Partners in Flight's Landbird Monitoring Strategy:
National Monitoring Working Group, October 1998
Abstract
PIF's National Monitoring Working Group has been working for
two years to develop a Landbird Monitoring Strategy. Several
national and regional meetings have been held and more than 50
suggestions have been received describing ways to improve
landbird monitoring. These suggestions have been studied and
consolided into three major recommendations: (1) A comprehensive
landbird monitoring program, involving all major groups that
produce or use information on the health of landbird populations,
should be created; (2) Quantitative goals for landbird monitoring,
endorsed by the scientific community and by organizations involved
in managing landbirds, should be developed, and (3) The comprehensive
monitoring program should include short-term projects, addressed
to specific management issues, as well as the activities needed to
meet the long-term goals. PIF's Joint Steering Committee is appointing
a multi-organizational working group to study the recommendations in
more detail, revise them as appropriate, and develop a "framework"
for an integrated monitoring program. Initial suggestions for defining
quantitative goals for long-term monitoring are contained in this
draft of the Strategy but will need much discussion. The Western
Working Group of PIF is taking the lead at present on developing
ways to combine long-term monitoring and short-term, targetted
investigations. Groups concerned with other avian taxa are encouraged
to consider joining with PIF, perhaps under the auspices of an
"umbrella" group, to develop a comprehensive avian monitoring program.
Introduction
The conservation community in the United States, working through the
grassroots organization, Partners in Flight (PIF), has made an unprecedented
commitment to the conservation of nongame birds. A network of biologists
and managers, including some of the most respected names in conservation,
and extending from bird watchers and field biologists to the highest
administrative levels of agencies and non-government organizations,
has begun work on Bird Conservation Plans for nongame birds throughout
the United States. The core group in this effort consists of four regional
coordinators, and a national level supervisor, who organize the efforts of
several dozen biologists involved in preparing the plans. Several national
level committees of PIF oversee the efforts of this group and are making plans,
along with numerous allies in conservation groups, to implement
the plans once they are completed in early to mid 1999.
The general goal of the Bird Conservation Plans is to reverse,
minimize, or avoid undesirable trends in avian populations. The
Plans are expected to be the primary basis for nongame bird
conservation efforts during the coming decades and it is widely
hoped that they will ultimately lead to efforts on the scale of
the North American waterfowl Management Plan.
As the Bird Conservation Plans are finished and implemented,
it will become essential to have avian monitoring programs to reveal
how effective the programs are at local, regional, and national levels.
Several avian monitoring programs have been in existence for decades,
and many more have been developed during the past few decades, but PIF
has been concerned that no comprehensive program exists at present that
could evaluate effectiveness of its Plans in restoring and maintaining
bird populations. Accordingly, the PIF National Monitoring Working
Group was asked, in March 1997, to develop a Landbird Monitoring
Strategy, building on existing programs, to help evaluate effects
of the Bird Conservation Plans. This report is a first draft of the Strategy.
Methods
An initial outline containing major components of the Strategy was
produced in March, 1997 when the project was first proposed. A final
version was distributed on May 20, 1997. During the fall of 1997 the
outline was expanded and a substantial effort was made to develop lists
of species that need to be monitored and criteria for deciding whether
coverage of them was adequate. This material was reviewed at another
meeting of the Monitoring Working Group in Sacramento in October 22-23, 1997.
Those in attendance suggested revisions and endorsed the general
approach being developed.
During November 1997 - January 1998 numerous people were recruited to
work on different aspects of the Strategy (Table 1). A Committee of
scientists was appointed to help develop and review major sections of
the Strategy. Liaisons were also appointed with other groups interested
in avian monitoring and with the major federal agencies that ultimately
will help implement these ideas.
