Research Articles, Systems/Circuits

Auditory Competition and Stimulus Selection across Spatial Locations from Midbrain to Forebrain in Barn Owls

Andrea J. Bae, Brian J. Fischer and José L. Peña
Journal of Neuroscience 11 December 2024, 44 (50) e1298242024; https://doi.org/10.1523/JNEUROSCI.1298-24.2024

Abstract

Barn owls enable investigation of neural mechanisms underlying stimulus selection of concurrent stimuli. The audiovisual space map in the optic tectum (OT), avian homolog of the superior colliculus, encodes relative strength of concurrent auditory stimuli through spike response rate and interneuronal spike train synchrony (STS). Open questions remain regarding stimulus selection in downstream forebrain regions lacking topographic coding of auditory space, including the functional consequences of interneuronal STS on interregional signaling. To this end, we presented concurrent stimuli at different locations and manipulated relative strength while simultaneously recording neural responses from OT and its downstream thalamic target, nucleus rotundus (nRt), in awake barn owls of both sexes. Results demonstrated that broadly spatially tuned nRt units exhibit different spike response patterns to competition depending on spatial tuning preferences. Modeling suggests nRt units integrate convergent inputs from distant locations across midbrain map regions. Additionally, STS within nRt reflects the temporal properties of the strongest stimulus. Furthermore, interregional STS between OT and nRt was strongest when spatial tuning overlap between units across regions was large and when the strongest stimulus location during competition was favorable for units in both regions. Additionally, though gamma oscillations synthesized within OT are weakly propagated within nRt, average gamma power across regions correlates with strength of interregional STS. Overall, we demonstrate that nRt integrates inputs across distant areas of OT, retains spatial information through differences in strength of inputs from various locations of the midbrain map across neurons, and prioritizes coding of identity features to the strongest sound.

Significance Statement

The brain strategically selects and preferentially processes salient stimuli. A critical function to this process involves transferring salient information across regions that may exhibit drastic transformations in coding schemes. Our study in barn owls investigates bottom-up signaling between the midbrain space map and its downstream thalamic target, which lacks spatial topography as also observed in mammalian auditory forebrain regions to elucidate general mechanisms underlying how spatial location information and other properties of the strongest sound are relayed between regions. Results show that the thalamus integrates neural responses widely across the midbrain map, retains coding of spatial location through varying strength of inputs of the map across neurons, and prioritizes further coding of identity features only to the strongest sound.

Introduction

The natural auditory scene is inundated with concurrent sounds, yet the brain is capable of segregating and prioritizing its computational resources toward salient sounds (Cherry, 1953; Bregman, 1990). As sound localization specialists, barn owls are an excellent model organism to investigate the neural mechanisms underlying stimulus selection because they localize the sound of prey among the many sounds in their environment (Payne, 1971).

Barn owls utilize binaural cues, interaural time difference (ITD) and interaural level difference (ILD), to infer locations of sounds in azimuth and elevation, respectively (Knudsen and Konishi, 1980; Moiseff and Konishi, 1981). Their localization abilities have been attributed to their midbrain maps of auditory space (Knudsen and Konishi, 1978; Knudsen, 1982) which represent sound source directions through topographically organized neurons tuned to combinations of binaural cues (Moiseff, 1989). The optic tectum (OT), avian homolog of the mammalian superior colliculus, contains an audiovisual space map and generates motor signals leading to orienting movements (Knudsen, 1982; du Lac and Knudsen, 1990; Masino and Knudsen, 1992). The OT, together with the isthmic nuclei, forms the midbrain stimulus selection network that is devoted to prioritizing the strongest stimulus among concurrent sounds (G. Marín et al., 2007; Mysore et al., 2010, 2011). Coordinated activity of OT and three isthmic nuclei enhances representation of the strongest source (G. Marín et al., 20071; Asadollahi et al., 2010; G. J. Marín et al., 2012) and suppression of responses for all other areas of the map (Mysore et al., 2010, 2011; Mysore and Knudsen, 2011). Our recent work demonstrated gamma oscillations in OT modulate interneuronal spike train synchrony (STS) to encode the relative strength of concurrent stimuli (Bae et al., 2024). However, functional consequences of OT interneuronal synchrony in interregional signaling remains unknown.

Significant open questions also remain regarding how OT transfers spatial information to its downstream forebrain thalamic target. Anatomical studies showed that output from OT is relayed along the tectofugal pathway (Benowitz and Karten, 1976; Straight et al., 2024). A thalamic region called the nucleus rotundus (nRt), comparable with the mammalian pulvinar (Shimizu and Bowers, 1999), receives direct excitatory and indirect inhibitory projections from OT (Karten and Hodos, 1970; Mpodozis et al., 1996; Hellmann and Güntürkün, 2001; Hu et al., 2003) and thus reads out OT's population activity that selects for the strongest stimulus. Electrophysiological studies have shown that nRt and its downstream region, entopallium, are multimodal in barn owls and exhibit broad tuning to ITDs corresponding to sounds at contralateral locations (Reches and Gutfreund, 2009). Critically, unlike OT, nRt exhibits functional organization based on visual features (Y. C. Wang et al., 1993). Previous studies have described a transformation in coding location from narrowly tuned and topographically organized neurons in OT to broadly tuned and nontopographically organized neurons in nRt (G. Marín et al., 2003; Hellmann et al., 2004); however, how location and other properties of the strongest sound selected by OT from concurrent sounds is conveyed and encoded in nRt is unknown. Investigating interregional communication in the context of competition may elucidate how spatial information is transferred between regions and uncover functional implications that result from convergence across regions.

To this end, we implanted chronic electrodes in OT and nRt and conducted dual region recordings during sound competition in awake owls. Our results show spike response patterns in broadly spatially tuned nRt units to auditory competition vary depending on spatial tuning selectivity. Modeling demonstrates that different response patterns result from nRt integrating opposing trends in spike responses across distant areas of the OT space map. Additionally, STS of nRt units reflects the temporal properties of the strongest stimulus. Lastly, interregional STS between OT and nRt is strongest when unit pairs across regions exhibit high spatial tuning overlap and when the location of the strongest stimulus during competition is preferred by units in both regions. Furthermore, while gamma oscillations in OT weakly propagate within nRt, average gamma power across regions correlates with strength of interregional STS.

Materials and Methods

Protocols utilized in this study were in compliance with guidelines set by the National Institutes of Health, Albert Einstein College of Medicine’s Institute for Animal Studies, and the Institutional Animal Care and Use Committee (IACUC) of the Albert Einstein College of Medicine.

Animal handling and surgery

Four adult North American barn owls (Tyto furcata) of both sexes (3 females, 1 male) were used in this study. For all animals, custom-built steel head plates were affixed to the skull with dental acrylic to enable head fixation during stereotaxic surgery and recordings. Prior to implanting electrode drives, owls were habituated to remain head fixed inside soundproof chambers over increasing amounts of time (up to 1.5 h) while undergoing sound stimulation. To enable awake recordings, chronic drives were implanted surgically. For these surgeries, owls were food deprived 12 h prior to surgery, and on the day of surgery, owls were anesthetized with intramuscular injections of ketamine (Ketaset: 20 mg/kg) and xylazine (Anased: 2 mg/kg). Prophylactic antibiotic (ampicillin: 20 mg/kg, i.m.) and lactated Ringer's solution (10 ml, s.c.) were administered to prevent infection and maintain hydration during surgery. Body temperature was maintained with a heating pad. Proper anesthetic level was monitored throughout surgery by checking for pedal and eyelid reflexes. To maintain anesthesia, doses of ketamine and xylazine were administered every 1.5–2 h. After confirmation of each target brain region, chronic drives were affixed to the skull using dental acrylic. Upon completion of the implant surgery, owls were given analgesic (Rimadyl: 3 mg/kg, i.m.) to minimize inflammation and pain, and animals were kept in a warm, quiet environment during recovery. Owls were returned to the main aviary once they were alert and capable of standing. Owls recovered for 5 d before awake recordings were initiated. During awake recordings, owls were monitored through an infrared camera, and recordings were terminated when the owl exhibited signs of stress.

