The Evergreen State College Institutional Repository

Prey Selection and Its Relationship to Habitat and Foraging Strategy of Molting White-winged (Melanitta fusca) and Surf Scoters (M. perspicillata) in Puget Sound, Wa, and the Strait of Georgia, B.C.

Item

Title (dcterms:title)

Eng Prey Selection and Its Relationship to Habitat and Foraging Strategy of Molting White-winged (Melanitta fusca) and Surf Scoters (M. perspicillata) in Puget Sound, Wa, and the Strait of Georgia, B.C.

Date (dcterms:date)

2010

Creator (dcterms:creator)

Eng Tschaekofske, Heather J

Subject (dcterms:subject)

Eng Environmental Studies

extracted text (extracttext:extracted_text)

Prey selection and its relationship to habitat and foraging strategy of molting Whitewinged (Melanitta fusca) and Surf Scoters (M. perspicillata) in Puget Sound, WA,
and the Strait of Georgia, BC.

by
Heather J. Tschaekofske

A Thesis submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
December 2010

i

© 2010 by Heather J. Tschaekofske. All rights reserved.

ii

This Thesis for the Master of Environmental Studies Degree
By
Heather J. Tschaekofske

Has been approved for
The Evergreen State College
By

____________________________
Alison Styring
Member of the Faculty
____________________________
Joseph Evenson
WDFW Wildlife Biologist
____________________________
Gerardo Chin-Leo
Member of the Faculty

____________________________
Date

iii

ABSTRACT
Prey selection and its relationship to habitat and foraging strategy of molting Whitewinged (Melanitta fusca) and Surf Scoters (M. perspicillata) in Puget Sound, WA, and
the Strait of Georgia, BC.
Heather J. Tschaekofske
This study provides an initial examination of the diet of molting scoters in the
Puget Sound, WA, and Strait of Georgia, BC region. Prey species consumption
preferences may indicate specific benefits and foraging strategies sought by molting
scoters. White-winged (Melanitta fusca) and Surf Scoter (M. perspicillata) diets differed
by location, and major taxonomic prey category. The dominant prey in most prior scoter
dietary surveys were bivalves, yet this study found that molting Surf Scoters consumed a
significant amount of non-bivalve prey. Additionally, prey size was an important factor
relating to habitat, and feeding technique, and small prey (< 5.0mm) were likely
consumed in accordance with scoter body size. Molting White-winged Scoters selected a
diet of bivalves almost exclusively, primarily mussels (Mytilus trossulus) and Varnish
clams (Nuttallia obscurata). Molting Surf Scoters consumed bivalves primarily at the
non-estuarine sites, while selecting for more gastropods, crustaceans, and polychaetes at
the estuary sites dominated by eelgrass habitat. The results of this study strongly indicate
that selection of smaller non-molluscan invertebrates by Surf Scoters likely relates to
their higher organic energy content. I conclude that the greater seasonal availability of
diverse prey species during the molting period for scoters allows for increased foraging
opportunities.

iv

Table of Contents:
Abstract …………………………………………………………………………………..iv
Table of Contents …………………………………………………………………………v
List of Figures ……………………………………………………………………………vi
List of Tables ……………………………………………………………………………vii
Acknowledgements …………………………………………………………………….viii
Introduction ……………………………………………………………………………….1
Methods …………………………………………………………………………………...6
Study Area ………………………………………………………………………..7
Fecal Sample Analysis ……………………………………………………………8
Statistical Analysis ………………………………………………………………10
Results …………………………………………………………………………………...12
Discussion ……………………………………………………………………………….20
Bivalve Consumption by Molting Scoters ………………………………………21
Non-Bivalve Consumption by Molting Scoters and Prey Size Selection ………26
Foraging Behavior and Nutrition in Relation to Scoter Size ……………………29
Literature Cited ………………………………………………………………………….36
Appendices ………………………………………………………………………………50

v

List of Figures:
Figure 1. Study site locations for molting White-winged and Surf Scoters in Puget
Sound, WA, and Strait of Georgia, BC …………………………………………………..3
Figure 2. Number of scoter samples with MPC /site …………………………………...16
Figure 3. Results of NMS ordination in three-dimensional space, showing MPC by
location for Surf Scoter samples ………………………………………………………...19
Figure 4. Photograph taken through dissecting microscope (10x power) of two Veneridae
clams, with likely Transennella species on the right, and possible Manila clam on the left
(several small gastropod species also pictured). Clams pictured < 3.0 mm in width (see
Appendix E for additional picture with mm calipers shown) …………………………...22

vi

List of Tables:
Table 1. Summary statistics for scoter samples collected at molting sites ……………..13
Table 2. Frequency of occurrence for prey items identified from Surf and White-winged
Scoter fecal samples at four molting locations in Northern Puget Sound, and Southern
Strait of Georgia, with sample totals in parenthesis …………………………………….14
Table 3. MRPP summary statistics for pair-wise comparisons of MPC consumption at
the four molting locations (rank-transformed data) …………………………………….17

vii

Acknowledgements:
Sincere thanks are due to my primary reader and graduate advisor, A. Styring, for
helpful comments and support throughout the thesis process. Equally important was the
assistance I received from J. Evenson, who was instrumental in project formation and
design, lead the capture efforts and fecal sample collection, and provided expert review,
and critique of thesis drafts. Additional thanks to my faculty reader G. Chin-Leo for
thesis critique, as well as M. Beagle, and the Evergreen State College Lab Stores for
equipment assistance. E. Anderson and J. Barrett were instrumental in the collection and
documentation of fecal samples during captures, and E. Anderson also assisted in the
study design, while J. Barrett coordinated the volunteers from British Columbia. T. Cyra
and B. Murphie were the primary WDFW capture personnel working with J. Evenson
during the banding season. I would also like to thank the numerous WDFW biologists,
and researchers from Simon Fraser University, as well as the volunteers from the public
who spent thousands of person hours during the capture efforts to whom I am thoroughly
in debt. D. Esler assisted in coordinating volunteers from Simon Fraser University, and
Ducks Unlimited Canada provided biologists to assist with the British Columbia scoter
captures. E. Thuesen, as well as personnel from the Pacific Shellfish Research Institute
provided assistance with taxonomic identification of problematic invertebrate species.
Additional thanks to A. Breault with the Canadian Wildlife Service for banding permit
assistance, and J. Cordell for the use of an epibenthic sled.

viii

Introduction:
Many seabird species have experienced dramatic declines in the North American
region over the last several decades (Hodges et al. 1996, Boersma et al. 2002, Dickson
and Gilchrist 2002). Speculation on the causes of decline are variable and diffuse, but
may include habitat modification, toxins, overharvest, disease, and invasive species
(Boersma et al. 2002, Shumway et al. 2003). Scoter (Melanitta spp.) populations in
North America have undergone significant declines, in particular along the west coast
(Hodges et al. 1996, Dickson and Gilchrist 2002, Nysewander et al. 2005, Bower 2009).
Many of the same pressures on seabirds worldwide are experienced by scoter populations
in North America, such as habitat changes on the breeding grounds, food stock depletion,
contaminants, and hunting pressure (Buchanan 2006). These pressures are exacerbated
by low recruitment rates for scoter populations which are dependent on periodic high
recruitment rate years and high adult survival to keep population numbers steady
(Evenson, J. pers. comm., Krementz et al. 1997). Until recently, most research focused on
the nesting biology of scoters, in particular of the White-winged Scoter (Melanitta fusca)
(Krementz et al. 1997, Savard et al. 1998). Recent investigations have begun untangling
the various facets of White-winged and Surf Scoter (Melanitta perspicillata) ecology
outside the breeding season. Much of this recent research has focused on winter and
spring foraging, as well as scoter prebreeding migratory behavior (Lewis et al. 2005,
Žydelis et al. 2006, Lewis et al. 2007a, Kirk et al. 2007, Kirk et al. 2008, Lewis et al.
2008, Anderson et al. 2009b, De La Cruz et al. 2009, Žydelis et al. 2009). The
postreproductive molting ecology of scoters has received very little attention until
recently (Evenson, J. unpubl. data, Savard et al. 2007, Dickson et al. 2010).
1

Sea duck (Mergini) species undergo a simultaneous molt in which all flight
feathers are replaced, resulting in a flightless period typically ranging from 4 – 7 weeks
during the summer and fall (Salomonsen1968, Jehl 1990, Hohman et al. 1992,
Guillemette et al. 2007). This results in a molting period that is compressed in time, and
energetically demanding. For example, Common Eiders (Somateria mollissima) are
flightless for an average of 36 days during late summer (Guillemette et al. 2007). Whitewinged and Surf Scoters are also typically flightless for four weeks approximately
(Anderson, E., pers. comm.). Scoters must compensate for their restricted movement
during the molting period by seeking out molting locations that provide predictable food
resources, as well as adequate predator protection (Salomonsen 1968, Storer and Jehl
1985, Hohman et al. 1992, Murphy 1996). Many scoters undergo an annual post
breeding molt migration from interior taiga/tundra breeding grounds to traditional sites in
saltwater habitats along the pathway to wintering grounds, with the exception of most
female White-winged and some female Surf Scoters that molt near the breeding grounds
(Salomonsen 1968, Johnson and Richardson 1982, Herter et al. 1989, Hohman et al.
1992, Savard et al. 2007). Many of these sites are remote and have been poorly
documented up until recently (Johnson and Richardson 1982, Herter et al. 1989, Brown
and Fredrickson 1997, Savard et al. 1998). In the Pacific Northwest, Surf and Whitewinged Scoters have been documented at molting locations in Northern Puget Sound and
the Southern Strait of Georgia (Savard 1988, Nysewander et al. 2005). From this region,
this study will focus on the following four sites: Boundary Bay, Padilla Bay, Forbes Point
and the Fraser River delta (Figure 1).

2

Figure 1. Study site locations for molting White-winged and Surf Scoters in Puget
Sound, WA, and Strait of Georgia, BC.

