Content uploaded by Josh M. London
Author content
All content in this area was uploaded by Josh M. London on Jan 24, 2014
Content may be subject to copyright.
Harbor Seals in Hood Canal: Predators and Prey
Joshua Michael London
A dissertation
submitted in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy
University of Washington
2006
Program Authorized to Offer Degree:
School of Aquatic and Fishery Sciences
University of Washington
Graduate School
This is to certify that I have examined this copy of a doctoral dissertation by
Joshua Michael London
and have found that it is complete and satisfactory in all respects,
and that any and all revisions required by the final
examining committee have been made
Chair of the Supervisory Committee:
Glenn R. VanBlaricom
Reading Committee:
Glenn R. VanBlaricom
John R. Skalski
Steven J. Jeffries
Date: ____________________________________
In presenting this dissertation in partial fulfillment of the requirements for the doctoral
degree at the University of Washington, I agree that the Library shall make its copies
freely available for inspection. I further agree that extensive copying of the dissertation
is allowable only for scholarly purposes, consistent with “fair use” as prescribed in the
U.S. Copyright Law. Requests for copying or reproduction of this dissertation may be
referred to Proquest Information and Learning, 300 North Zeeb Road, Ann Arbor, MI
48106-1346, to whom the author has granted “the right to reproduce and sell (a) copies
of the manuscript in microform and/or (b) printed copies of the manuscript made from
microform.”
Signature___________________________
Date_______________________________
University of Washington
Abstract
Harbor Seals in Hood Canal: Predators and Prey
Joshua Michael London
Chair of the Supervisory Committee:
Professor Glenn R. VanBlaricom
School of Aquatic and Fishery Sciences
The foraging ecology and population dynamics of harbor seals (
Phoca vitulina
richardsi
) were studied in Hood Canal, Washington from 1998 to 2005. Initial work was
conducted in response to concerns over the potential impact seals may have on
recovering populations of summer chum salmon. Direct observation of harbor seals
consuming salmon within the inter-tidal regions of four rivers in Hood Canal were
conducted from 1998-2001 and 2003. Seals were observed feeding on chinook, coho,
pink, summer chum and fall chum salmon. Estimates of summer chum consumption by
seals at each of the observation sites averaged 8.0% of returning adults across all sites
and all years. The maximum percentage of returning chum consumed was 27.7% and the
lowest was 0.84%. The number of seals observed foraging in the river for salmon
averaged from two to seven seals. Summer chum populations in the study streams have
increased over the time of the study to near historical highs. Because of this increase, the
levels of predation observed are not believed to significantly impact the recovery of
summer chum in Hood Canal. A protocol for extraction of DNA and identification of
seal sex from scats was developed to examine differential diets between male and female
harbor seals. Scats from both sexes contained similar levels of Pacific hake, but male
scats contained more salmon and female scats contained more Pacific herring. In 2003
and 2005, mammal-eating killer whales foraged exclusively within Hood Canal for 59
and 172 days respectively. Bio-energetic models and boat based observations were used
to estimate harbor seal consumption by killer whales and, in both years, the predicted
consumption was approximately 950 seals. However, aerial surveys conducted following
the two foraging events have not detected a significant decline in the harbor seal
population.
i
TABLE OF CONTENTS
List of Figures....................................................................................................................iii
List of Tables.....................................................................................................................iv
Introduction..........................................................................................................................1
Ballard Locks................................................................................................................ 2
Pinniped Food Habits ................................................................................................... 4
NMFS Investigation and the West Coast Pinniped Study ........................................... 5
Research Presented....................................................................................................... 6
Hood Canal - Harbor Seals and Summer Chum .......................................................... 6
Killer Whales in Hood Canal ....................................................................................... 7
Scat Genetics ................................................................................................................ 8
Chapter 1. Harbor Seals and Salmon in Hood Canal: Estimates of Predation by
Harbor Seals on Threatened Summer Chum Salmon ...................................................... 10
Introduction....................................................................................................................... 10
Methods ............................................................................................................................ 14
Surface Observations..................................................................................................14
Nighttime Predation.................................................................................................... 16
Calculation of Predation Estimates ............................................................................ 17
Allocation of Salmon Predations to Species ..............................................................17
Estimates of Summer Chum Abundance.................................................................... 19
Results............................................................................................................................... 20
Seal Behavior.............................................................................................................. 20
Predation Estimates .................................................................................................... 21
Variance Calculation and Confidence Intervals......................................................... 27
Discussion......................................................................................................................... 28
Assessing Population Impacts of Seal Predation ....................................................... 28
Chapter 2. Impacts of Two Extended Foraging Events by Mammal-Eating Killer
Whales on the Population of Harbor Seals in Hood Canal, Washington......................... 33
Introduction....................................................................................................................... 33
Methods ............................................................................................................................ 35
Behavioral Observations............................................................................................. 35
ii
Generalized Linear Model.......................................................................................... 38
Bio-Energetic Monte Carlo Simulation...................................................................... 39
Results............................................................................................................................... 41
Behavioral Observations............................................................................................. 41
Generalized Linear Model.......................................................................................... 44
Bio-energetic Monte Carlo Simulation ......................................................................46
Discussion......................................................................................................................... 47
Behavioral Observations............................................................................................. 47
Bio-Energetic Model .................................................................................................. 49
GLM Analysis of Harbor Seal Counts....................................................................... 50
Chapter 3. Pinniped Scat Genetics: Identification of Sex and Species from Feces
Collected for Food Habits Studies.................................................................................... 51
Introduction....................................................................................................................... 51
Methods ............................................................................................................................ 53
Sample Collection and DNA Extraction .................................................................... 53
Sex Specific Markers.................................................................................................. 54
Species Identification Markers ...................................................................................56
PCR Reactions............................................................................................................58
Sex and Species Determination ..................................................................................58
Sexual Differences in Diet.......................................................................................... 59
Results............................................................................................................................... 59
Discussion......................................................................................................................... 62
List of References............................................................................................................. 66
Appendix A: Prey Species Present in Scats Classified by Sex........................................ 72
iii
LIST OF FIGURES
Figure Number Page
Figure 1.1 Map of the Puget Sound and Hood Canal....................................................... 11
Figure 1.2 Summer chum salmon escapements to Hood Canal (1974-2002).................. 12
Figure 1.3 Map of Nort and South Hood Canal............................................................... 14
Figure 1.4 Tidal relationship of observed salmon predations in 1998 and 1999............. 21
Figure 1.5 Maximum number of foragers vs. number of observed salmon predations... 24
Figure 1.6 Estimated percentage of adult summer chum consumed by harbor seals...... 27
Figure 1.7 CV values for the estimates of summer chum consumed by harbor seals .....28
Figure 2.1 Box plot of observation times during 2005 .................................................... 42
Figure 2.2 Map of North and South Hood Canal showing whale tracklines ...................43
Figure 2.4 Frequency distribution of model outputs from the bio-energetic models ...... 47
Figure 3.1 PCR primers for the newly designed PvSRY2004 region.............................. 56
Figure 3.2 PCR primers for the newly designed PinID region........................................ 58
Figure 3.3 Example electropherogram for a male and female
P. vitulina
scat................ 60
Figure 3.4 Example electropherogram for male scats from three pinnipeds. ..................61
iv
LIST OF TABLES
Table 1.1 Maximum number of harbor seals actively foraging for salmon..................... 20
Table 1.2 Comparisons between paired day-time and night-time observations ..............23
Table 1.3 Estimates of predation by harbor seals on summer chum in Hood Canal ....... 25
Table 2.1
Parameter values and distributions used in the Monte Carlo simulation ......... 40
Table 2.2
AIC values from GLMs of harbor seal population response to killer whales.. 45
Table 2.3 Calorimetric values from recovered harbor seal carcasses.............................. 46
v
ACKNOWLEDGEMENTS
Funding Sources
Washington Department of Fish and Wildlife, Pacific States Marine Fisheries
Commission, National Oceanic and Atmospheric Administration (NOAA) National
Marine Fisheries Service, NOAA Alaska Fisheries Science Center National Marine
Mammal Lab, NOAA Northwest Fisheries Science Center, United States Marine
Mammal Commission, Washington Cooperative Fish and Wildlife Research Unit,
University of Washington School of Aquatic and Fishery Sciences, North Pacific
Universities Marine Mammal Research Consortium
Field Observations and Lab Assistance
Allison Agness, Kirstin Brennan, Robin Baird, Katy Calif, Kimber Cochrane, Volker
Deeke, Judy Dicksion, John Durban, James Grassley, Laura Hoberecht, Steve Jeffries,
Bernie Kauffman, Irene Kinan, Dyanna Lambourn, Kristin Laidre, Monique Lance,
Janell Majewski, Walter Major, Bryan Murphie, Orca Network Contributors, Anthony
Orr, Kim Parsons, Tammy Schmidt, Shannon Sewalt, Greg Shirato, Heather Smith,
Marissa Stratton, Robert Suydam, Glenn VanBlaricom, Paul Wade, Michelle Wainstein,
Alex Zerbini
Colleagues, Friends and Family
Sharon Aguilar, Carlos Alvarez, Paul Bentzen, Peter Boveng, Amanda Bradford, Robin
Brown, John Calambokidis, Michael Cameron, Mike Canino, Robert Delong, Marcus
Duke, Martin and Laila Grassley, Chris Grue, Jim Harvey, Donna Hauser, Laura
Hoberecht, Harriet Huber, Steve Jeffries, Pam Jensen, Jeff Laake, Kristin Laidre,
Monique Lance, Michelle Lander, Jeff and Darnelle London, Nicholas London, Shanna
London, Walter and Jennifer Major, Shannon McCluskey, Bryan Murphie, Anthony Orr,
Rolf Ream, Susan Reimer, Joe Scordino, Todd Seamons, Heather Smith, Robert
Suydam, Glenn VanBlaricom, Susan Wang, Alex Zerbini,
i
DEDICATION
Dwight Kertzman
My High School Biology Teacher, 1990 – 1993
Booker T. Washington High School
Tulsa, Oklahoma
1
Introduction
The foraging ecology of pinnipeds has been the subject of scientific investigations for
many years because pinnipeds are often perceived as being in competition with human
fisheries. Pinniped-fishery interactions are often classified as either operational or
biological (Baraff and Loughlin 2000). Operational interactions refer to activities that
result in a direct impact. Entanglement in fishing gear and other forms of incidental and
directed take would be considered ‘operational’ interactions. Pinniped behavior can also
result in a direct impact on a fishery. Seals and sea lions are known to remove fish from
set-nets, commercial long-lines and recreational fishermen. Grey seals (
Halichoerus
grypus
) in Britain have been implicated in a number of detrimental fishery interactions
ranging from predation on free swimming and net-pen salmon to more indirect effects of
driving fish away from nets and increasing the presence of cod worm in the Atlantic cod
(Furness 2002; Harwood and Greenwood 1985). As pinniped populations have increased
along the West coast of the United States, recreational fishermen commonly report loss
of catch due to seals and sea lions removing fish from their lines (NMFS 1999). Charter
boat companies in Washington, Oregon and California are often forced to change locales
during a trip because of the presence of seals and sea lions (NMFS 1999).
Biological interactions can be categorized as either exploitative competition or
interactive competition (Baraff and Loughlin 2000). Exploitative competition occurs
when pinnipeds are in direct competition with fisheries for the same prey. Seal and sea
lion populations in the northeast Pacific rely on ground-fish, herring, salmon and squid
as major components of their diet (Lowry and Frost 1985). These same species are also
large components of commercial fishery operations and pinnipeds are often perceived as
being in direct competition with these fisheries. Interactive competition is less direct and
examples include pinnipeds abandoning a foraging area due to disturbance, reduced
foraging efficiency or disruption of foraging patterns (Baraff and Loughlin 2000).
However, pinniped-fishery interactions must also be examined with some perspective.
During the late nineteenth and early twentieth centuries, many pinniped stocks
throughout the world were hunted to a small fraction of their former population levels
2
(Beddington et al. 1985) for both the fur trade and to reduce fishery interactions. From
early in the twentieth century until the passage of the federal Marine Mammal Protection
Act of 1972 (MMPA), all West coast states, including Washington, had an active
pinniped control unit and bounty program for the removal of pinniped species.
Significant conservation efforts in the last century and passage of laws such as the
MMPA have allowed many pinniped populations to rebound; however many remain
critically low or are declining (VanBlaricom et al. 2001).
In recent years, the west coast of the United States has seen the emergence of a new
category of pinniped-fishery interaction. Since the passage of the MMPA, populations of
California sea lions (
Zalophus californianus
) and Pacific harbor seals (
Phoca vitulina
richardsi
) have experienced dramatic increases throughout the West Coast, and may be
at their highest levels in several centuries (NMFS 1999). The increase in pinniped
populations has coincided with dramatic decreases of many marine and anadromous fish
populations (NMFS 1997b). A number of these populations have declined to a point
where they have been listed, or are under consideration for listing, as endangered or
threatened under the federal Endangered Species Act of 1973 as amended (ESA). There
has been a growing concern throughout the West Coast that pinnipeds, while not likely
the cause of any decline in salmonids, have the potential to affect the recovery of many
threatened and endangered salmonid stocks (NMFS 1999).
Ballard Locks
The importance of understanding the potential impact pinnipeds could have on declining
salmonid populations came to the forefront when California sea lions were consistently
observed foraging on returning winter steelhead at the Ballard Locks in Seattle,
Washington. The situation in Ballard was ongoing from the mid-1980s through 1995 and
was well documented (NMFS 1995; 1999). At the peak of predation activity in this
location, close to sixty percent of the returning winter steelhead run was being consumed
by sea lions. As the number of returning steelhead reached critically low levels, various
non-lethal mitigating actions were employed in an attempt to reduce the level of
predation on steelhead. By 1995, the non-lethal actions were determined to be
3
ineffective and, after a number of hearings and detailed investigations, the National
Marine Fisheries Service (NMFS) issued a permit to the Washington Department of Fish
and Wildlife (WDFW) for the lethal removal of individual pinnipeds under Section 120
of the MMPA.
Before lethal removal was implemented, however, arrangements were made to capture
and transport the predatory sea lions to Sea World in Orlando, Florida for display and
captive breeding. Since the individual sea lions were removed from the area, little or no
predation has been observed. However, while there was an initial increase in numbers of
returning steelhead in years directly following removal of predatory sea lions, the annual
number of returning spawners remains critically low (WDFW unpublished data).
The situation at Ballard is not unique. Predation by California sea lions on salmonids has
also been observed at the Willamette Falls area near Portland, OR (NMFS 1997a), below
and in the fish-way at Bonneville Dam on the Columbia River and below Merwin Dam
on the Lewis River (pers. comm. Steve Jeffries, Washington Department of Fish and
Wildlife, Tacoma, WA). Additionally, harbor seals have been observed feeding on out-
migrating salmon smolts in the Punteldge River on Vancouver Island in British
Columbia (Yurk and Trites 2000) and adult salmon in coastal rivers of Oregon,
Washington and California (NMFS 1997b).
Of the numerous lessons and knowledge gained from the situation at Ballard, three stand
out. First, managers and researchers need to take a proactive approach by initiating
studies to understand the impact of pinnipeds on a declining population before actually
reaching the level of threatened or endangered. Section 120 of the MMPA requires that a
“significant negative impact” to the threatened or endangered population be
demonstrated in a quantitative manner. This requires several years of research, and if not
done prior to the population reaching a critically low level, valuable time may be spent
investigating the impact instead of implementing management action that could lead to
recovery (NMFS 1997b).
4
The second lesson is, while there are certainly unique aspects to the Lake Washington
Ship Canal, and other locations that allow unprecedented levels of pinniped predation on
returning salmonids, it is reasonable to expect analogous situations to exist in a more
natural environment. Pinniped haul-outs throughout the west coast often occur at or near
the mouths of rivers that support a number of declining salmonid populations (NMFS
1997b). At any such estuarine or nearshore area where fish passage is constrained by
natural or artificial barriers, there are possibilities for predation to occur (NMFS 1999).