Sound stimulation

All recordings were performed in a double wall sound-attenuated chamber (Industrial Acoustics). The inner walls of the chamber were lined with acoustical foam to minimize echoes (Sonex). Tucker-Davis Technologies System 3 and custom written Python software were used to synthesize and deliver acoustic stimuli. Stimuli were presented through a free-field speaker array, as described in previous studies (Y. Wang et al., 2012; Beckert et al., 2017, 2020; Bae et al., 2024). The custom-built hemispherical free-field speaker array consisted of 144 speakers (Sennheiser, 3P127A) arranged to surround a stereotaxic device that was positioned such that the head-fixed owl faced 0° azimuth and 0° elevation. Speaker locations ranged between ±100° azimuth and ±80° elevation. The angular separation between speakers varied from 10° to 30° with the highest density of speakers located in frontal space at the center of the array. Speakers were calibrated using a Brüel and Kjær microphone (model 4190) positioned at the center of the array.

To measure spatial receptive fields of recorded units, broadband noise stimuli were presented from each of the 144 speakers, five times in random order (100 ms duration, 5 ms rise–fall times, 1.0 s interstimulus intervals). Preferred and nonpreferred locations were selected, respectively, by maximum and minimum response rates to sound stimulation from different directions and responses to preferred and nonpreferred locations were further characterized by measuring responses to changing stimulus levels at each of these locations. Either flat (not amplitude modulated) broadband noise or sinusoidally amplitude-modulated (modulation frequency, fAM 55 Hz or 75 Hz) broadband noise stimuli were presented at levels ranging from 27 to 72 decibels sound pressure level (dB SPL) in increments of 5 dB (1.0 s duration, 5 ms rise–fall times, 2.5 s interstimulus intervals, 20 repetitions per level in random order).

For competing sound paradigms, stimuli consisted of two sounds (1.0 s duration, 5 ms rise–fall times, 2.5 s interstimulus intervals, 20 repetitions per condition in random order) either two de novo uncorrelated flat broadband noise stimuli or two sinusoidally amplitude-modulated uncorrelated broadband noise stimuli with different modulation frequencies (either fAM 55 or fAM 75 Hz). Competing stimuli were presented simultaneously from two speakers at different azimuths and the same elevation. For each recording session, the preferred azimuth and elevation of all units were determined from the spatial receptive fields. Given that spatial tuning in the midbrain is more spatially circumscribable and the effects of competition in the midbrain are well characterized, the driver and competitor stimuli for units in both midbrain and forebrain were based on the spatial tuning of midbrain units. One stimulus, designated to be the driver, was presented from a speaker located within the receptive field of simultaneously recorded units (preferred location). The second stimulus, designated to be the competitor, was presented from a nonpreferred location relative to recorded midbrain units at least 50° away and within the same hemisphere, contralateral to the recorded midbrain hemisphere. For competition in the forebrain, driver and competitor stimuli remained fixed and varied in level across trials, respectively. However, because spatial tuning in the forebrain was broad, each unit's preferred location was determined from the tuning selectivity index. We focused the investigation on competition along azimuth because midbrain space map neurons exhibit ellipsoidal tuning curves that are wider in elevation than azimuth, consistent with twofold better discrimination in azimuth than elevation (Knudsen and Konishi, 1978; Bala et al., 2007). Thus, we could use a single elevation to evoke responses across simultaneously recorded units. For midbrain OT, we ensured that the distant competitor stimulus alone did not evoke significant responses by measuring spike response rates to increasing stimulus levels, and units with spike response rates that exceeded the baseline activity by more than 2 standard deviations (SD) in response to the competitor were excluded. Similarly, OT units that exhibited spike response rates exceeding 2 SD above baseline activity to the driver alone were included. However, because forebrain nRt units exhibit broad contralateral tuning, we included all units that exhibited significant response rates that exceeded 2 SD above the baseline for either the driver or competitor alone. The modulation frequencies 55 and 75 Hz for amplitude-modulated stimuli were chosen to match a previous study investigating the representation of concurrent sound sources in the external nucleus of the inferior colliculus (ICx; Keller and Takahashi, 2005). Amplitude-modulated stimuli were constructed by multiplying broadband noise with a sinusoidal envelope (100% modulation depth, meaning the final amplitude ranged between zero and the same amplitude of the flat broadband noise). Driver stimuli were presented at a fixed level in each recording block (flat noise, 47 dB SPL; amplitude-modulated noise, 43 dB SPL). Competitor levels varied across trials, ranging from 15 dB below to 10 dB above the level of the driver (relative levels −15 to +10 dB, increments of 5 dB). Both driver and competitor stimuli were presented at levels above threshold to ensure stimuli evoked sufficient neuronal responses.

Electrophysiology

Two chronic Neuronexus microdrives (d-Drive), each loaded with a single linear 16-channel probe (H16), were implanted in OT and nRt to record multiunit activity and local field potentials (LFPs) in awake owls. Recording sites along the 16-channel probe span 1.5 mm with 100 µm spacing between sites. Microdrives allowed electrodes to advance further up to 1.0 mm along the dorsoventral axis from the initial site of implantation. Proper placement of electrodes in OT and nRt was guided by stereotaxic coordinates and characteristic electrophysiological properties of neurons in each region. In OT, neurons exhibit unambiguous tuning to binaural cues and are spontaneously bursty, and multimodal, responding to both visual and auditory stimuli (Knudsen, 1982, 1984). In nRt, neurons exhibit broad, contralateral tuning to binaural cues and are spontaneously bursty, and multimodal like OT (Reches and Gutfreund, 2009). After each awake recording block, which spanned over 4 d, electrodes were advanced in 300 µm increments in search of new units. Recording signals were amplified, digitized, and stored using Tucker-Davis Technologies System 3 and custom written Python code. Similar to previous electrophysiology studies in OT and nRt reporting multiunit activity in recordings of these regions (Reches and Gutfreund, 2009), the characteristic spontaneous bursty activity of these regions made it difficult to isolate and distinguish bursty single unit waveforms from multiunit activity, and as such spike times of multiunit activity were extracted offline by thresholding voltage traces of each channel semimanually per recording session, blinded to which spikes were stimulus driven or spontaneous.

Data analysis

For all sound paradigms, we calculated the spike response rate from the mean spike count evoked by each stimulus within the stimulus presentation window across either five repetitions for spatial tuning or 20 repetitions for each stimulus level.

Two-sound competition

Relative changes in spike response rate to competing stimuli were calculated as the percent change in response to two concurrent stimuli relative to responses measured to a single stimulus at the position and level of the driver stimulus, similar to measures previously reported (Mysore et al., 2010, 2011; Mysore and Knudsen, 2011):ChangeinResponse=Rate(driver+competitor)−Rate(driver)Rate(driver)*100%. Responses to single stimuli at the driver position recorded during characterization were used to determine basic neuronal responses without sound competition.