Table 1. Organization for producing the PIF Landbird Monitoring Strategy
- Review of suggestions for improving landbird monitoring
- Brad Andres, FWS, Alaska
- Ed Arnett, Weyerhauser, Washington
- Greg Butcher, American Birding Association, Colorado
- Mike Carter, Colorado Bird Observatory, Colorado
- Geoff Geupel, Pt. Reyes Bird Observatory, California
- Fran James, Florida State University, Florida
- Doug Johnson, USGS, North Dakota
- Peter Kareiva, University of Washington, Washington
- Raymond O'Connor, University of Maine, Maine
- CJ Ralph, US Forest Service, California
- Ken Rosenberg, PIF, Cornell, New York
- Jerry Verner, US Forest Service, California
- Liaisons
- Canada: Connie Downes
- Mexico: Jorge Nocedal
- Raptors: Karl Bednarz, Keith Bildstein, Mark Fuller
- Seabirds: Dave Irons
- Shorebirds: Bob Gill, Brian Harrington, Gary Page (tentative)
- Ornithological Council: Mike Erwin, Mike Scott, Pat Heglund
- Ducks Unlimited: Keith McKnight
- US Environmental Protection Agency: David Davis
- US Bureau of Land Management: Chris Jauhola
- US Forest Service: Debbie Pressman
- US Geological Survey: Marshall Howe
- US National Park Service: Steve Fancy
- Implementation
- David Pashley, Gary Myers, Jonathan Bart
Numerous requests were circulated for suggestions about how avian
monitoring efforts can be improved. More than 50 suggestions were received
and posted for discussion on a home page. The suggestions were consolidated
and discussed at another meeting of the National Monitoring Working Group
in April 1998 at the Ornithological Societies meeting in Missouri. The
results were also presented at a special symposium on monitoring, organized
by the Ornithological Council and held at the Ornithological Societies
meeting, and to the PIF Joint Management and Steering Committees in September
1998 in Georgia. At the Joint Steering Committee meeting, representatives
of the major federal agencies endorsed the ideas presented in principle and
suggested that an interagency committee be appointed to review the findings
in more detail, revise them as necessary, and develop an Action Plan to
implement the revised Strategy. It was decided that initially, to keep the
task manageable, the Committee would consist of federal representatives but
that States and non-government organizations would be invited to participate
as soon as feasible.
This report briefly summarizes the major recommendations developed
during the past two years of work and suggests the major tasks for the
Committee that is being formed. It is expected that members of the
Committee will modify the recommendations of the Strategy. Thus, the
current document is best viewed as an early, rather than nearly complete,
draft of the forthcoming Landbird Monitoring Strategy.
Results
The suggested ways to improve landbird monitoring received
during this study are described below. No effort is made, in
this section, to present the ideas in the form of an action plan.
- Improve cooperation and coordination between monitoring programs
- Numerous programs monitor landbirds at present, and more
seem to appear every year, but there is little cooperation or
coordination between them. A central group is needed to decide
on the highest priority objectives, develop methods for data collection,
establish data centers to receive and disseminate information,
and help with analysis.
- Set quantitative goals for avian monitoring programs
- Agencies need help in deciding how much information,
and of what kind, they should collect about birds. The Landbird
Monitoring Strategy should provide state-of-the-art advice on this
topic and agencies should be encouraged to adopt the recommendations.
Doing so will help with coordination and protect the agencies from
critics as well as helping to insure that the most useful information
is collected and made available to decision makes.
- Develop targeted, hypothesis-driven programs
- Broad-based programs are poorly suited for evaluating effects
of management actions, such as effects of prescribed burning on nongame
birds. While such analyses are often viewed as "research" it should be
possible to use many of the same volunteers and infrastructure to address
these problems too. This, is already being done in some programs. Such
efforts should be expanded.
- Use monitoring programs to meet local and regional, as well as national, needs
- Numerous managers and biologists at local and regional levels
commented that participating in national programs does them little good.
They recognize the importance of meeting the national objectives but feel
that their objectives also deserve support.
- Use monitoring to promote adaptive management
- Monitoring programs are needed that will help managers
evaluate and modify their efforts to restore and protect nongame
birds. This is especially important because efforts of this sort
have been uncommon in the past, but this will change soon as the
PIF Bird Conservation Plans are completed and implemented.
- Develop programs for poorly covered species
- Many species are not well covered by existing programs. These
include whole taxa (e.g., most owls) and many species that have declined
seriously and are thus too rare to be well covered by the BBS. Giving more
attention to species of special concern is especially important.