The key question, however, is not whether predation is occurring, but rather if predation
is negatively affecting the ability of a particular salmonid population to recover from low
numbers. Additionally, any negative affect pinniped predation may be having on the
recovery of a particular salmonid stock should be considered within the context of a
myriad of other factors such as habitat degradation, harvest, climate change, pollution
and ocean conditions. Predator prey relationships between salmon and pinnipeds, such as
harbor seals, have been evolving for thousands of years and the complexity of that
relationship must be considered when looking to situations like Ballard as examples.
California sea lions, the salmonid populations, and any artificial barriers and human
modifications there now did not exist in the same location until relatively recent.
Another theme that emerged was the methods used to evaluate the food habits of
pinnipeds were inadequate to effectively estimate and predict the impact such predation
may or may not have on a particular salmonid population (NMFS 1997b). Scat
collection and analysis would have to be combined with other techniques and
technologies such as direct surface observation for predation events, genetic analysis,
and comprehensive population modeling approaches.
Pinniped Food Habits
Research on the food habits of California sea lions and Pacific harbor seals in the past
has shown that they are opportunistic consumers, with the majority of their diets
consisting of seasonally and locally abundant prey. For sea lions and harbor seals in the
greater Puget Sound and Straight of Georgia, this translates into a diet, while diverse,
made up mostly of Pacific hake (
Merluccius productus
) and Pacific herring (
Clupea
5
harengus pallasi
) (Calambokidis et al. 1989; NMFS 1995; 1997b; Olesiuk et al. 1990).
In some locations, significant pinniped predation has been reported on returning adult
salmonids (NMFS 1995; 1997a; b) or out-migrating salmon smolts (Yurk and Trites
2000). While salmonids have been found in the diet of local pinnipeds, the vast majority
of studies were not designed to address impacts of pinniped predation on specific
salmonid populations. Previous studies were mostly conducted on an opportunistic basis,
not necessarily within a period of high salmonid abundance, and focused on the
collection of scat and analysis of prey remains. Understanding the role of salmonids in
the foraging ecology of seals and sea lions is especially problematic because, until
recently, only otoliths (fish ear bones) were used to identify prey items. Otolith bones
from salmonids are more fragile than other bones and are not often recovered in
pinniped scats. In fact, recent analysis of the frequency of occurrence (FO) of salmonids
in the diet of harbor seals in Hood Canal has shown an approximate five-fold increase in
the percentage of scats containing salmonids when all structures are used compared to
only otoliths (Lance unpublished). Other factors, involving gut retention, travel time
between haul-outs and foraging locations, and the potential that scats collected are not a
representative sample of a population, may further limit the ability to interpret the role of
salmonids in the diet of pinnipeds with scat analysis alone.
NMFS Investigation and the West Coast Pinniped Study
In February of 1997, NMFS completed a review of scientific information on impacts of
California sea lions and Pacific harbor seals on West Coast salmonids (NMFS 1997b).
This report discussed themes previously mentioned as a result of the Ballard situation
and identified a number of locations where there was a potential for pinnipeds to impact
recovery of declining salmonids. This led to an initial allocation of resources to the
Oregon Department of Fish and Wildlife (ODFW) to begin evaluating the use of direct
surface observations as a way of estimating predation rates of pinnipeds on salmonids.
Studies were expanded to include researchers from WDFW, California Department of
Fish and Game (CDFG) and NMFS in 1998. All participating researchers coordinated
efforts through the Pacific States Marine Fishery Commission and adapted a similar
approach and methodology for each specific site and question of interest.
6
Research Presented
Research presented in this dissertation reflects much of the work conducted
collaboratively with the Washington Department of Fish and Wildlife between 1998 and
2005. Results are presented in three chapters:
(1) Harbor Seals and Salmon in Hood Canal: Estimates of Predation by Harbor
Seals on Threatened Summer Chum Salmon,
(2) The Impact of Two Extended Transient Killer Whale Foraging Events on the
Harbor Seal Population in Hood Canal.
(3) The Use of Genetic Scat Analysis in Pinnipeds for Determination of Sex and
Species Specific Food Habits,
Hood Canal - Harbor Seals and Summer Chum
Hood Canal is a fjord-like body of water that lies just east of the Olympic Peninsula and
makes up the western most portion of Puget Sound in Washington State. Five major
rivers (Quilcene, Dosewallips, Duckabush, Hamma Hamma and Skokomish rivers)
originate from headwaters in the Olympic Mountains and flow into Hood Canal. Each
river supports runs of various salmonid species including chinook (
Oncorhynchus
tshawytscha
), coho (
O. kisutch
), chum (
O. keta
) and pink (
O. gorbuscha
). Steelhead (
O.
mykiss
) and sea run cutthroat (
O. clarkii
) are present as well. In recent years, many
salmonid runs have declined sharply, with several runs (chinook, summer chum and
Dosewallips pinks) listed in the 1992 Salmon and Steelhead Stock Inventory (WDF et al.
1993) as critical or depressed.
Summer chum salmon in Hood Canal were listed as ‘Threatened’ under the ESA in 1999
(WDFW et al. 2000). Escapement and abundance estimates were precipitously low in
the late 1980s and 1990s and were at a fraction of historic levels (WDFW et al. 2000).
Summer chum salmon return to Hood Canal streams in August and September and are
genetically distinct from Fall chum runs that return in late October and November
(WDFW et al. 2003). The reasons for the decline in abundance are not fully understood,
7
but are likely related to by-catch, habitat loss and reduced ocean productivity (WDFW et
al. 2000).
Of the six extant native summer chum stocks within Hood Canal, four (Quilcene,
Dosewallips, Duckabush and Hamma Hamma) return to rivers that have harbor seal
haul-outs associated with the lower tidal areas. Harbor seal populations in Hood Canal
(approx. 1000 animals) and the rest of Washington state are considered abundant and
healthy (Jeffries et al. 2003). Each of the five main haul-out sites in Hood Canal can
range from approximately 50-250 seals during August and September (Calambokidis et
al. 1990). This temporal and spatial overlap has lead to concern over the impact seal
predation might have on the conservation and recovery efforts of summer chum in Hood
Canal. Previous studies in Hood Canal have examined harbor seal diet by identification
of otoliths found in scat (Calambokidis et al. 1978; Calambokidis and McLaughlin
1987). Pacific hake composed more than eighty percent of harbor seal diet based on scat
collected at the Skokomish, Duckabush, and Dosewallips rivers and Quilcene Bay. In
these studies, salmon was not found to be a significant portion in the diet of harbor seals
in Hood Canal, however; the use of only otoliths for identification of prey is known to
under-represent prey species with more fragile otoliths like salmonids.
Most studies of pinniped foraging behavior are limited because their foraging activity
occurs at depth and is unobservable by researchers. The typical solution is to rely on
archival tags to provide information on diving behavior and inferences to feeding
activities. The small, relatively shallow tidal streams in Hood Canal that were the focus
of our research efforts allowed an unprecedented view of this unique seal foraging
behavior. Information presented in this chapter will focus on seal behavior and
quantitative estimates of seal predation on summer chum based on surface observations
conducted in 1998, 1999, 2000, 2001 and 2003.
Killer Whales in Hood Canal
Prior to 2003, killer whales were considered a rare occurrence in Hood Canal. Acoustic
recordings from the US Navy submarine base at Bangor and reports from long time
residents of Hood Canal, suggest both resident and transient type killer whales have been
8
present in Hood Canal. However, the frequency of reports is extremely low.
Additionally, when whales have been observed in Hood Canal it has been for no more
than one or two days.
In January of 2003, 11 transient-type, mammal-eating killer whales arrived in Hood
Canal and remained exclusively within the canal for 59 days. The extended stay of such
a large group of transients was considered atypical. Anecdotal observations suggest
these whales were feeding on harbor seals and bio-energetic estimates suggested more
than half of the seal population should have been removed. Subsequent aerial surveys in
2003 and 2004 have not shown a significant decline in seal abundance, and, in January
of 2005, six different transient type killer whales arrived in Hood Canal and stayed for
172 days. Bio-energetic and observation estimates suggest a similar level of removal
occurred during the 2005 event, but harbor seal surveys in 2005 also do not exhibit a
sharp decline. This chapter will review details of these two extended foraging events,
parameters and predictions of the bio-energetic model and an evaluation of harbor seal
aerial surveys to determine the population impact of these killer whale predation events.
Scat Genetics
In recent years, analysis of prey remains found in scats has become the method of choice
for investigation of pinniped diets. In most locations, scats can be collected in large
numbers with relative ease and minimal disturbance. Scat analysis, however, does have
limitations. Biases associated with recovery and identification of otoliths and bones from
some prey species prevent reliable use of scats for more than generalized
characterization of diet (Cottrell et al. 1996; Harvey 1989; Lance et al. 2001).
Investigation of more detailed aspects of pinniped foraging, such as sexual variation in
diet, would require more intrusive actions such as enemas, lavaging, or examination of
stomachs from harvested individuals. Additionally, collections of scats from haul-outs
shared by more than one pinniped species are often confounded because scats cannot be
separated visually or based on collection location. Genetic analysis of scat material is an
alternative non-invasive technique that would allow individual scats to be classified
9
based on sex, species or individual identification of the source animal (Kohn and Wayne
1997).
Genetic scat analysis has been employed in a number of terrestrial mammalian studies
(Ernest et al. 2000; Farrell et al. 2000; Kohn et al. 1999; Morin et al. 2001; Wasser et al.
1997), yet its application to pinniped scat analysis has been limited (Reed et al. 1997).
This chapter will focus on development of an efficient and reliable protocol for
extraction of pinniped DNA from scats. While extracted DNA can provide the basis for a
variety of genetic investigations (Kohn and Wayne 1997), here the focus will be on
amplification of sex specific and species specific markers and their potential use for
examination of variation in diet between sexes and species. This methodology, when
combined with standard protocols for identification of prey remains from scat (Lance et
al. 2001), can provide researchers with new insights into the foraging ecology of
pinnipeds.
The chapters presented here are written as independent manuscripts that will be
submitted to peer-reviewed journals for publication. Therefore, there may be overlap and
repetition of themes covered in this introduction and other chapters.
10
Chapter 1. Harbor Seals and Salmon in Hood Canal: Estimates of
Predation by Harbor Seals on Threatened Summer Chum Salmon
Introduction
Since passage of the Marine Mammal Protection Act (MMPA) in 1972, many
populations of pinnipeds along the West Coast of the United States have rebounded to
historic highs. Harbor seals in Washington (Jeffries et al. 2003), Oregon (Brown et al.
2005) and California (Carretta et al. 2005) are all at or near estimated carrying capacity.
California sea lions throughout the West Coast are growing exponentially (Carretta et al.
2005) and expanding their range. The increasing numbers are likely a testament to the
adaptability and productivity of these pinniped species and protection afforded them
under the MMPA. From early in the twentieth century until the passage of the MMPA,
Washington and other states maintained active pinniped control and bounty programs for
the removal of pinniped species. Pinniped control programs were largely based on the
view that seals and sea lions were direct competitors to commercial and recreational
fisheries and removal of predators would benefit fisheries. Now that pinniped
populations have rebounded, they are, once again, a focus of concern among fisheries
managers. This time, however, the concern is over the impact increasing seal and sea
lion populations may be having on a major conservation effort to recover declining
populations of salmon.
Coincident to increases in pinniped numbers over the last 30 years has been a significant
decline in the number of returning salmon in Washington, Oregon and California.
Several of the declining populations have been listed as ‘Threatened’ or ‘Endangered’
under the Endangered Species Act (ESA). The extent to which pinnipeds are a hindrance
to recovery of these declining salmon populations is not known. There have been a few
case studies where pinniped predation on salmon has accounted for losses of a large
fraction of returning adults (Ballard, Willamette) (NMFS 1995; 1997a) or out-migrating
smolts (Puntledge River, British Columbia) (Yurk and Trites 2000). However, the
majority of pinniped diet studies in locales where salmon are present have shown salmon
to be a minor component of the year-round diet of seals and sea lions (NMFS 1997b).
11
Pinnipeds are likely not the reason for the widespread decline in salmon populations, but
could be responsible for inhibiting recovery in localized small populations.
Figure 1.1 Map of the Puget Sound and Hood Canal region in Washington State, USA.
Summer chum salmon in Hood Canal (Figure 1.1) were listed as ‘Threatened’ under the
ESA in 2001. Escapement and abundance estimates were precipitously low in the late
1980s and 1990s (WDFW et al. 2003) and were at a fraction of historic levels (Figure
12
1.2). Summer chum salmon return to Hood Canal streams in August and September and
are genetically distinct from Fall chum runs that return in late October and November
(WDFW et al. 2003). The reasons for the decline in abundance are not fully understood,
but are likely related to by-catch, habitat loss and reduced ocean productivity (WDFW et
al. 2000).
Figure 1.2 Summer chum salmon escapements to Hood Canal and Strait of Juan de Fuca streams, 1974
through 2002 (WDFW 2003).
Of six extant native summer chum stocks within Hood Canal, four (Quilcene,
Dosewallips, Duckabush and Hamma Hamma) return to rivers that have harbor seal
haul-outs associated with their lower tidal areas. The total population of seals in Hood
Canal is approximately 1,000 (Jeffries et al. 2003) and counts at each of these four haul-
out sites can range from approximately 50-250 seals during August and September
(Calambokidis et al. 1991). This temporal and spatial overlap has lead to concern over
13
impacts seal predation might have on conservation and recovery efforts of summer chum
in Hood Canal.
In most cases, the use of direct observation of harbor seal foraging behavior is limited
because the majority of foraging events take place several meters underwater. However,
in those situations where seals are taking advantage of high prey concentrations in a
limited area, direct observation can provide significant insight into consumption rates
and foraging behavior. Harbor seal predation on smolt and adult salmonids is one such
scenario and observation of surface predation events is the basis for estimates of adult
salmonid consumption at four river systems in Hood Canal, Washington between 1998
and 2003.
Harbor seal surface predation events are denoted as those times when a seal brings a
captured salmon to the surface for consumption. This is due, in large part, to the physical
size of the salmonid prey. Many returning salmon are close to one-third the body length
of an average harbor seal and significantly larger than other prey items found in the diet
of Hood Canal seals. Captured salmon are often brought to the surface for killing and
subsequent consumption, however; consumption underwater is possible and likely more
common for smaller sized species (e.g. pink salmon) that require less handling time.
Each of the four major river systems on the west side of Hood Canal (Figure 1.3) offers
unique access to the mouth and estuarine areas where salmonid surface predation events
can be observed. These areas are all less than one square kilometer in size and, for the
most part, can be effectively covered visually by one or two observers. Additionally, low
flow levels and relatively shallow waters that typically exist during the observation
season provide observers with the ability to track seals underwater by following their
characteristic surface wake. Major harbor seal haul-outs are located at the mouth or
within estuarine areas of each river. This, combined with concentrations of returning
adult salmon in the estuary, resulted in a significant proportion of salmonid predations
occurring within an observable area.
14
Quilcene Bay
Dosewallips
Duckabush
Hamma Hamma
Skokomish
Figure 1.3 Map of North Hood Canal (left) and South Hood Canal (right) showing the locations of the
observation sites and harbor seal haul-outs in Hood Canal.
Methods
Surface Observations
Field observations were conducted in Hood Canal to record surface predation events on
returning adult salmon from vantage points off the mouths of the Quilcene (Big and
Little), Dosewallips, Duckabush, and Hamma Hamma rivers. Observations were also
conducted at the Skokomish River in 1998 and 1999, but data from the Skokomish is not
presented here. Duration of field observation activity has varied from year to year. In
1998, observations were conducted from the first week of September through the second
week in November. 1999 observations started the third week in August and concluded
after the first week of November. 2000 observations started the same week as in 1999,
but were finished the last week of October. 2001 and 2003 followed similar time spans
as the 2000 season; however, observations were limited to the Dosewallips and
Duckabush rivers. Differences in observation schedules were due in large part to an
15
increased understanding of salmonid abundance and timing, weather and tidal conditions
at observation sites, and a desire to increase efficient use of research time and money.
In 1998 and 1999, a non-stratified cluster sampling regime was employed, consisting of
3 six-hour periods randomly sampled across 3 days each week (Sunday-Saturday).