Correlation analysis

To examine synchrony of temporal spiking patterns between nearby units, we computed cross-correlograms (CCGs) of spike trains for simultaneously recorded unit pairs (Bair et al., 2001; Kohn and Smith, 2005; Smith and Kohn, 2008; Beckert et al., 2017, 2020). Discrete spike trains were converted to binary sequences indicating the presence of spikes in time by constructing peristimulus time histograms (PSTHs) for each trial from 50 to 1,000 ms poststimulus onset in 1 ms bins. The spike times occurring during the onset response between 0 and 50 ms were excluded. The trial-averaged CCGs were smoothed using a rectangular 10 ms sliding window and normalized by the geometric mean spike response rate of neuron pairs within the analysis window to minimize the number of coincidences detected due to chance alone as would occur with increased spike response rates. To minimize changes in CCG magnitudes related to slow fluctuations and stimulus locked correlations, we computed the trial-averaged shifted CCG by first computing the shifted CCG for nonsimultaneous trials (Kohn and Smith, 2005). To obtain the shifted CCG, the spike train from one trial for a single neuron was cross-correlated with the spike train from the subsequent trial for the unit in the pair. The trial-averaged shifted CCG was then smoothed and subtracted from the original CCG to calculate the shift-corrected CCG. From this shift-corrected CCG, the noise level and standard deviation observed in each cross-correlogram at far time lags ±950 to ±900 ms was measured. STS was defined as the maximal value observed at lags within ±15 ms around zero minus the noise level. Peaks that did not exceed 5 standard deviations of the noise level were excluded from further analysis.

To average synchrony values across recording sessions, we determined the changes in STS induced by competition relative to STS observed for single stimuli and calculate the competition synchrony index (CSI):CSI=STS(driver+competitor)−STS(driver)STS(driver+competitor)+STS(driver).

Modeling OT and nRt responses to competition

We used a model to predict how OT responses at distant driver and competitor locations were integrated in nRt to match the observed nRt responses to competition. Assuming that nRt neurons integrate input from across the OT map (Hellmann and Güntürkün, 2001; G. Marín et al., 2003; Hellmann et al., 2004; Reches and Gutfreund, 2009), nRt responses will be driven by combinations of the driver level (DL) and competitor level (CL) , with relative level (RL) :RL=CL−DL. We modeled nRt responses r(DL,CL) to competing auditory stimuli as a linear summation of OT responses at the driver rD(DL)gD(RL) and competitor rC(CL)gC(RL) positions as follows:r(DL,CL)=rD(DL)gD(RL)+rC(CL)gC(RL). In this model, rD(DL) and rC(CL) represent responses of nRt neurons driven by the driver level and competitor level alone, respectively. The gain terms gD(RL) and gC(RL) model how OT responses change with relative level when competing stimuli are present simultaneously (Mysore et al., 2010, 2011; Mysore and Knudsen, 2011; Bae et al., 2024).

We used the model to estimate the gain change on OT responses required to produce the different types of observed nRt responses to the relative level of competing auditory stimuli. For each nRt response type observed with competition, the given response to the driver alone rD(DL) and competitor alone rC(CL) were taken from rate level curves observed in Figure 3B,C. The gain change observed with competition at OT's driver position gD(RL) was assumed to have a sigmoidal dependence on the relative level, as observed experimentally (Mysore and Knudsen, 2011; Fig. 1). We then computed the expected gain change of OT responses at the competitor position required to produce the observed nRt response as follows:gC(RL)=r(DL,CL)−rD(DL)gD(RL)rC(CL).

Figure 1.
Figure 1.

Concurrent auditory competition and regions of interest for simultaneous dual region recordings. A, Experimental design: The driver stimulus (blue) was played from a speaker located at the preferred location of recorded OT units at a fixed level. The competitor stimulus (green) was played from a speaker outside OT's receptive field >50° away from the driver speaker. Sound level of the competitor varied from −15 to 10 dB relative to the driver level. Chronic 16-channel linear electrodes were implanted in OT and a downstream forebrain thalamic region, nRt. B, C, Spatial receptive fields of example OT (B) and nRt (C) units that were simultaneously recorded. Color intensity in each circle represents the spike response rate measured for stimuli played from that speaker location.

nRt responses were given by observed from data for the different competition response types (Fig. 4D).

Measuring nRt tuning index and spatial tuning correlations

The tuning selectivity index of nRt units was calculated from responses observed to single sounds at either a frontal or lateral driver position as below, where values between 0 and 1 indicate stronger selectivity to the frontal location and values between −1 and 0 indicate stronger selectivity to the lateral position:TuningSelectivityIndex=ResponseFrontal−ResponseLateralResponseFrontal+ResponseLateral. Additionally, spatial tuning correlation between simultaneously recorded OT and nRt units were measured as Pearson’s correlation for responses observed at speaker locations across all 144 speakers in the free-field array.

LFP analyses

The LFP was derived from the wideband signal by first removing 60 Hz electrical line noise and its harmonic (180 Hz), using an infinite impulse response (IIR) notch filter (scipy.signal.iirnotch). Bleed-through effects from spiking activity were minimized by the following procedure adopted from previous studies (Sridharan et al., 2011). Briefly, the denoised signal was bandpass filtered between 300 Hz and 4.7 kHz (Butterworth filter, scipy.signal.butter, order 2, run forward-backward with scipy.signal.filtfilt) and subtracted from the original signal to get the remaining low-frequency signal. From the remaining low-frequency signal, 2 ms windows at each spike time were linearly interpolated. This spike-corrected remnant low-frequency signal was then added back to the original signal to arrive at the spike-removed signal. This signal was low-pass filtered at 200 Hz (Butterworth, as above) and resampled to 1 kHz (scipy.signal.resample_poly).

We conducted analyses on induced LFPs by subtracting the mean evoked response from the LFP of each trial. The induced response was filtered between 20–50 Hz (Butterworth, order 1), which is the low gamma range that corresponds to input from the isthmi pars parvocellularis (Ipc) nucleus of owl's midbrain stimulus selection network (Asadollahi et al., 2010), as well as the high gamma range (50–75 Hz) informed from the spike field coherence analyses. Gamma power was estimated as the root-mean-squared (RMS) magnitude of the analytic signal, calculated from the Hilbert transform of the bandpass filtered signal. Power during stimulus presentation was measured 50 ms after stimulus onset to the end of stimulus duration, and spontaneous baseline power was measured for an equivalent time without stimulation. The stimulus-related change in gamma power was estimated as follows:InducedγPower[dB]=20*logRMS(stimulus)RMS(spontaneous). Effects of competition on gamma power were expressed by subtracting the induced gamma power for a single stimulus from the induced gamma power measured during competition. Phases at which spikes occurred relative to either the low gamma (20–50 Hz) or high gamma (50–75 Hz) LFP were estimated by aligning spike times to the instantaneous phases of the analytic signal, calculated from the Hilbert transform of the bandpass filtered LFP. Spike phases were then binned in π/9 radian increments between −π and +π.

Spike field coherence

The coherence Cxy between spikes and the LFP was estimated as the cross-spectra between signals Sxy(f) over the geometric mean of the autospectra of each signal Sxx(f) and Syy(f) as follows:Cxy=Sxy(f)Sxx(f)Syy(f). The LFP between 50 and 1,000 ms poststimulus onset was considered a continuous signal and spikes were considered discrete signals, binned in 1 ms bins between 50 and 1,000 ms poststimulus onset. Because Cxy is described by complex coordinates, the magnitude squared coherence (scipy.signal.coherence) represents a value of coherence between 0 and 1, with 0 meaning the two signals were not coherent and 1 meaning the two signals are the same (Lepage et al., 2011).

Statistical analyses

To compare various measures across single stimulus levels and competing relative levels, we conducted one-way ANOVA and subsequent post hoc Tukey’s tests to distinguish significant differences across groups. To test for significant phase locking of spikes to amplitude-modulated stimuli, we performed a Rayleigh's test of uniformity on the distribution of spike-phase values for single units across trials or the distribution of mean spike-phases across units.

Code accessibility

Python code used for data analysis is available at https://github.com/penalab/Bae-et-al-2024 or available upon request.

