(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Ecology and conservation of the marbled murrelet"

United States 
Department of 
Agriculture 

Forest Service 

Pacific Southwest 
Research Station 

General Technical 
Report PSW-GTR-1 52 



Ecology and Conservation of the 

Marbled Murrelet 




Abstract: 



Ralph, C. John; Hunt, George L., Jr.; Raphael, Martin G.; Piatt, John F., Technical Editors. 1995. Ecology and 
conservation of the Marbled Murrelet. Gen. Tech. Rep. PSW-GTR-152. Albany, CA: Pacific Southwest 
Research Station, Forest Service, U.S. Department of Agriculture; 420 p. 

This report on the Marbled Murrelet (Brachyramphus marmoratus) was compiled and editied by the 
interagency Marbled Murrelet Conservation Assessment Core Team. The 37 chapters cover both original 
studies and literature reviews of many aspects of the species' biology, ecology, and conservation needs. It 
includes new information on the forest habitat used for nesting, marine distribution, and demographic analyses; 
and describes past and potential effects of humans on the species' habitats. Future research needs and possible 
management strategies for both marine and forest habitats are suggested. 

Retrieval Terms: Brachyramphus marmoratus, Marbled Murrelet, old-growth forests, habitat use, marine 
distribution, seabird. 



About This Report: 



Technical Editors: 

• C. John Ralph, research wildlife biologist, Pacific Southwest Research Station, USDA Forest Service, 
1700 Bay view Drive, Areata, CA 95521 

• George L. Hunt Jr., professor, Department of Ecology and Evolutionary Biology, 321 Steinhus Hall, 
University of California at Irvine, Irvine, CA 92717 

• Martin G. Raphael, chief research wildlife biologist, Pacific Northwest Resesrch Station, USDA 
Forest Service, 3625-93rd Ave. S.W., Olympia, WA 98512 

• John F. Piatt, research biologist, Alaska Science Center, U.S. Department of the Interior, National 
Biological Service, 101 1 East Tudor Road, Anchorage, AK 99503 

Cover: Late-winter-plumaged Marbled Murrelet, Auke Bay, Alaska — Photograph by Gus van Vliet 



Publisher: 



Pacific Southwest Research Station 
Albany, California 

(Mailing address: P. O. Box 245, Berkeley, California 94701-0245 
Telephone: 510-559-6300) 



February 1995 



From the collection of 



J 




k 



International 

Bird Rescue 

Research Center 

Cordelia, California 



in association with 













Prelin 

v Jj 



inger 

a 

ibrary 

p 



San Francisco, California 
2006 



Digitized by the Internet Archive 

in 2007 with funding from 

Microsoft Corporation 



http://www.archive.org/details/ecologyconservatOOpacirich 



3/*£* 



36 



Ecology and Conservation of the 

Marbled Murrelet 



Technical Editors: 

C. John Ralph George L. Hunt, Jr. Martin G. Raphael John F. Piatt 



Contents 



Part I: Introduction 

Chapter 1 3 

Ecology and Conservation of the Marbled Murrelet in North America: An Overview 
C. John Ralph, George L Hunt, Jr., Martin G. Raphael, and John F. Piatt 

Chapter 2 23 

The Asian Race of the Marbled Murrelet 
Nikolai B. Konyukhov and Alexander S. Kitaysky 

Part II: Nesting Ecology, Biology, and Behavior 

Chapter 3 33 

Comparative Reproductive Ecology of the Auks (Family Alcidae) 
with Emphasis on the Marbled Murrelet 
Toni L De Santo and S. Kim Nelson 

Chapter 4 49 

Nesting Chronology of the Marbled Murrelet 
Thomas E. Homer and S. Kim Nelson 

Chapter 5 57 

Nesting Biology and Behavior of the Marbled Murrelet 
S. Kim Nelson and Thomas E. Hamer 

Chapter 6 69 

Characteristics of Marbled Murrelet Nest Trees and Nesting Stands 
Thomas E. Hamer and S. Kim Nelson 



Contents 



Chapter 7 83 

Breeding and Natal Dispersal, Nest Habitat Loss and Implications for 
Marbled Murrelet Populations 
George J. Divoky and Michael Horton 

Chapter 8 89 

Nest Success and the Effects of Predation on Marbled Murrelets 
5. Kim Nelson and Thomas E. Hamer 

Chapter 9 99 

Molts and Plumages in the Annual Cycle of the Marbled Murrelet 
Harry R. Carter and Janet L. Stein 

Part III: Terrestrial Environment 

Section 1. Inland Patterns of Activity 

Chapter 10 113 

Marbled Murrelet Inland Patterns of Activity: Defining Detections and Behavior 
Peter W.C. Paton 

Chapter 1 1 1 17 

Patterns of Seasonal Variation of Activity of Marbled Murrelets in Forested Stands 
Brian P. O 'Donnell, Nancy L. Naslund, and C. John Ralph 

Chapter 12 129 

Daily Patterns of Marbled Murrelet Activity at Inland Sites 
Nancy L. Naslund and Brian P. O 'Donnell 

Chapter 13 135 

Interannual Differences in Detections of Marbled Murrelets in 
Some Inland California Stands 
C. John Ralph 

Chapter 14 139 

A Review of the Effects of Station Placement and Observer Bias in Detections of 
Marbled Murrelets in Forest Stands 
Brian P. O 'Donnell 

Section 2. Inland Habitat Use and Requirements 

Chapter 15 141 

Inland Habitat Suitability for the Marbled Murrelet in Southcentral Alaska 

Katherine J. Kuletz, Dennis K. Marks, Nancy L. Naslund, Nike J. Goodson, and Mary B. Cody 

Chapter 16 151 

Inland Habitat Associations of Marbled Murrelets in British Columbia 
Alan E. Burger 



11 USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Contents 



Chapter 17 163 

Inland Habitat Associations of Marbled Murrelets in Western Washington 
Thomas E. Homer 

Chapter 18 177 

A Landscape-Level Analysis of Marbled Murrelet Habitat in Western Washington 
Martin G. Raphael, John A. Young, and Beth M. Galleher 

Chapter 19 191 

Marbled Murrelet Habitat Associations in Oregon 
Jeffrey J. Grenier and S. Kim Nelson 

Chapter 20 205 

Relationship of Marbled Murrelets with Habitat Characteristics 
at Inland Sites in California 
Sherri L. Miller and C. John Ralph 

Part IV: The Marine Environment 

Section 1. Marine Setting 

Chapter 21 219 

Oceanographic Processes and Marine Productivity in Waters Offshore of 
Marbled Murrelet Breeding Habitat 
George L. Hunt, Jr. 

Section 2. Foraging Biology 

Chapter 22 223 

Marbled Murrelet Food Habits and Prey Ecology 
Esther E. Burkett 

Chapter 23 247 

Marbled Murrelet At-Sea and Foraging Behavior 
Gary Strachan, Michael McAllister, and C. John Ralph 

Chapter 24 255 

Monospecific and Mixed Species Foraging Associations of Marbled Murrelets 
George L. Hunt, Jr. 

Chapter 25 257 

Pollution and Fishing Threats to Marbled Murrelets 
D. Michael Fry 

Chapter 26 261 

Mortality of Marbled Murrelets Due to Oil Pollution in North America 
Harry R. Carter and Katherine J. Kuletz 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. iii 



Preface 



The Marbled Murrelet (Brachyramphus marmoratus) has 
long been regarded as a bird of mystery in the Pacific Northwest 
because its nesting habits have remained largely unknown to 
ornithologists, and its nearshore feeding habits made it difficult 
to survey. This small, dove-sized seabird inhabits coastal 
areas of North America from Alaska to central California. 
Throughout most of its range it nests in forests within about 
25 to 50 miles of the coast, and feeds in nearshore marine 
waters on small fish and invertebrates. In contrast to most 
alcids, which nest colonially on rocky cliffs or relatively 
barren islands, the Marbled Murrelet nests inland throughout 
most of its range in solitary pairs (or perhaps loose 
associations), on the wide, upper branches of old, coniferous 
trees. This retiring habit delayed the discovery of its nest in 
North America until 1974, when one was found in central 
California (Binford and others 1975). Since then, despite 
many thousands of person-days of effort over the past decade, 
fewer than 60 nests have been located through the 1993 
breeding season (Nelson and Hamer, this volume a). 

In the 1980s, field biologists discovered evidence 
suggesting that many, if not most, individuals nest in 
unharvested coniferous old-growth forests. Further research, 
much of it presented for the first time in this volume, has 
provided additional information on habitat use, on their 
relatively low reproductive rates, and on the high predation 
they experience at the nest. 

In at least some areas, evidence also began to accumulate 
that the Marbled Murrelet population has declined in recent 
years. This decline has been attributed to reduction and 
fragmentation of old-growth forests, increased predation, 
pollution (especially oil spills), and mortality from fishing 
nets. This potential decline heightened management sensitivity 
to assure the maintenance of healthy interacting populations 
throughout its range. At present, the murrelet is classified as 
threatened or endangered by the U.S. Fish and Wildlife 
Service in Washington, Oregon, and California, as well as 
by the State of California and the Province of British 
Columbia. For most land management agencies, these listings 
require inventories and analyses of potential impacts of 
proposed projects on the species. If adverse impact on murrelet 
habitat is found, it may result in mitigation measures, project 
modification, delays, and possible cancellation. 

Issues 

Several issues faced land management agencies in the 
United States and Canada in 1992 when the effort on this 
volume began. 



Timber harvest — The legal status of the species was 
beginning to prevent or delay timber harvest activities 
throughout most of its range on the Pacific Coast of North 
America. No forest management standards and guidelines to 
maintain murrelet habitats existed, because documentation 
of the full range of the species' habitat was unknown. 

Survey and monitoring efforts — Surveys to determine 
the species' presence or absence in forest stands throughout 
its range required substantial financial and personnel resources. 
Due to a lack of knowledge of its distribution and abundance, 
costly efforts often included surveys in areas that were 
unsuitable or of marginal value to the species. 

Other resources — It seemed probable that the species 
occupied habitats containing large amounts of economically 
valuable timber. 

These stands also functioned as reservoirs of biological 
diversity, and had great values as watersheds and as sources 
of a variety of wildlife and fishery resources. While at sea, 
the bird coexisted with large numbers of commercially im- 
portant fish, especially salmon, the harvesting of which may 
result in significant murrelet mortality. 

Consolidation of information — It was apparent that a 
need existed to consolidate available information, and to 
synthesize knowledge of population trends, distribution, 
habitat associations, and potential management alternatives. 
The U.S. Fish and Wildlife Service appointed a Marbled 
Murrelet Recovery Team early in 1993 to determine the 
status and mode of recovery of the species. They needed a 
rapid production of scientific background material for 
their deliberations. 

Goals of the Assessment 

To meet these issues, the USDA Forest Service began a 
"Marbled Murrelet Conservation Assessment" in late 1992 
with the following mandate. The Assessment would 
consolidate the available information concerning Marbled 
Murrelet ecology and evaluate current habitat conditions to 
determine the likelihood of long-term persistence of healthy 
populations throughout its current range. The Assessment 
would include monitoring and research recommendations, be 
a primary source of information for the Recovery Team, and 
provide information that would enable agencies to make 
management plans. 

This work would be accomplished by the following 
methods: 

1. Identify patterns of habitat use in the forests and 
marine environments occupied by the murrelet, and develop 
an understanding of the spatial and temporal dynamics of 
these habitats and murrelet populations, by using a compilation 
of existing survey data. 

2. Summarize and synthesize existing information from 
throughout the range about the life history, status, and trends 
of the murrelet and its utilized habitats, and provide the 
information gathered to all interested parties. 



vi 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



3. Identify additional inventory needs and methodology 
to facilitate statistically meaningful long term monitoring of 
both the species and its habitats, thus providing the information 
needed to develop sound strategies to provide for their 
maintenance and management. 

4. Identify additional research needs to fill information 
gaps preventing a full understanding of Marbled Murrelet 
ecology. 

5. Provide suggestions to improve the compatibility of 
data bases maintained by various entities. 

Organization 

The Assessment effort was organized into a set of working 
groups as follows: 

• Interagency Conservation Assessment Coordinating 
Group — The intent of this group was to coordinate and provide 
support to Conservation Assessment activities among the 
state, provincial, and federal agencies with Marbled Murrelet 
management responsibilities. These agencies and organizations 
were invited to participate by the two Group Leaders: Garland 
N. Mason, Pacific Southwest Research Station, Albany, 
California; and Hugh Black. Pacific Northwest Region, 
Portland, Oregon — both with the USDA Forest Service. 

• Conservation Assessment Core Team — The Core 
Team was headed by a Team Leader (C.J. Ralph), provided 
by the Pacific Southwest Station, and three senior scientists 
with established expertise in various aspects of ecology 
who, drawing on the knowledge provided by the Technical 
Working Group, provided the scientific expertise to formulate 
the Conservation Assessment. The Team Leader provided 
the overall technical and administrative leadership for 
assessment development and ensured good communication 
between the Coordinating Group, the Core Team, and the 
Technical Working Group. The scientists in the Core Team 
became the technical editors of the final volume. 

• Conservation Assessment Technical Working 
Group — This group was open to all persons with knowledge 
or abilities that could contribute to the formulation of the 
Conservation Assessment (see Appendix A in this volume), 
and provided the following functions: 

• Collected and provided technical information 
required by the Working Group. 

• Wrote chapters of the Assessment, as appropriate. 

• Provided assistance, advice, and input to other 
members of the Working Group as requested. 

• Informed respective agencies, organizations, or 
regions as to progress and findings of the 
Conservation Assessment. 

• Provided expertise to formulate inter-regional 
assessments. 

• Identified and overcame obstacles to gathering 
information for the Assessment. 



Members of the Working Group included: 

• Marbled Murrelet specialists from universities, 
agencies, private industry, and conservation 
organizations. 

• Regional representatives from USDA Forest Service 
Regions in Alaska, Washington, Oregon, and 
California. 

• Agency Representatives from three U.S. Department 
of the Interior agencies — Fish and Wildlife Service, 
National Biological Service, National Park Service — 
and Canadian Wildlife Service, among others. 

• Representatives from state and provincial fish and 
wildlife agencies not represented above. 

• Specialists from various disciplines useful to the 
process of the Assessment. 

• Line officers. 



Financial assistance was provided by various agencies 
and organizations, acknowledged in each chapter, and also 
by the Assessment itself that provided certain members of 
the Technical Working Group with funds to enable them to 
analyze their data in a more timely manner than would have 
been possible in the normal course of events. 

Working Environment 

Working sessions of the Core Team and the Working 
Group were open to all persons interested in the proceedings, 
with the Team Leader acting as chair. 

Working Group members participated fully with the 
Core Team and participated in all decisions. The Core Team 
provided direction and strived for consensus among the 
Team and Group members. Minority reports were possible 
and encouraged. Wildlife Society standards for authorship 
were used. In the final stages of compilation of the volume, 
the technical editors met and reviewed chapters which were 
then sent to authors for final approval of all contents. 

Products 

The primary product of the Assessment is this volume. 
Each chapter in the volume was reviewed by numerous 
researchers and biologists in appropriate fields, as well as by 
the Core Team. In addition, the entire document was reviewed 
by four persons appointed by the Presidents of learned 
societies: The Wildlife Society (David Marshall). American 
Ornithologists Union (Peter Conners), Ecological Society of 
America (Frank A. Pitelka), and the Cooper Ornithological 
Society (Douglas Bell). 

The report is organized into chapters addressing the 
various aspects of Marbled Murrelet biology and provide 
data and analyses. Some general management considerations 
are offered in the overview chapter, and are intended to 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



vu 



supplement those offered by the Recovery Team, appointed 
by the USDI Fish and Wildlife Service. 

Acknowledgments 

We express our appreciation to all the reviewers, members 
of the Technical Working Group, and the authors, who worked 
so smoothly together to assemble this compendium of 
knowledge of the murrelet. Behind the scenes, employees of 
the Pacific Southwest Station's Redwood Sciences Laboratory 
did the lion's share of the work in first assembling the data, 
and preparing the manuscripts. Sherri Miller, Deborah Kristen, 
Ann Buell, Tina Menges, Jennifer Weeks, Brian Cannon, 
Robin Wachs, Kim Hollinger, Jim Dahl, Brian O'Donnell, 
and Michelle Kamprath worked tirelessly in the "Murrelet 
House" in downtown Areata during 1993 to enable the authors 
to publish their data. John Young and Beth Galleher, Pacific 
Northwest Station, USDA Forest Service, Olympia, 
contributed to GIS data assembly and analysis. Garland 



Mason, Mike Lennartz, and Barry Noon were very supportive 
of the entire effort, and we are grateful to them. 

The final manuscripts were edited by technical 
publications editors B Shimon Schwarzschild, Sandra L. 
Young, and Laurie J. Dunn; and the layouts were designed 
and produced by visual information specialists Kathryn 
Stewart and Esther Kerkmann — all of the Pacific Southwest 
Research Station. 

Finally, we acknowledge the herculean effort that Linda 
Long provided at all stages of the manuscript preparation, as 
she directed all of us towards producing an excellent product. 

We hope that this effort will serve well the bird and the 
people charged with its management. Most importantly we 
dedicate this volume to the biologists who have spent so 
many cold, lonely, but exhilarating hours in pursuit of this 
sprightly, energetic bird, both on the ocean and in the forest, 
where it turns into a hurtling, small, dark shadow, as it 
enters the primeval forest in pursuit of its largely still 
mysterious habits. 



C. John Ralph George L. Hunt, Jr. 



Technical Editors: 
Martin G. Raphael John F. Piatt 



vni 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



I 



Introduction 




Chapter 1 

Ecology and Conservation of the Marbled Murrelet in 
North America: an Overview 



C. John Ralph 1 



George L. Hunt, Jr. 



Martin G. Raphael 3 



John F. Piatt 4 



Abstract: Over the past decade, the Marbled Murrelet has become 
a focus of much controversy. It was listed as threatened in Wash- 
ington. Oregon, and California by the U.S. Fish and Wildlife 
Service in February 1993. In order to aid the various agencies with 
management, the Marbled Murrelet Conservation Assessment was 
formed to bring together scientists, managers, and others to gather 
all the available data on this small seabird. This volume of research 
is the culmination of that effort. In this chapter, we integrate the 
results of the investigations and summaries on the past history, 
present status, and possible future of the species, based on the data 
presented in this volume and other published research. We also 
propose what we consider the most important research needs. 
Then, based on the findings of this volume, we suggest actions for 
management to help ensure the survival of the species. 



The recent decline and fragmentation of Marbled Murrelet 
(Brachyromphus marmoratus) populations in the southern 
portion of its range (California, Oregon, and Washington) 
resulted in an awareness that the species was in need of 
protection or it risked extirpation. In 1982 and 1986, the 
Pacific Seabird Group developed a set of resolutions that 
called attention to the Marbled Murrelet and the threats it 
faced. The Group requested that the appropriate agencies 
involved in management decisions consider research about 
the species. The response from the agencies was muted at 
best. On January 15, 1988, the National Audubon Society 
petitioned the U.S. Fish and Wildlife Service to list the 
California. Oregon, and Washington populations of the species 
as a threatened species. The Service's 90-day finding stated 
that the petition had presented substantial information to 
indicate that the requested action may be warranted. It was 
published in the Federal Register on October 17, 1988. Because 
of increased research efforts and the amount of new data 
available, several public comment periods were opened to 
receive additional information on the species and the potential 
threats to it. On the basis of the positive 90-day finding, the 
Marbled Murrelet was added to the Service's Notice of Review 
for Vertebrate Wildlife as a Category 2 Species for listing. 



1 Research Wildlife Biologist. Pacific Southwest Research Station. 
USD A Forest Service. Redwood Sciences Labratory.1700 Bay view Drive, 
Areata CA 95521 

: Professor, Department of Ecology and Evolutionary Biology, Univer- 
sity of California. Irvine, CA 92717 

3 Chief Research Wildlife Biologist, Pacific Northwest Research Sta- 
tion. USDA Forest Service. 3625 93rd Ave.. Olympia, WA 98512-9193 

4 Research Biologist. Alaska Science Center, U.S. Department of the 
Interior. National Biological Service, 1011 East Tudor Road, Anchorage, 
AK 99503 



In 1990, the Marbled Murrelet was proposed as a 
threatened species by the British Columbia Ministry of 
Environment, Lands, and Parks to the Committee on the 
Status of Endangered Wildlife in Canada. The species was 
designated as nationally threatened in June 1990. A recovery 
team was established in September of that year and was 
unique to Canada because it included representatives of both 
the federal and provincial governments, the forest industry, 
environmental non-governmental organizations, and 
academia. The species was listed as threatened mainly because 
of loss of nesting habitat, but also because of fishing-net 
mortality and the threat of oil spills. 

In 1991, the State of California listed the species as 
endangered because of the loss of older forests. On June 20, 
1 99 1 , the U.S. Fish and Wildlife Service published a proposed 
rule in the Federal Register to designate it as a threatened 
species in Washington, Oregon, and California. The main 
reason for listing was the loss of older forest nesting habitat. 
Secondary threats included loss due to net fisheries and the 
potential threat of oil spills. In July 1992, the U.S. Fish and 
Wildlife Service published another notice in the Federal 
Register announcing a 6-month extension for determining 
the status of Marbled Murrelets. However, the Service was 
taken to court for not meeting the legal time frames provided 
for in the Endangered Species Act and, in September 1992, 
published a final rule in the Federal Register, listing the 
Marbled Murrelet as a threatened species in the three States. 
A recovery team was established in February 1993 and is 
now in the final stages of a recovery plan for the three-State 
area (U.S. Fish and Wildlife Service, in press). 

The State of Washington is now reviewing a recommen- 
dation to classify the Marbled Murrelet as a threatened species. 
To date, the Marbled Murrelet has not been recommended 
for listing in Oregon. 

This chapter reviews the results of published research 
and new investigations presented in this volume, discusses 
the likely future of the species and its habitat in North 
America, and outlines the actions considered necessary to 
maintain viable populations. 

Background and Assessment of 
Available Information 

Distribution and Habitat 

Summary — Marbled Murrelets in North America occur 
from the Bering Sea to central California. During the breeding 
season, the majority of murrelets are found offshore of late 
successional and old-growth forests, located mostly within 60 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



km of the coast. In forests, most nest sites are on large 
diameter, often moss-covered, limbs. The small, relict 
populations at the limits of the range are particularly vulnerable 
to extirpation, and will require careful stewardship if they 
are to be preserved. At sea, foraging murrelets are usually 
found as widely spaced pairs. In some instances, murrelets 
join into flocks that are often associated with river plumes 
and currents. These flocks may contain sizable portions of 
local populations. Protection of foraging habitat and foraging 
murrelets will be necessary if adult mortality is to be minimized. 

Marbled Murrelets are secretive on land, but spend most 
of their lives at sea, where they are relatively easily observed. 
Data obtained at sea are at present the best source of 
information about the distribution and abundance of the 
species. Patterns of distribution provide information on the 
murrelet's geographic range, terrestrial nesting habitats, and 
the oceanographic features of foraging areas. 

Nest sites of the species were found only relatively 
recently. We can find no historical account that gives any 
credibility to the notion that the murrelet could nest in trees, 
although Dawson (1923) mentions (and then debunks) an 
apocryphal Indian account of them nesting inland in "hollow 
trees." Today, this seems easily interpretable as large, old 
trees containing hollows. In 1923, Joseph Grinnell (quoted 
in Carter and Erickson 1988) noted indirect evidence that the 
bird was associated with older forests. Since then, observers 
have noted links of the species with what has come to be 
called "old-growth" forests, that we define here for 
convenience as forests that have been largely unmodified by 
timber harvesting, and whose larger trees average over 200 
years old. This definition of old-growth is in general agreement 
with the ideas of Franklin and others (1986). In some places 
in this chapter we refer to old-growth trees as those with a 
diameter of more than 8 1 cm. 

In the following chapters, various authors discuss how a 
shift from efforts to find nest sites to broader surveys 
monitoring the presence of murrelets in forested tracts, 
especially those slated for timber harvest, have increased the 
knowledge of the use of inland sites by murrelets. These 
efforts have resulted in a more complete picture on current 
distribution and abundance which may lead the way for 
management for this species. 

Marine surveys remain the only method for estimating 
the size of Marbled Murrelet populations. These surveys have 
been carried out in a variety of intensities, and the most recent 
data are presented in the chapters to follow. Unfortunately, 
relatively little historical survey information is available. Early 
surveys were focused on species found in deeper waters, 
while the nearshore murrelet was generally ignored. Further, 
recent work has shown that to obtain useful data on murrelet 
distribution and abundance, surveys must be designed to 
focus on the nearshore waters where murrelets are found. 

Taxonomy and Range 

The species has been divided into two races, the North 
American (Brachyramphus marmoratus marmoratus) and 



the Asian (B. m. perdix). Recent evidence, not yet fully 
published in the literature (Friesen and others 1994a), strongly 
indicates that the North American race may be more distinct 
from the Asian race (referred to as the Long-billed Murrelet, 
B. perdix) than it is from the other North American 
Brachyramphus, the Kittlitz's Murrelet (B. brevirostris). 
Konyukhov and Kitaysky (this volume) contrast the Asian 
and North American races. 

From California to Alaska, the Marbled Murrelet nests 
primarily in old-growth coniferous forests and may fly up to 
70 km or more inland to nest. This is a radical departure 
from the breeding behavior of other alcids, but adaptation to 
old-growth conifers probably occurred early in its evolutionary 
history, perhaps in the mid-Miocene when enormous dawn 
redwoods (Metasequoia) blanketed the coast from California 
to the north slope of Alaska and Aleutian Islands. The other 
21 extant species of the family Alcidae, known as auks or 
alcids, breed on the ground, mostly on predator-free islands. 
In Alaska, a very small proportion of the Marbled Murrelets 
breed on the ground, usually on barren, inland slopes and to 
the west of the major rain forests along the Alaskan gulf 
coast. Initial divergence of perdix and marmoratus occurred 
in the mid-Pliocene, perhaps as cooling temperatures 
eliminated coastal old-growth forests in the exposed Aleutian 
Islands, leading to a gap in east- west distribution of murrelets 
and isolated breeding stocks (Udvardy 1963). The divergence 
of Marbled and Kittlitz's murrelets occurred at the onset of 
the Pleistocene (Friesen and others 1994), and the present 
strong association of Kittlitz's Murrelet with glacial ice 
clearly indicates the importance of the glacial landscape in 
determining the northeasterly distribution of Kittlitz's Murrelet 
and ecological segregation of brevirostris and marmoratus 
into subarctic and boreal species. 

Geographic Range 

At the broad scale, the distribution of the Marbled 
Murrelet is fairly continuous from the Aleutian Islands to 
California. The present geographic center of the North 
American populations is found in the northern part of southeast 
Alaska (fig. 1). Large populations are also found to the west 
around Prince William Sound and the Kodiak Island 
archipelago, and to the south along the British Columbia 
coast. In either direction, populations become more disjunct, 
with small, discrete sub-populations at the extreme ends of 
the range in the Santa Cruz Mountains of central California, 
and on Attu Island in the western Aleutians. In California, 
Oregon, and Washington, gaps in distribution between 
breeding populations may result largely from timber harvest 
practices. The disjunct distribution is a reflection of the 
remaining nesting habitat, primarily late-successional and 
old-growth forests on public land (Carter and Erickson 1992, 
Leschner and Cummins 1992a, Nelson and others 1992). 

The small, relict populations of murrelets at the limits of 
the species' range are particularly vulnerable to extirpation. 
Particular care will need to be exercised if they are to be 
conserved. Murrelets range along 4,000 km of coastline and it 
is possible that some populations have distinct genetic 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



c 



Upper Cook 
Inlet 



Prince William Sound 



Bering Sea 



I* 



J? 



\s- 



Kodiak Is. 


Alexander ^Kfc; 
Arch i pelago^K S&V 


Gu/f of Alaska 


British^P ; 

Columbia ^"fe 




Washington 




Oregon 



NUMBER OF BIRDS 
< 5,000 
''■ 5,000 - 10,000 
25,000 - 50,000 
50,000 - 100,000 



California ^ 



Figure 1 — Range of the Marbled Murrelet, which stretches from central California to southern Alaska, and 
population size along sections of the coast. See table 2 for further details. 



characteristics allowing for adaptation to variability in these 
environments. As an example, the waters between California 
and the Aleutian Islands are partitioned into several dramatically 
different regimes (Hunt, this volume a). The loss of these 
peripheral populations would likely reduce diversity in the 
population as a whole, and might reduce the capacity of the 
species to adapt to long-term environmental changes. 

Distribution in Relation to Nesting Habitat 

During the breeding season, the distribution of the Marbled 
Murrelet throughout its range is determined by the distribution 
and accessibility of old-growth and late-successional 
coniferous forests. Some evidence exists of a relationship 
between the estimates of Marbled Murrelet population size. 



based on at-sea surveys, and the amount of old-growth forest 
within a region. This relationship is most evident from 
California to southern Washington, a coastline that is relatively 
straight and contains disjunct pockets of old-growth forests. 
In this region, the largest concentrations of murrelets at sea 
during the breeding season are found along sections of coastal 
waters that are adjacent to inland breeding areas (Nelson and 
others 1992, Sowls and others 1980). Marine productivity is 
high along this entire coast during summer (Ainley and 
Boekelheide 1990), and access to suitable foraging areas 
does not appear to limit murrelet distribution. Circumstantial 
evidence is considerable that murrelet distribution is limited 
by nesting, rather than foraging, habitat. For example, 
murrelets concentrate offshore from old-growth areas during 



L'SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



the breeding season (April-August), but move elsewhere 
when not breeding, presumably in response to food 
availability, which becomes more problematic during winter. 
Murrelets do, however, have the ability to fly long distances 
to reach suitable foraging habitat or areas with high 
productivity, even during the breeding season. 

In northern Washington, British Columbia, and Alaska, 
the small-scale relationship between the at-sea distribution 
of murrelets and the presence of old-growth immediately 
adjacent to the coast is less clear. In this part of the murrelet' s 
range, the coastline is much more complex. The numerous 
islands, bays, fjords, and sheltered inside waters, the greater 
abundance of contiguous stands of mature, old-growth 
forests, and the lack of survey effort, all have hindered 
assessment of fine-scale spatial associations between nesting 
and foraging habitat. 

Inland, murrelets are detected almost exclusively in forest 
stands with old-growth characteristics (Burger, this volume 
a; Grenier and Nelson, this volume; Hamer, this volume; 
Kuletz and others, this volume; Paton and Ralph 1990; Rodway 
and others 1993b). All murrelet nests, south of Alaska, have 
been found in old-growth trees (>81 cm d.b.h.), therefore all 
nests have been in stands with old-growth trees. To our 
knowledge, essentially all stands with birds flying below the 
canopy (termed "occupied behaviors") have also been in 
stands with old-growth trees. Grenier and Nelson (this volume) 
found all occupied sites had at least one old-growth tree per 
acre. There are reports of possibly occupied inland sites in 
Oregon without old-growth trees, but Nelson (pers. comm.) 
had not verified occupancy in most of these areas. By contrast, 
there is a high probability that a few murrelets are nesting in 
coastal stands without old-growth trees in the Sitka spruce/ 
western hemlock (Picea sitchensislTsuga heterophylla) forest 
type in Oregon (Nelson, pers. comm.). This forest type may 
provide nesting habitat at younger ages because trees grow 
fast in this area and smaller trees may also be used because 
mistletoe deformations are abundant in the hemlock trees. 
Young Douglas-fir (Pseudotsuga menziesii) forests do not 
provide the same opportunities. 

Ground nesting by Marbled Murrelets has been 
documented in Alaska. Available information suggests that 
less than 5 percent of the total murrelet population in 
Alaska breeds on the ground in non-forested habitat in the 
western Gulf of Alaska and in the Aleutian Islands 
(Mendenhall 1992). There is also a small unknown 
percentage of the population that nests on the ground in 
old-growth forests; about five nests have been found to 
date (Kuletz, pers. comm.). It is important to recognize that 
despite these markedly different breeding habits, intermediate 
situations are generally not acceptable to murrelets. To our 
knowledge they do not breed in alpine forests, bog forests, 
scrub vegetation, scree slopes, and very rarely breed in 
second growth (e.g., trees <81 cm d.b.h.) (Rodway and 
others 1993b). In the farthest northern portion of the range 
in Alaska (Kuletz and others, in press; Naslund and others, 
in press), and in the habitat of the Asian taxon of the 



murrelet (Konyukhov and Kitaysky, this volume), the nest 
trees become relatively short in stature, as compared to 
trees in forests farther south in North America. In these 
areas, murrelets appear to nest in the largest trees of the 
oldest forests. On the basis of all the information available, 
we conclude that throughout their range in North America, 
the great majority of murrelets are strongly associated with 
old-growth forests for breeding. 

Distribution in Relation to Distance from the Coast 

The maximum distance that murrelets can occur inland 
from coastal foraging areas may result from several factors, 
including suitability of climate, availability of nesting habitat, 
the maximum foraging range, and rates of predation. Average 
and maximum summer temperatures increase as a function 
of distance from the coast and the decreased influence of 
cool maritime breezes. For a well-insulated, oceanic species 
spending more than 95 percent of its time on the cold waters 
of the Pacific, inland temperatures in the south of its range 
could be too hot for nesting. Greater distances to the coast 
would also require longer foraging flights. For other species 
of alcids, typical one-way foraging ranges are 10-40 km, 
with maximum extremes of 100-150 km (Ainley and others 
1990; Bradstreet and Brown 1985). For murrelets, studies of 
foraging range using radio-tagged birds have indicated that 
this species will fly up to 75 km from its nesting areas to 
forage, with most trips being considerably shorter (Burns 
and others 1994, Rodway and others, in press). The maximum 
distance inland at which murrelets have been found is about 
100 km although most appear to nest less than 60 km inland 
(Hamer, this volume; Miller and Ralph, this volume). Records 
for maximum inland distance based on the discovery of 
grounded fledglings may be misleading because of the 
possibility of misdirected birds flying inland from their nest. 
Average distances of inland nesting cannot be firmly 
ascertained until the distribution of inland detections of 
murrelets is documented with a consistent survey effort. We 
do not know how the potential for nest site predation may 
vary with distance from the coast, but certainly longer flights 
between the nest sites and at-sea foraging areas increase the 
chance of being taken by aerial predators. 

Although in some regions murrelets nest immediately 
adjacent to the coast, in most portions of their range studied 
the majority of nests are inland from the immediate coast. In 
Alaska, murrelets nest within 1 km of salt water (Kuletz and 
others, this volume; Naslund and others, in press), and in 
California the highest proportions of nesting stands are found 
within 10 km of the coast (Miller and Ralph, this volume). 
At least in the southern part of the range, we suspect that the 
readily-harvested trees on the coast were the first to be 
removed, leaving the more distant ones for future cutting 
and thereby influencing current patterns of murrelet nesting. 

Comparison of Habitat Correlates 

Several studies and surveys have documented behaviors 
at inland stands that are probably indicative of nesting (Nelson 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



and Hamer, this volume a; Nelson and Peck, in press); see 
Paton (this volume) for details on the survey method and its 
caveats. We compared for each region (table 1), the site, 
stand, and tree attributes that have been shown to be correlated 
with nesting behavior. Among the most consistently observed 
of these attributes are the presence of large diameter conifer 
trees and associated nest platforms and covering limbs. The 
use or presence of large diameter conifers is pervasive 
throughout the studies in this volume. Occasional sites with 
only a few old-growth trees have been found in Oregon to 
have murrelets present (Nelson, pers. comm.). Further, as 
Grenier and Nelson (this volume) point out, stand structure 
is probably more important than stand age itself. However, 
as stands mature, they generally gain the characteristics 
necessary for nesting. These observations support the idea 
that it is the presence of adequate nesting platforms that 
defines suitable nesting habitat. The species of conifer is less 
important than its structural ability to support nest platforms 
(Burger, this volume a). 

Limiting Factors and Relative Importance in Old-Growth 
Several factors appear related to the preference for 

old-growth, including temperature, predation, stand and tree 

structure, and accessibility. 

Temperatures in old-growth forests are lower than in 

open, second-growth areas. This may be very important for a 



thickly-feathered species primarily adapted for diving for 
food in cold ocean waters. 

Old-growth stands may also provide more protection 
from inclement weather by providing greater cover around 
branches. 

Predation apparently has a pervasive influence on murrelet 
reproductive success, as we detail below. Nelson and Hamer 
(this volume b) found that most studies of avian predator 
abundance or influence support the idea that modified forests 
have higher predator populations than older, undisturbed 
forests. In contrast, Rosenberg and Raphael (1986) found 
that predator populations were not greater in second growth, 
as compared to old-growth forests. Although more work 
needs to be done on this issue, it seems likely that predation 
could well be a principal limiting factor for selection of 
nesting habitat and reproductive success. 

The presence of old-growth in an area does not assure 
sufficient substrate for nesting. Though old growth appears to 
be a necessary condition, some old-growth stands may have 
relatively few deformed or broad-limbed trees, possibly limiting 
the availability of nest sites. The physical condition of a tree 
appears to be the important factor in determining its suitability 
for nesting. Specifically, the murrelet, a bird with high wing 
loading, prefers high and broad platforms for landing and 
take-off, and surfaces which will support a nest cup (see 
Hamer and Nelson, this volume b). Accessibility of the stand, 



Table 1 — Site, stand, and tree attributes important to Marbled Murrelets 



Region 


Important attributes 


Source, this volume 


Sesting stands in: 






Alaska 


Epiphyte cover, nesting platforms, 
large diameter trees, old-growth forests. 


Knletz and others 


British Columbia 


Old-growth forests, low elevation, large trees. 


Burger (a) 


Washington 


Old-growth forests, stand size, large sawtimber. 


Raphael and others 




In old-growth forests: nest platforms. 


Hamer 




moss cover, slopes, stem density, large 






d.b.h.. western hemlock, low elevation. 






lack of lichen, low canopy cover. 




Oregon 


Older forests and large diameter trees. 


Grenier and Nelson 


California 


Density of old-growth trees, lower 


Miller and Ralph 




elevation, topography, redwood. 


Paton and Ralph 1990 


Nest trees in: 






All areas 


Large diameter, old-growth forests, and 
decadent trees with mistletoe, deformations, 
and moss on limbs. 


Hamer and Nelson (b) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



either as a function of the distance from the coast, or the 
thickness of growth limiting the ability of birds to fly into the 
stand, is also likely a factor involved in nest site selection. 

Stand size has been suggested to be an important factor 
in the abundance of birds in a stand (Paton and Ralph 1990) 
and in the likelihood that a stand will be occupied (Raphael 
and others, this volume). Larger stands can contain more 
birds overall, but there is no evidence that density changes 
as a function of stand size (Miller and Ralph, this volume). 
Stand size, however, is probably importantly related to some 
of the other factors mentioned above, especially predation. 

Potential Biases in Determining Nest Sites 

Observers have usually chosen sites for nest search in 
areas with larger trees. Data from stratified samples for 
presence of birds in inland stands in California (Miller and 
Ralph, this volume), in Oregon (Nelson 1990), and in Alaska 
(Kuletz and others, this volume), as well as nests discovered 
from radio-tracking (Hamer and Nelson, this volume b), 
have probably been the most free of bias. Since these more 
randomly located sites have not differed markedly from the 
nest sites found in search areas that were potentially biased 
by the choice of area to be searched, we conclude that the 
documented nest sites described in most studies are 
representative of habitat selection by the species. 

Sources of Error in the Determination of Forest Use 

The most direct evidence of murrelet breeding is the 
finding of a nest, but nest detections are rare, due to the 
secretive nesting behavior of murrelets. Also, as Hamer and 
Nelson (this volume b) point out, locating nests is greatly 
dependent upon where observers have looked, making the 
habitat characteristics of nest sites subject to this bias. The 
majority of the conclusions about murrelet use of habitats 
relies upon detections of birds that have flown inland to 
presumed nesting areas (Naslund and O'Donnell, this volume; 
O'Donnell and others, this volume; Paton, this volume). 
Observations have been divided into two groups, those of 
birds flying over the canopy and those at or below the 
canopy level (Ralph and others 1993). It is suggested that the 
latter behavior (referred to as "occupied" behavior) is a 
strong indication that breeding is occurring in the stand, as 
this behavior is almost entirely restricted to the breeding 
season (O'Donnell, this volume). 

We believe that the most objective method of 
determining habitat relationships is the detection of birds in 
the forest. Detections are usually within a 100-m-radius 
circle surrounding the observer, and provide a larger and 
less potentially biased sample than the location of nest 
trees. Below-canopy behavior has been observed in the 
vicinity of known nest trees. Despite the lack of data 
demonstrating that this behavior occurs only when a pair is 
nesting or prospecting, we suggest that the presence of this 
behavior is a strong indication that murrelets are nesting or 
intending to nest in a given stand. Stands where murrelets 
exhibit this behavior should be treated as if they contain 



nesting murrelets. Circling above the canopy is also thought 
to be associated with nesting murrelets (e.g., Nelson and 
Hamer, this volume; Nelson and Peck, in press). Other 
species of alcids often circle above the breeding grounds as 
part of their social interactions. However, as Divoky and 
Horton (this volume) argue, the possibility that non-breeding, 
dispersing young could be prospecting in marginal stands, 
thus distorting the value of these stands to the observers, 
cannot be dismissed. 

Seasonality of the murrelets' visits may affect efforts to 
establish use of a stand. O'Donnell and others (this volume) 
describe the seasonal timing of forest visits, showing the 
peak of activity to be during the period of April through 
July, with peaks of activity in the more northern parts of the 
range occurring later in the summer. Naslund (1993) suggested 
that the winter visits of murrelets to stands, even though 
little or no below-canopy behavior is observed, might be a 
better indicator of nesting than those during the breeding 
season. However, we feel that until more compelling evidence 
is available, stand use during the breeding season should 
remain the criterion of breeding for management purposes, 
as suggested by Ralph and others (1993). 

Many land managers depend upon the protocol developed 
by the Pacific Seabird Group (Paton and others 1990; Ralph 
and others 1993) to determine if murrelets are present in 
their forests. The basis of the timing and frequency of the 
surveys has depended upon a firm foundation of research as 
summarized in O'Donnell and others (this volume), Naslund 
and O'Donnell (this volume), and O'Donnell (this volume). 
Active research and statistical analyses are underway to 
validate the method and to determine the number of surveys 
necessary to establish birds as breeding in a stand, and how 
many years of survey are necessary. At issue is the possibility 
of interannual variation in occupancy of a site that requires 
protection. Ralph (this volume) found no significant 
differences among years in detection levels at three sites in 
California during years when there was a range of sea 
temperature conditions, with both El Nino (the periodic 
warming of ocean waters) and non-El Nino years during the 
study. However, Nelson (pers. comm.) suggests that her data 
show consistent differences among these same years in 
Oregon. Burger (this volume a) also found higher inland 
detection rates with normal sea temperatures, and lower 
detections with high sea temperatures. Additional work needs 
to be done to determine if differences in offshore conditions, 
resulting in changes in food abundance and perhaps breeding 
frequency, are reflected in inshore detections during the 
breeding season. 

Local Distribution at Sea and Foraging 

Concentrations 

Patterns in the distribution of Marbled Murrelets at sea 
can be seen both at large scales (hundreds of kilometers) and 
at small scales of individual aggregations. The small-scale 
distribution of Marbled Murrelets at sea reflects their choice 
of foraging habitat. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



Although murrelets are often encountered as widely 
dispersed pairs, in some instances they gather into flocks 
that may contain a significant fraction of the local population 
(Strachan and others, this volume). Murrelets most often 
form flocks in the sheltered waters of Washington, British 
Columbia, and Alaska (Carter and Sealy 1990; Kaiser and 
others 1991; Piatt and Naslund, this volume; Prestash and 
others 1992), but they also occasionally aggregate along the 
open coasts of California (Ralph and Miller, this volume), 
and Oregon (Strong and others 1994). Information about 
where murrelets are likely to concentrate at sea is relevant to 
the prediction of where murrelet populations are likely to be 
particularly vulnerable to bycatch in gill nets, a local oil 
spill, other pollution event, or disturbance from good feeding 
areas by boat traffic (Kuletz 1994). Protection of these areas 
of aggregation may be important in reducing anthropogenic 
sources of adult mortality. 

There are relatively few data on the distribution of 
murrelet aggregations or their frequency. Several authors 
have noted the correspondence between murrelet 
distribution and certain physical processes in the ocean. 
For instance, some observations indicate that along the 
open coasts, aggregations may be more frequent in the 
vicinity of river plumes (Strong and others, this volume; 
Varoujean and Williams, this volume), although old-growth 
stands may also be most numerous in river valleys, thus 
confounding the cause of the aggregations. In the bays and 
sounds of Washington. British Columbia, and Alaska, 
aggregations of murrelets are common, but little is known 
about the environmental conditions causing these 
concentrations. Piatt and Naslund (this volume) suggest 
that murrelets prefer stratified as opposed to well-mixed 
waters, but they also report that murrelets often concentrate 
near the outflows of large rivers and in rip tides. Burger 
(this volume b) reviewed the available data for British 
Columbia, and found equivocal evidence linking densities 
with water temperature. Kaiser and others (1991) found 
some correlations with temperature which they attributed 
to the effects of local tidal rips. Also there were instances 
where murrelets aggregated at some tidal rips and upwelling, 
but they were scarce or absent at other tidal rips where 
other species aggregated. 

Murrelets have also been associated with particular marine 
habitats that are favored by prey, such as sand lance 
(Ammodytes hexapterus), and surf smelt (Hypomesus 
pretiosus). Burger (this volume b) suggested that murrelets 
aggregate in shallow bays of fjords, in estuaries, and off 
beaches because these locations are where prey such as sand 
lance might be common. In British Columbia, Carter (1984) 
found murrelets in waters over sand and gravel bottom, 
possibly because of the concentration of sand lance. Strong 
and others (1993) hypothesized that adjustments in local 
distribution off Oregon was in response to movements of 
surf smelt. Ainley and others (this volume) suggested that 
murrelets favor areas of upwelling. high productivity, and 
concentrations of prey along the more open coasts of 



California, and that local movements here were also in 
response to food availability. We need considerably more 
information before we will be able to predict the types of 
locations where murrelets are likely to be concentrated. Our 
success in identifying the factors responsible for aggregations 
is likely to depend on concerted efforts to investigate the 
issue of prey distribution, and also on our sensitivity to the 
underlying spatial and temporal scales of the various 
mechanisms involved. 

Seasonal Movements 

In some, if not all, areas of their range. Marbled 
Murrelets exhibit seasonal redistributions of their populations 
(Klosiewski and Laing 1994; Kuletz 1994; Piatt and Naslund, 
this volume; Ralph and Miller, this volume; Strachan and 
others, this volume; Strong and others 1993). The studies 
of Burger (this volume b) and Speich and Wahl (this volume) 
provide important data showing that in winter murrelets 
move from the outer, exposed coasts of Vancouver Island 
and the Straits of Juan de Fuca into the sheltered and 
productive waters of northern and eastern Puget Sound. 
Although the available data are sketchy, the possibility 
exists that a large portion of this murrelet population, which 
in summer is widely dispersed along remote coasts, is 
concentrated in winter in an area with heavy ship traffic, 
including the frequent movement of oil tankers to and from 
refineries. Less is known about seasonal movements along 
the outer coasts of Washington. Oregon, and California, 
but Speich and Wahl (this volume) suggest that birds from 
the outer coast of Washington move into Grays Harbor 
channel in winter. The potential for winter concentrations 
of murrelets to encounter industrial and oil pollution in the 
sheltered waters that they prefer is a conservation issue of 
considerable concern (Carter and Kuletz, this volume; Fry, 
this volume). 

Social Influences at Sea 

Association of murrelets in pairs, probably for foraging, 
is well documented (Strachan and others, this volume). The 
possible costs or benefits of interrelationships with other 
species, such as kleptoparasitism by gulls (Hunt, this volume 
b) or predation by Peregrine Falcons (Falco peregrinus) is 
more speculative. However, the possible effects of human- 
caused increases in gull populations may be of some concern. 

Estimates of Abundance and 
Historical Trends 

Summary — We estimate, based on information in this 
volume, that the total North American population of Marbled 
Murrelets is about 300,000 individuals. Approximately 85 
percent of this population breeds along the coasts of the 
Gulf of Alaska and in Prince William Sound. There are few 
murrelets in the Aleutian Islands and Bering Sea. Murrelet 
populations in both Alaska and British Columbia have 
apparently declined substantially over the past 10 to 20 



USDA Forest Sen ice Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



years. In Washington, Oregon, and California, population 
trends are downward, but the magnitude of decline over the 
past few decades is unknown. As a result of the small size of 
remnant populations, the species has been listed by various 
authorities as threatened or endangered in parts of its range. 

Counts of Marbled Murrelets at sea are currently the 
best method of estimating the size of regional populations. 
Nests are difficult to find, and although detections of calls at 
inland sites provide indices to local activity, numbers of 
detections can not be translated into absolute numbers of 
birds present. In contrast, surveys of birds at sea can be done 
from boats or airplanes, can cover large areas quickly, and 
can be standardized to provide repeatability. It is also possible 
to develop models for extrapolation of results from areas 
that have been surveyed thoroughly, and apply them to 
nearby areas that have received more cursory inspection 
(Ralph and Miller, this volume). 

Estimates of Population Size 

Based on the at-sea survey data, our best estimate of the 
Marbled Murrelet population in North America is on the 
order of 300,000 individuals (table 2). The major portion of 



this population is concentrated in northern Southeast Alaska 
and Prince William Sound. 

Population size diminishes rapidly north and west of 
there. Populations are relatively small and fragmented 
throughout Washington, Oregon, and California. 

The repeatability of survey results appears to vary 
considerably between location, methods, and researcher. In 
Alaska, overall population estimates were similar between 
summer and winter counts within the same decade (Piatt 
and Naslund, this volume). In contrast, population estimates 
for Prince William Sound varied considerably between 
those made in the 1970s and those made subsequent to the 
Exxon Valdez oil spill; the disparity is greater than can be 
explained by the oil spill alone, and probably is the result of 
different sampling methods in the 1970s or changes in food 
availability (Klosiewski and Laing 1994; Piatt and Naslund, 
this volume). In Washington, counts made from 1978 to 
1985 (Speich and others 1992), were similar in magnitude 
to those made in 1993 (Varoujean and Williams, this volume), 
with perhaps 5,000 in the entire state. Likewise, along the 
Oregon coast, Varoujean and Williams (this volume), using 
an aerial survey, found murrelet numbers in 1993 to be in 
the same range as their estimates of population size in the 



Table 2 — Estimated size of Marbled Murrelet populations by geographic regions 



Regions 


Estimated population' 


Source 


Alaska 






Bering Sea, Aleutians 


2,400 


Piatt and Naslund, this volume 


Gulf, Kodiak, Alaska Peninsula 


33,300 


Piatt and Naslund, this volume 


Prince William Sound 


89,000 


Klosiewski and Laing 1994 


Alexander Archipelago 


96,200 


Piatt and Naslund, this volume 


State total 


220,900 




British Columbia 


45-50,000 


Rodway and others 1992 


Washington 






Outer coast and Strait 


3,500 


Varoujean and Williams, this volume 


Outer coast only 


2,400 


Speich and Wahl, this volume 


Puget Sound 


2,600 


Speich and Wahl, this volume 


State total 


ca. 5,500 




Oregon 


6,600 


Varoujean and Williams, this volume 




15-20,000 


Strong and others, this volume 


California 






Northern California 


5,700 


Ralph and Miller, this volume 


Central California 


750 


Ralph and Miller, this volume 


State total 


6,450 




Total 


ca. 287,00 to 300,000 





Midpoints are usually used where ranges were given in the source 



10 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



1980s. Their estimate is lower than the ship-based survey 
of Strong and others (this volume) by a factor of three, and 
the causes of this difference are probably due to differences 
in methodology, spatial coverage, assumptions, and survey 
and model errors. 

Sources of Error 

Although we believe that at-sea surveys for estimation 
of Marbled Murrelet population size is necessary, there is 
still need for validation of the methodology. Few seabird 
species have population estimates based on at-sea counts, 
and the accuracy (as opposed to precision) of these techniques 
is only now being established. Population estimates of 
murrelets based on at-sea counts are subject to several 
sources of error, and these sources and their magnitudes are 
likely to vary with location and season. Three aspects of 
surveys that can affect accuracy are the way in which 
counts of flying birds are made, observation conditions, 
and observer competence. The possibility exists for double 
counting birds that are flushed from the survey track and 
then settle on a portion of the track yet to be surveyed. This 
problem is more severe for ship-based than aerial surveys, 
because the speed of the plane is great relative to that of the 
birds. Even in aerial surveys, double counting may occur if 
adjacent survey lines are sufficiently close. Strong and 
others (this volume) suggest that double counting, even on 
ship-based surveys, may be only a minor problem, with an 
estimated 5 percent of counted birds being vulnerable to 
recounting. There is also the likelihood of not counting 
murrelets because they are underwater, either foraging or 
diving in response to a vessel or airplane (Strachan and 
others, this volume; Strong and others, this volume). This 
source of error is greatest for aerial surveys because a given 
area is in sight only for an instant. The ability of an observer 
on a boat or plane to see birds will vary with the speed of 
the survey platform, height of the observer, use of binoculars, 
the area for which each observer is responsible, and observer 
competence. To minimize these sources of error, or 
uncertainty, it is necessary to either limit observations to a 
narrow band (e.g.. Varoujean and Williams, this volume), 
to correct for the diminishing v isibility of birds at greater 
distances (Ralph and Miller, this volume), to calibrate aerial 
versus boat surveys, or to calibrate observers (i.e., use only 
a limited number of persons). 

The patchy distribution of murrelets and their propensity 
for large daily shifts in distribution (Burger, this volume b: 
Rodway and others, in press: Speich and Wahl, this volume) 
further complicate the interpretation of survey data. Throughout 
their range, the largest numbers of Marbled Murrelets seen 
on the water are within a few kilometers, and often less than 
500 m, of shore. Data from British Columbia (Burger, this 
volume b; Morgan 1989; Sealy and Carter 1994; Vermeer 
and others 1983 ) suggest that along the outer, exposed coasts, 
murrelets may forage closer to shore (out to 500 m) than they 
do in sheltered bays and fjords where birds are often 1 to 5 
km offshore. Large-scale surveys by ship or airplane that fail 



to thoroughly survey this narrow, inshore strip are likely to 
underestimate local murrelet populations. Additionally, within 
this nearshore zone, murrelets are found concentrated in 
preferred foraging locales. A consequence of this small-scale 
patchiness is that surveys on different days must cover the 
same routes each time if they are to be comparable (they 
provide an index, not a sample), or they must be carefully 
stratified by foraging zone. In addition, variance of counts is 
large, so precise estimates of abundance require large samples 
(numbers of counts). 

Temporal variation in the use of marine habitats by 
murrelets further complicates the assessment of annual or 
decade-long changes in numbers. Data from Washington 
(Speich and Wahl, this volume), British Columbia (Burger, 
this volume b; Rodway and others, in press) and Alaska 
(Kuletz and others 1994a; Piatt and Naslund. this volume) 
show that murrelets exhibit considerable seasonal and daily 
variation in their use of specific foraging areas. During the 
breeding season, the portion of the population attending 
nests will change with time. In order for surveys to be 
strictly comparable, care should be exercised to conduct 
surveys in similar seasons and at the same time of day, or to 
make appropriate corrections to account for these sources of 
variation. These sources of error apply to all surveys in table 
2. More research is required to validate census techniques, 
to establish the accuracy of different survey methods, and to 
determine the time of year when the most comparable surveys 
should be done. 

Trends in Murrelet Populations 

Historical data for Marbled Murrelet populations are 
few, and no estimates can be made for populations before 
1900. It is at least possible that the murrelet was an abundant 
bird, nesting in old-growth forests all along the Pacific Coast 
in numbers commensurate with the abundant nearshore small 
fish it preys upon, and not limited, as it is today, by the 
availability of remnant stands of old-growth forests in the 
southern portion of its range. Circumstantial evidence to 
support this argument is the existence of large numbers of 
murrelets in very high densities where old-growth is still 
abundant (i.e., the Gulf of Alaska), or where it is the most 
abundant seabird in summer (i.e., Prince William Sound) 
(Kuletz, pers. comm.). 

Although the total population of Marbled Murrelets still 
appears large (table 2), there is reason for concern for the 
continued viability of this species in some regions. Numbers 
at the southern end of the range are small and concentrated 
geographically, thereby leaving subpopulations vulnerable 
to damage by stochastic events. More importantly, evidence 
is mounting that population trends are downward where they 
have been measured, even though short-term fluctuations in 
climate and longer-term variation in ocean currents can result 
in apparent or temporary increases. 

In Alaska (Piatt and Naslund, this volume), and in 
Clayoquot Sound, British Columbia (Burger, this volume b; 
Kelson and others, in press), populations have apparently 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



11 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



declined on the order of 50 percent in the last 10 to 20 years. 
Piatt and Naslund (this volume) based their conclusions on a 
decrease in the number of birds seen on small boat surveys of 
Prince William Sound in 1972-1973 compared to 1989-1991, 
as well as declines from Christmas Bird Counts during this 
period. Burger (this volume b) based his conclusions for 
Clayoquot Sound on density estimates from surveys made 
between 1979 and 1993. In Barkley Sound, British Columbia, 
Burger (this volume b) also found evidence of a decline from 
work in 1992 and 1993. However, in this area in the spring of 
1994, he recorded 2-3 times as many murrelets, leading to the 
possibility that the low numbers in 1992 and 1993 were due 
to El Nino-like effects in those years. 

The murrelet populations in Puget Sound, Washington, 
are apparently now lower than earlier this century. Few 
counts of offshore populations have been performed in the 
state, but Speich and Wahl (this volume) indicate some 
declines in recent decades. Both they, as well as Piatt and 
Naslund (this volume) in Alaska, suggest that some proportion 
of these declines may be linked to large-scale factors 
influencing the prey of marine bird populations over the past 
few decades, or that short-term environmental phenomena, 
such as El Nino events, may have caused local population 
declines or redistribution. They also identify a number of 
other factors that may have contributed to murrelet declines, 
including oil spills, gill netting, and timber harvest. 

Although quantitative evidence concerning population 
trends is not available for Oregon and California, it is our 
judgment that the long-term trends have been downward in 
these states, as well as in Washington. Murrelets require 
forests with old-growth characteristics for nesting, and with 
the loss of their nesting habitat and incidental take in fishing 
nets and oil spills, Marbled Murrelet populations in the three 
states are almost certain to have decreased, as they have in 
Alaska and British Columbia. The declines in these latter 
two regions appear to have coincided with the cutting of a 
large fraction of the old-growth forests. The cumulative 
effects of oil pollution, gill netting, and natural changes in 
the marine environment have undoubtedly played a role as 
well. We are not able to separate these potential causes of 
decline at this point, but the declines, whatever their origin, 
are at least a cause for concern. 

Beissinger (this volume) has estimated an annual 
decline of at least 4-6 percent throughout the species' 
range. These estimates are largely based on the observation 
of adult-to-young ratios at sea in the late summer, and 
inferences from other alcid species. However, the age 
ratio data are controversial, are from years when ocean 
conditions were warmer than usual, and may reflect a 
relatively temporary decline in reproduction. In addition, 
inferences from other species are fraught with danger. 
These estimates apply to past conditions and cannot be 
projected into the future, especially since implementation 
of the U.S. Government's Forest Plan would conserve 
most remaining nesting habitat on Federal lands of 
California, Oregon, and Washington. 



Demography of the Marbled Murrelet 

Summary — Based on the rate of successful fledging of 
young from observed nests, Marbled Murrelet populations 
in recent years have had one of the lowest reproductive rates 
of any alcid population thus far studied. For the population 
to be stable, these low rates of reproduction must be increased 
or balanced by higher than average rates of adult survival. 
Factors affecting these demographic parameters are the 
possible exclusion of a portion of the adult population from 
breeding due to lack of suitable nest sites, a decrease in the 
number of breeding attempts due to food limitation, loss of 
nest contents to avian predators, and mortality of adults 
from both avian predators and human activities, especially 
oil spills and entanglement in nearshore gill nets. 

Long-term demographic data on adult survival, chick 
production, and chick survival, would be useful for 
determining whether murrelet populations are decreasing, 
stable, or increasing. These data would also help in 
evaluating the significance of threats to different components 
of the population, such as reduced productivity, and chick 
and adult mortality. For example, a 50 percent increase in 
juvenile predation might not be as serious as a 10 percent 
increase in adult mortality from gill-net losses, depending 
on what would be considered the normal range of these 
population parameters. Some species of alcids, such as 
Common Murres (Uria aalge), can recover from relatively 
large population losses because they have, for alcids, 
typically high levels of annual production, with 0.5-0.9 
chicks fledging per pair (Hudson 1985). For species with 
low rates of reproduction, high rates of adult survival are 
essential for a stable population. 

It is exceptionally difficult to measure most of the critical 
population parameters for Marbled Murrelets. The traditional 
method of banding and resighting large numbers of seabirds 
at their colonies to estimate annual adult survival cannot be 
employed for murrelets because they are inaccessible. For 
example, a study of Common Murre breeding success at a 
single site in one year might include observations on hundreds 
of breeding pairs, and involve the banding of hundreds of 
chicks. At the end of the 1993 breeding season, after many 
years of dedicated effort, we have breeding success 
information on only 32 murrelet nests (Nelson and Hamer, 
this volume b). We do not know how representative these 
data are for the population as a whole. The only other source 
of demographic information is the ratio of juveniles to adults 
observed at sea during the post-breeding period (Ralph and 
Long, this volume; Varoujean and Williams, this volume). 
These are based on the identification of juveniles and adults 
on the water. As Carter and Stein (this volume) describe, this 
separation is fraught with difficulty. The extrapolation of 
these demographic data to longer time periods may be of 
limited value because many of the available data on 
juvenile:adult ratios were obtained in years when sea surface 
temperatures were unusually warm and prey availability 



12 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



may have been reduced. However, this type of demographic 
data provide perhaps our best hope of assessing this aspect 
of the species' life history. 

State of Knowledge of Marbled Murrelet Demography 

The rate of production of young by Marbled Murrelets 
appears to be one of the lowest of all alcids (De Santo and 
Nelson, this volume). Of the 32 nesting attempts for which 
we know the reproductive outcome, only 28 percent resulted 
in the fledging of young (Nelson and Hamer, this volume b). 
Data were gathered over several years throughout the 
murrelet' s range from the Gulf of Alaska to California. In 
Washington, Oregon, and California, the success rate was 
somewhat higher, 36 percent of 22 nests fledged young. 
Most of the known causes of nest failure were related to 
predation of nest contents. Analysis of counts of young at 
sea in the early stages of the species' fledging period in 
relation to numbers of adults indicate that the reproductive 
rate in recent years has been less than that needed for a 
sustainable population (Beissinger, this volume). 

The data used for determining productivity are based 
on a number of assumptions and may be biased. First, some 
of the data on nest success were gathered in years when 
ocean temperatures have been unusually high and prey 
availability may have been reduced. For most alcid species, 
breeding failures in warm water years are the result of 
adults forgoing breeding or chick starvation (Ainley and 
Boekelheide 1990). If warm water conditions during recent 
years depress the number of adults attempting to breed, the 
age ratio at sea would not be typical of years with high food 
availability. Secondly, the data on age ratios determined at 
sea are also based on assumptions about the ability of 
observers to separate adults from juveniles on the water 
(Carter and Stein, this volume). This can be further 
complicated later in the season when, as Hamer and Nelson 
(this volume a) indicate, young can be leaving the nest as 
late as September, and many adults have molted to a plumage 
indistinguishable from that of the young. Thus, the number 
of young may be underestimated. The data that Hamer and 
Nelson compiled can be used to correct for the proportion of 
young fledged at any given date (Beissinger, this volume; 
Ralph and Long, this volume), giving a more accurate picture 
of the proportion of young. 

We think it unlikely that a reduction of murrelet prey, if 
such was the case during the recent studies, would be 
responsible in some way for the high rate of predation of 
nest contents. The ratios of juveniles to adults at sea would 
be influenced by the proportion of the adult population that 
bred, a proportion that is likely to be sensitive to prey 
availability. In contrast, nesting success may be depressed 
due to the possible attraction of nest predators to activities of 
researchers (see below). 

Adult mortality rates are unknown. However, evidence 
is accumulating that fouling by oil and bycatch in gill nets 
may be locally significant (Carter and Kuletz, this volume; 
Carter and others, this volume; Fry, this volume). 



Inferences from Other Species 

In the absence of adequate data on most aspects of 
murrelet breeding, we must try to infer many of the 
population parameters of demography from more detailed 
studies of other alcids. We know that other small ( 1 50-500 
g), fish-eating alcids (three Cepphus guillemot species and 
four Synthliboramphus murrelet species) naturally suffer 
high juvenile and adult losses from predation. 

These species produce two eggs and often fledge two 
chicks. Thus, because they have high rates of reproduction, 
these species can experience high levels of adult mortality 
and still maintain stable populations (Hudson 1985). The 
larger (500-1000 g), fish-eating alcids (four puffin species 
[Fratercula sp.], two murre species, and the Razorbill Alca 
torda) produce only one egg, but under normal conditions 
have higher levels of chick production than most other alcids, 
due to low levels of juvenile and adult mortality and the long 
lifespans for some of these species (De Santo and Nelson, 
this volume). 

The relatively small (ca. 230 g) Marbled Murrelet differs 
from the other small fish-eating alcids by producing a 
single-egg clutch and having, at least in recent years, very 
low success in raising young to fledging. If these patterns of 
reproduction are typical, the Marbled Murrelet must have as 
high or higher levels of adult survival, compared to other 
alcids, if the murrelet populations are to be stable. The 
Marbled Murrelet may be more sensitive than other alcids to 
factors that increase adult mortality (Beissinger, this volume). 
In the absence of hard data, we must infer that murrelet 
demography is likely to be relatively more impacted than 
that of other alcids by adult losses to predation, oil pollution, 
gill nets, etc. Certainly, there is evidence of the pervasive 
influence of predation in shaping the breeding biology of the 
species (e.g., cryptic breeding plumage, crepuscular nest 
attendance, behavior at the nest, and nesting in trees) (Nelson 
and Hamer, this volume b; Ydenberg 1989). 

Factors Affecting Murrelet Demography 

The demography of Marbled Murrelets is influenced by 
age of first breeding, the proportion of the adult population 
that breeds, the rate of production of young that survive to 
breeding age, and adult and subadult mortality rates. In this 
section we evaluate these factors and their potential for 
influencing the population dynamics of Marbled Murrelets. 

Limits on the Proportion of Adults Breeding 

Limitation of Nesting Habitat — There is circumstantial 
evidence, including both distributional and observational 
data, that Marbled Murrelet populations are limited by the 
availability of suitable nesting habitat and that the habitat 
presently available is already occupied by breeding murrelets 
(at least south of Alaska). This evidence includes the 
following: 

(a) Concentrations at sea near suitable nesting habitat — 
Marine resources do not seem to determine the at-sea 
distribution of murrelets in the breeding season, at least in 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



13 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



Washington, Oregon, and California. The observations that 
murrelets redistribute themselves after young have fledged 
indicate that food may be more abundant or accessible 
elsewhere. We thus conclude that the large-scale at-sea 
distribution and abundance of murrelets during the breeding 
season is not primarily related to the distribution and 
abundance of prey. It is possible, however, that the amount 
of prey offshore of old-growth influences the number of 
murrelets that breed there. Additionally, prey abundance 
may be influenced by oceanographic events that cause 
widespread, as well as local, reduction of productivity and 
prey availability. 

(b) Winter visitation of nesting sites — Some murrelet 
populations continue to visit breeding areas during the winter 
(Naslund 1993), indicating that nest sites need to be defended 
year round. This is a behavior seen in other alcids when 
there is competition for nest sites (Ainley and Boekelheide 
1990), and site retention may require sustained occupancy 
through the winter. Winter visitation by murrelets, however, 
was not apparent in British Columbia because the birds 
leave offshore areas near nesting sites in many parts of the 
Province (southwest Vancouver Island, and the Queen 
Charlotte Islands) (Burger, this volume b). 

(c) Limitation of nest sites and habitat saturation — 
Spacing of nesting pairs might lead to unused nest sites in 
some areas, but in others, high quality nest sites might be 
relatively infrequent, even in old-growth forest (Naslund, 
pers. comm.). The short stature of most Alaskan old-growth 
and the forms of some old-growth tree species at lower 
latitudes (for instance, redwoods have few large or deformed 
limbs) result in a potential scarcity of usable nest sites. 

In areas where large amounts of habitat have been 
removed, it is likely that there is significant saturation of 
the habitat by murrelets. In Washington, Oregon, and 
California, approximately 85 percent of the historic 
old-growth has been removed. If the Marbled Murrelet was 
not limited by nesting habitat previously, certainly the 
chances of limitation have greatly increased today. If habitat 
is saturated, then the remaining stands in these three states 
should have maximum densities of murrelets. Data from 
Alaska suggest that murrelet density may be higher when 
the availability of suitable nesting habitat is restricted. For 
example, Kuletz and others (this volume) compared onshore 
dawn activity with offshore populations in the Kenai Fjords 
and in Prince William Sound. They found generally higher 
onshore populations in the Kenai than in the Sound, although 
the at-sea population in the Sound was much higher. They 
suggested the difference in numbers at sea was due to the 
relative abundance of good nesting habitat in the Sound, 
whereas the Kenai had relatively disjunct, smaller patches 
of large trees. We interpret the apparently higher number of 
detections on shore in the Kenai Fjords as a result of 
crowding into the limited number of sites available, rather 
than in a difference of the quality of the available nesting 
areas. More indirectly, evidence for packing into a habitat 
is found in an area of northwestern California, in the largest 



area of coastal old-growth forest that remains south of 
Puget Sound. That area, in the vicinity of Redwood National 
Park and Prairie Creek State Park, has the highest rate of 
murrelet detections of any area within the species' range, 
with detections often exceeding 200 per morning (Miller 
and Ralph, this volume). This may reflect packing into the 
remaining habitat, or it may reflect superior habitat that has 
always supported large numbers of birds, although we do 
not think the latter is the case. Even if nesting habitat is in 
general saturated, it is also probable that there will be years 
when suitable nest stands are unoccupied by murrelets. 
Absences could result from the temporary disappearance of 
inhabitants from the stand due to death or to irregular 
breeding, perhaps because of a temporary decline in prey 
resources. Under either of these circumstances, unoccupied 
stands would not necessarily indicate that, over a longer 
time scale, habitat was not limiting or that these stands 
were not part of the murrelet' s habitat. 

Behaviors — The behaviors that influence site fidelity 
and use, as well as the degree of coloniality, will affect the 
likelihood of occupying of new habitat, and both may 
influence the rate that birds displaced by habitat destruction 
will acquire new nesting grounds. Site fidelity is the 
propensity of breeding birds to return to the same nesting 
location year after year, whereas philopatry is the tendency 
of young birds to recruit to the area where they were raised. 
Coloniality, the clumping of nests in time and space, is a 
function of the number of nests likely to occur in a stand. 
Most seabirds show considerable site fidelity, and many 
individuals return to the same nest site annually (Divoky 
and Horton, this volume). The young of many alcid species 
recruit to their natal colonies, although the degree of 
philopatry can be as low as 50 percent. Previously unoccupied 
habitats are occupied and new colonies grow faster than can 
be accounted for by recruitment. As Divoky and Horton 
(this volume) discuss, from what we know of other seabirds, 
we can assume that Marbled Murrelets return to a stand 
once they have bred there and continue to use that stand at 
least as long as they breed successfully. Upon nest failure, 
they may change nest sites or mates, but they would be 
expected to remain in the stand. Thus, once a stand is 
occupied by murrelets, one would expect it to be used on a 
regular, if not annual, basis, so long as it is not modified. 

Marbled Murrelets do not form dense colonies as is 
typical of most seabirds. However, limited evidence suggests 
that they may form loose colonies or clusters of nests in 
some cases. We would expect to find that the species 
maintains low nest densities, commensurate with available 
habitat. Coloniality evolves either as a means of protection 
against predation, or as an adaptation to exploit shared 
resources (nesting or foraging). We have no evidence that 
murrelets engage in group defense against predators, and 
their reliance upon cryptic coloration to avoid detection 
would argue for a wide spacing of nests to prevent predators 
from using area-restricted search, or from forming search 
images for murrelet nests. Marbled Murrelets have a number 



14 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



of aerial calls and displays (Nelson and Hamer, this volume 
a; Paton, this volume), the functions of which are not 
understood. If they have the same function as songs of 
songbirds, they could affect spacing. We doubt that murrelets 
exchange information about food resources while in the 
vicinity of the nest 

The possibility that murrelets may nest in loose colonies 
is supported by data from Naked Island, Prince William 
Sound. Alaska, where 2-3 pairs were found using a 3.6- 
hectare stand, and 7-12 pairs used a 17.5-hectare stand 
(Naslund and others, in press). Two active Naked Island 
nests were <20 m apart and two were <300 m apart in 1991. 
In 1994, three inland locations of radio-tagged birds were 
within 1 km or less of each other (two were definite nests, 
one was uncertain) (Kuletz. pers. comm.). As Ainley (pers. 
comm.) points out, if these internest distances are typical, 
they might characterize the murrelet as being loosely colonial, 
as in the Pigeon Guillemot (Cepphus columba) or Xantus' 
Murrelet (Synthliboramphus hypoleucus). 

Food availability — Marbled Murrelets forage on a number 
of different species of small fish and macrozooplankton 
(Burkett. this volume). Several of these fish species are 
subject to commercial fishing. Although we suspect that 
food supplies do not limit murrelet populations at present, it 
is possible that the availability of fish to murrelets may be 
influenced by human fisheries activities. Fish species for 
which competition between fisheries and murrelets could 
occur include Pacific herring (Clupea harengus) (e.g., in 
Prince William Sound), rockfish (Sebastes sp.), and, more 
remotely, northern anchovy (Engraulis mordax). The stocks 
of both herring and rockfish are now depleted due to 
overfishing (Ainley and others 1994). Superimposed on any 
human-caused changes in food supply are short- and 
long-term natural fluctuations in marine productivity. El 
Nino events are well known to reduce food availability to 
seabirds (Ainley and Boekelheide 1990). Longer-term 
fluctuations in marine climate have apparently had major 
effects in the Bering Sea and on the reproductive performance 
of seabirds nesting on the Pribilof Islands (Decker and 
others 1994). Murrelets in central California generally forage 
in areas of upwelling (Ainley and others, this volume), and 
change their distribution in response to natural fluctuations 
in prey abundance, such as those ascribed to El Nino (Hunt, 
this volume b). 

Limitation of Reproduction by Predation 

Losses of eggs and chicks to avian predators were found 
to be the most important cause of nest failure in the 32 
Marbled Murrelet nests for which the reproductive outcome 
was known (Nelson and Hamer. this volume b). Forty-three 
percent of these known nesting failures were ascribed to 
predation, a figure equivalent to 31 percent of all nests. The 
extent of site bias in these nests and the effect of observer 
influence are not known, but most nests have been found by 
investigators looking for them in or near openings in the 
forest where risk of predation may be higher. 



The cryptic coloration and secretive, solitary (or loosely 
colonial) nesting behavior of Marbled Murrelets suggests 
that they have evolved under a regime of exposure to heavy 
predation. Only their ground-nesting congener, the Kittlitz's 
Murrelet, is equivalent in its cryptic coloration. Its nesting 
biology and behavior suggest that it is also subject to heavy 
predation. The apparently low levels of Marbled Murrelet 
reproductive success suggest that nest failure resulting from 
predation, if not higher than in the past, is certainly at 
present a significant factor in their demography (Nelson and 
Hamer, this volume b). It is therefore of interest to determine 
whether current forestry practices might be influencing the 
exposure of murrelet nests to predation. 

Exposure to avian nest predators (i.e., jays and corvids) 
may be influenced by the size of a stand, and the placement 
of a nest relative to the edge of the stand. Paton (1994) 
reviewed literature on songbirds and found that artificial 
nests are subject to greater predation within 50 m of the edge 
of forest stands than in the center, although none of the 
studies were in western coniferous forests within the range 
of the murrelet. In British Columbia, Bryant (1994; Burger, 
this volume a) showed artificial nests of songbirds placed on 
or near the ground near the edge of a stand were more 
frequently preyed upon than those in the center of the stand. 
Bryant (pers. comm. in Burger, this volume a) also found 
corvids on Vancouver Island to be more common along the 
edges of forests than in their interior. Nelson and Hamer 
(this volume b), from a literature review, showed that (1) 
loss of nest contents to avian predators increases in some 
forested areas with habitat fragmentation and an increase in 
the ratio of forest edge to center habitat; (2) successful nests 
were further from edges (more than 55 m) and were better 
concealed than unsuccessful nests; and (3) small stand size, 
fragmentation of forests, and the opening of roads and other 
clearings all increased the ratio of forest edge to center. 

The failure rate for Marbled Murrelet nesting attempts 
may have increased due to an increase in the numbers of 
avian, especially corvid, predators and their foraging 
effectiveness (Nelson and Hamer, this volume b). Corvids 
are well known camp followers in parks and other outdoor 
recreation areas, and frequently follow or approach people 
in forested areas. Activity by researchers in the area of 
murrelet nests may attract corvids and increase the likeli- 
hood of murrelet nesting failure. Murrelets have nested 
successfully in the vicinity of campgrounds (Naslund 1993, 
Singer and others 1991), but it would be useful to test 
whether predators are more common where human activity 
is present. It will also be important to review research 
procedures to ensure that predators are not gaining clues 
about the location of murrelet nests from researchers (see 
Nelson and Hamer, this volume a). 

Adult Mortality 

Mortality of adult Marbled Murrelets may occur from 
natural or human causes. Predation on adult murrelets by 
raptors occurs in transit to nest sites and at nest sites, but has 



L SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



15 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



not been documented at sea. Given the small number of nest 
sites that have been monitored, observations of the taking of 
adult murrelets by predators raises the possibility that this is 
not a rare event. In recently documented cases, a Sharp-shinned 
Hawk {Accipiter striatus) in Alaska attacked and killed a 
murrelet as it came to its nest (Marks and Naslund 1994), 
and a Peregrine Falcon was observed taking adults at Waddell 
Creek, California (Suddjian, pers. comm.). In Alaska, Marbled 
Murrelet wings were the most common prey remains found 
at coastal Peregrine Falcon nests (Jeff Hughes, pers. comm. 
to Kuletz), bones have been found at other Peregrine aeries 
(Campbell and others 1977), and the remains of unidentified 
alcids have also been found in goshawk nests (Iverson, pers. 
comm.). These anecdotal reports are primarily within the 
Gulf of Alaska region, where Ancient Murrelets were also 
found to form an important part of the Peregrine's diet 
(Gaston 1992). Therefore, it seems likely that Marbled 
Murrelets may also form a substantial part of the diet of 
avian predators. 

Marbled Murrelets are vulnerable to discharge of pollution 
from point sources on land, to fouling by spilled oil, and to 
bycatch in gill nets (Carter and Kuletz, this volume; Carter 
and others, this volume; Fry, this volume; Kuletz 1994; Piatt 
and Naslund, this volume). Pollution discharged from point 
sources on land, particularly when it enters partially enclosed 
shallow bays, is a potential problem (Fry, this volume). For 
example, Miller and Ralph (unpubl. data) observed an increase 
in murrelet use of the coast immediately north of Humboldt 
Bay in 1993 after pulp mill effluent ceased to be discharged 
into the ocean. This was likely a response to increased prey 
(Ainley, pers. comm.). Oil spills are also of considerable 
concern, and have caused numerous losses of murrelets. In 
Alaska, the Exxon Valdez oil spill is estimated to have killed 
about 8,400 murrelets, approximately 3.4 percent of the 
Alaska population (Piatt and Naslund, this volume). 

Nearshore gill-net fisheries are an important source of 
annual mortality in some regions. Murrelets are particularly 
vulnerable to entanglement in gill nets during the hours of 
darkness (Carter and others, this volume). Based on the 
compilation of DeGange and others (1993), an estimated 
2,000 to 3,000 Marbled Murrelets are killed annually in 
Alaskan gill-net fisheries. In Barkley Sound, British Columbia, 
Carter and Sealy (1984) estimated that a gill-net fishery for 
salmon (Salmo sp.) in 1980 killed 7.8 percent of the projected 
fall population of murrelets. The location of that fishery was 
in an area where high densities of murrelets overlapped with 
an area that was intensively fished. That fishery has not 
opened in every year since 1980 (Carter, pers. comm. in 
DeGange and others 1993), and the 1980 value might not be 
typical of a long-term average mortality. In Puget Sound, 
Washington, Wilson (pers. comm.), estimated that as many 
as several hundred murrelets are killed in gill nets annually. 
These numbers, if correct, are a large proportion of the 
estimated murrelet population in the Sound. Few, if any, 
murrelets are killed in gill nets in Oregon or California, 
although, prior to the ban of shallow water gill netting in 



California, murrelets were killed (DeGange and others 1993). 
The annual mortality rates of Marbled Murrelets projected 
for salmon gill-net fisheries of Washington, British Columbia, 
and Alaska are of a magnitude to cause concern because of 
overriding influence of adult survivorship on murrelet 
demographics (Beissinger, this volume). 

The Future Course of Habitat and 
Populations 

Habitat Trends 

We believe that the ultimate fate of the Marbled Murrelet 
is largely tied to the fate of its reproductive habitat, primarily 
old-growth forest or forest with an older tree component. It 
is clear that the amount of Marbled Murrelet nesting habitat 
has declined over the past 50 years, due primarily to timber 
cutting (Perry, this volume). Bolsinger and Waddell (1993) 
estimated that total acres of old-growth forest in California, 
Oregon, and Washington declined from nearly 33 million 
acres in the 1930s to about 10 million acres in 1992 (of a 
total forested area of 66 million acres), although their analysis 
was based on a broader region than the range of the Marbled 
Murrelet. Of the remaining 10 million acres of old-growth in 
this region, 85 percent is under federal ownership. Federal 
lands within the range of the northern race of the Spotted 
Owl (Strix occidentalis) in these three States contain an 
estimated 2.55 million acres of potential murrelet nesting 
habitat (U.S. Dep. Agric./U.S. Dep. Interior 1994). Some 
biologists, however, estimate that much of this land is too far 
inland and at too high an elevation to be used by murrelets 
(Hamer, pers. comm.). Assuming these federal lands represent 
about 85 percent of all murrelet nesting habitat on all lands, 
the future of current habitat heavily depends on management 
decisions on the federal lands. 

The U.S. Government's Forest Plan is projected to 
conserve 89 percent of current murrelet nesting habitat within 
various categories of reserves on Federal lands in California, 
Oregon, and Washington. This amount of land represents 
approximately 75 percent of present murrelet nesting habitat 
in the three States. In addition, the plan calls for protection 
of nesting habitat within half-mile circles around all occupied 
sites. Therefore, in the short term, we expect little further 
loss of current habitat on Federal lands if the plan is 
implemented (although some occupied sites have been 
released to logging). Over the long term, we expect the 
amount of habitat on Federal lands to increase, as younger 
forest within these reserves matures. 

In Alaska, about 90 percent of the coastal old-growth 
forests remain from Kodiak Island to northern Southeast 
Alaska. Approximately 93 percent of what is classified as 
productive (and of that, about 58 percent of the highly 
productive component) old-growth forests that represent 
Marbled Murrelet habitat remain on the Tongass National 
Forest in southeast Alaska (Perry, this volume). At this time 
there is no direct evidence that highly productive stands are 



16 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



used more than the lower volume productive stands in 
southeast Alaska. The results of Kuletz and others (in press, 
this volume) in Prince William Sound, Alaska, and Burger 
(this volume) in British Columbia do, however, suggest that 
stands with higher densities of old-growth trees have 
characteristics associated with high murrelet use. We cannot 
predict the trend of the remaining old-growth forests, as it 
will depend on the final outcome of National Forest land 
management plans. We expect further decline in area of 
murrelet nesting habitat in regions where terrestrial habitat 
loss continues, and we expect this decline to stabilize 
eventually. But when, and at what level, this stabilization 
will occur, is difficult to foresee. The apparent reduction of 
the species' population by 50 percent in Alaska must be 
viewed with concern. Similarly, in British Columbia, with 
only about 30 percent of original coastal old-growth forest 
remaining and a likelihood of further loss, we cannot predict 
when the amount of suitable habitat will stabilize. 

On State lands, the status and trend of murrelet habitat 
depends on state forest practice regulations and implementation 
of take guidelines or Habitat Conservation Plans in 
cooperation with the U.S. Fish and Wildlife Service under 
the Endangered Species Act. In Washington, the State may 
seek an incidental take permit in exchange for delineating 
and protecting most currently occupied suitable habitat. 
Future management is difficult to predict, as new information 
may lead to revised definitions of suitable habitat and new 
management strategies. On tribal lands, we do not have 
information on likely management direction. On private 
lands, reduction of habitat with apparent breeding behavior 
is likely in the short term, but Habitat Conservation Plans 
may be undertaken by larger land owners. These plans may 
result in agreements to harvest some habitat in exchange for 
deferral of harvest of other habitat. 

Population Trends 

As we have suggested, available evidence indicates that 
the population of murrelets has declined over most of its 
range. As more nesting habitat is lost, coupled with the adult 
mortality in some areas from gill-net fisheries and occasional 
oil spills, we expect continued decline in the population of 
murrelets. The rate of future population decline may exceed 
the rate of habitat loss because of cumulative effects on adult 
survival. At-sea counts do not necessarily reflect breeding 
density, as some lag is expected between reduction in the 
nesting habitat and a decline in the at-sea population. Thus, 
effects may not appear in the form of a declining population 
for a decade or more. Murrelets are suspected to be long-lived, 
and adults may survive at sea even if nesting habitat is 
removed, perhaps leading to the low ratio of juveniles to 
adults found in at-sea counts in recent years (Beissinger, this 
volume). Reduction in prey, as might be occurring in recent 
warm-water years (1992-93). may also lead to a lower 
proportion of adults nesting and to lower reproductive success 
among those birds that do nest. 



We do not have the necessary information to predict 
what proportion of the current population can be lost without 
irreversible consequences. The most prudent strategy for 
now is to conserve those forest stands (where the species is 
listed) that currently support murrelets within each 
physiographic region; between these conserved areas, 
additional areas should also be set aside to improve the 
likelihood of recolonization of unoccupied areas. 

Some provision for catastrophic habitat loss and other 
unpredictable events is a necessary component of a conser- 
vation strategy. We cannot count on all areas of habitat to 
persist indefinitely. The forests within the range of the Marbled 
Murrelet are subject to periodic wildfire, to insect or disease 
outbreaks, to large scale windthrow, and other catastrophic 
losses. Managers will need to apply active management to 
reduce risk of loss in some regions. We recognize, however, 
that not all of these effects are bad, as some of these events 
result in creation of nesting habitat by stimulating formation 
of nesting platforms. 

The following key points are clear: 

(1) Murrelet population trends will vary by region, in 
relation to changes in the amount, distribution, protection, 
and ownership of remaining forest habitat, catastrophic loss 
of breeding habitat, prey abundance, and extent of mortality 
factors, such as oil spills, gill netting, and predation on 
adults and young. 

(2) A need exists to establish the relative importance of 
nesting habitat versus other factors in causing population 
trends. We assume that the trend in amount and distribution 
of suitable nesting habitat is the most important determinant 
of the long-term population trends. 

(3) Existing demographic methods do not permit analysis 
of population trends in relation to variation in quality of 
habitat (measured by amount and pattern of appropriate 
forest structure), because of the cost of gathering such data. 

(4) Given current knowledge or demographic methods, 
we are unable to know the likelihood that any population of 
murrelets is approaching a demographic threshold from which 
recovery may not be possible. 

(5) Net change in amount of habitat is a function of loss of 
current habitat versus succession of potential habitat If other 
demographic characteristics prevent recovery as suitable habitat 
stabilizes or increases (that is, if murrelet populations continue 
to decline), then other factors are regulating the population. 

(6) Populations are relatively large in Alaska and British 
Columbia, perhaps allowing more time to evaluate trends 
than in other parts of the range. However, large population 
declines in Alaska are, at least, cause for concern. Certainly, 
throughout the range, immediate management efforts should 
be directed towards maintenance of the North American 
population at or near present levels. In Alaska and British 
Columbia, we need an accelerated effort to better understand 
murrelet ecology and habitat relationships through research 
and surveys. These need to be initiated immediately, and a 
conservative habitat management approach needs be adopted 
in the interim. 



I'SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



17 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



(7) The cumulative effects of further incremental loss of 
existing habitat, in addition to continued loss of adults at 
sea, must immediately be considered and dealt with by all 
relevant agencies. To this end, we strongly suggest that a 
prudent strategy would be to curtail further loss of occupied 
nesting habitat in at least Washington, Oregon, and 
California. Further, the sharp reduction, or preferably 
elimination, of night-time inshore gill netting at the earliest 
possible date, within the areas where murrelets are known 
to concentrate on the water, would greatly reduce the risk 
of adult mortality. 

Management 

The objectives of efforts to conserve the Marbled 
Murrelet should be to manage habitat and other factors to 
achieve a stable, well-distributed population of the species 
throughout its range. The U.S. Marbled Murrelet Recovery 
Plan (U.S. Fish and Wildlife Service, in press) has considered 
management alternatives, and most of our suggestions come 
from their findings. In some cases we further define potential 
management needs based on findings in this volume. 

We agree with the U.S. Recovery Plan (U.S. Fish and 
Wildlife Service, in press) that the next 50 years will be the 
most critical period for murrelet conservation. Assuming 
that there has recently been a severe and critical loss of 
breeding habitat, the lag from the longevity of the species 
will result in continued population decline, resulting from 
birds dying without replacement over the next decade or 
two. Further, the loss of suitable habitat will continue, albeit 
at a reduced rate for the coming decade, at least. While 
efforts to stem adult mortality can be successful, they do not 
increase productivity. Only with increased suitable habitat, 
will the population again increase. Some areas, peripheral to 
present nest stands, could mature and become at least 
marginally suitable in 50 or, more likely, 100 years. We 
would expect that such succession, augmented by creative 
silvicultural practices to mimic older forests, could result in 
increases in the breeding population within 50 to 100 years. 
The sooner that habitat loss can be stopped and replacement 
of suitable habitat begun, the sooner the species can begin to 
recover substantially. 

Management of Current Nesting Habitat 

The overall objective of managing current nesting habitat 
should be to stabilize the amount of habitat as quickly as 
possible. This objective is expected to have the long-term 
effect of stabilizing or increasing the proportion of breeding 
adults and stabilizing or increasing juvenile recruitment. 

Identify Management Units at Various Scales 

Broad objectives by management agencies should be 
based on biological processes, not on political or 
administrative boundaries. The overall goal should be to 
maintain a well-dispersed Marbled Murrelet population, with 
each segment of the species' range managed to maintain a 



viable population. We suggest that management should be 
on a zonal basis and that nine Zones be designated. The U.S. 
Recovery Plan (U.S. Fish and Wildlife Service, in press) 
suggests six Conservation Zones for management in 
Washington, Oregon, and California as the basis for 
maintenance of the population. We would add additional 
zones to include all populations in North America. They 
are: (1) the Aleutian Islands Zone; (2) Southcentral Alaska 
Zone, including Prince William Sound, and 40 miles inland; 
(3) Southeast Alaska and Northern British Columbia Zone, 
from Yakutat Bay, Alaska and coastal British Columbia, 
south to Vancouver Island, and 40 miles inland; (4) Vancouver 
Island and Puget Sound Zone, including the Olympic 
Peninsula, and 40 miles inland; (5) the Southwest Washington 
Zone to 40 miles inland; (6) Oregon Coast Range Zone, 
south to Coos Bay and 35 miles inland; (7) the Siskiyou 
Coast Range Zone of southern Oregon and northern California 
to the Humbolt County line, south of Cape Mendocino, and 
35 miles inland; (8) the Northcentral California Zone to 
include Mendocino, Sonoma, and Marin counties, and 25 
miles inland; and (9) the Central California Zone south to 
Point Lobos in Monterey County, and 25 miles inland. 
These Zones are smaller to the south where populations are 
more fragmented and at greater risk. At the Zone level, 
broad objectives can be based on large-scale distribution of 
murrelet populations. Within each Zone, forest management 
could be planned on a scale that is relevant to the biology of 
the murrelet. We suggest that a relevant scale is at least 
100-200 miles of coastline. 

The rationale for this scale of analysis is that individual 
birds are known to travel as far as 60 miles in one direction, 
so a given offshore group could range over an area twice as 
wide (120 miles plus). It would be best to consider that the 
size of protected stands be a minimum of 500- 1 ,000 acres or 
more. This does not imply ignoring smaller occupied stands; 
this would not be desirable. Rather, these small stands could 
be included within larger units when possible. It is critical to 
avoid the incremental loss of small units that could lead to a 
small core population of murrelets lacking viability. 
Management units would be most effective if tied to existing 
land classification systems such as USGS hydrological basins. 
In Southeast Alaska, individual islands might be useful 
management units. 

Identify Highest Priority Sites Within Management Units 

Where available, we suggest the use of multi-year inland 
survey results to identify areas of high use, as Burger (this 
volume a) suggests for British Columbia. If these data are not 
available, then managers could use at-sea survey results to 
infer habitats that might support the highest numbers of 
murrelets within each management unit. This is usually only 
useful on a large scale; for example, no correlation has been 
found between activity at inland sites and immediately adjacent 
waters. On the other hand, the foraging area of the Waddell 
Creek population near Santa Cruz, California, appears to be 
closely tied to the nesting area (Ainley and others, this volume). 



18 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



We suggest that for areas where nesting and foraging locations 
have not been identified, inland sites of best remaining nesting 
habitat could be selected, using information from studies of 
habitat requirements. These areas would have the highest 
likelihood of supporting adequate numbers of nesting platforms 
and other structural elements correlated with nesting. This 
effort should be supported by inland surveys conducted to 
protocol standards to verify occupancy, but it is not practical 
to expect all potentially suitable habitat to be surveyed. 

Determine Current Management Status 

High-priority sites could then be evaluated to determine 
their likely level of long-term protection (usually likelihood 
of being reserved from timber cutting). That is, they could 
be evaluated as to whether they are under protection as 
late-successional reserves, or whether they are publicly or 
privately administered lands. 

Develop Management Strategy for Each Management Unit 

We suggest that a management strategy then be 
developed for each unit, based on the potential to support 
nesting and the role within a broader landscape (e.g., is this 
an area of special concern due to gaps in distribution, or 
lack of adjacent similar habitat?). The most effective and 
least risky technique to slow the current population decline 
is to conserve all current occupied sites or high quality 
habitat in areas where it is a listed species. If appropriate, 
especially on private lands and over the longer term, 
guidelines should be developed for removing murrelet habitat. 
For this effort, it will be necessary to determine the proportion 
of some specified land area around a site that can be cut 
without jeopardizing suitability of that site. For example, 
Raphael and others (this volume) found that, in Washington, 
>35 percent of the 200 hectares surrounding occupied sites 
was late-successional forest. Similar analyses should be 
conducted in other regions to test whether more general 
guidelines can be developed. 

It is often an issue as to what effect the cutting of a tree 
or a partial harvest has on the birds. If nesting habitat is a 
limiting factor in an area, then the options for a bird to move 
to uncut habitat might be limited when a nesting stand or 
potential nest trees are removed. Although an individual 
might be able to move to an occupied stand, the increased 
risk of predation with increased density of nests could offset 
the advantages of this move. If evidence shows that nesting 
density is not at saturation, which would have prevented 
more pairs breeding, then this viewpoint could be changed. 
By the same logic, removal of non-nest trees could increase 
the risk of other factors, such as a resulting increase in 
predator populations (because of an increase in other prey 
populations), increased access of predators into the stand, 
and a decrease in hiding cover for murrelet nests. Such 
management activities could be interpreted as the biological 
equivalent of the removal of individuals from the reproductive 
population. For example, a tree hazard removal program in a 
state park could have the effect of remov al of old-growth 



trees. If continued over the next 50 years, there certainly 
could be a significant reduction of murrelet habitat. As an 
example, tree hazard removal is occurring in most of the 
old-growth forests that have recreational facilities in California 
(Strachan, pers. comm.). We recommend that managers 
consider removal of developments, such as campgrounds, 
that are currently in old-growth. 

Evaluate Potential for Disturbance 

In the case of disturbance due to human activity in 
forest stands, the timing of disturbance can be adjusted to 
avoid disruption of murrelet activity, such as courtship, mating, 
or nesting. Risks to perpetuation of these sites from effects 
of fire, insects, disease, windthrow and other catastrophic 
events, should be evaluated. Actions to reduce such risk may 
be appropriate. We assume there will always be loss of 
habitat through natural processes, and management actions 
should allow for such losses. We need additional information 
about the likelihood that human activity near nests has any 
detrimental effects. 

Management for Buffer and Future Suitable 
Nesting Habitat 

The objective of managing for buffer and future suitable 
habitat is to provide additional structural cover to reduce 
fragmentation of nesting habitat, and to provide for 
replacement of habitat that might be lost from catastrophic 
events. This would provide a hedge against stochastic events 
and uncertainties in knowledge. This secondary habitat may 
also support additional nesting. 

Identify Habitat for Buffer Secondary Stands 

Identification of secondary habitat should be based on 
proximity to known nesting habitat and its potential to develop 
as nesting habitat within an appropriate time, perhaps 25 to 
50 years. These secondary stands may serve as buffers around 
nesting stands to reduce risk of windthrow or other loss. 

Accelerate Habitat Development by Silviculture 

The potential (as yet untested and uncertain) exists to 
apply silvicultural techniques such as thinning and canopy 
modification that could accelerate the attainment of suitable 
habitat conditions in younger stands. These techniques need 
to be tested and fully evaluated in an adaptive management 
framework before being counted on to provide expected 
habitat conditions. If successful, such techniques might be 
used to produce trees with suitable nesting platforms and 
canopy characteristics. 

Managing At-Sea Habitats and Risks 

The management of marine habitats to reduce risks of 
mortality from human sources may be of equal importance to 
the management of terrestrial environments to maintain nesting 
habitats. It is essential for managers to identify at-sea areas 
where murrelets concentrate during both the breeding and 
non-breeding seasons. These areas should be designated as 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



19 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



critical habitat and managed to reduce harm to murrelets. 
Threats to murrelets at sea include entanglement in fishing 
nets (particularly nearshore gill nets), oil spills, the presence 
of other pollutants (especially those that might affect the 
availability of prey organisms), and other factors causing 
loss of forage fish. However, we see the greatest challenges 
in the marine habitat to be the reduction of human-caused 
mortality of adult murrelets, rather than the enhancement of 
prey availability. Managing at-sea conditions will require 
overcoming jurisdictional problems involving overlapping 
responsibilities of multiple agencies (NOAA, U.S. Navy, 
U.S. Coast Guard, National Marine Fisheries Service, 
Environmental Protection Agency, U.S. Fish and Wildlife 
Service, USDA Forest Service, marine sanctuaries, tribal 
agencies and groups, and various state agencies, among others). 
Any solution will require close coordination and cooperation 
among all relevant agencies, and will be most effective if 
coordination is started at the highest political level (e.g., 
between Secretaries of relevant departments, and with 
coordination amoung appropriate state agencies and tribes). 
There is also a need for international cooperation between 
the United States and Canada in marine management. Already 
in place is the British Columbia/Washington Environmental 
Cooperation Council with a Marine Science Panel, as well 
as the British Columbia-Alaska- Washington-Oregon- 
California Oil Spill Task Force. 

Research Needs 

We suggest a series of high-priority research needs for 
the species, as follows. We list these approximately in the 
order of what we consider their importance, although, in dif- 
ferent regions, different priorities would apply. 

Inland Range of the Species 

The protection of nesting habitat requires defining the 
inland extent of murrelet habitat use. This has been based on 
observations of birds at inland sites. At some distance from 
the coast, the abundance of birds drops dramatically. Agencies 
have required that surveys be conducted at and beyond the 
farthest inland records of the species. We suggest that surveys 
to determine habitat use be concentrated at distances from 
the coast where the great majority of the population lives. We 
see little virtue in surveys conducted where murrelets only 
rarely explore. It is our opinion that these extremely peripheral 
areas can contribute very little to the species' survival. We 
also suggest that surveys be conducted at a distance from the 
coast in which more than 99.9 percent of the individuals in a 
region have been detected. The U.S. Recovery Plan (U.S. 
Fish and Wildlife Service, in press), defines "critical habitat" 
as being within 40 miles of salt water in Washington, 35 
miles in Oregon and California north of Trinidad Head, and 
25 miles for the remainder of California. With limited resources 
available for surveys, it seems prudent, from the standpoint 
of the conservation of the bird, to concentrate the majority of 
murrelet survey effort to these zones. 



Inland Habitat Association Surveys 

Habitat association patterns have received much 
attention, but a greater information-gathering effort needs 
to be made in most areas. Especially needed are surveys in 
forests of Alaska and British Columbia. Also needed are 
systematic surveys throughout actual and potential habitats 
to determine relative abundances (as estimated by activity 
level) according to the variables described in the various 
chapters, as well as along coastal-inland transects. Among 
the most important variables are the size of stands, their 
structure, and landscape configuration. While we have a 
good idea of the correlation of some variables with abundance 
of murrelets, knowledge is lacking of the actual way that 
these variables are important to the reproductive success of 
the species. We do not suggest, however, that large-scale 
manipulative experiments be launched with the idea of using 
this worthy method, especially from Washington south, 
where the potential negative effects of experimentation on 
already tenuous populations would be great. Rather, humans 
and nature have provided a range of natural conditions that 
can give a retrospective view of the habitat suitability. 
These effects include partial harvesting of timber, as well as 
thinning due to disease, fire, and windthrow. 

Related to the above is the minimum stand size for 
occupancy. Part of the research involving stand size should 
include the gathering of data on the number of birds occupying 
a stand and the number of nests present. Using the number of 
detections in a stand (currently the only metric available), 
one could then estimate, at least in part, if bigger stands 
support more or fewer birds per unit area than smaller stands. 

Evaluate Importance of Human Causes of 
Mortality at Sea 

It is essential to obtain robust data on the take of murrelets 
in inshore gill nets and to relate that take to densities of 
murrelets in the area being fished, as well as the modes of 
fishing. Modifications of fishing techniques, such as limiting 
fishing to daylight hours or appropriate changes in mesh size, 
should be sought in areas where murrelets are killed, so as to 
reduce the bycatch. Gill-net fishing in inshore waters where 
murrelets are abundant should be prohibited at an early date, 
if less drastic measures are not successful. The concerns 
about loss to gill nets are particularly great in Washington and 
British Columbia, but apply throughout the species' range. 
Similar concerns apply to loss from oil spills and detailed 
knowledge of the distribution of murrelets could alert managers 
to potential areas of extreme risk to certain populations. 

Risk of Nest Predation Versus Forest Structures 

It is essential to determine the role of predation in 
populations by studying nesting success. We must also deter- 
mine the influence of forest stand structure, and in particular 
the importance of the ratio of forest edge to interior area, on 
the number of predators present and how these factors affect 
the probability that a nest will be lost to predation. Surveys 
of the populations of potential predators in forest stands of 



20 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



varying types and degrees of fragmentation will provide the 
information on the direct effects of forest management on 
this source of mortality. Predation rates can be altered by 
forest types also, as the exposure of nests becomes greater 
with a more open forest. These data can be taken at the same 
sites as the surveys for murrelets described below. 

Population Size and Trends 

The sizes of populations in most of the range have only 
been approximated. Intensive surveys by air or sea can 
provide at a minimal cost a reliable index of population 
size. This is especially critical in Washington and Oregon, 
but is needed in most other areas of the range as well. Since 
definitive long-term trend data are lacking in virtually all 
populations, and are absolutely necessary for comparing the 
effects of management, succession, stochastic events, or the 
aging of the murrelet population, immediate efforts should 
be initiated to establish long-term studies. Calibration of 
at-sea survey technit^ues, including determination of the 
time of year when surveys are best done to determine 
population size, are needed. As part of this study, the 
hypothesized relationships between numbers of murrelets 
seen offshore, the number of detections during dawn watches, 
and the number of murrelets nesting in a stand should be 
tested. We recommend convening a workshop to evaluate 
at-sea sampling and data analysis methods. 

Demographic Information 

The methods of determining the demographic parameters 
of murrelets need to be expanded and refined. At present, 
observations of nests and the finding of young at sea provide 
the only clues about the demography of the species. These 
methods need to be continued and expanded, and new 
methods devised. 

Limitations of Fish Stocks 

We do not know if the availability offish species important 
to murrelets has declined, because the relationship of the 
abundance and distribution of the several species taken by 
the bird and the interplay of the behavior and distribution of 
foraging birds is unknown. Use of bioacoustics could provide 
the data on fish abundance and distribution simultaneous 
with information on the birds' distribution, abundance, and 
foraging. We urge that these methods be implemented in at 
least two or three regions immediately. These methods would 
provide a basis for establishment of marine reserves to provide 
a source of abundant food fish for critical key areas of 
murrelet feeding, as well as providing a source offish stocks 
for surrounding areas. Part of this research would include 
studies on the food habits of the murrelet. 

Genetic Structure of Populations 

Determinations should be made about the size of the 
various gene pools, the relative divergence of the populations, 
and the importance of gaps in distribution. We need genetic 
samples taken from throughout the range of the species. 



Colonialit y and the Saturation of Nesting Sites 

The degree of clumping of murrelet nests should be 
determined on a stand, forest, and landscape basis, once 
sufficient data on nest locations are available. A determination 
should be made about the extent to which behavioral spacing 
mechanisms used by murrelets affect the density of birds in 
a stand and the potential for selective harvest of trees. 

Effects of Human Disturbance 

Both in forests and at sea, the effects of various types of 
human disturbance should be evaluated in controlled 
experiments. It is not necessary to conduct these experiments 
in areas where timber harvesting is being carried out, as the 
noise and traffic of such activities are easily simulated. 

Conclusions 

We conclude that the stabilization and recovery of 
murrelet populations will be aided by (1) provision of adequate 
nesting opportunities, (2) elimination of sources of adult 
mortality by human impact and development, and (3) 
management to minimize loss of nest contents to predators. 

Specifically, we suggest the following steps be taken: 

1. Maintain a well-dispersed Marbled Murrelet popu- 
lation, with each segment of the species' range managed to 
maintain a viable population. Nesting habitat appears to be a 
primary limiting factor in maintaining murrelet populations. 
We feel any futher reduction in nesting habitat or areas for 
the murrelet in Washington, Oregon, or California would 
severely hamper stabilization and recovery of those populations 
to viable levels. Occupied habitat should be maintained as 
reserves in large contiguous blocks and buffer habitat sur- 
rounding these sites should be enhanced. 

Progress in attaining population stablization or enhance- 
ment can be measured by an increase in the productivity of 
the population, by increases in the total breeding population, 
an increase in the ratio of juveniles to adults in offshore 
population, and an increase in nesting success. It is critical 
that relevant agencies move quickly to put in place monitoring 
programs suggested above which can provide at least some 
of these data. 

2. We suggest management for the murrelet on a regional 
basis, such as the Conservation Zones recommended by the 
U.S. Marbled Murrelet Recovery Team (U.S. Fish and Wildlife 
Service, in press). We strongly urge that objectives by 
management agencies be based on biological processes, not 
on political or administrative boundaries, as much as possible. 
The overall goal should be to maintain a well-dispersed Marbled 
Murrelet population, with each segment of the species' range 
managed to maintain a viable population. 

3. Draft a landscape-based habitat conservation plan 
within each of the nine zones described above to ensure the 
maintenance of a viable population. As a result of this step, 
the suggested reserves would likely need to be augmented to 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



21 



Ralph and others 



Chapter 1 



Overview of Ecology and Conservation 



promote conservation of the species. We feel that the reserves 
alone would be insufficient to reverse the decline and maintain 
a well-distributed population. 

4. Adoption of the U.S. Recovery Plan's (U.S. Fish and 
Wildlife Service, in press) strategy that Late-Successional 
Reserves, as defined (U.S. Dep. Agric./U.S. Dep. Interior 
1994), within the Conservation Zones of the murrelet in 
Washington, Oregon, and California, could be designated 
and serve as Marbled Murrelet Conservation Areas. 

5. Conduct inland surveys in all suitable habitat within 
55 miles of the coast. Most effort in surveys and research 
should be within the region of critical habitat defined by the 
U.S. Recovery Team (U.S. Fish and Wildlife Service, in 
press), within 40 miles of salt water in Washington, 35 miles 
in Oregon and California north of Trinidad Head, and 25 
miles for the remainder of California, to help define the 
known habitat components of the species. 

6. Accelerate efforts to better understand murrelet 
ecology and habitat relationships. Whereas the Alaskan 
and British Columbia populations are considered by many 
to be secure because of their large numbers, we have here 
reviewed evidence of a decline in populations in these 
regions and find that the evidence is sufficient to cause 
concern. Research efforts inland and at sea need to be 
started immediately, and a conservative habitat management 
approach be adopted in the interim. Otherwise, we believe 
that in Alaska and British Columbia, within the next 20 
years, the species could well decline markedly, requiring 
similar habitat protection actions to those needed for the 
southern three states, where significant loss of old-growth 
forests has minimized management flexibility. 

In British Columbia, the Canadian Marbled Murrelet 
Recovery Team has assembled guidelines for preservation 
of nesting habitat. Specifically, they have recommended 
preserving at least 10 percent of each watershed where logging 
is continuing, more if there is less habitat nearby, in minimum 
blocks of 200 ha. While this may be adequate, based on the 
experience in Washington, Oregon, and California, we do 
not believe that the literature is sufficient to support this 
level of harvesting. 

7. It is useful to distinguish between the probable cause 
of the decline, and additional major threats to persistence 
and recovery. We have little doubt that the loss of suitable 
old-growth habitat has caused a marked decrease in the 



number of murrelets in most of their range. Where loss has 
been recent (within the last 15 years), we would expect to 
find there are a number of displaced adults who are no 
longer able to find breeding sites. In those areas, we should 
expect murrelet numbers to continue to fall until these 
displaced adults die off, as they will not be replacing 
themselves. Recovery of murrelet populations depends on 
the survival of breeding adults and their ability to produce 
young. The greatest threat to the recovery, therefore, is 
continued loss of habitat, adult mortality, and causes of 
breeding failure, in that order. We stress that it is critical to 
maintain and enhance habitat, reduce adult mortality rates 
due to at-sea risks and predation, and the reduce loss of nest 
site contents to predators. Better knowledge of how habitat 
structure influences predation risk would be a useful first 
step in setting priorities for development or protection of 
existing nesting habitat. What habitat features affect predator 
numbers and success remains uncertain. 

We remain optimistic about the long-term survivablity 
of the species. The ability of the various agencies, 
organizations, members of the fishing and forestry industries, 
and others, to pull together in the survey and research efforts 
that are described in the chapters to follow, is strong evidence 
that many people of diverse opinions are interested in the 
maintenance of the Marbled Murrelet throughout its range. 

Acknowledgments 

Many people contributed insights into this chapter, 
including all of the authors of the other chapters in this 
volume. Gary S. Miller, of the U.S. Fish and Wildlife 
Service's Marbled Murrelet Recovery Team, was especially 
helpful and supportive, and provided the information on the 
listing process of the species. We thank Kelly Busse who 
created the range map. We also thank David Ainley, Alan E. 
Burger, Jack Capp, George Divoky, Ron Escano, Jeffrey 
Grenier, Thomas Hamer, Chris Iverson, Kathy Kuletz, Linda 
Long, Garland Mason, Sherri Miller, Nancy Naslund, S. 
Kim Nelson, Jim Space, Steve Speich, and Craig Strong for 
helpful comments on the manuscript. We appreciate their 
insight and clear thinking. The conclusions of this paper 
represent, however, our collective conclusions, and differing 
viewpoints can be found among many biologists and in the 
pages of this volume. 



22 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 2 

The Asian Race of the Marbled Murrelet 



Nikolai B. KonyukhoV 



Alexander S. Kttaysky 2 



Abstract: We present here an overview of the ecology, abun- 
dance, and distribution of the subspecies of the Marbled Murrelet 
inhabiting the coasts of Asia. In most regards, the species is similar 
to the North American race with respect to its feeding, breeding, 
molt, and habitat ecology. It is, however, a migratory species, 
moving into southern parts of its range in the Sea of Okhotsk 
during the winter. The population has not been censused. but may 
number in the tens of thousands. Populations in Asia are threat- 
ened by logging of breeding habitat, oil pollution, and gill nets. 



During the last two decades, the attention of ornithologists 
has turned to the North American subspecies of the Marbled 
Murrelet, Brachyramphus marmoratus marmoratus (Gemlin), 
and they have gathered many data on the status and distribution 
of this poorly-known bird. On the other hand, relatively little 
information is available on the Asian subspecies of the 
Marbled Murrelet, B. m. perdix (Pallas). Recent work has 
suggested that this subspecies may actually be a separate 
species, the Long-billed Murrelet (B. perdix)(Friesen and 
others 1994a, Piatt and others 1994). This paucity of data is 
due in large part to the difficulty in reaching the remote 
areas where the species breeds. This information gap could 
also be partially explained by the sparsity of marine 
ornithologists working within the large territory of the Far 
Eastern region of Asia. In virtually all areas where 
ornithological studies have been conducted, Marbled 
Murrelets have been recorded. 

Most data on the species comes from work on Sakhalin 
Island, the Kamchatka Peninsula, and the Low Amur River 
region. In most coastal areas of the Sea of Okhotsk (except 
for some small areas), ornithological studies have never 
been carried out. 

The broad outline of the breeding range can be designated 
quite clearly (fig. 1), although the status of the species in the 
Koryak uplands is not clear. In the north part of its breeding 
range, most Marbled Murrelet records coincide with the borders 
of boreal coniferous forest, as established by Kistchinsky 
(1968b). The Asian subspecies, in contrast to the North 
American subspecies, is a migratory bird and leaves almost 
all of its breeding range in winter. It returns in early May to 
the southern parts of its breeding range (Nechaev 1986) and 
into northern parts at the end of May (Lobkov 1986). 



• breeding 

o possibly breeding 

MUSEUM SKINS 
A Moscow 
▼ St, Petersburg 







'Research Fellow. Laboratory of Bird Ecology. Institute of Animal 
Evolutionary Morphology and Ecology. Leninsky pr., 33, Moscow 1 17071, 
Russia 

-Ph.D. Candidate. Ecology and Evolutionary Biology Department. 
University of California-Irvine. Irvine CA 92717 



Figure 1 — Distribution of the Asian race of the Marbted Murrelet 
{Brachyramphus marmoratus perdix) according to data obtained from 
literature and museum collections. Triangular symbols indicate collec- 
tion locales of museum skins now in Moscow and St Petersburg. 



Methods 

In order to collect data on the distribution of the Marbled 
Murrelet in the Russian Far East, the senior author has 
examined 36 specimens in collections of Zoological Museums, 
both of Moscow University (ZMMU) and Zoological Institute, 
St, Petersburg (ZMZI). We also have investigated the biology 
of this species by examining all available literature, and 
through field observations in the Sea of Okhotsk. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



23 



Konyukhov and Kitaysky 



Chapter 2 



The Asian Race of the Marbled Murrelet 



Breeding Biology 

Pre-Laying Period 

The wintering grounds of the Asian subspecies are thought 
to be situated around Hokkaido Island, Japan (excluding the 
shore of the Okhotsk Sea), and the northern part of Honshu 
Island (Brazil 1991). Although not identified with certainty 
as to subspecies, the Marbled Murrelet is rarely found in the 
waters of Kyushu, central Ryukyu, and the Amami Islands. 
Scattered birds have been noted inshore of the Korean 
Peninsula during the winter (Shibaev 1990). 