Prior studies of the molting ecology of ducks and geese primarily focused on
daily energy expenditures, and whether birds compensate for the high energy demands
during the molting period with changes in behavior or physiology (Ankney 1979, Sjöberg
1988, Thompson and Drobney 1997, Brown and Saunders 1998, Adams et al. 2000,
Guillemette et al. 2007). Few studies have directly examined diet composition during
3

molt for ducks and geese (Hohman et al. 1992, Thompson and Drobney 1997, Rodway
and Cooke 2002), and only one east coast study specifically examined scoter diets during
the molting period (Budge and Gilliland 2007). The high number of scoters that
congregate at saltwater molting locations indicates that these habitats are extremely
valuable, and require further study (Herter et al. 1989, Brown and Fredrickson 1997,
Savard et al. 1998, Savard et al. 2007). At a single molting site such as Padilla Bay
where breeding age female Surf Scoters are dominant, future environmental impacts
could have dramatic effects on Surf Scoter populations.
Diving ducks, such as scoters, have high daily energy expenditures related to
diving activities, and thermoregulation in cold-water environments (Nilsson 1972,
Goudie and Ankney 1986). When molting, these energy needs increase not only to
provide adequate protein synthesis for feather growth, but also to compensate for reduced
thermoregulatory and diving abilities with the loss of flight feathers (Hohman et al.
1992). There appears to be varied species responses to the demands of molt, as some
species select food items higher in protein content, while other species vary their
behaviors to offset the nutritional demands during molt (Hohman et al. 1992, Adams et
al. 2000, Guillemette et al. 2007, Dickson et al. 2010). Larger species have lower energy
demands per unit body mass, which allow for more flexible foraging strategies, while
smaller species are more susceptible to the increased nutritional costs during molt and
may suffer greater stress (Baker and Baker 1973, Hohman et al. 1992). Surf Scoters may
be subjected to greater nutritional demands as they are much smaller than White-winged
Scoters. Additionally, some molting ducks forage less and lose body mass over the
course of the molting period (Sjöberg 1988, Adams et al. 2000, Guillemette et al. 2007).
4

This occurs presumably as an adaptation for a quicker return to flight, as a lighter bird is
able to fly before all feathers re-grow to maximum remigial length (Sjöberg 1988,
Hohman et al. 1992, Brown and Saunders 1998).
Studies of scoter foraging ecology have primarily been conducted during the
winter and spring seasons (Grosz and Yocom 1972, Sanger and Jones 1982, Vermeer and
Bourne 1984, Hirsch 1980, Ryder 2007, Kirk et al. 2007, 2008, Lewis et al. 2007a,
2007b, 2008). Previous studies in North America documented the predominance of
bivalve species in both Surf and White-winged Scoter diets (Grosz and Yocum 1972,
Stott and Olsen 1973, Vermeer and Levings 1977, Hirsch 1980, Bourne 1984, Vermeer
1981, Lewis et al. 2007a). However, a number of studies from the Pacific Northwest
have illuminated greater variability in scoter diets, with alternative prey such as herring
roe, and soft bodied organisms showing increasing importance (Hirsch 1980, Lacroix et
al. 2005, Lewis et al. 2007b, Anderson et al. 2008, Anderson et al. 2009b). In some
cases, this variance may be related to sampling methodology and collection techniques
rather than regional or temporal differences (Anderson et al. 2008). Regardless, the
importance of non-bivalve prey items to scoters has been understated. Specifically,
several studies show Surf Scoter diets diverge from a bivalve dominated diet more often
than White-winged Scoters (Hirsch 1980, Anderson et al. 2008).
Many prior scoter dietary studies employed lethal methodologies in order to
evaluate the food items contained in the esophagus and/or gizzard of each bird (Grosz
and Yocom 1972, Stott and Olson 1973, Hirsch 1980, Sanger and Jones 1982, Vermeer
and Bourne 1984, Lovvorn and Baldwin 1996, Anderson et al. 2008). There is some
potential for differential retention of soft parts versus hard parts depending on whether
5

the esophagus or gizzard contents are analyzed, or whether the study took place in
freshwater versus saltwater (Swanson and Bartonek 1970, Stott and Olson 1973,
Anderson et al. 2008). Depending on how quickly samples are processed, there can be
varying degrees of digestion that occur in gastro-intestinal samples. However, the ability
to measure and weigh whole prey items is much more plausible with a lethal
methodology, versus fecal analysis. Despite the absence of volumetric information, fecal
analysis provides an appropriate snapshot of what the bird has recently consumed (Robert
and Cloutier 2001, Rodway and Cooke 2002). Fecal sample collection is a non-lethal
methodology that can yield large sample numbers, and does not negatively affect
population numbers for scoter species which are already in decline. The drawback with
utilizing fecal samples rather than intact gastro-intestinal samples is that prey break down
more completely, and soft parts may not be easily identifiable in fecal samples. This is a
debatable issue, and there is research utilizing both methodologies within the field of sea
duck dietary research.
The primary objective for this study was to determine to the finest level of detail
what prey species were being consumed by White-winged and Surf Scoters at the four
selected molting locations. This was done with a non-invasive methodology utilizing
fecal samples, and the final results of this study will be compared to prior dietary studies
that have primarily used gastro-intestinal samples. Comparative analysis between the
four molting locations, as well as between the two scoter species may indicate specific
preferences during the molting period. Final results will provide insight into the variable
nutritional benefits related to selected prey consumed by White-winged and Surf Scoters.
Methods:
6

Study Area:
This study was conducted in the Northern Puget Sound, and Southern Strait of
Georgia region at the following four locations: Padilla Bay, Boundary Bay, Fraser River
delta, and Forbes Point (Figure 1). Two of the sites, Padilla Bay, and Boundary Bay,
represent some of the largest estuary seagrass habitat on the Pacific coast of North
America, with 3200, and 5600 hectares of seagrass respectively (Bulthuis 1995, Lovvorn
and Baldwin 1996). Padilla Bay (48°30’N, 122°30’W) is located east of the San Juan
Islands, and is primarily intertidal with sandy substrates, with eelgrass (Zostera marina)
beds comprising the dominant habitat (Ray 1997). Padilla Bay is significantly smaller
than Boundary Bay, yet eelgrass densities are much greater in Padilla Bay. Molting Surf
Scoters were found primarily in the southern portion of Padilla Bay, where freshwater
influences are greatest. Boundary Bay (49°N, 123°1W), situated on the border between
British Columbia and Washington State, is characterized by extensive intertidal mudflats
and eelgrass beds. White-winged Scoters were most frequently found in the shallower
waters of Boundary Bay where eelgrass was sparse, and the substrates were sand/mud.
Surf Scoters primarily occupied the deeper waters at this site, with a thin transition zone
between deep and shallow waters in which both scoter species overlapped. The Fraser
River delta (49°7’N, 123°11W) also has a vast expanse of shallow intertidal mudflats that
are adjacent to Boundary Bay. The Fraser River delta is also estuarine in nature, yet the
large freshwater influence at this site prevents eelgrass from developing, differentiating it
from the two former estuarine sites dominated by eelgrass habitat. Scoters were captured
primarily between Sturgeon Bank and Roberts Bank north of the middle arm of the Fraser
River delta. White-winged and Surf Scoters were primarily found in mixed flocks on the
7

Fraser River delta, with White-winged occasionally occupying shallower waters. Forbes
Point (48°27’N, 122°62’W) separates Oak Harbor and Crescent Bay at the northern end
of Saratoga Passage, and is situated on the northwestern shore of Whidbey Island in
Northern Puget Sound. Forbes Point is characterized by rocky reefs, and sandy substrates
adjacent to the reef outcroppings. Aquacultural activities in the nearby Penn Cove area
have influenced the bivalve composition near Forbes Point, with abundant mussels
growing on the sandy bottom as well as the rock outcroppings. Boundary Bay and the
Fraser River delta are within the Western Hemisphere Shorebird Reserve Network, while
Padilla Bay is a National Estuarine Research Reserve. Each of these designations verifies
the importance of the bird and estuarine habitats at these scoter molting locations.
Fecal Sample Analysis:
Fecal samples were obtained from Surf and White-winged Scoters captured
during a Washington Department of Fish and Wildlife (WDFW) banding project from
July, August, and September 2009. (Breault and Cheng 1990, Paton et al. 1991, Kaiser et
al. 1995, Gilchrist and Heise 1997) WDFW began banding scoters at the study sites in
2007 to ascertain the molting demographics for these two species that have declined on
the west coast. The approximate age of each banded bird was determined through bursal
depth (Iverson et al. 2003), and various morphometric measurements such as 9th primary
length, culmen length, and body mass were also taken. Mass was used as a metric to
characterize the overall health of each bird.
Scoters were selected for fecal collection by securing the first 2-3 extracted birds
from each net set, and placed in plastic lined small animal crates. The other birds

8

captured in the set were placed in mesh bags, and then all birds were transported to a
larger processing vessel. Up to 8 of the bagged birds were placed in plastic lined small
animal crates on the processing boat, and all birds in the lined crates were processed last
to allow time for defecation. After all birds were released, the plastic liners that
contained fecal material were collected, bagged, kept in coolers, and then frozen at the
close of each field day for later analysis.
Sample identification began when samples were thawed, rinsed with water, and
examined under a dissecting microscope (10x power). Several test runs were conducted
sieving fecal samples with 500 μm mesh, and it was determined that important fragments,
such as worm setae, were lost in the sieving process. Prey fragments were identified to
the lowest taxonomic level possible by examining gross physical characteristics such as
interior and exterior shell color, texture, hinge structure, size, and presence of intact
appendages or body structures for non-molluscan prey. Where there were discernable
features, such as complete head parts or intact shells, individual prey species were
counted. Prey identification was facilitated by taxonomic keys (Griffith 1967, Kozloff
1987), as well as with qualitative reference samples collected from each of the study sites
during the same time period. These qualitative collections were categorized and stored in
70% isopropyl alcohol. Expert marine taxonomists were consulted for questionable
samples (Thuesen, E., pers. comm.; Pacific Shellfish Research Institute). Each fecal
sample represents a random sample unit, from an individual bird, and there were no
repeat samples taken. Samples were variable in the amount of fecal content, since some
birds may have defecated during the capture and handling process. Scoters placed
immediately in small animal crates after each net set would likely have more fecal
9