Results

To investigate the consequences of auditory competition in OT on downstream forebrain readout, we implanted two chronic 16-channel linear electrodes in OT and nRt for simultaneous dual region recordings of multiunit activity and LFPs in each region (Fig. 1A). To investigate auditory competition, we manipulated the relative strength of concurrent driver and competitor stimuli by varying the level of the competitor stimulus while maintaining the level of the driver stimulus fixed. Dual region recordings of neural responses were collected from four awake owls (three females, one male) and consisted of 125 OT units and 115 nRt units across 11 recording sessions for flat noise competition. If the recording function of implanted electrodes in a region was compromised at any time, we continued to record from the noncompromised single region as long as possible. These data from single regions were utilized for within-area analyses and comprised 229 OT units across 21 recording sessions and 143 nRt units across 16 recording sessions.

Representative spatial receptive fields from simultaneously recorded example units in OT (Fig. 1B) and nRt (Fig. 1C) illustrate the disparity in spatial tuning properties observed between the midbrain space map and forebrain thalamus. While most midbrain space map units exhibited focal spatial tuning, nRt units exhibited broad spatial tuning, consistent with previous reports of ITD tuning properties (Reches and Gutfreund, 2009), the large visual receptive fields reported in pigeons (Y. C. Wang et al., 1993), and the expectations based on anatomical evidence of convergence across widely distributed areas of the OT space map in nRt (G. Marín et al., 2003; Hellmann et al., 2004). Despite broad spatial tuning, most nRt units exhibited strong representation of frontal azimuths (Fig. 2A,B) and elevations (Fig. 2C,D), suggesting that the overrepresentation of frontal space in midbrain (Knudsen, 1982) may also have consequences for inputs transferred to nRt. Additionally, frontal representation was broader in elevation than in azimuth, reminiscent of the ellipsoidal spatial tuning observed in midbrain space map neurons where elevation tuning is broader than azimuth tuning (Knudsen and Konishi, 1978; Bala et al., 2007).

Figure 2.
Figure 2.

Despite broad spatial tuning, most nRt units represent frontal locations robustly. A, Heat map of azimuth tuning for all nRt units computed by taking the maximum normalized spike response at each azimuth across all elevations· B, Average azimuth tuning curve across the population. Positive and negative azimuths indicate contralateral and ipsilateral space respectively. C, Heat map of elevation tuning for all nRt units computed by taking the maximum normalized spike response at each elevation across all azimuths. D, Average elevation tuning curve across the population. For each unit, the spike response in each trial was normalized by the maximum spike response observed across all trials. Units on the y-axis of heat maps are sorted chronologically according to recording site date.

Changes in spike response rate and synchrony with flat noise competition in nRt

We first examined how nRt responses changed as the relative level of concurrent stimuli varied incrementally during auditory competition by measuring changes in spike response rate and STS (Fig. 3). For these experiments, we focused our investigation on the downstream consequences of competition in OT on nRt by presenting driver and competitor concurrent stimuli appropriate to the spatial tuning properties of simultaneously recorded midbrain units. Unlike midbrain OT where spike response rates and STS decreased as a distant competitor level grew in strength (Bae et al., 2024), nRt spike response rates significantly increased as the distant competitor grew in strength (Fig. 3A; F(1,5) = 11.08; p = 2.22 × 10−10; ANOVA). Raster plots showing spike times for two example nRt units show no changes in the number of coincident spikes observed when the driver is stronger compared with when the competitor is stronger (Fig. 3B). Consistent with the example raster plots, average cross-correlograms for all nRt unit pairs across relative level conditions did not show differences in peak heights (Fig. 3C) and the CSI across the population was not significantly impacted by competition (Fig. 3D; F(1,5) = 1.50; p = 0.187; ANOVA).

Figure 3.
Figure 3.

Spike response rate increases and STS does not change in nRt with competition of flat noise stimuli. A, Percent change in spike response rates for each relative level during competition. B, Example raster plots for two relative level conditions when the driver was stronger (−15 dB) and when the competitor was stronger (+10 dB). Spike times of two simultaneously recorded units (green and orange) with coincident spikes between the two units indicated as (black) ticks. C, Cross-correlograms from competing sounds, averaged across all simultaneously recorded unit pairs for each relative level condition (same order as in D). The number of unit pairs with significant STS is indicated above each correlogram. The horizontal colored line indicates the peak height of the average cross-correlogram computed from when a driver stimulus was presented alone. D, Population competition synchrony index (CSI) measuring the change in STS with competition relative to presentation of a single driver stimulus, across relative levels. Box plots show the interquartile range (25th–75th percentile) with whiskers indicating 5–95% confidence interval, and circles represent individual units. ANOVA p value and relevant post hoc p values reported for each pairwise comparison.

Because spatial tuning in nRt is broad and heterogeneous, we further investigated whether spatial tuning selectivity influenced the impact of competition observed on nRt spike responses. To assess tuning selectivity, a tuning selectivity index was computed from the difference of each unit's response to frontal and lateral stimuli alone, divided by the sum of the responses, such that positive index values indicated that spike response rates to frontal stimuli were greater than spike response rates for lateral stimuli and vice versa for negative values below zero (see Materials and Methods). From the tuning selectivity index distribution, we categorized units exhibiting selectivity index values above and below zero as frontal driver-selective and lateral driver-selective units, respectively. Overall, the distribution of tuning selectivity in nRt units centered around zero with values skewed positively, indicating that most nRt units in our population preferred frontal locations (Fig. 4A; mean ± SD: 0.07 ± 0.16). Average binaural level (ABL) curves for frontal driver-selective units (Fig. 4B) showed that these nRt units could respond to stimuli at both frontal and lateral locations but exhibited higher spike response rates for their preferred frontal location. Furthermore, our results revealed that nRt spike response patterns to competition depended on their tuning selectivity and whether the varying level competitor stimulus resided in the unit's preferred location (Fig. 4C). For most frontal driver-selective nRt units, frontal driver stimuli paired with lateral competitor stimuli led to invariant spike response rates with competition (Fig. 4C, yellow curve; F(1,5) = 0.54; p = 0.742). In contrast, lateral driver stimuli paired with frontal competitor stimuli for frontal driver-selective nRt units led to increasing spike response rates with competition (Fig. 4C, blue curve; F(1,5) = 15.70; p = 3.38 × 10−14). Altogether, these results suggest that the impact of competition by two concurrent stimulus locations is dependent on the tuning selectivity of individual nRt units, and the tuning selectivity index may reflect the strength of inputs from various parts of the midbrain map.

Modeling integration of OT inputs during competition matches observed diversity in nRt responses

We further investigated how different response types in nRt could result during competition by modeling nRt neurons integrating OT responses across distant areas of the map. We hypothesized that variations in response types to competition in nRt could result from simple linear summation of opposing changes in response to increasing competitor levels at different positions in OT. Here we assume that nRt units integrate inputs from neurons across OT's space map, which include distant driver and competitor positions. Given the observed changes in OT at preferred driver positions (Bae et al., 2024), we modeled the expected changes needed at OT's competitor position to give rise to the different response types in nRt. Overall, modeling OT responses predicts opposite changes in gain at the competitor position (Fig. 5). Symmetric and opposing changes in OT responses of neurons tuned to the driver and competitor position (Fig. 5A) result in invariant responses to competition in nRT (Fig. 5B). However, increasing responses to competition in nRt result when the opposing gain changes at the competitor position in OT increase asymmetrically relative to the decreasing gain changes at the driver position, such that input from OT neurons representing the competitor are stronger than input from OT neurons representing the driver stimulus (Fig. 5C,D). This modeling suggests that nRt's response to competition is largely inherited from integrating response patterns occurring across the space map in OT and that variations in nRt responses to competition may depend on the relative contribution of inputs from different locations of the space map.

