The species distribution maps in the Wildlife Viewer represent the third generation of maps developed by the Oregon Natural Heritage Information Center (ORNHIC). The maps were generated using several different data layers, and were built on the preceding mapping efforts. 

The initial effort was based on the creation of a database which attributed all wildlife species to a grid of 440 hexagons covering Oregon. The hexagons were originally developed for the Environmental Monitoring and Assessment Program (EMAP) of the US EPA. The dataset attributed the confidence we had that a species was breeding within a hexagon, based on data at the Oregon Department of Fish and Wildlife (ODFW) and the ORNHIC. Three levels of confidence were attributed:

To create the initial distribution maps, only hexagons that were attributed as confident or probable (C or P) were used. The funding to develop the hexagon database came from the Biodiversity Research Consortium and EPA, which also included resources to have all the hexagon distribution maps peer reviewed by taxonomic experts. The database was developed cooperatively with ODFW and NatureServe (at that time the National Heritage Program of The Nature Conservancy). In 1997, the updated hexagon distribution maps were intersected with appropriate habitat from on the 1992 Oregon Gap Analysis Map developed by Kagan and Caicco to create maps shown in the first edition of the Atlas of Oregon Wildlife. A detailed description of the methods used to create the first maps is found in the Introduction of the first version of the atlas.

The maps were updated using a 1999 vegetation map for the 2001 second edition of the atlas, using largely the same methods. In 2002, the hexagon dataset was updated with information from the Oregon Breeding Bird Atlas, which used these same EMAP hexagons for their long-term study of breeding bird distributions.

To update and improve the species distributions maps to be used to develop the Conservation Strategy, ORNHIC made two major changes to the methodology. First, we overlaid the modeled species distribution maps and the hexagon distribution information with 6th field HUCs to create a database showing the predicted distribution confidence of each species in each of Oregon’s 3,122 watersheds (Figure 1). This resulted in distribution maps with more ‘natural’ looking boundaries (watershed boundaries as opposed to hexagon boundaries). Also, these are much smaller areas, so can be used to better confine a species distribution. For each watershed, species were assigned a value (primarily from the hexagon data set), of “confident, probable or possible” using the same criteria as for the hexagons. However, these updated and finer resolution maps have not been peer reviewed. We hope the Wildlife Explorer will provide an opportunity to improve the accuracy of this database, by providing local experts the ability to send us updates and comments on species distributions.

 image citation

Figure 1. EMAP hexagons and Oregon watersheds (6th field hydrologic unit code or huc)

The second improvement was the way potentially occupied habitat was mapped. Unlike the initial atlas maps, which showed the species only as present or absent, new maps were created using a habitat suitability index (HSI) to show the relative quality of habitat. This was seen as important because the initial maps did not differentiate between poor quality habitat and very good quality habitat – which can make a big difference in how well a species can survive.

A revised vegetation map was put together with 30 meter pixel resolution and using the NatureServe Ecological System Classification. The vegetation map was crosswalked to a wildlife habitats map with 60 habitat types. The major difference between the initial 31 habitat types used in the first versions of the atlas, and the 60 used in the 2005 update was the inclusion of information on forest size classes. The majority of forest habitats were split into small, medium, and large forest stands, which tripled the number of forest types in the maps. To create the species distribution maps, we intersected the habitat map with the watershed-based distribution map. The WHR was used identify those habitats where the species could be expected to be found, and the watershed occurrence limits the predicted distribution to only the regions where species have a confirmed or probable occurrence.

In addition, the Wildlife Habitats Relationships matrices (WHR) from the first maps were converted to a Habitat Suitability Index, which was created for each ecoregion separately. Originally we used a HSI developed by Adamus et al. (2000) which assigned habitats to 11 levels of suitability as follows: Each habitat class was rated on a scale of 0 (not used) to 10 (most preferred or used) for each of the species that currently breed or historically occurred in the ecoregion. The suitability scores for a species reflected rankings of the habitat classes rather than intensity of use or preference. For example, a score of 8 should not necessarily be interpreted to mean that the species uses a particular habitat class twice as much as a habitat class scored 4 for the species.  In general, scores greater than 5 described habitats that, if sufficiently extensive and of a suitable spatial pattern and location, would be expected to serve as “sources” of potential breeders. Scores less than 5 described habitats expected to generally be population “sinks” (although they sometimes may comprise necessary dispersal corridors).

In most cases, we did not have adequate information to make such fine distinctions between habitat suitability. Also, multiple people were assigning HSI codes in different ecoregions, so we felt a more limited, structured system would work better. Therefore, we distilled the HSI categories to 4 as follows:

No habitat =  HSI code 0 = Seldom or never used habitat
Poor habitat = HSI codes 1 = poor or infrequently used habitat
Fair habitat = HSI code 2 = fair or mediocre habitat
Good habitat = HSI codes 3 = good or excellent quality habitat

Because the HSI codes were assigned by ecoregion, often by different biologists, there are often stark differences between ecoregions. This reflects differing opinions on what is, say, good versus poor habitat. These HSI codes have not been peer reviewed. Rather, we view this data set as a work in progress and hope that posting these maps will allow us to receive feedback from local experts who can comment on the suitability of different habitat types in their area. Ultimately we hope to improve these maps to closely reflect changing current distributions and habitat suitability for terrestrial vertebrates in Oregon.

Literature Cited:

Adamus, P.R., J.P. Baker, D. White, M. Santelmann, and P. Haggerty.  2000. Terrestrial Vertebrate Species of the Willamette River Basin:  Species-Habitat Relationships Matrix. Internal Report.  U.S. Environmental Protection Agency, Corvallis, OR.

Csuti, B., T. A. O’Neil, M. M. Shaughnessy, E. P. Gaines, and J. C. Hak. 2001. Atlas of Oregon Wildlife, second edition. Oregon State University Press, Corvallis, OR. 520 pp.

Csuti, B., A. J. Kimmerling, T. A. O’Neil, M. M. Shaughnessy, E. P. Gaines, and M. P. Huso. 1997. Atlas of Oregon Wildlife. Oregon State University Press, Corvallis, OR. 492 pp.

Kagan, J.S., J.C. Hak, Blair Csuti, C.W. Kiilsgaard, and E.P. Gaines. 1999. Oregon Gap Analysis Project Final Report: A geographic approach to planning for biodiversity. Oregon Natural Heritage Program, Portland, OR. 72 pp. + appendices.

Authored by: Jimmy Kagan, INR Information Program Manager and Eleanor Gains, INR Zoology Projects Manager (2008)