Sites : An Analytical Toolbox for Ecoregional Conservation Planning


Download toolbox package
View slideshow of example
Link to TNC Homepage
Link to UCSB Biogeography Lab
UCSB Biogeography Lab
Principal Investigator:
Dr. Frank Davis

Dr. Sandy Andelman
Dr. David Stoms

  TNC's Need for Ecoregional Planning Tools
Applications of Sites  revised October 11, 2001
Related Online Resources  revised October 11, 2001


The Nature Conservancy (TNC) recently initiated a new program of ecoregion-based conservation. The goal of this effort is to assemble regional portfolios of conservation lands that collectively represent viable examples of all native species and plant communities. This program aims to increase the efficiency and effectiveness of TNC's conservation activities through a systematic approach to regional portfolio assembly coupled to local, on-the-ground implementation. 

TNC's Need for Ecoregional Planning Tools

The initial steps in ecoregional conservation planning involve identifying an explicit conservation goal, selecting a set of target elements (i.e., species and communities), identifying levels of representation for each target, and then identifying a potential portfolio of sites. In many ecoregions where TNC works, there are hundreds of target elements to consider, as well as a variable number of potential conservation sites, each of which may differ in quality or suitability. To prioritize among potential conservation sites as the portfolio is assembled, often a hierarchical set of decision rules is applied. For each ecoregion where TNC works, the general process of portfolio assembly wll be similar, but the specific combination of goals, targets, sites, and decision rules likely will be unique. 

Given the large number of species, communities and potential sites in many ecoregions, without the aid of computer automation, it is virtually impossible for conservation planners to evaluate all possible alternative portfolios that potentially could meet a stated conservation goal. Computer automation also is needed because the portfolio assembly process is iterative, and once an initial portfolio is identified, there is an ongoing need to revise that portfolio. For example, the potential portfolio might change as a result of revisions to TNC's conservation goal for that ecoregion, or as a result of new data (e.g., for sites or elements not previously surveyed; or when the status of an element or site changes because of ownership changes). Without the aid of computer-based decision rules, it is difficult to develop an iterative portfolio assembly process that is repeatable, and the process of reviewing alternative site combinations can be labor intensive as well as unnecessarily slow.  In the last decade, a variety of computer-based algorithms and analytical approaches for reserve siting has been developed, however, at the time this project was initiated, these remained primarily research tools. Based on discussions with various TNC staff involved with ecoregional planning efforts, in their current form, none of the existing approaches was well-suited to TNC's planning needs. From TNC's perspective, the major limitations of current tools fell into three general categories: those that require high end computing power, specialized software, and/or a high level of GIS technical expertise (e.g., BMAS, Church et. al. 1996); those with insufficient documentation and/or inadequate testing of computer code; or those with overly simplistic decision rules (e.g., "greedy" or rarity-based heuristics). 


The following research and development effort for TNC was conducted: 
  1. Develop Sites 1.0, with an easy-to-use graphical interface in ArcView to select and display alternative portfolio solutions.
  2. Adapt the SPEXAN software for selecting reserve sites (written by Ian Ball and Hugh Possingham of the University of Adelaide in Australia) to accommodate TNC's portfolio design methodology.  The version used in Sites, called the Site Selection Module, like SPEXAN, explicitly incorporates spatial design criteria into the site selection process.
  3. Test and apply the planning approach in a structured decision process involving local TNC staff to develop hypothetical conservation portfolios in two different ecoregions (Idaho Batholith and the Northern Sierra Nevada). 
  4. Produce a web-based tutorial, including a worked example using TNC data and ecoregions; and conduct a training workshop for TNC staff interested in learning and applying the regional conservation planning tools to other TNC regions.

Applications of Sites 1.0

The following planning studies that have used Sites 1.0 have been presented at conferences or published in the peer-reviewed literature (as of October 11, 2001):
  1. Prioritizing Sites for Endangered Species Conservation along the Santa Clara River (large pdf file)
  2. TNC Conservation portfolio for the Middle Rocky Mountains - Blue Mountains Ecoregion (abstract of poster presented by TNC planning team at the 2000 Society for Conservation Biology conference)
  3. Conducting large-scale conservation evaluation and conservation area selection using a knowledge-based system (paper presented at GIS/EM4 in 2000 by Patrick Bourgeron of University of Colorado)
  4. Prioritizing Conservation Efforts in the Sierra Nevada Ecoregion.  Paper presented by Bo Wilmer at the 2001 ESRI User Conference. 
  5. Ecoregions and species assemblages: testing indicators for identifying areas of conservation concern.  Paper presented by Josh Lawler at the 16th Annual Symposium of the U.S. Chapter of International Association of Landscape Ecology, Tempe, Arizona, April 2001. 
  6. Leslie, Heather, Mary Ruckelshaus, Ian R. Ball, Sandy Andelman, and Hugh P. Possingham. Using siting algorithms in the design of marine reserve networks. In press. Ecological Applications
  7. Beck, M. W. and M. Odaya.  2001.  Ecoregional planning in marine environments: Identifying priority sites for conservation in the northern Gulf of Mexico.  Aquatic Conservation 11: 235-242.  

Related Online Resources

These web sites provide further background on the simulated annealing approach to reserve selection (as of October 11, 2001):
  1. MARXAN--the successor of SPEXAN--home page with descriptions, history, and more publications.  Hugh Possingham. 
  2. Mathematical methods for identifying representative reserve networks (pdf file of a book chapter by Hugh Possingham et al. 2000 that describes simulated annealing process and spatial design features used in Sites).