1 2 Seasonal plasticity of auditory saccular sensitivity in the vocal plainfin midshipman fish ,

33 34 The plainfin midshipman fish, Porichthys notatus, is a seasonally breeding species of marine teleost fish that 35 generates acoustic signals for intraspecific social and reproductive-related communication. Female midshipman 36 use the inner ear saccule as the main acoustic endorgan for hearing to detect and locate vocalizing males that 37 produce multiharmonic advertisement calls during the breeding season. Previous work showed that the 38 frequency sensitivity of midshipman auditory saccular afferents changed seasonally with female reproductive 39 state such that summer reproductive females became better suited than winter non-reproductive females to 40 encode the dominant higher harmonics of the male advertisement calls. The focus of this study was to test the 41 hypothesis that seasonal reproductive-dependent changes in saccular afferent tuning is paralleled by similar 42 changes in saccular sensitivity at the level of the hair cell receptor. Here, I examined the evoked response 43 properties of midshipman saccular hair cells from winter non-reproductive and summer reproductive females to 44 determine if reproductive state affects the frequency response and threshold of the saccule to behaviorally45 relevant single tone stimuli. Saccular potentials were recorded from populations of hair cells in vivo while sound 46 was presented by an underwater speaker. Results indicate that saccular hair cells from reproductive females 47 had thresholds that were approximately 8 to 13 dB lower than non-reproductive females across a broad range of 48 frequencies that included the dominant higher harmonic components and the fundamental frequency of the 49 male’s advertisement call. These seasonal reproductive-dependent changes in thresholds varied differentially 50 across the three (rostral, middle, and caudal) regions of the saccule. Such reproductive-dependent changes in 51 saccule sensitivity may represent an adaptive plasticity of the midshipman auditory sense to enhance mate 52 detection, recognition and localization during the breeding season. 53 54 55 56 57 58 59 60 61 Sisneros, p.3 Ver. 2a 6/17/2009 INTRODUCTION 62 Recently a novel form of auditory plasticity that is adaptive for encoding social and reproductive-related 63 communication signals was reported for females of the plainfin midshipman fish (Porichthys notatus) (Sisneros 64 and Bass 2003, Sisneros et al. 2004a). Females rely greatly on their auditory sense to detect and locate males 65 that “sing” during the breeding season to attract mates. The detection and localization of such vocal signals is 66 essential to the reproductive success of this nocturnal species and can evoke in reproductive females strong 67 phonotactic responses (McKibben and Bass 1998; for recent reviews regarding the neuroethology of this 68 species see Bass 2006, Bass and Ladich 2008, Sisneros 2009). Previous research showed that the frequency 69 sensitivity of auditory saccular afferents from wild-caught females change seasonally with reproductive state 70 such that reproductive females become better suited than non-reproductive females to encode the dominant 71 higher harmonic components of the male’s seasonal advertisement call (Sisneros and Bass 2003). This 72 enhanced sensitivity to the dominant higher harmonics of the advertisement call likely increases signal detection 73 by females because the higher harmonics will propagate further than the call’s fundamental frequency due to 74 environmental and physical constraints of the shallow-water breeding habitat that limit sound transmission (Fine 75 & Lenhardt 1983; Bass & Clark 2003). Furthermore, work by McKibben and Bass (2001) showed that the higher 76 harmonics of the male’s advertisement call can also potentially affect the encoding of the call’s fundamental 77 frequency, which could be important for mate localization when near the sound source. 78 The seasonal shift in midshipman hearing sensitivity is preceded by a seasonal change in gonadal 79 reproductive state, which leads to elevated levels of circulating gonadal steroids in female midshipman. 80 Approximately 2-3 months before the breeding season, females undergo seasonal recrudescence of the ovaries 81 and then subsequently exhibit brief peaks in circulating plasma levels of testosterone and 17β-estradiol 82 approximately one month before the beginning of the summer spawning season (Sisneros et al 2004b). These 83 seasonal peaks in gonadal steroid hormones have been shown experimentally to induce the female’s 84 reproductive auditory phenotype and enhance the phase-locking accuracy of auditory saccular afferents at 85 higher frequencies that correspond to the dominant frequency content of male advertisement calls (Sisneros et 86 al 2004a). In addition, midshipman-specific estrogen alpha receptor was discovered in the sensory epithelium of 87 the inner ear saccule, the main organ of hearing in the midshipman and most other teleost fish, which provides 88 additional support for a direct steroid effect on the inner ear (Sisneros et al 2004a). This novel form of 89 reproductive-state and steroid-dependent auditory plasticity likely represents an adaptable mechanism that 90 Sisneros, p.4 Ver. 2a 6/17/2009 increases the probability of mate detection and localization by enhancing the frequency coupling between 91 sender and receiver in this vocal-acoustic communication system. A prime candidate site where this novel form 92 of auditory plasticity may occur is at the level of the saccular hair cell. 93 The major goal of this study was to determine if female reproductive state influences the auditory 94 sensitivity of the sensory hair cell receptors in the saccule and represents the initial steps of ongoing 95 neurophysiological investigations to determine the potential site(s) of action for the related reproductive-state 96 and steroid-dependent plasticity observed in the midshipman peripheral auditory system. Here, I characterize 97 and compare the response properties of auditory evoked saccular potentials from reproductive and non98 reproductive females to determine whether there are differences related to reproductive state in the frequency 99 response, dynamic range and auditory threshold of midshipman saccular hair cells to behaviorally-relevant 100 single tone stimuli. An auditory evoked potential recording technique recently described by Sisneros (2007) is 101 used to determine the frequency response of hair cells in the saccule. Saccular potentials in the midshipman, as 102 in other teleost fish, are maximally evoked at twice the auditory stimulus frequency due to the presence of non103 linear and opposite oriented hair cell populations in the fish saccule and can be used to characterize the 104 response of saccular hair cells (Hama 1969, Furukawa and Ishii 1967, Fay 1974, Fay and Popper 1974, 105 Sisneros 2007). Using this recording technique, I test the hypothesis that seasonal variation in female 106 reproductive state can modulate the sensitivity of auditory saccular hair cells to the dominant higher harmonic 107 components in male midshipman advertisement calls. The findings are interpreted as they relate to possible 108 auditory adaptations for acoustic communication during the breeding season. 109 110 METHODS 111 Experimental animals 112 Forty-eight adult female plainfin midshipman fish (P. notatus) with standard lengths (SL) that ranged 113 from 9.0– 19.5 cm SL were collected during the years 2006 and 2007 in both the non-reproductive season 114 (January-March) and in the reproductive season (May-July). During the non-reproductive season, females with 115 regressed ovaries that contained only small (<1mm diameter) undeveloped oocytes were collected by otter trawl 116 (R/V Kittiwake, Bio-Marine Enterprises, and the R/V Centennial, Friday Harbor Marine Laboratories) at depths 117 from 60 to 130 m in Puget Sound near Edmonds, WA, USA and in Bellingham Bay, WA, USA. These trawl 118 collected animals showed no visible signs of stress and rapidly adjusted to the sudden change in water depth 119 Sisneros, p.5 Ver. 2a 6/17/2009 while in captivity. During the reproductive season, gravid reproductive females with ovaries that contained 120 relatively large (~ 5 mm diameter) yellow/orange, yolked eggs were collected by hand from the nests of parental 121 (type I) males at low tide from a natural breeding population in the Hood Canal at Seal rock in Brinnon, WA, 122 USA. The reproductive state of the animal was determined by measuring the gonadosomatic index (GSI, 123 defined here as 100 ∗ gonad mass/body mass – gonad mass; according to Tomkins and Simmons 2002). All 124 animals were maintained in saltwater aquaria at 14-16oC and fed a diet of goldfish every 2-4 days. Saccular 125 potential recordings were performed within 15 days after collection from trawls or nests to avoid any effects of 126 captivity on auditory saccular sensitivity (Sisneros and Bass 2003). All experimental procedures followed 127 National Institutes of Health guidelines for the care and use of animals and were approved by the University of 128 Washington Institutional Animal Care and Use Committee. 