habitats.
Pulse disturbance effects
Pulse disturbances are short discrete events. In regards to aquaculture, harvest
practices have been characterized as pulse disturbances with direct implications to sea
grass beds and nearshore sediments. The implications of these disturbances vary by
harvest method, scale of operation and environmental characteristics. Present studies of
harvest method reveal that on-bottom culture has more clearly demonstrated physical
effects during shellfish harvesting (Forrest et al. 2009), generally believed that alternative
off-bottom methods express less-intensive environmental implications. Additionally,
while cultured species can be harvested either by mechanical or physical hand method,
hand methods are considered less intensive to tidal ecosystems (Simenstad and Fresh
1995). Harvest activities have been shown to both directly affect seagrass and associated
organisms, and indirectly affect, through secondary implications, availability of epifaunal
and infaunal prey resources (Simenstad and Fresh 1995). A synthesis of west coast
aquaculture operations by Dumbauld et al. (2009) described that mechanical harvest
methods, including intensive mechanical dredging and suction, resulted in immediate
eelgrass declines of 42% and 96%, respectively. The greatest source of variation in these


 

18
 

studies was believed a result of site-specific recovery responses. whereby the first
dredging study reported a recovery time of 4 years, while the following suction study
reported a recovery time of 2 years. Across studies, recovery time varied as a result of
habitat characteristics, such as seagrass species, sediment attributes, disturbance size,
intensity and treatment responses by industry (Dumbauld et al. 2009).
Policy and Regulations
Following Washington State Authoritative Code definitions, aquaculture means
“the culture and/or farming of food fish, shellfish, and other aquatic plants and animals in
fresh water, brackish water or salt water areas. Aquaculture practices may include but are
not limited to hatching, seeding or planting, cultivating, feeding, raising, harvesting of
planted crops or natural crops so as to maintain an optimum yield, and processing of
aquatic plants or animals” (WAC 332-30-106). To reiterate, for the purposes of this
review, aquaculture is restricted to the cultivation and commercial harvest of shellfish in
marine nearshore tidelands. There are numerous organizations that play a role in the
monitoring and regulation of shellfish aquaculture, operating under the jurisdiction of
local, state, federal and tribal governments. As outlined in PCSGA (2011), the types of
regulation pertaining to shellfish aquaculture include: animal health, economic
development, trade and marketing; environmental quality and health; land use and
shorelines management; navigation and navigable waters; public health and food safety;
resource use; and wildlife and habitat conservation. These regulations can be largely
divided into those pertaining directly to the cultivation and commercial harvest of
shellfish, and those pertaining to activities associated with shellfish aquaculture.
National and Federal


 

19
 

In 2005, the US Food and Drug Administration in collaboration with the Interstate
Shellfish Sanitation Conference created the National Shellfish Sanitation Program
(NSSP). The NSSP is a federal/state cooperative program aimed to promote and improve
the sanitation of bivalve shellfish. The NSSP has since undergone multiple revisions to
maintain most up-to-date research and effective strategies. Another central force of
federal action pertaining to aquaculture comes from the US Army Corps of Engineers
(ACOE) operating under the Clean Water Act (Section 404) and Rivers and Harbors Act
(Section 10) (Dumbauld et al. 2009). Similarly, the National Oceanic and Atmospheric
Administration (NOAA) recognize the broad suite of economic, social and environmental
benefits provided by our nation’s coastal shellfish populations (NOAA 2011). In 2011,
NOAA released the National Shellfish Initiative in conjunction with National
Aquaculture Policy focused on increasing bivalve shellfish populations through
commercial production and conservation activities (NOAA 2011). To achieve this,
NOAA acknowledged and promoted increased collaboration with public and private
partners to address marine planning and permitting, environmental research, restoration
and farming techniques and coordinated and innovative financing (NOAA 2011).
State
In 1997, the Washington’s Shorelines Management Act was adopted to “prevent
the inherent harm in an uncoordinated and piecemeal development of the state’s land. “
Further, to ensure that “the interests of all the people be paramount in the management of
shorelines of statewide significance” including Puget Sound shorelines and waters
(WDOE 2012). Under this act shellfish aquaculture is defined as a water-dependent use,
and further described as a preferred use following that operations and activities aare


 

20
 

consistent with the control of pollution and prevention of damage to the natural
environment (WDOE 2012).
Succeeding national and federal initiatives, Washington State has also taken
additional action to protect and enhance shellfish resources with the creation of
Washington Shellfish Initiative in 2011. Like the national initiative, Washington
recognizes the benefits of shellfish protection, restoration and enhancement efforts to
increase recreation and jobs and ensure a healthier Puget Sound (WSI 2011). Dominant
jurisdiction operating over Washington’s shellfish resources and aquaculture is two-fold,
fulfilling both state wildlife and sanitation requirements. The Washington Department of
Fish and Wildlife (WDFW) operates under wildlife and habitat conservation policies to
regulate commercial fishery and aquaculture activities through aquatic farm registration
requirements. It is required that all aquatic farmers must first register their operations
with WDFW. However, prior to 2008, all operations could be registered under a single
application with no itemization of specific operations and activities (Galivan, P. pers.
communication). Additionally, the Washington State Department of Health authorizes
shellfish aquaculture activities to ensure safe growing and consumption of cultivated
product (DOH 2012). DOH issues shellfish operation license for acting farm in addition
to harvest site certificate per intended site of harvest. While the majority of shellfish
aquaculture activities occur on privately owned tidelands, still approximately 30% of
intertidal lands and nearly all subtidal lands are owned by the State and regulated by
Department of Natural Resources (DNR). DNR may hold contractual agreements of
leased aquatic lands with growers, and although the agency does not have a regulatory
role in aquaculture practices, it does actively enforce written lease conditions.


 

21
 

Policy Gaps and Needs
Consequently, there are many overlapping jurisdictions that may cause for
complex permitting pathways (PCSGA 2011) for both the subjective farm and regulatory
organizations. Acting agencies across regulatory levels acknowledge the necessity of
addressing the complexity of aquaculture policy and planning to effectively maintain the
viable economic contribution of shellfish aquaculture while continuing to meet the
various stewardship responsibilities (Dewey et al. 2008). In 2007, Washington State
Legislation passed house bill 2220 in regards to shellfish aquaculture in coastal waters.
This bill took action to confront three major needs in aquaculture policy and process: 1)
create a Shellfish Aquaculture Regulatory Committee representing a wide range of
perspectives and interests; 2) increase scientific research and advisory action on
expanding geoduck aquaculture in Puget Sound; and 3) expand information required of
aquaculture farms on WDFW registration documents (HB 2220).
Sea Ducks and Aquaculture
Shellfish aquaculture and sea duck populations occupy similar intertidal and
subtidal habitat within coastal ecosystems. The question reasonably follows, could the
significant and expanding aquaculture industry, such as in Puget Sound, influence
functional habitat use by wintering sea ducks? As a result, are trends in population
distribution and density of sea duck correlated to changing aquaculture landscapes?
Research demonstrates that aquaculture can directly modify habitat chemically,
biologically and physically and these modifications have the capacity to reverberate
indirectly through ecosystem processes. Further still, these implications may cascade to
higher trophic levels - influencing larger mobile epibenthic predators, such as sea ducks.


 

22
 

As a result, studying the response of sea ducks may provide an indicator of the
environmental response of estuarine ecosystems to commercial aquaculture. Previous
investigations into the relationship between sea ducks and shellfish aquaculture have
documented positive, negative, and neutral responses, as outlined in the next section.
Sea Duck Responses to Aquaculture
Positive
One key beneficial role of aquaculture for sea ducks is the addition of new prey
resources. Several studies show that off-bottom aquaculture farm structures may provide
novel recruitment surface for wild mussel species, as a result providing additional prey
populations for predatory sea ducks (Zydelis et al. 2009, Kirk et al. 2007). A study in
Desolation Sound, B.C., found that densities of wild mussels were much greater on
floating aquaculture structures than measurements taken in adjacent unstructured
intertidal areas (Tallis et al. 2009, Zydelis et al. 2009). What is more, structure – grown
mussel species displayed altered morphological traits, such as weaker byssal attachment,
and more fragile and thinner shells, making them a more advantageous foraging decision
for sea ducks (Kirk et al. 2007, Zydelis et al. 2009). This translated into observed positive
relationships between sea duck densities and off-bottom aquaculture sites in winter
habitats of sea ducks in coastal BC (Zydelis et al. 2009). For some industry members,
these events represent a mutually advantageous relationship where bird populations are
supplied valuable prey resources and in turn, clear aquaculture structures of nuisance
mussel species (Kirk et al. 2007). Additionally, one study by Caldow et al. (2003)
comparing bird assemblages in experimental areas of greater intertidal mussel density (to
represent areas of artificial increase in mussel species for cultivation) and adjacent


 

23
 

unaltered intertidal areas, demonstrated an increase in some opportunistic nearshore bird
species due to increased habitat complexity. In other cases, on-bottom aquaculture farms
(i.e. clams) provide prey by way of cultivated species (Price and Nickum 1995).
However, this negatively affects shellfish growers’ stock and is often prevented by
application of anti-predator nets or physical deterrents within culture areas (Lucas and
Southgate 2012).
Negative
Aquaculture could negatively impact sea duck populations in Puget Sound
through direct habitat exclusion or alterations (Dumbauld et al. 2009); introduction of
invasive and/or non-native marine invertebrate species (Tallis et al. 2009), displacing
historical prey choice (Caldow et al. 2003); and introduction of toxins, parasites and
disease with consequential contamination and bioaccumulation in sea ducks (Buchanan
2006). The most studied source of detriment to sea duck populations from aquaculture is
degradation or alteration of critical foraging habitat (Connolly and Colwell 2005).
Studies by Simenstad and Fresh (1995) and later supported by Tallis et al. (2009)
demonstrate the negative impacts of on-bottom aquaculture methods, including dredging,
harrowing and leveling of intertidal areas for farming. Artificial application of additional
substrate and subsequent mechanical harvest of cultured shellfish (i.e. dredging) can
displace nearshore birds by direct disturbance, and alteration or exclusion of valuable
eelgrass habitats and prey landscapes (Simenstad and Fresh 1995). Additionally, major
findings by Bendell-Young (2006) reported that areas along the coast of B.C. with
greatest intensity of aquatic farming demonstrated a decrease in species richness, altered
bivalve composition, abundance and distribution, and change in community intertidal


 

24
 

structure. Bendell-Young (2006) continued, that the consequences of aquaculture could
potentially restrict access to sea ducks during key over-wintering periods, exacerbating
already declining West Coast populations.
Neutral
In a study by Zydelis et al. (2006), efforts found that in Baynes Sound, supporting
50% of B.C.’s aquaculture, despite extensive clam and oyster farming, wintering scoter
densities were largely a variable of natural environmental attributes, including extent of
intertidal zone and substrate type. Zydelis et al. (2006), also noted that this could be
attributed to the significant and stable source of natural clam prey populations
documented in Baynes Sound, thus exclusion from some areas (i.e. anti-predator netting
and direct disturbance) did not result in significant displacement. This supports the
suggestion that the aquaculture industry and sea duck populations may be capable of
mutual sustainability. An earlier study by Caldow et al. (2003) also demonstrated neutral
effects of aquaculture on some marine bird assemblages where alternative habitat is
readily available. As demonstrated, some undetectable responses of sea duck populations
to nearshore aquaculture practices were a result of greater variables at play, covariation of
environmental and aquaculture variables and limited data availability. The abundance of
neutral responses seen in studies directly assessing the affect of aquaculture on nearshore
birds suggests a need to continue comprehensive monitoring efforts at greater spatial and
temporal scales to tease out the effect of natural environmental differences (Forrest et al.
2009) and variability among site-specific studies (Caldow et al. 2003).


