Woodcock Singing-Ground Counts and Habitat Changes in the Northeastern United States

Abstract

Aerial photography from the late 1960's and the late 1970's was used to study changes along 78 American woodcock (Scolopax minor) singing-ground routes in 9 northeastern states. The most noticeable changes were declines in the amount of abandoned field, cropland, shrubland, and field/pasture. The amount of land in the urban/industrial type increased 33.4% from the late 1960's to the late 1970's. We examined relationships between the woodcock call-count index and variables using multipleregression techniques. The abundance of calling male woodcock was positively correlated with the amount of abandoned field and alder (Alnus sp.) and negatively correlated with the amount of urban/industrial type. However, only the change in the urban/industrial type was significantly (P < 0.05) related to the change in the call-count index. Urban/industrial area increased, whereas the call-count index declined on average in our sample of routes by 1.4 birds/route (40.5%). J. WILDL. MANAGE. 47(3):772-779 Counts of occupied American woodcock singing grounds were first used as an index of population trends by Mendall and Aldous (1943) in Maine. By 1970, the U.S. Fish and Wildlife Service had established a system of randomly located woodcock singing-ground routes throughout the principal woodcock breeding range (Clark 1970). Each woodcock singing-ground route has 10 stops, 0.64 km apart, at which an observer listens for calling male woodcock for a 2-minute interval. The number of singing males per route is used as an index of population size. Owen (1977) reported that woodcock in most states was declining in woodcock breeding range due to successional trends, urbanization, and commercial development. Cushwa et al. (1977) used forest inventory data from U.S. Forest Service national summaries and attempted to relate these data to the woodcock base. They combined 3 types of forest land-area statistics (unproductive forest land, seedling-sapling, and nonstocked stands) into potential woodcock habitat within each of 30 states. Their findings did not agree with Owen (1977), perhaps because potential woodcock habitat as defined in their study did not closely agree with the requirem nts of the species. It is also possible that the subjective comments of state biologists about trends in woodcock (Owen 1977) might be erroneous. Dobell (1977), however, was able to show a small but statistically significant correlation between a woodcock index and the number of calling males heard by route surveyors in New Brunswick. This relationship, and a decline in his woodcock index from 1953 to 1974, caused him to conclude that good singing-ground in the form of old fields will be lost eventually because the peak of farm abandonment is past. The objectives of our study were to (1) examine changes from the late 1960's to the late 1970's for a sample of randomly located woodcock singingground routes in the eastern United States and (2) compare changes in the woodcock population index with changes in the base. This analysis seems warranted because the annual woodcock population index for the eastern region has been de772 J. Wildl. Manage. 47(3):1983 This content downloaded from 207.46.13.172 on Sat, 15 Oct 2016 04:24:08 UTC All use subject to http://about.jstor.org/terms WOODCOCK CALL COUNTS AND HABITAT CHANGES * Dwyer et al. 773 creasing at a significant rate (J. Tautin, unpubl. rep., U.S. Dep. Inter., Fish and Wildl. Serv., Off. Migratory Bird Manage., Laurel, Md., 1979) during the past several years. If the predictions of Owen (1977) and Dobell (1977) are correct, changes in the base may be implicated in the decline. We especially thank U.S. Fish and Wildlife Service pilot biologists J. Goldsberry and A. Novara for the effort they expended in acquiring the 1978-80 photography. We thank P. Geissler and E. Heinlein for statistical support and J. Tautin for guidance in the use of woodcock singing-ground survey files stored in the Office of Migratory Bird Management. We also benefited from discussion with R. Munro on several aspects of multipleregression analysis. Finally, we thank J. Nichols, P. Geissler, G. Sepik, and R. Munro for reviewing the manuscript.