129 130 Stimulus generation 131 Acoustic stimuli were generated via the reference output signal of a lock-in amplifier (SR830, Stanford 132 Research Systems, Sunnyvale, CA, USA) that was delivered to an audio amplifier and an underwater speaker 133 (UW-30, Telex Communications, Burnsville, MN, USA). The frequency response of the underwater speaker in 134 the experimental tank was measured using a mini-hydrophone (8103, Bruel and Kjaer, Norcross, GA, USA) that 135 was positioned 10 cm above the underwater speaker, which is the position normally occupied by the head of the 136 fish during the recordings. Relative sound measurements were then made using the mini-hydrophone and a 137 single channel FFT spectrum analyzer (SR780, Stanford Research Systems, Sunnyvale, CA, USA), calibrated 138 by peak-to-peak voltage measurements on an oscilloscope, and then adjusted with Matlab software so that the 139 sound pressures of the frequencies (75-385 Hz) tested were of equal amplitude within + 2 dB. Stimulus sound 140 levels were measured and described in terms of sound pressure. Although it is recognized that the midshipman 141 inner ear may be primarily sensitive to particle motion, the determination of sound level in terms of displacement 142 or particle motion is at best difficult. The relationship between particle motion and pressure in small tanks is 143 complex (Parvulescu 1967, Fay and Popper 1980) and the quantification and/or equalization of these two 144 measures is very difficult. However, previous studies have confirmed that the primary axis of particle motion in 145 this type of experimental tank setup is vertical and orthogonal to the surface plane of the underwater speaker 146 (McKibben and Bass 1999) and that the reflections from the tank walls and water surface did not alter the sound 147 pressure waveform of the acoustic signal (Bodnar and Bass 1997, 1999). 148 Sisneros, p.6 Ver. 2a 6/17/2009 Recent evidence indicates that many auditory saccular afferents of the midshipman respond to vertical stimuli or 149 dorsoventral acceleration and that the iso-level response curves based on pressure are similar in shape to iso150 intensity curves based on particle motion (Weeg et al. 2002). If the midshipman saccule is indeed primarily 151 particle motion-sensitive, then the two measures of sound (particle motion and pressure) will be proportional but 152 with a different proportionality at each frequency depending on the sound source and tank acoustics. The use of 153 sound pressure to describe the stimulus levels in this study should provide an interpretable measure of sound 154 stimuli that can be used to compare with previous midshipman auditory physiology studies and with other fish 155 species using a similar experimental setup (for extended discussion, see McKibben and Bass 1999, Weeg et al. 156 2002, also for recent review of underwater sound fields, see Bass and Clark 2003). 157 Basic auditory stimuli consisted of 8 repetitions of single 500-msec duration tones presented at a rate of 158 one every 1.5 seconds. Single tones were presented at 10 Hz increments from 75 to 145 Hz and at 20 Hz 159 increments from 165 to 385 Hz. The presentation order of single tone (frequency) stimulus was randomly 160 selected. To measure iso-level responses, the single tone stimuli were presented at a sound pressure of 130 dB 161 re 1 μPa, which is consistent with sound pressure levels for type I male midshipman calls recorded near their 162 nest (Bass and Clark 2003). To measure threshold tuning responses, single tone stimuli were presented at 163 sound pressure levels from 91 to 154 dB re 1 μPa in incremental steps of 3 dB. 164 165 Saccular potential measurements 166 Recording methods followed those used previously to characterize the evoked potentials from the 167 saccule in the midshipman fish (Sisneros 2007). Midshipman fish were first anesthetized by immersion in a 168 0.025% ethyl p-aminobenzoate (benzocaine) saltwater bath followed by an intramuscular injection of 169 pancuronium bromide (approximately 0.5 mg/kb) and 0.25% bupivacaine (approximately 1 mg/kg) for 170 immobilization and analgesia, respectively. The saccule was then exposed by a dorsal craniotomy. The cranial 171 cavity was filled with cold teleost Ringer‘s solution (Cavanaugh 1956) to enhance clarity and prevent drying of 172 the saccule. A 2-cm dam of denture adhesive cream was constructed around the exposed cranial cavity that 173 then allowed the animal to be lowered below the water line in the experimental tank. The fish was positioned in 174 the center of the tank such that the exposed saccule was below the water surface at a distance of 10 cm above 175 the underwater speaker, which was embedded in the sand/gravel on the bottom of a 30 cm diameter, 24 cm 176 high Nalgene tank (similar to that used by Fay 1990). The experimental tank was housed on a vibration 177 Sisneros, p.7 Ver. 2a 6/17/2009 isolation table inside an acoustic isolation chamber (Industrial Acoustics, New York, NY, USA). All the recording 178 and auditory stimulus generation equipment was located outside the chamber. Fish were perfused with chilled 179 seawater (14-15o) via a small plastic tube that was inserted into the fish’s mouth to provide a continuous stream 180 of recirculated seawater across the gills. 181 Evoked potentials from the saccule were recorded with glass microelectrodes filled with 3M KCl (1-7 182 MΩ). The recording electrode was visually guided into the endolymph of the saccule and positioned 183 approximately 2-4 mm away from the sensory bed of hair cells (macula) in one of the three recorded regions 184 (rostral, middle, and caudal) in either the right or left saccule (see Fig. 1). Analog saccular potentials were 185 amplified (x100) (model 5A, Getting Instruments), inputted into a digital signal processing Lock-in amplifier 186 (SR830, Stanford Research Systems), and then stored on a PC computer running a custom data acquisition 187 Matlab software program. The lock-in amplifier yields a DC voltage output signal that is proportional to the 188 component of the signal whose frequency is exactly locked to the reference frequency. The reference frequency 189 was set to the second harmonic of the stimulation frequency signal (i.e., 2x fundamental frequency) since the 190 maximum evoked potential from the teleost saccule occurs at twice the sound stimulus frequency due to the 191 non-linear response of opposite oriented hair cell populations within the saccule (Zotterman 1943, Furukawa and 192 Ishii 1967, Cohen and Winn 1967, Hama 1969). Noise signals at frequencies other than the reference frequency 193 were rejected by the lock-in amplifier and did not affect the saccular potential measurements. 194 To measure and compare the evoked iso-level responses of the saccule, the evoked saccular potentials 195 were first measured at 130 dB re 1μPa and then the response profiles were constructed using saccular potential 196 data that was normalized. The magnitude of the evoked saccular potentials varied depending on the distance 197 between recording electrode and the sensory bed of hair cells. In order to control for differences in the absolute 198 magnitude of the evoked saccular potentials recorded from different animals and from different recording regions 199 on the saccular macula (i.e., rostral, middle, and caudal areas of the macula), the signal averaged saccular 200 potential data were normalized and expressed relative to a value of 0 dB that was assigned to the maximum 201 evoked saccular potential recorded at the peak frequency sensitivity or best frequency (BF) for each record. The 202 BF was defined as the stimulus frequency that produced the greatest evoked saccular potential response. Iso203 level profiles were recorded from three regions of the saccule: rostral, middle, and caudal (Fig. 1). The response 204 profile of the evoked saccular potentials was determine for each saccular recording region and then summarized 205 Sisneros, p.8 Ver. 2a 6/17/2009 in plots. This analytical method allows for the relative comparisons of response profiles for the three recording 206 regions of the saccule (Sisneros 2007). 