material present than those birds placed in mesh bags prior to placement in the small
animal crates. Therefore, it was not possible to compare volumetric content between
samples. In addition, almost all hard prey items were broken down into fragments
approximately 5- 10 mm across, thus complete prey size was nearly impossible to
ascertain. For each individual fecal sample, prey presence was noted, and intact prey
were measured and counted. After identification, samples were preserved in 70%
isopropyl alcohol.
Statistical Analysis:
For descriptive statistical analysis, average number of species present, standard
deviation, frequency of occurrence, and Shannon-Wiener diversity indices were
calculated for all the prey species present at the four sites. Only major prey categories
(MPC) (bivalvia, gastropoda, crustacea, and polychaeta) , were utilized for ordination and
hypothesis testing, to rule out potential site specific environmental differences unrelated
to scoter dietary choices. For example, Varnish clams (Nuttallia obscurata) may be
abundant at one site, yet absent at another because of habitat differences, yet the choice
of an alternative bivalve species would indicate a similar scoter dietary preference.
Nonparametric statistical analyses were conducted on the MPC using PC-ORD (5.10)
software. Non-metric multidimensional scaling (NMS) ordination was performed to
graphically summarize the environmental patterns in the dataset, and thus the potential
similarity or dissimilarity between sample units (individual fecal samples) (see Appendix
A for description of NMS procedure). Since this study contains data that relate to a nonnormal scale, defined by scoter dietary preferences rather than prey availability, NMS
ordination was chosen because it avoids the assumption of linear relationships between
10

variables while using rank ordering to identify similarity between data points (Clarke
1993, McCune and Grace 2002). Multi-response Permutation Procedure (MRPP) was
utilized for significance testing (p < 0.05) of MPC differences based on location, and
scoter species, sex, and age. MRPP is also robust to datasets that exhibit nonlinear
relationships and is commonly used for ecological community datasets (Biondini et al.
1985, McCune and Grace 2002). Sørensen (Bray-Curtis) distance measures were used
for all MRPP calculations, to match the recommended settings for NMS analysis
(McCune and Grace 2002). Refer to Appendix B for a detailed description of MRPP
functions. Specific statistical differences in MPC consumption for Surf and Whitewinged Scoters at all four molting locations were determined with Fisher’s exact test
using R (2.11.0) software.
NMS is especially suited to the unique characteristics of community ecology
datasets which contain considerable zero values denoting species absence (McCune and
Grace 2002). However, overall comparisons between all four molting locations, and both
scoter species combined required a specific transformation to compensate for the
abundance of zero values within the total dataset. Beals smoothing is a multivariate
transformation which replaces all values in the dataset with new values based on the
probability of occurrence relative to co-occurring species in the dataset (McCune 1994,
McCune and Grace 2002, De Cáceres and Legendre 2008). Smoothing the dataset helps
make underlying patterns more evident, especially when dealing with an abundance of
samples that have few species. Beals smoothing was performed only for the NMS
ordination of both scoter species combined at all four molting locations. These results
revealed significant patterns in the dataset, but were unduly influenced by Surf Scoters at
11

Padilla Bay, thus ultimately each species was assessed separately. Subsequent analysis of
individual scoter species did not require Beals smoothing, and the remaining datasets
were rank-transformed. Rank-transformation involves conversion of the dataset distance
measures to ranks (see Appendix C for transformation procedures). This procedure tends
to produce results most closely aligned with those obtained from NMS ordination
(McCune and Grace 2002).
Results:
Ten samples from the 153 collected had no prey items present in the fecal sample. From
the 143 remaining (Table 1), there were 19 total prey items found in fecal samples from
all of the molting locations combined (Appendix D for total prey samples/site). Most
prey items found were identified to the family or species level. The two estuary sites
averaged more prey per sample, with 2.5 prey/sample (SD ± 1.7, n = 26) for Boundary
Bay, and 2.9 prey/sample (SD ± 1.4, n = 51) for Padilla Bay. Forbes Point averaged 1.4
prey/sample (SD ± 0.6, n = 20), and the Fraser River averaged 1.4 prey/sample (SD ± 0.8,
n = 46). The Shannon-Wiener index was calculated to determine which of the four
locations had greater prey diversity relative to the proportion of prey present. Prey
diversity followed a similar trend at the two estuary sites, with higher diversity indices
(H’) of 2.36 and 2.38 respectively for Boundary Bay and Padilla Bay. The Forbes Point
and Fraser River sites had lower H’ values of 0.94, and 1.15 respectively. The lower
diversity values are not unexpected when looking at the frequency of occurrence for prey
consumed at these two sites, which were both dominated by a single bivalve species
(Table 2).

12

Table 1. Summary statistics for scoter samples collected at molting sites.

Location
Boundary Bay
Forbes Point
Fraser River
Padilla Bay
All Locations

WWSC
Female Male
1
13
2
3
9
19
0
0
12
35

Total
14
5
28
0
47

SUSC
Female Male
9
3
7
8
8
10
48
3
72
24

Total
12
15
18
51
96

All Scoters
26
20
46
51
143

13

Table 2. Frequency of occurrence for prey items identified from Surf and White-winged Scoter fecal samples at four molting
locations in Northern Puget Sound, and Southern Strait of Georgia, with sample totals in parenthesis (F = female, M = male).

Bivalvia
Mytilidae (Mytilus trossulus)
Cardiidae
Macoma balthica
Macoma sp.
Psammobiidae (Nuttallia obscurata)
Veneridae
Unidentified Bivalvia
Gastropoda
Pyramidellidae
Potamididae (Batillaria atramentaria)
Unidentified Gastropoda
Crustacea
Idoteidae
Cancridae
Pinnotheridae
Caprellidea (Caprella sp.)
Unidentified Crustacean
Amphipoda
Echinodermata (Ophiuroidea)
Polychaeta
Plant material

Boundary Bay
SUSC
WWSC
(9F, 3M) (1F, 13M)
FO%
FO%
100
93
25
7
92
14
8
8
57
58
14
17
7
58
7
7
58
42
7
25
8
7
8
17
50
17
14

ForbesPoint
SUSC
WWSC
(7F, 8M)
(2F, 3M)
FO%
FO%
100
100
100
100
40
20
20
7
7
7
7
-

Fraser River
SUSC
WWSC
(8F, 10M)
(9F, 19M)
FO%
FO%
100
100
17
4
11
100
100
17
4
7
7
6
7
7
6
4
7

Padilla Bay
SUSC
(48F, 3M)
FO%
67
25
4
2
35
6
14
47
16
2
35
94
41
4
67
4
8
4
12
12
14

Overall, White-winged Scoters were more reliant on bivalves than Surf Scoters,
with 98% frequency of occurrence for bivalve prey species (Appendix D). Of these, the
Varnish clam was the most frequent bivalve species, found in 79% of White-winged
Scoter samples. Varnish clams have been determined to be an important food and habitat
attribute for both scoter species at other locations in the Strait of Georgia (Žydelis et al.
2006, Kirk et al. 2007, Lewis et al. 2007a, 2008). Additionally, mussels (Mytilus
trossulus) and cockles (Cardiidae family, Clinocardium spp.) occurred in 15%, and 9% of
all White-winged Scoter fecal samples, respectively. A small percentage of non-bivalve
prey species were eaten by White-winged Scoters at the Boundary Bay and Fraser River
sites. Of these, gastropods, and crustaceans occurred with 6% frequency, while
polychaetes were found in 2% of samples.
Bivalves also comprised a large percentage of prey consumed by Surf Scoters,
with 82% frequency of occurrence in samples from all molting locations (Appendix D).
Of these, the most frequent were Varnish clams, with 42% occurrence, followed by
mussels with 35% occurrence. Veneridae (likely Transennella spp.), and cockles were
both found in 14% of samples. Non-bivalve prey species were present in Surf Scoter
fecal samples with a much higher frequency of occurrence than in White-winged Scoters,
with gastropods in 33% of samples, crustaceans in 57%, and polychaetes in 13%. Over
half of the Surf Scoter samples were from the Padilla Bay site, which contributed the
greatest proportion of non-bivalve related prey samples (Figure 2). Removing the Padilla
Bay samples from analysis still results in higher Surf Scoter consumption of non-bivalve
prey than White-winged Scoters, with gastropods in 18% of samples, crustaceans in 16%,
and polychaetes in 13%.
15

Figure 2. Number of scoter samples with MPC /site (total samples above bar).

120

51

100
80
60

Polychaetes
Crustaceans

40
20

28

12
12

14

Gastropods

18

15

Bivalves

5
0
SUSC

WWSC

Boundary Bay

SUSC

WWSC

Forbes Point

SUSC

WWSC

Fraser River

SUSC
Padilla
Bay

Pooled scoter sample analysis between molting sites revealed site specific
variations in prey consumption. Geographically, there were significant differences
amongst the four molting locations (T = -38.24, A = 0.48, p < 0.00) in scoter consumption
of MPC items. The test statistic T reflects the difference between the grouping variables.
Larger negative T values indicate stronger separations between groups.
T = (observed δ – expected δ)/ standard dev. of expected δ
The distribution of δ (weighted mean within-group distance) describes whether or not
prey items are randomly defined by the grouping variables. The probability that this
result is greater than what is expected by chance is determined with Pearson type III
distribution analysis (Biondini et al. 1985, McCune and Grace 2002). Smaller δ values
indicate tighter within group clustering, while larger values indicate a more dispersed
16

arrangement. Finally, the A statistic describes the chance-corrected within-group
agreement, or the effect size (McCune and Grace 2002). An A value equal to one would
indicate absolute within-group homogeneity, with the observed δ equal to the expected δ.
An A value equal to zero indicates that heterogeneity within groups is due to chance.
Values of A greater than 0.3 can be considered fairly large in the context of community
ecology studies (McCune and Grace 2002).
Larger average within-group distances at Boundary Bay and Padilla Bay (δ =
0.14, and 0.10 respectively) indicate greater within-group variation, while the Forbes
Point and Fraser River sites (δ = 0.04, and 0.04 respectively) tend to be more clustered.
This likely reflects the greater prey diversity at the two estuary sites dominated by
eelgrass habitat. Pair-wise comparisons between all four molting locations revealed
significant differences between all sites, with the exception that prey consumption
between Forbes Point and the Fraser River sites was similar for both scoter species (T =
0.97, A = -0.01, p = 1.0) (Table 3). The greatest differences found in MPC consumption
were found when Padilla Bay was compared to the other three molting locations,
indicating the unique nature of this site.
Table 3. MRPP summary statistics for pair-wise comparisons of MPC consumption at
the four molting locations (rank-transformed data).