Approximately 300 hours of observation were conducted at each of the river mouths in
1998 and 1999. Additionally, a second observation site was added to the Duckabush
location in 1999 to address predation events occurring upstream from the Highway 101
Bridge that were not viewable from the lower river site. For each sampling week,
selection of specific sites for conducting surface predation observations was made
randomly and scheduled in advance. Each daily observation period was scheduled to
begin either 30 minutes after sunrise or end 45 minutes before sunset to allow adequate
ambient light for observations. Observations were made from either a 16 ft tower blind
(Dosewallips, Duckabush, and Hamma Hamma), or ground vantage point (Quilcene
Bay) which allowed viewing of predation events within and around the lower main
channel and tidal areas of each river.
In 1998 and 1999, daily observation periods lasted a total of 6 hours from arrival. The
first 20 minutes following arrival were spent organizing equipment, data forms and
setting up for observations. This was followed by three 100-minute observation periods
with a 20 minute break between each. Weather, overall visibility conditions, maximum
number of seals foraging in the river, and total number of salmonid predations was
recorded for each 100-minute observation.
The observation sampling-scheme was significantly altered for the 2000, 2001 and 2003
seasons. Preliminary data analysis from 1998 and 1999 indicated that predation rates
were not constant across the entire tidal cycle with a majority occurring on an incoming
tide. Stratified random sampling allowed concentration of observation effort on those
times during which predations were more likely. The 2.5-month observation season was
divided into five two-week periods. Two-week sampling periods were divided into two
strata. The first stratum consisted of four-hour periods during which a maximum
percentage of predations had been observed at a particular site during the previous two
16
field seasons. The second stratum consisted of the remaining time and was divided into
2-hour blocks. The first stratum consisted of those periods occurring during observable
daylight hours. As in 1998 and 1999, observable daylight hours were denoted as 30
minutes after sunrise to 45 minutes before sunset.
Once the two strata were determined, four maximum predation periods were chosen at
random for each two-week period. Each of these periods was observed in its entirety.
For the second stratum, two days were chosen at random and two two-hour periods were
randomly chosen for each day. These two-hour periods did not overlap and occurred
outside the maximum predation stratum.
The focus of the observer, during all years, was to cover the area encompassing each site
where predation by seals was possible. Binoculars and spotting scopes were used to scan
the area for pinniped presence and detection of predation events. Locations of predation
events were identified and recorded based on a gridded location map of each observation
site. Each observer documented any predation or foraging event on the data form.
Observers noted time, location, number of seals involved, species of salmon (if
possible), a confidence factor of 1-5 for prey identification, and a variety of possible
behaviors (e.g. chase, competition, partial consumption).
Nighttime Predation
Nighttime predation observations were conducted during the 2000, 2001 and 2003 field
seasons. Observations were only done at the Duckabush River and were paired with
scheduled daytime observations in order to provide statistically comparable information
on the potential differences in seal activity between daytime and nighttime. Nighttime
observation periods were selected to occur during four-hour high predation strata that
existed between sunset and sunrise, and within 24 hours of a similar daytime
observation. This allowed a paired analysis for comparison of mean number of
predations and number of foragers. Observers were positioned at the Mouth Site and at
the Highway 101 Bridge over the Duckabush River and observations were made using
an ITT 5001P head-mounted night vision goggle with a slip-on 3X magnifier lens. All
17
attempts were made to identify and record predation activity following the same
protocols used during daylight hours.
Night-vision equipment provided adequate clarity for observers to identify seal
movements and behavior within those areas of the river that were illuminated by light
from the moon or ambient sources. Areas covered by shadows (eg. bank edge areas
below the bridge, areas in the immediate vicinity of the bridge, and most area upstream
from the bridge) were too dark for detailed observations of predation events and other
seal behavior.
Calculation of Predation Estimates
Estimates of salmonid predation in 1998 and 1999 were determined through use of a
cluster sampling estimator. Each week served as an individual stratum with the primary
units being the random sample of three of seven days per week and the sub-units of three
100-minute observation periods on those days.
The additional up river site at the Duckabush River (1999-2001, 2003) was treated as
independent during calculation of predation estimates. Observers were in constant radio
communication to prevent any overlap in recorded predation events. After estimates
were analyzed, they were combined with the lower site (mouth) for comparison and
analysis with fish abundance numbers. Data collected in 2000, 2001 and 2003 relied on a
stratified random sampling estimator to calculate the number of predations for each bi-
week period. Weekly estimates from 1998 and 1999 were combined into bi-weekly
estimates for comparison across all years.
Allocation of Salmon Predations to Species
Observation of potentially significant predation of summer chum by harbor seals has
raised the importance of allocating salmonid predations to individual salmonid species.
Allocation of salmonid predations is especially problematic in Hood Canal because those
species with reduced populations (summer chum, chinook) overlap in timing with more
abundant species (coho, pink, fall chum). The extent to which seals are selective towards
one species over another is not known and likely not something that will be determined
18
without exceptional effort. Given these constraints, two scenarios with two different
analysis assumptions were explored for estimating predation impact on individual
salmon species. Each of these analysis scenarios is focused on determining the impact on
summer chum and is thus reflective of only the time during which summer chum were
present in each system (approximately August to mid-October).
Scenario I assumes there is no selection by harbor seals for or against summer chum in
relation to other salmonids, and the percentage of summer chum predations is equal to
the percentage of summer chum present. This scenario is the most objective and
parsimonious, however, estimates of availability derived from in-river spawner counts
likely do not fully represent the dynamic nature of species availability in the lower
reaches and tidal estuaries of these rivers.
Scenario II assumes predations identified by observers to species are reflective of all
salmonid predations and this percentage is used to estimate the impact of predations on
summer chum. This scenario relies heavily on the ability of observers to identify
predations to salmonid species, that their identifications are unbiased, and that each
species is equally identifiable. Most predations occur at a fair distance from the observer,
last only a few seconds, are mostly underwater and often provide little information that
would allow an observer to determine species. Additionally, the differences in size, color
and life history of each salmonid species (e.g. chum vs. coho) are variable and an
assumption that each species is equally identifiable is debatable.
Scenario II relies on the ability of each observer to identify predations to salmon species
consistently and without bias. All adult salmon undergo color and morphology changes
prior to spawning. Some species, such as chum and pink, begin the transformation to
their spawning phase at or just before their entry into freshwater. Coho and chinook
species tend to retain their non-spawning ‘bright’ coloration and morphology well after
their entry into freshwater. Salmon species in their ‘bright’ saltwater morphology and
coloration are virtually indistinguishable at any significant distance. Positive
identification of species while in this ‘bright’ phase usually requires careful examination
of gum coloration, spot patterns or scale size. Once salmon undergo transformation to
19
their spawning morphology, identification can be possible at distance because of
conspicuous coloration patterns or morphological changes. Male pink salmon acquire a
large hump on their back; chum salmon develop green and purple coloration patterns;
and coho salmon take on a deep red coloration. Because observation areas in this study
are tidally influenced, both phases are possible for each species along with various
transitional phases. This leads to a situation whereby a predation event involving a
‘bright’ salmon can only be identified as an ‘unidentified salmon’, a salmon in a
transitional phase may be identifiable to species, and a salmon in the spawning phase is
most easily identified to species. The different life histories and ecology of salmon
species present in Hood Canal result in unequal probabilities of identification in the
study observation areas. The unequal likelihood of positive species identification is
further complicated by differences in handling time for each species. Because of these
limitations with Scenario II and the more parsimonious nature of Scenario I, all
calculations of the percentage of summer chum consumed by seals will be based on the
relative percentage of summer chum available.
Estimates of Summer Chum Abundance
Detailed estimates of harbor seal predation on returning salmon do not provide any
quantifiable information on population level impacts without similarly detailed estimates
of summer chum escapement. WDFW generously provided bi-weekly estimates of
summer chum abundance for each river with predation observation sites. Estimates were
calculated based on spawning curves generated from spawner surveys done in each river.
Since spawner surveys were conducted upstream from the area covered during predation
observations, a correction for travel time to the spawning ground was included to
provide estimates that reflected abundance of each species in the lower river observation
areas. Details of the techniques and assumptions used to calculate escapement estimates
are presented in a WDFW technical report (Adicks et al. 2004).
20
Results
Seal Behavior
The Dosewallips and Duckabush rivers provided observers the best opportunity to record
seal behavior. Maximum numbers of seals foraging for salmon at each site for 1999-
2001 and 2003 were recorded for every observation period. The average value for the
maximum number of foragers at each site was calculated for each year (Table 1.1). The
mean value for the Dosewallips across all years was 2.75 (se=0.41) seals and 4.77
(se=0.68) seals in the lower reaches of the Duckabush river. The maximum number of
seals observed at the Dosewallips was eight and as many as fifteen were observed at the
Duckabush. The number of individual seals actively foraging for salmon in the lower
reaches of these rivers at any one time represents less than five percent of the total
population of seals that use nearby haul-outs.
Table 1.1 Average value for the maximum number of harbor seals actively foraging for salmon during an
observation. 2000, 2001 and 2003 values are calculated from the high-predation strata only. Standard
deviations based on the Poisson distribution are presented in parentheses.
Site 1999 2000 2001 2003
Dosewallips River 2.11 (1.45) 2.61 (1.62) 2.35 (1.53) 3.95 (1.99)
Duckabush River 2.81 (1.68) 4.85 (2.20) 5.67 (2.38) 6.45 (2.54)
The presence of seals foraging in the lower reaches is closely related to tidal stage. The
six-hour observations conducted in 1998 and 1999 were scheduled without respect to
tidal stage. Start times for all predations at each site were examined with respect to time
from the nearest high tide. At each site, the majority of salmon predations by harbor
seals occurred on an incoming tide and within a few hours of high tide. This relationship
between harbor seal behavior and tide cycles is a result of two key elements. First, the
small, low-flow nature of these streams restricts movement of seals into the lower
reaches where salmon are more vulnerable to predation. Seals are unable to move
upstream until tides reach one or two meters above sea level. Second, the movement of
salmon into these streams occurs mostly on the incoming tide.
21
Figure 1.4 Percentage of observed salmon predations occurring within overlapping four-hour bins relative
to high tide at four locations in Hood Canal. The ‘zero bin’ represents the percentage of predations
occurring from high tide to four hours after high tide.
The tidal cycle was divided into 30-minute segments that ranged from six hours before
to six hours after high tide. Recorded predations were binned into overlapping 4-hour
windows to determine the four hour window with the highest proportion of observed
predations (Figure 1.4). The window with the highest proportion was used as the
defining element for the stratified sampling design employed in 2000, 2001 and 2003.
Predation Estimates
Harbor seal predation on salmonids was observed at all five sites. On a few occasions,
California sea lion predation on salmonids was observed off the Hamma Hamma River.
Estimates of salmonid predations presented here only include predation attributed to
harbor seals. Harbor seals were observed preying on chum and coho salmon in Quilcene
22
Bay; chum, coho, and pink at the Duckabush River; chum, coho and pink at the
Dosewallips River; chum, pink and chinook at the Hamma Hamma.
Most predations at the Dosewallips and Duckabush rivers were within 50-75m of the
observation platform and provided observers with the best observation conditions for
identifying captured salmon to species. For 2000, 2001 and 2003, the average percentage
of predations identified to species was only 51.4% (n=106) for the Dosewallips and
52.3% (n=273) for the Duckabush. The low percentage of predations identified to
species indicates positive species identification is not common and, more importantly,
feedback from observers each field season suggests salmon species are not equally
identifiable.
Unequal probability of positive identification is related to the presence of conspicuous
characteristics for each salmon species and total handling time of the seals. Average
handling times for each species from the Dosewallips and Duckabush rivers across the
2000, 2001 and 2003 field seasons were calculated based on the recorded start and end
time for predation events. Seals averaged 8.89 minutes (s.d. = 10.23, n = 83) for chum
and 8.15 minutes (s.d. = 7.05, n = 37) for coho, while pink salmon predations lasted only
4.90 minutes (s.d. = 3.14, n = 92). Predations recorded as ‘unidentified salmon’ had an
average handling time of 5.16 minutes (s.d. = 5.87, n = 191). The shorter handling time
for pink predations would result in less opportunity for an observer to note
distinguishing characteristics for species identification. Pink salmon (present only in
2001 and 2003) may be under-represented in the subset of predation events with positive
species identification.
23
Table 1.2 Comparisons between paired day-time and night-time observations at the Duckabush River over
three years (2000, 2001, 2003) of observations. The presented p-values are from a two-tailed pairwise t-
test.
Salmon Predations Max Number of Foragers
Day-time Night-time Day-time Night-time
Mean per Observation 2.75 0.95 4.08 4.1
Standard Deviation 2.93 1.44 2.73 1.61
Paired Observations (n) 60 60 60 60
p-value < 0.001 0.96
The ability to discern detail during nighttime observations was significantly affected by
cloud cover, moon stage and angle of the moon. The ability of researchers to confirm
predation events was, therefore, significantly compromised compared to daytime
estimates. Differences in both the number of predations observed and maximum number
of foragers between day and night were evaluated with a paired t-test for sample means.
Results indicate a significant reduction in the number of predations observed at night
compared to the day (Table 1.2). Given the inherent differences in observability between
day and night, even with advanced night vision goggles, comparison of predation rates is
not informative. However, there was no significant difference in the number of foragers
present in the river.
Examination of telemetry data from animals tagged with VHF and sonic tags showed the
same animals were present in the river during both day and night. Additionally, a
logarithmic relationship (R
2
= 0.426, p=0) exists between the number of observed
salmon observations and the maximum number of foragers observed during an
observation (Figure 1.5). Comparing maximum number of foragers may provide a better
index for assessing predation activity between day and nighttime periods.
24
Figure 1.5 Scatter plot of the Maximum Number of Foragers vs. Observed Salmon Predations for day-time
observations during high predation strata in 2000, 2001 and 2003. The solid line indicates the logarithmic
regression model. The narrow and wide bands indicate the confidence interval and prediction band.
Estimates of harbor seal predation impacts on Hood Canal summer chum runs were
calculated at each site for each year using Scenario I for allocation of species (Table
1.3). Harbor seal consumption calculations are done with an assumption that predation
rates observed during daylight hours are consistent across the entire 24 hours. The
largest estimate of absolute consumption occurred at the Duckabush River in 2003 with
25
166 summer chum consumed. The lowest absolute consumption was in 1998 at the
Hamma Hamma River with only five summer chum consumed. The greatest and lowest
estimated percentage of returning summer chum consumed occurred at the Dosewallips.
In 1998, 27.7% of summer chum returning to the Dosewallips River were consumed by
harbor seals. Yet, in 2001, only 0.84% of the run was consumed. The average, across all
sites and years, was 8.0% (se=2.06) of the returning summer chum consumed.
Table 1.3 Estimates of predation by harbor seals on summer chum at sites in Hood Canal, assuming seals
consume salmon species in proportion to their relative abundance and that night predation rates are equal to
those observed during daylight hours.
Site Est. S. Chum Predations S. Chum Escapement % of Run CV
Year: 1998
Quilcene 66 1922 3.44 0.213
Dosewallips 91 329 27.7 0.178
Duckabush 26 226 11.3 0.264
Hamma Hamma 5 106 4.38 0.535
Year: 1999
Quilcene 43 2976 1.43 0.396
Dosewallips 19 351 5.46 0.271
Duckabush 10 93 10.4 0.209
Hamma Hamma 6 234 2.45 0.272
Year : 2000
Quilcene 92 5469 1.68 0.082
Dosewallips 87 1293 6.76 0.066
Duckabush 132 494 26.8 0.082
Hamma Hamma 21 202 10.2 0.067
Year: 2001
Dosewallips 8 990 0.84 0.064
Duckabush 40 944 4.22 0.031
Year: 2003
Dosewallips 159 7065 2.25 0.032
Duckabush 166 1873 8.84 0.034
The wide variation in seal consumption between sites and across years (Figure 1.6) is
due in large part to the differences in year to year salmonid escapement. In 1999, only 93
summer chum returned to spawn at the Duckabush River. That same year, nearly 3,000
summer chum returned to the Big Quilcene River. Furthermore, by 2003, the escapement
26
estimate for the Duckabush River had grown by twenty-fold. Variation in escapement of
other salmon species that overlap temporally with summer chum also plays an important
factor in each river. Under the Scenario I assumption of consumption in proportion to
abundance, species that return in high numbers during the summer chum run (i.e. pink
salmon) act as a buffer to summer chum consumption. Pink salmon only return to Hood
Canal streams during odd years (1999, 2001 and 2003 in this study). The estimates of
seal predation on summer chum tend to show higher percentage consumption in even
years when pink salmon were not present.