Figure 4.
Figure 4.

nRt units exhibit variations in their response patterns to competition depending on whether the varying competitor stimulus resides in their preferred location. A, Histogram of tuning selectivity indices measured for each unit across the nRt population based on responses to frontal and lateral stimuli alone. Values above and below zero (dashed line) indicate preference toward frontal and lateral driver locations, respectively. B, Normalized response rates to average binaural level (ABL) for single stimuli presented at either a frontal driver location (orange) or lateral driver location (blue) curves for frontal driver-selective units. The range of relative levels at which concurrent sounds were presented is highlighted in pink. C, Percent change in spike response rates for each relative level during competition for frontal driver-selective units when presented either frontal driver/lateral competitor pairs (yellow curve) or lateral driver/frontal competitor pairs (blue curve). ANOVA p value for each tuning selectivity category indicated in legend. Error bars indicate standard error of mean.

Figure 5.
Figure 5.

Modeling OT responses across the map during competition predicts that nRt linearly sums OT inputs from distant locations of the map. A, B, Modeling symmetric and opposing OT response changes (A) to competition at driver (orange) and competitor (blue) positions to observe invariant responses in nRt (B). C, D, Modeling asymmetric and opposing OT response changes (C) to competition at driver and competitor positions to observe increasing responses in nRt (D).

To further validate the hypothesis that responses in nRt during competition depended on the strength of their inputs from different locations of the space map, we examined whether nRt units with extreme spatial tuning selectivity would exhibit response patterns expected from the model. We hypothesized that nRt units with extreme frontal driver selectivity would have stronger inputs from OT's space map regions tuned to frontal positions and would exhibit reductions in their spike response rate as the lateral competitor sound grows louder. Similarly, nRt units with extreme lateral driver selectivity would have stronger inputs from regions of the OT map tuned to lateral positions and would exhibit increasing spike response rates as the lateral competitor sound grows louder. Although very few nRt units (n = 5) exhibited extreme tuning selectivity for either frontal or lateral drivers, extreme frontal driver-selective units exhibiting tuning selectivity values >0.30 (frontal location, Fig. 4A) decreased their spike response rate with frontal driver and lateral competitor competition (Fig. 6A). In contrast, extreme lateral driver-selective units exhibiting tuning selectivity values lower than −0.50 (Fig. 4A, lateral location) increased their spike response rate with frontal driver and lateral competitor competition (Fig. 6B).

Figure 6.
Figure 6.

nRt units with extreme selectivity exhibit response patterns expected from their dominant OT input. A, Percent change in spike response rates for each relative level during sound competition with a frontal driver and lateral competitor for nRt units with extreme frontal driver selectivity (tuning selectivity index >0.30, n = 3). B, Percent change in spike response rates for each relative level during sound competition with a frontal driver and lateral competitor for nRt units with extreme lateral driver selectivity (tuning selectivity index <−0.50, n = 2). Error bars indicate standard error of mean.

We further investigated how asymmetric responses predicted from the model could arise. Given that significant changes in spike response with competition in nRt were associated with frontal competitors (Fig. 4C), we analyzed whether frontal and lateral OT responses differed with competition. Our results show that lateral OT units presented with frontal competitors exhibit abrupt decreases in spike response rates at earlier transition points than frontal OT units presented with lateral competitors (Fig. 7A,B). The faster transition point in lateral OT units may reflect higher sensitivity to frontal competitors due to the outer facial ruff providing the highest gain for frontal locations (Keller et al., 1998; von Campenhausen and Wagner, 2006). Given that most nRt units also exhibit robust representation of frontal azimuths (Fig. 2A), the asymmetric gain changes predicted from the model (Fig. 5C) may be a natural consequence of frontal competitors (Fig. 7B) evoking earlier and stronger changes than lateral competitors in OT (Fig. 7A).

Figure 7.
Figure 7.

Lateral OT responses exhibit faster transitions than frontal OT responses. Percent change in spike response rates for each relative level during competition for OT units that prefer frontal (<45°, A) and lateral (>45°, B) azimuths. Box plots show the interquartile range (25th–75th percentile) with whiskers indicating 5–95% confidence interval, and circles represent individual units. ANOVA p value and relevant post hoc p values reported for each pairwise comparison.

Altogether, these results show that individual nRt units integrate spatial information from wide areas of the OT space map, with strong contributions coming from midbrain OT's overrepresented frontal locations. However, differences in the strength of input from various parts of the map lead to different responses from individual nRt units depending on location and competition. Therefore, activation of varying subpopulations of nRt units that receive their largest contribution of inputs from OT locations representing the salient sound source enables nRt to retain coding of spatial location conveyed from OT.

Changes in spike response rate and synchrony with amplitude-modulated noise competition in nRt

We also investigated the impact of competing sounds with different envelopes in nRt by presenting driver and competitor amplitude-modulated stimuli with different modulation frequencies. Amplitude-modulated noise stimuli exhibit time-dependent properties that approximate complex envelopes integral to the identity of natural stimuli (McDermott and Simoncelli, 2011). Consistent with the results observed for flat noise competition (Fig. 3A), the spike response rate in nRt induced by amplitude-modulated sounds increased as the competitor grew stronger with competition (Fig. 8A,B). These observations held true regardless of whether the amplitude modulation frequency (fAM) of the competitor was higher (Fig. 8A; F(1,5) = 4.75; p = 2.80 × 10−4; ANOVA) or lower (Fig. 8B; F(1,5) = 7.79; p = 4.04 × 10−7; ANOVA) than that of the driver.

Figure 8.
Figure 8.

Spike response rate increases but STS does not change in nRt with competition of amplitude-modulated stimuli. Competition between stimuli with different modulation frequencies (left column: driver fAM = 55 Hz, competitor fAM = 75 Hz; right column: driver fAM = 75 Hz, competitor fAM = 55 Hz). A, B, Percent change in spike response rates for each relative level during competition. C, D, Example raster plots for two relative conditions when the driver was stronger (−15 dB) and when the competitor was stronger (+10 dB). Spike times of two simultaneously recorded units (green and orange) with coincident spikes between the two units (black) indicated as ticks. E, F, Cross-correlograms from competing sounds, averaged across all simultaneously recorded unit pairs for each relative level condition. Number of unit pairs with significant STS indicated above each correlogram. Horizontal colored lines indicate the peak height of the average cross-correlogram computed from when a driver stimulus was presented alone. G, H, Competition synchrony index (CSI) measuring change in STS with competition relative to presentation of a single driver stimulus, across relative levels. Box plots show the interquartile range (25th–75th percentile) with whiskers indicating 5–95% confidence interval, and circles represent individual units. ANOVA p value and relevant post hoc p values reported for each pairwise comparison.

Similar to observations from flat noise competition (Fig. 3B), STS did not change significantly with amplitude-modulated noise competition, as evident from the raster plots of exemplary units (Fig. 8C,D) and average population competition synchrony measure (Fig. 8E–H). Instead, STS of nearby units in nRt was found to reflect the stronger stimulus’ modulation frequency (Fig. 9). Average cross-correlograms across different relative level conditions revealed oscillatory activity patterns (Fig. 9A,E). The average power spectrum across each unit pair's cross-correlogram confirmed that the oscillatory activity reflected the modulation frequencies of the driver and competitor stimuli (Fig. 9B,F). Interestingly, while both stimulus’ modulation frequencies are represented weakly in the oscillatory activity of CCGs for driver stronger conditions, only a single peak at the competitor's modulation frequency emerged as the competitor grew in intensity (Fig. 9B,F). We further investigated how this switch toward the competitor emerged for each relative level condition by plotting each unit's vector strength to the driver against the vector strength to the competitor (Fig. 9C,G). Unlike previously reported observations of midbrain ICx and OT units whose spike timing was influenced by both stimuli (Keller and Takahashi, 2005; Bae et al., 2024), nRt units exhibited strong preferences for either stimulus, as evident in the aggregation of data points toward the axes away from the line of unity (Fig. 9C,G). Additionally, the number of nRt units that exhibited significant phase locking to the driver stimulus decreased as the competitor increased in intensity during competition. Consistent with this observation, further analysis measuring the change in vector strength to the driver and competitor across relative levels showed that as the competitor level increased, the vector strength to the driver decreased and the vector strength to the competitor increased (Fig. 9D,H). Altogether these results suggest that spike timing in the nRt flexibly changes to reflect the identity of the strongest stimulus location, which may emerge from reading out properties of the winning stimulus from OT responses.