Timing of Breeding 

There are no direct observations of the length of the 
breeding season. The approximate time of hatching and 
fledging can be estimated from observations of fish-holding 
behavior by adults and of young birds at sea. Birds with fish 
were observed in the Amur River mouth in June (Shibaev 
1990). A collected bird (ZMZI 168716/220-987) taken in 
July 1979 had a brood patch beginning to refeather. 

Additionally, eggs and incubating birds have been found 
during June through the years (table 1). Specifically, eggs 
were found on 17 June 1961 and 23 June 1973 near the city 
of Okhotsk and to the south (Kuzyakin 1963, Yakhontov 
1979), on 21 June 1984 in southern Primorye Region 
(Primorye), on 19 June 1976 on Northern Sakhalin Island 
(Labzyuk, cited in Shibaev 1990; Nechaev 1986), and on 15 
June on Eastern Hokkaido Island (Brazil 1991). Using these 
data, Shibaev concluded (1990) that the breeding period is 
generally similar in different parts of the breeding range. 

There is no information available about the duration of 
either incubation period or the chick-rearing period. According 
to Shibaev (1990), the chick-rearing period in southern parts 
of the breeding region is in June-July. In the northern part of 
the Sea of Okhotsk, we have observed murrelets carrying 
fish from early July until late August. 

Characteristics of the Egg and Nest 

Only six Marbled Murrelet eggs have been found over 
its vast breeding range in Russia, and three have been found 
on the island of Hokkaido. All of the eggs in Russia were 
found in the middle of June (table 1). Two of these eggs were 
taken from oviducts of females. One was collected on the 



northwestern coast of the Sea of Okhotsk (58° 30'N, 141° 
20'E) on 23 June 1977 (Yakhontov 1979), and the other on 
Semyachik Spit (54° 08'N, 160° 02'E), Kamchatka Peninsula, 
on 13 June 1993 (Ladygin, pers. comm.). The ground color 
of most eggs is bluish-green or greenish-blue. Small markings 
of pale brown, brown, and dark-cream or brownish-gray, 
usually less than 2 mm in diameter, congregate near the blunt 
end of the egg (Kistchinsky 1968b, Nechaev 1986, Yakhontov 
1979). The pointed end is free of any marks (Kuzyakin 1963, 
Nechaev 1986, Yakhontov 1979). Three measured eggs had 
the following sizes: 63.6 x 39.3 mm and a weight of 48-50 g 
(Kuzyakin 1963); 63 x 38 (or 39) mm (Yakhontov 1979); 
66.2 x 39.0 mm and a weight of 53.7 g (Nechaev 1986). The 
murrelet lays only one egg (Kistchinsky 1968b). 

All four nests in Russia were found in larch trees (Larix 
daurica). The nest in the Okhotsk and Kukuchtui rivers area 
(in the vicinity of the city of Okhotsk) was 6-7 km from the 
shore, 7 m up on a branch with a broad base formed by 
growth of several small limbs (Kuzyakin 1963). The egg 
was laid on a piece of lichen (Bryopodori). The nest in the 
Olga Bay area was in a tree, in a rocky coastal cliff area, and 
2.5 m out on a branch similar to the above nest according to 
Labzyuk (Shibaev 1990). The nest on Sakhalin Island was 2 
km from Chay vo Inlet and 5 m up on the broken top of the 
tree. The fourth nest was found in the Koni Peninsula near 
Magadan city. Izmailov observed that the nest was 12 km 
from the shore and 7 m up in the tree (Kondratyev and 
Nechaev 1989). In Japan, three murrelet eggs were found on 
the ground in mixed coniferous/broad-leaved forest on Mt. 
Mokoto, 24 km inland from the Okhotsk Sea coast on the 
island of Hokkaido on 15 June 1961 (Brazil 1991). 

Nest Attendance 

During the breeding season, murrelets often fly over the 
forest and mountain summits uttering sharp shrill whistles. 
In the morning, birds call from 0500 until 0700, seldom as 
late as 0800. In the evening they call less often, beginning 
about an hour prior to dusk (Nechaev 1986). 

From Shibaev's observations (1990), adults visit the 
nest during the darker periods of the day in these high 
latitudes. But in the northern part of the Sea of Okhotsk, we 
have observed Marbled Murrelets carrying fish during both 
the morning period and at times during the day. 



Table 1 — Descriptions of Marbled Murrelet eggs found in Russia 







Egg 




Location 
found 


Height from 
ground 


Distance 
to sea 




Date 


Length 


Breadth 


Weight 


Source 




(mm) 


(mm) 


(g) 




(m) 


(km) 




17 June 1961 


63.6 


39.3 


48.0 


larch tree 


6.8 


6-7 


Kuzyakin (1963) 


23 June 1973 


63.0 


39.0 


— 


oviduct 


— 


— 


Yakhontov (1979) 


19 June 1976 


66.2 


39.0 


53.7 


larch tree 


5.0 


2 


Nechaev (1986) 


21 June 1984 


— 


— 


— 


larch tree 


2.5 





Labzyuk (1987) 


— 


— 


— 


— 


larch tree 


7.0 


12 


Kondratyev and Nechaev (1989) 


13 June 1993 


56.2 


39.3 


— 


oviduct 


— 


— 


Ladygin (pers. comm.) 



24 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Konyukhov and Kitaysky 



Chapter2 



The Asian Race of the Marbled Munelet 



Young Birds 

It is not clear how nestlings depart from the nest. As in- 
dicated earlier, nests can be situated close to or on the ground 
making it difficult for the young to take off. Some investigators 
believe that the young could use the nearest river or stream 
to reach the sea (Kistchinsky 1968a. Kuzyakin 1963). 

There have been no descriptions of murrelet chicks in 
Asia. Young birds were found in late July 1976 on the fresh 
water Ozhabachye Lake on the Kamchatka Peninsula (Vyatkin 
1981), in Mordvinov Bay, Sakhalin Island, on 6 August 
1972 (Nechaev 1986), and in South Kuril Strait, Kunashir 
Island. on 8 August 1963 (Nechaev 1969). Two young birds 
were taken in Avacha Bay, Kamchatka Peninsula, on 9 and 
18 August 1920 (Lobkov 1986). There are two skins of 
young birds in the Zoological Museum in St. Petersburg. 
These were taken from the Kamchatka Peninsula in the 
middle of August 1889 and in Aniva Bay, Sakhalin Island, 
on 23 September 1947. Four fledged young were collected 
close offshore near the foot of Mt. Mokoto, Hokkaido, in 
late August of 1982 (Brazil 1991). 

Foraging 

During the breeding period, the distribution of foraging 
Marbled Murrelets is linked to the estuarine ecosystems. 
They usually forage singly or in pairs, and rarely in groups 
of up to eight individuals. Near the southwest coast of the 
Sea of Okhotsk, birds fed from between 200-300 m to 
approximately 2-3 km offshore (Babenko and Poyarkov 
1987). In Amur Lagoon they fed up to 5-10 km from the 
shore in brackish water (Shibaev 1990). the depth of which 
is 1-10 m. In the Kamchatka Peninsula, birds congregate in 
bays, especially large ones (Lobkov 1986, Vyatkin 1981). 
Near the Kuril Islands, they have been observed opposite 
sandy beaches, very close to the shore (Velizhanin 1977). 
During observations in Tauy Liman area (in low-salinity 
water), birds foraged in shallow (5-20 m) inshore waters 
(Kitaysky. unpubl. data). 

In addition to feeding at sea, murrelets forage in large 
freshwater lakes on the Kamchatka Peninsula. On Sakhalin 
Island, murrelets were found regularly on two brackish 
lagoons. Kronotzkoe Lake (20-30 pairs) and Kurilskoe Lake 
» 15-20 pairs) (Lobkov 1986, Nechaev 1986). 

Only the remains of invertebrates were detected in the 
stomachs of birds collected in late June (the beginning of the 
breeding season) (Yakhontov 1979). Adult birds feed their 
chicks on fish (Kistchinsky 1968b, Shibaev 1990), though 
the exact composition of the diet is not known. We have 
observed birds feeding on both capelin (Mallorus villosus) 
and sand lance (Ammodytes hexapterus) in the northern Sea 
of Okhotsk. 

There are no direct reports on the availability of food 
resources, but according to indirect sources, they are 
ephemeral. For instance, near Baydukov Island, birds were 
absent on 26 July 1985, although during the previous day 
their density there was about 10 birds per kilometer of travel 



along the transect route (Babenko and Poyarkov 1987). High 
densities of murrelets in these areas are probably connected 
with aggregations of small fish, which are related to the 
complex dynamics of the oceanography in estuarine systems. 

Migration 

There is no information on visible observations of this 
murrelet's migration along the coastline. Birds disappear 
from the breeding range along the Kamchatka Peninsula in 
September (Lobkov 1986). In Ekaterina Bay, where they 
were common in summer, murrelets were gone by the end of 
October (Babenko and Poyarkov 1987). It is possible that a 
small number of birds reside in winter off Sakhalin Island 
(Nechaev 1986), but most of them migrate south to Japan 
(Brazil 1991, Shibaev 1990). It is possible that the species 
also winters in Alaskan waters based on a specimen (ZMZI 
5033) taken on or near Kodiak Island in January 1845. It is 
more likely, however, that this bird was a vagrant. At least 
1 3 specimens of this race of the Marbled Murrelet have been 
collected at various inland locations in North America in 
recent years (Sealy and others 1982, 1991). 

Molt 

We have little information on the molt of the Asian race 
of the Marbled Murrelet. According to data taken from 
collected birds, primary and rectrix molt of adults takes 
place between late July, when all birds have old feathers, 
and late October, when birds are in new plumage. Some 
birds, taken in early September from Avacha Bay, Kamchatka 
Peninsula, had begun their primary molt, but others had not. 

According to Koslova (1957), "A complete fall molt 
starts in cases with some adult murrelets in the first week of 
August. A female collected on 7 August in the Northern part 
of the Tatarskiy (Tartar) Strait (Taba Bay) had fresh feathers 
on the belly. Primaries, secondaries and tail feathers had not 
changed. Another female collected on 3 1 August in the Sea 
of Okhotsk had all its primaries, secondaries, and tail feathers 
fall out and the small feathers in the lower side were still in 
the tube [sheath] phase. Other adult individuals are delaying 
molt and have the full breeding plumage on the last week of 
August (date of collection: 18 August from the Litke Strait 
and on 24 August from the Ayan) without any signs of molt". 

The sequence of molt is variable, and birds do not lose 
all their primaries at once. They can possibly fly during 
early stages of primary molt. The primary molt begins from 
the inner end of the primaries. It is likely that greater coverts 
of the primaries are lost earlier than the primaries themselves. 
Specimens show a variable amount of loss of flight feathers 
during molt, individual specimens have lost between three 
(ZMZI 159931/425-974) and ten old primaries (ZMZI 159933/ 
425-974) during molL On the other hand, there was a specimen 
taken in the Sea of Okhotsk on 19 August 1845 which had 
lost all its primaries and secondaries (ZMZI 5047), and was 
obviously flightless. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



25 



Konyukhov and Kitaysky 



Chapter 2 



The Asian Race of the Marbled Murrelet 



Rectrix molt begins just after the beginning of primary 
molt. They are lost outwards from the inner pair. Contour 
feathers likely begin to molt in September. Two birds taken 
in early September in Avacha Bay had definite alternate 
plumage, but their throats and cheeks were almost white. 
A very pale bird in alternate plumage was taken on Sakhalin 
Island on 20 March 1969. The ground color was almost 
white, and all dark colors were replaced by yellow-brown. 

Shibaev (1990) describes a molting pattern in young 
birds as follows: "The change of nesting plumage into the 
first winter plumage takes place during September-October 
in young birds. The murrelets from the southern parts of the 
Far Eastern region of Asia, that had been collected in August, 
had a nesting plumage and had no signs of molt. The birds, 
collected in southern Primorye in the last week of October 
and in November, had changed from the nestling plumage 
into the first winter plumage already (excluding primaries, 
secondaries and tail feathers)." 



Body Measurements 

The Asian subspecies of the Marbled Murrelet is larger 
(tables 2 and 3) relative to the nominate race. This includes 
larger body mass (by about 50-70 g); larger culmen length 
(by about 5 mm); and larger wing length (by about 5-6 mm). 
See also Sealy and others (1982) and Piatt and others (1994) 
for discussion of body size characteristics. 

Besides these mensural differences (Sealy and others 
1982), these subspecies (or species) have differences in 
coloring of both their basic and alternate plumages. Jehl 
and Jehl (1981) noted that the North American Marbled 
Murrelet has a more rufous and darker alternate plumage 
than the Asian race, and has more pronounced spots near 
the eyes. In the basic plumage there is more contrast between 
the races (fig. 2). This can be seen by comparing pictures in 
field guides, for example Robbins and others (1983) and 
Wild Bird Society of Japan (1982). As with the basic 



Table 2 — Measurements of Brachyramphus marmoratus perdix, showing mean and range 





Shibaev 1990 




Stepanjan 1990 


Characteristic 


Male 


Female 


Both sexes 


Wing length (mm) 


141.2 


138.3 


143.5 




(136-147) 


(130-145) 


(133-150) 


Bill length (mm) 


20.2 


19.0 


— 




(18.9-22.2) 


(18-21) 




Tarsus length (mm) 


18.1 


18.0 


— 




(17-18.7) 


(16.8-19.0) 




Weight (g) 


295.8 
(258-357) 


— 


— 



Table 3 — Measurements of Brachyramphus marmoratus perdix from museum specimens 



Gender 


Mean 


s.d. 


n 


Minimum 


Maximum 


Characteristic 












Male 












Wing (mm) 


147.2 


3.7 


5 


141.0 


150.0 


Tarsus (mm) 


18.2 


- 


1 


- 


- 


Bill (mm) 


20.0 


1.4 


6 


19.1 


22.8 


Female 












Wing (mm) 


145.0 


7.2 


10 


130.0 


156.0 


Tarsus (mm) 


18.8 


0.6 


3 


18.4 


19.5 


Bill (mm) 


20.3 


1.3 


7 


19.0 


22.2 


Pooled data 












Weight (g) 


287.0 


41.7 


5 


258.0 


358.0 


Wing (mm) 


144.4 


7.3 


24 


126.0 


156.0 


Tarsus (mm) 


18.3 


0.6 


8 


17.4 


19.5 


Bill (mm) 


20.5 


1.3 


21 


18.4 


22.8 



26 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Konyukhov and Kitaysky 



Chapter 2 



The Asian Race of the Marbled Murrelet 




Figure 2 — Basic plumage head patterns of the Marbled Murrelet subspecies: (A) Asian (or Long-billed) subspecies and (B) North 
American subspecies. 



plumage, in the alternate plumage, the Asian race has more 
pronounced white eye spots. Also the border between white 
and dark brown on the head comes down to about the gape 
in the Asian race. Its upper mandible is dark and the lower 
is white. The chin is mainly white, but in some birds it is 
light gray. 

In the North American race, some white is always 
present on its upper mandible. In different birds, white on 
the face extends up to before the eye, forming a crescent 
patch. The chin, in contrast to the Asian race, is always gray 
and more extensive. 

The two races also differ in color of their tail feathers. In 
the Asian race, the outer vane of the outermost rectrix has a 
narrow, white marginal stripe. This stripe is especially 
pronounced in birds in fresh plumage. It might be absent in 
worn plumage because of abrasion. It is absent in the North 
American race. 

There are also similar differences in the juvenal plumage. 
In the first winter, the young look like adults, but the border 
between dark and white coloring is not as sharp, and on the 
entire ventral side a slight wavy pattern is present (Carter 
and Stein, this volume). 

Breeding Distribution and Abundance 

The Asian subspecies is widely distributed around the 
Sea of Okhotsk, on the Pacific Coast of the Kamchatka 
Peninsula, and in the Kuril Islands {fig. 1). The southern 
limit for breeding is on the island of Hokkaido in northern 
Japan. It is rare in eastern Hokkaido during summer, and 
more common on the Sea of Okhotsk coast, especially near 



the Shiretoko Peninsula (Brazil 1991). Observations of the 
murrelet from April- August in the area led to the assumption 
that it breeds in the region. This was confirmed when an 
incubating female and three eggs were collected on Mt. 
Mokoto in 1961. In late August 1982, four fledged young 
were collected close to shore near the foot of Mt. Mokoto 
(Brazil 1991). 

The southern limit of this species in Russia is in the 
southern Primorye Region where a bird was taken in Peter 
the Great Bay during the breeding period at the end of May 
(ZMZI 157639/6-971). Another location where murrelets 
have been taken at least twice is the middle reaches of the 
Bikin River (Gluschenko and others 1986). In the Primorye 
Region, a nest has been found in the forest on the shore of 
Olda Bay (Labzyuk 1987), and murrelets have been observed 
on the water there over many years. Birds, both single and in 
pairs, are found at sea there until the middle of June (Labzyuk 
1975). According to these authors, the Marbled Murrelet is 
quite uncommon near the southern limit of its distribution. 

At present, there are only a few areas where the Marbled 
Murrelet is considered common. One area is the lower Amur 
River area, on the southwestern coast of the Sea of Okhotsk. 
This coastline from Cape Lazarev to Aleksandra Bay was 
inventoried for seabirds in the summer months of 1980-1982 
and 1984-1986 (Babenko and Poyarkov 1987; Poyarkov, 
pers. comm.; Poyarkov and Budris 1991). Densities of 
murrelets averaged 0.5-2.0 birds per km of transect. Highest 
densities occurred between Baydukov Island and Aleksandra 
Bay. Lower densities were detected at the mouth of the 
Amur River and in Tatar Strait. The highest densities were 
found: (1) in Reynike Strait (300 birds on 2 km of transect) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



27 



Konyukhov and Kitaysky 



Chapter 2 



The Asian Race of the Marbled Murrelet 



on 3 September; (2) near the western point of Baydukov 
Island (50 birds on 5 km on 25 July, and 89 birds on 1 8 km 
on 1 August); and (3) in Schastye Bay, where the species 
congregates in great numbers every fall. The observers did 
not detect any between-year differences in the number of 
birds. Shibaev (1990) described it as relatively common in 
the Amur Liman, though apparently not many birds are along 
the Primorye coast (Elsukov 1 984) during the breeding season. 

On Sakhalin Island, Marbled Murrelets breed in different 
areas of the island, but are very patchy in distribution. Overall, 
Nechaev (1986, 1987) considered the species to be uncommon 
here, but he also stated that during the breeding season it 
often flies over forest and mountain peaks. For example, on 
Shmidt Peninsula, during the peak of breeding season, voices 
of 1 -2 birds could be heard frequently. 

One bird was taken at least inland 60 km from the sea in 
the mouth of the Maya River (ca. 54° 30'N, 134° 20'E), west 
of the Low Amur river area. 

Despite the fact that Gizenko (1955) and Nechaev (1969) 
wrote that the murrelet is rare in the Kurils, later publications 
indicate differently. Marbled Murrelets at least breed on all 
of the forested Kuril Islands: Shikotan, Kunashir, Iturup, 
and Urup (Velizhanin 1977). Nechaev and Kurenkov (1986) 
said that calls have been heard from time to time all over 
Kunashir Island in June and July. Several pairs and separate 
birds were detected over the forest near Korotky Stream at 
dawn on Kunashir Island on 17 July 1983 (Gluschenko 
1988). The species has not been recorded in the northern 
Kuril Islands (Podkovyrkin 1955). Since it also can nest in 
treeless areas (Hirsch and others 1981, Johnston and Carter 
1985, Simons 1980), and breeds both on the southern Kurils 
and on Kamchatka Peninsula where it is common, we think 
that murrelets may nest on all the larger islands through the 
Kuril chain (fig. 1). 

Near Magadan, this species is quite common in waters 
of Khmitievsky Peninsula, Tauyskaya Bay (Kondratyev and 
others 1992; Konyukhov, unpubl. data), and the Koni 
Peninsula (Leito and others 1991). The murrelet also is very 
common in Tauyskaya Bay in June through August. There 
we have found about 4 pairs per kilometer of coastline in 
June-July of 1991-1992. More than 200 birds were recorded 
in Nagaeva Bay (near Magadan) in June 1992 during a one- 
hour vessel trip along shore. In contrast, their abundance is 
very low along the northern coast of the Sea of Okhotsk 
(Kistchinsky 1968a, Yakhontov 1979). 

On the Kamchatka Peninsula the murrelets are quite 
common. It has been recorded within the inshore waters of 
the eastern coast, north to Karaginsky Bay, where they were 
more numerous in the larger bays. Eleven pairs were observed 
during a 50-km transect from Zhupanovo village to the 
Zhupanovo River, at a distance of 1-2 km from shore, on 27 
June 1973 (Lobkov 1986). Additional records are as follows: 
two in Asache Bay on 10 August 1972, four birds in Russian 
Bay on 13 August 1972; 52 birds in the southern half of 
Kronotzky Bay, from a ship over a 40-km transect on 16 
June 1974; and two pairs were seen daily in Ukinskaya Bay 



during May and early June (Vyatkin 198 1 ). Many specimens 
have been taken from Avacha Bay, near the city of 
Petropavlovsk-Kamchatskiy. 

It is possible that the species nests on the Komandorskie 
(Commander) Islands. Dementiev and Glagkov (1951) and 
Kuzyakin (1963) have mentioned that Dybovsky had found 
an egg on Medny (Copper) Island, which has been ascribed 
to the Marbled Murrelet. The size of the egg (62.5 x 41.2 
mm) was about the same as Marbled Murrelet eggs that have 
been described recently. One individual was also collected 
near that island in the spring. Later, Kartashev (1979) 
suspected that he had there a chick of the Ancient Murrelet 
(Synthliboramphus antiquus). He wrote: "The chick of the 
Ancient Murrelet, completely covered with down and with 
contour feathers beginning to erupt their sheaths, was found 
in a narrow crack of a cliff face near the southernmost tip of 
Medny (Copper) Island on 8 July 1960." One of us 
(Konyukhov 1990) had thought that this was a chick of the 
Kittlitz's Murrelet (Brachyramphus brevirostris), but because 
of the possibility of nesting Marbled Murrelets on land in 
crevices (Johnston and Carter 1985), and absence of records 
of enclosed Kittlitz's Murrelet nests, the senior author now 
thinks that the chick found by Kartashev was that of a 
Marbled Murrelet. Recent studies have shown presence of 
this species in that area during the breeding season. One bird 
was observed in Lisinskaya Bay, Bering Island, about 300 m 
from the shoreline on 17 June 1993 (Artyukhin, pers. comm.). 

Besides the nests noted previously, a breeding male was 
collected in Tayozhneya Bay in northern Primorye (Elsukov 
1984) and a female with a developing egg was collected 200 
km to the southwest of Okhotsk City (Yakhontov 1979). 
Adult birds carrying fish were recorded in the Amur Liman 
by Shibaev (1990), in the Tauyskaya Bay, and in the Tauy 
Liman (Kitaysky, unpubl. data). A fledgling with remains of 
downy feathers on its back was also observed by Kitaysky 
(unpubl. data) inshore of Zavyalova Island, in the northern 
part of the Sea of Okhotsk. An unidentified murrelet fledgling 
(perhaps B. m. perdix) was observed near Talan Island 
(Kondratyev, pers. comm.). 

Forest Habitat 

On Sakhalin Island, the species breeds in coniferous and 
mixed forests, both on the plains and in the mountains as 
follows: in the interior of the island, Nechaev (1986) recorded 
birds in flocks of 2-3, and sometimes four, in the upper 
reaches of the Onor River (30 km from the Sea of Japan) on 
29 June 1977; near the foot of Lapatin Mountain (30-40 km 
from the Sea of Okhotsk, at 600-700 m elevation) on 16-17 
July 1977; near the top of Krasnov Mountain (20-30 km 
from the Sea of Japan, at 500 m elevation) on 2 1 -22 July 
1987; and on the northern slopes of Nabil Mountain in the 
Shmidt Peninsula on 4-7 August. 

In the lower Amur River area, where Marbled Murrelets 
are numerous, the seaward slopes of the mountains are covered 
with coniferous forests, while the boggy level shore is covered 



28 



\ 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Konyukhov and Kitaysky 



Chapter 2 



The Asian Race of the Marbled Murrelet 



with larch forest (Babenko and Poyarkov 1987). It has been 
suggested (Kistchinsky 1968b) that the Marbled Murrelet 
breeding range is determined by taiga forest distribution in 
coastal areas of the region. Indeed, all nests and breeding 
birds observed to date have been in forested areas. 

Mortality and Population Trends 

Sources of mortality are rarely documented. There is 
one observation that Marbled Murrelets are occasionally 
shot by hunters. This occurred on the southwest coast of the 
Sea of Okhotsk (Babenko and Poyarkov 1987). There have 
also been records of plumage contamination by oil 
(Kondratyev and Nechaev 1989). 

Most authors have noted that the Marbled Murrelet is 
rare throughout its breeding areas. This may be a result of 
perspective, since the bird is small and relatively 
inconspicuous, as compared to other seabirds. No quantitative 
data exist, other than in small areas. Total population size is 
probably in the range of tens of thousands. No information 
exists to assess population trends. 

Conclusions 

Although Marbled Murrelets are widely distributed and 
relatively common in the Far Eastern region of Russia, to 
date we know little about the abundance, status, or main 
characteristics of the ecology of the Asian race of the species. 
Under these circumstances, it is impossible to say much 



about the murrelet' s population status or to make 
recommendations for management of this subspecies in the 
Sea of Okhotsk. 

Unfortunately, during the last few years the situation 
regarding the investigation and protection of wildlife in the 
former Soviet Union has taken a turn for the worse. It is 
important that Russia establish ecological control of natural 
resource exploitation, especially on the oceanic shelf. Intensive 
development of the oil industry on the Okhotsk and Bering 
Sea shelves is proceeding without appropriate control and is 
potentially threatening to shelf ecosystems in general. In 
particular, the overall breeding distribution of B. m. perdix 
matches the proposed areas for intensive oil development. It 
has long been suggested that increased murrelet mortality is 
quite possible because of oil pollution (Kondratyev and 
Nechaev 1989). Perhaps the greatest immediate threat to 
populations is from logging of forest habitats. Logging of 
prime old-growth forests has accelerated in recent years - 
particularly on Sakhalin Island and the Kamchatka Peninsula, 
where companies have recendy been granted logging rights 
over large tracts of virgin forest. This logging activity is 
apparently without regard to wildlife considerations. 
Ecological impacts of this industry are in need of investigation . 

Acknowledgments 

We thank John Piatt, Linda Long, and C. John Ralph for 
their hard work on the manuscript, including editing the 
final version into a coherent whole. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



29 



tr 

< 
a. 




Nesting Ecology, Biology, 

and Behavior 



Chapter 3 

Comparative Reproductive Ecology of the Auks 
(Family Alcidae) with Emphasis on the Marbled Murrelet 



Toni L. De Santo 1 2 S. Kim Nelson 1 



Abstract: Marbled Murrelets (Brachyramphus marmoratus) are 
comparable to most alcids with respect to many features of tbeir 
reproductive ecology. Most of the 22 species of alcids are colonial 
in their nesting habits, most exhibit breeding site, nest site, and 
mate fidelity, over half lay one egg clutches, and all share duties 
of incubation and chick rearing with their mates. Most alcids nest 
on rocky substrates, in earthen burrows, or in holes in sand, 
around logs, or roots. Marbled Murrelets are unique in choice of 
nesting habitat. In the northern part of their range, they nest on 
rocky substrate: elsewhere, they nest in the upper canopy of coastal 
coniferous forest trees, sometimes in what appear to be loose 
aggregations. Marbled Murrelet young are semi-precocial as are 
most alcids. yet they hatch from relatively large eggs (relative to 
adult body size) which are nearly as large as those of the precocial 
murrelets. They also share with precocial murrelets an early age of 
thermoregulation, as indicated by a short brooding period. Hatching 
success in monitored Marbled Murrelets nests was somewhat 
lower and fledging success was markedly lower than for other 
alcids. The lower rate of reproduction was attributed in part to 
egg and chick predation. Marbled Murrelet young raised in forest 
nests may incur additional mortality on their trips from inland 
nest sites to the ocean. El Nino effects may also decrease produc- 
tivity in this species. To document murrelet reproduction more 
fully, further study of individually marked, breeding Marbled 
Murrelets and their young conducted during periods without El 
Nino influences is needed. 



The family Alcidae is composed of 22 living species of 
marine diving birds representing 12 genera (table 1). These 
birds, commonly referred to as auks, murres, guillemots, 
murrelets. auklets, and puffins, inhabit oceans of the Northern 
Hemisphere (Nettleship and Evans 1985; Udvardy 1963). 
Although seabird research is logistically difficult, much 
information has been gathered on the reproductive biology 
of alcids. Such research has been facilitated by the colonial 
nature of most species and by the accessibility of some 
breeding areas to scientists (Birkhead 1985). Thorough 
reviews have been published on nearly half of the species. 
For instance, Birkhead (1985), Gaston (1985), Harris and 
Birkhead (1985), Hudson (1985), and Nettleship and Evans 
( 1985) present reviews of the reproductive biology of Atlantic 
alcids (Dovekie. Razorbill, Common Murre, Thick-billed 
Murre, Black Guillemot, Atlantic Puffin, and the extinct 
Great Auk [Plautus impennis]). Reviews of four auks that 



1 Research Wildlife Biologists. Oregon Cooperative Wildlife Research 
Unit. Oregon State University. Nash 104. Corvallis. OR 97331-3803 

: Current address: Postdoctoral Research Associate. Pacific Northwest 
Research Station. USDA Forest Service, 2770 Sherwood Lane. Suite 2A. 
Juneau. AK 99801-8545 



breed on the Farallon Islands in the Pacific Ocean (Common 
Murre. Pigeon Guillemot, Cassin's Auklet, Rhinoceros Auklet, 
and Tufted Puffin) are presented by Ainley (1990). Ainley 
and others (1990a, b, c) and Boekelheide and others (1990). 
Four inshore fish feeding alcids of the northern Pacific Ocean 
(Kittlitz's Murrelet, Pigeon Guillemot, Spectacled Guillemot, 
and Marbled Murrelet) are reviewed by Ewins and others 
(1993) (also see Marshall 1988a for a review of the Marbled 
Murrelet). The Ancient Murrelet, another inhabitant of the 
northern Pacific Ocean, has been reviewed by Gaston (1992). 

Alcids that nest in small, loosely-aggregated colonies, as 
isolated pairs, or in areas less accessible to researchers, have 
not been well studied. For instance, the reproductive biology 
of Craven's and Kittlitz's murrelets and Spectacled Guillemots 
is largely unknown. Although Marbled Murrelets have received 
considerable attention during the last two decades, the 
reproductive ecology of this species is not well understood. 
Unlike many other alcids. Marbled Murrelets do not nest in 
conspicuous colonies on cliffs, in rock crevices, or in burrows 
in the ground. Instead, this species nests on the alpine tundra 
or in the upper canopy of old-growth coniferous trees (Hamer 
and Nelson, this volume b; Marshall 1988a). Additionally. 
Marbled Murrelets are secretive around their nests and active 
during low light periods at dawn and dusk. Consequently, 
few nests have been located and observed, and few quantitative 
data have been collected. 

This paper summarizes the reproductive ecology of the 
auk family and specifically compares Marbled Murrelets to 
the other alcids. Such a comparison may allow for a better 
understanding of the reproductive strategy of Marbled 
Murrelets and should provide useful information for 
conservation and management of this species. 

Nest Sites and Coloniality 

The nest sites of all alcids have been described, although 
few nests of some species have been located (e.g., Kittlitz's 
and Marbled murrelets). Murres and Razorbills nest primarily 
on cliff ledges or in crevices or caves. The nests of Common 
and Thick-billed murres are in the open whereas those of 
Razorbills are typically partially or fully enclosed (Byrd and 
others 1993; Harris and Birkhead 1985). Puffins and 
Rhinoceros Auklets nest in burrows they excavate. Addition- 
ally, nests of these species are found in rock crevices (Tufted 
and Homed, on the level ground of forested islands (Rhinoceros 
Auklet), and among boulders and rocks of islands lacking 
soft substrate for burrowing (Atlantic Puffin) (Byrd and 
others 1993; Hatch and Hatch 1983). The guillemots nest in 
cracks and crevices of cliffs, among stones or boulders, in 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



33 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 1 — Extant members of the family Alcidae 



Common name(s) Scientific name 8 

Dovekie (Little Auk) Alle alle 

Razorbill (Razorbill Auk) Alca tarda 

Common Murre (Common Guillemot) Uria aalge 

Thick-billed Murre (Brunnich's Guillemot) Uria lomvia 

Black Guillemot Cepphus grylle 

Spectacled Guillemot (Sooty Guillemot) Cepphus carbo 

Pigeon Guillemot Cepphus columba 

Marbled Murrelet Brachyramphus marmoratus 

Kittlitz's Murrelet Brachyramphus brevirostris 

Xantus' Murrelet Synthliboramphus hypoleucus 

Craveri's Murrelet Synthliboramphus craveri 

Ancient Murrelet Synthliboramphus antiquus 

Japanese Murrelet (Crested Murrelet) Synthliboramphus wumizusume 

Crested Auklet Aethia cristatella 

Least Auklet Aethia pusilla 

Whiskered Auklet Aethia pygmaea 

Cassin's Auklet Ptychoramphus aleuticus 

Parakeet Auklet Cyclorrhynchus psittacula 

Rhinoceros Auklet (Horn-billed Puffin) Cerorhinca monocerata 

Tufted Puffin Fratercula cirrhata 

Horned Puffin Fratercula corniculata 

Atlantic Puffin Fratercula arctica 

"Nomenclature according to American Ornithologists' Union (1983) 



abandoned burrows, on covered ledges, or in self-excavated 
holes (Ewins and others 1993; Harris and Birkhead 1985). 
Nests of Dovekies are found most often in cracks in cliffs 
and among boulders (Harris and Birkhead 1985). Parakeet, 
Crested, Whiskered, and Least auklets nest under rocks in 
talus fields, whereas Cassin's Auklets excavate burrows in 
the soil (Springer and others 1993). The Synthliboramphus 
murrelets (Xantus', Craveri's, Ancient, and Japanese) nest 
in existing holes and hollows around tree roots, logs, or 
under rocks, or in crevices. Additionally, Japanese and Ancient 
murrelets may nest in self-excavated holes (Springer and 
others 1993). Kittlitz's Murrelets nest in the open on rocky 
ground. Marbled Murrelets nest in the open on rocky ground 
in the northern part of their range. In the southern part of 
their range, they nest on the large limbs of old-growth 
coniferous trees in forests up to 40 km from the ocean 
(Hamer and Nelson, this volume b; Marshall 1988a). 

Alcids are highly social birds, and most species are 
colonial in their nesting habits (table 2). Nineteen of the 22 
species can be found nesting in colonies consisting of 10 to 
several thousand pairs. Craveri's Murrelets probably nest in 
loose aggregations and as scattered pairs. The Kittlitz's 
Murrelet is the only species considered to be truly non- 
colonial (i.e., nesting only as isolated pairs). 

Marbled Murrelets have been described as solitary 
(Gaston 1985) and loosely colonial (Divoky and Horton, this 



volume), and may nest solitarily in some areas, but in loose 
aggregations in others. Simons (1980) reported a ground 
nest that appeared to be a solitary nest. There is also strong 
indirect evidence that murrelets nest in loose aggregations. 
In Washington and Oregon, two concurrently active nests 
were located 100 and 30 m apart, respectively, within a 
forest stand (Hamer and Cummins 199 1 ; Nelson, pers. obs.). 
In addition, in Oregon, multiple nests have been found in 
each of three trees located in different stands, and four trees 
within a small area (40-m radius) were found to contain 
nests (Nelson and others 1994). It is not known, however, if 
these nests were active concurrently. 

Breeding Site, Nest Site, and Mate Fidelity 

Studies of individually marked birds have provided 
information on the degree of breeding site and mate fidelity 
exhibited by alcids. Strong breeding site fidelity has been 
documented in the 15 species of alcids for which this aspect 
of reproductive ecology has been adequately investigated 
(Divoky and Horton, this volume) (table 2). For example, 96 
percent of Common Murres at one colony returned to breed 
at the same colony site the following year, and 90 percent 
used the same nest site (Birkhead 1976 as cited by Hudson 
1985). Similarly, Ashcroft (1979) as cited by Harris and 
Birkhead (1985) reported that 92 percent of Atlantic Puffins 



34 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter3 



Reproductive Ecology of Auks 



Table 2 — Breeding site fidelity, mate fidelity, and degree of coloniality in the ale ids 



Species 


Breeding site 

fidelity 


Mate fidelity 


Degree of coloniality 


Dovekie' 


yes 


yes 


small to large colonies, scattered pairs 


Razorbill 2 


yes 


probably 


small to large colonies 


Common Murre 3 


yes 


probably 


small to large colonies 


Thick-billed Murre 4 


probably 


7 


small to large colonies 


Black Guillemot 5 


yes 


yes 


small to large colonies, loose aggregations, scattered pairs 


Spectacled Guillemot 6 


? 


? 


small to medium colonies, solitarily 


Pigeon Guillemot 7 


yes 


probably 


small, loose aggregations, medium colonies, isolated pairs 


Marbled Murrelet* 


probably 


? 


probably in loose aggregations; probablysolitarily 


Kittlitz's Murrelet 9 


possibly 


? 


solitarily 


Xantus' Murrelet 10 


yes 


yes 


small to large colonies 


Craven's Murrelet 1 ' 


probably 


probably 


probably in loose aggregations and scattered pairs 


Ancient Murrelet 12 


yes 


possibly 


small to large colonies 


Japanese Murrelet 13 


? 


? 


small to medium colonies 


Crested Auklet 14 


yes 


yes 


small to large colonies 


Least Auklet 15 


yes 


yes 


small to large colonies 


Whiskered Auklet 16 


? 


J 


small to medium colonies 


Cassin's Auklet 17 


yes 


probably 


small to large colonies 


Parakeet Auklet 18 


yes 


? 


small, loose to large colonies 


Rhinoceros Auklet 19 


yes 


? 


small to large colonies, solitarily 


Tufted Puffin 20 


yes 


? 


small to large colonies 


Horned Puffin 21 


yes 


? 


large colonies 


Atlantic Puffin 22 


yes 


yes 


small to large colonies 



'Reviewed by Birkhead (1985), Freethy (1987), Harris and Birkhead (1985), Nettleship and Evans (1985); Evans (1981), 
Norderhaug(1968) 

2 Reviewed by Birkhead (1985), Freethy (1987), Hudson (1985), Nettleship and Evans (1985); Lloyd (1976) 

'Reviewed by Birkhead (1985), Freethy (1987), Harris and Birkhead (1985), Hudson (1985), Nettleship and Evans (1985); 
Sowls and others ( 1980), Speich and Wahl (1989) 

4 Reviewed by Birkhead (1985), Harris and Birkhead (1985), Hudson (1985), Nettleship and Evans (1985) 

5 Reviewed by Birkhead (1985), Freethy (1987), Harris and Birkhead (1985), Hudson (1985), Nettleship and Evans (1985) 

6 Reviewed by Birkhead (1985), Ewins and others (1993) 

'Reviewed by Birkhead (1985), Emms and Verbeek (1989), Ewins (1993); Ainley and others (1990b), Sowls and others 
(1980), Speich and Wahl (1989) 

8 Reviewed by Birkhead (1985), Ewins and others (1993); Divoky and Horton (this volume). Nelson and others (1994), 
Simons (1980), Strong and others (in press) 

9 Reviewed by Birkhead (1985); Day and others (1983), Naslund and others (1994) 

"•Reviewed by Birkhead (1985); Carter and McChesney (1994), Murray and others (1983), Sowls and others (1980), 
Springer and others (1993) 

"Reviewed by Birkhead (1985); DeWeese and Anderson (1976) 

1 2 Reviewed by Birkhead ( 1 985), Gaston ( 1 992); Gaston ( 1 990), Springer and others ( 1 993) 

''Reviewed by Birkhead (1985); Springer and others (1993) 

14 Reviewed by Birkhead (1985), Freethy (1987); Bedard (1969b), Jones (1993a), Konyukhov (1990a), Sealy (1968), 
Springer and others ( 1 993) 

1 Reviewed by Birkhead ( 1 985); BeViard ( 1 969b), Jones I. ( 1 992, 1 993b), Jones and Montgomerie ( 1 99 1 ), Roby and Brink 
(1986). Sealy (1968). Springer and others (1993) 

"•Reviewed by Birkhead (1985); Byrd and Gibson (1980), Byrd and others (1993), Springer and others (1993) 

17 Reviewed by Birkhead ( 1 985); Ainley and others ( 1 990a), Emslie and others ( 1992). Sowls and others ( 1 980), Speich and 
Manuwal (1974), Speich and Wahl (1989), Springer and others (1993), Vermeer and Lemon (1986) 

18 Reviewed by Birkhead (1985), Freethy (1987); Bedard (1969b), Sealy (1968), Springer and others (1993) 

19 Reviewed by Ainley and others (1990c). Birkhead (1985); Byrd and others (1993), Sowls and others (1980), Speich and 
Wahl (1989), Wehle (1980) 

Reviewed by Ainley and others (1990c), Birkhead (1985), Byrd and others (1993); Sowls and others (1980). Speich and 
Wahl (1989), Wehle (1980) 

2 'Reviewed by Birkhead (1985), Byrd and others (1993); Wehle (1980) 

-Reviewed by Birkhead (1985), Harris and Birkhead (1985), Nettleship and Evans (1985) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



35 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



returned to breed in the same burrow in consecutive years. 
Two studies reported at least 70 percent of Black Guillemots 
returned to use the same nest sites within the same nest 
colonies year after year (Asbirk [1979] and Petersen [1981], 
as cited by Harris and Birkhead 1985). Murray and others 
(1983) observed that 64 percent of Xantus ' Murrelets retained 
the same nest sites for two years, and Roby and Brink (1986) 
found that 9 1 percent of Least Auklets used the same nest 
entrance in two consecutive years. 

At least six alcids show mate fidelity (table 2). Divorce 
rates have been reported to be approximately 24 percent for 
Crested Auklets (Jones 1993a) and approximately 7 percent 
for Black Guillemots and Atlantic Puffins. These figures 
were confirmed by Harris and Birkhead (1985). Emslie and 
others (1992) have shown that mate retention has a positive 
influence on reproduction of Cassin's Auklets; both fledging 
and breeding success were higher for pairs that practiced 
mate retention. 

No studies have been conducted on individually marked, 
breeding Marbled Murrelets, but indirect evidence suggests 
that they show both mate and breeding site fidelity. Murrelets 
are primarily observed in groups of two throughout the year, 
and many groups include a male and female (Carter 1984, 
Sealy 1 975c). Strong and others (1993) observed at-sea groups 
of murrelets in spring and summer and reported that of 4918 
groups, 55 percent consisted of pairs. The possibility exists 
that these twosomes were mated pairs, although without 
observations of marked birds this is speculative. Marbled 
Murrelet activity has been documented in the same forest 
stands for periods up to 18 years (Divoky and Horton, this 
volume), and murrelet nests have been found in the same 
trees (Nelson and others 1994; Nelson and Peck, in press; 
Singer and others, in press), and on the same general location 
of tundra (Simons 1980), in consecutive years. These 
observations suggest breeding site fidelity. Reuse of nests 
has recently been documented for the ground nesting Kittlitz's 
Murrelet, a close relative of the Marbled Murrelet (Naslund 
and others 1994). 

Adult Life History Characteristics 

Historically, the Great Auk, which became extinct in 
the 1 800s, was the largest member of the Alcidae, ca. 5 kg 
(Harris and Birkhead 1985). At present, the murres are the 
largest alcids (ca. 1 kg). Fifteen alcids are small by comparison, 
having body masses less than half that of the murres {table 
3). The Marbled Murrelet has a mass of 220 g, approximately 
22 percent that of the murres. 

Adult annual survival has been estimated for ten species 
{table 3). The lowest estimates of this population parameter 
are 75 percent reported for both the Least Auklet, the smallest 
alcid (ca. 85 g), and 77 percent for the Ancient Murrelet, a 
relatively small alcid (ca. 200 g), (fig. 1, t 2 = 0.45, P < 0.05). 
The larger alcids, Common and Thick-billed murres, 
Razorbills, and Atlantic Puffins (ca. 1004, 941, 620, and 




400 800 

Adult mass (g) 



1 200 



Figure 1 — Relationship between mean adult body mass and percent 
annual adult survival for ten alcids (see table 3 for values). 



460 g, respectively), have higher survival estimates ranging 
from 89 to 94 percent. 

Adult annual survival has not been measured for Marbled 
Murrelets. However, based on the relationship between adult 
body mass and annual survival (fig. I), Marbled Murrelets 
(ca. 220 g) are predicted to have an annual adult survival of 
83 percent, comparable to alcids of similar size (i.e., the 
Ancient Murrelet, ca. 206 g, 77 percent survival, or the 
Crested Auklet, 272 g, 86 percent survival). 

Alcids are considered long-lived although this life history 
aspect has not been well studied. Longevity of several 
individuals of several species has been documented from 
recovery of marked birds or their bands. Values range from 
5 years for an Ancient Murrelet to 32 years for a Common 
Murre (table 3). Values determined from band returns may 
be indicative of band longevity, not bird longevity. These 
values should, therefore, be considered minimums (see Clapp 
and others 1982 for discussion). It is not known how long 
Marbled Murrelets live; no reports of recovered banded 
birds have been made. 

Alcids exhibit deferred maturity with most species 
beginning to breed between 2 and 8 years of age (table 3). It 
is not known at what age Marbled Murrelets begin to breed, 
but an estimate of 2 to 4 years is reasonable based on 
information available for other alcids. 

At least several alcid species breed annually once they 
reach sexual maturity (table 3). For example, over 80 percent 
of Least Auklets (Jones 1992) and 90 percent of Xantus' 
Murrelets (Murray and others 1983) bred in consecutive 
years. Cassin's Auklet is the only alcid known to lay a 
second clutch following the rearing of their first brood ( Ainley 



36 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 3 — Size, survival, longevity, age of first reproduction, and breeding frequency of adult alcids 



Species 


Mean body 


Annual 


Longevity 


Age first 


Breeding frequency 




masstgr* 


Survival of 

adults (pet.) 


(yr) b 


reproduction 




Dovekie 1 


164 


? 


7 


? 


1 /season 


Razorbill 2 


620 


90 


6,6,30 


4-6 


1 /season 


Common Murre 3 


1004 


89 


20, 26, 32 


4-6 


1/season 


Thick-billed Murre 4 


941 


91 


22 


? 


? 


Black Guillemot 5 


406 


84 


12,20 


2-8+ 


most annually and 1/season 


Spectacled Guillemot 6 


490 


? 


? 


? 


50 pet. annually 


Pigeon Guillemot 7 


487 


80-90 


9,11.14+ 


3-4 


1/season and probably not every year 


Marbled Murrcler 8 


221 


7 


? 


? 


? 


Kittlitz's Murrelet 9 


224 


? 


? 


? 


? 


Xantus' Murrelet 10 


167 


? 


9 


? 


most annually and probably 1/season 


Craveri"s Murrelet" 


151 


7 


? 


? 


7 


Ancient Murrelet 12 


206 


77 


5 


3-4 


1/season 


Japanese Murrelet 13 


183 


7 


? 


? 


7 


Crested Auklet 1 * 


272 


86 


? 


possibly 3 


1/season 


Least Auklet 15 


86 


75 


4.5 predicted 


3 


most annually and 1/season 


Whiskered Auklet 16 


121 


■» 


? 


? 


1/season 


Cassin's Auklet 17 


177 


86 


5, 10, 20 


2-4 


1-2/season 


Parakeet Auklet 18 


297 


7 


? 


? 


1/season 


Rhinoceros Auklet 19 


533 


7 


6,7 


? 


probably 1/season 


Tufted Puffin 20 


773 


? 


6 


? 


? 


Horned Puffin 21 


612 


? 


? 


? 


? 


Atlantic Puffin- 


460 


94 


13,20 


3-8+ (most at 5) 


probably breed annually and 1/season 



'Adult mass prior to chick rearing used for Ancient Murrelet, Crested Auklet, Least Auklet, Cassin's Auklet, and Atlantic Puffin; includes both males 
and females 

'Observed longevity of ringed or banded birds unless otherwise stated 

■Reviewed by Harris and Birkhead (1985); Norderhaug (1980) 

2 Harris and Birkhead (1985), Hudson (1985); Clapp and others (1982), Freethy (1987), Khmkiewicz and Futcher (1989), Lloyd (1979) 

'Reviewed by Harris and Birkhead (1985). Hudson (1985): Boekelheide and others (1990), Clapp and others (1982) 

"Reviewed by Harris and Birkhead (1985). Hudson 1985); Clapp and others (1982) 

-^Reviewed by Hudson (1985): Ainley and others (1990b), Clapp and others (1982), Cairns (1981, 1987), Divoky (1994, pers. comm.) 

6 Reviewed by Dunning (1992); Kitaysky (1994) 

"Re viewed bv Evvins (1993), Kuletz (1983); Ainley and others ( 1990b), Clapp and others (1982), Klimkiewicz and Fuicr*r( 1989). Nelswif 1987). Ewins 
and others (1993) 

8 Sealy(1975a.c) 

9 Sealy (1975b) 

10 Klimkiewicz and Futcher ( 1989), Murray and others (1983) 

"Reviewed by Dunning (1992) 

12 Clapp and others (1982), Gaston (1990), Gaston and Jones (1989), Jones (1990), Sealy (1975c, 1976). Vermeer and Lemon (1986) 

13 Kuroda (1967), Ono (1993) 

H Bedard (1969b), Jones (1992. 1993a), Piatt and others (1990) 

"Bedard (1969b), Jones (1992, 1993b, 1994), Piatt and others (1990), Roby and Brink (1986) 

16 Reviewed by Dunning (1992): Ainley and others (1990a), Knudtson and Byrd (1982) 

,7 Ainley and others (1990a). Clapp and others (1982), Jones P. (1992), Gaston (1992), Klimkiewicz and Futcher (1989), Manuwal (1979), Speich and 
Manuwal (1974), Thorensen (1964). Vermeer and Cullen (1982) 

18 Ainley and others ( 1990a), Bedard (1969b), Sealy (1972) 

19 Ainley and others ( 1990c). Clapp and others (1982), Klimkiewicz and Futcher (1989) 

^Klimkiewicz and Futcher ( 1989), Sealy (1972). Vermeer and Cullen (1979) 

21 Sealy (1973c) 

-Reviewed by Harris and Birkhead (1985). Hudson (1985); Barrett and Rikardsen (1992), Clapp and others (1982), Harris and Hislop (1978), 
Klimkiewicz and Futcher (1989). Kress and Nettleship (1988), Nettleship (1972) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



37 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



and others 1990a; Manuwal 1979). Within and between year 
breeding frequencies for Marbled Murrelets are unknown. 

The Egg and Incubation 

Most alcids, including Marbled Murrelets, lay a clutch 
consisting of one egg (table 4). The guillemots and the 
Synthliboramphus murrelets typically have clutch sizes 
of two. 

Alcid eggs range in size from less than 20 g to over 100 
g (table 4) and vary in proportion to adult mass (fig. 2,r* = 
0.92, P < 0.001). Alcid egg masses typically represent between 
10 and 23 percent of the laying female's body mass (table 4) 
with the precocial species laying the heaviest eggs relative 
to adult body size. Marbled Murrelet eggs (ca. 35 g) at 18 
percent of adult body mass are also large. 

The duties of incubation are shared by both members of 
the breeding pair (table 5). Incubation shifts can be as short 
as several hours (e.g., Pigeon and Black guillemots) or as 
lengthy as several days (e.g., Xantus', Ancient, and Japanese 
murrelets) (table 5). Incubation is completed within 27 and 
45 days (table 5). Overall, there is no significant correlation 
between incubation period and egg mass (fig. 3, r 2 = 0.10, P 
= 0.21). The eggs of four of the larger alcids (Rhinoceros 
Auklet and Tufted, Horned, and Atlantic puffins), however, 
require up to 45 days of incubation, while the small eggs of 
the Least Auklet are incubated for a much shorter period of 
time (ca. 30 days). 

Nine species of alcids are known to leave their eggs 
unattended for periods of 1 to 19 days, particularly during the 
early stages of incubation (table 5). Egg neglect is common in 
Xantus' (Murray and others 1983) and Ancient murrelets 
(Gaston and Powell 1989) occurring at nearly half of the nests 
studied. Egg neglect can lengthen the period from laying to 



hatching (Gaston and Powell 1989; Murray and others 1983; 
Sealy 1984), but can decrease the total number of days of 
actual incubation necessary (see Murray and others 1983). 

Compared to other alcids, Marbled Murrelets have a 
short incubation period (ca. 27-30 days) (table 5). Parents 
exchange incubation duties every 24 hours (table 5), typically 
during pre-dawn hours (Naslund 1993a, Nelson and Hamer, 
this volume a; Nelson and Peck, in press; Simons 1980). 
Simons (1980) noted a one-day period of egg neglect early 
in incubation at an exposed ground nest of a Marbled Murrelet. 
Additionally, at three tree nests, eggs were left unattended 
for up to 4 hours during the day and evening (Naslund 
1993a; Nelson and Hamer, this volume a; Nelson and Peck, 
in press). It is not known if egg neglect occurs commonly at 
Marbled Murrelet nests, but other alcid species which lay 
their eggs in open nests (e.g., Common and Thick-billed 
murres; see Gaston and Nettleship 1981) do not frequently 
leave their nests unattended. 

Average hatching success exceeds 70 percent for over 
half of the 19 alcids for which this parameter has been 
measured (table 6). The lowest value (33 percent) was reported 
for Xantus' Murrelets nesting on islands with high rates of 
mouse predation (Murray and others 1983). Avian and 
mammalian predation have been cited as a cause for clutch 
loss in other studies as well (Birkhead and Nettleship 1981; 
Drent and others 1964; Emms and Verbeek 1989; Evans 
1981; Ewins and others 1993; Gilchrist 1994; Harfenist 
1994; Jones 1992; Piatt and others 1990; Sealy 1982; 
Thorensen 1964; Vermeer and Lemon 1986). Additional 
causes of hatching failure include infertility and embryo 
death (Evans 1981; Knudtson and Byrd 1982; Thorensen 
1964), mechanical destruction of eggs or nests (Birkhead 
and Nettleship 1981; Thorensen 1964), and adverse weather 
(reviewed by Harris and Birkhead 1985). 



o> 



CO 
V) 



en 

HI 




400 800 

Adult mass (g) 



1 200 



Figure 2 — Relationship between mean adult body mass and mean egg 
mass for 21 alcids (see tables 3 and 4 for values). 




4 8 

Egg mass (g) 



1 20 



Figure 3 — Relationship between mean egg mass and incubation 
period for 19 alcids (see tables 4 and 5 for values). 



38 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 4 — Egg size, relationship of egg mass to adult body mass and clutch size ofalcids 



Species 


Mean egg 


Egg mass as pet 


Clutch size range 




mass(g) 


adult body mass 


(average) 


Dovekie 1 


31 (calculated) 


19 




Razorbill 2 


90 


14 




Common Murre 3 


ca.110 


12 




Thick-billed Murre 4 


100 


10-12 




Black Guillemot 5 


50 


12-13 


1-2(1.83) 


Spectacled Guillemot 6 


56 


11 


1-2(1.60) 


Pigeon Guillemot 7 


53 


11 


1-2(1.76) 


Marbled Murreler 8 


35 


18 




Kittlitz's Murrelet 9 


34 


15 




Xantus' Murrelet 10 


37 


22 


1-2(1.70) 


Craveri's Murrelet 11 


35 


23 


1-2(1.88) 


Ancient Murrelet 12 


46 


22 


1-2(1.99) 


Japanese Murrelet 13 


36 


22 


1-2(1.80) 


Crested Auklet 14 


36 


14 




Least Auklet 15 


18 


22 




Whiskered Auklet 16 


7 


? 




Cassin's Auklet 17 


27 


16 




Parakeet Auklet 18 


38 


ca.14 




Rhinoceros Auklet 19 


78 


ca.15 




Tufted Puffin 20 


93 


12 




Homed Puffin 21 


ca.60 


ca. 10 




Atlantic Puffin- 


61 


ca.13 





■Reviewed by Harris and Birkhead (1985), Evans (1981) 
2 Reviewed by Harris and Birkhead (1985) 
'Reviewed by Harris and Birkhead (1985) 
4 Reviewed by Harris and Birkhead (1985) 

5 Reviewed by Harris and Birkhead (1985); Cairns (1987), Divoky and others (1974) 
6 Reviewed by Ewins and others (1993); Kitaysky (1994), Thorensen (1984) 
7 Reviewed by Ewins (1993), Ewins and others (1993); Kuletz (1983), Nelson (1987) 
8 Hirsch and others (1981), Nelson and Hamer (this volume a). Sealy (1974, 1975b), Simons (1980) 
'Reviewed by Day and others (1983); Sealy (1975b) 
'"Murray and others (1983) 

"DeWeese and Anderson (1976), Schonwetter (1963) 

12 Reviewed by Gaston (1992); Gaston (1990), Gaston and Jones (1989), Jones (1992), Sealy (1975b, 1976), Vermeer 
and Lemon (1986) 

,3 Ono (1993), Ono and Nakamura (1993), Schonwetter (1963) 

14 Reviewed by Jones (1993a); Bedard (1969b) 

15 Jones (1993b), Piatt and others (1990), Roby and Brink (1986) 

,6 Freethy ( 1987), Williams and others (1994) 

■ 7 Ainley and others (1990a), Manuwal (1979), Vermeer and Lemon (1986) 

18 Sealy (1972), Bedard (1969b) 

19 Ainley and others (1990c), Freethy (1987), Sealy (1972), Wilson and Manuwal (1986) 

Reviewed by Boone (1986); Ainley and others (1990c), Sealy (1972) 

21 Freethy (1987), Sealy (1972) 

"Reviewed by Harris and Birkhead (1985) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



39 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 5 — Incubation patterns of the ale ids 



Species 


Incubating 


Incubation 


Mean duration of 


Egg neglect 




parent* 


shift (hours) 


incubation (days)" 




Dovekie 1 


both 


12 


29 


? 


Razorbill 2 


both 


12-24 


36 


? 


Common Murre 3 


both 


12-24 


33 


probably not 


Thick-billed Murre 4 


both 


12-24 


32 


very infrequently 


Black Guillemot 5 


both 


ca. 1-3 


29 


yes 


Spectacled Guillemot 6 


? 


? 


ca. 28 


? 


Pigeon Guillemot 7 


both 


2-4 but up to 17 


28 


yes 


Marbled Murrelet 8 


both 


ca. 24 


27-30 (probable range) 


yes for several hrs to 1 day 


Kittlitz's Murrelet 9 


7 


? 


? 


? 


Xantus' Murrelet 10 


both 


24-144, most at 72 34 


yes for 1-19 days 


Craveri's Murrelet" 


both 


? 


? 


7 


Ancient Murrelet 12 


both 


48-120 


34 


yes for 1-3 days 


Japanese Murrelet 13 


both 


24-72 


? 


yes for 5 days 


Crested Auklet 14 


both 


? 


35 


possibly 


Least Auklet 15 


both 


24 


32 


yes 


Whiskered Auklet 16 


both 


24 


ca. 35 


? 


Cassin's Auklet 17 


both 


24 


39 


very infrequently 


Parakeet Auklet 18 


both 


? 


35 


7 


Rhinoceros Auklet 19 


both 


24 


45 


yes for 1-3 days 


Tufted Puffin 20 


both 


? 


44 


7 


Horned Puffin 21 


? 


? 


41 


? 


Atlantic Puffin 22 


both 


2-50 


range 35-45 


yes frequently 



"Incubation refers to the period from clutch completion to egg hatching except for the Spectacled Guillemot for which this information 
was unavailable 

'Reviewed by Harris and Birkhead (1985) 

2 Reviewed by Harris and Birkhead (1985) 

3 Reviewed by Harris and Birkhead (1985), Boekelheide and others (1990), Hatch and Hatch (1990a) 

4 Reviewed by Harris and Birkhead (1985), Hatch and Hatch (1990a) 

5 Reviewed by Harris and Birkhead (1985) 

6 Ritaysky (1994), Kondratyev (1994) 

7 Reviewed by Harris and Birkhead (1985), Ewins (1993); Ainley and others (1990b), Drent and others (1964) 

'Carter ( 1 984), Hirsch and others ( 1 98 1 ), Naslund ( 1 993a), Nelson and Hamer (this volume a). Nelson and Peck (in press), Sealy ( 1 974, 
1975a), Simons (1980), Singer and others (1991, in press) 

^o information located 

10 Murray and others (1983) 

1 'Reviewed by DeWeese and Anderson (1976) 

12 Reviewed by Gaston (1992); Gaston and Jones (1989), Gaston and Powell (1989), Sealy (1976, 1984) 

13 Ono and Nakamura (1993) 

14 Reviewed by Freethy (1987); Jones (1993a), Piatt and others (1990), Sealy (1984) 

15 B6dard (1969b), Knudtson and Byrd (1982), Piatt and others (1990), Roby and Brink (1986), Sealy (1984) 

16 Reviewed by Freethy (1987); Knudtson and Byrd (1982), Williams and others (1994) 

"Reviewed by Manuwal and Thorensen (1993); Ainley and others (1990a), Manuwal (1974, 1979) 

18 B<5dard (1969b), Sealy and B&iard (1973) 

19 Leschner (1976), Wilson and Manuwal (1986) 

^Reviewed by Freethy (1987); Ainley and others (1990c), Boone (1986) 

21 Ainley and others (1990c), Leshner and Burrell (1977), Sealy (1973c) 

"Reviewed by Harris and Birkhead (1985) 



40 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter3 



Reproductive Ecology of Auks 



Table 6 — Hatching and fledging success, number of young produced per breeding pair and juvenile survival for ale ids 



Species 


Mean hatching 
success* 


Mean fledging 
success*- b 


Juvenile 
survival 










f~- 




Dovekie 1 


65 


77 


? 


Razorbill 2 


78 


93 


32 


Common Mums 3 


79 


88 


30 


Thick-billed Murre 4 


73 


85 


34 


Black Guillemot 5 


66 


68 


27 


Spectacled Guillemot 6 


? 


? 




Pigeon Guillemot 7 


70 


67 




Marbled Murreler 8 


67 


45 




Kittlitz's Murrelet 9 


? 


? 




Xantus' Murrelet 10 


33 


? 




Craven's Murrelet" 


? 


f 




Ancient Murrelet 12 


91 


>90 


ca.50 


Japanese Murrelet 13 


50 


76 




Crested Auklet 14 


63 


66 




Least Auklet 15 


82 


81 




Whiskered Auklet 16 


86 


100 




Cassin's Auklet 17 


75 


80 


65 


Parakeet Auklet 18 


65 


? 




Rhinoceros Auklet 19 


81 


69 




Tufted Puffin 20 


63 


70 




Homed Puffin 21 


76 


70 




Atlantic Puffin 22 


72 


73 


0.4-13 J observed. 
15-36 calculated 



"Includes replacement eggs for Common Murre, Razorbill. Thickbilled Murre, and Pigeon Guillemot, and possibly 
for Black Guillemot and Atlantic and Horned puffins: does not include second broods 

''Fledging is defined as departure from the nest to the ocean 

"Reviewed by Harris and Birkhead (1985); Evans (1981), Stempniewicz (1981) 

2 Reviewed by Harris and Birkhead (1985), Hudson (1985) 

Reviewed by Harris and Birkhead (1985), Hudson (1985); Ainley (1990), Boekelheide and others (1990), Murphy 
(1994): Hatch and Hatch (1990b); also see Byrd and others (1993) 

^Reviewed by Harris and Birkhead ( 1985), Hudson (1985); Hatch and Hatch (1990b); also see Byrd and others (1993) 

Reviewed by Harris and Birkhead (1985), Hudson (1985); Cairns (1981), Divoky (1994, pers. coram.) 

6 No information located 

7 Reviewed by Ewins and others (1993); Ainley and others (1990b). Kuletz (1983), Nelson (1987); also see summary 
by Ewins (1993) 

8 Nelson and Hamer (this volume b) 

*No information located 

10 Drost (1994), Murray and others (1983) 

n No information located 

l2 Gaston (1990, 1992). Rodway and others (1988), Venneer and Lemon (1986) 

,3 Ono (1993), Ono and Nakamura (1993) 

14 Knudtson and Byrd (1982). Piatt and others (1990), Scaly (1982); also see Jones (1993a) 

ls Jooes (1992), Piatt and others (1990), Roby and Brink (1986), Scaly (1982); also see Jones (1993b) 

16 Knudtson and Byrd (1982). Williams and others (1994) 

17 Ainley and others (1990a). Manuwal (1979), Thorensen (1964), Vermeer and Cullen (1982), Venneer and Lemon 
(1986) 

18 Sealy(1984) 

19 Verroeer and Cullen (1979), Watanuki (1987), Wilson and Manuwal (1986) 

Reviewed by Byrd and others ( 1993) 

21 Reviewed by ByTd and others (1993) 

-Reviewed by Barren and Rikardsen ( 1992), Harris and Birkhead ( 1985), Hudson ( 1985); Barrett and Rikardsen( 1992). 
Nettleship(1972) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



41 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



There is some indication that hatching success of the 
Marbled Murrelet is low compared to other alcids (Nelson 
and Hamer, this volume b). Combining observations 
throughout the range of the Marbled Murrelet, 67 percent (n 
= 20) of the eggs of 30 monitored nests successfully hatched. 
Egg predation was documented or strongly suspected to be 
the cause of failure for five of the 1 1 (45 percent) hatching 
failures (Nelson and Hamer, this volume b). 

When the clutches of alcids are lost or fail to hatch, 
some species (e.g., Razorbills, Common and Thick-billed 
murres, Atlantic Puffins, and Black Guillemots [see Harris 
and Birkhead 1985 for review], Pigeon Guillemots [Ainley 
and others 1990b], Cassin's Auklets [Ainley and others 1990a; 
Manuwal 1979], Horned Puffins [Wehle 1983]) lay 
replacement eggs. Egg replacement in murres has been 
reported to be between 15 and 43 percent (reviewed by 
Boekelheide and others 1990; Byrd and others 1993). Ten 
percent of Cassin's Auklet pairs replaced naturally lost eggs, 
and 54 percent replaced eggs removed by researchers 
(Manuwal 1979). Hatching and fledging success of 
replacement clutches was often lower than first clutches 
(Ainley and others 1990a; Byrd and others 1993; Manuwal 
1979; Murphy 1994). The incidence of egg replacement is 
low for Least and Crested auklets (Piatt and others 1990) 
and Xantus' Murrelets (Murray and others 1983) and 
apparently does not occur in Ancient Murrelets (Sealy 1976). 
Cassin's Auklet is the only alcid known to lay a second 
clutch following the rearing of their first brood (Ainley and 
others 1990a; Manuwal 1979). Hatching and fledging success 
of second clutches were usually lower than those of first 
clutches (Ainley and others 1990a). It is not known if Marbled 
Murrelets lay replacement eggs or if they attempt to raise 
more than one brood per season. 



Development and Survival of the Young 

Newly hatched alcids are downy (table 7) and are brooded 
by their parents for 1 to 10 days (table 8) until homeothermy 
has been achieved (table 7). Body mass of hatchling alcids is 
proportional to egg mass (fig. 4, r 2 = 0.98, P < 0.001) and 
adult body mass (fig. 5, r 2 = 0.91, P < 0.001). Alcid chicks 
are between 6 and 15 percent adult size at hatching (table 7). 
Newly hatched Marbled Murrelet chicks at 15 percent adult 
body mass, are large in comparison to the other alcid chicks 
(tables 7-9). 

Most alcid chicks are semi-precocial (table 7). Parents 
feed their semi-precocial young at the nest for 27-52 days 
until they reach at least 60 percent adult mass. Kittlitz's 
Murrelet may be an exception; the body mass of one fledgling 
was reported to be 40 percent that of an adult (Day and 
others 1983) (table 9). For most semi-precocial species, the 
young reach independence at the time of fledging (table 8). 

The Synthliboramphus murrelets have precocial young. 
For up to 2 days after hatching, precocial alcid chicks are 
brooded but are not fed at the nest. Following this time, they 
depart the nest at only 12-14 percent adult size and accompany 
their parents to the sea where they receive additional care 
until reaching independence at approximately 4 weeks of 
age (tables 7, 8, and 9). 

Murres and Razorbills are intermediate to these two 
patterns of development (Gaston 1985; table 7). Their 
young leave the nest at about 20 days of age, earlier than 
semi-precocial species, but much later than precocial species 
(table 9). At fledging, murre and Razorbill chicks are 
around 20 to 30 percent adult mass, lighter than semi- 
precocial young, yet heavier than precocial young (table 
9). The chicks accompany their male parents to the sea 




4 8 

Egg mass (g) 



1 20 



Figure 4— Relationship between mean egg mass and mean hatchling 
mass for 1 8 alcids (see tables 4 and 7 for values). 




400 800 

Adult mass (g) 



1 200 



Figure 5— Relationship between mean adult mass and mean hatchling 
mass for 18 alcids (see tables 3 and 7 for values). 



42 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 7 — Condition of ale id chicks at hatching and age at which homeothermy {uniform body temperature maintained nearly 
independent of environment) is achieved 



Species 


Developmental 


Plumage 


Mean body 


Pet. adult 


Age (days) of 




stage at hatching 




mass(g) 


mass at 
hatching 


homeothermy 


Dovekie 1 


semi-precocial 


downy 


21 


13 


2-5 


Razorbill 2 


intermediate 


downy 


ca.60 


9-10 


9-10 


Common Murre 3 


intermediate 


downy 


55-95 (range) 


6-10 


10 


Thick-billed Murre 4 


intermediate 


downy 


ca. 65 


7 


9-10 


Black Guillemot 5 


sem 


i-precocial 


downy 


ca. 40 


ca. 10 


1-4 


Spectacled Guillemot 6 


sem 


i-precocial 


downy 


40(n=l) 


8 


7 


Pigeon Guillemot 7 


sem 


-precocial 


downy 


38 


8 


1 


Marbled Murrelet 8 


sem 


-precocial 


downy 


33 


15 


probably 1-2 


Kittlitz's Murrelet 9 


semi-precocial 


downy 


? 


7 


? 


Xanrus' Murrelet 10 


precocial 


downy 


24 


15 


probably 1-2 


Craven's Murrelet" 


precocial 


downy 


? 


? 


? 


Ancient Murrelet 12 


precocial 


downy 


31 


13 


2 


Japanese Murrelet 13 


precocial 


downy 


7 


? 


7 


Crested Auklet 14 


sem 


-precocial 


downy 


ca.25 


10 


probably 4-5 


Least Auklet" 


sem 


-precocial 


downy 


11 


12-14 


probably 5 


Whiskered Auklet 16 


sem 


-precocial 


downy 


13 


11 


probably by 7 


Cassin's Auklet 17 


sem 


-precocial 


downy 


19 


11 


3^ 


Parakeet Auklet 18 


semi 


-precocial 


downy 


? 


? 


? 


Rhinoceros Auklet 19 


sem 


-precocial 


downy 


57 


10 


? 


Tufted Puffin 20 


sem 


-precocial 


downy 


64 


8 


? 


Homed Puffin 21 


sem 


-precocial 


downy 


59 


10 


? 


Atlantic Puffin 22 


semi-precocial 


downy 


47 


11 


6-7 



'Reviewed by Gaston (1985), Harris and Birkhead (1985), Ydenberg (1989); Evans (1981), Konarzewski and others (1993), 
Norderhaug(1980) 

2 Reviewed by Gaston (1985), Harris and Birkhead (1985), Ydenberg (1989) 

'Reviewed by Gaston (1985), Harris and Birkhead (1985), Ydenberg (1989); Birkhead (1976), Johnson and West (1975) 

4 Re vie w ed by Gaston ( 1 985), Harris and Birkhead ( 1 985), Ydenberg ( 1 989); Birkhead and Nettleship ( 1 98 1 ), Johnson and West 
(1975) 

'Reviewed by Gaston (1985), Harris and Birkhead (1985), Ydenberg (1989); Cairns (1981, 1987) 

6 Kitaysky (1994), Thorensen (1984) 

7 Reviewed by Freethy (1987), Gaston (1985), Ydenberg (1989); Ainley and others (1990b), Drent (1965) 

Reviewed by Gaston (1985), Ydenberg (1989); Hirsch and others (1981), Sealy (1975c), Simons (1980) 

Reviewed by Freethy (1987), Ydenberg (1989) 

""Reviewed by Freethy (1987), Gaston (1985), Ydenberg (1989); Murray and others (1983) 

"Reviewed by Gaston (1985), Ydenberg (1989); DeWeese and Anderson (1976) 

1 Reviewed by Gaston (1985, 1992), YdenbeTg (1989); Sealy (1976), Vermeer and Lemon (1986) 

"Reviewed by Gaston (1985) 

14 Reviewed by Freethy (1987), Gaston (1985), Ydenberg (1989); Sealy (1968), Jones (1993a), Piatt and others (1990) 

"Reviewed by Freethy(1987),Gaston(1985),Ydenberg(1989);Jones(1993b), Piatt and others ( 1990), Roby and Brink(1986) 

16 Reviewed by Byrd and Williams (1993); Williams and others (1994) 

1 7 Reviewed by Gaston ( 1 985), Ydenberg (1989); Ainley and others (1990a), Mamiwal ( 1 979), Thorensen ( 1 964), Vermeer and 
Lemon (1986) 

"Reviewed by Gaston (1985), Ydenberg (1989) 

"Reviewed by Freethy (1987), Gaston (1985), Ydenberg (1989); Wilson and Manuwal (1986) 

20 Reviewed by Freethy (1987), Gaston (1985), Ydenberg (1989); Boone (1986), Vermeer and others (1979) 

2 'Reviewed by Freethy (1987), Gaston (1985). Ydenberg (1989); Sealy (1973c) 

"Reviewed by Gaston (1985), Harris and Birkhead (1985). Ydenberg (1989); Barrett and Rikardsen (1992) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



43 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



Table 8 — Parental care in alcids 



Species 



Brooding 
parent 



Period of 
brooding (days) 



Feeding parent 



Time at which young 
reach independence 



Dovekie 1 
Razorbill 2 
Common Murre 3 
Thick-billed Murre 4 
Black Guillemot 5 
Spectacled Guillemot 6 
Pigeon Guillemot 7 
Marbled Murrelet 8 
Kittlitz's Murrelet 9 
Xantus' Murrelet 10 
Craveri's Murrelet 11 
Ancient Murrelet 12 
Japanese Murrelet 13 
Crested Auklet 14 
Least Auklet 15 
Whiskered Auklet 16 
Cassin's Auklet 17 
Parakeet Auklet 18 
Rhinoceros Auklet 19 
Tufted Puffin 20 
Horned Puffin 21 
Atlantic Puffin 22 



both 
both 
both 
both 
both 
? 

both 
both 
? 

both 
? 

both 
7 

both 
both 
? 

both 

both 

both 

? 

? 

both 



2-7 

5-10 

until fledging 

until fledging 

3-5 

? 

at least 3 

0.5-3.0 

? 

1-2 

? 

2 

? 

7 

7 

probably 7 

3-5 

? 

4 
? 
? 
9 



both at nest; probably neither at sea 


at fledging 


both at nest; male at sea 


several weeks following fledging 


both at nest; male at sea 


70-85 days 


both at nest; male at sea 


? 


both at nest; neither at sea 

? 

both at nest; neither at sea 


at fledging 

9 


at fledging 


both at nest; probably neither at sea 


at fledging 


both at nest 


? 


neither at nest; both at sea 


? 


both at sea 


? 


neither at nest; both at sea 


42-56 days 


neither at nest; both at sea 


? 


both at nest; neither at sea 


at fledging 


both at nest; neither at sea 


at fledging 


both at nest 


probably at fledging 


both at nest; neither at sea 


at fledging 


both at nest 


7 


both at nest 


probably at fledging 


? 


? 


? 


? 


both at nest; neither at sea 


at fledging 



■Reviewed by Gaston (1985), Harris and Birkhead (1985) 

2 Reviewed by Gaston (1985), Harris and Birkhead (1985) 

3 Reviewed by Gaston (1985), Harris and Birkhead (1985); Boekelheide and others (1990); also see Bayer and others (1991) 

"Reviewed by Gaston (1985), Harris and Birkhead (1985) 

5 Reviewed by Gaston (1985), Hudson (1985) 

6 No information located 

7 Reviewed by Ewins (1993), Freethy (1987), Gaston (1985) 

8 Naslund ( 1 993a), Nelson and Hamer (this volume a), Nelson and Hardin ( 1 993a), Nelson and Peck (in press), Simons ( 1 980), Singer and others ( 1 992, in press) 

*Naslund and others (1994) 

■"Reviewed by Freethy (1987); Murray and others (1983) 

"DeWeese and Anderson (1976) 

12 Reviewed by Gaston (1990, 1992); Sealy (1976) 

13 Ono and Nakamura (1994) 

'"Reviewed by Freethy (1987), Gaston (1985); Jones (1993a), Piatt and others (1990) 

15 Reviewed by Gaston (1985); Jones (1993b), Piatt and others (1990), Roby and Brink (1986), Sealy (1973a) 

16 Reviewed by Byrd and Williams (1993), Freethy (1987) 

17 Ainley and others (1990a), Manuwal (1979), Vermeer (1981) 

18 B(<dard (1969b) 

'"Reviewed by Vermeer and Cullen (1982); Wilson and Manuwal (1986) 

20, 21n i n f orma ti on located 

22 Reviewed by Gaston (1985), Harris and Birkhead (1985) 



44 



I USDA Forest Service Gen. Tech. Rep. PSW- 152. 1995. 



De Santo and Nelson 



Chapter3 



Reproductive Ecology of Auks 



Table 9 — Condition of akid young at time of fledging from the nest 



Species 



Mean fledging 
age (days) 



Mean body mass 
at fledging (g) 



Mean pet. 
adult mass 



Dovelbe 1 
Razorbill 2 
Common Murre-' 
Thick-billed Murre 4 
Black Guillemot 5 
Spectacled Guillemot 6 
Pigeon Guillemot 7 
Marbled Murrelet* 
Kittlitz's Murrelet 9 
Xantus' Muirelet 10 
Craven's Murrelet" 
Ancient Murrelet 12 
Japanese Murrelet 13 
Crested Auklet 14 
Least Auklet 15 
Whiskered Auklet 16 
Cassin's Auklet 17 
Parakeet Auklet 18 
Rhinoceros Auklet 19 
Tufted Puffin 20 
Homed Puffin 21 
Atlantic Puffin 22 



27 

18 

21 

22 

37 

ca.33 

38 

probably 27-40 

29 

1-2 

2-4 

2 

1-2 

33 

29 

probably 39-42 

43 

35 

52 

49 

38 

46 



120 

ca.170 

170-270 (range) 

180 

356 

545 (1 obs.) 

460 

149 

possibly 90 (lobs.) 

24 

26 
? 

ca.245 

87 

106 

153 

235 

329 

490 

400 

271 



67-80 

20-30 

18-28 

19 

86 

111 

95 

67 

possibly 40 

14 

? 

12-13 

? 

80-100 

104 

92 

90 

79 

61 

69 

65 

69 



'Reviewed by Gaston (1985), Harris and Birkhead (1985); Evans (1981) 

Reviewed by Gaston (1985), Harris and Birkhead (1985); Lloyd (1979) 

•Reviewed by Gaston (1985), Harris and Birkhead (1985); Hatch and Hatch (1990a) 

^Reviewed by Gaston ( 1 985 ). Harris and Birkhead (1985). Hudson (1985): Birkhead and Nenleship< 1981), Hatch and 
Hatch (1990a) 

'Reviewed by Gaston (1985), Harris and Birkhead (1985), Hudson (1985); Cairns (1981, 1987) 

6 Kitaysky (1994), Kondratyev (1994), Thorensen (1984) 

"Reviewed by Ewins (1993), Gaston (1985); Ainley and others (1990b), Drcnt and others (1964), Kuletz (1983) 

'Reviewed by Gaston ( 1985); Hirsch and others ( 198 1 ), Nelson and Hamer (this volume a). Nelson and Hardin (1993a), 
Nelson and Peck (in press), Sealy (1974, 1975a), Simons (1980), Singer and others (1992, in press) 

'Day and others (1983), Naslund and others (1994) 

10 Reviewed by Gaston (1985); Murray and others (1983) 

1 'Reviewed by DeWeese and Anderson (1976), Gaston (1985) 

12 Gaston (1992), Jones and Falls (1987), Sealy (1976). Vermeer and Lemon (1986) 

"Reviewed by Gaston (1985), Ono and Nakamura (1993) 

"Reviewed by Gaston (1985); Jones (1993a), Piatt and others (1990) 

"Reviewed by Gaston (1985); Piatt and others (1990), Roby and Brink (1986) 

16 Reviewed by Byrd and Williams (1993); Williams and others (1994) 

17 Reviewed by Gaston (1985); Ainley and others (1990a), Manuwal (1979), Thorensen (1964), Vermeer and CuDen 
(1982), Vermeer and Lemon (1986) 

18 Sealy and Bedard (1973) 

"Reviewed by Byrd and others ( 1 993), Gaston (1985); Ainley and others ( 1990c), Leschner ( 1976), Vermeer ( 1980). 
Vermeer and Cullen (1979), Wilson and Manuwal (1986); also see Bertram (1988) 

Reviewed by Gaston (1985); Ainley and others (1990c), Boone (1986), Vermeer and Cullen (1979X Wehle (1980) 

21 Reviewed by Gaston (1985), Wehle (1983); Sealy (1973c) 

^Reviewed by Gaston (1985), Harris and Birkhead (1985); Barrett and Rikardsen (1992), Harris and Hislop (1978), 
Nettleship (1972) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



45 



De Santo and Nelson 



Chapter 3 



Reproductive Ecology of Auks 



where they receive additional care for several weeks until 
independent (table 8). 

Marbled Murrelet chicks are semi-precocial and remain 
in the nest where they are cared for by both parents until 
fledging at 27 to 40 days of age (tables 7 and 8). The chick is 
apparently able to thermoregulate at an early age as continuous 
brooding by the parents ceases after 1-3 days (Naslund 
1993a; Nelson and Hamer, this volume a; Nelson and Peck, 
in press; Simons 1980; S.W. Singer, pers. comm.). The 
period of continuous brooding is shorter than most alcids 
raised in the nest (semi-precocial and intermediate species) 
and is comparable to that of the precocial murrelets (table 
7). Growth data have been collected for only four nestlings, 
the preliminary data suggest murrelets grow more rapidly 
than comparable alcids (Hamer and Cummins 1991; Hirsch 
and others 1981; Simons 1980). 

The incubation and nestling periods of semi-precocial 
alcids are related (fig. 6, r 2 = 0.68, P < 0.001), however, the 
precocial and intermediate species do not fit this pattern. (The 
relationship between the incubation and the nestling period 
including the alcids with precocial and intermediate 
developmental modes is not significant [r 2 = 0.19, P < 0.07]). 

Lengthy incubation and nestling periods have been 
attributed to slow rates of development (Manuwal 1979). In 
contrast, Marbled Murrelets appear to have a relatively short 
incubation and nestling period indicating a rapid rate of 
development. However, the nestling stage of the Marbled 
Murrelet can vary between 27 and 40 days and the extended 
growth period may reflect parental difficulty in provisioning 
the nestling (Nelson and Hamer, this volume a; Nelson and 
Hardin 1993a). Barrett and Rikardsen (1992) reported lengthy 
nestling periods of Atlantic Puffins during years of food 
shortages when parents delivered less food to their young. 

Estimates of mean fledging success range from 66 percent 
for Crested Auklets to over 90 percent for the Ancient 
Murrelets and Wiskered Auklet (table 6). Causes of pre- 
fledgling mortality include mammalian, avian, and reptilian 
predation (Emms and Verbeek 1989; Evans 1981; Ewins 
and others 1993; Gaston 1994; Jones 1992; Manuwal 1979; 
Sealy 1982; Thorensen 1964), food shortages or starvation 
(Ainley and Boekelheide 1990; Barrett and Rikardsen 1992; 
Manuwal 1979; Vermeer 1980), adverse weather (reviewed 
by Harris and Birkhead 1985), and injury inflicted by adult 
conspecifics (Birkhead and Nettleship 1981). 

Fledging success of Marbled Murrelets has been estimated 
to be 45 percent, a value lower than those of other species 
(table 6). Chicks in 19 nests were monitored in Alaska, 
California, Oregon, Washington, and British Columbia (Nelson 
and Hamer, this volume b). Nearly 25 percent of these young 
were documented or strongly suspected to have been taken 
by predators, three others fell from their nest trees, and one 
died of unknown causes. 

Although juvenile survival is difficult to observe and 
measure, banding studies have provided estimates of survival 
for seven alcid species ranging from below 1 percent for 
Atlantic Puffins to a high of 65 percent for Cassin's Auklets 



(A 

>• 
(0 

■o 



■a 
o 



a. 



c 



CO 

CD 




Incubation period (days) 

Figure 6 — Relationship between incubation and nestling periods for 1 9 
alcids (see tables 5 and 9 for values). 



(table 6). Juvenile survival has not been estimated for Marbled 
Murrelets. It is likely that recently fledged Marbled Murrelets 
experience some mortality on their trip from inland nest trees 
to the ocean. Forty-six juveniles in postfledging plumage 
have been found on the forest floor or in parking lots, 
presumably following unsuccessful attempts at fledging from 
inland nests (see Nelson and Hamer, this volume b). An 
indication of low fledgling success is also reflected in at-sea 
surveys conducted in California, Oregon, and Alaska in which 
only 1 to 5 percent of birds on the water were observed to be 
recently fledged young (Nelson and Hardin 1993a; Ralph 
and Long, this volume; Strong and others, in press; Strong 
and others, this volume; Varoujean and Williams, this volume). 

Although the average number of young produced by 
alcid pairs can be high in some years, it is common for 
productivity to be variable among years, and extremely 
low reproductive rates are not uncommon. For example, 
over a 12-yr period on the South Farallon Islands, Common 
Murre pairs produced an average of 0.86 young per season, 
but values over this time fluctuated from a high of 0.9 to a 
low of 0.1 fledglings (Boekelheide and others 1990). 
Complete nesting failures have been documented as well 
(Bergman 1971). 

Summarizing the available information on the repro- 
duction of Marbled Murrelets, it appears that this alcid has a 
low reproductive rate. This species lays only one egg, has 
relatively low hatching success and fledgling survival, and a 
low rate of recruitment of young into the population. However, 
some of the Marbled Murrelet reproductive data were collected 
during El Nino periods (Ainley 1990). Because reproduction 
of other alcids has been documented to be low during such 
times (Boekelheide and others 1990), values reported for 
Marbled Murrelets may reflect a similar depression. 
Reproduction in "good" years may be higher. On the other 



46 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



De Santo and Nelson 



Chapter3 



Reproductive Ecology of Auks 



hand, reproduction of the Marbled Murrelet is not likely to 
exceed that of other alcids with comparable reproductive 
traits. See Beissinger (this volume) for a discussion on the 
possible reproductive rates of Marbled Murrelets using general 
reproductive parameters. 

If the Marbled Murrelet does have a low rate of 
reproduction, then it is quite possible that this species will 
have difficulty recovering from significant population 
declines, and steps should be taken to minimize the impact 
of human activity on the production of murrelet young. To 
completely address this issue, however, thorough study of 
the reproductive biology of this species is needed. Long- 
term studies of individually marked, nesting Marbled 
Murrelets and their young need are required. Effects of 
natural and human-induced perturbations on the reproductive 
ecology of this species can then be better understood. 



Acknowledgments 

George Divoky, Harry Carter, Steve Singer, Barney 
Dunning, and Scott Hatch assisted us in locating information 
and references for this review. We thank Bob Peck, Dave 
D'Amore, George Divoky, Jeff Grenier, Scott Hatch, and 
George Hunt for reviewing earlier drafts of this manuscript. 
Support for preparation of this manuscript was provided by 
the Oregon Department of Fish and Wildlife, USDA 
Forest Service, USDI Bureau of Land Management, and 
U.S. Fish and Wildlife Service. Support for preparation of 
this manuscript was provided by the Oregon Department of 
Fish and Wildlife, USDA Forest Service, USDI Bureau of 
Land Management, and U.S. Fish and Wildlife Service. 
This is Oregon State University Agricultural Experiment 
Station Technical Paper number 10,539. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



47 



Chapter 4 

Nesting Chronology Of The Marbled Murrelet 



Thomas E. Hamer 1 



S. Kim Nelson 2 



Abstract: We compiled 86 breeding records of eggs, downy young, 
and fledgling Marbled Murrelets (Brachyramphus marmoratus) 
for which the fledging date could be estimated. Records were 
collected from California (n = 25), Oregon (n = 13), Washington 
(n = 13), British Columbia (n = 23), and Alaska (n = 12). The 
number of young fledging increased rapidly from 6 June to 19 July 
and peaked by the 10-day period beginning 19 July. A second peak 
in the number of young fledged was observed for the 1 0-day period 
beginning 18 August, with a rapid decrease in late August and 
early September. From these results, a gradual accumulation of 
fledglings on the ocean would be observed from 30 May until 16 
September. By 27 August, only 84 percent of the juveniles in a 
given year would be expected to be counted at sea. In California 
and Oregon, it is likely that two distinct periods of breeding 
activity result from some proportion attempting to lay a second 
clutch, or pairs renesting after nesting failure. The breeding season 
appears to be much longer and less synchronous than that of many 
other members of the alcid family. We conclude that egg-laying 
and incubation spanned a long period, beginning 24 March and 
ending 25 August, with the nestling period beginning 23 April and 
ending with a fledging record on 21 September, a breeding period 
of 182 days. 



Detailed information on the breeding chronology of the 
Marbled Murrelet (Brachyramphus marmoratus) has been 
limited. More recently, a large amount of unpublished 
information has been collected from research projects being 
conducted throughout the range of the Marbled Murrelet. In 
this paper we review the nesting chronology of the Marbled 
Murrelet using data from four published studies that 
specifically addressed the topic (n = 26 records), additional 
published breeding records (n = 26), and unpublished breeding 
accounts (n = 35) of the Marbled Murrelet from Alaska, 
British Columbia. Washington, Oregon, and California. This 
information was used to estimate the fledging dates for each 
record collected. We then summarized these fledging dates 
and used them to construct the timing of egg laying, 
incubation, and nestling period for each state and province 
to more accurately document the breeding chronology of 
the Marbled Murrelet. 

An understanding of the breeding chronology of the 
Marbled Murrelet is important for several reasons. To learn 
more about the nesting ecology of this species, it is important 
to understand the timing and lengths of breeding activities 
and what factors affect this timing. To avoid disturbance to 
nesting colonies from land management activities, land 



1 Research Biologist. Hamer Environmental. 2001 Highway 9, Mt. 
Vernon. WA 98273 

2 Research Wildlife Biologist, Oregon Cooperative Wildlife Research 
Unit, Oregon State University. Nash 104. Corvallis, OR 97331-3803 



managers will need to know the timing of the incubation and 
nestling periods for each geographic area. Biologists 
conducting nest searches and gathering information on nesting 
biology will want to know the optimum period to conduct 
these activities. In addition, biologists conducting marine 
surveys to collect information on the numbers of juveniles 
observed at sea, as an indication of reproductive success, 
will need nesting chronology data to determine the appropriate 
timing of these surveys. 

Several studies have addressed the breeding chronology 
of the Marbled Murrelet. In British Columbia, Sealy (1974) 
collected female specimens at sea and examined the 
maturation of the ovarian follicles, the size of the brood 
patch, and the date juveniles were first observed on the 
ocean. Carter and Sealy (1987b) used 41 records of downy 
young and grounded fledglings to estimate the timing of 
breeding. Carter and Erickson (1992) used additional records 
of grounded chicks and fledglings to estimate the timing of 
egg laying, incubation, and chick rearing for murrelets in 
California. In addition, the breeding phenology of the murrelet 
in British Columbia was reviewed by Rodway and others 
(1992), adding some records to the previous work of Sealy 
(1974) and Carter and Sealy (1987b). 

Methods 

We compiled unpublished breeding records from 
intensive field work conducted on murrelets over the last 
five years, and from published observations of breeding 
records, downy young and fledgling Marbled Murrelets (table 
1). Fledging dates were estimated using a 30-day incubation 
period and a 28-day nestling period (Sealy 1974, Simons 
1980, Hirsch and others 1981). For example, if a grounded 
chick was found and, from the description, was estimated to 
be 10 days old, we added 1 8 days to determine the approximate 
fledging date. Similarly, if an egg-laying date was available, 
but the egg was destroyed before hatching, we added 30 
days for incubation and a 28-day nestling period to estimate 
the fledging date. The initiation of egg laying, incubation, 
and hatching were estimated for each record in the same 
manner. In some cases, where the size and plumage of a 
chick were not described completely, a subjective estimate 
of the age of the chick was made. These records were given a 
higher error estimate. Fledging dates were used for the analysis 
only if the date could be estimated with an error of <8 days 
so that the results would accurately describe the nesting 
chronology. Records were not used when a description of 
the plumage or size of the chick was not available. Records 
derived from juveniles first observed at sea were used only if 
the researcher was conducting weekly boat surveys within 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



49 



Hamer and Nelson 



Chapter 4 



Nesting Chronology 



Table 1 — Inland and at-sea records of eggs, downy young, and fledglings of Marbled Murrelets in North America (n = 
86) where the fledging date could be estimated. The term "grounded" under "Record type" refers to chicks or fledglings 
that were found on the ground 



Location 


Sources 


Record type 


Estimated 
fledge date 


California 








Big Basin State Park 


Singer (pers. comm) 


Grounded chick 


5/20/89 


Big Basin State Park 


Binford and others (1975) 


Grounded chick 


8/20/74 


Big Basin State Park 


Carter and Erickson (1992) 


Grounded fledgling 


6/12/76 


Big Basin State Park 


Carter and Erickson (1992) 


Grounded fledgling 


6/14/79 


Portola State Park 


Anderson (1972) 


Grounded fledgling 


6/15/57 


Big Basin State Park 


Carter and Erickson (1992) 


Grounded fledgling 


6/17/73 


Portola State Park 


Desante and LeValley (1971) 


Grounded fledgling 


6/27/71 


Gasquet Ranger District 


Craig (pers. comm.) 


Grounded fledgling 


6/30/92 


Big Basin State Park 


Carter and Erickson (1992) 


Grounded fledgling 


7/04/76 


Sequoia Park 


Carter and Erickson (1992) 


Grounded fledgling 


7/04/24 


Memorial County Park 


Singer (pers. comm.) 


Grounded fledgling 


7/19/88 


Big Basin State Park 


Carter and Erickson (1992) 


Grounded fledgling 


8/11/82 


Prairie Creek State Park 


Carter and Erickson (1992) 


Grounded fledgling 


8/13/84 


Big Basin State Park 


Singer (pers. comm.) 


Grounded fledgling 


8/15/90 


Big Basin State Park 


Singer and Verado (1975) 


Grounded fledgling 


8/18/60 


Big Basin State Park 


Singer (pers. comm.) 


Grounded fledgling 


8/25/92 


Big Basin State Park 


Erickson and Morlan (1978) 


Grounded fledgling 


8/31/77 


LomaMar 


Carter and Erickson (1992) 


Grounded fledgling 


8/31/85 


Big Basin State Park 


Singer (pers. comm.) 


Grounded fledgling 


9/03/88 


Big Basin State Park 


Singer (pers. comm.) 


Grounded fledgling 


9/05/93 


Big Basin State Park 


Singer and Verado (1975) 


Grounded fledgling 


9/09/74 


Big Basin State Park 


Singer (pers. comm.) 


Grounded unknown 


5/18/84 


Big Basin State Park 


Singer (pers. comm.) 


Nest observed 


6/07/92 


Big Basin State Park 


Singer (pers. comm.) 


Nest observed 


7/03/91 


Elkhead Springs 


Chinnici (pers. comm.) 


Nest observed 


8/23/92 


Waddell Creek 


Naslund (1993a) 


Nest observed 


8/26/89 


Oregon 








Five Rivers 


Nelson and Peck (in press) 


Grounded chick 


7/07/90 


God's Valley 


Nelson and Peck (in press) 


Grounded fledgling 


9/13/90 


Powers Ranger District 


Nelson and Peck (in press) 


Grounded fledgling 


7/26/92 


North Fork Siuslaw River 


Jewett(1930) 


Grounded fledgling 


9/08/18 


Siletz 


Heinl (1988) 


Grounded fledgling 


9/21/87 


Five Rivers 


Nelson and Peck (in press) 


Nest observed 


6/22/91 


Boulder and Warnicke Creeks 


Nelson and Peck (in press) 


Nest observed 


7/08/92 


Iron Mountain 


Nelson and Peck (in press) 


Nest observed 


7/09/92 


Cape Creek 


Nelson and Peck (in press) 


Nest observed 


7/20/91 


Siuslaw River 


Nelson and Peck (in press) 


Nest observed 


8/29/91 


Valley of The Giants 


Nelson and Peck (in press) 


Nest observed 


8/30/90 


Valley of The Giants 


Nelson and Peck (in press) 


Nest observed 


7/09/91 


Siuslaw River 


Nelson and Peck (in press) 


Nest observed 


9/09/91 


Washington 








Rugged Ridge 


Leschner and Cummins (1992a) 


Grounded chick 


7/09/82 


Helena Creek 


Reed and Wood( 1991) 


Grounded chick 


7/22/89 


Baker Lake 


Hamer (pers. obs.) 


Grounded chick 


7/24/90 


Heart of the Hills Trail 


Hamer (pers. obs.) 


Grounded chick 


8/07/91 








continues 



50 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapcer4 



Nesting Chronology 



Table 1 — continued 



Location 


Sources 


Record type 


fledge date 


Aberdeen 


Leschner and Cummins (1992a) 


Grounded chick 


8/09/83 


Matheny Creek 


Leschner and Cummins (1992a) 


Grounded fledgling 


7/17/81 


North Fork Quinault 


Leschner and Cummins (1992a) 


Grounded fledgling 


7/23/86 


SedroWooUey 


Hamer (pers. obs.) 


Grounded fledgling 


7/24/90 


North Rosedale 


Leschner and Cummins (1992a) 


Grounded fledgling 


7/24/71 


Federal Way 


Leschner and Cummins (1992a) 


Grounded fledgling 


8/07/74 


Long Beach 


Ritchie (pers. comm.) 


Nest observed 


6/22/93 


Lake 22 


Hamer (pers. obs.) 


Nest observed 


7/1 8/90 


Lake 22 


Hamer (pers. obs.) 


Nest observed 


8/27/90 


British Columbia 








Langara Island 


Sealy (1974) 


Egg development 


7/20/71 


Langara Island 


Sealy(1974) 


Egg development 


7/26/71 


Langara Island 


Sealy (1974) 


Egg development 


7/30/71 


Langara Island 


Sealy (1974) 


Egg development 


8/02/70 


Langara Island 


Sealy (1974) 


Egg development 


8/02/70 


Langara Island 


Sealy (1974) 


Egg development 


8/04/71 


Langara Island 


Sealy (1974) 


Egg development 


8/06/71 


Langara bland 


Sealy (1974) 


Egg development 


8/08/70 


Langara Island 


Sealy (1974) 


Egg development 


8/19/70 


Langara Island 


Sealy (1974) 


Egg development 


8/30/71 


Vancouver Island 


Harris (1971) 


Grounded chick 


8/30/67 


Chilli wack 


Rodway and others (1992) 


Grounded fledgling 


7/07/87 


Hope Village 


Rodway and others (1992) 


Grounded fledgling 


7/12/47 


Queen Charlotte Island 


Sealy (1974) 


Grounded fledgling 


7/15/47 


Sayward 


Rodway and others (1992) 


Grounded fledgling 


7/1 5/86 


Karen Range 


Paul Jones (pers. comm.) 


Nest observed 


8/20/93 


Frederick Island 


Drent and Guiguet (1961) 


Sea observation 


6/28/61 


Frederick Island 


Drent and Guiguet (1961) 


Sea observation 


6/28/61 


Barclay Sound 


Carter (1984) 


Sea observation 


6/28/80 


Barclay Sound 


Carter (1984) 


Sea observation 


7/04/79 


Langara Island 


Sealy (1975a) 


Sea observation 


7/06/70 


Langara Island 


Sealy (1975a) 


Sea observation 


7/07/71 


Cox Island 


Brooks (1926b) 


Sea observation 


7/22/20 


Alaska 








Montague Island 


Mendenhall(1992) 


Egg development 


8/10/77 


Skagway 


Mendenhall (1992) 


Grounded fledgling 


7/18/87 


Afognak Island 


Carter and Sealy (1987b) 


Grounded fledgling 


8/18/76 


Cordova Airport 


Carter and Sealy (1987b) 


Grounded fledgling 


8/20/78 


Port Chatham 


Johnston and Carter (1985) 


Nest observed 


Unknown 


Naked Island 


Naslund and others On press) 


Nest observed 


7/23/92 


Kodiak Island 


Naslund and others (in press) 


Nest observed 


7/24/92 


Kodiak Island 


Naslund and others (in press) 


Nest observed 


8/03/92 


Naked Island 


Naslund and others (in press) 


Nest observed 


8/13/91 


East Amatuli Island 


Hirsch and others (1981) 


Nest observed 


8/16/79 


East Amatuli Island 


Simons (1980) 


Nest observed 


8/27/78 


Auke Bay 


Speckman (pers. comm.) 


Sea observation 


7/1 0/93 


Auke Bay 


Speckman (pers. comm.) 


Sea observation 


7/29/92 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



51 



Hamer and Nelson 



Chapter 4 



Nesting Chronology 



the same geographic area and surveys commenced before 
the fledging period of the breeding season. Therefore, only 
one datum was used for each boat survey in each year, 
indicating the first fledging date for that season. These 
observations made up a small portion of the records used. 

Birds were assumed to be juveniles when a grounded 
juvenile was reported, and a plumage description was not 
provided. Fortunately, the use of the word "juvenile" to 
describe a young bird that had lost its downy plumage was 
consistent throughout the literature. We assumed that grounded 
juveniles were recently fledged individuals. The majority of 
records of grounded juveniles included descriptions of 
remaining down on the back and head and the presence of an 
egg tooth, confirming recent fledging. 

Records of eggs collected or found and of incubating 
adults were not used because the development period of the 
embryo was unknown. But the presence of a postovulatory 
follicle, unshelled egg in the oviduct, or mature follicles 
from collected females indicated that egg laying would 
occur in 1-3 days (Sealy 1974). Therefore, egg-laying dates 
were estimated for these records by adding 2 days to the 
collection date and then estimating the fledging date by 
adding 58 days. 

Sealy (1974) obtained 12 breeding records in British 
Columbia. He collected murrelet specimens in weekly 
intervals during the breeding season between 30 April and 
10 August. He examined the size of the brood patch in male 
and females and the size and maturation of the largest 
follicle in each ovary of females to estimate the timing of 
egg laying, incubation, and chick rearing. Observations of 
adults carrying fish in their bills at dusk were used to 
estimate the hatching dates of eggs. The first fledglings 
observed on the water were used as an indication of the 
earliest fledging dates. For our summary, we used only 
records from Sealy in which the size and maturation of the 
ovarian follicles of females enabled an accurate estimate of 
the egg-laying date, and two cases in which juveniles were 
first observed at sea. Observations of brood patch 
development and fish-carrying behavior were not used 
because the accuracy of these methods in estimating the 
nesting stage of the Marbled Murrelet is unknown. 

Carter and Erickson (1992) reviewed the breeding 
chronology of the murrelet in California by examining inland 
records of downy young and grounded juveniles, molt 
conditions of museum specimens, and records of juveniles 
observed at sea from 1892 to 1987. Carter and Erickson used 
28 days for a nestling period and 30 days for an incubation 
period to estimate breeding chronology. In addition, 41 inland 
records of downy young and fledgling murrelets from 1918 
to 1986 were summarized in North America by Carter and 
Sealy (1987b). Records of grounded nestlings and juveniles 
from these studies, in which an accurate fledging date could 
be estimated, were used in this analysis. 

Fledgling dates were estimated using 86 breeding records 
from California (n = 25), Oregon (n = 13), Washington (n = 
13), British Columbia (n = 23), and Alaska (n = 12) (table 



1). Records used for this analysis included observations of 
the presence of a postovulatory follicle (n = 9) or unshelled 
egg in the oviduct of collected females (n = 2), known egg- 
laying dates (n = 2), known egg-hatching dates (n = 5), 
observations of young on nests (n = 4), grounded chicks (n = 
9), grounded fledglings (n = 35), juveniles observed to fledge 
(n = 10), and dates that juveniles were first observed at sea 
from marine census studies (n = 9). Fledging dates for a 
large proportion of the records (67 percent) could be estimated 
accurately because the egg-laying date or the hatching date 
was known, the young were accurately aged on the nest, a 
grounded fledgling was recorded, or a nestling was actually 
observed to fledge (n = 56). 

Results 

Records of the earliest and latest breeding records of the 
Marbled Murrelet were all collected from California and 
Oregon. The earliest fledging record in North America was 
of a grounded downy chick discovered in Big Basin State 
Park in central California on 20 May 1989. The chick was 
estimated to be at least 2 weeks old (S.W. Singer, pers. 
comm.). The next fledging was a nestling observed to fledge 
on 7 June 1992 from a nest also in Big Basin State Park 
(S.W. Singer, pers. comm.) (fig. 1). A record also exists of a 
grounded murrelet in Big Basin State Park on 18 May 1984, 
but it was not clear whether the bird was an adult or juvenile 
(S.W. Singer, pers. comm.). The next four earliest fledging 
dates, from 12 June to 17 June, were all from Big Basin and 
Portola State Parks in central California (Anderson 1972; 
Carter and Erickson 1992; S.W. Singer, pers. comm.). 

The latest fledging date was a record of a fledging found 
on 21 September 1987 in a parking lot in the town of Siletz, 
Oregon (Heinl 1988, Nelson and Peck, in press). The next 
four latest fledging dates, from 30 August to 9 September, 
were all recorded from California and Oregon (Carter and 
Erickson 1992, Erickson and Morlan 1978, Jewett 1930, 
Nelson and others 1992, Singer and Verardo 1975). 

The number of young observed or estimated to have 
fledged for all North American records was summarized for 
each 10-day period during the breeding season. Fledging 
rates increased rapidly from 6 June to 19 July, and peaked by 
the 10-day period beginning 19 July (fig. 1). A possible 
second peak in the number of young fledged was evident for 
the 10-day period beginning 18 August, with a rapid decrease 
in the number of young leaving nests in late August and 
early September. Egg laying and incubation began 24 March 
and ended 25 August, with the nestling period beginning 23 
April and ending with a fledging record on 2 1 September, a 
breeding period of 182 days. 

An analysis of the cumulative number of young fledged 
in North America for each 10-day period was used to predict 
the percentage of total juveniles that would be observed in 
the marine environment during different periods of the 
breeding season. This analysis also demonstrated the broad 
nesting chronology of the Marbled Murrelet (fig. 2). The 



52 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter4 



Nesting Chronology 




5/20 6/9 6/28 7/19 8/8 8/28 9/17 

5/30 6/19 7/9 7/29 8/18 9/7 

DATE 



Figure 1 — Fledging dates of Marbled Murrelet young from nests in North America 
(n = 86) in 10-day intervals. The date displayed for each histogram is the beginning 
of a 10-day period. Records were used only if the error in estimating the fledging 
date was <8 days. 




5/20 6/9 6/28 7/19 8/8 8/28 9/17 10/7 

5/30 6/19 7/9 7/29 8/18 9/7 9/27 

DATE 

Figure 2 — The cumulative number of Marbled Murrelet young fledged from nests in 
North America (n = 86) in 10-day intervals. The cumulative percentage of total young 
fledged in each 1 0-day interval is shown at the top of each histogram. The date displayed 
for each histogram is the beginning of a 10-day period. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



53 



Hamer and Nelson 



Chapter 4 



Nesting Chronology 



cumulative percentage change in the total number of young 
fledged from 20 May to 18 June was low, increasing only 1- 
6 percent between each 10-day period. A gradual accumulation 
of fledglings on the ocean would be predicted from 19 June 
until 27 August. During this period the cumulative percent 
change in the number of juveniles fledging between each 10- 
day period was very consistent, ranging from 9 to 15 percent. 
The latter part of the nesting season from 28 August to 26 
September was similar to the beginning of the season, with 
the cumulative percent change in the number of young fledged 
ranging from 1 to 5 percent. For all states and provinces 
combined, the results show that by 27 August, only 84 
percent of the juveniles in a given year would be expected to 
be counted at sea using marine census techniques. For 
California and Oregon, a census of all juveniles would not 
occur until the third week of September (fig. 3). In Washington, 
British Columbia, and Alaska, a full census of all juveniles 
would not occur until the third week of August (fig. 3). 

Discussion 

The breeding season of the Marbled Murrelet appears to 
be much longer than that of many other members of the alcid 
family. The long breeding period indicates that the 



synchronous nesting exhibited by many colonial and semi- 
colonial nesting seabirds is likely not a characteristic of the 
breeding biology of the Marbled Murrelet. Few active nests 
have been found within the same stand to verify this. In one 
instance, two active nests were found only 100 m apart in the 
same forest stand in Washington. The first nest fledged a 
young murrelet on 18 July with the second nest fledging on 
27 August, a 39-day difference (Hamer, pers. obs.). A trend 
toward a shorter breeding season in the northern range of the 
murrelet is apparent as one examines the fledging dates from 
California to Alaska (table 2, fig. 3). The longest breeding 
period was observed in California. Oregon had the next 
longest breeding period. The breeding season in Alaska was 
64 days shorter than California. 

For California and Oregon, a larger sample of breeding 
records is needed to further refine the breeding period and 
conduct statistical tests to determine whether two distinct 
breeding seasons exist. The fact that each breeding period 
was similar in total length supports the idea that there are 
two periods. When examined separately, the second breeding 
period was only 12 and 16 days shorter than the first breeding 
period for Oregon and California, respectively. The first 
breeding period in California (103 days) was similar to the 
total breeding season in Alaska (106 days). 



Mar Apr May Jun Jul Aug Sep 
1 15 31 15 30 15 31 15 30 15 31 15 31 15 



Incubation 

+ 



ALASKA 
n=12 



BRITISH COLUMBIA 
n=23 



WASHINGTON 
n=13 



OREGON 
n=13 



Nestling 

H 



Incubation 

-I 



Nestling 
-I 



Incubation 
1 



Nestling 

+ 



Incubation 

■» 



Nestling 

+ - 



Incubation 
+— 



CALIFORNIA 
n=25 



Nestling 
+— 



1 15 31 15 30 15 31 15 30 15 31 15 31 15 
Mar Apr May Jun Jul Aug Sep 



Figure 3 — Breeding phenology of the Marbled Murrelet in North America organized by state and 
province. The median for each incubation and nestling period is shown. 



54 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter 4 



Nesting Chronology 



Table 2 — Number of Marbled Murrelt young observed and estimated to have fledged from nests in North America in 10-day intervals and listed 
by state and province. Records (n = 86) were used only if the error in estimating the fledging date was <8 days. The total length of the breeding 
period is listed under each state or province. The date displayed is the beginning of each lO^lay period 



State or province 










Number of birds observed in each 10-day 


interval 












5/2 


5/30 


6/9 


6/19 


6/29 


7/9 


7/19 


7/29 


8/8 


8/18 


8/28 


9/7 


9/17 


9/27 


Alaska 

106 days 
British Columbia 

1 18 days 








3 


4 


2 
3 


2 
3 


2 
5 


3 

1 


3 

2 


2 








Washington 
124 days 








1 




3 


5 


2 


1 


, 










Oregon 
149 days 








1 


3 


1 


2 








2 


3 


1 




California 
170 days 


1 


1 


4 


1 


4 




1 




3 


5 


4 


1 







Explanations for the presence of two distinct breeding 
periods include: (1) small sample sizes, (2) variations in the 
timing of breeding of murrelets between years (as suggested 
by Carter and Erickson 1992), or (3) variation in oceanic and 
environmental conditions that promote breeding within these 
two distinct periods. Small sample sizes may not adequately 
explain this phenomenon, because it occurs only in the 
southern portions of the murrelet range, with sample sizes 
very similar to regions to the north. If the between-year 
variation in the timing of breeding was responsible for this 
trend, some young would have been expected to fledge over 
a period of years within the 1- to 2-week gap between 
breeding periods. A distinct proportion of the population 
nesting during each period is also possible. It is not clear 
what environmental or biological agents might cause this to 
occur. However, the most likely explanation is that some 
proportion of murrelets may attempt to lay a second clutch 
within the same breeding period. It is also possible that pairs 
with failed nests attempt to renest. The longer breeding 
season available for the murrelet in Oregon and California 
may make renesting more likely than in the northern regions 
of the range. The gap in breeding chronology may give 
females enough time to develop a new egg and select a 
different nest site. The shorter incubation and nestling period 
in the Marbled Murrelet, when compared to that of other 
species, such as the puffins and auklets, may make double 
brooding more feasible. 

Egg replacement is a regular occurrence in ledge-nesting 
Alcids (Johnsgard 1987). Tuck (1960) estimated that 44 
percent of Thick-billed Murres (Uria lomvia) pairs lost at 
least one egg during a 32-day period, with 30 percent of 
pairs laying one replacement egg; 1 1 percent laid two 
replacements, and the remaining 3 percent deserted or did 
not lay again. For higher arctic forms there are probably no 
opportunities to renest because of a short breeding season 
(Johnsgard 1987). The Xantus' Murrelet (Synthliboramphus 
hypoleucus) may lay two broods because egg laying has 
been observed until July, and Murray and others (1983) 



found evidence of occasional egg replacement in this species. 
Both Sealy (1975a) and Gaston (1992) found no evidence of 
replacement clutches in the Ancient Murrelet (S. antiquus). 
If renesting and egg replacement does occur in the Marbled 
Murrelet, it will affect the interpretation of inland survey 
data and at-sea census results and population modeling for 
this species. 

Breeding Phenology Dates 

From the information presented above, we propose the 
following dates for the breeding phenology of the murrelet 
by state and province (fig. 3). In California, we estimated 
that the total breeding season lasted approximately 170 days. 
The first breeding period was 103 days long while a possible 
second period was 87 days long. The breeding periods were 
separated by 8-11 days. Incubation commenced 24 March 
and ended 13 August. The nestling period began 23 April 
and ended 9 September. 

In Oregon, incubation was estimated to begin on 26 
April and last until 25 August (fig. 3). The nestling period 
was estimated to begin 26 May and end on 21 September. 
The total breeding season length was 21 days shorter than 
that in California and was approximately 149 days long. The 
two possible periods of breeding activity were separated by 
only 6 days. The earliest recorded fledging date in Oregon is 
of a nestling observed to fledge from a nest on 22 June 1993 
(Nelson, pers. obs.). 

North of California and Oregon, the length of the breeding 
season was more restricted (fig. 3) (table 2). In Washington, 
the breeding season might appear shorter because of the 
smaller sample of breeding records used to predict fledging 
dates. However, it is probably similar to that found in British 
Columbia. Incubation was estimated to begin 26 April and 
end 30 July. The nestling period began 26 May and ended on 
27 August. The total length of the breeding season was 124 
days long, 25 days less than Oregon. The earliest fledging 
record is a nestling observed to fledge on 22 June 1993 
(Ritchie, pers. comm.). The latest record is that of a nestling 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



55 



Hamer and Nelson 



Chapter 4 



Nesting Chronology 



recorded on video tape fledging from a nest on 27 August 
1990 (Hamer, pers. obs.). 

In British Columbia, incubation was estimated to 
commence on 2 May and ended by 4 July (fig. 3). The nestling 
period began 1 June and ended by 30 August. The total breeding 
season was approximately 1 18 days. The earliest fledging date 
recorded was for a juvenile collected by Drent and Guiguet 
(1961) on 28 June 1961. The next five earliest records are of 
all juveniles observed or collected at sea. The first grounded 
fledging was not recorded until 7 July 1987 (Rodway and 
others 1992). The latest two fledging dates occurred on 30 
August. One was of a chick that was discovered after a tree 
was felled (Harris 1971), and the second was a description of 
the follicle development of a female (Sealy 1974). 

