Background noise in the study area of Bunbury Bay measured as spectral noise levels (black, solid line) and one-third octave noise levels (squares) ±95% confidence intervals (grey area). Also included is the effective masking noise (crosses) given by the sum of spectral noise and bottlenose dolphin critical ratios (CR) (from Johnson et al., 1968).
Transmission loss coefficients estimated using sweeps (grey line) or pure tone (black squares) playbacks over a range of 250 m. The dip in transmission loss at 4 kHz is due to low S/N ratio of the playbacks at this particular frequency.
Acoustic localization of bottlenose dolphin whistle. A: Cross correlations of signals from four GPS synchronized recorders with time-of-arrival differences (t1−t3) given by the time difference of cross correlation peaks. B: Two-dimensional localization plot relative to the northernmost receiver (note the different scaling in the two axes). Each time-of-arrival difference gives rise to a hyperbola (h1−h3), and the source location S is estimated as the mean hyperbola intersection.
Whistle waveform (top), spectrogram (sample rate 48 kHz, FFT size 2048 samples, 95% overlap) (middle) and back-calculated cumulative acoustic energy radiated from the source (bottom).
(a) Source levels of whistles classified as either rise (upsweep), flat (constant frequency) or sine (loop) whistles. Lower and upper bounds on box indicate quartiles, and middle line and notch indicates the median and 95% confidence interval of the median. Whiskers show furthest point within 1.5 x interquartile range. No points lie outside the whiskers. (b) Histogram and smoothed density plot (2.5 dB Gaussian Kernel) of all whistles (N=180).
Source parameters of bottlenose dolphin (Tursiops sp.) whistles recorded in Bunbury Bay, West Australia.
Estimated costs of whistling for Bunbury Bay mother and newborn calf.
Article metrics loading...
Full text loading...