Boundary Bay vs. Forbes Point
Boundary Bay vs. Fraser River
Boundary Bay vs. Padilla Bay
Forbes Point vs. Fraser River
Forbes Point vs. Padilla Bay
Fraser River vs. Padilla Bay

T
-4.73
-8.01
-22.00
0.97
-38.22
-54.57

A
0.11
0.11
0.22
-0.01
0.47
0.57

p
0.00
0.00
0.00
1.00
0.00
0.00
17

Of the three locations with both scoter species present, significant differences
existed between White-winged and Surf Scoter diets at Boundary Bay (T = -8.80, A =
0.31, p < 0.00), where Surf Scoters consumed more gastropods (p = 0.009, two-sided
Fisher’s exact test), and polychaetes (p = 0.004, two-sided Fisher’s exact test). At the
remaining two sites with both scoter species present the difference between Whitewinged and Surf Scoter MPC consumption was not significant (Forbes Point, T = -0.33, A
= 0.01, p = 0.27) (Fraser River, T = 0.26, A = -0.00, p = 0.49).
Individual analysis of scoter diets at the four molting locations indicated specific
dietary preferences for Surf Scoters. NMS Ordination analysis revealed significant (p =
0.02) patterns in three-dimension space for Surf Scoter consumption of MPC at the four
molting sites (Figure 3). Ordination results for White-winged Scoters were inconclusive,
largely because of the overwhelming influence of bivalve prey.

18

Figure 3. Results of NMS ordination in three-dimensional space, showing MPC by location for Surf Scoter samples (MPC symbols
abbreviated: B, G, C, P).

19

MRPP analysis displayed similar trends in Surf Scoter consumption of MPC, with
significant differences amongst the four sites (T = -27.04, A = 0.43, p = 0.00). However,
there was no significant difference between Surf Scoter consumption of MPC at Forbes
Point and the Fraser River (T = 0.75, A = -0.02, p = 0.76). Male and female Surf Scoters
had significant differences in consumption of MPC across molting locations (T = -14.02,
A = 0.13, p = 0.00), specifically females ate more crustaceans (p < 0.00, two-sided
Fisher’s exact test). There were no differences in prey consumption by Surf Scoter age
classes. White-winged Scoters ate bivalves almost exclusively, thus there were no
significant differences in MPC consumption by sex, age, or location.
Discussion:
This study presents some of the first findings regarding the diet specific to the
molting period for Surf and White-winged Scoters in the Puget Sound, WA, and Strait of
Georgia, BC region. These findings may contribute to the broader investigation of scoter
declines in North America. By identifying critical habitat, and food resources, future
management of scoter resources can be appropriately directed. In general, White-winged
and Surf Scoters consumed a high proportion of bivalves at each of the study sites, with
the exception of Padilla Bay, where non-bivalve species were more prevalent for Surf
Scoters. Since the majority of Surf Scoters at Padilla Bay are breeding age females,
consumption of non-bivalve related prey in higher frequencies is noteworthy. At the two
estuary sites with eelgrass habitat, frequency of non-bivalve prey items rivaled bivalve
occurrence. Despite best efforts, these findings likely underestimate the proportion of
non-bivalve species since the amount of soft material that is retained in fecal samples is
lower than that obtained from gastro-intestinal sampling. Regardless, the specific bivalve
20

and non-bivalve prey that were consumed during the molting period undoubtedly relate to
specific nutritional and behavioral needs for each scoter species during this demanding
life stage.
Bivalve Consumption by Molting Scoters:
Prior scoter dietary studies emphasized the importance of bivalve species,
particularly Manila clams (Venerupis philippinarum), Littleneck clams (Protothaca
staminea), mussels, and Varnish clams in the Pacific Northwest (Vermeer and Levings
1977, Vermeer 1981, Vermeer and Bourne 1982, Bourne 1984, Ryder 2007, Žydelis et al.
2006, Lewis et al. 2007a, 2008). Interestingly, there were no adult Manila or Littleneck
clams present in any of the fecal samples, as almost all the Veneridae family prey items
were most likely Transennella species. One possible juvenile Manila clam was found
with the potential Transennella clams in a sample from Boundary Bay, and can be
visually compared in the photograph in Figure 4. Transennella tantilla clams were found
to be widespread throughout Padilla Bay habitats in a 1994 survey of the bay (Ray 1997).
These clams live on the surface of the sediments, generally not exceeding 5.0 mm in
length. The Veneridae bivalves consumed in this study were < 4.0 mm in size, and often
densely aggregated within samples. The small size of the Veneridae bivalves overall
suggests that scoters were likely siphon or suction feeding these bivalves from the sandy
sediments at the two estuary sites, as well as the Fraser River delta. Siphon feeding has
been found to be an efficient foraging strategy for sea ducks consuming small high
energy prey items, such as herring roe and small bivalve species (Sanger and Jones 1982,
De Leeuw and Van Eerden 1992, Lewis et al. 2007b). The soft sediments of the estuary

21

sites may also provide opportunities for consumption of other non-bivalve species in a
similar manner.
Figure 4. Photograph taken through dissecting microscope (10x power) of two Veneridae
clams, with likely Transennella species on the right, and possible Manila clam on the left
(several small gastropod species also pictured). Clams pictured < 3.0 mm in width (see
Appendix E for additional picture with mm calipers shown).

Transennella spp.

Of the larger bivalves consumed, Varnish clams and mussels dominated both Surf
and White-winged Scoter diets, with the exception of the Cardiidae species present in
Surf Scoter samples from Boundary Bay. Seasonal differences in mussel size and
availability often dictate consumption by scoters, and other sea duck species (Zwarts and
Wanink 1993, Bustnes 1998, Guillemette 1998, Kirk et al. 2007, 2008). Kirk et al.
(2008) found that seasonal depletion of mussels likely affected scoter movements later in
22

the winter. Many bivalve species are more abundant in the summer and fall, thus scoters
may not be limited by prey availability during the molting period as much as they would
during the winter (Zwarts and Wanink 1993). Various studies have inferred differential
energy gains from the consumption of mussels versus clams (Zwarts and Wanink 1993,
Bustnes 1998, Richman and Lovvorn 2003). In many instances, mussels are more
efficiently acquired, since they are not buried within the sediment and are often found
visibly anchored in dense clusters. However, the greater energy content of larger clams
such as Manila and Varnish clams offsets some of the expense from searching in
sediments (Kirk et al. 2007). Smaller mussels have higher organic energy content, and
less shell than larger mussels proportionally (Bustnes 1998). Mussels have historically
been an important prey item for both scoter species at the Fraser River delta, occurring in
19% of White-winged Scoter samples taken during the summer of 1974, and constituting
the primary prey in Surf Scoter samples annually from 1974 – 1977 (Vermeer and
Levings 1977, Vermeer 1981, Vermeer and Bourne 1982). The dependence on mussels
may be changing as the recent appearance and dominance of the Varnish clam in scoter
samples from this region would indicate.
Since the accidental introduction of the Varnish clam in the late 1980’s in British
Columbia waters, it has rapidly expanded throughout the Strait of Georgia and Puget
Sound (Gillespie et al. 1999). Recent studies of wintering scoters in this region have
found that both Varnish and Manila clams dominate Surf and White-winged Scoter diets
(Žydelis et al. 2006, Lewis et al. 2007a, Lewis et al. 2008). Varnish clams were found to
be a significant determinant for predicting scoter densities in the Baynes Sound, Strait of
Georgia region (Žydelis et al. 2006). Their presence and the absence of the more
23

regularly encountered Manila and Native Littleneck clams from molting scoter fecal
samples is noteworthy when comparing prior scoter dietary studies. All three of these
bivalve species can be found at the four molting sites throughout the year (Vermeer and
Levings 1977, Vermeer and Bourne 1982, Vermeer 1981, Dinnel 2000), yet scoters
appear to be selecting Varnish clams during the molting period. In general, Varnish
clams are more freshwater tolerant, and situate higher in the intertidal zone, and tend to
be located at similar or slightly deeper substrate depths when they co-occur with Manila
clams (Gillespie et al. 1999). Varnish shells are thinner than Manila clams, yet they
contain greater flesh ratios (Gillespie et al. 1999), making them an attractive prey
alternative for scoters. Additionally, it appears that this relatively new invasive species
may be linked to the recent appearance of substantial molting scoters at the Fraser River
during the summer. Coastal molting sea duck surveys had documented very few scoters
utilizing the Fraser River during the molting period prior to 1986 (Savard 1988).
Currently, approximately 10,000 or more molting scoters utilize the Fraser River
(Evenson, J., pers. comm.).
There have been other instances of ducks altering migratory behaviors to
congregate in dense flocks where invasive prey species have recently become prolific.
Specifically, Greater (Aythya marila) and Lesser Scaup (A. affinis) altered their diets in
the Great Lakes after the accidental introduction of the Zebra Mussel (Dreissena
polymorpha and D. bugensis), as well as in the San Francisco Bay area where the Asian
clam (Corbula amurensis) is now a favored prey species (Wormington and Leach 1992,
Petrie and Knapton 1999, Poulton et al. 2002). Invasive clam species such as the Zebra
mussel and Asian clam have been found to accumulate toxins at greater levels than other
24

native bivalve species (Hoffman et al. 1998, Petrie and Schummer 2002). The Zebra
mussel has prolific filtering capabilities that differ from the Varnish clam, yet Varnish
clams feed through both filter and deposit feeding mechanisms (Gillespie et al. 1999),
which may also permit greater toxin accumulation (Meacham 2010). The non-native
Manila clam has been studied in relation to Varnish clams in light of their similar
commercial benefits. Experimental treatments between Varnish and Manila clams show
that they tend to negatively affect each other when both are present in the same area
(Gillespie et al. 2001).
The Varnish clam has been proposed as a commercial clam species, yet there has
been little toxicological testing of contaminant or marine toxin accumulation with this
species to date. Testing done off Vancouver Island in September and October of 1998 by
the Northwest Fisheries Science Center division of the National Oceanic and
Atmospheric Administration, found Varnish clams accumulated slightly higher levels of
PSP toxin than Manila clams and oysters (NOAA-Northwest Fisheries Science Center
1998). They were tested twice during the fall season, and consistently retained higher
toxin levels than Manila clams. These levels were below the toxic indicator threshold for
human consumption, yet this may indicate higher retention of toxins in general for the
Varnish clam species. Gillespie et al. (1999) cited limited testing done by the Canadian
Food Inspection Agency in 1998 which found similar levels of toxicity between Varnish
clams and Manila clams. Without further testing, it is impossible to say whether there is
any concern for biomagnification of toxins in scoters from consuming large quantities of
Varnish clams. Despite this, scoters are certainly exposed to industrial and agricultural
contaminants that runoff from the Vancouver metropolitan and Fraser valley areas into
25

the Fraser River. In addition, the Vancouver International Airport is situated at the north
end of the Fraser delta near molting scoters, along with major shipping and coal
industries. Future research could compare the toxicity of molting scoters in these
industrial areas to those that molt on the outer coast at sites without urban influences, as
well as the toxicity of scoters present on the Fraser River year round. The relationship
between Varnish clams, and Surf and White-winged Scoters in the Pacific Northwest
deserves further scrutiny.
Non-Bivalve Prey Consumption by Molting Scoters and Prey Size Selection:
Molting Surf Scoters in this study consumed a greater variety of non-bivalve prey
items than White-winged Scoters. Consumption of gastropods, crustaceans, and
polychaetes was greatest at the two estuary sites with eelgrass habitat, highlighting the
importance of this habitat as it provides a diversity of epifaunal invertebrate prey. Surf
Scoters at Boundary Bay fed most frequently near eelgrass (Zostera marina) habitat
compared to White-winged Scoters, and this is evident when examining the significant
difference in prey species consumption from this site. Specifically, isopods and
pinnotheridae crabs were most frequently found in Surf Scoter fecal samples from
Boundary Bay and Padilla Bay. Two frequently encountered isopod species are the
Rockweed (Idotea wosnesenskii) and Eelgrass Isopod (Idotea resecata), commonly found
under rocks and amongst seagrasses. Isopod fragments in fecal samples were too small
to clearly identify to species. Pinnotheridae crabs are generally found commensally
within other invertebrates, most frequently in bivalves, or inside worm tubes. The pea
crab Pinnixa tubicola is the most commonly encountered pinnotheridae crab in sheltered
waters along the west coast, such as at Padilla Bay (Ricketts and Calvin 1978, Ray 1997,
26