207 Data analysis 208 Threshold tuning curves were constructed by characterizing the input-output measurements of the 209 evoked saccular potentials over the range of stimulus levels, in incremental steps of 3 dB, from 88 to 154 dB re 1 210 μPa at the tested stimulus frequencies (see above). In addition, background noise measurements (RMS) were 211 also recorded for 8-10 repetitions of the stimulus interval at each of the tested frequencies from 75 to 385 Hz 212 (see above) with no auditory stimulus present. These background noise measurements were then used to 213 establish the subthreshold saccular potential response levels and determine the auditory threshold at each 214 frequency. The measurement of background noise was performed prior to the recording of each threshold tuning 215 curve. The auditory threshold at each stimulus frequency was designated as the lowest stimulus level that 216 evoked a saccular potential that was at least 2 standard deviations above the background noise measurement. 217 The frequency that evoked the lowest saccular potential threshold was defined as the BF. 218 219 Statistical analysis 220 The effects of reproductive state (non-reproductive vs. reproductive females) and the recording region of 221 the saccule (rostral vs. middle vs. caudal) on the iso-intensity BF of saccular hair cells were determined by a 222 two-way ANOVA. Differences in body size (standard fish length) and response magnitude (relative gain) of the 223 saccular evoked potentials between non-reproductive and reproductive females were determined by t-tests. The 224 association between body size (standard fish length) and the response magnitude (relative gain) of the saccular 225 evoked potentials to iso-level tones of 130 dB (re 1 μPa) among non-reproductive and reproductive females was 226 analyzed using Pearson product-moment correlation. The overall effects of reproductive-state and stimulus 227 frequency on the auditory thresholds of saccular hair cells were analyzed using a repeated measures ANOVA 228 with thresholds for each of the 20 frequencies tested (75-385 Hz) as repeats (i.e. within-subject factors) and 229 reproductive-state of the animal as the between-subject factor. In order to determine the effects of reproductive 230 state and stimulus frequency based on saccular recording position, a separate repeated measure ANOVA 231 analysis was performed on data from each recorded region (ie., rostral, middle and caudal) with the saccular 232 thresholds for each of the 20 frequencies tested (75-385 Hz) as repeats (i.e. within-subject factors) and 233 reproductive-state of the animal as the between-subject factor. In the few cases where there were two positional 234 Sisneros, p.9 Ver. 2a 6/17/2009 recordings from the same individual fish (e.g., rostral threshold measurements from both the left and right 235 saccule), the two threshold tuning curves were then averaged so that a repeated measures ANOVA analysis 236 could be performed. P value corrections were made for all tests of within-subject effects based on the calculated 237 estimates of sphericity (equality of variances of the differences between measurements) (Girden 1992). Since 238 epsilon was < 0.75 in all analyses, the more conservative correction of Greenhouse-Geisser was used to 239 calculate P. The 95% confidence limits (CL) of the mean thresholds (Zar 1999) were calculated and used to 240 determine whether the mean evoked saccular thresholds differed between reproductive and non-reproductive 241 females at each frequency (i.e., overlapping 95% CL were considered not significantly different). For all 242 statistical analyses, α was set at 0.05. Statistical analyses were performed using the software programs 243 Statistica for Windows (StatSoft, Inc.) and Systat 7.0 (Systat software Inc). 244 245 RESULTS 246 Iso-level responses of the evoked saccular potentials 247 The evoked saccular potentials were recorded from a total of 48 adult female midshipman fish: 24 winter 248 non-reproductive females with a size range of 9.0 to 17.5 cm standard length (SL) (mean SL = 12.1 + 1.9 SD 249 cm, mean body mass (BM) = 22.0 + 9.3 SD g, mean gonadosomatic index (GSI) = 5.6 + 5.2 SD) and 24 250 summer reproductive females with a size range of 12.3 to 19.5 cm SL (mean SL = 16.8 + 1.6 cm, mean BM = 251 62.0 + 17.1 SD g, mean GSI = 18.9 + 9.8 SD). Iso-level response profiles of the evoked saccular potentials were 252 generated from the presentation of single tone stimuli that ranged from 75 to 385 Hz. Figure 2 shows 253 representative iso-level response curves of evoked saccular potentials to single tones of 130 dB (re 1 μPa) 254 recorded from the rostral, middle and caudal regions of the saccule. In general, the evoked saccular potentials 255 were much higher in the middle and caudal regions compared to that of the rostral region of the saccule (Fig. 2). 256 The iso-intensity response curves consisted of profiles that had best frequencies (BF, defined as the frequency 257 that evoked the greatest saccular potential) < 85 Hz with the evoked potentials declining rapidly above BF. BFs 258 ranged from 75 Hz to 145 Hz with the majority of BFs at 75 Hz (winter: 72%, summer: 51%), the lowest 259 frequency tested. The mode of BFs (75 Hz) of the evoked potentials to tones of 130 dB (re 1 μPa) did not differ 260 across position (rostral, middle, caudal) along the saccule (two-way ANOVA, effect of position, F = 0.44, df = 2, 261 61, p = 0.64) or between non-reproductive and reproductive females regardless of the recording position along 262 Sisneros, p.10 Ver. 2a 6/17/2009 the saccule (two-way ANOVA, effect of season, F = 0.21, df = 1, 61, p = 0.65; interaction of reproductive-state 263 and position, F = 1.36, df = 2, 61, p = 0.26). 264 Because BF did not differ among the rostral, middle, and caudal regions of the saccule, the iso-level 265 saccular potential data were first normalized and expressed relative to a value of 0 dB for the BF in each record 266 and then were pooled for both non-reproductive and reproductive females. The normalized iso-level response 267 data was averaged and then complied to construct a relative gain plot summarized in Fig 3, which shows the 268 relative dynamic range of the saccular potentials evoked from non-reproductive and reproductive females. In 269 general, the recorded iso-level saccular potentials evoked at 130 dB re 1μPa were greatest at 75 Hz with a rapid 270 decline in sensitivity (gain) of evoked potentials above 95 Hz to the lowest levels at 305-385 Hz. Although non271 reproductive females were smaller than reproductive females (t-test, t= -9.17, df = 46, p < 0.05), there was no 272 relationship between female size (SL) and the range of the response magnitude or relative gain of the evoked 273 saccular potentials based on female reproductive state (non-reproductive females, Ho: β= 0, t = 0.58, p = 0.56; 274 reproductive females, Ho: β= 0, t = -0.44, p = 0.96). However, the magnitude of the evoked saccular potentials 275 were greater in reproductive females compared to the non-reproductive females (t-test, t = 6.64, df = 74, p 276 <0.01). The dynamic range of relative gain from 75-385 Hz was approximately 13 dB greater in reproductive 277 females (dynamic range = 44 dB) than in non-reproductive females (dynamic range = 31 dB) (Fig. 3). Based on 278 saccular region, this seasonal difference in dynamic range of relative sensitivity (gain) from 75-385 Hz between 279 reproductive and non-reproductive females was approximately 16 dB, 18 dB, and 7 dB for the rostral, middle, 280 and caudal positions, respectively (Fig. 4). 281 282 Seasonal differences in auditory saccular sensitivity 283 Auditory threshold tuning curves were determined for whole populations of hair cells in the rostral, 284 middle, and caudal regions of the saccule in both non-reproductive and reproductive female fish. Figure 5 285 shows representative tuning curves recorded from the three different regions of the saccule. Threshold tuning 286 curves for the saccular potentials generally consisted of profiles with lowest thresholds at frequencies < 145 Hz 287 that steadily increased to the highest thresholds at frequencies > 305Hz (Fig. 5). Best frequencies (BFs, defined 288 as the frequency that evoked the lowest saccular potential threshold) ranged from 75 Hz to 145 Hz for non289 reproductive females and 75 Hz to 135 Hz for reproductive females with a mode of BFs occurring at 75 Hz for 290 both non-reproductive and reproductive females. The distribution of BFs based on the threshold tuning profiles 291 Sisneros, p.11 Ver. 2a 6/17/2009 did not differ by recording position (rostral, middle, caudal) along the saccule (two-way ANOVA, effect of 292 position, F = 0.99, df = 2, 61, p = 0.37) nor by reproductive-state of the animal regardless of recording position 293 along the saccule (two way ANOVA, effect of reproductive-state, F = 1.82, df = 1, 61, p = 0.18; interaction of 294 season/reproductive-state and saccular position, F = 1.58, df = 2, 61, p = 0.21). 