27
Estimated Percentage of Returning Adult Summer Chum
Consumed by Seals (Day + Night) Under Scenario 1
0%
10%
20%
30%
40%
1998 1999 2000 2001 2003
Year
Percent of Total Run
Quilcene
Dosewallips
Duckabush
Hamma Hamma
Figure 1.6 Estimated percentage of adult summer chum consumed by harbor seals at four sites in Hood
Canal across four years under scenario 1 and assuming night predation rates are equal to those observed
during daylight hours. Error bars indicate the upper 95% confidence interval.
Variance Calculation and Confidence Intervals
Variance calculations and 95% confidence intervals were calculated for the total
predation estimate during the summer chum run at each site in each year. It should be
noted that this does not include any variance that might be associated with salmonid
abundance estimates. Ranges presented here only represent variance associated with the
estimated total predation based on surface observations. Coefficients of variation were
calculated for each site in each year (Figure 1.7). The benefits of the stratified random
sampling approach used in 2000, 2001 and 2003 can be seen in reduced CV values for
those years compared to 1998 and 1999.
28
CV Values for Estimates of Summer Chum Consumed by
Harbor Seals 1998-2003
0%
10%
20%
30%
40%
50%
60%
1998 1999 2000 2001 2003
Year
Percent of Total Run
Quilcene
Dosewallips
Duckabush
Hamma Hamma
Figure 1.7 Coefficient of variation values for the estimates of summer chum consumed by harbor seals in
Hood Canal. 1998 and 1999 observation scheduling was done irrespective of tidal stage. 2000, 2001 and
2003 observations were scheduled under a stratified sampling protocol and observation effort was skewed
towards periods of higher predation rates.
Discussion
Assessing Population Impacts of Seal Predation
Seal predation on summer chum and other salmonids was observed at a higher rate than
initially expected, but predation estimates alone cannot provide enough insight to
evaluate impacts on recovery of Hood Canal summer chum populations. The estimates
of predation rates from this study should be evaluated in context with the productivity
potential of summer chum. Summer chum were not historically studied for critical
population parameters such as age composition and recruits per spawner. Lack of data
for these critical parameters has made it difficult to assess impacts of harbor seal
predation and other forms of mortality on the population dynamics of summer chum.
29
Under the summer chum conservation plan (WDFW et al. 2003), recovery criteria
include the achievement of an average of 1.6 recruits per spawner over the eight most
recent brood years, with values of 1.2 or less occurring in no more than two of the eight
years. Under the plan, fishery-related mortality is expected to be no more than 16.7%.
Using the highest reported predation estimate from this study (27.7% at the Duckabush
in 2000) and the maximum expected incidental-catch mortality, a population with a
spawner-recruit ratio of 1.6 would still increase by 15.6%. In reality, the likelihood of
these maxima occurring in the same year is minimal. Results from this study indicate
that in most years at most sites, harbor seal predation amounts to less than 8% of the
returning adults. Natural productivity would need to consistently fall below 1.1 recruits
per spawner before harbor seal predation could impact recovery. However, as summer
chum production and abundance fall, the functional response of harbor seals is likely to
result in reduced predation rates. Harbor seals are opportunistic, generalist predators and
likely have a type three functional response to the presence of salmonid prey
(Middlemas et al. 2006). Their main diet in Hood Canal is composed of more
consistently abundant and predictable prey such as hake and herring. With readily
available alternative prey, the few seals involved in salmon predation are not likely to
remain focused on salmon at low densities.
Examination of the annual escapement trend for summer chum in Hood Canal over the
time of the predation study does suggest the observed level of predation has minimal
impact on recovery of summer chum. Returns of summer chum to Hood Canal streams
increased each year from 1998-2003 and, in 2004, was the highest on record (Adicks et
al. 2004; WDFW et al. 2003). Extinction risk for summer chum in the four river systems
in this study is now listed as moderate to low (WDFW et al. 2003). The age of returning
chum salmon can be anywhere between two and five years with the majority returning as
age three or four (WDFW et al. 2000). Therefore, the increased returns in 2001-2004
represent the production from adults subjected to predation from 1998 to 2000. The
highest levels of predation observed during the study were 28% at the Dosewallips in
1998 and 27% at the Duckabush in 2000.
30
A natural broodstock supplementation program has been in employed in recent years for
some of the critically low populations of summer chum in Hood Canal (WDFW et al.
2003). However, only one of the study streams (Hamma Hamma) has been included in
the supplementation program. It is assumed increasing returns in the study streams in
recent years are a result of natural summer chum production. Additionally, it is
important to realize that results presented here account for only predation occurring
within the lower reaches of spawning rivers. While we feel this predation accounts for
the majority of harbor seal consumption in Hood Canal, it does not reflect any predation
that occurs in the open water region of Hood Canal.
It is also important to consider potential indirect impacts seal behavior may have on
production of summer chum in Hood Canal. Summer chum spawn in the lower reaches
of these rivers and some of the spawning area is accessible by harbor seals at higher tide
levels. Seal presence and harassment of spawning summer chum may have indirect
implications on reproductive behavior and success. Such indirect impacts were not
addressed by this study and no quantifiable data exists for proper examination.
However, we can say the frequency is high enough that future studies and population
viability exercises should include some low level of indirect effect.
The rebound of pinniped populations along the West Coast of the United States should
be considered a large conservation success. High profile situations like the Ballard Locks
in Seattle and Bonneville Dam on the Columbia River and, regular interactions between
pinnipeds and commercial and recreational fishermen, have led to a miss-
characterization of seals and sea lions as problem predators. This study demonstrates
that, at least for Hood Canal, harbor seal predation, while important for managers to
account for, is not an important priority in long term conservation efforts. The levels of
summer chum predation estimated are not consistently high enough to result in a
meaningful impact on summer chum production in Hood Canal.
The majority of seals in Hood Canal are relying on hake and herring as their stable diet.
Over two-thousand scats were collected across all the sites in Hood Canal from 1998-
2003 and 74% contained remains of Pacific hake and 43% Pacific herring; salmon was
31
present in 27% of the scats (Lance unpublished). When FO is normalized to add up to
100%, salmon only accounts for 14% of the harbor seal diet. Harbor seals are
opportunistic predators and when a large biomass of prey such as salmon is available
some individuals will choose to take advantage of the resource. In Hood Canal, an even
smaller subset of the population has adapted to focus almost exclusively on salmon
within the lower reaches of these small rivers.
The conclusion that harbor seal predation has minimal impact on the recovery potential
of summer chum in Hood Canal does not mean predation by harbor seals, and other
pinnipeds, at terminal areas can be ignored by resource managers. This is especially
important in areas where pinniped predation is known to be at consistently high levels
from year to year. Recent increases in the presence of California sea lions feeding on
salmon near dams along the Columbia River may be such a situation. It is essential, in
these areas, that some level of pinniped predation be incorporated into management
plans and harvest allocation scenarios. Without adequate set-asides for pinnipeds and
other natural predators, managers may unwittingly prescribe harvest levels beyond
sustainability. Unfortunately, accurate estimates of pinniped predation require long-term
studies and results are not likely applicable across different regions and species. In those
areas where pinniped predation is believed to be a critical issue, robust monitoring
programs should be implemented to estimate the level of predation and results from
other studies should be incorporated into a precautionary approach to resource
allocation.
Seals and salmon in Hood Canal have developed a balanced trophic relationship over
thousands of years. As long as higher priority efforts in salmon conservation to protect
and restore habitat and reduce by-catch and over-fishing are successful, there should be
enough salmon for harbor seals and other predators, including recreational, commercial
and tribal fisheries. The increasing trend in escapement of summer chum to Hood Canal
during the previous six years may be evidence of successful recovery planning efforts on
the part of resource managers and local citizens to restore a declining population to
historical levels.
32
33
Chapter 2. Impacts of Two Extended Foraging Events by Mammal-
Eating Killer Whales on the Population of Harbor Seals in Hood
Canal, Washington.
Introduction
The theory that transient-type, mammal-eating killer whales may be responsible for
declines of pinniped populations in the Eastern North Pacific has garnered significant
debate (Springer et al. 2003; Williams et al. 2004). This hypothesis stems from the
conclusion that predation by a few killer whales is responsible for a dramatic decline in
abundance of sea otters in the Aleutian Islands (Estes et al. 1998). Springer et al. (2003)
expanded the top-down effect of killer whale predation to other pinniped species in the
North Pacific. Their theories rely heavily on assumptions regarding daily metabolic
requirements of wild killer whales and their functional response to prey populations.
These assumptions and analyses were outlined in Williams et al. (2004). Unfortunately,
opportunities to verify these assumptions with empirical data from wild populations are
limited. Mammal-eating killer whales are characterized by discreet behavior and spend
much of their time in remote locations not frequented by researchers. Therefore,
knowledge of killer whale intake rates is limited to small datasets of mostly
opportunistic data or extrapolations from captive killer whales or other large terrestrial
carnivores.
One might imagine an ideal situation whereby daily requirements of killer whales could
be estimated from a wild population. Under this scenario, a group of whales would be
confined to a specific geographic area over a certain period of time. These whales would
then be provided with a known amount of prey. As long as no additional prey were
added to the area or removed by any means other than killer whale predation, differences
in prey abundance would provide an estimate for daily energetic requirements and
insights into the potential impact killer whales might have on isolated pinniped
populations throughout the Eastern North Pacific. Additionally, observations of killer
whales while in the area would provide an independent assessment of prey consumption.
34
Two recent extended foraging events by killer whales in Hood Canal, Washington are a
close representation of this ideal situation and have provided an unprecedented
opportunity to empirically measure the impact of mammal-eating killer whales on a
pinniped population. Hood Canal is an isolated 100km fjord on the west side of Puget
Sound and supported an estimated population of 1068 harbor seals in 2002. Between 2
January and 3 March, 2003, eleven mammal eating killer whales foraged exclusively
within Hood Canal. A second group of six mammal-eating whales were in Hood Canal
for 172 days in 2005.
Three separate killer whale ecotypes are present in the marine waters of the Pacific
Northwest. Fish-eating, resident-type orcas are believed to feed exclusively on fish and
predominantly on salmon. Fish-eating killer whales have a strong matrilineal social
structure and have been extensively studied throughout Washington and British
Columbia for the past 30 years. Mammal-eating, transient-type killer whales are known
to feed exclusively on other marine mammals (seals, sea lions, small cetaceans and some
large whales). Mammal-eating orcas are less frequently observed, although an extensive
photo-identification catalog does exist and many individuals have been photographed at
least once. Offshore killer whales are a third distinct ecotype and are most commonly
found away from coastal waters. Little is known about this ecotype, though preliminary
studies indicate they may be feeding on high trophic level fish species.
Killer whales have not had a significant presence in Hood Canal within the past thirty
years, although both mammal-eating and fish-eating killer whales have been previously
observed in Hood Canal. For both types, occurrences have been extremely rare and for
less than one or two days. A few acoustic recordings of killer whales from U.S. Navy
operations in Hood Canal have been confirmed and identified from their unique acoustic
dialect as fish-eating orcas.
Harbor seals are reported to be one of the preferred prey items for mammal-eating killer
whales, and harbor seals are the only consistently abundant resident marine mammal
species known to occur in Hood Canal. Regular aerial and ground counts of harbor seals
in Hood Canal have been conducted since the late 1970s and the population, as a part of
35
the larger population of seals within the semi-enclosed marine waters of Washington, is
believed to have stabilized at near carrying capacity in the mid-1990s (Jeffries et al.
2003). Tagging and telemetry studies conducted in Hood Canal and other areas of
Washington indicate no significant movement of seals between areas and, therefore, any
comparison of pre and post harbor seal relative abundance is likely not significantly
compromised by emigration or immigration.
The unique nature of these two killer whale incursions to Hood Canal has provided an
opportunity for empirical investigation into predation behavior of mammal-eating killer
whales and their impacts on localized pinniped populations. In order to maximize this
potential, a multi-faceted investigation was employed. The approach can be divided into
three key areas. First, behavioral observations, mostly from the 2005 event, have
provided opportunities to directly estimate killer whale consumption and document
foraging behaviors of mammal-eating killer whales. Second, a quantitative analysis of
harbor seal aerial survey counts over time provides a mechanism to evaluate expected
population responses given the presence of killer whales. Third, bio-energetic modeling
allows a more theoretical examination of the trophic impact of killer whales. Each facet
provides an independent evaluation of the impact of killer whale predation on the
population of harbor seals in Hood Canal. By comparing these estimates we not only
gain insight into the trophic ecology of killer whales, but also the benefits and limitations
of each approach.
Methods
Behavioral Observations
All behavioral observations were conducted under the authority of Scientific Research
Permit No. 782-1719, issued by the National Marine Fisheries Service under the
authority of the Marine Mammal Protection Act and the Endangered Species Act.
Opportunities to observe killer whales in Hood Canal during 2003 were limited to a few
days. Most of the 2003 field effort focused on documenting group structure through
photo identification and understanding the spatial use of Hood Canal. All whales were
36
photographed from the left and right side and individual identification was determined
from identification catalogs.
Hood Canal is populated with a number of shore-side residences and, because of its
narrow, fjord-like geography, provides ample opportunity for residents of Hood Canal to
observe killer whales. Observations by a few residents and a dedicated volunteer
provided the best information on movement and behavior of the whales in 2003.
In 2005, a coordinated effort between three research groups, in addition to observations
of local residents and volunteers, provided a better dataset for examining killer whale
foraging behavior and their spatial use patterns. As in 2003, all whales were
photographed from their left and right side and identified through comparisons with
photographic catalogs. Nineteen boat-based observations were conducted to document
predations and movements of the killer whales within Hood Canal. Each observation
was done opportunistically given weather and researcher availability. All observations
were conducted from 19-21 foot outboard powered vessels and all available resources
were used to locate the whales as soon as possible. Mobile phone coverage within Hood
Canal allowed local residents to quickly communicate sightings to researchers and recent
postings to internet distribution lists often provided critical information on sightings.
When recent sightings were not available, a search transect of Hood Canal was
conducted from the research boat until the whales were located.
Once whales were located, an initial GPS location was recorded and a trackline record
was initiated. Whales were counted and visually identified to confirm all individuals
were present. In general, the focus of the observation boat after first contact was to
follow and record confirmed predation events without altering whale behavior. Under
these circumstances, the general protocol was for the research boat to remain
approximately 100m behind the whales. Fast acceleration and ‘leap frog’ actions were
typically avoided and all attempts were made to minimize any effects the research boat
might have on the behavior of the whales. For some of the observation periods, other
objectives, such as collection of biopsy samples or prey remains, required temporary
departures from this protocol.
37
A strict protocol was employed for identification and confirmation of predation events.
Whales were closely observed for any changes in their behavior that might indicate
potential interactions with harbor seals or other prey. All predations were confirmed by
the presence of prey remains in the water column, an oil-slick on the surface of the water
or an observation of prey remains within the mouth of a whale. Additional behavioral
clues, such as observed interactions with live seals on the surface, and the presence of
diving gulls provided further evidence of predation activity but were not used as sole
confirmation of a predation event. The GPS location of all predation events was
recorded, and each predation event was considered complete when the whales returned
to their nominal travel behavior.
In order to extrapolate observed predations to an estimate of killer whale consumption,
observations would ideally be of equal length and scheduled randomly across time.
However, the opportunistic constraints of our effort negated the ability to plan
observations in advance. Additionally, time to first location for any given planned
observation trip was not predictable. Therefore, all attempts were made to approximate a
random and unbiased sample of time. When possible, the length of the observation
period was pre-determined on commencement. This was done to avoid any bias that
might occur if whale behavior was used as a determining factor. For instance, it would
not be advisable to consistently end observations after a predation event or to continue
an observation until a predation event occurred. Both situations would bias the final
estimates towards a higher consumption rate.