- Design monitoring programs to reveal causes of trends
- Monitoring programs like the BBS and Christmas Counts reveal
trends but have often not been very successful in revealing causes of
declines (or other trends of concern). New programs are needed that
can be focussed on species and areas of special concern and that will
reveal causes of undesirable trends.
- Incorporate habitat analyses in monitoring programs
- Nearly everyone is agreed on the value of comprehensive
land cover (i.e., habitat) information for interpreting results
from avian monitoring programs, and numerous programs are underway
to provide such information. Contacts should be made with the major
efforts of this kind to insure that they produce information that
will be of greatest use to ornithologists.
- Place more emphasis on productivity and survival
- Much useful work has been done during the past decade by
MAPS, BBIRD, and other programs that study birth and survival rates.
These efforts should be expanded and used to identify high
quality (i.e., source) habitats, especially for species of special concern.
- Explore ways of using birds as indicators of environmental health
- Study of birds has revealed serious environmental problems,
especially with pesticides and habitat loss, on many occasions (e.g., DDT).
More attention should be given to developing programs that focus specifically
on this goal. Doing so might lead to substantially greater support for avian
monitoring programs as well as revealing and documenting environmental
problems that we all need to become aware of.
- Place more emphasis on design issues such as objectives, power, and bias
- Although careful design is undoubtedly much more common now than in
past years, much room still exists for improvement. In particular, bias may
be a significant component of accuracy in trend estimates and thus should be
explicitly considered in analyses.
- Obtain two or more independent estimators for trends in population size
- Despite everyone's best efforts, bias of uncertain magnitude will
probably continue to affect trend estimates. We should therefore try to
obtain at least two estimators, believed to have negligible bias, for trends
of primary interest. If they are consistent with each other, we would have
more confidence in the estimates than will ever be possible with
a single program.
- Develop regional and national data centers
- The Patuxent wildlife Research Center, and more recently the Laboratory
of Ornithology at Cornell, have provided a valuable service by organizing and
supporting bird monitoring activities. Their work should be supported and expanded
and other data centers should be established. One approach would be to develop
a series of regional centers each of which would take on certain national
level work and would also provide personalized assistance for biologists
in their region.
- Prepare "report cards" for species that synthesize results from all programs
- Results from different programs, including ones to be initiated under
the Landbird Monitoring Strategy, should be presented in a "report card" format
that reveals how populations are performing with respect to birth and survival
rates as well as abundance. Such a comprehensive analysis would provide a much
more complete and useful picture of the species' health than is
currently available.
- Obtain long-term support for monitoring programs
- Assurance of long-term funding is needed for those programs most
likely to provide long-term benefits. Insuring that short-term projects -
providing frequent benefits - are embedded within the long-term programs
would increase the likelihood of continued funding as would incorporating
responsibility for monitoring more fully into agency mandates.
Discussion
Most or all of the suggestions described in Results indicate the
paramount need for a comprehensive, integrated avian monitoring program.
The program should involve all major organizations that need or provide
avian monitoring information. Clear goals for the program must be
established, and the rationale for each goal should be presented.
In addition, strong support was expressed for including short- and
long-term projects in the integrated program. This approach is already
being taken in some programs, and promises to greatly increase the
utility of the overall effort to local and regional managers. The need
for local utility of monitoring efforts was by the far the most common
comment received during this study.
If an integrated, multi-organizational program is established,
with clear goals and the ability to carry out short-term as well
as long-term projects, then most of the other recommendations
described in Results should be feasible. For example, short-term
projects can be used to promote adaptive management, and sound
design will be easier to incorporate as will hypothesis-driven,
targetted programs. Readers may verify for themselves that the other
suggestions in Results can be accommodated if the major tasks above
are accomplished. The major recommendations of the PIF Landbird
Monitoring Strategy are thus:
- An integrated landbird monitoring program, involving all major groups
that produce or use information on the health of landbird
populations, should be created.
- Quantitative goals for landbird monitoring, endorsed by the scientific
community and by organizations involved in managing landbirds,
should be developed.
- The comprehensive monitoring program should include short-term
projects, addressed to specific management issues, as well
as long-term monitoring.
