Conservation Area Design
Conservation Biology & Genetics
Conservation GIS
Conservation Education
About Us
Links
Join Us
The
mission of
the Institute is to increase humankind's understanding, appreciation,
and protection of our natural environment; particularly wildlife
populations and wild landscapes. Our goal is to enable human
beings to live in harmony with other species. |
|
A Spatially Explicit Decision-Support Toolkit For Yellowstone-to-Yukon
Conservation
Y2Y has long
recognized the need to integrate the concerns and interests of
many diverse constituencies in the region if it is to fulfill
its vision of sustaining vital communities, both human and wild,
in Yellowstone to Yukon ecosystems over the long term. A key
goal is to assure that long-term community planning efforts and
land-use decisions are based on sound ecological principles.
We propose to develop an interactive Conservation Area Design
Decision Support System (CAD DSS) that will improve the ability
of decision-makers in the region to access and apply the scientific
information generated by Y2Y and related researchers. The CAD
DSS will also actively guide the consensus-building process by
providing the leaders of various interest groups in the region
with a common base of knowledge, methods, and tools by which
to examine the ecological consequences and tradeoffs of a broad
range of conservation area design options.
The term "Decision Support
System" (DSS) has been used by a wide range of interests
to cover an even wider range of methods and applications. Ideally
a decision support system is a systematic, logical and repeatable
technique that can be used to make choices, often in areas of
conflicting demands. Quite commonly decision support systems
make use of computers for data storage and data analysis. In
some cases, the term decision support software is utilized more
or less synonymously with decision support system.
In the arena of Conservation Area Design (CAD), a DSS should
use the best available science concerning species, habitats and
landscapes to determine an "optimal" configuration
of core protected areas and connecting protected habitat. The
optimization goal should be the maintenance of viable units of
targeted biological elements of the system (the 'viable units'
question). That is, not only must the biodiversity be initially
captured by the CAD, it must also be sustained. Time frames of
100 to 500 years are commonly utilized.
A CAD DSS requires accurate spatially
explicit data on animal population demographics and distribution;
habitat extent, condition, and trend, and current land use and
human activities. A defensible DSS requires more than just data
inputs. It must be based on the best current knowledge in the
basic science areas that involve population dynamics, population
genetics, and so forth. In addition, a CAD DSS must explicitly
recognize conflicting demands and it must quantify such constraints
in a common currency. The design of a CAD DSS must begin from
the question to be answered. If not so structured, such systems
often degenerate into meaningless but colorful exercises in GIS
technology.
An overlooked aspect of a DSS
is education. A DSS cannot be some mysterious black box that
reduces the participants of the associated planning group into
confused submission. The workings of the DSS cannot be entirely
the province of some potentially biased guru who alone knows
how it works. A DSS must be open and flexible. All the participants
should ultimately come to understand how it works.
The SITES program is a CAD DSS
that some professional groups are using. SITES uses data about
the presence/absence of species in landscape polygons to determine
"optimal' configurations of habitat patches. It has two
goals. It tries to find the minimal number of polygon patches
that contains n instances of each species. It also employs
a heuristic for the clustering of patches based upon maximizing
the shared between-patch edges. SITES does not incorporate any
ecological or dynamical information about the species it is trying
to protect. It is not based on any biological science principals.
There is no utilization of information on numbers of animals,
their spatial population dynamics (source or sink), their genetics,
or their immigration/emigration to or from any of the patches.
SITES would be quite capable of designing a system doomed to
rapid decay and extinction.
SITES and other CAD efforts are
sometimes connected to a population viability analysis (PVA)
model, such as VORTEX or PATCH. When used in a limited sense
(number of model inputs limited to a small number of species),
with expert knowledge of patch dynamics, this combination can
usefully improve decision-making.
Our CAD DSS approach is based
on an external solution of the viable units question; whether
these units be carnivores, aquatic or avian. PVA models might
work with some species or elements, but simpler approaches might
be utilized for biodiversity layers with less information. These
efforts will be, and must be, coordinated through species experts.
They must also be defensible as the "best available science".
The output from each of the groups of viable units will be a
"choice group" layer that can be represented on the
landscape. For efficient downstream computation, these layers
will have to be represented as gridded units, not arbitrary polygons.
The actual DSS will manipulate
these units so as to assemble an "optimal" CAD that
simultaneously maximizes all layers for species viability, subject
to costs related to land use and land acquistion.
It is our goal to begin the development
of such a tool and to apply it to some local area of concern
to Yellowstone to Yukon scientists. Coordination, communication
and computation are indispensable to effective conservation area
design. CAD design cannot be some distant academic exercise presented
at the last minute as a 'fait accompli' to a surprised public.
