Peripheral auditory processing changes seasonally in Gambel's white-crowned sparrow

Abstract

Song in oscine birds is a learned behavior that plays important roles in breeding. Pronounced seasonal differences in song behavior and in the morphology and physiology of the neural circuit underlying song production are well documented in many songbird species. Androgenic and estrogenic hormones largely mediate these seasonal changes. Although much work has focused on the hormonal mechanisms underlying seasonal plasticity in songbird vocal production, relatively less work has investigated seasonal and hormonal effects on songbird auditory processing, particularly at a peripheral level. We addressed this issue in Gambel's white-crowned sparrow (Zonotrichia leucophrys gambelii), a highly seasonal breeder. Photoperiod and hormone levels were manipulated in the laboratory to simulate natural breeding and non-breeding conditions. Peripheral auditory function was assessed by measuring the auditory brainstem response (ABR) and distortion product otoacoustic emissions (DPOAEs) of males and females in both conditions. Birds exposed to breeding-like conditions demonstrated elevated thresholds and prolonged peak latencies when compared with birds housed under non-breeding-like conditions. There were no changes in DPOAEs, however, which indicates that the seasonal differences in ABRs do not arise from changes in hair cell function. These results suggest that seasons and hormones impact auditory processing as well as vocal production in wild songbirds.

Fig. 1

A representative white-crowned sparrow song from a breeding condition male. Songs typically consist of 5 syllables: a whistle, a warble, and 3 buzzes.

Fig. 2

Stimulus delivery schematics for three of the ABR paradigms. Horizontal arrows indicate passage of time. a) Schematic for the forward masking frequency resolution paradigm. A 100 msec pure tone masker is varied in frequency (Δf). The offset of the masker always occurs 10 msec before the onset of 3.3 kHz probe tone. b) Schematic for the forward masking temporal adaptation paradigm. The offset of a 100 msec band-limited (0.2–6kHz) white noise masker occurs at varying time intervals (Δt) before the onset of the 3.3 kHz probe tone. c) Schematic for the temporal variability paradigm. Clicks were presented at three different rates, with both fixed (left) and variable (right) inter-peak intervals.

Fig. 3

Representative frequency spectrum of a DPOAE recording from a breeding condition female. The primary tones (F1 and F2) were presented at the highest amplitudes (L1 = 90 dB SPL) to enable clear observation of the multiple distortion products. The distortion product with the largest amplitude is the cubic distortion tone (CDT), which corresponds to a frequency of 2F1-F2. F1 and F2 in this example are 7.4 and 8.5 kHz, respectively and the CDT is 6.3 kHz.

Fig. 4

Birds housed under breeding-like conditions have higher auditory thresholds than those housed under non-breeding-like conditions. a) Representative ABRs decrease in amplitude and increase in latency as stimulus intensity is decreased. Traces were elicited by a 4000Hz tone from a non-breeding female. The top 4 traces represent averages of 500 stimulus presentations. 35 and 30 dB SPL traces represent averages of 1000 presentations. The black arrow indicates stimulus onset. Scale bars = 2 μV/ 5 msec. Threshold was estimated to be 35 dB SPL and is indicated by the asterisk. One trace elicited by a 35 dB SPL stimulus is enlarged and shown over the original traces to more clearly demonstrate a response. For this trace only, the scale bar = 0.3 uV/5 msec. b) Representative ABR traces from a breeding female demonstrate an elevated threshold. Experimental parameters and figure notations are as in a. The top two traces represent averages from 500 stimulus presentations; the remaining traces were averaged over 1000 presentations. Threshold was estimated at 45 dB SPL. Scale bar is the same for a and b. c) Mean +/− S.E.M. ABR thresholds of birds exposed to breeding-like conditions (open circles) are higher than those housed under non-breeding-like conditions (closed circles) across all stimulus frequencies. Data are presented linearly (rather than logarithmically) for clarity. Thresholds to clicks are shown at the left most portion of each graph and are measured in dB peak equivalent (p.e.) SPL. Each experimental group had an n = 20 (except for clicks, where breeding birds n = 21).