For each observation, a predation rate (kills/hour) was calculated from the number of
confirmed predations and the length of the observation. An average predation rate was
extrapolated across the duration of killer whale presence in Hood Canal under two
scenarios. Scenario 1 assumes predations only occurred during daylight hours. All
observations were conducted during daylight hours only. Information on the behavior of
mammal-eating killer whales at night is limited and the only available study suggests
indications of lower activity levels at night (Baird et al.
in review
). For the daylight-only
scenario, the average predation rate was only extrapolated across hours between sunrise
38
and sunset. Scenario 2 was evaluated under the assumption that predation rates observed
during the day are representative of killer whale behavior across day and night. Under
this scenario, the average predation rate was extended across all hours of the day.
Generalized Linear Model
Aerial counts done between 1996 and 2004 were assembled and incorporated into a
generalized linear model (GLM) to evaluate the impact of killer whale consumption in
2003 on the harbor seal population of Hood Canal. Further details on the aerial survey
protocol can be found in Jeffries et al. (2003). When available, all counts were done
from photographs taken during the aerial survey. When photographs were not taken,
counts recorded by the aerial observer were used.
Harbor seal haul out patterns are known to be influenced by tidal height, tidal stage, time
of day and day of year. Historical observations in Hood Canal suggest harbor seals are
more likely to haul out at high tide stages in mid-afternoon and during the pupping
(August-October) and molting (September-November) seasons. All aerial surveys in
Hood Canal between 1996 and 2004 were flown between August and November and
within +/- 2 hours of high tide. Because surveys are limited to times when seals are
expected to haul out in the highest proportions, the inclusion of tidal factors (stage and
height) were not included in the final GLM analysis.
Four hypotheses on how the Hood Canal seal population has responded to killer whale
predation can be expressed as different GLMs. The first hypothesis suggests ‘no effects,’
and that aerial counts are correlated with only ‘day of year’ and ‘haul-out site.’ This
hypothesis also suggests the population is stable over time period from 1996 to 2002.
The second hypothesis predicts a ‘year effect’: the population of seals in Hood Canal is
changing on an annual basis. The third hypothesis is the ‘treatment’ and represents a
stable population between 1996 and 2000 that then changed in 2003 due to killer whale
predation. The final hypothesis is similar to the ‘treatment’ effect but allows for growth
in the population between 2003 and 2004. To evaluate whether there was a reduction in
seal abundance in 2003, the four model variants were compared using AIC model
selection.
39
At the time of writing, the aerial surveys of Hood Canal for 2005 have been completed
but counts from photographs have not been finalized. Once final counts are available, a
similar GLM analysis will be conducted.
Bio-Energetic Monte Carlo Simulation
A bio-energetic model of killer whale consumption was developed to estimate the
predicted number of seals consumed by killer whales during the extended stays in Hood
Canal. Parameters for metabolic requirements for killer whales were selected from
published literature and information on the caloric value of seals was derived from seals
captured in Hood Canal, caloric analysis of seals from Washington state and values from
published literature.
Caloric content of harbor seals was determined from two whole body carcasses collected
in the Grays Harbor and south Puget Sound regions of Washington. Both animals were
considered in healthy body condition at the time of death and were provided by the
Washington Department of Fish and Wildlife.
Carcasses were ground whole in the food preparation area at United Farms in Graham,
Washington. Homogenate was passed through the grinder twice to insure complete
homogenization. Four approximate four ounce aliquots were taken from each
homogenate and stored at -20 C. The grinder was washed and cleaned between each of
the carcasses to minimize any cross contamination.
Calorimetric content was determined with a Parr 1425 semi-micro bomb calorimeter
(Parr Instrument Company, Moline, Illinois). Two 10g sub samples from each specimen
were dried to constant mass at 50 C. Constant mass was reached when the percent
change in mass was less than 0.2% in a 24-hour period. Sub-samples were further
homogenized with mortar and pestle and an approximate 0.10g pellet was used in the
bomb calorimeter. Caloric content was determined and converted to wet weight values
based on sample moisture loss during drying.
The equation for determining total caloric requirements of the mammal-eating killer
whales in Hood Canal:
40
(Whale kcal/kg/day) x [(Adult Male Mass x N
m
whales) + (Adult Female -
Subadult Mass x N
f
whales) + (Juvenile Mass x N
j
whales)] x (t Days)
Where N
m
= number of Adult Males, N
f
= number of Adult Females and
Subadults, N
j
= number of Juveniles, and t = the number of days present in Hood
Canal
The caloric value of harbor seals in Hood Canal can be determined from
(Seal kcal/kg) x Seal Mass x Whale Assimilation Value
Whale requirements divided by the caloric value of harbor seals results in a predicted
number of harbor seals consumed. This value, however, does not accurately reflect
uncertainty around any of the parameters. Therefore, a Monte Carlo simulation was used
to include uncertainty in the final estimate.
Table 2.1
Parameter values and distributions used in the Monte Carlo simulation of Killer Whale
consumption of Harbor Seals in Hood Canal, Washington.
Parameter Source Range Distribution
Whale Requirements Williams et al. (2004) 55 kcal/kg/day Fixed
Adult Male Mass 4200-7000 kg Uniform
Adult Female-Subadult Mass 2100-3500 kg Uniform
Juvenile Mass 1365-2275 kg Uniform
Harbor Seal Caloric Content Perez 2500-3800 kcal/kg Uniform
Harbor Seal Mass WDFW unpub. data 50kg Normal with s.d. = 7
Assimilation Value Williams et al. (2004) 0.85 Fixed
Number of Days 59 Fixed
A range of values was used for each parameter in the model (Table 2.1). A value of 55
kcal/kg/day is the median value reported (51-59) by Williams et al. (2004) for metabolic
requirement of mammal eating killer whales. Ranges for killer whale mass were
determined from reported values and consultation with other killer whale biologists. The
harbor seal caloric content range includes values determined from the analysis of seals
from Washington as well as reported values by Perez (1990) for ringed seals. The harbor
seal mass value is a weighted average of non-pup, non-pregnant harbor seals captured in
41
Hood Canal between 1998 and 2002 (n=175). Non-pup, non-pregnant weights are used
to best represent the available prey between January and June. The assimilation value of
0.85 is similar to values reported by Williams et al. (2004).
The bio-energetic model was calculated 15,000 times with new parameter values chosen
from the listed ranges each time. A distribution of simulation outcomes and the median
outcome along with 2.5 and 97.5 percentiles were calculated.
Results
Behavioral Observations
The killer whales present in 2003 and 2005 represent different individuals that are of no
known relation. In 2003, the group consisted of 11 individuals (T14, T74, T73, T73a,
T73b, T73c, T77, T77a, T77b, T123, and T123a) of which 2 were adult males (T14 and
T74), 7 were sub-adults or females and 2 were juveniles (T73c and T77b). In 2005, six
whales were present (T71, T71a, T71b, T124a, T124a1, T124a2) and the group was
composed of two adult females (T71 and T124a) and their two offspring. With the
exception of whale T14 (2003), these whales have limited to no sighting history in
Washington state. The longest and most consistent sighting record of individuals from
both groups comes from areas of northern Southeast Alaska (pers. comm. Jan Straley,
University of Alaska Southeast, Sitka, AK).
Opportunities for detailed observations of the group in 2003 were limited to a few boat-
based observations and sighting reports from residents of Hood Canal. All eleven
whales were observed to use the entire expanse of Hood Canal and were most often
observed as either one large group or two smaller groups of 5-6 whales. No confirmed
predations were observed during boat-based research observations, however, several
residents did report sightings of harbor seal predations and a few of those observations
were confirmed with photographic documentation.
42
Figure 2.1 Box plot of observation times during 2005. The box extents represent the time when observers
were following the whales and the dark lines represent the mid-point of the observation.
Vessel based observation effort in 2005 was significantly greater than in 2003. Fourteen
observation periods were conducted between February 2, 2005 and July 1, 2005.. The
average observation period lasted 4.64 hours with a minimum of 1.7 hours and
maximum of 7.17 hours (Figure 2.1).
43
GPS track-lines and predation locations (Figure 2.2) clearly demonstrate how these
whales used the entire expanse of Hood Canal. Additional locations reported by
residents to Orca Network (not shown) present a similar spatial use pattern.
Figure 2.2 Map of North and South Hood Canal showing tracklines from each of the boat based
observations and locations of all confirmed harbor seal predations in 2005.
A total of 18 confirmed harbor seal predations were observed during the observation
periods. One unsuccessful predation attempt on a California sea lion was also observed,
but all other predation events were confirmed as harbor seals. It was not possible to
determine the level of individual consumption; therefore a group predation rate was
calculated. When adjusted for observation effort, the median consumption rate is 0.329
harbor seals per hour with boot-strapped 97.5 and 2.5 percentiles of 0.465 and 0.215
harbor seals per hour, respectively. The diurnal estimate for total consumption is 758
harbor seals consumed with a boot-strapped confidence interval of 495-1072. The
estimate of consumption across all hours is 1358 with boot-strapped confidence interval
of 887-1921.
44
Behaviors observed in Hood Canal appear to be typical of other mammal-eating killer
whales. Predation events occurred over deep water, away from any shoreline, and within
a few meters of the shoreline in relatively shallow water. The range of behaviors
observed was also variable between events. On those occasions when a predation event
was relatively short, an oil slick and small remains in the water column were often the
only indication of harbor seal presence. However, extended predation events were often
characterized by the presence of a seal at the surface. These longer predation events
often involved a number of tail slap and ramming attacks on the seal.
Generalized Linear Model
Counts from aerial surveys at five index haul-outs in Hood Canal do not exhibit obvious
signs of significant population reduction after either of the killer whale incursions
(Figure 2.3). The average count across years from 1996 to 2000 was 684. Huber et al.
(2001) have proposed a correction factor for seals in the water of 1.56 for the inland
waters of Washington. Thus, the pre-killer whale estimate of seals in Hood Canal is
1068.
45
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
200 400 600 800 1000
Aerial Survey Counts in Hood Canal 1996-2005
Year
Count
Median Photo Count
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
200 400 600 800 1000
Aerial Survey Counts in Hood Canal 1996-2005
Year
Count
Median Photo CountMedian Photo Count
Figure 2.3 Box and whisker plot of aerial survey counts in Hood Canal summed across five index haul-
outs for surveys flown between August and November from 1996 to 2004.
AIC values were determined for each GLM representing the four hypotheses (Table 2.2).
The ‘Treatment + Growth’ model was favored with the lowest AIC value of 1070.064.
However, the AIC values for the other models resulted in delta values of as little as
1.196 (‘Year Model’) and as much as 3.233 (‘Treatment Model’).
Table 2.2
AIC values from four GLMs evaluating hypotheses of harbor seal population response to killer
whale predation in Hood Canal, Washington
Model AIC
No Effect 1072.736
Year Effect 1071.260
Treatment Only 1073.297
Treatment + Growth 1070.064
The treatment and growth coefficients calculated from the GLM under the favored
‘Treatment + Growth’ model suggest a treatment reduction of 24% (95% CI: +3.5% to -
46
45.1%) after 2003 and a growth of 49% (95% CI: 0% to 123%) in 2004. Note, for
treatment effect and growth, the 95% confidence intervals include 0%.
Bio-energetic Monte Carlo Simulation
Moisture content values were approximately 42 to 51 percent in the two harbor seal
carcasses processed (Table 2.3). The yearling harbor seal carcass was recovered from
southern Puget Sound and had a mass of 19 kg. The sub-adult animal had a mass of 49
kg and was recovered from Gray’s Harbor, Washington.
The values of 2798 kcal/kg for the 49kg sub-adult and 3590 kcal/kg for the 19kg
yearling are lower values than reported for ringed seals (Perez 1990) and other pinnipeds
(Williams et al. 2004).
Table 2.3 Calorimetric values determined from whole body harbor seal carcasses recovered in Washington
State.
Age Class Mass (kg) % Moisture kcal/kg
SubAdult 49 kg 42.6 2798
Yearling 19 kg 50.8 3590
The bio-energetic Monte Carlo simulation for the 2003 event resulted in a median
outcome of 997 seals consumed (5th and 95th percentiles: 708, 1435). For the 2005
event, the median outcome determined from the model was 960 (2.5 and 97.5
percentiles: 685, 1383). The distributions of outcomes for both events are strikingly
similar (Figure 2.4). The bio-energetic model prediction compares with estimates of 758
and 1358 seals consumed for the diurnal only and all hour assumptions respectively. The
estimate from the bio-energetic model falls almost near the midpoint of these two
empirical estimates and the all-hour consumption estimate of 1358 is within the 95%
confidence range. The daylight only estimate falls just outside the 2.5 percentile.
47
Figure 2.4 Frequency distribution of model outputs from the bio-energetic Monte Carlo simulation for the
2003 and 2005 killer whale incursions.
Discussion
Behavioral Observations
The two extended foraging events by killer whales in Hood Canal differ in many
respects from the expected behavior of mammal-eating killer whales. Transient type
killer whales are thought to travel in small groups and spend only a few days in one
particular area. With stays of 59 and 172 days, the Hood Canal events represent some of
the longest reported stays by mammal-eating killer whales in one area. However, the
observed behaviors while in Hood Canal were not atypical of those seen in other
mammal-eating killer whales. In both 2003 and 2005, the killer whales appear to have
used the full expanse of Hood Canal as part of their regular movement and foraging
patterns. Neither group exhibited any abnormal behaviors that might be characteristic of
a group trapped or lost.
48
Predation locations were not necessarily associated with harbor seal haul-outs. Harbor
seal haul-out locations in Hood Canal are characterized by large, shallow-water, tidal
expanses. The physical characteristics of these haul-outs would provide refuge from
killer whale predation. Predation locations may better reflect harbor seal foraging
locations. The seals would be more vulnerable during foraging activities and some of the
predation locations do overlap with confirmed foraging locations from harbor seal
movement studies in Hood Canal (WDFW unpublished data).
The estimate of total harbor seal predation during the 2005 event relies on two key
assumptions. First, that every predation event occurring during an observation was
recorded accurately. Most of the predation activity occurs under water and out of sight of
the observer. This limitation is most obvious during shorter predation events when one
might only get a fleeting glimpse at the prey animal. During longer events, it was more
common to see a harbor seal at the surface and the final consumption was more easily
confirmed. To alleviate uncertainty in our predation estimates, we employed a strict
protocol for identification of predations. Given other reports of harbor seal predations
being very subtle and hard to detect, estimates presented here are probably conservative.
The second assumption critical to the calculation of total harbor seal predation is that the
predation rate observed during our observations is representative of the un-sampled
period. With the limited observation time and opportunistic nature of the study plan, the
robustness of our estimate is lower than would normally be desirable. However, we
made a concerted effort to minimize any activity that would contribute any significant
bias to the final outcome. It would be reasonable to expect the predation rate of the killer
whale group in 2005 to change over the nearly six month presence in Hood Canal due to
changes in prey availability or improved knowledge of the area. With only 18
observations, we are unable to examine this possibility. Fortunately, our observation
effort is spread relatively even across this period and our estimate of average predation
rate would not be overly influenced by any temporal changes.
All predation observations conducted were diurnal and little is known about nocturnal
behaviors of mammal-eating killer whales. Baird et al. (in review) present limited data
49
from time-depth recorders that suggest mammal-eating killer whales have a lower
activity level at night compared to daylight hours. Given the uncertainty about circadian
changes in foraging rates, we have chosen to present two estimates of total harbor seal
consumption. One estimate is extrapolated across just the diurnal period, while the other
is across all hours. Activity levels of killer whales are likely influenced by more than just
light-level. Actual predation rate likely falls within the bounds of these two estimates.
Bio-Energetic Model
Despite differences in the 2003 and 2005 events (number of days, number of whales,
individuals present, age and sex distribution), projected consumption of harbor seals in
each year based on the bio-energetic model is strikingly similar. A likely explanation for
the model consistency is the importance of prey density and corresponding functional
response of the killer whales. As the population of seals in Hood Canal drops due to
killer whale predation, so does prey density and prey availability. At some threshold, the
cost of finding and catching one more harbor seal in Hood Canal is no longer
energetically beneficial and the whales leave. In Hood Canal, this threshold level
appears to be a removal of just under 1000 seals and seems to have remained the same
across these two events. The longer presence in Hood Canal for the 2005 group is a
result of their smaller group size and the absence of large adult males, reducing their
combined daily foraging requirements.