Riggs 2003). P. tubicola resides within worm tubes, generally of terebellid polychaete
species. At Padilla Bay, these crabs have been found in Ulva spp. and Z. marina
dominated habitats, ranging from shallow to subtidal depths (Ray 1997). Riggs (2003)
documented P. tubicola ranging in size from 3.1 to 7.1 mm in above-ground traps at
Padilla Bay. In this study, Pinnotheridae crab fragments found in Surf Scoter fecal
samples from Padilla Bay were generally < 5.0 mm in size, and quite numerous in a
number of samples. Since this crab species resides in worm tubes, it is possible that Surf
Scoters were grabbing clumps of sediment in order to obtain worm and crab prey.
Another common pinnotheridae crab species that can be found in the Puget Sound
region is Fabia subquadrata, otherwise known as the mussel pea crab. They are
primarily commensal in mussels, but may inhabit other bivalves periodically (Pearce
1966, Ricketts and Calvin 1978), and have recently been found in Varnish clams (Cowles
2007). Additionally, Gillespie et al. (1999) found a similar species of pea crab, Pinnixa
faba, within many Varnish clams in British Columbia waters. The mussel pea crab has
several life stages in which it may periodically be found outside its host, during which the
shell hardens and its legs are densely fringed with setae for swimming (Wicksten 2009).
Some of the pinnotheridae leg fragments found in my Padilla Bay fecal samples had
dense setae, and the possibility of this crab being encountered more readily while outside
the commensal host makes this an additional prey species to consider for scoter
consumption.
Anderson et al. (2008) also reported that isopods and pinnotheridae crabs were the
dominant crustaceans found in Surf Scoters taken from Padilla Bay in 2005 and 2006. In
addition, Henny et al. (1991) found crab in 60% of Surf Scoter stomachs sampled in the
27

fall and winter from Alsea Bay, an isolated estuary on the outer Oregon coast. Other sea
ducks have displayed a preference for crabs as prey. Harlequin Ducks (Histrionicus
histrionicus) in the Strait of Georgia preferentially consumed crabs at the highest
frequency during their molting period (Rodway and Cooke 2002). Guillemette et al.
(1992) found that Common Eiders were selecting for crabs at a higher frequency relative
to their availability in the habitat, likely because crabs contained more organic energy
content than the alternative prey, mussels and urchins. Padilla Bay was the one site in
which bivalves were not the dominant prey item, instead crustaceans were found in 94%
of Surf Scoter samples. The high organic energy content, combined with the softer shell
of small crab species makes them an attractive prey item compared to many larger
bivalve species.
Anderson et al. (2008) found in their survey of prior Surf and White-winged
Scoter dietary studies that White-winged Scoters had higher mean consumption of
gastropods and crustaceans. Surveys of Boundary Bay in 1974, and 1976 revealed that
both Surf and White-winged Scoters ate more crustaceans during the summer than the
winter period (Vermeer and Levings 1977). In a number of selected studies, barnacles
were included as the dominant crustacean prey species (Vermeer and Levings 1977,
Vermeer and Bourne 1982). In these studies, mussels were also present, and as barnacles
are frequently found attached to mussels, their presence in scoter samples may have been
incidental. However, in both studies, the barnacles were consumed during the summer,
whole, and in large quantities which suggests they were consumed intentionally.
Vermeer and Levings (1977) suggested that these were likely consumed to serve as grit,
since there is little flesh value in barnacles. However, seasonal differences in prey
28

availability may foster greater reliance on barnacles as a prey species throughout the year,
in contrast to the greater diversity of prey items present during the molting period.
Barnacles were found infrequently in my samples, generally amongst mussel fragments,
and were not included in the final analysis.
Consumption of gastropods was greatest for Surf Scoters, despite prior evidence
that White-winged Scoters consume a greater proportion of gastropods (Anderson et al.
2008). Almost all gastropods consumed were < 4.0 mm in size. Energy savings during
the molting period can be achieved in a number of ways, one of which is to reduce the
amount of shell material that is processed. Consumption of less shell material by eating
smaller bivalves and gastropods relates to an increase in gizzard efficiency (DeLeeuw
and Van Eerden 1992, Bustnes 1998, Guillemette 1998), compensating for increased
metabolic demands during molt. DeLeeuw and Van Eerden (1982) found that Tufted
Ducks (Aythya fuligula) selected for smaller mussels on deeper dives, which could be
consumed below the surface rather than wasting time searching for larger mussels that
had to be swallowed above water. Since many birds compensate either behaviorally or
physiologically during the molt of flight feathers, scoters may be selecting for smaller,
energy dense prey which can be more efficiently processed by the gizzard.
Foraging Behavior and Nutrition in Relation to Scoter Size:
Surf Scoters are smaller than White-winged Scoters, and have smaller, more
flexible bills that may allow for greater feeding opportunities with prey that are smaller
and more mobile (Goudie and Ankney 1986). The small size of many prey species
consumed at the estuary sites with eelgrass habitat in this study highlights the diversity of

29

feeding mechanisms being employed by Surf Scoters. Small estuarine invertebrates, such
as the Eelgrass Isopod, as well as small crabs, and gastropods can be found attached to
eelgrass blades. Observational evidence suggests that Surf Scoters may strip prey from
the blades of eelgrass while underwater (Anderson, E., pers. comm.). Plant material was
found in 17% of Surf Scoter samples from Boundary Bay, and 12% of Padilla Bay
samples. These plants may be ingested incidentally while scoters fed on attached
invertebrates, or as debris while feeding off the substrate surface (Stott and Olson 1973).
The greater mobility of smaller sea ducks and the high energy content of small
invertebrate prey may offset the need to consume larger prey items. Goudie and Ankney
(1986) found that smaller sea ducks such as Harlequin Ducks and Oldsquaws (Clangula
hyemalis) consumed a varied diet, with more energy dense small invertebrates, such as
isopods and amphipods. In their study, despite the overlap in habitat, larger sea ducks
such as Black Scoters (Melanitta nigra) and Common Eiders, were not taking advantage
of the smaller prey items as much. Further studies of scoters during the molting period
may aim at quantifying foraging behavior. However care should be taken in observations
of above water feeding behavior, as evidence suggests that below surface feeding
techniques may be more complex (De Leeuw and Van Eerden 1992).
Molting sea ducks that undergo a simultaneous wing molt would be expected to
have high costs associated with the intensity of molting all flight feathers at one time.
Birds that do not compensate behaviorally for these demands must meet their nutritional
needs through available food resources. Scoters in the present study had body mass
averages that were comparable to taxonomic averages, thus it does not appear that birds
were lower in mass due to decreased foraging activities. Male and female White-winged
30

Scoters ranged from 1560 to 1180 g average mass, respectively, while Surf Scoters range
from 1050 to 900 g, respectively (Brown and Fredrickson 1997, Savard et al. 1998). In
this study, male and female White-winged Scoters averaged 1712 and 1502 g,
respectively, while Surf Scoters averaged 1090 and 959 g, respectively. Dickson et al.
(2010) found that molting White-winged and Surf Scoters from Puget Sound, WA, and
the Strait of Georgia, BC, were not losing mass during the molting period in 2008 and
2009.
Comparative analysis reveals that female Surf Scoters at Padilla Bay are
significantly lighter than the rest of their cohort at the other three molting locations
(Evenson, J. unpubl. data). This has been true for each year of the banding study (20072009) indicating that the Padilla birds have consistently been the lightest in this cohort.
Interestingly, a recent case study of Padilla Bay that synthesized prior seabird monitoring
results over the last several decades revealed overall declines for marine birds, yet while
most diving duck densities had declined, scoters had increased in the bay during the
winter and spring seasons (Anderson et al. 2009a). The availability of diverse food
resources at this site may be contributing to increased scoter presence, which in turn may
contribute to greater competition for food resources, possibly resulting in birds of lower
mass.
In general, since scoters do not appear to be losing mass on the molting grounds,
they are likely meeting their nutritional needs through available food resources (Dickson
et al. 2010). Hohman et al. (1992) reviewed studies of molting waterfowl diets during the
breeding and post-breeding period, and found very little evidence to suggest that diet
shifts tremendously during the molting period to meet protein requirements. Waterfowl
31

were generally eating the same proportion of plant or animal matter in both periods.
Waterfowl with less protein in their diets tend to lengthen the molting period, to provide
for adequate protein synthesis for feather re-growth (Hohman et al. 1992, Thompson and
Drobney 1997). Since scoters are relying on exogenous protein sources, the nutritional
content of specific prey species consumed during the molting period may be highly
informative.
Determination of the energetic content of invertebrate prey species is complicated
by various factors such as time of year, burial depth, density, age, thickness of shell, as
well as regional differences (Zwarts and Wanink 1993, Bustnes 1998, Larsen and
Guillemette 2000, Richman and Lovvorn 2003). Many bivalve species may provide
dramatically different energy estimates during bivalve spawning periods, as gonadal and
egg development add energetic content (Thayer et al. 1973, Zwarts and Wanink 1993).
Without quantitative sample analysis of prey species found at the molting sites, it is
impossible to state absolute energetic differences. Yet comparative analysis from other
studies tends to indicate that energetic density is generally greater for crustaceans and
polychaetes, than bivalves and gastropods (Thayer et al. 1973, Goudie and Ankney 1986,
Jorde and Owen 1988, Zwarts and Wanink 1993). The greater energy content is often
related to the lack of hard shell material that must be processed by the gizzard.
Guillemette et al. (1992) found energy content per gram live mass in mussels was about
two times greater than that in sea urchins, while crabs had three times the energy content
of mussels. Pinnotheridae crabs had some of the highest energy densities compared to
bivalves, gastropods, polychaetes, and other crustaceans at an estuary site in North