Fig. 5

Birds housed under breeding-like conditions have longer ABR peak latencies and inter-peak intervals than those housed under non-breeding-like conditions. a) Representative ABR traces from a breeding (thin line) and non-breeding (thick line) female in response to a 4 kHz tone. Traces are aligned in time and stimulus onset occurs at time zero. The breeding bird has a delayed response compared to the non-breeding bird; note that this temporal disparity increases between peak 1 and peak 2. b) Peak 1 latencies of birds exposed to breeding-like conditions (open circles) are longer than those housed under non-breeding-like conditions (closed circles). The same pattern was observed for peak 2 latencies (c) and inter-peak intervals (d). Data are means +/− S.E.M. generated in response to iso-intensity tones (70 dB SPL) and clicks (70 dB p.e. SPL). Missing data points were filled in with appropriate group averages before data were plotted and analyzed (see main text). Breeding birds n = 19; non-breeding birds n = 20.

Fig. 6

Breeding condition does not affect frequency tuning. a) Thresholds for a 3.3 kHz probe tone in a forward masking paradigm are similar for breeding males (open circles; n = 5) and non-breeding males (closed circles; n = 5) across all masker frequencies. b) Average Q₁₀ values (indicative of tuning sharpness) did not differ between breeding males (open bar) and non-breeding males (shaded bar). One subject in each group had tuning curves too broad to accurately measure Q₁₀. Thus, n = 4 for each group in b.

Fig. 7

Breeding condition only affects temporal adaptation at the longest masker-probe interval tested. a) The probe-elicited ABR response amplitude decreases by a similar amount for males in breeding (open bars) and non-breeding (shaded bars) conditions when the masker and probe are separated by 5 msec. Similar results were found for (b) 10 msec and (c) 25 msec masker-probe intervals. d) When the masker and probe are separated by 50 msec, breeding males show a significantly greater decrease in response amplitude than non-breeding males. Data are means +/− S.E.M. Breeding males n = 6; non-breeding males n = 5 (except n = 4 at 40 and 50 dB SPL masker levels for 10, 25 and 50 msec intervals, and 40 dB SPL masker level for 5 msec interval).

Fig. 8

Breeding condition does not affect processing of temporally variable stimuli. Clicks were presented at three rates with both fixed and variable inter-click intervals. a) Non-breeding males show similar peak 1 latencies to fixed (grey bars) and variable (black bars) inter-click intervals for all presentation rates tested. Though breeding males showed a trend towards longer latencies in general, their responses to fixed (open bars) and variable (striped bars) stimuli were also similar. Similar results were found for peak 2 latencies (b) and inter-peak intervals (c). Data are mean +/− S.E.M. Breeding males n = 6, non-breeding males n = 4.

Fig. 9

Breeding condition does not affect DPOAE amplitudes or thresholds. a) DPOAE amplitude changes systematically with F2 frequency, but no amplitude differences are observed between breeding (open circles) and non-breeding (closed circles) birds. DPOAEs were elicited by iso-intensity primary tones (L1 = 70 dB SPL) for all frequencies tested. b) DPOAE amplitude increases with increasing stimulus level, but no difference is observed between breeding (open circles) and non-breeding (closed circles) birds. The amplitude of the noise floor immediately surrounding the DPOAE frequency is indicated by the shaded grey area. DPOAEs were elicited by iso-frequency primary tones (F2 = 7 kHz) for all levels tested. c) DPOAE threshold decreases with increasing stimulus frequency, but no difference is observed between breeding (open circles) and non-breeding (closed circles) birds. Data are means +/− S.E.M. Breeding birds n = 11 (except n = 8 at 20 and 25 dB SPL in b); non-breeding birds n = 11 (except n = 7 at 1000Hz in c).