The length of the Alaska nesting season was greatly 
restricted and was estimated to be only 106 days. Incubation 
was estimated to begin on 14 May and end by 30 July. The 
nestling period ranged from 13 June to 27 August. The 
earliest fledging record was a juvenile observed at sea on 10 
July 1993. The next earliest was a grounded fledging observed 
on 18 July 1987 (Mendenhall 1992). The latest estimated 
fledging date of 27 August 1978 was from an active ground- 
nest observed by Simons (1980). 

Sealy (1974) discovered that murrelets laid eggs in 
British Columbia over a 6- to 7-week period beginning 15 
May and ending in late June or early July. Adults with fish in 
their bills were first observed on 16 June. Young birds were 
first observed on the water on 6 July 1970 and on 7 July 
1971. Sealy concluded that the period of egg laying and 
incubation began around 15 May and lasted until 3 1 July. He 
estimated that the period of hatching and chick rearing started 
15 June and ended around 15 August. Fledging began in the 
first week of July and continued to some time after 15 
August (Sealy 1974). Sealy's study took place in the most 
northern portion of coastal British Columbia and thus may 
be more representative of Alaska than British Columbia. His 
breeding dates closely resemble the breeding chronology we 
report for Alaska (fig. 3). Rodway and others (1992) reached 
similar conclusions. 

For North America, Carter and Sealy ( 1 987b) calculated 
that egg-laying dates began between 15 and 22 April. The 
latest fledging dates reported were 8 and 9 September. 
Additional breeding records that we collected extend these 
dates by several weeks in California and Oregon. However, 
observations by Carter and Sealy of adults holding fish at 
sea as late as 17 September and records of several young still 
in downy plumage on 4 and 13 September led them to 
believe that the nestling period of the murrelet may extend 
into late September. Carter and Sealy concluded that murrelets 
may nest earlier and have a longer breeding season south of 
British Columbia. Carter (1984) speculated that the breeding 
season was protracted in southern British Columbia as 
compared to northern British Columbia. 

Carter and Erickson (1992) estimated that egg-laying 
dates ranged from 15 April to 12 July in California, and 
hatching from 15 May to 10 August. New records that we 
examined extend these dates (fig. 3). Carter and Erickson 



found that fledging dates fell into two periods, 1 2 June to 4 
July, and 1 1 August to 9 September. They believed the two 
different fledging periods in California were due either to 
low sample size, unknown factors affecting the grounding of 
fledglings, or variation between years in the timing of 
breeding. They concluded that egg laying begins earlier in 
California than farther north. An earlier breeding chronology 
was further supported by an examination of 45 museum 
specimens which showed an earlier timing of prealternate 
body molt for birds in California. 

Juvenile/Adult Ratios 

Adults molting into winter plumage can make it difficult 
to discriminate between adults and juveniles after 15 August 
(Carter and Stein, this volume). Because of this, for all 
provinces and states combined, 29 percent of the juvenile 
population produced in a given year may go uncounted if 
surveys after 15 August cannot accurately census young 
birds (fig. 2). It is impossible, at this time, with the small 
sample sizes to calculate the percent of young expected to be 
counted at sea during the breeding season for each state or 
province. When collected, this information would be valuable 
to researchers attempting to calculate juvenile/adult ratios or 
model population trends. A full census of juveniles would 
not be possible until after 16 September for California and 
Oregon, and after late August in Washington, British 
Columbia, and California. Sealy (1974) collected an adult 
female in British Columbia on 9 July 1971 that had already 
undergone a nearly complete body molt and was nearly in 
winter plumage. He suggested that this female may have 
undergone a premature body molt after an unsuccessful 
breeding effort. A complete census of juveniles may not be 
necessary for year-to-year comparisons of reproductive 
success. But, if complete censusing is not done, researchers 
should be careful of variations in the timing of breeding 
between years when conducting any annual comparisons. In 
addition, ratios of juveniles to adults observed at sea can be 
adjusted for birds that have not yet fledged (Beissinger, this 
volume) which may aid population modelling efforts and 
annual comparisons of reproductive success. 

Acknowledgments 

We are grateful for the unpublished accounts of nest 
observations, grounded chicks, and grounded fledglings 
provided to us by Nancy Naslund of the U.S. Fish and 
Wildlife Service, U.S. Department of Interior, Steve Singer 
of the Santa Cruz Mountains Murrelet Group, Bill Ritchie of 
the Washington Department of Wildlife, Paul Jones, Phyllis 
Reed, and Brenda Craig of the USDA Forest Service. Ray 
Miller and Sal Chinnici of the Pacific Lumber Company 
provided information from a nest in northern California. We 
thank Craig Strong, C.J. Ralph, Kathy Kuletz, and Susan 
Speckman for providing records of juveniles first observed 
at sea during marine survey efforts. Joanna Burger, Anthony 
Gaston, and Frank Pitelka provided helpful comments on 
early drafts of this manuscript. 



56 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 5 

Nesting Biology and Behavior of the Marbled Murrelet 



S. Kim Nelson 1 



Thomas E. Hamer 2 



Abstract: We summarize courtship, incubation, feeding, fledging, 
and flight behavior of Marbled Murrelets {Brachyramphus 
marmoratus) using information collected at 24 nest sites in North 
America. Chick development, vocalizations given by adults and 
chicks at the nest, and predator avoidance behaviors are also 
described. Marbled Murrelets initiate nesting as early as March. 
Females lay a single egg and both adults participate in incubation, 
exchanging duties every 24 hours at dawn. Most incubation ex- 
changes occur before sunrise. Chicks hatch after 27-30 days. Adults 
feed chicks single fish up to 8 times daily, but most feedings occur 
at dawn and dusk. Dawn feeding visits occur over a wider time 
period than incubation exchanges, with some occurring as late as 
65 minutes after official sunrise. The timing of incubation ex- 
changes and feeding visits are affected by weather and light condi- 
tions, and adults arrive later on cloudy or rainy days. To minimize 
the attraction of predators, visits to the nest are inconspicuous, 
with adults entering and exiting the nest during low light levels, 
and primarily without vocalizations. Because of this seabird's 
secretive behavior, our understanding of murrelet demography, 
nest site selection, and social interactions remain limited. 



Marbled Murrelets (Brachyramphus marmoratus) are 
unique among seabirds in that they nest in older-aged 
coniferous forests throughout most of their range in North 
America. Little is known about their breeding biology because 
nest sites have only recently (1974) been discovered and 
described (Binford and others 1975; Hamer and Nelson, this 
volume b; Hirsch and others 1981; Nelson and Peck, in press: 
Quinlan and Hughes 1990; Simons 1980; Singer and others 
1991). Marbled Murrelet behavior at the nest has been 
monitored at 24 of 52 (35 tree and 17 ground) active nests 
since 1980; however, only a few accounts have been published 
(Nelson and Peck, in press: Simons 1980; Singer and others 
1991, in press). In this paper, we provide a synthesis of 
information on murrelet behavior patterns, chick development, 
and vocalizations recorded at these 24 nest sites. 

Methods 

We compiled all known data on Marbled Murrelet behavior 
at active nests in North America and combined them with our 
own studies of murrelet nests. Data were summarized from 
two ground and five tree nests in Alaska (Hirsch and others 
1981: Naslund, pers. comm.: Simons 1980), one tree nest in 
British Columbia (P. Jones, pers. comm.), two tree nests in 
Washington (Hamer and Cummins 1 99 1 ; Ritchie, pers. comm.), 



1 Research Wildlife Biologist, Oregon Cooperative Wildlife Research 
Unit, Oregon State University. Nash 104, Corvallis, OR 97331-3803 

: Research Biologist, Hamer Environmental. 2001 Highway 9, Mt. 
Vernon. WA 98273 



nine tree nests in Oregon (Nelson, unpubl. data; Nelson and 
Peck, in press), and five tree nests in California (Kems. pers. 
comm.; Naslund 1993a; Singer and others 1991, in press; 
S.W. Singer, pers. comm.) (table 1). Information on pah- 
bonding and courtship are also summarized. 

Active nests were located by observing murrelets land in 
trees, finding eggshells on the ground and subsequently locating 
the nest, using radio telemetry, or by incidental observations. 
Fifteen of the nests were found during the egg stage and 9 
during the nestling stage. Some nests were intensively 
monitored, others only intermittently. Data recorded at many 
nests included time and duration of incubation exchanges 
and feeding visits, behavior of chicks and adults, flight 
behavior, and vocalizations. Weather conditions (percent 
clouds, precipitation, temperature, wind) were also recorded 
for comparison with the timing and duration of murrelet 
activity at the nest. Means, standard errors, and ranges were 
calculated for numerical data such as the timing of incubation 
exchanges and feeding visits in relation to sunrise and sunset, 
and the length of these encounters at nests. 

Results 

Pair Bonding and Courtship Behavior 

Little is known about when and how Marbled Murrelets 
pair. Murrelets are primarily observed in groups of two 
throughout the year, both in the forest and on the water. 
Many pairs on the water have included a male and female, 
and were assumed to be mated (Carter 1984; Carter and 
Stein, this volume; Sealy 1975a). Some of these "pairs" 
could also be composed of adults in a temporary social 
association; this is known to occur on the water, especially 
when birds are not feeding (Carter, pers. comm.). However, 
we believe that Marbled Murrelets remain paired throughout 
the year based on these year-round pair groups and data 
from other alcids (e.g., Harris and Birkhead 1985). 

Courtship behavior has been observed on the water in 
early spring, when some adults are still in winter plumage, 
as well as throughout the summer. Participation in courtship 
behaviors while in winter plumage is expected because: (1) 
the monomorphic plumage in Marbled Murrelets in not a 
sexually selected trait; and (2) they probably maintain strong 
pair bonds throughout the year. During courtship, pairs join 
closely together (<0.5 m), point their bills in the air, partially 
lift their breasts out of the water, and swim rapidly forward 
(Byrd and others 1974; Nelson, unpubl. data; Van Vliet, 
pers. comm.). Pairs also dive synchronously and surface 
within 1-3 seconds next to one another, suggesting that they 
remain together under water (Van Vliet, pers. comm.). 
Preceding the dive or while swimming together in courtship 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



57 



Nelson and Hamer 



Chapter 5 



Nesting Biology and Behavior 



Table 1 — Marbled Murrelet tree and ground nests by state or province, site, year, and number of days of observation 



State/province 


Number of nests 


Number of observation days 


Reference 


Site/year 


Incubation 


Nestling 




Alaska 










Barren Islands 1978/1979 


2 


33 


56 


Hirsch and others 1981 
Simons 1980 


Naked Island 1991/1992 


5 


14 





Naslund, pers. coram. 


British Columbia 










Caren Range 1993 


1 





14 


P. Jones, pers. comm. 


Washington 










Lake 22 1991 


1 





24 


Hamer and Cummins 1991 


Nemah 1993 


1 





4 


Ritchie, pers. comm. 


Oregon 










Five Rivers 1990/1991 


2 


17 


14 


Nelson and Peck, in press 


Valley of Giants 1990/1991 


2 


26 


6 


Nelson and Peck, in press 


Cape Creek 1991 


1 


9 





Nelson and Peck, in press 


Siuslaw River 1991 


2 





23 


Nelson and Peck, in press 


Boulder Warnicke 1992 


1 





8 


Nelson and Peck, in press 


Copper Iron 1992 


1 





13 


Nelson and Peck, in press 


California 










Big Basin 1989 


2 


45 


4 


Naslund 1993a 
Singer and others 1991 


Father 1991/1992 


2 


13 


25 


Singer and others, in press 


Elkhead 1993 


1 





12 


Kerns, pers. comm. 



dances, birds frequently give soft, synchronous nasal 
vocalizations. Pairs also chase one another in flights just 
above the water surface throughout the spring and summer, 
in what may be courtship behavior (see below about similar 
behaviors exhibited at inland nesting sites). 

Copulation has rarely been observed. It is known to 
occur within trees (n = 1 observation in Alaska; Kuletz, pers. 
comm.) and on the water where it has been observed at least 
15 times (Kuletz, pers. comm.; Naslund, pers. comm; Nelson, 
unpubl. data; Van Vliet, pers. comm.). Preceding and 
following copulation, the birds often vocalize with an 
emphatic, nasal "eeh-eeh" call (Van Vliet, pers. comm.). We 
expect that copulation primarily occurs at the nest based on 
observations from other alcids (Sealy 1975a). 

Before they lay eggs, pairs probably visit the breeding 
grounds, not only to pair and copulate, but also to select nest 
sites. In Oregon, a pair was observed landing on a nest platform 
for 3 mornings in early May, two weeks prior to laying an egg 
at that site. Pre-laying visitation to nests, three to four weeks 
before egg-laying, has been observed in other alcids (Gaston 
1992; Nettleship and Birkhead 1985). 

Egg-Laying and Incubation Behavior 

Marbled Murrelets start to lay eggs as early as March 
(Hamer and Nelson, this volume a). They lay a single egg 



weighing approximately 36-41 g (16-18.5 percent of adult 
weight) (Hirsch and others 1981; Sealy 1975a; Simons 1980). 
The egg is subelliptical in shape, and measures an average of 
59.5 x 37.4 mm (n = 1 1 eggs) and 0.21 mm in thickness (Day 
and others 1983; Hirsch and others 1981; Kiff 1981; Sealy 
1975a; Simons 1980). The egg has a pale-olive green to 
greenish-yellow background color, and is covered with 
irregular brown, black, and purple spots which are more 
prevalent at the larger end of the egg (Becking 1 99 1 ; Binford 
and others 1975; Day and others 1983; Kiff 1981; Nelson 
1991, 1993; Nelson, and Hardin 1993a; Reed and Wood 
1991; Singer and others 1991). 

After the female lays an egg, the pair begins 24-hour 
shifts of incubation duty; one adult broods the egg while the 
other forages at sea (n = 12 nests) (Naslund 1993a, pers. 
comm.; Nelson and Peck, in press; Simons 1980; Singer and 
others 1991). The incubating adults sit on the egg in a 
flattened posture and remain motionless on the nest more 
than 90 percent of the time (n = 4 nests) (Naslund 1993a; 
Nelson and Peck, in press; Simons 1980). Other behaviors 
observed during incubation at most nests include turning the 
egg, re-arranging nest material, and preening. At nests in 
California (n = 1) and Alaska (n = 5), the average occurrence 
of these behaviors were 11, 8, and 1 time(s) per day, 
respectively (Naslund 1993a, pers. comm.). 



58 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter5 



Nesting Biology and Behavior 



At a given nest, the two adults appear to have distinct 
plumage colorations. A light brown and a dark chocolate 
brown adult (sex of each unknown) have been observed 
attending nests on 24-hour shifts, indicating possible sexual 
plumage dichromatism (n = 8 nests) (Fortna, pers. comm.; P. 
Jones, pers. comm.; Naslund 1993a; Nelson 1991,1992; 
Ritchie, pers. comm.; Singer and others 1991). In addition, 
the white patches on the nape of the neck and cheek have 
varied between adults at a single nest, and individuals at 
different nests (n = 7 nests) (Fortna, pers. comm.; Hamer and 
Cummins 1991; P. Jones, pers. comm.; Nelson 1991, 1992; 
Simons 1 980). The variations in these nape and cheek patches 
may provide a means for identifying individuals. 

Murrelets have been observed leaving their egg unattended 
for 3-4 hours during the morning, mid-day, and evening (n = 
4 nests; Naslund 1993a, pers. comm.; Nelson and Peck, in 
press; Simons 1980). Seabirds often leave their eggs unattended 
to maximize foraging time and accumulate sufficient energy 
reserves for lengthy incubation shifts (Boersma and 
Wheelwright 1979, Gaston and Powell 1989, Murray and 
others 1983). Murray and others (1993) have hypothesized 
that the benefits of increased foraging time during egg neglect 
often outweigh the disadvantages of leaving the egg 
unattended. Disadvantages of egg neglect include predation, 
heat loss, and exposure to the elements. In Oregon, an egg 
was believed to have been taken by a corvid when adults left 
their nest unattended (Nelson and Hamer, this volume b). 

Tuning of Incubation Exchanges 

Adults usually exchange incubation duties at dawn (n = 
12 nests), although Simons (1980) believed exchanges may 
have taken place at dusk at a ground nest in Alaska. Incubation 
exchanges generally occur before official sunrise, and often 
correspond with the first auditory detections of murrelets 
each morning (Naslund 1993a; Nelson and Peck, in press; 



S.W. Singer, pers. comm.) (table 2). The timing of exchanges 
were significantly affected by weather patterns and light 
levels; birds arrived later during overcast or rainy conditions 
(Naslund 1993a, pers. comm.; Nelson and Peck, in press). 
In addition, birds arrived earlier in areas of higher latitude 
likely because of longer periods of twilight In Prince William 
Sound, Alaska, incubation exchanges occurred from 37-82 
minutes prior to official sunrise ( x = -52, s.e. = 3.1, n = 14 
observations at 5 nests) (Naslund, pers. comm.). In Oregon 
and California, the timing of incubation exchanges ranged 
from 31 minutes before to 1 minute after official sunrise ( x 
= -18.5, s.e. = 0.7, n = 85 observations at 7 nests) (Naslund 
1993a; Nelson and Peck, in press; Singer and others 1991; 
S.W. Singer, pers. comm.) (table 2). No nocturnal incubation 
exchanges were observed during intensive observations in 
California (Naslund 1993a, Singer and others 1991); nocturnal 
surveys have not been conducted elsewhere. 

Incubating birds usually left immediately after the arrival 
of their mate. Most incubation exchanges lasted 3 to 60 
seconds ( x = 26.0 seconds, s.e. = 4.5, n = 76 observations at 
7 nests), although at one nest in California one exchange 
lasted 3 minutes and 40 seconds (Naslund 1993a; Nelson and 
Peck, in press; S.W. Singer, pers. comm.) (table 3). The 
arriving adult often remained motionless on the nest limb 
before occupying the nest and commencement of incubation; 
this waiting period lasted 14 to 357 seconds in California and 
Oregon (Naslund 1993a; Nelson and Peck, in press). 

Egg-Hatching, Brooding Behavior, and 
Chick Development 

The single murrelet chick hatches after 27 to 30 days of 
incubation (Carter 1984; Hirsch and others 1981; Sealy 1974, 
1975a; Simons 1980). Adults become active before the egg 
hatches, standing and turning more frequently than earlier in 
the incubation period (Naslund 1993a; Nelson and Peck, in 



Table 2 — Mean time of incubation exchanges in relation to official sunrise at Marbled Murrelet nests by state 1 



State 2 



Number Number 

nests observation days 



Mean time 

before sunrise 

(min) 



Standard 
error 



Range 



Alaska 




5 


14 


-523 


3.1 


-82,-37 


Oregon 




4 


49 


-18.5 


0.8 


-30. -8 


California 




3 


36 


-18.4 


1J 


-31. +1 


Total 




12 


99 


-23.2 


1.4 


-82, +1 


Oregon and California 


only 


7 


85 


-18.5 


0.7 


-31, +1 



! Data from Naslund, pers. comm.; Nelson and Peck, in press; S.W. Singer, pers. comm. 
2 Incubation exchanges were not observed in British Columbia and Washington. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



59 



Nelson and Hamer 



Chapter 5 
Table 3 — Mean length of incubation exchanges at Marbled Murrelet nests by state' 



Nesting Biology and Behavior 



State 2 ' 3 


Number 


Number 


Mean 


Standard 


Range 




nests 


observation days 


length 
(sec) 


error 


(min:sec) 


Oregon 


4 


42 


16.3 


2.8 


0:03-1:07 


California 


3 


34 


38.1 


1.1 


0:02-3:40 


Overall 


7 


76 


26.0 


4.5 


0:02-3:40 



1 Data from Nelson and Peck, in press; S.W. Singer, pers. comm. 

2 Incubation exchanges were not observed in British Columbia and Washington. 

3 Data from Alaska were not available. 



press; Simons 1980). Chicks are semi-precocial at hatching, 
and weigh approximately 32.0-34.5 g (n = 2 chicks) (Simons 
1980, Hirsch and others 1981). They are covered with a 
dense yellowish down, sprinkled evenly with irregular dark 
spots (brown and black), except on the head where spots are 
concentrated in large patches, and on their bellies which are 
covered with a dense, pale grey down (Binford and others 
1975, Simons 1980). 

Adults usually brood the chick for 1 to 2 days after 
hatching (n = 4 nests) (Nelson and Peck, in press; Simons 
1980; S.W. Singer, pers. comm.), possibly until the chick 
reaches homeothermy. However, Naslund (1993a) recorded 
intermittent brooding by adults after daytime and evening 
feedings at least 3 days after hatching. Naslund (1993a) 
suggested that the increased brooding may have occurred to 
protect the chick from predators in the vicinity of the nest. In 
addition, in British Columbia, Eisenhawer and Reimchen 
(1990) presented circumstantial evidence that adults returned 
at night to brood young chicks. 

During brooding, adults are active and restless, regularly 
standing, turning, and repositioning themselves on the chick. 
Adults do not remove the eggshell from the nest cup, therefore 
pieces that do not fall out accidentally remain in the nest cup 
and often are crushed into the nest material by adult and 
chick activity. 

During the first 6 days after hatching, droppings from 
the chick begin to accumulate around the perimeter of the 
nest cup (adults are not known to defecate at the nest). By 
the time the chick fledges, the fecal ring can be up to 5 1 mm 
thick. Odor (ammonia and fish) from fecal material can be 
detected by humans from up to 2 m away. 

Murrelet chicks grow rapidly compared to most alcids, 
gaining 5- 15 g per day during the first 9 days after hatching 
(n = 2 chicks) (Hirsch and others 1981; Simons 1980). As 
chicks age, the juvenal plumage begins to develop beneath 
the down; both feather types grow from the same sheath. 
By day 17, the wing coverts have emerged and down is 
missing from the forehead and around the mandibles (« = 5 
chicks). By day 21, chicks lose most of their belly down, 
and by day 26, up to 20 percent of body down disappears. 



Twelve to 48 hours prior to fledging, the murrelet chick, by 
preening, scratching, and wing flapping, removes the 
remaining down, revealing their black and white juvenal 
plumage (n = 10 chicks) (Hamer and Cummins 1991; 
Hirsch and others 1981; P. Jones, pers. comm.; Nelson and 
Peck, in press; Simons 1980; Singer and others 1992, in 
press). This pattern of down loss and feather development 
is unique among alcids, except the closely related Kittlitz's 
Murrelet (Brachyramphus brevirostris). 

Wing length increases rapidly in the last 4 days prior 
to fledging, and at fledging the chicks wings are 103-144 
mm long (86 percent of adult wing length) (Hamer and 
Cummins 1 99 1 ; Hirsch and others 1 98 1 ; Sealy 1 975a; Simons 
1980). Chicks fledge at age 27-40 days (Hirsch and others 
1981; Nelson and Peck, in press; Simons 1980). At this time 
they still possess an egg tooth, and weigh an average of 
146.8-157.0 g (s.e. = 3.6-9.5, n = 4-9), which is 63-70 
percent of adult (222 g) weight (Hamer and Cummins 1991; 
Hirsch and others 1981; Sealy 1975a; Simons 1980). Fledging 
takes place at dusk, between 11 and 55+ minutes after 
official sunset (Hamer and Cummins 1991 ; Hirsch and others 
1981; P. Jones, pers. comm.; Nelson and Peck, in press; 
Singer and others, in press) {table 4). 

Chicks are thought to fly directly from the nest to the 
ocean (Hamer and Cummins 1991; Quinlan and Hughes 
1990; Sealy 1975a). Hamer and Cummins (1991) radio- 
tagged a juvenile Marbled Murrelet on a nest in Washington, 
37 km inland, and monitored its flight to the ocean. The 
chick fledged in the evening and was found 1 8 hours later, 
100 m from shore and 2 km north of a direct east- west line 
between the nest and Puget Sound. The juvenile flew directly 
to the ocean and did not spend any time in the vicinity of the 
nest. However, several fledglings have been observed 
swimming in creeks in California and Washington (Hamer 
and Cummins 1991; Miller, pers. comm.). It is not known if 
these fledglings fell from nests, became grounded on their 
maiden flight to the ocean, or were actually trying to reach 
the ocean by swimming the creek. Numerous fledging birds 
in North America appear to have become grounded during 
flights to the Pacific (Nelson and Hamer, this volume b). 



60 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter5 



Nesting Biology and Behavior 



Table 4 — Dates and timing of observed Marbled Murreletfledgingsfrvm nests by state and 
province 1 



State/province 


Number 


Date 


Fledging 


Minutes after 




nests 




time 


sunset 


Alaska 


1 


8/16/79 


>2200 


+21 


British Columbia 


1 


8/20/93 


2051 


+30 


Washington 


3 


6/22/93 


2124 


+14 






8/07/91 


2046 


+11 






8/27/90 


2020 


+20 


Oregon 


1 


8/29/91 


>2050 2 


+55 


California 


2 


6/07/92 


2046 


+18 






7/03/91 


2054 


+19 



1 Data from Hamer and Cummins 1991; Hirsch and others 1981 ; P. Jones, pers. comm.; 
Nelson and Peck, in press: Singer and others, in press. 

2 Chick fledged between 2050 and 0700 hours. 



Many of these grounded fledglings may be unable to take 
flight again or make it to the ocean by other means. Once 
juveniles reach the ocean they are thought to be independent 
and not attended either parent contrary to the suggestion of 
Ydenberg(1989). 

Chick Behavior 

Chicks remain motionless or sleep 80-94 percent of the 
time on the nest (n = 8 chicks; (Hamer and Cummins 1991; 
Naslund 1993a; Nelson and Peck, in press). Other behaviors 
include standing, turning, shifting position, preening, 
stretching, flapping, pecking at the nest substrate or the tree 
limb, food begging in the presence of adults, and snapping at 
insects. Behaviors such as wing flapping and preening increase 
markedly in the week prior to fledging. 

On the two evenings prior to fledging, chicks are very 
active (Hamer and Cummins 1991: Ritchie, pers. comm.; 
Singer and others, in press). Behaviors during this time 
include continual rapid pacing on the nest platform, frequent 
vigorous flapping of the wings, repeated peering over the 
edge of the nest platform, rapid nervous head movements, 
and constant preening. After a vigorous session of wing 
flapping, young birds sometimes hold their wings outstretched 
and vibrate them rapidly, giving the appearance of shivering 
wings. These behaviors begin in late afternoon or minutes 
before sunset, and continue until dark or until the bird fledges. 

Low light levels may induce fledging. After a captive 
reared chick fledged from an artificial nest platform in the 
dark, it was placed back on the platform and the room 
brightened by artificial light (Hamer and Cummins 1991). 
The chick immediately sat motionless and ceased all activity. 



When the room was darkened again by turning off the light, 
the chick immediately began the pre-fledging behaviors 
described above and fledged a second time. 

Feeding Frequency, Behavior, and Prey Species 

Adults return to feed young up to eight times daily ( x = 
3.2, s.e. = 0.4, n = 10 nests) (Hamer and Cummins 1991; 
Hirsch and others 1981; P. Jones, pers. comm.; Kerns, pers. 
comm.; Nelson and Peck, in press; Simons 1980; S.W. Singer, 
pers. comm.) (table 5). Chicks are usually fed at least once a 
day for the 27-40 days they are on the nest, although the 
frequency is variable and sometimes decreases prior to fledging. 
The last feeding prior to fledging occurs between 5 minutes 
(Singer and others, in press) and 2.5 days (Hamer and Cummins 
1991 ) before the young murrelet leaves the nest. 

The timing of dawn feedings is more variable than 
incubation exchanges. First dawn feedings occur from 37 
minutes before to 65 minutes after official sunrise ( x = 6.0, 
s.e. = 3.7, n = 68 feedings at 1 3 nests) (Hamer and Cummins 
1991; Kerns, pers. comm.; Naslund 1993a; Nelson and Peck, 
in press; S.W. Singer, pers. comm.) (fig. 1, table 6). Similar 
to incubation exchanges, weather and light conditions 
influence the arrival times of the adults, and feedings often 
occur later on rainy or cloudy days (Naslund 1993a, Nelson 
and Peck, in press). Second morning feedings occur from 18 
minutes before, to 225 minutes (1009 hrs) (x = 53.7, s.e. = 
9.6, n = 40 observations at 13 nests) after, official sunrise. 
Other feedings take place during the day between the hours 
of 1100 and 1700 (Hamer and Cummins 1991; P. Jones, 
pers. comm.; Kems, pers. comm.; Naslund 1993a; Nelson 
and Peck, in press; Singer and others I99l)(fig. 1). Dusk 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



61 



Nelson and Hamer 



Chapter 5 



Nesting Biology and Behavior 



Table 5 — Mean number of feeding visits observed per day' at Marbled Murrelet nests by state and province 2 



State/province 3 


Number 


Number 


Number 


Mean 


Standard 


Range 




nests 


observation days 


feedings 


per day 


error 




British Columbia 


1 


11 


44 


4.0 


0.8 


1-8 


Washington 


1 


8 


23 


2.9 


0.4 


2-5 


Oregon 4 


5 


22 


61 


2.8 


0.3 


1-5 


California 4 


3 


21 


67 


3.4 


0.3 


1-6 


Overall 


10 


62 


195 


3.2 


0.4 


1-8 



1 Not all nests were monitored during mid-day or at night, thus some feeding visits may have been missed. Data 
include only days where nests were monitored at dawn and dusk on all observation days. 

2 Data from Hamer and Cummins 1991 ; P. Jones, pers. comm.; Kerns, pers. comm.; Nelson and Peck, in press; S.W. 
Singer, pers. comm. 

3 No tree nests with chicks were observed in Alaska; data from 2 ground nests in Alaska were not available. 

4 Two nests in Oregon and 1 in California were not monitored at dawn and dusk on the same day. 



CD 

z 

Q 

LU 
LU 



o 
cc 

LU 

CO 

Z) 

z 




0500 0700 0900 



1100 1300 
TIME 



1600 1800 2000 



Figure 1 — Number of feedings by time of day (0500-21 00 hrs) at ten Marbled 
Murrelet nests in British Columbia, Washington, Oregon, and California (n = 
206 feedings). 



feedings occur from 90 minutes before, to 7 1 minutes after, 
official sunset, with the last feeding visit occurring 40 minutes 
before, to 71 minutes after, official sunset (x = 18.4, s.e. = 
4.1, n = 41 feedings at 12 nests) (Hamer and Cummins 
1991; Naslund 1993a; Nelson and Peck, in press; Singer 
and others, in press) (table 7). No nocturnal (after dusk) 
feeding visits were recorded during all-night observations 
in Washington and California (« = 38 nights at 3 nests) 
(Hamer and Cummins 1991; Naslund 1993a). 

On several occasions (n = 7 of 68 visits at three nests), 
two adults arrived at the nest with fish at the same time 
(Kerns, pers. comm; P. Jones, pers. comm.; Nelson and Peck, 
in press). In Oregon, when this occurred, one adult flew 



away, and returned only after the other adult had left. In 
California and British Columbia, both adults left and returned 
individually at a later time, or both remained until the chick 
had eaten one of the fish. 

Adults usually carry single fish in their bills, holding it 
crosswise at the mid-point of the fish's body, or just posterior 
to the operculum. On several occasions, adults were observed 
arriving with 2 fish at nests in California and Oregon (n = 
3)(Buchholz, pers. comm.; Kerns, pers. comm). When adults 
arrive at the nest with a fish, they often remain in a motionless 
posture on the landing pad for up to 11 minutes before 
approaching the nest (n = 11 nests) (Hamer and Cummins 
1991; Kerns, pers. comm.; Naslund 1993a, pers. comm.; Nelson 



62 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter 5 



Nesting Biology and Behavior 



Table 6 — Mean time of first morning feeding visits in relation to official sunrise at Marbled 
Murrelet nests by state 1 



State" 


Number 


Number 


Mean 


Standard 


Range 




nests 


feedings 


time(min) 


error 




Washington 


2 


10 


-9.3 


11.1 


-37.+50 


Oregon 4 


7 


32 


+ 7.9 


5.1 


-36,+65 


California 


4 


26 


+ 9.5 


5.9 


-31, +62 


Overall 


13 


68 


+ 6.0 


3.7 


-37,+65 



1 Data from Hamer and Cummins 1991 ; Kerns, pers. comm.; Nelson and Peck, in press; S.W. 
Singer, pers. comm. 

2 No tree nests with chicks were observed in Alaska; data from 2 ground nests in Alaska not 
available. 

3 Data from British Columbia were not available. 

4 Does not include one late observation (104 min), assumed to be a second feeding. 



Table 7 — Mean time of last evening feeding visits in relation to official sunset at Marbled 
Murrelet nests by state 1 



State 2 - 3 


Number 


Number 


Mean 


Standard 


Range 




nests 


feedings 


time (min) 


error 




Washington 


2 


4 


+ 9.3 


22.3 


-39,+69 


Oregon 


7 


17 


+15.3 


5.7 


^M),+62 


California 


3 


20 


+23.0 


5.5 


-15.+71 


Overall 


12 


41 


+18.4 


4.1 


-40.+71 



1 Data from Hamer and Cummins 1991 ; Kerns, pers. comm.; Nelson and Peck, in press; S.W. 
Singer, pers. comm. 

2 No tree nests with chicks were observed in Alaska; data from 2 ground nests in Alaska not 
available. 

3 Data from British Columbia were not available. 



and Peck, in press; S.W. Singer, pers. comm.). At a nest in 
Washington, adults rested on the nest platform an average of 
2.2 minutes before approaching the chick with food. 

In Oregon and Washington, the chick sometimes gave 
begging calls just prior to the adults landing on the nest 
platform (x = 1.2 minutes before adults arrived, n = 8 
observations at t nest) and throughout the feeding visit (see 
vocalizations section) (Hamer and Cummins 1991; Nelson 
and Peck, in press). At a nest in Washington, the chick spent 
an average of 10.8 minutes begging during each feeding visit. 

After approaching the chick, the adult stands motionless 
as the chick energetically strokes or nudges the throat and 
beak of the adult with its beak (Hamer and Cummins 1991; 



Naslund 1993a; Nelson and Peck, in press). Adults at a nest 
in Washington held the fish over the chick for an average of 
9.7 minutes {s.e. = 1.4, n = 16 observations) before the food 
transfer took place. The adults occasionally give soft whistle 
or grunt-like vocalizations until the nestling takes the fish 
(Hamer and Cummins 1991; Nelson and Peck, in press). The 
time adults spent at nests during feedings ranged from 13 
seconds to 80 minutes (3c = 12.6 min, s.e. = 0.7, n = 16) 
(Hamer and Cummins 1991; Hirsch and others 1981; P. 
Jones, pers. comm.; Kerns, pers. comm.; Naslund 1993a; 
Nelson and Peck, in press; Simons 1980) (table 8). Fifty 
percent of feedings lasted > 1 1 (median) minutes. Chicks 
held the fish 5 seconds to 2 minutes before swallowing it 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



63 



Nelson and Hamer Chapter 5 

Table 8 — Mean length of feeding visits at Marbled Murrelet nests by state and province 1 



Nesting Biology and Behavior 



State/province 2 


Number 


Number 


Mean 


Standard 


Range 




nests 


feedings 


time (min) 


error 




Alaska 


2 


5 


5.0 


1.7 


1.43-10.0 


British Columbia 


1 


38 


13.2 


2.0 


1.00-80.0 


Washington 


2 


24 


10.4 


1.2 


0.30 3 -20.0 


Oregon 


7 


61 


16.7 


1.3 


0.18 3 -46.0 


California 


4 


82 


10.4 


0.8 


0.13 3 -48.0 


Overall 


16 


210 


12.6 


0.7 


0.13 3 -80.0 



1 Data from Hamer and Cummins 1 99 1 ; Hirsch and others 1 98 1 ; P. Jones, pers. comm. ; Kerns, pers. 
coram.; Nelson and Peck, in press; Ritchie, pers. comm.; Simons 1980; S.W. Singer, pers. comm. 

2 No tree nests with chicks were observed in Alaska. 

3 Adults may not have fed chicks fish on some of the shorter visits. 



head first and whole ( jc =1.4 minutes at a nest in Washington, 
n = 4 observations) (Hamer and Cummins 1991; Nelson and 
Peck, in press). Adults usually leave within 1 minute of the 
fish exchange. 

To provide chicks with fish at dawn, adults probably 
forage at night, perhaps taking advantage of fish that forage 
near the water surface during darkness (Carter and Sealy 
1987a, 1990). Fish species that have been fed to chicks at 
nests include Pacific sand lance (Ammodytes hexapterus), 
Pacific herring (Clupea harengus), and northern anchovy 
(Engraulis mordax) (P. Jones, pers. comm.; Nelson and 
Peck, in press). Other potential prey species that were not 
positively identified included capelin (Mallotus spp.), smelt 
(Osmeridae; probably whitebait [Allosmerus elongatus] or 
surf smelt [Hypomesus pretiosus]), and herring species 
(Clupeidae or Dussumieriidae) (Naslund 1993a; Nelson, 
unpubl. data; Simons 1980; but see Burkett, this volume). 

Flight Behavior 

Adult murrelets often use similar flight paths on 
approaches and departures from tree nests. Generally, they 
follow openings such as creeks, roads or other clearings that 
allow for direct approaches and departures from the nest 
(Kerns, pers. comm.; Nelson and Peck, in press; Singer and 
others 1991; Singer and others, in press). The directions that 
birds enter and leave nests appear to be related to openings 
in the canopy or forest around the nest tree, and gaps in the 
horizontal cover surrounding the nest limb (Naslund, pers. 
comm.; Nelson and Peck, in press; Singer and others 1991; 
Singer and others, in press). Birds approach nests below tree 
canopy at heights as low as 5 m, and usually ascend steeply 
to the nest in a "stall-out" fashion. Landings are sometimes 
hard and audible (Nelson and Peck, in press; S.W. Singer, 



pers. comm.). We have observed and heard murrelets crashing 
into tree limbs on some occasions during final approaches to 
nests (Nelson and Peck, in press). In addition, birds 
occasionally abandoned landings and circled around for second 
attempts. When leaving the nest, birds usually drop 5-30 m 
in height before ascending over the canopy to continue their 
departure flights. They have not been observed departing at 
nest height or flying upwards on take-off from the nest limb. 

When landing on the nest branch, murrelets splay out 
their webbed feet, lean backwards, and use their wings to 
slow their forward motion. They land hard enough on the 
nest limb to create a landing pad, or area where the moss or 
duff becomes flattened, removed, and worn by repeated 
landings. Toe nail markings are evident at some landing 
pads. Landing pads are most often located on the nest limb 
within 1 m of the nest cup, however they have also been 
located on adjacent limbs. In the latter case, murrelets hop 
to the nest limb. 

Subcanopy behaviors, including one or more birds flying 
through, into, or out of the tree canopy, and birds landing in 
trees, are flight behaviors indicative of nesting and have been 
noted in nest stands and around nest trees. Landings and 
departures from trees have been observed at nests, on other 
branches in nest trees, in trees adjacent to nest trees, and 
other trees in nest stands throughout the breeding season. 
These landings may indicate nesting, territorial behavior, 
searches for nest sites, or resting or roosting behavior (Naslund 
1993a). Singer and others (1991), and Naslund (1993a) 
described an additional four flight patterns observed near 
nest trees: (1) fly-bys and stall-flights, including single birds 
or pairs flying by or stalling out next to a known nest tree, at 
nest branch height; (2) flying-in-tandem and tail-chases, where 
pairs of birds fly in close proximity to known nest trees; and 



64 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter 5 



Nesting Biology and Behavior 



(3) buzzing, which includes single birds flying through the 
canopy making continuous low-pitched buzzing wing sounds. 
Several flight behaviors above the canopy are also 
indicative of nesting. Like other alcids, Marbled Murrelets 
are often observed circling singly or in groups above the 
nesting grounds (Gaston 1992). Nesting birds may join with 
others before returning to the ocean after incubation and 
feeding visits. Nonbreeders may also accompany nesting 
birds in these circling flights above the canopy. Murrelets 
also occasionally create a loud sound, like a jet, during a 
shallow or steep dive that often originates above the canopy 
and ends at or below canopy level (Nelson and Peck, in 
press; S.W. Singer, pers. comm.). In Oregon, this behavior 
has been observed most often (67 percent) associated with 
known nest trees. In California, the jet dive during encounters 
between two murrelets has been observed and may be an 
aggressive posture or territorial defense. 

Predator Avoidance Behavior 

The Marbled Murrelet's primary defense against predators 
at the nest is to avoid detection through their secretive behavior 
at or near the nest, morphological defense mechanisms, such 
as cryptic plumage, and location of nest sites in trees and 
stands with hiding cover. In direct response to calls, 
silhouettes, or the presence of predators, and other disturbances 
(e.g., airplanes) at nests, adults and chicks often flatten 
themselves against the tree branch, holding their backs and 
heads low and remaining motionless (Kerns, pers. comm.; 
Naslund 1993a; Nelson and Peck, in press; Quinlan and 
Hughes 1990; Simons 1980; Singer and others 1991). 
However, they may also attempt to defend themselves against 
predators that have located the nest. At a nest in California, a 
murrelet chick was observed to defend itself against a Steller's 
Jay (Cyanocitta stelleri) by standing erect, turning to face 
the intruder, and jabbing it with its slightly open bill (Naslund 
1993a; S.W. Singer, pers. comm.). In addition, Naslund 
(1993a) noted that occasionally when a raven flew by a nest, 
the adult assumed an erect posture as if readying itself to 
take flight. S.A. Singer (pers. comm.) observed an incubating 
adult lunge with open bill at a raven as it approached the 
nest, causing it to veer off instead of landing. 

Vocalizations and Wingbeats 

Marbled Murrelets primarily give soft or muted calls 
from the nest limb that are not audible from the ground. 
They rarely give loud vocalizations from stationary locations 
or in close proximity to a nest. When loud vocalizations are 
given at or near a nest, they can be heard from the ground 
depending on weather conditions and the location of the 
observer. However, because loud calls are uncommon, using 
them as a means for locating nests is not feasible with our 
current understanding of murrelet vocalizations. 

Below is a summary of the vocalizations heard from 
murrelet nests. Most of these vocalizations were heard with 
the aid of microphones and other recording equipment pointed 
directly at the nest branch. 



Adults and chicks were heard giving soft vocalizations 
at most nests (« = 14), but loud vocalizations were heard at 
only seven nests. These calls were given during incubation 
exchanges and feeding visits. Soft vocalizations include groan 
or grunt calls (duck-like quacks; previously referred to as 
alternate calls), whistle calls, and faint peeps. Loud 
vocalizations consist of keer and groan calls (Nelson and 
Peck, in press). 

During incubation exchanges in Alaska, Oregon, and 
California, vocalizations were primarily given at the nest as 
birds arrived or departed the nest limb or during the brief 
seconds when adults were on the nest limb together. However, 
an interesting long (13.5 sec) vocal sequence was recorded 
at one nest in Oregon. First, the incubating adult made soft 
groans from the nest branch, and at the same time a second 
adult flying nearby gave short, loud whistle calls. The 
incubating adult then emitted additional groans, which became 
increasingly louder and more emphatic. As the flying adult 
joined the other on the nest limb, one of these two birds gave 
loud whistle calls. 

The frequency of exchanges with vocalizations varied 
among nests. In Alaska, 10 of 11 incubation exchanges 
included soft groan and other (undescribed) calls; and adults 
gave 1-2 loud keer calls when arriving or departing during 
incubation exchanges on two of 12 mornings (Naslund, pers. 
comm.). In Oregon, only 10 percent of incubation exchanges 
included soft or loud vocalizations (n = 59). At a nest in 
California, adults gave loud, emphatic "keer" and groan 
calls just before leaving the nest branch on five of 17 
incubation exchanges (Naslund 1993a; Singer and others 
1991; S.W. Singer, pers. comm.). In addition, several soft 
grunt calls, sounding like "unh-unh-unh", were heard on one 
occasion after an adult landed on the nest branch. 

During feeding visits in Oregon, Washington, and British 
Columbia, adults occasionally gave loud keer calls and soft 
groan and "eeeuh" or "eeea" whistle calls as they flew from 
the nest branch or while bringing food to a chick at the nest 
(Hamer, unpubl. data; P. Jones, pers. comm.; Nelson and 
Peck, in press). The latter calls sounded like a muffled 
honking that adults gave while holding fish during feeding 
visits. In addition, in California, a series of soft "chip" notes, 
duck-like quacks, or short, soft grunts were given after the 
adult bird arrived to feed the chick (Singer and others 1991; 
S.W. Singer, pers. comm.). In British Columbia, a one-note 
bleating call (soft groan) was usually made when two adults 
were at the nest simultaneously (n = 4 occasions at 1 nest). 

Chicks emit a rapid, high pitched begging call during 
feeding sessions. This begging call was recorded from a 
captive chick, and heard or recorded from active nests in 
Oregon and Washington (n = 4). In addition, P. Jones (pers. 
comm.) described a soft peep or begging call (repeated 
"puli-puli") that may have been given by the chick during 
feedings at a nest in British Columbia. We believe begging 
calls occur during every food delivery, but this sound is not 
usually audible, especially without microphones placed at or 
near the nest. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



65 



Nelson and Hamer 



Chapter 5 



Nesting Biology and Behavior 



Calls and fly-bys that indicate the impending arrival of 
an adult may also be given at nest sites. On several occasions 
in Oregon and Washington, incubating adults or chicks became 
alert, and chicks gave begging calls, moments before the 
arrival of the (other) adult (Hamer and Cummins 1991; 
Nelson and Peck, in press). Naslund (1993a) and Eisenhawer 
and Reimchen (1990) also mentioned keer and groan 
vocalizations given by adults prior to incubation exchanges. 

Wingbeats have been heard during landings and take- 
offs from nest branches at all nests, and while murrelets were 
flying through the tree canopy. Murrelets appear to be able to 
purposely create the wing sounds, because they are not heard 
during all landings, take-offs, and flights through the canopy. 

Discussion 

Marbled Murrelet breeding biology, morphology, and 
behavior, like that of other alcids, is affected by distance of 
nest sites from food sources, risk of predation, and other 
physical and biological factors (Cody 1973; De Santo and 
Nelson, this volume; Vermeer and others 1987a; Ydenberg 
1989). The risk of predation may be the most significant 
factor in the development of alcid behavior, especially for 
Marbled Murrelets in their forest nesting environment. 

Exposure to predation has influenced the length of 
incubation shifts, chick feeding frequency, and fledging 
strategy of alcids (Ydenberg 1989). Predator avoidance may 
be the driving force behind the long incubation shifts of both 
the Marbled (24 hours) and the Ancient (48 to 120 hours) 
Murrelets (Synthliboramphus antiquus)(Gaston 1992), as 
frequent visits to a nest can increase the chances of being 
discovered by predators and endanger the parents and young. 
Because of this risk, some species of alcids only feed their 
chick once in a 24-hour period (nocturnal alcids with semi- 
precocial young), whereas others (Synthliboramphus spp.) 
produce precocial chicks that are not fed at the nest site. In 
addition, feeding frequency within a species can vary among 
nesting colonies, with young in safe sites receiving more 
food than those in unsafe sites (Ydenberg 1989). Young that 
receive multiple daily feedings grow faster and fledge earlier 
than those with lower provisioning rates (Gaston and 
Nettleship 1981). Early fledging helps to minimize nest 
mortality (Cody 1971). 

Marbled Murrelets have optimized their survival 
strategies by laying a single egg, feeding their chick relatively 
frequently, and concentrating most of their activity in the 
low light levels of dawn and dusk. With multiple daily 
feedings, murrelet chicks grow relatively rapidly and 
generally fledge earlier compared with most semi-precocial 
alcids (De Santo and Nelson, this volume). Despite this 
earlier fledging, Marbled Murrelet chicks are vulnerable in 
their open nest sites for 27 to 40 days. Therefore, selection 
of safe nest sites (Hamer and Nelson, this volume b) and 
secretive behaviors to avoid predation are also necessary for 
their survival. 



In response to pressures from predation at nesting sites, 
alcids have developed specific behavioral characteristics (flight 
behavior, nocturnal activity) and have selected nest sites in 
inaccessible areas (burrows and crevices). Whereas most 
alcids are diurnal, nine species, including Marbled Murrelets, 
are primarily nocturnal or crepuscular (Synthliboramphus 
murrelets, Cassin's [Ptychoramphus aleuticus] and Rhinoceros 
[Cerorhinca monocerata] Auklets, Kittlitz's Murrelet, Dovekie 
[Alle alle]). Activity during low light levels (or twilight 
hours in the high arctic) minimizes predation by diurnal 
avian predators like gulls and corvids ( Ainley and Boekelheide 
1990; Gaston 1992; Nettleship and Birkhead 1985). Most 
alcids nest in inaccessible areas (burrows, crevices) to hide 
from predators, however, some species nest in the open on 
rock ledges (Common Murre, Thick-billed Murre [Uria 
lomvia], and Razorbill [Alca torda]), and must protect their 
young by nesting in large colonies or by guarding them 
during the day (Nettleship and Birkhead 1985). 

The Brachyramphus murrelets also nest in the open, 
but they generally nest solitarily. For protection from 
predation, these murrelets have developed a cryptic plumage 
and secretive behaviors that allow them to remain hidden. 
For example, Marbled Murrelets have developed a variety 
of morphological and behavioral characteristics as defense 
mechanisms, some of which are shared by Kittlitz's Murrelet 
and other alcids: ( 1 ) concentrating activities in forests during 
crepuscular periods when light levels are low (i.e., incubation 
exchanges and feeding visits at dawn and dusk); (2) cryptic 
coloration of the egg, chick, and adult (breeding plumage); 
(3) rapid flight into and away from the nest; (4) visiting the 
nest briefly during incubation and less so during feeding of 
young; (5) "freezing" behavior exhibited by adults after 
landing at the nest during incubation exchanges and feeding 
visits; (6) remaining relatively silent on the nest branch 
(vocalizations are muted); (7) low, motionless posture of 
the incubating adult; (8) well developed thermoregulatory 
capabilities of the chick shortly after hatching allowing for 
minimal parental care; (9) chick remaining motionless for 
long time periods; (10) retention of down feathers by chick 
concealing bright juvenal plumage until just prior to fledging; 
(11) young fledging just after dusk; (12) long distance 
indirect flights through the forest canopy to access nests; 
(13) fly by inspections of nests and nesting area by adults 
before a nest visit; (14) flying in groups within and above 
the nesting grounds, which may provide protection from 
predators and serve as an important social function; and 
(15) selecting nest platforms with high levels of vertical or 
hiding cover (see Binford and others 1975; Hamer and 
Cummins 1 99 1 ; Hamer and Nelson, this volume b; Naslund 
1993a; Nelson and Peck, in press; Sealy 1974, 1975a; 
Singer and others 1991). The number and diversity of these 
adaptations suggests that predation has been a selective 
factor on Marbled Murrelets in the past. Given these predator 
avoidance strategies, one would expect predation at nests 
to be low. However, Marbled Murrelets are still vulnerable 



66 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelsoo and Haroer 



Chapter5 



Nesting Biology and Behavior 



to seemingly high rates of predation (see Nelson and Hamer, 
this volume b). Predation rates on alcid nests are often 
higher in areas where predators have been introduced, habitat 
has been modified, or where birds are disturbed by humans 
(Gaston 1992; Murray and others 1983; Nettleship and 
Birkhead 1985). 

Observations of Marbled Murrelet behavior at nest sites 
have provided us with a wealth of new information that was 
not available prior to 1980. Their secretive behavior, rapid 
flights in low light levels, and the inaccessibility of many of 
their nests, however, has limited our opportunities to study 
many aspects of their biology. The paucity of information on 
some aspects of Marbled Murrelet breeding biology minimizes 
the accuracy with which land managers can maintain or 
create suitable habitat for this species. In addition, their 
secretive behaviors limit our ability to identify nesting sites, 
especially in stands that contain few birds. Continued research 
on the biology, demography, and habitat selection of this 



species should be conducted, in addition to determining the 
effects of different forest management strategies on nesting 
success of this unique seabird. 

Acknowledgments 

We are grateful to the biologists who kindly shared their 
data with us; special thanks go to Dave Buchholz. Dave 
Fortna, Paul Jones, Steve Kerns, Kathy Kuletz, Nancy Naslund. 
Bill Ritchie, and Steve and Stephanie Singer for their time 
and generosity. We also thank Dan Anderson, Toni De Santo, 
George Hunt, Robert Peck, and C. John Ralph for reviewing 
earlier drafts of this manuscript. Support for preparation of 
this manuscript was provided by the Oregon Department of 
Fish and Wildlife, USDA Forest Service, USDI Bureau of 
Land Management and U.S. Fish and Wildlife Service. This 
is Oregon State University Agricultural Experiment Station 
Technical Paper 10,536. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



67 



Chapter 6 

Characteristics of Marbled Murrelet Nest Trees 
and Nesting Stands 



Thomas E. Hamer 1 



S. Kim Nelson 2 



Abstract: We summarize the characteristics of 61 tree nests and 
nesting stands of the Marbled Murrelet (Brachyramphus 
marmoratus) located from 1974 to 1993 in Alaska, British Colum- 
bia, Washington, Oregon, and California. Evidence of breeding 
30-60 km inland was common in California, Oregon, and Wash- 
ington. Nesting greater distances from the coast may have evolved 
to avoid nest predation by corvids and gulls which are more 
abundant in coastal areas. In California, Oregon, Washington, and 
British Columbia, murrelets nested in low elevation old-growth 
and mature coniferous forests, with multi-layered canopies (>2), a 
high composition of low elevation conifer trees ( x = 91 percent) 
and. on the lower two-thirds of forested slopes, with moderate 
gradients ( x = 23 percent slope). Stand canopy closure was often 
low ( x = 50 percent), suggesting use of canopy openings for 
access to nest platforms. Nests in the Pacific Northwest were 
typically in the largest diameter old-growth trees available in a 
stand ( x =211 cm); many nest trees were in declining conditions 
and had multiple defects. It is likely that western hemlock and Sitka 
spruce constitute the most important nest trees, with Douglas-fir 
important south of British Columbia. Many processes contributed 
to creating the nest platforms observed. Mistletoe blooms, unusual 
limb deformations, decadence, and tree damage, commonly ob- 
served in old-growth and mature stands, all appear to create nest 
platforms. Therefore, the stand structure and the processes within a 
stand may be more important than tree size alone in producing 
nesting platforms and suitable habitat. Moss cover was also an 
important indicator of suitable nesting habitat. 



We summarize the characteristics of 6 1 tree nests and 
nesting stands of the Marbled Murrelet (Brachyramphus 
marmoratus) located from 1974 to 1993 in Alaska, British 
Columbia, Washington, Oregon, and California (table 1). 
The majority of the nest site information was unpublished 
and obtained directly from field biologists who were 
conducting inland studies on the murrelet. The preponderance 
of unpublished nest information is due to the recent discovery 
of most nest sites. The only other summary was completed 
by Day and others (1983), based on two tree nests and five 
ground nests of the Marbled Murrelet. 

Because of the murrelet' s small body size, dense forested 
nesting habitat, cryptic plumage, crepuscular activity, fast 
flight speed, and secretive behavior near nests, its nests 
have been extremely difficult to locate. The first tree nest 



1 Research Biologist. Hamer Environmental, 2001 Highway 9, Ml 
Vernon, WA 98273 

2 Research Wildlife Biologist. Oregon Cooperative Wildlife Research 
Unit, Oregon State University, Nash 104, Corvallis, OR 97331-3803 



was located only in 1974 (Binford and others 1975), despite 
decades of searching by ornithologists in North America. 
Although a significant amount of nesting habitat information 
has been collected over the past four years, the efficiency 
of locating active nests is still low. Experiences gained 
from nest search efforts have led to the development and 
refinement of methodologies for locating new nests (Naslund 
and Hamer 1994). 

Fortunately, an increased understanding of murrelet 
nesting ecology has allowed biologists to locate nests that 
have not been used for several months or, in some cases, 
several years. This involves searching for old nest cup 
depressions, worn spots or "landing pads" created on moss- 
covered branches by visiting adults, old fecal rings, and 
habitat features commonly associated with suitable nesting 
platforms. In addition, biologists learned that eggshells could 
be located in the duff and litter of nest platforms unused for 
a year or more. 

Intensive search efforts by biologists across the Pacific 
Northwest have led to the discovery of 65 tree nests since 
1974, with 63 (95 percent) located since 1990. Although this 
is still a relatively small sample size considering the large 
geographic area these nests represent, the sample does allow 
a characterization of the tree nests and nesting stands. 

The two species of murrelets in the genus Brachyramphus 
(Kittlitz's and Marbled) display a complete dichotomy in 
their choice of nesting habitat. The Kittlitz's (B. brevirostris) 
murrelet nests up to 30 km inland on the ground on exposed 
rocky scree slopes, often at higher elevations. The Marbled 
Murrelet is unique among Alcids in selecting almost 
exclusively to nest on large limbs of dominant trees, which 
can be located long distances from the marine environment. 

Long considered a subspecies of the Marbled Murrelet, 
the Asian race of the Marbled Murrelet (B.m. perdix Pallas) 
is distributed from the Kamchatka Peninsula south to Japan. 
New genetic evidence (Friesen and others 1994a) indicates 
the it is most likely a distinct species from the Marbled 
Murrelet. From the little evidence collected to date, it may 
be an obligate tree nesting seabird (Konyukhov and 
Kitaysky, this volume), with its range coinciding closely 
with the coastal coniferous forests of Russia and Japan 
(Kuzyakin 1963). 

At a few sites in Alaska and Russia, at or beyond the 
margin of Pacific Coastal coniferous forests, the Marbled 
Murrelet nests on the ground. From an examination of the 
summer distribution of the species, approximately 3 percent 
of the Alaskan murrelet population may nest on the ground 
(Piatt and Ford 1993). These nests have been found at 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



69 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



Table 1 — Records of nest trees and nest stands of the Marbled Murrelet found in North America from 1974 to 1993 



State/province 


Location 


Date found 


Sources 


Record no. 








California 








1 


Big Basin Redwood State Park 


7 Aug. 1974 


Binford and others 1975 


2 


Big Basin Redwood State Park 


3 Jun. 1989 


S.W. Singer (pers. comm.) 


3 


Big Basin Redwood State Park 


28Jun. 1989 


S.W. Singer (pers. comm.) 


4 


Big Basin Redwood State Park 


5 May 1991 


S.W. Singer (pers. comm.) 


5 


Big Basin Redwood State Park 


24 May 1992 


S.W. Singer (pers. comm.) 


6 


Jedediah Smith State Park 


9 Aug. 1993 


Hamer (pers. obs.) 


7 


Prairie Creek State Park 


23 Jul. 1993 


Hamer (pers. obs.) 


8 


Bell-Lawrence 


14 Oct. 1993 


Chinnici (pers. comm.) 


9 


Elk Head Springs 


16 Sep. 1992 


Chinnici (pers. comm.) 


10 


Shaw Creek 


30 Sep. 1992 


Chinnici (pers. comm.) 


Oregon 








11 


Boulder and Warnicke Creeks 


17 Jun. 1992 


Nelson (pers. obs.) 


12 


Cape Creek 


23 May 1991 


Nelson (pers. obs.) 


13 


Iron Mountain 


30 May 1992 


Nelson (pers. obs.) 


14 


Five Mile Flume Creek 


28 Sep. 1993 


Nelson (pers. obs.) 


15 


Five Rivers 


19 May 1990 


Nelson (pers. obs.) 


16 


Five Rivers 


14 Jun. 1991 


Nelson (pers. obs.) 


17 


Five Rivers 


23 Sep. 1993 


Nelson (pers. obs.) 


18 


Green Mountain 


17 Jun. 1993 


Nelson (pers. obs.) 


19 


Green Mountain 


22 Sep. 1993 


Nelson (pers. obs.) 


20 


Siuslaw River 


13 Aug. 1991 


Nelson (pers. obs.) 


21 


Siuslaw River 


30 Aug. 1991 


Nelson (pers. obs.) 


22 


Valley of the Giants 


29 Jun. 1993 


Nelson (pers. obs.) 


23 


Valley of the Giants 


29 Jun. 1993 


Nelson (pers. obs.) 


24 


Valley of the Giants 


24 Aug. 1993 


Nelson (pers. obs.) 


25 


Valley of the Giants 


24 Aug. 1993 


Nelson (pers. obs.) 


26 


Valley of the Giants 


24 Aug. 1993 


Nelson (pers. obs.) 


27 


Valley of the Giants 


21 Sep. 1993 


Nelson (pers. obs.) 


28 


Valley of the Giants 


25 Aug. 1993 


Nelson (pers. obs.) 


29 


Valley of the Giants 


21 Sep. 1993 


Nelson (pers. obs.) 


30 


Valley of the Giants 


12 Jul. 1990 


Nelson (pers. obs.) 


31 


Valley of the Giants 


14 May 1991 


Nelson (pers. obs.) 


32 


Valley of the Giants 


14 Jul. 1992 


Nelson (pers. obs.) 


Washington 








33 


Nemah River 


7 May 1993 


Ritchie (pers. comm.) 


34 


Lake 22 Creek 


9 Jul. 1990 


Hamer (pers. obs.) 


35 


Lake 22 Creek 


2 Aug. 1990 


Hamer (pers. obs.) 


36 


Dungeness River 


10 Sep. 1990 


Holtrop (pers. comm.) 


37 


Heart of the Hills Trail 


26 Jul. 1991 


Hamer (pers. obs.) 


38 


Jimmey Come Lately Creek 


24 Jul. 1991 


Holtrop (pers. comm.) 



continues 



70 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter6 



Characteristics of Nest Trees and Nesting Stands 



Table 1— continued 



State/province 


Location 


Date found 


Sources 


Record no. 








British Columbia 








39 


August Creek, Vancouver Is. 


12 Sep. 1993 


Burger (pers. coram) 


40 


Carmanah Creek, Vancouver Is. 


2 Oct. 1992 


Jordan and Hughes (in press) 


41 


Walbran Creek, Vancouver Is. 


12 Oct. 1992 


Jordan and Hughes (in press) 


42 


Walbran Creek, Vancouver Is. 


3 Aug. 1990 


Manley and Kelson (in press) 


43 


Walbran Creek, Vancouver Is. 


24 Aug. 1991 


Manley and Kelson (in press) 


44 


Walbran Creek, Vancouver Is. 


25 Aug. 1992 


Jordan and Hughes (in press) 


45 


Caren Range 


1 Aug. 1993 


P. Jones (pers. comm) 


46 


Clayoquot River 


1993 


Kelson (pers. comm.) 


47 


Megin River 


1993 


Manley (pers. comm.) 


Alaska 








48 


Afognac Is., Alaska Peninsula 


26 Jul. 1992 


Nashmd and others (in press) 


49 


Afognac Is., Alaska Peninsula 


6 Aug. 1992 


Naslund and others On press) 


50 


Kodiak Is.. Alaska Peninsula 


17 Aug. 1992 


Naslund and others (in press) 


51 


Kodiak Is., Alaska Peninsula 


17 Aug. 1992 


Naslund and others (in press) 


52 


Naked Is., Prince William Sound 


13 Jim. 1991 


Naslund and others (in press) 


53 


Naked Is.. Prince William Sound 


25 Jun. 1991 


Naslund and others (in press) 


54 


Naked Is.. Prince William Sound 


6 Jul. 1991 


Naslund and others (in press) 


55 


Naked Is., Prince William Sound 


26JuL1991 


Nashmd and others (in press) 


56 


Naked Is., Prince William Sound 


1 Jul. 1991 


Naslund and others (in press) 


57 


Naked Is.. Prince William Sound 


25 May 1992 


Naslund and others (in press) 


58 


Naked Is.. Prince William Sound 


20 Jul. 1992 


Naslund and others (in press) 


59 


Naked Is., Prince William Sound 


5 Aug. 1992 


Naslund and others (in press) 


60 


Naked Is.. Prince William Sound 


6 Aug. 1992 


Nashmd and others (in press) 


61 


Naked Is., Prince William Sound 


9 Jun. 1991 


Naslund and others (in press) 



Augustine Island (Cook Inlet), Kodiak Island, the Barren 
Islands, and the Kenai Peninsula (Day and others 1983, 
Mendenhall 1992. Simons 1980). All of these nests were 
located in areas of talus where surrounding rocks formed a 
protected area for the nests, or in areas dominated by 
alder. The egg was laid on existing mat vegetation or bare 
soil. Whereas most of these sites were above the local tree 
line and had only low-lying mat vegetation, the Kenai site 
had a forested area on a nearby slope. An additional ground 
nest found on Prince of Wales Island in southeastern 
Alaska in 1993 was located on a platform of moss covering 
three intertwined roots of a western hemlock (Tsuga 
heterophylla) tree at the top of an 11 -meter high cliff 
(Ford and Brown 1994). The nest had many of the 
characteristics of a tree nest when approached from down- 
slope, but was similar to a ground nest when approached 
from up slope. 



Methods 

We compiled information from 61 nest stands and nest 
trees throughout the geographic range of the Marbled M urrelet 
in North America using published and unpublished 
information. Information from three additional tree nests in 
Alaska were not obtained for this review. We did not include 
data from ground nests in this summary. We summarized 
tree and stand characteristics from 14 tree nests in Alaska 
(Naslund and others, in press), nine nests in British Columbia 
(Burger, pers. comm.; P. Jones, pers. comm.; Jordan and 
others in press; Kelson, pers. comm.; Manley, pers. comm.; 
Manley and Kelson, in press), six nests in Washington (Hamer, 
unpubl. data; Holtrop, pers. comm.; Ritchie, pers. comm.), 
22 nests in Oregon (Nelson, unpubl. data), and 10 nests in 
California (Binford and others, 1975; Chinnici, pers. comm.; 
Folhard, pers. comm.; Hamer, unpubl. data; S.W. Singer, 
pers. comm.; Singer and others, 1991) {table 1). 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



71 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



The sample size for each nest characteristic varied because 
some variables were not measured at some nest sites, or the 
information was not available to us. A protocol that outlined 
a methodology for measuring the structure of nests was not 
available until 1993 (Hamer 1993), so some characteristics 
of earlier nests were not measured. Stands were delineated 
and stand sizes calculated generally by defining stands as a 
contiguous group of trees with no gaps larger than 100 m. 
Stand ages were derived from stand information data bases 
of the landowners or by aging individual trees in the stand 
using increment bores. Limb diameters were generally 
reported with the moss cover on the limb included in the 
measurement. Nest platform lengths were measured as the 
length of the nest branch until it was judged to be too narrow 
to support a nest (<10 cm). 

We calculated the range, mean, and standard deviation 
for each nest and stand characteristic for each state or province. 
In addition, we pooled the sample of nests for what we term 
the "Pacific Northwest", using data from nests located in 
California, Oregon, Washington, and British Columbia (tables 
2 and 3). Nests located in Alaska were treated as a separate 
sample (tables 2 and 3). 

We chose to segregate the data using state or provincial 
boundaries because different forest types generally occur 
within these boundaries. Forest types in California within the 
murrelet's breeding range were predominately coastal redwood 
(Sequoia sempervirens). Oregon had fire regenerated stands 
dominated by Douglas-fir (Pseudotsuga menziesii), and in 
Washington, mixed forests of western red cedar (Thuja plicata), 
western hemlock, Douglas-fir, and Sitka spruce (Picea 
sitchensis), created by the combined forces of fire and wind, 
covered the majority of the landscape. British Columbia was 
similar to Washington, with the addition of yellow cedar 
(Chamaecyparis nootkatensis), found in stands at higher 
elevations. Forest types in Alaska were very distinct, with 
many stands dominated by mountain hemlock (Tsuga 
mertensiana) which were small in stature and diameter. 

Results 

Landscape Characteristics 

Distance to Salt Water 

A sample of 45 nests in the Pacific Northwest were 
located a mean distance of 16.8 km inland (table 2, fig. 1). 
Nests in California were found a mean distance of 13 km 
from salt water; the farthest inland nest in California was 
located 28.9 km inland (table 2). The farthest inland nest in 
Oregon was located 40 km from the sea. This coincides with 
a historical record of a downy young found on the ground 40 
km inland on the South Fork of the Coos River in Coos 
County (Nelson and others 1992). In Washington, nests 
were located a mean distance of 16 km inland. Other 
information from Washington indicated nesting at stands 
further inland than known nest sites. A small downy chick 
was located by the senior author on the ground along a trail 
on the east shore of Baker Lake in 1991, 63 km from the 
ocean. Another downy chick was located 45 km inland in 



Helena Creek, in Snohomish County (Reed and Wood 1991). 
Six additional records of eggs, downy young, and fledglings 
found 29-55 km inland in Washington were compiled by 
Leschner and Cummins ( 1 992a), and Carter and Sealy ( 1 987b). 
In British Columbia, nest trees were located a mean 
distance of 1 1.5 km from the Pacific. In addition, there was a 
record of a fledgling found on the ground near Hope, British 
Columbia, 101 km from salt water (Rod way and others 
1 99 1 ). This is the farthest inland distance recorded for Marbled 
Murrelets in North America. Nest trees in Alaska were 
typically located close to the coast, with a mean distance of 
0.5 km (table 2), corresponding to the closer inland distribution 
of suitable nesting habitat. 

Elevation 

The mean elevation of nest trees from a sample of 45 
murrelet nests in the Pacific Northwest was 332 m (table 2). 
In Alaska nest trees were low in elevation with a mean of 96 
m and a maximum of 260 m (table 2). 

Aspect 

Nest stands in the Pacific Northwest occur on a variety 
of aspects. Twenty-six percent of the stands had northeast 
aspects, 12 percent southeast, 28 percent southwest, 12 percent 
northwest, and 21 percent were on flat topography with no 
aspect (table 2). In Alaska, 93 percent of the nest stands had 
westerly aspects (NW, W, or SW), with the majority (50 
percent) facing northwest. 

Slope 

Nests in the Pacific Northwest were located on slopes 
with moderate gradients, with a mean of 23 percent. Slope 
gradients for nests in Alaska were higher than nests for the 
Pacific Northwest with a mean slope of 69 percent. 

The majority of nests in the Pacific Northwest (80 percent) 
were located on the lower one-third or middle one-third of 
the slope. Nest stands in Alaska were located low in elevation, 
but were usually located on the top one-third of the slope, 
unlike nests in the southern part of the range. Nest stands in 
Alaska have been described as being located on gradual or 
moderate slopes (Naslund and others, in press). 

Forest Characteristics 

Age 

For a sample of 16 nests in the Pacific Northwest the 
mean stand age was 522 years with the youngest stand age 
reported as 180 years old (table 2). The oldest stand was 1,824 
years old located on the mainland coast of British Columbia, 
and was dated using nearby stumps from a recent clear-cut. To 
date, all 61 tree nests found in North America have been found 
in stands described as old-growth or mature forests. 

Tree Size 

The mean d.b.h. of trees in nest stands was not reported 
for many sites. Nest stands in Washington and Oregon were 
characterized by large diameter trees (x = 47.7 cm), a mean 
density of large trees (>46 cm d.b.h.) of 93.8/ha, an average 



72 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



Table 2 The mean, standard deviation, range, and sample size for the forest stand characteristics of Marbled Murrelet tree nests located in North America. 

The Pacific Northwest data include nests located in California, Oregon, Washington, and British Columbia For some characteristics, either no data were 
available for that state or province, or the sample size was too small to calculate the mean and range. Sample sizes for each variable are shown in parenthesis 




Pet. composition low elevation trees 2 



Total tree density- (number/ha; 



Canopy height (m) 



100±0 


100±0 


100-100 


100-100 


(10) 


(10) 


235±178 


I20±72 


92-504 


48-282 




(10) 


x8r<> 


59±g 


88-88 


48-75 


(5) 


(9) 



90±9 

78-100 

(5) 

84-162 



64±29 


91±19 


64114 


20-100 


20-100 


39-91 


(6) 


(31) 


(8) 


297±136 


1821132 


5751240 


148-530 


48-530 


295-978 


(5) 







Canopy layers (number) 



Canopy closure (pet) 



Distance to coast 




Distance to stream (m) 



1001165 

5-500 

(7) 



1591224 

5-1000 

(29) 



1091108 

2-325 

(9) 

















— 


15-300 


18-120 


— 


15-700 — 
(30) 


Stand age (yrs) 


— 


209148 


8791606 


— 


5221570 




— 


180-350 


450-1736 


— 


180-1824 — 






(10) 


(3) 




(16) 



'Slope position codes: (1) lower 1/3, (2) middle 1/3, and (3) upper 1/3. 

2 Measure of the percent of western hemlock, Douglas-fir, western red cedar, Sitka spruce, and coastal redwood in a stand. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



73 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



Table 3 — The mean, standard deviation, range, and sample size for platform and tree characteristics of Marbled Murrelet tree nests (n = 61) located in North 
America. The Pacific Northwest data include nests located in California, Oregon, Washington, and British Columbia. For some characteristics, either no data 
were available for that state or province, or the sample size was too small to calculate the mean and range. Calculations were rounded to the nearest cm for all 
measurements except nest substrate depth. Sample sizes for each variable are shown in parenthesis. 



Characteristics 









British 


Pacific 




California 


Oregon 


Washington 


Columbia 


Northwest 


Alaska 


n=10 


n = 22 


n = 6 


n = 9 


n = 47 


n=14 



Tree species 



Douglas-fir 
Western he 
Western red cedar 




Nest platform length (cm) 



"Jest platform width (cm) 




Nest platform moss depth (cm) 



~ 



platform duff 



6-23 

(10) 

2.912.7 

0.8-8.1 

(5) 



7-51 

(21) 

5.1±2.5 

0.6-12 

(17) 



10-39 
(5) 

2.7±0.7 

2.0-3.5 

(2) 



9-19 
(6) 

4.8±1.4 

2.7-7.0 

(9) 



7-51 

(42) 

4.512.4 

0.6-12 

(33) 



3.9±1.3 

2.0-6.0 

(12) 



JIU UUCl UCJHI! I.LIII^ 



Cover above nest (pet) 



2.5-20.0 

(4) 

90±28 

5-100 

(10) 



J.4I0.4 
3.0-3.8 

<2) 

79±14 

40-100 

(18) 



2.9+0.7 
2.0-3.8 

90±10 

70-100 

(5) 



10010 

100-100 

(2) 



5.015.2 
2.0-20.0 

(9) 

85+20 

5-100 

(35) 



89105 

81-95 

(8) 



74 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 




5 10 15 20 25 30 35 40 45 50 

DISTANCE FROM COAST (km) 

Figure 1 — Distances from the Marbled Murrelet nest trees (n = 35) to 
the nearest salt water for nests found in the Pacific Northwest. The 
number of nests was listed in 5-km increments beginning with nests 
found 0-5 km inland. 



total tree density (>10 cm d.b.h.) of 324/ha, multiple canopy 
layers (2-3), and the presence of snags (>10 cm d.b.h.) 
(mean density = 71/ha) (Nelson and others, in press). In 
Alaska, most nest trees were located in forests with 
significantly larger tree size classes (>23 cm d.b.h.) and 
higher volume classes (1883-5649 m 3 /ha) than other forest 
types (Kuletz and others, in press). 

Tree Species Composition and Stem Density 

Conifer species that typically grow at higher elevations 
in the Pacific Northwest include mountain hemlock, silver 
fir (Abies amabilis), and yellow cedar. Conifer species 
most abundant at lower elevations include Douglas-fir, 
western red cedar, Sitka spruce, western hemlock, and 
coastal redwood. Nest stands in the Pacific Northwest were 
composed primarily of low elevation conifer species ( x = 
91 percent) (table 2). In Alaska, the composition of low 
elevation trees was much lower, with a mean of 64 percent. 
The total mean tree density for nest stands in the Pacific 
Northwest was 1 82 trees/ha; total density was about three 
times greater in Alaska (table 2). 

All nest trees in the Pacific Northwest were recorded in 
stands characterized as old-growth and mature forest. These 
stands were dominated by either Douglas-fir, coast redwood, 
western hemlock, western red cedar, or Sitka spruce. The 
one exception was a higher elevation nest stand found in the 
Caren Range of British Columbia which was dominated by 
old-growth mountain hemlock (60 percent) with smaller 
percentages of yellow cedar (20 percent) and silver fir (20 
percent). In California, nest stands were dominated by coast 
redwood and Douglas-fir, with a component of western 
hemlock and Sitka spruce in some nest stands. In both 
central and northern California, all nest sites had a higher 
percentage of redwood trees than Douglas-fir. Nest stands 



in Oregon were dominated by Douglas-fir and western 
hemlock, with one site dominated by Sitka spruce. Forest 
types in Washington included stands dominated by western 
hemlock, Douglas-fir, and Sitka spruce. These stands 
commonly had a large component of western red cedar. 
Silver fir made up a smaller component of some of the nest 
stands in Washington. 

In British Columbia, six nest stands were dominated 
primarily by Sitka spruce and western hemlock, with four 
stands also having a component of silver fir, and one stand 
with western red cedar. One nest stand in the Caren Range 
was dominated by mountain hemlock. For a sample of eight 
nests located in Alaska, mountain hemlock was the dominant 
tree species at five nests, and western hemlock was the 
dominant species at three nest stands (Naslund and others, in 
press). Sitka spruce were reported as an important component 
at most of these nest sites. 

Canopy Characteristics 

Nest stands in the Pacific Northwest had a mean canopy 
height of 64 m with the redwood zone included in this sample 
(table 2). The mean canopy height for stands located in Oregon, 
Washington, and British Columbia was 61 m. The canopy 
height of Alaska nest stands were lower (3c =23 m), reflecting 
the small stature of the trees in this geographic area. 

For nest stands in the Pacific Northwest, the mean canopy 
closure was 49 percent, and all nest stands were reported to 
have 2-4 tree canopy layers where this variable was recorded 
(table 2). Canopy closures below 40 percent were reported 
for 40 percent of the nest stands (fig. 2). Mean canopy 
closures were especially low in California and Oregon. Canopy 
closures for a typical old-growth stand in Washington 
generally average 80 percent. Canopy closures reported from 
Alaska were similar to nest stands in the Pacific Northwest 
(table 2) with a mean of 62 percent. 

The presence of dwarf mistletoe (Arceuthobium) in the 
nest stands or within the canopy of nest trees was not reported 
consistently enough to determine its importance to murrelets. 
Mistletoe was reported at 13 of 20 nest stands, where its 
occurrence was evaluated. 

Stand Size 

Mean nest stand size for the Pacific Northwest was 206 
ha. Several nest stands were only 3, 5, and IS ha in size. In 
Alaska, stands were naturally fragmented in many cases, 
and averaged 3 1 ha. Stand sizes were generally smaller in 
Alaska because of the naturally fragmented nature of the 
coastal forests in this region. 

Distance to Openings 

Distance of nest trees to streams for nests in the Pacific 
Northwest was variable, with a mean of 159 m. Nest trees 
were located a mean distance of 92 m from natural or man- 
made openings (table 2). A combined analysis indicated that 
the mean distance to an opening or stream was 123 m (n = 
68, s.d. = 177). Sixty-six percent of the nest trees were <100 
m from an opening (fig. 3). 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



75 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



8 



w7 
Q 

|. 

w 

h5 

CO 

LU 

Z4 

Li- 

°3 

LU 
CD O 

2 Z 
21 



10 20 30 40 50 60 70 80 90 100 
CANOPY CLOSURE OF STAND AROUND NEST(%) 

Figure 2 — Canopy closure of the stand surrounding the nest tree for 34 
Marbled Murrelet nests found in North America. The number of nests 
was listed in 10-percent increments beginning with nests with 0-10 
percent canopy closure. 










-Ill 



50 150 250 350 450 'i»500 

DISTANCE FROM STREAM OR OPENING (m) 

Figure 3 — Distances from the Marbled Murrelet nest trees (n = 68) to the 
nearest stream, creek, or opening for nests found in North America. 
Some nests had two measurements, one to the nearest opening and 
one to the nearest stream. 



Tree Characteristics 

Nest trees used by murrelets in the Pacific Northwest 
included Douglas-fir (57 percent), Sitka spruce (15 percent), 
western hemlock (13 percent), coast redwood (11 percent), 
and western red cedar (2 percent) (table 3). In one exception, 
a nest in British Columbia was found in a yellow cedar (2 
percent). Western hemlock was the only nest tree species 
reported used by Marbled Murrelets throughout their 
geographic range. Although Sitka spruce was only reported 
from Alaska, British Columbia, and Oregon, it is likely this 



species is also used throughout the range of the murrelet 
since it is common in coastal coniferous forests of Washington 
through California. Douglas-fir nest trees were only located 
in Washington, Oregon, and California. Nests in cedar trees 
were reported only from Washington and British Columbia, 
but this was probably due to a small sample size. Mountain 
hemlock nest trees were only reported from Alaska. 

In the Pacific Northwest, the mean nest tree diameter 
was 211 cm, with the smallest diameter nest tree reported 
from Washington, which was a western hemlock 88 cm in 
diameter (table 3). Nest tree diameters were normally 
distributed with a maximum number of trees found between 
140 and 160 cm, and 85 percent of the trees ranging between 
120 and 280 cm (fig. 4). Nest tree diameters were much 
smaller in Alaska ( x = 63 cm) due to the small stature of the 
trees in this region. 

Mean nest tree heights were highest in California and 
Oregon where the majority of nest trees were in redwood 
and Douglas-fir trees which can grow to great heights. Mean 
tree heights were similar between Washington and British 
Columbia where more of the nest trees were in cedar, spruce, 
and hemlock. Mean tree heights in the Pacific Northwest 
were 66 m (table 3). Nest tree heights in Alaska were low, 
with a mean of 23 m, with one nest tree measured at 16 m. 

The mean diameter of the tree trunk at nest height was 
88 cm in the Pacific Northwest, with minimum trunk diameters 
of 36 cm and 40 cm reported for Oregon and Washington 
respectively. Trunk diameters at the nest height were not 
reported for nests in Alaska (table 3). 

The condition of nest trees in the Pacific Northwest 
varied, with 64 percent recorded as alive/healthy and 36 
percent as declining (n = 44). No nests were reported in 
snags. Nest trees with declining tops (8 percent), broken 
tops (37 percent) and dead tops (8 percent) were commonly 
reported, with only 47 percent of the nest tree tops recorded 
as alive/healthy. In Alaska (n = 14), 57 percent of the nest 
trees were reported as declining, and one nest tree was 
recorded as dead. 

In the Pacific Northwest, mean nest branch height was 
45 m (table 3). Mean nest branch height was highest in 
California and Oregon, where the mean tree height was also 
the highest. Mean nest branch height was lowest in Alaska 
(13 m), with one nest located only 10 m above the ground. 

The mean diameter of nest branches measured at the 
tree trunk and at the nest varied little between each state or 
Province for the Pacific Northwest (table 3). Mean nest 
branch diameters at the nest for each state or province ranged 
from 27-34 cm with a mean diameter of 32 cm for the Pacific 
Northwest. The distribution of limb diameters at the nest in 
the Pacific Northwest were normally distributed, with a 
maximum number (22 percent) of nests located on limbs 35- 
40 cm in diameter (fig. 5). In Alaska, the smallest branch 
diameters at the nest were 12, 14, and 16 cm, with a mean 
diameter of only 1 9 cm. The length of the nest branches in 
the Pacific Northwest ranged from 1 m to 14 m, with a mean 
length of 5.3 m (n = 42). 



76 



USD A Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter6 



Characteristics of Nest Trees and Nesting Stands 



10 



co 8 

UJ 
UJ 

or 



co 6 

UJ 

u. 

LU 
CD 

i 2 







l: 








- 


1 


1 




1... 1 


80 120 160 


200 240 280 320 360 400 440 480 520 560 



NEST TREE DIAMETER (cm) 

Figure 4 — The diameter at breast height for 46 nest trees of the Marbled Murrelet found 
in Califorid, Oregon, Washington, and British Columbia. The number of nest trees was 
listed 1 1 20-cm increments beginning with trees 70-80 cm in diameter. 



CO 

r- 

co 

HI 



O 
a: 

UJ 

m 



o . 






o 






A - 










9 








[III 

III 


| III 



5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 
LIMB DIAMETER AT NEST (cm) 

Figure 5 — The diameter of the tree limbs under or next to 41 nests of the Marbled 
Murrelet found in California, Oregon, Washington, and British Columbia The 
number of nests was listed in 5-cm increments beginning with limbs 0-5 cm 
in diameter. 



The condition of the nest branches for nests in the Pacific 
Northwest varied from healthy limbs (70 percent) to those 
reported as declining (27 percent) or dead (3 percent) (n = 37). 
Nest limbs with broken ends were reported in 16 percent of 
the records (n = 37). In Alaska, 50 percent of the nest branches 
were recorded as declining, 7 percent were reported with 
broken ends, with 1 nest located on a dead branch (« = 14). 



The position of the nest on the tree bole was calculated 
by dividing the nest height by the total tree height. Nests in 
the Pacific Northwest were located an average of 68 percent 
up the bole of the nest tree (table 3). Fifty-nine percent of the 
nests were located in the top one-third of the tree bole, and 
87 percent of the nests were located in the top one-half of the 
tree. No nests were located lower than 40 percent of the total 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



77 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



tree bole height. Nests in Alaska were also located high up 
the tree bole with a mean of 59 percent. Positions of the nest 
on the tree bole for all nests throughout the range of the 
Marbled Murrelet showed that the top 10 percent of the tree 
was not utilized to any great degree, with a maximum number 
of nests located 70-80 percent up the tree bole (fig. 6). 

The majority of nest limbs in the Pacific Northwest (n = 
44) were oriented toward the south or the north. Forty-four 
percent of the limbs faced a southerly direction ranging 
between 136 and 225 degrees (table 3). Another group of 
nests (26 percent) were oriented in a northerly direction 



ranging between 316 and 45 degrees. Nest limbs oriented 
toward the east or west consisted of 14 percent and 16 
percent of the sample respectively. 

Nest Characteristics 

Nest cups were located a mean distance of 89 cm from 
the tree bole for nests in the Pacific Northwest (table 3). Here, 
a total of 7 1 percent of the nests were located within 1 m of 
the tree bole. This relationship was also true for nests located 
throughout the North American range (fig. 7), as 51 percent 
of the nests were located within 40 cm of the tree trunk. 




6 8 10 12 

NUMBER OF NESTS 



Figure 6— The relative vertical positions of Marbled Murrelet nests in relation to 
the heights of the tree bole for 59 tree nests found in North America. 



24 + 
22 



20 
CO 

W16 
u_ 14 - 

2 12 

a: M 

£ 10 



3 
Z 



I, 



HIM in 



MM 



i — i — t- 



-™- 



20 60 100 140 180 220 260 300 340 
NEST DISTANCE FROM TRUNK (cm) 



762 



Figure 7 — Nest distances from the tree trunk for 57 Marbled Murrelet nests found 
in North America. The number of nests was listed in 20-cm increments beginning 
with nests found 0-20 cm from the tree trunk 



78 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



Nest platforms in the Pacific Northwest had a mean 
length of 32 cm and a mean width of 22 cm. The mean total 
platform area was 842 square cm (table 3). In the Pacific 
Northwest, moss (Isothecium) formed the major proportion 
of the substrate for 67 percent of the nests. Litter, such as 
bark pieces, conifer needles, small twigs, and duff, was 
substrate in 33 percent of the nests. For nests found throughout 
North America, moss formed 49 percent of the substrate, 
moss mixed with lichen or litter formed 30 percent of the 
nests, and litter 21 percent (n = 37). All nests found in 
Alaska had moss as a component of the nest substrate. 

Mean moss depth at, or directly adjacent to, the nest cup 
was 4.5 cm (table 3). Mean litter depth was 5 cm for nests in 
the Pacific Northwest. Mean moss depths in Alaska were 3.9 
cm. The majority (86 percent) of nests in North America (n 
= 52) had substrates that were >2 cm in depth with a large 
number of nests (n = 16) having substrate depths between 
3.1 and 4.0 cm (fig. 8). 

Nest platforms in the Pacific Northwest (n = 44) were 
created by large primary branches Y& 32 percent of the cases. 
In addition, 23 percent of the rests were located on tree 
limbs where they became larger in diameter when a main 
limb forked into two secondary limbs, or a secondary limb 
branched off a main limb. In many instances, branches were 
also larger in diameter where they were attached to the tree 
bole. Locations where a limb formed a wider area where it 
grew from the trunk of a tree formed 1 8 percent of the nest 
platforms. Cases of dwarf mistletoe infected limbs (witches' 
broom) (9 percent), large secondary limbs (7 percent), natural 
depressions on a large limb (7 percent), limb damage (2 
percent), and an old stick nest (2 percent) were also recorded 
as forming platforms. Multiple overlapping branches at the 



point where they exited the trunk of a tree were sometimes 
used as a nest platform. Many of the tree limbs creating nest 
platforms had grooves or deformations forming natural 
depressions on the surfaces of the limb. 

Cover directly above the nest was high in almost all 
cases in the Pacific Northwest, with a mean of 85 percent. 
Eighty-seven percent of all nests had >74 percent overhead 
cover. Cover above the nest platforms in Alaska was similar 
to that in the Pacific Northwest (table 3). 

Discussion 

Marbled Murrelets have a limit on their inland breeding 
distribution because of the energetic requirements of flying 
inland to incubate eggs and feed young. They forage at sea, 
carrying single prey items to the nest and feed their young 
several times per day during the late stages of nesting. To 
some extent, the inland distance information presented here 
is biased towards lower values, because nest search and 
survey efforts have been more intensive closer to the coast 
in all regions, where higher murrelet detection rates make 
locating nests an easier task. Even with the potential problems 
of energetic expenditure, murrelets displayed a great tolerance 
for using nesting stands located long distances from the 
ocean. Evidence of breeding was common in California, 
Oregon, and Washington, in areas located 30-60 km inland. 
Unlike many other alcids, the Marbled Murrelet forages in 
near-coastal shallow water environments. The use of tree 
limbs as a nesting substrate may have developed because 
older-aged forests were the only habitats that were abundant 
and commonly available close to the foraging grounds of 
this seabird. Areas of brush-free open ground or rocky talus 




1 2 3 4 5 6 7 8 9 10 >10 
MOSS AND DUFF DEPTH AT NEST CUP (cm) 

Figure 8 — The depth of moss and litter under or directly adjacent to the nest cup 
for 52 nests of the Marbled Murrelet in North America. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



79 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



slopes that are commonly used by other alcids as nesting 
habitat, are not commonly available along the forested coasts 
of the Pacific Northwest. Old-growth and mature forests 
also provided large nesting platforms on which to raise 
young. Nesting greater distances from the coast may have 
developed over time to avoid higher nest predation by corvids 
and gulls whose population numbers may be much higher in 
food-rich coastal areas. In addition, much of the near-coastal 
nesting habitat has been eliminated in the Pacific Northwest 
which may cause birds to nest further inland. Nest search 
efforts and surveys for the presence of murrelets should be 
conducted in areas farther inland in order to refine the 
abundance and distribution of this seabird away from the 
coast. We currently have no information to determine what 
proportion of the population nests in these inland areas, or 
any data to compare the reproductive success of far versus 
near-coastal nesting pairs. 

In Washington, inland detection rates of Marbled 
Murrelets did not show declines until inland distances 
were >63 km from salt water (Hamer, this volume). In 
Oregon, most detections occurred within 40 km of the 
ocean (Nelson, pers. obs.). In British Columbia, murrelet 
detection rates in Carmanah Creek on Vancouver Island 
decreased with increasing distance from the ocean (Manley 
and others 1992). Savard and Lemon (1992) found a 
significant negative correlation between detection frequency 
and distance to saltwater on Vancouver Island in only 1 of 
3 months tested during the breeding season. Inland distances 
for all nests in Alaska were low because rock and icefields 
dominate the landscape a few kilometers from the coast in 
most regions. 

We found that all nest trees throughout the geographic 
range were located in stands defined by the observers as old- 
growth and mature stands or stands with old-growth 
characteristics. The youngest age reported for a nesting stand 
was 1 80 years. Marbled Murrelet occupancy of stands, and 
the overall abundance of the species has been related to the 
proportion of old-growth forest available from studies 
conducted in California, Washington, and Alaska (Hamer, 
this volume; Kuletz, in press; Miller and Ralph, this volume; 
Raphael and others, this volume). Carter and Erickson ( 1 988) 
reported that all records of grounded downy young and 
fledglings (young that have fallen from a nest or unsuccessfully 
fledged) (n = 17) that they compiled were associated with 
stands of old-growth forests in California. All records of 
nests, eggs, eggshell fragments, and downy chicks in 
Washington have been associated with old-growth forests (n 
= 17) (Hamer, this volume; Leschner and Cummins 1992a). 

Marbled Murrelets consistently nested in low elevation 
(<945 m) old-growth and mature forests. Tree species that 
are most abundant at lower elevations (<945 m) such as 
Douglas-fir, western hemlock, Sitka spruce, redwood, and 
cedar, may have a higher abundance of potential nest platforms 
than the higher elevation conifers such as silver fir and 
mountain hemlock. 



Marbled Murrelets were found nesting in stands of very 
small size in some instances, although the reproductive success 
of these nests compared to stands of larger sizes was not 
known (but see Nelson and Hamer, this volume b). A wide 
range of canopy closures were reported for nest stands and 
around nest sites. A study conducted in Washington and 
Oregon compared random plots within a stand to plots 
surrounding the nest tree (Nelson and others, in press). They 
found that canopy closures were significantly lower around 
nest trees in Oregon compared to random plots adjacent to 
the nest tree, but the relationship was not significant in 
Washington. It is unknown how stand size and canopy closure 
affect nest success, but stands with lower canopy closures 
might have less visual screening to conceal adult visits to the 
nest tree (see Nelson and Hamer, this volume b). Therefore, 
it is possible that low canopy closures within a stand will 
make locating nests easier for visual predators such as corvids. 
In addition, smaller stands will have fewer nesting and hiding 
opportunities for Marbled Murrelets. They may be choosing 
lower canopy closures immediately around the nest to improve 
flight access, but select nest platforms with dense overhead 
cover for protection from predation, as indicated by the 
extremely high cover values found directly over the nest. 

The majority of nests in the Pacific Northwest were 
located within 100 m of water, but a few nest sites were 
found at much longer distances (fig. 3). Small streams and 
creeks commonly bisect stands in the Pacific Northwest, 
creating larger openings and long travel corridors. Murrelets 
are often observed using these features to travel through a 
stand. This may be one reason nest sites were often in close 
proximity to streams. Many nests were also located near 
openings such as roads or clear-cuts, but there may be an 
observer bias to finding nests located in areas with better 
access and viewing conditions. 

A variety of processes contributed to producing potential 
nest platforms within the forest including deformations and 
damage sustained by trees. This is probably why a measure 
of potential nest platforms, and not tree size, was the best 
predictor of stand occupancy by murrelets in Washington 
(Hamer, this volume), as larger diameter trees alone were 
often not responsible for the majority of available platforms 
within a stand. Mistletoe blooms, unusual limb deformations, 
decadence, and tree damage commonly observed in nest 
stands, all appear to create a large number of nest platforms. 
Therefore, the structure of a stand and the processes occurring 
within a stand may be more important than tree size alone in 
producing nesting platforms and suitable habitat for the 
Marbled Murrelet (see Grenier and Nelson, this volume). 

It would still be desirable to know when trees, in general, 
begin producing potential nest platforms. In Washington, 
Hamer (this volume) measured potential nest platform 
abundance using a sample of 2,035 conifers, and found 
platforms were generally available when tree diameters 
exceeded 76 cm. The mean number of platforms/tree was 
found to increase rapidly with an increase in tree diameter 



80 



I'SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



from 50-200 cm. No increase in the mean number of platforms 
was evident for larger trees that ranged from 220-300 cm in 
diameter. These results explain why all the nest trees found 
in the Pacific Northwest were >88 cm in diameter, although 
mistletoe brooms on smaller trees may also provide habitat. 
In southcentral Alaska, the minimum d.b.h. associated with 
a tree having at least one platform ranged from 29-37 cm 
(Naslund and others, in press). 

In a study completed in 1993, nest tree and stand 
characteristics in Washington and Oregon were compared 
between 15 murrelet nests and randomly located dominant 
trees and plots within the same nest stand (Nelson and others 
in press). Nest sites were similar to the forest stands in which 
they were located, except that a significantly higher number 
of potential nest platforms were recorded at nest trees, than 
at random trees. They also found that Marbled Murrelets 
selected trees at nest sites that had >4 potential nest platforms, 
and trees with <3 platforms were avoided. In Alaska (Naslund 
and others, in press), one study compared nest tree char- 
acteristics (n = 14) to a sample of random trees surrounding 
each nest tree, and found nest trees ' /ere larger in diameter, 
had more potential nest platforms, and had greater epiphyte 
cover. This study also concluded that Sitka spruce appeared 
to be the most suitable tree for nesting when compared to 
western hemlock and mountain hemlock, because of its high 
number of platforms and greater epiphyte cover. They also 
found that nest and landing trees tended to be larger in 
diameter than surrounding trees, and nest trees were more 
likely to contain at least one potential nest platform with 
moderate to heavy epiphyte cover when compared to nearby 
trees. Stands with high potential nest platform densities may 
reduce competition for nest branches and provide a high 
diversity of nest site choices. 

Nests located high in the canopy may provide better 
access by adults to the nest site in dense, old-growth stands. 
Nesting as high in the canopy as possible may also help in 
avoiding predation. Although positioning the nest as high 
off the ground as possible would likely reduce the incidence 
of mammalian predators, we have also observed that the 
Steller's Jays (Cyanocitta stelleri), predators of nestlings 
and eggs, often forage in the lower portions of the canopy. 
Better horizontal and vertical cover is available in the top 
portions of the tree crown which may help reduce predation. 
Data needs to be collected on the positioning of nests 
within the live crown of the tree, not just the tree bole, to 
determine if murrelets prefer certain areas of the tree crown 
foliage for nesting. 

Murrelets may choose to place nests near the trunk of 
the tree for a variety of reasons. First, overhead and horizontal 
cover is higher around the nest cup due to the position of the 
tree crown directly overhead. Second, the tree trunk itself 
provides a large amount of cover and visual screening and 
branches are typically larger in diameter near the tree bole. 
Also, more duff and litter, which often form the nest substrate, 
is trapped near the tree bole, and the percent cover of moss 
on the limbs of trees is higher, often forms a more complete 



coverage, and forms a deeper layer near the tree bole. Some 
conifer species typically have little or no moss available on 
their limbs, so that platforms created by accumulations of 
duff and debris are the only nest choices available for murrelets 
in these forest types. 

Murrelets nest on large limbs. The smallest limb used at 
the nest cup throughout the range of the murrelet was 10 cm 
in diameter, which is likely the smallest diameter branch 
that could support a successful nest. Nests located on smaller 
limbs would probably have a higher likelihood of losing 
chicks or eggs from accidental falls, an occurrence that is 
well documented (Hamer and Nelson, this volume a). Nests 
located on limbs <16 cm diameter all had moss as a nest 
substrate, except in one instances where a 1 3 cm nest branch 
had litter and lichen as a substrate. Small limb diameters 
without a moss covering may be avoided by nesting birds 
because the hazards of raising eggs and young are increased 
without the moss to help stabilize and insulate the egg on 
the limb, increase the diameter of the nest limb/platform, 
and provide a substrate on which to create a nest cup 
(depression). In addition, moss and litter may help insulate 
eggs and chicks during cold weather and may help drain 
water from eggs and chicks helping thermoregulation 
(Naslund and others, in press). An abundance of mosses 
creates a multitude of nest platform choices by providing 
substrate on many locations throughout a single limb. In 
addition, the presence of dwarf mistletoe in stands can 
increase the number of nesting opportunities for murrelets 
and may be important in providing nest platforms in areas 
with low moss abundance and dryer conditions. 

The nest site selection of the Marbled Murrelet may 
have evolved primarily to reduce predation. Selection of 
nest sites away from the coast, in dense old-growth and 
mature forests with multi-layered canopies, high in the forest 
canopy, on limbs with high overhead and horizontal cover, 
and near the tree bole where the tree bole itself provides a 
large degree of cover, may help reduce nest predation. Results 
from studies of murrelet habitat use to date have been derived 
from comparisons of stands occupied by murrelets to 
unoccupied stands, comparisons of stands receiving high 
use versus low use, or comparisons of nest trees and nest 
plots to random trees and plots. Although these can provide 
extremely useful descriptions and definitions of suitable 
habitat, they do not provide information on the habitat 
characteristics associated with successful nests. Information 
on the landscape and within-stand habitat characteristics 
that influence reproductive success is needed to fully 
understand murrelet nesting ecology and to model optimum 
habitat suitability for this species. Such studies may find that 
stand size analyzed in conjunction with the number of nesting 
and hiding opportunities within the stand (habitat quality), 
may greatly influence reproductive success because of 
predation pressures at the nest site. Habitat factors that could 
influence reproductive success may include stand fragmen- 
tation, stand canopy closure, and the amount of overhead 
and horizontal cover surrounding the nest. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



81 



Hamer and Nelson 



Chapter 6 



Characteristics of Nest Trees and Nesting Stands 



Acknowledgments 

We would like to acknowledge the contributions of 
unpublished information made available to us by researchers 
and wildlife biologists throughout the range of the Marbled 
Murrelet. Their generous contributions of nest information 
made this summary possible. For the California data, we 
are extremely grateful to Steve Singer for the information 
he has collected over the last two decades. Additional 
information was kindly provided by Sal Chinnici of the 
Pacific Lumber Company in association with Dave Fortna 
and Steve Kerns. Lee Folliard of Areata Redwood Company 



provided information from an additional 2 nests. We would 
like to thank Karen Holtrop for her efforts and successes in 
locating nests in Washington and her contributions of nest 
data. Stephanie Hughes, Kevin Jordan, Irene Manley, John 
Kelson, Paul Jones, and Alan Burger provided valuable 
nest information from British Columbia. We are also grateful 
to Nancy Naslund and Kathy Kuletz for contributing the 
majority of nest site information from Alaska. We thank 
Alan Burger, Martin Raphael, Nancy Naslund, Evelyn Bull, 
Kimberly Titus, and C. John Ralph for reviewing earlier 
drafts of this paper. 



82 



UCDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 7 

Breeding and Natal Dispersal, Nest Habitat Loss and 
Implications for Marbled Murrelet Populations 



George J. Divoky 1 



Michael Horton 2 



Abstract: Evidence of breeding and natal dispersal in alcids is 
typically provided by the resightings of banded birds, the establish- 
ment of new colonies, and/or evidence of immigration to established 
colonies. The difficulties in banding, observing, and censusing 
Marbled Murrelets at nesting areas preclude using any of these 
methods for this species. Based on the limited number of nests 
observed in consecutive breeding seasons, breeding site fidelity 
(birds breeding in the same nest as the previous year) may be lower 
than most other alcids. This is likely due to low breeding success 
associated with high levels of nest predation. By contrast, annual 
use of nest stands suggests fidelity to a nesting area may be high. 
Natal dispersal, the breeding at locations away from their fledging 
site, is likely similar to that of other alcids. Loss or degradation of 
previously occupied nesting habitat will result in the displaced 
breeders prospecting for new nest sites. In areas with no unoccu- 
pied available habitat, this could result u birds being prevented 
from breeding, attempting breeding in ^uboptimal habitat, or in- 
creasing the distance dispersed from the previous breeding sites. 
Each of these is likely to result in a decrease in reproductive 
output. Dispersal patterns need to be considered when assessing 
the importance of stands and the status of populations. The small 
population size and fragmented nature of the remaining breeding 
habitat could increase the time required for prospecting birds to 
locate recently matured old-growth forest, resulting in underesti- 
mating the importance of a stand. Additionally, birds could be 
dispersing from regions of high production of young to areas with 
low production but where recruitment opportunities are higher, 
partially hiding the low reproduction of the latter population. 



The ability of Marbled Murrelets to disperse from natal 
sites, and their fidelity to breeding sites or stands, has important 
implications for the potential of the species to respond to 
habitat loss and colonize or reestablish breeding areas when 
habitat has been altered. With knowledge of these factors, 
we could more accurately assess the effects of habitat 
destruction on the viability of populations throughout the 
species' range. In the discussion below, we examine what is 
known about dispersal in other alcid species and the possible 
implications for the Marbled Murrelet. 

Dispersal of birds can occur both by established breeders 
changing breeding sites (breeding dispersal) and by birds 
nesting away from their natal nesting area (natal dispersal) 
(Greenwood and Harvey 1982). The degree of nest-site fidelity 
by established breeders can be expected to be related to 
previous breeding success and the frequency of change in 
availability of suitable nest sites and prey resources. Nest 



1 Wildlife Biologist, Institute of Arctic Biology. University of Alaska, 
Fairbanks, AK 99705 

2 Wildlife Biologist, U.S. Fish and Wildlife Service. U.S. Department 
of Interior, 2800 Cottage Way, Room El 803, Sacramento, CA 95825 



site availability can be decreased both through the destruction 
of nest sites and through chronic predation. An increased 
rate of natal dispersal should be related to the potential to be 
more successful in finding mates or nest sites away from the 
natal nest site or colony. 

Breeding Dispersal 

Breeding site fidelity in a long-lived species, which the 
Marbled Murrelet is presumed to be (Beissinger, this volume), 
can provide benefits in increased breeding success and 
lifetime fitness. Site fidelity can reduce potential reproductive 
effort by (1) increasing the chances of breeding with the 
previous year's mate, (2) eliminating or reducing the need 
to locate a suitable nest site, and (3) allowing the development 
of familiarity with the marine and terrestrial environment. 

The rate of breeding dispersal is low for most alcid 
species that have been studied. Rates of nest-site fidelity of 
previously breeding alcids are: 91 .5 percent Razorbills (Alca 
torda) (Lloyd 1976); 96 percent Common Murres (Uria 
aalge) (Birkhead 1977); 93.2 percent Atlantic Puffins 
(Fratercula arctica) (Ashcroft 1979), 57-95 percent Black 
Guillemots (Cepphus grylle) (Divoky, unpubl. data; Petersen 
1981); 86 percent Pigeon Guillemots (C. columba) (Drent 
1965); 78 percent Ancient Murrelet (Synthliboramphus 
antiquus) (Gaston 1992). 

The degree of breeding dispersal displayed by an alcid 
should be related to the rate that nesting habitat is created 
and destroyed, the level of mortality of breeding birds, and 
the availability of nest sites. Species with a high probability 
of returning to a nest site destroyed over the winter would 
have fewer reasons to have evolved site tenacity. Harris and 
Birkhead (1985) suggested that the Thick-billed Murre {Una 
lomvia) might show less site tenacity than other Atlantic 
alcids because rockfalls destroy or create nest sites in their 
colonies more frequently than for other species. Burrow 
nesting alcids could be expected to show higher rates of 
breeding dispersal than talus nesters due to the higher 
frequency of collapse of burrows. 

Annual overwinter mortality could be expected to 
influence breeding site fidelity. High overwinter mortality 
would decrease the chances of a surviving bird being able to 
breed with the previous year's mate and, by creating more 
vacancies at established nest sites, increase the opportunities 
for dispersal for species that are nest site limited. 

For those alcid species in which breeding site fidelity 
has been examined, and for birds in general (Greenwood and 
Harvey 1982), changes in nest site are more frequent after a 
breeding failure. For Black Guillemots, nest-site fidelity 
was 92 percent for successful pairs and 48 percent for failed 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



83 



Divoky and Horton 



Chapter 7 



Dispersal, Habitat Loss, and Implications 



pairs (Petersen 1981). For Ancient Murrelets, reoccupancy 
rates of burrows that supported successful breeding the 
preceding year was 80 percent, and only about 50 percent for 
unsuccessful burrows (Gaston 1992). Nest changes caused 
by simple breeding failure typically result in small scale 
movements (usually tens of meters) to nearby sites (Divoky, 
unpubl. data; Petersen 1981). 

Chronic disturbance at the nest site can cause estab- 
lished breeders to move to a new breeding location thousands 
of meters away. A Pigeon Guillemot that experienced 
persistent disturbance at its nest site was found breeding on 
an island 7.7 km away 3 years later (Drent 1965). At a Black 
Guillemot colony where any movement of established 
breeders is typically to an adjacent nest site (<10 m), one 
bird moved approximately 1 km and another over 5 km, 
after Horned Puffins (Fratercula corniculata) using the 
same nest site had repeatedly disrupted nesting (Divoky 
1982 and unpubl. data). 

Essentially all information on breeding dispersal in alcids 
has been obtained through the banding and resighting of 
individuals. The difficulty of capturing and observing Marbled 
Murrelets at the nest site has prevented the collection of 
similar information for this species. The old-growth nesting 
habitat of the Marbled Murrelet is relatively stable. Natural 
destruction of old growth forests through fire or wind storms 
is rare enough, and the degradation of nest trees is slow 
enough, that high site fidelity could have evolved. 

Observations of murrelets engaging in "occupied behavior," 
strongly suggesting nesting (Ralph and others 1993), indicate 
that Marbled Murrelets, as a species, exhibit high fidelity to 
a nesting area. Marbled Murrelets have been recorded in the 
same forest stands for a minimum of 20 years in northern 
California (Strachan, pers. comm.; Miller, pers. comm.), 18 
years in central California (S.W. Singer, pers. comm.), 7 
years in Oregon (Nelson, pers. comm.), and 3 years in 
Washington (Hamer, pers. comm.). These results are in part 
a function of the duration of survey effort. While these 
observations indicate that the species exhibits high fidelity 
to forest stands, no direct information is available on stand 
or nest-site fidelity of individual birds. 

For species having high annual survival and site fidelity, 
the occupation of the same nest site in consecutive years is 
strongly suggestive of individual nest-site fidelity. Re- 
occupation of the same nest site has occurred only once in 
the 13 instances where Marbled Murrelet nests have been 
examined in the breeding season following a year of known 
occupancy (P. Jones, pers. comm.) and nesting occurred in 
the same tree four times (P. Jones, pers. comm.; Naslund, 
pers. comm.; Nelson, pers. comm.; Singer, in press). 
Additional evidence of fidelity to a nest tree is provided by 
Nelson's (pers. comm.) finding of three nest cups on three 
platforms in a single tree, although we do not know if it was 
the same individuals. While the sample size is small, the 
observed fidelity to the same nest depression in consecutive 
years appears to be lower than for other alcids. This could be 
related to the high rate of predation recorded for murrelet 



nests (Nelson and Hamer, this volume b). It also indicates 
that while breeding habitat for this species is reduced (Perry, 
this volume), and may be limiting, the number of nest 
platforms apparently is not. If the high predation rate is a 
recent phenomenon, nest-site fidelity may have been higher 
in the past. As previously mentioned, breeding dispersal 
increases with increased rates of nesting failure (Greenwood 
and Harvey 1982). The high rates of observed nest failure 
(Nelson and Hamer, this volume b) may explain murrelets 
not reoccupying a nest site in subsequent years. 

Natal Dispersal 

The primary benefit that a bird derives from breeding at 
its natal colony may be that the natal area is a known 
location where conspecifics of a similar genetic background 
successfully bred in the past (Ashmole 1962). However if a 
breeding location is near those of related individuals, there is 
the possibility of kin selection occurring and a moderate 
level of inbreeding (Shields 1983). 

Philopatry (chicks returning to their natal colony or 
nesting location to breed) is more difficult to study than the 
fidelity of breeders to a nest site. It had been assumed that 
the majority of alcids surviving to breeding are recruited 
into their natal nesting area (Hudson 1985). More recent 
information, however, shows that prospecting by prebreeders 
at non-natal colonies is a regular occurrence in Common 
Murres (Halley and Harris 1 993) and Atlantic Puffins (Harris 
1983, Kress and Nettleship 1988). Until recently, the instances 
of banded birds initiating breeding at a non-natal colony 
were limited (Asbirk 1979, Lloyd and Perrins 1977). However, 
recent information indicates that, at least in the Atlantic 
Puffin, half the chicks that survive to breeding emigrate to a 
new colony (Harris and Wanless 1991). 

Other evidence of natal dispersal is provided by the 
establishment of new colonies and growth rate of existing 
colonies that could only be explained by immigration (Divoky, 
unpubl. data; Gaston 1992; Petersen 1981). The frequency 
with which new alcid colonies have formed on the west 
coast of North America in the short period that systematic 
censusing has been conducted (table 1) proves that natal 
dispersal is common in the alcidae. 

The distance that birds will breed from their natal site 
can be great. Banding returns show that the distance dispersed 
can be as great as 420 km (by sea) for the Common Murre 
(Halley and Harris 1993) and over 450 km for the Atlantic 
Puffin (Harris and Wanless 1991). The rate of increase of 
some breeding populations, and the establishment of new 
colonies, indicates that Ancient Murrelets are being recruited 
into breeding populations at least 30 km from their natal site 
(Gaston 1992), Black Guillemots from over 500 km, and 
Horned Puffins from over 200 km (Divoky, unpubl. data). 

Because of the difficulties of marking and subsequently 
resighting Marbled Murrelets, any direct evidence of natal 
dispersal would have to come from observations of range 
expansion, occupation of previously unoccupied breeding 



84 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Divoky and Horton 



Chapter7 



Dispersal. Habitat Loss, and Implications 



Table I — Ale id species that have recently formed new colonies in western Sorth America 



Species 


Alaska 


British Columbia 


Washington 


Oregon 


California 


Common Murre 




Campbell and 
others 1975 


Speich and Wahl 1989 


USFWS'.unpubl.data 
Newport, OR 


Sowls and others 1980 
CarteT and others 1992 


Thick-billed Murre 


Sowls and others 
1982 


Vallee and Cannings 
1983 









Pigeon Guillemot 



Sowls and others 

1978 

USFWS unpubl. data. 

Anchorage, AK 



Campbell 1977 Speich and Wahl 1989 USFWS, unpubl data 

Newport OR 



Sowls and others 1980 
Carter and others 1992 



Black Guillemot 



^assin's Auklet 



Divoky and others 
1974 




Carter and others 1992 



Rhinoceros Auklei 



Campbell and 
others 1975 



Speich and Wahl 1 989 USFWS, unpubl. data 
Newport, OR 
Scott and others 1974 



Sowls and others 1980 
Carter and others 1992 




Byrd and others 198C 



Divoky 1982 
Divoky, unpubl. data 




USFWS. unpubl. data 
Newport. OR 



Sowls and others 1980 



i USFWS — U.S. Fish and Wildlife Service 



areas, or growth of local populations that could only be 
accounted for by immigration. The nesting habits of the 
species makes the detection of any of these difficult, as does 
the short period that the species has been the focus of research. 
In addition, the high rate of habitat destruction recently 
experienced (Perry, this volume) adds to these difficulties. 

Natal dispersal can be expected to be high in Marbled 
Murrelets compared with other alcids for several reasons. 
The winter distribution is extensive, with the species wintering 
in the nearshore waters of the breeding range, as well as in 
areas where breeding does not occur. The distance that 
individual birds disperse from either their breeding or natal 
area can be great, as murrelets are regularly found in southern 
California some 300 km south of the closest known breeding 
area (Briggs and others 1987). Because murrelets attend 
inland breeding areas during the winter (Naslund 1993b), 
information on breeding areas is provided to prospecting 
nonbreeders at all times of the year. The prebreeding period 
for this species is probably between 2 and 5 years (Beissinger. 
this volume), allowing sufficient time to prospect for a 
suitable nesting area. Additionally, the area where Marbled 
Murrelets might discover suitable nesting habitat is a 60-km 
band adjacent to the coast. This extensive area of potential 
breeding habitat may have selected for more extensive 



prospecting behavior than in other alcids where potential 
breeding sites are largely linearly distributed in a narrow 
shoreline band. 

Methods of Dispersal 

The manner in which alcids coalesce into breeding pairs 
can have implications for the level of breeding and natal 
dispersal. The vast majority of breeding dispersal in alcids 
consists of birds moving to sites either immediately adjacent, 
or close to, the previously occupied nest site (Divoky, unpubl. 
data). This occurs even when an established breeder initiates 
a new pair bond with another established breeder (Divoky, 
unpubl. data), indicating that pairing for most, if not all, 
alcids occurs near the breeding site. If pairing occurs on the 
water when birds are staging near the breeding location, one 
would expect to see almost random movement of the 
established breeders that lose or change mates. Additionally, 
if established breeders paired on the water, the pair would 
have affinities to two sites. 

