Sampling designs for bird population monitoring and habitat surveys are being increasingly implemented at landscape and regional levels. Some of the statistical methods used to sample natural resources over large areas involve subdividing the sampling area into smaller equal area units. This creates a mesh or grid of cells that are used to conduct systematic, random or stratified sampling in each cell of the grid. For example, grids are often generated prior to establishing Generalized Random Tessellation Stratified (GRTS) spatially-balanced survey designs. These kinds of sampling schemes are becoming increasingly popular within the U.S. Department of Agriculture’s Forest Service, the National Park Service, the Environmental Protection Agency, and more recently the U.S. Fish and Wildlife Service’s Landscape Conservation Cooperatives.
In bird monitoring, grids not only provide a means to identify points, but most importantly to identify areas where sampling should, has, or has not occurred, and to describe results of monitoring projects in a spatially comparable way. Grids may be generated with cells of different shapes (e.g. squares, rectangles, hexagons, triangles, etc.), and as hierarchical systems that allow aggregation of cells at different scales. The cells (units of subdivision of a grid) can be named and described in terms of size, relative location within the grid, and other attributes (e.g. field data) occurring at the same location. This allows users to stratify, summarize and query grid cells for many purposes.
Although grid sampling could facilitate generation of natural resource samples across scales, grids for bird monitoring are often generated on a project-by-project basis using an arbitrary reference point and cell naming system. The result is that most sampling grids used in bird population and habitat studies are independent from one another. This lack of coordination has resulted in repeated expenditures of time and resources to generate grids and to incorporate cell attributes into those grids. Furthermore, cells of independent grids often do not spatially overlap or nest within each other, and hence hinder development of location-based services for these data.
Examples of use
Rocky Mountain Bird Observatory is using the U.S. National Grid for developing sampling designs across the Intermountain West.