Figure 9.
Figure 9.

STS in nRt reflects the amplitude modulation frequency of the strongest stimulus. Competition between stimuli with different modulation frequencies (left column: driver fAM = 55 Hz, competitor fAM = 75 Hz; right column: driver fAM = 75 Hz, competitor fAM = 55 Hz). A, E, Cross-correlograms from competing sounds, averaged across all simultaneously recorded unit pairs for each relative level condition. Number of unit pairs with significant STS indicated above each correlogram. Horizontal colored lines indicate the peak height of the average cross-correlogram computed from when a driver stimulus was presented alone. B, F, Power spectra of cross-correlograms, averaged across all simultaneously recorded unit pairs. Colored arrows indicate the frequency corresponding to the maximum peak observed in the power spectra. C, G, Scatterplots of vector strength (VS) to the competitor as a function of the VS to the driver with competition across relative levels. Each dot represents one unit, gray dots show units that did not phase lock to the driver (Rayleigh's test of uniformity, percentage of units that significantly phase lock to the driver indicated within each subplot, green and blue dots), black dashed lines indicate unity. D, H, Mean VS to the driver and competitor as a function of relative level.

Interregional STS between OT and nRt during competition

This study also explored the downstream consequences of competition in OT on interregional signaling. Previous investigation of auditory competition in OT demonstrated that modulations in interneuronal synchrony are an emergent coding scheme for representing relative stimulus strength in OT and may have a functional role in downstream read out (Bae et al., 2024). Given that spatial tuning in nRt is broad, we hypothesized that instances of interregional STS within our simultaneously recorded unit pairs across regions would be highest when spatial tuning between OT and nRt units overlapped and when the relative strength of the driver stimulus that was favorable for both regions was strongest during competition.

First, we quantified the distribution of spatial tuning correlation values observed across our population of simultaneously recorded OT and nRt units (Fig. 10A; mean ± SD: 0.07 ± 0.14). Figure 10, B and C, illustrates an example OT and nRt unit pair that exhibited a low spatial tuning correlation value belonging to the bottom 25th percentile, and Figure 10, D and E, illustrates an example unit pair that exhibited a high spatial tuning correlation value belonging to the top 75th percentile. We further segregated the data by frontal (preferred location magnitude <45 degrees) and lateral (preferred location magnitude >45 degrees) OT units to observe how these different populations were contributing to spatial tuning correlation values in nRt. Frontal OT units exhibited larger spatial tuning correlation values with nRt units than lateral OT units (Fig. 10F; p = 1.73 × 10−13; t test), which could be a consequence of the overrepresentation of frontal space in OT influencing inputs to nRt (Knudsen, 1982).

Figure 10.
Figure 10.

Distribution of spatial tuning correlation values between simultaneously recorded OT and nRt units. A, Histogram of spatial tuning correlation values between OT and forebrain units where colors indicate interquartile ranges. Example OT (B) and nRt (C) unit pair exhibiting low spatial tuning correlation below the 25th percentile range (gray bars in A). Example OT (D) and nRt (E) unit pair exhibiting high spatial tuning correlation above the 75th percentile range (navy bars in A). F, Histogram of spatial tuning correlation values between OT and nRt contributed by frontal OT units with preferred location magnitudes less than 45° (blue) and lateral OT units with preferred location magnitudes greater than 45° (red). p value from t test on the frontal and peripheral histogram distributions reported in the top left corner indicates frontal OT units exhibited larger spatial tuning correlation values with nRt units than lateral OT units.

We then examined interregional STS by computing cross-correlograms from spike times of simultaneously recorded OT and nRt unit pairs and observed the relationship between spatial tuning overlap and competition (Fig. 11). First, we examined the impact of competition on interregional signaling across relative level conditions and found that cross-correlograms showed the largest peak heights for conditions when the driver was stronger than the competitor (Fig. 11A, columns). The peak of the mean cross-correlograms observed when the driver is strongest (Fig. 11A, relative level = −15 column) centered at +3 ms, suggesting that spike timing in nRt followed spike timing in OT at latencies expected of direct inputs. Consistent with this general observation, CSI across the population confirms that competition has a significant impact on interregional STS, such that signaling between regions is strongest when the driver is stronger (Fig. 11B; F(1,5) = 43.06; p = 1.63 × 10−43; ANOVA). Furthermore, cross-correlograms with the largest peaks were asymmetric and broad with higher spike train correlation values for positive time lags than negative time lags, suggesting stronger feedforward activity from OT to nRt than feedback activity from nRt to OT. We next investigated whether communication between regions was entirely dependent on feedforward OT activity regardless of spatial tuning overlap between units or whether shared inputs between OT and nRt also determined the strength of competition dependent signaling. To this end, we examined CSI across spatial tuning correlation quartiles (Fig. 11A, rows), specifically for competition conditions where cross-correlograms exhibited significant peaks (relative level = −15, −10). While STS correlations at long time scales across several hundreds of milliseconds contributed to the overall peak height of all cross-correlograms, a sharp peak associated with correlations at fast time scales across tens of milliseconds was observed only for unit pairs exhibiting spatial tuning correlation values above the top 50th percentile. Consistent with this observation, CSI is significantly modulated by the spatial tuning correlation quartile (Fig. 11C; F(1,5) = 6.53; p = 2.17 × 10−4; ANOVA), suggesting that signaling between regions is also dependent on the strength of shared inputs between OT and nRt units.

Figure 11.
Figure 11.

Interregional STS between OT and nRt during competition is most prominent when unit pairs across regions exhibit similar spatial tuning and the driver is strongest. A, Cross-correlograms computed from spike trains of units in OT and nRT for competing sounds, averaged across all simultaneously recorded unit pairs belonging to spatial tuning correlation quartiles, as indicated in Figure 10A colors and color bar (rows), and each relative level condition (columns). Number of unit pairs indicated in each subplot. B, C, Competition synchrony index (CSI) measuring change in STS with competition relative to presentation of a single driver stimulus across all relative levels (B) and for unit pairs exhibiting spatial tuning correlation values across all quartiles for relative level conditions −15 and −10 (left 2 columns) (C). Box plots show the interquartile range (25th–75th percentile) with whiskers indicating 5–95% confidence interval, and circles represent individual units. ANOVA p value and relevant post hoc p values reported for each pairwise comparison.