Because ownership of a quality nest site or territory is 
an important prerequisite for breeding, pairing at the nest 
site allows a bird to find out whether a prospective mate 
owns a site and to determine the quality of that site. Pairing 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



85 



Divoky and Horton 



Chapter 7 



Dispersal, Habitat Loss, and Implications 



with a bird that owns a nest site increases the chances that a 
bird will pair with an experienced breeder. 

Nonbreeding birds, with no previous experience, also 
probably form pairs near the nest site. Observations of Black 
Guillemots in northern Alaska (Divoky, unpubl. data) show 
that nonbreeders are present at the colony throughout the 
breeding season, and many display a high level of mate and 
site fidelity. Although nonbreeders form pairs with each other, 
when one member of an established nest site owning pair dies, 
the vacancy is typically filled by a nonbreeder of the appropriate 
sex. Nonbreeding pairs can be recruited as a unit should a new 
site be created or should two vacancies occur at an established 
site. However, the low annual mortality rates of breeding 
alcids indicates that most recruitment occurs through a single 
vacancy in an established pair. With recruitment occurring at 
or near the nest site, the established breeder and the individual 
being recruited, can pair with a familiar bird. Recruitment in 
murrelets could occur in the same manner. Those birds 
prospecting new nesting areas could pair on the water before 
prospecting potential nest sites. 

Implications of Habitat Loss and 
Fragmentation of Populations 

The final rule listing Marbled Murrelets as threatened 
(U.S. Fish and Wildlife Service 1992) regards loss of older 
forests and associated nest sites as the main cause of decline 
in murrelet populations. When nest sites are limiting, the 
loss of nesting habitat has both immediate and long term 
impacts on the reproductive potential of a murrelet population. 
While alcid populations have been shown to recover in a 
relatively short period from episodic anthropogenic mortality 
events, such as gill net and oil spill mortality (Piatt and 
others 1991; Carter and others 1992), loss of nesting habitat 
directly affects the long term reproductive potential of a 
population. This is especially true for tree-nesting Marbled 
Murrelet populations where the creation of nesting habitat is 
extremely time-consuming, perhaps 200 years. 

Fragmentation of old-growth also has the potential of 
reducing murrelet breeding success by increasing the densities 
of predator populations. Corvids are "edge species" that 
have been found to increase in numbers with increased 
forest fragmentation (Andren and others 1985, Wilcove 
1985, Small and Hunter 1988). Similar findings have been 
reported in central Oregon regarding Great Horned Owls 
(Johnson 1992). In addition, corvid predation on small bird 
nests has been found to increase with increased forest 
fragmentation, decreased distance of nests from a forest 
edge or both (Gates and Gysel 1978, Andren and others 
1985, Small and Hunter 1988, Yahner and Scott 1988). 
Factors that increase fragmentation, such as a wildfire or 
timber harvest, could reduce murrelet breeding success both 
through the reduction of cover and the increase in predator 
densities. This reduced breeding success could be expected 
to increase the rate, and possibly the distance, of breeding 



dispersal. The distances moved would probably relate to the 
level of disturbance and the threat that the predators pose to 
adult birds. The reduction and fragmentation of habitat 
would also act to increase the distance prospecting prebreeders 
would have to travel to find a suitable nest site. 

Habitat loss could be expected to result in the 
displacement of breeding birds, while fragmentation could 
lead to both displacement and decreased breeding success. 
In cases where stands used for nesting are destroyed, the 
birds previously breeding in the stand would have to locate 
a new nesting area. If all available nest sites in adjacent 
habitats are occupied, the displaced birds could attempt to 
breed in suboptimal sites with a decreased chance of 
successful reproduction, prospect more distant areas, or not 
breed at all. There are no conclusive indications of higher 
densities of murrelet nesting in stands remaining after timber 
harvests (Ralph and others, this volume). The ease and 
rapidity with which displaced murrelets seek out new 
breeding areas could be expected to be related to how 
frequently murrelets normally change sites. If the level of 
individual nest-site fidelity is as low as observations indicate, 
then murrelets may be able to readily move at least short 
distances to new nest sites. The fidelity birds show to a 
previously used breeding area or site that no longer can 
support breeding, should be related to the rate and magnitude 
of habitat destruction. There is evidence of murrelets visiting 
remnants of newly harvested stands before disappearing 
from the area (Folliard, pers. comm.), thus indicating that 
murrelets might not immediately abandon the unsuitable 
nest stand. This is consistent with observations in other 
alcid species. Pairs have shown fidelity to previously 
occupied, and recently destroyed, nest sites for two years in 
the Black Guillemot (Divoky, unpubl. data), and a minimum 
of two years in the Least Auklet (Aethia pusilla) (I. Jones, 
pers. comm.). This type of nest loss would be similar to the 
loss of a previously used murrelet nest platform branch and 
not the removal of a nesting stand. 

Management Implications of Dispersal 

High levels and extensive distances of natal dispersal 
could result in source areas with high productivity producing 
young that will be incorporated into sink regions with low 
productivity, or high adult mortality, or both. This could 
result in populations in sink areas showing little change in 
numbers. Without monitoring breeding success, the inability 
of the sink population to produce enough young to balance 
adult mortality would not be evident. The maintenance of 
such a population would be dependent on the continued 
production of a surplus of young by the source population. 
The true reproductive status of the sink population would be 
masked until immigration declines. Such immigration could 
explain the ability of the central California murrelet population 
to lose an estimated 150 to 300 birds in the early 1980s 
(Carter and Erickson 1988) and not show any signs of decline 
(Carter and others 1992). 



86 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Divoky and Horton 



Chapter7 



Dispersal, Habitat Loss, and Implications 



The secretive nature of murrelet nesting has precluded 
the determination of breeding areas solely by the discovery 
of nests, eggs or chicks. Biologists and managers have had 
to identify breeding areas based on the birds engaged in 
activities included in "occupied behavior" as strongly 
indicative of nesting (Ralph and others 1994). Relying on 
instances of occupied behavior as an indication of the 
importance of a stand to Marbled Murrelets has a number of 
potential weaknesses. 

First, recently matured forests that are able to support 
nesting could not be expected to be immediately discovered 
and occupied by prospecting murrelets. The ability of alcids 
to occupy areas where suitable breeding habitat is made 
available is evident from the rapid colonization of islands in 
the Aleutian Islands where fox have been eliminated (Bailey 
and Kaiser 1993). The occupation of newly available suitable 
habitat by Marbled Murrelets in Washington, Oregon, and 
California may be delayed by the small stand size, high 
fragmentation and disjunct distribution of the old growth 
forest. The small size and apparently low breeding success 
(Nelson and Hamer, this volume b) of the population can be 
expected to further slow occupation i f newly available habitats. 
Because almost all prospecting of currently unoccupied suitable 
habitat would occur through natal dispersal, low productivity 
would reduce the potential of a population to disperse. This 
would result in a lack of detections in stands that have the 
potential of supporting murrelet breeding, but have not yet 
been discovered by murrelets. The importance of this 
apparently suitable but currently unoccupied habitat to the 
future of the species needs to be recognized. 

In regions where a large nonbreeding population is 
prevented from breeding by lack of nest sites, prospecting 
birds might investigate areas and habitats that do not support 
breeding. This could result in "occupied" behavior being 
recorded in areas where nesting is not occurring. Prospecting 
alcids can be present in apparently suitable habitat (Divoky 
1982, unpubl. data; Kress and Nettleship 1988; Carter and 
others 1992), although no breeding is occurring. If loss of 
old-growth habitat has both increased the level of dispersal 
and limited potential nest sites, substantial numbers of 
murrelets could be displaying "occupied behavior" in habitats 
where breeding is not currently being attempted or where 
successful breeding could not occur. Such could be the case 
in central California where Carter and Erickson (1988) 
believed that all remaining nesting habitat is occupied and 
because the population is nest site limited, nonbreeding 
birds may be present over land and sea in a greater percentage 
than elsewhere. While this may result in overestimating the 
use of stands, it is unlikely that murrelets would be repeatedly 
encountered in stands that do not have some present or 
future potential for supporting successful breeding. 



Discussion 

The coastal old-growth forest utilized for breeding by 
Marbled Murrelets would have selected for relatively high 
rates of breeding and natal dispersal. Based on the behavior 
and cryptic coloration of the breeding adults and chicks, and 
the high rate of nest predation for observed nests (Nelson 
and Hamer, this volume b), the risk of nest predation appears 
to be higher than for other alcids. The assumed high rate of 
nest predation would have selected for frequent short distance 
movements, while the extensive time required for old growth 
stands to be destroyed or degraded under natural conditions 
would have selected for individual fidelity to a nesting stand. 
There is no indication that the distance that breeding murrelets 
typically disperse would be any greater than the conservative 
movements (usually <1 km) that have been observed for 
other alcids. 

Most dispersal in alcids is probably due to natal dispersal, 
and Marbled Murrelets appear to have the capacity for extensive 
natal dispersal given the extent of the breeding range, the 
overlap between the wintering and breeding areas, and the 
distance individuals are known to move from breeding areas 
in winter. It would not be unreasonable to assume the percentage 
of birds that initiate breeding at a non-natal locality (natal 
dispersal) is as high or higher than has been reported for other 
alcids (approximately 50 percent) (Harris and Wanless 1991). 
The ability to prospect for breeding localities should be well 
developed in Marbled Murrelets. Unlike the potential breeding 
area of most alcids, which is linearly distributed in a narrow 
band on the shoreline, murrelet nesting habitat is found in a 
wide (as much as 60 km) band adjacent to the coast. 

Breeding habitat fragmentation and loss can be expected 
to have affected the rate and extent of murrelet dispersal. 
In Washington, Oregon, and California, high predation 
rates apparently associated with fragmentation would select 
for increasing the rate and extent of breeding dispersal. 
However, the small size and highly fragmented and disjunct 
nature of the old-growth remaining in this area can be 
assumed to have decreased the potential distance for breeding 
dispersal (at least in areas where stand size is small). Natal 
dispersal rates and extent may have been increased as 
habitat in the natal locality was reduced and the distance to 
the location of suitable habitat is increased. These changes 
in dispersal may have the overall effect of depressing 
reproductive output. 

Acknowledgments 

We thank George Hunt, Linda Long, Phil Detreich, 
and Edward Murphy for helpful comments and work on 
this manuscript. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



87 



Chapter 8 

Nest Success and the Effects of Predation on 

Marbled Murrelets 



S. Kim Nelson 1 



Thomas E. Hamer 2 



Abstract: We summarize available information on Marbled Murrelet 
(Brachyramphus marmoratus) productivity and sources of mortality 
compiled from known tree nests in North America. We found that 
72 percent (23 of 32) of nests were unsuccessful. Known causes of 
nest failure included predation of eggs and chicks (n = 10), nest 
abandonment by adults (n = 4), chicks falling from nests (n = 3), 
and nestlings dying (n = 1). The major cause of nest failure was 
predation (56 percent: 10 of 18). Predators of murrelet nests in- 
cluded Common Ravens (Corvus corax) and Steller's Jays (Cyanocitta 
stelleri): predation of a nest by a Great Horned Owl < Bubo virginianus) 
was also suspected. We believe that changes in the forested habitat, 
such as increased amounts of edge, are affecting murrelet produc- 
tivity. Successful nests were significantly further from edges ( x = 
155.4 versus 27.4 m) and were better concealed ( x = 87.2 versus 
67.5 percent cover) than unsuccessful nests. The rate of predation 
on Marbled Murrelet nests in this stud appear higher than for many 
seabirds and forest birds. If these pre< ation rates are representative 
of rates throughout the murrelet*s range, then the impacts on 
murrelet nesting success will be significant. We hypothesize that 
because this seabird has a low reproductive rate (one egg clutch), 
small increases in predation will have deleterious effects on popu- 
lation viability. Rigorous studies, including testing the effects of 
various habitat features on recruitment and demography, should be 
developed to investigate the effects of predation on Marbled Murrelet 
nesting success. 



Nesting success in seabirds is influenced by a variety of 
physical and biological factors, including food availability, 
habitat quality, energetics, predation, and climatic conditions 
(Croxall 1987, Nettleship and Birkhead 1985, Vermeer and 
others 1993). Because the effects of these factors can vary 
spatially and temporally, seabird nesting success can be 
highly variable among years (Birkhead and Harris 1985; 
Boekelheide and others 1990; De Santo and Nelson, this 
volume). For example, in some years, anomalous warm 
oceanographic conditions (El Nino) cause a decrease in prey 
availability, thus impacting nesting attempts and nest success 
(Ainley and Boekelheide 1990, Hodder and Greybill 1985, 
Vermeer and others 1979). In addition, disturbance to nesting 
habitat (e.g., habitat loss, modification) and associated 
cumulative impacts can affect the ability of seabirds to 
successfully reproduce (Evans and Nettleship 1985; Gaston 
1992. Reville and others 1990). 



1 Research Wildlife Biologist. Oregon Cooperative Wildlife Research 
Unit. Oregon State University, Nash 104, Corvallis. OR 97331-3803 

; Research Biologist. Hamer Environmental. 2001 Highway 9, Mt. 
Vernon, WA 98273 



The influence of these biological and physical factors 
on the nesting success of Marbled Murrelets {Brachyramphus 
marmoratus) is not fully known. In order to completely 
address this issue, well designed studies investigating the 
conditions that directly influence murrelet reproduction are 
needed. However, data are available on murrelet nesting 
success from tree nests that have been located and monitored 
in North America. In this paper, we summarize this 
information on murrelet productivity and sources of mortality. 
In addition, because predation was the major cause of nest 
failure, we discuss the implications of predation on this 
threatened, forest-nesting seabird. 

Methods 

We compiled information on nest success and failure 
from published and unpublished records of 65 Marbled 
Murrelet tree nests found in North America between 1974 
and 1993. The sample size of tree nests were distributed by 
state and province as follows: Alaska (n = 18), British 
Columbia (n = 9), Washington (n = 6), Oregon (n = 22), 
and California (n = 10) {table 1). Success and failure of 
nests were determined through intensive monitoring of 
nesting activity, or evidence collected at the nest. The 
outcomes of nests were compared between regions (Alaska 
versus British Columbia, the Pacific Northwest and northern 
California). Nests were considered to fail if: (1) the chick 
or egg disappeared, fell out of the nest, or was abandoned; 
(2) the chick died; (3) unfaded eggshell fragments were 
found during the breeding season in nest cups without 
fecal rings; or (4) predation was # documented. Nests were 
considered or assumed to be destroyed by a predator based 
on one or more of the following: ( 1 ) predation was observed, 
(2) the egg or chick disappeared prematurely between nest 
observations and neither were located on the ground after a 
thorough search of the area, and (3) evidence, such as 
puncture marks on eggs, or albumen or blood on eggshell 
fragments, was discovered and predators were aware of the 
nest location or seen in the immediate area. In addition to 
data from active nests, information on eggs, nestlings, and 
hatch-year birds found on the ground were compiled from 
published and unpublished records between 1900 and the 
present. 

We used a Mann-Whitney U-test to compare the 
characteristics of nests that were successful with those of 
nests that failed because of predation. Variables used in the 
analysis were those that could have an effect on nest exposure 
or concealment: distance to edge, canopy cover, stand size, 
percent cover above the nest cup, nest height, distance of the 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



89 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



Table 1 — Marbled Murrelet tree nests by state or province, site, year, and outcome. 



State or province 




Nest outcome 




Reason for failure 1 


Predator 2 


Nest site/year found 


Successful 


Failed 


Unknown 




Alaska 












Kelp Bay 1984" 


- 


1 


- 


Abandoned egg 


- 


Naked Island 1991/92 b 




7 


3 


?Predation of egg (n = 1) 
Abandoned egg (n = 3) 
Unknown/egg stage (n ■ 2) 
Unknown/chick stage (n = 1) 


?Steller's Jay 
TCommon Raven 3 


Kodiak 1992" 


- 


- 


2 


— 


— 


Chugach 1992 b 


- 


- 


1 


— 


— 


Afognak 1992 b 


- 


- 


2 


— 


— 


Prince of Wales 1992 c 


- 


- 


1 


— 


— 


SE Alaska 1993 d 


- 


1 


- 


Predation of egg or chick 


— 


British Columbia 












Walbran 1990/91 e 


- 


- 


2 


— 


— 


Carmanah 1992 f 


- 


- 


1 


— 


— 


Walbran 1992 f 


- 


- 


2 


— 


— 


Clayoquot 1993* 


- 


- 


2 


— 


— 


Carmanah 1993* 


- 


- 


1 


— 


— 


Caren 1993 h 


1 


- 


- 


— 


— 


Washington 












Lake 22 199P 


2 


- 


- 


— 


— 


Jimmy come lately 1991' 


- 


- 


1 


— 


— 


Heart of Hills 1991' 


- 


1 


- 


Chick fell out 


— 


Olympic 1991' 


- 


- 


1 


— 


— 


Neman 1993* 


1 


- 


- 


— 


— 


Oregon 












Five Rivers 1990* 


- 




- 


Chick fell out 


— 


Valley of Giants 1990 k 


- 




- 


Predation of chick 


TGreat Horned Owl 


Five Rivers 1991 k 


1 


- 


- 


— 


— 


Valley of Giants 1991 k 


- 




- 


Predation of egg 


?Common Raven 


Cape Creek 1991 k 


- 




- 


Predation of egg 


TCommon Raven 


Siuslaw #1 1991 k 


1 


- 


- 


— 


— 


Siuslaw#2 1991 k 


- 




- 


Predation of chick 


?Steller's Jay 


Boulder Wamike 1992 k 


- 




- 


?Predation of chick 


? 


Valley of Giants 1992 k 


- 




- 


Predation of egg 


?Common Raven 


Copper Iron 1992 k 


1 


- 


- 


— 


— 


Valley of Giants 1993 1 


- 


- 


8 


— 


— 


Green Mountain 1993' 


- 


- 


2 








Five Rivers 1993 1 


- 


_ 


1 








Five Mile Flume 1993' 


- 


- 


1 


i 


— 


California 












"J"Campl974 m 


- 


1 


- 


Chick fell out 


— 


Waddell Creek 1989° 


- 


1 


- 


Predation of chick 


Steller's Jay 


Opal Creek 1989° 


- 


1 


- 


Predation of egg 


Common Raven 


Father 1991/92° 


2 


- 


- 


— 





Palco 1992? 


- 


1 


2 


Chick died 


— 


Prairie Creek State Park 1993' 


- 


1 


- 


Unknown 





Jedediah Smith State Park 1993' 


- 


1 


- 


Unknown 


— 



90 



USD A Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter8 



Nest Success and Effects of Predation 



Table 1 — continued 



1 ?Predation = predation known or suspected based on available evidence. 

2 TPredator = suspected predator, species seen in vicinity of nest 

3 Common Raven flushed adult off one of these nests; this may have had an impact on its abandonment which occurred 2 days later. 
> Quinlan and Hughes, 1990 

b Naslund and others, in press 

c Twelve Mile Arm nest; Brown, pers. comm. 

d Unusual ground level nest located on tree roots above 11 m cliff in Log Jam Creek; Brown, pers. comm. 

■ Manley and Kelson, in press 
f Jordan and Hughes, in press 
* Hughes, pers. comm. 

h P. Jones, pers. comm. 

1 Hamer. unpublished data 

I Ritchie, pers. comm. 

k Nelson, unpublished data; Nelson and Peck, in press 

1 Nelson, unpublished data 

■ Binford and others 1975. 

■ Singer and others 1991. 

Singer and others, in press. 
p Kerns, pers. comm. 



nest from the trunk, limb dia leter at the nest, and nest 
substrate type (i.e., moss or diff). Edges were defined as 
unnatural openings, including, but not limited to, roads and 
clearcuts. Differences in the mean characteristics (ranks) 
were considered significant at P < 0.05. 

Results and Discussion 

Nest Success and Failure 

Nesting success or failure was documented at 49 percent 
(32 of 65) of the nests (table 2). Timing of discovery (after 
the nesting season), limited evidence, or inadequate monitoring 
prevented conclusions about the outcomes at the remainder of 
nests. Therefore we limit our discussion to these 32 tree nests. 

Seventy-two percent (23 of 32) of the nests were 
unsuccessful (tables 1 and 2). Known causes of nest failure 
included predation of eggs and chicks, nest abandonment 
by adults, chicks falling from nests, and nestlings dying 
(tables 1 and 3). Nesting success of 28 percent is lower than 
reported for 17 other alcid species ( x = 57 percent, range = 
33-86) (De Santo and Nelson, this volume), and for 11 
species of sub-canopy and canopy nesting neotropical 
landbird migrants (1=51 percent, range = 20-77) (Martin 
1992). However, some species of seabirds (e.g., Xantus' 
Murrelet [Synthliborampkus hypoleucus]) and forest nesting 
neotropical migrants (e.g.. Western Kingbird [Tyrannus 
verticalis] ), also experienced low nesting success (33 and 
20 percent, respectively) in some years (Martin 1992; Murray 
and others 1983). Hatching and fledging success of Marbled 
Murrelet nests were 67 and 45 percent, respectively. Fledging 
success was also lower than reported for all other alcid 
species ( x = 78 percent, range = 66-100, n = 16) (De Santo 
and Nelson, this volume). 

For all nests, 52 percent of the failures occurred during 
the egg stage, whereas in Washington, Oregon, and California 



most (62 percent) failed during the chick stage (table 3). The 
difference in stage of failure between the southern portion of 
the murrelet' s range and all known nests can be explained by 
greater abandonment of eggs at nests in Alaska (Naslund, 
pers. comm.). The high incidence of abandonment in eggs in 
Alaska between 1991 and 1994 may have been related to 
limited food resources (Kuletz, pers. comm.). 

Failure during the egg stage was caused by abandonment 
and predation. Failure during the chick stage occurred because 
of predation, death from a burst aorta (Palco nest in California), 
and falling from the nest. Chicks may fall from nests because 
nests are located on small platforms, or in response to 
unfavorable weather conditions, such as high winds, or other 
natural and unnatural disturbances. In Oregon, a 6-day-old 
chick may have fallen from its ridgetop nest tree (Five 
Rivers) because of gusty winds that occurred during a midday 
storm. Chicks are also occasionally very active on the nest, 
picking at nesting material, changing positions, snapping at 
insects, exercising their wings, and pacing on the nest limb 
(see Nelson and Hamer, this volume a). They could easily 
fall from the nest platform during these times of activity. In 
addition, predator activity could cause chicks to fall from the 
nesting platform. 

In addition to failure documented at active nests, nestlings, 
fledglings, and eggs have been found on the ground during 
the breeding season at numerous sites throughout North 
America (table 4). Chicks and eggs located on the ground 
probably fell from nests as indicated above. However, eggs 
could also be carried by predators and dropped in locations 
distant from nest sites. 

Fledglings have been discovered on the ground at varying 
distances from the ocean during the breeding season (up to 
101 km inland). Many of these birds still retained an egg 
tooth and small traces of down on their head and back, 
indicating recent fledging. Marbled Murrelet hatch-year birds 



LSDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



91 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



Table 2 — Summary of Marbled Murrelet nest success and failure by state and province 



State/province 


Nest outcome 




Successful 


Failed 


Unknown 


Alaska 

British Columbia 

Washington 

Oregon 

California 



1 
3 
3 
2 


9 


1 
7 
6 


9 
8 
2 
12 
2 


Overall total 


9 
(14 pet.) 


23 
(35 pet.) 


33 
(51 pet.) 


Total for Washington, 
Oregon, and California 


8 
(22 pet.) 


14 

(39 pet.) 


14 
(39 pet.) 



Table 3 — Type and stage of Marbled Murrelet nest failure 



Type of failure 


Number (pet.) 


Stage of failure 




Egg 


Chick 


All nests 








Predation 


10 1 (43) 


5 (56) 


4 (44) 


Unknown 


5 1 (22) 


2 (50) 


2 (50) 


Abandonment 


4 (17) 


4(100) 





Chick fell out 


3 (13) 


- 


3(100) 


Chick died 


1 (4) 


- 


1(100) 


Total 


23 2 (100) 


11 (52) 


10 (48) 


Nests in Washington, 








Oregon, and California 








Predation 


8 (57) 


5 (62) 


3 (38) 


Unknown 


2 1 (14) 





1(100) 


Abandonment 











Chick fell out 


3 (21) 


- 


3(100) 


Chick died 


1 (7) 


- 


1(100) 


Total 


14* (109) 


5 (38) 


8 (62) 



1 One nest failed at unknown stage. 

2 Two nests failed at unknown stage. 



92 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



Table 4 — Marbled Murrelet chicks, eggs, and juveniles found on the ground by 
state and province - an indication of additional nest failure 1 



State/province 


Number 

grounded 

chicks 


Number 
whole 

eggs 


Number 
grounded 
juveniles 


Alaska 


1 


1 


5 


British Columbia 


3 





6 


Washington 

Oregon 

California 


3 
2 
3 


2 
1 



9 

4 
22 


Overall 


12 


4 


46 



1 Data from Atkinson, pers. comm.; Confer, pers. coram.; Carter and Erickson 
1992; Carter and Sealy 1987b; Hamer, unpublished data; Kuletz, pers. comm.; 
Leschner and Cummins 1992b; Mendenhall 1992; Nelson, unpublished data; 
Nelson and others 1992; Rodway and others 1992; S.W. Singer, pers. comm. 



are believed to fly directly from .nland nest sites to the ocean 
after fledging (Nelson and Hamer, this volume a; Quinlan 
and Hughes 1990). Their travel to the ocean may be 
unsuccessful, however, because of navigational problems or 
exhaustion. Unlike other alcids, hatch-year Marbled Murrelets 
must fly relatively long distances to reach the sea without 
the benefit of past flight experience, wing muscle development 
that comes with flight, or adult guidance. The large number 
of juveniles found on the ground while dispersing from nest 
sites raises questions about the relationship between murrelet 
energetics, location of the nest in relation to the ocean, and 
nesting success. Given that some hatch-year birds become 
grounded each year, and may be unable to take flight again, 
nest success may actually be much lower than our estimates 
from nest observations. 

Failure because of predation 

The major cause of nest failure was predation. Forty- 
three percent of all nests and 57 percent of nests in Washington, 
Oregon and California failed as a result of predation (table 
3). Predation rates were higher (56 and 67 percent, respectively) 
when excluding unknown causes of failure, which could 
have included predation. Known predators of murrelet nests 
include Common Ravens (Corvus corax) and Steller's Jays 
(Cyanocitta stelleri) (Naslund 1993; Singer and others 1991) 
(table 1). Predation of a nest by a Great Horned Owl (Bubo 
virginianus) is also suspected. Other potential predators in 
forests include several species of forest owls, accipiters and 
American Crows (Corvus brachyrhynchos). No Marbled 
Murrelet nests are known to have been destroyed by 
mammalian predators, although raccoons (Procyon lotof), 
marten (Martes americana), fisher (Martes pennanti), and 
several species of rodents are potential predators. 

Predation rates on murrelet nests appear higher than 
other alcids, perhaps with the exception of areas with 



introduced or high numbers of predators. For example, 44 
percent of the eggs laid by a population of Xantus' Murrelets 
on Santa Barbara Island in California were taken by deer 
mice (Peromyscus spp.) during periods of egg neglect (Murray 
and others 1983). Rates of predation on murrelet nests also 
appear higher than those observed for many forest birds, 
with the exception of some species of sub-canopy and canopy 
nesting neotropical migrants (e.g., x = 42 percent, range = 
18-67 percent) (Martin 1992). However, the impacts of 
predation on the nesting success of species that lay clutches 
of two or more eggs (e.g., Xantus' Murrelets, Yellow-rumped 
Warbler [Dendroica coronata]) may be less than on species 
that lay only one egg, such as Marbled Murrelets. 

Predation on Marbled Murrelet nests has been observed 
or documented during both the egg and nestling stages, but 
most (56 percent) occurred during the egg stage (table 3). 
Predation during the egg stage is most likely to occur if an 
incubating adult neglects or abandons the nest. Seabirds are 
known to completely abandon their nests during years in 
which prey availability is limited (i.e., during El Nino events) 
(Ainley and Boekelheide 1990, Hodder and Greybill 1985, 
Vermeer and others 1979). In addition, seabirds may neglect 
their eggs for short periods to maximize foraging time and 
accumulate sufficient energy reserves for the lengthy 
incubation shifts (Boersma and Wheelwright 1979, Gaston 
and Powell 1989, Murray and others 1983). During this 
time, the eggs are subject to a variety of negative factors 
including predation, heat loss, and exposure to the elements. 

Murrelets have been observed leaving their eggs 
unattended for short periods of time (2-3 hrs on several 
days) (Naslund 1993; Nelson and Peck, in press), and during 
such a time in Oregon (Cape Creek nest), an egg was taken 
by a predator (most likely a Common Raven). In addition, 
murrelets regularly left their egg unattended in the afternoon, 
evening, and early morning hours during a 5-day period at a 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



93 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



nest in Alaska (Naked Island 1992), and the nest subsequently 
failed (Naslund and others, in press). Eggs were also 
abandoned when adults were flushed from the nest by a 
predator in California (Opal Creek) and Alaska (Naked Island) 
(Naslund 1993; Naslund and others, in press; Singer and 
others 1991). The eggs from these nests were later observed 
or believed to have been destroyed by a Common Raven and 
Steller's Jay, respectively. 

In Oregon, additional egg predation was determined by 
finding blood and albumen on eggshell fragments. The egg 
disappeared from the 1991 Valley of the Giants nest after 
three weeks of incubation. Upon climbing the nest tree, a 
large eggshell fragment with blood stains was found in the 
nest cup. The suspected predator was a Common Raven that 
flew directly adjacent to the nest branch on its daily foraging 
forays. At the 1992 Valley of the Giants nest, eggshell 
fragments with blood and albumen were found at the base of 
a large Douglas-fir (Pseudotsuga menziesii) tree. An empty 
nest cup was subsequently discovered. The predator was 
most likely a Common Raven observed near the nest tree on 
several occasions. 

In Oregon, chicks disappeared or were killed by predators 
at three nests during the 1991 and 1992 breeding seasons. A 
3-week-old chick at the Siuslaw #2 nest was killed when its 
skull was pierced by a predator. Two species of corvids 
(Steller's Jay or Gray Jay [Perisoreus canadensis]) detected 
in the nest tree and adjacent area are the suspected predators. 
At the Boulder Warnicke nest, a 3-week-old chick disappeared 
from the nest. The predator could have been any one of the 
corvids that were present in the area or landed in the nest 
tree: Steller's Jays, Gray Jays, or Common Ravens. A 6-day- 
old chick disappeared at the Valley of Giants 1990 nest 
between 2100 and 0600 hrs on 6 August. A Great Horned 
Owl was heard calling from an adjacent tree (within 10 m) 
during this time period, and is the suspected predator. 

Marbled Murrelets have limited defenses and their 
primary protection against predation at the nest is to avoid 
detection (Nelson and Hamer, this volume a; Nelson and 
Peck, in press). Therefore, the nestling depends on its cryptic 
plumage and the location of the nest for safety. If a predator 
discovers the nest, the chick will attempt to defend itself 
with aggressive behaviors as witnessed by Naslund (1993) 
and Singer and others (1991), when a Steller's Jay attacked a 
4-day-old chick at the Waddell Creek nest in California. The 
chick rotated its sitting position on the nest to constantly 
face the predator, reared up its body and head, opened its 
beak, and jabbed at the predator. The chick was unable to 
ward off the jay and was carried away. 

Nesting attempts also may fail because adults have been 
killed on their way to or at nest sites. In forests of southeast 
and southcentral Alaska, Sharp-shinned Hawks (Accipiter 
striatus) and Northern Goshawks {Accipiter gentilis) are 
known to prey on adult murrelets (Marks and Naslund 1994; 
Naslund, pers. comm.). In addition, Peregrine Falcons (Falco 
peregrinus) and Common Ravens have been observed chasing 
Marbled Murrelets just above and within the forest canopy, 



respectively (Hamer, unpubl. data; Hunter, pers. comm.; 
Suddjian, pers. comm.). A Peregrine Falcon was successful 
in capturing a Marbled Murrelet at one such site in central 
California (Suddjian, pers. comm.). 

Predation of adults at the nest site also can occur. There 
are two known records from California and Alaska. A 
Common Raven flushed an adult murrelet from a nest in 
California (Opal Creek), and was later seen carrying what 
appeared to be a partial carcass (Naslund 1993, Singer and 
others 1991). In Alaska, an adult was killed by a Sharp- 
shinned Hawk seconds after it landed on a suspected nest 
limb (Naked Island) (Marks and Naslund 1994). 

Potential for Bias 

The Marbled Murrelet nests at which predation has 
been studied may not be an unbiased sample. The high 
predation rates recorded at these nests could be biased because 
many of the nests were located in fragmented areas and near 
forest edges (table 5) rather than in the centers of large, 
dense stands. Thus, there is the possibility that nest sites 
located by researchers are also those more easily located by 
predators (see below). At present we lack information to 
evaluate this source of potential bias. 

In addition, it has been suggested that researchers 
studying these nests had an impact on their success (see 
Gotmark 1992; Martin and Geupel 1993). We believe the 
disturbance to the nests was minimal, except at two. In 
southeast Alaska, researchers approached very close to an 
unusual murrelet nest located on tree roots near ground 
level (Brown, pers. comm.). The adult was flushed or 
disturbed on five occasions, which may have contributed to 
its failure (egg or newly hatched chick disappeared). The 
"J" Camp nest in California also failed from direct human 
intervention (Binford and others 1975). No human impacts 
are suspected at nests where the chick fell out (n = 1 in 
Oregon) or died (n = 1 in California), or where nests were 
found after they had failed (n = 1 each in Washington and 
Oregon, n = 2 in California). At all other nests, human 
impacts were also limited because: (1) some nests were 
monitored infrequently (n = 8 in Alaska and n = 2 in Oregon); 
(2) predators knew the location of the nest on day of and 
probably prior to discovery, and, additionally, precautions 
(e.g., limiting noises and number of observers near nest; see 
Martin and Geupel 1993) were implemented to minimize 
disturbance and predator attraction (n = 1 in Oregon, n = 2 in 
California); and (3) nests were monitored from >25 m 
horizontal distance from the nest and precautions (see above) 
were implemented (n = 17). For (2) and (3) above, predators 
were occasionally attracted to the observer's location on the 
ground (especially Steller's Jays), but not to the nest site, 
>18 m above the ground. In contrast, intensive disturbance 
occurred at three successful nests. In Oregon, the only nest 
tree that was climbed while active was successful, and in 
Washington, chicks at two nests fledged despite regular 
climbing (approximately once a day for 9-20 days) to collect 
nestling growth and development data. 



94 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predabon 



Table 5 — Characteristics of successful Marbled Murrelet tree nests compared with those that failed because of predarion 



State/province 




Distance 


Canopy 


Stand 


Nest 


Nest 


Limb 


Distance 




Siie/ye«r 


Outcome 1 


to edge 2 


cover 


size 


concealment 


height 


diameter 


from 


Substrate 






(m) 


.pet i 


(ha) 


(pet) 


(m) 


(cm) 


trunk (cm) 




British Columbia 




















Carenl993* 


1 


700 


70 


800 


100 


18.0 


20.0 





moss 


Washington 




















Lake221991 b 


1 


55 


ss 


405 


70 


31.4 


10.7 


45.6 


moss 


Lake 22 1991 b 


1 


65 


74 


405 


95 


27.7 


363 


57.0 


duff 


Nemahl993 c 


1 


10 


65 


142 


80 


. .3 






moss 


Oregon 




















Valley of Giants 199^ 





20 


44 


149 


70 


56.0 


343 


33.0 


moss 


five Rivers 1991* 


1 


75 


49 


46 


80 


503 


38.0 


116.2 


moss 


Valley of Giants 1991* 





28 


50 


149 


80 


503 


41.0 


17.1 


duff 


Cape Creek 1991 d 





16 


65 


138 


95 


44.2 


10.0 


762.0 


moss 


Siuslaw*! 1991 d 


1 


56 


60 


89 


85 


60.3 


233 


152.0 


moss 


Siuslaw#2 1991 d 





64 


52 


47 


80 


513 


13.0 


230.0 


duff 


Boulder Wamicke 1992 d 





32 


19 


3 


80 


61.0 


21.6 


46.0 


moss 


Valley of Giants 1992" 1 





15 


66 


149 


70 


52.0 


47.0 


35.0 


moss 


Copper Iron 1992 d 


1 


300 


93 


542 


75 


49.0 


34.0 


1.0 


moss 


California 




















Waddell Creek 1989 s 





10 


40 


1700 


25 


383 


363 


61.0 


moss 


Opal Creek 1989* 





34 


40 


1700 


40 


43.7 


47.7 


122.0 


moss 


Father 1991 f 


1 


69 


40 


1700 


100 


41.1 


61.0 





duff 


Father 1992 f 


1 


69 


40 


1700 


100 


53.2 


42.0 





duff 



1 1 = successful. = failed. 

2 Edge = Distance to nearest unnatural edge (road or cleared). 
' Data not available. 

* P. Jones, pers. comm. 

b Hamer. unpublished data. 

c Ritchie, pers. comm. 

' Nelson, unpublished data; Nelson and Peck, in press. 

e Singer and others 1991. 

1 Singer and others, in press. 



Habitat Characteristics and Predation of Nests 

The effect of predators on avian nesting success can 
vary significantly with geographic location, and is dependent 
upon the species of predators present accessibility of nests, 
type and dimension of the habitat, topography, and vegetative 
complexity (vertical and horizontal diversity) (Chasko and 
Gates 1982: Martin 1992; Marzluff and Balda 1992; Paton 
1994: Reese and Ratti 1988; Yahner 1988; Yahner and others 
1989). For example, alcids nesting on islands relatively free 
of mammalian predators, or on cliffs inaccessible to terrestrial 
predators, experience lower predation rates than species 
nesting in accessible sites and with abundant predators (Ainley 
and Boekelheide 1990; Hudson 1985). Because many species 
of birds have evolved in association with predators, the long 
term impacts of predation on avian nesting success are 
expected to be minimal in natural situations. However, rapid 



and unnatural changes, such as the introduction of mammalian 
predators (cats, goats, mice, pigs, raccoons, rats) and habitat 
modification, can have significant impacts on nesting success 
of seabirds (Bailey and Kaiser 1993; Ewins and others 1993; 
Gaston 1992; Murray and others 1983), and neotropical 
migrants (Chasko and Gates 1982; Martin 1992), respectively. 
In these cases, predation can be a major factor affecting 
avian population viability (Martin 1992). 

Significant changes have occurred in the forested 
landscapes of the United States over the past century, including 
loss of late-successional forests, habitat fragmentation, and 
increases in the amount of edge (Hansen and others 1991; 
Harris 1984; Morrison 1988; Perry, this volume; Thomas 
and Raphael 1993). These changes have affected the 
abundance and distribution of many avian predators and 
forest nesting birds. For example, populations of corvids and 



USD* Forest Service Gen. Tech. Rep. PSW-152. 1995. 



95 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



Great Homed Owls are increasing in numbers throughout 
the western United States, especially in response to increases 
in habitat fragmentation and human disturbance (Johnson 
1993; Marzluff 1994; Marzluff and Balda 1992; Robbins 
and others 1986; Rosenberg and Raphael 1986; Yahner and 
Scott 1988). In contrast, numerous neotropical migrant species 
are declining in numbers because they are unable to adjust to 
fragmentation and rapidly changing habitat conditions (Hagan 
and Johnson 1992; Hansen and others 1991, Hejl 1992, 
Martin 1992, Morrison and others 1992, Rosenberg and 
Raphael 1986). The Marbled Murrelet was listed as a 
threatened species in 1992 as changes in the forested landscape 
appear to be negatively impacting their populations (U.S. 
Fish and Wildlife Service 1992). 

Although the relationship between predator numbers, 
habitat fragmentation, and predation on Marbled Murrelet 
nests has not been specifically studied, we believe, based on 
the following data, that changes in their habitat, such as 
increased amounts of edge, may significantly affect their 
nesting success. First, evidence from murrelet nests in this 
study suggests that distance to edge, stand size, canopy 
closure, percent cover above the nest cup (nest concealment), 
and distance of the nest from the tree trunk may be affecting 
predation rates (table 5). In a comparison of these habitat 
characteristics between successful nests (n = 9) and nests 
that failed because of predation (n = 8, excluding Alaska), 
we determined that successful nests were located significantly 
farther from edges (U = 2.9, n = 16 trees, P < 0.05) (table 5). 
All successful nests were located >55 m (x = 166.3, n = 8 
trees, s.e. = 82.3) from an edge (road or clearcut), with the 
exception of the Nemah nest in Washington, which was 
located within 10 m of an old road near the center of a 142 ha 
forest. In contrast, all nests that failed because of predation 
were located within 64 m ( 3c = 27.4, s.e. = 6.0) of an edge. In 
a review of numerous artificial nest predation studies, Paton 
(1994) found evidence that predation of bird nests is higher 
within 50 m of edges. This result supports our hypothesis 
that murrelet nests near edges may be more vulnerable to 
predation than those located in the stand interior. In addition, 
nest concealment was significantly greater at successful nests 
( x = 87.2 percent, s.e. = 3.9) compared with nests that failed 
because of predation (x = 67.5 percent, s.e. = 8.2) (U = 2.3, 
n = 17, P < 0.05) (table 5). Nest concealment has been 
shown to decrease predation rates (Chasko and Gates 1982; 
Marzluff and Balda 1992; Martin and Roper 1988). Stand 
size (532.0 versus 407.4 ha, n = 11 stands) and canopy 
closure near nests (63.6 versus 47.0 percent, n = 16 plots) 
were higher and nests located closer to the trunk (46.5 versus 
163.3 cm) at successful sites, but were not significantly 
different from nests that failed because of predation. 

Second, it has been suggested that changes in forests 
where boundaries are contiguous with secondary succession 
may not create the same predation problems as those observed 
in static, simple forests in urban and agricultural areas that 
are defined by distinct boundaries (Rosenberg and Raphael 
1986; Rudnicky and Hunter 1993). However, numerous 



studies in the eastern United States provide empirical evidence 
that edge effects in a forest dominated landscape (forest/ 
clearcut edge) are similar to those in forest/urban or 
agricultural settings. For example, in studies of eastern 
neotropical migrants, predation was lower in the forest interior 
(>50 m from the edge) compared with edge habitat (Chasko 
and Gates 1982; Yahner and Scott 1988). Predation was also 
lower in areas with high vegetative heterogeneity and 
concealing cover (Chasko and Gates 1982). 

Evidence from artificial nest studies in forests of the 
Pacific Northwest also suggests that predation of birds' nests 
may be affected by habitat fragmentation and forest 
management. On Vancouver Island, British Columbia, Bryant 
(1994) demonstrated that artificial ground and shrub nests 
located along forest/clearcut edges (within 100 m) were subject 
to higher predation rates than those in the forest interior 
(100-550 m from the edge). In the Oregon Coast Range, 
predation on artificial shrub nests was higher in clearcuts and 
shelterwood (20-30 green tress >53 cm d.b.h./ha) stands than 
in stands with group selection cuts (1/3 volume removed in 
0.2 ha openings) and unmanaged (control) stands (Chambers, 
pers. comm.). Additionally, in the Oregon Cascades, Vega 
(1994) found that predation on ground nests was significantly 
greater in clearcuts compared with retention stands (12 trees/ 
ha and 7.5 snags/ha), and predation on shrub nests was highest 
in retention stands compared to the other treatment types 
(clearcuts and mature stands). Steller's Jays, the suspected 
predator of the shrub nests, were more abundant in the retention 
stands, where they probably used the remnant trees for perching 
(see Wilcove 1985; Yahner and Wright 1985). 

Third, despite differences in results among nest predation 
studies (e.g., Rudnicky and Hunter 1993 versus Yahner and 
Scott 1988), existing evidence strongly indicates that avian 
nesting success declines near edges (Paton 1994). In addition, 
regardless of the type of edge, fragmentation of forests often 
reduces structural complexity and heterogeneity of stands, 
and exposes remnant patches to edge effects (Hansen and 
others 1991; Harris 1984; Lehmkuhl and Ruggerio 1991). 
Because of increases in the amount of edge, productivity of 
interior forest species is generally impacted (Lehmkuhl and 
Ruggerio 1991; Reese and Ratti 1988; Yahner and others 
1989), and generalist species, which benefit from the ecotone, 
usually increase in numbers (Yahner and Scott 1988). In 
addition, as vegetative complexity and canopy volume are 
reduced through fragmentation, bird nests (especially those 
located in shrubs or trees) may be more conspicuous and 
easier for avian predators to locate (Rudnicky and Hunter 
1993; Vega 1994; Wilcove 1985; Yahner and Cypher 1987; 
Yahner and others 1989; Yahner and Scott 1988). 

The rates of predation on Marbled Murrelet nests in this 
study appear higher than for many seabirds and forest birds. 
If the observed predation rates are representative of predation 
rates throughout the murrelet' s range, then the impacts of 
predation on murrelet nesting success is significant and of 
concern (Wilcove 1985). Even if these high predation rates 
are localized to certain states or areas within states, the 



96 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Nelson and Hamer 



Chapter 8 



Nest Success and Effects of Predation 



combination of low annual nesting success, low fecundity 
rates (Beissinger, this volume), and low or declining 
population sizes (Carter and Erickson 1992; Kelson and 
others, in press; Kuletz, 1994), could impact the survival and 
recovery of this threatened seabird. 

Conclusions 

Results from this study suggest that predation on murrelet 
nests may be relatively high compared with many alcids and 
forest nesting birds. Because Marbled Murrelets have no 
protection at nest sites other than the ability to remain hidden 
(Nelson and Hamer, this volume a), the availability of safe 
nest sites will be imperative to their survival. If logging and 
development (e.g., clearing land, creating patches of habitat, 
thinning stands) within the murrelet' s range has resulted in 
increased numbers of predators or predation rates, and has 
made murrelet nests easier to locate because of increased 
amounts of edge and limited numbers of platforms with 
adequate hiding cover, then predation on murrelet nests 
could be significantly higher in such situations. In addition, 
areas heavily used by humans fci recreational activities (i.e., 
picnic and camping grounds) c&'i attract corvids (Marzluff 
and Balda 1992, Singer and others 1991) and may increase 
the chance of nest predation within these areas. Therefore, 
we hypothesize that because this seabird has low reproductive 
rates (one egg clutch), small increases in predation will have 
deleterious effects on murrelet population viability. 



Rigorous studies should be developed to investigate the 
effects of predator numbers, predator species, predator 
foraging success, landscape patterns, habitat types, and forest 
structural characteristics on Marbled Murrelet nesting success. 
In implementation of these studies, hypotheses on the effects 
of various habitat features on fitness components (recruitment 
and demography) should be tested (Martin 1992, Paton 1994). 
At the same time, the effects of these hypotheses on coexisting 
species and the interacting effects these species have on one 
another should be evaluated (Martin 1992). 

Acknowledgments 

We are grateful to the biologists who kindly shared their 
data with us; special thanks go to Jim Atkinson, Alan Burger, 
Stephanie Hughes, John Hunter, Paul Jones, Kevin Jordan, 
John Kelson, Steve Kerns, Kathy Kuletz, Irene Manley, Ray 
Miller, Nancy Naslund, Bill Ritchie, Steve and Stephanie 
Singer, and David Suddjian for their time and generosity. 
We also thank Alan Burger, George Hunt, John Marzluff, 
Robert Peck, Steve Speich, and several anonymous reviewers 
for providing valuable comments on earlier drafts of this 
manuscript. Support for preparation of this manuscript was 
provided by the Oregon Department of Fish and Wildlife, 
USDA Forest Service, USDI Bureau of Land Management, 
and the U.S. Fish and Wildlife Service, U.S. Department of 
the Interior. This is Oregon State University Agricultural 
Experiment Station Technical Paper 10,540. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



97 



Chapter 9 

Molts and Plumages in the Annual Cycle of the 
Marbled Murrelet 



Harry R. Carter 1 



Janet L. Stein 2 



Abstract: Marbled Murrelets have distinct basic, alternate and 
juvenal plumages. In after-hatching-year (adult) birds, the incom- 
plete pre-altemate molt occurs rapidly over a period of about one 
month per individual between late February and mid-May. The 
complete pre-basic molt occurs between mid-July and December. 
At this time, individuals are flightless for about two months. In late 
summer, it is difficult to distinguish adult birds undergoing pre- 
basic molt from juveniles at sea. Field methods for separating 
these age categories at sea at this time of the year are presented. By 
early fall, older juveniles are not distinguishable in the field from 
after-hatching-year birds in basic plumage. The timing of pre- 
basic and pre-altemate molts were closely related to the timing of 
breeding, movements and other aspects of the annual cycle of 
Marbled Murrelets in Barklev Sound, British Columbia. 



Little has been published on the plumages and molts of 
the Marbled Murrelet (Brachyramphus marmoratus). 
Although the general pattern of molt and plumages has been 
documented, many details that are important for interpreting 
aspects of the biology of this enigmatic species have remained 
undescribed. Adults, also referred to as after-hatching-year 
birds (i.e.. breeding adults and subadults. including first- 
year birds in their second calendar year), have distinct alternate 
versus basic plumages that they wear during summer and 
winter periods, respectively. Subadults have not attained full 
maturity and have not yet bred. The mottled-brown alternate 
plumage is certainly responsible for the English name 
"Marbled" Murrelet. In addition, juveniles less than 6 months 
old, also known as hatching-year birds, wear a distinct juvenal 
plumage during late summer. Murrelets replace their alternate 
plumage with a basic plumage during a complete pre-basic 
molt (involving flight and body feathers) in the late summer 
and fall. Similarly, during an incomplete pre-altemate molt 
(involving only body feathers), they replace their basic 
plumage w ith the alternate plumage in spring. These general 
plumage stages and molts are similar for many other seabirds 
and birds in general (Welty and Baptista 1988). The juvenal, 
alternate, and basic plumages of the Marbled Murrelet are 
illustrated in many reputable bird identification field guides 
(e.g.. Harrison 1983. National Geographic Society 1983). 

Many past studies of Marbled Murrelets have not required 
a detailed knowledge of the stages of molts and plumages. 
Workers quantifying distribution and abundance of murrelets 



1 Wildlife Biologist. National Biological Service. U.S. Department of 
Interior. California Pacific Science Center, 6924 Tremont Rd. Dixon, CA 
95620 

2 Wildlife Biologist. Washington Department of Fish and Wildlife, 
16018 Mill Creek Blvd., Mill Creek. WA 98012 



at sea have usually lumped all murrelets together regardless 
of plumage, or they conducted their studies in summer or 
winter when most or all birds were in the same plumage. 
Plumages of birds observed at inland nesting areas have not 
been distinguished because individuals fly high overhead 
under low light conditions or darkness during censuses. 
Interest in the relationship of plumage and molt to other 
aspects of the murrelet 's life history has grown rapidly since 
1992. Researchers in Alaska, British Columbia, Washington, 
Oregon, and California have recently attempted to census 
juveniles at sea in the late summer and early fall to indirectly 
determine breeding success. These efforts were prompted by 
concerns that the very low numbers of juveniles compared to 
adults (1-5 percent) observed during recent surveys in Oregon 
and California represent very low breeding success (Nelson, 
pers. comm.; Hardin, pers. comm.; Ralph and others, this 
volume; Strong and others 1993). Such low levels of breeding 
success could indicate that murrelet populations in 
Washington, Oregon and California can no longer maintain 
themselves. However, surveys at this time of the year have 
difficulties that can lead to undercounting or overcounting 
juveniles in relation to adult birds from the same breeding 
population, including: (1) the degree that researchers can 
accurately separate the plumages of juveniles and adult birds 
in the field, even under adequate viewing conditions; (2) 
possible post-breeding season movements of adults, juveniles, 
or both into or out of the area studied; (3) differential use of 
at-sea habitats by various age classes and during different 
stages in the annual cycle: and (4) the timing and degree of 
natural and anthropogenic mortality of juveniles and adult 
birds. Thus, the adult:juvenile ratio is complex and must be 
interpreted with caution. 

To address these difficulties, especially the first, we 
reviewed available information on plumages and molt from 
published and unpublished sources with three main objectives 
in mind. First, we summarized information on plumages and 
molt and identified gaps. Second, we summarized some 
other aspects of murrelet biology during the molt period that 
may be important for assessing the adult:juvenile ratio. Third, 
we developed field criteria for separating juveniles from 
adult birds at sea during the late summer and early fall. This 
method, based on current knowledge, will require modification 
as new results are obtained. Our goal has been to provide 
workers with sufficient information to gather more data to 
confirm and expand on known patterns. This summary is not 
complete and we refer the reader to other chapters in this 
volume for additional information on murrelet biology during 
the breeding and non-breeding seasons. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



99 



Carter and Stein 



Chapter 9 



Molts and Plumages 



Methods 

We relied heavily on studies involving collected birds 
that allowed a close examination of plumages and molt 
condition. Sealy (1972; 1974; 1975a,b) studied breeding 
phenology, diet and body condition of murrelets at Langara 
Island, British Columbia, March-July 1970-1971. Carter 
(Carter 1984, Carter and Sealy 1990, Rodway and others 
1992) studied at-sea distribution and foraging behavior of 
murrelets, as well as breeding phenology, diet and body 
condition, in Barkley Sound, British Columbia, May-October 
and December 1979-1980. Carter (unpubl. data) collected a 
complete series of birds undergoing pre-basic molt, as well 
as some juveniles, from July to October. These birds were 
preserved as study skins by Sealy and are housed at the 
University of Manitoba Zoology Museum, Winnipeg, 
Manitoba. In addition, Carter (unpubl. data) observed Marbled 
Murrelets off Victoria, British Columbia, during November- 
March 1978- 1980 (see Gaston and others 1993). These studies 
were collated to present a general picture of murrelet plumages 
and molts throughout the year for southern Vancouver Island, 
British Columbia. 

To confirm plumage and molt patterns identified from 
other studies, we examined a total of 106 specimens from 
the late summer and fall periods in the Royal British Columbia 
Museum (Victoria, British Columbia) and in the California 
Academy of Sciences (San Francisco, California). We 
examined total length, the ratio of dark: light coloration, 
ventral coloration and patterning, dorsal coloration, and 



primary wing molt. Total length was measured from 46 adult 
and 30 juvenile (including recently-fledged and older juvenal 
plumages) specimens that had been collected during June 
through September. The ratio of dark:light coloration was 
determined by placing a grid marked with 0.5 inch x 0.5 inch 
quadrats over the dorsal, left and right sides of museum 
specimens and tallying the number of quadrats filled with 
mainly dark or mainly light plumage. Only the dorsal surface 
and sides of the specimens were examined in order to 
determine the dark:light ratio for the area of the bird most 
often seen when they are sitting on the water. Notes on the 
ventral coloration and patterning and dorsal coloration were 
also recorded for 67 adult and 35 juvenile specimens. 

Plumages 

Basic and Alternate Plumages 

Kozlova (1957) provides good general descriptions of 
the basic and alternate plumages of the Marbled Murrelet. 
The following is a summary of Kozlova (1957) with a few 
added comments. In basic plumage, adults are dark brownish 
above, with bluish grey margins to the back feathers and 
largely white scapulars. The sides of the head and band 
around the neck, extending almost to the nape, are white. 
The underparts are white with some brown feathers still 
sprinkled on the flanks (figs. 1 and 2). In alternate plumage, 
the upper body parts are brownish black with rusty-buff 
margins to the back feathers. The sides of the head, front and 




Figure 1— Plumage similarities during fall between older juvenile Marbled Murrelets (top) and adult birds (bottom). Collection dates of juveniles: 5 
October 1907 (left), 8 November 1907 (right). Collection dates of adult birds: 23 September 1895 (left), 16 November 1895 (right). Specimens are 
housed at the California Academy of Sciences, San Francisco, California. Photo taken by H.R. Carter. 



100 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Carter and Stein 



Chapter 9 



Molts and Plumages 




Figure 2 — Pre-alternate molt sequence in Marbled Murrelets. Museum specimens are ordered to reflect changes in ventral plumage during pre- 
alternate molt. The far left specimen is in basic plumage and the far right specimen is in alternate plumage. Collection dates of specimens from left 
to right: 22 February 1900, 27 February 1 907, 18 February 1907, 30 March 1907, 18 February 1907, 20 February 1896, 27 February 1 907, 26 March 
1907. Specimens are housed at the California Academy of Sciences, San Francisco, California. Photo taken by H.R. Carter. 



sides of the neck, and underparts are covered with white 
feathers that are edged with broad dark-brown margins (fig. 
2). These dark margins take up about half of each feather. 
The flanks are almost entirely dark brown, the upper wing 
coverts are dark brown with occasional narrow white edges, 
and the under wing coverts and axillaries are brownish grey. 
The rectrices are brownish black, occasionally with narrow 
white margins and brownish dots on the outer rectrices. 
There are no known differences in plumage appearance 
between sexes or ages of adult birds. However, in some 
European alcids, first-year birds may retain certain upperwing 
coverts, leading to a visible contrast between older, retained 
feathers against newer, replaced feathers (Pyle, pers. comm.). 
Such detailed examinations are required for the Marbled 
Murrelet to unveil such possible distinctions when examining 
birds in the hand. 

Murrelets in basic plumage closely resemble the plumage 
of several other alcids, being "dark above" and "light below." 
The basic plumage is often considered closer to an older, 
ancestral plumage. The evolution of the cryptic alternate 
plumage is an obvious adaptation for nesting solitarily in 
old-growth forests (Binford and others 1975). It is likely 
that the Marbled Murrelet originally evolved its cryptic 



plumage by using similar nesting habitats as the closely- 
related Kittlitz's Murrelet (B. brevirostris). The latter species 
also attains a very cryptic alternate plumage for nesting 
solitarily on mountain scree slopes in Alaska and Russia up 
to 100 km inland from the ocean (Day and others 1983). 
However, the alternate plumage of the Marbled Murrelet is 
darker overall and, unlike the Kittlitz's Murrelet, the rust- 
tipped back feathers of Marbled Murrelets closely match 
the bark of typical nest trees. While about 3 percent of the 
Alaskan population of Marbled Murrelets nests solitarily on 
the ground (Day and others 1983, Mendenhall 1992, Piatt 
and Ford 1993), it is unclear whether they represent remnant, 
ancestral ground-nesting behavior or a more recent 
redevelopment of such behavior. In any case, the cryptic 
alternate plumage was one preadaption that may have allowed 
Marbled Murrelets to originally colonize and nest in old- 
growth forests. 

Distinctions between the plumages and other char- 
acteristics of the American Marbled Murrelet (B. m. 
marmoratus) and the Asian Marbled Murrelet (B. m. perdix) 
can be found in several papers (Erickson and others 1994; 
Kozlova 1957; Sealy and others 1982, 1991; Sibley 1993). 
Recent evidence indicates that the Asian Marbled Murrelet 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



101 



Carter and Stein 



Chapter 9 



Molts and Plumages 



should be considered to be a separate species (Friesen and 
others 1994a, Piatt and others 1994). 

Nestling Plumage 

Binford and others (1975) described the downy chick in 
detail. Newly-hatched chicks are covered by a thick layer of 
natal down. Generally, the yellowish down is interspersed 
with irregular dark spots that cover the upper parts and are 
more prevalent on the head. A paler grey down covers the 
belly (Simons 1980). The down covers the developing ju venal 
plumage and is retained for a relatively long period of time, 
until just prior to fledging. At this time, the down appears to 
be preened or scratched off and may be ingested by the chick 
(Simons 1980). At fledging, juvenile birds fly to the ocean 
(Carter and Sealy 1987b, Hamer and Cummins 1991). Most 
juveniles arrive at sea in juvenal plumage, although some 
individuals may still retain some down, especially on the 
neck and crown (Sealy 1975a). 

The cryptic downy nestling plumage of the Marbled 
Murrelet is also an obvious adaptation for nesting in old- 
growth forests (Binford and others 1975) or on mountainous 
scree slopes. Chicks of precocial alcids have more dense 
down coverings and resemble the adult plumage in pattern 
and coloration. Other semi-precocial alcid nestlings (like the 
Marbled Murrelet) have unmarked grey down. The late 
retention of this downy nestling plumage, in association 
with nest placement, tree bark or rock color, adult activities, 
and chick behavior, is probably important for reducing 
predation at the nest site. 

Juvenal Plumage 

Recently-fledged juveniles are uniformly dark brownish 
above with white scapulars. The underparts and sides of the 
head are white and speckled with blackish brown which 
does not fully conceal the white ground color of the feathers 
(fig. 3). The under wing coverts are brownish grey with 
some white. White bars are present on the outer rectrices and 
the inner vanes are pale brownish. Recently-fledged juveniles 
also retain the egg tooth for some time after fledging (Sealy 
1970), although it is almost impossible to see the egg tooth 
in the field. The late retention of the egg tooth is probably 
related to the late retention of nestling down, early fledging 
(i.e. when less than fully grown), or both. 

The juvenal plumage of recently-fledged juveniles differs 
from older juveniles that have been at sea for a longer period 
of time. Recently-fledged juveniles appear darker overall 
with most feathers on the sides of the head, neck, breast and 
abdomen edged with thin dark margins (fig. 3). This pattern 
gives juveniles a "speckled" appearance, especially on the 
breast and upper abdomen. Thicker dark margins occur on 
the side and flank feathers (similar to adults). Recently- 
fledged juveniles often exhibit a neckband formed by a 
greater density of feathers with dark margins in the upper 
breast region. The plumage of recently-fledged juveniles is 
often referred to as the "juvenal plumage" in such field 
identification guides as the National Geographic Society 



guide (1983). Older juveniles appear to become whiter and 
lose any neck band and most or all of the dark margins that 
characterize typical juvenal plumage (fig. 1). This transition 
may occur as early as a few weeks after fledging. In addition, 
the uniform dark brown to black feathers on the upperparts 
of recently-fledged juveniles are replaced with feathers edged 
with thick grey margins in older juveniles (similar to adult 
birds). It is unclear how these plumage changes occur during 
this period (see below). Once older juveniles have completed 
this plumage transition, they are impossible to separate from 
adult birds in full basic plumage in the field (fig. 1). However, 
in the hand, remnant speckling of the juvenal plumage can 
be seen on the ventral parts of some birds as late as February. 
One hypothesis that explains the plumage transition 
between recently-fledged and older juveniles is that murrelets 
have not achieved their full juvenal plumage at fledging. 
Chicks fledge at 70 percent adult weight (Sealy 1975a) and 
grow to attain full adult size at sea. For instance, recently- 
fledged juveniles often still have sheathed outer primaries. 
The dark margins on recently-fledged juveniles may represent 
a stage of feather growth between the shedding of natal 
down and the full attainment of juvenal plumage when full 
adult size is reached. The dark-margined ventral feathers 
and/or the grey back feathers may be replaced near the end 
of the "nestling" growth period that occurs at sea. 
Alternatively, the thin and fragile dark margins of the ventral 
feathers may wear off quickly when exposed to salt water 
and swimming and diving activities. A second explanation 
for the plumage transition between recently-fledged and 
older juveniles is that a separate partial body molt occurs, 
causing loss and replacement of dark-margined ventral feathers 
and dark back feathers with completely white and grey- 
margined feathers, respectively. Kozlova (1957) stated that 
the juvenal plumage is exchanged for the first winter plumage 
in the fall. She did not provide the basis for this statement, 
and it is unclear if actively molting feather tracts were observed 
on specimens examined. If such a molt did occur, it would 
probably occur some time well after fledging. We cannot 
currently determine which mechanism best explains this 
transition because the actual fledging dates of specimens 
examined is not known and could vary by several months 
due to protracted breeding. Some form of feather replacement 
could be supported by finding actively molting feather tracts 
on juveniles collected in late summer and early fall. 

Annual Cycle of Molts and Plumages 

Pre-alternate and pre-basic molts are controlled by levels 
of sex and other hormones, which change throughout the 
year. The pre-alternate molt precedes breeding and is 
associated with egg-laying and/or associated nesting behaviors. 
However, the onset and progression of molt probably also is 
modified by several environmental factors. Molt imposes 
high energetic demands within the annual cycle of the Marbled 
Murrelet. In particular, the replacement of flight and body 
feathers during the pre-basic molt requires significant changes 



102 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Carter and Stein 



Chapter9 



Molts and Plumages 




Figure 3 — Ventral plumage differences between adult Marbled Murrelets undergoing pre- 
basic molt (top) and recently-fledged juveniles (bottom). Note the "blotchy" appearance of 
adult birds versus the "speckled" appearance of juveniles. Collection dates of adult 
specimens: 1 July 1 965, 1 1 August 1 964, 1 1 September 1 969. Collection dates of juvenile 
specimens: 1 8 July 1 920, 8 August 1 961 , 24 September 1 924. Specimens are housed at the 
Royal British Columbia Museum, Victoria, British Columbia. Photo taken by J.L Stein. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



103 



Carter and Stein 



Chapter 9 



Molts and Plumages 



in behavior and biology. Flightless murrelets must select 
molting areas which provide adequate prey resources within 
swimming distance for about two months. Clearly, it is 
impossible for Marbled Murrelets to overlap breeding with 
the flightless pre-basic molt because they would be unable to 
fly to nests. In contrast, the gradually molting auklets retain 
flight during the pre-basic molt and do overlap pre-basic 
molt with breeding (Bedard and Sealy 1984, Emslie and 
others 1990, Payne 1965). 

It is likely that the timing of molt varies between years 
and between different parts of the breeding range, in concert 
with variation in the timing of breeding and variation in 
local prey resources (Ewins 1988, Emslie and others 1990). 
It is clear that the hormonal integration of molt, breeding 
and other aspects of the annual cycle of the Marbled Murrelet 
is complex and our understanding of these processes is 
limited. In southern parts of the breeding range in North 
America where murrelets are largely resident, visitation of 
nesting areas does not occur during the flightless pre-basic 
molt, does occur during the winter period (when birds are in 
basic plumage), is reduced during pre-alternate molt (prior 
to egglaying), and then occurs throughout the breeding season 



by birds in alternate plumage (Carter and Sealy 1 986, Naslund 
1993b). Some birds that nest farther north in parts of British 
Columbia and Alaska appear to winter in different areas or 
habitats than where they breed. While a portion of the 
population may visit nesting areas for most of the year, a 
significant portion or the majority may visit nesting areas 
only during the breeding season (Rodway and others 1992). 
Such major differences in the annual cycles of differing 
populations undoubtedly results in complex patterns of molts 
and plumages in different geographic areas. 

Timing of Breeding and Pre-Basic Molt 

In Barkley Sound, British Columbia, Carter (1984) found 
that the asynchronous or protracted timing of breeding within 
this population of Marbled Murrelets appeared to lead to a 
protracted pre-basic molt period (fig. 4). Breeding occurred 
mainly from early April to the end of July, although it extended 
as late as mid-September. The first fledglings were observed 
on 4 July 1979 and 28 June 1980 and the last fledgling (a 
recently-fledged juvenile with an egg tooth) was collected on 
5 October 1980. The last bird in alternate plumage was 
observed flying and carrying a fish on 17 September 1980. 




M 



M 



Figure 4 — Annual cycle of molts, plumages, breeding phenology and attendance of at-sea feeding areas 
for Marbled Murrelets in southern Vancouver Island, British Columbia, in 1 979-1 980 (Carter 1 984; Carter, 
unpubl. data; Sealy 1975b). Codes are: AHY (attendance by adult, after-hatching-year birds); IP 
(incubation period); NP (nestling period); HY (attendance by juvenile, hatching-year birds); BP (basic 
plumage); PAM (pre-alternate molt); AP (alternate plumage); PBM (pre-basic molt); WM (wing primary 
molt); RFJ (recently-fledged juvenile); and OJ (older juvenile). Thick portions of ranges indicate timing for 
a large proportion of the population. Thin lines indicate usual range. Dots indicate extremes. 



104 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Carter and Stein 



Chapter 9 



Molts and Plumages 



Most recently-fledged juveniles occurred at sea in July and 
August in Barkley Sound (Carter 1984, Guiguet 1971), 
although recently-fledged and older juveniles occurred there 
into early October when observations ceased (fig. 4). 

To project possible timing of molt for other populations 
in relation to Barkley Sound, we have summarized the earliest 
and latest possible fledging dates for Marbled Murrelets in 
different areas from British Columbia to California. Less is 
known about the average and latest fledging dates (but see 
Hamer and Nelson, this volume a). At Langara Island, British 
Columbia, Sealy (1974, 1975a) reported the first young on 
the water on 6 and 7 July in 1970 and 1971, respectively. In 
all of British Columbia, juveniles have been observed at sea 
between 28 May and 5 October (Rodway and others 1992). In 
Washington, the earliest known nest fledging date is 22 June 
1993 (Ritchie, pers. comm.). A juvenile collected on 3 August 
1950 in the San Juan Islands, Washington, still had an egg 
tooth (Leschner and Cummins 1992a). In Oregon, juveniles 
have been observed at sea as early as 15 June (Hardin, pers. 
comm; Nelson, pers. comm.; Strcr.g and others 1993). Inland 
records of fledglings in California occur from 12 June to late 
September whereas recently-fledged juveniles have been found 
at sea as early as 1 June (Carter and Erickson 1988, 1992; 
Carter and Sealy 1987b). In general, nesting appears to occur 
slightly earlier, but over the same general period from late 
April to September, in the southern part of its range. Thus, the 
timing of molt would not be expected to vary much throughout 
this area in relation to the timing observed at Barkley Sound, 
British Columbia, in 1979-1980 {fig. 4). 

In Barkley Sound, British Columbia, pre-basic molt 
extended over a long period from mid-July to at least late 
November {fig. 4). The first bird undergoing pre-basic wing 
molt was collected on 24 July 1980 (Carter 1984). Whereas 
some collected birds had almost completed wing molt by 
mid-September, others that were still molting in early October 
would not have completed remigial molt until November 
(Carter, unpubl. data). Murrelets examined by Sealy (1975a) 
on 20 July at Langara Island had begun body molt on their 
capital and spinal tracts, but the remiges and rectrices had 
not begun to molt when observations ceased on 12 August. 
Kozlova (1957) stated that the complete molt of adult 
American Marbled Murrelets occurs in September and October 
and may extend into November, but she did not give the 
geographic locations of the specimens examined. She also 
noted that an Asian Marbled Murrelet collected in the Sea of 
Okhotsk on 31 August had already shed its flight and tail 
feathers but that other birds obtained in late August on the 
east coast of Kamchatka showed no traces of molt. Stresemann 
and Stresemann (1966) noted a rapid molt of the flight 
feathers that occurred between early August and late October, 
after examining specimens mainly from California. The 
closely related Kittlitz's Murrelet also undergoes a flightless 
pre-basic molt in Alaska between August and October (Sealy 
1977). Only a few other references to molting Marbled 
Murrelets have been made. Smith (1959) noted a bird "in 
changing plumage" drowned in a fisherman's net at Cohoe 



Beach, Alaska, on 22 August 1959. DeBenedictis and Chase 
(1963) noted one bird "in molt" on 27 July 1963 between 
Santa Cruz and Pigeon Point, California. Gill and others 
( 1 98 1 ) noted two flightless adults in Nelson Lagoon, Alaska, 
on 3 September 1977. On 1 September 1992, eight murrelets 
were collected in Mitrofania Bay, Alaska (Piatt, pers. comm.; 
Pitocchelli, pers. comm.): four birds were in alternate plumage 
(three with bare brood patches and one with a regressing 
brood patch), two birds were well into pre-basic molt and 
two birds were recently-fledged juveniles (with neck bands 
and egg teeth). In general, it appears that the timing of pre- 
basic molt follows breeding phenology throughout their range 
in North America. Large numbers of molting birds occur in 
museum collections which still need to be summarized to 
confirm this generalization (Carter, unpubl. data; Becking, 
pers. comm.). 

Failed breeders or stressed adult birds may initiate an 
unusually rapid body molt much earlier than the rest of the 
population. At Langara Island, British Columbia, Sealy 
(1975a) collected an adult female on 9 July 1971 with a fully 
developed brood patch and a flaccid ovary. This bird had 
already undergone a nearly complete body molt into basic 
plumage, without having yet started wing molt 

Timing of Pre- Alternate Molt 

The timing of pre-alternate molt is more poorly known 
than for pre-basic molt and appears to vary between breeding 
adults and subadults. For the American Marbled Murrelet, 
Kozlova (1957) stated that the incomplete pre-altemate molt 
began in April and is completed by late May. Molt may be 
delayed until June in first-year birds. One male, collected on 
31 May in the Diomede Islands, had many growing alternate 
plumage feathers (evident through active blood-filled papillae) 
on the upper parts, whereas most of the rest of the body was 
in basic plumage. This bird was collected north of the current 
breeding range for the species (Sealy and others 1982). It is 
possible that this bird was not molting in the usual pattern. 
Sealy (1975a) noted a slight delay in the pre-altemate molt 
in subadult murrelets at Langara Island, British Columbia. 
Both adults and subadults returned to Langara Island in late 
April. Most adults were in alternate plumage whereas 
subadults were still in basic plumage, although actively 
molting on their capital and spinal tracts. All subadults 
eventually achieved alternate plumage by late May (Sealy, 
pers. comm.). In Barkley Sound, British Columbia, two of 
45 birds in alternate plumage were considered to be subadult 
non-breeders because they lacked brood patches and had 
small gonads in June and July (Carter 1984). No birds in 
basic plumage were observed in Barkley Sound from early 
May to late July (Carter, unpubl. data). Occasional summer 
sightings of murrelets in basic plumage have been reported 
to Carter from various areas along the west coast of North 
America but none have been confirmed with specimens or 
photographs. Museum specimens must be examined to further 
confirm that all adult birds (including first-year birds) attain 
the full alternate plumage during the breeding season. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



105 



Carter and Stein 



Chapter 9 



Molts and Plumages 



Pre-Basic Molt Duration and Sequence 

The length of time required to complete the pre-basic 
molt is not well known because individuals have not been 
followed in captivity or in the wild throughout this period. In 
Barkley Sound, British Columbia, Carter (unpubl. data) 
determined that the relatively synchronous molt of the 
primaries, secondaries and rectrices in each individual required 
about 65 days but ranged between 45 and 75 days, based on 
a regression of molt scores and date (Pimm 1976). The entire 
pre-basic molt (body and remiges) probably requires about 
2-3 months per individual. In adult birds, pre-basic molt 
occurs almost simultaneously in all body tracts. Body molt 
begins slightly before and ends slightly after remigial molt. 
In the field, body molt is visible first in the throat area, as the 
dark feathers are lost and replaced with white feathers. The 
completion of body molt proceeds from anterior to posterior 
in ventral feather tracts from the breast to the vent area. In 
some ventral areas, thick dark-margined feathers are not all 
lost simultaneously and some are retained for a period of 
time. Remnant feathers from the alternate plumage were 
visible mainly in abdominal areas on museum specimens we 
examined as late as December. The grey-edged, dark back 
feathers (typical of the basic plumage) gradually replace the 
rust-edged feathers as the molt progresses. Certain museum 
specimens that had not yet shed their primaries already 
showed some grey-edged back feathers, suggesting that molt 
starts earlier in this region. 

During pre-basic molt, murrelets are flightless (Carter 
1984), as is expected during a synchronous wing molt. Such 
molts are considered to be adaptive by shortening the period 



of feather replacement in birds with aerodynamically 
inefficient wings such as loons (Savile 1957, Woolfenden 
1967), alcids, and diving petrels (Storer 1971, Stresemann 
and Stresemann 1966, Watson 1968). Stresemann and 
Stresemann (1966) considered the Marbled Murrelet to have 
an "accelerated" pre-basic molt where they incorrectly 
assumed that birds could barely fly during molt. Whereas 
murrelets are in fact flightless, they do have a less than 
synchronous pattern of primary replacement. Carter (unpubl. 
data) found that the first six primaries are lost in order and 
almost simultaneously; the outer four primaries are lost later. 
The order of feather loss and replacement is similar to gradual 
molting auklets and to most birds. The delay in the molt of 
the outer primaries also was evident in birds examined by 
Stresemann and Stresemann (1966). Eventually, all primaries 
are shed and growing at the same time. However, due to the 
delay in the shedding of the outer primaries, the growth of 
the inner primaries are completed first, leading to a rounded 
wing tip in birds later in the molt (fig. 5). Regardless of the 
delay in the outer primaries, pre-basic molt still occurs 
relatively rapidly. Molt duration is similar to Common Murres, 
Uria aalge (mean = 63 days in nine captive birds; Birkhead 
and Taylor 1977) but takes longer than for ducks (e.g., 18-29 
days; Bailey 1980, Balat 1970). 

Pre-Alternate Molt Duration and Sequence 

The duration and sequence of pre-alternate molt is even 
less well known. It is likely that this molt occurs more 
rapidly than the pre-basic molt. Carter and Erickson (1988, 
1992) noted that museum specimens from California collected 




Figure 5— Wing tracings of juvenile, hatching-year (HY) and adult, after-hatching-year (AHY) Marbled Murrelets, illustrating differences 
between non-molting and molting birds. Molting adult birds have "stubby" wings (bottom right) if all primaries have been recently lost, 
or "paddle-shaped" wings (top right) as the new inner primaries grow out before the outer primaries. All birds were collected on 1 
September 1 992 in Mitrofania Bay, Alaska by J. Pitocchelli. Tracings by H.R. Carter. 



106 



USDA. Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Carter and Stein 



Chapter 9 



Molts and Plumages 



as early as 18 February already had some white body feathers 
with broad dark margins on their underparts (fig. 2). It is not 
likely that these represent remnant feathers that were not 
replaced during pre-basic molt because several specimens 
exhibited a similar pattern in late February. By March, many 
specimens were well into alternate plumage. The first bird in 
full alternate plumage was collected on 26 March, as were 
several birds on the same date. Without further information, 
pre-alternate molt appears to occur rapidly and requires 
about one month. Additional field work and examination of 
more specimens will better establish the full sequence of the 
pre-alternate body molt. However, the highest density of 
dark, thick-margined feathers were seen in the neck area on 
several spring specimens, suggesting that molt proceeds 
from anterior to posterior in the ventral tracts. 

Behavior and Diet of Murrelets During 
Pre-Basic Molt 

In Barkley Sound, British Co'onbia, Carter (1984) noted 
that most adult birds departed f.om the Sound after breeding 
in early August (fig. 4) and presumably underwent the pre- 
basic molt elsewhere. However, the smaller numbers of 
adult birds that remained, moved into nearshore areas and 
underwent molt from late July to November. During this 
period, they occurred with juvenile birds which also did not 
appear to leave the Sound until at least early October. Even 
the smaller numbers of remaining adult and juvenile birds 
were mostly gone by late December 1979 (Carter, unpubl. 
data). Stresemann and Stresemann (1966) asserted that 
Marbled Murrelets molt after reaching their wintering areas. 
We presume that they reached this conclusion after examining 
molting birds from California where, at that time, murrelets 
were not known to breed. Undoubtedly, the proportion of 
birds that remain to molt near breeding areas rather than 
molt at wintering areas will depend on a variety of factors, 
including the timing of breeding, degree of winter residency, 
the timing of winter dispersal or migration, and other 
environmental parameters. McAllister (pers. comm.) reported 
that most adult birds remained in the general vicinity of 
summer feeding areas in southeastern Alaska but tended to 
occur in somewhat different areas and closer to shore during 
molt in September. 

In Barkley Sound, flock sizes of adult birds during pre- 
basic molt were difficult to obtain since few birds were 
present and it was difficult to separate molting and juvenile 
birds from a distance (Carter, unpubl. data). There were 30 
flocks from which molting birds were collected in 1979— 
1980. Of these flocks, 10, 15, 2, 2 and 1 contained 1, 2, 3, 5 
and 6 birds, respectively. The larger flocks also contained 
juveniles. McAllister (pers. comm.) noted the tendency for 
juveniles to occur very close to shore in southeastern Alaska, 
although he found juveniles in different areas than molting 
adults. Most molting and juvenile birds were observed very 
close to shore, usually within 200 m, in Barkley Sound 
(Carter, unpubl. data). Most birds were observed in the Deer 



Group islands (south of Fleming Island, mainly in Satellite 
Passage and near Seppings Island) and in the Broken Group 
islands (Sechart Channel, Coaster Channel and between 
Gibraltar and Nettle islands) (Carter, unpubl. data). In contrast, 
birds were found both in nearshore, inshore and offshore 
habitats in many parts of the Sound during the breeding 
season (Carter 1984, Sealy and Carter 1984). One adult bird 
that was collected on 24 July 1980 about 1.5 km SE of Cree 
Island was just beginning primary molt. Birds must swim 
into nearshore areas if they become flightless farther offshore. 

Carter (unpubl. data) collected five pairs of molting 
Marbled Murrelets in Barkley Sound in the pre-basic molt 
period. All were male-female pairs and were probably mated. 
One pair had started body molt but not wing molt and the rest 
were all actively undergoing wing molt. All mates were almost 
synchronized and had very similar molt scores, despite the 
generally asynchronous timing of molt within the population. 

During molt in Barkley Sound, adult and juvenile 
murrelets fed primarily on small fish of size classes U and HI 
(fish length classes: I, <30 mm; II, 30-60 mm; and HI, 60- 
90 mm), primarily Pacific Herring (Clupea harengus). Pacific 
Sand lance (Ammodytes hexapterus) and Northern Anchovy 
(Engraulis mordax) (Carter 1984). On average, similar size 
classes (II and HT) were eaten by wintering birds in December, 
but smaller size classes (I and U) of the same species were 
eaten by adults during the main breeding season. 

Guide for Differentiating Juvenile 
Murrelets from Adult Birds at Sea in 
Late Summer and Early Fall 

It is not difficult to differentiate a recently-fledged 
juvenile from an adult bird at sea in alternate or basic plumage, 
given adequate viewing conditions. Due to the protracted 
breeding season, however, many complete or partial plumages 
of both juveniles and adult murrelets may be encountered at 
sea during the late breeding and early post-breeding seasons. 
As mentioned previously, many factors affect counts of 
juveniles at sea during this time, including timing of fledging, 
at-sea mortality after fledging, timing of post-breeding 
dispersal and ocean habitats used. Highest counts will occur 
in suitable habitats in the post-breeding season when most or 
all juveniles have fledged but have not yet dispersed. However, 
during this period, confusion in identification occurs because 
adult birds undergoing pre-basic molt become difficult to 
separate from juveniles. In addition, older juveniles attain a 
first-winter plumage that is inseparable in the field from 
adult birds in basic plumage. Despite the difficulties of 
determining breeding success indirectly from surveys of 
juveniles at sea, such surveys are still one of the only measures 
of breeding success, unless larger numbers of nests can be 
located and monitored. 

To help prevent misidentification of juveniles at sea, 
Stein and Carter (1994) examined museum specimens and 
reviewed literature and unpublished data from British 
Columbia, Washington, and California in order to evaluate 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



107 



Carter and Stein 



Chapter 9 



Molts and Plumages 



five main criteria for use as an efficient, standardized method 
in identifying juveniles versus adult murrelets in the late 
breeding and early post-breeding season from June to 
November. The criteria examined included: (1) relative size; 
(2) darkrlight coloration ratio; (3) ventral coloration and 
patterning; (4) dorsal coloration; and (5) primary molt and 
wing shape. 

Recently-fledged juveniles are smaller than adult birds 
(70 percent adult mass at fledging) but become more similar 
to adult birds in size after foraging at sea during the first few 
months after fledging. Total length differed significantly 
between adult and juvenile birds (t = 7.52, P < 0.001) but 
some juveniles collected in August and September were as 
long as adults collected during these months. In the field, 
size may be a useful criterion for differentiating juveniles 
from adults when mixed flocks are encountered early in the 
post-breeding season, but will be less useful during August 
and September when many juveniles have already grown at 
sea for at least a month. 

Recently-fledged juveniles are lighter in color than adults 
in alternate plumage. In alternate plumage, a murrelet was 
estimated to have a 90:10 dark:light (D:L) ratio. In basic 
plumage, the ratio changed to 55:45 D:L. One molting adult 
collected on 19 June 1985 was 75:25 D:L. By August, more 
molting adults were found in the collections. Flightless adults 
collected on 11, 22, and 31 August measured 70:30, 65:35, 
and 79:21 D:L, respectively. By September, there was a 
larger range of color ratios. Some birds were close to 
completing the pre-basic molt while others were still in 
alternate plumage. Birds collected on 1 1, 16 (two specimens), 
20, and 30 September varied from 53:47, 99: 1, 47:53, 85: 15, 
and 52:48 D:L, respectively. Juvenile specimens were also 
examined. Generally, older juveniles appeared to be lighter 
overall than recently-fledged juveniles. The coloration ratio 
for juveniles with egg teeth averaged 65:35 D:L (n = 15), 
whereas juveniles without egg teeth averaged 58:42 D:L (n 
= 14). Due to protracted breeding, juveniles varying in age 
by 2-3 months may be present on the water in August and 
September and be easily confused with molting adults that 
may have similar color ratios. 

Recently-fledged juveniles possess a "speckled" 
appearance, resulting from many of the feathers on the sides 
of the head, neck, breast and abdomen being edged with thin 
dark margins (fig. 3). However, older juveniles without egg 
teeth appear lighter overall as many of the dark margins are 
lost. In contrast to the thin dark margins of juvenal feathers, 
the dark margins of the feathers of adult birds are much wider 
and have a "blotchy" appearance (fig. 3). As adult pre-basic 
molt progresses from anterior to posterior in ventral tracts, 
the density of blotchy feathers decreases. During boat survey 
work in Washington, "blotchy" feathers were often visible 
on the posterior ventral surface of molting adults as they 
dove in front of the boat (Stein, unpubl. data). This criterion 
was often helpful in separating adult birds that were more 
advanced in their pre-basic molt from juveniles. Remnant 
blotches or "speckled" feathers were noted on some specimens 



as late as November and December, but they probably would 
not have been noticeable in the field (fig. 1). 

As pre-basic molt progresses in adult birds, the rust- 
edged back feathers are gradually replaced by the grey- 
edged, dark back feathers, typical of basic plumage. Some 
grey-tipped back feathers were observed on adult specimens 
collected as early as June although specimens in which all 
orange-tipped feathers had been replaced had not been 
collected until July. Although recently-fledged juveniles are 
uniformly dark brown to almost black above, the upperparts 
of older juveniles possess grey margins similar to adult 
birds. Grey-tipped feathers were not noted on juvenile 
specimens collected earlier than September, but were more 
common after this time. Dorsal surface coloration appears to 
be an unreliable criterion for separating older juveniles from 
adult birds during the pre-basic molt. 

About 35 percent (n = 20) and 78 percent (n = 9) of adult 
birds collected in August and September, respectively, 
appeared flightless. Except for their flightless condition, 
these specimens were similar in appearance to juveniles. 
During August 1993 boat surveys in Washington, the condition 
of the molted primaries was the most reliable criterion for 
differentiating confusing birds (Stein, unpubl. data). Many 
adult birds were advanced in their pre-basic body molt by 
this time and could not be differentiated from older juveniles 
on the basis of the other four criteria mentioned above. 

In summary, the number of criteria useful for 
differentiating juveniles from adult murrelets decreases as 
the post-breeding season progresses. Of the five criteria 
evaluated, the first four would be useful in June and July. 
Recently-fledged juveniles are smaller than adults (70 percent 
body weight at fledging), lighter overall, appear speckled on 
the throat, breast, abdomen, and are uniformly dark brown to 
black on the upper body parts. In comparison, most adults in 
June and July are still in their alternate plumage and are 
much darker overall with rust-edged back feathers still 
apparent. By August and September, most adult birds are 
undergoing pre-basic molt, have lost the dorsal rust coloration, 
have replaced many of the dark-edged ventral body feathers 
with totally white feathers, and appear much lighter overall. 
Many juveniles, which have grown at sea for at least a month 
by this time, lose many of the characteristic speckled feathers, 
either by wear or replacement. During these months, the 
most reliable criterion for differentiating juvenile from adult 
birds is the condition of the molted primaries that can be best 
assessed when birds flap their wings while sitting on the 
water. Molting adults have "stubby" wings, if all primaries 
have recently been lost. Later, wings appear more rounded 
and have a "paddle-shaped" appearance when the new inner 
primaries become fully grown before the outer primaries 
(fig. 5). Juveniles lack gaps and have more pointed wing tips 
than molting birds at all times. In October and November, it 
is not practical to separate juveniles from adult birds in the 
field (fig. 1), although some late breeders, late molters and 
late fledglings may still be encountered and differentiated on 
the basis of all criteria. 



108 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Carter and Stein 



Chapter9 



Molts and Plumages 



Acknowledgments 

Murrelet research by Carter in Barkley Sound, British 
Columbia, was conducted as part of a Master of Science 
degree at the University of Manitoba, under the supervision of 
Spencer G. Sealy. In particular, Sealy's encouragement, interest 
and involvement with regard to studying alcid molt patterns 
has made this paper possible. Funding was provided through 
Canadian Wildlife Service Scholarships to Carter and Natural 
Sciences and Engineering Research Council grants to Sealy. 
The Bamfield Marine Station provided facilities and support. 
The Washington Department of Fish and Wildlife and the 



National Biological Service, U.S. Department of Interior, 
also provided support We thank Michael McNall and Andrew 
Nieman of the Royal British Columbia Museum and Karen 
Cebra of the California Academy of Sciences for assistance 
with museum specimens. Deborah Jory Carter and Leigh K. 
Ochikubo assisted with the examination of museum specimens 
and Michael Casazza assisted with figure preparation. We 
are grateful for the comments provided on this manuscript by 
Rudolph Becking, John Kelson, Linda L. Long, Irene Manley, 
Michael L. McAllister, Sherri Miller, Jay Pitocchelli, Peter 
Pyle, C. John Ralph, Sievert Rohwer, Steven M. Speich, and 
especially for those by Spencer G. Sealy. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



109 




Terrestrial Environment 






Chapter 10 

Marbled Murrelet Inland Patterns of Activity: Defining 
Detections and Behavior 



Peter W. C. Paton 1 

Abstract: This chapter summarizes terminology and methodology 
used by Marbled Murrelet (Brachyramphus marmoratus) biolo- 
gists when surveying inland forests. Information is included on the 
types of behaviors used to determine if murrelets may be nesting in 
an area, and the various types of detections used to quantify 
murrelet use of forest stands. Problems with the methodology are 
also discussed. 



Censusing Marbled Murrelets (Brachyramphus mar- 
moratus) at inland forest sites presents a relatively unique 
problem to avian ecologists attempting to assess population 
trends, determine current population densities, or merely 
quantify the presence or absence of birds on a specific tract 
of land. In contrast to most avian species which tend to be 
relatively sedentary and territorial on their breeding grounds 
(see Ralph and Scott 1981). murrelets are considerably more 
difficult to detect near their nests (e.g., Naslund 1993a; 
Nelson and Hamer, this volume a; Singer and others 1991). 

Murrelets tend to be detectable at inland forested sites 
only at dusk and dawn, and most observations are auditory 
detections of birds vocalizing while flying overhead (e.g., 
Naslund and O'Donnell, this volume; O'Donnell 1993; 
Rodway and others 1993b). In addition, murrelets are non- 
vocal near their nests (e.g., Naslund 1993a; Nelson and Hamer. 
this volume a; Singer and others 1991 ). suggesting that birds 
heard calling are often not near their own nest. Murrelets 
have been recorded as far inland as 84 km, with downy 
chicks found up to 64 km inland (Hamer and Nelson, this 
volume b; Ralph and others 1994). Therefore, murrelets 
observed flying overhead may be great distances from their 
breeding stands. Finally, virtually nothing is known about 
what percentage of birds visiting inland sites is non-breeding 
birds: this can be greater than 25 percent at Ancient Murrelet 
(Synthliboramphus antiquus) colonies (Gaston 1990). 

Detections provide a relative index to murrelet 
abundance, and presently have not been used to calculate 
density estimates. This is because individual murrelets will 
often circle over the forest canopy for long periods of time, 
vocalizing (Hamer and Cummins 1990, 1991; Naslund 1993a; 
Nelson 1989; Rodway and others 1993b). Therefore, a series 
of calls could represent a single bird or several birds. Unless 
a bird is under constant observation, it is usually extremely 
difficult to determine how many birds a series of detections 
actually represents. 



1 Biologist. Utah Cooperative Fish and Wildlife Unit, Utah State 
University. Logan. UT 84322 



Although biologists have attempted to quantify murrelet 
use patterns on inland forested sites for about eight years, 
the biological significance of these data has yet to be 
determined. Only when an actual nest site is found can one 
be certain that murrelets are breeding in a particular forest 
stand. All other types of observations only suggest, with 
varying degrees of certainty, that murrelets may be nesting 
in a specific tract of land forest tract. There is no definitive 
evidence that Marbled Murrelet use inland sites for night 
roosts. Birds in some areas can be detected at inland sites 
virtually year-round (Naslund 1993b). Only within the past 
five years have detailed behavioral observations taken place 
at nest sites. This information may aid in pinpointing nest 
sites by determining if murrelets give any unique behavioral 
clues near nests sites (e.g., Naslund 1993a). Many of those 
data are summarized for the first time by Nelson and Hamer 
in this volume a. 

The primary objective of this chapter is to give some 
sense of the types of data ornithologists have collected over 
the past eight years to quantify murrelet activity levels at 
inland forested sites. It is hoped that these data, specifically 
detection rates, can eventually be converted to a relative 
index to determine the approximate number of murrelets 
using a forest stand. Given the current state of the art 
concerning murrelet detection rates, comparisons between 
forest tracts are best done with data that were collected at the 
same time of year using similar methodology (e.g., fixed- 
point count for the entire morning survey period). Given 
those criteria, areas that have an order of magnitude difference 
in detection rates (e.g., 10 detections versus 100 detections) 
probably have different numbers of birds using each area, 
but exacdy how many birds a specific detection rate represents 
remains uncertain. 

Given this brief summary of the problems with surveying 
murrelets at inland sites, the following summarizes the 
methodology used by most ornithologists to quantify murrelet 
activity levels at inland sites: 

Definition of Detection 

The primary method for censusing Marbled Murrelets 
at inland forested stands is surveying from fixed points for 
varying amounts of time: 10 minutes (Paton and Ralph 
1990) to 2-3 hours (Naslund 1993a, ODonnell 1993). The 
sampling unit of inland surveys is a Detection, defined as the 
sighting or hearing of one or more murrelets acting in a 
similar manner (Paton and others 1990, Ralph and others 
1994). Therefore, only when the observer is certain that 
vocalizations are coming from the same bird or flock of 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



113 



Paton 



Chapter 10 



Defining Detections and Behavior at Inland Sites 



birds, is the observation classified as a single detection. Two 
birds under observation simultaneously, but behaving 
differently, are categorized as two separate detections. Hamer 
and Cummins (1990) required a minimum time interval 
between call notes to classify an observation as two detections. 
The total number of birds in a detection represents the total 
group size. Therefore, biologists can quantify detection rates 
between study sites (e.g., Naslund 1993b), and also determine 
annual fluctuations in mean total group size at the same site 
(e.g., Rodway and others 1993b). 

Rodway and others (1993) also suggested an alternative 
method to quantify murrelet activity patterns would be to 
count all vocalizations and visual detections, rather than 
keep track of total detections. 

Type of Detection 

Murrelet detections are generally classified into one 
of three categories: (1) the bird was only heard and not 
seen (i.e., an audio detection); (2) the bird was seen and 
not heard (a visual detection); or (3) the bird was both seen 
and heard (see Ralph and others 1994). Audio detections 
are usually subdivided into separate types of vocalizations 
and mechanical sounds, in the hope that future researchers 
will be able to determine the context when a specific 
vocalization is given. As far as I know, there is no unique 
vocalization given only at the nest that would aid researchers 
in finding nests (Nelson and Hamer, this volume a). Listed 
below are the current categories for types of vocalizations 
and visual detections. 

Types of Audio Detections 

(1) Keer calls — two-syllable, high-pitched vocalization, 
similar to the vocalizations of many gulls (Larus spp.) 
(O'Donnell 1993). When properly trained, there appears to 
be little observer bias in quantifying the number of keer calls 
given by murrelets (Rodway and others 1993b). During the 
summer months in northern California, 91.1 percent of the 
detected birds vocalized, compared to 98.7 percent during 
the winter months (O'Donnell 1993). In addition, O'Donnell 
(1993) found in the summer that murrelets flying above the 
canopy were significantly more likely (P < .001) to vocalize 
than birds flying below the canopy. Rodway and others (1993) 
found the number of detections increased on cloudy days, but 
the number of calls per detection was not affected by weather. 

(2) Non-keer calls — A low, two-part, guttural vocal- 
ization, which some researchers believe is associated with 
reproductive behavior. However, O'Donnell (1993) heard 
murrelets give non-keer vocalizations all months of the year, 
although at a reduced rate from December through February. 
In addition, O'Donnell (1993) found in his study of nine 
forest stands, that an average of 12 percent of murrelet 
detections had one or more non-keer vocalization (range = 
7.5-2 1 .9 percent). For further details, see Nelson and Hamer 
(this volume a) who have subdivided non-keer vocalizations 
into whistle- and groan-like calls. 



(3) Stationary calls — Detections with three or more calls 
that are 100 m or less from the observer, where the observer 
believes the bird has not moved, are classified as a Stationary 
Detection (Ralph and others 1994). 

(4) Wing beats — A tremulous, fluttering sound presumably 
generated by movement of a murrelet' s wings through the air. 
Singer and others (1991) heard wing beats near active nest 
sites, and wing beats were also heard every morning near an 
active nest in northern California (Fortna, pers. comm.). Wing 
beats were heard on 0.5 percent of detections at nine sites 
in northwestern California (O'Donnell, pers. comm.). 

(5) Jet dive — Little is known about the origin or function 
of the jet dive, or power dive, which makes a sound somewhat 
similar to the roar of a jet engine. It is heard rarely, comprising 
only 10 of 21,437 detections at nine sites in northwestern 
California (O'Donnell, pers. comm.). This sound is 
presumably a mechanical sound made by murrelet' s feathers 
while in a steep dive above the forest canopy. Nelson and 
Hamer (this volume a) report in Oregon on the rare occasions 
when this sound is heard, it is usually near a nest tree. 

Visual Detections 

Rodway and others (1993) found significant variation 
between observers in the proportion of murrelets that were 
visually detected. This suggests that biologists doing field 
work should be screened and trained to insure that there is 
minimal observer bias (see also Ralph and others 1 994 for 
training details). Categories for visual detections include: 

(1) Birds flying above the canopy — This includes both 
straight-line flight and circling over a forest stand. This was 
the most frequently observed type of detection in a study of 
nine study areas in northwestern California, ranging from 8 
to 33 percent of all detections (O'Donnell 1993). In British 
Columbia, 75-89 percent of all detections were birds flying 
over the canopy (Rodway and others 1993b). 

(2) Birds flying below the canopy — This refers to 
murrelets both flying through a forest stand and adjacent to 
the stand. O'Donnell (1993) found that at Lost Man Creek in 
northern California, 25 percent of murrelet detections during 
the summer months (April-August) were birds flying below 
the canopy, compared to 0.4 percent during the winter months 
(September-March). Rodway and others (1993b) found that 
more birds flew below the forest canopy in June than during 
other times of the year. 

(3) Landing and perching in a tree — O'Donnell (1993) 
found 0.4 percent of the total summer detections (April- 
August) at Lost Man Creek were of birds landing in trees (n 
= 10,154), although no nests were found in this area. At two 
active nests, Naslund (1993a) observed birds flying to the 
nest for incubation exchanges 3 1 minutes before sunrise to 3 
minutes after sunrise (see also Nelson and Hamer, this volume 
a). Adults typically took predictable flight paths to the nest 
(Nelson and Hamer, this volume a). Murrelets incubate for 
24-hour bouts (Naslund 1993a, Nelson and Hamer, this volume 
a, Singer and others 1991). Nest exchanges and feedings 
generally took place 17-24 minutes before sunrise, with two 



114 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 10 



Defining Detections and Behavior at Inland Sites 



daytime feedings 82 and 150 minutes after sunrise (Naslund 
1993a, see also Nelson and Hamer. this volume a). 

Below are some relevant definitions useful to biologists 
studying Marbled Murrelets, based on Ralph and others (1994): 

Potential Nesting Habitat — (1) mature or old-growth 
coniferous forests: mature forest can be with and without an 
old-growth component (see Ralph and others 1994, Raphael 
and others, this volume); (2) younger coniferous forests that 
have large, deformed trees or structures suitable for nesting. 

Forest Stand — a group of trees that forms a continuous, 
relatively homogeneous, potential nesting habitat with no 
gaps >100m. 

Survey — The process of determining the presence, 
absence, and occupancy of Marbled Murrelets in a forest 
stand. Surveys generally are conducted during the morning 
hours, when detection rates are greatest (Paton and others 
1990: Ralph and others 1994; Rodway and others 1993b). In 
addition, surveys generally occur from May through July 
when detection rates peak (e.g., Rodway and others 1993b); 
however, murrelets are known to visit inland forest stands 
throughout the year (Naslund 1993b; O'Donnell 1993; 
O'Donnell and Naslund, this volume). 

Intensive Survey — Designed to determine the probable 
presence, absence, or occupancy of Marbled Murrelets in a 
specific tract of land. When conducting an intensive survey, 
the observer surveys from one point for the entire morning 
survey period. Under most forest conditions, observers can 
see murrelets within 100 m. and hear them within 200 m 
(Ralph and others 1994). Therefore, approximately 12 ha (n 
x [200 m] 2 = 12.6 ha) can be adequately surveyed from a 
single point for auditory detections, while only 3.14 ha can 
be monitored for visual detections. Under certain conditions, 
visual and auditory ranges are reduced (e.g., next to a stream 
or under a dense forest canopy). Surveys generally are 
conducted from 45 minutes before sunrise to 75 minutes 
after sunrise (Paton and others 1990, Ralph and others 1994), 
although surveys at northern latitudes start and end earlier 
(e.g., Kuletz and others, this volume; Rodway and others 
1993b). The exact methodologies for Intensive and General 
Surveys are detailed in Ralph and others (1994). 

General Survey — A survey designed to determine the 
geographic distribution of Marbled Murrelets over large 
tracts of land (e.g., states, counties, basins). General surveys 
are exploratory in nature and cannot be used to confirm 
murrelet absence from specific forest stands. These surveys 
consist of a transect of 8-10 stations surveyed during a 2- 
hour period each morning. Stations are spaced 0.5-1.0 km 
apart, depending on terrain, with each station surveyed for 
10 minutes. 

Survey Area — the entire area being surveyed. 

Survey Visit — a single morning's visit 

Survey Site — an area containing >1 survey station. 

Survey Station — the exact location where an observer 
stands to survey murrelets. 

Occupied Stand — a forest stand, consisting of potential 
nesting habitat, where murrelets were observed exhibiting 



subcanopy behaviors associated with nesting. Quantitative 
information on murrelet behavior near nests is scarce; 
however, some data are available from Naslund (1993a), and 
Nelson and Hamer (this volume a). Data collected by Naslund 
(1993a) suggests that only 6-21 percent of the detections 
<100 m from known active nests represent "occupied 
behaviors" (see below), while most detections near nests 
were birds flying above the canopy. The proportion of 
detections which were categorized as occupied behaviors 
was not affected by weather conditions (i.e. cloud cover, 
ceiling), although the total number of detections increased 
significantly on cloudy days (Naslund 1993a, Rodway and 
others 1993b). 

Evidence for Nesting: 

Seven different categories are considered indicators of 
nesting. They are listed below with varying degrees of certainty 
that murrelets are nesting in a particular forest stand. Only 
categories 1-3 listed below provide confirmation of breeding, 
whereas categories 4-7 are occupied behaviors, which are 
behaviors that suggest that murrelets could be nesting in a 
specific forest stand. 

Confirmation of breeding: 

( 1 ) Discovery of an active nest — either with an incubating 
adult, brooding adult and chick, or pre-fledged chick. 

(2) Obvious signs of recent nesting activity — such as 
fecal rings surrounding the nest or eggshell fragments in a 
nest scrape. 

(3) Discovery of a chick or eggshell fragments on the 
forest floor — see Becking 1991, and Ralph and others 
1994 for detailed information on the characteristics of 
murrelet eggs. 

Occupied behaviors: 

(4) Birds flying below the top of the forest canopy (also 
called subcanopy behaviors; Ralph and others 1994) — This 
refers to murrelets either flying through the stand, into or out 
of the stand, or adjacent to a forest stand, the weakest evidence 
in this category (O'Donnell and Naslund, this volume; Rodway 
and others 1993b). Because tree heights can vary, a bird 
observed at or below the height of the top of the tallest tree 
visible to the observer would be classified as a subcanopy 
detection. Based on observations at active nests, only silent 
birds are probably near an active nest (Naslund 1993a, but 
see Nelson and Hamer, this volume a). This category includes 
birds flying over or along roads, young stands, or recently 
harvested areas adjoining potential nesting habitat. In these 
latter two instances, only the adjacent potential nesting habitat 
should be classified as occupied. Subcanopy behaviors are 
currently thought to be the strongest indirect evidence of 
nesting in a stand (Ralph and others 1994). 

(5) Birds circling above the forest canopy at any radii — 
Circling is common flight behavior over occupied sites. 
However murrelets have also been observed circling over 
young or non-forested habitats (Hamer and Cummins 1990, 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



115 



Paton 



Chapter 10 



Defining Detections and Behavior at Inland Sites 



1991; Nelson 1989, 1990a). In most instances, these areas of 
apparently unsuitable nesting habitat were near or adjacent 
to potential nesting habitat. Circling is currently believed to 
be fairly strong evidence that a stand is occupied (Ralph and 
others 1994). 

(6) Birds seen perching, landing, or attempting to land 
on tree branches — There are a total of six flight behaviors 
recorded near known active nests (Naslund 1993a; Nelson 
and Hamer, this volume a; Singer and others 1991). Birds 
landing in trees could indicate nest sites, although I know of 
no evidence to suggest that murrelets commonly perch in 
trees near active nests. Therefore, perching is currently not 
definitive evidence there is a murrelet nest in the area. 
During observations of two nests in Big Basin State Park, 
California, Naslund (1993a) found that, during incubation 
exchanges, the adults always flew directly to the nest branch 
without vocalizing (with one exception), landed directly on 
the nest branch, and then walked to the nest (see also Nelson 
and Hamer, this volume a). 

(7) Birds calling from a stationary location within the 
stand. — This category only applies to detections with >3 
calls heard and a bird <100 m away. Adult and juvenile 
murrelets are generally quiet while on the nest limb (Nelson 
and Hamer, this volume a). Naslund (1993a) never heard 
adults give loud vocalizations from the nest while incubating 
or brooding. Because adults and juveniles tend to be relatively 
quiet on the nest, this category is probably weak evidence 
for an active nest in the area, at least for the murrelet giving 
the vocalizations. Further research is needed to quantify the 
types of behaviors given at active nests. 

Presence 

When murrelets are detected, but no occupied behaviors 
are observed, then observation is categorized simply 
as "presence". 



Discussion 

Most biologists conducting murrelet surveys use 
detections, defined as the sightings or hearing of individuals 
or flocks behaving similarly, as the independent sampling 
unit. The primary variable when comparing studies is the 
amount of time observers remain at survey stations, which 
can range from 10 minutes to 3 hours. Most inland surveys 
conducted to date have concentrated on the breeding season 
(April through August). However, a recent paper by Naslund 
(1993b) suggests that surveys during the winter months may 
be more useful for monitoring long-term population trends. 
This was because variability in detection rates is relatively 
low in the winter months compared to breeding season surveys. 
Currently, we have no basis to convert detection rates into 
density estimates, and it is unclear when ornithologists will 
be able to determine an accurate conversion factor. However, 
Ralph (pers. comm.) and Miller (pers. comm.) recently have 
been working on determining a conversion factor, using a 
combination of offshore survey data and intensive inland 
surveys. Data that have been gathered to date will provide 
baseline data for future researchers, and can be used for 
comparative purposes across studies to provide relative indices 
to murrelet activity patterns. 

Acknowledgments 

I thank Peter Connors, Steve Courtney, Dave Fortna, 
Anne Harfenist, Gary Kaiser, Brian O'Donnell, C.J. Ralph, 
Lynn Roberts, Michael Rodway, and Fred Sharpe for useful 
comments on earlier drafts of this paper. 



116 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 11 

Patterns of Seasonal Variation of Activity of Marbled 
Murrelets in Forested Stands 



Brian P. O'Donnell 1 



Nancy L. Naslund 2 C. John Ralph 3 



Abstract: Determining the annual cycles of Marbled Murrelet 
(Brachyramphus marmoratus) behavior is crucial both for under- 
standing the life history and for management of this species. In 
this paper we review available information on the annual patterns 
of behavior in forests throughout its range, with special emphasis 
on California. Data were derived from standardized forest 
surveys. Murrelet activity peaks during the summer (breeding 
season), is lower during the winter (non-breeding season), and 
absent or very low during transition periods (pre-altemate and 
pre-basic molts). Murrelets are regularly detected at some breed- 
ing stands even in the non-breeding season, however, birds are 
rarely observed flying through or below the forest canopy during 
this period. Vocalizations and flock size exhibit seasonal varia- 
tion as well. While certain aspects of seasonal activity and behav- 
ior patterns conform with our limited understanding of its life 
history, much of the species' behavior within the forest remains a 
mystery. Current guidelines for monitoring the Marbled Murrelet 
at inland sites restrict surveys for management purposes to the 
breeding season. 



Determining the annual cycles of Marbled Murrelet 
{Brachyramphus marmoratus) activity and behavior at inland 
sites is important for an understanding of this species' life 
history and its management. In order to assess the probable 
presence of nesting murrelets in a forest stand, we must 
first know how their behavior in the forest changes through 
the year, and what these seasonal changes tell us about its 
biology. With this information in hand, we can then 
determine how best to develop a survey protocol. In this 
chapter we review available information on the annual 
cycles of activity and behavior in the forest. We draw 
heavily from the results of two studies in California (Naslund 
1993,b; O'Donnell 1993). Data in these studies were 
collected using intensive survey techniques (Paton, this 
volume; Paton and others 1990; Ralph and others 1994). 
Additional information, derived from general and intensive 
survey techniques, are reported from studies throughout 
the range of the species. As the measure of activity we use 
the "detection": the observation of one or more birds acting 
in a similar manner. 



1 Wildlife Biologist. Pacific Southwest Research Station. Redwood 
Sciences Laboratory, USDA Forest Service, 1700 Bay view Drive, Areata, 
CA 95521 

2 Wildlife Biologist, U.S. Fish and Wildlife Service, U.S. Department 
of Interior, 1011 E. Tudor Road, Anchorage, AK 99503 

3 Research Biologist, Pacific Southwest Research Station, Redwood 
Sciences Laboratory, USDA Forest Service, 1700 Bay view Drive, Areata, 
CA 95521 



Variation in Detection Levels 

Numbers of detections vary dramatically through the 
year, and in general, are greatest during the summer months. 
Detection levels were compared between months through 
the year in two studies in California. Naslund (1993a) 
compared detections at two sites in central California between 
three periods: (1) breeding season — April through July; (2) 
transition periods — March, and August through October; 
and (3) winter — November through February. She found 
detections were significantly greater during the summer period 
{table 1). O'Donnell (1993) also found significant differences 
between months at each of three sites in northwestern 
California. He found that mean numbers of detections per 
survey were greatest in July at all sites {fig. 1). Mean numbers 
of detections per survey in April through June ranged from 
28 to 62 percent of those in July, and mean detections per 
survey in May and June were always intermediate between 
those in April and July. 

Murrelet detection levels also tended to peak during 
July and early August in most locations to the north of 
California. In Oregon, Nelson (1989) found the greatest 
detection levels from 12 July to 9 August at sites {fig. 2). She 
also noted an early period of high activity in late May and 
early June. The two activity peaks were detected during both 
dawn and evening surveys. During the 1990 breeding season 
in northwestern Washington, Hamer and Cummins (1990) 
noted 60 percent of all detections occurred between 25 June 
and 27 July {fig. 3). Numbers of detections per survey were 
greatest from mid- July through the end of the month. In the 
following summer, 77 percent of all detections were recorded 
between 8 July and 1 1 August, with the greatest numbers of 
detections per survey occurring in the week of 22 July 
(Hamer and Cummins 1991) (fig. 4). During 1990, weekly 
mean numbers of detections of murrelets peaked in the last 
week of July at two sites in the Queen Charlotte Islands, 
British Columbia (Rodway and others 1993b) {figs. 5, 6). 
However, detection levels at sites on Vancouver Island, 
British Columbia, reached their greatest levels in late June 
(Manley and others 1992) (fig.7). At a site on Mitkof Island, 
in southeastern Alaska, where Doerr and Walsh have 
conducted one to three morning surveys each month since 
December 1992, the numbers of detections peaked in July 
(Doerr, pers. comm.; Walsh, pers. comm.). Kuletz and others 
(1994c) found that seasonal peaks of murrelet activity in 
Prince William Sound, Alaska, were similar to those reported 
for the more southerly areas of the species' distribution (figs. 
8, 9). They also noted an early period of greater activity in 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



117 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 






Table 1 — Variations in Marbled Mur relet detections by season at the Waddell Creek and Opal Creek nest sites. Results are given for each ANOVA comparing 
total detections, occupied detections, and percent of occupied detections between seasons 





n 




Total detections 


Occupied Detections 


Percent Occupied* 


Variable 


X 


s.d. 


(range) 


X 


s.d. 


(range) 


X 


s.d. 


(range) 


Waddell Creek 






















Season 






















Summer 


32 


50 


32.8 


(18-176) 


5 


8.6 


(0-45) 


18.6 


17.1 


(0-55.6) 


Transition 


16 


12 


17.1 


(0-49) 





0.6 


(0-2) 


8.7 


25.5 


(0-100.0) 


Winter 


2 


17 


2.8 


(15-19) 


1 


0.7 


(0-1) 


16.7 


23.6 


(0-33.3) 


ANOVA 




F=14.6, 


df=2, 


P = 0.0001 


F = 4.11, 


df = 2, 


P = 0.0230 


F = 2.79, 


df=2, 


P = 0.0723 


Opal Creek 






















Season 






















Summer 


18 


100 


28.5 


(59-159) 


2 


2.1 


(0-7) 


6.3 


6.1 


(0-21.7) 


Transition 


11 


25 


41.5 


(0-121) 





0.3 


(0-2) 


0.2 


0.7 


(0-2.4) 


Winter 


2 


80 


22.6 


(64-96) 





0.0 










ANOVA 




F= 16.51, 


df = 2. 


P = 0.0001 


F = 2.5l, 


df=2, 


/> = 0.1019 


F=7.92, 


df=2, 


P = 0.0022 



1 Percent of detections < 100 m from the observer 



SITE 


Mar 


Apr 


May 


Jun 


Jul 


Aug 


Sep 


Oct 


Nov 


Dec 


Jan 


Feb 


n 


10 


17 


19 


16 


17 


10 


6 


4 


7 


3 


8 


8 


•Tno. 


JI2.8_ 


70.3 

A 


113.8 
A 


140.6 
A 


222.9 
A 


87.2 


.7 


25.8 


52.4 


29.0 


39.1 


22.9 


Lost Man 
Creek 




B 
C 


B 

C 


B 




b 
C 






B 
C 


C 


8 
C 








D 








D 




D 





D 


D 


D 




E 












E 












n 


9 


13 


18 


16 


12 


10 


7 


6 


7 




■ : '7 


i 4 


If no. 


3.0 


36.9 


58.8 
A 


60.4 
A 


134.3 
A 


54.0 


.1 


17.8 


32.1 




12.3 


14.5 


James Irvine 




B 


8 


B 




B 






B 








Trail 


E 


C 








C 

D 


F 


D 

E 


C 




D 
E 


C 
D 


n 


4 


5 


7 


7 


8 


7 








3 






7 no. 


