Biases
In some species, unpaired birds sing more. For example, unpaired male ovenbirds Seiurus aurocapillus and Kentucky warblers Oporonis formosus sing 3.5 and 5.4 times more often than paired males, respectively (Gibbs & Wenny 1993). In the same study, all unpaired males were detected but only 50% of paired male ovenbirds and 65% of paired male Kentucky warblers were located over ten visits. Male sedge warblers Acrocephalus schoenobaenus cease to sing as soon as they have attracted a mate, whereas males of the congeneric reed warbler Acrocephalus scirpaceus continue to sing once mated (Catchpole 1973). These differences in behaviour can make it unclear whether the breeding population is being censused.
The method assumes that birds live in pairs in fixed, discrete and non-overlapping ranges, which is often not the case (e.g. polygynous species, polyterritorial species). Despite standard guidelines for map interpretation, there is nevertheless a good degree of subjectivity involved and this can lead to variation, which makes comparison between studies difficult. The method can be unreliable at high densities, if birds are not readily visible, if registrations are plotted inaccurately,or if it is difficult to obtain many simultaneous registrations.