295 Although there was no seasonal difference in BF, there were significant threshold differences at the 296 tested frequencies between non-reproductive female and reproductive female midshipman. In general, auditory 297 thresholds were significantly lower at every frequency (75-385 Hz) in reproductive females than in non298 reproductive females (repeated measures ANOVA: between-subject factor ‘reproductive-state’: F = 117.20, df = 299 1,65, P < 0.001) with no significant interaction of frequency between females of different reproductive states 300 (repeated measure ANOVA: within-subject contrast ‘frequency*reproductive-state’: F = 1.96, df = 19,1235, P = 301 0.058; assumptions of homogeneity of variances were met for each of the tested frequencies: Levene’s test, P 302 values > 0.09). The threshold tuning curves for non-reproductive and reproductive females are summarized in 303 Fig. 6. In addition, there were also significant threshold differences at various frequencies of the tuning curves 304 from the three saccular regions that varied by reproductive-state of the animal and saccular position. In general, 305 the threshold tuning curves for each region of the saccule were similar in shape with lowest thresholds at 75 and 306 85 Hz that rapidly increased to highest thresholds at 365 and 385 Hz (Fig. 7). Auditory thresholds recorded from 307 the rostral region of the saccule were lower in reproductive females at 165 Hz, 185 Hz, 205 Hz and 225 Hz 308 compared to non-reproductive females (repeated measures ANOVA: between-subject factor ‘reproductive309 state’: F = 17.48, df = 1, 14, P < 0.001; within-subject contrast ‘frequency*reproductive-state’: F = 0.68, df = 19, 310 266, P = 0.69). In contrast, the threshold recorded from the middle saccular region were lower in reproductive 311 females at every frequency except 95 Hz, 125 Hz, 345 Hz, and 365 Hz compared to non-reproductive females 312 (repeated measure ANOVA: between-subject factor ‘reproductive-state’: F = 26.48, df = 1, 19, P < 0.001; within313 subject contrast ‘frequency*reproductive-state’: F = 1.30, df = 19, 361, P = 0.25). Similarly, thresholds recorded 314 from the caudal saccular region were lower in reproductive females at every frequency except 75 Hz compared 315 to non-reproductive females (repeated measure ANOVA: between-subject factor ‘reproductive-state’: F = 79.82, 316 df = 1, 28, P < 0.001; within-subject contrast ‘frequency*reproductive-state’: F = 1.03, df = 19, 532, P = 0.41). 317 Thus, auditory saccular thresholds were lower in reproductive females compared to non-reproductive females 318 with thresholds varying differentially across the three (rostral, middle, and caudal) regions of the saccule. 319 320 Sisneros, p.12 Ver. 2a 6/17/2009 DISCUSSION 321 The goal of this study was to determine if seasonal variation in gonadal reproductive state can modulate 322 the sensitivity of auditory saccular hair cells to behaviorally-relevant single tone stimuli in female midshipman 323 fish. The results from this evoked potential recording study indicate that the auditory saccular hair cells from 324 reproductive females had lower tuning thresholds and were more sensitive to all the frequencies tested (75-385 325 Hz) compared to that of saccular hair cells from non-reproductive females. This study represents the first 326 attempt to determine the potential site of action for the related reproductive-state and steroid-dependent 327 plasticity observed in the previous saccular afferent studies (Sisneros and Bass 2003, Sisneros et al. 2004a). In 328 this discussion, I interpret the results as they relate to the encoding of vocal communication signals for the 329 midshipman fish and discuss possible adaptations for the seasonal plasticity of auditory tuning and how this 330 seasonal shift in tuning may facilitate acoustic communication during the midshipman breeding season. 331 332 Seasonal changes in the iso-level response of the evoked saccular potentials 333 In general, the iso-level saccular potential records indicate that hair cells from the midshipman saccule 334 were most sensitive to frequencies < 145 Hz at the behaviorally relevant sound level of 130 dB re 1 μPa, a level 335 that is consistent with known sound levels for midshipman advertisement calls recorded near the entrance of 336 type I male nests (Bass and Clark 2003). These iso-level responses of midshipman saccular hair cells were 337 characterized to pure tones at 130 dB re 1 μPa so that they could be compared to the responses of primary 338 auditory afferents characterized at similar levels in previous studies (McKibben and Bass 1999, Sisneros and 339 Bass 2003). Results indicate that the iso-level response profiles of the saccular potentials (Fig. 2) were similar to 340 the profiles plotted for auditory afferent spike rate at the same sound level in a previous study (see figure 3 in 341 Sisneros and Bass 2003). McKibben and Bass (1999) and Sisneros and Bass (2003) previously reported a 342 secondary mode of BFs at approximately 140 Hz for auditory afferents in midshipman fish which is congruent 343 with the iso-level profiles of the saccule in non-reproductive and reproductive females reported here that show a 344 secondary peak around 145 Hz (Fig. 3, but also see individual representative profiles in Fig. 2). Remarkably the 345 distribution of BFs for the midshipman saccule reported here and in a previous study (Sisneros 2007) is similar 346 to the distribution of BFs for the saccule in the toadfish (Opsanus tau), which has bimodal peaks at 74 Hz and 347 140 Hz (Fay and Edds-Walton 1997a). In O. tau, there is no evidence for regional differences in frequency 348 sensitivity along the saccule (Fay and Edds-Walton 1997b), which is also consistent with the frequency 349 Sisneros, p.13 Ver. 2a 6/17/2009 sensitivity data reported here that shows no difference in peak sensitivity (75 Hz) along the three recorded 350 regions of the saccule for females in a given reproductive state (i.e., non-reproductive vs. reproductive). 351 The iso-level response profiles for saccular potentials evoked at 130 dB re 1μPa indicate that saccular 352 hair cells from reproductive females had a greater magnitude and relative gain of evoked potentials compared to 353 that of non-reproductive females. Reproductive females had a dynamic range (expressed in relative gain or 354 sensitivity) that was 14 dB greater than the range of evoked saccular potentials from non-reproductive females 355 (Fig. 3). Although there was no relationship between female size (SL) and the range of the response magnitude 356 or relative gain of the evoked saccular potentials based on female reproductive state, there was considerable 357 variation in the range of the evoked saccular potentials from both non-reproductive and reproductive females 358 (see supplementary figure). Nevertheless, this seasonal increase in the sensitivity of the evoked saccular 359 response was found to occur across the entire saccule in reproductive females having a range that was 16 dB, 360 18 dB and 9 dB greater than that of non-reproductive females for the rostral, middle and caudal saccular 361 positions, respectively. This difference in seasonal sensitivity could be potentially due to changes in hair cell 362 responsiveness to stimuli in the vertical axis. Although the hair cell orientation patterns of the saccule are not 363 known for P. notatus, many auditory primary afferent neurons that innervate midshipman saccular hair cells are 364 known to respond to vertical stimuli or dorsoventral acceleration (Weeg et al. 2002, Sisneros and Bass 2003). 