Previous work showed that gamma power in OT modulated STS, which may have implications for downstream readout (Bae et al., 2024). Additionally, previous electrophysiological studies in pigeons that recorded from OT and nRt simultaneously demonstrated that nRt responses synchronized to the spiking responses in OT representing salient novel visual stimuli (G. J. Marín et al., 2012), and this synchronization between OT and nRt relied on input from the isthmic nucleus, Ipc (G. Marín et al., 2007). This motivated further investigation into whether gamma oscillations in the barn owl OT, induced by the strongest stimulus by input from Ipc (Bae et al., 2024), may be involved in facilitating interregional signaling. We first examined the relationship between gamma power and interregional STS within and across regions (Fig. 12). Consistent with previous reports, gamma power was highest within OT when its driver stimulus was strongest (Fig. 12A, −15 relative level column) and decreased as the competitor level increased during competition. However, within nRt, gamma power was strongest only for units that exhibited high spatial tuning correlation values with simultaneously recorded OT units (Fig. 12B, >50th percentile rows). The average gamma power between both OT and nRt for each relative level condition and spatial tuning correlation quartile revealed the strongest gamma power condition when the driver stimulus was strongest and the spatial tuning correlation was highest between regions (Fig. 12C). Additionally, our results show a strong positive correlation between cross-region STS and the average gamma power observed within OT (Fig. 12D) and within nRt (Fig. 12E). However, the strongest correlation was observed between cross-region STS and the average gamma power between regions (Fig. 12F), suggesting that gamma power between regions reflects a measure of interregional signaling.

Figure 12.
Figure 12.

Cross-regional power of gamma oscillations and STS between OT and nRt is strongest during competition when the driver is strongest. A–C, Heat map of average gamma power within OT (A), within nRt (B), and between OT and nRt (C) for each spatial tuning correlation quartile, as indicated in Figure 10A (rows), as a function of relative level conditions (columns). Color bars indicate the average gamma power in each condition observed within OT (A), nRt (B), and between OT and nRT for simultaneously recorded units (C). D–F, Scatterplot of cross-regional STS as a function of average gamma power within OT (D), within nRt (E), and between OT and nRt (F). Pearson’s correlation coefficient indicated top left.

To further determine how gamma in these regions relate to spiking activity, we conducted spike field coherence analyses within each region and examined whether spikes occurred at a preferred phase of the gamma LFP (Fig. 13). Consistent with previous reports, OT exhibits large coherence values in the low gamma range (20–50 Hz) as well as a smaller bump within the high gamma range (50–75 Hz; Fig. 13A). On the other hand, nRt exhibited a small bump only in the high gamma range (50–75 Hz; Fig. 13B). Spikes in OT occurred at specific phases in the low gamma range (Fig. 13C) and exhibited stronger vector strengths to the low gamma compared with the high gamma range (Fig. 13E). In contrast, spikes in nRt did not exhibit strong preferences for any phase in the low gamma range (Fig. 13D) and exhibited weak preferences for a specific phase in the high gamma range (Fig. 13F). These results reinforce previous observations that low gamma is a specific signature of input from Ipc that influences spiking activity in OT and plays a local role in modulating synchrony of nearby units representing similar azimuths. Nevertheless, gamma oscillations in the low gamma range in OT likely supports interregional signaling indirectly by influencing strong synchronous activity in a local manner that impacts relay to nRt, even though spike timing in nRt is weakly determined by gamma oscillations.

Figure 13.
Figure 13.

Midbrain and forebrain spikes are coherent with different frequency bands of LFP gamma oscillations. A, B, Spike field coherence in OT (A) and nRt (B) during competition when the driver was strongest (−15), the driver and competitor were equal (0), and when the competitor was strongest (+10). C, D, Histogram of the difference between spike phases and preferred phase relative to the LFP bandpass filtered within the low gamma range (20–50 Hz) in OT (C) and nRt (D) during competition. E, F, Histogram of the difference between spike phases and preferred phase relative to the LFP bandpass filtered within the high gamma range (50–75 Hz) in OT (E) and nRt (F) during competition. Vector strengths (VS) indicated in each subplot. Gray curve in all plots constructed from estimating spike field coherence and spike phase from trial shuffled data.

Discussion

We examined the consequences of auditory competition on cross-regional transfer of spatial information across midbrain and forebrain regions where there is a transformation in the representation of auditory space. Overall, our findings suggest that nRt integrates responses from distant areas across OT's space map yet encodes spatial location through selective activation of subpopulations of nRt units that receive strong inputs from portions of OT's space map that represent the location of the strongest sound and selectively encodes sound identity features of the stronger stimulus selected by the midbrain. Additionally, interregional STS between OT and nRt is most prominent when unit pairs exhibit high spatial tuning overlap and when the strongest stimulus’ location during competition is preferred by units in both regions. Furthermore, though gamma oscillations weakly influence spike timing in nRt, average gamma power between regions correlates with strength of interregional STS.

Nrt responses to competition and spatial location are well explained by the consequences of convergence

General properties of OT's auditory space map influence the representation of space in nRt. Despite broad spatial tuning of nRT neurons, overrepresentation of frontal locations in OT (Knudsen 1982) leads to strong representation of frontal space for many nRt units (Fig. 2A), and frontal OT units exhibit higher spatial tuning correlation values with nRt units (Fig. 10F). These results are consistent with prior anatomical studies in lateral eyed pigeons that demonstrated pars dorsalis anterior and pars centralis subdivisions of the nRt exhibited stronger ventral tectal inputs corresponding to the lower visual field (G. Marín et al., 2003). Because other subdivisions in nRt, such as the pars posterior, exhibited more homogenous representation of tectal retinotopic map (G. Marín et al., 2003), future investigation could examine whether subdivisions in barn owl nRt also exhibit variations in biases of tectal inputs.

Our study also demonstrates nRt spike response rate and STS of nearby units diverge from responses observed in OT with auditory competition. While OT spike response rate and STS decrease as a distant competitor grows in strength (Mysore et al., 2010, 2011; Bae et al., 2024), nRt spike response rate either increases or remains unchanged (Fig. 4C). Because analyses in this study were conducted on multiunit activity, we cannot rule out the possibility that increasing responses of some nRt units as the competitor grows louder may be attributed to an overall increase in multiunit activity. However, this alternative explanation is unlikely because variations in nRt responses depend on whether the varying competitor stimulus occurs at the position of the units’ preferred location (Fig. 4C) and an overall increase in multiunit activity is expected from either combination of driver and competitor position pairs. Instead, modeling integration of OT responses at the driver and competitor position revealed that nRt units linearly combine opposing changes of OT responses (Fig. 5) across distant positions of the map, and variations in nRt responses result from varying strengths of input.

Furthermore, we demonstrate that nRt STS does not change during competition (Figs. 2C, 6E,F), corroborating that projection patterns from OT to different subdivisions in nontopographic nRt do not organize according to spatial tuning properties (Hellmann and Güntürkün, 2001; G. Marín et al., 2003; Hellmann et al., 2004). Instead, cross-correlograms of nRt units reveal oscillatory patterns that reflect the temporal properties of the strongest stimulus (Fig. 7A–F), consistent with nRt's role in encoding identity features (Y. C. Wang et al., 1993) and expected from reading out responses from the minority of OT units that phase lock to stimuli (Bae et al., 2024). Interestingly, spike timing of nRt units encode the modulation frequency of the competitor stimulus even when the competitor stimulus is weaker than the driver stimulus (Fig. 9D,H). This early transition point suggests that competition in nRt is biased toward the varying competitor level, which may either be attributed to enhanced sensitivity toward frontal locations due to the acoustic effects of the outer facial ruff (Fig. 7; Keller et al., 1998; von Campenhausen and Wagner, 2006), or consequence of weakening OT inputs at the driver position with adaptation toward the fixed driver stimulus level (Reches and Gutfreund, 2008; Reches et al., 2010). Future research could distinguish these circumstances by investigating stimulus-specific adaptation in OT toward amplitude-modulated noise stimuli.