 

25
 

Conclusions
Puget Sound is an unparalleled resource in Washington and along the US West
Coast, ecologically and economically, providing for the productive and expanding
shellfish aquaculture industry, and supporting vital nearshore ecosystems. Consistent
with expansion, there is increasing evidence of threatened components of Puget Sound’s
nearshore ecosystems (Fresh et al. 2011, PSP 2012), such as declining sea duck
populations (Buchanan 2006, WDFW 2010). Although documentation still remains
limited in explorations linking the ecological role of aquaculture to nearshore bird
populations, concurrent use of nearshore habitats is evident. The question remains, how
and to what degree does shellfish aquaculture play in determining habitat use and
distribution of winter sea duck populations in Puget Sound (Evenson 2012).
It is important to recognize the inherent variability of outcomes in studies
addressing environmental effects of shellfish aquaculture and its role in sea duck
population patterns. Continued work is required to characterize and project the ecological
importance of the intertidal zone (Bendell-Young 2006), the role of aquaculture in those
coastal ecosystems (Dumbauld et al. 2009) and the wider implications of aquaculture for
wintering sea duck populations. There is no single driver of documented declines in
winter bird populations, but instead investigations must explore all facets of dynamic
estuarine ecosystems and assess the role of human impacts. For example, research efforts
must address historically underrepresented over-wintering areas (SDJV 2012), as a
substantial portion of a species’ annual cycle is spent in non-breeding areas (Gaydos and
Pearson 2011). Additionally, monitoring and assessment efforts must continue to address
smaller scale local systems, such as critical habitat areas, while maintaining connectivity


 

26
 

to larger scale systems, such as regional population trends (Anderson et al. 2009).
Anderson et al. (2009) elaborates “identifying causes of population decline is especially
difficult for species that migrate substantial distances among distinct habitat used over an
annual cycle because it is unclear how changes in resources versus non-local factors have
contributed to declines”.
There is an increasing demand for integrated approaches to managing coastal
ecosystems. Decisions affecting coastal resources are fragmented among many
organizations, often proving inefficient, resulting in conflict among sectors, and
contributing to management gaps and overlaps (PCSGA 2011). Alternatively,
collaborative work among different organizations including governmental (e.g. WDFW,
DNR, NOAA) tribal, non-governmental, universities, independent research institutions
and aquaculture industry members (e.g. SDJV, PSP, PSNERP, PSI, PCSGA), addressing
aquacultures impact to nearshore ecosystems are necessary to effectively create and
execute productive and sustainable outcomes. In a similar vein, collaborative working
relationships among organizations have also been promoted in sea duck conservation and
management objectives outlined by SDJV (2012) and WDFW (2010), to better
coordinate monitoring and assessment efforts and identify key research links throughout
regional populations. In this way, the best solutions to define sea duck – aquaculture
relationships can only be achieved through interdisciplinary, interorganizational and
international approaches to research, conservation, management, and industry actions.


 

27
 

CHAPTER TWO
MANUSCRIPT
Introduction
The Puget Sound region is an unparalleled resource in Washington State –
supporting productive marine ecosystems, providing thriving coastal economies and
defining many social and cultural identities and traditions. In particular, the nearshore
environments that blanket Puget Sound shorelines contain some of the most abundant
marine resources and highlight the complexity of balancing productive market economies
with maintaining functional coastal ecosystems. Shellfish aquaculture is a historically
significant and expanding industry in Puget Sound, dependent on these nearshore
environments. Further, the Sound’s complex of bays and inlets provide important habitat
to more than 70 species of bird critically dependent on marine resources for a significant
portion of their life-histories (Essington et al. 2009). Consequently, this interface of
overlapping, and at times competing use necessitates comprehensive monitoring and
assessment efforts to identify and evaluate wildlife responses to a variable network of
natural and anthropogenic landscapes.
Aquaculture – as used here, the aquatic farming of marine bivalve shellfish – is a
long-standing coastal activity in Puget Sound. The extensive sheltered bay and inlets
provide productive grounds for the artificial cultivation of shellfish, predominantly
comprised of clams, oysters, geoducks and mussels. Over the last several decades, in
response to growing human population demands and aided by the advancements of
culture method and operation (Magoon and Vining 1981), aquaculture has seen a
substantial growth in industry. Currently, aquaculture in Washington provides the


 

28
 

Nation’s top production of cultured clams, oysters and mussels, and further supports the
state with an annual $270 million industry (PSP 2012). The valued economy and tradition
of shellfish aquaculture in Washington is clearly recognized with National and State
initiatives in place to encourage the growth and vitality of industry (PSP 2012). Since the
farming of shellfish as a water-dependent use relies on functional marine environments, it
is recognized that this may also provide industry incentive to adhere to and encourage
issues of water quality and pollution in their tidelands (PCSGA 2011). However, as the
support and growth of shellfish aquaculture continues to expand in Puget Sound, it is
important that industry and resource management agencies are able to clearly define the
ecological role of operations on nearshore habitat and wildlife.
Sea ducks in Puget Sound are a characteristic marine bird – as a species intimately
associated with the nearshore environment. The complex of estuaries that make up Puget
Sound define important habitat for over-wintering populations. However, varying levels
of long-term decline in many wintering sea ducks have sparked concerns by conservation
and management agencies (Essington et al. 2009). This concern is exacerbated “given the
number of sea duck species for which basic biology is poorly known or unknown, their
broad and remote distributions, and the unique challenges of studying sea ducks” (SDJV
2012). Consequently, reasons for these declines are largely unknown. A key step in
understanding and mitigating declining populations is the development of a
comprehensive assessment of habitat use, needs and availability, in particular in Puget
Sound (WDFW 2012). Because sea ducks that utilize Puget Sound do so for a significant
portion of their annual cycles (Gaydos and Pearson 2011), the importance of quantifying
the extent of impacts to non-breeding areas is a vital first step to describing population


 

29
 

dynamics and habitat use (SDJV 2012). One facet of defining the habitat use, needs and
availability of sea ducks is to identify and evaluate possible limiting factors. As the
nearshore environments of Puget Sound provide for both thriving aquaculture industries
and critical winter sea duck habitats, the probability for interaction is high. As a result,
concern to identify overlapping distributions and competing uses is of reasonable
consequence.
Past studies evidencing implications of this overlap have documented varying
levels of positive, negative and neutral sea duck – aquaculture relations. Shellfish
aquaculture may negatively influence marine birds species by direct exclusion or
modification to habitat (Simenstad and Fresh 1995), alteration of biological communities
(Bendell-Young 2006), or by the physical presence and disturbance associated with
culture operations (Forrest et al. 2009) However, several studies documenting positive
associations of sea ducks and shellfish aquaculture describe advantageous bird responses
to novel prey sources provided by industry operations through the introduction of
structure-grown mussels (Kirk et al. 2007), natural dispersal of planted shellfish seed
(Caldow et al. 2003, Zydelis et al. 2006), or direct consumption of cultivated species
(Zydelis et al. 2009). Alternatively, lack of response to aquaculture by overlapping bird
species suggests alternative forces describing species behavior and fitness and overall
population dynamics (Zydelis et al. 2006, 2009). Overall, several studies suggest
responses to aquaculture are species specific (Caldow et al. 2003), most likely a
reflection of variable behavior, feeding and habitat associations among subjects
(Connolly and Colwell 2005).


 

30
 

Previous studies suggest sea ducks respond in varied ways to shellfish
aquaculture; these findings highlight the need to continue and advance research efforts.
As this valued industry further expands its role in nearshore ecosystems of Puget Sound,
it is important to assess the wider ecological influences to overlapping marine bird
populations. Further, it is advantageous to expand the scope of past studies to incorporate
broad-scale spatial and temporal variability to better understand sea duck population
dynamics and assist in the mitigation of observed population declines. Our research
explored how location and extent of an expanding shellfish aquaculture industry affects
winter sea duck populations in South Puget Sound. Our objectives were to 1) define the
nature and degree of sea duck-aquaculture relations; 2) describe how responses vary
among sea duck species, and 3) explore how these relations trend over time in response to
a changing aquaculture landscape.


 

31
 

Methods
Study Area
Puget Sound is the second largest estuary in the United State, draining 19 major
river basins and comprising 2,800 square miles of marine waters. Extending inland from
the Pacific Ocean on the west to the Canadian border to the North, the sound is bordered
by roughly 2,500 miles of shoreline. The Puget Sound is also home to nearly two-thirds
of the total state population of Washington, and consequently human activities heavily
impact these coastal environments (Fresh et al. 2011). This interface of terrestrial and
aquatic ecosystems makes up the Sound’s nearshore environments, including the coastalmost upland riparian area, through the intertidal and extending just beyond the subtidal
zone (Fresh et al. 2011). Due to its vast and complex structure, the Sound is often divided
into six subregions for monitoring and research efforts. Following delineations outlined
by the Puget Sound Assessment and Monitoring Program (PSAMP), our research is
focused on describing aquaculture and sea duck characteristics in South Puget Sound,
encompassing the southern-most inland waters, south of the Tacoma Narrows to
Olympia, WA. With such a large estuarine system, variation in the biophysical
environment exists, often extending in a North-South gradient. South Puget Sound is
more sheltered from the influence of the ocean than its corresponding central and
northern regions, characterized by shallower waters, weaker circulation and less saline
conditions than the northern exposed region (Gustafson et al. 2000).
These shallow bays and inlets that make up South Puget Sound provide suitable
muddy to sandy sediments for both the cultivation and natural propagation of bivalve
shellfish (Simenstad and Fresh 1995). Currently, Washington State shellfish aquaculture


 

32
 

operations produce the nation’s greatest number of farmed oysters, clams and mussels
(PCSGA 2011). Puget Sound holds the second greatest concentrations of shellfish
aquaculture operations in Washington, following closely behind the outer coast (PCSGA
2011). Since its industry inception in the mid 1900s, aquaculture operations have
expanded substantially to meet growing human demand (Magoon and Vining 1981).
Within Puget Sound, individual shellfish operations can vary by species cultured and
method of culture as a result of local regulations, environmental conditions, and market
demands (PCSGA 2011).
Study Organisms
Shellfish
Cultured clam species predominantly include Manila (Venerupis philippinarum),
with additional farming of Butter (Saxidomus gigantean) and Littleneck (Protothaca
staminea). Two main forms of clam cultivation are in operation in Puget Sound, ground
and bag culture. Ground culture involves species grown directly in the intertidal
substrate, and often covered by protective netting to prevent loss by predation (i.e. crabs
and ducks) (PCSGA 2011). Alternatively, bag culture includes clams grown in bags set
directly in the intertidal beaches or suspended from racks or trays within both the
intertidal and subtidal zones (PCSGA 2011). Harvesting of clams in South Puget Sound
is most often conducted by hand rake methods. More recently, Geoduck clams (Panopea
generosa) have also become a significant subject of aquaculture activity in Washington
State. Geoduck farms are typically located low in intertidal zones, utilizing a network of
PVC pipes buried vertically into the substrate. Seeds are then added to each pipe and
protective netting is placed individually over each tube or completely over the entire


 

33
 

operation. Geoducks are harvested using a pump and hose, pressure-injecting water into
each pipe to loosen the clam from the substrate for collection (PCSGA 2011). Cultivated
mussel species include Blue (Mytilus trossulus) and Mediterranean or Gallo (Mytilus
galloprovincialis). The majority of West Coast mussels are grown suspended from rafts
or surface long-lines in the subtidal zones. These rafts are typically constructed of
lumber, galvanized steel and plywood and afloat via plastic barrels or foam. Raft
structures may also periodically utilize protective netting. Surface long-lines are
commonly made of heavy plastics or nylon suspended by floats or buoys anchored and
attached at intervals. Bags or lines are removed and harvested when appropriate market
size is reached (PCSGA 2011). Oyster culture provides the greatest source of variation in
species farmed, including Pacific (Crassostrea gigas), Native (Ostrea lurida), Kumanoto
(Crassostrea sikamea), Eastern or American (Crosostrea virginica) and European flat
(Ostrea edulis). Consequently, oyster culture also demonstrates greatest variability in
operational methods among both ground and suspended culture, including bag, rack and
bag, long-line and stake culture. Variation in oyster cultivation method most often
depends on the target market, farmed for half-shell, ornamental or shucked meat (PCSGA
2011).
Sea Ducks
In addition to the economic values provided, Puget Sound provides abundant
habitat for marine birds of the US West Coast, and is one of the most important wintering
areas in the eastern Pacific (Nysewander et al. 2005). The Puget Sound supports
approximately 70 species of marine bird that depend on the marine environment for all or
a significant portion of their life histories (Essington et al. 2009). In particular, the


