Informing ecosystem management: science and process for landbird conservation in the western United States Biological Technical Publication BTP-R1014-2011

              U.S. Fish & Wildlife Service
Informing Ecosystem
Management: Science
and Process for Landbird
Conservation in the
Western United States
Biological Technical Publication
BTP-R1014-2011
U.S. Fish & Wildlife Service
Informing Ecosystem
Management: Science
and Process for Landbird
Conservation in the
Western United States
Biological Technical Publication
BTP-R1014-2011
Jaime L. Stephens1
Kimberly Kreitinger2
C. John Ralph3
Michael T. Green4
1 Klamath Bird Observatory, Ashland, OR
2 PRBO Conservation Science, Petaluma, CA
3 U.S.D.A. Forest Service, Redwood Sciences Laboratory, Arcata, CA
4 U.S. Fish and Wildlife Service, Portland, OR
Photo credit: © Jim Livaudais
Cover Design: From Nyberg (1999)
Cover birds, clockwise from top: Bell’s Vireo (Vireo bellii), Black-throated Gray Warbler
(Dendroica nigrescens), Nashville Warbler (Vermivora ruficapilla), Purple Finch
(Carpodacus purpureus), Winter Wren (Troglodytes troglodytes), Yellow-breasted Chat
(Icteria virens)
ii Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Editor Contact Information:
Jaime L. Stephens
Klamath Bird Observatory
P.O. Box 758
Ashland, OR 97520
Phone: (541) 282-0866
E-mail: jlh@klamathbird.org
Kimberly Kreitinger
PRBO Conservation Science
3820 Cypress Drive #11
Petaluma, CA 94954
Phone: (415) 265-9153
E-mail: K.Kreitinger@gmail.com
C. John Ralph
USFS Redwood Sciences Laboratory
1700 Bayview Drive
Arcata, CA 95521
Phone: (707) 825-2992
E-mail: cjralph@humboldt1.com
Michael T. Green
U.S. Fish and Wildlife Service
Division of Migratory Birds and Habitat Programs
911 Northeast 11th Ave
Portland, OR 97232
Phone: (503) 872-2707
E-mail: Michael_Green@fws.gov
For additional copies or information, contact:
Jaime L. Stephens
Klamath Bird Observatory
P.O. Box 758
Ashland, OR 97520
Phone: (541) 282-0866
E-mail: jlh@klamathbird.org
Recommended Citation:
Stephens, J. L., K. Kreitinger, C. J. Ralph, and M.
T. Green (editors). 2011. Informing ecosystem
management: science and process for landbird
conservation in the western United States. U.S.
Department of Interior, Fish and Wildlife Service,
Biological Technical Publication, FWS/ BTP-R1014-
2011, Washington, D.C.
Series Technical Editor:
Stephanie L. Jones
U.S. Fish and Wildlife Service
Nongame Migratory Bird Coordinator
P.O. Box 25486
Denver Federal Center
Denver, CO 80225-0486
Table of Contents iii
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Integrating Partners in Flight Bird Conservation and Priority Land Management Objectives
John D. Alexander. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Partnering to Conserve Avian Biodiversity in National Parks of the Klamath Region
Daniel Sarr, Sarah McCullough, and Sean Mohren. . . . . . . . . . . . . . . . . 6
Use of Bird Conservation Plans for Development of Management Plans for National Wildlife Refuges
in Washington, Oregon, and California
Michael T. Green, Kevin Kilbride, and Fred Paveglio .  . . . . . . . . . . . . . 10
Risk Analysis of Birds Associated with Older Forests of the Pacific Northwest
Martin G. Raphael. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
A Watershed Analysis for Establishing Local Population Objectives for Pacific-slope Flycatcher
and a Suite of Mid- to Late-Successional Pacific Northwest Landbirds
Bob Altman, Michael T. Green, Barb Bresson, Erin Stockenberg, Daniel Casey, and
Susannah Casey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Demographic Monitoring, Modeling, and Management of Landbird Populations in Forests of the
Pacific Northwest: An Application of the MAPS Dataset
M. Philip Nott .  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Integrating Avian Monitoring into Forest Management: Pine-Hardwood and Aspen Enhancement
on the Lassen National Forest
Ryan D. Burnett .  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Partial Harvesting Can Enhance Foraging Habitat for Birds Associated with Understory Vegetation
in Western Oregon Forests
Joan C. Hagar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Success in Recovery Efforts of the Least Bell’s Vireo in Southern California
Barbara E. Kus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Fighting Fire with Fire: Bird Responses to Ponderosa Pine Treatments
Steve Zack. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Literature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Appendix: List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
iv Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
1
Preface
Jaime L. Stephens, Kimberly Kreitinger, C. John Ralph, and Michael T. Green
Integrating bird conservation with land
management
Recent advances in bird conservation are marked
by the integration of science and land management.
Information gained from past research can now
be used to develop user-friendly management
tools. Partners in Flight (Rich et al. 2004) as well
as shorebirds (Brown et al. 2001), waterbirds
(Kushlan et al. 2002), and especially waterfowl
(North American Waterfowl Management Plan
Committee 2004) initiatives use their respective
conservation plans to catalyze this process and
influence land management planning across the
landscape. Using these conservation plans within a
broader monitoring framework, managers can glean
pertinent information about ecosystem dynamics.
Why monitor birds?
Land managers work in a setting where change is
continuous and unpredictable (Bosch et al. 2003).
Within this dynamic environment, they often are
faced with making management decisions without
any scientific support to guide them. Management
activities need to be linked to the scientific process
in order to better understand potential influences on
the surrounding ecosystem. One scientific tool that
will help to forge this link is monitoring. Monitoring
measures population and habitat change and often
elucidates the causes of change. Performed in
concert with management actions, monitoring can
help to evaluate the effectiveness of management
prescriptions (Alexander et al. 2007) and provide
assurance that management efforts are focusing on
agreed-upon goals (Keough and Blahna 2006).
Land managers and biologists commonly monitor
birds, both to track bird populations themselves,
and as a tool to measure ecosystem health as a
whole. Birds are relatively easy and cost-effective
to monitor and standardized methodologies exist
to allow comparisons across sites (Ralph et al.
1993). Birds occupy a wide diversity of ecological
niches and respond quickly to changes in their
environment. While bird monitoring is common,
it is not always clear exactly what is gained by this
monitoring. Primarily, bird monitoring is integral in
answering the immediate questions about the effects
of land management on an ecosystem. In addition,
the value of monitoring data could increase with
time as it contributes to answering longer and larger
scale questions. However, monitoring data are only
as valuable as the extent to which they are applied.
It is therefore important that we step back and
evaluate the influence that bird monitoring projects
have had on management. With this, we can learn
from the past and inform others of how to implement
successful, meaningful monitoring projects for the
future.
How do adaptive management and monitoring
interact?
This volume highlights bird conservation
successes resulting from the integration of science,
management, and learning within a collaborative
framework, i.e., adaptive management (Jacobson
et al. 2006). The adaptive management process
consists of six stages: assessment, design,
implementation, monitoring, evaluation, and
adjustment. Land management projects are
implemented one stage at a time and tested at each
step, allowing for detection and correction of any
deleterious effects (Moir and Block 2001). Ideally,
information from one stage is incorporated into
subsequent stages and an informational feedback
loop or “adaptive management circle” is created.
When properly integrated, the process is continuous,
cyclic, and constantly evolving (Haney and Power
1996).
Examples from the western United States
In this publication, we present ten examples
illustrating both the process and science behind bird
conservation throughout the western United States.
We begin with a series of papers that describe
integrating bird conservation and effectiveness
monitoring into land management guidelines and
emphasize the importance of partnerships. This is
followed by a series of case studies which highlight
bird monitoring within the adaptive management
framework. We emphasize the science of monitoring
and the process of its integration into land
management because both are necessary in order
for effectiveness monitoring to fully impact decision
making.
Acknowledgments
We thank John D. Alexander, Bob Altman, Barb
Bresson, Geoff R. Geupel, Aaron L. Holmes, Melissa
Pitkin, and Terry D. Rich who were integral in
the workshop that was the inspiration for this
publication, “Tools for Bird Conservation in
Conifer Forests: A Joint California and Oregon-
Washington Partners in Flight Workshop,” held in
2 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Ashland Oregon in April 2005. Many thanks to Bob
Altman, Carol J. Beardmore, Joe Buchanan, Ryan
D. Burnett, Barb Bresson, Dan Casey, David Craig,
Joe Fontaine, Thomas Gardali, Joan C. Hagar, Rob
Holbrook, Gary Ivey, Dave Mauser, Larry Neel,
Nadav Nur, Arvand O. Panjabi, Scott F. Pearson,
Hildy Reiser, Jon Robinson, Christopher Rustay,
Steve I. Rothstein, Paul Roush, Nathaniel E. Seavy,
Tom Will, Julian K. Wood, and Jock Young for their
review of these papers. We would also like to thank
Danielle M. Morris for her assistance with editing
and formatting the compiled works. In addition, we
are grateful to the authors for their contribution to
this publication. Partial funding for this publication
was granted by the National Fish and Wildlife
Foundation, the M. J. Murdock Charitable Trust,
and U.S. Fish and Wildlife Service, Region 1,
Division of Migratory Birds and Habitat Programs.
3
Integrating Partners in Flight Bird Conservation
and Priority Land Management Objectives
John D. Alexander
issues and develop management objectives; 2)
design management actions to achieve objectives
(e.g., desired conditions); 3) implement management
actions; 4) monitor the results of management
actions; 5) use monitoring results to evaluate the
efficacy of the management actions in achieving the
objectives; and 6) adjust treatments, prescriptions,
plans, and policies accordingly.
PIF’s conservation planning strategy is a process
that uses science-based information about
birds to link bird conservation objectives and
management issues. Using results from research
and monitoring efforts in the Klamath-Siskiyou
Region, I demonstrate how PIF’s conservation
planning strategy can be implemented within the
adaptive management framework to integrate
bird conservation objectives with priority land
management challenges.
Assessing populations and designing
conservation objectives
Bird conservation plans present a synthesis
of priorities and objectives to guide landbird
conservation actions (Rich et al. 2004). To design
and implement meaningful bird conservation plans,
conservation issues must be assessed at multiple
scales. Traditional conservation efforts based
on a single-species approach, often driven by the
Abstract
Using results from ongoing research and monitoring
studies in the Klamath-Siskiyou Region of northern
California and southern Oregon, I demonstrate how
a Partners in Flight conservation planning strategy
can be implemented using an adaptive management
approach. Partners in Flight’s planning strategy
involves: 1) species and habitat assessment to derive
population and habitat objectives for focal species;
2) working with land managers to integrate these
objectives into management plans and implementing
conservation actions on the ground; and 3)
monitoring the effectiveness of these actions as an
evaluation component of the conservation strategy.
These conservation strategy components allow land
managers to design projects that simultaneously
meet priority management objectives (e.g., fire
hazard reduction) and achieve bird conservation
objectives. Monitoring bird community response to
such projects leads to refinements or adaptations
to future management actions, a critical step
for managers concerned with achieving certain
desired conditions within an adaptive management
framework.
Introduction
Partners in Flight (PIF) has developed a
conservation planning strategy (Bonney et al.
1999) that serves as a model for integrating bird
conservation objectives into land management
programs through the adaptive management
framework (Fig. 1; Nyberg 1999). This strategy
involves: 1) assessing the conservation status of
bird species at continental and regional scales;
2) identifying habitat characteristics important
for species of concern; 3) implementing land
management actions that improve habitat
characteristics for those species; and 4) monitoring
the response of those species to evaluate the
effectiveness of management actions.
Adaptive management is a systematic approach for
improving resource management by learning from
management outcomes (Williams et al. 2009). It
has been traditionally conceptualized as a circular
feedback loop with six components (Fig. 1). Working
through this framework land managers: 1) assess
Figure 1. The traditional circular model of the
adaptive management framework from Nyberg
(1999).
4 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Endangered Species Act (Endangered Species
Act 1973; ESA), are not adequate for addressing
continent-wide bird population declines (Bonney
et al. 1999). The PIF approach to proactive
conservation considers a suite of focal species (Chase
and Geupel 2005) with the ultimate goal of reversing
population declines before ESA listing becomes
necessary (Rich et al. 2004).
A continental assessment of all landbirds was
completed in 2004 (Rich et al. 2004). Population
trends generated from the Breeding Bird Survey
(BBS), a continent-wide bird monitoring program
(Sauer et al. 2008), and species distribution
information, were used to identify species of high
conservation concern at a continental scale (Panjabi
et al. 2001).
To assess the status of bird species at regional
scales, the Oregon-Washington and California PIF
chapters instituted multiple regional monitoring
programs. The Klamath Bird Monitoring Network
(Network) is an example of such a program
(Alexander et al. 2004). The Network was designed
to: 1) monitor regional bird population trends for
comparison with BBS results; 2) determine the
distribution of species of concern in southern Oregon
and northern California; and 3) develop habitat
relationship models.