-6_'°_ 


26.4 


29.3 


33.1 


84.0 


16.7 








4.3 






Experimental 




A 
B 


A 
B 


A 

B 


A 


B 














Forest 


c 










C 








C 







Mar Apr May Jun Jul 



Aug Sep Oct Nov Dec Jan Feb 



Figure 1 — Results of Ryan-Einot-Gabriel-Welsh multiple range test comparing numbers of detections of 
Marbled Murrelets per survey between months at three sites in northwestern California, 1989-1991. 
Months with the same letter indicate that the mean numbers of detections were not significantly different 
from each other, "n" indicates the number of surveys in the respective month and " x no." is the mean 
number of detections per survey. Means presented are untransformed values. Surveys from all years were 
combined for the analysis, and months with less than three surveys were not included in the analysis. From 
O'Donnell 1993. 



118 



USDA Forest Service Gen. Tech. Rep. PSW-I52. 1995. 



O'Donnell and others 



Chapter I 



Panems of Seasonal Variation of Activity 



CO 

| 

I- 
O 

LU 
I- 
LU 

Q 
LL 

o 

DC 

LU 
CD 

z 




MAY 



JUNE 



JULY 



AUGUST 



MONTH 



Figure 2 — Number of detections of Marbled Murretets by survey month on 20 
general surveys (Paton, this volume) along roads, central Oregon Coast Range. 
1988. From Nelson 1989. 



300 



CO 

z 
g 

H 
U 
LU 



o 

CO 

cr 

LU 
CD 




MAY 16-30 JUN 12-25 

MAY 31 - JUN 11 



JUL12-27 
-JUL11 JUL28-AUG15 



TWO-WEEK PERIODS 

Figure 3 — Total number of Marbled Murrelet detections for each 
two-week period from 1 6 May to 1 5 August, 1 990. at 41 sites (245 
survey mornings) in northwestern Washington. From Hamer and 
Cummins 1990. 



CO 

z 
o 



E 



LL 

o 

CO 

et- 
ui 
m 

5 

Z 



300 



250 " 



200 - 




MAY 



JUNE 



Figure 4 — Total number of Marbled Murrelet detections for each one- 
week period from 9 May to 9 August 1991, at 75 sites (287 survey 
mornings) in northwestern Washington. From Hamer and Cummins 
1991. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



119 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 



100 
80 






,11 


60 




I 


40 




I 

I 




20 


. ' I 










I I 





,1 1 — 


• ■ ■ 


I I I 1 1 



* / /p y/ + y.*s 



Figure 5 — Weekly mean (±s.e.) numbers of Marbled Murrelet detections per survey at 
Phantom Creek, British Columbia, in 1990. A total of 49 morning surveys were 
conducted. From Rodway and others 1993b. 



100 














V) 














Z 80 
O 


- 








I 


I 


F 












o 














UJ 60 

z m 

W": 40 

s o 


■ I 


I 


] 


; ■ 






BE 

UJ 




I 








m 20 














1 














D 














Z 

















■ 


" 


i 


■ 


i 


i 



At 



* 



<? 



£ 



# 






.# 



& 



# 



$> # 






# 



9 



<? 



Figure 6 — Weekly mean (±s.e.) numbers of Marbled Murrelet detections per 
survey at Lagins Creek, British Columbia, in 1 990. A total of 33 morning survey 
were conducted. From Rodway and others 1993b. 



late May, similar to that reported by Nelson (1989). At two 
sites, detection levels during this period were approximately 
equal to or even exceeded those during July. 

Patterns of Detections in Winter 

Winter attendance at breeding stands is well documented 
for California (figs. 10, 11). For example, Sander (1987) 
detected murrelets on 66 percent of 53 mornings surveyed 
between January and mid-March at a site in northwestern 
California. Carter and Erickson (1988) also report on the 
detection of murrelets from January through March at Big 
Basin State Park (central California) over several years. 



Murrelets have also been detected at forest stands during the 
winter in Oregon (Nelson, pers. comm.), Washington (Hamer, 
pers. comm.), and southeastern Alaska (Naslund, unpubl. 
data; Walsh, pers. comm.). 

At the three sites in northwestern California, O'Donnell 
(1993) found that mean numbers of detections per survey 
during the winter months (November through February) ranged 
from nine to 24 percent of mean levels in July, with detection 
numbers in November consistently the greatest in this period. 
Naslund (1993b) found that mean numbers of detections for 
winter surveys ranged from 35 to 80 percent of mean summer 
detection levels for five sites in central California (fig. 12). 
Doerr and Walsh noted similar differences between winter 



120 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



O'Donnell and others 



Chapter 11 



Patterns of Seasonal Variation of Activity 



Off 

UJ CO 

£55 



70 



60 " 



50 



40 



< UJ 

il 



¥ 30 



20 - 



10 - 




«» w 



CM 



3S88SS8S5 



WEEK BEGINNING 

Figure 7 — Weekly mean numbers (bar segment) of Marbled Murrelet detec- 
tions per survey per week along upper Carmanah Creek, British Columbia, 28 
May to 27 August 1990. The standard error (line segment) and number of 
surveys in each week are also given. From Manley and others 1992. 



and summer detection levels on Mitkof Island, Alaska (Doerr, 
pers. comm.; Walsh, pers. comm.). 

The remaining two months, March and September, had 
low to no activity (Naslund 1993b, O'Donnell 1993). In 
northwestern California, detections in March ranged from 
four to seven percent of those in July, and September levels 
were less than one percent of July levels. Detection levels in 
March and September were usually significantly lower than 
all other months (O'Donnell 1993). 

Patterns of Absence from Stands 

Marbled Murrelets are most often absent, or in much 
reduced numbers, from breeding stands during the two 
transition periods: (1) March, and (2) mid- August through 
October (Naslund 1993b, O'Donnell 1993) (figs. 10, 11). 
Doerr and Walsh failed to detect murrelets from 27 August 
to 2 October 1992, at their study site on Mitkof Island, 
Alaska (Doerr, pers. comm.; Walsh, pers. comm.). In central 
California, from 1989 through 1991 , the proportion of surveys 
in August through October with murrelets present was 
significantly lower than for surveys in summer, in winter, or 
in March (Naslund 1993a). Similarly, murrelets were observed 
on a significantly smaller proportion of surveys in March 
than during summer or winter (Naslund 1993a) (fig. 13). At 
nine sites in northwestern California from 1989 through 
1991, murrelets were not detected on 33 percent of 30 surveys 



conducted in March, nor on 3 1 percent of 80 surveys conducted 
in August through October (O'Donnell 1993). In addition, 
surveys with no detections occurred with significantly greater 
frequency from November through February than from April 
through July. Murrelets were not detected on 10 of 71 counts 
conducted from November through February, while birds 
were detected during all but one of 227 surveys from April 
through July (O'Donnell 1993). 

Variation in Frequency of Behaviors and 
Flock Sizes 

Flight Altitude 

The behavior classes recorded during murrelet surveys 
differentiate between two classes of behaviors, those occurring 
above the forest canopy, and those at the top, below, and 
within the forest canopy. These latter behaviors occurring at 
or below the canopy we will refer to as "below canopy 
behaviors", and are considered most indicative of probable 
nesting (Ralph and others 1993). They have also been referred 
to as "occupied behaviors", that is, indicative of birds occupying 
a given stand for nesting. Studies in California (Naslund 1993b, 
O'Donnell 1993) found that the numbers of different behaviors, 
both above and below the canopy, differed significandy between 
months through the year (table 1). O'Donnell (1993) detected 
murrelets flying above the canopy throughout the year at three 
sites in northwestern California. The patterns of behaviors 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



121 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 






200 



150 



100 




MAY JUNE 



JULY 



AUG 



ZOO 


SITE 2 




Z 






O 






m 






111 






1 




/ A \ 


Q 100 
U. 




/ ' \ \ 


O 
OC 
! 50 


ao\-°- -/ 


/ j/ \\ 


1 






Z 









t 1 ■ ■ 


H 1 1 1 • ' ' 



MAY JUNE 



JULY 



AUG 



200 



150 



100 - 




• 1991 

— □ 1992 



MAY 



JUNE 



JULY 



AUG 



Figure 8 — Numbers of detections of Marbled Murrelets per survey at three sites on Naked 
Island in Prince William Sound, Alaska, during the 1991 and 1992 breeding seasons. From 
Kuletz and others 1994c. 



above the canopy at Lost Man Creek (fig. 14a) are representative 
of those at the two other sites. Numbers of these behaviors 
were greatest during the breeding season, reaching a peak in 
July, and lowest during the non-breeding season, with a small 
winter peak in November (fig. 14b). 

Below canopy behaviors showed a more pronounced 
seasonality (fig. 14a) at Lost Man Creek and are representative 
of the two other sites. Naslund (1993a) found that only a 
small percentage of detections recorded near two nest trees 
during the winter, non-breeding season (October through 
March) consisted of below canopy behaviors (table 1). 
Similarly, at Lost Man Creek, below canopy behaviors made 
up 24.7 percent of detections from April through August, 



while from September through March, of 1 , 1 85 detections, 
only five were of murrelets flying below the canopy. Numbers 
of occupied behaviors were segregated by a multiple range 
test into three periods at Lost Man Creek (fig. 14b), reflecting 
peak levels in July, lower levels during the remainder of the 
breeding season, and their absence in the non-breeding season. 

Vocalizations 

O'Donnell (1993) examined seasonal differences in 
the number of calls per detection at two sites in northwestern 
California, Lost Man Creek and James Irvine Trail. 
Detections with greater then 9 calls were assigned a value 
of ten for the analysis. The number of calls per detection 



122 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



O'DonneU and others 



Chapter 11 



Patterns of Seasonal Variation of Activity 



</> 



UJ 

tu 

o 
- 
o 

cc 
tu 
m 



250 



200- 



150- 



100- 



50- 




i — " — r 

AUGUST 



Figure 9 — Numbers of detections of Marbled Murrelets throughout western Prince 
William Sound, Alaska, during the 1992 breeding season. Data was collected at 67 
randomly selected (boat-based) sites. From Kuletz and others 1994c. 



was greater during the winter, from October through 
February, than from March through August at both sites. 
For instance, at Lost Man Creek the mean numbers of calls 
per detection ranged from 7.4 to 9.3 from October through 
February. From March through August mean calls per 
detection ranged from 3.7 to 6.1 at this site. Numbers of 
calls were significantly different between months only at 
Lost Man Creek. 

Rodway and others (1993b) compared levels of vocal- 
izations between months at two sites in the Queen Charlotte 
Islands. British Columbia. Months compared were May 
through July at the Lagins Creek site, and May through 
August at the Phantom Creek site. They examined changes in 
both the number of calls per detection (all calls counted), as 
well as number of calls per survey (detections with "multiple" 
calls assigned a value of 25). Number of calls per detection 
were similar in May, June, and July at both sites. At Phantom 
Creek, vocalization levels dropped significantly after July 
24. Number of calls per survey increased through July, reaching 
peaks in the last week of July at both sites, and falling rapidly 
in the second week of August at Phantom Creek. 

O'DonneU (1993) also looked at the occurrence of grunt 
or groan calls (previously referred to as alternate calls) for 
evidence of seasonal patterns (Paton. this colume; Nelson 
and Hamer. this volume a). He compared between months, 
for above and below canopy detections, the proportion of 
detections which included these calls, and found seasonality 
was not marked. Only at Lost Man Creek, where the sample 
was greatest, was there significant differences between 
months in the proportions of detections above the canopy 
with alternate calls. These calls occurred in lower proportions 
in December through February than in the remainder of the 
year. However, a subsequent multiple range test did not 
distinguish any significantly different months. The 
percentage of detections of murrelets giving grunt calls 
below the canopy also showed seasonality only at Lost 



Man Creek, where they occurred in significantly greater 
proportions during July. 

Flock Size 

Changes in flock size through the breeding season have 
been noted in two studies (O'DonneU 1993, Rodway and 
others 1993b). O'DonneU (1993) found that both flocks 
observed above and below the canopy were smallest during 
May and June at each of three sites in California. Above 
canopy flocks at the Experimental Forest site had significantly 
fewer birds in June, and were also smaller in June at James 
Irvine TraU, though not significantly so {fig. 15). At Lost 
Man Creek, flocks above the canopy had significantly fewer 
murrelets in May, and significantly more birds in July. 
Reduction in the size of flocks below the canopy was especially 
marked at James Irvine Trail, where flock size was 
significantly less in May and June than during the remaining 
summer months {fig. 15). Below canopy flocks with the 
fewest numbers occurred in June at all three sites. 

Rodway and others (1993b) similarly detected smaller 
flocks during May and June at their two sites in British 
Columbia. Single birds were observed most frequenUy in 
these two months, and flocks of two were most common in 
July at both sites. 

Discussion 

Seasonal Patterns of Behavior and Activity 

Marbled Murrelets show consistent, seasonal patterns of 
activity and behavior. Throughout their range they exhibit the 
greatest levels of inland activity from April through August, 
with peak levels usually occurring from about the second 
week of July through early August. Hamer and Cummins 
( 1 990) suggested greater detection rates in late July may reflect 
increased food needs of nestling murrelets and the consequent 
increase in foraging trips by parent birds. Paton and Ralph 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



123 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 



>- 
LU 

> 

cc 
cc 

LU 

a. 






< 

LU 




EXPERIMENTAL FOREST 



l*l*ff*| f ?' ■ l^t |*f ■ » " | ''*» M I * 't " 



JAMES IRVINE TRAIL 



It, m v I II III ,■ i.i,i!i, 



LOST MAN CREEK 



■ilh.i iiiii, 



MAR MAY JUL SEP NOV JAN 

APR JUN AUG OCT DEC FEB 

ONE-THIRD MONTH PERIODS 

Figure 10 — Mean numbers of detections of Marbled Murrelets per survey (±s.e.) for one-third- 
month periods at three sites in northwestern California, 1989-1991 . Data from all years was 
combined for calculations. Asterisks (*) denotes periods in which no birds were detected during 
surveys; bars without standard error (vertical line segments) indicates only one survey was 
conducted during the period; all other blank columns represent weeks in which no data was 
collected. From O'Donnell 1993. 



(1988, 1990) felt the summer peak was likely the result of 
increased activity by breeding birds in the stand, perhaps in 
association with the fledging period, as opposed to an influx 
of non-breeding birds. However, many investigators have found 
that it is common among long-lived seabirds that defer sexual 
maturity for immatures to visit breeding sites later in the 
season in years prior to their first breeding attempt (Lack 
1968, Sealy 1976, Gaston 1990). Sealy (1976) found increasing 
numbers of subadult Ancient Murrelets (Synthliboramphus 
antiquus) visiting nesting colonies later in the breeding season. 
He found that attendance by subadults peaked by about one 
month after 90 percent of adults and newly-hatched young 



had departed to sea. Increased activity at breeding stands in 
July by Marbled Murrelets may indeed involve non-breeders 
investigating potential breeding sites. An increased presence 
by non-breeding birds later in the breeding season might also 
contribute to the increase in flock size noted by both O'Donnell 
(1993) and Rodway and others (1993b). 

In California, regular visitation at forest stands outside 
of the breeding season has been established. Fall and winter 
attendance has also been documented at several alcid colonies 
(e.g., Common Murre, Uria aalge, Razorbill, Alca torda, 
Black Guillemot, Cepphus grylle, Atlantic Puffin, Fratercula 
arctica, and Cassin's Auklet, Ptychoramphus aleuticus), 



124 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



O'Donnell and others 



Chapter 11 



Patterns of Seasonal Variation of Activity 



M 

X 140 



| 120 
H 

g 100 

M 

M SO 
Q 



O 



CO 
40 

d 

2 20 



a 



u 



1 i 



• NO DETECTIONS 






• • • • 



i i r 




I_J 



i-*-i i r 



APR MAT JUN JUL ADO SEP OCT MOV DEC JAH FEB MAR 
1989 1990 



^ 140 

a 

Z 

O 130 

H 

K 

U 100 

B 

I 



O 



o 

z 



80 
60 
40 
20 




• HO DETECTIONS 




T 1 l~ I " T 

* APR MAY JOH JUL AOO SEP OCT MOV DEC 



Ui 



Ll 



1990 



MONTH 



T 

JAM FEB MAR 
1991 



Figure 11— Mean number of Marbled Murrelet detections per week at Bloom's Creek Campground, 
California, in 1989-1990 (upper) and 1990-1991 (lower). Each weekly mean was calculated from 1-3 
intensive dawn inventories. Asterisks (*) indicate that a survey was conducted but no murrelets were 
detected; all other blank columns represent weeks in which no data was collected. From Naslund 1 993b. 



generally at the southern end of each species' range (Ainley 
and Boekelheide 1990, Greenwood 1987, Harris 1985, Harris 
and Wanless 1989, Taylor and Reid 1981, Sydeman 1993, 
Thoreson 1964). Harris and Wanless ( 1989) found a positive 
correlation between winter visitation and breeding success 
in the previous and following breeding season for a 
population of marked Common Murres. Winter attendance 
at breeding stands by murrelets may similarly relate to 
prior reproductive success, and might also enhance pair 
bond maintenance, facilitate earlier breeding (Carter and 
Erickson 1988), and reinforce familiarity with flight paths 
to the breeding stands. 

Two periods of very low (or no) activity occurred during 
March and from mid-August through early October. The 
pre-altemate molt period in California may begin as early as 
mid-February and extend through March (Carter and Stein, 
this volume). The relatively low level of March detections 
levels probably reflects this molt. Although murrelets do 



not molt flight feathers at this time (Carter and Stein, this 
volume), the increased energetic demands of molting body 
feathers could limit inland visits. The second period, from 
mid-August through early October coincides with the 
cessation of nesting and the molt into basic plumage. The 
more extensive nature of this prebasic molt (full body and 
simultaneous wing molt of the adults) is reflected in the 
longer period of time murrelets are absent from the forest. 
Numbers of above canopy behaviors closely mirror 
the patterns of total detection levels through the year, with 
the greatest levels occurring in the summer months and 
lower levels during the remainder of the year. Detections 
of birds below the canopy, however, coincide with the 
breeding season (April through September), and are largely 
absent outside of this period. Investigations of murrelet 
behavior around nest sites have consistently reported on 
observations of single birds and pairs flying below the 
canopy in the vicinity of nest trees (Nelson and Hamer, this 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



125 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 







BCCO HHRD JCAM 

STUDY SITE 



OCPA 



RLMF 



Figure 12 — Mean number (histogram bar) and standard error (line segment) of Marbled Murrelet 
detections by season at five study sites in central California. "Summer" includes April-July and 
"winter" includes November-February. Sample sizes (number of surveys) are indicated in histo- 
gram bars. From Naslund 1993b. 



100 




SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL ADO 

MONTH 

Figure 13 — Percent of surveys with detections of Marbled Murrelets by month for five study sites 
in central California, 1 989-1 991 . Surveys from all sites were pooled. Sample sizes (number of 
surveys) are indicated in the histogram bars. Missing histogram bar in September denotes that no 
murrelets were detected during 14 surveys. From Naslund 1993b. 



126 



USD\ Forest Service Gen. Tech. Rep. PSW-152. 1995. 



O'Donnell and others 



Chapter 11 



Patterns of Seasonal Variation of Activity 




(a) yo 



•2x3^<slo2o3u- 

MOMTH 



Cfi LANDNG OR STATIONARY CALLING 

% ORCLE BELOW CANOPY 

9 FLY THROUGH CANOPY 

$$ ORCLE ABOVE CANOPY 

■ FLY ABOVE CANOPY 



(b) 



5S 



n 


10 -7 t« 16 17 10 S 4 7 3 « 8 1 


* number* 


J.W0 3K!!I.I«A?I.S .2 1.1 VST ».! AT S.J 


ABOVE 


AAA A A A 
I B W I B I 3 B 3 

1 - D D D D D 
£ E E E E 


n 


10- 17 -1S' 16 V7 : !0 S ; 4 7 3 as 


x nuRibars 


a i s.2 t*.r*o.» rrjt zi.i .z o .40 .1 


BELOW 


i • A A A • e ; i e 



figure 14— (a) Percemtfdetectic<is of Marbled Murrelets in each riK>nto 

1 991 . The number of total detections in each month is shown above the bars, and the percent of "unknown" behaviors is not 
shown; (b) Results of Ryan-Einot-Gabriel-Weish multiple range tests comparing numbers of above and below canopy behaviors 
by marbled murrelets between months. Months with the same letter indicate that the mean number of detections were not 
significantly different from each other. The term "n" indicates the number of surveys in the respective month and " x numbers" 
is the mean number of detections per survey. Means presented are untransformed values. Surveys from all years were combined 
for the analysis, and months with less than three surveys were not included in the analysis. From O'Donnell 1993. 



volume a). Several studies (Naslund 1993a, Nelson and 
Hamer, this volume a; Nelson and Peck, in press; O'Donnell 
1993, Rodway and others 1993b, Singer and others 1991) 
have also noted the tendency of birds below the canopy to 
fly silently without vocalizing. The sharp seasonality of 
below canopy behaviors, in conjunction with these 
behavioral observations gathered at nest sites, strongly 
reinforces the relationship between below canopy behaviors 
and breeding activity. It should be noted, however, that 
murrelets have on rare occasions been observed flying 
below the canopy in habitat not considered suitable for 
nesting (e.g., Habitat Restoration Group 1992; Keitt 1991; 
Singer and others 1991, 1992). These were usually in areas 
adjacent to suitable habitat. 

Monitoring 

Current guidelines, as recommended in Ralph and others 
(1993), restrict surveys for management purposes to the 
breeding season. The survey season in California begins on 



15 April, in Oregon, Washington, and British Columbia on 1 
May, and in Alaska on 15 May. The survey season ends on 5 
August in all regions. The later start of the season at more 
northerly latitudes reflects a later breeding season in these 
areas (Kuletz and others 1994c; Hamer and Nelson, this 
volume a; Sealy 1974, 1975a). The timing of the survey 
season should of course maximize survey goals. Based on 
data collected in northwestern California (O'Donnell 1993), 
the recommended survey season for this state is a reasonable, 
if not slightly conservative, window for monitoring murrelets. 
Murrelets were detected during all surveys conducted at 9 
sites in April ( 16 of which were conducted before 15 April) 
throughout the study. Mean detection levels were slightly 
higher, however, during the last two-thirds of the month. It 
has been clearly established that numbers of detections, as 
well as above and below canopy behaviors, usually reach 
peak levels during July throughout the range of the species. 
To minimize the likelihood of failing to detect murrelets 
when they are actually present, Ralph and others (1993) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



127 



O'Donnell and others 



Chapter 1 1 



Patterns of Seasonal Variation of Activity 



(a) 



■ ABOVE CANOPY 

■ BELOW CANOPY 




APR 



MAY JUN 

MONTH 



JUL 



AUG 




(b) 



Figure 15 — (a) Mean size of Marbled Murrelet flocks observed above and below the 
canopy during the breeding season at James Irvine Trail, California, 1989-1991. 
Surveys from all years were combined. The numbers of surveys in each month are 
shown above the bars. An asterisk (*) denotes a significant difference (P = 0.05) 
between above and below canopy flock sizes in the respective month; (b) Results of 
Ryan-Einot-Gabriel-Welsh multiple range tests comparing above and below canopy 
flock sizes between months. Months with the same letter indicate that the mean flock 
sizes were not significantly different from each other. 



recommend that of four surveys conducted within a summer, 
two be conducted after 20 June, and at least one be conducted 
during the last three weeks of July. The earliest that birds 
were no longer detected at a stand in northwestern California 
was on 17 August (O'Donnell 1993). Detections of murrelets 
below the canopy, however, are absent earlier than this, and 
therefore 5 August is a reasonable termination date for the 
murrelet survey season in California. 

Naslund (1993b) speculates that the population of birds 
visiting breeding stands during the winter may consist of a 
higher proportion of resident breeders than during the summer. 
Therefore, she suggests that surveys conducted during the 



winter may actually monitor, for management purposes, the 
most important segment of the population (i.e., breeding 
birds). Until the relevance of winter numbers is established, 
however, surveys should continue in the breeding season. 

Acknowledgments 

We thank Jim Baldwin, Ann Buell, Alan E. Burger, 
Peter Connors, George Hunt, Debbie Kristan, S. Kim 
Nelson, Lynn Roberts, Michael Rodway, Jean-Pierre Savard, 
Fred Sharpe, and Sherri Miller for helpful comments on 
this manuscript. 



128 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 12 

Daily Patterns of Marbled Murrelet Activity at Inland Sites 

Nancy L. Naslund 1 Brian P. O'DonnelP 



Abstract: Patterns in the daily activity of Marbled Mumelets 
(Brachyramphus marmoraius) at inland sites has been studied through- 
out their range from California to Alaska. Murrelets are most active 
at inland sites around dawn, and to a lesser degree, at dusk. Throughout 
their range, peak levels of activity (detections) occur in the hour 
around dawn, but detections begin progressively earlier as one 
moves from south to north, corresponding to changing daylight 
regimes (e.g.. California: 45 minutes before to 75 minutes after 
sunrise: Alaska: 90 minutes before to 40 minutes after sunrise). 
Timing of dawn detections also varies seasonally in relation to 
changing sunrise times. The duration of morning activity periods 
varies seasonally, being longest (2 hours) during summer and short- 
est <<1 hour) in winter. In all areas, weather conditions affect the 
timing, duration, and level of murrelet activity. In general, activity 
tends to begin later, last longer, and reach peak levels on cloudy or 
foggy mornings. The frequency of different behaviors varies through- 
out the morning period of activity. Murrelets tend to fly below the 
canopy more before sunrise than after, and group sizes become 
larger after sunrise. Early detections tend to include more silent 
birds, solitary calls, and wing sounds than later detections. 



The relatively predictable changes in diurnal activity of 
birds have been well documented. Patterns of daily activity 
and behavior can vary widely between species. Knowledge 
of these activity patterns helps us understand avian ecology, 
develop appropriate survey techniques, and ultimately to 
manage threatened or endangered species. Relevant research 
questions include: How do activity levels and behaviors change 
between different times of the day? What are reasonable 
interpretations of temporal variation in behaviors? How should 
factors influencing variation in daily activity be used to 
interpret survey results? In this chapter, we examine the daily 
patterns of Marbled Murrelet (Brachyramphus marmoratus) 
activity at inland sites and how they are influenced by season, 
geographic location, and environmental conditions. Where 
applicable, patterns of subcanopy behaviors (i.e., murrelets 
occurring below canopy level), thought to be indicative of 
nesting, are also examined (see Ralph and others 1994). 

Methods and Results 

Data were collected primarily using general and intensive 
survey techniques (see Paton and others 1990. Ralph and 
others 1994). During these surveys, each time one or more 
murrelets were seen or heard, the event was recorded as a 



1 WUdlife Biologist. U.S. Fish and Wildlife Service, U.S. Department 
of Interior, 101 1 E. Tudor Road. Anchorage. AK 99503 

2 Wildlife Biologist, Pacific Southwest Research Station, USDA Forest 
Service. Redwood Sciences Laboratory, 1700 Bayview Drive. Areata. 
CA 95521 



"detection" (see Paton and others 1990). Additional results 
from other research activities in murrelet nesting habitat 
(e.g., nest searches, observations at nests) are presented. For 
example, data from Alaska are from intensive surveys and 
from "stake-outs." During the latter, only those murrelet 
detections that occurred within 100 m of the observer were 
recorded (see Kuletz and others 1994c, Naslund and Hamer 
1994). Only visual observations of murrelets were used in 
analyses of behavior (i.e., flying above or below canopy) and 
murrelet group size. All detections were used in other analyses. 

General Patterns of Daily Activity 

Murrelets are primarily active at inland sites around 
dawn and dusk. However, activity levels in the evening are 
lower and more sporadic than those during the morning. 
Nelson (1989) recorded that 12 percent of detections occurred 
at dusk and that murrelets were present on only 36 percent of 
dusk surveys in Oregon during the breeding season. In northern 
California, dawn activity was about five to six times greater 
than at dusk (Paton and Ralph 1988). Similar trends have 
been observed during the breeding season in British Columbia, 
Alaska, and at known nest sites in central California 
(Eisenhawer and Reimchen 1990, Kuletz 1991, Manley and 
others 1992, Naslund 1993a, Rodway and others 1993b). 
Anecdotal evidence for California and Alaska indicates that 
dusk activity also occurs during winter but may be less 
frequent than during the breeding season (Naslund, unpubl. 
data; Piatt, pers. comm.; Westphal, pers. comm.). 

In central California, two murrelet nests were monitored 
using video equipment and night viewing devices. Murrelets 
were not observed visiting nests during the night (i.e., 1 hour 
after sunset through 1 hour before sunrise; Naslund 1993a). 
Radar studies on Vancouver Island found no detectable flight 
activity by murrelets through the middle of the night (Burger 
1994. Burger and Dechesne 1994). 

Timing and Duration of Morning Activity 

In California and Oregon, murrelets were generally active 
between 45 minutes before and 75 minutes after official 
sunrise although most activity occurred during the hour 
around sunrise (Nelson 1989, Paton and Ralph 1988, Sander 
1987). Murrelets occasionally were detected prior to 45 
minutes before sunrise, but rarely more than an hour before 
(fig. 7; Naslund 1993a, Nelson 1989, Paton and Ralph 1988). 
Activity in Washington probably began earlier and lasted 
later than activity further south, and the peak activity period 
also occurred slightly earlier (Hamer and Cummins 1990, 
Hamer and others 1991). Murrelets in British Columbia 
typically became active up to about 75 minutes before sunrise 
(fig. 1; Manley and others 1992, Rodway and others 1993b). 
In southeast Alaska, most murrelets were detected between 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



129 



Naslund and O'Donnell 



Chapter 12 



Daily Patterns of Activity at Inland Sites 



20 



C/3 

Z 

o 

H 
U 

H 
- 
Q 

- 
O 

H 
Z 
W 
U 

- 

=- 



15- 



10- 



d 



l 



r 

-100-90 -80 -70 -60 -50 



f 
•30-20-10 



■ ALASKA 

E2 BRITISH COLUMBIA 

□ CALIFORNIA 





r ' ~r ■ t ' t ' T r ■ — r- — r — r— 
30 40 50 60 70 80 90 100110 



MINUTES BEFORE AND AFTER SUNRISE 

Figure 1— Timing of Marbled Murrelet detections relative to sunrise in California (n = 9764; Big Basin Redwoods State 
Park, 1989-1991; Naslund, unpubl. data), British Columbia (n = 2142; Phantom Creek, May-August 1990; Rodway and 
others 1991), and Alaska (n = 1649; Naked Island, May-August 1991; Kuletz and others 1994c) 



1 hour before and 1 hour after sunrise (Walsh, pers. comm.). 
In southcentral Alaska, murrelets were generally active 
between 90 minutes before and 40 minutes after sunrise, 
although the majority were detected during the 75 minutes 
before sunrise (fig. 1; Kuletz 1991, Kuletz and others 1994c). 
However, they sometimes were detected 120+ minutes before 
sunrise (Kuletz 1991; Naslund, unpubl. data). In Alaska, the 
timing of first detections varied during the breeding season. 
Murrelets were active earliest around the beginning of summer 
(fig. 2; Kuletz and others 1994c). 

In California, the duration of murrelet activity was longest 
during the breeding season (fig. 3; Naslund, unpubl. data; 
O'Donnell, unpubl. data). Conversely, their winter activity 
period was compressed and typically ended before sunrise 
(fig. 3). Murrelet activity occasionally began slightly earlier 
during winter than during summer (fig. 3). Murrelets tended 
to be active later in the morning and for shorter periods of 
time during August than in other summer months, in both 
California and Alaska (fig. 3\ Kuletz, pers. comm.; Naslund, 
unpubl. data; O'Donnell, unpubl. data). 

Timing and Duration of Evening Activity 

In California and Oregon, most evening detections 
occurred from about 20-30 minutes (but up to 90 minutes) 
before, through about 20-30 minutes after, official sunset 
(Fortna, pers. comm.; Nelson 1989; Paton and Ralph 1988). 
Murrelets have been detected up to 45 minutes past sunset 
in Oregon, and were rarely heard during the middle of the 
night in California (Nelson 1989; Strachan, pers. comm.). 
Timing of evening activity in British Columbia was slightly 



later than that observed farther south, with 95 percent of 
detections occurring between sunset and 45 minutes after 
sunset (Rodway and others 1993b). Elsewhere in British 
Columbia, murrelets were most active >45 minutes after 
sunset in early June (Eisenhawer and Reimchen 1990). 
Virtually all evening activity in Alaska has been detected 
after sunset, and murrelets occasionally fly inland throughout 
the relatively bright nights around the summer solstice 
(Kuletz, pers. comm; Naslund, unpubl. data). 

Weather Effects on Timing and Levels of Activity 

Weather has been observed to affect the timing and 
duration of activity throughout the murrelet' s range. Murrelet 
activity tends to begin later and last longer on cloudy or 
foggy mornings than on clear mornings (Kuletz, pers. comm; 
Manley and others 1992; Naslund, unpubl. data; Nelson 
1989; Nelson and Hardin 1993a; Paton and Ralph 1988; 
Rodway and others 1993b; Sander 1987). However, Nelson 
(1989) noted that murrelet activity in Oregon also began 
earlier and lasted longer on clear mornings than on mornings 
with intermediate cloud cover, though not longer than on 
mornings with 100 percent cloud cover. In Alaska, activity 
several hours after sunrise was associated with heavy fog at 
ground level or mist (Kuletz 1991; Walsh, pers. comm.). 

Environmental conditions can also affect levels of 
murrelet activity. At nest sites in central California, total 
numbers of detections and numbers of subcanopy behaviors 
tended to be higher when cloud cover was >80 percent, but 
was variable between sites (Naslund 1993a, unpubl. data). 
Rodway and others (1993b) also found that activity levels 



130 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Naslund and O'Donnell 



12 



Daily Patterns of Activity at Inland Sites 



-50 



- 




VI 




— 




X 


-60 


Z 




3 




s. 






-70 


o 




- 




- 




oa 






m 


V) 




- 




H 




3 




Z 


-% 




■ V I I | I I ll|lffffTT-rT7 »Y» T t I f T T'T" 

JUNE JULY AUGUST 

Figure 2 — Timing of first Marbled Murrelet detections relative to official sunrise on Naked Island, 
Prince William Sound, Alaska, in May-August 1991 (Kuletz, pers. comm.) 



LU 
LU 0) 

Pi 
8s 

is 



BO 

70 

60 

50 " 

40 " 

30 

20 

10 

-10 - 
-20 " 
-30 
-40 
-50 




v ♦ ♦ 



♦V 



r v^s v »*n 



--♦-• MEAN TIME OF FIRST DETECTION 

■ MEAN TIME OF MEDIAN DETECTION 
— •-- MEAN TIME OF LAST DETECTION 



/ \ 



sunrise 



•s 




v^*» 



i * * i * ' i ' ■ i ■ ■ i ■ ■ i ' ' i ■ ■ i ' ■ i ■ ■ i ■ ■ i ■ ■ i i i i 

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC 



Figure 3— Timing of first, median, anolasraeiectid 
northwestern California, 1989-1991 (O'Donnell, unpubl. data) 



relative to sunrise at Lost Man Creek in 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



131 



Naslund and O'Donnell 



Chapter 12 



Daily Patterns of Activity at Inland Sites 






were higher on cloudy (>80 percent cloud cover) than on 
clear (<80 percent cloud cover) days in British Columbia. 
Conversely, highest activity levels and detection rates have 
been recorded during clear (<25 percent cloud cover) and 
mostly cloudy (>75 percent cloud cover) mornings in 
Washington and Oregon (Hamer and Cummins 1990, Nelson 
1989, Nelson and Hardin 1993a). However, Hamer and 
Cummins also noted that activity (mean number of detections) 
was greatest during conditions of light drizzle. 

Other environmental factors also affect murrelet activity. 
Activity levels were high during periods of low cloud ceiling 
and decreased with increased wind speed and decreased 
temperatures during summer in Alaska (Kuletz and others 
1994c). Mean detection rates were highest during conditions 
of poor visibility (i.e., low visibility ratings corresponded to 
days with low cloud ceilings) (Hamer and Cummins 1990). 
When examining weather effects in more detail, it was found 
that murrelets in Oregon were most active when it rained at 
the beginning of the survey, or when it was foggy at the end 
(Nelson and Hardin 1993a). In Alaska, murrelets sometimes 
exhibited high activity levels during snowstorms, when cloud 
ceilings were low and wind was negligible (Naslund, unpubl. 
data; Piatt, pers. comm.). 

Weather may also influence the occurrence of activity 
around dusk. Although activity at dusk is infrequent in Alaska, 
murrelets were detected circling inland on two extremely 
foggy evenings (Kuletz 1991). 



Variation in Behaviors, Vocalizations, and Group Size 
During the Morning Activity Period 

Murrelet detections below canopy were more frequent 
than those above canopy, early in the morning activity period 
during the breeding season in northern California {fig. 4; 
O'Donnell, unpubl. data). The opposite was true after sunrise. 
It appears that this pattern remains intact throughout the 
year. The mean time that murrelets were seen in nest stands 
below canopy was significantly earlier than flight activity 
seen above canopy year-round in central California and 
during the breeding season in Alaska {table 1). 

Group size of murrelets seen flying in forested stands 
varies with time of day. In central California and Alaska, the 
mean time at which different group sizes were observed 
varied throughout the morning during the breeding and 
nonbreeding (California only) seasons {table 1). Individuals 
occurred earliest, pairs somewhat later, and groups (i.e., >3) 
latest. However, this trend was not apparent during the 
transitional period. Similar trends have been noted in British 
Columbia and Oregon. In these regions, most observations 
before 20 minutes before sunrise were single birds, whereas 
later detections included larger groups (Manley and others 
1992, Nelson and Hardin 1993a). 

Numbers of vocalizations made by murrelets also exhibit 
temporal variation during the morning activity period. Single 
calls were heard earlier than calls involving >6 calls/detection, 
although this was only significant during the breeding season 



z 

O 

I- 

o 

LU 



U_ 

o 



UJ 

o 

DC 
UJ 

a. 



20 -| 



15 - 



10 - 



5 ~ 



ABOVE CANOPY, n = 2,636 
BELOW CANOPY, n = 2,519 




I I I I I I IT 



MINUTES BEFORE AND AFTER SUNRISE 

Figure 4 — Timing of detections of Marbled Murrelets above and below the canopy relative to 
sunrise. Data presented are from Lost Man Creek in northwestern California, 1 989-1 991 . n is 
the number of detections for each behavior class (O'Donnell, unpubl. data) 



132 



USDA Forest Service Gen. Tech. Rep. PSW-I52. 1995. 



Naslund and CTDonnell 



Chapter 12 



Daily Patterns of Activity at Inland Sites 



Table 1 — Mean detection times < x ) and their standard deviation <sd), in minutes, relative to sunrise of Marbled Murrelet detection categories at Big Basin 
Redwoods State Park, California from 1989-1991, and at Saked Island, Alaska during 1992. Results of Tukey-Kramer range tests are shown. Means with 
different letters (in parentheses) are significantly (P < 0.05) different from each other 









Breeding 








Nonbreeding 




Transitional 








California 1 






Alaska 2 






California 1 






California 1 




Van able 


n 


X 


sd 


n 


X 


sd 


n 


X 


sd 


n 


X 


sd 


Bird height 


























Below 


714 


-«a) 


19.9 


252 


-37(a) 


21.5 


7 


-28(a) 


8.0 


63 


-3(a) 


14. 


Above 


1251 


8(b) 


20.4 


297 


-18(b) 


25.4 


33 


-17(b) 


10.1 


100 


3(b) 


15.1 


Group size 


























1 


1005 


-1(a) 


19.7 


226 


-34(a) 


22.3 


16 


-25(a) 


9.8 


94 


2(a) 


16.7 


: 


737 


6(b) 


22.5 


258 


-26(b) 


24.2 


13 


-19(ab) 


8.0 


55 


-Ha) 


12.4 


>3 


259 


10(c) 


18.2 


68 


-8(c) 


29.2 


13 


-13(b) 


li.: 


17 


Ka) 


1Z4 


No. calls 





























1248 


3(a) 


21.8 


126 


-38(a) 


18.9 


9 


-27(a) 


12.1 


106 


2(a) 


16.4 


1 


1761 


-4(b) 


23.6 


231 


-38(a) 


22.5 


291 


-27(a) 


15.4 


299 


13(b) 


16.3 


:-5 


2376 


-2(bc) 


23.5 


490 


-33(ab) 


25.6 


332 


-25(a) 


13.8 


274 


lOfbc) 


15.2 


6-9 


824 


0(c) 


23.6 


190 


-29(b) 


27.0 


148 


-23(a) 


12.8 


83 


-8(c) 


15.4 


>9 


4171 


0(c) 


22.6 


1004 


-29(b) 


24.5 


859 


-25(a) 


12.1 


457 


10(bc) 


14.7 


Type 


























Wings 3 


74 


-13(a) 


14.2 


36 


-48(a) 


13.0 


7 


-32(a) 


7.0 


7 


-19(a) 


8.6 


Heard 4 


8415 


-Kb) 


23.4 


944 


-33(b) 


21.8 


1602 


-25(ab) 


13.2 


1054 


-12(abi 


153 


Both 5 


751 


4(bc) 


19.5 


80 


-18(c) 


26.6 


31 


-16(ab) 


8.9 


60 


-l(bc) 


11.8 


Seen 6 


1174 


4(c) 


21.8 


83 


-32(b) 


17.5 


2 


-1Kb) 


14.1 


99 


3(c) 


15.9 



1 Naslund (unpubl. data) 
1 Kuletz (pers. comm.) 

3 Wings heard only, not seen 

4 Heard calling, not seen 

5 Seen and heard calling 

6 Seen, not calling 



(table I). In British Columbia, solitary calls were most frequent 
before sunrise (Manley 1992). Murrelets making only wing 
sounds were heard earlier than those heard vocalizing or 
those seen (table 1 ). This pattern was consistent year-round 
but was significant only during the breeding seasons in 
California and Alaska. Silent murrelets were also seen 
relatively earlier in Alaska than in California. This may 
partially be a function of greater light levels before dawn in 
Alaska, thereby making murrelets easier for observers to 
see. In British Columbia. Manley (1992) found that the 
occurrence of silent murrelets (including both single birds 
and pairs) peaked 20 minutes before sunrise. 

Discussion 

Daily Patterns of Activity and Behaviors 

Murrelets exhibit a primary period of inland activity 
around dawn and a secondary period around dusk. That 
murrelets are most active during the low light levels of dawn 
and dusk presumably reflects adaptation to predation pressures 
in the forest. Nesting murrelets and their chicks and eggs are 



vulnerable to a variety of avian predators including corvids 
and raptors (Brown, pers. comm.; Marks and Naslund 1994; 
Naslund and others, in press; Nelson and Hamer 1992, this 
volume b; Singer and others 1991 ). Crepuscular activity also 
allows for maximum diurnal foraging time. 

Variation in activity levels during the day appears to 
mirror aspects of murrelet nesting biology. Murrelets exchange 
incubation duties and exhibit peak feeding rates of young 
chicks around dawn (Hamer and Cummins 1991; Naslund 
1993a; Nelson and Hamer, this volume a; Nelson and Hardin 
1993a; Nelson and Peck, in press; Singer and others 1991, 
1992). Murrelets also sometimes exhibit flight behaviors 
around nests and feed chicks around dusk. They visit nests 
with young chicks infrequently mid-day, though diurnal 
feedings increase when chicks get older (Forma, pers. comm.; 
Hamer and others 1991; Naslund 1993a; Nelson and Hamer, 
this volume a; Singer and others 1991). 

Low detection levels at dusk may result from temporal 
differences in the composition and behavior of murrelets at 
inland sites. Fewer nonbreeders may fly inland during the 
evening activity period. Murrelets appear to fly silently while 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



133 



Naslund and O'Donnell 



Chapter 12 



Daily Patterns of Activity at Inland Sites 



carrying fish and are generally silent when visiting or flying 
around nests during the evening and are thus less easily 
detected (Naslund 1993a, unpubl. data). 

Activity levels relative to sunrise are notably earlier at 
northern latitudes (i.e., British Columbia and Alaska) than at 
more southern latitudes. This difference in activity periods 
results from differing light regimes. Pre-dawn light levels 
are greater and occur earlier, relative to sunrise, in Alaska. 
In this region, the seasonal variation in timing of first murrelet 
detections appeared to track changes in light levels. Murrelets 
were heard earliest, and occasionally throughout the "night", 
around the summer solstice when light levels were greatest 
(Kuletz and others 1994c). As summer advanced and light 
levels decreased, murrelet activity occurred increasingly later. 
Similarly, early activity in Washington and British Columbia 
is thought to result from longer twilight periods (Eisenhawer 
and Reimchen 1990; Hamer and Cummins 1990; Rodway 
and others 1991, 1993b). 

Cloudy or foggy weather results in lower light levels than 
clear mornings and may thus be affecting the timing of murrelet 
activity similar to changes in twilight regimes. In addition, 
murrelets may respond to periods of low fog or clouds, light 
rain, or snow by flying lower and calling more frequently and 
are thus detected more frequently under these conditions. 
However, on at least some occasions, murrelets fly above the 
fog, then drop below the fog just before entering the forest 
canopy (Kristan, pers. comm.). The influence of weather on 
murrelet activity is further evidenced by observations of 
murrelets exchanging incubation duties later on cloudy mornings 
and mornings with low cloud ceilings than on clear mornings, 
as well as changes in behaviors at nests with changes in 
weather conditions (Naslund 1993a, Nelson and Peck, in press). 

Although weather conditions apparently affect many 
aspects of murrelet activity, murrelets exhibit variable 
responses to conditions observed inland. This variability 
may reflect differences between weather conditions at survey 
sites and conditions that murrelets respond to down drainages 
and other flight corridors, or at the coast. Timing and 
duration of activity inland also reflects seasonal variation 
in environmental conditions. For example, activity is earlier 
and shorter in winter when days are shorter and 
environmental conditions more extreme than in summer. 
This presumably reduces the time available to murrelets for 
foraging, and may increase the effort required to obtain 
food. Consequently, less time and energy may be available 
for inland flights. Differences may also correspond to changes 
in social behavior or reduced numbers of birds in winter 
(see Naslund 1993a,b; O'Donnell and others, this volume). 
The late and reduced duration of activity observed in August 
corresponds to a time when detections become sporadic 
and decrease overall (Kuletz and others 1994c, Naslund 
1993a, Nelson and Hardin 1993a). 

Temporal variation in behavior, group size, and 
vocalization patterns of murrelets during the morning activity 
period reflects features of nesting biology. The early timing 
of single birds and birds flying below canopy coincides with 
the typical times that murrelets exchange incubation duties 



and display around nest sites (Naslund 1993a; Nelson and 
Hamer, this volume a; Nelson and Peck, in press; Singer and 
others 1991, 1992). Similarly, murrelets make single calls 
and wing sounds early in the morning. These behaviors have 
also been associated with incubation exchanges, chick 
feedings, and possible displays in nesting territories (Naslund 
1993a; Naslund and Hamer 1994; Nelson and Hamer, this 
volume a; Nelson and Hardin 1993a). Conversely, the larger 
and more vocal groups that are more frequent later in the 
morning may represent murrelets engaged in social 
interactions or joining together for flights to sea. 

Survey Implications 

Based on the daily activity patterns described here for 
murrelets, it is clear that current guidelines, which recommend 
that surveys be conducted during the dawn activity period, 
will provide the most consistent information on use of inland 
habitat by nesting murrelets (see Ralph and others 1993, 1994). 
Evening surveys may furnish additional information useful 
for interpreting stand-use or furthering our understanding of 
murrelet biology. It is evident that survey start-times should 
be shifted earlier as one moves north to compensate for changes 
in light levels relative to sunrise. Exact timing for some areas 
(e.g., southwest Alaska) may require further evaluation. 

It is difficult to standardize surveys in a manner which 
eliminates the contribution of weather conditions to daily 
variation in activity patterns. Variability in activity is further 
confounded by the effects of weather conditions on the ability 
to detect murrelets. For example, fog and rain may reduce 
observers' abilities to see or hear murrelets. However, Rodway 
and others (1993b) found no evidence that some weather 
conditions (e.g., cloud cover) affect the proportion of detections 
that are seen. Avoiding surveys during certain conditions 
(e.g., heavy rain), as recommended by current guidelines 
(Ralph and others 1993, 1994), will reduce variation in recorded 
activity due to differences in visibility. This can be particularly 
important when evaluating subcanopy behaviors, which relies 
primarily on the visual detection of murrelets. In Alaska, 
where inclement weather prevails, surveys may be conducted 
on all days except those with high winds and extreme rain. 
Weather effects should be considered accordingly when 
making temporal and spatial comparisons between surveys. 

Collection of data on group size, behaviors, and 
vocalizations during surveys provides information that is 
important for interpreting stand-use by murrelets. These 
data may also prove useful for unraveling various aspects of 
the ecology of this enigmatic species. 

Acknowledgments 

We thank Kathy Kuletz, Peter Walsh, and Michael 
Westphal for generously allowing us access to their 
unpublished data. This manuscript was greatly improved by 
the insightful comments of Jim Baldwin, Alan Burger, Peter 
Connors, David Fortna, Anne Harfenist, Gary Kaiser, Debbie 
Kristan, S. Kim Nelson, Peter Paton, John Piatt, Michael 
Rodway, Jean-Pierre Savard, and Fred Sharpe. 



134 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 13 

Interannual Differences in Detections of Marbled Murrelets 
in Some Inland California Stands 



C. John Ralph 1 

Abstract: I compared the mean level of detections of Marbled 
Murrelets by month over five years at three inland sites in northern 
California. These areas all have relatively high levels of detec- 
tions. There were no significant differences in mean detection 
levels year to year at any site, and for any month with the excep- 
tion of April at one site. This lack of evidence for significant 
interannual variation in the number of detections of birds suggests 
that data from any one of the years would have been sufficient to 
detect occupancy of these stands by Marbled Murrelets. Caution 
must be used in applying this result, as interannual variation in 
detection rates may be greater at sites with relatively few birds, 
and only three sites were investigated in this study. 



Most species of birds vary in the proportion of birds 
breeding among years, with profound effects upon the 
demography of the species. In the case of the Marbled 
Murrelet, it would be useful to know the proportion of the 
population breeding. This knowledge would help determine 
if surveys taken in different years are comparable for purposes 
of determining the occupancy status of stands proposed for 
timber harvest. Changes in the number of murrelets detected 
in a stand during the breeding season are assumed to be 
related to changes in the number actually breeding in the 
stand. In this study, I compared the detection rates of murrelets 
at three sites for evidence of year-to-year variation. Finding 
a significant difference would indicate that surveys in any 
one year might not detect birds in a stand that would have 
had birds in another year, especially a stand with a relatively 
low detection rate. Although detection rates are not equivalent 
to numbers of birds actually breeding in a stand (Paton, this 
volume), I make the assumption that they are analogous. 

Methods 

I examined the among-year variations for three areas 
with moderate and high detection levels (table 1) in northern 
California: Lost Man Creek, in Redwood National Park, 
Humboldt County; James Irvine Trail, in Prairie Creek 
Redwoods State Park, Humboldt County; and Redwood 
Experimental Forest, near Klamath, Del Norte County. These 
three survey sites all are located within large contiguous 
stands of old-growth redwood in a natural reserve and parks. 

Data used in this analysis were total number of detections 
(both audio and visual) per survey for each study site for the 



1 Research Wildlife Biologist, Pacific Southwest Research Station, 
USDA Forest Service, Redwood Sciences Laboratory, 1 700 Bayview Drive. 
Areata. CA 95521 



years 1989-1993. Surveys at these sites were conducted 
according to Marbled Murrelet survey protocol (Ralph and 
others 1993). Only data from April through August of each 
year were used, the recommended murrelet survey period in 
the protocol. 

Data were analyzed using one-way ANOVAs (a < 
0.05), for each month and year with surveys (table 1). The 
number of birds detected in a morning's survey were log 
(count + 1) transformed to approximate normality of the 
distribution of detections. 

Results 

The detection rate was highest at Lost Man Creek with 
monthly means ranging up to 240 detections in July 1990 
(table 1). James Irvine Trail had fewer detections with a 
maximum average of 146 in July 1990. Experimental Forest 
had the lowest rate, with a maximum average detection rate 
of 111 in July 1993. 

I first compared each site separately by month. An 
inspection of the average number of detections of murrelets 
(table 1) shows that months in a given year, even with only a 
few samples, were generally very similar to the averages for 
that month in the other years with more robust samples. 
Monthly means were not significantly different at any site, 
with the exception of April at Lost Man Creek (P = 0.004). 
This month had a larger range of mean detections than in 
other months or at other sites. Comparing among years at 
Lost Man Creek in April, I found that 1990 and 1991 were 
similar, but that 1989 and 1992 were both different from 
each other, as well as from other years (Ryan-Einot-Gabriel- 
Welsch multiple comparison test). 

Discussion 

Though only one month was significantly different over 
a five-year period at three sites, it is quite likely that further 
data would show that detections are lower in certain years at 
specific sites. 

Particularly unseasonable weather during the breeding 
season could impact numbers of inland detections at specific 
sites. Fluctuations of prey fish populations may also be a 
factor in inland murrelet detection levels. Warmer ocean 
temperatures associated with an El Nino event are 
responsible for changing local and global weather cycles 
that affect many species of marine animals, including 
nesting seabirds and their food (Ainley and Sanger 1979). 
The ocean temperature events may also affect Marbled 
Murrelet prey (Burkett, this volume), although this has not 
been documented. The effects of warmer offshore water 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



135 



Ralph 



Chapter 13 



Interannual Differences in Detections 



5 
a 



2 



■ 

e 



•Sf 

! 

I 

M 



"5 






l 



3 
1 



e 
| 

r 

I 



85 



CN 

Sn 



85 



•j 
S 



on 

00 
On 



2 


w-> 

NO 

* 




co 


ON 


o 


o 


o 


o 


o 


3 


On 


o 

CO 


On 


00 


no 


o 


~- 


— 


o 



5 



o 

CO 
NO 



» 3 



t> a 



O p 
<S CO 

CO — ' 



o <-. 

si § 



o r- no 
~ no t~ 

CN ■* CO 



- p o 

si 5 3 



m 


!■» 


ro 




CO 


C 


~- 


~* 


n 



P o 
cs 



r- ~ 



•5 wo 

no — 
CO — ' 



o 

CN 



O 



« CM 
CN CM 
— CN 



n 

o> 


NO 

cn 


NO 
00 


5 


■a 

n 
CO 


o 


o 


o 


© 


d 


a 


r- 


oo 




cn 

CO 


o 


J 


o 


o 


w 



wo — 



CN 

© 

CO 



— **> 



"5 

CO 



■* 



o 

On 



o 

NO 



CO 

d 



OO WO 00 wo 



o 
e 


o 


s 


d 


o 


c 
-t 


00 
CO 


CO 

00 


o 


q 



wo c— 

. •* ON 

I CM NO 



ON On 

— o 



I — ■* 



9 



s 

d 



— °n 



O oo ^ 

1 On r^ 2 

* <*> — 



CN 



I I I 



o 
d 



l I l l l 



l I l l l 



ro 

d 


wi 


o 
•>* 

CO 


wo 

NO 

CN 


CO 
CO 


On 

NO 


2 


CnI 

00 


CO 
CM 


r-- 


00 


w-i 
wi 


* 

00 


CO 

d 

CO 


* 

CN 


nC 
CO 
00 


00 

s 


00 


OO 
ON 
CO 

cn 


i 


•<* 
00 
CO 


wi 

Cn) 
MO 


wo 




q 

wS 

CO 


00 

t-' 

CM 


NO 


o 
cn 


o 
r- 


d 



o ■* 

CN CO 



W0 NO — 

NO « NO 



o 



c 
o 



■5 
a 
o 

2 














t« 


1 


B 


>> 


•U 




1 


a 


o. 

< 


3 


3 

— » 


3 


3 
< 



1 fr 8 * 

< S I 5 



3> 



I 

3 
c 
<u 

S 
•c 

I 

x 



£ 

f 



•C 
o- 
< 



a 

3 
< 



B 

3 

5 " 

_ t ■ 
u u B 



— (S m 









136 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Ralph 



Chapter 13 



Interannual Differences in Detections 



temperatures during an El Nino event may cause a reduction 
in murrelet breeding effort, and thus influence inland 
detection levels. The current El Nino has become the 
longest on record, beginning in early 1991, or perhaps 
even earlier. 

All of the sites studied had relatively high murrelet 
activity, as compared to many sites elsewhere in the Pacific 
Northwest. This may have had an effect of moderating 
differences if social facilitation is a factor in levels of 
murrelet activity. However, we have no data at present to 
support such a supposition, although Shaughnessy (pers. 
comm.) and Nelson (pers. comm.) found differences 
between years when comparing murrelet use at a site. 
Also, there is some evidence that detections vary as a 
function of weather (Naslund and O'Donnell, this volume). 
For example, there are frequently more detections on foggy 
mornings. Thus, a year in which low detection rates would 
have been expected might instead have normal detection 
rates because of unusually foggy weather in that year. 
However, the amount of daily variation induced by clouds 
in our studies has been less than 20 percent (O'Donnell, 
pers. comm.). 

The great variation between mornings at most sites 
might be the key to the lack of significant difference among 
years. However, the fact that the monthly average values 
were quite similar indicates that no differences exist. 



I was unable to find any evidence that would suggest 
that the number of detections of birds was consistently 
lower or higher in any one of the five years. Therefore, 
results of inland surveys used to determine presence or 
absence of Marbled Murrelets in proposed timber harvest 
stands would likely have been valid in any of these years in 
this area of California. Caution must be used in applying 
these data to other sites and regions, however, as only three 
sites were surveyed, and the variance was large. 

I suggest that we need continued monitoring of murrelets 
at established sites over several years, combined with careful 
quantification of the many influences on inland detection 
levels, to fully resolve the indications derived from this 
study. This effort would greatly increase our understanding 
of this bird and its use of inland habitats. 

Acknowledgments 

I am very grateful to the biologists who have worked in 
the early dawn over the years to put together this data set. 
Especially noteworthy are Sherri Miller, Brian O'Donnell, 
and Linda Long. I thank Robin Wachs for her excellent help 
in tabulating and analyzing these data. I also thank Jim 
Baldwin, Ann Buell, George Hunt, Debbie Kristan, Kim 
Nelson, Peter Paton, and Meg Shaughnessy for helpful 
comments on the manuscript. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



137 



Chapter 14 

A Review of the Effects of Station Placement and Observer 
Bias in Detections of Marbled Murrelets in Forest Stands 



Brian O'Donnell 1 



Abstract: A variety of factors influence the results of surveys 
conducted for Marbled Murrelets (Brachyramphus marmoratus) 
in the forest. In this paper we examine observer variability and 
survey station placement as factors influencing murrelet survey 
data. A training and evaluation protocol (Ralph and others 1993) 
was developed to insure high field abilities and comparability of 
data among and between observers. Site characteristics which may 
limit the hearing and sighting of murrelets (e.g., wind, road, or 
stream noise, visual obstructions) can largely be controlled through 
the sensible placement of survey stations. 



In order to interpret data from inland surveys of Marbled 
Murrelets (Brachyramphus marmoratus), we must be aware 
of those factors influencing numbers and types of detections 
at a site. Topography, stand size and shape, and other factors 
that may influence murrelet densities and habitat use are 
examined elsewhere in this volume, as are temporal influences 
on detections and behavior levels. If two areas, each having 
equivalent populations of murrelets, were surveyed, why 
might the survey data differ between the two sites? Variability 
within and among persons conducting surveys is clearly one 
factor influencing survey data. Levels of extraneous noise or 
visibility at survey stations may also differ between sites. 
This chapter examines the influences of observer and survey 
station placement on murrelet survey data. 

Observer Variability 

There has been little quantitative analysis of observer 
effects on murrelet survey results. Rodway and others (1993b) 
found that the numbers of detections recorded by pairs of 
observers at the same sites showed significant positive 
correlation. However, they also found significant differences 
between observers in the proportion of visual detections. In 
Ralph and Scott ( 1 98 1 ), there are studies on observer effects 
on landbird censusing results. Ralph (pers. comm.) compared 
murrelet surveys by many observers from a high-activity 
site in northwestern California to find the area around the 
observer that is effectively surveyed. Kuletz and others 
( 1 994c ) found significant observer effects on detection levels 
at sites in Alaska. 



1 Wildlife Biologist, Pacific Southwest Research Station. USDA Forest 
Sen-ice. Redwood Sciences Laboratory, 1700 Bayview Drive. Areata, 
CA 95521 



Site Characteristics 

Visibility at Survey Station 

A high percentage of murrelets remain unseen to the 
observer during surveys (Paton and Ralph 1988; Nelson 
1989). The numbers of visual sightings of murrelets are 
strongly influenced by the location of the observer, yet 
they are critical for determining murrelet use of a stand 
(Paton, this volume). Nelson (1989) reported the highest 
percent of visual detections (55 percent) occurred at a 
survey station with the greatest view of open sky. Rodway 
and others (1993b) detected murrelets visually in 19 percent 
and 26 percent of detections at two sites in British Columbia. 
They speculate that greater visibility accounted for more 
visual detections at the latter site. Ralph (pers. comm.) 
examined the effect of canopy cover upon detection and 
behavior levels to assess the level of survey effort needed 
to determine occupancy by breeding murrelets in a stand. 
Detections indicating probable nesting in a stand are of 
murrelets below or within the canopy, and require visual 
sightings of the birds (Paton, this volume). At a dense 
canopy site, only 3 percent of detections indicated 
occupancy, while at moderate and open canopy sites, these 
detections were 14 percent and 30 percent, respectively. I 
(O'Donnell 1993) found that, in general, sites with greater 
visibility had more detections of murrelets below the canopy 
(fig. 1). Visibility at each site was quantified by estimation 
of the percent of clear view to the horizon in all directions. 
Including only sites in or near old-growth stands, I found 
that visibility had a significant positive relationship (r = 
0.73, P = 0.04, n = 8) with the numbers of below canopy 
behaviors observed. 

Environmental Noise 

While there are no studies which examine the effects of 
extraneous noise (e.g., wind, road, or stream) specifically on 
murrelet survey results, it seems very likely that any noise 
impairing an observer's ability to hear murrelets will be 
detrimental to survey goals. The effects of environmental 
noise on the audio detection of landbirds are discussed in 
papers in Ralph and Scott (1981). 

Environmental Acoustics 

Acoustical properties in the environment will degrade 
bird song and calls in a variety of ways (Richards 1981). 
Attenuation, the decrease in intensity of sound with distance, 
can be affected by habitat type. Kuletz and others (1994c) 



USDA Forest Service Gen. Tech. Rep. PSW-I52. 1995. 



139 



O'Donnell 



Chapter 14 



Effects of Station Placement and Observer Bias 



• ABOVE CANOPY ACTIVITES 
D BELOW CANOPY ACTIVITIES 



35-i 



(A 

Z 


30 


o 




O 


25 


UJ 




1- 




UJ 

Q 


20 


u. 




O 


15 


H 




T« 




LU 


10 


U 




rr 




Ol 


5 


0. 



B 

D 



I* I I I | I I I I | I I I I | I I I I | I I I 1 | I I I I | 

10 20 30 40 50 60 

PERCENT OPEN SKY 



Figure 1 — Relationship between percent of open sky at survey stations and 
observations of marbled murrelets flying above and below the canopy. Data from 
nine sites in northwestern California. 



detected murrelets by sound at greater distances from stations 
placed in open meadow, than at stations closely surrounded 
by forest, accounting for difference in numbers of detections 
between sites. Detection distances for murrelets that were 
heard and not seen varied considerably between sites in 
northwestern California (O'Donnell, unpubl. data). The 
locations of stations ranged from closed canopy forest to 
large, open prairies. The maximum detection distances at 
stations in more open areas was generally greater than for 
stations within the forest. 

Discussion 

Observer Variability 

There can be little doubt that variability exists between 
observers in their ability to see and hear murrelets. While 
some differences between observers cannot be eliminated, 
adequate training and evaluation can greatly improve 
individual abilities and increase comparability between 
observers. A training and evaluation protocol (Ralph and 
others 1993) was developed to accomplish this goal. The 
training program helps trainees to develop their ability to 
detect murrelets in the forest and to accurately record 
observations according to protocol. The evaluation, a 
simultaneous survey conducted by trainees and a qualified 



evaluator, insures that trainees are able to survey murrelets 
at acceptable levels of proficiency. In California, Oregon, 
and Washington, all persons conducting murrelet surveys 
for management or research purposes must successfully 
complete an evaluation process. Since inception of the training 
and evaluation program in 1991, approximately 500 persons 
have been evaluated in California alone (Burkett, pers. comm.), 
providing a large pool of qualified observers. 

Site Characteristics 

Factors influencing an observer's ability to hear and see 
murrelets can largely be controlled by sensible placement of 
survey stations. Because such a high proportion of murrelets 
are detected by call alone, survey stations should not be 
placed near sources of loud noise. Similarly, since behaviors 
suggestive of breeding activity are determined primarily 
from visual observations of murrelets, it is important to 
place survey stations in areas that have the greatest view of 
open sky (Ralph and others 1993). 

Acknowledgments 

I thank Steve Courtney, Dave Fortna, Debbie Kristan, S. 
Kim Nelson, Peter Paton, C. John Ralph, and Sherri Miller 
for comments on this manuscript. 



140 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 15 

Inland Habitat Suitability for the Marbled Murrelet in 
Southcentral Alaska 



Katherine J. Kuletz Dennis K. Marks Nancy L. Naslund 



Nike J. Goodson Mary B. Cody 1 



Abstract: The majority of Marbled Murrelets (Brachyramphus 
marmoratus) nest in Alaska, where they sometimes nest on the 
ground, and their nesting habitat requirements are not well under- 
stood. The inland activity of murrelets was surveyed, and habitat 
features measured, between 1991 and 1993, in Prince William Sound. 
Kenai Fjords National Park and Afognak Island, Alaska (n = 262 
sites). We used these data to develop statistical models that explain 
variation in murrelet activity levels and predict the occurrence of 
occupied behaviors (indicative of nesting), based on temporal, geo- 
graphic, topographic, weather, and habitat characteristics. Multiple 
regression analyses explained 52 percent of the variation in general 
murrelet activity levels (P < 0.0001). The best model included sur- 
vey date, location relative to the head of a bay, elevation, slope, 
aspect, percentage of forest cover, tree diameter, and epiphyte cover 
on tree branches. The highest activity levels were associated with 
late July surveys at the heads of bays where there was high epiphyte 
cover on trees. Stepwise logistic regression was used to identify 
variables that could predict the probability of detecting occupied 
behaviors at a survey site. The best model included survey method 
(from a boat, shore, or upland), location relative to the head of a bay, 
tree diameter, and number of potential nesting platforms on trees. 
The best predictors for observing occupied behaviors were tree 
diameter and number of platforms. In a jackknife procedure, the 
logistic function correctly classified 83 percent of the occupied 
sites. Overall, the features indicative of murrelet nesting habitat 
include low elevation locations near the heads of bays, with exten- 
sive forest cover of large old-growth trees. Our results were derived 
from surveys designed to estimate murrelet use of forested habitat 
and may not accurately reflect use of nonforested habitat. There- 
fore, caution should be exercised when extrapolating observed trends 
on a broad scale across the landscape. 



The reliance of Marbled Murrelets (Brachyramphus 
marmoratus) on mature and old-growth forest for nesting 
has been well established in the southern portion of the 
species' range (see Carter and Morrison 1992; Hamer and 
Nelson, this volume b). Yet, the majority of Marbled Murrelets 
breed in Alaska, where nesting habitat requirements are not 
clearly understood (Mendenhall 1992). Offshore surveys 
suggest that about 97 percent of the population within Alaska 
occurs offshore of lands with at least some old-growth forest 
cover (Piatt and Ford 1993). These forested areas extend 
from southeast Alaska, north along the Gulf of Alaska, and 
throughout southcentral Alaska. However, the extent of 
forested habitat is variable in this region. "Forested" areas 
include unforested habitat, and tree line may extend only 
200 m above sea level and a few kilometers inland. 



1 Wildlife Biologists, Migratory Bird Management, U.S. Fish and 
Wildlife Service. U.S. Department of Interior, 101 1 E. Tudor Road. An- 
chorage, AK 99503 



The choice of nesting habitat for murrelets appears 
superficially to be broader in Alaska, where murrelets nest 
both in trees and on the ground, than at lower latitudes. 
Before the early 1980's, only six Marbled Murrelet ground 
nests had been found (Day and others 1983). Since then, three 
tree nests have been documented in southeast Alaska, and one 
nest was found on a tree root overhanging a cliff (Brown, 
pers. comm.; Ford and Brown 1994; Quinlan and Hughes 
1990). In southcentral Alaska, 15 tree nests and seven additional 
ground nests were found between 1989 and 1993 (Balogh, 
pers. comm.; Hughes, pers. comm.; Kuletz and others 1994c; 
Mickelson, pers. comm.; Naslund and others, in press; Rice, 
pers. comm.; Youkey, pers. comm.). The apparent importance 
of ground nesting by murrelets in Alaska is partially an artifact 
of effort. Ground nests are more easily discovered than tree 
nests, inflating their relative numbers. Additionally, it is 
possible that ground nests of the Kittlitz's Murrelet (B. 
brevirostris) can be mistaken for those of Marbled Murrelets 
(Day and others 1983). Therefore, it was unclear how important 
ground nesting was to the Marbled Murrelet population. 

Following the 1989 Exxon Valdez Oil spill, the protection 
of habitat was identified as a means of restoring injured resources 
such as the Marbled Murrelet. Our goal was to provide 
information on murrelet nesting habitat in the spill zone to 
guide protection and land acquisition decisions. Between 1990 
and 1993, we examined aspects of murrelet nesting behavior 
and habitat use in Prince William Sound and Kenai Fjords 
National Park (Kuletz and others 1994b, c). Concurrently, in 
1992, murrelet surveys were conducted on Afognak Island, 
north of Kodiak Island (Cody and Gerlach 1993, U.S. Fish 
and Wildlife Service 1993). Although there were differences 
in study design among the studies, they provided a substantial 
data base for relating habitat variables to murrelet activity 
throughout the spill zone. Data from these four studies were 
combined to develop a broad-based model of murrelet activity 
in relation to weather, season, and habitat variables that 
would apply throughout southcentral Alaska. We also 
developed a statistical model of site characteristics where 
occupied behavior, indicative of nesting birds, was observed. 

Methods 

Study Area 

The study area encompasses the Naked Island group in 
central Prince William Sound, western Prince William Sound, 
the Kenai Fjords National Park, and two parcels on Afognak 
Island (fig. 1). Brachyramphus murrelets comprise a large 
portion of the avifauna in these areas. The estimated 
Brachyramphus murrelet population for Prince William 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



141 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 




Naked L 
Area^c^ 




western Prince 
William Sound 



Afoqnak I 



^l 



Q^g^KodMc.. 



Kilometers 

o 20 40 



SCz Alaska \ 


^ ^ 






O^L* 


/Cffi* 7 




y 


d? 




^STUDY AREA 



Figure 1 — The four study areas of southcentral Alaska surveyed for inland murrelet activity between 1 991 and 1 993: Naked Island, western Prince 
William Sound (PWS), Kenai Fjords National Park (KFNP), and Afognak Island (in two parcels). 



Sound is approximately 100,000 birds (Klosiewski and Laing 
1994). Within 5 km of the Naked Island group (Naked, 
Peak, and Storey islands), there are an estimated 3,000 
Marbled Murrelets (Kuletz and others 1994a). At-sea surveys 
of Kenai Fjords National Park have been restricted to shoreline 
surveys (within 200 m of shore) and complete counts in 
some bays. In 1989 the estimates ranged from 2,000 
Brachyramphus murrelets in June to 6,500 in August (Tetreau, 
pers. comm.). At-sea surveys off Afognak Island in summer 
1992 produced estimates of 2200 murrelets off the northern 
section, and 2000 murrelets off the southwest section (Fadely 
and others 1993). Brachyramphus murrelet population 
estimates include a small percentage of Kittlitz's Murrelets 
in Prince William Sound (approximately 7 percent; Laing, 
pers. comm.) and Kenai Fjords National Park (between 7- 
12 percent; Tetreau, pers. comm.). 



General Habitat 

Prince William Sound, the northernmost portion of the 
study area, is characterized by protected waters, numerous 
islands and bays, and deep-water fjords, including some 
with tidewater glaciers. Forested areas of mixed hemlock- 
spruce forests (Tsuga mertensiana, T. heterophylla, and Picea 
sitchensis) are interspersed with muskeg meadows, alpine 
vegetation, and exposed rocks. Tree line ranges from 30 to 
600 m (see Isleib and Kessel 1973). Naked Island is in the 
center of Prince William Sound, and vegetation is a mix of 
forest and muskeg meadow, but lacks other habitat types 
(Kuletz and others, in press). 

The Kenai Fjords National Park, on the southern Kenai 
Peninsula, is characterized by steep, rugged coastline and 
numerous islands on the outer coast. There are protected 
waters and tidewater glaciers at the heads of fjords, and 



142 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



exposed coasts near fjord mouths bordering the Gulf of 
Alaska (Bailey 1976). Glaciers cover more than 50 percent 
of Kenai Fjords National Park (Selkregg 1974). Because of 
receding glaciers, forested portions of the coast are primarily 
in the outer, more exposed headlands and islands. Tree line 
is typically 300 m, and few areas beyond 500 m from shore 
are forested. Tree species are similar to those in Prince 
William Sound, and alder is the dominant vegetation in 
unforested areas. 

There were two study sites on Afognak Island. The 
northern parcel faces north into the Gulf of Alaska and is 
heavily forested. The southwest parcel faces west into Shelikof 
Strait and is primarily unforested, except along river valleys 
and around the heads of bays. There are no glaciers. Tree 
line ranges from 100 to 300 m and the only conifer is Sitka 
spruce (Picea sitchensis), which tends to be larger than on 
the mainland. 

Data Collection 

Dawn Watch Surveys 

In Alaska, surveys are limited by logistic considerations 
due to inaccessibility of coastal habitats, and by the relatively 
short time available for breeding surveys (mid-May through 
early August). Therefore, intensive surveys (hereafter referred 
to as "dawn watches"; Paton and others 1990, Ralph and 
others 1993) were conducted from land-based ("upland") 
sites and from boats anchored near shore. The basic unit of 
measure was the 'detection' which is defined as "the sighting 
or hearing of a single biid or a flock of birds acting in a 
similar manner" (Paton and others 1990). We assume that 
dawn activity (i.e., numbers of detections) is positively related 
to nesting activity. We recognize, however, that no quantitative 
relationship between dawn activity and numbers of nesting 
murrelets has been defined, and conclusions about relative 
use of different habitats are tentative. 

Dawn watches were modified for southcentral Alaska 
(for more details see Kuletz 1991b, Kuletz and others 1994c). 
Modifications included: (1) earlier start and finish times 
relative to sunrise (i.e., usually beginning 105 min before 
official sunrise and lasting until 15 min after sunrise, or 15 
min after the last murrelet detection) to compensate for 
greater light levels in Alaska; (2) addition of behavior 
categories not observed further south; and (3) some watches 
were conducted from boats and shore to allow sampling of 
shoreline habitat. Using landmarks, we designated each 
detection as <200 m or >200 m from the observer. When the 
dawn watch was conducted near the water, a bird passing 
over land at any time during the observation was designated 
a land detection. 

Behaviors indicative of murrelet nesting are referred to 
as "occupied behaviors." These included flying below canopy, 
emerging from or flying into trees, landing on or departing 
from a branch, or calling from a stationary point in the forest 
(Paton and others 1990). In unforested areas we considered 
flights into hillsides or brush or <3 m above ground cover to 
be occupied behaviors. Occupied sites were those with at 



least one recorded occupied behavior. We considered other 
sites to be of "unknown status" since a single visit was not 
sufficient to determine whether a site was unoccupied (Ralph 
and others 1993). 

Habitat Variables 

A 50-m vegetation plot was sampled at each dawn watch 
site. When the dawn watch was conducted from shoreline or 
from a boat, the vegetation plot center was placed within the 
habitat most visually representative of the area adjacent to 
the dawn watch site. Within the plot we measured the diameter 
at breast height (d.b.h.) of the 10 nearest upper canopy trees, 
the percentage of epiphyte cover on the branches of each 
tree, and the number of platforms per tree (horizontal surfaces 
>15 cm diameter and >10 m above the ground). Data on 
epiphyte cover and platforms were not collected for the 
Naked Island group. We also made visual estimates of overall 
canopy height, percentage canopy closure, and percentage 
of forested area. Slope grade, aspect, and elevation were 
measured on site or from topographic maps. Distance from 
the ocean was measured from aerial photographs. Each site 
was classified as either exposed coastline, semi-protected in 
a bay, or at the head of a bay. 

Study Design Sampling and Analyses 

The Naked Island group was surveyed between 10 June 
and 11 August 1991 (n = 69 sites). Sites in western Prince 
William Sound were surveyed between 15-18 July 1991 (n = 
9) and 12 June-3 August 1992 (n = 68). Afognak Island was 
surveyed from 4 June-5 August 1992 (n = 76). Kenai Fjords 
National Park was surveyed from 8-29 July 1993 (n = 40). 
We surveyed Marbled Murrelet activity and recorded weather, 
survey period, and topographic and vegetation variables at 
each survey site in the four study areas. Murrelet activity is 
highly seasonal and generally exhibits a pattern of peak 
activity during the breeding season (Hamer and Cummins 
1991, Nelson 1989, Rodway and others 1993b). Therefore, 
survey period was categorized as early and late (before or 
after 10 July, respectively), based on activity patterns 
previously documented in Prince William Sound (Kuletz 
and others 1994c). Study designs and survey methods varied 
among areas (for details see Kuletz and others 1994b, c). At 
Naked Island, sites were randomly selected equally among 
four forest types (Kuletz and others, in press), with 69 of the 
sites having sufficient habitat data to include in this study. In 
western Prince William Sound, 77 sites were randomly 
selected from available habitat, although sample sizes among 
habitat types were not equal. Forty-six surveys were done 
from an anchored vessel, 23 from shore locations, and eight 
upland. An additional nine upland sites were surveyed 
opportunistically in 1991 . These sites were located in forested 
and nonforested habitat, and occurred in areas of western 
Prince William Sound not previously surveyed. Sampling at 
Kenai Fjords National Park was randomly stratified by forested 
versus unforested and bay head versus not bay head. The 38 
survey sites were equally distributed among the strata; 21 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



143 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



sites were surveyed from shore, eight from boats, and nine 
from upland sites. At Afognak Island, 76 dawn watch sites 
were arbitrarily selected with efforts to sample equally 
throughout the north and southwest parcels. Two sites were 
surveyed from shore and 74 upland. 

Sites were not randomly located within the entire spill 
zone. Therefore, our statistical results apply directly only to 
the sampled sites, and caution should be used when making 
inferences about other areas. Application of results to the 
entire area is based on the assumption (supported by our 
observations) that the study sites were representative of 
habitat types throughout the spill zone. 

Because epiphyte cover and platforms were not recorded 
at Naked Island, we used Naked Island data for preliminary 
analyses, but not for the final multivariate analyses. For 
analyses, we used detections over land <200 m from the 
observer because it produced stronger relationships with 
predictor variables in preliminary analysis of portions of the 
data set. Data from boat- and shore-based surveys were 
combined with upland survey data because these data are 
highly correlated (Marks and others, in press). Data from all 
areas were grouped because preliminary analyses indicated 
that within-site trends were similar to trends exhibited for all 
sites combined. 

Multiple Regression Analyses ofMurrelet Activity Levels 

We used multiple regression analyses to examine the 
continuum of murrelet activity levels relative to independent 
variables, to examine the interactive effects of those variables, 
and to describe the amount of variation explained by the 
model. Although season and weather affect inland activity 
level, we incorporated all these variables into the model 
rather than attempting to develop standardization factors. 
Our initial set of 19 predictor variables were factors known 
or suspected to be associated with high levels of activity or 
nesting of Marbled Murrelets, based on previously conducted 
analyses (Kuletz and others, in press; Marks and others, in 
press; Naslund and others, in press), and on univariate statistics 
across the four study areas. We used Kruskal-Wallis 
nonparametric analysis of variance to test categorical variables 
for significant effects on the number of detections. We 
calculated Pearson correlation coefficients between continuous 
variables and the number of detections, and between each 
pair of continuous variables. To control for colinearity, only 
one of a pair of variables with r > 0.80, whichever had the 
strongest correlation with the number of detections, was 
included in the same regression analysis. 

Because categorical and continuous variables were 
included in the multiple regression model, we used a General 
Linear Model procedure (SAS Institute 1988) to examine 
variation in murrelet activity levels. We transformed the 
number of detections by using natural logarithms and the 
percent data (canopy cover, forest cover, alder cover, and 
slope) by using square roots to stabilize residuals. We ran 
our initial regression model with all sites, and included all 
significant (P < 0.05) categorical variables and those 



continuous variables which were measured across all four 
study areas. We ran a second regression model for the three 
areas for which variables more directly related to Marbled 
Murrelet nest site selection (epiphyte cover and platforms 
per tree) were estimated. For this model we included all 
variables in the initial regression and epiphyte cover, which 
was highly correlated with platforms per tree. We reduced 
the model to include t probabilities for parameter estimates 
where P < 0.25 in the original model. This criterion was 
selected because our objective was to include all variables 
that explained variation in murrelet activity. Standardized 
parameters (parameter estimates divided by their standard 
error) were used to determine the relative importance of 
variables included in the models. 

Discriminant Analyses ofMurrelet Occupancy 

We used univariate tests and stepwise logistic regression 
to identify variables that could predict the probability of 
detecting occupied behavior at a survey site. This analysis 
included a test of how well the logistic model performed in 
classifying individual observations. For all four areas 
combined, we tested frequencies of classes of categorical 
variables for differences between occupied sites and sites 
of unknown status by using chi-square; and for differences 
in rank sums of continuous variables between occupied 
and unknown status sites by using the Wilcoxon 2-Sample 
Test (procedure NPAR1WAY; SAS Institute 1988). 
Significant variables (P < 0.05) in these tests were entered 
into a stepwise logistic regression model (procedure 
LOGISTIC; SAS Institute 1990; Naked Island group 
excluded). Inclusion and retention of variables in the 
stepwise logistic analysis were allowed at P < 0.10. We 
included platforms per tree in the model because it performed 
marginally better than one including epiphyte cover. 
Standardized parameter estimates were estimated by dividing 
the parameter estimate by the ratio of the standard deviation 
of the underlying distribution to the sample standard 
deviation of the explanatory variable (SAS Institute 1990), 
and were used to determine the relative importance of 
variables in the model. The classification error rate was 
calculated using a jackknife approach to reduce the bias of 
classifying the same data from which the classification 
criterion was derived (SAS Institute 1990). 

Results 

Marbled Murrelet Activity Levels 

Activity of Marbled Murrelets differed by study area 
(P = 0.018), with the greatest level of activity occurring at 
Afognak Island, the least at Naked Island, and intermediate 
levels in western Prince William Sound and Kenai Fjords 
National Park (table 1). Activity was greater during late 
summer than during spring and early summer (table 1). 
Activity was greater when the cloud ceiling was low than 
when there was a high ceiling or clear conditions (table 1). 
Activity was also greater at survey sites located at the heads 



144 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Sourhcentral Alaska 



Table 1 — The number of detections for categorical variables considered for inclusion in multiple regression analyses 
relating activity of Marbled Murrtlets to survey period, weather, topographic, and vegetation variables. A Kruskal-Wallis 
no nparametric analysis of variance tested the null hypotheses that murrelet activity did not differ between (or among) classes 
of each variable 



Variable regression 


Classes (n) 


Number of detections 


Chi-square 


df 


P 




Mean 


(") 




Area 


Naked Island (69) 
Prince William Sound (77) 
Kenai Fjords (38) 
Afognak Island (76) 


15.8 
23.8 
29.9 
38.4 


(2^7) 
(3.11) 
(5-78) 
(5.27) 


10.12 


3 


0.0175 


Survey period 


Early (May 1-Jul 10X1 13) 
Late (Jul 11-Aug 31X147) 


18.1 
33.6 


(2.84) 
(2.96) 


11.03 


1 


0.0009 


Survey method 


Boat (54) 
Shore (67) 
Upland (139) 


28.0 
23.6 
28.0 


(3-62) 
(432) 
(3.10) 


2.48 


2 


0.2890 


Cloud ceiling 


None (26) 
Above ridge (103) 
Below ridge (68) 


15.4 
35.1 
18.6 


" (4.05) 
(4.09) 
(2.90) 


6.44 


2 


0.0398 


Windspeed 


0km/h(123) 
l-8km/h(103) 
9-16km/h(15) 
>16km/h(18) 


31.1 
23.6 
113 
28.6 


(331) 
(2.86) 
(4.14) 
(8*6) 


631 


3 


0.0893 


Headbay 


Exposed shore (59) 
Bay (106) 
Headbay (95) 


16.6 
21.1 
39.6 


(3.45) 
(2.62) 
(4.28) 


27.75 


2 


0.0001 






of bays than elsewhere in bays or on exposed shorelines 
(table 1). Windspeed did not significantly affect murrelet 
activity and activity did not vary significantly among survey 
methods (by boat, from shore or upland; table 1). 

Correlation coefficients between Marbled Murrelet 
activity and continuous weather, topographic, and vegetation 
variables measured in all four areas varied from -0.16 for 
alder cover to 0.39 for d.b.h. (table 2). The largest correlation 
coefficients were between murrelet activity and variables 
directly related to nest site selection (epiphyte cover, platforms 
per tree; table 2). 

Our reduced model explained 52 percent of the total 
variation in murrelet activity ( table 3). Parameters for survey 
period, location relative to the head of a bay. and epiphyte 
cover were highly significant. Based on ratios of parameters 
to their standard errors (table 3), epiphyte cover, survey 
period, and location relative to the head of a bay were the 
most important predictors of murrelet activity. 

Across all four study areas combined, tree d.b.h. (x 2 = 
7.58, df = 2, P = 0.02), number of potential nesting platforms 
(X 2 = 7.08, df = 2, P = 0.03), and percent epiphyte cover (x 2 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Table 2 — Pearson correlation coefficients between continuous variables 
considered for inclusion in multiple regression model and murrelet activity 
(Overland detections <200 mfrom observer) 







Pearson 






correlation 


Variable 


Units 


coeffjeent 


Cloud cover 


Percent 


0.14 


Elevation 


Meters 


-0.14 


Slope 


Percent 


0.08 


Degrees from north 


Degrees 


-0.03 


Degrees from east 


Degrees 


-0.03 


Forest 


Percent 


0.24 


Canopy cover 


Percent 


0.12 


Canopy height 


Meters 


0.24 


Diameter at breast height 


Centimeters 


0.39 


Alder cover 


Percent 


-0.16 


Epiphyte cover 1 


Percent 


0.48 


Platforms 1 per tree 


Number 


0.43 



Not estimated at Naked Island 



145 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



Table 3 — Multiple regression model relating activity of Marbled Murrelets 1 to survey period, weather, topographic, 
and vegetation variables at three study areas: western Prince William Sound, Kenai Fjords National Park, and 
Afognak Island. Categorical variables were entered into the regression as dummy variables 







Levels of 


Estimate (s.e.) 


Parameter 






t 2 


P 


Standardized 


Model 


Variable categorical variables 








estimate 


F= 15.21 


Intercept 




2.326(0.421) 


5.53 


0.0001 




df= 10,140 














R 2 = 0.52 


Period 


(Early) 

1 (Late) 


-0.851 (0.19) 


^.38 


0.0001 


4.39 


P =0.0001 


Headbay 


(Exposed) 


-1.028(0.281) 


-3.66 


0.0004 


3.66 






1 (Bay) 


-0.820 (0.200) 


-4.10 


0.0001 


4.10 






2 (Headbay) 












Elevation 




-0.005 (0.002) 


-3.03 


0.0029 


2.50 




Slope 3 




0.131 (0.053) 


2.47 


0.0148 


2.47 




Degrees from north 




-0.003 (0.002) 


-1.86 


0.0648 


1.50 




Forest cover 3 




0.121 (0.070) 


1.72 


0.0700 


1.73 




Canopy cover 3 




-0.120 (0.072) 


-1.67 


0.0964 


1.70 




D.b.h. 




0.010 (0.006) 


1.73 


0.0863 


1.67 




Epiphyte cover 




0.018 (0.004) 


4.73 


0.0001 


4.50 



1 Variable was natural log transformed 

2 Tested null hypothesis that coefficient estimate = 

3 Variable was square root transformed 



= 6.73, df = 2, P = 0.03) were greater at sites located at heads 
of bays, than at more exposed sites. 

Identification of Occupied Sites 

The probability of observing occupied behavior was 
greater: (1) at Afognak Island than at other areas; (2) during 
upland surveys than during boat or shore surveys; (3) during 
days with a high percentage of clouds than during clear 
days; and (4) at bays (especially at heads of bays) than at 
exposed sites (table 4). The probability of observing occupied 
behaviors did not vary with survey period or windspeed. 
Occupied sites had greater levels of cloud cover, forest 
cover, canopy cover, canopy height, d.b.h., epiphyte cover, 
and platforms per tree, than other sites (table 5). Alder cover 
was greater at other sites than at occupied sites. 

Tree size (d.b.h.) and location relative to the head of a 
bay entered the model at the P < 0.10 level; survey method 
and platforms per tree were also included. Standardized 
parameter estimates (table 6) indicated that d.b.h. and 
platforms per tree were the most important predictors of 
occupied sites. The logistic function correctly classified 78.9 
percent of observations in a jackknife procedure; 82.7 percent 
of occupied sites, and 74.6 percent of sites of unknown 
status were correctly classified. 



Discussion 

Habitat Predictors Of Murrelet Use 

Murrelet Activity Levels 

Several variables were consistent predictors of high 
murrelet activity. Allowing for survey period, activity was 
highest at the heads of bays, at low elevations, and in areas 
with a high percentage of forest cover and large diameter 
trees. The most important habitat variables across all study 
areas were location relative to heads of bays, tree size (d.b.h.), 
and epiphyte cover on trees (excluding the Naked Island 
group for which there was no data on epiphyte cover). The 
number of platforms per tree was also important because it is 
highly correlated with epiphyte cover. 

The importance of tree size and the number of platforms 
per tree was consistent with results from other studies and 
with attributes of nest trees found in southcentral Alaska 
(Hamer and Cummins 1991; Hamer, this volume; Naslund 
and others, in press). The importance of location relative to 
heads of bays was noted in earlier analyses of Prince William 
Sound data (Kuletz and others, in press; Marks and others, in 
press) but has not been reported elsewhere. Further, the 
trend for a bay effect in Kenai Fjords National Park was not 
significant in prior analyses (Kuletz and others 1994b). It is 



146 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



Table 4 — Univariate tests for differences in frequencies of classes of categorical variables between occupied 
sites (where behaviors indicating nesting were observed) and other sites (where behaviors indicating nesting were 
not observed 







Proportion of 








Variable 


Class (n) 


occupied sites 


Chi-square 


df 


P 


Area 


Naked Island (69) 
Prince William Sound (77) 
Kenai Fjords (38) 
Afognak Island (76) 


0.22 
0.22 
0.32 
0.66 


42.08 


3 


0.0001 


Survey period 


Early (May 1 -Jul 10) (113) 
Late(Julll-Augll)(147) 


0.34 
0.37 


0.23 


1 


0.629 


Survey method 


Boat (54) 
Shore (67) 
Upland (139) 


0.24 
0.24 
0.47 


14.56 


2 


0.001 


Cloud ceiling 


None (68) 
Above Ridge (103) 
Below Ridge (63) 


0.23 
0.44 
0.41 


7.74 


2 


0.021 


Windspeed 


0Km/h(123) 
l-8Km/h(103) 
9-16Km/h(15) 
>16Km/h(18) 


0.37 
038 
0.33 
0.22 


1.704 


3 


0.636 


Headbay 


Exposed shore (59) 
Bay (106) 
Headbay (95) 


0.22 
0.35 
0.46 


9.42 


2 


0.009 



Table S — Means, standard errors, and univariate tests for differences in rank sums of continuous variables between sites 
where one or more occupied behaviors (behaviors indicating nesting of marbled murrelets) were observed (occupied sites) 
and sites where no behaviors indicating nesting of Marbled Murrelets were observed (other sites) 



Variable 




Occupied sites 






Other sites 




Z> 


P 




n 


Mean 


(s.e.) 


n 


Mean 


(s.e.) 




Cloud cover 


94 


80.85 


(3.76) 


166 


68.75 


(3.33) 


2.06 


0.04 


Elevation 


87 


51.65 


(4.61) 


140 


71.70 


(6.81) 


-0.62 


0.53 


Slope 


88 


21.25 


(18.52) 


140 


22.15 


(12.89) 


-1.30 


0.19 


Degrees from north 


88 


91.25 


(5.53) 


140 


91.29 


(4.51) 


-O.05 


0.% 


Degrees from east 


88 


99.77 


(6.00) 


140 


99.14 


(4.61) 


0.12 


0.90 


Forest cover 


88 


74.64 


(2.64) 


136 


60.34 


(3.00) 


2.69 


0.008 


Canopy cover 


88 


63.26 


(2.46) 


134 


49.69 


(2.86) 


2.54 


0.01 


Canopy height 


88 


26.71 


(1.25) 


135 


17.31 


(119) 


7.94 


0.0001 


D.b.h. 


87 


57.11 


(198) 


140 


33.70 


(1.77) 


7.94 


0.0001 


Alder cover 


86 


3.03 


(0.70) 


132 


10.90 


(1.86) 


-3.08 


0.002 


Epiphyte cover 


72 


54.57 


(3.88) 


82 


16.78 


(2.18) 


7.06 


0.0001 


Platforms per tree 


72 


7.36 


(0.67) 


82 


2.06 


(0.38) 


6.95 


0.0001 



'Wilcoxon 2-Sample Test 






USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



147 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



Table 6 — Logistic regression model to predict probability of occupied sites of Marbled Murrelets (sites where one 
or more behaviors indicating nesting were observed) for the three study sites: western Prince William Sound (1992), 
Kenai Fjords National Park (1993) andAfognak Island (1992), Alaska (n = 152 sites total) 



-2LogL 


df 


P 


Variable 




Parameter 






Chi-square 


Estimate (s.e.) 


Chi-square 


P 


Standardized 
















estimate 


73.513 


4 


0.0001 


Intercept 


4.918(0.903) 


29.633 


0.0001 










Method 


-0.679 (0.257) 


6.970 


0.0083 


0.31 








Headbay 


-0.559 (0.306) 


3.331 


0.0680 


-0.26 








D.b.h. 


-0.040(0.012) 


11.320 


0.0008 


-0.56 








Platforms 


-0.138(0.057) 


5.776 


0.0162 


-0.41 



possible that high detection rates result from murrelets 
funneling through bay heads and using them as flyways. 
However, the consistency of high activity at bay heads for 
the study areas overall, combined with the high proportion 
of occupied sites at bay heads, suggests otherwise. 

Marks and others (in press) found that murrelet activity 
was positively correlated with stand size in western Prince 
William Sound. High activity at bay heads may be a result of 
larger contiguous forests at bay heads, although stand size 
relative to landform has not been investigated in these areas. 
Microclimate and minimal exposure to weather at bay heads 
may foster characteristics associated with known murrelet 
nesting habitat, including large tree size and mossy platforms 
on trees. This may explain the larger tree d.b.h., greater 
number of potential nesting platforms, and higher percentage 
of epiphyte cover at sites located at heads of bays relative to 
more exposed sites. However, these trends were not evident 
at Kenai Fjords National Park in earlier analyses (Kuletz and 
others 1994b). This is likely due to the recent deglaciation of 
many of the bay heads. 

The importance of tree size and elevation in predicting 
murrelet activity has been suggested by other studies. Murrelets 
typically nest in old-growth stands where trees tend to be 
relatively large (see Hamer and Nelson, this volume b). 
Hamer and Cummins (1991) and Rodway and others (1991) 
found that murrelet activity was highest in low elevation 
forests in Washington and British Columbia. In northern 
latitudes, larger trees are found at lower elevations (Viereck 
and Little 1972). Kuletz and others (in press) found a significant 
negative correlation between tree d.b.h. and elevation on the 
Naked Island group, even though the highest elevation was 
<460 m. Thus, the contribution of elevation to the model is 
likely due to its effect on patterns of vegetation growth. 

Conversely, it is also possible that murrelets are detected 
more frequently at low elevations, as they move from marine 
to terrestrial areas, because low elevation habitat tends to be 
closer to shore. Murrelets must pass over the shoreline to 
reach sites further inland. However, in some areas, murrelets 
leave the water and rapidly gain altitude before flying to 
distant inland sites (Van Vliet, pers. comm.), and would not 
be detected along the shoreline. 



Responses of murrelet activity to variation in slope, 
aspect, and canopy cover were not consistent, and may have 
been influenced by local geography. Activity was positively 
related to northerly aspect in preliminary regression models, 
similar to findings of earlier analyses for Naked Island data. 
At Naked Island, there was a non-significant positive trend 
of higher murrelet activity on northerly slopes, possibly due 
to more high- volume forests on these slopes or the prevalence 
of southeast winds, that murrelets may seek to avoid (Kuletz 
and others, in press). 

Occurrence of Occupied Behaviors 

The influence of habitat features on the occurrence of 
occupied behaviors was similar to their influence on murrelet 
activity levels. In particular, the size of trees and the number 
of potential nest platforms were good predictors of murrelet 
occupied behavior. This is consistent with Alaskan tree nests 
that have been documented; most were located on large 
moss-covered platforms, often on the largest trees in an area 
(Naslund and others, in press). However, our results could 
be biased in that occupied behaviors in non-forested habitats 
have not been adequately defined. 

Epiphyte cover, number of potential nest platforms, and 
tree size were clearly related. The importance of these habitat 
features to nesting murrelets may vary geographically. For 
example, epiphyte cover may be more important in Alaska 
than in other areas; moss was not the primary nest substrate 
of some nests at lower latitudes (Hamer and Nelson, this 
volume b; Singer and others 1991). Naslund and others (in 
press) suggested that moss is more important as insulation in 
Alaska's severe climatic conditions. Additionally, moss 
increases platform size, which could be important where 
small trees predominate. 

Nesting clearly occurs in non-forested areas (Day and 
others 1983). However, the extremely low levels of general 
activity and of occupied behaviors at non-forested sites suggest 
that nesting activity in non-forested areas is less common than 
in forested areas. We believe that our results indicate that 
murrelet nesting density is low in sparsely forested or non- 
forest areas and that such habitat is of less importance to the 
population. However, it is possible that differences in murrelet 



148 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Kuletz and others 



Chapter 15 



Inland Habitat Suitability in Southcentral Alaska 



activity levels and behaviors in non-forested and forested habi- 
tats may not reflect actual differences in murrelet abundance. 
For example, murrelets may be more vulnerable to predation 
in open areas and therefore less active around ground nests. 

Effects of Survey Methods 

Levels of murrelet activity did not vary among survey 
methods. However, significantly more occupied behaviors 
were observed when surveys were done from upland sites 
rather than from the shoreline or a boat. Occupied behaviors 
may be hard to detect during surveys conducted from a boat 
because the observer is often 50-100 m from forest habitat. 
However, occupied behaviors were equally low in frequency 
when surveys were done from the shoreline. Thus, our results 
may reflect real differences in habitat use. Although murrelets 
sometimes nest within a few hundred meters of the shore 
(Cody, unpubl. data; Kuletz, unpubl. data; Marks, unpubl. 
data; Naslund and others, in press), they may use areas along 
the shoreline less frequently than those further inland (Hamer, 
this volume). The effect of survey method was confounded 
with effect of survey area, because boat and shore-based 
surveys predominated at Prince William Sound and Kenai 
Fjords National Park, whereas upland surveys predominated 
at Naked Island and Afognak Island. The latter had very 
high activity levels, large trees and high epiphyte cover 
(Naslund and others, in press), and the high occupied status 
rate could have been due to truly higher nesting densities. 

Sources of Unexplained Variation 

Our best multiple regression model explained 52 percent 
of the variation in murrelet activity. There were many potential 
sources of unexplained variation. Because sites were surveyed 
only once, day-to-day variation within the same area could 
have contributed to incorrect estimation of general activity 
level of a given site. We did not account for observer 
variability, which can introduce additional bias to murrelet 
surveys (Kuletz and others 1994c; Ralph, pers. comm.). 
Because each area was generally surveyed by different 
observers, area effects could be due partially to observer 
variability. In addition, differences in sampling design may 
have contributed to area effects or other variation. For 
example, all forest was treated equally in our analyses, yet 
forest characteristics (e.g., age structure, volume, tree species) 
are quite variable. The Naked Island group was the only area 
for which specific forest types were stratified and sampled. 

Prevailing winds, local topography and vegetation patterns 
varied throughout the study area. Therefore, the geographic 
range of study sites likely contributed to the variation in 
murrelet activity we observed. In addition, murrelet nesting 
distribution may vary with availability of suitable habitat. For 
example, murrelets may be more dispersed in Prince William 
Sound if prime nesting habitat is abundant and widespread, 
whereas nesting density may be higher in good habitat on the 
Kenai Peninsula if suitable habitat is sparse. Thus, the lower 
activity levels in Prince William Sound, relative to the Kenai 
Peninsula, may reflect differences in habitat availability, rather 
than habitat suitability, between the two areas. 



An important factor not considered in our models was 
the adjacent marine environment and the availability of 
foraging habitat. These factors must ultimately determine 
the use of suitable nesting habitat. Thus, the apparent 
increase in murrelet activity from Prince William Sound to 
Afognak Island may also reflect large-scale differences in 
prey availability. 

Conclusions 

These models primarily serve as descriptive tools until 
they can be tested with independent data. However, we were 
able to explain 52 percent of the total variation in Marbled 
Murrelet activity levels based on temporal, topographic, and 
habitat characteristics. Further, our results suggest an 83 
percent success rate of classifying murrelet nesting habitat 
in the areas examined on the basis of occupied behavior. The 
features indicative of murrelet nesting habitat include low 
elevation locations near the heads of bays, with extensive 
forest cover of large old-growth trees. In some areas, such as 
the Kenai Fjords, location relative to bay heads may be less 
important. The best predictors of nesting habitat in forested 
areas are high epiphyte cover and large numbers of potential 
nesting platforms on trees. 

Our results were derived from surveys designed to 
estimate murrelet use of forested habitat. Potential variation 
in murrelet behavior associated with habitat type (i.e., forest 
or non-forest) has not been adequately examined and could 
influence accurate interpretation of survey results. Therefore, 
caution should be exercised when extrapolating observed 
trends on a broad scale across the landscape. 

Acknowledgments 

The contribution of several studies was integral to this 
paper. We thank the USDA Forest Service (Chugach National 
Forest), who was a cooperative partner in the studies in 
Prince William Sound and Naked Island areas, and the U.S. 
Fish and Wildlife Service, Division of Realty who conducted 
murrelet surveys on Afognak Island. This research was funded 
through the U.S. Fish and Wildlife Service, Division of 
Migratory Bird Management as part of the Exxon Valdez oil 
spill restoration program and the Division of Realty. For 
assistance on projects we thank G. Esslinger, A. Belleman. I. 
Manley, S. Anderson, D. Huntwork, K. Rausch, K. Fortier, J. 
Maniscalco, E. Tischenor, L. Fuller, B. Grey, J. Fadely, B. 
Fadely, D. Zwiefelhofer, G. Johnson, V. Vanek, M. Nixon, 
T. Nelson, G. Landua, D. Goley and D. Kaleta. For the GIS 
support, we also thank T. Gerlach of the Division of Realty 
and T. Jennings and C. Wilder, all of them with U.S. Fish and 
Wildlife Service. From the USDA Forest Service we thank 
C. Hubbard, R. DeVelice, Z. Cornett, and B. Williams. S. 
Klosiewski provided guidance on study design and data 
analysis. The comments of R. Barrett, P. Connors, Chris 
Iverson, Michael McAllister, C. John Ralph, and an anonymous 
reviewer greatly improved this manuscript. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



149 



Chapter 16 

Inland Habitat Associations of Marbled Murrelets in 
British Columbia 



Alan E. Burger 1 

Abstract: Most Marbled Murrelets (Brachyramphus marmoratus) 
in British Columbia nest in the Coastal Western Hemlock 
biogeoclimatic zone. In this zone, detection frequencies were highest 
in the moister ecosections and in low elevation forests. Nests and 
moderately high levels of activity were also found in some forest 
patches in the subalpine Mountain Hemlock zone. There was no 
evidence of nesting in subalpine scrub forest, lowland bog forest, 
or alpine tundra. Studies on the Queen Charlotte Islands and 
Vancouver Island reported consistently higher detection frequen- 
cies in old-growth than second growth forests (20-120 years old). 
Detections in second-growth were usually associated with nearby 
patches of old-growth. Within low elevation old-growth, detection 
frequencies were sometimes positively correlated with mean tree 
diameter, but showed weak or no associations with tree species 
composition and minor variations in forest structure. Sitka spruce 
(Picea sitchensis) and western hemlock (Tsuga heterophylla) were 
important components of many high-activity sites. High murrelet 
activities were associated with well-developed epiphytic mosses, 
but mistletoe seemed less important. A study on Vancouver Island 
showed higher predation of artificial nests and eggs at forest edges, 
which suggests problems for Marbled Murrelets in fragmented 
forests. The use of detection frequencies in the selection and 
preservation of potential nesting habitat is discussed and the limi- 
tations of single-) ear studies are exposed. 



British Columbia supports a significant portion of 
the North American population of Marbled Murrelets 
(Brachyramphus marmoratus). Over the past century, 
evidence accumulated that the birds nested in large trees 
in British Columbia (Campbell and others 1990), and at 
least one early biologist made the connection between 
declining numbers of murrelets and the reduction of old- 
growth forests on eastern Vancouver Island (Pearse 1946). 
In recent decades the pace of logging of coastal old- 
growth forests has greatly increased. Between 1954 and 
1990 about half of the large-tree old-growth forest on 
Vancouver Island (75 percent in the southern island) was 
logged ( Husband and Frampton 1 99 1 ). Out of 354 forested 
w atersheds larger than 5,000 ha in coastal British Columbia, 
only 20 percent are pristine and 67 percent have been 
significantly changed by industrial activity, primarily 
logging ( Moore 1 99 1 ). Concerns over the effects of logging 
on Marbled Murrelet populations were raised by Sealy 
and Carter (1984), but there were no intensive inland 
studies until the species was listed as threatened in Canada 
in 1990. Loss of nesting habitat by logging was considered 



1 Associate Professor (Adjunct). Department of Biology , University of 
Victoria, Victoria. British Columbia, V8W 2Y2. Canada 



the greatest threat (Rodway 1990, Rodway and others 
1992). The listing stimulated several inland studies, 
including reconnaissance surveys in many watersheds of 
the Queen Charlotte Islands (Rodway and others 1991, 
1993a) and Vancouver Island (Savard and Lemon in press) 
and intensive surveys at several sites. 

Identification and mapping of potential nesting habitats 
was identified as a high priority for research in the National 
Recovery Plan for the Marbled Murrelet. prepared by the 
Canadian Marbled Murrelet Recovery Team (Kaiser and 
others 1994). Detailed 1:50,000 maps of coastal old-growth 
forests are being prepared (Derocher. pers. comm.). There 
are still very few data available for either landscape- or 
stand-level analyses of habitat associations. I review the 
available data and point out research topics that urgently 
need to be addressed. 

Methods and Sources of Data 

The studies reviewed here followed the Pacific Seabird 
Group survey protocols for general (road) and intensive 
(fixed station) surveys (Paton and others 1990, Ralph and 
others 1994), with the exception of Eisenhawer and Reimchen 
(1990) and Reimchen (1991). 

Rodway and others (1991; 1993a,b) did intensive 
sampling through the 1990 season in Lagins Creek and 
Phantom Creek on Graham Island, and less frequent general 
surveys in 1 2 other watersheds on the Queen Charlotte Islands. 
Savard and Lemon (in press) analyzed data from 382 surveys 
at 151 fixed stations and 88 road surveys in 82 watersheds 
on Vancouver Island in 1991. Relatively few surveys were 
made at each station (mean 1.6, range 1-5), and large numbers 
of observers were used with variable degrees of training. 
Savard and Lemon (in press) warned that their data could 
not present an accurate picture of murrelet activity in any of 
the watersheds surveyed. Nevertheless, some significant 
patterns emerge at the landscape scale. 

The remaining studies focussed on fine-scale temporal 
and spatial variations within single watersheds during one 
season (Eisenhawer and Reimchen 1990; MacDuffie and 
others 1993; Manley and others 1992, 1994) or 3-4 seasons 
(Burger 1994; Jones 1992, 1993). Only three studies combined 
repeated intensive surveys with detailed habitat analysis at a 
variety of sites (Burger 1994. Manley and others 1994, 
Rodway and others 1993a). These data are insufficient for a 
thorough examination of habitat patterns at stand and 
landscape scales in British Columbia, but some trends are 
apparent and are reviewed here. Figure 1 shows the location 
of the study sites. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



151 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



Phantom Creek 
Lagins Creek 




BRITISH COLUMBIA 



Kitlope Valley 
.Mussel Inlet 




scale 



50 
km 



100 




Megin Valley 



Carmanah Valley 
Walbran \ 



Caren Range 
Sechelt Peninsula 



Vancouver 



»^L Gulf Islands 



Victoria 



Figure 1 — Coastal British Columbia showing the location of inland studies of Marbled Murrelets (open stars). 



Biogeoclimatic Zones 

Marbled Murrelets have access to four biogeoclimatic 
zones (Meidinger and Pojar 1991). The Coastal Western 
Hemlock Zone covers most of coastal British Columbia at low 
to mid elevations (0-900 m on windward and 0-1050 m on 
leeward slopes on the south and mid-coast; and 0-300 m on 
the north coast). Dominant trees are western hemlock (Tsuga 
heterophylla), western red cedar (Thuja plicata), and Amabilis 
fir (Abies amabilis), with yellow cedar (Chaemaecyparis 
nootkatensis) in higher elevations and Douglas-fir (Pseudotsuga 
menziesii) in drier habitats. Lodgepole pine (Pinus contorta) 



occurs in dry shoreline areas and bogs. Sitka spruce (Picea 
sitchensis) is an important component on floodplains in the 
southern forests, and in many older forests in the Queen 
Charlotte Islands and the northern mainland, and is an 
important nest site for Marbled Murrelets. Most Marbled 
Murrelets in British Columbia appear to nest in this zone 
(see below). 

The Coastal Douglas-fir Zone covers a small area on 
southeastern Vancouver Island, the Gulf Islands, and a narrow 
strip of the adjacent southern mainland at elevations below 



152 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



1 50 m. It characterizes relatively dry forest in the rain shadow 
of the Vancouver Island and Olympic Mountains. Very little 
old-growth remains in this heavily populated zone. Douglas- 
fir is the dominant tree, with other conifers and broad-leaved 
trees sometimes common. There has been no research on 
Marbled Murrelets in this zone, but nesting is likely, because 
the birds are often seen nearby on the ocean. 

The Mountain Hemlock Zone occurs at 900-1800 m in 
southern British Columbia (lower on windward slopes) and 
400-1000 m in the north. It is most common above the 
Coastal Western Hemlock Zone on the mainland Coast 
Mountains and the insular mountains of Vancouver Island 
and the Queen Charlotte Islands. Dominant trees are mountain 
hemlock (Tsuga mertensiana), amabilis fir, and yellow cedar. 
Much of this forest occurs as a mosaic among areas of 
subalpine heath, meadow, and ferns. Nesting has been recorded 
in these forests on the southern mainland (see below). 

The Alpine Tundra Zone occurs on high coastal 
mountains, above 1650 m in the south and 1000 m in the 
north, and is dominated by shrubs (willows and birch), herbs, 
bryophytes, and lichens. Marbled Murrelets have been reported 
flying over such habitats (Rodway and others 1993a), but 
there is no evidence that they nest there in British Columbia. 



Landscape Attributes 

Old-Growth Compared with Second-Growth 

Two studies compared detection frequencies in old- 
growth and second-growth. Rodway and others (1993a) 
recorded high densities of activity in intensive surveys in 
old-growth on the Queen Charlotte Islands (details below), 
but had only one detection in five intensive surveys in second- 
growth stands (60-120 years old). In road surveys, detections 
were reported at 76 percent (n = 25) of old-growth stations, 
but only at 27 percent (n = 101) of second-growth stations 
(20- 1 20 years old). In 85 percent of the cases where detections 
were recorded in second-growth forest, there were stands of 
old-growth within 500 m. Detection frequencies were 
significantly higher in old-growth than second-growth, and 
within second-growth they were significantly higher if there 
was old-growth nearby (fig. 2). 

Savard and Lemon (in press) reported significantly fewer 
detections from stations in watersheds with less than 50 
percent remaining old-growth, compared to more intact 
watersheds (fig. 3). At fixed stations in May and July, fewer 
detections were recorded when the proportion of old-growth 
fell below 75 percent of the watershed. In addition, stations 



□ >500 □ 151-500 H 51-150 



0-50 



4 T 



5 3 



& 



c 
o 



I 2 



1 - 



n = 12 




Distance from station (m) 



n = 13 




n = 57 



n = 44 



Spruce-Hemlock Cedar-Hemlock 
Old-growth 



<500 m from old- >500 m from old- 
growth growth 
Second-growth 



Figure 2 — Mean number of Marbled Murrelet detections per road transect station in relation to adjacent 
habitat type in the Queen Charlotte Islands (from Rodway and others 1 993a) . The sample size (n) is the 
number of surveys. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



153 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



A: Fixed station intensive surveys 
Percentage of old-growth in watershed 



40 



0-25 M 25-50 D 50-75 □ 75-100 



31 



34 




3 T 



eg 



S.2 



</> 

c 
o 

t3 

I 

O 

c 

c 

CD 

cu 

5 



'. 1 



B: Road surveys 



No surveys 





15 


30 


14 










15 j 




16 | 

— t-^* 



May 



June 




Figure 3 — Mean numbers of Marbled Murrelet detections in intensive fixed station (A) 
and general road surveys (B), in relation to the percentage cover of remaining old- 
growth forest in the sampled watersheds on Vancouver Island (from Savard and 
Lemon, in press). Sample sizes (n) shown above columns are numbers of surveys. 



close to old-growth (within 200 m in fixed stations and 
within 500 m in road transects) had higher detection rates 
than those further away. 

These studies confirm that murrelets avoid second-growth 
forests, even those 60-120 years old. Furthermore, the 
Vancouver Island results tentatively suggest that murrelets 
do not pack into the remaining old-growth with increased 
density; reduced habitat leads to reduced populations. 

Relationship Between Landscape 
and Stand 

Distance to Salt Water and Location Within the Watershed 

Savard and Lemon (in press) found no significant 
correlation between detection frequency and distance from 
salt water (using intervals of 0-5, 5-15, and >15 km) at 151 



stations on Vancouver Island in May and July, but found a 
negative correlation in June. They found no effects of distance 
to open ocean (beyond the inlets) in any month. The location 
of fixed stations within each watershed did not affect detection 
rates (each watershed was divided into four zones, from 
mouth to headwaters), although road surveys showed 
significantly higher detections in the centers of the watersheds. 
These data indicate that Marbled Murrelets are able to access 
all of Vancouver Island, although only a small portion might 
be suitable nesting habitat. 

The effect of distance from the ocean was tested in the 
Carmanah and Walbran watersheds in which unbroken old- 
growth forest extends from the ocean almost to the headwaters 
for 21 and 18 km, respectively. Manley and others (1992) 
reported a significant negative correlation between detection 
rates and distance from the ocean at six stations in Carmanah- 



154 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



Walbran in 1990. A larger data set (1 1 stations in 1991 and 13 
in 1992) produced no significant correlations when occupied 
detections (Pearson correlation, r = -0.081 and -0.271, 
respectively) or total detections (r = -0.140 and -0.267, 
respectively; P > 0.05 in all cases) were considered (fig. 4; 
Burger 1994). The highest detection frequencies were found 
at sites 8-17.5 km inland. All six nests found in Carmanah- 
Walbran were more than 10 km from the ocean (Burger 1994). 