 

34
 

sheltered inland marine waters of South Puget Sound provide substantial nearshore
habitat for over-wintering populations of sea ducks – a group of diving duck intimately
associated with the marine environment. Nine species commonly occur within Puget
Sound, including: black scoter (Melanitta nigra), surf scoter (M. perspicillata), whitewinged scoter (M.fusca), common goldeneye (Bucephala clangula), Barrow’s goldeneye
(B. islandica), bufflehead (B. albedo), common merganser (Mergus merganser), hooded
merganser (Lophodytes cucullatus), and red-breasted merganser (M. serrator). Further
description of these nine species is provided in Table 1, Appendix A. Sea ducks, as game
species managed under state and federal migratory waterfowl regulations are subject to
monitoring and management efforts to maintain sustainable population numbers (WDFW
2010). However, as described, habitat assessments are limited, thus identification of
possible sources of limitation and stress are lacking (Evenson 2012).
Data Sets
Bird Survey Data
The Washington Department of Fish and Wildlife conducts annual aerial surveys
of marine birds throughout Puget Sound, as part of the Marine Bird and Mammal
Component of the Puget Sound Ambient Monitoring Program (PSAMP). Surveys began
in 1992, including both parallel nearshore (<20m) and zig-zag off-shore (>20m)
transects, estimated to cover 16% to 19% and 3% to 4.5% of total area, respectively
(Nysewander et al. 2005). Efforts were taken to minimize possible sources of variation
inherent in aerial surveys of waterfowl (Butler et al. 1995) by using experienced pilot
biologists and observers, consistent timing and trajectory, and using the same aircraft
throughout. The floatplane flew at 80-90 knots at an altitude of approximately 65 meters


 

35
 

above sea level. Two observers positioned on each side of the aircraft recorded
occurrence, species, GPS location and time within a 50 m wide search area (Nysewander
et al. 2005). Separate databases were obtained through DFW for each winter survey
season: 1994-95, 1995-96, 1996-97, continuous to 2012-13. For analysis purposes, each
season was identified by its December year (eg. 1994-95 renamed to 1994). Within each
year, anecdotal observations of bird counts while off transect were removed from
representation and analysis. Bird surveys from the first 2 years were removed from
analysis to eliminate variation, as method and operations were still developing at its
onset. Therefore, bird survey data was pulled from the 1994/1995 winter season to the
most current 2012/2013 winter season. Survey data was used to examine several
parameters of sea duck populations, including counts, density indices and associated
confidence limits. WDFW identifies our focal sea duck species by the following codes:
Surf Scoter (SUSC), Black Scoter (BLSC), White-winged Scoter (WWSC), Unidentified
Scoter (UNSC), Common Goldeneye (COGO), Barrow’s Goldeneye (BOGO),
Unidentified Goldeneye (UNGO), Common Merganser (COME), Hooded Merganser
(HOME), Red-breasted Merganser (RBME), Unidentified Merganser (UNME), and
Bufflehead (BUFF) (Table 1, Appendix A). Final data relevant to GIS and subsequent
analysis included: survey year, latitude – longitude coordinates, on/off transect
designation, species observed, and count per species observation.
Aquaculture Data
A long-term inclusive database of shellfish aquaculture operations and activities
in SPS was non-existent before this research. With multiple agencies exercising
permitting and regulatory authority over shellfish aquaculture operations, information


 

36
 

exists across State agencies and agency departments. Additionally, these multiple
agencies lack a concerted effort in data collaboration and sharing. Consequently, efforts
to compile and construct a central database required cross-referencing of multiple
resources. All information was acquired from public document sources provided by the
State. Initial data was obtained from Washington Department of Fish and Wildlife
(WDFW) aquatic farm registration forms, required under WAC 220-76 to be completed
by aquaculture farms prior to commencement of culture activities. Ideally, this
application would describe aquatic farm information per individual operation including:
location - by parcel number, site address and/or section, township and range; size amount of acres under cultivation; species cultivated; and method of culture.
Realistically, information provided by WDFW public records request regarding shellfish
aquaculture operations and activity in South Puget Sound over the last 20 years produced
a database with substantial gaps (explained in further detail in the Discussion section).
Therefore, to fill in data gaps, additional information was obtained from Department of
Health (DOH) harvest certificate applications, required under WAC 246-282 by
companies or individuals who harvest a commercial quantity of shellfish or any quantity
for human consumption. Through DOH harvest certificates, information provided
included: location by parcel number, site address, and/or section, township and range;
acres harvested; and species harvested. Available information received more often than
not remained in raw form, unsuitable for analysis requirements of my research.
Consequently, extensive quality assurance/control (QA/QC) was undertaken to convert
information to usable data for statistical analysis. Significant gaps existed in WDFW data
provided including: lack of parcel ID (as this is optional yet preferred method of location


 

37
 

identification); incorrect parcel ID as a result of inaccuracy provided by organization,
transferring of data to digital record, or computer formatting errors; and missing acreage
information. The final collated database relevant to supplementary analysis consisted of
the following: year of operation, parcel ID, total acres cultivated, acres cultivated by
species (i.e. clam, geoduck, oyster, mussel). Data was formatted for conversion to a GIS
geodatabase for spatial representation and analysis (operations described in detail below).
Spatial Analysis Using Geographic Information Systems
The development and use of spatial databases in studies of both bird populations
and aquaculture is increasingly recognized as a valuable method of identifying and
analyzing spatial and temporal patterns and trends. As described by Simms (2002), a GIS
1) provides the capability to integrate, scale, organize and manipulate spatial data from
many differences sources; 2) can be manipulated, updated, extracted and mapped
efficiently; and 3) permits quick and repeated testing of models which could be used to
aid in decision-making processes. In particular, use of ArcGIS 10 (ESRI, Inc. 2013) for
analysis allowed for the identification and evaluation of the location and extent of
shellfish aquaculture activities and its influence on sea duck distribution and abundance
over an extended period of time.
Shorezone Sampling Designation
To create a base layer for analysis begin, a sampling framework was created to
partition South Puget Sound nearshore areas into manageable units of observation. The
entirety of Washington state shorelines has been mapped and made available by the
Department of Natural Resources via online GIS data center, as part of an effort to
provide a baseline measurement system for any coastal assessments. This shoreline arc


 

38
 

was first clipped to our study area, South Puget Sound, comprising sheltered marine
waters south of the Tacoma Narrows Bridge. For the purposes of this study, projection
and analysis did not require the intricate representation of the many small bays and inlets
drawn in the DNR shoreline. As a result, the simplify line tool was used to create a
coarser scale map while retaining basic geometry. Using the simplified shoreline, a 500
meter buffer polygon was drawn, as the estimated relevant extent of interaction between
our variables, established in past methods of analyzing PSAMP bird count data in
response to shoreline attributes (Rice 2007). This buffer was used to identify nearshore
parcels for joining of aquaculture sites, discussed in further detail below. In establishing a
sampling grid for South Puget Sound, to be uniformly sampled each year, the buffer
polygon was first restricted to the 500 m portion extending into the Sound to eliminate
irrelevant upland area. This operation was performed using the split polygons tool, where
all resulting island and upland polygons were deleted, leaving only the tidal 500 meter
polygon. Efforts to further split the shoreline polygon to subsets of roughly equal areas
were performed manually. First by generating points every 2 km along the simplified
shoreline arc, and second by manually drawing lines extending perpendicular from these
points snapping to the edge of the 500 m tidal polygon. Concerted effort was taken within
narrow bays and inlets to insure as uniform area as possible. This resulted in the
production of 192 shoreline polygons, regarded as our sampling areas where summaries
of variable values would be calculated (Figure 2).
Aquaculture Projection
County parcel polygons were obtained from the three counties surrounding South
Puget Sound: Mason, Thurston and Pierce. Rather than operating across three separate


 

39
 

county assessor offices, a single shapefile representing county 2012 parcel data was
provided by the Parcels Working Group at the University of Washington (UWP 2012).
This is a collaborative project aimed at promoting the development and coordination of
federal, state tribal and local governments to produce a statewide parcel framework
accessible to various participating agencies and interested parties (UWP 2012). Still, the
parcel shapefile contained many upland parcel entities unnecessary for my work, and
parcels were clipped along the 500 m shoreline buffer, to aid in focused and efficient
processing. A separate parcel layer clip was created for each year, 1994-2012. Each layer
was then joined to its corresponding year of aquaculture activity by county parcel ID,
choosing to keep only those records where a join was successful. The function failed to
join 152 of the total 888 aquaculture parcels identified in the database. This is an error we
accepted as a result of misidentification on the aquatic farm registrations, or changing
parcel IDs (eg. splitting, moving, or renaming) over the last 19 years. Representing a
minimal proportion of total activity, the data we retained still accurately represented
activities. Next, all years were merged, and polygons centroids were converted to points
using the feature to point tool. Because parcel polygons often extend perpendicular to
shorelines, this feature tool placed some points upland of the shoreline where no
aquaculture activity is actually occurring. To remedy this, we used the near tool to
generate new coordinates of each point nearest to the simplified shoreline arc. The near
distance traveled for each point was checked to ensure location was not altered by more
than half of our shoreline polygon size.
Bird Survey Projection


 

40
 

To project bird surveys into a GIS, an x,y feature class was created from geodetic
observation coordinates for each year. Data originally represented transects flown
throughout the entire Puget Sound Region, so the same 500m shoreline buffer was used
to clip points to our study area within South Puget Sound. Complete marine bird
observations were further concentrated to our focal nine sea duck species. Like the
aquaculture data, all 19 years of bird survey data were then merged to a single point
shapefile (unidentified spp. included). For each year, anecdotal observations made while
off transect were removed from representation and analysis. Additionally, several small
inlet areas identified by WDFW were excluded from representation and analysis as these
zones were surveyed only during initial years but later excluded from subsequent aerial
transects for safety reasons. Sea duck data existed in stacked columns, with species
identified in one and corresponding count adjacent. Therefore, new columns were added
to the final merged table to combine our nine species and three unidentified species into
four species groups and/or species. These species groups, and included species, are as
follows: Scoters – Black, Surf, White-winged and Unidentified Scoter; Mergansers –
Common, Hooded, Red-breasted and Unidentified Merganser; Goldeneyes – Common,
Barrow’s and Unidentified Goldeneye; and Bufflehead. Summed count values were
transferred to their appropriate species group row.
Data Overlay Operations
The uniform sampling grid constructed initially from the buffered shoreline
polygon was used to integrate our aquaculture and bird survey data for further analysis.
Using the framework of the 189 polygons, an identity analysis was performed to append
a unique corresponding polyID to each aquaculture point. As bird points did not always


 

41
 

align directly with ground point shoreline boundaries (an inherent effect of aerial
surveys), the spatial join tool was used to append polyIDs to each bird points using the
‘closest’ specification, where each observation point joined to the closest shoreline
polygon. As a result, each aquaculture and bird point observation over the last 19 years
was associated with a specific polygon ID. The two point features, aquaculture and bird,
were merged a final time under common year and polyID fields, where each grid cell per
year was now correlated with information regarding its use by shellfish aquaculture and
by sea duck populations. Using the summary statistics tool, a table was generated to
summarize activities per grid cell by year, quantifying the following attributes: 1)
frequency and total counts of each four species groups, 2) count and total acreage of
shellfish aquaculture, and 3) total acreage of each cultivated species (geoduck, clam,
oyster, mussel). Not every grid cell had observations of both aquaculture and bird counts,
therefore NULL values were given to those cells and/or years where no sea ducks were
observed. However, under the assumption that our data represents the best estimate of
shellfish aquaculture activities within South Puget Sound, those grid cells with no
observed aquaculture were converted to zero acreage. Summaries identified 3 study site
polygons with limited annual observations at 3, 6 and 9 out of the total 19 years of
observation, while all other polygons identified 17-19 years of observation. These 3 study
sites were consequently removed from any further analysis (Figure 2). This tabular data
was then exported from GIS for further statistical analysis.
Statistical Analysis
Research Questions:


 

42
 

1) Do winter sea duck distributions and abundances demonstrate a relationship to
the location and extent of shellfish aquaculture in South Puget Sound?
a) If so, what is the direction and magnitude of associations?
b) How do different species groups respond to aquaculture?
c) How do sea duck-aquaculture relationships vary over the last 19 years?
Tabulated data generated in ArcGIS was exported to Excel as an intermediate
before further statistical analysis was conducted using JMP Pro 10 (SAS Institute, Inc
2012). Final data was charted in a stacked format, with individual rows representing
observations by each unit polyID, further regarded as study site, for each year 1994-2012.
Due to the limitations of PSAMP method, data produced from surveys cannot be used to
explicitly determine population numbers, but instead can provide index values to
characterize and evaluate observed spatial and temporal trends (Nysewander et al. 2005).
Indices are often used in studies of animal populations where absolute values are rarely
measured. Therefore, relative abundances per year by site for each species group were
calculated as a proportion of total abundance. Descriptive analyses were run to explore
overall summaries by year, by study site and by species group to uncover overarching
trends and identify possible sources of errors. To eliminate unnecessary extreme variation
and aid in analysis, aquaculture acreages were defined in four levels, based on frequency
of observation: zero, where no aquaculture is present; small, 0 < x ≤ 5 acres; medium, 5 <
x ≤ 25 acres , and; large, x > 25 acres.
To understand and predict animal-landscape relationships over longer time
periods and larger spatial areas, our analytical approaches must incorporate temporal
variation in explicit and robust ways (Gutzwiller and Riffell 2007). Analyzing


 

43
 

aquaculture-sea duck associations within a single time frame, or summarizing values
across a specified time span would not allow for the interpretation of chronic influences
and spatial variability of an expanding aquaculture industry on changing sea duck species
populations. Several statistical methods to incorporate time measures are available,
however the nature and restrictions of our data made a mixed model repeated measures
analysis the ideal choice for several reasons. Importantly, mixed models for repeated
measures allow opportunity to incorporate simultaneous inferences about time and space
in studies of animal-landscape relations (Bissonette and Storch 2007), where our analysis
is evaluating the temporal variations within and among our different study sites
throughout South Puget Sound. In part, this is because repeated measures controls for
non-independence among temporally repeated observations, often where varying levels
of data are subsampled within experimental units. Further, mixed model approaches offer
greater flexibility in analysis on multiple scales, in particular unbalanced data where
some observations may be missing (Gutzwiller and Riffell 2007). This was applicable in
our case where some study sites lacked a complete 19 years of observation values.
A mixed model repeated measures analysis was used to determine the effects of
year and aquaculture acreage on abundance and distribution of each of our species groups
– Bufflehead, Goldeneye, Scoter, Merganser (JMP Pro 10). To set the foundation for our
analysis, a description of units, factors, and effects are described below. There are two
experimental units that must be accounted for in our analysis, both study site and year.
Again, the 189 polygons created in GIS act as our permanent study sites, or subjects. Our
analysis aims to identify both within-subject and between-subject factor. In a repeated
measures analysis, the effect of year acts as our within-subject factor, with 19 levels. Our


 

44
 

between subject factor is acreage size class, with 4 levels. Both acreage and year act as
our main fixed effects. Additionally, under a mixed model approach, random effect(s)
need to also be identified, under the assumption that this effect represents a sampled
estimate of total values. Our experimental unit for acreage size class is our individual
study sites. Therefore, acreage size classes cannot be randomized to sites, as they are
specific designations of each polygon. Consequently, we assume sites selected are a
random sample from the corresponding acreage size classes representative of South Puget
Sound. In our analysis, model effects of interest are as follows:
acreage size class
site[acreage size class] & Random
year
year*acreage size class
Where acres, year and the interaction of acres x year (time) are our fixed predictor effects
of interest and, to account for spatial variation, where site nested within acres acts as our
random effect. This is due to the assumption that sample sites represent our entire South
Puget Sound study area, thus providing structure for our estimated effect of aquaculture
acreage. The inclusion of a random effect consequently runs our analysis using restricted
maximum likelihood (REML), a method of parameter estimation restricted to
maximizing the likelihood function over the random effects portion of the model
(Gutzwiller and Riffell 2007). This random effect calculates subject effects, accounting
for variation in acreage both within and among sites. Our model tests how species
abundances respond to aquaculture acres, to temporal variability, and how species
respond to aquaculture over time.


 

45
 

Using the above analysis, we tested the following specific hypotheses for each of
our species groups:
The test for interaction is:
H: there is no interaction between the effects of aquaculture acreage and year on
the relative abundance of species groups
A: there is an interaction between the effects of aquaculture acreage and year on
the relative abundance of species groups
The test for our main effect of year is:
H: the mean relative abundance averaged over aquaculture acreages is equal
across all years
A: the mean relative abundance averaged over aquaculture acreages is not equal
across as years
The test for the main effect of acreage is:
H: the mean relative abundance averaged over years is the same for all acreage
values.
A: the mean relative abundance averaged over years differs between the acreage
values.
Across our developed model set assessing for the implications of shellfish aquaculture on
sea duck population dynamics, support for our null hypotheses would suggest that factors
other than those addressed in our valuation might determine the variability of observed
sea duck population abundance.


 

46
 

Figure 1. Simplified flow chart depicting method of operations to collate, merge, summarize and export data for statistical analysis.

 

47
 

Figure 2. Map depicting study site polygons along South Puget Sound shorelines, generated in a GIS. 192
total sites.189 utilized in analysis. Boundaries follow approx. 2 km of shoreline and extend out 500 meters.

 

48
 

RESULTS
Shellfish Aquaculture
Figure 3. illustrates that total area under cultivation increased steadily since 1994,
increasing at a rate of 128 acres annually (Table 2). Further, number of study sites under
cultivation increased steadily at a rate of nearly 3 sites per year (Table 2), growing from
37 to 80 total occupied sites (Figure 3). Subsequent comparison of temporal trends in our
acreage size classes reveal that medium farms (5 < x ≤ 25), continued to define that
greatest number of our study sites, increasing at a rate of 1.64 sites per year to 34 sites by
2012 (Table 2., Figure 3). Close behind medium-classed study sites, those defined by
large operations (x > 25), increased by 1.16 sites annually (Table 2), growing from only 5
sites to currently represent 24 in 2012 (Figure 3). Alternatively, small acreages classes (0
< x ≤ 5) once defining the greatest number of sites with aquaculture, demonstrated a
relatively stable trend at an annual growth rate of 0.15 to a current count below historic
numbers (Table 2, Figure 3). Further comparisons of species cultivated identify clams as
the dominant species cultured over the 19-year study period (Figure 4). Oysters
represented the next largest proportion, followed by geoduck and mussels, respectively
(Figure 4).
Bufflehead
Testing first for the interaction effects, we reject the stated null hypothesis and
conclude that there is a significant interaction between the effects of aquaculture acreage
class*year on the relative abundance of Bufflehead, F(54,3283)=1.59, p<0.0043 (Table
3). Subsequent analyses demonstrated that there were simple effects for year at the zero
[F(18, 3226)=5.40, p<0.0001], small [F(18,3317)=2.04, p=0.0058] and large


 

49
 

[F(18,3245)=2.75, p<0.0001] levels of the aquaculture acreage size class factor. As
Illustrated in Figure 5., across each acreage size class, Bufflehead showed steady
increases in relative abundance from 1994 to 2012. Initial divergence of abundance
values among acreage size classes spiked in 1997-1998, with subsequent abundances
across size classes exhibiting varying degrees of difference in means (Figure 5). Overall,
Figure 5. shows that Bufflehead demonstrated higher marginal mean relative abundances
at study sites classified by large aquaculture acreage than zero classified sites.
Scoter Species Group
For Scoter species, effects of both year and acreage class individually resulted in
significant relationship with Scoter spp. relative abundances. From this, we can first
conclude that relative abundance averaged over aquaculture acreage sizes is not equal
across all years, F(18,3285)=8.25, p<0.0001 (Table 4). Further analysis show a gradual
decline of sample means over time, with significantly lower abundances in later years
(2008-2012) compared to initial year (1994-1998) (Figure 6). Secondly, analysis rejected
the null hypothesis of acreage and suggests that mean relative abundance averaged over
years differs between acreage size classes F(3,314)=4.3, p=0.0095 (Table 4). Tukey HSD
comparisons shows that mean relative abundances at medium acreage class sites were
significantly greater than at zero class sites (Table 5, Figure 7). However comparisons of
small and large acreage class sites suggest no significant difference in Scoter species
group relative abundance (Table 5, Figure 7). Analysis of the interaction effect of acreage
class*year detected no significance, F(54,3289)=1.18, p=0.2599 (Table 5). The relative
conformity of temporal trend lines among varying acreage size classes support evidence
of non-significance of interaction (Figure 14, Appendix A).


 

50
 

Goldeneye Species Group
Testing for the effect of aquaculture size class in determining Goldeneye relative
abundance values indicate significant differences among levels of acreage,
F(3,334)=10.51, p<0.001(Table 6). Supplemental Tukey comparisons of LSMeans
reveal that relative abundances at zero acreage sites are significantly greater than those at
small, medium and large acreage sites at 0.3050 compared to 0.2396, 0.1870, and 0.1868,
respectively (Table 6, Figure 7). However, comparisons between varying levels of
aquaculture acreage revealed no significant difference in determining abundance values
(Table 6, Figure 7). Testing for the effect of time revealed no significant influence on
Goldeneye relative abundances, F(18, 3302)=1.34, p=0.1521. Further, testing for the
interaction effect of acreage*year provided no significant results, F(54, 3306)=0.93,
p=0.6210 (Table 6).
Merganser Species Group
First examining for our interaction effect, acreage*year, analysis failed to reject
the null hypothesis that there is no interaction between the effects of aquaculture acreage
and year on the relative abundance of Merganser Spp, F(54,3327)=0.82, p=0.8214 (Table
8). However, individual main effects of year and acreage both exhibited a significant
interaction with Merganser population abundances. Analysis determines that the mean
relative abundance averaged over aquaculture acreage size classes is not equal across all
years, F(18,3322)=1.66, p=0.0398 (Table 8). Figure 8. demonstrates that while relatively
stable, Merganser relative abundances are showing a slight increase over time. Testing
for the effect of acreage concludes that mean relative abundance averaged over years
differed significantly between acreage size classes, F(3, 359)=3.32, p=0.0201 (Table 8).


 

51
 

Additional Tukey HSD of sample mean differences show that relative abundance at sites
classified by zero aquaculture acreage, 0.086, are significantly different than abundance
values at large acreage classes, at 0.033 (Table 9, Figure 8).


 

52
 

Figure 3. Growth in shellfish aquaculture in South Puget Sound by total acres under
cultivation (black line)(p<0.001, R2=0.96) and number of study sites under cultivation
(shaded bars) by acreage size class - small (0>x≥5), medium (5>x≥25) and large (x>25).