The Network facilitated regional assessment using
mist-netting and point count data collected with
standard protocols (Ralph et al. 1993) employed
at different spatial and temporal scales. Regional
data from the Network’s long-term (>10 year)
constant-effort mist-netting stations corroborated
BBS data that suggest declines for Swainson’s
Thrushes (Catharus ustulatus), Orange-crowned
Warblers (Vermivora celata), Black-throated Gray
Warblers (Dendroica nigrescens), MacGillivray’s
Warblers (Oporornis tolmei), and Purple
Finches (Carpodacus purpureus) (Klamath Bird
Observatory pers. comm.). Point count data refined
our knowledge of the distribution and habitat
relationships of bird communities in the Klamath-
Siskiyou Region. We confirmed that elevation,
plant species composition (i.e., habitat type) and
vegetation structure are important factors for
determining species distribution (Alexander 1999,
Seavy 2006).
Results from analyses of population status and
habitat requirements of bird species of concern can
guide the land management process in the Klamath-
Siskiyou Region. They provide a foundation for
regional habitat-based conservation plans (Altman
2000a, California Partners in Flight 2002b) and
contribute to continental bird conservation planning
(Rich et al. 2004). Variables used to describe the
distribution of birds (e.g., vegetation structure and
volume; Alexander 1999, Seavy 2006) are the same
variables used to describe current and desired
conditions in the land management planning process.
Effectiveness monitoring results and adaptive
management
Land management agencies are required to monitor
the effectiveness of their management actions to
determine if they are meeting desired ecological
conditions (Forest Ecosystem Management
Assessment Team 1993). Birds can serve as useful
tools when evaluating management actions and
designing conservation efforts because they occupy
a diversity of ecological niches (Riparian Habitat
Joint Venture 2004) and respond to a wide variety
of habitat conditions (Hutto 1998). In addition,
compared to other taxa, birds are inexpensively
detected using standardized sampling protocols
(Alexander et al. 2007). Thus, birds serve as “focal
species” whose requirements define different spatial
attributes, habitat characteristics, and management
regimes of healthy ecosystems (Chase and Geupel
2005).
We evaluated the ecological effects of fuel reduction
projects in oak woodland and chaparral habitats
of the U. S. Bureau of Land Management (BLM)
Medford District in the Klamath-Siskiyou Region
using point counts, comparing the abundance of
PIF focal species in treated and adjacent untreated
habitats (Alexander et al. 2007). Our results
suggested that small-scale treatments that retained
shrub patches benefited edge-associated birds,
including regionally declining Purple Finches (Fig.
2). These results corroborated information in the
PIF regional bird conservation plan for landbirds in
lowlands and valleys (Altman 2000a) regarding the
importance of edge habitats for some species. Our
data also suggested that the fuel reduction efforts
retained shrub patches resulting in no measurable
decline in shrub-associated bird species. However,
our results did raise a concern about negative
impacts of treatments on species that use small
snags.
Figure 2. Mean abundance (individuals per station)
of Purple Finches detected in hand-pile and burn
treatment (51 stations clustered in 9 units) and
untreated (49 stations clustered in 7 units) oak
woodland and chaparral plots of the Applegate
Valley, Oregon, from Alexander et al. (2007).
5
The BLM Medford District’s multi-disciplinary
management team incorporated these results into
subsequent treatment projects, altering treatment
prescriptions to include the retention of small snags
(V. Arthur pers. comm.). These revised prescriptions
not only addressed the needs of edge and shrub
associated species, they also maintained key features
for snag associates. Monitoring bird response to
land management continues to play a crucial role
in the management of oak-shrub-conifer matrix on
BLM’s Medford District.
Extending the PIF strategy to land managers
throughout the Klamath-Siskiyou Region
Federal agencies manage the majority of forested
and shrubland landscapes across the west and
therefore offer some of the best opportunities to
implement bird conservation objectives at large
scales. PIF has a long history of partnership
with these agencies; however land management
decisions do not consistently consider or align
with PIF conservation objectives. Increased
effectiveness monitoring which uses PIF focal
species as indicators of current and desired
ecological conditions would result in better informed
management decisions with regards to bird
conservation.
Encouragingly, in the Klamath-Siskiyou Region,
land management agencies are beginning to use
the information from the analyses of the Network’s
data to design oak woodland treatments to be more
consistent with PIF habitat-based conservation
objectives. Additionally, increased collaborations
within the PIF conservation strategy are engaging
land managers to evaluate the impacts of other land
management projects, including larger-scale fuel
reduction treatments in oak woodlands (Seavy et
al. 2008) and small-scale fuel reduction treatments
in riparian habitats (Klamath Bird Observatory
and U.S. Bureau of Land Management 2009).
Furthermore, as landscape level fuel reduction
programs are being planned regional land managers
are consulting with PIF conservation planners to
design the spatial distribution and replication of
treatments that serve as a frame for well designed
effectiveness monitoring studies (Klamath Bird
Observatory and U.S. Bureau of Land Management
2009). Thus, the PIF strategy is being more widely
incorporated into land management throughout the
Klamath-Siskiyou Region.
By integrating the PIF conservation planning
strategy within local land management planning
processes, the PIF strategy can serve as a catalyst
for sustainable land management within the adaptive
management framework. Such integration results in
three conditions that Williams et al. (2009) suggest
are ideal for successful implementation of adaptive
management:
(1) Because the use of bird monitoring, as a cost
effective tool to monitor the ecological effects of
management, is integral to the PIF conservation
strategy, it works well within ecosystem manage-ment;
(2) PIF conservation planners are engaging man-agement
leadership by identifying conservation
opportunities within priority land management
objectives; and
(3) Broad stakeholder consensus is being built
around resulting land management actions that
meet both land management and bird conserva-tion
objectives.
As a means for supporting land management agency
efforts to implement adaptive management the
integration of PIF’s conservation planning strategy
within local land management planning should
result in more opportunities to implement bird
conservation objectives within land management
programs.
Acknowledgments
The Klamath Bird Observatory has worked with the
U.S. Forest Service Redwood Sciences Laboratory
and many partners to implement the Partners in
Flight conservation planning strategy described
here through a series of Oregon-Washington and
California Partners in Flight projects. These have
been supported with funding from U.S. Forest
Service, U.S. Bureau of Land Management, and
the Joint Fire Sciences Program. Support from
Jackson and Klamath counties was provided through
the Secure Rural Schools and Community Self-
Determination Act of 2000 (Public Law 106-393).
Additional support was provided from the National
Fish and Wildlife Foundation and the M.J. Murdock
Charitable Trust. This success story resulted from
long-running collaborative relationships with local
Bureau of Land Management and Forest Service
partners. Comments from Paul G. Sneed, Rick
Medrick, C. John Ralph, Jaime L. Stephens, Michael
T. Green, J. Michael Scott, Victoria Sturtevant and
three anonymous reviewers greatly improved this
manuscript.
6 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Partnering to Conserve Avian Biodiversity in
National Parks of the Klamath Region
Daniel Sarr, Sarah McCullough, and Sean Mohren
Abstract
National Park lands are often believed to contribute
towards the habitat-based objectives outlined in
the Partners in Flight Bird Conservation Plans
by protecting large tracks of contiguous land
holdings where natural processes predominate.
However, a paucity of accurate data to evaluate such
assumptions has left the National Park Service’s
contributions to regional conservation initiatives
open to question.
The Klamath Network, a confederation of six
National Park Service units in southern Oregon
and northern California, launched its Inventory
and Monitoring Program in 2000. Since then,
the Network has taken four sequential steps to
explore patterns of avian biodiversity and to lay the
groundwork for long-term landbird monitoring. The
steps include: 1) conducting inventories to determine
distribution and abundance of relatively common
species in the parks; 2) updating the bird species
list for each park; 3) designating landbirds as vital
signs for the Network; and 4) developing landbird
monitoring protocols to guide long-term monitoring.
In 2002, the Klamath Network approached the
Klamath Bird Observatory with a request to partner
for inventory and monitoring of landbirds. Since
then, Klamath Bird Observatory has provided
assistance with each of the network steps for
the development of its inventory and monitoring
program. Through this collaboration, the Klamath
Network has been able to meet park management
objectives and become an active contributor to
Partners in Flight conservation objectives at
regional and continental scales.
Background
The National Park Service Inventory and
Monitoring Program.—When President Woodrow
Wilson signed The Organic Act of 1916, he
authorized the formation of a National Park Service
(NPS) dedicated to “conserve the scenery and the
natural and historic objects and the wild life therein
and to provide for the enjoyment of the same in
such manner and by such means as will leave them
unimpaired for the enjoyment of future generations”
Early park service administrators often assumed
that the exclusion of logging, grazing, and mining
would ensure, in the words of Horace Albright,
second Director of the NPS, that national parks
would persist in “everlasting wildness” (Sellars
1997). As early as the 1930s, however; scientific
studies showed that this was an invalid presumption
(Sellars 1997). Declines in native species (especially
predators), introductions of exotic plants and
animals, and impacts from roads were noted in the
earliest investigations of national parks in California
(Sellars 1997). It became apparent that there was a
need for quantitative information about the status
of park ecosystems, their intrinsic variability, and
potential threats. A scarcity of information made it
difficult to assess the contributions of the national
parks to other regional conservation initiatives such
as Partners in Flight (PIF).
To address internal and external demands for
scientific information, NPS developed a nationwide
Inventory and Monitoring Program (National
Park Service 2006a) to yield scientifically sound
information on the status and long-term trends
of park ecosystems and to determine how well
current management practices are sustaining those
ecosystems (National Park Service 2008a). As a
critical step in the development of the Inventory
and Monitoring Program, 270 national park
units nationwide were grouped into 32 networks
linked by geographic similarities, common natural
resources, and resource protection challenges. The
network approach was chosen to facilitate staffing,
collaboration, information sharing, and economies of
scale in natural resource monitoring.
The Klamath Network encompasses six units
managed by NPS in northern California and
southern Oregon: Crater Lake National Park,
Lassen Volcanic National Park, Lava Beds
National Monument, Oregon Caves National
Monument, Redwood National and State Parks, and
Whiskeytown National Recreation Area (National
Park Service 2008b). Collectively, the six units
comprise nearly 200,000 ha and range considerably
in size (196–73,775 ha), relief, and character (Fig. 1).
The parks of the Klamath Network span a region
of exceptional complexity, where steep climatic,
geologic, and topographic gradients and varied
disturbance regimes yield biological diversity that is
exceeded in few similarly sized areas of the continent
7
songbird declines have been reported throughout
North America (Ballard et al. 2003) and recent
analyses suggest this is also the case in the
Klamath-Siskiyou ecoregion (Trail 2004), which is
central to the Klamath Network parks. Within this
ecoregion, there are a number of potential factors
contributing to population declines. Limiting factors
include habitat loss and alteration, land uses that
compromise the integrity of natural systems such
as overgrazing, development, and suppression
of natural processes (e.g., fire, flooding), nest
parasitism by Brown-headed Cowbirds (Molothrus
ater), competition from invading species (e.g. Barred
Owls (Strix varia) supplanting Northern Spotted
Owls (Strix occidentalis caurina)), and predation
by both native and non-native predators (Sarr et al.
2007). For these reasons, the parks in the Klamath
Network desired a better understanding of the
current status of landbirds within their boundaries
and at a regional scale. In addition, the Network
desired baseline landbird data to support potential
monitoring in the future.
Partnering with Klamath Bird Observatory.—
When confronted with the need to inventory
landbirds in the parks within the Klamath Network,
partnering with KBO was a logical choice. KBO
(DellaSala et al. 1999, Sarr et al. 2004). The parks
in the Klamath Network contain a diverse mosaic
of climates, landforms, and ecosystems, from moist
redwood forests near the coast to xeric sagebrush
steppe inland, and from oak woodlands to alpine fell
fields (Sarr et al. 2004). This paper describes four
steps the Klamath Network, with Klamath Bird
Observatory (KBO), has taken to inventory and
better understand avian biodiversity in the parks
and to lay the groundwork for long-term landbird
monitoring.
Inventory needs in the Klamath Network Parks.—
The Klamath Network Inventory Program was
a five year project funded by NPS from 2000
- 2004 (Ackers et al. 2002, McCullough 2006b).
The intent was to develop a current species list
of at least 90% accuracy for vascular plants and
vertebrates (i.e., birds, mammals, amphibians,
reptiles, and fish), and to determine distribution
and abundance of taxa of special concern in each
park. During initial scoping prior to the launch of
the inventory program, the parks of the Klamath
Network identified Neotropical migrant birds as
a taxon of special concern and primary emphasis
for field sampling (Sarr et al. 2007). Neotropical
Figure 1. Six National Park Service units in southern Oregon and northern California constitute
the Klamath Network.
8 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
has been operating a network of bird monitoring
stations throughout the Klamath-Siskiyou region
of California and Oregon since 1993. More
importantly, their dedication to providing science-based
bird monitoring to further bird conservation
and help make land management decisions is a
central interest to NPS, as well as to other Federal
agencies. Both the NPS and KBO are involved
in PIF, a conservation initiative dedicated to
increasing the cooperative efforts of public and
private organizations involved in bird conservation.
The lands within the Klamath Network parks
are extraordinarily diverse, falling under six
regional PIF Bird Conservation Plans, the Oak
Woodland, Riparian, Coniferous Forest, Coastal
Scrub, Sagebrush, and the East-slope Cascades
conservation plans and two biomes (Pacific and
Intermountain West) listed in the PIF North
American Landbird Conservation Plan (Altman
1999, 2000c; California Partners in Flight 2002a,
2002b; Riparian Habitat Joint Venture 2004;
Rich et al. 2005). Accurate information about the
distribution and abundance of landbirds in the
Klamath Network parks was considered essential to
helping the parks meet their inventory goals while
contributing to PIF efforts to conserve and improve
our understanding and conservation of focal bird
species.
