Bird Atlas - Methods

Methods

In the oldest and most popular sampling approach, the region to be covered is gridded and volunteers are expected to visit representative locations within each grid cell and provide information that is collated. Often the method of collecting data, time and season in which to obtain the sample information are pre-decided as part of a protocol. In some cases birds that are found to be breeding are recorded, others may use point sampling or transects within the grid cells to obtain quantitative estimates of abundance. In some countries the grid cells follow the latitudes and longitudes - cell intervals of 1 degree, 30 and 15 minutes are often chosen for convenience. In higher latitudes where such an approaches leads to grid cells with large differences in area, sizes are more often fixed using grid distances of 1, 2, 5, 10 or even 50 km grid intervals. A disadvantage with grids is that boundaries rarely match those of habitats, making them unsuitable for ecological studies. Another problem is that the data collected in one project cannot be readily reused with new grid alignments that may be needed for instance when combining information with other projects. Repeat atlases made after a decade or two have helped in identifying long term range changes.

Another approach that does not need pre-defined grids makes use of the coordinates of individual points. Coordinates may be determined from maps or using GPS devices, and the point densities can be interpolated to generate grid or contour maps. The Summer Atlas of North American Birds (1995) is one such example that makes use of such point data (not using grids) collected by the North American Breeding Bird Surveys and Christmas bird counts. Others such as the EPOQ atlas for Quebec in Canada use 'trip lists', lists of birds seen at a place on a trip. A problem in atlas projects is the unequal distribution of available observers with some grids having very few visits. Such problems are often corrected for observer effort. In order to obtain a quantitative indicator, 'reporting rate' is sometimes used, although this requires multiple visits or samples to arrive at.