Figure 4. Growth in total acres cultivated across South Puget Sound aquaculture study
sites delineated by shellfish species cultured – mussel, geoduck, oyster and clam.
Table 2. Rate of change in acres and count of study sites under cultivation in South Puget
Sound from 1994 to 2012 by aquaculture size class – zero, (0) small (0>x≥5), medium
(5>x≥25) and large (x>25).
Size Class
Large
Medium
Small
Zero
Total

 

Rate of Change
Acres
Count
105.51
1.16
21.56
1.64
0.71
0.15
-2.92
127.79
2.95
53
 

Bufflehead
LS Means


 

 

 

 

 

 

1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0

LARGE
MEDIUM
SMALL
ZERO


 
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

 
Year

 
Figure 5. Sample means plot depicting significant interaction effect of aquaculture by year on Bufflehead relative abundance of winter
populations in South Puget Sound. Levels of aquaculture acreage defined as large (open red circle), medium (green plus), small (open
blue diamond) and zero (brown x) aquaculture acreage size classes. F(54,3283)=1.59, p=0.0043.
Table 3. Results of test slices report for significant interaction effect of acreage x year on Bufflehead relative abundances. Shows the
effect of time on each acreage size class
Prob >
NumDF DenDF F Ratio F
Large
18
3245
2.75 <.0001
Medium
18
3295
1.46 0.0958
Small
18
3317
2.04 0.0058
Zero
18
3226
5.4 <.0001


 

54
 

Table 4. Standard least squares REML results of fixed effects analyzing Scoter species (BLSC, SUSC, WWSC, UNSC) winter
relative abundances in South Puget Sound from 1994 to 2012.
Source
Nparm DF DFDen
F Ratio Prob > F
Acreage Class
3
3 314.8235123
3.394
0.0183
Year
18 18 3285.446328 8.2477
<.0001
Acreage Class*Year
54 54 3289.479811 1.1172
0.2599


 

 

 

 

 

 

 

 

 

Scoter LS Means

Table 5. Tukey HSD crosstab report for interaction effect of aquaculture acreage size on Scoter species group (BLSC, SUSC,
WWSC, UNSC) winter relative abundance in South Puget Sound averaged over 1994-2012. Levels not connected by the same
letter are significantly difference (p<0.05).
Level
Least Sq Mean Std Error
MEDIUM
A
0.43680579 0.02725981
 
SMALL
A B
0.41841467 0.02493732
 
LARGE
A B
0.39141436 0.03743206
 
ZERO
B
0.35492628 0.01362282
 

 
1
0.8
0.6
0.4
0.2
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 6. Sample means of Scoter species (SUSC, BLSC, WWSC, UNSC) winter relative abundance in South Puget Sound from
1994-2012. Standard error bars included. Results from standard least squares REML, F(18, 3285), p<0.0001.


 

55
 

Table 6. Standard least squares REML results of fixed effects analyzing Goldeneye
species (COGO, BAGO, UNGO) winter relative abundances in South Puget Sound
from 1994 to 2012.
Source
Nparm DF DFDen
F Ratio
Prob > F
Acreage Class
3
3 334.8209456 10.5101
<.0001
Year
18 18
3301.93743
1.3399
0.1521
Acreage Class*Year
54 54 3306.296044
0.9297
0.621
Table 7. Tukey HSD crosstab report for interaction effect of aquaculture acreage size
on Goldeneye species group (COGO, BOGO, UNGO) winter relative abundance in
South Puget Sound averaged over 1994-2012. Levels not connected by the same letter
are significantly difference (p<0.05).
Level
ZERO
SMALL
MEDIUM
LARGE

A
B
B
B

Least Sq Mean Std Error
0.30502524 0.01130776
0.23962439 0.02097505
0.18704016 0.02302209
0.18682118 0.03155346


 

 
Figure
 7.
 A
 comparison
 of
 Tukey
 HSD
 results
 by
 species
 group.
 Relative
 abundances
 
by
 acreage
 size
 class
 averaged
 over
 time.
 In
 each
 species
 group,
 results
 not
 
connected
 by
 the
 same
 letter
 are
 significantly
 different
 (p<0.05).


 

56
 

Table 8. Standard least squares REML results of fixed effects analyzing Merganser species (COME, HOME, RBME, UNME)
winter relative abundances in South Puget Sound from 1994 to 2012.
 
Source
Acreage Class
Year
Acreage Class*Year

Nparm DF
DFDen
F Ratio Prob > F
3
3
359 3.3173
0.0201
18
18
3322 1.6572
0.0398
54
54
3327 0.8211
0.8214

Table 9. Tukey HSD crosstab report for interaction effect of aquaculture acreage size on Merganser species group (COME,
HOME, RBME, UNME) winter relative abundance in South Puget Sound averaged over 1994-2012. Levels not connected by the
same letter are significantly difference (p<0.05).
Level
Least Sq Mean Std Error
ZERO
A
0.08573273
0.00654795
MEDIUM
A B
0.06654402
0.01377676
SMALL
A B
0.05824106
0.01246492
LARGE
B
0.03256016
0.01880374

Merganser
LS Means

1
0.8
0.6
0.4
0.2
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 8. Sample means of Merganser species (COME, HOME, RBME, UNME) winter relative abundance in South Puget Sound
from 1994-2012. Standard error bars included. Results from standard least squares REML, F(18,3322)=1.66, p=0.0398.


 

57
 

Discussion
Sea Duck – Aquaculture Relations
This study is one of the first to examine sea duck population dynamics in response
to changing nearshore aquaculture landscapes across two decades in South Puget Sound.
While many studies of animal-habitat relations incorporate measures of either time or
space (Biossonette and Storch 2007), our mixed model approach to analyzing repeated
measures data permitted simultaneous inference of both spatial and temporal factors. In
studying the interface of condition and time-based effects we expand our frame of
inference to better address how winter sea duck populations are responding to changing
aquaculture landscapes. This research directly addressed recommendations to include
broader spatial and temporal scales of analysis to monitor and evaluate possible factors
contributing to the decline of Pacific coast sea duck populations (SDJV 2012, WDFW
2010). Furthermore, we evaluate the ecological role of shellfish aquaculture on sea duck
populations and on coastal marine habitats by explicitly incorporating the influence of
time, space and intensity of aquaculture (Simenstad and Fresh 1995). Our study
concluded significant responses to aquaculture by all four species groups – Bufflehead,
Goldeneye, Scoter and Merganser; however, in concordance with past studies of sea
duck-aquaculture relations, our results depict a range of direction and degree of responses
among species. Our results show varied responses in regards to the interaction effect of
acreage by time, and individual effects of acreage and year.
Contrary to a negative disturbance definition of aquaculture, both Bufflehead
species and the Scoter species group (BLSC, SUSC, WWSC, UNSC) exhibited a positive
association with shellfish cultivation operations in the South Puget Sound. However,


 

58
 

variation in degree and significance of response to different model parameters was
observed between species groups. In testing for our interaction of acreage and year, only
the Bufflehead demonstrated a statistically significant relationship of relative abundance
responses to a changing aquaculture environment. Results specified greater abundances
values within study sites under large levels of cultivation as opposed to those under small
or no levels of cultivation. While Scoter species failed to exhibit a significant response to
interaction effects, averaging mean relative abundance over time showed Scoter species
abundance values were clearly greater at medium level compared to zero level
aquaculture sites. Several indications, which could explain these trends, have been
identified in past studies. Because food quality and quantity strongly influence habitat use
in birds (Palm et al. 2012), observed population dynamics may reflect spatial and
temporal patterns in food resources. One commonly identified driver of positive
associations between sea ducks and aquaculture can be attributed to the addition of food
resources provided by aquaculture operations. Bufflehead and Scoter species are both
characterized as omnivorous diving ducks, predating largely on bottom-dwelling marine
invertebrates in coastal bays and inlets, consisting predominantly of bivalve mollusks and
crustaceans (SDJV 2003b,c,j,k). In a study by Kirk et al. (2007) comparing mussels on
natural and aquaculture-structured habitats found that mussel density and morphology
differed dramatically between artificial structures and intertidal habitats. Specifically, that
mussel densities were considerably higher within aquaculture facilities by providing new
substrate on which to attach. Further, that mussels grown on structures tended to be
larger, thinner shelled and attached more weakly than comparative intertidal specimens


 

59
 

(Kirk et al. 2007), thus providing a more profitable resource to local sea duck predators
(Zydelis et al. 2009).
Foraging theory suggests that animals respond to changes in food abundance and
quality, encouraging species to optimize net energy intake when faced with variation in
prey attributes or abundance (Kirk et al. 2008, Lewis et al. 2008). Therefore, changes in
sea duck distributions and abundances can reflect underlying prey resource availability.
This may be particularly true for Bufflehead and Scoter species as dominantly
molluscivorous sea ducks feeding largely on wild mussels, and further documented to
severely deplete structure-grown mussels in the presence of off-bottom aquaculture (Kirk
et al. 2007). It is of interest to mention that some shellfish farmers actually welcome
predatory sea ducks - by which the predation of mussels fouling aquaculture structures is
alleviated without the costly efforts of doing so manually (Kirk et al. 2008). The
possibility of a mutually sustainable or even positive wildlife-industry interaction is all
too rare (Zydelis et al. 2009), and consequently should inspire further collaborative
research. Although we did not directly observe mechanisms driving population dynamics,
these positive associations support use theories that sea duck species may exploit novel
and advantageous prey populations provided by aquaculture operations.
Although these two species groups documented similar responses to our model
effect of aquaculture, their observed overall population trends diverge considerably.
Bufflehead species were one of two groups (the other, Merganser) to exhibit overall
increases in relative abundance values over time, regardless of acreage class
specifications. Past studies of WDFW PSAMP marine bird data have also found that
Bufflehead, representing the second most numerous diving duck in Puget Sound,


 

60
 

demonstrated more stable populations patterns among wintering sea duck species
(Nysewander et al. 2005). This may suggest the overall stability of Bufflehead
populations to adjust to changing nearshore landscapes and advancing aquaculture
operations. In contrast, Scoter species, although comprising the greatest proportion of sea
duck populations also represent one of the most extreme declines, at an estimated Soundwide decline of 53% (WDFW 2010). This significant declining trend was similarly
identified in my analysis (Figure 6, 10). However, lack of negative relations with
shellfish aquaculture, may suggest that alternative sources may be at play leading to longterm declines, such as: natural environmental attributes (Zydelis et al. 2006), availability
of herring spawn (Anderson et al. 2009), cumulative levels of nearshore urbanization
(Rice 2007), or predation of scoters (Anderson et al. 2012),
Our Goldeneye species group (BAGO, COGO, UNGO) exhibited a unique
response to aquaculture compared to other species groups of interest. Despite the relative
temporal stability of species group abundances, with no significant influence recognized
by the effect of year, Goldeneye showed a significant negative response to any level of
aquaculture. Figure 16. illustrates the relatively constant temporal trend of greater
population abundance in zero-classed study sites compared to small, medium and large.
Results suggest that presence of shellfish aquaculture may be displacing wintering
Goldeneye species populations throughout South Puget Sound. However, the lack of
significance between small medium and large operations could indicate that greater
intensity of aquaculture do not invoke significantly increased responses by Goldeneye.
Comparable negative associations to shellfish aquaculture were reflected in analysis of
the Merganser species group (COME, HOME, RBME, UNME). Although Mergansers


 

61
 

comprised the smallest proportion of sea duck abundances over all, this species group
demonstrated a significant increase in abundance values over time. This increasing trend
is representative of Sound-wide trends in Merganser populations, as one of few sea duck
species with some degree of increase (Nysewander et al 2005). Added analysis of
abundance values associated with varying aquaculture acreage levels indicate that
populations occurred in significantly lower numbers in large-class sites compared to
those of zero-class delineations. This illustrates that even under increasing population
trends, species are adversely responding to extensive shellfish aquaculture operations.
Similar deleterious impacts of shellfish aquaculture have been identified in past
studies to both directly and indirectly drive observed marine bird habitat use and patterns.
Similar to mechanisms lending to utilization of choice resources, observed absence in
certain areas may suggest lack of availability to vital habitat. In regards to sea ducks, the
most recognized source of negative implications of aquaculture is the degradation or
alteration of critical foraging habitat. This is most likely a response to change of prey
landscapes due to nearshore fauna alterations (Caldow et al. 2003), declines to seagrass
communities (Tallis et al. 2009) and sediment modifications (Connolly and Colwell
2005). One study by Caldow et al. (2003), found that the effects of intertidal mussel
cultivation were associated with decreases in over-wintering nearshore bird overall
abundance and species richness. A similar study on the effects of oyster cultivation in
three major estuaries in Washington found that aquaculture harvest practices had
identifiable and distinct impacts on eelgrass density and growth, with lower densities
observed in all oyster culture areas (Tallis et al. 2009). Negative disturbance of seagrass
communities could have further implications to dependent infaunal benthic and


