The possibilities for species misidentification based on vocalization or habitat association are high. However, the magnitude of the errors actually perpetrated is generally within an acceptable range in most types of bird survey work. Examples of problems discussed are: congeners that are similar in appearance or in song (such as Chimney and Vaux's Swifts, Chaetura pelagica, C. vauxi; Hammond's, Dusky and Gray Flycatchers, Empidonax hammondii, E. oberholseri, E. wrightii; Willow and Alder Flycatchers, E. traillii, E. alnorum; Common and Fish Crows, Corvus brachyrhynchos, C. ossifragus); birds that are misidentified because they are not expected by the observer (House Finches, Carpodacus mexicanus, invading new areas of eastern U.S.); birds that imitate other species (especially Starling, Sturnus vulgaris, and Mockingbird, Mimus polyglottos); birds in mixed flocks; birds with geographic differences in vocalizations (Solitary Vireo, Vireo solitarius); woodpeckers that are only heard drumming; and nests or eggs that are misidentified. Equally serious problems are the errors resulting from undetected species and from careless recording or failure to check manuscripts against original data. The quality of published count work can be improved considerably by (1) recognizing the problems that exist, (2) standardizing techniques for dealing with situations where not all birds can be identified, and (3) routinely applying all appropriate safeguards such as verification by mist netting and measuring, photography, tape recording or playback, additional observations, and careful verification of all entries in the final manuscript.
