Oregon Breeding Bird Atlas
Breeding bird atlases have been published in at least 23 states, 5 counties, and 6 provinces in North America. In all instances, a basic feature has been the reporting and compilation of bird observations using a grid consisting of standard-sized geometric units. Grid units are used rather than county, watershed, ecoregion, town, or land management unit boundaries because (1) the equal sizes of grid units make statewide patterns of species occurrence easier to interpret and allow fairer comparisons of species diversity among units, (2) grid unit boundaries are permanent and not subject to reinterpretations or redefinition (as sometimes occurs with ecoregion and political boundaries), and (3) grid units can easily be aggregated or divided into larger or smaller equal-sized units.
For the Oregon Breeding Bird Atlas Project, two types of grids were used. One type of grid blankets the state with a series of contiguous hexagon-shaped units. Each hexagon covers an area of 634.5 km2 (245 mi2, or 156,000 acres), has six sides each with a length of 15.8 km (9.8 mi), and measures 27.4 km (17.0 mi) perpendicularly from side to side. Where some of these hexagons overlapped partly into adjoining states, only the Oregon portion was surveyed. Geographic coordinates of the corners of all these hexagons are provided in a table in the Hexagons section as well as in a data file.
The total number of hexagons that include any part of Oregon is 438. Of these, 337 are entirely within Oregon’s land area whereas 101 include parts of adjoining states or offshore areas. Of the 101 incomplete hexagons, we excluded 6 that were only barely in Oregon. Thus, the number of hexagons surveyed was 432. Only the part of a hexagon actually located in Oregon was surveyed.
The hexagon grid and its orientation as used in this project is part of a global grid described by White et al. (1992). The Oregon part of the grid is contiguous to the hexagon grid currently used for species mapping in 14 states. At the project outset we chose to use hexagons of this size as one of our two types of grid because:
1. When stretched across a spherical surface (the Earth) hexagons have less distortion than squares or rectangles, so are virtually the same size regardless of latitudinal position (White et al. 1992);
2. The number of hexagons that cover Oregon contiguously (432) was felt by the project steering committee to be a number that could be surveyed rather completely by the anticipated pool of volunteers over a 5 year period. We felt strongly that a comprehensive survey (rather than a partial survey focused on noncontiguous sample units chosen randomly or systematically) was needed. Had the project used smaller units to divide the state into more than a few hundred contiguous units, it might not have achieved acceptable coverage of all units during 5 years;
3. The relatively large area of the hexagons (435 km2) makes it easy to find access opportunities to each hexagon, i.e., constraints from lack of roads and private property restrictions are less likely to be limiting;
4. This hexagon grid was being used as a framework for organizing data on species distribution by the Oregon Natural Heritage Program (ONHP) and some other agencies. This increases the utility of our data and facilitates comparison with other data sets.
The second type of grid was comprised of squares distributed rather evenly throughout the state. Each square was 25 km2 (9.7 mi2, or 6178 acres) with dimensions 5 km x 5 km (3.1 x 3.1 mi). This grid did not cover the state contiguously. Rather, one square was located in each hexagon, but some of the hexagons that included only a small part of Oregon did not contain squares. Thus, the total number of squares was 414, and each comprised about 4% of its hexagon. Geographic coordinates of the corners of all these squares are provided in a table in the Hexagons section.
In general, squares were located in the southeastern "corner" of each hexagon, with their southeastern corner being at the same location as the southeastern point of the hexagon. However, in 98 instances (24 % of the 414 squares) we shifted the square’s location before beginning the atlasing because otherwise no part of the square could be covered, due to private land restrictions and lack of public roads. In general, the square was shifted to a location nearest its intended position but including at least 1 mile of public road or other access.
The decision to use squares as the second of our two types of grid was made because:
1. The smaller area of the squares (25 km2) allows for intensification of the search effort. At least in theory, a larger portion of each unit and its varied habitats can be surveyed;
2. Geographic locations of species can be represented more precisely;
3. Breeding bird atlas projects in other states (including the atlas in adjoining Washington) have commonly used a grid of squares with dimensions 5 km x 5 km (or 3 mi x 3 mi), or rectangles of about the same area. Using the same grid in Oregon facilitates comparison with these data.
The hexagons and squares collectively are termed "atlas units" (some other state atlas projects call these "blocks.") The efforts made to survey species in these units are termed "atlasing;" and the volunteers who participated are termed "atlasers."