 

62
 

epibenthic prey species on which predatory sea ducks rely (Dumbauld et al. 2009). These
influences to seagrass communities could be especially suggestive of our observed
negative response in Merganser species, as the only piscivorous diving duck of focus
feeding largely on small fish, crustaceans and aquatic insects (SDJV 2003e,g.i), which
rely on functional seagrass habitats.
Despite varying results documented to nearshore ecosystems under shellfish
aquaculture activity, the overall conclusion of shift in community structure and
ecosystem processes demonstrate varying directions and degrees of disturbance
(Simenstad and Fresh 1995). The concern then becomes, how to quantify changes that
provide for some species while displacing other. Caldow et al. (2003) suggests then, that
evaluations of aquaculture may require assessments on a case-by-case basis. This issue is
evident in our study, where different species groups are documented relating to
aquaculture in varying ways, coinciding with past results suggesting that responses to
aquaculture may be species-specific (Connolly and Colwell 2005). Further, one might ask
how different sea duck species themselves relate, whereby past studies have suggested
the sheer mass of feeding scoter populations in certain coastal estuaries, as a top predator,
may influence community dynamics of competing predators (Lewis et al. 2007).
Future Considerations
The dynamics of evaluating wildlife-landscape relations coupled with the inherent
difficulty involved in marine bird population studies highlight the complexity of
quantifying sea duck responses to shellfish aquaculture. As a result, many additional
questions have developed out of this study, further identifying several key areas of
expansion for future studies. Our efforts focused only on identifying sea duck relations to


 

63
 

one key anthropogenic habitat feature - shellfish aquaculture. By expanding future studies
to incorporate for additional natural habitat features, this may provide identification of
alternative sources of variation in sea duck population observations. We recommend
using two key natural habitat features in future analysis: nearshore substrate type and
intertidal width. In a study addressing movements of foraging winter scoter populations,
Kirk et al. (2008) found that populations demonstrated significantly different feeding
behaviors between soft-bottom intertidal flats and rocky intertidal shores. Supported in
additional studies as a response to differing prey landscapes, such as species composition
(Kirk et al. 2008) and density (Lewis et al. 2008). Furthermore, past studies by Zydelis et
al. (2006, 2009) have identified intertidal width as an important predictor in scoter
population distributions, even under varying substrate type. This is not surprising, as the
intertidal zone constitutes the majority of habitat utilized by sea ducks.
Second, it would be valuable to further explore varying levels of spatial and
temporal scales of inference. For example, supplementary integration of efforts directed
at specific bays and inlets of South Puget Sound with documented heavy uses by both sea
duck species and aquaculture industry. Because temporally repeating focused
observations within nearly 200 study sites is unreasonable, these additional direct
inferences on specific interactions of sea ducks and aquaculture could uncover underlying
mechanisms of observed broad-scale relations. Likewise, because many sea ducks that
utilize South Puget Sound during non-breeding periods do so for a large portion of their
annual cycle (Gaydos and Pearson 2011) further addressing seasonal variations may help
clarify site-specific directions and degrees of disturbance. This may be especially true for
sea ducks, such as scoters, where within-season variation of foraging behaviors have been


 

64
 

observed (Palm et al. 2012). Specific to aquaculture, while artificial farm structures may
provide novel foraging habitat, the pulse of prey that draw opportunistic sea duck species,
may lack the stability to provide a lasting resource and this variability may have further
ecological implications to wintering sea ducks (Kirk et al. 2008).
Finally, future studies of this nature would benefit from additional information
regarding specific aquaculture method and activity. In this analysis, by characterizing
each of the study sites by total acres cultivated, inferences could be drawn on the
intensity of operation - whereby larger operations using greater portions of tidal area
would require greater degree of cultivation effort, thus industry activity. However,
bivalve species are not all cultivated in the same manner, and still within each species,
method of cultivation varies by tidal and industry resources. Therefore, it would be
beneficial to address specific responses of sea duck populations to differing cultivation
operations. Effect of aquaculture method has been employed in past studies,
demonstrating greater degree of negative implications due to on-bottom methods (eg.
Dumbauld et al. 2009) versus neutral or beneficial implications of off-bottom methods
(e.g. Zydelis et al. 2009). Method-specific evaluations in South Puget Sound would allow
greater ability to tease out certain deleterious implications associated with certain culture
operations, while also identifying those where mutual sea duck – industry inhabitance is
occurring.


 

65
 

Conclusions
Long-term population declines in many sea duck species wintering in Puget
Sound have sparked concerns of local wildlife conservation and management agencies.
Given the limited understanding of sea duck basic biology and the inherent challenges of
studying species with such broad and remote distributions, current knowledge is lacking
in its ability to identify sources of decline. In Puget Sound in particular, an understanding
of habitat requirements, and identification and evaluation of possible limiting factors,
remains insufficient to best understand and mitigate these declines. From this need, our
research contributes to this limited body of knowledge by providing an assessment of sea
duck habitat use and availability in response to a variable aquaculture landscape.
This study is the first of its scale to address the influence of shellfish aquaculture
on winter sea duck populations. As my study supports, by explicitly incorporating for
both spatial and temporal variability in assessments of sea duck-aquaculture relations,
efforts may provide a better understanding of the direction and degree of suggested
associations. However, this is only one aspect of defining sea duck interactions with
aquaculture, where further efforts could benefit from integrating scales of inference
across different life stages, to better understand underlying mechanisms.
Findings suggest that the location and extent of shellfish aquaculture plays a
significant role in defining winter sea duck population distribution and abundance in
South Puget Sound. However, that sea duck responses to aquaculture differ in the nature
and degree according to species or species group. Whereby Bufflehead and Scoter
species (Black Scoter, Surf Scoter, White-winged Scoter) exhibited varying levels of
positive associations with aquaculture; however, Goldeneye species (Common


 

66
 

Goldeneye, Barrow’s Goldeneye) and Merganser species (Common Merganser, Hooded
Merganser, Red-breasted Merganser) demonstrated differing degrees of negative
responses to aquaculture. This study highlights the complexity of analyzing sea duck
populations in an anthropogenically charged landscape - that aquaculture induced
disturbances are divergent at times, that sea duck habitat use patterns are dynamic, and
that consistent future monitoring and assessment efforts are needed to clearly evaluate
changing ecosystems.


 

67
 

CHAPTER THREE
GENERAL CONCLUSIONS
Observed long-term declines in many Puget Sound winter sea duck populations
have driven the scientific community to address the significant lack of knowledge
defining basic sea duck biology and ecology. Conservation and management agencies
have recognized that without sustained and expanded effort to identify sea duck habitat
use, needs and availability of Puget Sound winter grounds, effective measures to mitigate
or reverse observed declines remains unlikely. By investigating the implications of
shellfish aquaculture on winter sea duck populations in South Puget Sound, our findings
suggest:
• Aquaculture location and extent plays a significant role in determining sea duck
distribution and abundance.
• Sea duck responses to aquaculture vary in nature and degree by species or species
group
o Bufflehead respond positively over time to a changing aquaculture
landscape, with greatest abundances at study sites with greater than 25
acres of cultivation.
o Scoter species (BLSC, SUSC, WWSC) respond positively overall to study
sites subject to between 5 and 25 acres of cultivation compared to those
with no cultivation. Though, species trends also show overall declines in
relative abundances over time within our study area.
o Goldeneye species (BAGO, COGO) remained stable in their overall
abundance trends; however, these species were negatively associated with


 

68
 

aquaculture. Relative abundances were significantly greater at study sites
under zero cultivation activity compared to any level of cultivation.
o Merganser species group (COME, HOME, RBME), while showing
significant increases in abundance over time, demonstrated negative
association with aquaculture. Relative abundances were clearly greater at
sites under zero cultivation compared to those with greater than 25 acres
of cultivation.
The presence of sea duck-aquaculture associations may be clear, however the
variability in direction and degree of relations suggest that industry in South Puget
Sound, while proving deleterious for some species, may provide for others. These
findings highlight the dynamic nature of sea duck-aquaculture relationships and the need
for continued investigations to inform industry practices and resource management
agency options to minimize the impacts of aquaculture on nearshore ecosystems and sea
duck populations. To provide a comprehensive understanding of this animal-landscape
relationship, monitoring and assessment efforts should strive to:
•

Maintain broad spatial and temporal scales of inference to account for the
heterogeneity of habitat and resources

•

Further integrate local scale research to link underlying mechanisms
defining regional populations trends
o Thereby better understanding species-specific responses

•

Connect efforts across different stages throughout their annual cycles –
among molting, breeding and wintering habitats


 

69
 

Our determinations to collate a database of activities in South Puget Sound have
identified the abundance of information available in characterizing aquaculture in the
State. However, the exhaustive efforts required to provide a representative and analytical
resource should prompt regulatory agencies to readdress their permitting requirements,
documentation methods and collaborative capacities. To succeed at understanding sea
duck-aquaculture relationships, it is important those agencies develop and maintain a
complete representation of shellfish aquaculture operations and activities in the state.
Promoting this will require:
•

Redevelopment of permitting requirements, regulation and documentation
to improve analytical capacities
o To include detailed spatial location
o To include specific harvest methods

•

Communication and collaboration among regulatory agencies (WDFW,
DOH, DNR) to make productive and efficient use of shellfish aquaculture
information
o Integrated application process and central database

•

Communication and collaboration within agencies (WDFW licensing,
records, wildlife) to effectively allocate resources and advance research
capacities

•

Communication and collaboration with independent research
organizations (SDJV, PCSGA, PSI) and shellfish industry to build
productive and cooperative relationships to address sea duck-industry
conflict
 


 

70
 

This study was the first in Puget Sound to build a comprehensive database of
shellfish aquaculture at this temporal and spatial capacity, and integrate this into winter
sea duck habitat assessments. Our efforts have provided novel findings elucidating sea
duck-aquaculture relationships in Puget Sound, a foundation for growth and management
of a functional aquaculture database, and suggestions to advance effective agency and
industry cooperative research efforts. To effectively address complex issues pertaining to
both declining sea duck populations and advancing aquaculture industry, it is necessary
to develop dynamic methods to meet the challenges of a dynamic system. Ultimately, the
objective of conservation, management and industry is to meet the pressures of balancing
the functional requirements of sea duck habitats with the economic values of a shellfish
aquaculture in Puget Sound.


