Short-term temporal avoidance of interspecific acoustic interference among forest birds

Abstract

-We investigated whether birds of deciduous forests avoid temporal overlap in their songs. Singing patterns of four common species were analyzed. All species showed a strong tendency to avoid singing during the song of another species. One tactic used to avoid song overlap was to sing immediately after another species had stopped. Individuals sang more often and more regularly (less variance in the interval between songs) when they sang alone than when they sang with other species. The hypothesis that short-term adjustments in temporal patterning of songs minimize interspecific acoustic interference is supported. Received 26 November 1984, accepted 11 April 1985. BIRDS giving long-range vocalizations face several sources of noise. Sound is both degraded and attenuated as part of the natural transmission process (Wiley and Richards 1976, Marten and Marler 1977). A source of noise given little attention in the literature is interference between vocalizations given simultaneously. Interference may mask the signal carried in vocalizations. Acoustic overlap should be avoided, so as not to waste time or energy in vocalizations whose message will be lost. Several studies suggest that birds actively avoid both intraspecific and interspecific acoustic overlap through short-term changes in the temporal patterning of song delivery (Ficken et al. 1974, Wasserman 1977). While a single pair of species has been shown to adjust the temporal pattern of song delivery to avoid acoustic overlap (Ficken et al. 1974), this mechanism has not been investigated in an assemblage of birds, nor is there any supporting experimental evidence. This study investigates (1) whether interspecific avoidance of acoustic overlap occurs through shifts in the temporal patterning of songs in assemblages of chorusing forest birds; (2) if such avoidance occurs, how overlap is avoided; and (3) how the presence of singing individuals of another species affects the timing of singing. In addition, playbacks were used to experimentally test for avoidance of acoustic interference. METHODS Chorusing passerines were recorded in deciduous forest at the University of Wisconsin-Milwaukee Field Station, Saukville, Ozaukee County, Wisconsin and at Harrington Beach State Park, Lake Church, Ozaukee County, Wisconsin. The Field Station site was in an upland beech-maple forest; the dominant tree species were sugar maple (Acer saccharum), white ash (Fraxinus americana), beech (Fagus grandifolia), and basswood (Tilia americana) (Dunnum 1972). The state park site was a second-growth forest characterized by white ash, basswood, northern white-cedar (Thuja occidentalis), and paper birch (Betula papyrifera). The date, time, location, and species for each recording period are given in Table 1. A Sennheiser MKH 104 omnidirectional condenser microphone connected to a Nagra 4.2L tape recorder was used about 1.5 m off the ground. Portions of tapes containing high levels of background noise (e.g. planes, trains) were not analyzed. Three thousand seventythree songs were obtained from the remaining 173 min of recording. To determine relative timing of vocalizations, the tapes were analyzed using a Briiel and Kjaer 2305 Graphic Level Recorder with a 2,000Hz high-pass filter. The length of each song and interval between songs were then measured. Predicted number of song overlaps was determined using the method of Ficken et al. (1974). The predicted number of species A songs begun during