Implementing the landbird inventory.—During a
two year field effort conducted in 2002 – 2003, KBO
scientists established study areas in riparian and
adjacent upland habitats and used multiple avian
survey methods that varied by park in order to
maximize the inventory data in each park. Methods
were determined based on park size, variability in
park habitat, and to align with monitoring methods
used in the past. The objective of the inventory
was to obtain baseline data on the distribution
and abundance of target species during both
the breeding and migration seasons. KBO also
conducted constant-effort mist netting at Oregon
Caves National Monument, summarized datasets
from previous bird monitoring efforts in three parks,
and compared the results from the 2002 and 2003
field seasons to existing species lists for each park.
During the two year inventory, a total of 234 new
landbird inventory stations were established in
Crater Lake National Park and Whiskeytown
National Recreational Area. At Lava Beds National
Monument, 36 fall migration bird area search
inventory stations were established (overlapping
with existing breeding season stations) to create a
baseline for fall migration data. For each station
that was established, habitat composition and
structure data were collected, and GPS data were
recorded. Standardized methodologies were
used to facilitate replication by future inventory
or monitoring efforts. In addition, KBO added to
available baseline data by summarizing previous
point count efforts in Lassen Volcanic, Crater Lake,
and Redwood national parks.
At Oregon Caves National Monument, a constant-effort
mist netting station with 10 nets added to
existing baseline breeding season and migration
season data. It was anticipated that multiyear data
would assist potential monitoring, so mist netting
was funded in each year since 2003, with a five year
summary report completed in 2007 (Frey et al. 2007).
Documenting Avian Biodiversity in the Klamath
Network Parks.—The inventories conducted
by KBO recorded several new species for the
parks and confirmed many species considered
hypothetical (Sarr et al. 2004). In addition to
its role in implementing field inventories, KBO
assisted the Network in a certification process
whereby species lists were reviewed for accuracy.
Once field inventories and certification processes
were complete, the parks had current information
about the presence, distribution, and abundance of
many of the more common landbird species in the
parks (National Park Service 2009). These data,
together with park-specific historical data and
knowledge, provided an excellent inventory for the
Klamath Network that has been of direct relevance
to management and subsequent monitoring
development efforts. In addition, standardized
survey methods were field tested on-the-ground in
parks for potential use in long-term monitoring.
Developing a landbird monitoring program
Upon completion of the five year inventory
programs, each of the 32 networks in the NPS
Inventory and Monitoring Program was provided
with base funding to support the development
of a long-term Vital Signs Monitoring Program.
Development of a monitoring program tasked
each Inventory and Monitoring network with
convening the parks, regional universities, and other
conservation science organizations to identify “vital
signs” to monitor as a way to gauge the health of the
park ecosystems. Ongoing examples of such vital
signs monitoring include tracking air and water
quality, climate, and population dynamics of small
mammals or waterfowl, and studying historical
photographic records.
During the initial scoping meetings, the Klamath
Network park representatives recognized landbirds
as a key resource that could provide valuable
information about the park’s ecosystems through
long-term monitoring. In addition, they recognized
that landbird conservation requires a perspective
that extends to the regional and continental scale,
well beyond park boundaries.
The Vital Signs Process
The Klamath Network vital signs selection process
began in 2004. The process involved 130 experts
representing a broad array of scientific disciplines
and required them to rank candidate vital signs
(biological communities or ecological components
of the parks) based on ecological and management
significance (National Park Service 2006b). The
final selection of vital signs was accomplished at a
9
workshop in Redding, California on 27-28 April 2005,
where NPS staff reviewed and approved the final
list: bird communities ranked fourth in importance
to the parks, out of over 100 vital signs (Table 1).
Landbird communities were selected as a focal
community important to maintaining and measuring
ecological integrity in terrestrial ecosystems. Bird
communities are species-rich, easy to monitor
compared to other kinds of communities, present
in most park habitats, and can serve as indicators
of environmental change (Temple and Wiens 1989).
Long-term monitoring of species composition,
population trends, and distributions of landbird
communities will provide valuable information on
population responses to natural and anthropogenic
influences within and outside of park boundaries.
Developing a Landbird Monitoring Protocol.—
In 2007, KBO began assisting the Network with
the development of a landbird monitoring protocol
for the parks. Under the protocol process, KBO
and the Network have developed spatial and
temporal sampling designs for each park, standard
data analysis and reporting practices, and a
comprehensive data management system that
contributes information for local, regional, national,
and continental needs (Stephens et al. 2009).
Implementation of the protocol began in the spring
of 2008.
Collaborative conservation in the Klamath
Region
Since 1993, KBO and the U.S. Forest Service’s
Redwood Sciences Laboratory have been
coordinating bird monitoring efforts in the
Klamath-Siskiyou region. Known as the Klamath
Demographic Monitoring Network, this effort has
yielded a substantial regional dataset (Alexander
et al. 2004). The NPS vital signs bird monitoring
program, although designed to answer park-specific
questions, will contribute to monitoring
bird distribution and population trend information
being gathered by KBO at the regional scale. The
nesting of the NPS vital signs monitoring program
within the larger Klamath Demographic Monitoring
Network provides an opportunity to explore
questions about the effects of habitat management
and environmental conditions on landbird
populations across a large landscape.
Moreover, the NPS Vital Signs Monitoring Program
will complement the PIF goals both materially and
conceptually; their approaches are complementary.
The overall goal of PIF bird conservation planning
is to ensure long-term conservation of native
landbirds (Rich et al. 2004). The vital signs process
is intended to provide a broad view of the integrity
of park ecosystems. Vital signs monitoring of
landbirds in the Klamath Network parks will work
toward both these broad and interdependent goals.
Quantitative information about landbird distribution
and abundance, managed in high quality databases
developed in partnership with KBO, will allow the
NPS to meet its local management needs and to
make substantial contributions to regional and
continental bird conservation.
Table 1. Top ten vital signs (biological communities or ecological components of the parks) of the Klamath
Network based on ecological and management significance (National Park Service 2006).
Ranking Vital Signs
1 Non-native species
2 Keystone and sensitive plants and animals
3 Terrestrial vegetation
4 Bird communities
5 Intertidal communities
6 Freshwater aquatic communities
7 Cave collapse / entrance communities
8 Water quality (aquatic, marine and subterranean)
9 Land cover, use, pattern
10 Environmental conditions in caves
10 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Use of Bird Conservation Plans for Development of
Management Plans for National Wildlife Refuges
in Washington, Oregon, and California
Michael T. Green, Kevin Kilbride, and Fred Paveglio
Abstract
The National Wildlife Refuge System is in
the midst of developing and revising resource
planning documents, including Comprehensive
Conservation Plans to guide long-term management,
Habitat Management Plans which add detail
to the Comprehensive Conservation Plans,
and Environmental Assessments for specific
activities. Each of these planning documents
offers opportunities for setting specific biological
targets for management. Partners in Flight and
shorebird and waterbird initiatives have developed
conservation plans that provide land managers with
information to improve habitat conditions for birds.
Increasingly, refuge staff and planners with the U.S.
Fish and Wildlife Service in California, Oregon,
and Washington are using objectives from the
bird conservation plans to develop detailed refuge
objectives in resource planning documents. Using
focal species from bird conservation plans to guide
the development of habitat objectives will enable
land managers to recreate functioning ecosystems
in priority habitats. In addition, monitoring the
habitat and landbird responses to the conservation
recommendations provides feedback for assessing
their effectiveness. We present examples from
four National Wildlife Refuges that incorporated
Partners in Flight plans into refuge planning
documents: Little Pend Oreille, Klamath Marsh,
Sacramento River, and San Joaquin River.
Introduction
Under the jurisdiction of the U.S. Fish and Wildlife
Service, all National Wildlife Refuges (refuges)
are developing, or have recently developed,
Comprehensive Conservation Plans (CCPs) to guide
long-term management in accordance with the
National Wildlife Refuge System Improvement Act
(1997). In addition, refuges are updating Habitat
Management Plans (HMPs). CCPs typically have
a 15-year planning horizon and updated HMPs
add detail to management prescriptions that are
presented as strategies within CCPs. Furthermore,
the National Environmental Policy Act (1970)
requires Environmental Assessments (EAs) for
some refuge activities. Each of these resource
planning documents offers opportunities for setting
specific biological targets for management. At the
same time, Partners in Flight (PIF) and initiatives
to conserve waterbirds and shorebirds have
developed conservation plans that strive to provide
land managers with information that will translate
into improved habitat management for birds.
Increasingly, refuge staff and planners in California,
Oregon, and Washington are using bird conservation
plans from PIF and the other bird initiatives to
develop detailed refuge objectives in CCPs, HMPs,
and EAs. While many refuges were established
for the purposes of conserving species other than
birds (e.g., Hart Mountain National Antelope
Refuge), many others have purposes related directly
to migratory birds through the Migratory Bird
Conservation Act (1929). In addition, each refuge
has at least a secondary responsibility to consider
the needs of birds on their lands through the trust
responsibility endowed upon the U.S. Fish and
Wildlife Service for the protection, conservation,
and management of migratory birds through
the Migratory Bird Treaty Act (1918). Thus,
management for migratory birds is a prominent
feature of many refuge planning documents.
The PIF plans for California, Oregon, and
Washington provide detailed strategies to meet life
history requirements for high-priority (e.g. focal)
landbirds in priority habitats, habitats which have
generally been substantially altered relative to pre-
European settlement. The habitat requirements of
focal species represent spatial attributes, habitat
conditions, and management regimes characteristic
of healthy ecosystems (Riparian Habitat Joint
Venture 2004). Thus, by using focal species to guide
the development of habitat objectives on refuges,
land managers can recreate functioning ecosystems
in these priority habitats. Monitoring the habitat
response and responses of bird populations to
the PIF conservation recommendations provides
invaluable feedback for assessing their effectiveness.
The following are examples from refuges in
Washington, Oregon, and California that have used
PIF bird conservation plans for the development of
recent refuge planning documents.
11
Little Pend Oreille National Wildlife Refuge
Established in 1939 as a “refuge and breeding
ground for migratory birds and other wildlife,”
(U.S. Fish and Wildlife Service 2000) Little Pend
Oreille National Wildlife Refuge comprises 16,268
ha of cold, moist, and dry forests along with alluvial
riparian and some meadow habitat. It lies 100 km
north of Spokane, Washington, and is surrounded
by U.S. Forest Service lands, including the Colville
National Forest.
The CCP for this refuge was developed in 2000
(U.S. Fish and Wildlife Service 2000), and describes
long-term habitat management and restoration
goals, objectives, and strategies for its forested,
riparian, and wetland habitats. The 2005 HMP
further refines the CCP objectives (U.S. Fish and
Wildlife Service 2005a) and draws heavily from
the Oregon-Washington PIF landbird plan for the
northern Rocky Mountain region (Altman 2000b).
The HMP used all of the focal landbird species and
their habitat objectives in this bird conservation
plan except for Upland Sandpipers (Bartramia
longicauda) and Vesper Sparrows (Pooecetes
gramineus), which lack appropriate habitat on the
refuge.
Habitat objectives in the HMP are derived from
habitat requirements for several focal species in
the PIF plan, but the most striking use of the PIF
plan for the HMP is the refuge’s long-term habitat
objective for ponderosa pine (Pinus ponderosa).
Ponderosa pine dominated, late-seral dry forest is a
habitat type considerably reduced in the Northwest
due to logging and fire suppression (O’Neil et
al. 2001). White-headed Woodpeckers (Picoides
albolarvatus) are the focal species representing
healthy ponderosa pine forests in late-seral condition
in the PIF plan. It has also shown local population
declines, and is a conservation priority in this region
(Rich et al. 2004, U.S. Fish and Wildlife Service
2008). The prescription for the habitat attributes
in the HMP, as described for White-headed
Woodpeckers in the PIF plan (Altman 2000b), are
to provide 1821 ha in patches larger than 142 ha
through periodic thinning and burning of mid-seral
stage forest (Fig. 1), and that these late-seral dry
forest stands have:
“10 or more trees per acre larger than 53 cm dbh,
with at least two of those exceeding 79 cm dbh;
10–40% tree canopy cover; and more than 1.4 snags
per acre that are greater than 20 cm dbh.”
By achieving this habitat objective, refuge lands
would provide protected habitat needed by 5 to 12
pairs of White-headed Woodpeckers where there are
none now (calculated from home range estimates;
Garrett et al. 1996). This objective is striking not
only because of its required 100–200 year time
frame, but also for its degree of specificity. The long
time frame is appropriate for developing stands
of old-growth ponderosa pine, but unusually far-sighted
for a refuge and well beyond the 15-year
scope of most CCPs.
Point counts were conducted by refuge staff from
2000 – 2002, and will continue periodically as
restoration continues. A long-term monitoring
strategy will allow for the evaluation of the
effectiveness of the habitat restorations, and of the
habitat recommendations in the bird conservation
plan. If White-headed Woodpeckers do not respond
as expected, habitat restorations will be examined,
simultaneously with the habitat prescriptions
suggested for this species in the bird conservation
plan.