Parameters included in the bio-energetic model are reasonable given the current state of
knowledge with respect to the ecology and biology of mammal-eating killer whales. By
incorporating parameter uncertainty into the model we can better represent our
understanding of killer whale bio-energetics and the range of population impacts that
could be expected. Validation is a key aspect of any modeling exercise, and observations
conducted during the 2005 incursion have provided an opportunity to compare the model
predictions with empirical field data. Overall, the bio-energetic model is not inconsistent
with empirical estimates determined from field observations. The bio-energetic model
and range of parameters used appear to be appropriate predictors of killer whale
consumption of harbor seals in Hood Canal.
50
GLM Analysis of Harbor Seal Counts
The difference between the predicted impact of killer whales on the harbor seal
population in Hood Canal and the observed population change is unexpectedly large.
The reason for the disparity is unknown at this time. While many of the parameter values
and ranges used in the bio-energetic model are not based on empirical data, we feel these
values are reasonable based on all current knowledge of killer whale and large mammal
bio-energetics. The aerial surveys do exhibit a large amount of variability within and
between years. It is not known how much of the variability is due to natural variation
and how much is related to sampling variability. A significant factor in this variation
may be the influence of human disturbance on haul-out patterns of seals in Hood Canal.
The fact such a rare behavior has happened twice in the same location within two years
suggests there may be something especially attractive about Hood Canal. The harbor seal
population in Hood Canal is relatively naïve to killer whale predation. Hood Canal is a
long and narrow fjord with deep water areas that may provide a situational advantage to
the predator. Warmer temperatures and relative quietness of the environment in Hood
Canal may also be of importance to killer whales. Any attempt to explain why whales
have chosen Hood Canal for these extended stays is mostly speculative at this point. It
does, however, seem clear from the bio-energetic models that prey density plays a
critical role in determining the timing of departure from Hood Canal.
51
Chapter 3. Pinniped Scat Genetics: Identification of Sex and Species
from Feces Collected for Food Habits Studies
Introduction
Analysis of prey remains found in scats has become the method of choice for
investigation of pinniped diets. Scats are pinniped feces deposited on haul-out sites. In
most locations, scats can be collected in large numbers with relative ease and minimal
disturbance. Interpretation of data from cat analysis, however, does have limitations.
Biases associated with recovery and identification of otoliths and bones from some prey
species prevent the reliable use of scats for more than generalized characterization of
diet (Cottrell et al. 1996; Harvey 1989; Lance et al. 2001; Tollit et al. 2003).
Investigation of more detailed aspects of pinniped foraging, such as sexual variation in
diet, would require more intensive, potentially intrusive, and expensive methods such as
direct observations, enemas, lavaging, or examination of stomachs from harvested
individuals.
The potential for significant inter-sexual variation in the foraging ecology of pinnipeds is
substantial given there are behavioral and physiological differences that often exist.
Strong sexual dimorphism is evident across a number of pinniped species. This,
combined with the different energetic demands between males and reproductive females,
accounts for differing energetic requirements. Additionally, known cases of seasonal
geographic differences between sexes in foraging location and range seem to indicate
divergent foraging ecologies in some species (Le Boeuf et al. 2000; Thompson 1989;
Thompson et al. 1998).
Scat collection for food habits research is also confounded in some locations where more
than one pinniped species share the same haul-out location. Scats are, typically, not
reliably distinguished between species and any information from such locations has
limited application. Genetic analysis of scat material is a non-invasive technique that
would allow individual scats to be classified based on species, sexual or individual
identification of the source animal (Kohn and Wayne 1997; Reed et al. 1997).
52
Genetic scat analysis has been employed in a number of terrestrial mammalian studies
(Ernest et al. 2000; Farrell et al. 2000; Fedriani and Kohn 2001; Flagstad et al. 1999;
Kohn et al. 1999; Morin et al. 2001; Taberlet and Luikart 1999; Wasser et al. 1997), yet
its application to pinniped scat analysis has been limited (Reed et al. 1997). This study
presents the development of an efficient and reliable protocol for the extraction of
pinniped DNA from scats. While the extracted DNA can provide the basis for a variety
of genetic investigations (Kohn and Wayne 1997), here the focus is on amplification of
sex specific and species specific markers and their potential use for examination of
variation in diet. This methodology, when combined with standard protocols for
identification of prey remains from scat, can provide researchers with new insights into
the foraging ecology of pinnipeds.
Genetic scat analysis, as discussed here, refers to the use of DNA extracted from fecal
samples for investigation of genotypic attributes (e.g. sexual, species or individual
identification). Epithelial cells from an animal’s intestinal lining are sloughed off during
scat deposition and DNA can be extracted from these cells. Various loci (e.g.,
microsatellites, mitochondrial DNA regions, SRY) can then be amplified through a
Polymerase Chain Reaction (PCR) and used to establish sex, species or individual
identity. Reed et al. (1997) used genetic scat analysis and microsatellites to assign
species of seal to each scat in those areas of the Moray Firth, Scotland where harbor
seals and grey seals (
Halichoerus grypus
) share haul-out sites. Of 82 samples collected
and analyzed, seal species identification was possible in all but four scats.
In many locations along the Washington and Oregon coastlines, haul-outs are shared by
California sea lions and Steller sea lions. Research into the diet of both species in these
locations can be confounded because the segregation of the animals on the haul-out is
not consistent and the scats cannot be reliably identified to species from visual cues.
Therefore, any scats collected from haul-outs with mixed species composition must be
discarded or classified as ‘generic sea lions,’ a category with little interpretive value.
This approach ignores the significant possibility that these two species, while using the
same haul-out, are utilizing very different areas of the marine ecosystem. Genetic scat
53
analysis would provide a means for reliable identification of scats to pinniped species so
proper segregation and analysis is possible.
Methods
Sample Collection and DNA Extraction
The extraction protocol presented here was adapted from Wasser et al.’s (1997) method
for bear scats. A total of 834 scats were collected from five haul-out locations in Hood
Canal, Washington in 1999 and 2001. Each scat was collected and stored in individual
whirl-packs. Disposable, sterile tongue depressors and latex gloves were used to assist
with collection and prevent cross-contamination between scats. Scats were placed on ice
while in the field and stored at -20° C as soon as possible.
An extraction buffer, consisting of 500mM Tris-HCL, 16mM EDTA and 100mM NaCl
at a pH of 6.0, was added to each scat bag in approximately equal volume to the amount
of scat present. Each bag was resealed and the scat-buffer mixture (aka “poop soup”)
was massaged gently until the scat was thoroughly dissolved. The bag was then set aside
to settle for 10-15 minutes. Four 1.5mL tubes were labeled with the scat collection
number and four 1mL aliquots were removed from each scat with a pipette and barrier
tips. Given the consistency of the “poop soup” it was sometimes necessary to cut a 1/8”
section off the pipette tips. Negative control samples containing just the extraction buffer
were included in the workflow to monitor any cross-contamination. Tubes were stored in
a -20° C freezer. For characterization of diet, remaining scat material was poured
through a series of nested sieves and all prey remains were removed and identified in
accordance with standard pinniped food habits protocols (Lance et al. 2001).
The use of a liquid medium for separation of sample aliquots differs significantly from
other procedures for extraction of DNA from fecal material. The liquid medium not only
serves as a buffer, but also provides an efficient means of collecting multiple,
independent samples from each scat. The extraction and amplification of multiple tubes
in parallel provides further verification of genotypic characteristics in the final analysis
(Taberlet and Luikart 1999; Taberlet et al. 1999). The thorough mixing and
54
incorporation of the buffer and scat also alleviates any concerns that target epithelial
cells may not be consistently present throughout each scat. Without incorporation of the
buffer, thorough mixing would not be possible without risk of damaging fragile prey
remains present in each scat.
The remainder of the genetic extraction involves the use of a Qiagen QiAmp Stool
Extraction Kit (Qiagen catalog# 51504). Standard extraction protocol outlined in the
stool kit for extraction of human DNA was used with the exception that two 150ul
elutions were done at the final stage of the protocol. All extracted samples were stored at
-20° C.
Sex Specific Markers
Sex specific loci have been employed in a number a genetic studies for identification of
sex. The SRY (sex determining region) and ZFY/X (zinc-finger region) have been used
for sexual identification in mammals (Berube and Palsboll 1996; Escorza-Trevino and
Dizon 2000; Fedriani and Kohn 2001; Kohn and Wayne 1997; Reed et al. 1997; Wasser
et al. 1997). The SRY (Berta et al. 1990) is found on the Y-chromosome and thus only in
males, while the ZFY and ZFX loci are homologous regions present on the Y and X
chromosome, respectively. Thus, males are identified by the positive amplification of
both the SRY and ZFY/X loci, while in females only the ZFY/X loci amplify. A
GenBank search revealed a number of primers, in addition to those provided in the
literature, designed for amplification of these sex-specific regions. There are, however,
three drawbacks to using most of these listed primers. Many were originally designed on
the basis of either human or mouse sequences. Successful amplification is possible in
pinnipeds, however, the primers and sequence may not match entirely. In the case of
degraded DNA or poor PCR conditions, this may result in reduced amplification
success. Second, the vast majority of the published primers will amplify both seal and
human DNA. This is a considerable problem given the potential for human
contamination and the extra time and cost involved in effectively preventing
contamination from human sources. Lastly, many of the published primers amplify
products greater than 200bp in length. This is only a problem when dealing with highly
55
degraded and/or fragmented template DNA, as with scats, where small fragment lengths
would be ideal (Frantzen et al. 1998).
Given these constraints, new primers were developed based on complete sequence
information for the SRY region of harbor seals (GenBank Accession Number:
AY424662). The PvSRY2004 primer set (Figure 3.1) was designed with three goals: 1)
reliable amplification in pinniped scats, 2) no amplification of human DNA
contaminants, and 3) in the case where human DNA did amplify, the ability to detect it
as contamination. The primer sequences were designed to match conserved regions in
Zalophus californianus
and
Phoca vitulina
that were differentiated from
Homo sapiens
sequences. The region also included an extra 8-bp that is not present in the human
sequence. Successful amplification of the PvSRY2004 region should result in a 91-bp
fragment when amplified from pinniped DNA and an 83-bp fragment when amplified
from human DNA contaminant. Primer 3 (Rozen and Skaletsky 2000) was used for the
final design and evaluation of primer sequences for optimum PCR amplification.
10 20 30 40 50
....|....| ....|....| ....|....| ....|....| ....|....|
PvSRY2004 Primers T GGGCGGAGAA ATGAGTATTT
Zalophus californianus TCAGGGG--G CGGGTTTGAG GCAAGGTGCT GGGCGGAGAA ATTAGTATTT
Phoca vitulina TCGGGGG--G CGGGTTTGAG GCAAGGTGCT GGGCGGAGAA ATGAGTATTT
Homo sapiens CGGAGAAATG CAAGTTTCAT TACAAAAGTT AA-CGTAACA AAGAATCTGG
Clustal Consensus * * * **** * * * * ** * * * * * *
60 70 80 90 100
....|....| ....|....| ....|....| ....|....| ....|....|
PvSRY2004 Primers TAG---> <---CTTA
Zalophus californianus TAGAAACAAA AGTTACAGCA CCAGAGTGTA GATAATTTTT CGAACGCTTA
Phoca vitulina TAGAAACAAA AGTTACAGCA CCAGAGTGTA GATAATTTTT CGAACGCTTA
Homo sapiens TAGAAGTGAG TTTTG----- ---GATAGTA AAATAAGTTT CGAACTCTGG
Clustal Consensus ***** * ** ** *** * * *** ***** **
110 120 130 140 150
....|....| ....|....| ....|....| ....|....| ....|....|
PvSRY2004 Primers CACCTTCCAG CTTTGCTACC
Zalophus californianus CACCTTCCAA CTTTGCTACC CACCCACCCT TTTTTTTTTC CCACCGCTGT
Phoca vitulina CACCTTCCAG CTTTGCTACC CACCCACCCT TTTTTTCCCC CCACCGCTGT
Homo sapiens CACCTTTCAA TTTTGTCGC- -ACTCTCCTT GTTTTTGACA ATGCAATCAT
Clustal Consensus ****** ** **** * ** * ** * ***** * *
___________________
Z. californianus sequence excerpted from GenBank AY424650.1
P. vitulina sequence excerpted from GenBank AY424662.1
H. sapiens sequence excerpted from GenBank NM_003140.1
56
Figure 3.1 Aligned DNA sequences and PCR primers for the newly designed PvSRY2004 region of the Y-
chromosome for Z. californianus, P. vitulina and H. sapiens
A GenBank search, unfortunately, did not reveal complete sequence data for the ZFY/X
region so that harbor seal specific markers could also be developed as a positive control
for reliable identification of female scats. Instead, a microsatellite marker (Pvc63;
GenBank Accession Number: L40985) that amplifies reliably in tissue samples collected
from seals in Hood Canal was used. Initial screening of the Pvc63 locus indicated an
allelic range of 104-108bp.
Non-specific amplification is a common problem when working with DNA extracted
from fecal material. Non-specific bands were present on occasion and due to their
similarities in size, could have resulted in false sexual identification. In order to prevent
ambiguous results, the forward primer of each marker was labeled with different
fluorescent labels (PvSRY2004 with FAM and Pv63 with HEX).
Testing and optimizing scat DNA extraction and amplification is often difficult because
of the limited opportunity for simultaneous collection of scat samples and higher quality
DNA samples (flipper punch, blood, etc) from the same individual. We were able to
collect such paired samples on a few occasions during studies of wild harbor seals
conducted by the Washington Department of Fish and Wildlife (WDFW) in Puget
Sound, Washington, USA. Extractions from these samples were used to confirm
appropriate amplification of the primer sets and to optimize PCR conditions for large
scale application. Each sample was run on a Molecular Dynamics MegaBACE capillary
automated sequencer.
Species Identification Markers
Mitochondrial DNA sequences of Steller’s sea lions, California sea lions and harbor
seals were extracted from a GenBank search. A multiple alignment was performed with
ClustalW 1.8 to determine conserved and differentiated sequence regions. A PinID
primer set was designed with four separate primer sequences (Figure 3.2). The reverse
primer was designed to correspond with a conserved region across all three species.
Three forward primers were designed to match species specific sequence
57
differentiations. The four primer set will result in amplification of fragments of different
size depending on the species of source DNA. Harbor seals will amplify a 170bp region,
California sea lions a 129bp region and Steller’s sea lions a 134bp region. To further aid
in proper species identification, each of the forward primers could be labeled with a
different fluorescent label. Testing and optimization of the primer set was done using
tissue and scats from each of the species.
10 20 30 40 50
....|....| ....|....| ....|....| ....|....| ....|....|
PinID_Rev Primer
PinID_Zc Primer CTAT TCCCTGACAT GATTAAACTC C--->
PinID_Pv Primer CAATC CCCCTTTCAC TCCTCA--->
PinID_Ej Primer CCCTGACAT GACTAGGCCC TC--->
Zalophus californianus ATTAAACTAT TCCCTGACAT GATTAAACTC CC-----CAC ------ATTC
Phoca vitulina ATTAAACTAT TCCCTGACGC CCGCCCAATC CCCCTTTCAC TCCTCAATTC
Eumetopia jubata ATTGGACTAT TCCCTGACAT GACTAGGCCC TC-----CAC ------ATTC
Clustal Consensus *** ***** ******** * * *** ****
60 70 80 90 100
....|....| ....|....| ....|....| ....|....| ....|....|
PinID_Rev Primer
PinID_Zc Primer
PinID_Pv Primer
PinID_Ej Primer
Zalophus californianus ATATA-TACC ACTACCCCTA CTGTGCCACC ATAGTATCT- ----------
Phoca vitulina ATATAATAAT ATCACCT-TA CTGTGCTATC ACAGTATTCA CGCACACTGG
Eumetopia jubata ATATA-TACC ACTACACCCA CTGTACCACC ACAGTATCTC ----------
Clustal Consensus ***** ** * ** * **** * * * * *****
110 120 130 140 150
....|....| ....|....| ....|....| ....|....| ....|....|
PinID_Rev Primer
PinID_Zc Primer
PinID_Pv Primer
PinID_Ej Primer
Zalophus californianus ---------- ---------- ----TTTTTT CCC------- -------CTA
Phoca vitulina CCTATGTACT TCGTGCATTG CATGTCCCCC CCC-ATCCTC GGACCCCCTA
Eumetopia jubata ---------- -------TTT CTTTTTTTTT ---------- ----------
Clustal Consensus *
160 170 180 190 200
....|....| ....|....| ....|....| ....|....| ....|....|
PinID_Rev Primer <---TTGCCCCA TGCATATAAG CACTGTACAT
PinID_Zc Primer
PinID_Pv Primer
PinID_Ej Primer
Zalophus californianus TGTACATCGT GCAGTTGATG GTTTGCCCCA TGCATATAAG CACTGTACAT
Phoca vitulina TGTATATCGT GCA-TTAATG GTTTGCCCCA TGCATATAAG CA-TGTACAT
Eumetopia jubata --TATATCGT TNACTTAATG GCTTGCCCCA TGCATATAAG CA-TGTACAT
Clustal Consensus ** ***** * ** *** * ******** ********** ** *******
___________________
Z. californianus sequence excerpted from GenBank
P. vitulina sequence excerpted from GenBank
E. jubata sequence excerpted from GenBank
58
Figure 3.2 Aligned mtDNA sequences and PCR primers for the newly designed PinID region for Z.
californianus, P. vitulina and E. jubatus
PCR Reactions
Each extracted sample was included in three separate PCR reactions for each primer set
(PvSRY2004, Pvc63 and PinID). Muli-plex reactions were tried with the primer sets, but
discontinued because there was a significant decrease in reliable amplification and an
increase in non-target fragment amplification. Multiple negative and positive controls
were included in each PCR plate in addition to negative controls introduced in the
extraction procedure. All PCRs were conducted in a separate building and laboratory
from the location of the sample processing and DNA extraction to eliminate potential
cross-contamination.
