Population Estimates of Breeding Birds in North Dakota

Abstract

REASONABLY accurate measurements of breeding bird populations on large land areas are needed if the overall effects of man's impact on the environment are to be measured and understood. Through use of these data, we will be able to detect the average response of these populations to generalized environmental changes caused by diverse and widespread land-use practices. As this kind of information becomes available, it will serve as the basis for developing appropriate management programs for the purpose of maintaining or increasing breeding bird populations. Estimates of the summer population of birds in an entire state were first attempted for Illinois during 1907 (Forbes, 1908) and 1909 (Forbes and Gross, 1922), followed by a similar comparative study during 195658 by Graber and Graber (1963). These estimates were derived primarily from census data of summer birds on selected straight-line transects. Each of these routes contained segments of representative habitats that were considered to be typical of northern, central, and southern portions of the state. Projected statewide populations of summer birds were extrapolated by relating population densities per habitat type to known acreages of each habitat and expanding these data for the state as a whole. The Illinois studies represent a pioneering effort in the development of methods for censusing breeding birds over large areas. The resulting population estimates appear to be of reasonable magnitude. However, as the estimates were based on censuses of birds on sample units that were not randomly selected, their statistical reliability cannot be determined. The total breeding land bird populations of Great Britain (Fisher, 1940) and Finland (Merikallio, 1958) have also been estimated. In the British study the population data were derived from bird censuses on selected types of land including croplands, grazing lands, and forest. In Finland bird censuses were conducted along transects selected according to biological provinces based on soil productivity and features of vegetation. Apparently a statistical evaluation of the reliability of the population data was not considered feasible in either study. A more recent method for sampling breeding bird populations over extensive parts of the United States and Canada was developed by Robbins and Van Velzen (1967, 1969). This method is a cooperative venture, requiring large numbers of competent observers. The primary purpose is to obtain population indexes for each breeding species in