|
Résumé :
|
Acoustic signals are important communication means in birds, especially in nocturnal owls where calls may constitute the only option for individual recognition. Studies of a number of owl species have now confirmed that their territorial calls are highly individual even though these calls are very stereotyped. This study aims at investigating whether territorial calls of the Little Owl vary in a distinct way among individuals. If so, the calls may provide important conservation tools since the species is declining all over its European range and may be sensitive to the time-consuming conventional marking methods.The study was conducted in Northern Jutland on 19-23 March and 17-22 April 2003. The calls were recorded between 20:30 and 02:25 on calm nights with temperatures of 0-5 degrees C, at distances of 10-45 meters from the owls. We did not use playback stimulation. A total of 247 calls of sufficiently high quality to allow spectrographic analysis were recorded using a SONY TC D5M audiotape recorder and an AUDIO-technica LO-Z condenser microphone. The calls came from five different males; four were recorded on one occasion only, while one (the Gulbaek male) was recorded in both March and April. Spectrograms of the calls were produced using AVISOFT SASLab Pro, sampling frequency 8000 Hz, FFT-length 256, frame size 50%, Hamming window, Bandwidth 81 Hz, overlap 75%, time and frequency resolution 8 ms and 3 Hz, respectively. Eight sound variables were measured: duration of call, peak frequency, 25%,50% and 75% energy quartiles, minimum and maximum frequency, and bandwidth.A Mann-Whitney W-test showed that the calls of the Gulbaek male in March and April, respectively, differed in six of the eight variables, and in four of variables the difference remained statistically significant after a Bon-ferroni correction of significance levels. The changing call may have been related to season or stage within the breeding cycle, so the comparison between males was only performed for the April calls, where calls from four males were available. All eight variables varied significantly among the four individuals (Kruskal-Wallis one-way ANOVA, df = 3, K = 102.3-130.5, P
|