32

Carolina (Thayer et al. 1973). This particular pea crab also has both soft and hard
skeleton stages, in which energy densities would undoubtedly be highest in the soft stage.
It appears that Surf and White-winged Scoters may be partitioning their molting
habitat to obtain preferred prey species, likely in accordance with scoter size. Both scoter
species are capable of eating similar prey, yet Surf Scoters are selecting for non-bivalve
prey species with high frequency in the eelgrass estuary habitats during the molting
period. It would appear that there are strong energetic gains from this prey selectivity,
which may range from the overall energetic content of non-bivalve prey, to the lighter
shell processing required from a sea duck that is already physiologically compensating
for feather re-growth. The dominance of mussels and Varnish clams at the remaining
molting sites likely indicates that these bivalve species are readily available in these
habitats. As such, it is important to maintain these molting habitats, as they provide the
food resources necessary during this important life stage, as well as safety during the
flightless period. Each of these sites has historically experienced varying levels of
developmental pressure, from aquaculture operations, to pollution from upland sources.
Just as scoter populations were likely affected by recently declining herring runs in the
Puget Sound region, a single large incident or habitat modification at molting locations
could drastically affect these sea ducks. For example, a major oil spill at the Padilla Bay
estuary from local refineries could have devastating effects on the large numbers of
waterfowl that utilize this habitat. Conversely, it is also possible that the appearance of
an invasive species, such as the Varnish clam, could have positive ramifications for the
needs of scoter species, as it is evident that both Surf and White-winged Scoters are
consuming it in large quantities.
33

Have scoters selected traditional molting sites because food resources are
abundant at these sites, or because specific food resources exist at these sites to provide
for unique nutritional needs during this life stage? Based on the recent appearance of
numerous molting scoters in the Fraser River delta concurrently with the initial expansion
of Varnish clams, it would appear that specific food resources may be more influential.
As waterfowl adjusted their migratory behavior to the appearance of Zebra mussels as a
novel food resource in the Great Lakes, it would appear that scoters have done the same
in the Pacific Northwest. I would argue that the specific prey species occurring with the
greatest frequency in scoter samples from each of the molting locations are likely the
reason that scoters return every year, just as herring spawn influence scoter movements
during spring migration (Lewis et al. 2007b, Anderson et al. 2009b). Vermeer and
Levings (1977) documented approximately 10,000 scoters, primarily Surf, congregating
in Boundary Bay during the molting period in 1974. In Vermeer and Levings (1977)
study, both scoter species ate more snails and crustaceans, and fewer bivalves in summer
and fall, versus during the winter. The diversity of alternative prey consumed at the two
estuary sites with eelgrass habitat in this study, Boundary Bay and Padilla Bay, mussels
at Forbes Point, and Varnish clams at the Fraser River, are likely providing for the
nutritional needs of molting scoters in a unique way. The absence of any adult Manila or
Littleneck clams in any molting scoter fecal samples likely indicates their less desirable
status as a prey species during this life stage.
If molting habitat selection is based on the preferential selection of certain prey
species, then it is extremely important to protect and manage these habitats accordingly.
Protection of eelgrass habitat in particular may require further emphasis, as many prior
34

scoter studies have potentially over-emphasized the importance of bivalves as a dominant
prey species, for Surf Scoters specifically (Anderson et al. 2008). Estuarine habitat is
known to provide substantial resources for marine plants and animals, and the nutritional
linkages in this habitat directly influence the productivity of various sea duck species
(Stott and Olson 1973, Baldwin and Lovvorn 1994). Unfortunately, eelgrass habitat has
undergone significant declines worldwide due to habitat change, pollution, invasive
species, and decreased water clarity (Ralph et al. 2006). As annual stopover sites for
migrating scoters, habitat conditions at these molting sites relate directly to the future
fitness of scoter population. Additional quantitative analysis of prey species density,
distribution, and biomass may lead to improved management for Surf and White-winged
Scoters at these migratory stopover sites.

35

Literature Cited:
Adams, P. A., G.J. Robertson, and I. A. Jones. 2000. Time-activity budgets of Harlequin
Ducks molting in the Gannet Islands, Labrador. The Condor 102: 703-708.
Anderson, E. M., J. R. Lovvorn, and M.T. Wilson. 2008. Reevaluating marine diets of
Surf and White-winged Scoters: interspecific differences and the importance of softbodied prey. The Condor 110:285-295.
Anderson, E. M., J. L. Bower, D. R. Nysewander, J. R. Evenson, and J. R. Lovvorn.
2009a. 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., J. R. Lovvorn, D. Esler, W. S. Boyd, and K. C. Stick. 2009b. Using
predator distributions, diet, and condition to evaluate seasonal foraging sites: sea ducks
and herring spawn. Marine Ecology Progress Series 386: 287-302.
Ankney, C. D. 1979. Does the wing molt cause nutritional stress in Lesser Snow Geese?
The Auk 96: 68-72.
Baker, M. C. and A. E. M. Baker. 1973. Niche relationships among six species of
shorebirds on their wintering and breeding ranges. Ecological Monographs 43: 193-212.
Baldwin, J. R., and J. R. Lovvorn. 1994. Habitats and tidal accessibility of the marine
foods of dabbling ducks and brant in Boundary Bay, British Columbia. Marine Biology
120: 627-638.

36

Biondini, M. E., C. D. Bonham, and E. F. Redente. 1985. Secondary successional
patterns in a sagebrush (Artemisia tridentate) community as they relate to soil disturbance
and soil biological activity. Vegetatio 60: 25-36.
Boersma, P.D., J. A. Clark, and N. Hillgarth. 2002. Seabird conservation. Pages 559579 in E.A. Schreiber and J. Burger, editors. Biology of Marine Birds. CRC Press, Boca
Raton, Louisiana, USA.
Bower, J. L. 2009. Changes in marine bird abundance in the Salish Sea: 1975 to 2007.
Marine Ornithology 37: 9-17.
Bourne, N. 1984. Clam predation by scoter ducks in the Strait of Georgia, British
Columbia, Canada. Canadian Technical Report of Fisheries and Aquatic Sciences No.
1331.
Breault, A. M., and K. M. Cheng. 1990. Use of submerged mist nets to capture diving
birds. Journal of Field Ornithology 61: 328-330.
Brown, P.W., and L. H. Fredrickson. 1997. White-winged Scoter (Melanitta fusca).
Pages 1-27 in A. Poole and F. Gill, editors. The Birds of North America, No. 274. The
Academy of Natural Sciences, Philadelphia, Pennsylvania, and The American
Ornithologists’ Union, Washington, D.C., USA.
Brown, R. E., and D. K. Saunders. 1998. Regulated changes in body mass and muscle
mass in molting Blue-winged Teal for an early return to flight. Canadian Journal of
Zoology 76: 26-32.

37

Buchanan, J.B. 2006. Nearshore birds in Puget Sound. Puget Sound Nearshore
Partnership Report number 2006-05. Published by Seattle District, U.S. Army Corps of
Engineers, Seattle, Washington.
Budge, S., and S. Gilliland. 2007. Diet of Surf Scoters molting in Eastern North
America (SDJV#92). Sea Duck Joint Venture annual project summary for endorsed
projects, FY 2007. Available: http://www.seaduckjv.org/
Bulthuis, D. A. 1995. Distribution of seagrasses in a North Puget Sound estuary: Padilla
Bay, Washington, USA. Aquatic Botany 50: 99-105.
Bustnes, J. O. 1998. Selection of Blue Mussels, Mytilus edulis, by Common Eiders,
Somateria mollissima, by size in relation to shell content. Canadian Journal of Zoology
76: 1787-1790.
Clarke, K. R. 1993. Non-parametric multivariate analyses of changes in community
structure. Australian Journal of Ecology 18: 117-143.
Cowles, D. (2007). Fabia subquadrata (Dana, 1851). In Key to invertebrates found at
or near Rosario Beach Marine Laboratory, Fidalgo Island, Anacortes, Washington
[online]. Available: http://www.wallawalla.edu/academics/departments/biology/rosario
/inverts/index.html [2010, October 10].
De Cáceres, M. and P. Legendre. 2008. Beals smoothing revisited. Oecologia 156: 657669.
De La Cruz, S. E. W., J. Y. Takekawa, M. T. Wilson, D. R. Nysewander, J. R. Evenson,
D. Esler, W. S. Boyd, and D. H. Ward. 2009. Spring migration routes and chronology of
38

Surf Scoters (Melanitta perspicillata): a synthesis of Pacific coast studies. Canadian
Journal of Zoology 87: 1069-1086.
De Leeuw, J. J., and M. R. Van Eerden. 1992. Size selection in diving Tufted Ducks
Aythya fuligula explained by differential handling of small and large mussels Dreissena
polymorpha. Ardea 80: 353-362.
Dickson, D. L., and H. G. Gilchrist. 2002. Status of marine birds of the Southeastern
Beaufort Sea. Arctic 55: 46-58.
Dickson, R., D. Esler, J. Hupp, E. Anderson, J. Evenson, and J. Barrett. 2010. Remigial
moult of Surf Scoters (Melanitta perspicillata) and White-winged Scoters (M. fusca) on
the Pacific coast of North America. 1st World Seabird Conference, Victoria, British
Columbia.
Dinnel, P. A. 2000. Padilla Bay mollusks: a review, with emphasis on the bivalves.
Washington State Department of Ecology, Padilla Bay National Estuarine Research
Reserve Technical Report No. 24 (Padilla Bay Estuarine Profile Contribution No. 1).
Gilchrist, H. G. and K. Heise. 1997. Studies on the demography and reproductive
ecology of common and king eiders breeding at East Bay, Southampton Island. Canadian
Wildlife Service.
Gillespie, G. E., M. Parker, and W. Merilees. (1999). Distribution, abundance, biology
and fisheries potential of the exotic Varnish Clam (Nutallia obscurata) in British
Columbia. In Canadian Stock Assessment Secretariat Research Document 99/193