Klamath Marsh National Wildlife Refuge
The Klamath Marsh National Wildlife Refuge is
one of six refuges in the Klamath Basin National
Wildlife Refuge Complex located in southern
Oregon and northern California. The Klamath
Marsh refuge lies about 50 km north of Klamath
Falls, Oregon. The refuge was established in 1958
to provide migration and production habitat for
migratory birds, particularly waterfowl and Sandhill
Cranes (Grus canadensis). The 16,502 ha refuge
is 90% permanent and seasonal marsh, with a 1376
ha fringe of forest characterized by lodgepole pine
(Pinus contorta), ponderosa pine, and relict quaking
aspen (Populus tremuloides). Winema National
Forest and private lands border the refuge, and
nearby farms and ranches grow hay and livestock.
A fuels reduction EA (U.S. Fish and Wildlife Service
2003) was developed to protect refuge structures
and neighboring residences from wildfires, and
to restore and maintain the condition of wildlife
habitats including old-growth ponderosa pine and
lodgepole pine, aspen stands, and seasonally-wet
meadows. As elsewhere in the West (Covington
and Moore 1994, Fleischner 1994), the condition
of pine forests and aspen woodlands on the refuge
have declined due largely to fire suppression and
grazing pressure. Several bird species would benefit
from proper aspen management, e.g., removing
heavy grazing pressure (Earnst et al. 2005, Heltzel
and Earnst 2006). Decadent aspen groves also
regenerate rapidly when challenged with controlled
burns and cutting of competing species of conifer
(Jones and DeByle 1985). Bird species likely to
benefit from management for aspen include Western
Screech-Owls (Otus kennicottii), Northern Pygmy
Owls (Glaucidium gnoma), Williamson’s Sapsuckers
(Sphyrapicus thyroideus), Red-naped Sapsuckers
(Sphyrapicus nuchalis), Northern Flickers
(Colaptes auratus), Tree Swallows (Tachycineta
bicolor), House Wrens (Troglodytes aedon), and
Mountain Bluebirds (Sialia currucoides) (Altman
2000c). Appendix 1 of the EA describes in detail the
desired conditions for each of those habitats, their
associated focal bird species, and treatment options
(thinning and burning) to achieve those conditions.
In aspen, for example, the desired future condition
in the EA is “large aspen trees and snags with
regeneration” to benefit Red-naped Sapsuckers.
12 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
The habitat objective includes the following habitat
attributes (Altman 2000c):
“… to maintain or provide some areas with natural
(e.g., fire) or mechanical disturbance regimes to
ensure proper successional development… > 10%
cover of sapling aspen in the understory to provide
adequate representation of younger seral stages for
replacement; > 4 trees and > 1.5 snags/ac > 12 m
in height and 25 cm dbh; mean canopy cover 40-80%
—either clumped with patches and openings or
relatively evenly distributed.”
Klamath Bird Observatory conducted baseline bird
monitoring in future aspen restoration sites from
2003 – 2005 including 140 point count stations during
the spring and 70 area search plots during the fall
(Stephens and Alexander 2006). Monitoring will
continue periodically after habitat management
strategies commence to evaluate the efficacy of
the treatments in achieving the desired habitat
conditions and to assess the recommendations in the
bird conservation plan for creating habitat for Red-breasted
Sapsuckers and other species associated
with aspen.
Figure 1. On Little Pend Oreille National Wildlife Refuge, the extent of old growth ponderosa pine and
potential habitat of White-headed Woodpeckers (>80 ha contiguous forest) now (above) and in 100-200 years.
13
Sacramento River and San Joaquin River
National Wildlife Refuges
These two refuges conserve riverine and floodplain
habitats along the Sacramento and San Joaquin
rivers in California’s Central Valley. Sacramento
River National Wildlife Refuge currently
manages approximately 4654 ha in 26 units along
the Sacramento River from Red Bluff south to
Princeton, California, and could expand to 7284
ha based upon the approved boundary. The San
Joaquin River National Wildlife Refuge lies in
the historic floodplain of the confluence of the San
Joaquin, Stanislaus, and Tuolumne rivers and
comprises 2428 ha west of Modesto, California in
Merced County; the approved boundary includes
5180 ha. Both refuges are important foci for riparian
restoration in California, and are identified as
conservation portfolio sites in the Riparian Bird
Conservation Plan (Riparian Habitat Joint Venture
2004).
The staff of the Sacramento River National Wildlife
Refuge drew upon nearly 15 years of riparian
restoration experience for development of the CCP
(U.S. Fish and Wildlife Service 2005b). Since 1993,
the refuge has restored approximately 1335 ha
(mostly in recently acquired orchards) of riparian
vegetation within the historic Sacramento River
floodplain. The Riparian Bird Habitat Conservation
Plan (Riparian Habitat Joint Venture 2004) provided
significant guidance on appropriate restoration
techniques to address the habitat needs of riparian
focal species. PRBO Conservation Science (PRBO)
is monitoring the bird response to the restoration
to direct future management and restoration
efforts in an adaptive management framework.
Approximately 1214 additional hectares are
planned for restoration efforts through 2015 with
management strategies to be derived directly from
the Riparian Bird Conservation Plan (Riparian
Habitat Joint Venture 2004, U.S. Fish and Wildlife
Service 2005b, Gardali et al. 2006).
In completing the CCP for the San Joaquin
River National Wildlife Refuge, the staff was
able to include the results of riparian restoration
efforts guided by the Riparian Bird Conservation
Plan (Riparian Habitat Joint Venture 2004) and
monitoring provided by PRBO (U.S. Fish and
Wildlife Service 2007). Riparian restoration at
this refuge resulted in the first recorded nesting
of endangered Least Bell’s Vireos (Vireo bellii
pusillus) in the Central Valley in over 60 years (U.S.
Fish and Wildlife Service 2005c). The restoration
incorporated native riparian vegetation such as
mugwort (Artemisia douglasiana), California
wild rose (Rosa californica), arroyo willow (Salix
lasiolepis), and valley oak (Quercus lobata);
plant species known to benefit riparian-associated
birds. The restoration design also integrated the
Riparian Bird Conservation Plan recommendation
to promote a dense, shrubby understory, an
important component in the breeding habitat of
Least Bell’s Vireos (Kreitinger and Wood 2005).
The documentation of Least Bell’s Vireos breeding
at the San Joaquin River National Wildlife Refuge
underscores the role that proper habitat restoration
and management can play in conserving biodiversity.
Conclusion
The use of PIF plans to facilitate the development
of long-term management plans on refuges in
Oregon, Washington, and California is a PIF success
story. The mission of the Fish and Wildlife Service
is “Working with others to conserve, protect, and
enhance fish, wildlife, and plants and their habitats
for the continuing benefit of the American people…”
(U.S. Fish and Wildlife Service 1999) and the
Service has primary conservation and management
responsibilities for the nation’s migratory
birds. Thus, the adoption of PIF management
recommendations into their own planning documents
is a natural union. However, PIF bird conservation
plans, and plans from waterbird and shorebird
initiatives, provide solutions not just for National
Wildlife Refuge managers, but for all land managers
tasked with meeting agency requirements for
wildlife management and conserving focal species or
birds of high conservation priority.
The responses of birds to management in quick-growing
riparian habitats can be measured within
a few years; Sacramento and San Joaquin river
refuges are good examples. Projects designed to
create old-growth conditions in younger forests, such
as at Little Pend Oreille National Wildlife Refuge,
will take much longer to measure. Regardless, it is
important to incorporate a pre- and post-treatment
effectiveness monitoring into any major project, at
the very least to measure changes in habitat and
bird abundance. The iterative loop linking planning
to management and monitoring, fundamental to
good land management and bird conservation, will
only be powerful with all three components
14 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Risk Analysis of Birds Associated with Older
Forests of the Pacific Northwest
Martin G. Raphael
conduct an assessment of risk to population viability
for each species under each of the proposed options.
The compilation of a list of birds potentially
associated with older forest began with the naming
of a terrestrial science team as part of the overall
FEMAT. This team comprised biologists from
the Forest Service, Bureau of Land Management,
National Park Service, Fish and Wildlife Service,
National Marine Fisheries Service, Environmental
Protection Agency, and various universities.
The terrestrial team compiled a list of 119 species
of birds that were thought to be associated with
forests in the plan area. The team then applied a set
of criteria to judge whether each species was closely
associated with older forest (Thomas et al. 1993,
Forest Ecosystem Management Assessment Team
1993). These criteria included:
(1) The species was statistically more abundant in
older forest than in younger forest in any part of
its range.
(2) The species reached highest abundance in older
forest (but not necessarily statistically so) and re-quired
habitat components that are contributed
by older forest.
(3) The species was associated with older forest and
was on a federal or state threatened, endangered,
or sensitive species list.
(3) Strength of association with older forest was
unknown, but the species was listed as threat-ened
or endangered and the team had reason
to suspect the species was associated with older
forest.
Of the original list of 119 species, 38 met one or
more of these criteria and were thereby classified as
closely associated with older forest (Thomas et al.
1993).
Management Goals
The FEMAT developed a set of 10 land management
options that varied in the size and distribution of
blocks of land reserved from timber harvest (Fig.
1) as well as specifications for logging and other
silvicultural procedures. The terrestrial science
team was tasked with assessing the likelihood that
Abstract
A team of scientists and managers used research
data on the relative abundance of birds in relation to
structural stage and forest attributes to list species
associated with older forest and to evaluate the
likelihood of long-term persistence of those species
under a range of forest management alternatives.
This knowledge helped craft the final design of the
Northwest Forest Plan. Research and monitoring
have been essential to the adaptive management
process, which is an inherent component of the forest
plan. Although monitoring of the two federally
listed species Marbled Murrelets (Brachyramphus
marmoratus) and Northern Spotted Owls (Strix
occidentalis caurina) is ongoing, there remains
a need to evaluate whether the plan has been
successful in meeting the needs of other forest birds.
Introduction
During the years leading to the implementation of
the Northwest Forest Plan in 1994, timber cutting
and other operations on federally managed lands
had largely been brought to a halt by federal court
orders. At issue was concern for the conservation
of biological diversity, especially for those species
that might be closely associated with older forests.
In response, President Clinton formed the Forest
Ecosystem Management Assessment Team
(FEMAT), and gave the team an objective to craft
land management options (including harvesting)
that would maintain or enhance biological diversity,
particularly that of late-successional and old-growth
ecosystems. To meet this objective, the team was
chartered to develop options that would maintain
and/or restore habitat conditions to support viable
populations, well-distributed across their current
ranges, of species known (or reasonably expected)
to be associated with old-growth forest conditions
(Forest Ecosystem Management Assessment Team
1993). This project covered federal lands within the
range of Northern Spotted Owls (Strix occidentalis
caurina), a total area of about 23 million ha, of
which 10 million ha is federal land mostly west
of the Cascade crest in Washington, Oregon, and
California. The challenges the team faced were to
first compile a list of species that were associated
with older forest within the project area, and then to
15
habitat conditions would support stable and well-distributed
populations of each species of bird under
each of the land management options. Detailed
assessments were completed for seven of the ten
options; the remaining three options (options 2, 6,
10) were relatively minor variations of other options
and did not require full assessments. These viability
assessments, conducted for birds as well as for
other vertebrates and invertebrates, were used to
help rank the relative contributions of the seven
options to overall biodiversity (Fig. 2). Results of
this assessment had a key influence on the final
decision by the Secretaries of the Departments of
Interior and Agriculture to adopt Option 9, which
ultimately was implemented as the Northwest
Forest Plan (U.S. Department of Agriculture and
U.S. Department of the Interior 1994a).
Monitoring Regime
Application of the four criteria cited above required
that FEMAT gather information on relative
abundance of forest birds in relation to structural
stage and on specific habitat elements used by each
species. Fortunately, several large scale habitat
relationships summaries and sampling programs
had been completed recently (Thomas 1979, Marcot
1984, Brown 1985, Raphael et al. 1988, Ralph et
al. 1991, Ruggiero et al. 1991), which FEMAT
relied upon to make the determinations of species’
association with older forest.
The field studies (Marcot 1984, Raphael et al. 1988,
Ralph et al. 1991, Ruggiero et al. 1991) employed
approximately comparable sampling strategies. In
each study, a large number of plots were replicated
within a range of early to late seral stages, including
both managed and unmanaged stands. Within each
plot, a set of sample stations was established and the
investigators conducted variable-radius point counts
during the breeding season to estimate relative
density of each bird species by seral stage. Studies
were carried out over 3-5 years. The combination
of studies incorporated locations throughout the
Northwest Forest Plan area.
Separate assessments were conducted for the two
listed species, Marbled Murrelets (Brachyramphus
marmoratus) and Northern Spotted Owls. For
these assessments, the species experts relied
on published and unpublished studies, including
ongoing monitoring results, to make their
determinations (Forest Ecosystem Management
Assessment Team 1993).
Response to Management
The FEMAT organized a panel of ornithologists to
perform a subjective evaluation of the likelihood that
each land management option would provide habitat
conditions to support stable and well-distributed
populations over the life of the plan (the next 100
years). Panelists relied on information about each
option (e.g., extent of reserve system, special
management provisions, projected habitat trends),
as provided by the FEMAT. They also relied on data
from the bird counts cited above. After reviewing
available materials and publications, the panelists
discussed each species in turn, and arrived at a
consensus score for each option, distributing 100
“points” among four possible outcomes:
(A) The species is stable and well-distributed on
federal lands;
(B) The species is stable but with significant gaps in
distribution with some limitation on population
dispersal;
Figure 1. Comparison of amounts of federal land in
various allocations in each of 10 land management
options considered by Forest Ecosystem
Management Assessment Team (1993). Lands
designated as matrix and adaptive management
areas are generally available for timber harvest,
whereas all other allocations are generally reserved
from harvest.