 

71
 

REFERENCES
Anderson, E.M, Lovvorn, J.R., & Wilson, M.T. (2008). Reevaluating marine diets of surf
and white-winged scoters: interspecific differences and the importance of soft-bodied
prey. The Condor, 110(2), 285-295.
Anderson, E.M., Bower, J.L., Nysewander, D.R., Evenson, J.R., & Lavvorn, J.R. (2009)
Changes in avifaunal abundance in a heavily used wintering and migration site in
Puget Sound, Washington during 1966-2007. Marine Ornithology, 37, 19-27.
Anderson, E.M., Lovvorn, J.R., Esler, D., Boyd, W.S., & Stick, K.C. (2009). Using
predator distributions, diet and condition to evaluate seasonal foraging sites: sea
ducks and herring spawn. Marine Ecology Progress Series, 386, 287-302.
Anderson, E.M., & Lovvorn, J.R. (2011). Contrasts in energy status and marine foraging
strategies of white-winged scoters (Melanitta fusca) and Surf Scoters (M.
perspicillata). The Auk. 128(2) 248-257.
Anderson, E.M., Esler, D., Boyd, W.S., Evenson, J.R., Nysewander, D.R., Ward, D.H.,
… & Hupp, J.W. (2012) Predation rats, timing, and predator composition for Scoters
(Melanitta spp.) in marine habitats. Canadian Journal of Zoology, 90, 42-50.
Bendell-Young, L.I. (2006) Contrasting the community structure and select geochemical
characteristics of three intertidal regions in relation to shellfish farming.
Environmental Conservation.
Bissonette, J.A. & Storch I. (Eds.) (2007) Temporal Dimensions of Landscape Ecology:
wildlife responses to variable landscapes. New York, NY: Spring Science+Business
Media, LLC
Bower, J.L. (2009). Changes in marine bird abundance in the Salish Sea: 1975-2007.
Marine Ornithology, 37, 9-17.
Brook, R.W., Abraham, K.F., Middel, K.R., & Ross, R.K. (2012) Abundance and habitat
selection of breeding scoters (Melanitta spp.) in Ontarios Hudson Bay Lowlands.
Canadian Field-Naturalist, 126(1), 20-27.
Buchanan, J.B. (2006). Nearshore Birds in Puget Sound. Puget Sound Nearshore
Partnership Report No. 2006-05. Seattle, WA: Seattle District, U.S. Army Corps of
Engineers.
Butler, W.I., Stehn, R.A., & Balogh, G.R. (1995). GIS for mapping waterfowl density
and distribution from aerial surveys. Wildlife Society Bulletin, 23, 140-147


 

viii
 

Caldow, R.W.G, Beadman, H.A., McGororty, S., Kaiser, M.J., Goss-Custard, J.D.,
Mould, K., & Wilson, A. (2003). Effects of intertidal mussel cultivation on bird
assemblages. Marine Ecology Progress Series, 259, 173-183.
Chadwick, M. (Eds) (2006). Proceedings of CSAS ’06: Assessing habitat risks associated
with bivalve aquaculture in the marine environment. Moncton, N.B: DFO.
Chamberlain, J., Weise, A.M., Dowd, M., & Grant, J. (2006). Modeling approaches to
assess the potential effects of shellfish aquaculture on the marine environment.
(Research document 2006/032) Retrieved from http://www.dfo-mpo.gc.ca/csas/,
Fisheries and Oceans Canada.
Connolly, L.M. & Colwell, M.A (2005). Comparative use of longline oysterbeds and
adjacent tidal flats by waterbirds. Bird Conservation International, 12, 237-255.
De la Cruz, S.E.W., Takekawa, J.Y., Wilson, M.T., Nysewander, D.R., Evenson, J.R.,
Esler, D., Boyd, W.S., & Ward, D.H. (2009). Spring migration routes and chronology
of surf scoters (Melanitta perspicillata): a synthesis of Pacific coast studies.
Canadian Journal of Zoology, 87, 1069-1086.
Dewey, W., Bunsick, S., Moyer, J., & Plauche, S. (2007). Proceedings of Coastal Zone
07: National Trends in Shellfish Aquaculture Production, Policy and Regulation –
Planning and Regulation Tools for Coastal Planner, Portland, OR.
Dumbauld, B.R., Ruesink, J.L., & Rumrill, S.S. (2009) The ecological role of bivalve
shellfish aquaculture in the estuarine environment: a review with application to oyster
and clam culture in West Coast (USA) estuaries. Aquaculture, 290, 196-223.
Essington, T., Klinger, T., Conway-Cranos, T., Buchanan, J., James, A., Kershner, J.,
Logan, I., & West, J. (2011). Marine Birds. In The Biophysical Condition of Puget
Sound: Biology (Section 2.9) Retrieved from http://eopugetsound.org/science-review.
Evenson, J. (2012) Pacific Coast Joint Venture Discretionary Funds 2012 Washington
Application. Lakewood, WA: WDFW
Gaydos, J.K. & Pearson, S.F. (2011). Birds and Mammals that depend on the salish sea: a
compilation. Northwestern Naturalist, 92, 79-94.
Gustafson, R.G., Lenarz, W.H., McCain, B.B., Schmitt, C.C., Grant, W.S., Builder, T.L.,
and Methot, R.D. (2000). Status review of Pacific Hake, Pacific Cod, and Walleye
Pollock from Puget Sound, Washington. NOAA Tech. Memo NMFS-NWFSC-44
Esler, D., Bowman, T.D., O’Clair, C.E., Dean, T.A., & McDonald, L.L. (2000). Densities
of Barrow’s Goldeneyes during winter in Prince William Sound, Alaska in relation to
habitat, food and history of oil contamination. Waterbirds, 23(3) 423-429


 

ix
 

FAO (2006). Fishery and Aquaculture Statistics: FAO Yearbook. Rome: Food and
Agriculture Organization of the United Nations.
FAO (2012, January). The State of World Fisheries and Aquaculture (SOFIA): Topics
Fact Sheets. Text by Jean- Francois Pulvenis. Retrieved from
http://www.fao.org/fishery/topic/13535/en
Forrest, B.M., Keely, N.B., Hopkins, G.A., Webb, S.C., Clement, D.M., (2009). Bivalve
aquaculture in estuaries: review and synthesis of oyster cultivation effects.
Aquaculture, 298, 1-15.
Fresh, K.L., Dethier, M.N., Simenstad, C.A., Logsdon, M., Shipman, H., Tanner, C.D.,
… Newton, J.A. (2011). Implications of observed anthropogenic changes in the
nearshore ecosystem in Puget Sound. Puget Sound Nearshore Restoration Project
Report No. 2011-03. Seattle, WA.
Gaydos, J.K, & Pearson, S.F. (2011). Birds and mammals that depend on the Salish Sea:
A compilation. Northwestern Naturalist, 92, 79-94.
Gutzwiller, K.J.& Riffell, S.K. (2007). Using statistical models to study temporal
dynamics of animal-landscape relations. In Bissonette, J.A. & Storch, I. (Eds),
Temporal dimensions in landscape ecology (93-118). New York, NY: Spring
Science+Business Media LLC.
Harley, C.D.G., Hughes, A.R., Hultgren, K.M., Miner, B.G., Sorte, C.J.B., Thornber,
C.S., … Williams, S.L. (2006) The impacts of climate change in coastal marine
systems.
Helmuth, B., Kingsolver, J.G., & Carrington, E. (2005). Biophysics, physiological
ecology, and climate change: does mechanism matter? Annual Review of Physiology,
67, 177-201
Kirk, M., Esler, D., & Boyd, W.S. (2007) Morphology and density of mussels on natural
and aquaculture structure habitats: implications for sea duck predators. Marine
Ecology Progress Series, 346, 179-187.
Kirk, M., Esler, D., & Iverson, S.A. (2008). Movements of wintering surf scoters:
predator responses to different prey landscapes. Oecologia, 115, 859-867.
Lewis, T.L., Esler, D., & Boyd, W.S. (2008). Foraging behavior of surf scoters
(Melanitta prespicillata) and white-winged scoters (M. fusca) in relation to clam
density: inferring food availability and habitat quality. The Auk, 125, 149-157.
Lucas, J.S. & Southgate, P.C. (Eds.) (2012). Aquaculture: Farming aquatic animals and
plants. Aimes, IA: Wiley-Blackwell


 

x
 

Magoon, C. & Vining, R. (1988) Introduction to Shellfish Aquaculture in the Puget
Sound Region. Olympia, WA: Washington State Department of Natural Resources.
McLoughlin, P.D., Morris, D.W., Fortin, D., Vander Wal, E., & Contasti, A.L. (2010).
Considering ecological dynamics in resource selection functions. Journal of Animal
Ecology, 79, 4-12.
National Oceanic and Atmospheric Administration (NOAA) (2011). National shellfish
initiative fact sheet. Retrieved from
http://www.nmfs.noaa.gov/stories/2011/12/shellfish.html
Nysewander, D., Evenson, J., Murphie, B., & Cyra, T. (2005). Report of marine bird
mammal component, Puget Sound Ambient Monitoring Program, for July 1992 to
December 1999 period. Olympia, WA: WA Department of Fish and Wildlife
Pacific Coast Shellfish Growers Association (PCSGA) (2011) Washington’s Shellfish
Growers: Who We Are, What We Do. Retrieved from http://pcsga.net/farmingscience/issue-papers/
Pacific Coast Shellfish Growers Association (PCSGA) (2011). Environmental Codes of
Practice for Pacific Coast Shellfish Aquaculture. Retrieved from
http://www.pcsga.net/wpcontent/uploads/2011/07/ECOP_Final_June_2011_Section1.pdf. Olympia, WA.
Palm, E.C., Esler, D., Anderson, E.M., & Wilson, M.T. (2012). Geographic and temporal
variation in diet of wintering white-winged scoters. Waterbirds, 35(4), 577-589.
Price, I.M., & Nickum, J.G. (1995) Aquaculture and birds: the context for controversy.
Colonial Waterbirds, 18, 33-45.
Puget Sound Action Team (PSAT) (2007). 2007 Puget Sound Update: Ninth Report of
the Puget Sound Ambient Monitoring Program. Olympia, WA.
Puget Sound Nearshore Ecosystem Restoration Project (PSNERP) (n.d.) What is the
Puget Sound nearshore and what is the problem with it? Retrieved from
http://www.pugetsoundnearshore.org/what.htm
Puget Sound Parternship (PSP) (2003a, July). A Heritage of Harvest. Retrived from
psparchives.com/publications/
Puget Sound Parternship (PSP) (2003b, July). Keystone Species of the Estuary. Retrived
from psparchives.com/publications/
Puget Sound Partnership (PSP) (2003c, July). Treasures of the Tidelands. Retrieved from
psparchives.com/publications/


 

xi
 

Puget Sound Partnership (PSP) (2012). 2012 state of the sound: a biennual report on the
recovery of Puget Sound. Tacoma, WA.
Rice, C.A. (2007). Evaluating the Biological Condition of Puget Sound. (unpublished
doctoral dissertation). University of Washington. Seattle, WA.
Savard, J.P.L., Reed, A., & Lesage, L. (2007). Chronology of breeding and molt
migration in surf scoters (Melanitta perspicillata). Waterbirds, 30 (2), 223-229.
Sea Duck Joint Venture (2003a). Sea duck information series: barrow’s goldeneye
(Bucephala islandica). Info sheet #1 of 15. Unpublished report. U.S. Fish and
Wildlife Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville,
New Brunswick, Canada.
Sea Duck Joint Venture (2003b). Sea duck information series: black scoter (Melanitta
nigra). Info sheet #2 of 15. Unpublished report. U.S. Fish and Wildlife Service,
Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New Brunswick,
Canada.
Sea Duck Joint Venture (2003c). Sea duck information series: bufflehead (Bucephala
albeola). Info sheet #3 of 15. Unpublished report. U.S. Fish and Wildlife Service,
Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New Brunswick,
Canada.
Sea Duck Joint Venture (2003d). Sea duck information series: common goldeneye
(Bucephala clangula). Info sheet #5 of 15. Unpublished report. U.S. Fish and
Wildlife Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville,
New Brunswick, Canada.
Sea Duck Joint Venture (2003e). Sea duck information series: common merganser
(Mergus merganser). Info sheet #6 of 15. Unpublished report. U.S. Fish and Wildlife
Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New
Brunswick, Canada.
Sea Duck Joint Venture (2003f). Sea duck information series: harlequin duck
(Histrionicus histrionicus). Info sheet #7 of 15. Unpublished report. U.S. Fish and
Wildlife Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville,
New Brunswick, Canada.
Sea Duck Joint Venture (2003g). Sea duck information series: hooded merganser
(Lophodytes cacullatus). Info sheet #8 of 15. Unpublished report. U.S. Fish and
Wildlife Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville,
New Brunswick, Canada.
Sea Duck Joint Venture (2003h). Sea duck information series: long-tailed duck (Clangula
hyemalis). Info sheet #10 of 15. Unpublished report. U.S. Fish and Wildlife Service,


