Link to Data Download page

Northwest Gap Analysis Project

Northwest GAP home

NW GAP Status

Land cover Mapping Status

A completed draft land cover dataset is available by clicking the above link.

Stewardship Mapping Status

A draft map of Northwest Stewardship will be available Winter 2008.

If you have data that you believe would be important to include in these data, or have any questions, please contact:

Lisa Audin
Stewardship Coordinator
530 S. Asbury Street, #1
Moscow, ID 83843
laudin@uidaho.edu
(208-885-3901).

Species Modeling Status

This part of the NW project started in September 2005. Currently, we have gathered all the species occurrence records totaling over 2 million from all 5 Northwest Natural Heritage Program. The records are being filtered for duplicate records and region-wide data are also being incorporated.

Each of the 5 Natural Heritage Programs is assisting us with collecting species occurrence data, providing biological expertise, and building review teams within their respective states. However, The Wyoming Natural Diversity Database at the University of Wyoming in Laramie, WY, is coordinating the species modeling efforts.

The five Natural Heritage Programs are:

Our intent with this approach was to divvy up the modeling work such that primary experts and data holders on some species are responsible for those species (e.g., Natural Heritage biologists model species they know and track). And divvy up point data compilation such that Natural Heritage biologists aggregate data for species they would model in their states (i.e., they already know the sources and have the contacts if not the data; other species will be modeled centrally ( Laramie) and museum sources of data will be compiled centrally and distributed).

The Natural Heritage Programs have existing occurrence data, expertise, and infrastructure that cannot be replicated. Each program acts as a central clearinghouse for occurrence data. Furthermore, each program only has occurrence data for one state, which eliminates duplication of records and makes data compilation easier. All programs use the same software, standards, and methodologies including a common database. This ensures individual records match in type and format. The database is continually reviewed, quality checked, and updated which minimizes duplicate records and keeps the database available for re-analysis at any time.

We intend to map the range (total areal extent occupied by a species), distribution (spatial arrangement of environments suitable for occupation by a species, and habitat (environments with the combination of resources and conditions that promote occupancy, survival, and reproduction by a species) of each species. We are attempting to make regional maps more useful to local land managers by taking this approach.

We are taking 2 modeling approaches for the Northwest Gap Project. First a deductive modeling approach, which was the standard approach used by state-based Gap projects and the Southwest Gap Analysis Project. And second an inductive modeling approach, which is being used in both Southeast and Northwest Gap Analysis projects.

A deductive modeling approach s ynthesizes information from experts and literature reviews regarding habitat associations. Then uses land cover data to predict a species’ distribution based on habitat associations. The species’ distribution is then refined using species occurrence records. This modeling approach works well for species with a bundant information regarding habitat associations and limited occurrence records. It t ends to over predict when h abitat associations and land cover types are too general and when species occur in habitat difficult to define using satellite imagery (e.g., r iparian habitat).

An inductive modeling approach u ses empirical observations to derive objective conclusions. A predicted species distribution is based on environmental parameters (e.g., elevation, climate gradients) at known points of occurrence. This modeling approach works well when p resence and absence data are available and there are numerous occurrence records for a species with good spatial distribution. It tends to under predict i f only l imited or unevenly distributed occurrence data are available and if false negatives exist in data.

Please check back for future updates.

2007 progress report.