The current Strategy is restricted to landbirds because that
is PIF's area of concern. Ultimately, however, a general avian
monitoring strategy should be developed, and it is possible that
development of the landbird monitoring strategy would provide a
useful vehicle for developing a more general avian monitoring
strategy. This issue was discussed in the Joint Steering Committee
meeting in September 1998 and it was agreed that groups concerned
with other taxa should be informed of the PIF effort and invited
to join with PIF. If this occurs, then PIF will provide input on
landbirds and other groups will contribute expertise and recommendations
for their taxa. Such an effort might be carried out under the
auspices of a group with broader interests than PIF such as the
Ornithological Council, the Biological Resources Division of USGS,
or the Commision for Environmental Cooperation which is convening
a conference to begin work on a North American Bird Conservation
Initiative. Alternatively, such an effort might be undertaken by
a multi-organizational committee established expressly for the
purpose of formulating an avian monitoring program. Thus, at
present, the Strategy is being developed for landbirds under the
auspices of PIF's National Monitoring Working Group, but PIF would
be receptive to joining forces with other groups.
Progress towards implementing the major recommendations
that comprise the Strategy at present is discussed below.
- Develop an integrated landbird monitoring program which
involves all major groups that produce or use information on
the health of landbird populations.
This recommendation was reviewed by the PIF Joint Steering
Committee in its September 1998 meeting in Georgia. Those in
attendance expressed general approval for the recommendation
and decided that the best way to investigate it thoroughly was
to create a multi-organizational Committee to develop a "framework
for landbird monitoring". The Committee, which is currently being
formed, will make recommendations on the best way to develop the
proposed integrated landbird monitoring program.
Their specific tasks are as follows:
- Review the draft Landbird Monitoring Strategy and work with
the PIF National Monitoring Working Group to revise it as appropriate.
Seek endorsement by each participating agency of the Strategy as
representing a sound approach for landbird monitoring.
- Provide descriptions of each agency's current monitoring
programs and how well they meet the monitoring goals.
Describe the work needed to carry out the Landbird Monitoring Strategy.
- Reach consensus on the roles of each agency in carrying out the Strategy.
- Determine which tasks for each agency can be carried out with
existing resources and what new resources would be needed to fully
implement the Strategy.
- Transmit a description of the needed resources to the Implementation
Committee for the Landbird Monitoring Strategy.
- Develop an action plan for implementing those portions
of the Strategy that can be undertaken immediately.
Liaison with the States has been discussed with the PIF State
Directors Committee. It was decided that the PIF Regional Monitoring
Committees would each appoint a person to work with the National
Monitoring Working Group on development of the Strategy and that
these people would have responsibility for liaison with the States.
The National Monitoring Working Group will thus handle liaison with the States.
Non-government organizations active in landbird monitoring will
be reviewing the Landbird Monitoring Strategy and will provide their
input to it via the PIF National Landbird Monitoring Working Group.
- Develop quantitative goals for landbird monitoring and obtain
endorsement of them by the scientific community and by organizations
involved in managing landbirds.
A first effort to develop quantitative goals for landbird monitoring
is provided below. It will need review and revision by the scientific
community and the agencies but the following material provides one
possible approach.
The approach follows the widely accepted process of defining a
quantitative "criterion" for describing or measuring the quantity
of interest (in this case, accuracy of the trend estimate) and then
setting a "standard" that defines the target level of the criterion
to be achieved. One possibility for the criterion, which is consistent
with the recommendation that effects of bias be considered in accuracy
assessments, is a "bias corrected, 90% relative confidence interval."
This quantity is defined as the half-width of the 90% confidence interval,
increased if necessary to account for possible bias, divided by the point
estimate and expressed as a percent. For example, suppose the point estimate
was 100 and the usual 90% confidence interval was +3 (i.e., 97 to 103),
but bias of +1 was judged to be possible. The bias-corrected 90% confidence
interval would then be 96 to 104 and the half-width would be 4.
The relative confidence interval would be 4% (4/100). The interpretation
of this measure is that the true trend is probably (i.e., within 90% confidence)
within 4% of the estimated trend. The number of years on which the estimate
is based affects accuracy and thus must also be specified. Long-term trends
are difficult to detect without 10-15 years of data (refs. to be added).