365 Based on the evoked potentials recorded from each of the three saccular regions (rostral, middle and caudal), it 366 is likely that the hair cells found in each of the three regions have hair cell orientation patterns that respond to 367 stimuli in the vertical axis. In the closely related toadfish Opsanus tau, the hair cell orientation pattern in the 368 saccule gradually changes from patches of rostrally oriented hair cells in the rostral region to primarily vertical 369 oriented hair cells in the middle region to caudal oriented hair cells in the caudal region (Edds-Walton and 370 Popper 1995). In this study, the evoked potentials recorded from the rostral region of the saccule were the 371 lowest in magnitude and more variable compared to evoked potentials from the middle and caudal regions of 372 the saccule. Such differences are likely related to hair cell orientation patterns found in the midshipman saccule. 373 It would be interesting to know in future work whether hair cells found in the middle and caudal saccular regions 374 are oriented to receive particle motion stimuli from the vibrations of the fish’s swim bladder produced by the 375 pressure component of the sound during sound reception. Future studies that examine the distribution, 376 morphology and orientation patterns of saccular hair cells in the midshipman will be insightful in determining 377 Sisneros, p.14 Ver. 2a 6/17/2009 how hair cell orientation patterns relate to the evoked saccular potentials recorded from different regions of the 378 saccule. 379 One possible explanation to account for such seasonal differences in the magnitude and range of the 380 evoked potentials could be due to a differential seasonal increase in the relative number of saccular hair cells 381 that response to vertical axis stimulation in reproductive females compared to that of non-reproductive females. 382 A number of studies have demonstrated that fish continue to add hair cells for a number of years post383 embryonically (Platt 1977, Corwin 1983, Lombarte and Popper 1984, Popper and Hoxter 1984), but only a few 384 studies have examined the relationship between hair cell addition and auditory sensitivity of the fish inner ear. 385 Based on multiunit recordings of primary afferents from the macula neglecta in the skate Raja clavata, Corwin 386 (1983) demonstrated that an increase in auditory sensitivity was correlated with an increase in the number of 387 hair cells innervated by individual auditory afferents. Similarly, Sento and Furukawa (1987) also found that the 388 sensitivity of auditory saccular afferents in the goldfish, Carassius auratus, was correlated with the number of 389 hair cells innervated by individual afferent neurons. In addition, age/size-related increases in behavioral auditory 390 threshold sensitivity have been reported using behavioral conditioning techniques for two species of teleost fish, 391 the damselfish (Pomacentrus sp.) and the Red Sea bream Pagrus major (Kenyon 1996, Iwashita et al. 1999). 392 In contrast to previous findings, Higgs et al. (2002, 2003) using the auditory brainstem recording 393 technique showed that auditory sensitivity did not increase with age and size of the zebrafish, Danio rerio. 394 Similarly, Popper (1971) showed that behavioral auditory sensitivity in the goldfish (C. auratus) did not change 395 with age/size (and presumably with hair cell addition which is correlated with age/size) between two subadult 396 groups of goldfish. Although the study by Popper (1971) did not directly test the effects of hair cell addition on 397 auditory sensitivity of the goldfish saccule, the results are congruent with one model of fish hearing that predicts 398 hair cell addition with fish growth will maintain sound detection and processing capabilities as the relative sizes 399 and positions of different structures associated with fish hearing change during ontogeny (Popper et al. 1988, 400 Rogers et al. 1988). 401 Alternatively, the seasonal differences in the magnitude and relative gain (sensitivity) of the evoked 402 midshipman saccular potentials could be related to reproductive-state dependent effects on the central input 403 from the hindbrain efferent nucleus and the efferent neurons that innervate the inner ear of the midshipman and 404 other teleost fishes (Bass et al. 1994). Efferents to the saccule provide inhibitory inputs to saccular hair cells and 405 can directly modulate auditory sensitivity (gain) to auditory stimuli (Furukawa and Matsura 1978, Lin and Faber 406 Sisneros, p.15 Ver. 2a 6/17/2009 1988). Recent evidence suggests that auditory efferent feedback can modulate the signal to noise ratio of 407 auditory responses from saccule to brain in teleost fish such that the efferent regulatory effects can be either 408 excitatory or inhibitory depending on the level of background noise (Tomchik and Lu 2006). Future studies that 409 investigate the seasonal addition of hair cells in the saccule and the potential seasonal effects of efferent 410 activation and modulation of auditory hair cell sensitivity will be needed in order to determine the mechanism for 411 the observed seasonal differences in relative sensitivity (gain) of saccular potentials found between reproductive 412 and non-reproductive female midshipman fish. 413 414 Seasonal plasticity of auditory saccular sensitivity and its functional significance 415 Comparison of the tuning curves revealed a dramatic seasonal difference in the auditory thresholds of 416 saccular hair cells between reproductive female and non-reproductive females. The auditory thresholds of 417 saccular hair cells from reproductive females were approximately 8 to 13 dB lower than non-reproductive 418 females at frequencies from 75 to 385 Hz compared to (Figs. 6 and 8). Furthermore, the auditory thresholds of 419 saccular hair cells from reproductive females were at least 9.5 dB or lower (a sensitivity increase equal to 3X or 420 greater) than non-reproductive females at frequencies that corresponded to the fundamental frequency (~ 100 421 Hz) and the dominant second harmonic (~ 200 Hz) component of the male’s advertisement call (Fig. 8). It is 422 important to note the differences of the evoked saccular responses reported in Figures 3 and 6. The iso-level 423 gain data (Fig. 3) illustrated that the evoked saccular potentials of reproductive females stimulated at 130 dB re 424 1μPa showed progressively more relative gain starting at 200 Hz up to 385 Hz compared to non-reproductive 425 females. These results suggest that the plasticity may potentially be frequency dependent, which would be 426 consistent with previous single unit data of saccular afferents that showed an increase in phase-locking 427 accuracy within the same frequency range (Sisneros and Bass 2003). In contrast, the threshold data (Fig. 6) 428 showed a near parallel shift in sensitivity with reproductive females having a greater sensitivity over the same 429 frequency range compared to non-reproductive females, which suggests an overall increase in sensitivity that is 430 not frequency dependent. This apparent discrepancy is noteworthy and should be investigated in future work. 431 An analysis of the general receptor potential dynamics with a description of the slopes from the input-output 432 functions of the evoked saccular potentials would be very informative in future work in determining any possible 433 frequency dependent effects that could not be resolved in this study. 434 Sisneros, p.16 Ver. 2a 6/17/2009 An important finding reported in this study was that the seasonal change in auditory threshold occurred 435 across a broad range of frequencies that included the dominant higher harmonic components and the 436 fundamental frequency of the male’s advertisement call. This seasonal change in auditory threshold may 437 function to increase the probability of conspecific mate detection and localization during the summer breeding 438 season as previously proposed by Sisneros and Bass (2003). The seasonal shift in auditory saccular sensitivity 439 may be adaptive for females to enhance the acquisition of auditory information needed for mate detection and 440 localization when both far and near the sound source during the breeding season, especially in shallow water 441 environments like those where midshipman fish court and spawn. The dominant harmonics of the advertisement 442 call that range up to 400 Hz often contain as much or more spectral energy than the fundamental frequency and 443 are hypothesized to be important for the detection and sound source localization of the advertisement call by 444 females during the midshipman breeding season. The mate call’s harmonics likely increase signal detection by 445 the receiver because the higher frequency harmonics will propagate further than the mate call’s fundamental 446 frequency due to the inverse relationship between