Each PCR reaction was 15 µl and included 2 µl of template DNA from the extracted
sample. Amplitaq Gold taq polymerase was used at a final concentration of 0.035 U/µl
along with a 1x final concentration of Amplitaq Gold PCR buffer. All locus primers
were included at a final concentration of 0.3 µM with the exception of the three species
diagnostic PinID forward primers which were included at a concentration of 0.2 µM
each. Final dNTP concentration was 0.3 mM, 0.6 mM and 0.8 mM respectively for the
PvSRY2004, Pvc63 and PinID reactions. Corresponding MgCl
2
concentrations were 2.5
mM, 2.75 mM and 2.5 mM.
All reactions were carried out on an MJ Research thermocycler. Thermocycling
conditions all involved an initial ten minute denature at 94°C to activate the AmpliTaq
Gold. Each reaction also involved a 10-cycle touchdown procedure with an initial
annealing temperature five degrees above the target annealing temperature. Each
subsequent denaturing, annealing and extension step lasted fifteen seconds. The PCR
program was scheduled to run for 35 cycles. A different annealing temperature was used
for each locus (PvSRY2004: 56, Pvc63: 52, PinID: 54).
Sex and Species Determination
All samples were initially classified by sex and species based on a strict protocol.
Species was determined for those samples that exhibited positive amplification for only
59
one of the PinID loci. Any sample with product for more than one of the PinID loci was
removed from analysis as inconclusive. Sexual identification was determined only for
those samples with positive amplification of the Pvc63 marker and the PinID loci
specific to
P. vitulina
. Those samples with amplification of the PvSRY2004 fragment
were classified as males and those without were classified as females.
A second, less stringent, protocol for classification of sex was employed to include those
samples with positive amplification of the PvSRY2004 region and the PinID locus for
P.
vitulina
. Since the PvSRY2004 will only amplify in the presence of DNA from males, it
is reasonable to conclude these samples are from scats deposited by male seals. The use
of these criteria, however, does increase the probability of identifying a male scat
compared to a female scat.
Sexual Differences in Diet
Pearson Chi-Square contingency tables were used to test for differences in diet between
sexes for each of the top three prey species found in scats in Hood Canal. The top three
prey species were determined from a combined frequency of occurrence table across all
sites and years for which scats have been collected in Hood Canal (Lance, unpublished
data). A 2x2 contingency table was constructed for each of the prey items. The rows
represent the presence or absence of the prey species and the columns indicate those
scats identified as male or female. This analysis provides initial indications of diet
separation between male and female harbor seals in Hood Canal. However, the approach
will not account for any species interactions that are likely to exist when comparing
complex diets.
Results
For both the PvSRY2004 primer set and PinID primer set, positive amplification was
always a correct identification for sex and species when amplified from a known source.
All tissue samples resulted in consistent amplification across multiple, independent
reactions. Scat samples tended to either work reasonably well, or not at all.
Amplification of the mitochondrial DNA PinID marker was more successful than the
nuclear DNA Pvc63 and PvSRY2004 markers.
60
Figure 3.3 Example electropherogram for a male (top) and female (bottom) P. vitulina scat. The green
peak at 90bp represents a positive amplification of the PvSRY2004 fragment and confirms the male
identification. The blue peaks represent the microsatellite Pvc63 and the diagnostic PinID fragment for P.
vitulina is shown in black at 169bp.
The product size, as measured by the MegaBace Fragment Analysis software, differed
slightly from the predicted size (Figure 3.3 and Figure 3.4). This is likely an artifact of
the software algorithms used to determine fragment lengths, and not an indication of
actual differences in the sequence region amplified. Both the PvSRY2004 and PinID
markers often exhibited a secondary peak 1bp larger or smaller in size. The Pvc63
marker exhibited stereotypical microsatellite two base pair stutter patterns.
61
Figure 3.4 Example electropherogram for male scats from Z. californianus, E. jubata and P. vitulina. The
green peak at 90bp represents a positive amplification of the PvSRY2004 fragment and confirms the male
identification. The blue peaks represent the microsatellite Pvc63 and the diagnostic PinID fragments are
shown in black.
Amplification success in scats collected from Hood Canal was greatest for the mtDNA
PinID fragment. The success ratio for identification of sex based on the PvSRY2004 and
Pvc63 markers differed depending on the level of stringency applied to the classification
process. Only 244 of the 831 scats extracted had positive amplification of the Pvc63 and
PinID markers (29.4%). Of those 244 scats, 99 (40.5%) were determined to be female
and 145 (59.4%) were identified as male. The number scats for which a positive sexual
identification was possible increased when those scats exhibiting positive amplification
62
of the PvSRY2004 and either the Pvc63 or the PinID marker were used. This resulted in
an overall success in 370 (44.5%) of the 831 scats extracted.
Frequency of occurrence tables (Lance unpublished data) for all scats collected in Hood
Canal between 1998 and 2003 show the top three prey species found in scats to be
Pacific hake, Pacific herring and salmon. When compared between males and females,
Pacific hake is present in 69% and 66% of the scats respectively and a chi-square
comparison found no significant difference (p=0.268). There were, however, significant
differences in FO values for Pacific herring and salmon. Pacific herring was found in
54% of the female scats, while only 28% of the male scats contained herring (p<0.001).
The situation was reversed for salmon which was present in only 12% of the female
scats compared to 37% for male scats (p<0.001).
Discussion
High variability in amplification quality is common among scats (Taberlet et al. 1999).
This variability is a result of multiple factors such as degraded template DNA, low
concentration of DNA from the extract and large quantities of PCR inhibitors. An
adaptation of the multiple tubes approach (Taberlet and Luikart 1999) is suggested to
prevent false amplification and the misinterpretation of final results. Ideally, all samples
should be amplified in multiple, independent reactions. Funding and time constraints
prevented such an approach in this study. This study, however, has focused specifically
on the use of presence/absence markers as a means for reducing our susceptibility to
errors common in the use of fecal DNA. A large sample size of scats is also a key factor
in reducing the impact of any ambiguous amplification, as a higher standard can be set
without reducing the number of usable samples, and thereby the statistical power of
group comparisons, by a significant percentage. While the overall percentage of scats for
which we can confidently assign a sex to is relatively low, the absolute number of scats
represents one of the larger sample sizes to date for the examination of sex specific diets
in pinnipeds.
Recent terrestrial studies have had greater success amplifying fecal nuclear DNA for
genotypic and sexual identification. The reason for the reduced performance in our study
63
is probably a combination of several factors. A comparison of amplification success
rates of nuclear and mitochondrial DNA from scats of captive brown bears on controlled
diets was conducted to compare the influence of diet on the amplification of DNA from
feces (Murphy et al. 2003). They showed a significantly lower success rate for those
scats coming from bears on a salmon diet (26% vs. >60% for other diets). The
mechanism for the lower amplification success in scats containing salmon is not clear,
especially since mtDNA amplification was not influenced by diet. Murphy et al. (2003)
suggest the poor performance may be due to a lower slough rate for intestinal cells as a
result of the high lipid content or interference of salmonid by-products with the
extraction chemistry. Pacific hake, pacific herring and salmon are the three major
components of the harbor seal diet in Hood Canal. All are high lipid prey species that
could have a similar impact on amplification success as seen in the Murphy et al. (2003)
study.
Amplification success in this study may have also been influenced by the collection
protocol, procedural constraints required to maintain the integrity of the prey remains
and the time between collection from the haul-out and extraction and amplification.
Many of the more successful studies involving the use of fecal DNA employ a
desiccation step in the process to impede degradation of the genetic material over time
(Nsubuga et al. 2004; Wasser et al. 1997). A concerted effort was made to incorporate
the genetic component of this study into the normal practices and procedures of pinniped
scat collection and analysis. The high moisture content in pinniped scats and the need to
efficiently separate prey remains for identification eliminates the possibility of
desiccating the sample for preservation. While there were no strong differences in
success between scats collected in 1999 and 2001 (all were extracted in 2002 and
amplified in 2004), other studies have demonstrated decreased PCR performance with
increased time between collection and extraction/amplification.
The successful development and implementation of the PinID marker set is directly
applicable for use in the study of pinniped diets on the West Coast of the United States.
California sea lions and Steller sea lions commonly share haul-out locations between
64
Oregon and British Columbia. Researchers now have a reliable tool for examining
differences in diet between these two sea lions species.
The diet differences observed in the initial comparison of male and female scats in Hood
Canal are surprising given the minimal sexual dimorphism and generalist diet of harbor
seals. Without further details on sex-specific spatial use and foraging behavior, it is
difficult to determine the mechanism for this divergence. However, the timing of scat
collections overlaps with periods of peak pupping and weaning in harbor seal
populations of Hood Canal. The diets may reflect the differences in parental care
investment. Salmon are large, mobile prey that may require mor effort and longer trip
durations by foraging seals. During this time period, females are likely to be more
focused on parental care and to restrict movements to areas near the haul-out. The
similar FO values for Pacific hake may reflect an overwhelming availability or
preference for this prey species. Further examination of multi-variate relationships and
interactions between species present in harbor seal diet should be fully explored.
The study of sexual variation in diet among pinnipeds has been limited mainly to the
examination of stomachs from known-sex carcasses (Antonelis et al. 1994; Daneri et al.
2000; Murie and Lavigne 1992). Sexual variation in diet cannot be determined from
standard scat analysis protocols. Understanding the role of pinnipeds in various
ecosystems is crucial from both a management and conservation perspective.
Application of genetic scat analysis for sexual identification can provide researchers
with insights into pinniped foraging ecology that were previously unattainable without
excessive or highly intrusive efforts. This knowledge is of key importance when
applying food habits information from scat to various bio-energetic models to improve
conservation measures or understand potential conflicts with commercial fisheries or the
recovery of imperiled prey species. The striking inter-sexual variation that characterizes
adult anatomy, energy metabolism, and foraging geography in many pinnipeds require
that pinniped trophic ecology be understood at the level of gender. As the results from
this study show, even species with minimal size differentiation between genders can
have observable dietary preferences. We hope the techniques outlined here will be
65
applied to a variety of pinniped species and potentially lead to a greater understanding of
scat analysis and pinniped foraging ecology.
66
List of References
Adicks, K., J. Ames and J. Haymes. 2004. The Relative Abundance of Adult Salmon in
Four Hood Canal Streams During Pinniped Predation Studies Spanning the 1998 - 2001
Return Year. Washington Department of Fish and Wildlife.
Antonelis, G.A., S.R. Melin and Y.A. Bukhtiyarov. 1994. Early Spring Feeding-Habits
of Bearded Seals (Erignathus-Barbatus) in the Central Bering Sea, 1981. Arctic 47:74-
79.
Baraff, L.S. and T.R. Loughlin. 2000. Trends and Potential Interactions Between
Pinnipeds and Fisheries of New England and the U.S. West Coast. Marine Fisheries
Review 64.
Beddington, J.R., R.J.H. Beverton and D.M. Lavigne. 1985. Marine mammals and
fisheries. George Allen and Unwin, London.
Berta, P., J.R. Hawkins, A.H. Sinclair, A. Taylor, B.L. Griffiths, P.N. Goodfellow and
M. Fellous. 1990. Genetic-Evidence Equating Sry and the Testis-Determining Factor.
Nature 348:448-450.
Berube, M. and P. Palsboll. 1996. Identification of sex in cetaceans by multiplexing with
three ZFX and ZFY specific primers (vol 5, pg 283, 1996). Molecular Ecology 5:602-
602.
Brown, R.E., B.E. Wright, S.D. Riemer and J. Laake. 2005. Trends in abundance and
current status of harbor seals in oregon: 1977-2003. Marine Mammal Science 21:657-
670.
Calambokidis, J., K. Bowman, S. Carter, J. Cubbage, P. Dawson, T. Fleischner, J.
Schuett-James, J. Skidmore, B. Taylor and S.G. Herman. 1978. Chlorinated hydrocarbon
concentrations and the ecology and behavior of harbor seals in Washington state waters.
Cascadia Research Collective.
Calambokidis, J., J.R. Evenson and G.H. Steiger. 1991. Monitoring of harbor seals at
Dosewallips state park, Washington. Cascadia Research Collective.
67
Calambokidis, J. and B. McLaughlin. 1987. Harbor seal populations and their
contrbution to fecal coliform pollution in Quilcene Bay, Washington. Cascadia Research
Collective, Final Report.
Calambokidis, J., B.D. McLaughlin and G.H. Steiger. 1989. Bacterial contamination
related to harbor seals in Puget Sound, Washington. Cascadia Research Collective.
Calambokidis, J., G.H. Steiger, B. McLaughlin and J.R. Evenson. 1990. Harbor seal
haul-out habitat and the feasibility of shifting haul-out locations at Dosewallips state
park, Washington. Cascadia Research Collective.
Carretta, J.V., K.A. Forney, M.M. Muto, J. Barlow, J. Baker, B. Hanson and M.S.
Lowry. 2005. U.S. Pacific Marine Mammal Stock Assessments: 2005. U.S. Department
of Commerce, Technical Memorandum NOAA-TM-NMFS-SWFSC-375.
Cottrell, P.E., A.W. Trites and E.H. Miller. 1996. Assessing the use of hard parts in
faeces to identify harbour seal prey: Results of captive-feeding trials. Canadian Journal
of Zoology-Revue Canadienne De Zoologie 74:875-880.
Daneri, G.A., A.R. Carlini and P.G.K. Rodhouse. 2000. Cephalopod diet of the southern
elephant seal, Mirounga leonina, at King George Island, South Shetland Islands.
Antarctic Science 12:16-19.
Ernest, H.B., M.C.T. Penedo, B.P. May, M. Syvanen and W.M. Boyce. 2000. Molecular
tracking of mountain lions in the Yosemite Valley region in California: genetic analysis
using microsatellites and faecal DNA. Molecular Ecology 9:433-441.
Escorza-Trevino, S. and A.E. Dizon. 2000. Phylogeography, intraspecific structure and
sex-biased dispersal of Dall's porpoise, Phocoenoides dalli, revealed by mitochondrial
and microsatellite DNA analyses. Molecular Ecology 9:1049-1060.
Estes, J.A., M.T. Tinker, T.M. Williams and D.F. Doak. 1998. Killer whale predation on
sea otters linking oceanic and nearshore ecosystems. Science 282:473-476.
Farrell, L.E., J. Romant and M.E. Sunquist. 2000. Dietary separation of sympatric
carnivores identified by molecular analysis of scats. Molecular Ecology 9:1583-1590.
Fedriani, J.M. and M.H. Kohn. 2001. Genotyping faeces links individuals to their diet.