Figure 2. Relationship between species viability
(number of species of all taxa with > 60% likelihood
of habitat of sufficient quality to support stable and
well-distributed populations over 100 years) and
amount of land allocated outside of reserves (matrix,
see Fig. 1) for 7 of the 10 land management options
considered in Forest Ecosystem Management
Assessment Team (1993).
16 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
(C) The species is restricted to smaller, isolated
refugia with significant limitations on population
interactions among refugia;
(D) The species is extirpated from federal lands.
Outcomes for Marbled Murrelets (Fig. 3A) and
Northern Spotted Owls (Fig. 3B) were poorest for
options 7 and 8. Option 1, in which virtually all older
forest was protected from logging, had the highest
likelihood of outcome A for Marbled Murrelets
and the second highest likelihood for Northern
Spotted Owls. Option 9, which formed the basis of
the Northwest Forest Plan, had an intermediate
likelihood of outcome A for both species. Each
of the options was projected to support stable
and well-distributed populations (outcome A) of a
majority of the other (not listed as either threatened
or endangered) bird species (Fig. 3C). None of
these birds was projected to have any likelihood of
outcomes C or D. Options 7 and 8, which had the
lowest amount of land in reserves, had four and
nine species, respectively, with likelihoods of 20%
or greater in outcome B. Option 9, the Northwest
Forest Plan, had only one species, Black-backed
Woodpeckers (Picoides arcticus), with 20% or
greater likelihood of outcome B.
Implementation of results
Option 9 was selected as the preferred alternative in
the Environmental Impact Statement that followed
the FEMAT plan (U.S. Department of Agriculture
and U.S. Department of the Interior 1994a, 1994b).
Specific provisions to augment or meet habitat
requirements of forest birds were added to the
original design of option 9 during the transition from
the FEMAT to the Record of Decision. For Marbled
Murrelets and Northern Spotted Owls a rigorous
monitoring program was implemented (Lint et
al. 1999, Madsen et al. 1999) and both habitat and
population monitoring continues to this day (Lint
2005, Haynes et al. 2006, Huff et al. 2006, Miller et
al. 2006, Noon and Blakesley 2006, Raphael 2006a,
Falxa et al. 2009). Results of this monitoring have
been essential to managers in their evaluation of
the success of the forest plan in meeting its original
objectives for species and habitat conservation. For
Marbled Murrelets, monitoring has indicated that
populations over the bird’s range in Washington,
Oregon, and California have declined from 2000
to 2008 (Fig. 4). Monitoring shows that Northern
Spotted Owl populations have declined from 1985
to 2003 but that the rates vary across the range
(Fig. 5). Northern Spotted Owl populations are
declining at the greatest rate in the northern part
of the range, at intermediate rates in the middle
of the range, and may be stable in the southern
range (Fig. 5). For both Marbled Murrelets and
Northern Spotted Owls the forest plan has been
successful in conserving most of the higher-quality
nesting habitat within its reserve system on federal
lands. For both species, however, conditions outside
Figure 3. Predicted outcomes of each management
option for distribution and persistence of populations
over 100 years. 3A: Marbled Murrelets; 3B:
Northern Spotted Owls; 3C: the 38 other forest
birds (Forest Ecosystem Management Assessment
Team 1993). Likelihood, as indicated on the x-axis,
is the mean likelihood score calculated from the data
recorded by individual panelists.
17
Figure 4. Population estimates and 95% confidence
intervals from rangewide at-sea Marbled Murrelet
surveys (Falxa et al. 2009).
Figure 5. Estimates of mean lambda (, finite rate of
population change, with 95% confidence intervals)
for Northern Spotted Owls on 13 study areas in
Washington (WEN, CLE, RAI, OLY), Oregon
(WSR, COA, HJA, TYE, KLA, CAS), and California
(NWC, HUP, SIM). The dashed line indicates the
level for a stable population; values below that line
denote a declining population and values above that
line are increasing (modified from Anthony et al.
2006).
of the control of federal land managers (such as
oceanic conditions in the case of Marbled Murrelets,
competition from increasing Barred Owl (Strix
varia) populations for Northern Spotted Owls,
and management of forests in state or private
ownership for both species) may also be important
in determining the likelihood of species persistence.
The FEMAT envisioned a monitoring program for
other forest birds, but one was never implemented
primarily due to competing demands for limited
funding. Instead, managers rely on a variety of
other shorter term studies to evaluate the status of
birds associated with older forest.
Conclusion
Research data on the relative abundance of birds
in relation to structural stage and forest attributes
proved essential in refining a list of species
associated with older forest and evaluating the
likelihood of persistence of those species under a
range of forest management alternatives. This
knowledge helped craft the final design of the
Northwest Forest Plan, which remains one of the
world’s most comprehensive attempts to conserve
biological diversity. Research and monitoring have
been essential to the adaptive management process,
which is an inherent component of the forest plan
(Haynes et al. 2006, Raphael 2006b). Although
monitoring of the two federally listed species is
ongoing, there remains a need to evaluate whether
the plan has been successful in meeting the needs of
other forest birds.
18 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
A Watershed Analysis for Establishing
Local Population Objectives for Pacific-slope
Flycatchers and a Suite of Mid- to
Late-Successional Pacific Northwest Landbirds
Bob Altman, Michael T. Green, Barb Bresson, Erin Stockenberg, Daniel Casey, and
Susannah Casey
Abstract
We provide an example of how modeling bird-habitat
relationships with geospatial analyses can be used to
assess the capacity of the landscape to establish local
bird population objectives in support of Partners
in Flight continental population objectives, and
also provide an accounting tool for assessing the
impact of forest management on bird populations.
We initially focus on the process and outcomes for
Pacific-slope Flycatchers (Empidonax difficilis)
within the U.S. Forest Service boundaries of the
Hamma Hamma watershed in western Washington.
We then do the same analysis for a suite of mid- and
late-successional focal bird species as an example of
optimizing conservation efforts for several species
at once. Our 30-year scenario of natural succession
includes 10% harvest of the 61–80 year age class, and
100% thinning of the 41–60 year age class, in order to
increase the populations of Pacific-slope Flycatchers
by 12%, Winter Wrens (Troglodytes troglodytes)
by 11%, Varied Thrushes (Ixoreus naevius) by 8%,
and Townsend’s Warblers (Dendroica townsendi) or
Hermit Warblers (Dendroica occidentalis) by 3%.
Introduction
Forest land managers must balance the needs of
a variety of biological and non-biological factors
when making management decisions. Landscapes
that have been designed and managed to meet
these diverse needs result in an efficient use of
resources. One of the potential management targets
is bird conservation. A recent emphasis in landbird
conservation is the modeling of bird populations and
habitat relationships to provide quantitative habitat
objectives. These habitat objectives are directly
linked to bird population abundance objectives and
provide the avian component of conservation design.
One challenge for forest managers interested in
bird conservation is designing optimal landscapes
to meet the needs of multiple bird species. As an
example of how this challenge can be addressed, we
modeled bird-habitat relationships and conducted
geospatial analyses in the 16,793 ha Hamma Hamma
watershed in the Hood Canal Ranger District of the
Olympic National Forest (Fig. 1) first for Pacific-slope
Flycatchers (Empidonax difficilis) and then
three mid- and late-successional forest focal bird
species. The three additional species are Winter
Wrens (Troglodytes troglodytes), Varied Thrushes
(Ixoreus naevius), and Townsend’s (Dendroica
townsendi) or Hermit (Dendroica occidentalis)
Warblers (these two species are treated together in
this paper because of range overlap, hybridization,
and difficulties with vocal identification).
Pacific-slope Flycatcher
Population objectives.—Partners in Flight (PIF)
North American Landbird Conservation Plan
(Rich et al. 2004) used range-wide Breeding Bird
Survey (BBS) trend data (Sauer et al. 2008) to
establish an ideal (i.e., not based on potential
or capacity to achieve it) population abundance
objective to maintain the continental population of
Pacific-slope Flycatchers at the current level over
the next 30 years. These continental population
objectives were set to stimulate dialogue and
action towards conservation of continental priority
bird species. The expectation was that regional
and local assessments would be conducted to
establish habitat-based population abundance
objectives at those scales that reflect the practical
realities of those areas to contribute towards the
continental objective. Often within a species range
there is substantial variation in BBS trends from
significantly declining to significantly increasing,
and substantial variation in the problems and
opportunities for trying to maintain or increase the
species population. Thus, the variability of local
and regional conditions and the projections of how
those conditions might change over time, warrant
a habitat-based approach to developing local or
regional population objectives that are realistic
within the context of current and projected future
land uses.
Habitat relationships.—In western Washington,
Pacific-slope Flycatchers are primarily associated
19
with mesic coniferous forest, mixed coniferous-deciduous
forest, and especially deciduous
forest (Leu 2000, Pearson and Manuwal 2001).
Additionally, they are most abundant in late-successional
forest (Manuwal 1991), and occur
mostly at low to moderate elevations (generally
<1250 m; Smith et al. 1997).
Vegetation classifications.—We used the Olympic
National Forest Total Resource Inventory (TRI)
GIS layer. This layer includes over 40 forest
habitat classifications and six different forest age
classifications. The only TRI classification in the
Hamma Hamma watershed consistent with suitable
breeding habitat for Pacific-slope Flycatchers is
Figure 1. Hamma Hamma watershed, Hood Canal Ranger District, Olympic National Forest, Washington.
20 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
western hemlock (Tsuga heterophylla). Four of the
six age-classes of forest were considered suitable
habitat; 41–60 years (young forest), 61–80 years
(young/mature forest), 81–160 years (mature forest),
and > 160 years (old-growth forest).
Bird densities.—We assigned Pacific-slope
Flycatcher density values to each suitable habitat
classification based on studies that provided actual
density estimates from spot-mapping or program
DISTANCE (Thomas et al. 2003). We only used
densities from the ecological region of the Hamma
Hamma watershed (i.e., southwestern British
Columbia, western Washington, and northwestern
Oregon), and from the same habitat type (i.e.,
western hemlock) and age classes (Table 1).
Population estimates.—Using bird densities and
area of suitable habitat by age class, we estimated
the current population of Pacific-slope Flycatcher
within the study area to be 11,293 birds (Table 2).
Future population projections.—We modeled
the future population at 30-years to be consistent
with the time frame used in the PIF Continental
Plan for setting continental population objectives.
We assumed both natural succession and forest
management. Natural succession results in a gain
in population because Pacific-slope Flycatcher
densities increase in older forests (Table 2).
We used an example management scenario of 10%
harvest (i.e., clear-cut) of young/mature forest
(61–80 year age class) and 100% thinning of young
forest (41–60 year age class) based on general
knowledge of current forest management activities
in the region. In our models, harvest results in an
immediate and complete loss of habitat suitability
(and birds) in harvested stands during our time
frame of 30 years. Thinning results in an immediate
reduction of the quality of the habitat for Pacific-slope
Flycatchers (and hence densities of birds),
although returns to original densities would be
expected in the later half of our 30-year time frame
Table 1. Pacific-slope Flycatcher (Empidonax difficilis) density estimates by forest classification for the
Hamma Hamma watershed on the Olympic National Forest, Washington. Density is the mean density
(range) from various studies and is reported as birds ha-1 but equated to pairs ha-1 because the detections
are almost always singing males and presumably maed birds since the studies were conducted during the
breeding season. Sample size is the number of reported density estimates (BA).
Forest Classification Years Old Density (pairs ha-1) Sample Size
Young Forest 41-60 0.27 (0.19-0.35) 8
Young/Mature Forest 61-80 0.70 (0.27-1.09) 9
Mature Forest 81-160 0.80 (0.37-1.11) 10
Old-Growth Forest >160 1.09 (0.62-1.19) 6
Table 2. Pacific-slope Flycatcher (Empidonax difficilis) population estimates for the Hamma Hamma
watershed on the Olympic National Forest, Washington. WH = western hemlock; numbers indicate the
dominant age of the stand in years; 0 – 40 years are not presented because that age class is not considered
suitable habitat. Population (# individuals) calculated by multiplying area of habitat x bird density x two (to
account for the second individual of each pair in the population).

Forest Classification Habitat (ha) Bird Density (pairs ha-1) Population
(# individuals)b
WH 41-60 369 0.27 199
WH 61-80 1817 0.70 2544
WH 81-160 240 0.80 384
WH >160 3746 1.09 8166
Total 11,293
21
(Altman and Hagar 2006). To establish a single
density estimate covering the changes over time,
we used the percent difference of the cumulative
mean density between thinned stands versus stands
not thinned in four studies representing 1–24 years
post-thinning (Artman 1990, Hagar et al. 1996, Muir
et al. 2002, Hagar et al. 2004). This resulted in a
mean density that was 30% less in thinned habitat,
or a density of 0.19 birds ha-1. When population
losses from harvest and thinning are combined
with population gains from natural succession, the
outcome is a population of 12,600 birds (Table 3) or a
gain of 1307 birds (approximately 12%).
Alternatives to increase the population.—We
assessed two alternatives to increase the population.
A change in our management scenario to include no
thinning and no harvest results in modest population
gains (255 birds or 2% for the no harvest and 125
birds or 1% for the no thinning). However, it is
unrealistic on managed public lands to project no
harvest and no thinning.