There must be coordination between multiple levels of government
and between various public interest groups. All forms of clear
communication need to be utilized in this endeavor. Communication
will not be simple because competent and efficient biological
conservation must meet sets of criteria defined in different
ways for different biota; and these need to be overlayed on specific
landscape. The very difficult, and as yet unsolved, problem of
accomplishing this is the focus of this project. The DSS science
will involve the creation of a decision support tool which can
effectively integrate science with spatially-explicit landscapes.
We propose to perform a pilot demonstration for a particular
planning region (yet to be determined) within the Yellowstone
to Yukon planning efforts. We recognize the following stages
in our effort: with each, we associate a deliverable product.
Stage One: to be completed by
1 August 2002. A rapid but critical status review of current
Y2Y science, knowledge, GIS layers, and so forth. It is necessary
to know what information we have and where we are lacking. No
conservation planning effort ever has complete or even adequate
information. However, conservation planning cannot afford to
wait for perfect knowledge. The only requirement is that the
best available science be utilized. Mike Gilpin will probably
have to visit Canmore to access some of the data and documents
and will confer with GIS analysts involved in Y2Y. A review document
will be compiled.
Stage Two: to be completed by
1 October 2002. A scoping document is required for the initial
feasibility study. Some circumscribed planning arena needs to
be identified and the general strategy for our approach needs
to be written and communicated to as many of the stakeholders
as possible. There should also be some public presentation of
this approach after completion of the document. At a minimum,
we would present this at the annual CERI meeting at the B-Bar
Ranch in November.
Stage Three: to be completed
by 1 March 2003. This stage will involve the core computer software
for the DSS. There are two somewhat independent research packages.
The first involves the creation of the choice group layers. Such
an effort involves participation of scientists expert in the
taxon to be represented. It is not reasonable to expect that
all layers could be put together in so limited a time and with
such limited funds. One layer should suffice for the demonstration
effort. At this point, we feel that the aquatics layer could
be so produced. This would principally involve Chris Frissel.
Secondly, independent of the creation of the choice group layers,
a feasibility demonstration for the linking of the GIS layers
and Java-based optimization is required. Hopefully, this can
be demonstrated at the annual Y2Y science meeting, where interested
parties will be able to criticize the approach and add their
own insights. The deliverables are thus GIS layers representing
the best available science on status and distribution of the
chosen taxon and other model inputs, and a beta version of the
DSS model itself which will be run and peer-reviewed.
Stage Four, the Final Stage:
to be completed by 1 May 2003. We want to bring everything together
at a one-day workshop where we apply our approach to some design
arena (as determined in Stage Two). Participants in the process
would form an ad hoc planning team and would be prepared ahead
of time, perhaps via the Internet. At the meeting, the DSS tool
would be presented and made available for inspection on computer
workstations. In the afternoon session, the actual decision-making
would be carried out by the ad hoc planning team. We would observe
this process and would analyze the experience in a final report
to the Y2Y team.
On the technical
side, three domains of expertise will be required: biological,
GIS, and reserve design. Y2Y has the working groups to carry
out the research needed on the biological side. However, these
groups will need guidance regarding the character of the biological
information required to support a CAD DSS. Michael Gilpin, the
principal investigator on this proposal, has the conservation
background and the computer modeling skills to carry out the
reserve design. The GIS modeler must sit in the middle of this
process, which involves the integration of the biological data
layers with the requirements of reserve design. Thus, the GIS
position requires someone who understands the entire process.
The core of
the CAD DSS design and development would occur in Bozeman, Montana,
and would be carried out under the auspices of Craighead Environmental
Research Institute, which would provide, among other things,
computers and GIS software. We anticipate that the GIS technician
would be employed by CERI. The second interlocking requirement
of CAD DSS development, as mentioned above, involves the human
context. It is vitally important that biologists, conservation
managers, government planners, and economic representatives in
the Y2Y region - that is, all those who will ultimately take
part in the reserve design process -- be educated about this
tool and have the opportunity for input during its development.
Their involvement must begin at the earliest point in the development
process. This would require, at a minimum, several multi-day
workshops.
In summary, the process of developing and beta-testing the DSS
will take one year. The pilot application will be targeted at
some ecologically important region of Y2Y for which adequate
data is available. Possible pilot areas include the Gravelley-Centennial
corridor between the Greater Yellowstone and Salmon-Selway Ecosystems,
the Big Belt corridor between the GYE and Northern Continental
Divide Ecosystem, the Crowsnest corridor between the Northern
Continental Divide and the Canadian Southern Rockies, and other
regional areas of connectivity with a complex matrix of land
use and ownership. Stage two will identify the area for which
the best data is currently available. The results of the pilot
study will provide a solution to guide conservation efforts to
ensure the maintenance of biodiversity: through the protection
of native forests and other habitat in a network of core areas
and connecting movement corridors sufficient to maintain long-term
wildlife population viability for the target area. The DSS tool,
once developed, will be applicable to any region of conservation
concern throughout the world. |
Craighead
Environmental Research Institute