The habitat-induced degradation of the full song of the blackbird (Turdus merula) was quantified by measuring excess attenuation, reduction of the signal-to-noise ratio, and blur ratio, the latter measure representing the degree of blurring of amplitude and frequency patterns over time. All three measures were calculated from changes of the amplitude functions (i.e., envelopes) of the degraded songs using a new technique which allowed a compensation for the contribution of the background noise to the amplitude values. Representative songs were broadcast in a deciduous forest without leaves and rerecorded. Speakers and microphones were placed at typical blackbird emitter and receiver positions. Analyses showed that the three degradation measures were mutually correlated, and that they varied with log distance. Their variation suggests that the broadcast song could be detected across more than four, and discriminated across more than two territories. The song's high-pitched twitter sounds were degraded more rapidly than its low-pitched motif sounds. Motif sounds with a constant frequency projected best. The effect of microphone height was pronounced, especially on motif sounds, whereas the effect of speaker height was negligible. Degradation was inversely proportional to microphone height. Changing the reception site from a low to a high position reduced the degradation by the same amount as by approaching the sound source across one-half or one-whole territory. This suggests that the main reason for a male to sing from a high perch is to improve the singer's ability to hear responses to its songs, rather than to maximize the transmission distance. The difference in degradation between low and high microphone heights may explain why females, which tend to perch on low brush, disregard certain degradable components of the song.
Acoustical Society of America. Journal, 1993, Vol 93, Issue 4, p. 2206-2220