Oilbirds produce echolocation signals beyond their best hearing range and adjust signal design to natural light conditions

Abstract

Oilbirds are active at night, foraging for fruits using keen olfaction and extremely light-sensitive eyes, and echolocate as they leave and return to their cavernous roosts. We recorded the echolocation behaviour of wild oilbirds using a multi-microphone array as they entered and exited their roosts under different natural light conditions. During echolocation, the birds produced click bursts (CBs) lasting less than 10 ms and consisting of a variable number (2-8) of clicks at 2-3 ms intervals. The CBs have a bandwidth of 7-23 kHz at -6 dB from signal peak frequency. We report on two unique characteristics of this avian echolocation system. First, oilbirds reduce both the energy and number of clicks in their CBs under conditions of clear, moonlit skies, compared with dark, moonless nights. Second, we document a frequency mismatch between the reported best frequency of oilbird hearing (approx. 2 kHz) and the bandwidth of their echolocation CBs. This unusual signal-to-sensory system mismatch probably reflects avian constraints on high-frequency hearing but may still allow oilbirds fine-scale, close-range detail resolution at the upper extreme (approx. 10 kHz) of their presumed hearing range. Alternatively, oilbirds, by an as-yet unknown mechanism, are able to hear frequencies higher than currently appreciated.

Keywords: biophysical constraint; biosonar; convergent evolution; multi-modal integration; vision.

Figure 1.

Oilbirds emit sequences of echolocation click bursts (CBs) in low-light conditions. (a) Oscillogram of a CB sequence emitted by an oilbird flying away from its roost. In the raw sound recordings, oilbird echolocation signals were often partly masked by ambient noise obscuring CBs and clicks within CBs (grey trace; raw sound recording). Applying a spectral subtraction algorithm (black trace; see Material and methods) facilitated CB extraction. Vertical lines indicate the start (green) and end (red) of each CB as detected by our automated CB extraction procedure. Inset in the upper right of (a) shows that each CB consists of multiple separate clicks. To find clicks within CBs, we calculated the product of two spectral subtraction methods ([35,36]; see Material and methods). Normalizing and integrating this product (conditioned signal, light blue trace) provided a robust signal to reliably extract click time stamps within CBs. (b) Spectrogram of CBs in (a) displaying sound pressure power spectra (grey scale, dB re 20 µPa² Hz⁻¹; FFT length: 1024, overlap: 1000, Hamming window).

Figure 2.

Temporal parameters of oilbird echolocation signals change with lighting condition. We recorded oilbird echolocation signals under five natural conditions, one in full moon, the other four on dark nights (either around new moon or in overcast conditions): (i) birds flying solo out of Dunstan's Cave in moonlight, (ii) birds flying solo out of Dunstan's Cave in darkness, (iii) birds flying with one or more conspecific out of Dunstan's Cave in darkness, (iv) birds flying solo out of Aripo Cave (no water) in darkness, and (v) birds flying solo into Dunstan's Cave in darkness. One-way ANOVAs and Tukey HSD post hoc tests were performed (α = 0.05 for all tests). Bars illustrate group means with 95% confidence intervals; groups that differed significantly from one another do not share the same letter (Photo: Roger Ahlman—www.pbase.com/ahlman).

Figure 3.

Source levels of oilbird echolocation signals change with lighting condition. (a) Illustrated crescendo and decrescendo of consecutive click source level (peak-to-peak, ptp) over time within each CB (aligned to the loudest click per CB at time = 0). Note that the loudest clicks appear towards the middle of a given CB. (b) The loudest clicks in darkness have a significantly higher source level than those produced in moonlight (mean ± 1 s.d.). (c) Source level of the entire CB (root-mean-square, RMS) increases significantly with number of clicks per CB in both light and dark conditions. See Results for further information on statistical analyses.

Figure 4.

Spectral composition of oilbird echolocation signals misaligns with hearing range. (a) The sound amplitude spectrum (dB SPL) of 71 CBs (blue) and concurrent ambient noise segments (grey) show that CBs overcome ambient noise masking in a range from 2 to 35 kHz. The overlaid audiograms depict a best-frequency (BF) in oilbird hearing of approximately 2 kHz as measured by evoked potentials (range 1–4 kHz, +10 dB from BF; the black line depicts potentials from the inner ear, dashed black lines potentials from the forebrain nucleus, adapted from [40]). (b) Illustrates the spectral content of CBs in darkness and in moonlight after subtraction of ambient noise and using a relative dB range which is by convention set to zero for the lowest energy frequency components (blue; darkness (N = 71), yellow, moonlight (N = 65), bandwidth −6 dB from peak: darkness: 8–23 kHz; moonlight: 7–18 kHz). CBs therefore comprise an unexpected maximum energy plateau from 10 to 20 kHz at sound pressure levels approximately 10 dB higher than a CB local maxima at around approximately 2 kHz. Note, however, that the observed misalignment between plateau and the best frequency of oilbird hearing does not indicate oilbirds are deaf to their own echolocation signals, at least not at frequencies less than or equal to 8 kHz.