Another consideration is to increase suitability
of existing habitat by increasing bird densities
greater than the mean density we assumed. Two
alternatives are to: 1) encourage mature deciduous
tree growth in appropriate places by creating small
openings or plantings; and 2) emphasize larger
patches of forest because the species is considered
a forest interior species with increased densities in
larger patches (Rosenberg and Raphael 1986, Brand
and George 2001, George and Brand 2002). In order
to achieve significant gains in population from these
alternatives they would have to be implemented
extensively across the landscape, and that is simply
unrealistic. Additionally, the time to achieve these
habitat conditions is well beyond our 30-year time
frame.
Optimization with a Suite of Focal Species
Our analysis so far assumes management in the
Hamma Hamma watershed only for the habitat
needs of Pacific-slope Flycatchers, an unlikely
scenario because management for a single species is
generally not conducted unless it is a federally-listed
threatened or endangered species. Additionally,
there are many other management considerations
that would likely need to be applied to the region,
including consideration of Late Successional
Reserves (i.e., mature and old-growth forests
designated for conservation under the Northwest
Forest Plan; Forest Ecosystem Management
Assessment Team 1993) and harvest targets for
timber management, as well as management for
other bird species of interest.
In the interest of developing a more inclusive and
realistic model, we assessed the effects of this
Pacific-slope Flycatcher management scenario on
Table 3. Pacific-slope Flycatcher (Empidonax difficilis) population projections in 30 years with natural
succession and management (10% harvest of 61–80 year age class and 100% thinning of 41–60 year age
class) in the Hamma Hamma watershed on the Olympic National Forest, Washington. WH = western
hemlock; numbers indicate the dominant age of the stand in years. New habitat assumes equal distribution
of hectares among age classes when adding 30 years (which moves old habitat into one or two new habitat
age classes) thus proportioning of hectares into new age classes is necessary. Population (# individuals)
calculated by multiplying area of habitat x bird density x two (to account for the second individual of each
pair in the population).


Forest
Classification
Old Habitat (ha) New Habitat (ha) Density (pairs ha-1) Population
(# individuals)
WH 0-20 819a
WH 21-40 775a
WH 41-60 369 798b 0.19c 303
WH 61-80 1817 573d 0.70 802
WH 81-160 240 2002e 0.80 3203
WH >160 3746 3804f 1.09 8292
Total 12,600
a Not considered suitable habitat, but presented because these numbers figure in the calculation of future suitable habitat due to natural
succession.
b Calculated by adding 50% of the 21–40 year age class + 50% of the 0–20 year age class.
c Percent difference of the cumulative mean density between thinned versus unthinned in four studies (see text) representing 1–4 years
post-thinning (i.e., 30% lower density in thinned) applied to the existing mean density in the 41–60 year age class.
d Calculated by adding 50% of the 41–60 year age class + 50% of the 21–40 year age class.
e Calculated by adding all of the 61–80 year age class (after 10% harvest) + 50% of the 41–60 year age class + 76% (prorated) of the
existing 81 – 160 year age class that remains as 81 – 160.
f Calculated by adding all of the > 160 age class + 24% (prorated) of the 81–160 year age class that advances to > 160.


22 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
three other forest species to see whether or not we
could maximize (optimize) bird conservation. The
additional birds are focal species (Lambeck 1997)
in the Oregon-Washington PIF Bird Conservation
Plan (Altman 1999) that represent a suite of desired
habitat conditions within mid- and late-successional
forests including Winter Wrens (complex
understory), Varied Thrushes (multi-layered
midstory), and Townsend’s and Hermit Warblers
(high canopy cover). These four focal species
complement Pacific-slope Flycatchers’ habitat
(deciduous tree component) and capture the desired
habitat conditions of most bird species in mid- and
late-successional forests.
Continental population objectives.—All four focal
species are species of continental importance in the
PIF Continental Plan, and all have objectives to
maintain the current population abundance at the
continental level over the next 30 years (Rich et al.
2004).
Habitat relationships.—Among all four bird
species, suitable breeding habitat only occurs in
stands > 40 years old, and includes four TRI habitat
classifications; silver/noble fir (Abies amabalis/
procers), western hemlock, mountain hemlock
(Tsuga mertensiana), and silver fir/mountain
hemlock. There were no forest classifications in
which all four bird species occurred (i.e., the same
habitat and age class). Winter Wrens had the
broadest habitat range including all four habitats
and all elevations; however, there are areas of
overlap in habitat among the four species (Table 4).
Bird densities and population estimates.—For
each focal species, we assigned density values for
each forest type and age class as described earlier
for Pacific-slope Flycatchers (Appendix). Existing
population estimates were derived by multiplying
bird densities by area of suitable habitat.
Future population projections.—We modeled
the future population for each focal species under
the same scenario as described earlier for Pacific-slope
Flycatchers. We used the same method for
calculating overall 30-year mean densities in the
41–60 year age class thinned stands as we did
for Pacific-slope Flycatcher. The quantitative
differences in mean densities between stands that
were thinned and not thinned over the 30-year
period were: Winter Wrens, a 21% higher density in
thinned; Varied Thrushes, a 16% higher density in
thinned; and Hermit and Townsend’s Warblers, a 7%
lower density in thinned. When population losses
from harvest and losses or gains from thinning
are combined with population gains from natural
succession, the predicted outcome is population
gains of 1724 (11.1%) for Winter Wrens, 433 (8%)
for Varied Thrushes, and 71 (3%) for Hermit or
Townsend’s Warblers (Appendix).
Assessing impacts on bird populations.—In
addition to establishment of population objectives,
our bird-habitat modeling, geospatial analyses, and
optimization provides forest managers a process
for efficient bird conservation design and assessing
outcomes of management on bird populations. We
provide few example scenarios within the Hamma
Hamma watershed that maximize bird conservation
through natural succession, minimize the negative
population impacts of harvest, and manage species-specific
population losses and gains resulting from
thinning (Table 5).
Discussion
Population objectives.—The future management
options we described within the Forest Service
lands of the Hamma Hamma watershed results
in objectives to increase the population by
approximately 12% for Pacific-slope Flycatchers,
11% for Winter Wrens, 8% for Varied Thrushes,
and 3% for Hermit and Townsend’s Warblers.
These are modest gains over a 30-year period,
but since much of this part of the watershed is
already in late-successional forest there are limited
opportunities for increasing populations of late-successional
bird species. If the analyses were
conducted for the entire watershed, the remainder
of which is comprised of private forest lands and
likely in much younger age classes, there would be
more possibilities to increase populations with some
Table 4. Habitat compatibility among four focal species in the Hamma Hamma watershed on the Olympic
National Forest, Washington.
Species Combinations Habitats and Elevations
Winter Wrens and Hermit and Townsend’s Warblers Silver/noble fir < 500 m
Pacific-slope Flycatchers and Winter Wrens Western hemlock < 500 m
Winter Wrens and Varied Thrushes All habitats > 1250 m and mountain
hemlock and silver fir/mountain hemlock
500–1250 m
Winter Wrens, Varied Thrushes, and Hermit and
Townsend’s Warblers
Silver/noble fir 500–1250 m
Pacific-slope Flycatchers, Winter Wrens, and Varied Thrushes Western hemlock 500–1250 m
23
targeted management for mid- and late-successional
forests. Conversely, much of this land is intensively
managed for timber production and harvested
before achieving mid- to late-successional status,
so opportunities for increasing populations would
be negated to some degree by the realities of land
ownership and management.
It is noteworthy that Pacific-slope Flycatchers, the
species most negatively affected by thinning over a
30-year time frame, shows the highest population
increase (i.e., the highest population objective). This
is because it occurs in the highest densities of the
four species, and its only suitable habitat, western
hemlock, is the dominant forest type in the study
area. Thus, despite losses in population due to
thinning, it benefits greatly from the large amount of
natural succession in western hemlock and the most
birds per unit area in that habitat. This exemplifies
the need to consider all management scenarios and
long-term objectives, including natural succession,
rather than just assessing short-term impacts based
on a species response to one management activity.
Our analysis is presented as an example of how using
geospatial data and bird-habitat data can be used
to develop bird population objectives. These same
types of analyses should be routinely done as part of
forest planning throughout western Washington and
elsewhere to determine cumulatively what a region
can contribute towards the continental population
objectives of these and other bird species.
Management impacts.—Our process of using
bird-habitat data and geospatial analyses can
be a valuable “accounting” tool for assessing
management impacts directly on bird populations
rather than indirectly on bird habitat. The results
of the analyses allow for comparative accounting
of impact on bird populations among alternatives,
and thus can be used to advance strategic bird
conservation. This tool has many additional
potential applications for use in projects such
as environmental assessments, land acquisition
evaluations, and restoration proposals.
It is important to recognize that our example
optimization analysis is not complete. Our example
needs to be integrated with a similar analysis of
a suite of early-successional focal bird species
to balance their habitat needs and population
objectives. Additionally, there are many non-bird
considerations that would need to be applied. These
comprehensive types of analyses will be necessary
across regional landscapes not only to determine
optimal bird conservation, but efficient management
and conservation of all natural resources. Finally,
we did not conduct an analysis of demographic data
to provide complementary population objectives
for primary population parameters such as
reproduction, survivorship, or recruitment into the
population. This should be done in concert with the
analysis described herein for population abundance
to provide population objectives for both primary
and secondary population parameters.
Table 5. Example management objectives to maximize bird focal species conservation in the Hamma Hamma
watershed on the Olympic National Forest, Washington.
Management Ideal Focal Species Scenario Example Objective Focal Species Rationale
Natural
Succession
Manage where most species
occur, where their densities
are high, and where most
habitat occurs
Allow succession
to occur in western
hemlock 500-1250
meters
Benefits 3 of the 4 species
Thinning Conduct least where Pacific-slope
Flycatchers and
Hermit/Townsend’s Warblers
occur, and most where Winter
Wrens and Varied Thrushes
occur
Thin in silver/noble
fir and in western
hemlock >500 meters
Limits negative population
effects on Pacific-slope
Flycatchers and Hermit/
Townsend’s Warblers, while
enhances positive population
effects on Winter Wrens and
Varied Thrushes
Harvest Conduct where fewest
number of species occur, and
where their densities are low
Harvest in silver/noble
fir > 1250 meters
Harvest in western
hemlock < 500 meters
Affects only 2 of 4 species
including Varied Thrushes which
has lowest densities
Affects only 2 of 4 species and
limits negative effects on Pacific-slope
Flycatchers which has
highest densities
24 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Acknowledgments
Funding for this work was provided by the Bureau
of Land Management, Oregon State Office. We
thank Arvind O. Panjabi, C. John Ralph, Jaime L.
Stephens, and an anonymous reviewer for comments
on earlier drafts.
Appendix.
Existing and 30-year population estimates for Winter Wrens, Varied Thrushes, and Hermit and Townsend’s
Warblers with natural succession and management (10% harvest of 61-80 age class and 100% thinning of 41-60
age class) in the Hamma Hamma watershed on the Olympic National Forest, Washington. WH = western
hemlock; SN = silver/noble fir; MH = mountain hemlock; SM = silver fir/mountain hemlock. Densities (pairs
ha-1) are mean densities from reported studies (sample size). Densities without sample sizes were projected
based on known densities from other age classes. Population (# individuals) calculated by multiplying area
of habitat x bird density x two (to account for the second individual of each pair). Under Future Projections,
New Habitat (ha) assumes equal distribution of hectares among age classes when adding 30 years which
moves old habitat into one or two new habitat age classes; thus, proportioning of hectares into new age classes
is necessary.