 

xii
 

Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New Brunswick,
Canada.
Sea Duck Joint Venture (2003i). Sea duck information series: red-breasted merganser
(Mergus serrator). Info sheet #11 of 15. Unpublished report. U.S. Fish and Wildlife
Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New
Brunswick, Canada.
Sea Duck Joint Venture (2003j). Sea duck information series: surf scoter (Melanitta
perspicillata). Info sheet #14 of 15. Unpublished report. U.S. Fish and Wildlife
Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New
Brunswick, Canada.
Sea Duck Joint Venture (2003k). Sea duck information series: white-winged scoter
(Melanitta fusca). Info sheet #15 of 15. Unpublished report. U.S. Fish and Wildlife
Service, Anchorage, Alaska, USA and Canadian Wildlife Service, Sackville, New
Brunswick, Canada.
Sea Duck Join Venture (2012). Sea Duck Join Venture implementation plan for April
2012 through March 2015. Report of Sea Duck Join Venture. Available at U.S. Fish
and Wildlife Service, Anchorage, Alaska, and Environment Canada, Sackville, New
Brunswick. Also available at seaduckjm.org. 38 pp.
Sequeira, A., Ferreira, J.G., Hawkins, A.J.S., Nobre, A., Lourenco, P., Zhang, X.L., Yan,
X., & Nickell, T. (2008). Trade-offs between shellfish aquaculture and benthic
biodiversity: A modeling approach for sustainable management. Aquaculture, 274,
313-328.
Simenstad, C.A., & Fresh, K.I. (1995) Influence of intertidal aquaculture on benthic
communities in Pacific Northwest estuaries: scales of disturbance. Estuaries, 18, 4370.
Simms, A. (2002). GIS and aquaculture: assessment of soft-shell clam sites. Journal of
Coastal Conservation, 8(1), 35-48.
Tallis, H.M., Ruesink, J.L., Dumbauld, B., Hacker, S., Wisehart, L.M. (2009). Oysters
and aquaculture practices affect eelgrass density and productivity in a pacific
northwest estuary. Journal of Shellfish Research, 28 (2), 251-261.
Tidewell, J.H. (Ed.) (2012). Aquaculture Production Systems. Aimes, IA: WileyBlackwell
University of Washington Parcels Working Group (UWP) (Accessed March, 2012).
Retrieved from http://depts.washington.edu/wagis/projects/parcels/group/


 

xiii
 

WA Department of Ecology (accessed Feb. 2013). Shoreline Management: Introduction
to the Shoreline Management Act. Accessed at
http://www.ecy.wa.gov/programs/sea/sma/st_guide/intro.html
WA Department of Health (2012). 2011 Annual Report: Commericial and Recreational
Shellfish Areas in Washington State. Olympia, WA: Office of Shellfish and Water
Protection.
WA Department of Fish and Wildlife (2010). WDFW sea duck management strategies:
draft report to the Washington Fish and Wildlife Commission. Olympia, WA:
WDFW.
WA State Blue Ribbon Panel on Ocean Acidification (2012). Ocean Acidification: From
Knowledge to Action, Washington State’s Strategic Response. H. Adelsman, and L.
Whitley Binder, Eds. Olympia, WA: WA Department of Ecology.
WA State Legislation (n.d.) WAC 332-30-106: definitions. Accessed at
http://apps.leg.wa.gov/wac/default.aspx?cite=332-30-106
Wahl, T.R. (1981). Marine bird populations of the Strait of Juan de Fuca, Strait of
Georgia, and adjacent water in 1978 and 1979. NOAA; EPA. WA: Seattle.
Weston, D.P. (1986) The environmental effects of floating mariculture in Puget Sound.
Seattle, WA: UofW School of Oceanography.
Zydelis, R., Esler, D., Boyd, W.S., Lacroix, D.L., & Kirk, M. (2006) Habitat use by
wintering surf and white-winged scoters: effects of environmental attributes and
shellfish aquaculture. Journal of Wildlife Management 70(6), 1754-1762.
Zydelis, R., Esler, D., Kirk, M., & Boyd, W.S. (2009) Effects of off-bottm shellfish
aquaculture on winter habitat use by molluscivorous sea ducks. Aquatic
Conservation: Marine and Freshwater Ecosystems. 19, 34-42.


 

xiv
 

APPENDIX A.
 
 

 

 

 

 

Figure 9. Raw means analysis of average count (light grey) and average relative
abundance (dark grey) of Goldeneye species (BAGO, COGO, UNGO) across
South Puget Sound study sites (n=189) from 1994 to 2012.

Figure 10. Raw means analysis of average count (light grey) and average relative
abundance (dark grey) of Scoter species (BLSC, SUSC, WWSC, UNSC) across
South Puget Sound study sites (n=189) from 1994 to 2012.


 

xv
 

Figure 11. Raw means analysis of average count (light grey) and average relative
abundance (dark grey) of Merganser species (COME, HOME, RBME, UNME)
across South Puget Sound study sites (n=189) from 1994 to 2012.

Figure 12. Raw means analysis of average count (light grey) and average relative
abundance (dark grey) of Bufflehead species across South Puget Sound study
sites (n=189) from 1994 to 2012.


 

xvi
 

Bufflehead
LS Means

1
0.8
0.6
0.4
0.2
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year

Scoter LS Means

Figure 13. Sample means of Bufflehead winter relative abundance in South Puget Sound from 1994-2012. Calculated using standard
least squares REML. F(18, 3280), p<0.0001.
 

1
0.8
0.6
0.4

LARGE
MEDIUM
SMALL
ZERO

0.2
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 14. Sample means of Scoter species (SUSC, BLSC, WWSC, UNSC) winter relative abundance in South Puget Sound by study
sites defined as large (open red circle), medium (green plus), small (blue diamond) and zero (brown x) aquaculture acreage size
classes Non-significant interaction calculated using REML based repeated measured analysis testing for acreage class*year
(p=0.2599).


 

xvii
 

Goldeneye
LS Means

1
0.8
0.6
0.4
0.2
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 15. Sample means of Goldeneye species (COGO, BAGO, UNGO) winter relative abundance in South Puget Sound from 19942012. Calculated using standard least squares REML. (p=0.1521).

Goldeneye
LS Means

1

LARGE
MEDIUM
SMALL
ZERO

0.8
0.6
0.4
0.2
0

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 16. Sample means of Goldeneye species (COGO, BAGO, UNGO) winter relative abundance in South Puget Sound by study
sites defined as large (open red circle), medium (green plus), small (blue diamond) and zero (brown x) aquaculture acreage size
classes Non-significant interaction calculated using REML based repeated measured analysis testing for acreage class*year
(p=0.6210).


 

xviii
 

Merganser
LS Means

1
0.8

LARGE
MEDIUM
SMALL
ZERO

0.6
0.4
0.2
0

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Year
Figure 17. Sample means plot depicting non-significant interaction effect of aquaculture by year on Merganser species group (COME,
HOME, RBME, UNME) relative abundance of winter populations in South Puget Sound. Levels of aquaculture acreage defined as
large (open red circle), medium (green plus), small (blue diamond) and zero (brown x) aquaculture acreage size classes.
F(54,3327)=0.82, p=0.8214.
Table 10. Standard least squares REML results of fixed effects analyzing Bufflhead winter relative abundances in South Puget
Sound from 1994 to 2012.
Source
Nparm DF DFDen F Ratio
Prob > F
Acreage Class
3
3
309
5.1104
0.0018
Year
18
18 3279
4.6693
<.0001
Acreage Class*Year 54
54 3283
1.5853
0.0043
Table 11. Tukey HSD crosstab report for interaction effect of aquaculture acreage size on Bufflehead winter relative abundance in
South Puget Sound averaged over 1994-2012. Levels not connected by the same letter are significantly difference (p<0.05).
Level
LARGE
MEDIUM
SMALL
ZERO


 

A
A

B
B
B

Least Sq Mean
0.39263543
0.30085448
0.28431502
0.25549815

Std Error
0.03425058
0.02492296
0.02284394
0.01255311

xix
 

Table 12. Standard least squares REML variance component estimates measuring Bufflehead winter relative abundance across 189
study sites in South Puget Sound.
Random Effect
Var Ratio Var Component
Std Error
95% Lower
95% Upper Pct of Total
Site[Acreage Class]
0.5028
0.02065
0.00212
0.01650
0.02481
33.46
Residual
0.04107
0.00102
0.03914
0.04316
66.54
Total
0.06173
0.00231
0.05743
0.06654
100
Table 13. Standard least squares REML variance component estimates measuring Scoter species (BLSC, SUSC, WWSC, UNSC)
winter relative abundance across 189 study sites in South Puget Sound.
Random Effect
Var Ratio Var Component
Std Error
95% Lower
95% Upper Pct of Total
Site[Acreage Class]
0.4399
0.02388
0.00249
0.01900
0.02877
30.553
Residual
0.05429
0.00135
0.05173
0.05705
69.447
Total
0.07818
0.00278
0.07300
0.08394
100
Table 14. Standard least squares REML variance component estimates measuring Goldeneye species (COGO, BAGO, UNGO)
winter relative abundance across 189 study sites in South Puget Sound.
Random Effect
Site[Acreage Class]
Residual
Total

Var Ratio
0.33448

Var Component
0.01575
0.04709
0.06284

Std Error
0.00168
0.00117
0.00201

95% Lower
0.01244
0.04487
0.05907

95% Upper
0.01905
0.04947
0.06698

Pct of Total
25.065
74.935
100

Table 15. Standard least squares REML variance component estimates measuring Merganser species (COME, HOME, RBME,
UNME) winter relative abundance across 189 study sites in South Puget Sound.
Random Effect
Site[Acreage Class]
Residual
Total


 

Var Ratio
0.2091

Var Component
0.00478
0.02286
0.02764

Std Error
0.00055
0.00056
0.00077

95% Lower
0.00363
0.02179
0.02619

95% Upper
0.00587
0.02402
0.02923

Pct of Total
17.299
82.701
100

xx
 

Table. 1. Nine species common in Puget Sound, including WDFW species code, feeding characteristics (Rice 2007) and abundance in
and dependence on marine environment (Gaydos and Pearson 2011). O = omnivore; C = carnivore, R = rare, M = medium, H = high.
Scientific
Name
Bucephala
albeola
Bucephala
clangula
Bucephala
islandica
Lophodytes
cucullatus
Melanitta
fusca
Melanitta
nigra
Melanitta
perspicillata
Mergus
merganser
Mergus
serrator


 

Common
Name
Bufflehead

Species
Code
BUFF

Diet
O

Primary
Food
Invertebrate

Common
Goldneye
Barrows
Goldeneye
Hooded
Merganser
Whitewinged scoter
Black Scoter

COGO

O

Invertebrate

BAGO

O

Invertebrate

HOME

C

Fish

WWSC

C

Invertebrate

BLSC

O

Invertebrate

Surf Scoter

SUSC

C

Invertebrate

Common
Merganser
Red-breasted
Merganser
Unidentified
Goldeneye
Unidentified
Merganser
Unidentified
Scoter

COME

C

Fish

RBME

C

Fish

UNDD

O

Invertebrate

UNME

O

Fish

UNSC

O

Invertebrate

Feeding
Behavior
Surface
Dive
Surface
Dive
Surface
Dive
Surface
Dive
Dive
Surface
Dive
Surface
Dive
Surface
Dive
Surface
Dive
Surface
Dive
Surface
Dive
Dive

Marine
derived
food
H

Winter Spring
H
H

Summer
R

Fall
H

Marine
habitat
H

H

H

R

H

H

H

H

H

R

H

H

H

M

M

R

M

M

M

H

H

H

H

H

H

M

M

R

M

M

M

H

H

H

H

H

H

M

M

R

M

M

M

H

H

R

H

M

H

xxi