AbstractThe Brown Thrasher, like his fellow mimic thrushes, sings a highly variable song. So variable are the individual notes of the song that field identification on the basis of a single burst of song is hazardous. Reliable field identification traditionally has depended on the number of utterances of each sound rather than on the note qualities of the sounds themselves: Brown Thrashers (Toxostoma rufum) seem to utter sounds in pairs, whereas Catbirds (Dumetella carolinensis) do not repeat sound and Mockingbirds (Mimus polyglottos) repeat each sound several times. The purposes of this study were two: first, to establish that the thrasher does in fact distinguish his own song from that of the other mimic thrushes. Second, to develop evidence of the cues used to make the discrimination. Two techniques were used, playback trials and song recording analysis. In playback trials, Brown Thrashers were played samples of Brown Thrasher, Catbird, and Mockingbird song. Some samples were natural recordings and some were artificial samples in which the numerical properties of the songs had been distorted. For comparison purposes, an analogous set of recordings was played to Catbirds. The song recording analyses entailed sonagraphic reproduction of extensive segments of the songs of four thrashers and a careful study of the form, temporal properties, and statistical characteristics of the sounds thus represented. 1) Brown Thrashers clearly discriminate Brown Thrasher song from Catbird song and the basis of the discrimination is the number of utterances of each sound. Thrashers respond more vigorously to Brown Thrasher song than to Catbird song, to doubled Catbird song than to normal Catbird song, and to normal thrasher song than to halved thrasher song. 2) The results do not show that thrashers discriminate their own songs from Mockingbird song. Thrashers do respond more vigorously to normal Brown Thrasher song than to artificially lengthened Brown Thrasher song, but do not respond more vigorously to normal Brown Thrasher song than to normal Mockingbird song nor do they respond more vigorously to shortened Mockingbird song than they do to the normal form. 3) Catbirds can discriminate their songs from the songs of the other mimic thrushes, but they do not appear to make this discrimination on the basis of number. Catbirds respond more vigorously to the normal songs of their own species than to normal songs of the other two species but they do not respond more vigorously to shortened versions of the other species songs nor less vigorously to lengthened versions of their own song. 4) The sonagraphic analysis of thrasher song produced paradoxical results. Despite the importance of the two-ness in the field identification of thrasher song, no reliable property of two-ness was discernible in the sonagrams. Brown Thrashers did not reliably repeat each utterance nor pair utterances. In fact, the numerical properties of their songs, like all of the properties observed, overlapped extensively with the numerical properties of the other two mimic thrushes. However, these properties, like other properties studied, did differ on the average between the song of the thrasher and the song of the other mimic thrushes. The results suggest that the search for a single parameter that immediately distinguishes the songs of the species may be misguided. The songs of the mimic thrushes may differ only in the average value of several parameters and the birds like human observers in the field may have to hear several units of song before they can make a definite species identification. Despite the inefficiency of this process, the birds sing so rapidly and constantly that species identification of their songs would take no longer than species identification of slower but more consistent song.
Fast Software Implementation of Serial Test and Approximate Entropy Test of Binary Sequence