Precipitation Amount and Form 

Most of the old-growth forests in which high densities 
of murrelets have been reported receive high rainfall (most 
in winter) and relatively little snow. On Vancouver Island, 
detection frequencies were significantly higher in the two 



moist ecosections (Western Island Mountains and Northern 
Island Mountains; Demarchi and others 1990) than in the 
drier Nahwirti Lowland and Nanaimo Lowland ecosections 
(Savard and Lemon, in press). Overall, detections were 
significantly higher on the moister western side of Vancouver 
Island than on the eastern side, but the latter area has also 
been far more extensively logged and urbanized, which might 
contribute to this difference. 

Rodway and others (1993a) reported no detections at 
apparently suitable forest with large Sitka spruce at Gray 
Bay, Queen Charlotte Islands. The spruce trees there had 
virtually no moss development on their limbs, apparently as 
a result of sea spray, which might have made them less 
attractive to murrelets. 



100 - 



Occupied detections □ Other detections 




FRD HEA STR SIS SW AC SH W90 RT BP HUM LCC UCC 



100 - 



8- 80- 

2 

3 
■ 

8. 60 - 

■ 
c 
o 

n 



O 

5 20 



1S92 



4 

4 3 

4 



FRD 
Km from sea: 6.2 



HEA 

7 



STR 
8 



SIS 
9 



_ 



SW 
11.5 



5 


















10 




9 


















^^ 










10 




+ 


8 






-H 





AC 
12.5 



SH 
15.1 



W90 
17.5 



RT 
17.8 



BP 
18.2 



HUM 
20 



LCC 
20.3 



UCC 
21.1 



Figure 4 — Mean frequencies of occupied and other detections reported from 13 intensive survey stations 
(arranged in increasing distance from the ocean) in the Carmanah-Walbran watersheds, Vancouver Island, in 
the period 15 May through 16 July in 1991 and 1992 (from Burger 1994). Sample sizes (n) above columns are 
numbers of surveys. The x-axis is labelled with the codes for each station. Codes for each station are: FRD = Ford, 
HEA ■ Heaven Canp, STR = Stream Site. SIS = Three Sisters, SW = South Walbran Bridge, AC = August Creek, 
SH = Sleepy Hollow, W90 = West Walbran 1990 Nest Site, RT = Research Tree, BP = Bearpaw Camp, HUM = 
Hummingbird Camp, LCC = Lower Ctearcut, UCC = Upper Ctearcut. 



L'SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



155 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



Stand Attributes and Relative 
Murrelet Densities 

Elevation 

Eisenhawer and Reimchen (1990) found no evidence of 
Marbled Murrelets in high elevation (to 700 m) subalpine 
scrub forest of lodgepole pine above Coates Lake, Queen 
Charlotte Islands. At Lagins Creek, Queen Charlotte Islands, 
Rodway and others (1991, 1993a) found a significant 
difference in mean detection rates in May through July 
between low elevation forests (90- 1 50 m), high forests (230- 
460 m), and alpine areas (720-1000 m): 32.4 ± 4.1 (s.e.), 
17.5 ± 3.0, and 3.0 ± 0.7 detections per survey, respectively. 
About 98 percent of the old-growth forest occurred below 
500 m in this area. A few birds passed over alpine ridges in 
this area, but 84 percent of the detections in high altitude 
stations were of birds 500-1500 m distant, flying in the 
valleys below. Ground searches in alpine areas yielded no 
sign of nesting. 

Marbled Murrelets do nest in some high altitude forests 
above fjords on the mainland coast. Murrelets have been 
reported flying over the steep slopes, mostly covered in 
scrubby sub-alpine forest with patches of taller trees, which 
surround fjords (Burns, pers. comm.; Kaiser, pers. comm.; 
Prestash, pers. comm.). One radio-tagged bird was tracked 
to a sub-alpine stand of large conifers above Mussel Inlet 
(Prestash and others 1992b; see details below). Similar habitat 
appears to support Marbled Murrelets in the Kitlope drainage 
on the north-central mainland (Kelson, pers. comm.). 

Fairly high rates of activity (details below) were reported 
from sub-alpine forest at 750-1200 m, dominated by 
mountain hemlock and yellow cedar in the Caren Range, 
Sechelt Peninsula (Jones 1992; P. Jones, pers. comm.). An 
active nest was found here in 1993 at 1088 m (Jones 1993). 
A fledgling Marbled Murrelet was found alive on the ground 
by a tree faller at Downing Creek, near Furry Creek on the 
east side of Howe Sound in 1985. The suspected nest was at 
the top of a "red cedar" (sic) at an altitude of 1064 m 
(Morgan 1993). 

Marbled Murrelets nest as high as 1000 m, and these 
somewhat meager data suggest that vegetation development, 
specifically the absence of large trees at high altitudes, affects 
Marbled Murrelets more than altitude per se. 

Aspect, Slope and Stand Position on Slope 

The effects of slope and aspect have not been adequately 
investigated in British Columbia. High elevation stations on 
side slopes in two watersheds in the Queen Charlotte Islands 
(see above for altitudes) had lower detection rates than those 
in the valley bottoms, but this might be a consequence of 
elevation, rather than slope or aspect (Rodway and others 
1991, 1993a). These authors pointed out that if birds circled 
over narrow valleys, they would probably pass over observers 
on the valley floor more often than observers on the side 
slopes, causing differences in detection frequencies. 



Vegetation Classification and Tree Size 

Intensive surveys in Lagins Creek, Queen Charlotte Island, 
by Rodway and others (1993a) yielded the highest densities 
of detections in stands of large Sitka spruce and western 
hemlock. These preferred stands included the following site 
associations: ( 1 ) valley bottom, western red cedar/Sitka spruce 

- foamflower (mean diameter at breast height [d.b.h.] = 162 
cm); (2) valley bottom, western red cedar/Sitka spruce - 
Conocephalum (d.b.h. = 104 cm); and (3) slope forest, western 
hemlock/Sitka spruce - lanky moss (d.b.h. = 93 cm). Within 
these associations, vegetation groups with the largest trees 
(mean d.b.h. 141 cm vs. 60 cm for all other plots) had 
significantly higher rates of murrelet detections. These 
differences disappeared when only low-altitude sites were 
considered. Lower detections rates were found in these site 
associations: ( 1 ) valley bottom, western red cedar/Sitka spruce 

- skunk cabbage (d.b.h. = 40.4 cm); (2) higher altitude, 
western red cedar/western hemlock - blueberry (d.b.h. not 
measured); and (3) lodgepole pine/yellow cedar - sphagnum 
(d.b.h. not measured) found in low-elevation bog-forest. 

Reimchen (1991) made informal observations of flight 
activity of Marbled Murrelets (not following the Pacific 
Seabird Group protocol) at 49 lakes on Graham and Moresby 
Islands (Queen Charlotte Islands) between 25 May through 
25 July over a 1 2 year period. The birds were absent or rare 
(<2 calls per 15 minute survey) at 40 lakes, most of which 
were surrounded by unforested scrubby vegetation or "poorly 
forested" terrain. The nine lakes at which there was extensive 
murrelet activity were distributed primarily in old-growth 
forest with mossy boughs. Sitka spruce appeared to be an 
important component of the vegetation at active sites. Intensive 
observations by Eisenhawer and Reimchen (1990) at Coates 
Lake, Graham Island from 1 June to 3 August 1986 yielded a 
mean of 12.9 (range 1-50, n = 42) detections per dawn 
survey, as well as records of birds carrying fish, landing on 
trees, and possibly copulating on a branch. The old-growth 
forests here were mixtures of western hemlock, Sitka spruce, 
western red cedar, and yellow cedar, with canopies 40-70 m 
tall. No detailed habitat plots were made. 

Murrelet activity was reported over the steep forested 
slopes overlooking Mussel Inlet, a northern mainland fjord 
(Prestash and others 1992b; Prestash, pers. comm.; Burns, 
pers. comm.). The forests were primarily within the Very 
Moist Coastal Western Hemlock (CWHvml and CWHvm2) 
and Moist Maritime mountain hemlock (MHmml) biogeo- 
climatic subzones. Two radio-tagged murrelets were 
repeatedly tracked to forest stands here (the third radio- 
tagged bird reported by Prestash and others [1992b] appeared 
to have lost its transmitter or died in the forest). Vegetation 
characteristics of these stands were derived from forest 
inventory maps. One stand was in sub-alpine hemlock/amabilis 
fir forest (400 m asl) with large mountain hemlock trees (37- 
46 m tall, estimated age >250 years), and the second in a low 
altitude (80 m) moss-covered bog-forest dominated by western 
red cedar (28-37 m, estimated 141-250 years old). 



156 



USDA forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



Murrelets were also studied in subalpine forests in the 
Caren Range. Sechelt Peninsula (Jones 1992). Dominant 
trees were mountain hemlock and yellow cedar. This is very 
old forest and one cedar stump was 1717 years old. Detection 
frequencies from scattered stations in June and July in 1991, 
1992 and 1993 averaged 13.9 ± 13.8 (s.d.; n = 27; range 1- 
61). 17.6 ± 16.7 (17; 0-45), and 20.3 ± 13.7 (54; 0-57), 
respectively (P. Jones, pers. comm.). Vegetation was not 
analyzed in detail. A nest was found here in a yellow cedar 
in 1993 (Jones 1993). 

High densities of murrelet detections (mean 24.4 ± 20.7 
s.d.. range 9-85, n = 12) were obtained at Tsitika Creek 
station between 29 June and 15 July 1991 in the lower Tsitika 
Valley, northeastern Vancouver Island (MacDuffee and others 
1993). A second station nearby, affording less visibility, 
yielded only 1-4 detections in two surveys in this period. 
Western hemlock (mean d.b.h. = 73 cm), western redcedar 
(117 cm), amabilis fir (75 cm) and Sitka spruce (112 cm) 
made up 60 percent. 18 percent, 16 percent and 7 percent, 
respectively, of the trees with d.b.h. >7.5 cm in this stand. 

Vegetation analysis has been done in Carmanah-Walbran. 
Vancouver Island in conjunction with murrelet surveys in 
1990-1993 (Burger 1994, Manley 1992, Manley and others 
1992). This is an area of relatively unfragmented valley- 
bottom old-growth, dominated by western hemlock (47 percent 
of all sampled stems >10 cm d.b.h.; 37.7 percent of combined 
basal area), amabilis fir (41.8 percent; 19.2 percent), Sitka 
spruce (8.4 percent; 33.3 percent), western red cedar (2.6 
percent; 9.7 percent) with a few red alder. Six nests have 
been found in this area, five in large Sitka spruce (d.b.h. 



range 1.33-3.7 m) and one in a large western hemlock 
(d.b.h. 2. 1 m). Manley (1992) found that murrelet detections 
at six stations were positively correlated with combined 
basal areas of hemlock and spruce, and negatively correlated 
with combined fir and cedar. Burger (1994) used a larger 
sample (11 stations in 1991, 12 in 1992) and considered a 
wider range of habitat variables, including stem densities 
and basal areas of all species, combinations of species, snags 
and trees >1 m d.b.h.. He found the same patterns as Manley, 
but the only significant correlation was a negative relationship 
between detection rate and stem density of hemlock in 1991 
(and not 1992). Burger (1994) concluded that the habitat 
variables measured were too coarse, and detection rates too 
variable, to detect subtle variations in suitability in this 
relatively homogeneous watershed. All of the stations were 
clearly in suitable nesting habitat, and occupied behaviors 
had routinely been recorded at all stations (fig. 4). 

Manley and others (1994) sampled 14 sites in old- 
growth forest in the Megin Valley, central Vancouver 
Island. These were grouped into sites dominated by western 
hemlock (4 sites), western red cedar (4), Sitka spruce (5) 
and amabilis fir (1), although all sites supported a variety 
of these large trees. Analysis of detection frequencies in 
June and July 1993 showed that the spruce sites had 
significantly lower detection rates than either cedar or 
hemlock, but cedar and hemlock did not differ significantly 
(table 1). The differences disappeared when only occupied 
detections were considered, because spruce sites had higher 
proportions of occupied detections (14 percent) than 
hemlock (4 percent) and cedar (3 percent). Average tree 



Table 1 — Mean (s.d.) detection frequencies of Marbled Murrelets in three forest types in the Megin Valley, central 
Vancouver Islands in June and July 1993 (from Manley and others 1994) 





Mixed forests dominated by: 




Parameters 


Spruce 


Cedar 


Hemlock 


Significant differences* 


Total detections 










June 


12.75 (8.75) 


38.0 (35.29) 


27.56(13.61) 


Cedar>Spruce (Z = 2.28, P < 0.02) 
Hemlock>Spruce (Z = 2.65, P < 0.01) 


July 


13.36(8.3) 


27.13(9.08) 


19.56(10.1) 


Cedar>Spruce (Z = 1 .96, P < 0.02) 
HemlocloSpruce (Z = 3.33, P < 0.01 ) 


Occupied detections 










June 


1.44(2.37) 


2.00 (4.50) 


1.11(1.76) 


None 


July 


1.82(2.74) 


0.25 (0.46) 


0.56(1.13) 


None 


No. of stations 












4 


4 


5 




No. of surveys 










June 


16 


8 


9 




July 


10 


8 


9 





* Multiple Kruskal- Wallace comparisons 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



157 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



diameter and total basal area of trees ranged from 46 to 
123 cm, and 5.9 to 25.3 m 2 per 0.9 ha plot, respectively. 
Frequencies of occupied detections were positively 
correlated with both mean tree diameter (r * 0.729, n = 1 5, 
P < 0.01) and basal area (r = 0.585, n = 15, P < 0.05), but 
frequencies of all detections showed no significant 
correlations (Manley and others 1994). These data suggest 
that the murrelets were more sensitive to tree size than to 
tree species composition in these old-growth forests. 

There have been no analyses of the effects of stand 
size, edge effects or stand isolation on Marbled Murrelets in 
British Columbia. 

Effects of Epiphytic Mosses and Mistletoe 

All nine nests known for British Columbia were on 
platforms of epiphytic mosses. Dense mosses were associated 
with the large trees in those vegetation groups in which 
detection frequencies were highest in the Queen Charlotte 
Islands (Rodway 1993a). In Carmanah-Walbran watersheds, 
Burger (1994) found no correlation between murrelet detection 
frequency and estimated moss cover per site, but the trees in 
all of the sample plots were well endowed with mosses and 
this was not a limiting factor for the murrelets here. 

None of the nine nests found in British Columbia were 
associated with mistletoe. Murrelet detection frequencies 
were not correlated with mistletoe index (Hawksworth 1977) 
in Carmanah-Walbran in 1991 (11 sites) or 1992 (12 sites), 
and moss-covered boughs provided many more potential 
nest sites than mistletoe in these large trees (Burger 1994). 

Predator Abundance 

I found no records of predation of Marbled Murrelets 
from British Columbia, but did not review all the raptor 
literature. Marbled Murrelets were absent from prey remains 
of Bald Eagles (Haliaeetus leucocephalus) found beneath 
35 nests (which included 145 bird carcasses) in the Gulf 
Islands (Vermeer and others 1989a) and 17 nests (33 bird 
carcasses) in Barkley Sound (Vermeer and Morgan 1989). 
Jones (1992) reported that murrelets fell silent and 
disappeared for 10 minutes when a large owl (probably 
Barred Owl [Strix varia]) appeared. 

Bryant (1994) tested the effects of egg predators in 
montane western hemlock-mountain hemlock forest in 
central Vancouver Island, using 120 artificial nests, each 
with three quail eggs, placed on the ground or in trees at 
eye level. He found that 43 percent of nests (52 percent of 
eggs) were damaged or removed in the first week, and 87 
percent (91 percent eggs) after two weeks. The survival of 
both nests and eggs placed in trees was significantly higher 
with increasing distance from the forest edge, after both 7 
and 14 days (fig. 5). Nests of Marbled Murrelets are much 
higher in trees and better camouflaged than these 
experimental nests, and so would not necessarily experience 
the same levels of predation. Nevertheless, these results 
indicate a strong edge effect of nest predation, suggesting 
that fragmentation of forests exposes Marbled Murrelet 



nests to increased predation. Steller's Jays (Cyanocitta 
stelleri), Gray Jays (Perisoreus canadensis) and Common 
Ravens (Corvus corax) were likely predators of tree nests 
in this experiment. These corvids did not appear in Bryant's 
census transects often enough to determine their distribution 
(Bryant, pers. comm.). 

These results are consistent with the conclusions reached 
by Paton (1994). In a critical review of 14 studies, he found 
strong evidence that avian nest success was reduced by 
predation and parasitism near habitat edges. Increased 
predation of natural and artificial (experimental) nests was 
most marked within 50 m of forest edges. In addition, nest 
success was consistently correlated with habitat patch size. 
There were apparently no studies in old-growth forest in the 
Pacific Northwest, nor did any studies consider nests as 
high in trees as those of the Marbled Murrelet. Studies on 
the effects of edges and habitat fragmentation on nest success 
of Marbled Murrelets are clearly a priority in areas with 
intensive logging. 

Assessing Marbled Murrelet Habitat 
Quality in British Columbia 

Conservation and Management Requirements 

Marbled Murrelets appear to nest in scattered forest 
locations over a vast area in coastal British Columbia 
(Campbell and others 1990, Rodway 1990, Rodway and 
others 1992). There is a growing need to identify and preserve 
nesting habitat, particularly in the many areas facing clearcut 
logging. Unlike the situation to the south in the United 
States, identification of occupied stands has not guaranteed 
protection in British Columbia because Canada lacks an 
Endangered Species Act to enforce strict protection of habitat, 
and neither federal nor provincial governments are likely to 
block all commercial logging in occupied stands. Only the 
most valuable nesting habitat is likely to be preserved outside 
parks, and measures to identify such habitat are urgently 
needed. At least two categories of forest need to be considered 
for immediate preservation: areas supporting many breeding 
birds which make up a significant proportion of the provincial 
murrelet population; and forest patches supporting remnant 
populations in areas severely affected by habitat loss. The 
first is important for maintaining a large, viable breeding 
population of murrelets and the second to maintain a wide 
breeding range and genetic diversity. 

Efforts to identify high quality habitat in British Columbia 
are at a very early stage. The huge areas involved and 
paucity of resources for surveying murrelets make it unlikely 
that the intensive multi-year surveys covering 12-30 ha, 
which are recommended for identifying occupied stands 
(Ralph and others 1994) will be widely implemented for 
short term management in British Columbia. As an interim 
measure, forest and wildlife managers will need general 
guidelines on the quality of forest stands being considered 
for logging. Intensive surveys can then be focused on the 
forest stands with greatest potential as nest sites. 



158 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



After 7 days 



After 14 days 



Survival of nests (N = 10 per transect 




15 100 250 400 

Distance from forest edge (m) 



550 



Survival of eggs (N = 30 per transect) 




15 100 250 400 

Distance from forest edge (m) 



550 



Figure 5 — Survival of artificial nests, each containing three quail eggs, placed at 
eye level in trees in transects laid out at various distances from the forest edge in 
montane western hemlock-mountain hemlock forest in central Vancouver Island, 
1992 (data from Bryant 1994). Nest "survival" meant the nest was in good 
condition with at least one undamaged egg, egg survival was the count of 
undamaged eggs. 



Use of Detection Frequency to Delineate Marbled 
Murrelet Habitat 

Standardized pre-dawn surveys provide indications of 
relative nesting density (Ralph and others 1994), although 
the relationship between the number of detections per survey 
and the density of nesting pairs has not been established and 
is likely to vary among sites and through the season (Rodway 
and others 1993a,b). As a first approach I have compared the 
frequency of detections among a wide range of survey stations 
from three sources: (1) the Queen Charlotte Islands (158 
surveys at 50 sites in 1990; Rodway and others 1991, 1993a) 



(2) a large sample of watersheds throughout most of 
Vancouver Island (471 surveys at 151 sites in 1991; Savard 
and Lemon in press); and (3) intensive surveys made over 
four years (1990-1993) at 12 sites in Carmanah Valley, two 
in the Walbran Valley and one at Nitinat Lake (Burger 
1994). At each site (in some of the Queen Charlotte Islands 
surveys, a site included several stations), the mean frequency 
of detections per morning survey was calculated for the 
period 1 May through 31 July. Occupied detections (Ralph 
and others 1994) could not be analyzed separately since 
these were not given in all reports. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



159 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



The percentage of the sampled sites in which the 
mean frequency of detections exceeded a given threshold 
was then plotted (fig. 6). This should facilitate ranking a 
particular site, relative to other sites, or guide decisions on 
how important surveyed sites might be on a provincial or 
regional basis. The trends in the Queen Charlotte Islands 
and on Vancouver Island were surprisingly similar. These 
indicate, for example, that about 18 percent of all sites in 
these areas had mean densities exceeding 40 detections 
per survey. If a manager decided to preserve all sites 
above this threshold, then one would expect about 18 



percent of the potential sites to be included. These trends 
should obviously only be used as guides, since some low- 
density sites might be important in places where there are 
few high quality sites. 

These data were derived from relatively few surveys 
(means for Queen Charlotte Islands and Vancouver Island 
were 3.2 and 1.6 surveys per site, respectively), made in a 
single year (1990 and 1991, respectively). By contrast, the 
surveys made in Carmanah-Walbran-Nitinat used fewer sites, 
but were much more intensive (mean 3 1 .4 surveys per site) 
and covered four years. Not surprisingly, the threshold pattern 



100 
90 
80 
70 
60 - 
50 
40 
30 + 
20 
10 





Queen Charlotte Is. 
Vancouver Island 
Carmanah-Walbran 



i i i i i i i i i i i i i i i i i i > i i i t i i i i i i 1 t ~T i i r»i i i i i 



10 20 30 40 50 60 

Threshold (mean no. detections per survey) 



70 80 



Queen Charlotte Is. 
Vancouver Island 



Carmanah-Walbran 

1990-1991 

Carmanah-Walbran 

1992-1993 




20 30 40 50 60 

Threshold (mean no. detections per survey) 



80 



Figure 6 — A: plot of the percentage of sites in which the mean frequency of Marbled Murrelet detections 
exceeded the thresholds on the x-axis. Data from the period 1 May through 31 July in the Queen Charlotte 
Islands (1 58 surveys at 50 sites in 1 990; Rodway and others 1 991 ) , Vancouver Island (209 surveys at 1 51 
sites in 1 991 ; Savard and Lemon in press), and Carmanah-Walbran-Nitinat (471 surveys at 1 5 sites in 1 990- 
1 993; Burger 1 994). B: the same plot as A, but with the Carmanah-Walbran-Nitinat data separated into two 
periods: 1990-1991 (176 surveys at 12 sites) and 1992-1993 (297 surveys at 14 sites). 



160 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burger 



Chapter 16 



Inland Habitat Associations in British Columbia 



differed from the previous studies, showing a smaller 
proportion of sites at each extreme (fig. 6a). These results 
emphasize that the single-year Queen Charlotte Islands and 
Vancouver Island surveys provide only rough guides to the 
expected patterns in a specific area. 

The effect of year-to-year variability in detection 
frequency can be clearly seen when the Carmanah-Walbran- 
Nitinat data are split into two periods (fig. 6b). The first 
(1990-1991) was a period of normal sea temperatures and 
high murrelet detections in the Carmanah-Walbran-Nitinat 
forests, whereas the second (1992-1993) covered two years 
with unusually high inshore sea temperatures and low murrelet 
activity in parts of the forest (Burger 1994). The resultant 
threshold patterns are quite different, showing that variable 
factors affecting murrelets (such as El Nino effects) must be 
considered when habitats are assessed on the basis of detection 
frequency. If, for example, forest managers set a threshold 
of 30 detections per survey to delineate optimal habitat, then 
this would cover 50 percent of all sites sampled in the good 
years (1990-1991), but only 7 percent of the same sites in 
poor years (1992-1993). 

In order to avoid such problems, managers would need 
to be very conservative and use relatively low thresholds 



(e.g., means of 10 or 20 detections per survey) to delineate 
high-quality habitat requiring preservation. Comparisons 
among sites of the mean detection frequencies provides only 
a crude estimation of the quality of a stand, particularly if 
only one or two intensive surveys are made in a single 
season. A more meaningful analysis would use the relative 
frequency of occupied behaviors recorded over at least two 
years (Ralph and others 1994), and surveys in British 
Columbia should be directed towards this goal. 

Acknowledgments 

Preparation of this chapter was funded by the British 
Columbia Ministries of Forests (Research Branch) and 
Environment, Lands, and Parks (Wildlife Branch); I thank 
Brian Nyberg and Don Eastman for their support. I thank 
Rick Burns, Andy Derocher, Andrea Lawrence, Moira Lemon, 
David Manuwal, Ken Morgan, Lynne Prestash, Martin Raphael 
for valuable comments. Unpublished material was provided 
by Andrew Bryant, Rick Burns, Paul Jones (Friends of Caren), 
Moira Lemon (Canadian Wildlife Service), Irene Manley, 
Misty MacDuffee (Western Canada Wilderness Committee), 
and Lynne Prestash. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



161 



Chapter 17 

Inland Habitat Associations of Marbled Murrelets in 
Western Washington 



Thomas E. Hamer 1 



Abstract: Little research has been done to quantify and describe 
the structural characteristics of forest stands that are associated 
with Marbled Murrelet (Brachyramphus marmoratus) nesting in 
the Pacific Northwest. Vegetation measurements and murrelet 
surveys to determine occupancy were conducted in stands located 
throughout western Washington. I used logistic regression to con- 
trast stand attributes between occupied (n = 64) and unoccupied (n 
= 87) stands. The probability of occupancy of an old-growth stand 
increased with increasing total number of potential nest platforms, 
percent moss coverage on the limbs of dominant trees (>81 cm 
d.b.h.). percent slope, the stem density of dominant trees, and the 
mean d.b.h. of western hemlock. The probability of occupancy of 
a stand decreased as lichen coverage on the limbs of dominant 
trees, stand elevation, and canopy closure increased. Mean detec- 
tion rates and the percent of stands surveyed and verified as 
occupied declined sharply with an increase in elevation over 1,067 
m. and for stands >63 km from salt water. The relationship of the 
number of potential nest platforms and elevation to the probability 
of occupancy was best explained by comparing the structural 
characteristics of old-growth trees for the five conifer species 
available for nesting. Land management activities that reduce or 
affect the number of potential nest platforms/ha. composition of 
low elevation conifers, moss cover on tree limbs, stem density of 
dominant trees (>81 cm d.b.h.), or canopy closure, would reduce 
the quality of a site as nesting habitat for murrelets. Reproductive 
success should be used as a measure of habitat suitability in future 
studies by intensively studying occupied stands that have high 
detection rates of Marbled Murrelets and locating a sample of 
active nests to observe. 



The research attempts to quantify and describe the 
structural characteristics associated with Marbled Murrelet 
{Brachyramphus marmoratus) nesting habitat have examined 
the general relationship between murrelet abundance and 
stand age. stand size, and tree size. A more specific model 
describing habitat is needed for a variety of reasons. A 
model would help ( 1 ) assess the relative impacts that forest 
management practices and associated activities will have on 
the quality of murrelet nesting habitat, (2) evaluate the relative 
suitability of a forest stand as nesting habitat for murrelets, 
(3) more accurately map suitable habitat, (4) understand 
how to speed the development of suitable habitat to meet 
long-term objectives for maintaining or increasing murrelet 
populations, (5) attempt to fashion habitat enhancement 
techniques or mitigation measures, and (6) plan future habitat 
research studies. 



' Research Biologist, Hamer Environmental. 2001 Highway 9, Ml 
Vernon. WA 98273 



Studies specifically addressing the forest habitat 
associations of Marbled Murrelets in Washington were 
initiated in 1990 and continued through 1993. A 1990 study 
examined the association of murrelets to four broad habitat 
categories and recorded the distribution and abundance of 
murrelets within an entire drainage basin, beginning at the 
Cascade crest and ending at its terminus with the Puget 
Sound (Hamer and Cummins 1990). An analysis of murrelet 
detection rates relative to the percent of old-growth forest 
available on the landscape was also conducted in this study. 
Studies from 1991 to 1993 focused on describing and 
analyzing the structural differences between old-growth stands 
occupied by murrelets and unoccupied old-growth stands. 
The results of these structural analysis are presented in this 
paper. In addition, a landscape analysis examining the 
attributes associated with stands occupied by Marbled 
Murrelets was completed in Washington in 1994 (Raphael 
and others, this volume). 

Methods 

Landscape Characteristics 

Detection Rate Comparisons 

For the comparison of Marbled Murrelet detection 
(murrelet detections/survey morning) and occupancy rates 
(number of stands surveyed and verified as occupied/number 
of stands surveyed) with respect to elevation, inland distance, 
and physiographic province, 262 old-growth stands were 
used. To investigate the effect of elevation on murrelet 
detection and occupancy rates, the mean detection rate and 
the percent of old-growth stands found occupied by murrelets 
were averaged for each 150-m interval in elevation ranging 
from to 1 ,500 m. To determine the effect of inland distance 
on habitat use by murrelets, the mean detection rate and 
percent of old-growth stands verified as occupied were 
averaged for inland distances using 15 -km intervals ranging 
from to 95 km. The mean detection rate and the percent of 
stands surveyed and verified as occupied were also used as 
measures of the use of a region by murrelets. The 
physiographic provinces we used for data comparisons are 
those described by Franklin and Dyrness (1973). Of the 262 
old-growth stands in this analysis, 132 stands occurred in the 
North Cascades Province, 32 in the South Cascades, 80 on 
the Olympic Peninsula, 8 in the Coast Range (southwest 
Washington), and 10 stands in the Puget Trough Province. 

Inland surveys for Marbled Murrelets were conducted 
using standardized survey techniques developed by the Pacific 
Seabird Group Marbled Murrelet Technical Committee (Ralph 



USDA Forest Service Gen. Tech. Rep. PS W- 152. 1995. 



163 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



and others 1994). Single observers visited each stand three 
or more times during the breeding season ( 1 May-5 August) 
recording observations during a 2-hour dawn survey period 
each visit. Mean detection rates for each stand were calculated 
by dividing the total number of detections by the number of 
survey visits. To standardize this calculation, stands with <3 
visits were not used in the analysis. These stands would not 
have had enough survey effort to determine occupancy with 
sufficient likelihood. For sites with >4 visits, survey visits 
were removed by selecting those four visits that best 
represented the seasonal timing of surveys recommended by 
the Pacific Seabird Group survey protocol. This helped 
standardize the selection of surveys in order to equalize the 
survey effort between stands. Therefore, survey effort was 
standardized by using only three or four visits for each stand 
used in the analysis. 

Occupied sites were defined as those stands with birds 
observed flying through the canopy, in or out of the canopy, 
birds observed landing or perched in trees, or stands with 
murrelets observed circling over the canopy (Ralph and 
others 1994). Occupied sites also included stands where nest 
platforms, murrelet egg shells, or juveniles had been found. 
Unoccupied sites included stands with birds present, but 
where no occupied or below canopy behaviors were observed, 
and stands where birds were not detected. 

Stand Characteristics 

Old-growth stands were included in the study if they 
met the definition of old-growth developed by the Washington 
Department of Wildlife Remote Sensing Program. Old-growth 
stands were defined as having at least 20 dominant overstory 
trees per hectare that were >8 1 cm diameter at breast height 
(d.b.h.). Co-dominant trees were >40 cm d.b.h. The presence 
of at least 2 canopy layers was also required. 

Vegetation Quantification 

A total of 38 attributes describing forest characteristics 
were used in the analysis (table 1). Observers were trained 
during a 3-day period to ensure forest variable measurements 
and estimates were performed consistently by all crew 
members. Vegetation data was not obtained from all the 
stands that were surveyed, therefore the sample size for the 
vegetation analysis was less than the number of stands used 
in the comparisons of mean detection and occupancy rates. 
Vegetation measurements were obtained from 64 occupied 
and 87 unoccupied old-growth stands located throughout 
western Washington for a total sample size of 151 stands. 

The sample size of stands where vegetation data was 
collected was variable in each physiographic province (table 
2). Old-growth stands in the North Cascades and Puget 
Trough Physiographic Provinces were selected systematically 
to represent a range of elevations, forest zones, and geographic 
areas. One to several stands were selected from each drainage 
depending on drainage size and access. Old-growth stands in 
the Olympic, South Cascades, and Coast Range Physiographic 
Provinces were selected in a opportunistic manner, primarily 



from a need to conduct surveys for Marbled Murrelets in 
certain stands because of impending forest harvest plans or 
other land management projects. The North Cascades and 
Olympic Peninsula physiographic provinces contained the 
largest proportion of sites because these provinces were 
areas where research had been conducted earlier and more 
intensively. 

Because murrelet detection rates were found to decline 
with increasing inland distance, not all stands that were 
surveyed were used in the statistical analysis. Some stands 
may have possessed all the appropriate structural features 
required to produce suitable nesting habitat, but were 
unoccupied because the inland distance was too great. To 
avoid misinterpreting study results, only stands <61 km 
from salt water were used in the vegetation analysis. I arrived 
at this value by examining the relationship between murrelet 
abundance and the inland distance of stands. 

Sites <0.8 km from salt water were not used in any 
analysis. Over the last three years a total of nine unoccupied 
sites have been located in Washington <0.8 km from salt 
water, with what appears to be excellent murrelet nesting 
habitat. This included five sites from southwest Washington 
and the Puget Trough, and four sites from the San Juan 
Islands. Murrelets may avoid using these stands because of 
their exposure to wind and coastal storms, or because of the 
presence of a higher number of predators such as gulls 
(Larus spp.), and crows and ravens (Corvus spp.). 

Survey stations were located in or adjacent to old-growth 
stands with a minimum stand size of 50 ha. This is an area 
encompassed by a circle with a 0.4 km radius and was 
therefore the sampling unit used for the study. From field 
experience I felt that this would be the approximate area an 
observer could detect murrelets on the landscape, and also 
prevented the surveying of small old-growth stands in heavily 
fragmented areas. These smaller stands may have a lower 
abundance of murrelets because they lack a sufficient amount 
of habitat, rather than a deficiency in any particular structural 
feature of the forest. Stand size was not included as a variable 
in the study design. This was due to the large number of 
observation stations per stand needed to successfully measure 
this effect, and the large sample of stands required for the 
statistical design. Although the influence of stand size on 
murrelet abundance is a vital piece of information required 
by land managers, more extensive research will be needed to 
evaluate this variable. The mean stand size or age of the 
stands sampled was not determined. 

The forest vegetation was measured using one 25-m 
radius plot for each old-growth stand being surveyed. The 
exact location of the plot was chosen by placing it in an area 
where flight behaviors below the canopy indicated possible 
nesting or, in other stands, in an area with the highest murrelet 
activity. For stands with no activity, the plot was located in 
an area with the highest stem density and largest basal area 
of old-growth trees. Therefore, even in areas with no activity, 
the highest quality old-growth available was selected to 
represent the stand thus establishing a conservative analysis. 



164 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



Table 1 — Definitions and units of measurement for each habitat variable used in the statistical comparison of occupied versus 
unoccupied murrelet stands in western Washington, 1991-92. A dominant tree was >S1 cm diameter at breast height t d.b.h.) 



Variable 


Definition and units of measurement 


Aspect 


Major aspect of the plot in degrees 


Basal area 


Basal area (m 2 ) of all dominant trees (>81 cm d.b.h.) in a 25-m radius plot 


Canopy closure 


Percentage of plot occupied by the crowns of live trees over 10 m in height 


Canopy height 


Mean tree height (m) of 10 trees measured per plot 


High comp. 


Percent composition of silver fir and mountain hemlock 


Low comp. 


Percent composition of Douglas-fir, western hemlock, western red cedar and Sitka spruce 


Nest comp. 


Percent composition of those tree species selected for nesting by murrelets in Washington and 




Oregon including Sitka spruce, Douglas-fir, and western hemlock 


Silver fir comp. 


Percent composition of silver fir 


Sitka spruce comp. 


Percent composition of Sitka spruce 


Douglas-fir comp. 


Percent composition of Douglas-fir 


Western red cedar comp. 


Percent composition of western red cedar 


Western hemlock comp. 


Percent composition of western hemlock 


High d.b.h. 


Mean db J>. (cm) of silver fir and mountain hemlock 


Low d.b.h. 


Mean d.b.h. (cm) of of Douglas-fir, western hemlock, western red cedar and Sitka spruce 


Mean d.b.h. 


Mean d.b.h. (cm) of all dominant trees measured per plot 


Nest d.b.h. 


Mean d.b.h. (cm) of tree species selected for nesting by murrelets in Washington and Oregon 


Silver fir d.bJL 


Mean d.b.h. (cm) of silver fir 


Sitka spruce d.b.h. 


Mean d.b.h. (cm) of Sitka spruce 


Douglas-fir d.b.h. 


Mean d.b.h. (cm) of Douglas-fir 


Western red cedar d.b.h. 


Mean d.b.h. (cm) of western red cedar 


Western hemlock d.b.h. 


Mean d.b.h. (cm) of western hemlock 


Mountain hemlock d.b.h. 


Mean db.h. (cm) of mountain hemlock 


Distance to saltwater 


Closest distance (km) from the plot to salt water 


Ecozone 


Geographical areas of similar environments (Henderson and others 1989, 1991 ) 


Elevation 


Plot elevation (m). 


Forest zone 


A classification method for determining plant association based on vegetation series of tree species 




present (Henderson and others 1989, 1991) 


Latitude 


Latitude of the plot to the nearest minute 


Mean lichen 


The mean amount of lichen per plot based on an index of lichen coverage on the limbs of all 




dominant trees 


Mean mistletoe 


The mean amount of mistletoe per plot based on an index of mistletoe abundance 




(Hawksworth 1977) 


Mistletoe number 


The total number of trees/hectare infected with mistletoe 


Mean moss 


An index of moss coverage on the platforms of all dominant trees 


Percent moss 


The percent moss coverage on the limbs of all dominant trees in a plot 


Platforms/ha 


The total number of potential nest platforms/ha over 15 m in height and 18 cm in diameter 


Platform total 


The total number of platforms from all dominant trees measured within and outside the plot 


Platforms/tree 


Mean number of potential nest platforms per tree 


Percent slope 


Percent slope of plot 


Slope position 


Position of stand on slope: (1) lower 1/3: (2) middle 1/3; and (3) upper 1/3 


Stem density 


The number of dominant trees/hectare 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



165 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



Table 2 — Sample size of stands used in the stepwise logistic regression analysis, listed by physiographic province and stand status. 





Number of sites 






Stand status codes 1 






Total 


occupied 


Total 




Physiographic province 





1 


2 


3 


4 


5 


unoccupied 


North Cascades 


84 


20 


1 


3 


16 


28 


16 




40 




44 


South Cascades 


17 


1 








1 


6 


9 




2 




15 


Olympic Mountains 


45 


3 


1 


1 


13 


20 


7 




18 




27 


Southwest Coast 


5 








1 


3 





1 




4 




1 


Puget Trough 

































Total 


151 


24 


2 


5 


33 


54 


33 




64 




87 



1 Stand status codes were: = Marbled Murrelets observed circling the stand; 1 = nest platform was located; 2 = juveniles, eggs, or eggshell fragments were 
located; 3 = murrelets were observed flying in the canopy; 4 = murrelets were detected in the area; and 5 = no murrelets were detected. Occupied stands included 
status codes 0-3 



Only dominant trees > 81 cm in diameter were included 
in all vegetation measurements except for canopy closure and 
forest vegetation series. In addition, only conifer trees were 
included in the measurement for each variable, except canopy 
closure. To ensure that a large sample of tree measurements 
for each variable were recorded from each site, at least 20 
trees were measured at each plot. If 20 trees were not available 
within the plot, the nearest dominant trees to plot edge were 
selected to be measured until 20 trees were recorded. Trees 
selected outside the plot were included in the calculations for 
mean tree d.b.h., total number of potential nest platforms, 
potential nest platforms/tree, lichen coverage, dwarf mistletoe 
(Arceuthobium spp.) infestation, moss (Isothecium spp.) 
coverage on potential nest platforms, and all tree species 
composition variables. Trees within the plot were used to 
calculate basal area, forest zone, vegetation series, canopy 
closure, mean canopy height, and all other measurements. 

Ecozones, geographical areas of roughly similar 
environments, were delimited on the basis of the abundance 
and distribution of plant indicator species and are a general 
measure of the amount and kind of precipitation an area 
received. Ecozones were mapped in Washington by the USDA 
Forest Service (Henderson and others 1989, 1991). Ecozone 
represented the wettest part of the study area (457 cm or 
more of annual precipitation), whereas ecozone 13 was the 
driest (less than 203 cm). Each plot was given an ecozone 
classification based on its location. The vegetation series 
and forest zone were identified for each plot using standard 
protocol and field guides (Henderson and others 1989, 1991). 
The latitude and distance to nearest salt water for each site 
was measured using topographic maps with a scale of 
1:250,000. Latitude was measured to the nearest minute and 
distance to salt water to the nearest 0.4 km. 

The number of potential nest platforms (platform total) 
for each tree was estimated from one point near the tree 



where the maximum number of limbs could be seen. The 
observer counted the number of limbs or structures >15 m in 
height and >18 cm in diameter directly along the tree bole. 
All structures were counted; the observers did not make 
judgments as to the suitability of the platforms for nesting. 
These measurements were chosen because all of the 1 8 nests 
found at the time the index was developed were >27 m in 
height with the majority of nest limbs >20 cm in diameter. 
Therefore, limbs >18 cm seemed a reasonable threshold to 
use for the index. To practice estimating whether tree limbs 
were >18 cm, limbs of known diameters were observed from 
a 30-m distance. A total count of all potential nest platforms 
in a tree was not possible, so this measurement was treated 
as an index. Mistletoe blooms located away from the tree 
bole were not counted as platforms, since their abundance 
was measured using another index. Mistletoe infestation 
was rated for each tree following an index developed by 
Hawksworth (1977). The number of trees infected with 
mistletoe (mistletoe number) were summed for each plot. 

The percent cover of all epiphytes (moss and lichens 
separately) on the surface of the limbs of dominant trees was 
recorded for each tree by estimating the average cover for all 
limbs using five categories, including 0-20 percent, 21-40 
percent, 41-60 percent, 61-80 percent, and 81-100 percent 
cover. Each tree was placed in a category and an average 
calculated for all trees in the plot for both lichen and moss 
coverage. Moss cover (mean moss) was estimated for potential 
nest platforms only. Lichen cover (mean lichen) on the 
surface of the limbs of dominant trees was estimated by 
averaging all the limbs of the tree. The average percent moss 
coverage (percent moss) on all the limbs of dominant trees 
in the plot were also estimated to the nearest 5 percent, as an 
additional measure of moss abundance. 

Canopy closure was measured in a smaller 17.8-m plot 
by physically measuring all gaps in the canopy >4 m 2 in size. 



166 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Harner 



Chapter 17 



Inland Habitat Associations in Western Washington 



This was accomplished by estimating the distance between 
gap edges as if the canopy created vertical shadows on the 
ground. Trees <9 m tall were not considered a part of the 
canopy. Mean canopy height was calculated from 10 dominant 
trees in the plot using a clinometer. 

The percent composition and mean values for mean 
d.b.h.. height, basal area, number of potential nest platforms, 
moss cover, lichen cover, and mistletoe abundance were 
calculated for each tree species present on each plot. 

Statistical Model 

Stepwise logistic regression was used to compare the 
structural characteristics of occupied and unoccupied old- 
growth stands in Washington. A predictive model for the 
binary dependent variable, defined as occupied and 
unoccupied stands, was developed to help define those forest 
characteristics associated with murrelet nesting habitat. 

Logistic regression methods (SAS Institute, Inc. 1987) 
were used to develop a model for the binary dependent 
variable which was defined as occupied and unoccupied 
stands (Hosmer and Lemeshow 1989). Candidate independent 
variables were selected for inclusion in the model using the 
stepwise selection procedure. The P-value chosen for allowing 
a candidate variable to enter the model was 0.05. This value 
was also used as the criteria for retaining an independent 
variable in the model at the conclusion of each step. 

For the statistical analysis, all 38 forest variables were 
treated as continuous variables except for forest zone. Forest 
zone was divided into two categories, high-elevation, and 
low-elevation zones. High-elevation zone included stands 
located in silver fir (Abies amabilis) and mountain hemlock 
(Tsuga mertensiana) zones. Low-elevation zone included 
stands located in the western hemlock (Tsuga heterophylla), 
western red cedar (Thuja plicata). Douglas-fir (Pseudotsuga 
menziesii). and Sitka spruce (Picea sitchensis) zones. In 
addition, the variable ecozone was analyzed as a separate 
logistic stepwise model, because at a few sites the ecozone 
value could not be determined. 

Principal Components Analysis (PCA) (SAS Institute, 
Inc. 1987) was used to create a correlation matrix of all 
variables and to consider more complex interdependencies 
among the independent variables. The correlation matrix 
was used to gauge the degree of association and 
interdependence between pairs of variables. This helped 
determine if one variable could be used in the logistic 
regression model as a substitute for another highly correlated 
variable. The PCA was not definitive in identifying higher 
order dependencies in these data. 

Importance of Independent Variables 

Four methods were used to subjectively evaluate the 
relative importance of each variable to the model's ability to 
predict occupancy and the importance of each variable in 
describing the differences between occupied and unoccupied 
sites. The first method was to examine the initial chi-square 
values of each variable before they entered the model. The 



second technique involved examining the step in which a 
variable was selected by the model. Variables selected earlier 
in the stepwise selection procedure had more power in 
explaining the variation between occupied and unoccupied 
sites than variables selected later in the procedure or variables 
not selected at all. The third method involved examining the 
final chi-square values for each variable used in the model. 
The last technique examined the stability of a variable as the 
stepwise selection procedure of the model progressed. 
Unstable variables experienced large fluctuations in chi- 
square value as each new variable was selected in the stepwise 
procedure, because of high colinearity with other variables 
used in the model. 

Tree Characteristics 

The mean structural characteristics of old-growth trees 
for the six conifer tree species available for nesting by 
Marbled Murrelets in Washington were calculated by pooling 
the values for each variable measured for each tree species 
across all plots. These variables included mean d.b.h.. mean 
tree height, basal area, potential nest platforms/tree, percent 
moss coverage on limbs, percent lichen cover on limbs, and 
mistletoe abundance. This analysis was used to subjectively 
compare the structure and suitability of tree species in 
providing murrelet nesting habitat. 

Results 

Landscape Characteristics 

Distance to Salt Water 

Highest detection rates (5.9-9.5 detections/survey 
morning) in Washington occurred in intervals between 16 
km and 64 km inland, but declined to 0.85 detections/ 
morning at distances >63 km from salt water (fig. 1). To 
date, 98.5 percent of all detections have been recorded <64 
km inland, but this is partly due to the extensive survey 
effort that has occurred in this zone. The maximum distance 
at which birds were detected inland was at an occupied 
stand 84. 1 km from salt water, located on Irene Creek near 
the Cascade River Drainage in 1992 and 1993. The next 
farthest occupied stands were located 72 km and 74 km 
inland. Of the known occupied stands, 36 percent (n - 31) 
were located more than 47 km from the ocean. Nests were 
located an average of 1 6 km inland, with a maiimnm distance 
of 34 km (n = 6). Of the old-growth stands located between 
and 63 km inland, 20-54 percent were occupied (fig. 1). 
The percentage of occupied stands declined sharply after 63 
km. with only 13 percent of stands occupied >63 km from 
the ocean. 

Elevation 

In Washington, detection rates declined sharply with an 
increase in elevation over 1,067 m (fig. 2). The highest 
detection rates, which ranged from 4.3 to 9.2 detections/ 
survey morning, were recorded between sea level and 1,067 
m. Stands located above 1 ,067 m had mean detection rates 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



167 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 




0-15 16-31 32-47 48-63 64-79 80-95 
DISTANCE TO SALT WATER (KM) 



PERCENT OF STANDS OCCUPIED II MEAN DETECTIONS/MORNING 



Figure 1 — The percent of stands surveyed and verified as occupied, and the mean number of 
murrelets detected/survey morning, in relation to the distance of the stand from salt water. The 
sample of stands is from all the physiographic provinces in western Washington, 1991-93. 
Mean detection rates corresponded closely to occupancy trends. 




152 305 457 610 762 914 106712191372 1524 
ELEVATION (M) 



PERCENT STANDS OCCUPIED 



NUMBER OF STANDS SURVEYED 



Figure 2 — The percent of stands surveyed and verified as occupied in relation to stand 
elevation. The sample of sites is from all the physiographic provinces in western Washington, 
1991-93. 



168 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



<1.1 detections/morning. The highest occupied stand in 
Washington was located at 1,105 m in elevation, in the 
North Cascades Physiographic Province, near the upper 
headwaters of Crevice Creek. The highest occupied stand in 
the Olympic Peninsula Physiographic Province was located 
1,025 m in elevation near Spot Lakes on the Hood Canal 
Ranger District, Olympic National Forest The South Cascades 
physiographic province had an occupied stand 1,051 m in 
elevation, located 13 km south of Alder Lake in Lewis 
County, near the East Fork Little Creek drainage. More than 
98 percent of all detections in Washington were recorded 
below 1 ,067 m in elevation. 

Forest Type and Physiographic Province 

Forest types surveyed in Washington included stands 
dominated by western hemlock, Douglas-fir, Sitka spruce, 
silver fir, and mountain hemlock. These stands commonly 
had a large component of western red cedar. 

The mean detection rates for 229 old-growth stands 
were compared between the five physiographic provinces in 
Washington. The North Cascades Province had a mean 
detection rate of 23.5 detections/survey morning (n = 117 
sites, s.d. - 7.9) and 40 percent of the old-growth stands 
surveyed were verified as occupied. The Olympic Peninsula 
had a similar rate of 18.3 detections/survey morning (n = 67, 
s.d. = 8.2) and a 37 percent occupancy rate of old-growth 
stands. The South Cascades had a detection rate of 12.5 
detections/survey morning (n = 30, s.d. = 15.7), and 10 
percent of the stands surveyed were occupied. The Puget 
Trough had the lowest detection rate of any province, with 



1.3 detections/survey morning (n = 7, s.d. = 0, and no 
occupied stands, but the number of stands sampled was 
small. The Southwest Coast had the highest detection rate 
(90.1 detections/survey morning; n = 8, s.d. = 7.4) and the 
highest occupancy rate (50 percent), but the number of old- 
growth stands surveyed was also small. 

Stand Characteristics 

Statistical Model 

The results of the logistic regression model results from 
the 1994 study gave a total accuracy rate of 74.2 percent for 
a predicted probability of occupancy for each stand analyzed. 
The classification accuracy of occupied stands was 67.2 
percent. The classification accuracy of unoccupied stands 
was 79.2 percent. Of the 32 stands with a predicted probability 
of occupancy >0.75, 74 percent were occupied. Of the 54 
stands with a predicted probability of occupancy of <0.25, 
93 percent were unoccupied. A total of 65 stands (43 percent) 
had probability values >0.25 and <0.75. 

Eight forest variables were included in the model by the 
stepwise logistic regression procedure. These variables best 
predicted occupancy of a stand by murrelets ( table 3). The 
stepwise selection procedure was completed in 10 steps. 

The probability of occupancy of an old-growth stand 
increased with increasing percent topographic slope, total 
number of potential nest platforms/ha, stem density of 
dominant trees, mean d.b.h. of western hemlock, and the 
moss coverage (percent moss) on the limbs of dominant 
trees (table 3). The probability of occupancy of a stand 
decreased with increasing stand elevation, canopy closure. 



Table 3 — Stepwise Logistic Regression Model of Marbled Murrelet habitat in western Washington. The eight variables are listed 
in order of their probability values 



Variable 


Regression coefficient 


Standard error 


Wald Chi-square 


Prob. > Chi-square 


Intercept 


-1.31820 


1.6352 


0.65 


0.41 


Percent slope 


0.04360 


0.0123 


12.47 


<0.01 


Platform total 


0.04330 


0.0151 


8.19 


•cO.01 


Elevation 


-0.00083 


0.0003 


8.16 


<0.01 


Stem density 


0.18590 


0.0668 


7.75 


<0.01 


Canopy closure 


-0.03340 


0.0130 


6.63 


0.01 


Western hemlock d.b.h. 


0.01610 


0.0065 


6.18 


0.01 


Percent moss 


0.02220 


0.0093 


5.72 


0.02 


Mean lichen 


-0.58700 


0.2726 


4.64 


0.03 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



169 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



and lichen coverage (mean lichen) on the limbs of dominant 
trees. Sites with a high probability of occupancy had a 
mean canopy closure of 86 percent. 

Importance of Independent Variables 

The step in which each variable was selected, the stability 
of variables through the stepwise procedure, the final chi- 
square values of variables used in the model, and the 
relationship between variables were used to subjectively 
assess the relative contribution of variables in predicting the 
probability of occupancy {table 3). The variables most 
correlated with occupancy of old-growth stands, included 
total potential nest platforms/ha, total percent moss cover on 
tree limbs, percent slope, mean d.b.h. of all dominant trees, 
mean lichen cover on tree limbs, stem density of dominant 
trees, elevation, canopy closure, mean d.b.h. of western 
hemlock, and percent composition of low elevation conifers. 

Describing Low- and High-Quality Habitat 

To begin to define what values would be considered to 
be the lower and upper thresholds for describing murrelet 
nesting habitat, the minimum, mean, and average values for 
each forest variable were calculated for occupied and 
unoccupied stands (table 4). Suitable murrelet nesting habitat 
was defined as sites with a high probability of occupancy. 
These stands had a mean topographic slope of 50 percent 
and were found at a mean elevation of 152 m. Stands with a 
high probability of occupancy also had a mean of 92 platforms/ 
ha, a stem density of 50 dominant trees/ha (>81 cm d.b.h.), 
83 percent canopy closure, 101 cm mean d.b.h. of western 
hemlock, 49 percent moss coverage on tree limbs, and a low 
index of lichen cover (table 4). 

Stands with a high probability of occupancy (>0.76) had 
minimum values of 10 platforms/ha, 29 dominant trees/ha, 
29 percent canopy closure, 85 cm mean d.b.h. of western 
hemlock, 5 percent moss cover, and 97 cm mean tree d.b.h.. 
These occupied stands were found at a maximum of 288 m 
in elevation. 

Tree Characteristics 

A comparison of old-growth tree characteristics for 
different conifer species in Washington indicated that old- 
growth Sitka spruce had most of the characteristics associated 
with known nest sites (Hamer and Nelson, this volume b). 
Sitka spruce had a higher mean d.b.h., taller height, higher 
number of platforms/tree, and higher moss coverage of the 
limbs than any of the five other conifers (table 5). On 
average, this species had more than two times as many 
platforms/tree than any other conifer species except Douglas- 
fir. Douglas-fir was second in having characteristics deemed 
suitable for murrelet use, with a similar number of platforms/ 
tree as Sitka spruce, a large height, high mean d.b.h., but a 
low moss coverage on the limbs. Western red cedar ranked 
third as a suitable nest tree choice with a large mean d.b.h., 
high basal area, 1 .4 platforms/tree, and one of the highest 
moss cover indexes. Western hemlock ranked fourth in the 



comparison but, as expected, has one of the highest mistletoe 
indexes of any tree species. Mountain hemlock ranked third 
and silver fir last. Both silver fir and mountain hemlock had 
a low mean d.b.h., low basal area, low number of platforms/ 
tree, and a higher lichen index. Silver fir had an average of 
only 0.81 platforms/tree. 

Discussion 

Landscape Characteristics 

Distance to Salt Water 

Because murrelets forage at sea and only carry single 
prey items to the nest, but can nest at long distances from the 
coast, the energetic requirements of flying inland to incubate 
eggs and feed young, places a limit on their inland breeding 
distribution and use of inland forests. Even with the potential 
problems of energetic expenditure, Marbled Murrelets 
displayed a great tolerance for using nesting stands located 
up to 63 km inland from the ocean. Almost all the habitat in 
the North Cascades and South Cascades Physiographic 
Provinces is located >42 km inland because of rural 
development and intensive forestry practices within the Puget 
Trough. Even with these long flight distances, some birds 
were passing occupied stands to fly farther inland. 

Breeding records also indicated that nesting is 
occurring at stands located long distances from salt water. 
A small downy chick was located on the ground along a 
trail on the east shore of Baker Lake in 1 99 1 , 63 km from 
the ocean (pers. obs.). Another downy chick was located 
45 km inland at Helena Creek, in Snohomish County (Reed 
1991). Six additional records of eggs, downy young, and 
fledglings found 29-55 km inland in Washington were 
compiled by Leschner and Cummins (1992a), and Carter 
and Sealy (1986). 

Elevation 

In general, stands found at higher elevations had a lower 
composition of conifer species reported to be used as nest 
trees. Murrelet nests have not been located in the higher 
elevation conifers such as silver fir or mountain hemlock in 
British Columbia, Washington, Oregon, or California, (Hamer 
and Nelson, this volume b). A negative association of murrelet 
abundance and stand occupancy to the occurrence of silver 
fir and mountain hemlock (high elevation tree species) is 
best explained by these species low mean d.b.h. and low 
number of platforms/tree (see Tree Characteristics). In 
addition, silver fir branches generally exit the trunk at sharp 
downward angles creating few level platforms. 

Forest Type and Physiographic Province 

All records of nests, eggs, eggshell fragments, and downy 
chicks in Washington have been associated with old-growth 
forests (n = 17) (Leschner and Cummins 1992a). In North 
America, fledglings have been found in a variety of unusual 
habitat types such as roads, airports, and rural areas (Carter 
and Sealy 1987b; Hamer and Nelson, this volume a). These 



170 



USDA Fprest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



Table 4 — Mean values for occupied murrelet stands in Washington calculated using stands with a predicted 
probability of occupancy >76 percent (o m 25). Mean values for unoccupied stands were calculated using sites with 
a predicted probability of occupancy <J 4 percent in = 44). Variables are listed in order of their initial c hi- square score 
before step 1 of the model, with the eight variables used to develop the logistic regression model listed first Final chi- 
square scores for the eight variables used by the model are listed in table 3 



Variable 


Predicted 

probability 

of occurrence 


Mean 


Minimum 


Maximum 


Probability > 
Chi-square 


Percent slope 


>0.76 


49.9 


3.0 


90.0 


0.05 




<0.14 


35.1 


3.0 


75.0 




Platform total 


>0.76 


212 


4.0 


65.0 


<0.01 




<0.14 


13 


0.0 


29.0 




Elevation 


>0.76 


152.4 


29.6 


288.0 


<0.01 




<0.14 


271.4 


27.9 


445.8 




Stem density 


>0.76 


50.0 


29.0 


89.0 


0.05 


(trees/ha) 


<0.14 


39.0 


0.0 


84.0 




Canopy closure 


>0.76 


82.6 


29.0 


98.0 


0.02 




<0.14 


81.1 


50.0 


100.0 




Western hemlock 


>0.76 


100.8 


84.7 


136.2 


0.02 


dUkfc. 


<0.I4 


983 


563 


135.2 




Percent moss 


>0.76 


49.1 


5.0 


82.0 


<0.01 




<0.14 


14.1 


0.0 


75.0 




Meand-b-h. 


>0.76 


131.7 


973 


169.7 


<0.01 




<0.14 


103.3 


58.7 


183.0 




Lowd-b.h. 


>0.76 


133.0 


97.3 


169.7 


<0.01 




<0.14 


107.0 


58.8 


183.0 




Platforms/ha 


>0.76 


92.0 


10.0 


183.0 


<0.01 




<0.14 


25.0 


0.0 


89.0 




Western redcedar 


>0.76 


153.9 


91.4 


2473 


<0.01 


d.b.h. 


«0.14 


122.0 


98.0 


1773 




Mean lichen 


>0.76 


1.4 


1.0 


3.1 


<0.01 




<0.14 


23 


1.0 


4.8 




Western red cedar 


>0.76 


45.1 


5.9 


100.0 


<0.01 


composition 


<0.14 


20.6 


6.7 


40.0 




Slope position 


>0.76 


1.6 


1.0 


3.0 


<0.01 




<0.14 


22 


1.0 


3.0 




Basal area 


>0.76 


14.3 


A2 


28.4 


<0.01 




<0.14 


8.1 


1.4 


17.7 




Platforms/tree 


>0.76 


2.0 


03 


5.7 


<0.01 




<0.14 


0.8 


0.0 


3.2 




Canopy height 


>0.76 


53.6 


37.2 


69.3 


0.01 




<0.14 


45.8 


26.1 


713 




Low composition 


>0.76 


92.4 


50.0 


100.0 


0.01 




<0.14 


773 


6.7 


100.0 




High composition 


>0.76 


26.7 


63 


50.0 


0.02 




<0.14 


49.7 


7.0 


100.0 




Mistletoe number 


>0.76 


15.0 


5.0 


39.0 


0.03 


(trees/ha) 


<0.14 


10.0 


5.0 


25.0 




Distance to 


>0.76 


38.2 


13 


62.8 


0.94 


saltwater 


<0.14 


38.5 


13 


62.5 





USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



171 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



Table 5 — Summary of seven characteristics measured for six species of conifers available as nest trees by the 
murreletin Washington state. Only trees >Hl cmd.b.h. were measured. The mean, range, and sample size are shown. 
See text for moss and lichen cover categories 









Tree 


Species 






Variable 


Sitka 


Douglas- 


Western 


Western 


Mountain 


Silver 




spruce 


fir 


red cedar 


hemlock 


hemlock 


fir 




n = 55 


« = 552 


n = 347 


n = 793 


n = 54 


n = 234 


D.b.h. (cm) 


163.1 


131.7 


143.0 


106.7 


103.0 


100.4 




91-326 


55-268 


81-290 


51-268 


55-140 


52-184 


Height (m) 


57.2 


58.3 


49.3 


47.4 


40.5 


50.8 




27-73 


18-85 


26-72 


15-76 


18-73 


23-69 


Basal area (m 2 ) 


2.3 


1.9 


4.6 


4.3 


2.2 


0.8 




0.6-8.4 


0-5.6 


0.5-6.6 


0.2-4.4 


0.2-1.5 


0.2-2.7 


Platforms/tree 


2.9 


2.3 


1.4 


1.2 


1.0 


0.8 




0-18 


0-13 


0-10 


0-19 


0-6 


0-5 


Moss index 


2.8 


1.5 


2.2 


2.0 


1.1 


2.4 




1-5 


1-5 


1-5 


1-5 


1-2 


1-5 


Lichen index 


2.5 


2.0 


1.2 


1.7 


2.6 


2.2 




1-5 


1-5 


1-5 


1-5 


1-5 


1-5 


Mistletoe index 


0.2 


0.1 


0.0 


1.1 


0.1 


0.2 




0-5 


0-2 


0-3 


0-6 


0-3 


0-4 



records indicate that fledglings may travel some distance 
before becoming grounded. 

Detection and stand occupancy rates increased with 
more older forest available on the landscape. For all provinces, 
the low detection and occupancy rates near the coast were 
probably due to the presence of large amounts of unsuitable 
or marginal habitat in the Puget Trough and near coastal 
lowland areas of the Olympic Peninsula. In a study 
encompassing the entire South Fork of the Stillaguamish 
River basin in northern Washington, significantly higher 
numbers of murrelets were observed in old-growth and 
mature forests than either rock/talus, clear-cut/meadow, or 
small saw/pole cover types (Hamer and Cummins 1990). 
Murrelet detection rates increased rapidly when the 
percentage of old-growth and mature forest cover types 
found within a 2,000-m-radius circle around each survey 
station made up more than 30 percent of the landscape. 
Mean detection rates for sites located in these areas ranged 
between 1 and 20 detections/morning (3c = 5.7; s. d. = 5.8). 
All sites with <30 percent old-growth and mature forest 
cover had <1.5 detections/morning ( x = 0.2; s.d. = 0.4). An 
analysis of the landscape features associated with occupied 
and unoccupied stands in Washington found that the amount 



of old-growth and large sawtimber available best predicted 
murrelet occupancy at the stand level (Raphael and others, 
this volume). Sites with a higher proportion of these mature 
forest classes were more likely to have evidence of nesting 
or occupancy than unoccupied sites. 

Stand Characteristics 

Statistical Model 

Overall the model correctly predicted occupancy on 
about 74 percent of the sites. However, this success rate 
may be biased because the same sites that were used to 
build the model were used to test it. Because the model 
treats occupancy as a categorical variable, individual sites 
that scored near 0.5 were difficult to judge. In these cases it 
was more convenient to think of occupancy as a continuous 
variable where the higher probability scores indicated more 
suitable habitat and a higher probability of being occupied 
by murrelets. Errors in the classifications of stands could 
be due to several factors: (1) some stands determined to be 
unoccupied from field surveys may have actually been 
occupied; (2) it is possible in some instances that birds may 
be occupying stands of marginal habitat and; (3) the 
vegetation sampling for some stands may have been 



172 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



inadequate to accurately reflect the true structure of the 
stand. These potential problems could be avoided by 
increasing the number of survey visits to a stand used to 
determine occupancy and increasing the vegetation sampling 
effort. More vegetation information from a larger number 
of independent occupied and unoccupied stands needs to be 
collected to validate the model. 

The results of the statistical model suggested that any 
land management activity that reduced or affected the number 
of potential nest platforms/ha, composition of low elevation 
conifers, moss cover on tree limbs, stem density of dominant 
trees, or canopy closure, would reduce the probability of 
occupancy of old-growth, and thus the suitability of an old- 
growth stand as nesting habitat for murrelets. Results from 
studies of murrelet habitat use to date have been derived 
from comparisons of stands occupied by murrelets to 
unoccupied stands, comparisons of stands receiving high 
use versus low use, or comparisons of nest trees and nest 
plots to random trees and plots. Although these can provide 
extremely useful descriptions and definitions of suitable 
habitat, they do not provide information on the habitat 
characteristics associated with successful nests. Information 
on the landscape and within-stand habitat characteristics 
that influence reproductive success is needed to fully 
understand murrelet nesting ecology and to model optimum 
habitat suitability for this species. Reproductive success should 
be used as a measure of habitat suitability in future studies 
by intensively studying occupied stands that have high 
detection rates of Marbled Murrelets and locating a sample 
of active nests to observe. A discussion of each variable used 
by the model follows. 

Total Platforms — Results suggest that if any variable 
were to be used solely to assess habitat quality, total platforms 
would be the best indicator. More potential nest platforms 
within a stand mean more nesting and hiding opportunities 
and a higher diversity of nest choices for the murrelet. Although 
the total number of platforms was important, I currently have 
few measures of platform quality. A examination of the limb 
diameters of Marbled Murrelet nests indicated higher use and 
possible selection for platforms >35 cm in diameter (Hamer 
and Nelson, this volume b). Some stands may have an 
abundance of smaller potential nest platforms that are only 
10-20 cm in diameter. These stands may be marginal nesting 
habitat because of the limitations of platform size. Future 
studies should include a measure of mean platform size when 
quantifying forest vegetation. 

The total number of potential nest platforms would be 
especially important if nest platforms within a stand were 
limited, the number of nesting stands available on the 
landscape were limited, or intraspecific competition occurred 
for nest platforms within a given area. It is unknown whether 
platforms meeting all the requirements for nesting are 
limited in availability in a typical old-growth stand. It has 
been assumed that nest platforms may be unlimited in old- 
growth stands (Sealy 1974), but an understanding of the 
structural requirements needed for a platform to be used by 



murrelets is required before an analysis of platform 
availability is possible. 

Total Moss — The presence of moss in the tree canopy 
was another important indicator of murrelet habitat. Although 
murrelets do not absolutely require moss as a nest substrate, 
the majority of nests have been located on moss (Hamer and 
Nelson, this volume b); the presence of moss may increase 
the number of potential platforms within a stand. Limbs with 
little or no moss coverage result in nest locations close to the 
trunk of a tree, which is usually the only area on a tree where 
debris such as needles and duff collect in sufficient quantities 
to form a thick substrate suitable for nesting, or where 
branches are large enough in diameter to create suitable nest 
platforms (pers. obs.). Other areas on the tree are usually too 
exposed to wind and other environmental influences to collect 
enough substrate to form a platform of suitable size. Thick 
mistletoe blooms are sometimes the exception to this 
observation. A high cover of moss creates a multitude of 
nest platform choices by providing substrate on many locations 
throughout a single limb, especially where there is suitable 
overhead cover and the limb is large enough to support a 
nest. In addition, the presence of a moss carpet essentially 
thickens the diameter of limbs, transforming limbs of marginal 
size into suitable nesting platforms. Moss is therefore related 
to the number of potential nest platforms of a stand. It is not 
known if one species of moss is preferred over others. 

Mean D.b.h. — Although not selected by the final 
regression model, mean tree d.b.h. had one of the highest 
initial chi-square values (16.2) and the chi-square values 
showed high stability through the selection process. The 
mean number of platforms/tree increased rapidly with an 
increase in tree diameter from 50 to 200 cm (fig. 3). No 
increase in the mean number of platforms was evident for 
larger trees that ranged from 220 to 300 cm in diameter. 
Suitable platforms were most commonly found in stands 
with larger tree sizes, as evidenced by a correlation of total 
platforms to mean tree d.b.h. (r = 0.60), but the relationship 
of these two variables was complex. The presence of larger 
trees alone did not always explain the presence of nest 
platforms. In Washington, there were abundant examples of 
large trees > 176 cm in diameter that contained no platforms. 
Other factors that can create platforms may include wind 
and insect damage, mistletoe brooms or other plant parasites, 
moss or larger quantities of duff, multiple overlapping tree 
limbs, natural limb deformities, and disease. Examples of 
80-year-old stands of western hemlock that are heavily infested 
with mistletoe and occupied by murrelets have been found in 
Oregon (Nelson, pers. comm.). Therefore, total platforms 
was the best indicator of suitable murrelet nesting habitat 
because it directly measures the nesting structures required 
by this alcid, whereas mean tree diameter measures the 
availability of platforms indirectly and with less accuracy or 
predictability. Still, most agencies and private timber 
companies have measures of mean tree diameter available 
for their stands, but no measures of platforms or structure. In 
attempts to force the model to use mean tree diameter, the 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



173 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 




101 126 151 176 201 226 
TREE DIAMETER (CM) 

Figure 3 — The mean number of potential nest platformsAree in relation to tree 
diameter (25-cm intervals) for western hemlock, western red cedar, Douglas-fir, and 
Sitka spruce trees in western Washington, 1 991 -93. Trees (n = 1 ,860) were sampled 
from 151 stands. 



model always re-selected a platform variable to replace mean 
tree diameter. Total platforms accounted for all the variation 
of mean tree d.b.h., but mean tree d.b.h. could not account 
for all the variation of total platforms. These results indicate 
that the structure of a stand is more important in predicting 
stand occupancy by murrelets than the size of the trees 
within the stand. 

Mean Lichens — The percent cover of lichens on tree 
limbs was negatively correlated with the percent cover of 
moss (r = -0.23). Some common moss species such as 
Isothecium spp. require mild and wet conditions. These 
conditions are usually found at lower elevations in the Sitka 
Spruce and Western Hemlock Zones. Lichens such as Alectoria 
spp. and Bryoria spp. are most abundant at higher elevations 
where conditions are colder and dryer (Henderson and others 
1989). These stands usually have a high percent composition 
of silver fir and mountain hemlock, which are not known to 
be used as nest trees by Marbled Murrelets in the Pacific 
Northwest. Therefore, it was not surprising that lichen cover 
was negatively related to the probability of occupancy. 

Stem Density (trees >81 cm d.b.h.) — This variable was 
not correlated to any other variable to any great degree except 
basal area (0.63), but it can be assumed that in general, stands 
with a higher stem density of trees >81 cm d.b.h. would have 
a larger number of potential nest platforms/ha and higher 
canopy closures. A larger sample of stands in Washington 
with lower stem densities is needed to fully understand this 
variable and its effect on the probability of occupancy. Occupied 
stands with stem densities of only 5 trees/ha have been 
documented in Oregon (Nelson, pers. comm.) 



Canopy Closure — It may be difficult for murrelets to 
locate and access nest platforms in stands with extremely 
high canopy closures, and the results of the analysis may 
reflect this because occupied old-growth stands still had 
mean canopy closures of 86 percent. A larger sample of 
stands with lower canopy closures is needed to fully 
understand this variable and its effect on the probability of 
occupancy. Nests located in stands with very low canopy 
closures may be subject to higher predation rates since corvids 
are the most common nest predator and locate prey almost 
entirely by sight. Stands with low canopy closures and low 
tree densities would be expected to have longer sight distances 
through the canopy. In these cases, murrelet nests would be 
easier to locate by visual predators. 

Mean Diameter of Western Hemlock — Because the 
majority of trees infected with mistletoe were western hemlock 
and the mean d.b.h. of low-elevation trees was useful in 
assessing suitable habitat, the mean d.b.h. of western hemlock 
appears to combine the variation of these two factors into 
one variable. 

Mistletoe Number — Stands that are infested with mistletoe 
may provide a higher number of nest platforms for murrelets. 
Mistletoe infects the branches of living trees, causing swelling, 
deformation, and brooming, which acts to thicken smaller 
diameter branches. This process can create suitable nest 
platforms from otherwise marginally-sized limbs. Thick 
secondary branching is characteristic of these mistletoe brooms 
that create dense overhead cover, a characteristic found at 
many murrelet nest platforms in Washington and Oregon 
(Hamer and Nelson, this volume b). In addition, mistletoe 



174 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hamer 



Chapter 17 



Inland Habitat Associations in Western Washington 



blooms help trap debris falling from the upper canopy, creating 
additional nesting platforms and platforms of larger size. 

Describing Low and High Quality Habitat 

In order to use the model to predict the probability of 
occupancy of an old-growth stand by murrelets, and thus 
judge the suitability of a stand as nesting habitat, it is necessary 
to obtain values for the 8 variables used by the model from 
the stand needing evaluation. The values for these variables 
can then be compared to the mean, minimum, and maximum 
values calculated for stands with a high probability of 
occupancy and stands with a low probability of occupancy 
(table 4). Using this comparison, a general sense of the 
suitability of a stand as nesting habitat can be obtained. In 
addition, by entering the values for the 8 forest characteristics 
into the formula shown below, the probability of occupancy 
can be calculated. Elevation should be entered in feet, stem 
density as the number of trees/25 m plot, mean d.b.h. of 
western hemlock in cm; and lichen, moss and canopy cover 
as percent total cover. First, the logistic regression model is 
used to predict g(x) as follows: 



g(x) = b + b x x x + b^c 2 + .... + bjc % 
where. 



(1) 



b Q is the intercept and, b v ...., ft g are the logistic regression 
coefficients for each variable. These values are listed under 
the Regression Coefficients in table 3. 

x. , jc 8 are the values for the independent variables 

measured at the stand in question and, 

g(x) is the predicted value of the logistic transformed 
probability of occupancy. 

Then g(x) is retransformed to estimate the probability of 
occupancy as follows: 

P = EXP (g(x))/[ 1 + EXP (g(x))] where, (2) 

P is the predicted probability of occupancy, 

g(x) is as defined in equation (1). 

EXP is the exponentiation function, i.e. 

EXP 3 = e 3 where e = 2.7183..., the base of natural 

logarithms. 

It is important to recognize that this model was developed 
from a sample of old-growth stands and its reliability in 
other stands has not been evaluated. 

Tree Characteristics 

Because western red cedar ranked third in producing 
potential nest platforms and was indicated by the regression 
analysis to be helpful in assessing suitable habitat, nest- 
search parties should pay closer attention to this conifer. 



Western hemlock was rated lower as a suitable nest tree 
because of a lower platform abundance. Because observers 
did not count mistletoe brooms on the outer limbs of trees as 
potential nest platforms, the actual number of potential nest 
platforms/tree for western hemlock may be much higher. 

Because murrelet surveys are often conducted in stands 
containing a mix of conifer species, it is difficult to use 
detection trend information from different stand types to confirm 
a preference for nesting in one type of conifer. In addition, not 
enough murrelet nests have been located, or located in a 
random manner, to determine whether birds are selecting 
particular tree species for nesting, especially since greater 
nest-search and survey effort have occurred in the Douglas-fir 
and Western Hemlock zones than in the Sitka Spruce zone. 
This comparison provides evidence that certain tree species 
are more likely to be used by murrelets than others. 

Acknowledgments 

These studies were funded by the Washington Department 
of Wildlife Nongame Program and Pacific Northwest Region 
Office, USDA Forest Service. I am grateful to Eric Cummins 
and Bill Ritchie of Washington Department of Fish and 
Wildlife for their considerable contributions of time and 
energy in developing and carrying out this research. Additional 
funding and field personnel were obtained from the 
Washington Department of Natural Resources Forest Land 
Management Division. The Endangered Species and 
Migratory Bird Programs of the U.S. Fish and Wildlife 
Service, U.S. Department of Interior, in Portland, Oregon, 
also contributed valuable funding. Dick Holthausen and Grant 
Gunderson (USDA Forest Service) were both instrumental 
in coordinating the participation of the Forest Service in the 
project. I thank Lenny Young and Chuck Turley of the 
Washington Department of Natural Resources for their support 
of the research and Tara Zimmerman of the U.S. Fish and 
Wildlife Service for her efforts in providing additional funding. 
I thank Phyllis Reed and Charlie Vandemoer of the USDA. 
Forest Service, Mt. Baker-Snoqualmie National Forest for 
their cooperation and help with logistical needs. 

I also acknowledge the large number of field personnel 
from the Washington Department of Fish and Wildlife, Mt. 
Baker-Snoqualmie, Olympic, and Gifford Pinchot National 
Forests, Washington Department of Natural Resources, private 
biological consulting companies, and the timber industry for 
their major contributions to data collection and willingness 
to share information. I thank statisticians Tim Max and Don 
Bachman of the USDA Forest Service Forestry Sciences 
Laboratory biometrics group for their help in study design, 
analysis, and interpretation of a complex habitat model. 

Helpful reviews of this manuscript were provided by 
Alan Burger, Martin Raphael, C. John Ralph, Peter Conners, 
Dean Stauffer, Bill Block, and Jim Baldwin. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



175 



Chapter 18 

A Landscape-Level Analysis of Marbled Murrelet Habitat in 
Western Washington 



Martin G. Raphael 1 



John A. Young 2 



Beth M. Galleher 3 



Abstract: Relationships between landscape-level patterns of for- 
est cover and occupancy by Marbled Murrelets in the state of 
Washington where state-wide forest-cover information was avail- 
able were investigated. Using a geographic information system, a 
203-hectare circular area surrounding each of 261 previously 
surveyed locations was delineated. Within each area, we calcu- 
lated the amount, distribution, and pattern of various classes of 
late-seral forest. Proportions of old-growth forest and large saw- 
timber were greater at sites that were occupied by murrelets than 
at sites where they were not detected. Mean size of patches 
(contiguous cover) of old growth and large sawtimber were also 
greater among occupied sites than among detected and undetec- 
ted sites. On average, old growth and large sawtimber combined 
comprised about 36 percent of occupied sites (203-ha areas) vs. 
30 percent and 18 percent on detected and undetected sites, 
respectively. Various indices of landscape pattern were less use- 
ful in distinguishing these sites, but in general, occupied sites had 
more complex patterns with more edge, a greater variety of cover 
types, and more complex shapes (greater lengths of edge relative 
to area of patches). Broader patterns, evaluated within large river 
basins, are also described, but lack of consistent survey effort 
among these basins precluded analyzing rates of occupancy in 
relation to forest cover at that scale. 



Studies of murrelet nesting behavior in the Pacific 
Northwest have shown that breeding birds select stands 
of old-growth forest or stands that provide platforms for 
nests and suitable protection from predators in California 
(Paton and Ralph 1988), Oregon (Grenier and Nelson, this 
volume), and Washington (Hamer and Cummins 1990, 1991). 
All murrelet nests found in these states have been located in 
old-growth conifer forests (Hamer and Nelson, this volume 
b). Whereas nesting habitat requirements of murrelets at the 
individual tree or nest platform and the stand level have been 
examined in some detail, characteristics of murrelet nesting 
habitat at the landscape level are less understood (Hamer 
and Cummins 1990). 

Recently-completed studies by Hamer and others (1993) 
have provided much needed information on suitable nesting 
habitat characteristics within forest stands in Washington 
that can be used as predictors of murrelet occupancy from 
ground-based surveys or forest inventories. No studies have 



1 Chief Research Wildlife Biologist, Pacific Northwest Research 
Station, USDA Forest Service, 3625 93rd Ave., Olympia, WA 98512-9193 

2 Geographer, Pacific Northwest Research Station, USDA Forest 
Service, 3625 93rd Ave., Olympia, WA 98512-9193 

3 Program Analyst, Pacific Northwest Research Station, USDA Forest 
Service, 3625 93rd Ave., Olympia, WA 98512-9193 



as yet considered whether landscape-level characteristics of 
nesting habitat such as shape, size, or configuration among 
forest stands have predictive capabilities for occupancy by 
murrelets. To determine if broad-scale patterns of habitat 
distribution influence murrelet occupancy, we initiated a study 
of relationships between amount and configuration of habitat 
and occupancy of murrelets at previously surveyed locations. 
Information on relationships between habitat characteristics 
and occupancy by murrelets at broader scales could be of 
value in planning conservation strategies and guidelines for 
management at the regional level. Assessments of habitat 
requirements across all scales — nest, stand, site, and landscape 
— are necessary to determine the proper mix of management 
guidelines to assure adequate amounts and configuration of 
nesting habitat for the murrelet in the Pacific Northwest. 

Methods 

Analysis of landscape attributes of Marbled Murrelet 
habitat selection proceeded at two scales. A broad scale 
analysis within major river basins considered the distribution 
of potential habitat among land owners (Federal and non- 
Federal) over the species' range in Washington. A more site- 
specific analysis considered the influence of landscape 
characteristics immediately adjacent to survey sites on 
occupancy status of murrelets. We generated statistical 
measures for both scales of analysis using Geographic 
Information Systems (GIS) and landscape pattern programs. 

Data Sources 

We obtained a database of all murrelet survey locations 
(through 1992) from the Washington Department of Fish 
and Wildlife (WDFW). This database was used previously 
in regional conservation planning efforts for the Northern 
Spotted Owl, Marbled Murrelet, and other species associated 
with late-successional forests (Thomas and Raphael 1993). 
Murrelet survey locations (n = 708) are represented by x,y 
coordinate locations and associated attributes mapped in 
GIS form (fig. 1). Survey points were coded by the WDFW 
into five levels of murrelet detection (table 1) following 
protocols and definitions of the Pacific Seabird Group (Ralph 
and others 1993). Many of the locations were collected 
before the currently accepted survey protocol was developed. 
In addition, some of the database records represent multiple 
sites clustered around a single survey station. For purposes 
of this analysis, we analyzed only those surveys conducted 
following protocol standards, and we eliminated any additional 
multiple sites around a single station. The number of resulting 
sites were n = 261. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



177 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Murrelet Survey 
Locations 

• Protocol Survey 

a Non-Protocol Survey 

N WRIA Basin 

/f 50 Mile Buffer 




54000 



Data Sources: 

Basins - Washington Oept. of Natural Resources 

Murrelet - Washington Dept of Wildlife, 1993 

06 May 94 



Figure 1 — Marbled Murrelet survey locations in western Washington. Murrelet surveys are identified as those conducted 
following accepted protocols (Ralph and others 1 993) or otherwise. The heavy dashed line indicates a 50-mile zone from marine 
water, an area considered by Washington Department of Fish and Wildlife as the range of the Marbled Murrelet for management 
purposes. Map is divided into Washington Department of Natural Resources' designated Water Resource Inventory Areas 
(WRIA) identifying corresponding river basins. Numbers within WRIAs indicate identifications assigned to each WRIA by the 
Washington Department of Natural Resources. 



178 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 1 — Status of inland sites where Marbled Murrelets have been surveyed 
in Washington 1 , through 1992 





code 


Status 2 


Number of surveys 


Survey 


All 
surveys 


Surveys following 
protocol standards 2 


1 




Active nest 


5 





2 




Nest site 


19 


3 


3 




Occupied site 


141 


66 


4 




Presence 


308 


108 


5 




No detection 


235 


84 


Total 






708 


261 



'Source: Washington Department of Fish and Wildlife. 

2 Protocol developed by Pacific Seabird Group (Ralph and others 1993). 
Multiple records from the same station are also excluded. See this document for 
definition of status categories. 



We obtained two maps of forest vegetation from 
Washington State natural resource databases for this analysis. 
These maps represent the only sources of forest cover 
classified across both Federal and non-Federal lands in 
Washington. A digital map of old-growth and other cover 
classes was obtained from the WDFW (Eby and Snyder 
1990, Collins 1993). This map was updated by Washington 
Department of Natural Resources (WDNR), Forest Practices 
Division, using 1991 Landsat Thematic Mapper (TM) imagery 
to account for timber cutting since 1988 (Collins 1993). The 
map displays old-growth and other forest conditions in western 



Washington from the Pacific coast to 50 miles inland on 
lands below 3200' elevation (table 2). The 50-mile limit was 
defined by WDFW as the inland extent of murrelet activity, 
even though their database contains three records at greater 
distances (to 53 miles). 

The WDFW forest-cover map was used for both a basin- 
level analysis and a site-level analysis. We received the data 
as 1 : 100,000 vector (polygon) maps. We converted the vector 
maps into a raster (grid) format using the ARC/INFO GRID 
software (Environmental Systems Research Institute, Inc., 
Redlands, CA) at a cell resolution of 50 by 50 meters. We 
projected the maps from a State Plane coordinate system 
into a Universal Transverse Mercator (UTM) map projection 
and joined the individual 1:100,000 scale maps together to 
form one seamless map that we could use with our existing 
GIS databases. 

We used a second source of vegetation data for basin- 
level analysis to compare against the WDFW data. The WDNR, 
Forest Practices Division provided a map of forest serai 
stages that was developed for the state from 1988 Landsat 
TM imagery (Green and others 1993). This map is in a raster 
(grid) format with a cell resolution of 147 by 147 meters and 
was classified into six classes (table 3). To match the WDFW 
map, we created a murrelet zone map by drawing a boundary 
50 miles inland from the Washington Pacific and Puget Sound 
coasts. This map was used as the geographic extent for all 
subsequent analyses; maps of vegetation, river basins, and 
elevation were subset to coincide with this zone map. 

We used other GIS data sources in conjunction with the 
above sources of forest vegetation data to analyze murrelet 



Table 2 — Washington Department of Fish and Wildlife old-growth classification 1 



Class name 


Description 


Old growth 


Coniferous forest stands, dominant trees > 30" d.b.h ?, co-dominant trees > 16" d.b.h., 8 




or more dominant trees per acre, multi-layered canopy, several snags per acre > 20" 




d.b.h.. many down logs > 24" diameter 


Large sawtimber 


Coniferous forest stands, dominant trees 20-30" d.b.h., co-dominant trees > 14" d-b.h.. 10 




or more dominant trees per acre, 2-3 layer canopy, few snags or downed logs 


Small sawtimber 


Coniferous forest from sapling/pole stands to large sawtimber, < 20" d.b.h.. closed single 




layer canopy, very Utile dead wood 


Other 


Non-forested, or non-vegetated Also includes closed mature deciduous stands 


Above 3,200 feet 


All areas above 3200 feet were masked out of the updated version of Eby and Snyder's 




(1990) map 


Cleared 


Clear-cut since 1988 


Partial harvest 


Partial harvest since 1988 


Salt water 


Ocean. Puget Sound, other marine waters 


FreshwateT 


Inland lakes, rivers 



1 Source: Washington Department of Fish and Wildlife, Eby and Snyder (1990), Collins (1993). 

2 D.b-h. = diameter at breast height 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



179 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 3 — Washington Department of Natural Resources serai stage classification (Green and others 1993) 



Class 



Description 



Late serai 

Mid-seral 

Early serai 

Cleared/Other 

Water 

Non-forested 



Coniferous forest stands, > 10 pet tree crown closure in trees > 21" d.b.h., with > 70 
pet total crown closure, and < 75 percent of the crown in hardwoods or shrubs 

Coniferous forest stands, < 10 pet tree crown closure in trees > 21" d.b.h., with > 70 
pet total crown closure and < 75 percent of the crown in hardwoods or shrubs 

Coniferous forest stands, 10-70 pet total crown closure and < 75 pet of the crown in 
hardwoods or shrubs 

< 10 pet crown closure conifers and/or > 75 pet of the crown in hardwoods or shrubs 

Open water bodies 

Non-forest land (agriculture, urban, rock, etc.) 



occurrence against measures of landscape pattern and 
composition. A map of major river basins depicting WDNR' s 
Water Resource Inventory Areas (WRIA's) was obtained 
from WDNR and used to subdivide the vegetation maps into 
analysis units based on river drainages (fig. 1) for the basin- 
level analysis (Green and others 1993). 

Accuracy of Forest-Cover Maps 

Forest-cover maps used for the Marbled Murrelet 
landscape analysis were developed by WDFW and WDNR. 
The WDFW data set was developed from 1984 and 1986 
Landsat Multi-Spectral Scanner (MSS) imagery. This imagery 
has a minimum spatial resolution of approximately 80 m 2 
and collects information in four spectral bands. Digital 
elevation models were used by WDFW to compensate for 
shadowing on north-facing slopes (Eby and Snyder 1990). 
The stated accuracy of this data source for mapping old- 
growth cover is 80 percent for the Cascades (20 percent 
error of commission and 7 percent error of omission) and 85 
percent for the Olympic Peninsula (15 percent error of 
commission and 7 percent error of omission) (Eby and Snyder 
1990). Errors of commission are areas that are mapped as 
old-growth forest, for example, but are found to be some 
other type upon field inspection. Errors of omission are 
areas of old-growth forest that are missed in the mapping but 
are found to exist on the ground. Accuracy was assessed by 
WDFW by checking mapped interpretations against field 
observations (Eby and Snyder 1990). Other potential errors 
in this data set are large sawtimber stands mapped as old- 
growth forest, wind-throw or fire regenerated stands mapped 
as old growth, and the omission of small, narrow features 
and stand edges (Eby and Snyder 1990). In addition, areas of 
mature deciduous forest and sapling conifer are lumped into 
the "other forest" class, which causes difficultly in determining 
actual stand boundaries in areas with little older forest, such as 
in southwest Washington (Snyder, pers. comm.). In addition, 



because the focus of the mapping effort was to determine 
areas of old-growth forest, errors associated with other types 
of land cover were not distinguished. 

The WDNR data set was developed from 1991 TM 
imagery. This imagery has a minimum spatial resolution of 
30 m 2 and collects information in seven spectral bands. 
High altitude aerial photography, field reconnaissance, and 
WDNR maps were used to guide the classification (Green 
and others 1993). The stated overall accuracy of this data 
set within the range of the Marbled Murrelet is 92 percent, 
with the lowest accuracy in the Puget lowland (87 percent) 
and the highest in the North Cascades (97 percent) (Green 
and others 1993). No information is given on errors of 
commission or omission. Potential confusion in this dataset 
may be caused by the grouping of stands with >75 percent 
crown closure in hardwoods and young conifer in the "other 
forest" category. 

GIS Processing 

We subdivided both habitat maps (WDFW and WDNR) 
into WRIA river basins by using ARC/INFO GRID commands 
for the basin level analysis. Attributes from each basin were 
then input to the DISPLAY landscape pattern program. 
DISPLAY is a package of statistical routines that calculates 
indices of landscape pattern from GIS maps (Flather and 
MacNeal 1993). Landscape pattern indices calculated by 
DISPLAY (table 4) are based on pattern indices discussed in 
O'Neill and others (1988), Milne (1991, 1992), and Krummel 
and others (1987). 

For the site-level analysis, we subsetted the WDFW 
forest condition map into 0.5-mile radius circles around 
survey locations (fig. 2). We calculated indices of pattern on 
each resulting circular landscape using the FRAGSTATS 
program (Marks and McGarigal 1993). FRAGSTATS is a 
set of routines that calculates indices of pattern on landscapes. 
FRAGSTATS calculates many of the same indices as 



180 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 4 — Landscape pattern indices output by DISPLA Y and used in the basin level analysis 



Pattern index 


Possible values' 


Description 


Landscape diversity 


0-°° 




Measures proportion of landscape in different types; = lowest 
diversity (only 1 type); larger value indicates more diverse 
landscape 


Landscape dominance 


0-~ 




Extent to which 1 or few types dominate the landscape; as 
value approaches 0, all types are present in equal proportions; 








max. value depends on number of types in landscape 


Landscape contagion 


0-ea 




Extent to which landscape is aggregated or clumped; as value 
approaches 0, many small patches exist; max. value depends 
on number of types in landscape 


Number of different types 


WDFW 2 = 


= 9 


Number of types possible in landscape, also termed 




WDNR 3 = 


6 


"patch richness" 


Proportion of each type 


0-1 




Percent of total area 


in landscape 








Number of patches of each 


0-<x> 




Count of patches by type 


type in landscape 








Mean patch size by type 


0- total 




Sum of patch area by type divided by total area 




landscape 


area 




Perimeter/area fractal 


1.0-2.0 




Index of patch edge complexity, contrasts log (patch 


dimension 






perim.) with log (patch area) 


Grid based fractal 


1.0-2.0 




Index of patch edge complexity, calculated using a 


dimension 






grid-cell counting method 



1 Values reported are theoretical limits, not actual ranges. 

2 WDFW = Washington Department of Fish and Wildlife 

3 WDNR = Washington Department of Natural Resources 



DISPLAY and also calculates additional landscape-level 
and patch-level indices (table 5). We attempted to use 
FRAGSTATS for the basin level analysis, but these landscapes 
were too large to process using this program. We tabulated 
indices of pattern for each of the 261 circular areas and 
compared site-level attributes among survey-status attributes 
(those sites where murrelets were not detected, were detected, 
or classified as occupied). 

We computed additional site-level variables using the 
GIS to add other environmentally related measures to the 
multivariate comparison of site-level pattern and occupancy 
status. Distance to closest coastline (meters) was calculated 
for each murrelet survey location using the NEAR function 
in ARC/INFO. This represents a straight-line distance between 
a survey location and the closest body of salt water. 

We identified patch size and type for each survey location 
by recording the contiguous patch on the overall landscape 



(whether or not that patch was outside of the 0.5-mi radius 
circle) directly underneath each survey point. The definition 
of patch used here differs significantly from the concept of a 
stand typically used by foresters. In this case, a patch is 
defined in terms of the GIS map as each unique set of 
contiguous cells of the same cover class type. Some of these 
patches can be quite large (up to 25,000 hectares) and should 
not be considered equivalent to typically defined stands in 
forest management. Rather, these are areas defined by pixels 
sharing the same class value. 

We determined survey-site elevations by overlaying the 
map of survey locations on a digital elevation model and 
interpolating the elevation at each point using GIS operations. 
United States Geological Survey 1:250,000 scale digital 
elevation models were used to derive an elevation surface 
for the state of Washington. These elevation models are a 
regular (grid) sample of elevations and have a vertical accuracy 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



181 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 






Sample Circular 

Landscape Around 

Survey Location 



I Old-Growth 

| Large Sawtimber 

| Small Sawtimber 

I Non-forest 
or unknown 



LANDSCAPE INDICES: 
Total area: 203 ha 
Number of patches: 29 
Mean patch size: 7.0 ha 
Shannon's Diversity Index: 
Contagion: 123 % 
Total edge: 20,100 meters 



1.20 



OLD-GROWTH INDICES: 
Total number of patches: 6 
Total area: 46 ha 
Mean patch size: 7.7 ha 
Mean Shape Index: 1.37 
Mean Nearest Neighbor: 134 meters 

LARGE SAWTIMBER INDICES: 
Total number of patches: 10 
Total area: 30 ha 
Mean patch size: 3.0 ha 
Mean Shape Index: 1.27 
Mean Nearest Neighbor: 161 meters 



Meters 



Data Source 



Eby and Snyder (1990) 
06 May 94 



Figure 2 — Example of forest cover classification within a 203-ha circular area surrounding a Marbled Murrelet 
survey location. Forest cover from classification by Washington Department of Fish and Wildlife (Eby and Snyder 
1990). See table 5 for explanation of landscape indices. 



182 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 5 — Landscape pattern statistics output by the FRAGSTA TS program and used in the site-level analysis, Washington Department of Fish and W .ildlife 
old-growth data set 





Range of values for 203-ha 




Pattern index 


circles (n = 261) 


Description 


Number of patches 


1-46 


Count of number of patches 


Mean patch size 


44.1-203 


Average size (ha ) of all patches in landscape 


Patch size std dev 


0-143 


Standard deviation of patch sizes in landscape (ha ) 


Patch size coeff var 


0-321.81 


Coefficient of variation of patch sizes in landscape 


Mean shape index 


1.06-1.98 


Average shape index (complexity) of all patches in landscape 


Area weighted mean shape index 


1.12- 3 36 


Average shape of patches standardized by patch area 


Landscape shape index 


0.98-4.85 


Overall complexity of landscape 


Mean patch fractal dimension 


1.0-1.1 


Fractal edge complexity for all patches in landscape 


Patch richness 


1-9 


Maximum number of different types in landscape 


Shannon's diversity index 


0-1.77 


An index of patchiness. dependent on proportion of landscapes of different types 


Simpson's diversity index 


- 0.83 


Another index of patchiness, 1 minus the squared sum of the proportion of the 
landscape in different types 


Modified Simpson's diversity index 


0-1.75 


The Simpson index modified by taking the negative log of the sum of landscape 
proportion of patch types 


Shannon's evenness index 


0-1 


Index relating the proportion of landscape in each type to the number of 
different types 


Simpson's evenness index 


0-1 


Index relating 1 minus the proportion of landscape in each type to 1 minus the 
inverse of the number of different types 


Modified Simpson's evenness index 


0-1 


Modified evenness index . relates the negative log of squared proportion of 
landscape in different types to log of the number of types 


Mean nearest neighbor 


0-1304 


Average distance (m) to closest patch of similar type 


Nearest neighbor std dev 


0-721 


Std. deviation of nearest neighbor distance by type 


Nearest neighbor coeff var 


0-124 


Coefficient of variation (m) for nearest neighbor distances 


Contagion 


0-154 


Extent to which landscape is aggregated or clumped; as value approaches 0, 
many small patches exist: maximum value depends on number of types in 
landscape 


Contagion(2)' 


0-80 


Extent to which landscape is aggregated or clumped; as value approaches 0, 
many small patches exist; maximum value depends on number of types in 
landscape (excludes landscape border) 


Total edge 


5600 - 27650 


Total length of edge (m) between patches of different types 



1 Varies from contagion in that landscape border is excluded. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



183 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



of (•*/-) 30 m (U.S. Geological Survey 1990). The original 
cell resol ution of 90 m was resampled to 200 m to create a 
statewide elevation grid. Elevations were recorded as height 
above se a level in meters. 

Results 

Basin-I ,evel Analysis 

Laindscape characteristics for the WDFW data by major 
river basin (table 6) show the majority of the basins' area is 
in the "other forest/unknown" category (3c = 84 percent). 
Two basins have 100 percent of their area in this class. 
Mean proportion of old growth was only 5 percent over all 
25 bflisins, although one basin (number 21) had 21 percent of 
its area in old growth (table 6). Pattern statistics calculated 
on these basins show a low diversity of types ( x = 0.57), a 
high dominance (x = 0.94) or influence of one or a few 



types, and a high contagion ( x = 4.55) or "dumpiness" in 
the data (table 6). 

In contrast, the WDNR seral-stage data (table 7) show 
a more even distribution of classes. Late-seral classes 
averaged 15 percent of the basin's area, and had higher 
mean patch sizes ( Jc = 69 ha) than the mean patch sizes of 
old growth from the WDFW data set (10 ha) (table 6). 
Pattern indices show the WDNR serai stage data by basin 
as relatively less "clumpy" ( x contagion = 5.58), and with 
a greater diversity ( x = 1 .40) than the WDFW classification. 
Basins classified using the WDNR serai stages also have a 
greater proportion of area in mid-seral ( x = 27 percent) 
and cleared ( x = 32 percent) classes. 

The range outlined in figure 1 encompasses about 5.2 
million ha, over half of which is privately managed (2.9 
million ha, 56 percent). Another 0.6 million ha (12 percent) 
are managed by the Washington Department of Natural 



Table 6 — Water Resource Inventory Area (WRIA) basin characteristics: Landscape pattern indices and proportion of basin area in cover classes, 
Washington Department of Fish and Wildlife cover classification (Eby and Snyder 1990, Collins 1993Y 



WRIA 








Old 


Mean patch size (ha) 


Large 


Small 


Cleared/ 


Other forest/ 


Basin 


Diversity 


Dominance 


Contagion 


growth 


old growth 


sawtimber 


sawtimber 


thinned 


unknown 


1 


0.31 


1.30 


4.61 


0.01 


6.5 


0.01 


0.03 


0.01 


0.94 


2 








0.01 

















1.00 


3 


0.27 


1.34 


4.63 


0.01 


4.2 


0.01 


0.03 


0.01 


0.95 


4 


0.74 


0.86 


4.62 


0.08 


12.7 


0.04 


0.06 


0.01 


0.80 


5 


0.85 


0.76 


4.59 


0.08 


10.9 


0.05 


0.09 


0.03 


0.76 


6 








0.01 

















1.00 


7 


0.84 


0.77 


4.98 


0.05 


7.9 


0.09 


0.07 


0.02 


0.77 


8 


0.36 


1.25 


4.70 


0.02 


6.4 


0.02 


0.03 


0.01 


0.92 


9 


0.56 


1.05 


4.65 


0.03 


6.0 


0.03 


0.05 


0.02 


0.87 


10 


0.43 


1.18 


4.74 


0.02 


8.6 


0.01 


0.03 


0.02 


0.91 


11 


0.61 


1.18 


6.65 


0.04 


9.9 


0.03 


0.06 


0.02 


0.85 


12 


0.56 


1.05 


4.19 


0.05 


7.3 


0.02 


0.03 


0.02 


0.87 


13 


0.40 


1.39 


5.96 


0.01 


2.8 


0.01 


0.04 


0.03 


0.91 


14 


0.46 


1.15 


4.58 





1.8 


0.02 


0.07 


0.02 


0.89 


15 


0.66 


0.95 


4.46 


0.03 


5.5 


0.05 


0.08 


0.01 


0.83 


16 


1.13 


0.48 


4.95 


0.17 


22.8 


0.12 


0.11 


0.01 


0.60 


17 


1.02 


0.77 


6.33 


0.09 


10.8 


0.07 


0.13 


0.02 


0.69 


18 


1.00 


0.79 


6.65 


0.16 


29.0 


0.07 


0.09 


0.01 


0.68 


19 


0.90 


0.71 


4.61 


0.06 


9.8 


0.04 


0.14 


0.03 


0.73 


20 


1.02 


0.59 


4.80 


0.16 


28.6 


0.05 


0.11 


0.02 


0.67 


21 


0.96 


0.65 


4.92 


0.21 


50.9 


0.04 


0.06 


0.02 


0.68 


22 


0.60 


1.01 


4.77 


0.04 


15.3 


0.02 


0.07 


0.02 


0.85 


23 


0.29 


1.32 


4.39 





2.5 


0.01 


0.04 


0.01 


0.94 


24 


0.16 


1.45 


4.22 





1.7 





0.02 





0.97 


25 


0.42 


1.19 


4.45 





2.2 


0.02 


0.06 


0.02 


0.90 


26 


0.31 


1.30 


4.46 


0.01 


4.5 


0.01 


0.02 


0.02 


0.94 


Mean 


0.57 


0.94 


4.55 


0.05 


10.4 


0.03 


0.06 


0.02 


0.84 


Std. dev. 


0.33 


0.39 


1.53 


0.06 


11.6 


0.03 


0.04 


0.01 


0.12 



1 Set figure 1 for location of each basin; table excludes fragments of WRIA basins 38, 39, and 45 along western boundary of range. 



184 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 7 — Water Resource Inventory Area (WRIA) basin characteristics: landscape pattern indices and proportion of basin area in cover classes, 
Washington Department of Natural Resources serai stage data (Green and others 1993) 1 



WRIA 








Late 


Mean patch 


Mid- 


Early 






Other/non- 


Basin 


Diversity 


Dominance 


Contagion 


serai 


size (ha) LS 2 


seral 


serai 


Cleared 


Water 


forest 


1 


1.55 


0.24 


6.06 


0.16 


88.6 


0.13 


0.09 


031 


0.01 


0.30 


2 


1.44 


0.35 


4.67 


0.02 


10.2 


0.37 


0.06 


0.27 


0.04 


0.23 


3 


1.44 


0.35 


6.20 


0.03 


41.8 


0.15 


0.10 


0.36 


0.03 


0.34 


4 


1.38 


0.41 


6.00 


0.43 


149.0 


0.15 


0.01 


0.26 


0.02 


0.13 


5 


1.48 


0.31 


5.81 


0.26 


117.7 


0.23 


0.05 


0.33 





0.11 


6 


1.29 


0.50 


5.27 





4.6 


0.15 


0.07 


0.43 


0.02 


033 


7 


1.54 


0.26 


533 


0.24 


703 


0.25 


0.05 


031 


0.01 


0.14 


8 


1.46 


0.33 


6.26 


0.05 


43.4 


0.19 


0.03 


0.21 


0.08 


0.44 


9 


1.44 


0.35 


5.73 


0.11 


34.4 


0.24 


0.03 


0.28 


0.01 


034 


10 


1.48 


031 


6.04 


0.20 


97.1 


0.27 


0.02 


0.29 


0.01 


0.20 


11 


1.43 


0.36 


6.07 


0.07 


162.6 


0.36 


0.06 


0.35 


0.01 


0.15 


12 


1.05 


0.56 


4.17 





0.0 


0.14 


0.01 


0.23 


0.02 


0.60 


13 


1.33 


0.28 


3.77 





0.0 


0.29 


0.08 


0.36 


0.01 


0.26 


14 


1.41 


0.38 


5.36 


0.05 


108.5 


0.40 


0.09 


0.35 


0.03 


0.09 


15 


1.33 


0.28 


3.35 





0.0 


0.43 


0.09 


0.26 


0.02 


0.19 


16 


1.42 


0.37 


5.57 


0.47 


126.4 


0.15 


0.04 


0.21 


0.01 


0.12 


17 


1.50 


0.29 


5.20 


0.20 


723 


0.27 


0.06 


0.36 


0.01 


0.10 


18 


138 


0.41 


5.49 


0.46 


202.0 


0.10 


0.04 


0.17 





0.23 


19 


1.22 


0.57 


5.91 


0.14 


38.5 


0.46 


0.01 


035 


0.02 


0.02 


20 


1.49 


0.30 


5.99 


0.29 


84.9 


0.28 


0.08 


0.28 


0.02 


0.06 


21 


1.48 


0.31 


5.70 


0.38 


150.8 


0.22 


0.23 


0.12 


0.01 


0.05 


22 


1.32 


0.47 


5.71 


0.17 


34.4 


033 


0.02 


0.40 


0.01 


0.07 


23 


1.28 


0.51 


5.92 





12.8 


0.34 


0.09 


0.39 





0.17 


24 


1.25 


0.54 


532 


0.03 


10.7 


0.44 


0.05 


0.37 


0.01 


0.10 


25 


1.34 


0.45 


6.07 





11.4 


0.34 


0.12 


0.41 


0.03 


0.10 


26 


1.38 


0.41 


6.33 


0.02 


51.0 


0.27 


0.10 


0.45 


0.02 


0.13 


Mean 


1.40 


037 


5.58 


0.15 


69.0 


0.27 


0.0 


032 


0.02 


0.18 


Std. dev. 


0.09 


0.09 


0.72 


0.16 


57.8 


0.10 


0.05 


0.08 


0.02 


0.11 



1 See figure I for locations of each basin: table excludes fragments of WRIA basins 38, 39, and 45 along western boundary of range. 

2 LS = late sera) 



Resources and 0.9 million ha ( 1 8 percent) by the National 
Park Service. Based on the WDNR classification, private 
and state lands are predominantly mid-seral and other forest, 
whereas National Forest and Park Service lands are 
predominantly late-seral (fig. 3). An analysis based on the 
WDFW classification (fig. 4) shows a similar distribution of 
forest age classes among land managers. However, the amount 
of late-seral forest (old growth and large sawtimber) is much 
lower than that estimated from the WDNR classification. 
This difference reflects the elevation cutoff (3200 feet) used 
by the WDFW (table 2). 

The WRIA basin is too large an area relative to the 
number of surveys conducted within each basin (fig. 7) to 
detect relationships among landscape pattern variables and 
detection rate. Analysis of smaller basins with greater sampling 
intensities may help to clarify what, if any relationship exists 
between broad landscape pattern and likelihood of murrelet 
detection. However, the description of amount and config- 



uration of forest vegetation within river basins given here 
may help to determine those areas in Washington that are in 
need of closer examination at finer scales of analysis and 
with greater surveying effort. 

Site-Level Analysis 

Stand Characteristics 

Most (59 percent) of the Marbled Murrelet survey sites 
were centered within the various other forest categories 
(WDFW forest-cover map). Most of the remaining sites 
were located within old-growth stands (table 8). The 
proportion of sites within the various forest-cover classes 
differed significantly among detection classes (chi-square = 
40.2, P = 0.000). Patch area did not differ significantly 
among occupied, detected or undetected sites, nor did it 
differ among forest-cover classes (table 9). Survey sites 
averaged 30.6 km from nearest saltwater; mean distance did 
not significantly vary among occupied, detected, and un- 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



185 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



2,000.000 



Private ^^ m ^m State 

■ 250,000 
1,500,000 ■ 

200,000 - 

1,000,000 iSO.000 • 

,-v 100.000 

CD 500.000 ^^^ I I ^^^ I 

o — ■ ■■ ^ Hi ■ 



500,000 



400,000 



300,000 



200,000 



100.000 




300.000 



250,000 



200,000 



150,000 



100,000 



50,000 



National Park 



Late-Seral Mid-Serai Early-Serai Other 



Late-Serai Mid-Serai Early-Serai Other 



Figure 3 — Classification by Washington Department of Natural Resources of the distribution of forest- 
cover classes among Federal, state, and private lands within the range of the Marbled Murrelet in 
Washington (Green and others 1993). "Other" includes all remaining cover classes from table 3. See 
figure 1 for map. 



3,000,000 



2,500,000 



2,000,000 



1.500,000 



600,000 



1,000.000 



500,000 



CO 



co 

s> 

<£ 700,000 
600,000 
500,000 
400,000 
300,000 
200,000 
100,000 




Old Growth Small Sawtimber 

Large Sawtimber 



Other 



Old Growth Small Sawtimber 

Large Sawtimber Other 



Figure 4 — Classification by Washington Department of Fish and Wildlife of the distribution of forest- 
cover classes among Federal, state, and private lands within the range of the Marbled Murrelet in 
Washington (Eby and Snyder 1990, Collins 1993). "Other" includes all remaining cover classes from 
table 2. See figure 1 for map. 



186 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others. 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table & — Frequency of Marbled Murrelet survey sites among forest-cover 
classes by detection class, western Washington 





Detection class 


Forest-cover class 1 


Occupied 


Detected 


Undetected 


Total 


Old growth 


27 


17 


8 


52 


Large sawtimber 


9 


17 


4 


30 


Small sawtimber 


10 


10 


5 


25 


All other classes 


23 


64 


67 


154 


Total 


69 


108 


84 


261 



'Cover classes from Washington Department of Fish and Wildlife (Eby and 
Snyder 1990. Collins 1993). updated by Washington Department of Natural 
Resources (Collins, pers. comm.). See table 2 for cover class descriptions. 



Table 9 — Analysis of variance of patch size in relation to survey status 
(occupied, detected, unoccupied) and cover class (old-growth, large 
sawtimber, small sawtimber) of Marbled Murrelet survey sites, western 
Washington 1 



Source of variation 



Status 
Cover class 
Interaction 



df 



Significance 



0.40 
2.35 
0.69 



0.671 
0.100 
0.603 



'Patch area was estimated for contiguous cover surrounding each survey 
site as classified from cover maps of Eby and Snyder ( 1 990), Collins ( 1 993). 



detected sites (31.2, 30.3, 30.6 km, respectively). Elevation 
of survey sites averaged 482 meters and mean elevation did 
not significantly differ among occupied, detected, and 
undetected sites (520, 467, and 473 meters, respectively). 
Maximum elevation for all surveys was 1,455 meters, 
minimum elevation was sea-level (0 meters). 

Site Characteristics 

We investigated two general characteristics in describing 
the 203-ha area surrounding each site — amount and pattern 
of forest-cover classes (WDFW forest-cover map). The 
relative amounts of each of four general forest cover classes 
varied significantly among each of the detection classes 
(table 10). Over the entire sample of 261 survey sites, old- 
growth forest averaged 18 percent of the 203-ha landscape 
surrounding each site. Percentages of large sawtimber, small 
sawtimber and other averaged 9 percent, 1 1 percent, and 6 1 
percent, respectively. Percentage of old-growth forest was 
significantly greater on occupied sites compared to undetected 
sites (table 10). Similarly, the proportion of large sawtimber 
was greater on occupied sites than on undetected sites. 
Proportion of other forest land was greater on undetected 
sites than occupied sites. 



Many of the landscape partem indices are correlated. 
Rather than report estimates for each of the 21 different 
indices we computed, we used principal components analysis 
to produce composite landscape shape index variables. This 
analysis resulted in four factors that contained about 88 
percent of the variation inherent in the original set of variables. 
The first factor contained about 61 percent of the variation 
in the original variables and was used in subsequent analyses. 
This composite factor was highly correlated (r > 0.80) with 
10 of the original variables. Values of this composite index 
increased with increasing number of patches, landscape 
shape index, Shannon's diversity index, Simpson's diversity 
index, modified Simpson's density index, Shannon's and 
Simpson's evenness indices, modified Simpson's evenness 
index, contagion index, and total edge. Mean values of this 
composite landscape pattern index (table 11) varied 
significantly among detection classes (F= 14.88, P = 0.000), 
and was significantly greater among occupied sites than in 
either of the other detection classes (planned contrast, t = 
5.17, P = 0.000). 

We also investigated the influence of shape and size of 
old-growth and large sawtimber patches (table 11). These 
attributes are correlated with the amount of each cover 



Table 1 — Forest cover (mean percentage) within 203-ha circles centered on Marbled Murrelet survey sites, western Washington 1 





Forest-cover class- 




Other forest 




Small sawtimber 


Large sawtimber 
x min max 


Old growth 


Status 


X 


min 


max 


X 


min 


max 


x min max 


Occupied 3 

Detected 

Undetected 


51.7A 
57.8AB 

72.4B 


2.5 
4.4 



100 
100 
100 


12.1A 

12.1A 

9.6A 








46.7 
51.5 
61.7 


11.4A 

10JAB 

6.4B 


55.4 
70.3 
91.8 


24.7A 76.3 
19.8AB 73.9 
11. 6B 54.2 



1 Letters indicate results of pairwise comparisons among means; experiment-wise P <0.05, using Tukey 's test Means with same 
letter (within columns) did not differ significantly. 