water depth and the cutoff frequency of sound transmission 447 (Fine and Lenhardt 1983, Roger and Cox 1988, Bass and Clark 2003). According to this relationship, as water 448 depth decreases the frequency below which sound transmission is negligible (cutoff frequency) increases. Thus, 449 in very shallow water (< 5m) low frequency signal components such as the mate call’s fundamental frequency 450 should attenuate rapidly while the higher frequency harmonics that are above the cutoff frequency should 451 propagate more readily away from the sound source. However, the enhanced detection of the fundamental 452 frequency (~100 Hz) would also be equal important in localizing the advertisement call when the female receiver 453 is very close or near the sound source. In addition, the dominant higher harmonics of the advertisement call may 454 also affect the encoding of the call’s fundamental frequency when near the sound source. Previously McKibben 455 and Bass (2001) showed that the encoding of the advertisement call’s fundamental frequency was enhanced 456 when harmonics were added to tonal stimuli that were similar to the fundamental frequencies of the male’s 457 advertisement call. Interestingly, recent evidence also suggests that the lateral line system of the midshipman 458 fish is also sensitive to the fundamental frequency component of the male’s advertisement call (Weeg and Bass 459 2002) with supporting neuroanatomical evidence for the central integration of lateral line and auditory 460 information (Weeg and Bass 2000, Edds-Walton and Fay 2003). In sum, the seasonal increase in auditory 461 saccular sensitivity may represent an adaptive plasticity of the midshipman auditory sense for the enhancement 462 of mate detection and localization during the breeding season. 463 Sisneros, p.17 Ver. 2a 6/17/2009 Currently it is not known whether the observed peripheral auditory plasticity also extends to the male 464 morphs (types I and II) of the plainfin midshipman fish. Type I males build nests, acoustically court females, and 465 provide parental care for fertilized eggs during the breeding season, whereas type II males neither build nests 466 nor acoustically court females, but instead sneak or satellite spawn to steal fertilizations from type I males 467 (Brantley and Bass 1994). There is no a priori reason to expect that the auditory plasticity of peripheral 468 frequency sensitivity be limited to only females. The seasonal enhancement of frequency sensitivity for 469 conspecific detection and localization would also be adaptive for type I or nesting males during mate competition 470 in the establishment of nest sites and in the case of type II or sneaker males in the selection of cuckoldry sites 471 for sneak or satellite spawning. Thus, future studies will be needed to determine whether seasonal reproductive472 state and/or steroid-dependent plasticity occurs in the male peripheral auditory system. In addition, similar 473 mechanisms of auditory plasticity may also be operative in other vertebrate groups were studies have 474 suggested either seasonal or steroid-related changes in audition, which includes recent studies of birds (Lucas 475 et al. 2002, Lucas et al. 2007), amphibians (Penna et al. 1992, Goense and Feng 2005, Gordon and Gerhardt 476 2009) and humans (Guimaraes et al. 2006, Hultcrantz et al. 2006). 477 478 Mechanisms for the plasticity of saccular sensitivity 479 As demonstrated for the auditory saccular afferents, the mechanism responsible for the observed 480 changes in auditory sensitivity of midshipman saccular hair cells is most likely dependent on seasonal changes 481 in circulating levels of gonadal steroids, specifically testosterone (T) and 17β-estradiol (E2). Approximately 1 482 month before the beginning of breeding season, female midshipman fish exhibit annual seasonal peaks in 483 circulating levels of T and E2 (Sisneros et al. 2004b). Experimental implants of T and E2 in female midshipman 484 confirmed a steroid-dependent effect and induced an increased phase-locking accuracy of the auditory saccular 485 afferents at higher frequencies within the midshipman’s hearing range, especially at frequencies that 486 corresponded to the dominant frequency content of the male’s advertisement call (Sisneros et al. 2004a). 487 Concurrent with these findings, midshipman-specific estrogen receptor alpha was demonstrated to be 488 expressed in the saccular epithelium and in the saccular afferent branches that were proximal to the saccular 489 epithelium (Sisneros et al. 2004a, Forlano et al. 2005). It was the discovery of the estrogen receptor alpha in the 490 midshipman saccule that was impetus to first determine in this study if related reproductive-state dependent 491 changes in auditory sensitivity occur at the level of the hair cell in female midshipman fish. Results from this 492 Sisneros, p.18 Ver. 2a 6/17/2009 study now show that seasonal changes in auditory tuning do occur at the level of the saccular hair cell and thus 493 will now be used to guide future studies that examine the potential role of T and E2 in modulating the response 494 properties of midshipman saccular hair cells. 495 The potential cellular mechanism(s) responsible for the steroid-dependent plasticity of peripheral 496 auditory frequency sensitivity in the plainfin midshipman is still unknown. One proposed mechanism for the 497 observed changes in midshipman is the action that steroid hormones may have on the ion-channel current 498 kinetics of auditory hair cells by genomically up regulating the differential expression ion channels such as 499 calcium-dependent BK or Kv channel types (and/or their related subunits), which then in turn could potentially 500 influence the biophysical properties of hair cells and their electrical resonance. Such a mechanism has been 501 posited for similar steroid-related changes in the frequency sensitivity of electroreceptors in weakly electric 502 fishes (Zakon 1987, Zakon et al. 1991). The electrical resonance that arises from ion-channel current kinetics of 503 the basolateral membrane of auditory hair cells is thought to be the major contributing factor that establishes 504 hair cell low frequency (< 1kHz) tuning in non-mammal auditory systems (Fettiplace and Fuch 1999) including 505 the toadfish, Opsanus tau (Steinacker and Romero 1991, 1992). Hair cell electrical resonance originates from 506 the interaction between inward calcium and outward Ca+-dependent K+ currents that produce an electrical 507 oscillation of the receptor potential along the hair cell epithelium (Lewis and Hudspeth 1983, Roberts et al. 508 1988). Perhaps the observed changes in auditory threshold in reproductive females is due changes in the 509 electrical tuning properties of hair cells that result in the electrical “detuning” of the saccular hair cells to the 510 effect that widens the restricted frequency response of hair cells observed in non-reproductive females (Fig. 3) 511 via changes in non-linear input-out functions of the hair cells. Thus, the characterization of the hair cell ion512 channel current kinetics in non-reproductive and reproductive females would help to elucidate the potential role 513 female reproductive plays in affecting the electrical tuning properties of saccular hair cell. 514 Alternative sites of action for steroid hormones and their effects on the response properties of the 515 midshipman auditory system include the auditory saccular afferents and hindbrain efferent nuclei. Gonadal 516 steroids may have direct effects on the saccular afferents that innervate hair cells in the saccule. Forlano et al. 517 (2005) shows evidence for aromatase-ir ganglion cells in the saccular afferent branches of the VIII auditory 518 nerve, which reveals potential sites for the conversion of T into E2 in areas of the midshipman auditory nerve 519 that are contiguous to hair cells in the saccular epithelium. Other steroid sensitive sites that may potentially 520 affect peripheral auditory processing include hindbrain efferent nuclei and the efferents that directly innervate 521 Sisneros, p.19 Ver. 2a 6/17/2009 the saccule in the midshipman inner ear (Bass et al. 1994). Efferents from the hindbrain nuclei provide inhibitory 522 input from the CNS to the auditory periphery and modulate the gain or sensitivity of hair cells in the saccule 523 (Furukawa and Matsura 1978). Work by Xiao and Suga (2002) more recently showed that neurons in the 524 mammalian auditory cortex modulate the frequency sensitivity of cochlear hair cells via olivocochlear efferents in 525 the corticofugal (descending) auditory system. Future studies that examine the potential cellular effects of 526 steroid hormones on auditory saccular afferent neurons and the potential seasonal reproductive-state and 527 steroid-dependent effects on the efferent activation and modulation of saccular hair cell sensitivity will be 528 instrumental in the understanding of how peripheral auditory sensitivity is modulated in the midshipman fish. 529 In summary, the results reported in this study indicate that seasonal plasticity of auditory sensitivity 530 previously reported in the female midshipman fish does occur at the level of the saccular hair cell. This novel 531 form of seasonal reproductive-state dependent plasticity of auditory saccular sensitivity provides an adaptive 532 mechanism that enhances the coupling between sender and receiver in this communication system and should 533 enhance the probability of mate detection and localization by female midshipman during the breeding season. 