Habitat (ha) Densities (pairs ha-1) Population (# individuals)
Forest
Class
Winter
Wrens
Varied
Thrushes
Hermit/
Townsend’s
Warblers
Winter
Wrens
Varied
Thrushes
Hermit/
Townsend’s
Warblers
Winter
Wrens
Varied
Thrush
Hermit/
Townsend’s
Warblers
Existing Conditions and Population Estimates
WH 0-20 819.00a 785.00a
WH 21-40 790.42a 750.42a
WH 41-60 369.29 227.28 0.29 (8) 0.05 (12) 214.19 22.73
WH 61-80 1826.86 778.23 0.38 0.14 (4) 1388.41 217.90
WH 81-160 240.43 240.43 0.49 (9) 0.18 (11) 235.62 86.55
WH >160 3848.74 3181.27 0.94 (5) 0.21 (8) 7235.63 1336.13
SN 0-20 240.00a 240.00a 238.00
SN 21-40 90.31a 90.31a 89.80
SN 41-60 28.13 28.13 4.25 0.27 (2) 0.12 (2) 0.86 (14) 15.19 6.75 7.31
SN 61-80 744.51 744.06 686.53 0.39 0.28 0.78 580.72 416.67 1070.99
SN 81-160 92.51 92.51 92.47 0.55 0.30 (2) 0.67 (3) 101.76 55.51 123.91
SN >160 2615.31 2611.91 2522.79 0.72 (2) 0.47 (3) 0.29 (9) 3766.05 2455.20 1463.22
MH 61-80 108.88 108.88 0.45 0.28 97.99 60.97
MH 81-160 232.32 232.32 0.59 0.31 (4) 274.14 144.04
MH >160 972.85 972.85 0.77 (2) 0.47 (3) 1498.19 914.48
SM 61-80 87.48 87.48 0.51 0.26 89.23 45.49
SM 81-160 0.75 (4) 0.31 (4)
Totals 15,497.12 5762.43 2665.43
Continued on next page
25
New Habitat (ha) Densities (pairs ha-1) Population (# individuals)
Forest
Class
Winter
Wrens
Varied
Thrushes
Hermit/
Townsend’s
Warblers
Winter
Wrens
Varied
Thrushes
Hermit/
Townsend’s
Warblers
Winter
Wrens
Varied
Thrush
Hermit/
Townsend’s
Warblers
Future Projections of Habitat and Population Estimates
WH 41-60 804.71b 767.71b 0.35c 0.06c 563.30 92.12
WH 61-80 579.86d 488.85d 0.38 0.14 (4) 440.69 136.88
WH 81-160 2011.55e 996.77e 0.49 (9) 0.18 (11) 1971.32 358.84
WH >160 3906.44f 3238.97f 0.94 (5) 0.21 (8) 7344.11 1360.47
SN 41-60 165.16b 165.16b 163.90b 0.33c 0.14c 0.80c 109.61 46.25 262.24
SN 61-80 59.22d 59.22d 47.03d 0.39 0.28 0.78 46.19 33.16 73.37
SN 81-160 754.43e 754.03e 690.28e 0.55 0.30 (2) 0.67 (3) 829.87 452.42 924.98
SN >160 2637.51f 2634.11f 2544.98f 0.72 (2) 0.47 (3) 0.29 (9) 3798.01 2476.06 1476.09
MH 61-80 54.44d 54.44d 0.45 0.28 49.08 30.49
MH 81-160 274.56e 274.56e 0.59 0.31 (4) 323.98 170.23
MH >160 1028.61f 1028.61f 0.77 (2) 0.47 (3) 1584.06 966.89
SM 61-80 43.74d 43.74d 0.51 0.26 43.74 22.75
SM 81-160 78.73e 78.73e 0.75 (4) 0.31 (4) 118.10 48.81
Totals 17,221.46 6195.37 2736.68
Number of birds gained in population 1724.34 432.94 71.25
Percent population gain (i.e., population objective) 11.1% 7.5% 2.7%
a Not considered suitable habitat, but area presented because these numbers figure in the calculation of
future suitable habitat due to natural succession.
b Calculated by adding 50% of the 21 – 40 year age class + 50% of the 0 – 20 year age class.
c Densities are different from existing conditions densities due to thinning. Calculation is the percent
difference of the cumulative mean density between thinned versus unthinned in four studies (see text)
representing 1 – 24 years post-thinning (i.e., 30% lower density in thinned) applied to the existing mean
density in the 41– 60 year age class.
d Calculated by adding 50% of the 41 – 60 year age class + 50% of the 21 – 40 year age class.
e Calculated by adding all of the 61 – 80 year age class (after 10% harvest) + 50% of the 41 – 60 year age
class + 76% (prorated) of the existing 81-160 year age class that remains as 81 – 160.
f Calculated by adding all of the > 160 year age class + 24% (prorated) of the 81–160 year age class
Continued from previous page
26 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
Demographic Monitoring, Modeling, and
Management of Landbird Populations in Forests
of the Pacific Northwest: An Application of the
MAPS Dataset
M. Philip Nott, and Nicole L. Michel
Abstract
Pacific Northwest forests support over a hundred
breeding landbird species (including many
Neotropical migrants) in a variety of forested,
meadow, shrub, and riparian habitats. With the
need for increased management to both maintain
the health of those habitats and reduce the risk of
wildfire managers need tools to assess the effect
of their management. Additionally, these habitats
and the birds that breed in them face increasingly
variable environmental conditions due to recent and
extreme fluctuations in weather patterns driven by
cyclical phenomena associated with the Pacific (e.g.
the El Nino Southern Oscillation) and Atlantic (e.g.
the North Atlantic Oscillation) oceans. Demographic
monitoring of the avifauna can help determine the
proximal causes of population change (i.e., whether
changes are linked to survival rates and/or to
reproductive effort). Survival rates are likely mostly
influenced by conditions during the non-breeding
season whereas reproductive effort is likely most
influenced by conditions just prior to and during the
breeding season and by the pattern and health of the
forested landscapes.
The Institute for Bird Populations, monitored 21
landbird species in six national forests and calculated
their survival rates and annual reproductive indices.
Of these 21 species, we identified 13 species of
conservation concern that were listed in federal,
regional, and state conservation plans. For these
13 species, we constructed species-landscape
models from which we formulated management
guidelines to maintain or create landscapes
that support healthy productive populations.
GIS-based simulations can be used to generate
post-management landscapes, the spatial statistics
of which can be used to populate multiple species-landscape
models. In this way, managers can assess
the effects of alternate management scenarios
(or natural disturbances) on breeding landbird
populations.
Introduction
The U.S. Forest Service Pacific Northwest Region
manages 19 national forests that provide timber,
forage for cattle and wildlife, and numerous
recreational opportunities. These and similar
activities on lands surrounding national forests
affect avian communities through alteration or
removal of their preferred habitats.
In 1993, the Pacific Northwest Forest Plan emerged
for coordinating forest management actions
with federal agencies and state, local, and tribal
governments across Oregon, Washington, and
California. The plan includes strategies for adaptive
forest management, conservation and restoration
of riparian habitat, and the protection of sensitive
species on federal forestlands (U.S. Department of
Agriculture and U.S. Department of the Interior
1994a).
In addition, Partners in Flight formulated avian
conservation plans (Rich et al. 2004) at the federal,
regional, and state levels that list species of
conservation concern and the critical habitats that
they require to successfully breed. These plans call
for adaptive management guidelines to maintain or
improve habitats for species of conservation concern.
It is essential, therefore, to construct appropriately
scaled ecological models that can quantify the effects
of changing landscape pattern and structure on
avian population dynamics. Such models could be
used by land managers as decision-making tools to
enable them to predict the effects of proposed forest
management activities on avian demographics,
including population densities, population
trajectories, and reproductive success.
Developing Species-Habitat Models from
Monitoring Avian Productivity and Survivorship
Data
The Institute for Bird Populations (IBP), through
collaboration with (and funding from) U.S. Forest
Service, Pacific Northwest Region Six established
27
36 demographic monitoring stations under the
Monitoring Avian Productivity and Survivorship
(MAPS; DeSante et al. 1995, DeSante and Nott 2001)
program (Fig. 1; Table 1). Since 1992, these stations
have effectively monitored 21 landbird species on six
national forests of the Pacific Northwest. Of these
21 species, we constructed species-landscape models
for 13 species.
We collected breeding season mist-netting and
banding data from 36 constant-effort monitoring
stations (Nott et al. 2005). In 1992, six stations
were established on each of six national forests
(Fig. 1; Table 1): two in Washington, and four in
Oregon. We collated and analyzed banding data
(1992 - 2001) from each station to obtain study-wide,
forest-specific, and station-specific demographic
parameters for 21 species (Nott et al. 2005). Of
these, species-landscape models were constructed
for 13 species of management concern whose
demographics could be modeled (minimum of eight
stations each capturing 2.5 adult birds per year) and
that were also included in federal, regional, or state
conservation plans.
We defined two sets of MAPS stations in this
investigation. A “Northwest Forests” set included
those 36 MAPS stations operated on national forest
lands with the financial and logistical support of
the U.S. Forest Service Region 6 (Fig. 1; Table
1). A more spatially extensive “Pacific Northwest
Regional” set (not shown) included the Northwest
Forest set as well as 150+ “independent” stations
operated by public agencies, academic institutions,
private organizations, and individual bird banders.
We used the Pacific Northwest Regional dataset
to correct the raw MAPS data for missed banding
effort (Nott and DeSante 2002a) as defined by
the MAPS constant-effort mist netting protocol
(DeSante et al. 2010) and effort correction algorithm.
The diurnal- and seasonal-correction models (Nott
and DeSante 2002b) were then applied to the less
extensive Northwest Forests dataset to determine
the forest-specific avian demographics subsequently
Figure 1. Clusters of MAPS stations (red circles) located on six named national forests (green) in Washington
(2), and Oregon (4), where landbird species of conservation concern have been monitored by the Institute for
Bird Populations (IBP) since 1992. Other MAPS stations that have been active for four or more years but
not operated by IBP are shown as black dots. MAPS stations are superimposed upon federally-managed
lands as denoted by yellow (Tribal Land), light tan (Bureau of Land Management), brown (Bureau of
Reclamation), gray (Department of Defense), green (Forest Service), orange (Fish and Wildlife Service), and
blue (National Park Service).
28 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
used to parameterize landscape management models
for birds of management concern. This process
resulted in station- and species-specific annual
numbers of adult and young birds. Reproductive
success was expressed as the ratio of young to
adults.
For each species of management concern, we
analyzed MAPS banding data and combined
demographic estimates with five regional spatial
datasets: USGS National Land Cover Dataset
(Vogelmann et al. 2001), USFS Region 6 canopy
cover, USGS National Elevation Dataset (NED),
Streamnet, and the USFS Forest Health Protection
Aerial Survey (McConnell et al. 2000). From
these we constructed landscape-scale (1000’s of
hectares) “species-landscape” models that describe
demographic parameters as functions of the land
cover (e.g. coniferous cover), canopy cover, edge type
(e.g. forest-grassland), topography, water features
(e.g. permanent stream density), and defoliation
indices that represent the frequency and intensity of
spatial pest outbreak data.
These species-landscape models can be used to
predict the likely effect of forest management on
adult population density and reproductive success
for multiple species.
Management Goals
Forest management can change landscape patterns
and structures that in turn can change avian
diversity and local population trajectories (Mitchell
et al. 2006). The species-landscape models provide
tools that allow managers to assess the effects of
management on species of conservation concern.
In order to validate the models we must also,
where possible, monitor the “effectiveness” of that
management.
Accordingly, the next stage in the adaptive
management cycle was to identify stations at which
particular management could be applied that I
expect to benefit species of conservation concern and
to reorganize our network of monitoring stations
to monitor the effectiveness of past or future
management. In 2004 - 2005, we discontinued five
Table 1. The direction of the forest-wide trend for each of 13 species (eight Neotropical and five short-distance
migrants) of regional conservation concern is indicated as decreasing (-) or increasing (+), and
significance is indicated by multiple plus or minus characters (e.g. + = non-significant, ++ = 0.05 ≤ P
< 0.10, +++ = 0.01 ≤ P < 0.05). The species of forest-specific management concern for which adult
populations are declining significantly at one or more stations are shown shaded. For each national forest,
the number of species of management concern (declining significantly at one or more stations) is given
with the numbers of species with declining or increasing trends.




Species of regional conservation
concern Baker Wenatchee Umatilla Willamette Siuslaw Fremont
Neotropical migrants
Hammond’s Flycatchers - + —- +++ ++
“Western” Flycatchers - - —- +
Warbling Vireos + - —— + -
Swainson’s Thrushes ++ + —- + +
MacGillivray’s Warblers - - —- - +
Wilson’s Warblers + + - +++ + -
Chipping Sparrows - —-
Lincoln’s Sparrows —- - - -
Short-distance migrants
Chestnut-backed Chickadees - +++ - -
Winter Wrens - +++ ++ -
Song Sparrows + — +++ +
Dark-eyed Juncos - ++ —- - ++
Pine Siskins + - —- -
Total management concern 3 4 8 4 3 2
Total declining 6 5 9 6 3 4
Total increasing 4 6 1 6 3 4
29
stations and reestablished them in other parts of
the forest to better monitor species of conservation
concern, and measure the effects of thinning
practices on their avian populations by locating new
stations in similarly treated forests. We continue to
operate the remaining 30 stations as control stations;
they effectively monitor a number of species of
concern in areas that are not managed.
Monitoring Regime
We used the MAPS monitoring protocol (DeSante
et al. 2008). Each station consists of 10 nets located
in the same place each year and, every ten days
for three months, opened for six hours following
sunrise. Birds are identified to species, age, and
sex and marked with a federal band; in addition,
morphometric (e.g. wing length, weight, etc.) and
molt pattern data were recorded (DeSante et al.
2008).
Response to Management
Analyses of the demographic data revealed the
direction and significance in adult population trends
from MAPS data pooled by two national forests in
Washington (Mount Baker and Wenatchee), and four
in Oregon. Few stations were affected by nearby
management during the period 1992 - 2001, so we
can assume that these trends (Table 1) are the
result of species response to historical (pre-1992)
management or prevailing abiotic conditions. We
hypothesized that the density and reproductive
success of the species breeding there are a
response to the landscape pattern resulting from
historical management at the level of the landscape
surrounding each MAPS station. By quantifying
these responses we can construct models that can be
used to reverse the declines.
Results of Models
The species-landscape models can be used to predict
the likely effect of forest management on adult
population density and reproductive success for
multiple species. For example, the models can be
used in the following manner to assess the effect of
small clear cut:
(1) Select a 2 km radius of the landscape centered on
the proposed cut.
(2) Gather relevant spatial statistics (to populate
parameters of each model (e.g. percent cover of
deciduous forest) using FragStats (McGarigal
and Marks 1995) or equivalent.
(3) Estimate pre-management numbers of birds and
reproductive indices.
(4) Simulate proposed management in multiple lay-ers
of a GIS application.
(5) Repeat spatial analysis to populate parameters of
each model (repeat 2).
(6) Estimate post-management numbers of birds
and reproductive indices.
(7) Compare pre- and post-management predictions
of population density and reproduction to assess
the impact of the proposed management on each
species.