: Forest cover map from Washington Department of Fish and Wildlife (Eby and Snyder 1990, Collins 1993). 
3 Includes status codes 1 . 2, and 3 from table 1. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



187 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washingtoi 



n 



Table 11 — Attributes of forest cover within 203-ha circles centered on Marbled Murrelet survey sites, western 
Washington 







Detection class 


Univariate 
significance 1 


Correlation with 


Site attribute 


Occupied 


Detected 


Undetected 


discriminant function 


Old growth: 












Proportion 


0.247 


0.198 


0.116 





0.90 


Mean patch size 


18.600 


13.900 


8.500 





0.96 


Mean shape index 


1.500 


1.400 


1.100 





0.65 


Large sawtimber: 












Proportion 


0.115 


0.103 


0.064 





0.73 


Mean patch size (ha) 


4.100 


3.800 


4.100 





0.58 


Mean shape index 


1.300 


1.200 


1.000 





0.69 


Small sawtimber: 












Proportion 


0.121 


0.121 


0.096 





0.49 


Other forest proportion 2 


0.517 


0.578 


0.723 





...2 


Landscape pattern index 


0.413 


0.062 


-0.419 





0.76 


Sample size 


69 


108 


84 







'Significance of univariate analysis of variance, based on transformed variables where appropriate. 
2 Variable was not included in the discriminant analysis. 



type; as the amount increases, the values of the pattern 
indices increase. Therefore, using planned contrasts we 
found that mean patch size of old-growth (t = 4.67, P = 
0.000) and large sawtimber (t = 3.03, P = 0.003) was 
greater among occupied sites than among detected and 
undetected sites and that mean shape index was greater as 
well (t = 3.64, P = 0.000 for old growth; t = 4.24, P = 0.000 
for large sawtimber). 

To evaluate the relative contributions of the amounts of 
various forest cover classes and the pattern of those classes 
over the 203-ha landscapes, we used discriminant analysis to 
compare attributes among the three detection classes. For 
this analysis, we used all of the attributes listed in table 11 
with the exception of proportion other forest (because all of 
the proportions sum to 1.00 within any 203-ha area, the 
proportion of other forest is directly implied by the sum of 
the remaining proportions). This analysis resulted in a single 
significant discriminant function (chi-square = 48.8, df= 16, 
P = 0.000); each detection class differed significantly from 
each of the other classes. The variables that best discriminated 
among the classes were old-growth proportion, landscape 
pattern index, old-growth patch size, large sawtimber 
proportion, and large sawtimber shape index (table 11). 
Although the average differences among the detection classes 
were significant, there was considerable overlap among the 
sites; R 2 was only 17.5 percent and only about 44 percent of 
the sites could be correctly classified based on the discriminant 
function (table 12). 



Discussion 

Landscape-level analysis of amount and configuration 
of forest vegetation can be a valuable tool for assessing the 
nesting habitat requirements of murrelets. However, the scale 
of analysis influenced our ability to predict occupancy in a 
given landscape. We found the forest-cover attributes within 
a 203-ha circular area surrounding each survey location 
were useful predictors of occupancy by the Marbled Murrelet. 
Both the amount and the pattern of various forest-cover 
classes differ among occupied, detected, and undetected 
203-ha sites. Given the strong correlations among the forest 
pattern and amount attributes, the variables describing the 
amounts of the various cover classes are probably most 
useful in describing Marbled Murrelet habitat as it occurs in 
this sample from western Washington. Among the forest- 
cover classes, old-growth cover, and to a lesser extent, large 
sawtimber, seem best to predict murrelet occupancy. Sites 
occupied by murrelets, as evidenced by nests or circling 
behavior, have a higher proportion of these mature forest 
classes than do non-occupied sites. 

More definitive analyses must await completion of 
additional surveys. The present database is not the result of a 
survey designed to understand the statewide distribution of 
the species. Instead, it is heavily influenced by one intensive 
study (Hamer and Cummins 1990, 1991) and by sites selected 
at the location of proposed timber sales. Therefore, the set of 
survey sites we analyzed may be biased. Until more systematic 



188 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Raphael and others 



Chapter 18 



Landscape-level Analysis of Habitat in Washington 



Table 12 — Predicted and observed classification of detection-status of. Marbled 
Murrelet survey locations based on discriminant analysis using forest-cover 
attributes within 203-ha acres surrounding each site, western Washington 





Predicted status 


(pet) 1 




Actual status 


Occupied 2 


Detected 


Undetected 


Locations 


Occupied 2 


71 


16 


13 


69 


Detected 


49 


21 


30 


108 


Undetected 


29 


19 


52 


44 



1 Predicted from discriminant function (see table 11). 

2 Includes status codes 1, 2, and 3 from table 1. 



surveys are completed, it will be difficult to judge the reliability 
of estimates of habitat selectivity. 

Until such surveys are completed, we offer the following 
tentative guidelines. For purposes of identifying potential 
habitat, areas composed of at least 35 percent large sawtimber 
and old-growth forest (as classified by WDFW) are most 
likely to be occupied. Landscapes on the order of 200-300 ha 
should be examined to determine proportion of potential habitat 

In evaluating areas of about 200 ha, we conclude that the 
amount and configuration of old-growth or large sawtimber 
forest (Eby and Snyder 1990) are important components of 
murrelet nesting requirements, as has been previously 
demonstrated in analyses at the stand (Hamer 1993) and the 
nest level (Hamer and Cummins 1991) in Washington. 
Quantifying the amount and pattern of late-seral forests in 



larger landscapes can help to determine areas that may be at 
risk to loss of suitable nesting habitat for murrelets. Further 
landscape analysis at a basin level between the small landscapes 
and broad river basins we used here may help to determine 
the appropriate configurations and amounts of nesting habitat 
necessary to support murrelets, assuming adequate surveying 
has been conducted. This information would be a useful 
component of local or regional conservation planning for the 
murrelet and other old-growth associated species. 

Acknowledgments 

This study would not have been possible without the 
full cooperation of the Washington Department of Fish and 
Wildlife and the Washington Department of Natural 
Resources. These organizations (and their cooperators) 
conducted the Marbled Murrelet surveys and developed the 
rangewide habitat maps that were the basis of our analyses. 
Tom Hamer, who collected much of this data, helped us 
assemble the information. Additional help was provided by 
Eric Cummins, William Ritchie, James Eby, Michelle Snyder, 
Doretta Collins, Steve Bernath, Randy Kreuziger, and Tom 
Owens. We thank Kurt Flather for assistance with DISPLAY 
software. We appreciate the helpful comments from Tom 
Hamer, Jeffrey Granier, David Hays, Kevin McKelvey, and 
Gordon Orions. This study was funded in part by the 
Ecosystem Processes Research Program and by the Ecological 
Framework for Management Research, Development and 
Application Program of the Pacific Northwest Research 
Station. We thank Kevin Peeler for assistance with GIS 
analysis and Janet Jones for help preparing the manuscript. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



189 



Chapter 19 

Marbled Murrelet Habitat Associations in Oregon 



Jeffrey J. Grenier 1 - 2 



S. Kim Nelson 1 



Abstract: We described habitat associations of Marbled Murrelet 
(Brachyramphus marmoratus) nesting (n = 10) and occupied 
(n = 184) sites in Oregon. We compared habitat characteristics of 
177 occupied sites to a random sample of 9.625 sites (n = 531) of 
unknown murrelet status. In addition, we briefly described the char- 
acteristics of 22 nests and compared 10 of the nest sites to adjacent 
sites. In general, occupied sites were older, had larger midstory 
trees, and had larger and greater densities of dominant 
i or remnant) trees than random sites. In addition, dominant tree 
height and density, midstory and understory tree diameter and per- 
cent cover, and percent canopy closure were important habitat 
components for predicting murrelet occupancy. All nests were in 
old-growth Douglas-fir (Pseudotsuga menziesii). western hemlock 
(Tsuga heterophylla). and Sitka spruce (Picea sitchensis) trees > 127 
cm in diameter and > 36 m tall. Murrelet nest sites had fewer trees/ha 
and less canopy closure compa«d to adjacent sites. Our results 
support previous studies that concluded murrelets use stands with 
old-growth characteristics and that stand structure is more important 
than stand age. Knowledge of habitat associations does not imply 
habitat quality, which should be quantified through studies on repro- 
ductive success in relation to habitat and landscape characteristics. 



Marbled Murrelets (Brachyramphus marmoratus) use 
old-growth and mature forests, or forests with old growth 
components, nearly year-round (Naslund 1993b, Nelson 
1990b. Paton and Ralph 1990, Rodway and others 1991). 
Characteristics of these forests have been analyzed at the 
stand, landscape, nest, and nest-site levels (Burger, this volume 
a; Hamer. this volume; Hamer and Nelson, this volume b; 
Naslund and others, in press; Nelson and Hamer 1992; Rodway 
and others 1991. Singer and others 1991, in press). From 
these studies we know that murrelets nest in large diameter 
trees and may be selecting stands based on number of potential 
nest platforms, and density and diameter of dominant trees. 

Data on the distribution and habitat associations of 
Marbled Murrelets have been collected in Oregon since 
1988. This paper provides a synthesis of murrelet habitat 
associations by using existing data on occupied stands and 
nest sites, and summarizes new stand-level habitat data from 
state and federal agencies. Our objectives were to: (1) 
summarize habitat characteristics of occupied sites, (2) 
determine habitat associations by comparing occupied sites 
to other sites, and (3) identify the key habitat components of 
murrelet habitat. Knowledge of Marbled Murrelet habitat 
associations may assist in designing and implementing 
successful habitat management plans for this species. 



1 Research Wildlife Biologists. Oregon Cooperative Wildlife Research 
Unit Oregon State University, Nash 104. Corvallis. OR 97331-3803 

2 Present address. 1402 Cedar Street. Philomath. OR 97370 



Study area 

Study sites were located in the Coast Range and Klamath 
Mountain (Siskiyou Mountains) Provinces in Oregon (Franklin 
and Dyrness 1973). These areas consisted of rugged, 
mountainous terrain, with steep slopes and deeply cut river 
and creek drainages. Elevations ranged from SO m along the 
coast of Oregon, to more than 1 200 m in the central mountains. 
The climate consists of cool, wet winters and warm, dry 
summers. Mean temperatures range from 0° C in winter to 
24° C in summer, and annual precipitation varies from ISO 
to 300 cm (Franklin and Dyrness 1973). 

These areas are primarily forested, although they have 
been intensively managed for timber since the early 1900s, 
and many stands are <200 years old. In addition, natural and 
man-caused fires have altered many stands. Relatively small, 
isolated patches of mature and old-growth tree species remain. 
Douglas-fir (Pseudotsuga menziesii) was the dominant tree 
species in the north and mixed-evergreen species, including 
Douglas-fir and tanoak (Uthocarpus densiflorus), were 
dominant in the south. 

Methods 

Between 1990 and 1993, murrelets were surveyed on 
state and federal lands throughout the Coast Range and 
Siskiyou Mountains, primarily within 50 km of the coast. 
Surveys included intensive research surveys, and intensive 
and general surveys for agency monitoring projects. Forest 
stands were surveyed to existing protocols (Paton and others 
1990. Ralph and Nelson 1992, Ralph and others 1993) and 
were classified as occupied (birds exhibiting nesting or 
below canopy activity), with murrelets present (presence), 
or without murrelets (undetected), based on murrelet behavior 
patterns. In addition, we searched for nests using three 
methods: watching murrelets land in trees, searching for 
eggshells on the forest floor, and climbing trees to examine 
branches for nest cups. 

Characteristics of Occupied Sites 

Four databases were examined. The characteristics of 
occupied sites were summarized using one state lands database 
from Oregon Department of Forestry (Reagan, pers. comm.), 
two U.S. Forest Service databases from the Siuslaw National 
Forest (McCain, pers. comm.; Wettstein, pers. comm.). and 
one research database from Oregon State University. Habitat 
variables differed among databases. Similar habitat variables 
were used in the analyses where possible. We used Spies 
and Franklin's (1991) definition for stand age (i.e., young 
stands = 40-80 years, mature = 80-200 years, and old- 
growth = 200+ years). Remnant trees were defined as those 



LSDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



191 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



that survived recent fires and were >66 cm d.b.h., except in 
one Forest Service database where remnant trees were 
classified as >100 cm d.b.h.. 

State Lands Database 

Data were compiled from Oregon Department of 
Forestry's (ODF) OSCUR Inventory System (Ownership, 
Soils, forest Cover, land Use, and operation /fating). The 
OSCUR database was comprised of habitat variables 
collected between the mid-1970s and 1993 (appendix 1). 
ODF delineated and described habitat characteristics in 
forest sites through one of the following: (1) photographic 
interpretation; (2) stand examinations (fixed plot cruising, 
variable plot inventory, or timber sale appraisal); and (3) 
reconnaissance (walk-through) (ODF 1991). The OSCUR 
database included data from 6,409 sites. A site was defined 
as a uniform, homogeneous tree community that usually 
was a portion of a larger, contiguous, heterogeneous stand. 
Sites were characterized by approximately 160 habitat 
and geographic variables. In addition, comments from 
original data sheets were included to supplement data for 
some sites. 

We selected 34 key habitat variables (appendix 2) and 
forest sites >40 years old for analyses. Sites of this age were 
chosen because the youngest occupied site on state lands 
was classified as 42 years old (although the site included 
remnant old-growth trees). In addition, we were interested in 
examining differences of habitat characteristics between 
occupied and random sites within a sample universe of only 
suitable habitat i.e., sites containing large trees with adequate 
branch sizes and moss coverage to accommodate nesting 
(Grenier and Nelson 1994). Using maps and databases, we 
found that 72 occupied sites existed on ODF lands (Allen, 
pers. comm; Goggans, pers. comm.; Nelson 1990b; Nelson 
and Shaughnessy 1992; Piatt and Goggans 1992; Shaughnessy 
and Nelson 1991). Characteristics of these 72 occupied sites 
were compared to a random sample of 216 sites of unknown 
murrelet status. 

National Forest Land Databases 

Vegetation Resource and Structure Examination 
Databases (VSE) — Older-aged forests or those with multiple 
canopy layers were monitered on the Siuslaw National Forest 
in 1990. These study sites were located in areas proposed for 
timber harvest and in Northern Spotted Owl (Strix occiden- 
talis) Habitat Areas. In addition, Vegetation Resource 
Examinations (VRE) were conducted in 1991 and 1992 to 
ground truth satellite imagery of old-growth and mature 
forests with multi-layered canopies. Overlapping data from 
the two databases (1209 VRE and 1210 VSE plots) were 
used for analyses. Forty-seven habitat variables were common 
to both databases (appendix 2). Data from 120 sites (Wettstein, 
pers. comm.), 30 occupied sites and 90 other sites of unknown 
murrelet status, were used for analyses. 

Ecological Habitat Sampling — Ecologists at the Siuslaw 
National Forest collected habitat data in intensive and 



reconnaissance plots throughout the forest from 1981 to 
1984 (Hemstrom and Logan 1986; USDA 1983, 1985). The 
database provided to us included 974 forested sites and 162 
habitat variables (McCain, pers. comm.). We used ArcView 
(1992) to determine that 75 occupied sites overlapped with 
plots in this database. We selected 13 of the 162 habitat 
variables for our analyses (appendix 3). To be consistent 
with the ODF database, we used data from sites >40 years 
old. The characteristics of the 75 occupied sites were 
compared with a random sample of 225 sites of unknown 
murrelet status. 

Research Database 

In 1992, 40 small (<12 ha), isolated, mature and old- 
growth stands were selected (Nelson and Hardin 1993a). 
Using protocol surveys, we determined that 10 of these 
stands were occupied. Habitat characteristics were measured 
in two 25-m-radius circular plots randomly located within 
each of these stands. Variables included number of trees and 
snags by species, tree and snag diameter at breast height 
(d.b.h.), heights (m) of five dominant conifers (measured by 
triangulation with a Suunto optical clinometer), height (m) 
and decay class of snags (Cline and others 1980), forest zone 
(Franklin and Dyrness 1973), ecozone (average precipitation 
levels), plant associations (Hemstrom and Logan 1986), 
number of canopy layers, canopy cover (visual estimate of 
percent crown closure), ground cover (percent and species 
composition), abundance of moss and dwarf-mistletoe 
(Arceuthobium sp.), number of suitable nest platforms (> 18 
cm d.b.h., > 15 m height), slope (percent), aspect (degrees), 
position on slope (canyon bottom, lower 1/3, middle 1/3, 
upper 1/3, ridgetop), distance to water (m), and distance to 
opening (m; opening defined as road, river, clearcut, or 
vegetation type without trees but not forest gap). Calculations 
made from these data included density (number/ha) of trees 
(>46 and <80 cm d.b.h.) and dominant trees (>81 cm d.b.h.), 
mean diameter (d.b.h., cm) of all trees and dominant trees, 
mean dominant tree height (m), and tree species composition. 
Percent cover of epiphytes (moss and lichens) were recorded 
in four categories: (1) trace, (2) 1-33 percent, (3) 34-66 
percent, (4) 67-100 percent. Average mistletoe infestation 
was calculated for each plot using an index of to 6 developed 
by Hawksworth (1977). Distance inland (km), latitude, 
elevation (m), and stand size (ha) were determined from 
topographic maps (1:250,000) and aerial photos (1:1,000). 

Nest Site Characteristics 

Nests were located using ground-based and tree climbing 
techniques, most (15 of 22) in areas where likelihood of 
finding nests was considered to be high. Ground-based 
methods consisted of observing the flight of individual birds 
during dawn and dusk activity periods, and searching for 
eggshell fragments on the forest floor. Flight behaviors 
suggesting the presence of nesting birds (e.g., landing in or 
departing from trees and flying silently below the canopy) 
were identified at survey stations established in areas where 



192 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Grenier and Nelson 



Chapter 19 



Inland Habitat Association 5 in Oregon 



potential nest trees were located or where general activity 
levels were high, and at previously unsurveyed sites. Eggshell 
searches were conducted around all trees where birds were 
observed landing or taking off, and around numerous other 
trees that had potential nest platforms (platforms >18 cm in 
diameter and >15 m above ground). When potential nest 
trees were found, surveys were conducted on 1 to 3 successive 
mornings to confirm the presence of an active nest and 
identify its location. 

Nests located by tree climbing were found during an 
intensive tree climbing study at a single site or while 
reclimbing trees previously known to support nests (Nelson 
and others 1994a). The tree climbing study consisted of 
climbing all trees (Perry 1978) within a 40-m-radius plot and 
examining all platforms for nests. In addition, seven trees 
containing nests found between 1990 and 1992 were climbed 
in 1993 to determine if the nests had changed over time and 
to determine if nests or nest trees were reused. 

Characteristics of nests and nest trees were measured at 
the 22 nest sites. Nest tree measurements included diameter 
at breast height (d.b.h., cm), height (m), diameter at nest 
limb (cm), and nest branch height (m), diameter at the trunk 
and at the nest (cm), branch length (m), and position in 
crown (percent tree height). Nest measurements included 
distance from the trunk (cm) and cup dimensions (cm). In 
addition, the moss depth adjacent to the nest (cm), nest 
platform dimensions (cm), and canopy closure above the 
nest (percent) were measured. 

Micro-site habitat features of nest sites and 2 to 3 adjacent 
sites were measured at ten of the nests in 0.2-ha (25-m 
radius) plots (as described previously). Plots for nest sites 
were centered on the nest tree. Adjacent plots, centered 
around a dominant canopy-forming tree, were a minimum of 
75 m from nest trees, and were located the same distance 
from forest edges as nest trees to minimize any edge biases. 
Micro-site habitat characteristics were compared between 
nest sites and adjacent sites within the same stand to determine 
if the location of nests was associated with specific micro- 
site characteristics. 

Data Analyses 

Occupied Sites and Habitat Associations 

We used the two-sample Kruskal-Wallis test (Zar 
1984:138) to compare habitat characteristics of occupied 
sites to a random sample of other sites (with unknown murrelet 
status). The number of random sites selected equalled three 
times the number of occupied sites (Breslow and Day 1980:27; 
Ramsey, pers. comm.; Schafer. pers. comm.). The Chi-square 
goodness-of-fit test (Zar 1984) and Bonferroni Z-statistic 
(Byers and others 1984. Neu and others 1974) were used for 
categorical data. 

Logistic regression was used to determine key habitat 
components of occupied and random sites for each database 
(Manly and others 1993, Ramsey and others 1994). The 
following two steps were used in our analyses: ( 1 ) habitat 
variables (continuous and categorical) were divided into 



groups of related variables that described cae or two 
biological aspects of the site. Logistic regressi ion was used 
to test statistical significance of each variab le within the 
group. (2) habitat variables that were statistica" ily significant 
within the groups (P < 0.05) were then used ii a the stepwise 
procedure to determine the final model. V ariables were 
excluded if P > 0.05. 

Logistic regression helps select a set of key habitat 
variables that represent the probability of s ite occupancy. 
We chose to use logistic regression as a tool to identify key 
components of murrelet habitat, rather than < Jetennining the 
predictive probabilities of the occupancy rat es of murrelets. 
This was due to: ( 1 ) the use of retrospective sa mpling (Ramsey 
and others 1994); (2) the limitations of the; databases (i.e., 
data were collected over many years so sa mpling methods 
and data collectors may have changed ycirly and the data 
were not collected with the murrelet in m ind); aijd (3) the 
murrelet status of random sites was unkno wn. 

Nest Sites 

Habitat characteristics within nest plots were compared 
to average values for adjacent plots usir lg a Wilcoxon test 
(paired-sample signed-rank; Snedecor and Cochran 1980:140). 
Each nest site was treated as a block, ;and t'ne overall test 
statistic was based upon the cumulative differences among 
plots. A Chi-square test using a Bonfer.Toni. Z-statistic was 
used to compare snag decay class. 

Results 

Occupied Site Characteristics 

State Lands Database 

Tree Species — Douglas-fir was 'the dominant tree 
species ( SPECIES 1) in 67 percent of occupied sites (n = 
72) and 83 percent of random sites (n - • 2. 1 6). The codominant 
trees (SPECIES2) were generally a cc mibination of Douglas- 
fir, western hemlock (Tsuga heterophylla), Sitka spruce 
{Picea sitchensis), and red alder (Aln us oregona) in occupied 
and random sites. 

Age, Tree Diameter, and Tre* Density — Twenty-two 
percent (16 of 72) of occupied sit es were <80 years of age 
(AGE1993) whereas 60 percent (1 30 of 216) of random sites 
were <80 years old. Ninety four j >ercent (15 of 16) of these 
young occupied sites had remna m trees (TPH66, >66 cm 
d.b.h.), averaging 19.5 trees/ha (range: 2.5-75.0 trees/ha; 
s.e. = 5.0), and 15 of these sit es contained mature trees 
(TPH46, >46 cm d.b.h.; x = 73.< 0; s.e. = 1 1.6; range = 12.3- 
140.8). Random sites <80 year s of age averaged only 10 
remnant trees/ha (s.e. = 1.4; range = 2.5-59.3) (table 1, 
appendix I). All occupied sites >81 years old had remnant 
trees, averaging 46.9 trees/ha (range - 2.5-116.1 trees/ha; 
s.e. = 2.8). Similarly, 97 perce nt (83 of 86) of random sites 
>8 1 years old had remnants avt ;raging 45.9 trees/ha (range = 
2.4-79.0 trees/ha; s.e. = 2.3). 

Mistletoe, Platforms, and Moss Abundance — Variables 
in the database did not include information on nest platforms. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



193 



Grenierand Melson 



Chapter 19 



Inland Habitat Associations in Oregon 



Table 1 — Habitat characteristics of 72 Marbled MurreUt occupied sites and 216 random sites on State Lands, Oregon. 
Data are from Oregon Department of Forestry's OSCUR database, 1993. See appendix I for variable definitions 




HMBFAC 



0.9120 



'Significant > difference between occupied and random sites (P < 0.05, Kruskal-Wallis test) 



194 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



However, data concerning tree damage or abnormalities were 
collected and may indicate presence of nesting platforms. 
Sixty-three percent (45 of 72) of the occupied sites had tree 
damage or abnormalities, 33 percent (15 of 45) of which 
included mistletoe infestations. Damage or abnormalities 
observed in random sites was similar (59 percent; 128 of 216 
trees), however mistletoe infestation was less common in 
random sites (12 percent; 15 of 128 trees). 

National Forest Land Databases 

Vegetation Resources and Structure Examination 
Databases — Douglas-fir was the most common dominant tree 
in occupied and random sites. However, western red cedar 
(RC_DBH3, Thuja plicata), which occurred in approximately 
33 percent of occupied sites, had the largest mean diameter 
(x = 114.3 cm, s.e. = 6.65, n = 10). Occupied sites had an 
average of 13.4 remnant trees/ha (REM, trees >100 cm d.b.h.), 
and 63.7 percent canopy closure (table 2, appendix 2). 

Ecology Databases — Occupied sites were mid-seral 
stage stands with large tall conifer trees (table 3, appendix 
3). Most (75 percent) occupied sites were located in western 
hemlock climax forest types. The most common (17 percent; 
13 of 75) plant association in occupied sites was western 
hemlock/Oregon oxalis (Oxalis oregona). This plant 
association only occurred in 5 percent (11 of 225) of 
random sites, and the frequency of occurrence was 
significantly less than in occupied sites (% 2 = 30.2, df= 9, 
P - 0.0004). The most frequently occurring plant association 
in random sites was western hemlock/vine maple (Acer 
circinatum)/westem swordfern (Polystichum munitum) (12 
percent; 27 of 225). The western hemlock/Oregon oxalis 
sites occur on moist, shaded upper slopes and benches or 
alluvial terraces and have high conifer basal area (79.7 m 2 / 
ha. s.e. = 4.6, n = 9 compared to 71.0 m 2 /ha average for 22 
plant associations, s.e. = 2.0, n = 178; Hemstrom and 
Logan 1986). 

Research Database 

Ten occupied sites from this study were located in the 
western hemlock (n = 6) and Sitka spruce (n = 4) Zones 
(Franklin and Dyrness 1973), and included one or more of 
the following plant associations: vine maple (n = 4), salal 
(Gaultheria shallon) (n = 3), sword fern (n = 5), salmonberry 
(Rubus spectabilis) (n = 2), Pacific rhododendron (Rhodo- 
dendron macrophyllum) (n = 1 ), and Oregon grape (Berberis 
aquifolium) (n = 2). All sites had two or more canopy layers 
( x= 2.2. s.e. = 0.1), were located between middle of the 
slope and the ridgetop. had > 25 percent moss ( x index = 
3.8, s.e. = 0.2), contained some mistletoe or witches brooms 
( x index = 2.2, s.e. = 0.3), and had platforms for nesting ( x 
= 1.3 per plot, s.e. = 0.3) (table 4). The number of suitable 
platforms was correlated with the number of canopy layers 
(r = 0.39, P = 0.02, n = 34) and with mistletoe abundance (r 
= 0.43, P = 0.01, n = 34). 



Habitat Associations 

State Lands 

Twenty-one of 34 habitat variables were significantly 
different between occupied (n = 72) and random (n = 216) 
sites (P < 0.05) (table 1). In general, occupied sites were 
older (AGE 1 993 ), contained larger diameter conifers (DB H 1 - 
4, CDBH, CBA) and hardwoods (HDBH), had more large 
remnant trees/ha (TPH66), and had fewer small/medium- 
sized trees (TPH 15-36) and hardwoods/ha (HTPH) than 
random sites (table 1, appendix 1). 

Stepwise logistic regression suggested a model with 
only large remnant tree density (TPH66) as a significant 
predictor of murrelet habitat (P = 0.0164; n = 46 occupied 
and 80 random sites). It was estimated that the odds of use 
were 4.9 percent higher for each unit increase in large remnant 
tree density (95 percent Confidence Interval [CI] = 0.8 percent 
to 9.1 percent; s.e. = 0.02). 

National Forest Land Databases 

Vegetation Resources and Structure Examinations — 
Seven variables differed (P < 0.05) between occupied (n = 
30) and random (n = 90) sites. Occupied sites had less slope, 
less canopy closure (CC), more large Douglas-fir/ha 
(DF_TPH3), larger midstory western hemlocks (WH_DBH 1 ) 
and western redcedars (RC_DBH1, RC_DBH2), and fewer 
shade tolerant trees/ha >41 cm (STH_40) than random sites 
(table 2, appendix 2). 

Stepwise logistic regression suggested a model with 
larger diameters of midstory western redcedar (RC_DBH 1 ) 
and less canopy closure (CC) as significant predictors of 
murrelet habitat. It was estimated that the odds of use were 
48.3 percent higher for each unit increase in diameter of 
midstory western red cedar (95 percent CI = 9. 1 percent to 
100 percent; s.e. = 0.1565, n = 12 occupied and 28 random 
sites). Also, it was estimated that the odds of use were 16.5 
percent higher for each unit decrease in canopy closure 
when comparing occupied to random sites (95 percent CI = 
2.8 percent to 32.1 percent; s.e. = 0.064). 

Ecology — Four habitat variables were significantly 
different (P < 0.05) between occupied (n = 75) and random 
sites (n = 225). Occupied sites were in later serai stage 
classes (SERAL), and had taller (HT), older (AGE), and 
larger diameter trees (DIA) than random sites (table 3, 
appendix 3). 

Stepwise logistic regression suggested a model with 
tree height (HT) and percent cover of understory trees 
(TREER) as significant predictors of murrelet habitat. It 
estimated that the odds of use were 2.3 percent higher for 
every unit increase in tree height (95 percent CI = 0.8 
percent to 3.7 percent; s.e. = 0.0022; n = 34 occupied and 86 
random). Also the odds of use were 4.3 percent higher for 
every unit increase in percent cover of understory trees (95 
percent CI = 0.3 percent to 8.5 percent; s.e. = 0.0202). 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



195 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



Table 2 — Habitat characteristics of 30 Marbled Murrelet occupied sites and 90 random sites on the Siuslaw 
National Forest, Oregon, 1990-1992. Data are from forest stand inventories (Vegetation Structure and Vegetation 
Resource Examination). See appendix 2 for variable definitions 







Occupied 






Random 






Variable 


X 


(s.e.) 
Range 


n 


X 


(ml) 

Range 


n 


P-value 1 






SLOPE 


35.0 


(3.63) 
7-80 






5-90 




0.0053 


ELEV 


255.0 


(18.68) 
122-488 


30 


268.5 


(15.79) 
91-823 


89 


0.9606 


REM 


13.4 


(3.87) 
1-64 


16 


10.4 


(2.42) 
1-58 


29 


0.2626 


CC 


63.7 


(3.20) 

16-83 


30 


76.1 


(116) 

36-94 


90 


0.0004 


C_HT 


55.4 


(1.53) 
37-69 


30 


55.7 


(0.85) 
35-72 


90 


0.9999 


DF_DBH3 


88.4 


(3.12) 
41-122 


29 


94.7 


(2.18) 
64-160 


84 


0.3472 


DF_DBH2 


57.7 


(3.63) 
33-99 


23 




25-185 


49 


uj iH 


DF_DBH1 


27.9 


(2.46) 

17-46 


10 


30.7 


(4.01) 
13-137 


31 


0.6054 


DF.TPH3 


66.0 


(8.70) 
11-224 


29 


45.0 (3.34) 


84 


0.0267 


DF_TPH2 


17. X 


(4.08) 

1-69 


23 


16.8 


(1.66) 

0-43 


49 


0.4249 


DF_TPH1 


22.2 


(10.28) 


10 


16.3 


(2.30) 

2-56 


31 


0.8021 


WH_DBH3 


74.9 


(2.95) 

61-90 


14 


76.2 


(1.93) 
51-107 


53 


0.7991 


WH_DBH2 


52.6 


(4.09) 
26-94 






(152) 
25-94 


55 


0.1090 


WH_DBH1 


31.5 


(3.25) 
19-61 


15 


23.4 


(1.14) 

13-56 


53 
53 


0.0021 


WH_TPH3 


37.8 


(11.19) 
2-115 


14 


44.0 


(5.14) 
2-162 


0.3468 


WH_TPH2 


25.0 


(4.94) 

2-79 


19 


39.0 


(4.25) 

3-136 


56 


0.1038 


WH_TPH1 


42.7 


(9.29) 
4-130 


15 


52.4 


(6.25) 
3-185 


53 


0.5483 


RC_DBH3 


114.3 


(6.65) 
82-150 


10 


105.2 


(7.62) 

5-211 


34 


0.2263 


RC.DBH2 




(3.99) 
58-105 


12 


59.7 


(4.80) 
8-156 


32 


0.0081 


RC_DBH1 


39.9 


(4.01) 
23-71 


13 


26.7 


(1.80) 
3-56 


29 


0.0031 


RC_TPH3 


3.5 


(1.38) 
0-15 


10 


5.4 


(1.26) 
1-28 


23 


0.1265 


RC TPH2 


10.6 


(1.85) 

2-27 


12 


17.3 


(1.16) 

2-69 


32 


0.2626 


RC_TPH1 


27.4 


(7.41) 
2-94 


13 


25.0 


(3.83) 
4-113 


31 


0.9384 


DF_81 


27.9 


(2.82) 
3-57 


30 


23.0 


(1.70) 

1-66 


82 


0.1046 


STH_40 


47.4 


(8.10) 
1-149 


26 


67.5 


(5.14) 
1-194 


69 


0.0270 


1 Significant < 


lifference between occupied and random sites (P 


< 0.05, 


Kruskal-Wallis test) 





196 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



Table 3 — Habitat characteristics of 75 Marbled Murrelet occupied sites and 225 random sixes on the Siuslaw Sational 
Forest. Oregon, 1981-1984. Data are from the Ecology Program Intensive and Reconnisance Plots database, 1993. 
See appendix 3 for variable definitions 



Variable 



Occupied 



RiTijory. 



P-vaiue 



198.4 



(12.06) 
6-358 



75 



194.1 



(6.86) 
0-361 



225 0.6602 



TREER 


9.5 


(2-37) 
1-75 


50 


4.7 


(0.60) 
1-35 


126 


0.1737 


SHRUBH 




1-95 






1-99 






HERB 


52.8 


(4.02) 
3-99 


60 


45.7 


(2^8) 
1-98 


181 


0.0972 



SERAL 


2.9 


(0.07) 

2-4 


75 


2.7 


(0.04) 
2-5 


218 


0.0414 






9-156 






14-134 




0.3299 


TBAD 


17.8 


(169) 
9-46 


34 


173 


(110) 
0-64 


103 


0.5677 




1 Significant difference between occupied and random sites ( P < 0.05, Kruskal-Wallis test) 



Nest Sites 

Twenty-two nests were found between 1990 and 1993. 
Ten nests were located by observing adult Marbled Murrelets 
landing in or departing from a nest tree, three were found 
after finding eggshell fragments on the forest floor, six were 
located during the intensive tree climbing study, and three 
were found while reclimbing known nest trees. Overall, nine 
nests were confirmed active when discovered (Nelson and 
Peck, in press ). Of these, four were in the egg stage and five 
were in the nestling stage. Nestlings were thought to have 
fledged from three of these nests (see Nelson and Hamer. 
this volume a: Nelson and Peck, in press). 

Nest and Nest Tree Characteristics 

All nests were located in trees >127 cm in d.b.h. ( x = 
187.9. s.e. = 9.4) and >36 m tall ( x = 65.5, s.e. = 2.4; table 
5). Twenty nests were found in Douglas-fir, one was found 
in a Sitka spruce, and one was found in a western hemlock. 
Nests generally consisted of depressions in a moss mat 



(average moss depth = 5.3 cm), but compacted duff (needles, 
lichen, debris) substrates were also used. Nest cups averaged 
1 2.0 x 1 1 . 1 cm (length x width) and were located on platforms 
considerably larger than the nest ( x = 42.2 x 31.7 cm, s.e. = 
4.2 and 2.9, respectively). The placement of nests on nest 
branches was variable, ranging from 1 to 762 cm from the 
trunk of the tree (x = 101.3, s.e. = 35.7). The nest in the 
Sitka spruce was furthest from the tree trunk (at 762 cm); all 
other nests were located <230 cm from the trunk. Nest 
height ranged from 18 to 73 m above the ground ( x = 50.3, 
s.e. = 2.4; table 5). 

Nest Stand Characteristics 

Nest stands generally consisted of 2 to 3 canopy layers 
( x = 2.2, s.e. = 0.09, n = 10). Trees <45 cm d.b.h. were most 
numerous (94.3 trees/ha), followed by trees >81 cm d.b.h., 
and trees 46-80 cm d.b.h. (64.3 and 39.7 trees/ha) (table 6). 
There were 7. 1 (s.e. = 1 .7) standing snags/ha in nest stands, 
with a mean height of 15.5 m (s.e. = 2.1) and mean d.b.h. of 



LSDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



197 



Grenier and Nelson Chapter 19 

Table 4 — Overall habitat characteristics of occupied stands, Oregon Coast Range, 1992 



Inland Habitat Associations in Oregon 



Characteristics 


X 


s.e. 


Range 


." 


Stand size (ha) 


6.5 


1.1 


2.0-12.1 


10 


Elevation (m) 


298.8 


56.6 


61-610 


10 


Distance inland (km) 


15.9 


3.5 


2.3-33.0 


10 


Distance to stream (m) 


310.0 


98.3 


0-1000 


10 


Aspect (°) 


194.3 


15.7 


65-328 


20 


Slope (pet) 


53.7 


7.0 


3-126 


20 


Position on slope 1 


3.9 


0.2 


2-5 


20 


Canopy closure (pet) 


48.9 


4.3 


13-83 


20 


Canopy height (m) 


56.9 


3.4 


32-88 


20 


Canopy layers (no.) 


2.2 


0.1 


2-3 


20 


Number of platforms 


1.3 


0.3 


0-5 


20 


Moss index 2 


3.8 


0.2 


3-5 


20 


Mistletoe index 3 


2.2 


0.3 


1-5 


20 


Tree d.b.h. (cm) 


43.8 


0.9 


10-206 


1374 


Tree d.b.h. ^46 cm 


80.6 


1.3 


46-206 


518 


Tree d.b.h. >81 cm 


109.3 


1.7 


81-206 


208 


Tree density (no./ha) 


349.8 


24.1 


208.8-565.2 


20 


Tree density >46 cm (no./ha) 


131.9 


15.0 


10.2-269.9 


20 


Tree density >81 cm (no./ha) 


55.7 


6.8 


15.3-127.3 


19 


Snag d.b.h. (cm) 


57.8 


2.7 


10.5-187.0 


212 


Snag density (no./ha) 


54.0 


9.3 


5.1-142.6 


20 


Tree basal area (mVha) 


56.3 


5.2 


12.6-86.3 


20 


Tree basal area ^46 cm (mVha) 


69.9 


5.9 


4.6-111.5 


20 


Tree basal area ^8 1 cm (m 2 /ha) 


53.1 


7.0 


8.0-101.1 


19 


Snag basal area (mVha) 


16.8 


4.3 


0.5-66.8 


20 



1 Position on Slope: 1 = canyon bottom, 2 = lower 1/3, 3 = middle 1/3, 4 = upper 1/3, 5 = ridge top. 

2 Moss Index: = none, 1 = trace, 2 = 1-24 percent, 3 = 25-49 percent, 4 = 50-100 percent. 

3 Mistletoe Abundance: Divide tree into thirds. Rate each section; for no mistletoe, 1 if less than 1/3 of branches 
are infected, and 2 if more than 1/3 are infected. Score range is 0-6. 



81.4 cm (s.e. = 7.4). Mistletoe and witches brooms were 
found on dominant trees in most stands. Nests were located 
between 1 .6 and 40 km from the ocean. 

Three habitat variables at 10 nest sites differed from 
adjacent plots (table 6). The density of live trees in the 
largest size class (d.b.h. >81 cm) was greater in adjacent 
plots than in nest plots (64.3 versus 43.3 trees/ha; P = 0.03). 
Snag height was greater and snags were less decayed in 
adjacent plots compared with nest plots (15.5 versus 7.5 m; 
P = 0.03; 3.3 versus 3.7; P < 0.001, respectively). In addition, 
canopy closure was marginally greater in adjacent plots than 
nest plots (61.3 percent versus 41.2 percent, P = 0.06). 



Discussion 

Occupied Sites 

In Oregon, sites occupied by Marbled Murrelets were 
characterized by large diameter conifers and hardwoods, tall 
trees, high densities of dominant trees, and low densities for 
small diameter trees (conifers and hardwoods combined). In 
addition, these sites were older, located on gentler slopes, 
and had less percent canopy cover than random sites. Important 
habitat components for predicting occupancy were height 
and density of dominant trees, diameter and percent cover of 
midstory and understory conifers, and canopy cover. 



198 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



Table 5 — Summary of nest and nest tree characteristics from 22 nests found in Oregon between 1990 and 1993 



Characteristic 


X 


s.e. 


Range 


n 1 


Nest tree 








1 


D.b.h. (cm) 


187.9 


9.4 


127-279 


19 


Diameter at nest limb (cm) 


80.1 


6.2 


36-122 


15 


Height (m) 


65.5 


2.4 


36-85 


19 


Nest branch 










Height (m) 


50.3 


2.4 


18-73 


21 


Diameter at trunk (cm) 


31.1 


2.6 


15-56 


19 


Diameter at nest (cm) 


29.4 


2.7 


10-50 


20 


Position in crown (pet.) 










( nest ht/tree ht) 


74.7 


2.6 


50-9221 




Length (m) 


5.1 


0.7 


1-12 


17 


Distance from trunk to nest 












101.3 


35.7 


1-762 


21 


Nest cup 










Length (cm) 


12.0 


1.0 


6.0-26.0 


19 


Width (cm) 


11.1 


0.7 


7.0-17.8 


19 


Depth of cup (cm) 


3.1 


0.3 


0.5-5.1 


17 


Depth of moss on branch (cm) 


5.2 


0.6 


0.6-12.0 


18 


Nest platform 










Length (cm) 


42.2 


4.2 


11-66 


14 


Width (cm) 


31.7 


2.9 


10-51 


21 


Canopy closure above nest (pet.) 


78.6 


3.5 


40-100 


18 



1 Characteristics were not measured at some nests. 

2 Excludes a nest in a Sitka spruce 7.6 m from the trunk. This nest was 3.3 times farther than the next most distant 
nest and 1 1 .2 times farther than the mean. 



Tree species composition of occupied sites was consistent 
with composition across the landscape. The western hemlock/ 
Oregon oxalis plant association, within the western hemlock 
zone, was especially important. These sites are very fertile 
and moist and may produce larger trees (Hemstrom and 
Logan 1986). In addition, moisture in these sites may decrease 
the likelihood of intense fires, thereby allowing higher 
densities of remnant trees. 

Historically, extensive, and sometimes catastrophic, fires 
occurred in the Oregon Coast Range (Agee 1994). These 
fires created diverse forests with attributes of older-aged 
forests that are not found in intensively managed plantations. 
For example, natural stands generally have more tree species, 
less uniform tree sizes, more random spacing of trees, and 
larger remnant overstory trees, as compared to even-aged 
stands of the same age (Spies and Franklin 1991). Many of 
the occupied sites in Oregon were created naturally, and 
many have not been managed (i.e., thinned or partially 
harvested). Thus, these sites were uneven-aged forests and 
they included a variety of tree sizes and ages. Spacing of the 
dominant trees was not uniform, allowing midstory and 
understory trees to fill in the gaps in the canopy. The structure 
of these forests, in most cases, was similar to old-growth 



forests, although tree density was lower and average tree 
size smaller than "classic" old-growth (as defined in Franklin 
and others 1986). It is the structure of these stands, the large 
trees with nesting platforms, hiding cover (vertical canopy 
cover), and variable canopy cover, that are important to 
murrelets. Mean tree age alone and low canopy closure do 
not indicate the quality of the habitat. For example, some 
sites on state lands were typed as young (<80 years old), yet 
all of these sites had remnant trees (>66 cm d.b.h.), except 
one, and all had other older forest structures that survived or 
were created by fire (snags, woody debris). The single young 
site without remnant trees was located adjacent to and 
contiguous with a stand that contained 42.0 remnant trees per 
ha. In addition, while low canopy closure may allow murrelets 
access to nests, most (70 percent) nests near openings or 
edges have been unsuccessful (Nelson and Hamer, this volume 
b). Therefore, suitable murrelet habitat likely includes complex 
structure, high densities of large trees, large nesting platforms, 
and hiding cover. 

The key components of occupied sites in this study 
were similar to occupied sites throughout the Pacific 
Northwest and California, and to other studies in Oregon 
(Nelson 1989, 1990a). Most sites used by murrelets have 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



199 



Grenier and Nelson 



Chapter 19 



Inland Habitat Associations in Oregon 



Table 6— Comparison' of vegetation characteristics between 2 5-meter plots around nest trees and within three plots 2 
adjacent to 10 nest trees found in Oregon between 1990 and 1992 





Nest 


plots 


Adjacen 

X 


plots 
(s.e.) 






X 


(s.e.) 


P-value 


Characteristic 




Range 




Range 




Canopy closure (pet.) 


41.2 


(9.2) 


61.3 


(6.6) 


0.06 




12-99 




22-91 


■■1 


Tree density (no./ha) 
d.b.h. < 45 cm 


87.6 


(12.3) 
25.5-147.7 


94.3 


(12.7) 
5.0-134.1 


0.42 


d.b.h. = 46-80 cm 


44.8 


(9.2) 
10.2-91.6 


39.7 


(3.0) 
28.9-56.0 


0.65 


d.b.h. > 81 cm 


43.3 


(7.1) 
15.3-91.6 


64.3 


(7.5) 
20.4-88.3 


0.03 


Nest platforms (no./tree) 


6.7 


(10) 
00-110 


4.7 


(0.8) 
3-8 3 


0.10 


Mistletoe abundance 














2.2 


(0.4) 
0.0-4.0 


2.3 


(0.5) 
1.0-5.0 


0.61 


Snag density (no./ha) 


5.4 


(0.9) 
2.0-11.0 


7.1 


1.3-18.7 




Snag height (m) 


7.5 


(2.0) 
2.0-21.2 


15.5 


(2.1) 
4.9-24.8 


0.03 


Snag d.b.h. (cm) 


75.4 


(9.0) 




(7.4) 






30.8-108.3 




33.2-113.0 




Snag decay class 


3.7 


(0.2) 
1-5 


3.3 


(0.1) 
1-5 


<0.001 



1 Comparisons based upon a Wilcoxon paired-sample test for all characteristics except decay class of snags which 
is compared using a Chi-square test and a Bonferroni Z-statistic (Neu and others 1974) for the distribution of 
observations within five decay-class categories. 

2 Characteristics measured within two adjacent plots at one nest site. 



been in older-aged forests, with high densities of dominant 
trees (Burger, this volume a; Hamer, this volume; Miller 
and Ralph, this volume; Nelson 1989, 1990a; Paton and 
Ralph 1990). In addition, in British Columbia, the presence 
of murrelets was correlated with the presence of Sitka spruce 
and western hemlock, low elevations, and large trees which 
had platforms for nesting (Burger 1994, this volume a; 
Manley and others 1992; Rod way and others 1991). In 
Washington, number of platforms and moss abundance were 
the most important variables for predicting murrelet 
occupancy (Hamer, this volume). Elevation (low) and canopy 
closure (moderate) were also key characteristics of occupied 
sites (Hamer, this volume). 

Nest Sites 

Murrelets used mature or old-growth forests and large 
diameter Douglas-fir, Sitka spruce, and western hemlock trees 



for nesting. Nests (n = 22) were on large limbs ( x = 29 A cm) 
usually 1 8 m or more above ground level, and were located 
in trees > 127 cm ( x= 187.9) in diameter. Other structures at 
the nest were also important, including high vertical canopy 
cover above the nest cup ( x = 78.9 percent). In addition, all 
nests were located in forests with multi-layered canopies (2- 
3) with a wide range of both tree densities (121.0-718.8 ha) 
and canopy cover (12-99 percent). 

In general, the forest immediately around the nest trees 
was open, with fewer dominant trees. The density of 
dominant trees at nest sites was lower than at adjacent sites. 
However, occupied sites had a higher density of dominant 
trees than random sites. Apparently, it is important for 
murrelets to have numerous dominant trees throughout a 
stand to provide nesting opportunities, but at the nest site, 
a lower density of dominant trees may facilitate access to 
the nest tree for a bird with limited flying maneuverability. 



200 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 

» 



Grenier and Nelson 



19 



Inland Habitat Associations in Oregon 



However, nests may have been easier to find by observers 
in situations where tree densities were low. In addition, 
very low canopy cover may allow higher predation rates 
(see Nelson and Hamer, this volume b). Conclusions about 
appropriate canopy cover levels for nesting murrelets cannot 
be made without further research on nesting success and 
with larger sample sizes. 

The characteristics of nests in Oregon were similar to 
those described by others in Alaska, British Columbia, 
Washington, and California (Hamer and Nelson, this volume 
b; Jordan and Hughes, in press; Manley and Kelson, in 
press; Naslund 1993b; Naslund and others, in press; Nelson 
and Hamer 1992; Singer and others 1991, in press). Excluding 
Alaska, all nests have been located in trees > 88 cm in 
diameter and > 18 m in height (n = 47) (Hamer and Nelson, 
this volume b). 

Conclusions 

Our results support previous studies and observations 
that murrelets use older forests or forests with old-growth 
characteristics. Key habitat characteristics of occupied sites 
were tree height, density of dominant (or remnant) trees, 
diameter and percent cover of midstory and understory trees, 
and canopy cover. Nests were located in large trees with 
large platforms and high vertical canopy cover. 

Additional detailed information on the characteristics 
(platform availability, abundance of mistletoe and moss) 
of Marbled Murrelet habitat are needed to refine the 
definition of suitable habitat in Oregon. Studies designed 
to collect habitat data specific to murrelets from plots in 
the forest (including the key variables listed above) are 
needed. In addition, further investigations into plant 
associations are recommended. 

There are limitations to describing murrelet habitat on 
the basis of occupancy, presence, or abundance. We can 
describe the general features of habitat used by this species; 
however use of these measures as a means for determining 
habitat quality, suitability, or preference may not be valid, 
especially in patchy habitats (Fretwell and Lucas 1969, Hanski 
1982, Van Home 1983). Models developed to measure habitat 
quality and suitability have included components of density, 
reproductive rates, genetic contribution of adults to the next 
generation, and survival of adults and juveniles (Fretwell 
and Lucas 1969. Van Home 1983). A more adequate means 
of evaluating habitat suitability will be to explore the 
relationship between reproductive success, and habitat and 



landscape characteristics. This should be an emphasis of 
future research projects. 

In addition, caution is advised in using our description 
of occupied sites from the state and federal databases. These 
data were not collected to describe murrelet habitat, and our 
analysis was retrospective. Occupied sites also were not 
selected randomly. Therefore, occupied site characteristics 
may be an artifact of how stands were chosen for timber 
harvesting and thus murrelet surveying, and which habitat 
variables were measured. Future research should include 
collecting data specific to biology of murrelets, e.g., 
availability of platforms, moss, and mistletoe. 

Acknowledgments 

Funding for this project was provided by the Oregon 
Department of Fish and Wildlife, Nongame Wildlife Research 
Program; the USDI Bureau of Land Management, Salem 
and Coos Bay Districts; the U.S. Fish and Wildlife Service, 
U.S. Department of Interior; and the USDA Forest Service. 
Special thanks go to G. Gunderson, W. Logan, L. Mangan, 
M. Nugent, C. Puchy, M. Raphael for their cooperation, 
support, and advice. 

We also thank S. Andrews, C. Bickford, R. Davis, S. 
Madsen, C. McCain, C. Wettstein of the Siuslaw National 
Forest, B. Reagan, B. Reich, K. Graham of the Oregon 
Department of Forestry, and N. Allen, C. Bruce, G. Sieglitz of 
the Oregon Department of Fish and Wildlife for access to their 
databases. Data collection and management were conducted 
by D. Elliott, J. Hardin, A. Hubbard, B. Peck, M. Pope, M. and 
J. Raisinghani. J. Reams, J. Rosenthal, T. Ross, M. Shaughnessy, 
and J. Wells. Additional logistical support was provided by 
D. Crannell, J. Guetterman, J. Heeney, B. Hill, S. Hopkins, 
K. Kritz, from the U.S. Bureau of Land Management, and D. 
Gutherie and S. Livingston of the U.S. Forest Service. 

F. Ramsey and D. Schafer from Oregon State University, 
Statistics Department provided assistance with data analyses. 
We thank J. Baldwin. B. Block, A. Burger, P. Connors, T. 
De Santo, R. Mannan, C.J. Ralph, M. Raphael, R. Steidl and 
J. Weeks for reviewing earlier drafts of this manuscript and 
making numerous suggestions that greatly improved its 
quality. Support for preparation of this manuscript was 
provided by the Oregon Department of Fish and Wildlife, 
USDA Forest Service, USDI Bureau of Land Management, 
and U.S. Fish and Wildlife Service, U.S. Department of 
Interior. This is Oregon State University Agricultural 
Experiment Station Technical Paper Number 10,537. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



201 



Grenier and Nelson Chapter 19 Inland Habitat Associations in Oregon 

Appendix 1 — Variables used to describe Marbled Murrelet occupied sites using data from Oregon 's Depart- 
ment of Forestry OSCUR database. Note that the 1, 2, and 3 assigned to DBH are reversed compared to federal 
lands database 



DBH1 Average diameter (centimeters) at breast height (1.4 m) for the dominant species 

BA1 Basal area (m 2 /hectare) for dominant species 

MBFAC 1 Cubic meters per hectare of dominant species ' 

DBH2 Average diameter (centimeters) at breast height (1.4m) for the codominant species 

BA2 — Basal area (m 2 /hectare) for codominant species 

MBFAC2 Cubic meters per hectare of codominant species 

DBH3 Average diameter (centimeters) at breast height ( 1 .4 m) for other species (midstory) 

BA3 Basal area (m^ectare) for other species 

MBFAC3 Cubic meters per hectare of other species 

DBH4 Average diameter (centimeters) at breast height (1.4 m) for other species (understory) 

BA4 Basal area (m 2 /hectare) for other species 

MBFAC4 Cubic meters per hectare of other species 

TPH15 Trees per hectare in the 15.2-20.3 cm d.b.h. class 

TPH20 Trees per hectare in the 20.3-25.4 cm d.b.h. class 

TPH25 Trees per hectare in the 25.4-30.5 cm d.b.h. class 

TPH30 Trees per hectare in the 30.5-35.6 cm d.b.h. class 

TPH36 Trees per hectare in the 35.6-45.7 cm d.b.h. class 

TPH46 Trees per hectare in the 45.7-66.0 cm d.b.h. class 

TPH66 Trees per hectare in the 66.0+ cm d.b.h. class 

AGE 1993 Age of the tree type calculated from 1993; age was estimated from several large trees in each plot. 

CDBH Average diameter (centimeters) at breast height ( 1 .4 m) for conifers 

HDBH Average diameter (centimeters) at breast height (1.4 m) for hardwoods 

CTPH Trees per hectare for conifers 

HTPH Trees per hectare for hardwoods 

CBA Basal area (m 2 /hectare) for conifers 

HBA Basal area (mVhectare) for hardwoods 

CMBFAC Cubic meters per hectare of conifers 

HMBFAC Cubic meters per hectare of hardwoods 



1 1,000 board feet = 5.83 m 3 /hectare (ha) 



202 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Grenier and Nelson 



19 



Inland Habitat Associations in Oregon 



Appendix 2 — Variables used to describe stands of mature and old-growth forests on the Siuslaw National Forest, Corvallis, 
Oregon. Data were collected using Vegetation Resource and Structure Examination methods, 1990-1993. Note that the 
1, 2, and 3 assigned to DBH are reversed compared to state lands database. 



SLOPE Slope of the plot 

ELEV Elevation of the plot (meteis) 

REM Number of remnant trees (> 100 cm d.b.h.) 

CC Canopy closure as a percentage 

C_HT Canopy height (meters) 

DF_DBH3 Average diameter (centimeters) at breast height for dominant (canopy) Douglas-fir 

DF_DBH2 Average diameter (centimeters) at breast height for codominant Douglas-fir 

DF_DBH1 Average diameter (centimeters) at breast height for early-seral (midstory) Douglas-fir 

DF_TPH3 Number of dominant Douglas-fir per hectare 

DF_TPH2 Number of codominant Douglas-fir per hectare 

DF_TPH 1 Number of early-seral (midstory) Douglas-fir per hectare 

WH_DBH3 Average diameter (centimeters) at breast height for dominant western hemlock 

WH_DBH2 Average diameter (centimeters) at breast height for codominant western hemlock 

WH_DBH1 Average diameter (centimeters) at breast height for early-seral (midstory) western hemlock 

WH_TPH3 Number of dominant western hemlock per hectare 

WH_TPH2 Number of codominant western hemlock per hectare 

WH_TPH 1 Number of early-seral (midstory) western hemlock per hectare 

RC_DBH3 Average diameter (centimeters) at breast height for dominant red cedar 

RC_DBH2 Average diameter (centimeters) at breast height for codominant red cedar 

RC_DBH1 Average diameter (centimeters) at breast height for early-seral (midstory) red cedar 

RC_TPH3 Number of dominant red cedar per hectare 

RC_TPH2 Number of codominant red cedar per hectare 

RC_TPH 1 Number of early-seral (midstory) red cedar per hectare 

DF_81 Number of Douglas-fir per hectare with an average diameter > 81 .3 cm at breast height 

STH_40 Number of shade tolerant trees per hectare with an average diameter > 40.6 cm at breast height 

(midstory) 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



203 



Grenier and Nelson Chapter 19 Inland Habitat Associations in Oregon 

Appendix 3 — Variables used to describe forest stands >7S years old on the Siuslaw National Forest. Data were 
collected by the Ecology Program (USDA 1983, VSDA 1985), Siuslaw National Forest, Corvallis, Oregon, 
1982-1986 



ELEV The elevation of the plot (meters) 

ASPECT The aspect toward which the plot faces recorded as degrees azimuth 

SLOPE The slope of the plot in percent 

TREER-- The percent cover of understory trees (<3.7 meters in height) 

SHRUBH The percent cover of high woody perennial plants greater than 0.9 meters in height 

HERB The percent cover of herbaceous plants 

MOSS The percentage of the ground area which is covered by moss or lichens 

SERAL The serai stage of the site [1 = very early, pioneer (1 to 30 years) brush and small trees; 2 = early 

serai (30 to 100 years) young stand; 3 = mid-seral (100 to 00 years) mature stand; 4 = late serai, 
highly stable (200 until late serai species are gone) old growth; 5 = climatic, topographic, or 
edaphic climax (serai dominants gone)] 

TBAL The total basal area in m 2 /hectare of living trees 

TBAD The total basal area in m 2 /hectare of standing dead trees, including snags in all stages of deacy 

greater than 3.7 meters in height 

DIA The diameter of the site dominant trees to the nearest centimeter 

HT The total height measured on the site dominant trees 

AGE The breast height age of the site dominant trees from bore count 



2 *** USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 20 

Relationship of Marbled Murrelets with Habitat Characteristics 
at Inland Sites in California 



Sherri L. Milter 



C. John Ralph 1 



Abstract: We examined the range and the relationships of Marbled 
Murrelet {Brachyramphus marmoratus) behavior with habitat and 
landscape characteristics in isolated old-growth and residual forest 
stands from 2 to 400 ha in California. In large contiguous stands of 
old-growth forest in parks, we examined relationships of murrelet 
detections with elevation and topography. In isolated stands we 
found higher murrelet detection levels in stands with higher domi- 
nant and codominant crown cover and >50 percent coast redwood 
(Sequoia sempervirens). Surveys also were more likely to detect 
occupied behaviors at stands with higher crown cover and a greater 
proportion of redwoods. Density of old-growth cover and species 
composition may be the strongest predictors of murrelet presence 
and occupancy in California. Contrary to previous studies, we did 
not find that larger stands were more likely to have murrelets 
present. In the large park stands, we found that mean detection 
levels and the number of occupied stations were highest in the 
major drainages and at lower elevations. Major ridges tended to 
have lower detection levels and fewer occupied behavior stations. 



In recent years, much has been learned about the occurrence 
of Marbled Murrelets {Brachyramphus marmoratus) at inland 
forest sites. Throughout most of its range, the murrelet nests 
in old-growth forests within 50-75 miles of the coast (Carter 
and Morrison 1992). In California, Paton and Ralph (1990) 
conducted general surveys (Paton, this volume) to determine 
the distribution of murrelets in coastal old-growth and mature 
second-growth forests. Concentrations were found in regions 
containing large, contiguous, unharvested stands of old-growth 
redwood, mostly within state and federal parks, with the 
highest detection numbers in stands >250 ha. In excess of 
200 detections for single-survey mornings have been recorded 
at some survey stations in remaining unharvested stands 
within parks in California, including Redwood National Park 
and Prairie Creek State Park in Humboldt County (Ralph 
and others 1990); and Big Basin State Park in San Mateo 
County (Suddjian, pers. comm.). 

Federal listing of the Marbled Murrelet as threatened 
(U.S. Fish and Wildlife Service 1992) has created a need for 
information about the role of habitat and landscape features 
for the murrelet. 

We conducted two studies to examine the relationships 
of the murrelet to habitat and landscape characteristics within 
old-growth forests, as defined by Franklin and others (1986). 
In isolated stands in fragmented landscapes (the Stand Study), 
we compared murrelet detections with stand size, structure, 



1 Wildlife Biologist and Research Wildlife Biologist, Pacific South- 
west Research Station, USDA Forest Service, Redwood Sciences Labora- 
tory, 1700 Bayview Drive. Areata, CA 95521 



and landscape characteristics. In large contiguous stands of 
old-growth in state and federal parks (the Park Study), we 
examined murrelet detections with landscape features, such 
as elevation and topography. We confined our study to old- 
growth forests, because previous studies indicate murrelets 
nest only in forests with these characteristics. 

Methods 

The survey methods followed the intensive survey 
protocol of Ralph and others ( 1 993). To maximize the number 
of visual detections, we selected station positions at the 
edges of the isolated stands or at interior locations with 
openings in the canopy whenever possible. Observers could 
move within a 50-m radius of the station. 

We estimate that, for an individual forest stand, four 
surveys are needed to determine with a 95 percent probability 
that murrelets are present {appendix A). If below canopy 
behaviors were observed, we categorized the stand as 
Occupied (see below) for analyses. During 1992 and 1993 
for the Stand Study, we attempted to survey each isolated 
stand at least four times between 15 April and 7 August. 
Surveys at each stand were distributed throughout the survey 
period whenever possible. However, due to difficult access 
for some stands, surveys in some areas were temporally 
aggregated. To eliminate potential effects from aggregated 
surveys, detection levels were standardized for seasonal 
variation (see Analyses below). 

For the 1993 Park Study, within the boundaries of the 
large stands of old-growth forests in national and state parks 
(fig. 1), stations were placed in a matrix over the landscape, 
as illustrated in figure 2. We surveyed all sections of park 
stands with adequate accessibility. We placed stations 400 
meters apart on roads and trails, and 400 meters out 
perpendicular to trails, creating a matrix. Ralph and others 
(1993) found that observers detect few birds at distances 
>200 m, therefore, we assumed each station covered a 200- 
m radius circle, approximately 12.5 ha. Due to safety 
considerations for observers hiking to stations in pre -dawn 
hours, we limited stations to within 400 meters of a trail or 
road. Stations were surveyed once during the survey season. 
We attempted to avoid surveys at adjacent stations on the 
same morning. 

The species' range in northern California was determined 
by examining the results of inland surveys conducted from 
1988 through 1992 by government agencies and private 
landowners. Murrelet use for each stand or station was 
determined by the number and type of detections. All survey 
stations were digitized into a Geographic Information System 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



205 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



-a 



County 
Locations 




Stale and National Parks within 
Redwood National Park boundaries 
(Del Norte and Humboldt Counties) 



Humboldt Redwoods State Park 
(Humboldt County) 



San Francisco Bay 



Monterey Bay 




Big Basin Redwoods State Park 
(San Mateo County) 



Figure 1— Location of state and national parks surveyed during the summer of 1993. Shaded areas represent 
distribution of old-growth forests within the parks. 



(GIS) database (ARC/INFO 6.1.1) and grouped by distances 
from the ocean by 10-km bands from to 60 km (fig. 3). 

Definition and Selection of Isolated Study Stands 

Isolated stands were located by examining habitat maps 
of private lands, state and federal parks, and national forests. 
The maps were drawn from interpretation of aerial photographs. 
For the stand selection process, stand size was estimated from 



measurements on the maps. Stands were randomly selected 
from size categories of 2 to 20 ha, 21 to 40 ha, 41 to 100 ha, 
and greater than 100 ha. If the stand was accessible, it was 
visited and visually inspected. If the stand was old-growth or 
residual forest, the stand was surveyed, if not, then another 
stand was selected. Upon completion of field work, station 
locations and stand perimeters were adjusted on maps according 
to ground-truthing, then digitized into a GIS database. 



206 



USDA Forest Servicemen. Tech. Rep. PSW-152. 1995. 



Miller and Ralph 



Chjp;e: 20 



Inland Habitat Relationships in California 





| 








^/ Marbled Murrelet 












J , Survey 


Stations - 1993 


• 
• 














Prairie Creek 


o 
O 










HV V r^-^^ 




State Park 


3 

■ 

0. 












V 


• 






sD i 


Wv J 

tlfl 

A ^ 
A • A 


v" 
V 


• □ 

/ A ^ 

D/ * 
A ° A 
\ A 


D 


□ j 

k A 

UP A 

^JT A 
V v 

■^ V A 

' A 








A 


A 


*^LV A J 


mm 

W V v 


O 


Tributary drainage 




© 


Major 


drainage 




D^^^ 


□ 


# \_ 




D 


General slope 




/ D 




Location \ I f 




V 


Minor 


ridge 




/ D 






A 


Major 


ridge 




1 • 




V\ 




• 


Occupied site 






• 


- NJ 










Hetara 



Figure 2 — Spatial and topographical distribution of a subset of Marbled Murrelet stations surveyed 
at Prairie Creek Redwoods State Park during the summer of 1993. Occupied sites are shaded in 
groups to illustrate possible associations with topographical features. 



Stand area, perimeter length, and distance from salt 
water were derived from the GIS database. For stands with 
inclusions of non-forested area within the stand, we added 
the length of the lines around the stand and around the 
inclusions for the total perimeter measurements. Perimeter, 
therefore, is a measure of the amount of forest edge in and 
around the stand. 

Stand type was characterized as residual or old-growth. 
This variable is a measure of harvest history for the stand, 
but is not a direct measure of years since the last disturbance. 
Old-growth stands contained trees greater than 90 cm diameter 
at breast height (d.b.h.) with no history of timber harvest and 
some evidence of decadence in the canopy. Residual stands 



had some history of partial removal of large trees with the 
remaining dominant trees greater than 90 cm d.b.h.. Some 
stands with contiguous areas of old-growth and residual 
were classified as mixed. 

Stands also were classified by density as determined by 
interpretation of aerial photographs. Density was defined as 
the percent of the old-growth canopy cover (dominant and 
codominant trees): sparse, <25 percent; low, 25-50 percent; 
moderate, 51-75 percent; and dense, >75 percent. Species of 
dominant trees (>50 percent) was determined from aerial 
photography and verified by vegetation information after 
visiting the stand. For the purpose of this study, a stand was 
a single, isolated group of old-growth trees surrounded by 



L SDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



207 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



Kilometers From Coast 10 20 30 40 50 60 



Pacific Ocean 




10 km 




Figure 3— Distribution of Marbled Murrelet survey stations in northern California. Stations are 
located on private and public lands and surveys were conducted one or more seasons from 1 988 to 
1 994. Open circles represent one survey station or a group of stations in one isolated stand. In areas 
with high concentrations of stations, open circles appear filled in or shaded. 



non-forested or harvested habitat. If groups of trees were 
less than 160 meters apart they were considered one stand. 
Stands that met all of the following criteria were 
included in the group of potential survey sites: old-growth 
or residual stands with dominant and codominant trees that 
comprised at least 20 percent canopy cover; size between 2 
ha and 400 ha; distance from coast less than 40 km (25 
miles); dominant vegetation type of coast redwood {Sequoia 
sempervirens) or Douglas-fir (Pseudotsuga menziesii) at 



elevations of less than 1,000 m; and safely accessible by 
road or well-defined trail. 

Analyses 

Standardization for Seasonal Variation 

Various factors may influence the numbers of detections 
of murrelets at inland locations, including environmental 
conditions, time of year (O'Donnell and Naslund, this 



208 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



volume), and observer (O'Donnell, this volume). To help 
eliminate the effects of observer bias, all stands were surveyed 
by two or more observers. The influence of weather on 
numbers of detections appears to be highly variable (Naslund 
and O'Donnell, this volume). The effect of weather is 
probably stochastic with respect to survey days, and we 
assumed it did not have an overall impact at a site because 
surveys were distributed throughout the breeding season. 
The seasonal variation in detection levels, however, has 
been well documented and quantified at several sites in 
California (O'Donnell and Naslund, this volume). To identify 
differences in murrelet use (detection levels) of stands in 
our study, we first accounted for the effect of season on 
detection levels. 

Morning surveys were conducted throughout the breeding 
season in multiple years at three sites in Humboldt County. 
The sites at Lost Man Creek (Redwood National Park) and 
James Irvine Trail (Prairie Creek Redwoods State Park) 
were surveyed from 1989-1993. The Experimental Forest 
site was surveyed in 1989, 1990, 1992, and 1993. We 
attempted to monitor each site weekly. Data from these three 
sites was used to calculate standardization factors. 



Standardization 

The following method was used to calculate a factor to 
standardize the number of detections for seasonal differences. 

1. We examined the distribution of detections (fig. 4) 
over all years for the three sites and used a Kruskal- 
Wallis test to determine that the distributions by 
season were similar for the three sites (P < 0.0001). 
Surveys from all sites and years then were pooled. 

2. We calculated the mean number of detections per 
survey for the period 15 April to 12 August, that we 
refer to as the summer mean. 

3. We then calculated the mean numbers of detections 
per survey for each 10-day interval, the interval mean. 
Detection levels for periods longer than 10 days 
began to show the effects of seasonal variation. 

4. The ratio of each of the 12 interval means and the 
summer mean was calculated (interval mean/summer 
mean = standardization factor). 

The 10-day intervals and corresponding standardization 
factors calculated for the data from the three sites are presented 
in table 1. 



250 -r 



200 -- 



O 1 50 - - 



o 
Q 



o 

Z 



o 

1 



:. -- 



— All sites combined 

' " Experimental Forest 

— James Irvine Trail 
" " Lost Man Creek 



A. 



^•o 




a. 


Q. 


> 


< 


< 


? 


in 




m 



> 
- 


C C 

3 3 


c 


in 

CM 


■4 Tt 

10 Day Period 


4 

CM 



4 



4 

CM 



3 
< 



Figure 4 — Mean Marbled Murrelet detections from forest surveys at three sites in northern California: James Irvine 
Trail, Prairie Creek Redwoods State Park: Lost Man Creek, Redwood National Park; and the USDA Forest Service 
Experimental Forest Klamath. Means for the three sites combined by 10 day intervals also are presented. Surveys 
were conducted 3-4 times per month most years from 1 989-1 993 and points represent the means for 1 0-day intervals. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



209 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



Table 1 — Ten-day intervals and corresponding standard- 
ization factors for seasonal variation of mean Marbled 
Murrelet levels at three sites in northern California 



Interval 


Standardization factor 


April 15 to April 24 


0.86 


April 25 to May 4 


0.51 


May 5 to May 14 


0.82 


May 15 to May 24 


1.01 


May 25 to June 3 


0.95 


June 4 to June 13 


0.77 


June 14 to June 23 


0.68 


June 24 to July 3 


1.04 


July 4 to July 13 


1.22 


July 14 to July 23 


1.59 


July 24 to August 2 


1.04 


August 3 to August 12 


1.03 



Thus for surveys conducted at the three sites from 14 
July to 23 July, numbers of detections per survey were on 
average 1 .59 times greater than the summer mean; surveys 
conducted from 15 May through 24 May had numbers of 
detections which were about equivalent to the summer mean; 
and numbers of detections for surveys from 25 April to 4 
May averaged about half of the summer mean. 

In applying the standardization, we made the assumption 
that the relationship between detections at any site on a 
given day and the mean detection levels for the summer 
period at that site would be the same as the relationship we 
found at the three test sites. We have compared data with 
one site with very low activity and found the seasonal curves 
were similar. Standardized mean detection levels were 
calculated for all stands and stations and this measure used 
for all analyses. 

Stand Study: Isolated Stands 

Multiple Regression 

We examined the relationship between standardized mean 
detection levels for the stand, referred to as the dependent 
variable, and the following independent variables: stand 
size, Patton's index of perimeter to area (Patton 1975) which 
was used as a measure of the edge or shape, distance from 
salt water, density of old-growth trees, type of stand, and 
dominant tree species. As a transformation of the standardized 
mean detection level, we used the square root of the mean 
for the multiple regression. 

Logistic Regression 

For each stand we summarized the detections and 
behaviors for all surveys conducted during the study to 
determine the status of the stand. If no murrelets were detected 



during any of the surveys, then the status was "Undetected." 
Stands with murrelet detections were assigned a status of 
"Present" or, if occupied behaviors (Paton, this volume; 
Ralph and others 1 993) were observed, a status of "Occupied." 
Using logistic regression (SAS Institute, Inc. 1991) 
with maximum likelihood analysis of variance, we examined 
the relationship between a selection of independent variables, 
and status. We compared response variables Present 
(including Occupied stands) and Undetected, and response 
variables Occupied and Unoccupied (all stands with a status 
of Undetected or Present). For the stands with murrelets 
present we compared Occupied stands, with stands with a 
status of Present. 

Park Study: Large Contiguous Stands 

Elevation and position on the landscape were estimated 
from topographic maps to give a measure of topography for 
each station. Landscape position was described as one of 
five categories: (1) in the bottom of a major drainage, a 
drainage covering a large length of the landscape and isolated 
by parallel ridges; (2) in the bottom of a tributary (or minor) 
drainage, a drainage flowing into a major drainage, or a 
short, steep drainage flowing directly into the ocean; (3) on 
top of a major ridge, a ridge running parallel to a major 
drainage; (4) on top of a minor ridge, a ridge line that 
originated from the major ridge and was generally 
perpendicular to a major drainage; and (5) on a general 
slope, a station not on a ridge nor in a drainage. 

When stations were located on slopes or ridges, it was 
possible to detect murrelets calling in the drainages. The 
topography within 100 m of the stations was similar to the 
topography at the station itself. To help isolate the effects 
of topography, we included only detections within 100 m 
of the observer. 

Results 

Stand Study: Isolated Stands 

We identified 286 potential study stands in Del Norte, 
Humboldt, Trinity, San Mateo, and Santa Cruz counties 
meeting the criteria in the four size categories 2 to 20 ha (n = 
184); 21 to 40 ha (n = 39); 41 to 100 ha (n = 35); >100 ha (n 
= 28). We located few stands >21 ha, therefore, we surveyed 
all accessible stands in those categories. From these potential 
study stands we selected and surveyed 152 stands as follows: 
2 to 20 ha (n = 86); 21 to 40 ha (n = 22); 41 to 100 ha (n = 
23); >100 ha (n = 21). Due to weather conditions, three 
stands were surveyed only three times. 

Density of the combined dominant and codominant tree 
cover and presence of redwood trees were positively and 
significantly (F 005 = 2.428, df model = 10, P = 0.0105, R 2 = 
0.1625) related to mean murrelet detection levels in the 
multiple regression model. Because only 16 percent of the 
variation in the system was explained by the model, the 
predictive ability was limited. Other variables examined 
were not related to mean detection levels. 



210 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



The logistic regression model included density of old- 
growth (dominant and co-dominant) tree cover, tree species, 
and stand size as variables explaining the differences between 
sites with no detections and those with murrelets present 
(table 2). Stands with higher density classifications, and 
with redwood as the dominant tree species, were more likely 
to have murrelets present. Results also indicated a very 
minor effect of smaller stands increasing the likelihood of 
murrelet presence. We found, however, no significant effect 
of stand size on the status of murrelets in the stands 
(Undetected, Present, or Occupied), when tested by Chi- 
square contingency table (df = 6, % 2 = 3294, P = 0.7721) 
(table 4). Using these variables accounts for virtually all of 
the variability in the model. 

For stands with a status of Occupied (n = 37), compared 
with all Unoccupied stands (n =115), old-growth tree density 
and tree species were significant variables (table 3) for 
predicting observations of occupied behaviors. Stands in 
higher density classes with redwood as the dominant species 
were more likely to be classified as Occupied. 

Among stands with murrelet detections (n = 62), we 
found no differences in habitat variables between stands 
with a status of Occupied (n = 37) and Present (n = 25). 

Park Study: Large Contiguous Stands 

Central California 

Big Basin Redwoods State Park was surveyed in a matrix 
of 37 survey stations. The elevation ranged from 240-500 m 
and we divided stations into four equal categories (table 5). 
We found the mean detection levels and the number of 
Occupied stations higher for stations in lower elevation 
categories. The proportion of Occupied stations was not 
significantly different (P > 0.05) among topography categories 
(table 5). Occupied behaviors were observed in all topography 
categories, and the only station with a status of Undetected 
was on a major ridge. 



Table 2 — Results of logistic regression analysis for stands in California • n = 
152) with a status of murrelets Present (Present and Occupied) (n = 62) and 
Undetected fn = 90). Only variables with significant contribution to the 
model are presented 



Variable 



Tree species' 
Cover density 2 
Stand size 



Regression 
coefficient 



Chi-square 



Chi-square 
probability 



1.8101 
0.8755 
-0.0206 



9.43 
5.76 
5.45 



0.0021 
0.0164 
0.0195 



'Coast redwood (Sequoia sempervirens) or Douglas-fir (Pseudotsuga 
menziesii) >50 percent of stand. 

2 Percent dominant and codominant tree cover. 



Table 3 — Results of logistic regression analysis for stands in California 
(n m 152) with status of Occupied (n=37) and stands with murrelets Present 
or Undetected (L'noccupiedtfn = 115). Only variables with significant 
contribution to the model are presented 



Variable 



Tree species' 
Cover density 2 



Regression 
coefficient 



Chi-square 



Chi-square 
probability 



1.9243 
1.0831 



5.86 
6.64 



0.0155 
0.0100 



'Coast redwood (Sequoia sempervirens) or Douglas-fir (Psuedotsuga 
menziesii) >50% of stand. 

2 Percent dominant and codominant tree cover. 



Northern California 

We surveyed 352 stations in the 8 stands within northern 
California parks. We found that topography had a major 
influence on murrelet use (P < 0.0001). The mean detection 
levels were three times higher in major drainages (table 6) 
than on the major ridges. 



Table 4 — Percent of stands by murrelet use or status in each size category of stands surveyed in California for 
the Stand Study. Stands with a designation of Present had murrelet detections, but no observations of below 
canopy, or Occupied behaviors 





(ha) 


n 






Percent of stands by 


murrelet use (status) 






Not detected 




Present 


Occupied 


Stand size 


n 


Percent 


n 




Percent 


n 


Percent 


2-20 




86 


55 


63.9 


14 




16.3 


17 


19.8 


21-40 




22 


12 


54.6 


3 




13.6 


7 


19.8 


41-100 




23 


12 


522 


5 




21.7 


6 


26.1 


>100 




21 


11 


52.4 


3 




14J 


7 


33.3 


Totals 




152 


90 


59.2 


25 




16.4 


37 


24.3 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



211 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



Table 5 — For central California: Summary of detections' and status for Marbled Murrelet stations surveyed in old-growth forests within state and national 
parks during the summer, 1993 





Mean 
number of 








Number of stations (n) 














Landscape variable 


detections 2 


s.d. 


Range 


Occupied 


Present 


Absent 


Total 


Topography 
















Tributary drainage 


55 


42 


30-104 


3 








3 


Major drainage 


74 


53 


1-177 


10 


3 





13 


General slope 


58 


31 


1-97 


7 


1 





8 


Minor ridge 


34 


31 


1-83 


5 


2 





7 


Major ridge 


11 


14 


0-37 


3 


2 


1 


6 


Elevation 
















240-305 m 


70 


53 


1-177 


10 


2 





12 


306-360 m 


64 


36 


13-122 


10 


1 





11 


361-420 m 


35 


31 


1-946 


4 





10 


10 


421-500 m 


4 


6 


0-122 


1 


1 


4 


4 



'Includes only detections within 100 meters of observer 
Standardized detections 



Table 6 — For northern California: Summary of detections 1 and status of Marbled Murrelet stations surveyed in old-growth forests within the state and national 
parks during the summer, 1993 





Mean 
number of 








Number of stations (n) 














Landscape variable 


detections 2 


s.d. 


Range 


Occupied 


Present 


Absent 


Total 


Topography 
















Tributary drainage 


22 


33 


0-134 


18 


19 


54 


91 


Major drainage 


30 


28 


0-160 


67 


25 


17 


109 


General slope 


14 


17 


0-83 


40 


67 


22 


129 


Minor ridge 


16 


19 


0-107 


19 


29 


18 


66 


Major ridge 


10 


13 


0-51 


14 


27 


6 


47 


Elevation 
















21-100 m 


28 


30 


0-160 


83 


53 


27 


163 


101-200 m 


16 


18 


0-83 


46 


66 


36 


148 


201-300 m 


12 


13 


0-56 


19 


37 


19 


75 


301-500 m 


4 


6 


0-22 


10 


11 


18 


39 



'Includes only detections within 100 meters of observer 
Standardized detections 



The proportion of Occupied stations was significantly 
higher at stations of less than 100-m elevation than at stations 
>200 m (P < 0.0001) (table 6). The proportion of stations 
with no detections was significantly higher in the >300 m 
category and significantly lower in the <100 m category. 

Inland Range 

We found highest frequencies of presence (89.05 percent) 
and occupancy (21.91 percent) at stands and stations within 
10 km of the coast (table 7). The proportion of Occupied 
sites decreased in the 10- to 20-km band. The number of 
stations with detections declined by more than 99 percent 
from the 30- to 40-km to the 40- to 50-km band, although 



four times the number of stations were surveyed in the 40- to 
50-km band. The proportion of Occupied stations declined 
rapidly beyond 30 km from the coast. 

Discussion 

Stand Study 

The most important factor in indicating Occupied stands 
was density of the old-growth cover, that is, the percent of 
the area covered by the crowns of old-growth trees. Occupied 
stands had a higher percentage of old-growth cover than 
stands with murrelets only present, or in stands with no 
detections. These relationships are consistent with those 



212 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



Table 7 — Marbled Murrelet use of forest stands in northern California. 
\umbers represent individual stands for isolated stands surveyed surveyed 
four times during the Stand Study or stations for surveys conducted in each 
12.5 ha of a large contiguous stand for the Park Study or in preparation for 
timber harvest 



Distance 


Number of 
stations 




Number of stations by use 


band km 










from coast 


surveyed 


Detected 1 


Percent 


Occupied 


Percent 


0-10 


283 


252 


89.05 


62 


21.91 


10-20 


133 


38 


28.57 


6 


4.51 


20-30 


144 


52 


36.11 


24 


16.67 


30-40 


100 


36 


36.00 


6 


6.00 


40-50 


428 


1 


0.23 


1 


0.23 


50-60 


95 


2 


ill 





0.00 


Totals 


1183 


379 


32.04 


98 


8.28 



1 All stations or stands with murrelet detections, including occupied behaviors 



found in Oregon (Grenier and Nelson, this volume) and 
Washington (Hamer. this volume). 

We found the presence of redwood as the dominant tree 
species to be a factor for predicting higher mean detection 
levels and stand occupancy. In Washington. Hamer and 
others (1993) also found tree species composition to be an 
important factor for murrelet occupancy. Within the range of 
our study, stands dominated by Douglas-fir often were in 
drier areas with higher summer temperatures. Sites very 
close to the coast are usually dominated by Douglas-fir and 
Sitka spruce (Picea sitchensis) and, for unknown reasons, 
also lack murrelets. 

Contrary to previous studies we did not find larger 
stands more likely to have murrelets present or to be occupied. 
Other factors, such as, stand history and juxtaposition to 
other old-growth stands may mask the effects, if any, of 
stand size on murrelet presence and use. 

Although in the Stand Study we did not find a significant 
relationship between distance from the ocean and murrelet 
detections or behaviors, this possibly was related to the limited 
range of distances for stands surveyed. Our examination of 
all surveys from 1988 through 1992, however, indicates a 
strong partem of declining murrelet presence with distance 
from the coast (table 7). The number of stations more than 40 
km inland with murrelet detections was only about 2 percent. 
One factor which may have biased the bands >40 km inland 
was the selection of the survey sites. Many of these sites 
were located in forest habitat selected for timber planning 
and not considered optimal for murrelets. A lack of murrelet 
detections would then allow timber harvesting on some of 
these lands. Further studies inland in California at sites selected 
by unbiased methods would provide needed information on 
the murrelet' s distribution in these areas. 

It is unlikely that one factor alone will best describe 
murrelet habitat. Density of old-growth cover and species 
composition are included as important factors in more than 



one analysis. These variables may be the strongest predictors 
of murrelet presence in California. 

Large Contiguous Stands 

Within the large stands of old-growth in the parks, most 
stations with observations of occupied behaviors occurred in 
the major drainages and, correspondingly, at low elevations. 
Occupied behaviors were observed at 69 (73 percent) of the 95 
stations in the major drainages. Trees in these drainages tend 
to be larger, and experience less limb breakage from wind 
(Tangen, pers. comm.). Both of these factors could contribute 
to larger diameter branches and more potential nest platforms. 

Acknowledgments 

This research was a cooperative effort with private 
landowners in California who formed the Marbled Murrelet 
Study Trust including: The Pacific Lumber Company, Areata 
Redwood Company, Miller-Rellim Company, Big Creek 
Lumber Company, Simpson Timber Company, Sierra-Pacific 
Industries, Barnum Timber Company, Eel River Sawmills, 
Louisiana-Pacific Corporation and Schmidbauer Lumber 
Company. We especially would like to thank Lloyd Tangen 
and Lee Folliard who gave frequent help and encouragement. 
Jim Brown, Sal Chinnici, Joe Dorman, Steve Kerns, Ray 
Miller, and Rob Rutland also gave of their assistance and 
insight. We appreciate the guidance of members of the 
project's advisory committee, which included members from 
all cooperators and agencies, and the U.S. Fish and Wildlife 
Service, especially Mike Horton. Financial support for the 
study was provided by the Marbled Murrelet Study Trust of 
the timber companies. USDA Forest Service. California 
Department of Fish and Game, California Department of 
Forestry, California Department of Parks and Recreation, 
and Redwood National Park. We thank Jennifer Weeks, 
Brian O'Donnell, Deborah Kirstan, and Ann Buell for their 
assistance with data preparation, analysis and manuscript 
reviews and James Baldwin for his statistical advice. We 
appreciate also the reviews provided Martin Raphael, Alan 
Burger, Frank Thompson, Mark Huff, Marty Berbach, Valerie 
Elliot, Sal Chinnici, Lee Folliard, Todd Sloat, and Steve 
Kerns. We appreciate the efforts of the many field personnel 
who gathered data for the study. 

Appendix A 

Designing a study to examine the relationship of Marbled 
Murrelets with forest habitats requires first determining if 
the birds are present or absent from individual forest stands. 
Here, we outline the methods used to determine the appropriate 
number of surveys required when the objective is to determine 
murrelet presence or absence. 

For our study, we wished to know how many survey 
mornings were necessary to determine presence in a stand of 
murrelets with a 95 percent probability of being correct. We, 
therefore, set the level of probability of a false negative at 5 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



213 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



percent. That is, murrelets are present, but we accept a 5 
percent probability that they are not detected. Data from 
previous surveys have been used in the discussion below 
(table 1). From the data provided by Rob Hewlett, Steve 
Kerns, Kim Nelson, and our studies, we determined the 
number of survey mornings needed to meet this level of 
confidence at sites having various levels of detection rates. 

In the following example, we assumed murrelets are 
present in the relatively homogeneous stand of old-growth 
timber to be surveyed. Each survey consists of one person 
observing from a station for one morning. 

The method for examining our data was: 

P=\-(\-pY 
where — 

P is the probability of at least one detection, 

p is the proportion of surveys with at least one detection, 
that is, the number of surveys with at least one detection, 
divided by the number of surveys, and 

n is the number of surveys required to detect at least 
one bird. 

To determine the number of surveys needed if we want 
to be 95 percent certain (P = 0.95) we are not missing birds 
which are present, we solve for n: 



Table 1 — Detection rate at stations with low rates, and the percent of surveys 
with detections 



ln(l-P) 



n> 



ln(l-p) 
where — 

In is the natural log. 

We tested our survey sample size from 19 sites (table 1) 
with relatively low average detection rates and a minimum 
of seven survey mornings. The mean detection rate per 
morning was divided into four categories, 0.4 to 2.5, 2.6 to 
5.0, 5.1 to 7.5, and 9.4 to 16.6 detections. We used the 
average percent of surveys with detections within each 
category to estimate p. 

In the 0.4 to 2.5 category, the percent of survey mornings 
with detections varied from 13 percent to 75 percent, with an 
average of 48 percent of the mornings with detections. The 
calculation is as follows: 

In (1-0.95) 

n > = 4.58, or 5 surveys. 

In (1-0.48) 

In the 2.6 to 5.0 detection range, the percent of surveys 
with detections varied over a smaller range, from 63 percent 
to 91 percent, an average of 81 percent. Using the average 
number, the calculation is: 



n > 



In (1-0.95) 



In (1-0.81) 



= 1.80, or 2 surveys. 



Station name 


Number of 


Mean 


Percent of 




surveys 


detection 


surveys with 






rate 


detections 


SiteF 


8 


0.4 


13 


ALCR6 


8 


1.0 


75 


FRNO 


7 


1.3 


57 


SiteE 


8 


2.1 


25 


ALCR3 


8 


2.5 


75 


ALCR9 


8 


2.6 


63 


ALCR4 


8 


3.0 


88 


ALCR1 


8 


3.1 


75 


FRSO 


11 


4.7 


91 


PATM 


8 


5.0 


88 


ALCR10 


8 


5.1 


75 


ALCR 12 


8 


5.1 


88 


ALCR13 


8 


5.6 


88 


KLMO 


11 


6.2 


65 


SFYA 


13 


6.5 


77 


EHSP 10 


8 


7.5 


75 


ALCR 11 


8 


9.4 


75 


ALCR 8 


8 


13.0 


88 


CUPE 


13 


16.6 


92 



In the 5.1 to 7.5 detection range, the percent of surveys 
with detections varied from 65 percent to 88 percent, an 



average of 78 percent. The calculation as above was 1.98 or 
a minimum of 2 surveys. 

The highest detection range used for this calculation 
was 9.4 to 16.6 birds per morning, an average of 85 percent 
of survey mornings with at least 1 detection. The calculation 
resulted in 1.75, or 2 surveys. 

From these data we can conclude that in areas with 
mean detection rates as low as 0.4 to 2.5 per survey (and 
presumably low occupancy rates as well), a minimum of 
five survey mornings will detect birds if they are present, 
with a 95 percent probability. In areas of detection rates 
from 9.4 to 16.6, the number of surveys necessary to 
prevent a false negative is about two. Using this formula, 4 
surveys would be required to detect birds in areas with a 
mean of 1.0 to 2.5 detections per survey. We can then 
conclude that a suggested survey rate of four surveys per 
stand, will detect birds in excess of 95 percent of the time, 
and will likely detect all but the smallest populations 99 
percent of the time. 

Assumptions 

There are several assumptions we have made in using 
these methods. We list them below and discuss each. 

We assume that the amount of canopy cover at a station 
will have no effect on detection probability (P). 



214 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Miller and Ralph 



Chapter 20 



Inland Habitat Relationships in California 



In most forests, the majority of detections are audio and 
are not affected by canopy cover. Though the number of 
visual detections decreases with increased canopy cover, 
there should be a compensating effect as we have found 
higher numbers of total detections (e.g., Paton and Ralph 
1990) as forest age and canopy cover increase. 

In calculating P, the probability of at least one detection 
in a stand, we assume that murrelets are present in the stand 
when the sun'ey is conducted. 

The effects of this assumption are discussed in detail 
in Azuma and others (1990), and the situation with the 
murrelet is similar. Since there is some probability that 
murrelets will be present in a stand and not be detected, the 
result would be an underestimate of the number of stands 
with murrelets present. Following data collection, bias 
adjustments presented in Azuma and others (1990) could 
be used to estimate the number of stands with murrelets in 
each stand category. 



We assume that P is constant and independent of stand 
size and habitat type. 

It is possible that as stand size increases and habitat 
matures, the number of birds in a stand will increase. 
Increased numbers will likely increase P as individuals 
may call in response to other birds as a result of social 
facilitation. Therefore, stands with few birds will have 
fewer detections than stands with many birds. We will be 
examining this assumption, and it forms the basis of the 
null hypothesis that stand size and habitat type have no 
effect on detection rate. 

Frequency of surveys 

If the habitat is homogeneous and we assume that 
the birds are distributed essentially evenly throughout the 
stand, the stations can be positioned throughout the stand 
and all stands, regardless of size, would be surveyed four 
survey mornings. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



215 






IV 



The Marine Environment 






— 



Chapter 21 

Oceanographic Processes and Marine Productivity in Waters 
Offshore of Marbled Murrelet Breeding Habitat 



George L. Hunt, Jr. 1 

Abstract: Marbled Murrelets (Brachyramphus marmoratus) oc- 
cupy nearshore waters in the eastern North Pacific Ocean from 
central California to the Aleutian Islands. The offshore marine 
ecology of these waters is dominated by a series of currents roughly 
parallel to the coast that determine marine productivity of shelf 
waters by influencing the rate of nutrient flux to the euphoric zone. 
Immediately adjacent to the exposed outer coasts, wind driven 
Ekman transport and upwelling in the vicinity of promontories and 
other features create zones of enhanced primary production in 
which primary and secondary consumers may aggregate. In the 
more protected waters of the sounds, bays and inlets of British 
Columbia and Alaska, tidal processes dominate the physical mecha- 
nisms responsible for small-scale variation in primary production 
and prey aggregations. 



In North America, Marbled Murrelets (Brachyramphus 
marmoratus) occupy coastal marine waters from central 
California to the Aleutian Islands of Alaska. To understand 
factors controlling marine resources in the habitats occupied 
by Marbled Murrelets, it is useful to review the coastal 
oceanography of the region between California and Alaska. 
For the purposes of this review. I focus on three types of 
habitat: shelf waters, influenced primarily by the major long- 
shore current systems: inshore waters of the open coasts; and 
the relatively sheltered waters of sounds, inlets and bays. 

This chapter provides an overview for the non-marine 
specialist of the types of habitats, and the processes that 
determine the distribution and abundance of marine resources 
used by Marbled Murrelets. 

Determinants of the Shelf Circulation 

The major offshore currents off the west coast of northern 
North America originate as eastward flowing currents crossing 
the North Pacific Ocean. One of these, the North Pacific 
Current, divides into two branches west of the continental 
shelf off the British Columbia coast (Reed and Schumacher 
1987. Thomson 1981). The northern branch curves northeast 
as the Alaska Current, and forms a counterclockwise rotating 
gyre in the Gulf of Alaska (fig. 1). The second branch of the 
North Pacific Current turns southeast as the California Current 
and flows along the edge of the continental slope off 
Washington. Oregon and California. The division of the 
North Pacific Current is seasonally variable; it is most abrupt 
in winter, and most diffuse and spatially variable in summer 
(Thomson 1981). 



1 Professor, Department of Ecology and Evolutionary Biology, Uni- 
versity of California. Irvine. CA 92717 



The Alaska Current is relatively wide (400 km) and 
slow (30 cm/s) as it moves through the eastern Gulf of 
Alaska (Reed and Schumacher 1987). As the Alaska Current 
passes Kayak Island in the northern Gulf of Alaska, it forms 
a strong (>50 cm/s), clockwise rotating gyre in the island's 
lee (Royer and others 1979). A branch of the Alaska Current 
the Alaska Coastal Current, diverges from the gyre and 
approaches the Kenai Peninsula coast (fig. 7). In fall, the 
Alaska Coastal Current shows a marked increase in velocity, 
apparently as a result of both increased freshwater runoff 
and easterly winds that constrain the current in a narrow 
coastal stream and produce coastal convergence (movement 
of water toward the coast, with attendant downwelling) (Royer 
1979, 1983; Schumacher and Reed 1980). Much of this flow 
passes through Kennedy Entrance, south of the Kenai 
Peninsula, and thence into either Cook Inlet or westward 
into Shelikof Strait between Kodiak Island and the Alaska 
Peninsula. The main Alaska Current exits the Alaska Gyre 
to the west as the Alaska Stream, flowing along the Alaska 
Peninsula and the south side of the Aleutian Islands. West of 
Kodiak Island, it becomes narrow (100 km) and swift (-100 
cm/s) (Reed and Schumacher 1987). Although these currents 
are for the most part seaward of the distribution of Marbled 
Murrelets in the Gulf of Alaska (Piatt and Ford 1993), the 
currents are important to marbled murrelets because they 
influence the transport of plankton into coastal waters and 
also because they can play an important role in the transport 
of oil slicks when spills occur (Piatt and others 1990). 

The California Current varies in its intensity, definition, 
and direction of flow geographically and seasonally (fig. 2) 
(Mooers and Robinson 1984; Thomson 1981 ). It is relatively 
weak off the Washington and Oregon coasts, where it has a 
southward flow only 20 percent of the time. In contrast, off 
California, the current is usually well defined and flows 
southward about 50 percent of each month. The California 
Current is most often southward and strongest between March 
and September. 

Changes in the direction and intensity of flow of the 
California Current have important effects on offshore marine 
production (Chelton 1981. Chelton and others 1982). When 
the current moves strongly southward, water throughout the 
water column moves away from the coast (offshore transport) 
due to the Coriolis Effect. In addition, offshore transport of 
surface water, also related to the Coriolis Effect (Ekman 
transport), results when north and northwest winds force 
increased surface flow to the south. Water transported offshore 
is replaced by the upwelling of deep, cold, nutrient rich 
water that supports enhanced productivity. These seasonal 
and interannual fluctuations in the California Current system 
and its productivity have been linked to changes in the 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



219 



Hunt 



Chapter 21 



Oceanographic Processes and Marine Productivity 




170 160 150 140 

Figure 1— Major features of ocean circulation in the Gulf of Alaska. From Reed and Schumacher (1987), by permission. 



breeding success of seabirds (Ainley and Boekelheide 1990, 
Ainley and others, in press) and in the numbers and distribution 
of seabirds at sea (Briggs and others 1987). 

Inshore of the California Current, the Davidson current 
flows northward seasonally from about 32° to about 50° N 
(fig. 2). The onset of the Davidson Current usually occurs 
in October, when the overall average movement of water in 
the California Current system shifts toward the north until 
March (Thomson 1981). When the northward flowing 
Davidson Current prevails, upwelling is suppressed because 
northward flowing water is deflected by the Coriolis Effect 
toward the shore and downwelling is likely to prevail 
(McLain and others 1985). The seasonal shifts in the flow 
of the California Current system are largely the result of 
changes in the direction of the prevailing winds. In spring 
and summer, the winds blow from the northwest and move 
the surface water southward, whereas in winter, prevailing 
winds are from the southwest and surface water movements 
are to the north. 

Off Vancouver Island, a northwestward coastal current 
flows inshore of the southeastward flowing southern branch 
of the North Pacific Current (Thomson 1981). This inshore 
current originates in the outflow of the Strait of Juan de Fuca 
and is confined in summer to within 15-20 km of the coast. 
The speed of the coastal current is determined by the velocity 
of the winds. In winter, the coastal flow merges with that of 
the Davidson Current. 

Strong El Nino-Southern Oscillation events cause a 
reversal of flow in the California Current System, the presence 



of a surface layer of warm, nutrient-depleted water, and the 
replacement of coastal upwelling with downwelling (Johnson 
and O'Brien 1990; Norton and others 1985; Rienecker and 
Mooers 1986). A consequence of these events is a marked 
reduction in primary production, followed by a reduction in 
zooplankton populations and reduced survival of at least 
some larval fish (Barber and Chavez 1984, MacCall 1986, 
Pearcy and Schoener 1987). These events result in a marked 
decrease in seabird reproductive success and in striking 
changes in the offshore distribution and abundance of seabirds 
(Ainley and Boekelheide 1990; Ainley and others, in press; 
Briggs and others 1987). 

Inshore Waters of the Open Coasts 

Large oceanic currents determine regional marine habitat 
types and are responsible for a major portion of the seasonal 
variation in production on the shelf. However, marine waters 
within a few kilometers of the shore are where Marbled 
Murrelets spend most of their time. In these areas, currents 
interacting with bathymetry can create fronts (boundaries 
between water masses where convergences or upwelling 
may occur) and upwellings that either enhance productivity, 
or cause organisms to accumulate because of behavioral 
responses to physical gradients. For example, upwelling 
results when a current passes a promontory and draws away 
surface water that is then replaced by water from depth 
(Pingree and others 1978; Thomson 1981). Fronts associated 
with these processes provide foraging sites for seabirds. 



220 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Hunt 



Chapter 21 



Oceanographic Processes and Marine Productivity 




B 




Figure 2 — Schematic of the circulation of the California Current in (a) February and 
(b) August. From Ingmanson and Wallace (1989), by permission. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



221 



Hunt 



Chapter 21 



Oceanographic Processes and Marine Productivity 



In coastal waters, strong winds cause upwelling by two 
mechanisms. In the first, water is displaced from near the 
coast by winds blowing parallel to the coast from the north. 
In the northern hemisphere, if water depths are sufficient, 
surface waters will move at approximately 90 degrees to the 
right of the direction of the surface wind because of the 
Coriolis Effect (Ekman transport). When this occurs near the 
coast, the displaced water is replaced by nutrient rich water 
from depth. In the second, winds blowing from the shore 
cause inshore upwelling. If the water is sufficiently shallow, 
the Ekman transport is effectively canceled by friction with 
the bottom, and surface flows will be in the same direction 
as the wind. When strong land breezes blow surface water 
away from a lee shore, inshore surface waters are replaced 
with water from greater depth. 

Along the open coasts of California, Oregon, and 
Washington, localized nearshore upwelling due to Ekman 
transport and offshore winds blowing water away from lee 
shores provides regions of enhanced primary and secondary 
production. These upwelling processes, and fronts associated 
with river discharges are expected to be the most important 
physical features in determining murrelet foraging 
opportunities. Ainley and others (this volume) provide one of 
the only examples of the sort of mesoscale studies needed to 
link murrelet foraging distributions to physical and biological 
processes that result in exploitable concentrations of prey. 

Sheltered Waters of Sounds, Inlets 
and Bays 

The physical and chemical oceanographic processes 
controlling primary production in the fjords and estuaries of 
the Gulf of Alaska and the British Columbia coasts are 
reviewed by Burrell (1987) and Reeburgh and Kipphut (1987). 
In these fjords, freshwater input, primary production, and 
other biogeochemical processes are highly seasonal. 
Freshwater, less dense than saltwater, forms a surface layer 
in the fjords and is discharged from these upper layers into 
the Gulf of Alaska; waters from the Gulf of Alaska circulation 
episodically penetrate the fjords to replace intermediate and 
deep resident waters (Burrell 1987). These exchanges 
influence the availability of nutrients to, and the residence 
time of, phytoplankton. Both factors also affect the timing 
and magnitude of primary production in the fjords. Coastal 
frontal zones associated with shallow areas with increased 
water flow can be the site of elevated primary production 
because of enhanced vertical flux of nutrients (Parsons and 
others 1983, 1984). High tidal ranges present in British 
Columbia and along the coast of the Gulf of Alaska would 
promote these enhanced vertical fluxes in the vicinity of sills 
at the mouths of fjords (Burrell 1987). In late summer and 
early fall, turbidity from river inflows may progressively 
limit primary production in the upper ends of fjords (e.g., 
Goering and others 1973). 



Fjords may support one of two generalized trophic 
pathways (Burrell 1987, Matthews and Heindal 1980). In 
shallow fjords or those with shallow sills, the pathway may 
lead from small phytoplankton to small copepods to jellyfish. 
In deeper fjords, and fjords with deep sills, the trophic 
pathway may include large net phytoplankton (primarily 
diatoms), large copepods and finfish. Apparently, the depth 
of the sill is a critical feature; if it intercepts the pycnocline 
(the layer in which water density changes rapidly with 
depth, and which inhibits vertical mixng of water), the 
upper layer of out-flowing fresh water inhibits the recruitment 
of large calanoid copepods from outside the fjord. The 
ontogenetic migration to the upper water column of 
Neocalanus plumchrus and related oceanic copepod species 
in the North Pacific occurs at the same time as the coastal 
convergence mentioned above (Burrell 1987). Their presence 
in the upper water column at this time allows them to be 
transported into adjacent fjord environments according to 
observations by R. T. Cooney, as cited by Burrell (1987). 
These large copepods are likely to be important prey for the 
small fish taken by Marbled Murrelets. One might 
hypothesize, then, that murrelets would be more likely to 
forage in fjords supporting populations of large copepods 
than in fjords lacking this component of trophic transfer. 
Additionally, one might expect that Marbled Murrelets 
would be more likely to forage at the seaward ends and near 
the sills of these fjords, rather than at their inner ends. 

In the inland waters of the sounds and channels of 
Washington, British Columbia, and Alaska, tidal processes 
are likely the most important determinants of localized 
foraging opportunities for marbled murrelets and other 
seabirds. Upwellings can be caused by currents impinging 
on an obstruction and being driven to the surface, such as 
when strong tidal currents encounter a sill and flow over it. 
In these circumstances planktonic organisms are driven to 
the surface (Brown and Gaskin 1988; Vermeer and others 
1987), or may be concentrated at depth where their ability to 
swim against a gradient is matched by an opposing current 
(e.g., Coyle and others 1992). 

Superimposed on the physical mechanisms that enhance 
primary production and concentrate prey are the seasonal 
variations in production and the movements of prey of suitable 
size. We know of few studies of physical processes and fish 
movements at temporal or spatial scales appropriate for 
understanding murrelet foraging, and none for which murrelets 
were a focus of the study. This paucity of data makes it 
difficult to assess, in oceanographic terms, the characteristics 
of habitats critical for foraging murrelets. 

Acknowledgments 

I thank Dan Anderson, Larry Spear and C. John Ralph 
for helpful comments on an earlier version of this manuscript. 



222 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 22 

Marbled Murrelet Food Habits and Prey Ecology 



Esther E. Burkett 1 

Abstract: Information on food habits of the Marbled Murrelet 
(Brachyramphus marmoratus) was compiled from systematic stud- 
ies and anecdotal reports from Alaska to California. Major differ- 
ences between the winter and summer diets were apparent, with 
euphausiids and mysids becoming more dominant during winter and 
spring. The primary invertebrate prey items were euphausiids. mysids, 
and amphipods. Small schooling fishes included sand lance, an- 
chovy, herring, osmerids. and seaperch. The fish portion of the diet 
was most important in the summer and coincided with the nestling 
and fledgling period. Murrelets are opportunistic feeders, and 
interannual changes in the marine environment can result in major 
changes in prey consumption. Site-specific conditions also influ- 
ence the spectrum and quantity of prey items. More information on 
food habits south of British Columbia is needed. Studies on the 
major prey species of the murrelet and relationships between other 
seabirds and these prey are briefly summarized. Short-term phe- 
nomena such as El Nino events would not be expected to adversely 
affect murrelet populations over the long term. However, cumula- 
tive impacts in localized areas, especially in conjunction with El 
Nino events, could cause population declines and even extirpation. 



An understanding of Marbled Murrelet (Brachyramphus 
marmoratus) food habits is needed for effective conservation 
of this threatened seabird. Many seabirds are known to be 
affected by prey availability, though human activities induce 
and compound impacts (Croxall 1987: 377-378; Furness 
and Monaghan 1987: 35^45, 98-99; Gaston and Brown 
1991; Jones and DeGange 1988; Tyler and others 1993). 
Ainley and Boekelheide (1990: 373-380) discuss the interplay 
of factors affecting seabird reproduction and total population 
size, especially as related to different marine systems. 

The dramatic loss of old-growth forest nesting habitat 
(Marshall 1988b) has resulted in a fragmented distribution of 
the murrelet at sea, especially during the breeding season 
(Carter and Erickson 1988, Piatt and Ford 1993). Proximity 
of nesting habitat to an oceanic prey base is important for 
energetic reasons (Cody 1973, Sealy 1975c, Carter and Sealy 
1990). but the bird's capabilities are not understood, and 
fluctuations in prey populations and variability in prey 
distribution have not been studied relative to murrelet nesting 
success or inland distribution. Nevertheless, much of the 
work on food habits conducted thus far is useful for 
management purposes and can be used to direct further research. 

Six systematic studies on food habits of the murrelet 
have been conducted in North America. Two occurred during 
the breeding season in British Columbia (Carter 1984, Sealy 
1975c) and one in the non-breeding season (Vermeer 1992). 



1 Associate Wildlife Biologist, California Department of Fish and 
Game, 1416 Ninth Street. Sacramento. CA 95814 



In Alaska, two studies have been conducted in the non- 
breeding season (Krasnow and Sanger 1982, Sanger 1987b), 
and one took place during the breeding season (Krasnow and 
Sanger 1982). These studies form the basis for much of the 
knowledge of murrelet food habits and are discussed below 
along with anecdotal information on murrelet diet. 

Recent genetic analysis has indicated that the North 
American Marbled Murrelet warrants full specific status 
(Friesen and others 1994a). For this reason, and since this 
chapter was written primarily to aid in management action 
and recovery planning in North America, information on the 
diet of the Long-billed Murrelet (Brachyramphus marmoratus 
perdix) has been omitted. 

Overall, murrelet food habits in the Gulf of Alaska and 
British Columbia have received the most attention. Very 
little information is available on food habits of murrelets in 
Washington, Oregon, or California, and systematic stomach 
analyses have never been conducted in these states. 

Methods 

Because so few studies with large sample sizes have 
been conducted and the geographic scope of the studies to 
date is limited, an attempt was made to assemble information 
on food habits from Alaska to California, even though 
many of the records are anecdotal or represent field studies 
with small sample sizes. In addition to a literature review, 
murrelet biologists from Alaska to California were contacted 
for information. 

An attempt was made to separate adult and nestling food 
items and to distinguish between foods used in the breeding 
and non-breeding seasons. However, in some cases the 
researcher's "winter" collection period continued into the early 
part of the breeding season (March and April), and the data 
were not analyzed separately. Also, at times the age class of 
the murrelet specimens was not stated in the literature. Even if 
such information were known, the small sample sizes, large 
geographic differences, and separation of time scales would 
confound the interpretation of results. Prior to this work, four 
summaries of murrelet diet were produced (Ainley and Sanger 
1979, Ewins and others 1993, Sanger 1983, Carter 1984). 

Results 

Systematic Studies of Food Habits 

Sealy (1975c) 

Sealy (1975c) was the first to systematically study 
murrelet feeding ecology, along with work on the diet of the 
Ancient Murrelet (Synthliboramphus antiquus) near Langara 
Island, British Columbia. Langara Island is part of the Queen 
Charlotte Islands and is approximately 500 kilometers 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



223 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



northwest of Vancouver Island. The study spanned two 
breeding seasons (1970 and 1971), and 86 adult and subadult 
Marbled Murrelets were collected between March 25 and 
August 10 (years combined). The diets were essentially 
the same for both sexes, and samples from subadults and 
adults were identical, so the data were pooled for a total 
sample of 75 individuals. Additionally, six newly fledged 
murrelets were taken between July 10 and August 4, 1 97 1 , 
and their food habits were analyzed separately. The 
percentage of murrelets collected that contained prey ranged 
from 87 to 100 percent. 

Sand lance (Ammodytes hexapterus) made up 67 percent 
of the food items in the diet of the adults and subadults. 
Euphausiids were the next most important food item and 
contributed 27 percent of the items. Two species of euphausiids 
were consumed, Euphausia pacifica and Thysanoessa 
spinifera, with relative importance values of 2 percent and 
25 percent, respectively. The next most important food item 
was the viviparous seaperch (Cymatogaster aggregata), with 
a value of 3 percent. Overall, sand lance, euphausiids, 
seaperch, scorpaenids, and osmerids made up 98 percent of 
the murrelet diet. Including the less common food items 
which occurred in very small amounts, at least nine different 
types of prey were identified (table 1). 

The six samples of newly fledged young selected different 
prey than adult/subadult murrelets (table 1). Sand lance still 
dominated the diet at 65 percent (similar to 67 percent for 
adult/subadult murrelets), but the seaperch was the next 
most important prey species, rather than euphausiids, with a 
value of 35 percent. The euphausiid, T. spinifera, and 
amphipods made up trace amounts of the remainder of the 
fledgling diet. 

The difference in adult and juvenile diets can be partially 
explained by looking at the difference in abundance of prey 
items taken by the adult/subadult murrelets over the course 
of a breeding season. The euphausiid, T. spinifera, was 
found more commonly in the adult/subadult diet during the 
mid-April to mid-May period and was more important than 
the sand lance at this time, but euphausiids diminished greatly 
in the diet after the early part of the breeding season. However, 
T. spinifera remained important in the diet of adult Ancient 
Murrelets through mid- July when the study concluded. Sealy 
attributed this difference in diet to the offshore movement of 
E. pacifica (affinity for deeper water than T. spinifera) and, 
to some extent, offshore movement of T. spinifera as the 
spring progressed and water temperature rose. He also 
attributed the diet change to reduced abundance of T. spinifera 
due to loss of females after reproduction. Additionally, he 
noted that adult Ancient Murrelets feed further offshore than 
Marbled Murrelets or juvenile Ancient Murrelets, and he 
believed the food supply of the Ancient Murrelet was spotty 
and unpredictable. 

Sealy tested for a measurable change in prey avail- 
ability mid-summer by examining the stomach contents 
of 13 individuals of seven species, including the Ancient 
and Marbled Murrelet, from six mixed-species feeding 



assemblages. Between 9 May and 26 June 1971 he conducted 
plankton hauls where collected birds had been foraging. The 
results indicated that only Thysanoessa was available and 
taken by those individuals examined in May, and later samples 
in June found only Ammodytes available and being consumed. 
He concluded that fishes such as Cymatogaster and 
Ammodytes tend to spend the winter and early spring in mid- 
water offshore, but migrate to the surface and move inshore 
in late spring, thus possibly becoming available to murrelets 
at this time. 

Plankton hauls made in 1971 also indicated that the 
murrelets were more selective in their feeding habits when 
compared to prey availability (Sealy 1975c). Organisms such 
as ctenophores, amphipods, and polychaetes were obtained 
in the plankton hauls, but none of these organisms were 
found in the food samples analyzed. Zooplankton sampling 
by Project NorPac (Dodimead 1956) during summer 1955 
(primarily in August) resulted in a similar difference in prey 
availability; copepods were by far the most numerous 
organisms with a total volume of more than 65 percent, 
while euphausiids composed less than 10 percent of the total 
volume (LeBrasseur 1956). 

Sealy (1975c) concluded that murrelets seldom feed 
more than 500 m from shore, usually in water less than 30 m 
deep. His work demonstrated that euphausiids made up only 
a small part of the overall diet during the breeding season, 
but were dominant during the early part of the breeding 
season. He thought the breeding season was possibly 
ultimately controlled by the cycles of abundance of fishes 
near shore, especially the sand lance, which were taken by 
the murrelet in great quantities in the study area. 

Krasnow and Sanger (1982) 

Krasnow and Sanger (1982) collected murrelets at sea 
in the vicinity of Kodiak Island in the winter of 1976/ 
1977. They collected 18 murrelets (all with food) between 
December 1976 and April 1977 at Chiniak Bay, a large 
bay on the northeast end of Kodiak Island; a second sample 
of 19 murrelets (16 with food) was collected from Chiniak 
in February 1978. Two other sites were sampled during 
the breeding season of 1978. At Izhut Bay, a small bay 
north of Chiniak Bay, Krasnow and Sanger collected 34 
murrelets (25 with food) between April and August 1978 
and from Northern Sitkalidak Strait, which is located on 
the southeast end of Kodiak, they collected 26 murrelets 
(17 with food) between May and August 1978. The 
percentage of murrelets collected which contained prey 
ranged from 65 to 100 percent. 

Krasnow and Sanger calculated an Index of Relative 
Importance (IRI) value for the foods consumed by murrelets 
according to Pinkas and others (1971). During the 1976/ 
1977 winter, fish, primarily of the family osmeridae, were 
the most important prey, followed by euphausiids of the 
genus Thysanoessa, and mysids (table 2). A total of 1 1 
different prey items were identified (table 2), compared to 
nine from Sealy's (1975c) breeding season study (table 1). 