534 This adaptive plasticity of the female midshipman’s auditory system may act to increase the probability of mate 535 detection/localization and potentially enhance the acquisition of auditory information needed for mate choice 536 decisions during the breeding season. 537 538 539 540 Acknowledgements 541 I thank M. Marchaterre for the photograph use in Figure 1 and K-S Leon for field assistance in animal collections 542 and for her laboratory assistance in the collection of the initial pilot data for this study. I also thank the three 543 anonymous reviewers for their helpful comments on the manuscript, Drs James Ha, Brian Flaherty and Kevin 544 King for statistical consultation and assistance, and the Captain Charlie Eaton and his crew of the R/V Kittiwake 545 and the Captain Mark Anderson and his crew of the R/V Centennial from the Friday Harbor Laboratories for their 546 field assistance. This research was supported by a Royalty Research Fund grant from the University of 547 Washington. 548 549 550 Sisneros, p.20Ver. 2a 6/17/2009 REFERENCES551Bass AH. Shaping brain sexuality. Amer. Scientist 84: 352-363, 1996.552Bass AH, Clark CW. The physical acoustics of underwater sound communication. In: Simmons AM, Fay RR,553Popper A, eds. Springer handbook of auditory research, Springer, New York, pp.1-64, 2003.554Bass AH, Ladich F. Vocal-acoustic communication: From behavior to neurons. In: Fish Bioacoustics. (A.555 Popper, R. Fay, J. Webb, eds). Springer, New York, pages 253-278, 2008.556Bass AH, Marchaterre MA, Baker R. Vocal-acoustic pathways in a teleost fish. J. Neurosci. 14, 4025-4039,5571994.558Bodnar D, Bass AH. Temporal coding of concurrent acoustic signals in auditory midbrain. J Neurosci 17:7553-5597564, 1997.560Bodnar D, Bass AH. Midbrain combinatorial code for temporal and spectral information in concurrent acoustic561signals. J Neurophysiol 81:552-563, 1999.562Brantley RK, Bass AH. Alternative male spawning tactics and acoustic signals in the plainfin midshipman fish,563Porichthys notatus (Teleostei, Batrachoididae). Ethology 96: 213-232, 1994.564Cavanaugh GM. Formulae and methods VI of the Marine Biological Laboratory. Woods Hole, MA: The Marine565Biological Laboratory, 1956.566Cohen MJ, Winn HE. Electrophysiological observations on hearing and sound production in the fish, Porichthys567notatus. J Exp Zool 165:355-369, 1967.568Corwin JT. Postembryonic growth of the macula neglecta auditory detector in the ray, Raja clavata: continual569 increases in hair cell number, neural convergence and physiological sensitivity. J. Comp. Neurol. 217: 345-570356, 1983.571Edds-Walton PL, Fay RR. Directional sensitivity and frequency tuning of midbrain cells in the oyster toadfish,572Opsanus tau. J. Comp. Physiol. A 189:527-543, 2003.573Edds-Walton PL, Popper AN. Hair cell orientation patterns on the saccules of juvenile and adult toadfish,574Opsanus tau. Acta Zool. 76:257-265, 1995.575Fay RR. Sound Reception and processing in the carp: saccular potentials. Comp Biochem Physiol 49A:29-42,5761974.577Fay RR. Suppression and excitation in auditory nerve fibers of the goldfish, Carassius auratus. Hear Res 48:93-578110, 1990.579 Sisneros, p.21Ver. 2a 6/17/2009 Fay RR, Edds-Walton PL. Diversity in frequency response properties of saccular afferents of the toadfish,580Opsanus tau. Hear. Res. 113:235-246, 1997a.581Fay RR, Edds-Walton PL. Directional sensitivity of saccular afferents of the toadfish, Opsanus tau. Hear Res.582111:1-21, 1997b.583Fay RR, Popper AN. Acoustic stimulation of the ear of the goldfish (Carassius auratus). J Exp Biol 61:243-260,584 1974.585Fay RR, Popper AN. Structure and function in teleost auditory systems. In: Popper AN, Fay RR (eds)586Comparative studies of hearing in vertebrates. Springer, Berlin Heidelberg New York, pp 3-42, 1980.587Fettiplace R, Fuch PA . Mechanisms of hair cell tuning. Ann. Review Physiol. 61: 809-834, 1999.588Fine ML, Lenhardt ML. Shallow-water propagation of the toadfish mating call. Comp. Biochem. Physiol. A 76:589225-231, 1983.590Forlano PM, Deitcher DL, Bass AH. Distribution of estrogen receptor alpha mRNA in the brain and inner ear of a591vocal fish with comparisons to sites of aromatase expression. J. Comp. Neurol. 483: 91-113, 2005.592Furukawa T, Ishii Y. Neurophysiological studies on hearing in goldfish. J Neurophysiol 30:1377-1403, 1967.593Furukawa T, Matsura S. Adaptive rundown of excitatory postsynaptic potentials at synapses between hair cells594and eighth nerve fibers in goldfish. J. Physiol. (London) 276: 193-209, 1978.595Girden ER. ANOVA: repeated measures. Sage University Papers Series on Quantitative Applications in the596Social Sciences, 84. Thousand Oaks, CA: Sage, 1992.597Goense JBM, Feng AS. Seasonal changes in frequency tuning and temporal processing in single neurons in the598 frog auditory midbrain. J. Neurobiol. 65: 22-36, 2005.599Gordon NM, Gerhardt HC. Hormonal modulation of phonotaxis and advertisement-call preferences in the gray600treefrog (Hyla versicolor). Horm. Behav. 55: 121-127, 2009.601Guimaraes P, Frisina ST, Mapes F, Tadros SF, Frisina DR, Frisina RD. Progestin negatively affects hearing in602aged women. Proc.Nat. Acad. Sci. 103: 14246-14249, 2006.603Hama K. A study on the fine structure of the saccular macula of the goldfish. Z. Zellforsch Mikrosk Anat 94:155-604171, 1969.605Higgs DM, Souza MJ, Wilkins HR, Presson JC, Popper AN. Ageand size-related changes in the inner ear and606hearing ability of the adult zebrafish (Danio rerio). JARO 3: 174-184, 2002.607 Sisneros, p.22Ver. 2a 6/17/2009 Higgs DM, Rollo AK, Souza MJ, Popper AN. Development of form and function in peripheral auditory structures608of the zebrafish (Danio rerio). J. Acoust. Soc. Am. 113: 1145-1154, 2003.609Hultcrantz M, Simonoska R, Stenberg AE. Estrogen and hearing: a summary of recent investigations. Acta.610Otolaryngol. (Stockh.) 126: 10-14, 2006.611Iwashita A, Sakamoto M, Kojima T, Watanabe Y, Soeda H. Growth effects on the auditory threshold of Red Sea612 bream. Nippon Suisan Gakkaishi 65: 833-838, 1999.613Kenyon TN. Ontogenetic changes in the auditory sensitivity of damelfishes (pomacentridae). J. Comp. Physiol.614A 179: 553-561, 1996.615Lewis ER, Hudspeth AJ. Voltage-dependent and ion-dependent conductances in solitary vertebrate hair-cells.616Nature 304: 538-541, 1983.617Lin JW, Faber DS. An efferent inhibition of auditory afferents mediated by the goldfish Mauthner cell. Neurosci.61824: 829-836, 1988.619Lombarte A, Popper AN. Quantitative analyses of postembryonic hair cell addition in the otothic endorgans of620the inner ear of the European hake, Merluccius merluccius (Gadiformes, Teleostei). J. Comp. Neurol. 345:621419-428, 1994.622Lucas JR, Freeberg TM, Krishnan,A, Long G. A comparative study of avian auditory brainstem responses:623correlations with phylogeny and vocal complexity, and seasonal effects. J. Comp. Physiol. A 188: 981-992,6242002.625Lucas JR, Freeberg TM, Long GR, Krishnan A. Seasonal variation in avian auditory evoked responses to tones:626 a comparative analysis of Carolina chickadees, tufted titmice, and white-breasted nuthatches. J. Comp.627Physiol. A 193: 201-215, 2007.628McKibben JR, Bass AH. Behavioral assessment of acoustic parameters relevant to signal recognition and629preference in a vocal fish. J. Acoust. Soc. Amer. 104: 3520-3533, 1998.630McKibben JR, Bass AH. Peripheral encoding of behaviorially relevant acoustic signals in a vocal fish:single631tones. J Comp Physiol A 184:563-576, 1999.632McKibben JR, Bass AH. Peripheral encoding of behaviorally relevant acoustic signals in a vocal fish: harmonic633and beat stimuli. J. Comp. Physiol. A 187: 271-285, 2001.634Parvulescu A. The acoustics of small tanks. In: Tavolga WN (ed) Marine bioacoustics. Pergamon Press,635Oxford, pp 7-14, 1967.636 Sisneros, p.23Ver. 2a 6/17/2009 Penna M, Capranica RR, Somers J. Hormone-induced vocal behavior and midbrain auditory-senstivity in the637green treefrog, Hyla cinerea. J. Comp. Physiol. A. 170: 73-82, 1992.638Platt C. Hair cell distribution and orientation in goldfish otolith organs. J. Comp. Neurol. 