We suggest 12 years as a reasonable standard. The criterion for describing
accuracy thus becomes the bias corrected, 90% confidence interval,
when based on 12 years of data.
One approach for setting standards is to identify the species
of interest and then specify the proportion of them that should
receive "adequate coverage" in the monitoring program. The PIF
prioritization scheme provides a list of species for both the United
States and for regions within it. The US list has approximately 600
species. Adequate coverage would be defined as achieving a specified
level of accuracy, as defined above.
Previous work has shown (refs. to be added) that apparently healthy
populations often exhibit 12-year rangewide trends of up to about 3%,
and regional (e.g., multi-state) trends of up to about 5%. This suggests
that +3% for rangewide trends and +5% for regional trends might be useful
threshholds. This criterion is similar in principle to quantities used in
the past such as having 90% probability of detecting a 50% decline over
a 25-year period (refs. to be added), and a crosswalk between the two
criteria could be developed. Preliminary analysis (Bart, unpub.) suggests
that meeting the 3% and 5% standards for 60% of the species of interest
is feasible and would be cost-effective in the sense that the efforts to
obtain this information would be more than re-paid in greater agency
efficiency. The goals for the monitoring program, under this approach,
would thus be:
- Achieve bias-corrected, 90% relative confidence intervals for annual,
proportional change in rangewide population size, during the most recent 12 years,
of no more than +3% for at least 60% of the species regularly
occurring in the United States.
- Achieve bias-corrected, 90% relative confidence intervals for annual,
proportional change in regional population size, during the most recent 12 years,
of no more than +5% for at least 60% of the species regularly
occurring in any region.
The suggestion above is just one of many that could be made.
A full discussion is needed of the various possibilities, along with
assessments of how difficult - and how beneficial - achieving them would be.
In addition, the suggestions above were developed by biologists in the
United States, thinking primarily about the United States. It is recognized
that the approach may not be appropriate in Mexico and Canada. Representatives
from these countries are encouraged to develop goals for their countries.
- Integrate long-term monitoring programs and short-term assessments
for species or areas of special concern.
PIF's Western Working Group is taking the lead at present on
developing programs that include both long and short-term projects.
A workshop was held in September, 1998, to discuss initiating a
"West Wide Inventory and Monitoring Project" (WWIMP). Speakers from
existing programs and agencies that use monitoring information concurred
both that the time seemed right for an integrated effort and that it
should include short-term as well as long-term projects. Long-term goals are
being developed and a call has been issued for suggested short-term projects
that the WWIMP could help organize. Several suggestions have been received.
The underlying theme seems to be that the WWIMP should help achieve consensus
on several generic issues (e.g., the utility of "reference sites" and how many
should be established) and that it should organize a 3-5 year effort to obtain
better information on the distribution and abundance of species of special
concern that are poorly known in the west at present. This effort will
include assessing effects of selected management actions (e.g., prescribed burning)
on abundance of these species and will probably include an attempt to develop one
or more habitat classification systems that can be applied westwide using GIS
methods. A meeting to initiate these activities has been tentatively
scheduled in Portland, Oregon during late January or early February, 1999.
This effort should provide valuable experience in how best to embed short-term
projects within long-term projects.
Conclusion
After two years of sometimes faltering effort, a coherent plan
is emerging for development of the PIF Landbird Monitoring Strategy.
The next steps are to obtain wide review of the suggested ways to
improve monitoring (see Results) and the three major recommendations
developed from them (see Discussion); to form the multi-organizational
working group, proposed by the PIF Steering Committee; to begin
discussion of the quantitative goals for monitoring proposed
in this document; and to experiment more widely with ways to
embed short-term, targetted projects within longer term monitoring
efforts. Groups concerned with other taxa are encouraged to consider
joining with PIF to develop a comprehensive avian monitoring program
that serves both landbirds and other taxa.
Research is being conducted by:
Maintaining Faunal Diversity in Forested Ecosystems of the Coastal and Intermountain West
(RWU-4251)
|