224 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



% 



% X 



& 



"%>. 



X 

4^ 

\ 



fa 



^b 



€. 



V 



V 

ft 



IS 



V 



'i. 



** 



si 



% 



v\ 



/5> 



\ % 






^ 



^ 



V 



% 



^ 






X 



^ 



^rs 



& 



<« 









e 



%„ <V 



1 



\ 



% X ° 
X % 



i 



'* 



\ 



c* 



'a. 









% VI 



% 



■ J /, 



VV 

\\ 

X 
% 



\. 



< 



< 
u 



a: 

o 



< 



< 



< 



u 

03 



o 

03 



O 

03 



O 
03 



(J 

03 



03 



-J 
< 



H 
as 



Chapter 22 



Food Habits and Prey Ecology 



03 









O 03 
— Z 



< 2 



OS 
03 



0C 
09 



03 



OS 
03 



OS 
03 



OS 
03 



03 <N 



os 

03 



0! 

03 



03 



- 
03 



£ - 



03 

z 



a: 

03 



o - 

03 



-03 ^ 



O - 

03 r- 



03 



£ - 






03 r- 



Oi 

03 



g - 



03 
Z 



< 2 



_ <* 
03 



OS 
03 



1 

•3 

o o <— 
c * 



c/5 c/3 



s 

I 



a 

■ 



90 















































































































































































































































































































































































































































































— H •—■ 




- 




- 




o 




- 












































1 2 

li.il 

D 1 -2 


1 

- 3 




1 = 

11 




— 


Acanlhomysis sp. 
(Mysidae) 


I - 


1 ft 

E | 


3 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



225 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



a 
1 
a 



1 




**. \ 





H Ol. 




1 < 




1 < 


1 "*■ 




"I 










I < 




1 ■*■ 


0. 












* 1 




■ 








< 1 








° 1 




< 1 


1 < 


* 




* 1 


1 " 


























i ! i 

2F -— v Or ■— 

■* 8. §■ 8. 

I - J - 


ill 

1 1 1 
a -a 3 

3 uj 



226 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



Chapter 22 



Food Habits and Prey Ecology' 




USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



227 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



Table 2 — Comparison of winter diet of Marbled Murrelets in Chiniak Bay, 
Alaska, between December 1976-April 1977, and February 1978" 





Year 


Prey 


1976/1977 


1978 


Nereidae 


3" 





Chaetognatha 


1 





Mysidacea 


23 


447 


Acanthomysis sp. 


4 


10,548 


Neomysis sp. 





870 


N. rayii 


2 


27 


Thysanoessa sp. 


74 





T. inermis 


1,169 





T. spinifera 


5 





T. raschii 





4 


Gammaridea 





58 


Decapoda 





8 


Pandalidae 





6 


Pandalus goniurus 





4 


Osteichthyes 


3 


62 


Osmeridae 


1,584 


33 


Mallotus villosus 


526 


41 


Theragra chalcogramma 





4 




n= 18 


n= 16 



a Data from Krasnow and Sanger (1982) 

b Values are Index of Relative Importance values calculated after Pinkas 
and others (1971). 



In contrast to the results of Sealy (1975c), no Ammodytes 
were present, but, similar to Sealy's (1975c) study, 
Thysanoessa was an important prey item. 

The results from the February 1978 collections were 
extremely different from the 1976/1977 winter data. Mysids 
dominated the prey items with a cumulative IRI value of 
11,892 {table 2). Osteichthyes were second, followed by 
gammarids and capelin {Mallotus villosus). A total of 13 
different prey items were identified {table 2). Once again, no 
Ammodytes were noted, and even Thysanoessa was reduced 
to an IRI value of 4. Sealy's (1975c) breeding period study 
did not detect mysids and gammarids, but these prey items 
appear to be more important in the winter diet of murrelets, 
at least in the Gulf of Alaska (Sanger 1987b, Sanger and 
Jones 1982). The lack of Thysanoessa consumption in 
February 1978 by the murrelets is particularly interesting in 
light of Sealy's (1975c) work. Krasnow and Sanger (1982) 
reported that murrelets fed primarily in shallow water but 
obtained their prey throughout the water column. Sanger 
(1987b) noted that the ability of murrelets to forage at least 
part of the time near the bottom assures a broader trophic 
spectrum than a food supply originating with phytoplankton 
productivity in the water column alone. 



The reduction of capelin in the winter diet of murrelets 
between the study periods may be due to the dynamic nature 
of capelin populations. Because capelin live only 3 or 4 
years and most spawn only once, poor recruitment of a 
given year class can lead to cycles of abundance and near 
absence [Warner and Dick in Krasnow and Sanger (1982)]. 
Fisheries data indicated that the distribution of capelin was 
different in the 2 years, with most fish being caught in deep 
troughs in 1 978 [Rogers and others in Krasnow and Sanger 
(1982)]. Additionally, fewer capelin and more sand lance 
were fed to Black-legged Kittiwake {Rissa tridactyla) chicks 
in Northern Sitkalidak Strait during 1978 than in 1977. 
Productivity of kittiwakes declined from 0.74 young fledged 
per nest attempt to 0.17, suggesting that the availability of 
food was depressed below some "critical level" [Baird and 
Hatch in Krasnow and Sanger (1982)]. Productivity of 
kittiwakes in Chiniak Bay also decreased, from 1 .23 young 
fledged per nest attempt in 1977 to 0.77 in 1978 [Nysewander 
and Barbour in Krasnow and Sanger (1982)]. Food samples 
were not collected at the breeding colonies of kittiwakes in 
Chiniak Bay in 1978, and thus the assumption that fewer 
capelin were brought to chicks than during the previous 
years could not be substantiated. 

If euphausiids were scarce or, for some reason, 
unavailable to murrelets in early 1978 in Chiniak Bay, then 
the ability of the murrelet to feed so heavily on detritivores 
such as mysids and gammarids likely demonstrates prey- 
switching capability. This adaptive and opportunistic behavior 
illustrates the result of natural selection pressure due to 
dynamic prey populations. Alternatively, two factors, small 
sample size and a difference in the collection period (5 
months compared to 1 month), could be complicating the 
results. However, given the information on kittiwake 
reproduction and capelin being found in deeper waters cited 
above, it would appear that changes in the marine food web 
in Chiniak Bay between years and prey-switching behavior 
by the murrelet are more plausible explanations. 

The results of Krasnow and Sanger's (1982) study of 
breeding-season diet at Izhut Bay and Northern Sitkalidak 
Strait in 1978 pointed to the importance of local differences 
in the relative availability of major prey species within the 
same year. The diets from the two different study areas 
included a high proportion of unidentified osteichthyes {table 
3), with ten different prey items identified in the summer 
diet, comparing with 9 from Sealy (1975c). Euphausiids 
were more common in the murrelet diet at northern Sitkalidak 
Strait. For the murrelets and most other seabird species in 
the Kodiak area, distinct seasonal trends were apparent from 
spring through late summer 1978. Marbled Murrelets, Tufted 
Puffins {Fratercula cirrhata). Sooty Shearwaters {Puffinus 
griseus), and Black-legged Kittiwakes exploited a similar 
suite of prey. Sand lance and euphausiids were taken during 
spring, capelin during early summer, and sand lance during 
late summer. The authors attributed this chronology to the 
probable seasonal occurrence and distribution of prey as did 
Sealy (1975c) and Carter (1984) in their study areas. 



228 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burketl 



Chapter 22 



Food Habits and Prey Ecology 



Table 3 — Comparison of May 1978 breeding season diet of Marbled 
Murrelets between Izhut Bax and Sorthern Sitkalidak Strait, Alaska" 





Location 


Prey 


Izhut Bay 


Northern Sitkalidak Strait 


Crustacea 


60" 







Thysanoessa inermis 







18,910 


Osteichthyes 


316 




82 


Osmeridae 


326 







Mallotus villosus 


5,957 




190 




/i = 3 




H = 4 



' Data from Krasnow and Sanger (1982) 

b Values are Index of Relative Importance values calculated after Pinkas 
and others (1971) 



The difference between the two areas in the May diet 
(table 3) may be due to the small sample sizes or may 
represent a local difference in prey abundance as discussed 
above relative to winter diet. The two study areas showed 
similarity in murrelet diet in June, with fish (primarily capelin) 
the most important food item. The July samples indicated 
the importance of sand lance and fish in murrelet diet during 
that period: three birds collected at Izhut Bay had only sand 
lance in their stomachs, while four birds collected at Sitkalidak 
were full of sand lance and other unidentified osteichthyes. 

Sanger (1983) 

Sanger's compilation of data from throughout the Gulf 
of Alaska, and across all seasons, provides an overview of 
the broad spectrum of the murrelet's diet (table 1). Data were 
derived from multiple Outer Continental Shelf Environmental 
Assessment Program (OCSEAP) studies (including Krasnow 
and Sanger 1982. Sanger and Jones 1982) and from the 
National Marine Fisheries Service (n = 129). At least 16 prey 
species were identified. This broad spectrum of prey species 
from different trophic levels is a good indication that the 
murrelet is an opportunistic feeder, though preferences have 
been documented (Sealy 1975c). Generally, murrelets seem 
to prefer euphausiids in spring and fish in summer though 
prey availability and energetic requirements during these 
seasons are also important factors in prey selection (Carter 
and Sealy 1990, Cody 1973, Sealy 1975c). 

Additionally, "food-chain pathways that include detritus 
may result in a more stable food supply than non-detrital 
food chains. This could be reflected in demersal-benthic 
feeders like Pelagic Cormorants [Phalacrocorax pelagicus] 
and Pigeon Guillemots [Cepphus columba] showing stable 
productivity over the years, compared with midwater and 
surface feeders. Winter survival of species like Common 
Murres [Una aalge] and Marbled Murrelets may be enhanced 
by their ability to alter their "normal' diet of pelagic fishes 



to include demersal crustaceans, thus seasonally linking 
themselves to a detrital-based food chain" (Sanger 1987a). 

Sanger (1987b) 

One last example of the importance of local conditions 
on murrelet diet from the OCSEAP work in Alaska comes 
from a summary of work done in Kachemak Bay during the 
winter of 1978 (Sanger 1987b). Twenty-one murrelets were 
collected from January to April 1978, and 18 stomachs were 
used for the analysis. Capelin and osmerids dominated the 
diet, followed by euphausiids (Thysanoessa sp.), mysids, 
unidentified gammarid amphipods, and sand lance. Compared 
to the work of Krasnow and Sanger (1982) in Chiniak Bay, 
euphausiids were more important, and sand lance were taken. 
Thus, although the sample sizes are similar, the relative 
importance of prey species is variable. This disparity is 
another example of the importance of local and interannual 
conditions in determining murrelet food habits. 

Carter (1984) 

Carter's intensive study occurred in Barkley Sound, on 
the southwest coast of Vancouver Island. Field work was 
conducted from 10 May to 7 September 1979, 18-19 
December 1979, and 8 June to 13 October 1980. Eighty- 
seven murrelets were obtained during the study and examined 
for diet information. Carter (1984) noted that small fish 
larvae (<31 mm) were apparently digested quickly, and 
therefore this size class was under-represented in the results. 
Food samples from both sexes were taken throughout the 
day in both years and were combined for analysis. Carter 
also separated the diet of breeding, molting, hatching-year, 
and winter birds and calculated a relative importance value 
in the same way of Sealy (1975c), though he referred to this 
percent value as frequency. 

Breeding adults fed primarily on sand lance and Pacific 
herring (Clupea harengus), including larval and juvenile 
fish (table 1). Molting and hatching-year birds also fed 
primarily on herring and sand lance, and four juvenile northern 
anchovy (Engraulis mordax) were found in the stomach of 
one molting bird. Carter (1984) noted that molting murrelets 
consumed more herring (90 percent) than sand lance (7 
percent), and the same was true for the hatching-year 
murrelets, with herring consumption at 81 percent and sand 
lance at 13 percent. By contrast, the breeding murrelets 
consumed more sand lance (63 percent) and less herring (36 
percent) (table 1). 

In contrast to the work of Sealy (1975c), euphausiids 
were absent in the diet of murrelets in Barkley Sound. Though 
Carter's (1984) work began approximately one month later 
than Sealy' s (1975c), euphausiids in minor amounts should 
have occurred at least in May and throughout the summer at 
least as a minor component of the diet. Additionally, the 
overall diversity of prey species in the summer diet of 
murrelets from Barkley Sound was low (4 different prey 
items) compared to 9 from Sealy' s (1975c) study and 10 
from Krasnow and Sanger (1982). 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



229 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



Five murrelets collected in winter had eaten scorpaenid 
rockfish and squid (Loligo opalescens), as well as large 
juvenile herring and sand lance (table 1). Scorpaenids and 
Loligo were also found in the murrelet diet at Langara Island 
by Sealy (1975c) during the breeding season (table 1). 

Carter (1984) also made observations at sea of adults 
holding fish for nestlings; Ammodytes, Clupea, and Engraulis 
were documented as nestling food (table 1). 

The importance of herring in the diet of the murrelet in 
Carter's (1984) study correlates with the local abundance 
and availability of juvenile herring. He suggested that 
murrelets fed opportunistically on available prey and noted 
that juvenile herring were abundant only in localized areas 
near spawning grounds (Hourston in Carter 1984). This 
conclusion is further strengthened by the work of Vermeer 
(1992) discussed below. 

Vermeer (1992) 

Winter food habits of murrelets from Quatsino Sound, 
British Columbia, were studied for the period from October 
1981 through March 1982 (Vermeer 1992). Quatsino Sound 
is located approximately 270 kilometers northwesterly of 
Barkley Sound where Carter's (1984) work was conducted. 

Twenty-five murrelets were collected, and all birds (100 
percent) contained food. Most fish were digested, but Pacific 
herring were identified in 15 of the 25 murrelets. All 
invertebrates eaten consisted of euphausiids, of which T. 
spinifera and E. pacifica were the main species. The fish 
portion of the diet constituted 71.2 percent of the wet weight 
of the prey items, and the invertebrate portion was 28.7 percent; 
thus, the murrelets ate mostly fish, primarily herring, during 
the non-breeding season in Quatsino Sound (table 1). Sand 
lance were not consumed, and the diversity of prey items (at 
least 3) was low compared to that found in the winter diet 
work by Krasnow and Sanger (1982) and Sanger (1987b). 

Vermeer (1992) did point out that the study location 
has one of the largest herring spawn areas along the west 
coast of Vancouver Island and that herring spawn constitutes 
a major food source for piscivorous as well as nonpiscivorous 
birds, such as diving ducks. The massive presence of herring 
in March for spawning and the predictable nature of this 
occurrence has resulted in annual utilization of this resource 
by many seabirds and other animals (Vermeer 1992). 
Therefore, it seems apparent that the high use of herring in 
Vermeer's (1992) study is another example of the 
opportunistic foraging behavior of the murrelet and another 
demonstration of the importance of local differences in 
availability of prey as noted by Krasnow and Sanger (1982). 
Of further interest, four male murrelets collected in Departure 
Bay on the southeast coast of Vancouver Island during 
February and March (in 1928 and 1929) did not contain any 
identifiable herring in their stomachs even though the study 
area was also known as a major spawn location for herring 
during March (Munro and Clemens 1931). Results and 
implications of the Munro and Clemens (1931) collection 
effort are described in more detail below. 



Freshwater Feeding 

The studies described previously were conducted to 
assess murrelet food habits in the marine environment. To 
assess the importance of freshwater lakes in the feeding 
ecology of murrelets, Carter and Sealy (1986) summarized 
records of year-round use of coastal lakes for the period 
1909 to 1984 from Alaska to California. No records were 
found for California. Three of the 67 records included small 
collections of murrelets at lakes in British Columbia during 
late April and early May. Five stomachs of adults were 
examined, and three were found to contain yearling Kokanee 
salmon (Onchorhynchus nerka kennerlyi), while the fourth 
contained two fingerling sockeye salmon (O. nerka). The 
examiner of the fifth murrelet, R.M. Stewart, noted, "The 
stomach was full of small fish which looked like salmon fry" 
[Onchorhyncus or Salmo sp.] (Brooks 1928) (table 1). Carter 
and Sealy's (1986) work contains numerous anecdotal 
sightings of murrelets feeding at inland lakes and references 
which document many of the lakes as large nurseries for 
juvenile salmon. The discussion includes evidence for 
nocturnal feeding by murrelets and winter-time use of inland 
lakes. The relative lack of inland lakes near known nesting 
sites south of British Columbia, along with a lack of census 
effort for murrelets at inland lakes, could lead to an 
underestimate of the importance of lakes and freshwater fish 
species as a food source for the murrelet. The effect of the 
reduction of salmonid stocks on the use of lakes by murrelets 
is unknown. This aspect of the murrelet' s life history needs 
further investigation throughout its range. 

Isotopic Analysis of Diet 

Stable carbon and nitrogen isotopic analyses were 
performed on tissues of Marbled Murrelets collected from 
July to December 1979 (Carter 1984), in Barkley Sound 
(n = 18), and in June 1985 on Johnston Lake, British Columbia 
(n = 3) (Hobson 1990). Most murrelets showed stable carbon 
isotopic values (pectoral muscle) between -15.5 and -17.5, 
and males and females were the same. These values compare 
favorably to the value of - 1 7.9 for a sample of five Ammodytes 
sp. taken from coastal British Columbia for comparison. 
However, three individuals, an adult male from Barkley 
Sound and two adult males from Johnston Lake, differed 
significantly from the group. On the basis of a model, Hobson 
concluded that the three individuals had short-term freshwater- 
derived protein inputs to their diets ranging from 50 to 100 
percent. Hobson (1990) suggested that while some murrelets 
may feed exclusively on freshwater prey for a short but 
important period of several weeks, freshwater protein did 
not appear to be a significant long-term dietary component. 
However, he concluded that he was unable to ascertain the 
relative importance of freshwater feeding in different murrelet 
populations without additional analysis. He suggested that 
tissues from murrelets found dead or collected for other 
studies be analyzed by isotopes of stable carbon. 

Analysis by isotopes of stable nitrogen cannot be used 
for separating dietary differences between freshwater and 



230 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burken 



Chapter 22 



Food Habits and Prey Ecology 



marine protein contributions, because nitrogen isotope ratios 
in the muscles of fish species in coastal lakes may overlap 
with those of marine fish (Hobson 1990). Lower trophic- 
level fish such as fingerling salmonids also overlap with 
marine invertebrates. Thus, nitrogen isotope analysis may 
be better suited than carbon to delineating the trophic levels 
of murrelets and other seabirds. The results of this analysis 
(Hobson 1990) showed Marbled Murrelets in the middle of 
a spectrum ( 10 species) from Dovekies (Alle alle) to Pigeon 
Guillemots; the Marbled Murrelet was between the Ancient 
Murrelet and the Common Murre. This isotopically 
intermediate position is consistent with the results of the 
studies described above which document murrelet 
consumption of invertebrate prey as well as marine fish. 
The trophic-level approach also has the value of being less 
biased against the soft-bodied invertebrates which are not 
easily detected in conventional studies. 

A further analysis of the variability of stable nitrogen 
isotopes in wildlife showed that tissue can be enriched because 
of fasting or nutritional stress (Hobson and others 1993). 
Thus, studies using analysis by stable nitrogen isotopes to 
infer diet or trophic position must take into account the 
nutritional history of the individual specimen. Fasting should 
not be a factor for the murrelet because both sexes incubate 
the egg and feed the nestling, but nutritional stress could 
affect the results in a year of severe prey shortage. 

Ecological Studies and Anecdotal Information 

Alaska 

Food habits of the murrelet were described by Bent 
(1963). "The food of the marbled murrelet seems to consist 
largely of fish which it obtains by diving in the tide rips and 
other places where it can find small fry swimming in schools." 
It appears he derived this information from observations 
contained in Grinnell (1897) and Grinnell (1910) {table 1). 
In the summer of 1896. during a visit to Sitka Bay, Alaska. 
Grinnell (1897) noted, "Small fish caught by diving seemed 
to be the standard article of food, but dissection of the 
stomachs also showed remains of some small mollusks. A 
shoal of candle-fish [Thaleichthys pacificus] was sure to 
have among its followers, besides a cloud of Pacific kitti wakes 
[Rissa sp.], several of the Murrelets" (table 1). Grinnell 
(1910: 366) noted fish as a prey item in a collected specimen 
and during an observation by Joseph Dixon of a foraging 
murrelet, but the species offish were not recorded (table 7). 

Observations at the first documented ground nest of a 
murrelet indicated capelin as a food source for the nestling 
(Simons 1980) (table 7). An adult murrelet delivered a 
single fish about 8 cm long. Simons (1980) noted, "The 
fish appeared to be a capelin (Mallotus sp.)..." [emphasis 
added]. This observation would appear valid given the 
documented importance of capelin in murrelet diet in 
Alaska (Sanger 1983). Simons (1980) also noted that the 
pattern of weight gain was variable from days 2 to 12. and 
he suggested the possibility of multiple feedings. He 
concluded that predation and the distribution of the food 



resource were important selective agents acting upon 
ground-nesting murrelets. 

British Columbia 

Food habits of "water fowl" during the spawning season 
of herring in the vicinity of Departure Bay, British Columbia, 
were studied between 1928 and 1930 (Munro and Clemens 
1931). Four male murrelets were collected in late February 
to mid-March, and the stomachs contained Cymatogaster, 
larval fish, mysids, and schizopods (table 1). The archaic 
group schizopoda included euphausiids and mysids because, 
superficially, the members of these two orders appeared so 
similar. These two groups are now separated into different 
tribes based on characteristics of the carapace and the 
distinguishing luminescent organs of the euphausiids (Hardy 
1965: 171-172). 

The results from Munro and Clemens ( 1 93 1 ) differ from 
the winter results of Carter (1984) and Vermeer (1992), in 
that identifiable herring are absent (table 7). This difference 
could be due to the small sample size. Alternatively, it could 
result from differences in availability of herring age classes 
and in herring distribution relative to murrelets, and differences 
in the magnitude and duration of the herring spawn between 
the three study areas (McAllister, pers. comm. >. A number 
of herring stocks aggregate close to the spawning area for 
some time before actually moving on to the grounds to 
spawn (Lambert 1987). 

An anecdotal account of murrelet diet by Guiguet was 
published in 1956. He spent many summers on zoological 
exploration in coastal British Columbia and stated that the 
murrelet "...eats small Crustacea such as euphausid [sic] 
shrimps, and fishes such as the sand launce [sic]...." He also 
described watching murrelets foraging off the Queen 
Charlotte Islands in July 1946 and noted, "all were feeding 
on sand launces [sic]...." When darkness had almost 
descended that day, the murrelets disappeared inland to the 
west. Guiguet (1956) noted, "All of them were 'packing 
feed' in their bills, and the silvery sand launce [sic] showed 
up in the darkness" (table 1). 

Between 6 June and 8 August 1991, Manon and others 
(1992) conducted 27 at-sea surveys to determine the composition 
and density of mixed feeding flocks. They observed 126 feeding 
flocks, 100 of which contained only murrelets and Glaucous- 
winged gulls (Lams glaucescens). Murrelets were seen to 
feed on schools of sand lance by driving the fish to the surface. 
First-year sand lances were the only prey identified in feeding 
flocks (table 7). In the evenings, murrelets were seen holding 
larger sand lance. Pacific herring, and shiner perch as prey for 
nestlings (table 7). The nestling prey items closely match the 
juvenile diet reported by Sealy (1975c), and two of the nestling 
items, herring and sand lance, reported by Carter (1984) and 
Guiguet (1956), respectively. 

Additional anecdotal information on nestling food habits 
in British Columbia comes from a nest which was monitored 
in summer 1993 (Jones and Dechesne 1994). Sand lance was 
noted as a prey item for the nestling (table I). 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



231 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



Washington 

During the summers of 1968 and 1969, Cody (1973) 
collected information on seabird breeding activity, prey 
species, and foraging patterns off the west coast of the 
Olympic Peninsula in Washington State. Murrelets holding 
fish before their evening flights inland were observed at 
close range from a boat. The birds were seen to carry only 
anchovy (Engraulis) and sand lance (Ammodytes) in their 
bills, and it was presumed these fish were for nestlings 
(Carter and Sealy 1987a) (table 1). The murrelets showed 
great similarity in chick diet with the Common Murre, 
Tufted Puffin, and Rhinoceros Auklet (Cerorhinca 
monocerata), though smelt (Hypomesus) and sea-bass 
(Sebastoides) were also recovered from 54 fish loads for 
these latter three species of alcids. 

Similar to Sealy's (1975c) study of sympatric Ancient 
and Marbled Murrelets, Cody (1973) concluded that differences 
in foraging areas at sea reduced interspecific competition 
between alcids off the west coast of the Olympic Peninsula, 
though prey species consumed were similar. Lacking specific 
knowledge of murrelet nesting areas, neither of these 
researchers were able to compare foraging areas with nesting 
habitat distribution, though Cody (1973) concluded that the 
zonation of alcid feeding areas with respect to distance from 
the nest was the most important factor affecting coexistence. 
He contrasted this to other studies which have found differences 
in diet between similar seabird species to be the isolating 
mechanism. He also pointed out that foraging zonation which 
is optimal while adults feed nest-bound young is relaxed and 
expanded when young leave their nests and accompany the 
parents. Cody (1973) found that murrelets fed within a few 
kilometers of the shore. He observed that in the evenings 
they were often seen carrying food within a half kilometer of 
the Hoh and Quilleute Rivers and that adults and partially- 
grown, non-flying young were observed close to these same 
river mouths in August. 

Cody presumed these rivers provided transportation 
for the young murrelets from inland nesting sites (Day and 
others 1983, Nechaev 1986). The discovery of a young 
murrelet at a freshwater marsh close to the sea in British 
Columbia is described by Brooks (1926a). The bird appeared 
unable to fly, and it was noted that the primaries were in 
sheaths at their bases and there was a good deal of down on 
the head, back, and flanks. Another similar young was with 
it. Brooks (1926a) noted another juvenile murrelet, collected 
off Langara Island, British Columbia: "...the bases of its 
quills still in the sheath was taken some 200 yards out to 
sea...". Young fledglings would consume available prey 
resources in freshwater environments as they gained 
sustained flight capabilities and made their way to the 
ocean (Carter and Sealy 1986). It is thought that the majority 
of murrelets fledge by direct flight to the ocean (Nelson and 
Hamer, this volume a). Diving behavior is an escape response 
and does not necessarily indicate an inability to fly (Carter 
and Sealy 1987b); however, repeated harassment of the 
juveniles by Cody (1973) resulted in no flight attempts, 



though adults would take wing when continually harassed 
by boat (Cody, pers. comm.). 

Additional work by Cody in Carter (1984) at the San 
Juan Islands again revealed anchovy as nestling prey from 
fish held in the bill by murrelets on the water (table 1). 

One other observation on murrelet food habits from 
Washington was provided by Hunt (pers. comm.). He observed 
murrelets foraging in August in mixed-species flocks in the 
San Juan Islands. He dip-netted (approximately 7.5 cm mesh) 
for surface fish in this foraging area and captured only 
herring (table /). 

Oregon 

At-sea surveys for murrelets during 1992 off the coast 
of Oregon resulted in some anecdotal information on nestling 
food items (Strong and others 1993). A total of six murrelets 
carrying fish were observed from 15 June to 11 August 
(table 1). The first two observations occurred on 15 June, 
and the prey type was judged to be "smelt sp." (osmeridae). 
The next four observations, on 1 August, 2 August, and 1 1 
August (two observations), were of sand lance. On the basis 
of additional observations of other seabirds with prey over 
the same time period, the authors thought a switch in prey 
occurred from smelt in late July to sand lance thereafter. 

Video footage from an active nest site in 1992 documented 
sand lance as nestling food, and during at-sea surveys, 
observers noted osmerids, sand lance, and a possible herring 
as nestling food items being held by murrelets (Nelson, pers. 
comm.) (table 1). 

California 

A report on the population status and conservation 
problems of the murrelet in California was produced in 1 988 
as the Department of Fish and Game began gathering 
information on the species (Carter and Erickson 1988). Field 
notes from work by R. H. Beck in the vicinity of Point Pinos, 
Monterey County, were included in Carter and Erickson 's 
(1988) report and are repeated here (Museum of Vertebrate 
Zoology; see also Beck 1910): "...the Marbled Murrelets 
yesterday [had in their stomachs] 2, 3, 4, or 5 small sardines 
[Sardinops sagax] about 3 inches long" (November 24, 1910); 
four days later, 13 murrelets were collected (November 28, 
1910), and Beck noted, "Sardines 2 to 3 inches long in 
stomachs"; then, on February 16, 1911, Beck reported, "A 
six [inch] needle fish? [Strongylura exilis] swallowed by 
Marbled Murrelet inside bill when picked up fish just caught"; 
and finally, on March 1, 1911, a Marbled Murrelet was 
collected with a "...6 1/2 [inch] fish in stomach" (table 1). 

The reference to the possible needlefish (California 
needlefish = Strongylura exilis) is interesting because the 
northern distribution limit for this species is San Francisco 
(Miller and Lea 1972). Carter and Erickson (1988) thought 
the fish may have been a sand lance. 

Carter and Erickson (1988) also reported on the food 
habits of 10 murrelets which were collected in early fall 
from northern Monterey Bay in the late 1970's. The murrelets 



232 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkeit 



Chapter 22 



Food Habits and Prey Ecology 



were noted as feeding mainly on anchovy and to a lesser 
extent on sand lance (table 1). 

Another instance of anchovy in murrelet diet came from 
mist netting of murrelets in Redwood National Park for 
radio-telemetry purposes during summer 1989 (Ralph and 
others 1990). During this work, on July 3, 1989, one murrelet 
hit the mist net and bounced out (05:30 p.d.t.), leaving a 
whole northern anchovy at the base of the net. The anchovy 
weighed 10.0 grams and was 113 mm in length. It seems 
most likely that this prey item was destined for a murrelet 
nestling (table 1). 

It is unfortunate that systematic studies of murrelet food 
habits in this region of California did not occur before and 
after the great sardine fishery (mid 1930 to mid 1940). The 
anecdotal information from above mirrors the documented 
change in prey abundance over time, from sardine to anchovy. 
The interesting history of sardine and anchovy population 
fluctuations and their fisheries are briefly summarized below 
under the prey ecology section of this chapter. The fact that 
murrelets have persisted in the central California region 
after a decline in the largest fishery in the Western Hemisphere 
is probably another indication of the opportunistic feeding 
behavior of the bird. This flexibility in prey choice has 
probably helped to sustain the murrelet population in this 
geographic region in spite of massive loss and deterioration 
of inland nesting habitat. 

Anecdotal information on nestling diet was obtained 
from video footage recorded during observation of an active 
nest site in Big Basin State Park in the Santa Cruz Mountains 
(Naslund 1993a). Three fish carried to the nestling were 
identified (table 1). Two of the fish appeared to be either 
northern anchovy or possibly of the clupeidae. The third fish 
was judged to be a smelt (osmeridae). 

Rockfish make up an important component of seabird 
diet in California, and if more intensive studies of murrelet 
diet were conducted it is possible that these fish would be 
found to be eaten by murrelets (Ainley and others, this 
volume). Both Sealy (1975c) and Carter (1984) documented 
scorpaenids in the murrelet's diet (table 1). 

Food Habits Summary 

The sand lance is the most common food of the murrelet 
across its range (table 1). For the fish species, records of 
sand lance represent 52 percent of the compiled information 
(11 occurrences per 21 studies/anecdotal observations) on 
murrelet food habits. The next most commonly recorded 
species are anchovy and herring at 29 percent, followed by 
osmerids at 24 percent, and by Cymatogaster at 14 percent. 

Euphausiids as a group represented 24 percent of the 
compiled information (table 1). They were generally not a 
dominant component of murrelet diet during the breeding 
season: however, euphausiids were an important prey source 
for murrelets in the spring (Sealy 1975c) and during the 
breeding season in some years (Krasnow and Sanger 1982). 
Euphausiids were also important during the winter in the 
Gulf of Alaska (Krasnow and Sanger 1982) and in British 



Columbia (Vermeer 1992). Mysids and gammarids were 
another component of murrelet diet, especially in winter 
(Krasnow and Sanger 1982, Munro and Clemens 1931, 
Sanger 1987b). 

Studies under the OCSEAP program revealed the 
importance of seasonal and interannual variation in prey 
abundance (Krasnow and Sanger 1982, Sanger 1983, Sanger 
1987b). The OCSEAP compilation (Sanger 1983) revealed a 
broader prey spectrum compared to systematic studies (Carter 
1984, Sealy 1975c, Vermeer 1992), though this may have 
been partially because of the larger time period and larger 
geographic extent of collection (table I). It may also have 
been a function of the larger sample size compared to these 
other studies (table 1). 

Comparison of results from Sealy ( 1 975c), Carter (1984), 
and Vermeer (1992) reveals the influence of site-specific 
conditions on prey availability and selection by murrelets 
(table 1). Differences between adult, nestling, and fledgling 
diet were also apparent (Carter 1984, Mahon and others 
1992, Sealy 1975c) (table 1). 

Though much work needs to be done on food habits in 
different geographic regions and seasons, in general it can 
be said that murrelets feed on invertebrates such as 
euphausiids, mysids, decapods and amphipods, and small 
schooling fishes including sand lance, anchovy, herring, 
smelt, and seaperch. The fish portion of the diet is most 
important in the summer and coincides with the nestling and 
fledgling period (Carter 1984, Carter and Sealy 1990, Sealy 
1975c). 

Prey Ecology 

Because few systematic studies of murrelet food habits 
have taken place and the murrelet occupies such a large 
geographic area with a wide variety offish species potentially 
available, the rest of this chapter will focus on selected 
prey species considered most important in murrelet diet at 
this time. Due to the long-standing commercial value of 
anchovies, herring, and sardines, there is a large body of 
information on life history and factors affecting their 
abundance and distribution. The following overview is not 
an attempt to compile the rich literature on these or the 
other known prey species, but instead focuses on interesting 
aspects of their life history and the interrelationship between 
prey species, murrelets, humans, and the marine 
environment. The relationship between other seabirds and 
these same prey resources, along with the marine 
environment, will be discussed. Sand lance and euphausiids 
have been little studied compared to the commercially 
valuable fish species, but are discussed first because of 
their position and interaction in the marine food web, and 
their importance in the murrelet's diet. 

Euphausiids 

Euphausiids are a group of small crustaceans which 
make up part of the zooplankton ("krill") found in the 
marine environment. Euphausiids are more or less transparent 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



233 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



and phosphorescent and closely resemble shrimps in form 
though they are often not more than 25 mm long. The 
phosphorescent organs are along the sides of the body. 
Their purpose is not known. Attached to the thorax are the 
eight pairs of two-branched legs which give rise to the name 
"schizopoda," as this order was formerly called (Johnson 
and Snook 1967: 293-294). 

Zooplankton are found in greater abundance during cold- 
water years in California waters. Many of the zooplankton 
are predators on fish eggs and larval fish, and their abundance 
was sometimes twenty times greater during the colder periods 
(Reid and others 1958). Accordingly, not only would the 
lowered temperature affect survival of fish eggs and larvae 
directly, it would also add to the hazards of being eaten by 
providing conditions for the rapid increase of zooplankton 
(Ricketts and Calvin 1962: 394). 

Komaki (1967) summarized information obtained from 
fishermen on the phenomenon of surface swarming of 
euphausiids {E. pacified) in the Sea of Japan. This phenomenon 
differs from the usual vertical migratory behavior because it 
occurs in the daytime, independent of light intensity. The 
swarming season in the Kinkazan waters ranged between 
late February and late May. Water temperature was determined 
to be the most important factor, with swarming starting at a 
slightly higher temperature than the local minimum (7 degrees 
Celsius), continuing with increasing temperature, and then 
terminating as the temperature exceeded 16 degrees Celsius. 
Because swarming did not occur earlier in the year when 
temperatures were favorable, Komaki (1967) concluded that 
the swarming was related to reproduction. Also, it appeared 
that the population was composed of several stocks, and that 
as stocks reached a certain degree of maturity, they approached 
the coast in succession. 

The daily phenomenon of vertical migration was noted 
as early as 1872 during the Challenger Expedition. Many 
plankton animals actively move towards the surface of the 
ocean at night and sink or swim away to the depths in the 
daytime. Vertical climbing requires much energy and has 
been developed so frequently in the animal kingdom that it 
was thought to clearly be of some significance in the lives of 
such animals (Hardy 1965: 199-200). The main proximate 
factor for daily vertical migration appears to be light intensity 
(Cushing in Raymont 1963: 435). 

Both E. pacifica and T. spinifera were found to undergo 
vertical migration off Washington State in summer 1967 
(Alton and Blackburn 1972). High catch rates were sustained 
from near-surface water throughout the late evening and 
early morning hours, approximately 2200 to 0500 hours. 

Hardy (1965: 212-215) advanced a general theory for 
the value of diurnal vertical migration. Because the uppermost 
layers of the sea generally move at higher speeds than lower 
levels and bottom topography results in currents which may 
differ from surface layers, the regular movement of plankton 
between these layers allows the animals to be carried over 
greater distances than would otherwise be the case. Thus, the 
plankton population can be distributed over a much larger 



area of the ocean than if continually moved by only one 
body of water. This large-scale movement has the advantage 
of putting the animals in contact with more food source 
patches. Individual variation in the degree of vertical migration 
and the amount of time spent at any one layer further promote 
the patchy distribution of plankton. 

A genetic theory has also been proposed (David in 
Raymont 1963: 466). Marine planktonic species may tend to 
become divided into relatively small, separate populations if 
continually drifting in one stratum and not normally 
encountering directional stimuli to encourage horizontal 
migration. However, the broader distribution caused by 
vertical migration would help to encourage interchange of 
zooplankton populations and thus promote gene flow. 

There are many other theories regarding vertical 
migration. The fact that both Raymont (1963) and Hardy 
(1965) devoted an entire chapter to the subject attests to the 
complexity of factors which operate in the marine 
environment. As summarized by Hardy (1965: 217): "There 
can be no doubt that the patchy distribution of the plankton 
must be due to a great variety of causes." Raymont (1963: 
466) ended his chapter by recognizing the need for more 
research on the subject: "At this stage no conclusive answer 
can be given to the question as to the value of diurnal 
vertical migration, but the tremendously wide occurrence of 
this phenomenon in the seas is one of the most challenging 
aspects of marine plankton study." 

With the variability in zooplankton distribution and 
abundance, the way in which murrelets find such prey resources 
warrants attention. The interannual variability in euphausiid 
consumption (tables 2 and 3) noted by Krasnow and Sanger 
(1982) could demonstrate differences in zooplankton 
distribution (rather than abundance) and the corresponding 
inability of murrelets to locate and use the resource. However, 
some of the distribution patterns should be predictable at 
least between "normal" years, and thus the learning of foraging 
areas by murrelets would indeed be important for minimizing 
energy expenditure as suggested by Carter (1984). Komaki 
(1967) demonstrated that E. pacifica and sand eels (A. 
personatus) fluctuated in parallel on the basis of data on 
fishery harvest, and that sand eels were taken in almost the 
same area as the euphausiid fishery. The traditional euphausiid 
swarming areas were known to the fishermen, though the 
density of the swarm varied between years (Komaki 1967). 
Thus, conservation of the murrelet and its food web will be 
aided by identification and appropriate management of 
important euphausiid swarming areas, especially in the vicinity 
of known murrelet nesting areas. 

Pacific Sand Lance 

Sand lance are slim, elongated, usually silver fishes 
especially abundant in northern seas. They belong to the 
ammodytidae and are sometimes called sand eels, but they 
are not true eels even though eel-like in shape and 
movement. The Pacific sand lance is distributed from 
southern California to Alaska and to the Sea of Japan. 



234 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



They grow to about 20 centimeters in length (Miller and 
Lea 1972). There has been much confusion over the 
taxonomy of the sand lances throughout the world since 
they are similar in external appearance (Hardy 1965: 209- 
210; McGurk and Warburton 1992). Of this confusion 
Hardy (1965: 210) wrote, "It all goes to show how 
elusive. ..these sand-eels are. They are all very much alike; 
little silvery eel-like fish which occur in large shoals in 
sandy parts of the sea and escape from their predators by 
diving like a flash into the sand and becoming completely 
covered." They are most abundant in the shallow regions 
around the coast, but may also be found on sand banks far 
out on the continental shelf (Hardy 1965: 21 1). 

The most interesting characteristic of the sand lances is 
their ability to burrow into sand or gravel and remain there 
for long periods. Both burrowing and emergence are extremely 
rapid, the fish entering and leaving the surface almost vertically 
at swimming speed. Coastal sand lance may bury themselves 
above low-water mark and remain buried as the tide recedes 
and until it covers the area again. This habit demands a 
loose, porous substrate in which respiratory water maintains 
sufficient oxygen to support life (Scott 1973). 

Food habits of 486 specimens ( 1 5-3 1 cm) of northern 
sand lance (A. dubius) taken at various localities and seasons 
from Nova Scotia Banks revealed copepods as the most 
frequent food item, followed by crustacean larvae, invertebrate 
eggs, and polychaete larvae. Volumetric analysis showed 
copepods to comprise the bulk of the food (65 percent), 
followed by polychaete larvae (15 percent) and euphausiids 
(14 percent). The latter two food items were selected for in 
greater volume when compared to availability, since 
euphausiids made up less than 4 percent of the volume of 
simultaneous plankton tows (Scott 1973). 

McGurk and Warburton (1992) conducted an intensive 
study of environmental conditions and the effects on sand 
lance larvae in the Port Moller estuary in Alaska. They 
found that three waves of spawning sand lance entered the 
estuary from mid- January to late May. Peak spawning occurred 
in January. March, and April. Eggs incubated for a period of 
45 to 94 days. Slow growth was directly responsible for the 
reduced number of cohorts and the long time periods between 
peak hatch dates compared to other demersally-spawning 
fish such as herring or capelin, because first-feeding sand 
lance larvae took longer to vacate their feeding niches. The 
larvae fed primarily during the day on a diet of copepod eggs 
and nauplii, copepodites, and small adult copepods. This 
type of prey and its average length and width were similar to 
that of herring larvae, indicating that the larvae of these two 
species shared the same food resource. 

McGurk and Warburton (1992) concluded that the stock 
of sand lance that spawns in Port Moller belongs to a class of 
stocks that have an entirely estuarine or coastal early life 
history, in contrast to some stocks of sand lance whose 
larvae disperse offshore from inshore spawning sites. This 
life history strategy may have evolved in response to the 
unique physical conditions of the Port Moller estuary — a 



shallow, well-mixed site with sandy substrate that is suitable 
for incubation of demersal eggs next to a deep, stable fjord 
with a rich zooplankton community that is suitable for rearing 
of larval and juvenile sand lance. 

Variation in physical factors, particularly, storm events, 
local wind-forced surface currents, baroclinic surface currents, 
and regional downwelling events at the boundary of the 
estuary cause annual variation in recruitment. Additionally, 
density-dependent factors such as competition for food between 
sand lance, between sand lance and other plankti vorous fish 
larvae such as herring, and between sand lance and invertebrate 
planktivores such as chaetognaths may play as important a 
role as density-independent physical factors (McGurk and 
Warburton 1992). McGurk and Warburton noted that the 
small scale of dispersal in the Port Moller stock also leaves it 
more vulnerable to industrial development such as dredging 
or release of toxic chemicals. 

Sherman and others (1981) summarized research in the 
North Sea which documented an increase in sand eel 
{Ammodytes sp.) as a result of depleted herring and mackerel 
(Scomber sp.) stocks. In the absence of a sand lance fishery 
on the east coast of North America from which to estimate 
population trends, researchers in this area used ichthyoplankton 
surveys. As in the North Sea, population explosions of small, 
fast-growing sand eel coincided with depletions of larger 
tertiary predators, including herring and mackerel. From 
1974 to 1979 the percentage of sand eel increased from less 
than 50 percent of the total mid-winter ichthyoplankton 
community to more than 85 percent (Sherman and others 
1981). This change followed significant fishing stress of the 
northwest Atlantic ecosystem, where fish biomass in the 
region was reduced by 50 percent from 1968 to 1975 (Clark 
and Brown in Sherman and others 1981). 

Clark and Brown concluded that reductions in herring 
and mackerel on both sides of the Atlantic in response to 
heavy fishing mortality, followed by increases in sand eel 
and other small, fast-growing fish, made unlikely the 
hypothesis that the changes were due to environmental factors. 
They concluded that, when a large biomass of mid-size 
predators is removed, it can be replaced by smaller, faster- 
growing, opportunistic species (Sherman and others 1981). 

This relationship was further evaluated by Fogarty and 
others (1991) with a mathematical model. A significant 
negative interaction between sand lance recruitment and an 
integrated measure of herring and mackerel biomass was 
indicated. However, since both herring and mackerel feed 
on sand lance, it was impossible to distinguish the relative 
roles of the two predators. The authors concluded that 
direct evidence of predation by mackerel and herring was 
available to support the inference of interactions between 
sand lance and pelagic predators, though alternative 
hypotheses could be formulated. 

Recent changes in the population of sand eels (Ammodytes 
marinus) at Shetland were studied in relation to estimates of 
seabird predation (Bailey and others 1991). Since 1974 there 
has been a sand eel fishery in inshore waters around Shetland, 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



235 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



and landings have decreased. Simultaneously, there was a 
decrease in consumption of sand eels by seabirds. These 
findings indicate that the switching of seabirds from sand 
eels to other prey is in approximate proportion to the 
abundance of sand eels. However, Bailey and others (1991) 
concluded that more data were needed to significantly refine 
the analysis. It was noted that different seabirds respond 
differently to changes in stock. Surface feeders must forage 
close to the colony and make many fishing trips per day; 
thus they are especially sensitive to reduction in food 
availability. This is in agreement with evidence that Arctic 
Terns (Sterna paradisaea) showed the earliest and most 
severe breeding failures at Shetland (Heubeck in Bailey 
and others 1991). By contrast, some of the larger seabirds 
with generalist feeding abilities took sand eels when these 
were abundantly available but switched diet as the sand eel 
stock declined. 

The relationship between British [Black-legged] 
Kittiwake breeding success and the Shetland stock of sand 
eels (Ammodytes) was studied by Harris and Wanless (1990). 
The evidence that food shortage was responsible for low 
breeding success was mostly circumstantial but, taken as a 
whole, compelling. However, the authors concluded that 
natural factors could have caused the decline in the sand 
eels, rather than overfishing (Kunzlik in Harris and Wanless 
1990). They also suggested that herring predation was 
responsible for the fishery decline. As in the other studies of 
seabirds and sand lance described above, the authors concluded 
that more comprehensive studies were needed to allow 
definitive interpretation of the results. 

More studies on the Pacific sand lance are needed on the 
west coast of North America, especially on environmental 
effects and predator influence on survival and abundance. 
Spawning areas of sand lance need to be identified and 
managed. The trophic links between sand lance and two 
other murrelet prey items, euphausiids and herring, indicate 
a need for comprehensive, long-term study and management. 

Northern Anchovy 

These fish belong to the engraulidae family. They 
have no adipose fin or lateral line and are closely related 
to herring. 

The following life history information (for anchovies 
and sardines) was taken from a draft document (Anonymous 
1993) which was not completed or published because of a 
change in Pacific coastal pelagic species management policy 
between regulatory agencies (Wolf, pers. comm.). 

Northern anchovy are distributed from the Queen 
Charlotte Islands, British Columbia, to Magdalena Bay, Baja 
California. The population is divided into northern, central, 
and southern subpopulations, or stocks. The central 
subpopulation, which supports significant commercial 
fisheries in the United States and Mexico, ranges from 
approximately San Francisco, California, to Punta Baja, Baja 
California. The northern subpopulation supports a small but 
locally important bait fishery in Oregon and California. 



Anchovies are small, short-lived fish typically found in 
schools near the surface. The fish rarely exceed 4 years of age 
and 18 cm in total length. They have a high natural mortality; 
approximately 45 to 55 percent of the total stock may die each 
year of natural causes in the absence of fishing. Northern 
anchovy eat plankton either directly or by filter feeding. 

Anchovy spawn during every month of the year, but 
spawning increases in late winter and early spring and peaks 
from February to April. The eggs, found near the surface, are 
typically ovoid and translucent and require two to four days 
to hatch, depending on water temperature. Anchovy are all 
sexually mature at age 2. The fraction of one-year-olds that 
is sexually mature in a given year depends on water 
temperature and has been observed to range from 47 to 100 
percent (Methot in Anonymous 1993). 

Northern anchovy in the central subpopulation are 
harvested by commercial fisheries in California and Mexico 
for reduction, human consumption, live bait, dead bait, and 
other nonreduction commercial uses. Anchovy landed in 
Mexico are used primarily for reduction although small 
amounts are probably used as bait. Small quantities of the 
northern subpopulation are taken off Oregon and Washington 
for use as dead bait. 

Anchovy landed by the reduction fisheries are converted 
to meal, oil, and soluble protein products sold mainly as 
protein supplements for poultry food and also as feed for 
pigs, farmed fish, fur-producing animals, laboratory animals, 
and household pets. Meal obtained from anchovy is about 65 
percent protein. 

Anderson and others (1980) compared estimates of 
anchovy biomass and catch statistics to Brown Pelican 
(Pelecanus occidentalis californicus) reproductive success. 
Brown Pelican diet was composed of 92 percent anchovies 
in the Southern California Bight (SCB) study area. Mean 
SCB anchovy biomass (square miles of anchovy schools) 
and mean pelican reproductive rate (number of fledglings 
per nesting attempt) were highly correlated. It was estimated 
that a minimum anchovy biomass of 43 square miles was 
necessary for maintaining the existing pelican reproductive 
rate, though it was recognized that the rate would have to 
increase in order to at least maintain the pelicans in the SCB. 
Secondly, the minimum biomass estimate was almost twice 
the forage reserve which was recommended at the time in 
the Anchovy Management Plan. They regarded the 
information as preliminary and concluded that better estimates 
of the forage reserve were needed. 

A similar relationship between anchovies and Elegant 
Terns {Sterna elegans) was described by Schaffner (1986). 
Breeding pairs of Elegant Terns and estimates of anchovy 
spawning biomass were significantly correlated for the period 
of 1979 through 1983. Additionally, extensive overlap in age 
compositions of the tern and fishery samples suggested they 
were using similar resources and the potential for competition 
existed. Anchovies constituted more than 86 percent of the 
chick regurgitations when population size peaked. Schaffner 
(1986) pointed out the similarities to the Brown Pelican 



236 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



study and advised that a close watch of the situation was in 
order because of declining anchovy populations. 

Anderson and others (1980) proposed the establishment 
of protected foraging zones as critical habitat under the 
federal Endangered Species Act in order to assure adequate 
pelican reproduction and conservation. However, they 
recognized that because of the unpredictable nature of anchovy 
distribution, such areas could be difficult to define between 
seasons and between years. Protection of marine habitat as 
critical habitat for the murrelet has also been recommended 
by researchers, the Marbled Murrelet Recovery Team, and 
the California Department of Fish and Game. 

Pacific Sardine 

These small pelagic clupeids occur in the California 
Current system from southern Baja California to southeastern 
Alaska, and in the Gulf of California. In the northern portion 
of the range, occurrence is seasonal. It has been generally 
accepted that the sardine population off the west coast of 
North America consists of three subpopulations. A northern 
subpopulation (northern Baja California to Alaska), a 
southern subpopulation (off Baja California), and a Gulf of 
California subpopulation were distinguished on the basis of 
serological techniques (Vrooman in Anonymous 1993). 

Historically, the sardines migrated extensively, moving 
north as far as British Columbia in the summer and returning 
to southern California and northern Baja California in the 
fall. The migration was complex, and timing and extent of 
movement were affected to some degree by oceanographic 
conditions (Hart in Anonymous 1 993). 

Sardines reach about 41 cm in length, but usually are 
shorter than 30 cm. They live as long as 13 years, although 
most sardines in the historical and current commercial catch 
are 5 years and younger. They spawn in loosely aggregated 
schools in the upper 50 meters of the water column probably 
year-round, with peaks from April to August. Spawning 
has been observed off Oregon, and young fish have been 
seen in waters off British Columbia, but these were probably 
sporadic occurrences (Ahlstrom in Anonymous 1993). The 
spatial and seasonal distribution of spawning is influenced 
by temperature. 

Sardines prey on crustaceans, mostly copepods, and 
consume other phytoplankton, including fish larvae. Larval 
sardines feed extensively on the eggs, larvae, and juvenile 
stages of copepods, as well as on other phytoplankton 
and zooplankton. 

The fishery began in central California in the late 1800's 
and developed in response to a demand for food during 
World War I (Schaefer and others in Wolf 1992). The 
Pacific sardine supported the largest fishery in the Western 
Hemisphere during the 1930's and 1940's, with landings in 
British Columbia, Washington, Oregon, California, and 
Mexico. The fishery declined, beginning in the late 1940's 
and with some short-term reversals, to extremely low levels 
in the 1970's. There was a southward shift as the fishery 
decreased, with landings ceasing in the northwest in 1947- 



1948, and in San Francisco in 1951-1952. The regulatory 
history of the sardine fishery might best be described as 
"too little too late." Regulatory authority for the sardine 
fishery in California rested with the legislature, which 
delegated only limited authority to the Fish and Game 
Commission. State biologists had expressed concern about 
the size of the fishery as early as 1930. Industry opposed 
any regulation of total catch, and an intense debate began 
over whether the decline of the sardine fishery and population 
was due to overfishing or environmental factors (Clark and 
Marr in Wolf 1992). 

It was not until 1967, well after the fishery had collapsed, 
that the California legislature passed an "emergency" bill 
declaring a 2-year moratorium on fishing sardines, and in 
1974 another bill was enacted which established a complete 
moratorium on directed fishing for sardines, though an 
incidental catch provision continued. A small directed fishery 
was first allowed in 1986 and the directed quota has recently 
been enlarged (Wolf 1992). 

Since the early 1980's, sardines have been taken 
incidentally with Pacific {Scomber japonicus) and jack 
mackerel (Trachurus symmetricus) in the southern California 
mackerel fishery and primarily canned for pet food, although 
some were canned for human consumption. Sardines landed 
in the directed sardine fisheries off California are primarily 
canned for human consumption and sold overseas. 

Management of the sardine is difficult in the absence of a 
large fishery since a precise, direct estimate of a relatively 
small biomass is difficult and expensive to obtain (Wolf 
1992). Integrated methods of stock assessment will be necessary 
to manage this resource (Barnes and others 1992). 

Baumgartner and others (1992) presented a composite 
time series of anchovy and Pacific sardine fish-scale- 
deposition rates which they developed from sampling the 
anaerobic layered sediments of the Santa Barbara Basin off 
southern California. Other researchers (Soutar; and Soutar 
and Isaacs in Baumgartner and others 1992) had previously 
collected information on the deposition rates of these species, 
but their sample sizes were limited and there was uncertainty 
in the underlying chronology because of imperfect 
preservation of the annually deposited layers. The new 
sardine and anchovy series provide significantly more 
reliable estimates of the scale-deposition rates (SDR's) 
(Baumgartner and others 1992). An overriding lesson from 
the Santa Barbara records is that in the past both sardines 
and anchovies experienced large natural fluctuations which 
were clearly unrelated to fishing, and that abrupt natural 
declines, similar to the collapse of the sardines during the 
1940's, are not uncommon. 

The scale-deposition record shows nine major recoveries 
and subsequent collapses of the sardine population over the 
past 1,700 years. The average time for a recovery of the 
sardine is 30 years. Sardines and anchovies both tend to vary 
over a period of approximately 60 years. In addition, the 
anchovies fluctuate at a period of 100 years. There is a 
moderate correlation between sardines and anchovies over 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



237 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



long time scales of several centuries or more, but the 
correlation of shorter-period components in the time series 
is virtually nil. 

Baumgartner and co-authors say that caution in 
interpreting the data should be exercised on two fronts: (1) 
sample size; they acknowledge that additional samples are 
needed to capture the complete range of variability of the 
SDR's over the basin. (2) The collapse and recovery 
demonstrated for the sardine do not necessarily mean that 
the current cycle of collapse and recovery has no relation to 
the application/release of fishing pressure or change in ocean 
climate, or both. They infer that even though the causes may 
vary (biological interaction, environmental change) for 
different recoveries or collapses, the sustained reproductive 
consequences are similar from one event to another 
(Baumgartner and others 1992). 

Analysis of fish scales in sediments of the central Gulf 
of California resulted in similarities with the Santa Barbara 
Basin work (Holmgren-Urba and Baumgartner 1993). The 
reconstructions show a strong negative association between 
the presence of sardines and anchovies, with anchovies 
dominating throughout the 19th century, and with only two 
important peaks of sardine scale deposition. The two episodes 
of sardine scale deposition occur virtually 1 80 degrees out 
of phase with anchovy scale deposition. This suggests an 
overall coherent pattern in changing ecosystem structure 
that operates over a period of about 120 to 140 years. The 
collapse of the sardine population in the Gulf of California 
was very similar to the collapse in the California Current 
during the late 1940's and 1950's. Both populations declined 
under heavy fishing pressure (Barnes and others 1992) 
superimposed on broad, natural, decadal-to-centennial-scale 
biomass fluctuations (Soutar and Isaacs in Holmgren-Urba 
and Baumgartner 1993). Both declines appear to be 
accompanied by an increasing population of northern anchovy 
(MacCall and Praeger in Holmgren-Urba and Baumgartner 
1993). The relationship to climate was not entirely clear, but 
suggested a mediating effect on population sizes. However, 
the process is still subject to strong filtering through biological 
interaction among species. 

Butler and others (1993) modeled anchovy and sardine 
populations to examine how natural variation of life-history 
parameters affected per capita growth. The greatest change 
in growth for both species occurred during larval stages. A 
number of important life history parameters of marine fish 
are directly affected by changes in temperature, and 
temperature and food densities affect growth at all stages. 
For anchovies, there is some evidence that reproduction is 
drastically reduced during major El Nino events. Under such 
conditions, the anchovy stock declined. For the sardine, high 
fishing mortality reduces the abundance of the oldest age 
classes, which have the highest reproductive potential because 
of their larger size and greater number of spawnings. Density- 
dependent factors such as cannibalism on eggs may also be 
important (Valdes and others and Valdes Szeinfeld in Butler 
and others 1993). The results of this modeling exercise 



parallel the results and conclusions of McGurk and Warburton 
(1992) described earlier in the section under sand lance. 

Structural changes over time in the California Current 
ecosystem between sardines and anchovies are similar to 
changes between herring and sand lance described previously 
for the North Sea and the Atlantic, though different factors 
were probably operative. Additionally, most researchers have 
found it difficult to separate the effects of humans from 
natural influences on the fish stocks. The fact that both 
mechanisms will continue to operate dictates that managers 
conduct effective monitoring programs and adaptive 
management to allow prompt remedial action to be taken 
where necessary (Wilson and others 1991). 

The low occurrence of sardines in the diet of murrelets 
is interesting given the wide geographic distribution of this 
fish (table 1). This low occurrence may be due to fewer 
studies in the southern end of the murrelet's geographic 
range where sardines are more abundant. Alternatively, it 
may represent an overall lower abundance due to overfishing, 
competition, and natural influences. Anderson and Anderson 
in Anderson and others (1980) suggested that past breeding 
populations of Brown Pelicans in the Southern California 
Bight probably had a larger prey base than the existing 
anchovy-dominated diet, perhaps also importantly involving 
Pacific sardines and Pacific mackerel. Recent increased 
abundance of sardines off southern California was followed 
by increased breeding success and abundance of Brown 
Pelicans (Ainley and Hunt in Anonymous 1993). 

Because of the natural fluctuations in anchovies and 
sardines as shown from the scale-deposition studies, murrelets 
probably evolved to use this resource in proportion to 
availability. Thus, the periodic lows in anchovy and sardine 
populations would probably not adversely affect the murrelet 
as long as alternative forage fish remained available. 
Development of new fisheries (sand lance or euphausiids) 
and escalation of harvests for rockfish and herring would be 
expected to affect murrelets, especially in conjunction with 
a low period of anchovies and sardines, and El Nino events. 

Pacific Herring 

Herring belong to the clupeidae as do the Pacific sardine. 
Adults range up to 45 cm in length (Miller and Lea 1972: 
54). Herring are one of the most abundant species of fishes 
in the world and prey upon copepods, pteropods, and other 
planktonic crustaceans, as well as fish larvae. They travel in 
vast schools, providing food for larger predators. 

The Pacific herring ranges from Baja California to Alaska 
and across the north Pacific to Japan. Within this range, 
abundance generally increases with latitude and the largest 
populations are centered off Canada and Alaska (Spratt 1 98 1 ). 

Currently, all herring commercially harvested in California 
and Oregon are taken as sac -row for Japanese markets. In 
British Columbia and Alaska, herring are primarily harvested 
for sac -row, and as longline bait (McAllister, pers. comm.). 

Spawning begins during November in California and 
ends during June in Alaska, becoming progressively later 



238 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkert 



Chapter 22 



Food Habits and Prey Ecology 



from south to north. During the spawning season, herring 
congregate in dense schools and migrate inshore where they 
deposit their sticky eggs on vegetation found in intertidal 
and shallow subtidal areas of bays and estuaries. The eggs 
hatch in about 2 weeks. After spawning, herring return to the 
open ocean where their movements are largely unknown 
(Spratt 1981). The large herring fisheries are subject to great 
fluctuations in their annual catches because the survival of 
young herrings varies widely from year to year, with a heavy 
dependence on copepods (Hardy 1965: 62). The fish mature 
in about 4 years and may live 20 years. 

Information on the age structure of spawning herring 
was analyzed by Lambert (1987). He noted that it is an 
underappreciated fact that herring often arrive at spawning 
grounds in runs or waves. This phenomenon has been reported 
in both the Atlantic and Pacific Oceans for C.h. harengus 
and C.h. palasii. It is suggested that spawning proceeds 
consecutively through year classes from oldest to youngest 
due to differential maturation. Discrete batches of eggs 
deposited by these waves of spawning herring give rise to a 
succession of larval cohorts. The more age classes involved 
in spawning, the longer will be the spawning season and the 
spawning will be more widespread since different age groups 
tend to spawn in different areas. Therefore, it would appear 
that the maintenance of a wide, well-balanced age structure 
tends to promote a resilient or more stable population 
(Lambert 1987). 

Near the Queen Charlotte Islands in British Columbia, 
shoals of immature herring occur frequently at the surface, 
where they often jump clear, making a calm sea suddenly 
erupt in a tiny "boil." Herring boils are often associated with 
swarms of euphausiids which provide food for the herrings 
(Gaston 1992: 74). Many seabird species will be found 
feeding at such prey concentrations. 

In his chapter on the herring. Hardy (1965: 61) wrote: 
"Early in the year, in March and April, the North Sea herring 
is feeding very largely on young sand-eels [Ammodytes sp.]; 
and often at this season you will find the stomach of the 
herring crammed full of them, lying neady side by side like 
sardines in a tin." 

McGurk and Warburton (1992) found that herring and 
sand lance larvae consumed prey of similar lengths and 
widths. They concluded that herring and sand lance larvae 
compete for substantially the same prey resource. More than 
99 percent of the prey items found in the guts of sand lance 
larvae were various life history stages of copepods (McGurk 
and Warburton 1992). 

The work of Carter (1984) and Vermeer (1992) 
indicated the importance of herring in the diet of murrelets 
{table 1). Lid ( 1 98 1 ) suggested that the breeding failure of 
Puffins [Atlantic Puffins] (Fratercula arctica) in Norway 
was due to over-harvesting of herring and. to some extent, 
over-fishing of sand eels (Ammodytes sp.). Many puffin 
chicks died, and adult weights were lower during the study 
period. Spawning stock size in weight-of the Norwegian 
spring-spawning herring declined frorr^approximately 9.5 



million tons to less than 0.5 million tons between 1950 
and 1980 (Lid 1981). 

Commercial fishing harvest of herring should be 
monitored for effects on murrelet reproductive success. In 
the absence of a sand lance fishery on the west coast of 
North America, it may be that sand lance populations will 
respond positively to reduction in herring as documented 
elsewhere. However, murrelet use of either of these resources 
will depend on temporal and spatial distribution of the prey 
relative to murrelet nesting and foraging habitat. The patchy 
distribution of prey during different seasons must be 
considered along with changes in offshore distribution of the 
murrelet between seasons. 

Smelt 

The osmeridae are closely related to salmon and trout, 
and like trout, have a small, adipose fin. They are confined 
to arctic and north temperate waters and are best represented 
in the north Pacific basin. All spawn in fresh water or along 
the seashore (Hart and McHugh 1944). Among related Pacific 
species are the surf smelt or silver smelt (Hypomesus 
pretiosus), capelin, and eulachon or candlefish. 

The silversides (atherinidae) and other unrelated fishes 
are sometimes also called smelts, sand smelt, or whitebait. 
The atherinidae also include grunion (Leuresthes tenuis) 
which occurs north only to the San Francisco area. 

The eulachon has been called candlefish because the 
flesh is so oily that the dried fish, when provided with a wick 
of rush-pith or strip from the inner bark of cedar, burns with 
a steady flame and was used as a candle by the natives. This 
fish gave rise to the famous "grease trails" which roughly 
follow the courses of the great northern rivers (Hart and 
McHugh 1944). The only record of eulachon in murrelet diet 
was the anecdote by Grinnell (1897) described previously 
under the section on Alaska. Eulachon are distributed from 
northern California to the Bering Sea. They seem to feed 
primarily on euphausiids. Eulachon are important as an 
intermediate step in the food chain between the euphausiids 
and larger fish (Hart and McHugh 1944). 

The range of the silver smelt extends from southern 
Alaska to central California. Some of the smelt may spawn 
at the end of the first year as has been indicated for Puget 
Sound fish. They spawn under a great variety of conditions 
and in most months of the year. Summer spawnings take 
place both on exposed beaches and at the head of sheltered 
bays. Usually the fish spawn where there is a certain amount 
of seepage of fresh water through the fine gravel to which 
the eggs adhere. Euphausiids seem to be the main food item 
consumed by silver smelt (Hart and McHugh 1944). 

The capelin is an arctic species with its center of 
abundance in the Bering Sea or Arctic Ocean (Hart and 
McHugh 1944). In the Pacific, capelin occur from Alaska to 
Juan de Fuca Strait. Their distribution in the coastal zone 
varies seasonally but peaks in June and July when beach 
spawning occurs. At other times of the year, capelin can be 
found in large concentrations in the offshore waters (Jangaard 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



239 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



in Carscadden 1984). At spawning time, capelin appear in 
schools of considerable size along the shores of gravelly 
beaches. Spawning occurs in the evening at high tide right at 
the water's edge. Studies of the beach both during and after 
spawning indicate that a specific type of ground is selected, 
the fish tending to avoid both rocky and sandy patches. The 
eggs are extremely adhesive and immediately become firmly 
cemented to the gravel (Hart and McHugh 1944). 

Capelin mature at 3 or 4 years of age with faster growing 
fish maturing earlier (Winters in Carscadden 1984). In the 
spawning populations, 3- and 4-year-olds usually predominate. 
Spawning mortality is high, usually greater than 80 percent 
(Carscadden and Miller in Carscadden 1984). 

Like other pelagic fish species, capelin populations exhibit 
large variations in abundance of year classes, and natural 
fluctuations in abundance are often complicated by the 
presence of fishing mortality. Carscadden (1984) evaluated 
fluctuations in capelin biomass in the northwest Atlantic and 
concluded they were the result of natural variation in year- 
class strength. The causes of the variation were not well 
understood, but temperature and onshore wind-induced wave 
action have been correlated with emergence of larval capelin 
(Frank and Leggett in Carscadden 1 984). 

Carscadden (1984) considered the relationship between 
Atlantic Puffins and capelin as described by Brown and 
Nettleship (1984) and concluded that a complex of natural 
environmental and biological factors would probably affect 
the abundance and behavior of capelin predators, rather than 
a single one such as abundance of capelin. Brown and 
Nettleship (1984) concluded that the management of the 
capelin fishery in the northwest Atlantic should "proceed 
cautiously" until the relationships between the capelin and 
its predators were better understood. 

Vader and others (1990) evaluated the relationship 
between Common Murres, Thick-billed Murres (U. lomvia), 
and capelin in Norway. A complete collapse of the Barents 
Sea stock of capelin occurred between 1985 and 1987, and 
in 1987 fishermen noted a near-complete absence of sand 
lance. The low sand lance population resulted in a complete 
breeding failure of Shags (Phalacrocorax aristotelis) in West 
Finnmark, where Shags are normally totally dependent on 
sand lance during the breeding season. A sudden drop in 
breeding Common Murres also occurred in 1987. The authors 
concluded that the capelin and sand lance food shortage 
caused the large drop in Common Murres and the reduced 
breeding of Thick-billed Murres. The authors thought the 
larger prey spectrum utilized by the Thick-billed Murres 
allowed that population to fare better than the Common 
Murres in the face of the food shortage. The causes of the 
decline in capelins probably included overfishing, 
uncommonly large year-classes of the predatory cod (Gadus 
morhua), and a reduction in recruitment due to changes in 
the physical oceanography of the Barents Sea (Hamre; 
Ushakov and Ozhigin in Vader and others 1990). 

The importance of capelin in the diet of the murrelet in 
the Gulf of Alaska (Sanger 1983) indicates the need to 



monitor and manage carefully this resource. Other smelt 
species may be important in murrelet diet; unidentified 
osmerids have been documented as murrelet prey over a 
broad geographic range {table 1). Further research is needed 
on the importance of smelt in the diet of the murrelet, 
especially in Washington, Oregon, and California. 

Prey Ecology Summary 

The marine environment, especially in an eastern boundary 
current system, is not static (Ainley and Boekelheide 1990: 
376). In his book on the Ancient Murrelet, Gaston (1992: 74) 
wrote of a diagram of the food web of Reef Island: "A 
complete diagram of the food webs of Hecate Strait would 
probably cover a baseball field at this scale, and would take 
several lifetimes of research to construct." Sanger's (1983) 
compilation contains numerous food web diagrams which 
depict the complex interactions in the marine environment. A 
food web and a model of the trophic-level interactions 
influencing murrelets at any site in North America would be 
complex indeed, but much information on life history of prey 
species and the murrelet at sea must be gathered. 

From the studies discussed above, some variability in 
reproductive success of the murrelet can be expected because 
of the naturally dynamic nature of their prey base and the 
marine environment. Anthropogenic influences can compound 
prey fluctuations; thus, marine research and management 
should be designed to minimize or avoid adverse changes in 
seabird reproduction and marine trophic-level interactions. 
Anthropogenic and environmental influences will continue 
to affect marine ecosystems. Management must therefore 
entail monitoring and the ability to change course in response 
to observed effects. Cumulative impacts in localized areas of 
murrelet abundance should be anticipated and averted. 

Size of Prey Items 

A compilation of prey item size in the diet of adult and 
subadult murrelets from systematic studies indicates the 
majority of fish taken ranged from 30. 1 to 60.0 mm {table 
4). The largest combined sample size was for sand lance, 
and the distribution indicated a heavy reliance on fish up to 
60.0 mm, although fish greater than 90.0 mm were also 
taken. Sanger (1987b) calculated a mean value of 45 mm 
(total length) for sand lance which correlates well with the 
distribution of prey size revealed in table 4. Smaller size 
classes (0.1-30.0 mm) of scorpaenids and Cymatogaster 
aggregata were taken by murrelets; this could be a function 
of availability or preference. Larval and juvenile fish (0.1- 
60.0 mm) appear to be the main size classes eaten by adult 
and subadult murrelets. Larval fish are underrepresented in 
murrelet diet because they are digested quickly (Carter 1984), 
therefore, the overall importance of larval fish for murrelets 
is difficult to assess. 

The size of prey items in the diet of hatching-year and 
nestling murrelets is markedly different (table 5) though a 
comparison of fish lengths in tables 4 and 5 reveals adult/ 
subadult and hatching-year murrelet prey size to be similar. 



240 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



Chapter 22 



Food Habits and Prey Ecology' 



Table 4 — Size of prey items for adult and subadult Marbled Murrelets 





Sample size 


Mean length 


Range or size class 


Pre> 


n 


mm 


mm 


Loligo opalescens 


5* 


- 


0.1-30.0 



Cymatogaster aggregata 




•Carter ( 1984): length for invertebrates is total length, and fork length for fish 
b Seal y ( 1 975c): length same as Carter ( 1984) except as noted for Loligo opalescens 
-"Sanger ( 1987): length for all specimens is total length 



0.1-30.0 
30.1-60.0 
60.1-90.0 



Tirfciritir 6» 30.1-60.0 


Ammodytes hexapterus 


I3F 


45 


29-135 




528* 


- 


0.1-30.0 




596* 


- 


30.1-«0.0 




88* 


- 


60.1-90.0 




6* 


- 


>90.0 



As noted by Carter (1984) and Mahon and others (1992), 
murrelet nestlings are fed much larger fish than the adults 
consume. Most nestling prey items were >60.1 mm, and 
sand lance prey were >90. 1 mm {table 5). 

Schweiger and Hourston in Carter (1984) concluded 
that second-year herring fed to nesdings were much less 
abundant than the juvenile herring that adult murrelets ate 
for themselves. Second-year sand lance and anchovy were 
also not considered very abundant in Carter's (1984) study 
area, which suggested that murrelets selected larger prey to 
carry to nestlings, even though such fish were less abundant. 
This behavior is consistent with optimal foraging theory 
(Carter and Sealy 1990). and other seabirds have exhibited 



this same adaptive trait (Gaston and Netdeship; and Slater 
and Slater in Carter 1984). 

The lengths of nestling prey probably represent second- 
year fish (Hart in Carter 1984), thus, murrelet adults, subadults, 
and hatching-year birds feed primarily on larval and juvenile 
fish, whereas nestlings are most commonly fed second-year 
fish. Therefore, both of these cohorts of the principal prey 
species should be monitored and managed to assure maximum 
productivity of murrelets in any one year. 

Energetics and Energy Values of Some Prey Items 

Energy values of prey items also help explain why 
murrelets select certain prey species for themselves and their 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



241 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



Table 5— Size of prey items for hatching-year and nestling Marbled Murrelets 



Prey 



Hatching-year prey size 



Nestling prey size 



ea harengus 




Engraulis mordax 

Mallotus villosus 
Ammodytes hexapterus 



Sample size 
n 



3» 

38" 

7 a 



2 e 
25 a 

2 e 



Range 
mm 



0.1-30.0 
30.1-60.0 
.1-90.0 



Sample size 


Length of specimen or range 


n 


mm 


16 a 


60.1-120.0 b 



2" 
l c 



0.1-30.0 
30.1-60.0 
60.1-90.0 



70" 
Unknown' 



90.1-120.0 b 
113 




Cymatogaster aggregau 
Scorpaenidae 
Unidentified fish 



a Carter (1984); 16 June - 6 July 1980, n = 144 fish observed. 

b Sizes of prey estimated while held by murrelets in their bills when on the water 

c Ralph and others (1990); observation during mist-netting operation, 3 July 1989. 

d Simons (1980); observation of a feeding at a ground nest. 

e Sealy (1975c); from 6 newly-fledged murrelets collected between 10 July and 4 August 1971. 

f Mahon and others (1992); murrelets observed on the water in the evenings, 6 June - 8 August 1991 . 



90.1-120.0 b 
140-180 b 



nestlings. Energy values of seabird prey items have been 
little studied (Hislop and others 1991), but a gross comparison 
from some closely related species reveals marked differences 
in food energy, protein, and total lipids {table 6). Especially 
when considering the feeding of nestlings at inland sites, 
optimal foraging theory would predict that the largest and 
most energy -rich food items would be brought to the nestlings. 
This would be adaptive by reducing energy demand on the 
adults, and by increasing the chances of a successful fledging. 
However, prey availability and competition with other seabirds 
also affects prey selection. The small size of the murrelet 
also limits its prey load, and a long flight time inland with a 
heavy prey load would be energetically costly and would 
subject the bird to an increased period of vulnerability to 
inland predation. Prey also loses water during transport by 
seabirds. Montevecchi and Piatt in Hislop and others (1991) 
simulated transport of capelin by tying fish to a drying rack 
mounted on a pick-up truck which was driven at 60 km/h. 
After one hour, weight loss averaged 9 percent for male 
capelin and 1 1.5 percent for females. 

A detailed analysis of variation in the calorific value 
and total energy content of the lesser sand eel (A marinus) 



and other fish preyed on by seabirds was conducted in north 
Scottish waters by Hislop and others (1991). They found the 
calorific values and body weights of sand eels larger than 10 
cm showed marked seasonal trends, and thus the total energy 
content of a sand eel of given length in summer was 
approximately double the spring value. Calorific values of 
Atlantic herring also varied from month to month, but seasonal 
cycles were less obvious. Seasonal cycles in fat content and, 
consequently, in calorific value are generally associated with 
the annual reproductive and feeding cycles of the fish, and 
tend to be greater among the larger, mature members of the 
population. Since different species offish spawn at different 
times, their condition cycles are out of phase to some extent. 
And, since herring spawn in different waves, their condition 
is not uniform at any one point in time. 

Hislop and others (1991) concluded that because fish 
demonstrate intraspecific length-related and seasonal changes 
in calorific value and energy content, it is unwise to generalize 
about the relative food values of different prey species to 
predators. They noted that sand eels have maximum calorific 
values intermediate between those of gadoids and clupeoids. 
Of interest, Hislop and others also noted that juvenile sand 



242 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkett 



Chapter 22 
Table 6 — Mean food values of selected invertebrates and fishes 1 



Food Habits and Prey Ecology 



Prey 



squid 



Food Energy 



Protein 



Total Lipids [fat] 



kcal 



Mixed species 



Shrimp 

Mixed species 

(Penaeidae and Pandalidae) 



15.58 



138 



106 



20 - ; I 



.73 




Mixed species 
fSinhmri <pp ■ 



1 Amounts in 100-g raw samples, edible portion (Exler 1987) 






eels (<10 cm) which have low body weights and high water 
content seemed, on purely energetic grounds, to be low- 
quality food. Because many different seabird species use 
sand lance in their diet, it may be that the overall abundance 
and availability of these fish compensates for the low energy 
value. Sand lance may also contain essential nutrients which 
seabirds have a need for. and the higher water content may 
also be important physiologically. The estimation of total 
energy content is complicated by dehydration of fish 
specimens: thus, Montevecchi and Piatt in Hislop and others 
( 1 99 1 ) urged seabird biologists to compare dry weight energy 
densities across studies. Both sets of researchers also noted 
the value of including wet calorific values as well. 

The work of Hislop and others (1991) provides data for 
comparison of energy values between sand eel and herring 
which indicates that herring have much higher total energy 
value than sand eel (table 7). Unfortunately, there is no data 
available for both herring and sand eel of murrelet nestling 
prey size (60. 1-120.0 mm) in July or August to allow a more 
relevant comparison. 

Roby (1991) studied the diet and postnatal energetics in 
three species of plankton-feeding seabirds. Lipid-rich diets 
were associated with shorter brooding periods, higher rates 
of nestling fat deposition, and larger lipid reserves at fledging. 
The energy cost of growth was a relatively minor component 
of nestling energy budgets; most assimilated energy was 
allocated toward maintenance and fat deposition. Once growth 
requirements for protein had been met, any additional 
assimilated protein was metabolized to meet maintenance 



costs, and the energy saved was stored as fat. High lipid diets 
were associated with higher rates of lipid deposition by chicks, 
but not higher growth rates. Instead, constraints operating at 
the level of tissues are apparently responsible for most of the 
variation in growth rate among seabirds. Large lipid reserves 
at fledging presumably enhance post-fledging survival. 

Discussion 

Conservation and recovery of the murrelet will depend 
in part on a better understanding of the interaction between 
the factors affecting the species inland and at sea. The 
studies described in this chapter have shed some light on 
this relationship and have indicated the need for 
comprehensive management of marine resources and inland 
nesting habitat. 

There is a need for additional study of murrelet diet, 
especially in the southern end of its range. Winter diet 
studies are also needed to help understand why some murrelet 
populations disperse to other locales during the non-breeding 
season. Comparison of prey abundance and composition 
between breeding and non-breeding foraging areas may help 
explain these movements. 

Additionally, more research on the use of inland lakes 
and estuaries as foraging sites is needed, across all seasons. 
This aspect of the murrelet' s life history has not received 
adequate attention except by Carter and Sealy (1986) and 
Hobson (1990). Though there are few large inland lakes in 
the coastal area of Washington, Oregon, and California, 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



243 



Burketl 



Chapter 22 



Food Habits and Prey Ecology 



Table 7— Calorific values (kj/g) and total energy content (kj) of Lesser sand eel (Ammodytes marinus) and 
Atlantic herring (Clupea harengus)' 











Dry 


Wet 




Total 






Mean 


Collection 


calorific 


calorific 


Wet 


energy 






length 


month 


value 


value 


weight 


content 


Fish 


cm 




U/g 


U/g 


g 


U 


Lesser sand eel 


16.5 


June 


25.8 


6.9 


13.5 


93.2 


Atlantic 


herring 


15.5 


August 


28.9 


7.5 


41.6 


312 



1 Data from Hislop and others ( 1991 ) 



there are numerous coastal lagoons and estuaries which may 
be important to murrelets. 

Intensive studies on factors affecting sand lance 
distribution and abundance are needed, as well as further 
exploration of the food web of this species. The importance 
of this little-studied fish in the diet of murrelets from Alaska 
to California certainly indicates a need for further investigation 
of predator-prey interactions. Monitoring of sand lance 
populations may prove useful for comparisons with murrelet 
population and productivity estimates from at-sea surveys. 
Sand lance recruitment could potentially serve as an indicator 
of murrelet reproductive success. A strong correlation was 
reported between number of tern chicks available for banding 
and recruitment of sand lance (Monaghan and others 1989). 
The effect of pollution and physical disturbance (dredging) 
on sand lance populations needs management attention (Auster 
and Stewart 1986, Nakata and others 1991, Pinto and others 
1984). Identification of sand lance spawning areas could aid 
conservation of the murrelet through directed management 
of these sites. 

The threatened and endangered status of the murrelet, 
coupled with the low productivity estimates, indicates the 
need for intensive field work in order to determine food 
habits without sacrificing birds. Long hours of observation 
of murrelets at sea catching and holding fish will be necessary, 
and intensive, systematic searches for beached birds could 
yield specimens for studies of food habits. Plankton hauls 
along with traditional methods of assessing marine fish can 
be used in areas where murrelets are actively foraging to at 
least determine prey abundance and composition. Video 
footage of prey items along with collections of fish parts 
from nest sites can contribute to knowledge of murrelet diet. 
Specimens can also be obtained from gill netting operations 
and oil spill events. Stomach pumping or emetics could 
possibly be employed, especially in conjunction with radio- 
telemetry studies and banding or marking operations. 

The identification of important foraging areas near known 
murrelet nesting sites will help in the conservation of this 
species (Ainley and others, this volume). Human activities or 
influences which are detrimental to the murrelets or their prey 
resources could then be appropriately managed in such areas. 



Clark and others ( 1 990) compared habitat structure and 
the number of active nests for Red-tailed Tropicbirds 
(Phaethon rubricauda) before and after an El Nino event. 
An increase in availability of quality habitat post-El Nino 
resulted in an increase in the number of active nest sites 
relative to pre-El Nino breeding seasons. The data of Clark 
and others supported the hypothesis that suitable nest sites 
may limit short-term reproductive opportunities of tropicbirds 
and, hence, influence the rate of population growth and time 
course of recovery from catastrophic events such as El Nino. 
This has important management implications for a threatened 
species such as the murrelet. Inland nesting habitat becomes 
a very important management consideration even though the 
murrelet relies on the marine environment for food. A study 
on the Red-throated Loon (Gavia stellata) indicated that as 
distance from the ocean to the nest site increased, both 
density and nesting success of loons decreased (Eberl and 
Pieman 1993). The authors suggested that the higher density 
of breeding loons in areas near the ocean reflected a preference 
by these birds for nesting grounds that are closer to their 
foraging areas. 

Since the murrelet is a forest-nesting seabird, it is 
imperative to consider multiple factors when devising research 
and management strategies. Because of its secretive nesting 
habits, it has been difficult to document nest success relative 
to prey abundance as has been done with other seabirds. 
Even if adult murrelets can easily choose alternate prey 
species for their own diet, having abundant forage fish 
available during the nestling period may significantly reduce 
the energy demand on the adults by requiring less foraging 
time and fewer trips inland for feeding nestlings (Carter 
1984, Carter and Sealy 1990, Cody 1973, Sealy 1975c). The 
juxtaposition of nesting areas and foraging areas is probably 
most critical as one determinant of reproductive success in 
years of low prey abundance. Increased foraging time of 
adults, long flights inland, and more numerous trips inland 
with small prey items would potentially reduce both adult 
and chick survival. Competition with other seabirds for 
available food is also an important factor in foraging patterns 
and prey selection (Ainley and Boekelheide 1990: 380; Cody 
1973; Mahon and others 1992). 



244 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Burkelt 



Chapter 22 



Food Habits and Prey Ecology 



Marine communities have been altered by the activities 
of humans in conjunction with natural influences (Caims 
1992a: 39). As early as 1886. declines in fish populations in 
heavily fished areas seemed apparent. The notes of an 
expedition to Puget Sound in July 1895 contained the 
following anecdote regarding herring: "Exceedingly abundant. 
J. P. Hammand (American Angler. December 18, 1886) states 
that from 1 8-25 years ago it was not an uncommon occurrence 
for a gang' of fishermen to catch from 200-300 barrels of 
herring in a night on Puget Sound. Now the largest night's 
work is 20 barrels" (Jordan and Starks 18%). The sardine 
fishery in California which was discussed above is another 
example. The relatively new fishery went from the "palmiest 
days" of Cannery Row in the mid- 1 930' s to a catastrophic 
drop in 1947. In response to this drastic decline, there emerged 
what is known today as the California Cooperative Oceanic 
Fisheries Investigations (CalCOFI) program to help better 
manage marine resources (Ricketts and Calvin 1962: 382). 
The observations of Radovich (1961) are helpful: "The mere 
fact one can demonstrate the environment has a large effect 
on the catch does not imply man's effect is inconsequential. 
To understand man's effect, one must study the effect of 
man. However, a satisfactory understanding may never be 
achieved so long as one fails to recognize the existence of 
some of the other factors constantly confusing his data. The 
effects of environment and man on fish populations are not 
mutually exclusive." The long history of fishing activity in 
the North Sea produced steep declines followed by increases 
when fishing pressure diminished. Such events resulted in 
Sir Alister Hardy's remark: "Certainly no one can deny that 
over-fishing exists: we must find the best way to remedy it" 
(Hardy 1965: 247-248). 

The scale-deposition studies described above provide 
evidence that abundance of coastal pelagic fish species varied 
considerably before the inception of modern fisheries. 
Environmental factors and trophic-level interactions contribute 
to the naturally dynamic state of marine ecosystems. Fishing 
has. however, probably exacerbated the natural variability in 
recent decades because reduced stock size and loss of old 
fish, which is an inescapable result of fishing, increase the 
speed and magnitude of population decreases during periods 
of poor reproduction. Approaches to fishery management 
based on equilibrium or steady-state concepts that ignore 
variability in abundance have a long history of failure for 
coastal pelagic species in many regions of the world (Troadec 
and others in Anonymous 1993). Managers should expect 
considerable interannual variation in abundance and yields 
and should curtail fisheries to protect the long-term health of 
the stock when necessary. Chaotic ecosystems appear to 
require reliance on management that is beneficially adaptive 
rather than manipulative. The possibility of detailed predictions 
is effectively ruled out, and many factors, including 
socioeconomic ones, must be used when modeling populations, 
ecosystems, and fishery impacts (Wilson and others 1991). 

Throughout its range, the murrelet consumes a very 
diverse group of prey resources, especially when one considers 



the limited studies which have been done to date. This 
indicates great flexibility in prey choice and a high capability 
for prey-switching behavior. This would make adaptive sense 
given the multiple factors affecting prey availability each 
year and the oceanographic differences found offshore from 
forest nesting habitat throughout the range of the species. It 
also indicates that El Nino events would not be expected to 
cause catastrophic population fluctuations or declines, 
especially in the long term. Given the variability in frequency 
and intensity of El Nino events, murrelet production could 
be lower than "normal" in some years as has been 
demonstrated for many other seabirds. But. like other seabirds, 
the murrelet has evolved with this phenomenon and can 
likely change its foraging behavior and food preferences to 
some degree in order to utilize available resources (Carter 
1984, Croll 1990, Krasnow and Sanger 1982, Sanger 1987b, 
Sealy 1975c). Additionally, the long life span of the species 
allows for adequate reproduction and dynamic equilibrium 
of the population, even in the face of low reproduction in 
some years. However, cumulative impacts in localized areas 
over a short time period could cause serious population 
declines or possibly even extirpations. 

Research should continue to identify bottlenecks to 
recovery; "scientifically approachable" and "practically 
realizable" studies should be done along with attempts at 
"integrated management of the marine ecosystem as a whole" 
(Holt 1993). A lack of information on the functioning of 
"natural systems" (Willers 1993) should not prevent 
comprehensive research or recovery actions in the future, 
but instead should help guide more unified study efforts. 
Biologists have long recognized the need to integrate seabird 
and marine science (Ainley and Sanger 1979, Furness 1984, 
Munro and Clemens 1931, Sealy 1990, and others) and the 
excellent treatise on the matter by Cairns (1992b) should 
help guide marine ecosystem research and management in 
the future. 

Managers and researchers today are faced with the listed 
or sensitive management status of the murrelet and limited 
financial resources to conduct the necessary studies. It is 
now more important than ever to pool resources and seek 
innovative ways to conduct the necessary research. Mitigation 
banking policies imposed on commercial fishing and timber 
industries, coupled with damage assessment rewards, could 
help gather research funds and support the large-scale studies 
proposed by Nisbet (1979) and Vermeer (1992). Research 
monies alone are not recognized as adequate mitigation for 
negative impacts to natural resources, but funds derived 
from such policies could certainly play a stronger role in the 
conservation and recovery of the murrelet than has occurred 
up to this point 

These ideas are not new (Drury 1979, Nisbet 1979), but 
implementation has yet to occur on a meaningful scale. In 
the words of Drury (1979: 136): "Experience in Europe and 
in New England suggests that if reasonable limitations are 
set on human activities and that if adequate money charge is 
made against those who profit by economic development to 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



245 



Burkett 



Chapter 22 



Food Habits and Prey Ecology 



defray full social costs, wildlife can continue to do well. In 
most cases where damage has occurred it is because those 
who administer the public institutions have failed to include 
consideration of the common property resources". 

Acknowledgments 

I thank my friends and colleagues Jack Fancher and Bob 
Hoffman for involving me in fish assessment work, and 
Charlie Collins and Mike Horn for stimulating my interest in 
seabird feeding ecology during the Bolsa Chica days. Special 
thanks to Charlie for being such a good teacher and enthusiastic 
field partner. Thanks also to Dick Zembal for involving me 
in a project on a coastal wetland food web; that project 
cemented my interest in predator-prey relationships. Patty 



Wolf was especially helpful in providing much needed 
literature and background information from all the work 
done on the Pacific Coastal Pelagic Species Fishery 
Management Plan. I thank C.J. Ralph, Mike McAllister, 
George Hunt, and an anonymous reviewer for their 
encouragement and helpful comments on an earlier version 
of this manuscript. I especially appreciate George's interest 
in the energetic aspects of seabird ecology. The dedicated 
work of Harry Carter and Gerry Sanger provided inspiration. 
They also provided much help and guidance. This project 
would not have been possible without the capable assistance 
of Vikki Avara who searched for and obtained the copious 
and obscure literature upon my sometimes fickle demands. 
Thanks also to Shannon Sayre and Heather Johnson for 
assistance with the tables. 



246 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Chapter 23 

Marbled Murrelet At-Sea and Foraging Behavior 



Gary Strachan' 



Michael McAllister 2 C. John Ralph 3 



Abstract: The behavior of Marbled Murrelets (Brachyramphus 
marmoratus) at sea while foraging for small fish and inverte- 
brates is poorly known. This murrelet forages by pursuit diving 
in relatively shallow waters, usually between 20 and 80 meters in 
depth. We have also observed it diving in waters less than 1 
meter and more than 100 meters deep. The majority of birds are 
found as pairs or as singles in a band about 300 to 2000 meters 
from shore. Pairs tend to dive simultaneously when foraging, and 
we suggest that pairing has some benefit to foraging efficiency. 
At times they are in small flocks and in aggregations. Larger 
aggregations are found in the northern part of its range, probably 
due to the denser population. Foraging dive times averaged about 
16 seconds. Murrelets generally forage during the day, and are 
most actively in the morning and late afternoon hours. Some 
foraging occurs at night. Vocalizations during foraging occur 
after individuals of a pair surface apart from each other. The 
majority of the birds" surface time is spent loafing, preening, and 
wing stretching. We feel that adults holding fish are usually 
about to depart inland to feed a young, and are potentially a very 
useful measure of reproductive rate. Murrelets are not generally 
associated with interspecific feeding flocks, except in the north- 
ern part of its range. 



The at-sea behavior of the Marbled Murrelet {Brachy- 
ramphus marmoratus) is relatively little known, with the 
exception of the work of Carter and Sealy (1990). Under- 
standing the relationship between the species, its foraging 
habitat, and its prey species are important so that appropriate 
decisions are made concerning future recovery efforts. We 
have spent many thousands of hours observing murrelets 
on the ocean and this paper brings together these 
observations, contributions from colleagues, and the 
published literature, to give a perspective on the life history 
of the species in its marine environment. 

Foraging Range 

Nearshore feeding — During the breeding season, the 
Marbled Murrelet tends to forage in well-defined areas along 
the coast in relatively shallow marine waters (Carter and 
Seals 1 990 ). Part of their distribution is related to availability 
of nesting habitat, as discussed in other chapters in this 
volume. Murrelets generally forage within 2 km of the shore 
in relatively shallow waters in Washington, Oregon, and 



1 Supervising Ranger. Ano Nuevo Stale Reserve, New Year's Creek 
Road. Pescadero. CA 94060 

: Wildlife Biologist. Wildland Resources Enterprises. 60069 Morgan 
Lake Road. La Grande. OR 97850 

-' Research Wildlife Biologist. Pacific Southwest Research Station. 
USDA Forest Service. Redwood Sciences Laboratory. 1 700 Bayview Drive. 
Areata. CA 95521 



California. The species does occur farther offshore than 2 
km (Carter, pers. comm.; Piatt and Naslund. this volume; 
Ralph and Miller, this volume; Sealy 1975a). but in much 
reduced numbers. Ainley and others (this volume) reported a 
few murrelets up to 24 km offshore in central California. 
Their offshore occurrence is probably related to current 
up welling and plumes during certain times of the year (Hunt 
this volume a). Off Alaska and British Columbia, the bird 
occurs more frequently further offshore; they occur quite 
regularly out 40 km in the Gulf of Alaska in the relatively 
shallow waters of that region (Piatt and Naslund, this volume; 
McAllister, unpubl. data). During the non-breeding season, 
murrelets disperse and can be found farther from shore, as is 
the case with some other alcids. 

Murrelet prey species mostly include small inshore fish 
and invertebrate species such as sand lance {Ammodytes 
hexapterus), smelt (Hypomesus spp.). Pacific herring (Clupea 
spp.), capelin (Mallotus spp.). and various other fish (Burkett, 
this volume). Invertebrates such as Euphausia pacifica and 
Thysanoessa spinifera are also important prey (Sanger 1987b. 
Sealy 1975a). 

Winter distribution — In some locations, after the 
breeding season, birds appear to disperse, and are less 
concentrated in the immediate nearshore coastal waters. 
This has been observed in Ano Nuevo Bay in central 
California {fig. /), as birds move away from this protected 
bay from November through April. Similar movements have 
been observed in Clarence Strait in Southeast Alaska 
(McAllister, unpubl. data), where the birds are greatly reduced 
in numbers and probably have moved to the south. In the 
southern portion of their range, murrelets are reported in 
winter as far south in central California as San Luis Obispo 
County, and at times to the southern portion of the state. In 
many areas, however, individuals maintain an association 
with the inland nesting habitats during the winter months 
(Carter and Erickson 1988). 

Fresh water lake use — Carter and Sealy (1986) found 
67 records of birds on 33 fresh water lakes; 78.6 percent of 
those recorded were in British Columbia, 12.1 percent in 
Alaska, 6. 1 percent in Washington, and 3 percent in Oregon. 
Foraging on lakes had been suspected because salmon fry, 
fingerlings, and yearlings that have been found in birds' 
stomachs (Carter and Sealy 1986). A few observations of 
birds presumably feeding in lakes have been recorded (Munro 
1924, Carter and Sealy 1986). Carter and Sealy (1986) 
speculated that murrelets feed at night on these lakes when 
fish are available closer to the surface. Hobson ( 1 990) found 
evidence, based on isotope analysis of murrelet muscle 
tissue, that birds collected on Johnston Lake. British 
Columbia, may feed in fresh water lakes for several weeks 
at a time. 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



24" 



Strachan and others 



Chapter 23 



At-Sea and Foraging Behavior 




Figure 1— Average number of Marbled Murrelets/census by biweekly periods on Ano Nuevo Bay during 1990. Figure shows 
the mean, standard error, and minimum and maximum values of from-shore censuses, n = number of censuses/time period. 



Foraging Behavior 

Pairing and Group Size 

Frequency of pairs — Murrelets forage mostly in pairs 
throughout the year. This is an important aspect of their life 
at sea, as we have often observed murrelets vocalizing on the 
water while foraging, apparently attempting to locate the 
other member of a pair when coming to the surface, or after 
a disturbance. The call usually used is the typical "keer" in 
rapid succession or singly. The percentage of birds in an area 
that are foraging or loafing in pairs varies, but not greatly. 
Mean group size from Oregon was 1.8 birds, with about 70 
to 80 percent of the birds observed in pairs (Nelson 1990). 
Along the central Oregon coast, Strong and others (1993) 
observed that murrelets almost always occurred as single 
birds or pairs. In Alaska, pairs made up 45 percent of the 
population (Kuletz, pers. coram.). Carter and Sealy (1990) 
found in Trevor Channel, Alaska, that pairs were 40 percent 
of the birds seen. During the summer of 1993, Ralph and 
Long (this volume) reported 63 percent of groups were pairs 
and 27 percent were single birds in northern California. 

In central California, 75 to 80 percent of birds foraged 
as pairs during the breeding season (fig. 2). Single birds are 
more common in the winter, when the populations are low at 
this location (fig. 2). Sealy (1975c) suggested that, during 
the incubation period, a daily pairing of birds occurred as 
birds flew around in the forested nesting area after an 
incubation exchange. We have observed many single birds 
circling and calling at inland sites until joined by a second 
bird, when both headed west to the ocean. We have also 
observed at times many hundreds of birds arriving at the 
ocean in the morning from inland nesting sites, usually in 



pairs, threes, or fours. Observations at the nest would suggest 
that the birds should arrive singly (Naslund 1993a; Nelson 
and Hamer, this volume a), as pair members are rarely at the 
nest simultaneously, which might suggest that the birds pair 
with non-mates enroute to the sea. 

Composition of pairs — In British Columbia, Sealy 
(1975c) found that 11 out of 13 pairs collected in late April 
were composed of an adult male and adult female. After egg 
laying occurred, more single "off duty" birds were encountered 
at sea. He surmised that both adults stay together during the 
day and returned to the nest site at night to feed their chick. 
The subadults (birds one or two years old who have not yet 
bred, as determined by collecting) also returned in late April, 
but were encountered only as single individuals until late 
June and early July when mixed groups of "off duty" adults 
and subadults, were observed. During late July newly fledged 
young were frequently seen in these groups. 

Reason for foraging in pairs — Sealy (1975c) stated "I 
believe that the occurrence of these pairs can be adequately 
explained on the basis of pair bond maintenance and that an 
advantage to feeding need not be involved." Possible 
evidence of pair bonding is found in observations of pairs 
separated by boats. Ralph (unpubl. data) and Miller (pers. 
comm.) have noted that about two-thirds of these pairs call 
and attempt to reunite, while the remaining birds simply 
disperse. However, we feel that foraging plays the major 
role in pairing, and probably involves some sort of cooperative 
foraging technique. Evidence of this includes the observation 
that the vast majority of actively foraging paired murrelets 
consistently dive together (Carter and Sealy 1990). Laing 
(1925) stated that the "birds of this genus work in winter 
and summer in pairs, but not as a defensive measure, for 



248 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Strachan and others 



Chapter 23 



At-Sea and Foraging Behavior 



100 



co 80 
a. 

3 

o 

§ 60 



LU 
O 

rr 

UJ 

a. 



40- 




PAIRS 




SON 
702 451 115 



D 
15 



Figure 2 — Group size of Marbled Murrelets by month on Ano Nuevo Bay during 1990, detailing 
percentage of groups observed consisting of single birds, pairs, or groups of >3 birds, n = number 
of groups observed. 



they dive almost together". Carter and Sealy (1990) reported 
that pairs were often seen swimming towards each other 
before diving, and that three or more birds never dove 
together in a coordinated fashion. They also stated that 
foraging by singles and pairs may prevent foraging 
interference, competition, and kleptoparasitism that would 
be more likely in foraging flocks. 

Flock size and frequency — Carter and Sealy (1990) sug- 
gested that murrelets are most aggregated during the nesting 
period. Aggregations of large numbers have been reported in 
the northern range (Carter 1984; Carter and Sealy 1990; 
Hunt. pers. comm.; McAllister, unpubl. data). Foraging 
aggregations were probably related to concentrations of prey. 
McAllister (unpubl. data) observed an aggregation of 4,000 
to 6.000 individuals at Point Adolphus on Icy Strait, in 
southeast Alaska, on 3 May 1991. 

Observers have noted great variation in size of flocks 
(defined as three or more birds in close proximity and 
maintaining that formation when moving). In southeast Alaska, 
Quinlan and Hughes (1984) reported flock sizes up to 50 
birds in Kelp Bay. Kuletz (1991a) found in another Alaskan 
population that flock sizes greater than three birds made up 
about 8 percent of the birds, 7 percent of the birds were 
found in groups of four birds, 3 percent of the birds in groups 
of five, and 1 percent were found in groups larger than five. 
The largest number in a concentrated flock was 22 birds. In 
British Columbia, Carter (1984) found larger, non-feeding 
flocks of up to 55 birds. The larger flocks usually occur 
during the later part of the breeding season, and may be 
made up of juveniles and subadults. Sealy (1975c) found 
that flocks would feed together at Langara Island, British 
Columbia, with the mean flock size of eight. 

Flock sizes in the southern populations of California. 
Oregon, and Washington, rarely number more than 10, 



according to our and others' observations. Nelson (pers. 
comm.) recorded groups greater than 3 as very uncommon in 
Oregon, with a maximum of 10 birds in a flock. Also in 
Oregon, Strong (pers. comm.), found similar flock sizes 
during his 1992 study. The largest flock that he observed 
was 15 birds. In California, Ralph and Long (this volume) 
found two was the most frequent group size (63 percent), 
while less than 10 percent of flocks contained more than 
three birds. The largest flock seen was 12 birds at Santa 
Cruz. At Afio Nuevo Bay, in central California, flocks are 
similar in size {fig. 2). Here, at the southern end of the 
species' range, during late summer and early fall, flocks of 
over three would often contain juvenile birds. Groups of 
three or more were found during the summer, when the 
population is highest (fig. 1), and may be a function of 
density, rather than flocking. 

Behavior in flocks — Sealy (1975c) observed that the 
flocks would tend to dive against the current, and soon 
become spaced in a linear fashion with the main axis of the 
flocks paralleling the direction of the current. Carter and 
Sealy (1990) observed that larger flocks do not appear to be 
foraging. Sealy (1975a) stated that birds foraging during the 
breeding season "invariably occur in pairs or as single 
individuals." Early in spring adults feed in pairs while the 
subadults feed singly, but in early July, when pairs are still 
feeding young at the nest, mixed flocks of adults and subadults 
begin to form. 

Foraging of Juveniles 

When the first juveniles reach the water during the 
breeding season, usually by early July (Hamer and Nelson, 
this volume a), they are distinctive in plumage from adults, 
making identification of individuals in a small flock possible 
(Carter and Stein, this volume). From this we can learn 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



249 



Strachan and others 



Chapter 23 



At-Sea and Foraging Behavior 



about the need for foraging in groups. McAllister (unpubl. 
data) found in Alaska, between mid-July and mid-August, 
that more than 80 percent of the young were observed without 
adults present. By contrast, in California, Ralph and Long 
(this volume) observed that half of the juveniles observed 
were accompanied by one or more adults, while the remaining 
juveniles occurred alone. 

By mid-August, it becomes difficult to differentiate 
juvenal plumage from molting adults. In our observations of 
juveniles on the water, we found that juveniles foraged 
without the assistance of the adults. They were seen as single 
birds, in pairs, and in small flocks. The largest flock was one 
of 12 young seen together in Peril Strait (McAllister, unpubl. 
data). Juveniles were most common within 100 m of 
shorelines, particularly where bull kelp (Nereocystis spp.) is 
present (McAllister, unpubl. data). At this time of year, 
adults were generally farther from shore in this area, at the 
sharp tidal interfaces, e.g. rips. However, in a 1993 study, 
Ralph and Long (this volume) found no difference between 
the distribution of adults versus juveniles in California. 

Behaviorally, the fledglings are generally less wary, 
more curious, and much more approachable by boat. In 
flight, they are weak and slow (McAllister, unpubl. data), as 
compared to adults. 

Interspecific Relations During Foraging 

In the southern part of the range, from Washington 
south, murrelets rarely forage in mixed seabird flocks. Pairs 
or small flocks will usually forage away from other species. 
In California and Oregon, murrelets have been reported 
foraging close to Pigeon Guillemots (Cepphus columba) and 
Common Murres (Uria aalge), but seldom within any major 
mixed species flocks. Murrelets have been observed by Strong 
and others (1993) to avoid large feeding flocks of murres, 
cormorants (Phalacrocorax spp.), and other species in Oregon. 
He presumed that the small size of the murrelet may render 
them vulnerable to kleptoparasitism or predation in mixed 
species flocks. In addition, if the murrelets forage in some 
cooperative effort, the confusion of a large flock of birds 
might reduce foraging efficiency. 

In the northern part of the range of the murrelet, from 
Puget Sound north, the literature has more records of the 
bird mixing with other seabirds when foraging (e.g., Hunt, 
this volume b). In this region, Marbled Murrelets were less 
common than the other species in the flocks, and rarely 
initiated the feeding flock (Porter and Sealy 1981; Chilton 
and Sealy 1987). Porter and Sealy (1981) found in Barkley 
Sound, British Columbia, that the murrelet had the lowest 
flocking tendency (0.2 percent) of the birds seen participating 
in multispecies feeding flocks, although there they did 
appear to initiate feeding flocks. Mahon and others (1992) 
observed that murrelets participate frequently in mixed 
species feeding flocks in the Strait of Georgia, British 
Columbia. They found a correlation between the number 
of feeding flocks observed in the area and the number of 
murrelets present. Chilton and Sealy (1987) suspected that 



murrelets enter small flocks to minimize disturbance from 
larger, more numerous, and aggressive individuals of other 
species that would find single birds easy to intimidate. 
Mixed flocks would occur after murrelets drove a school 
of sand lance to the surface. Other species participating in 
these feeding flocks in order of relative occurrence were 
Glaucous- winged Gulls (Larus glaucescens), Bonaparte's 
Gulls (Larus Philadelphia), Common Mergansers (Mergus 
merganser). Pigeon Guillemots, Mew Gulls (Larus canus), 
and Pelagic Cormorants (Phalacrocorax pelagicus). They 
felt that several factors encouraged a higher level of 
interspecific flocking behavior by murrelets: (1) larger 
and more aggressive alcids, such as Common Murres 
were absent; (2) the area had a high density of Marbled 
Murrelets; and (3) prey were locally concentrated, as the 
fish balled up at the surface when attacked, likely facilitating 
flock formation. 

In Alaska, the foraging flock of 4,000-6,000 Marbled 
Murrelets on 3 May 1991 in Icy Strait contained an equal 
number of Bonaparte's Gulls (McAllister, unpubl. data). 
Both species were feeding actively on what was suspected to 
be the hatch from a recent herring spawn. In southeast Alaska, 
McAllister (unpubl. data) found that Marbled Murrelets were 
rare in the areas where Common Murres and Rhinoceros 
Auklets (Cerorhinca monocerata) from the Forrester Island 
colony foraged around Prince of Wales Island. This area 
contains much suitable nesting habitat for murrelets, including 
large, contiguous stands of old-growth trees, but murrelets 
apparently avoid the region. He has also observed this at 
colonies near Saint Lazaria Island, in Sitka Sound, and Hazy 
Islands group. 

In the Gulf of Alaska, where the range of the Kittlitz's 
Murrelet (Brachyramphus brevirostris) overlaps with that of 
the Marbled Murrelet, the two species often share common 
foraging areas (McAllister, unpubl. data). However, the two 
species were not found to interact as pairs or in flocks. 

Diving 

Marbled Murrelet foraging is by pursuit diving (Ashmole 
1971). Depth and time of murrelet dives are little known. 

Dive times — We have recorded dive times of birds using 
birds with transmitters that were monitored by an observer 
on shore. When birds are underwater, the transmitter can no 
longer be heard. We also present some data from birds 
observed from shore through telescopes. 

Dive times were obtained from six birds fitted with 
transmitters in studies in 1989 and 1991 in northern 
California. The birds were followed on 13 occasions by a 
monitor on shore. The median dive times averaged 14 
seconds, with the longest at 69 seconds. The mean length of 
pauses between dives averaged 17 seconds in each year. 
Rest times were naturally more variable, with as long as 1 8 
minutes between dives. 

From-shore observations at Afio Nuevo Bay in California, 
birds were observed with dive times ranging from 7 to 42 
seconds. The depth of water for the 7-second dive was 1-2 



250 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



Strachan and others 



Chapter 23 



At-Sea and Foraging Behavior 



meters, in a rocky substrate, and the longer dives were 
observed in about 40 meters of water. Over a 4-year 
observation period, Strachan (unpubl. data), observed dive 
times in 20 to 50 meters deep water averaging about 31 
seconds. Also at Ano Nuevo, Strachan (unpubl. data) observed 
four birds, with a combined average dive time of 17 seconds, 
ranging from 6-39 seconds. The shortest average time (11.2 
seconds ) was a pair of possible juveniles in shallow waters 
of 2 to 5 m depth, up against the edge of a cliff. Those dives 
in the deepest water (40-45 m) were also the longest, and 
averaged 20.0 seconds. 

Pairs of birds resurface together on most dives, suggesting 
that they likely keep in visual contact underwater. Carter and 
Sealy ( 1 990) found that dive times of individual birds averaged 
27.8 seconds. Thoresen ( 1989), in a Washington study, observed 
the mean time for a dive was 44 seconds (range 15-115). 

Dive depths — Carter and Sealy (1984) found that murrelets 
killed in gill nets at night were probably feeding near the 
surface, as they were caught within 3 to 5 meters of the 
surface. Sealy (1974) stated that they usually foraged in 
areas that were sheltered from the prevailing winds and were 
relatively shallow (<30 m in depth). In southeast Alaska, 
Quinlan and Hughes (1984) found them most often in water 
less than 100 meters in depth and along steep, rocky coastline. 
In Prince William Sound, Alaska, Kuletz (1991a) found the 
highest densities of foraging birds in waters less than 80 
meters deep. Also in Alaska, Sanger (1987b) collected birds 
in January and estimated that most birds had been feeding in 
water of 18 to 45 meters deep. The birds had apparently 
foraged from the mid depths, to occasionally at or near the 
bottom, based on the prey species found in their stomachs. 
In Aiio Nuevo Bay. California, Strachan (unpubl. data) found 
the murrelets generally foraged in waters that ranged from 
20-30 meters. 

Fish Holding 

Few observations have been published of birds on the 
water holding fish. Carter and Sealy (1990) observed that 
most murrelets seen holding fish were observed near dusk, 
just before they fly to their nest to feed nestlings. A few 
birds were observed holding fish at dawn and later in the 
morning. They inferred that some individuals may feed chicks 
during the day because they felt that adults holding fish can 
not usually capture more fish. Carter and Sealy (1990) felt 
that increased fish holding by birds toward dusk coincided 
with the decrease in overall numbers of birds in the foraging 
area. Larger flocks sometimes included birds holding fish 
that were not feeding, although most birds that held fish 
were alone or in pairs. McAllister (unpubl. data) has recorded 
pre-dusk flyways where hundreds of fish-holding murrelets 
are counted as they leave foraging areas in Icy Strait, Sumner 
Strait, and in Frederick Sound in Southeast Alaska, heading 
towards their presumed nesting areas. At numerous locations, 
McAllister (unpubl. data) has recorded continuous fly way 
activity (averaging more than 20 birds per minute), with the 
majority of birds holding fish. 



On a few occasions, birds have been reported holding 
more than one fish in their bill. Thoresen (1989) observed a 
bird with two fish and a bird with three fish held crosswise, 
both on the water's surface and flying. Other observations of 
multiple fish in the bill include Carter (pers. comm.), Cody 
(1973), Fortna (pers. comm.), and Savile (1972). 

Foraging Influences 

Adjacent inland habitat — Densities of the Marbled 
Murrelet in specific geographic areas during the breeding 
season appear to be related to the adjacent nesting habitat 
(Carter and Sealy 1990; Ralph and Miller, this volume). It is 
also very probable that foraging locations are dependent 
upon prey habitat or availability, but no research has been 
conducted on this subject to date. 

Weather — Throughout their range, murrelets have been 
observed foraging in all weather conditions normal for that 
habitat. They have also been seen foraging in extreme weather 
conditions. McAllister (unpubl. data) has recorded foraging 
at night in sub-freezing conditions, with 40-60 knot easterly 
winds blowing out from the Taku River Valley. The birds 
were foraging on the herring schools that were feeding in the 
interface between marine and fresh water. Due to the 
topography, nearby waters within 4 km were relatively calm, 
yet the birds chose to be active at night in the rough weather 
and seas. 

Times of day — Birds appear to forage at all times of the 
day, and in some cases during night hours, presumably when 
there is enough ambient light to capture prey. Some observers 
have hypothesized that murrelets move from one feeding 
area to another during the early morning and late afternoon 
periods (Carter 1984, Carter and Sealy 1984, Prestash and 
others 1992). On the other hand, they may be staging in an 
area in the early morning near the nesting area, then moving 
out into foraging areas. Off the California coast, six birds 
with radio transmitters did not forage during the night in 
June or July (Ralph, unpubl. data), rather, the foraging was 
confined to the daylight hours. 

Topography — We have observed consistent densities of 
birds utilizing the lee of protected headlands in California, 
as has Kuletz (pers. comm.) in Alaska, We have noticed, but 
not quantified, that the wind conditions could be a factor for 
greater bird densities in the lee of headlands. Carter and 
Sealy (1990) speculated that prey also concentrates in sheltered 
waters. Certainly concentrations of birds are likely due to 
the availability of prey at the rip-current lines and in the tidal 
eddies that are established to the downwind of such features. 
In Oregon, Strong and others (1993) found that the highest 
densities of murrelets were found adjacent to beaches or 
mixed beach and rocky shore areas. 

Non-Foraging Behavior 

Coalescence 

An interesting phenomena that has been noted by a few 
researchers is that during the breeding season, about an hour 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



251 



Strachan and others 



Chapter 23 



At-Sea and Foraging Behavior 



before dusk, birds that are both loafing or foraging will 
coalesce into loose aggregations with much preening and 
wing stretching (Carter and Sealy 1990; Nelson, pers. comm.; 
O'Donnell, pers. comm.). We and Sealy (1975) have noted 
that specific sites are consistently used for these gatherings. 
Carter (pers. comm.), Kuletz (pers. comm.), and we have 
observed many times that a few minutes before dark the 
birds will begin to take off and fly inland in pairs or singly. 
In southeast Alaska, McAllister (