172: 283-297, 1977.639Popper AN. The effects of fish size on auditory capacities of the goldfish. J. Aud. Res. 11: 239-247, 1971.640Popper AN, Hoxter B. Growth of a fish ear: 1. Quantitative analysis of sensory hair cell and ganglion cell641 proliferation. Hear Res. 15: 133-142, 1984.642Roberts WM, Howard J, Hudspeth AJ. Hair cells: transduction, tuning, and transmission in the inner ear. Ann.643Rev. Cell Biol. 4: 63-92, 1988.644Roger PH, Cox M. Underwater sound as a biological stimulus. In: Atema, J., Fay, R.R., Popper, A.N., Tavolga,645W.N. (eds), Sensory Biology of Aquatic Animals. New York: Springer-Verlag, pp. 130-149, 1988.646Roger PH, Popper AN, Cox M, Saidel WM. Processing of acoustic signals in the auditory system of bony fish. J.647Acoust Soc Am 83: 338-349, 1988.648Sento S, Furukawa T. Intra-axonal labeling of saccular afferents in the goldfish, Carassius auratus: correlations649between morphological and physiological characteristics. J. Comp. Neurol. 258: 352-367, 1987.650Sisneros JA. Saccular potentials of the vocal plainfin midshipman fish, Porichthys notatus. J. Comp. Physiol. A.651193: 413-424, 2007.652Sisneros JA. Adaptive hearing in the vocal plainfin midshipman fish: getting in tune for the breeding season and653implications for acoustic communication. Integr. Zool. 4:33-42, 2009.654Sisneros JA, Bass AH. Seasonal plasticity of peripheral auditory frequency sensitivity. J. Neurosci. 23: 1049-655 1058. 2003.656Sisneros JA, Forlano PM, Deitcher DL, Bass AH. Steroid-dependent auditory plasticity leads to adaptive657coupling of sender and receiver. Science 305: 404-407, 2004a.658Sisneros JA, Forlano PM, Knapp R, Bass AH. Seasonal variation of steroid hormone levels in an intertidal-659nesting fish, the vocal plainfin midshipman. Gen. Comp. Endocrinol. 136: 101-116, 2004b.660Sokal RR, Rohlf FJ. Biometry. New York: WH Freeman and Company, 1981, p. 242-245.661Steinacker A, Romero A. Characterization of voltage-gated calcium-activated potassium currents in toadfish662saccular hair cells. Brain Research 556: 22-32, 1991.663Steinacker A, Romero A. Voltage-gated potassium current and resonance in the toadfish saccular hair cells.664Brain Research 574: 229-236, 1992.665 Sisneros, p.24Ver. 2a 6/17/2009 Tomchik SM, Lu Z. Modulation of auditory signal-to-noise ratios by efferent stimulation. J. Neurophysiol. 95:6663562-3570, 2006.667Tomkins JL, Simmons LW. Measuring relative investment: a case study of testes investment in species with668alternative male reproductive tactics. Animal Behaviour 63: 1009-1016, 2002.669Weeg M, Bass AH. Midbrain lateral line circuitry in a vocalizing fish. J. Comp. Neurol. 418:841-864, 2000.670 Weeg M, Fay RR, Bass AH. Directionality and frequency tuning of primary saccular afferents of a vocal fish, the671plainfin midshipman (Porichthys notatus). J Comp Physiol A 188:631-641, 2002.672Xiao Z, Suga N. Modulation of cochlear hair cells by the auditory cortex in the mustached bat. Nat. Neurosci. 5:67357-63, 2002.674Zakon HH. Hormone-mediated plasticity in the electrosensory system of weakly electric fish. Trends Neurosc.67510: 416-421, 1987.676Zakon HH, Mills AC, Ferrari MB. Androgen-dependent modulation of the electrosensory and electromotor677systems of a weakly electric fish. Sem. Neurosci. 3, 449-457, 1991.678Zar JH. Biostatistical Analysis, 4 edition, New Jersey: Prentice Hall, 1999, p.98-104.679Zotterman Y. The microphonic effect of teleost labyrinths and its biological significance. J Physiol 102: 313-318,6801943.681682683684685686687688689690691692693694 Sisneros, p.25Ver. 2a 6/17/2009 Figure Legends695696Figure 1. Dorsal view of the brain and inner ear of the plainfin midshipman fish, Porichthys notatus. The inner697ear saccule (S) is marked by a dotted line that defines the border and indicates the region of the saccule698(rostral, middle and caudal) that was used to record the evoked saccular potentials. Note that in this photo the699 saccule has been deflected laterally to expose the dorsal view of the nerves. C cerebellum, M midbrain, T700telencephalon.701702Figure 2. Representative examples of iso-level curves of the evoked saccular potentials recorded from the703 rostral, middle and caudal regions of the saccule in response to single tones at 130 dB (re 1 μPa) for non-704 reproductive and reproductive female midshipman. Note that the scale of the y-axis for the representative705evoked saccular potentials in the plots are different to emphasis the overall relative shape of the iso-level706response curves. All saccular potentials are plotted as mean + 1 SD and that most of the SD bars are obscured707by the symbols.708709Figure 3. The evoked potentials recorded from the saccule of non-reproductive and reproductive female710 midshipman fish based on the response to iso-level tones of 130 dB (re 1μPa). Peak frequency sensitivity of 75711 Hz was the lowest frequency tested. In order to control for absolute sensitivity of the saccule from different712recording positions and compare across different animals, the iso-level response data were normalized to a713relative valve of 0 dB assigned to the peak response for each record and then expressed in relative dB re Best714Frequency Sensitivity. All data are plotted as mean + 1 SD. Note the number of animals and records per group715(reproductive vs. non-reproductive) are indicated in parentheses.716717Figure 4. Iso-level response curves of the evoked saccular potentials recorded from the rostral, middle and718 caudal regions of the saccule in response to single tones at 130 dB (re 1 μPa) for non-reproductive (NR) and719 reproductive (R) female midshipman. Iso-intensity response data were normalized to a relative valve of 0 dB720assigned to the peak response for each record and then expressed in relative dB re Best Frequency Sensitivity.721All data are plotted as mean + 1 SD. Note the number animals and records are indicated in parentheses.722723 Sisneros, p.26Ver. 2a 6/17/2009 Figure 5. Representative examples of individual auditory threshold tuning curves for non-reproductive and724reproductive females based on the evoked potentials recorded from the rostral, middle and caudal regions of the725midshipman saccule. The auditory threshold at each stimulus frequency was determined as the lowest stimulus726 intensity in dB (re 1 μPa) that evoked a saccular potential that was at least 2 SD above the background noise727 measurement.728729Figure 6. Auditory threshold tuning curves for non-reproductive and reproductive female midshipman based on730the evoked potentials recorded from the saccule. All data are plotted as mean + 95% confidence limit (CL) and731the number of animals and records are indicated in parentheses. Auditory threshold at each stimulus frequency732 was determined as the lowest stimulus intensity in dB (re 1 μPa) that evoked a saccular potential that was at733 least 2 SD above the background noise measurement.734735Figure 7. Auditory threshold tuning curves for non-reproductive and reproductive female fish based on the736evoked potentials recorded from the rostral, middle and caudal regions of the midshipman saccule. All data are737plotted as mean + 95% CL and the number of animals and records are indicated in parentheses. Auditory738 threshold at each stimulus frequency was determined as the lowest stimulus intensity in dB (re 1 μPa) that739 evoked a saccular potential that was at least 2 SD above the background noise measurement.740741Figure 8. Comparison between the vocal characteristics of the male advertisement call and the change in742auditory threshold saccular tuning in female midshipman fish based on seasonal reproductive state. Shown here743is a combined plot of the difference in auditory saccular thresholds between non-reproductive and reproductive744females at each frequency tests (Δ Threshold, left y axis) and the power spectrum of the male midshipman745advertisement call showing the first 4 harmonics (right y axis, in relative dB values); inset show the temporal746 waveform of the advertisement call from a nesting male midshipman fish recorded at 16° C at a nest site (scale747 bar = 50 ms).748