39

[online]. Available: www.dfo-mpo.gc.ca/csas/Csas/DocREC/1999/pdf/99_193e.pdf
[2010, Oct 10].
Gillespie, G. E., B. Rusch, S. J. Gormican, R. Marshall, and D. Munroe. (2001). Further
investigations of the fisheries potential of the exotic Varnish Clam (Nutallia obscurata) in
British Columbia. In Canadian Science Advisory Secretariat Research Document
2001/143 [online]. Available: www.dfo-mpo.gc.ca/csas/Csas/DocREC/2001/
RES2001_143e.pdf [2010, Oct 10].
Goudie R. I., and C. D. Ankney. 1986. Body size, activity budgets, and diets of sea
ducks wintering in Newfoundland. Ecology 67: 1475-1482.
Griffith, L. M. 1967. The intertidal univalves of British Columbia. Handbook No. 26.
British Columbia Provincial Museum, Victoria, British Columbia, Canada.
Grosz, T., and C. F. Yocom. 1972. Food habits of the White-winged Scoter in
Northwestern California. The Journal of Wildlife Management 36: 1279-1282.
Guillemette, M. 1998. The effect of time and digestion constraints in Common Eiders
while feeding and diving over Blue Mussel beds. Functional Ecology 12: 123-131.
Guillemette, M., R. C. Ydenberg, and J. H. Himmelman. 1992. The role of energy
intake rate in prey and habiat selection of Common Eiders Somateria mollissima in
winter: a risk-sensitive interpretation. Journal of Animal Ecology 61: 599-610.
Guillemette, M., D. Pelletier, J.-M. Grandbois, and P. J. Butler. 2007. Flighlessness and
the energetic cost of wing molt in a large sea duck. Ecology 88: 2936-2945.

40

Henny, C. J., L. J. Blus, R. A. Grove, and S. P. Thompson. 1991. Accumulation of trace
elements and organochlorines by Surf Scoters wintering in the Pacific Northwest.
Northwestern Naturalist 72: 43-60.
Herter, D. R., S. M. Johnston, and A. P. Woodman. 1989. Molt migration of scoters at
Cape Peirce, Alaska. Arctic 42: 248-252.
Hirsch, K. V. 1980. Winter ecology of sea ducks in the inland marine waters of
Washington. MS Thesis. University of Washington, Seattle, Washington, USA.
Hodges, J. I., J. G. King, B. Conant, and H. A. Hanson. 1996. Aerial surveys of
waterbirds in Alaska 1957-94: population trends and observer variability. National
Biological Service Information and Technology Report.
Hoffman, D. J. , H. M. Ohlendorf, C. M. Marn, G. W. Pendleton. 1998. Association of
mercury and selenium with altered glutathione metabolism and oxidative stress in diving
ducks from the San Francisco Bay region, USA. Environmental Toxicology and
Chemistry 17:167-172.
Hohman, W. L., C. Davison Ankney, and D. H. Gordon. 1992. Ecology and
management of postbreeding waterfowl. Pages 128-165 in Batt, B. D., A. D. Afton, M.
G. Anderson, C. Davison Ankney, D. H. Johnson, J. A. Kadlec, and G. L. Krapu, editors.
Ecology and Management of Breeding Waterfowl. University of Minnesota Press,
Minneapolis, Minnesota, USA.
Iverson, S. A., D. Esler, and W. S. Boyd. 2003. Plumage characteristics as an indicator
of age class in the Surf Scoter. Waterbirds 26: 56-61.

41

Jehl, J. R. 1990. Aspects of the molt migration. Pages 102-113 in Gwinner, E., editor.
Bird migration: physiology and ecophysiology. Springer Verlag, Berlin, Germany.
Johnson, S. R., and W. J. Richardson. 1982. Waterbird migration near the Yukon and
Alaskan coast of the Beaufort Sea: moult migration of seaducks in summer. Arctic 35:
291-301.
Jorde, D. G., and R. B. Owen Jr. 1988. Efficiency of nutrient use by American Black
Ducks wintering in Maine. The Journal of Wildlife Management 52: 209-214.
Kaiser, G. W., A. E. Derosher, S. Crawford, M. J. Gill, and I. A. Manley. 1995. A
capture method for marbled murrelets in coastal inlets. Journal of Field Ornithology 66:
321-456.
Kirk, M. K., D. Esler, and W. S. Boyd. 2007. Foraging effort of Surf Scoters (Melanitta
perspicillata) wintering in a spatially and temporally variable prey landscape. Canadian
Journal of Zoology 85: 1207-1215.
Kirk, M., D. Esler, S. A. Iverson, and W. S. Boyd. 2008. Movements of wintering Surf
Scoters: predator responses to different prey landscapes. Oecologia 155: 859-867.
Kozloff, E. N. 1987. Marine invertebrates of the Pacific Northwest. University of
Washington Press, Seattle, Washington, USA.
Krementz, D. G., P. W. Brown, F. P. Kehoe, and C. S. Houston. 1997. Population
dynamics of White-winged Scoters. Journal of Wildlife Management 61: 222-227.

42

Lacroix, D. L., S.Boyd, D. Esler, M. Kirk, T. Lewis, and S. Lipovsky. 2005. Surf
Scoters Melanitta perspicillata aggregate in association with ephemerally abundant
polychaetes. Marine Ornithology 33:61-63.
Larsen, J. K., and M. Guillemette. 2000. Influence of annual variation in food supply on
abundance of wintering Common Eiders Somateria mollissima. Marine Ecology
Progress Series 201: 301-309.
Lewis, T. L., D. Esler, W. S. Boyd, and R. Žydelis. 2005. Nocturnal foraging behavior
of wintering Surf Scoters and White-winged Scoters. The Condor 107: 637-647.
Lewis, T. L., D. Esler, and W. S. Boyd. 2007a. Effects of predation by sea ducks on
clam abundance in soft-bottom intertidal habitats. Marine Ecology Progress Series 329:
131-144.
Lewis, T. L., D. Esler, and W. S. Boyd. 2007b. Foraging behaviors of Surf Scoters and
White-winged Scoters during spawning of pacific herring. The Condor 109:216-222.
Lewis, T. L., D. Esler, and W. S. Boyd. 2008. Foraging behavior of Surf Scoters
(Melanitta perspicillata) and White-winged Scoters (M. fusca) in relation to clam
density: inferring food availability and habitat quality. The Auk 125:149-157.
Lovvorn, J. R., and J. R. Baldwin. 1996. Intertidal and farmland habitats of ducks in the
Puget Sound region: a landscape perspective. Biological Conservation 77:97-114.
Meacham, P. (2010). Aquatic Invasive Species: Varnish Clam. In Washington
Department of Fish and Wildlife [online]. Available: http://wdfw.wa.gov/
ais/species.php?Name=nuttalia_obscurata [2010, Oct10].
43

McCune, B. 1994. Improving community analysis with the Beals smoothing function.
Ecoscience 1: 82-86.
McCune, B. and J. B. Grace. 2002. Analysis of ecological communities. MjM Software
Design, Gleneden Beach, Oregon, USA.
Murphy, M. E. 1996. Energetics of molt. Pages 158-198 in Carey, C., editor. Avian
energetics and nutritional ecology. Chapman and Hall, New York, USA.
Nilsson, L. 1972. Habitat selection, food choice, and feeding habits of diving ducks in
coastal waters of South Sweden during the non-breeding season. Ornis Scandinavica 3:
55-78.
NOAA - Northwest Fisheries Science Center. 1998. Test Results Indicate Varnish
Clams are Safe. West Coast Marine Biotoxins & Harmful Algal Blooms Newsletter
Nysewander, D. R., J. R. Evenson, B. M. Murphie, and T. C. Cyra. 2005. Report of
marine bird and marine mammal component, Puget Sound Ambient Monitoring Program,
for July 1992 to December 1999 period. Washington State Department of Fish and
Wildlife, Olympia, Washington, USA.
Paton, P. W. C., C. J. Ralph, and J. Seay. 1991. A mist net design for capturing marbled
murrelets. North American Bird Bander 16:123-126.
Pearce, J. B. 1966. The biology of the mussel crab, Fabia subquadrata, from the waters
of San Juan Archipelago, Washington. Pacific Science 20: 3-35.

44

Petrie, S. A., and R. W. Knapton. 1999. Rapid increase and subsequent decline of Zebra
and Quagga Mussels in Long Point Bay, Lake Erie: possible influence of waterfowl
predation. Journal of Great Lakes Research 25: 772-782.
Petrie, S., and M. L. Schummer. 2002. Waterfowl responses to Zebra Mussels on the
lower Great Lakes. Birding 34: 346-351.
Poulton, V. K., J. R. Lovvorn, and J. Y. Takekawa. 2002. Clam density and scaup
feeding behavior in San Pablo Bay, California. The Condor 104: 518-527.
Ralph, P. J., D. Tomasko, K. Moore, S. Seddon, and C. M. O. Macinnis-Ng. 2006.
Human impacts on seagrasses: eutrophication, sedimentation, and contamination. Pages
567-593 in Larkum, A. W. D., R. J. Orth, and C. M. Duarte, editors. Seagrasses: biology,
ecology, and conservation. Springer, Netherlands.
Ray, G. 1997. Benthic Assemblages of the Padilla Bay National Estuarine Research
Reserve, Mount Vernon, Washington. Washington State Department of Ecology
Publication No., Padilla Bay National Estuarine Research Reserve Technical Report No.
21, Mount Vernon, Washington, USA.
Richman, S. E., and J. R. Lovvorn. 2003. Effects of clam species dominance on nutrient
and energy acquisition by Spectacled Eiders in the Bering Sea. Marine Ecology Progress
Series 261: 283-297.
Ricketts, E. F., and J. Calvin. 1968. Between Pacific tides. Stanford University Press,
Stanford, California, USA.