Adjustments to the simulated management can
be made to selectively benefit one or more species
or guild. For instance, to minimize the effect of
clear-cutting (e.g. 100 ha of 1250 ha) upon species
that requires large forest patches (e.g. Swainson’s
Thrush) you might cut a single 100 ha block and
orient that cut to leave the largest uniformly shaped
contiguous patch of low canopy cover coniferous
forest possible. However, to maximize habitat for a
species that prefers forest-shrub edge habitat many
small narrow cuts should be made. In this way the
models can be applied to multiple species and act
as decision-making tools for managers to create or
maintain high quality breeding habitat for species
of regional or local conservation concern. Similarly,
these models can be used to assess the consequences
of proposed management upon local avifauna, or
used in a “what if ” sense to formulate management
plans that maximize the benefits to multiple species.
Continued monitoring of demographic performance
measures (Nott and Morris 2007) in managed and
unmanaged areas provides the ability to assess the
efficacy of management or track the consequences of
natural disturbances.
We summarize the general interpretations of species
landscape models for each species and demographic
for which statistically significant and interpretable
models emerged. Overall, selected models for
forest-dwelling species suggest that management
plans should aim to conserve large areas of
contiguous forest, upwards of 900 ha (72%), in a 2
km radius landscape covering 1250 ha. Within those
forested areas, canopy cover, as well as the density of
undergrowth and ground cover, should be managed
in a manner consistent with published habitat
management procedures for each target species.
Riparian, deciduous, and edge habitat also emerged
as important components of several species’ habitat
requirements.
Hammond’s Flycatchers (Empidonax
hammondii).—To maintain healthy and productive
Hammond’s Flycatcher populations, land managers
should create a shifting mosaic of successional or
low canopy cover habitat (covering 10 – 20% of the
landscape) within extensive stands of uniformly
shaped coniferous forest or woodland covering
80-90% of the landscape. Because reproductive
success responds negatively to stream density, such
management would best be applied to the drier
higher elevation (600 - 1800 m) coniferous stands.
30 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
“Western” Flycatcher - The term “Western”
Flycatcher refers to the occurrence of Pacific-slope
(Empidonax difficilis) or Cordilleran
(Empidonax occidentalis) flycatchers which cannot
be distinguished from one another in the “hand”
where their ranges overlap. “Western” Flycatchers
as a group are sensitive to proximal edges (i.e., patch
size) of coniferous habitat; smaller patch sizes might
result in higher risk of nest predation and parasitism
(Robinson et al. 1995). The numbers of young and
reproductive success are higher at those stations
associated with a high total core area of coniferous
forest habitat totaling 72% of the landscape. Large
tracts of old-growth forests (large core areas of
coniferous forest) and dry-upland and riparian
sites (thinner canopy and some mixed habitats) are
beneficial to the reproductive success of “Western”
Flycatchers.
Warbling Vireos (Vireo gilvus).—Warbling Vireos
are associated with large tracts of coniferous forest,
and with forest-successional and forest-grassland
edge components. This suggests that creation
of regeneration gaps could create productive
habitat. However, the pattern of the logging may
be important. The results suggest that at high
elevations, large tracts of open coniferous forest
interspersed with larger patches of successional
habitat create good habitat for successful breeding.
Chestnut-backed Chickadees.—Chestnut-backed
Chickadees are best managed through the creation
or maintenance of open (thin-canopied) forest and
forest-successional habitat edge, especially at higher
elevations where pest damage was high. However,
extensive riparian habitat, as reflected by stream
density, was associated with increasing trends in the
numbers of young and with reproductive success.
Other research suggests that pest infestation is
a natural process that benefits bird populations
(Torgersen et al. 1990, Crawford and Jennings 1989);
while increased magnitude and extent of damage due
to several species of bark beetle at higher elevations
is likely a result of recent climate change and results
in reducing the core area of forest and thinning
canopy cover (Raffa et al. 2008). Our results show
strong positive correlations between Chestnut-backed
Chickadee demographics and elevation
(spatial mean), the extent of successional habitat,
and cumulative bark beetle damage.
Winter Wrens (Troglodytes troglodytes).—Higher
populations and greater reproductive success of
Winter Wrens were associated with large areas
of evergreen forests. However, population sizes
and reproductive success seem to be increasing
over time in areas that were classified as thinner
forest with successional habitat and a deciduous
component. These results suggest that the best
way to manage for Winter Wrens would be to
maintain large uniformly shaped patches of thinner-canopy
evergreen forests in stream-dense areas.
In addition, smaller patches of mixed or deciduous
forests (associated with riparian areas and covering
greater than 10% of the area) should be maintained.
Swainson’s Thrushes (Catharus ustulatus).—
Within coniferous forests, adult populations of
Swainson’s Thrushes required large patches
(representing 10% or more of the landscape) of
dense, low-elevation, deciduous and mixed-deciduous
forests, with high canopy cover (i.e. mature lowland
forests). However, numbers of young and increased
reproductive success benefit from large patches
(> 16% of the landscape) of more open deciduous
and mixed habitat forests. The selection of highly
correlated core area variables in these models
supports previous findings of “edge sensitivity”
for this species (Brand and George 2001). This
emphasizes the need to conserve large tracts of
contiguous forest in lowland areas where moister
forests and riparian areas occur. The presence of
grassland and successional habitat is deleterious to
population dynamics. These results suggest that the
riparian management, currently being implemented
across the region, should lead to increases in
Swainson’s Thrush populations.
Inspection of the landscape data associated with
the 25 MAPS stations used in Swainson’s Thrush
analyses reveals that coniferous forest was the
dominant habitat type covering 50-90% of the 1250
hectares within a 2-kilometer radius of each station.
Deciduous and mixed forest coverage, combined,
accounted for up to 500 hectares (approx. 40%) of the
remaining areas (e.g. station 11166) and averaged
13% of the cover. The coverage of successional
habitat was consistently under 35 hectares (approx.
3%) except for stations 11143 (~9%) in Mount
Baker N.F., 11154 (~40%), 11155 (~15%), and 11156
(~35%) in Umatilla N.F. We reported statistically
significant correlations between demographics and
landscape variables. At this sampling level (n = 25)
two-tailed critical values of Pearson’s correlation
coefficient (r) lie at 0.337 (P < 0.10), 0.462 (P <
0.05) and 0.505 (P < 0.01). Figure 2 shows the
forest fragmentation patterns associated with three
Willamette MAPS stations; a fragmented high
elevation station where thrushes’ adult abundance
and reproductive index were low; two lower
elevation stations which were less fragmented and
supported higher abundances and productivity
levels.
MacGillivray’s Warblers (Oporornis tolmiei).—
MacGillivray’s Warblers at higher elevations are
best managed by maintaining large patches of
successional habitat interspersed among low to
medium canopy cover coniferous forest. Such a
coarsely grained habitat should feature extensive
successional habitat-forest edge. Although no
strong correlations were found between stream
density (indicative of the extent of riparian or
meadow habitat) and demographic variables, stream
density was generally high at the stations included in
this study.
31
Wilson’s Warblers (Wilsonia pusilla).—Adult
Wilson’s Warblers abundance are most closely
associated with deciduous habitats with successional
habitat edge. However, the models also suggest
that reproductive success was higher in successional
habitats where the adults were less common.
Therefore, riparian management zones do not
appear to be as important to Wilson’s Warblers
as extensive high canopy cover deciduous forests.
If riparian management zones include areas of
deciduous forest, we predict that they will be
beneficial to this species. We recommend the
maintenance of high canopy cover deciduous or
mixed forest in excess of 60% of the landscape and
narrow successional habitat cover in excess of 4%.
Chipping Sparrows (Spizella passerina).—Chipping
Sparrow models were weak but suggested that the
maintenance of a coarse grained, heterogeneous
forested landscape featuring larger patches of
successional habitat and grassland should benefit
Chipping Sparrow populations.
Song Sparrows (Melospiza melodia).—Song
Sparrows appear to be edge-sensitive; thus,
maintaining or creating large patches of low canopy
cover evergreen forest in stream-dense areas should
benefit adult and young populations and lead to high
reproductive success. The results also suggest that
defoliation events may help create suitable habitat
for Song Sparrows by thinning the canopy. The
extent of successional habitat should be held at
less than 3%. It is possible that mechanical canopy
thinning may also benefit Song Sparrow populations.
Grazing exclusion and creek restoration will help
restore higher elevation habitat of Song Sparrows.
Lincoln’s Sparrows (Melospiza lincolnii).—
Maintaining coarse grained habitat heterogeneity
(meadow and successional) among high elevation
moist coniferous forests is beneficial to Lincoln’s
Sparrow populations. At high elevations, frequent
natural disturbances such as defoliation events may
be responsible for the development of dense scrubby
patches and edge habitats where Lincoln’s Sparrows
prefer to breed. Adults responded negatively to
grassland area but young responded positively.
Larger patches appear to represent better quality
habitat in which individuals produce more offspring,
whereas smaller patches are available to non-breeders
or less fit individuals. This pattern fits
an ideal despotic distribution which is commonly
associated with the population dynamics of sparrows
and other species (Moller 1991).
Dark-eyed Juncos (Junco hyemalis).—Maintaining
coarse grained heterogeneity among drier, higher
elevation coniferous forests benefits Dark-eyed
Juncos. At high elevations, frequent natural
disturbances such as defoliation events may be
responsible for the development of dense scrubby
patches and edge habitats where Dark-eyed Juncos
populations appear to thrive. However, some
Figure 2. Aerial land cover images of 2 km radius
(oval due to projection) landscapes derived from
the National Land Cover Dataset (2001), associated
with three MAPS stations that monitor Swainson’s
Thrush on Willamette NF. The forested (green)
landscape surrounding the Clear Cut (#11160)
station is more fragmented by shrub/successional
habitat (tan) than that surrounding the stations
Major Prairie (#11161), and Brock Creek (#11162).
The latter two stations are at ~700m elevation
and support stable and abundant adult population
(10 and 13 adults per year, respectively) with high
productivity indices (0.15 and 0.29, respectively),
whereas Clear Cut at ~1300m elevation supports
half the adult abundance (6 adults) with a
productivity index of only 0.06.
32 Informing Ecosystem Management: Science and Process for Landbird Conservation in the Western United States
populations thrived in areas where a mosaic of larger
regeneration cuts had been created.
Pine Siskins (Spinus pinus).—Maintaining large
contiguous (low levels of fragmentation) tracts
of drier, high-elevation, coniferous forests is
beneficial to Pine Siskins. Although populations
declined at 11 of 13 stations they declined slower
at stations dominated by high canopy cover
forest. Interestingly, cumulative pest damage
was significantly (P < 0.05) higher, by a factor of
approximately 2.4, among the stations used in the
Pine Siskins study than they were at the other 23
stations. Possibly, canopy cover reduction by insects
helped cause the declines.
Conclusion
Healthy productive populations of 13 species
of management concern depend upon differing
landscape-scale factors. Some species, like
Hammond’s Flycatchers, depend upon the presence
of contiguous coniferous forest with varying
degrees of canopy cover. Other species, such as
“Western” Flycatchers and Winter Wrens, depend
upon sensitive forested riparian habitats. At higher
elevations moist forest-meadow complexes are
critical to species like MacGillivray’s Warblers,
and Lincoln’s and Song sparrows. Also, at higher
elevations, forests affected by defoliating insects
and beetles appear to benefit Chestnut-backed
Chickadees, Song Sparrows, and Dark-eyed Juncos
reproductive success.
At higher elevations, a coarse-grained, habitat
heterogeneity of forest, successional-shrubland, and
grassland-meadow occurs naturally. This provides
quality breeding habitat for several species including
Chipping Sparrows and Pine Siskins. Habitat
edges in these and other managed landscapes are
ecologically important components in the population
dynamics of several species. More importantly,
specific pairs of habitats that make an edge may
be a preferred habitat component. For example,
Warbling Vireo reproductive success responded
positively to forest-successional and forest-grassland
edges. Other species, including Swainson’s
Thrushes and Chestnut-backed Chickadees,
responded negatively to forest-grassland edge.
In this study, long-term demographic monitoring
and species-landscape modeling have revealed
important ecological relationships for demographics
among 13 species of conservation concern. We can
use these models to predict the effects of proposed
forest management on populations of multiple
breeding species, thereby providing useful decision-making
tools. Furthermore, it is possible to spatially
extend these models to map potential habitat for
a particular species and forest type throughout an
entire forest.
As monitoring continues on the newly established
(and/or managed) stations through future breeding
seasons, we will begin to compare observed numbers
with predictions of my models and be able to
validate this approach. A similar study, based on
data collected from a network of stations located
on Department of Defense lands in the eastern and
south-central U. S., is used to predict the effects
of management (Nott and Michel 2005). Recently,
decision support tools were provided online for
both the Department of Defense network (Nott and
Chambers 2008) and this study (Nott and Kaschube
2007).
Finally, there are factors affecting the productivity
and survival of forest birds that have nothing to
do with management actions, especially shifting
climates and regional variation in weather patterns,
effects that are being detected globally (Root et
al. 2003). The data used in this study were also
used to reveal that climate and weather are strong
influences upon avian population dynamics in the
Pacific Northwest (Nott et al. 2002) and may mask
the effects of habitat management on avifauna.
To remove the bias of climate and/or weather,
it is important to quantify such relationships,
especially in the light of global warming. In some
regions it is increasingly valuable to quantify the
variable patterns of precipit