Ecology Letters 4:477-483.
68
Flagstad, Q., K. Roed, J.E. Stacy and K.S. Jakobsen. 1999. Reliable noninvasive
genotyping based on excremental PCR of nuclear DNA purified with a magnetic bead
protocol. Molecular Ecology 8:879-883.
Frantzen, M.A.J., J.B. Silk, J.W.H. Ferguson, R.K. Wayne and M.H. Kohn. 1998.
Empirical evaluation of preservation methods for faecal DNA. Molecular Ecology
7:1423-1428.
Furness, R.W. 2002. Management implications of interactions between fisheries and
sandeel-dependent seabirds and seals in the North Sea. Ices Journal of Marine Science
59:261-269.
Harvey, J.T. 1989. Assessment of Errors Associated with Harbor Seal (Phoca-Vitulina)
Fecal Sampling. Journal of Zoology 219:101-111.
Harwood, J. and J.J.D. Greenwood. 1985. Competition between British grey seals and
fisheries. Pages 153-169
in
J. Harwood, J.R. Beddington and D.M. Lavigne, eds. Marine
Mammals and Fisheries. George Allen and Unwin, London.
Huber, H.R., S.J. Jeffries, R.F. Brown, R.L. DeLong and G. VanBlaricom. 2001.
Correcting aerial survey counts of harbor seals (Phoca vitulina richardsi) in Washington
and Oregon. Marine Mammal Science 17:276-293.
Jeffries, S., H. Huber, J. Calambokidis and J. Laake. 2003. Trends and status of harbor
seals in Washington state: 1978-1999. Journal of Wildlife Management 67:207-218.
Kohn, M.H. and R.K. Wayne. 1997. Facts from feces revisited. Trends in Ecology &
Evolution 12:223-227.
Kohn, M.H., E.C. York, D.A. Kamradt, G. Haugt, R.M. Sauvajot and R.K. Wayne.
1999. Estimating population size by genotyping faeces. Proceedings of the Royal
Society of London Series B-Biological Sciences 266:657-663.
Lance, M.M., A.J. Orr, S.D. Riemer, M.J. Weise and J.L. Laake. 2001. Pinniped food
habits and prey identification techniques protocol. U.S. Department of Commerce AFSC
Processed Report 2001-04.
Le Boeuf, B.J., D.E. Crocker, D.P. Costa, S.B. Blackwell, P.M. Webb and D.S. Houser.
2000. Foraging ecology of northern elephant seals. Ecological Monographs 70:353-382.
69
Lowry, L.F. and K.J. Frost. 1985. Biological interactions between marine mammals and
commercial fisheries in the Bering Sea. Pages 41-61
in
J.R. Beddington, R.J.H. Beverton
and D.M. Lavigne, eds. Marine mammals and fisheries. George Allen and Unwin,
London.
Middlemas, S.J., T.R. Barton, J.D. Armstrong and P.M. Thompson. 2006. Functional
and aggregative responses of harbour seals to changes in salmonid abundance.
Proceedings of the Royal Society of London Series B-Biological Sciences 273:193-198.
Morin, P.A., K.E. Chambers, C. Boesch and L. Vigilant. 2001. Quantitative polymerase
chain reaction analysis of DNA from noninvasive samples for accurate microsatellite
genotyping of wild chimpanzees (Pan troglodytes verus). Molecular Ecology 10:1835-
1844.
Murie, D.J. and D.M. Lavigne. 1992. Growth and Feeding-Habits of Gray Seals
(Halichoerus-Grypus) in the Northwestern Gulf of St-Lawrence, Canada. Canadian
Journal of Zoology-Revue Canadienne De Zoologie 70:1604-1613.
Murphy, M.A., L.P. Waits and K.C. Kendall. 2003. The influence of diet on faecal DNA
amplification and sex identification in brown bears (Ursus arctos). Molecular Ecology
12:2261-2265.
NMFS. 1995. Environmental assessment on protecting winter-run steelhead from
predation by California sea lions in the Lake Washington Ship Canal. U.S. Department
of Commerce
Washington Department of Fish and Wildlife.
NMFS. 1997a. Environmental assessment on preventing California sea lion foraging and
predation on salmonids at Willamette Falls, Oregon. U.S. Department of Commerce.
NMFS. 1997b. Investigation of scientific information on the impacts of California sea
lions and Pacific harbor seals on salmonids and on the coastal ecosystems of
Washington, Oregon , and California. U.S. Department of Commerce NOAA Technical
Memorandum NMFS-NWFSC-28.
NMFS. 1999. Report to Congress: Impacts of California sea lions and Pacific harbor
seals on salmonids and West Coast ecosystems. U.S. Department of Commerce, NOAA
National Marine Fisheries Service.
Nsubuga, A.M., M.M. Robbins, A.D. Roeder, P.A. Morin, C. Boesch and L. Vigilant.
2004. Factors affecting the amount of genomic DNA extracted from ape faeces and the
identification of an improved sample storage method. Molecular Ecology 13:2089-2094.
70
Olesiuk, P.F., M.A. Bigg and G.M. Ellis. 1990. Recent Trends in the Abundance of
Harbor Seals, Phoca-Vitulina, in British-Columbia. Canadian Journal of Fisheries and
Aquatic Sciences 47:992-1003.
Perez, M.A. 1990. Review of marine mammal population and prey information for
Bering Sea ecosystem studies. National Oceanic and Atmospheric Administration
(NOAA) Technical Memorandum National Marine Fisheries Service F/NWC-186.
Reed, J.Z., D.J. Tollit, P.M. Thompson and W. Amos. 1997. Molecular scatology: The
use of molecular genetic analysis to assign species, sex and individual identity to seal
faeces. Molecular Ecology 6:225-234.
Rozen, S. and H.J. Skaletsky. 2000. Primer3 on the WWW for general users and for
biologist programmers. Pages 365-386
in
S. Krawetz and S. Misener, eds.
Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press,
Totowa, NJ.
Springer, A.M., J.A. Estes, G.B. van Vliet, T.M. Williams, D.F. Doak, E.M. Danner,
K.A. Forney and B. Pfister. 2003. Sequential megafaunal collapse in the North Pacific
Ocean: An ongoing legacy of industrial whaling? Proceedings of the National Academy
of Sciences of the United States of America 100:12223-12228.
Taberlet, P. and G. Luikart. 1999. Non-invasive genetic sampling and individual
identification. Biological Journal of the Linnean Society 68:41-55.
Taberlet, P., L.P. Waits and G. Luikart. 1999. Noninvasive genetic sampling: look
before you leap. Trends in Ecology & Evolution 14:323-327.
Thompson, P.M. 1989. Seasonal-Changes in the Distribution and Composition of
Common Seal (Phoca-Vitulina) Haul-out Groups. Journal of Zoology 217:281-294.
Thompson, P.M., A. Mackay, D.J. Tollit, S. Enderby and P.S. Hammond. 1998. The
influence of body size and sex on the characteristics of harbour seal foraging trips.
Canadian Journal of Zoology-Revue Canadienne De Zoologie 76:1044-1053.
Tollit, D.J., M. Wong, A.J. Winship, D.A.S. Rosen and A.W. Trites. 2003. Quantifying
errors associated with using prey skeletal structures from fecal samples to determine the
diet of Steller's sea lion (Eumetopias jubatus). Marine Mammal Science 19:724-744.
VanBlaricom, G.R., L.R. Gerber and R.L.J. Brownell. 2001. Extinctions of Marine
Mammals. Pages 37-69 Encyclopedia of Biodiversity. Academic Press.
71
Wasser, S.K., C.S. Houston, G.M. Koehler, G.G. Cadd and S.R. Fain. 1997. Techniques
for application of faecal DNA methods to field studies of Ursids. Molecular Ecology
6:1091-1097.
WDF, WDW and W.W.T.I. Tribes. 1993. 1992 Washington State Salmon and Steelhead
Stock Inventory (SaSI). Washington Department of Fish and Wildlife.
WDFW and P.-N.-P.T. Tribes. 2000. Summer Chum Salmon Conservation Initiative: An
Implementation Plan to Recover Summer Chum in the Hood Canal and Strait of Juan de
Fuca Region. Washington Department of Fish and Wildlife.
WDFW and P.-N.-P.T. Tribes. 2003. Summer Chum Salmon Conservation Initiative: An
Implementation Plan to Recover Summer Chum in the Hood Canal and Strait of Juan de
Fuca Region. Washington Department of Fish and Wildlife, Supplemental Report
Supplemental Report Number 5.
Williams, T.M., J.A. Estes, D.F. Doak and A.M. Springer. 2004. Killer appetites:
Assessing the role of predators in ecological communities. Ecology 85:3373-3384.
Yurk, H. and A.W. Trites. 2000. Experimental attempts to reduce predation by harbor
seals on out-migrating juvenile salmonids. Transactions of the American Fisheries
Society 129:1360-1366.
72
Appendix A: Prey Species Present in Scats Classified by Sex
Listed is a table of all scats for which a sex classification was possible using the genetic
techniques described in Chapter 3. The sites codes correspond to Quilcene Bay (QB),
Dosewallips (DO), Duckabush (DK), Hamma Hamma (HH) and Skokomish (SK). The
prey species present in the scat are indicated by the first three letters of the genus
followed by the first three letters of the species. When species is not known ‘SPP’
indicates ‘unknown species.’ Prey species were identified from all possible diagnostic
structures. The diet data is provided by Monique Lance and the Washington Department
of Fish and Wildlife (WDFW). These scats are only a subset of all scats collected in
Hood Canal between 1998 and 2003 and any characterization of the overall diet of
harbor seals in Hood Canal should reference the full dataset available from WDFW.
Sample ID Sex Site Year Species Present in Scat
do 99-001 F DO 1999
CLUPAL
do 99-002 F DO 1999
MERPRO
do 99-018 F DO 1999
MERPRO CLUPAL
do 99-020 F DO 1999
LOLOPA MERPRO CLUPAL
do 99-061 F DO 1999
SALSPP CLUSPP CYMAGG
do 99-080 F DO 1999
MERPRO
do 99-087 F DO 1999
OCTSPP MERPRO CLUSPP
do 99-094 F DO 1999
CYMAGG
CLUPAL
do 99-104 F DO 1999
ENGMOR
MERPRO
do 99-105 F DO 1999
MERPRO
do 99-106 F DO 1999
CLUPAL MERPRO
do 99-128 F DO 1999
MERPRO
do 99-130 F DO 1999
MERPRO SALSPP CLUPAL
do 99-136 F DO 1999
CYMAGG
CLUPAL
73
Sample ID Sex Site Year Species Present in Scat
do 99-140 F DO 1999
MERPRO CLUPAL
do 99-142 F DO 1999
MERPRO CYMAGG
do 99-153 F DO 1999
MERPRO CLUPAL
do 99-166 F DO 1999
MERPRO CLUPAL
do 99-167 F DO 1999
CLUPAL
do 99-172 F DO 1999
CYMAGG
MERPRO RHAVAC
do 99-173 F DO 1999
MERPRO
do 99-176 F DO 1999
RHAVAC
do 99-179 F DO 1999
CLUPAL
do 99-193 F DO 1999
MERPRO CLUPAL
do 99-003 M DO 1999
PORNOT MERPRO
do 99-010 M DO 1999
MERPRO
do 99-015 M DO 1999
MERPRO
do 99-019 M DO 1999
MERPRO CLUSPP
do 99-021 M DO 1999
MERPRO CLUPAL
do 99-024 M DO 1999
MERPRO CLUSPP
do 99-027 M DO 1999
MERPRO ENGMOR
do 99-029 M DO 1999
MERPRO
do 99-031 M DO 1999
CLUPAL MERPRO
do 99-032 M DO 1999
CLUPAL MERPRO
do 99-033 M DO 1999
MERPRO MICPRO UNKID
do 99-034 M DO 1999
MERPRO SALSPP
do 99-035 M DO 1999
MERPRO CLUPAL AMMHEX
do 99-038 M DO 1999
MERPRO
do 99-043 M DO 1999
MERPRO
do 99-050 M DO 1999
MERPRO CLUSPP
do 99-064 M DO 1999
MERPRO
do 99-071 M DO 1999
MERPRO
74
Sample ID Sex Site Year Species Present in Scat
do 99-095 M DO 1999
MERPRO
do 99-097 M DO 1999
MERPRO ENGMOR
do 99-102 M DO 1999
ENGMOR
do 99-103 M DO 1999
MERPRO CLUPAL ENGMOR
UNKID
do 99-107 M DO 1999
RAJSPP
do 99-113 M DO 1999
MERPRO CLUPAL ENGMOR
do 99-118 M DO 1999
SALSPP MERPRO
do 99-137 M DO 1999
MERPRO PORNOT SALSPP
do 99-149 M DO 1999
MERPRO
do 99-169 M DO 1999
MERPRO
do 99-174 M DO 1999
CLUPAL MERPRO
do 99-180 M DO 1999
CLUPAL MERPRO
do 99-181 M DO 1999
SALSPP MERPRO
do 99-184 M DO 1999
MERPRO CLUPAL
do 99-186 M DO 1999
MERPRO CLUPAL SALSPP
do 99-188 M DO 1999
SALSPP CLUSPP
do 99-199 M DO 1999
MERPRO SALSPP SEBSPP
do 99-200 M DO 1999
SALSPP MERPRO
do 99-004 M DO 1999
MERPRO CLUPAL
do 99-016 M DO 1999
MERPRO
do 99-022 M DO 1999
MERPRO CLUSPP SALSPP
do 99-030 M DO 1999
MERPRO CLUSPP
do 99-037 M DO 1999
MERPRO
do 99-039 M DO 1999
MERPRO MICPRO SALSPP
do 99-040 M DO 1999
MERPRO
do 99-041 M DO 1999
MERPRO CLUPAL SALSPP
do 99-042 M DO 1999
MERPRO
do 99-044 M DO 1999
MERPRO
75
Sample ID Sex Site Year Species Present in Scat
do 99-054 M DO 1999
MERPRO CLUSPP
do 99-057 M DO 1999
MERPRO CLUSPP
do 99-058 M DO 1999
MERPRO CLUPAL
do 99-062 M DO 1999
LOLOPA MERPRO
do 99-066 M DO 1999
SALSPP
do 99-070 M DO 1999
MERPRO
do 99-075 M DO 1999
MERPRO CLUPAL
do 99-076 M DO 1999
MERPRO
do 99-079 M DO 1999
MERPRO MICPRO CLUPAL
do 99-086 M DO 1999
SALSPP
do 99-091 M DO 1999
MERPRO
do 99-092 M DO 1999
MERPRO CLUPAL
do 99-093 M DO 1999
SALSPP
do 99-100 M DO 1999
MERPRO
do 99-101 M DO 1999
CLUPAL MERPRO
do 99-109 M DO 1999
MERPRO
do 99-114 M DO 1999
CLUPAL MERPRO ALOSAP
do 99-115 M DO 1999
MERPRO
do 99-122 M DO 1999
MERPRO
do 99-124 M DO 1999
MERPRO CLUPAL
do 99-131 M DO 1999
MERPRO
do 99-132 M DO 1999
MERPRO CLUPAL
do 99-139 M DO 1999
MERPRO
do 99-141 M DO 1999
SALSPP
do 99-145 M DO 1999
MERPRO
do 99-146 M DO 1999
MERPRO
do 99-147 M DO 1999
SALSPP
do 99-154 M DO 1999
CYMAGG
76
Sample ID Sex Site Year Species Present in Scat
do 99-157 M DO 1999
CLUPAL MERPRO CYMAGG
do 99-159 M DO 1999
MERPRO CLUPAL
do 99-161 M DO 1999
MERPRO CLUPAL
do 99-171 M DO 1999
OCTSPP MERPRO
do 99-198 M DO 1999
MICPRO CYMAGG
CLUPAL MERPRO SALSPP
do 99-201 M DO 1999
MERPRO CLUPAL
do 99-025 M DO 1999
MERPRO
dk 99-002 F DK 1999
LOLOPA MERPRO MICPRO LEPARM CYMAGG
hh 99-010 F HH 1999
MERPRO PORNOT
hh 99-023 F HH 1999
LOLOPA CLUPAL
hh 99-038 F HH 1999
CLUPAL
hh 99-008 M HH 1999