45

Riggs, S. R. 2003. Monitoring for invasive crabs in seagrass habitat in the Padilla Bay
National Estuarine Research Reserve, Skagit County, Washington. Washington State
Department of Ecology, Padilla Bay National Estuarine Research Reserve Technical
Report No. 28.
Robert, M., and L. Cloutier. 2001. Summer food habits of Harlequin Ducks in eastern
North America. Wilson Bulletin 113:78-84.
Rodway, M. S., and F. Cooke. 2002. Use of fecal analysis to determine seasonal
changes in the diet of wintering Harlequin Ducks at a herring spawning site. Journal of
Field Ornithology 73:363-371.
Ryder, J. M. 2007. Scoters and other birds feeding on purple varnish clams: some
observations at Spanish Banks, Vancouver, British Columbia. British Columbia Birds
16:24-27.
Salomonsen, F. 1968. The moult migration. Wildfowl 19: 5-24.
Sanger, G. A., and R. D. Jones Jr. 1982. Winter feeding ecology and trophic
relationships of Oldsquaws and White-winged Scoters on Kachemak Bay, Alaska. Pages
20-28 in D. N. Nettleship, G. A. Sanger, and P. F. Springer, editors. Marine birds: their
feeding ecology and commercial fisheries relationships. Proceedings of Pacific Seabird
Group Symposium, Seattle, Washington, USA. Canadian Wildlife Service, Ottawa,
Ontario, Canada.
Savard, J.-P. L. 1988. A summary of current knowledge on the distribution and
abundance of moulting seaducks in the coastal waters of British Columbia. Technical

46

Report Series No. 45. Canadian Wildlife Service, Pacific and Yukon Region, British
Columbia, Canada.
Savard, J.-P. L., D. Bordage, and A. Reed. 1998. Surf Scoter (Melanitta perspicillata).
Pages 1-27 in A. Poole and F. Gill, editors. The Birds of North America, No. 363. The
Birds of North America, Inc., Philadelphia, Pennsylvania, USA.
Savard, J. L., A. Reed, and L. Lesage. 2007. Chronology of breeding and molt migration
in Surf Scoters (Melanitta perspicillata). Waterbirds 30: 223-229.
Shumway, S. E., S. M. Allen, and P. D. Boersma. 2003. Marine birds and harmful algal
blooms: sporadic victims or under-reported events? Harmful Algae 2:1-17.
Sjöberg, K. 1988. The flightless period of free-living male Teal in northern Sweden.
Ibis 130: 164-171.
Storer, R. W., and J. R. Jehl. 1985. Moult patterns and moult migration in the Blacknecked Grebe Podiceps nigricollis. Ornis Scandinavica 16: 253-260.
Stott, R. S., and D. P. Olson. 1973. Food-habitat relationship of sea ducks on the New
Hampshire coastline. Ecology 54: 996-1007.
Swanson, G. A., and J. C. Bartonek. 1970. Bias associated with food analysis in
gizzards of Blue-winged Teal. The Journal of Wildlife Management 34: 739-746.
Thayer, G. W., W. E. Schaff, J. W. Angelovic, and M. W. LaCroix. 1973. Caloric
measurements of some estuarine organisms. Fisheries Bulletin 71: 289-296.

47

Thompson, J. E., and R. D. Drobney. 1997. Diet and nutrition of male Canvasbacks
during postreproductive molts. Journal of Wildlife Management 61: 426-434.
Vermeer, K. 1981. Food and populations of Surf Scoters in British Columbia. Wildfowl
32: 107-116.
Vermeer, K. 1983. Diet of the Harlequin Duck in the Strait of Georgia, British
Columbia. The Murrelet 64:54-57.
Vermeer, K., and N. Bourne. 1982. The White-winged Scoter diet in British Columbia
waters: resource partitioning with other scoters. Pages 30-38 in D. N. Nettleship, G. A.
Sanger, and P. F. Springer, editors. Marine birds: their feeding ecology and commercial
fisheries relationships Proceedings of Pacific Seabird Group Symposium, Seattle,
Washington, USA. Canadian Wildlife Service, Ottawa, Ontario, Canada.
Vermeer, K., and C. D. Levings. 1977. Populations, biomass, and food habits of ducks
on the Fraser Delta intertidal area, British Columbia. Wildfowl 28: 49-60.
Wicksten, M. K. (2009). Decapod Crustacea of the Californian and Oregonian
Zoogeographic Provinces. In UC San Diego: Scripps Institution of Oceanography
Library [online]. Available: http://escholarship.org/uc/item/7sk9t2dz [2010, October 10].
Wormington, A., and J. H. Leach. 1992. Concentrations of migrant diving ducks at
Point Pelee National Park, Ontario, in response to invasion of Zebra Mussels, Dreissena
polymorpha. The Canadian Field-Naturalist 106: 376-380.
Zwarts, L., and J. H. Wanink. 1993. How the food supply harvestable by waders in the
Wadden Sea depends on the variation in energy density, body weight, biomass, burying
48

depth, and behavior of tidal-flat invertebrates. Netherlands Journal of Sea Research 31:
441-476.
Žydelis R., D. Esler, W. S. Boyd, D. L. Lacroix, and M. Kirk. 2006. Habitat use by
wintering Surf and White-winged Scoters: effects of environmental attributes and
shellfish aquaculture. The Journal of Wildlife Management 70: 1754-1762.
Žydelis R., D. Esler, M. Kirk, and W. S. Boyd. 2009. Effects of off-bottom shellfish
aquaculture on winter habitat use by molluscivorous sea ducks. Aquatic Conservation:
Marine and Freshwater Ecosystems 19: 34-42.

49

Appendix A:
Nonmetric Multidimensional Scaling (NMS) Ordination Technique
Procedure is based on an n x n distance matrix, calculated from the n x p original species
data matrix, in which n is the number of rows and p is the number of columns in the
species data matrix. NMS describes the best combination of n entities (or sample units)
on k dimensions in sample space, while minimizing stress. Stress is determined by the
difference in the original p dimensional space and the distance described by the k
dimensional ordination space. NMS allows for multiple distance measures, while rankordering of species space distances and ordination space distances linearizes their
relationship. McCune and Grace (2002) provide the following variable definitions:
X = coordinates of n sample units or entities in a k-dimensional space. The element xil is
the coordinate of sample unit i on dimension (axis) l.
Δ = matrix of dissimilarity coefficients from the original data. Any distance measure can
be used. We recommend using the quantitative version of the Sørensen coefficient for
most ecological community data.
δij = elements of Δ.
D = matrix of interpoint distances in the k-space. This matrix of n x n Euclidean
distances is calculated from X.
dij = elements of D.
Raw stress:
S* =

dij-ḓij)²

S* measures the departure from monotonicity. If S* = 0, the relationship is perfectly
monotonic.
Stress is then standardized:
S = S*/

ij²

Final stress is rescaled:
SR = 100 S
(See McCune and Grace (2002) for additional detailed methodology).

50

Appendix B:
Multi-response Permutation Procedure Technique
To determine whether or not there is a difference between two or more test groupings,
McCune and Grace (2002) outline the following procedure:
1. Calculate distance matrix D, utilizing distance measures such as Euclidean, or
Sørensen distance measures.
2. Calculate the average distance xi within each group i.
3. Calculate delta
δ=

ixi

for g groups, where C is a weight that depends on the number of items in the groups
(normally Ci = ni / N, where ni is the number of items in group i and N is the total number
of items). Note that all ni ≥ 2. For a given mean overall distance, smaller values of δ
indicate tighter clustering within groups.
4. Determine probability of a δ this small or smaller. A reasonable method that is also
applicable to medium or large data sets is to approximate the distribution of δ from a
continuous distribution (Pearson type III). The Pearson type III distribution incorporates
three parameters, the mean m, standard deviation s, and gamma g (skewness of δ under
the null hypothesis).
The test statistic, T is
T = (δ – mδ) / sδ
where mδ and sδ are the mean and standard deviation of δ under the null hypothesis. Or,
in other words, the test statistic is the difference between the observed and expected
deltas divided by the square root of the variance in delta.
T = (observed δ – expected δ) / s.dev of expected δ
The test statistic, T, describes the separation between the groups. The more negative is T,
the stronger the separation. The p-value associated with T is determined by numerical
integration of the Pearson type III distribution.
5. A description of the effect size that is independent of the sample size is provided by
the chance-corrected within-group agreement (A).
A = 1 – δ / mδ = 1 – observed δ / expected δ

51

Appendix C:
Beals smoothing transformation procedure
This transformation provides a quantitative value that indicates the “favorability of a
given sample for species i, based on the whole data set, using the proportions of joint
occurrences between the species that do occur in the sample and species i”(McCune and
Grace 2002).
bij = (1 / Si)

Mjk / Nk)

Si = number of species in sample unit i
Mjk = number of sample units with both species j and k
Nk = number of sample units with species k
See McCune and Grace (2002) for an example of a data matrix before and after the Beals
smoothing transformation.

52

Appendix D: Sum of total samples with prey items identified from Surf and White-winged Scoter fecal samples at four molting
locations in Northern Puget Sound, and Southern Strait of Georgia, with total scoters in parenthesis (F = female, M = male).

Bivalvia
Mytilidae (Mytilus trossulus)
Cardiidae
Macoma balthica
Macoma sp.
Psammobiidae (Nuttallia obscurata)
Veneridae
Unidentified Bivalvia
Gastropoda
Pyramidellidae
Potamididae (Batillaria atramentaria)
Unidentified Gastropoda
Crustacea
Idoteidae
Cancridae
Pinnotheridae
Caprellidea (Caprella sp.)
Unidentified Crustacean
Amphipoda
Echinodermata (Ophiuroidea)
Polychaeta
Plant material

Boundary Bay
SUSC
WWSC
(9F, 3M) (1F, 13M)
12
13
3
1
11
2
0
0
1
0
1
8
7
2
2
1
7
1
0
0
0
1
7
0
5
1
3
0
1
1
0
0
0
0
1
0
2
0
0
0
6
0
2
2

Forbes Point
SUSC
WWSC
(7F, 8M)
(2F, 3M)
15
5
15
5
0
2
0
0
0
0
3
1
0
0
0
0
1
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0

Fraser River
SUSC
WWSC
(8F, 10M) (9F, 19M)
18
28
3
1
0
0
2
0
0
0
18
28
3
1
0
0
0
2
0
0
0
0
0
2
1
2
0
2
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
2

Padilla Bay
SUSC
(48F, 3M)
34
13
2
0
1
18
3
7
24
8
1
18
48
21
2
34
2
4
0
2
6
6

53

Appendix E: Wide-angle view above dissecting microscope of several Veneridae clams,
with likely Transennella species on the right, and possible Manila clam on the left
(several small gastropod species also pictured). Clams pictured < 3.0 mm in width, see
calipers.

54

Item sets

The Evergreen State College Masters Theses: Environmental Studies

Media

Tschaekofske_HMESThesis2010.pdf