Although nRt exhibits functional subdivisions based on sensitivity to various visual features (Y. C. Wang et al., 1993), as well as both direct excitatory and indirect inhibitory projections from OT (Mpodozis et al., 1996), our study was unable to determine the subdivisions of our nRt recording sites due to the lack of reports describing functional subdivisions based on auditory features. Nevertheless, our interregional STS analyses provide strong evidence of functional connectivity suggestive of direct inputs from OT, consistent with previous reports that demonstrated strength of interregional synchronization between OT and nRt did not change across different nRt subdivisions (G. J. Marín et al., 2012). Furthermore, we demonstrate that the strength of interregional STS depended on both spatial tuning overlap and competition conditions that favored signaling between strongly connected unit pairs (Fig. 11), underscoring that despite loss of topography, spatial information is transferred to subpopulations of nRt neurons that receive strong inputs from OT representing the salient sound source (Fig. 4). The broad and heterogeneous spatial tuning of individual nRt units may contribute to a distributed code that ensures preservation of spatial information throughout the nucleus.

Gamma oscillations and cross-regional synchrony

Previous dual region electrophysiological recordings of OT and nRt with single electrodes in pigeons showed that novel moving visual stimuli-induced synchronization of spiking activity between regions, which depended on burst firing input from Ipc to OT (G. Marín et al., 2007; G. J. Marín et al., 2012). More recent studies that utilized multielectrode arrays to record across the retinotopic map in pigeons’ OT demonstrated that concurrent visual stimuli at different locations elicited oscillatory bursts that alternated between tectal locations corresponding to the concurrent stimulus locations (Reynaert et al., 2023), suggesting that oscillatory bursts in OT, evoked by re-entrant signals from Ipc, reflect an underlying selection signal that potentially relays OT output to nRt sequentially.

The present study demonstrated nRt encodes identity features of the strongest sound, thereby extending Reynaert et al. (2023)’s findings that stimulus selection across OT's space map gates relay to nRt. Additionally, elevated gamma power in both regions correlated to the strength of interregional STS (Fig. 12D). Low gamma range oscillations in barn owls’ OT are spatially tuned (Sridharan et al., 2011) and modulate interneuronal synchrony for salient sound locations (Bae et al., 2024) by organizing spiking activity to a preferred phase (Fig. 13C,E). However, gamma's modulation of spike timing is much weaker in nRt than in OT and spike field coherence in nRt is greatest at a higher gamma frequency range (Fig. 13D,F), suggesting high-frequency gamma relays more strongly from OT to nRt than low-frequency gamma, and is consistent with previous studies that demonstrated higher priority stimuli were associated with higher oscillatory burst frequencies in OT (Reynaert et al., 2023).

Differences in results between our study and Marin et al. (2012) may be due to several factors such as differences related to visual and auditory sensory modality and the use here of multichannel recordings. Previous characterization of gamma oscillations in barn owls showed gamma induced by auditory stimuli was weaker than gamma induced by visual stimuli (Sridharan et al., 2011) and as such may explain the local extent of gamma observed within OT for auditory competition. Additionally our study utilized linear multichannel electrodes to record subset populations of units between regions and observed interregional STS was dependent on the extent of spatial tuning overlap (Fig. 11) and as such strength of connectivity between simultaneously recorded unit pairs (Bair et al., 2001; Atencio et al., 2016). Although the use of chronic multichannel linear electrode implants may have limited our search for highly connected units between regions, the weakening role of gamma oscillations in determining spike timing in nRt could be a natural consequence of convergence from OT to nRt, consistent with primate cortical studies where modulations of spike timing by gamma decreased along the visual pathway hierarchy (Jia et al., 2013). Altogether, we corroborate previous findings that gamma oscillations in OT play a local role within the space map to organize spiking activity of nearby neurons tuned to similar regions of space and may reflect an underlying selection signal. Given that average gamma power in both OT and nRt correlates with strength of interregional STS, gamma's role in facilitating strong coherent activity within the OT map may support interregional signaling, similar to how synchronous activity in primate V1 increases the probability of spikes occurring in downstream V2 (Zandvakili and Kohn, 2015).

Overall, neural coding transformations across regions are an important principle observed in many interconnected regions across species. This investigation of the barn owl's midbrain to forebrain junction along the tectofugal pathway in the context of auditory competition has uncovered general mechanisms in which convergence allows transfer and prioritization of salient spatial information between regions.

Footnotes

  • This work was funded by National Institute on Deafness and Other Communication Disorders (R01DC007690 and F30DC020109) and CRCNS-US-Israel R01NS135851 projects. We thank Yoram Gutfreund for his helpful comments for this manuscript.

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Andrea J. Bae at andrea.bae{at}einsteinmed.edu.

References

  1. ↵
    1. Asadollahi A,
    2. Mysore SP,
    3. Knudsen EI
    (2010) Stimulus-driven competition in a cholinergic midbrain nucleus. Nat Neurosci 13:889–895. https://doi.org/10.1038/nn.2573 pmid:20526331
    OpenUrlCrossRefPubMed
  2. ↵
    1. Atencio CA,
    2. Shen V,
    3. Schreiner CE
    (2016) Synchrony, connectivity, and functional similarity in auditory midbrain local circuits. Neuroscience 335:30–53. https://doi.org/10.1016/j.neuroscience.2016.08.024 pmid:27544405
    OpenUrlCrossRefPubMed
  3. ↵
    1. Bae AJ,
    2. Ferger R,
    3. Pea JL
    (2024) Auditory competition and coding of relative stimulus strength across midbrain space maps of barn owls. J Neurosci 44:1–15. https://doi.org/10.1523/JNEUROSCI.2081-23.2024 pmid:38664010
    OpenUrlCrossRefPubMed
  4. ↵
    1. Bair W,
    2. Zohary E,
    3. Newsome WT
    (2001) Correlated firing in macaque visual area MT: time scales and relationship to behavior. J Neurosci 21:1676–1697. https://doi.org/10.1523/JNEUROSCI.21-05-01676.2001 pmid:11222658
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Bala AD,
    2. Spitzer MW,
    3. Takahashi TT
    (2007) Auditory spatial acuity approximates the resolving power of space-specific neurons. PLoS One 2:e675. https://doi.org/10.1371/journal.pone.0000675 pmid:17668055
    OpenUrlCrossRefPubMed
  6. ↵
    1. Beckert MV,
    2. Fischer BJ,
    3. Pena JL
    (2020) Effect of stimulus-dependent spike timing on population coding of sound location in the owl's auditory midbrain. eNeuro 7:1–21. https://doi.org/10.1523/ENEURO.0244-19.2020 pmid:32188709
    OpenUrlCrossRefPubMed
  7. ↵
    1. Beckert MV,
    2. Pavão R,
    3. Peña JL
    (2017) Distinct correlation structure supporting a rate-code for sound localization in the owl's auditory forebrain. eNeuro 4:1–18. https://doi.org/10.1523/ENEURO.0144-17.2017 pmid:28674698
    OpenUrlCrossRefPubMed
  8. ↵
    1. Benowitz LI,
    2. Karten HJ
    (1976) Organization of the tectofugal visual pathway in the pigeon: a retrograde transport study. J Comp Neurol 167:503–520. https://doi.org/10.1002/cne.901670407
    OpenUrlCrossRefPubMed
  9. ↵
    1. Bregman AS
    (1990) Auditory scene analysis: the perceptual organization of sound. Cambridge, Massachusetts: The MIT Press.
  10. ↵
    1. Cherry EC
    (1953) Some experiments on the recognition of speech, with one and with two ears. J Acoust Soc Am 25:975–979. https://doi.org/10.1121/1.1907229
    OpenUrlCrossRef
  11. ↵
    1. du Lac S,
    2. Knudsen EI
    (1990) Neural maps of head movement vector and speed in the optic tectum of the barn owl. J Neurophysiol 63:131–146. https://doi.org/10.1152/jn.1990.63.1.131
    OpenUrlCrossRefPubMed
  12. ↵
    1. Hellmann B,
    2. Güntürkün O
    (2001) Structural organization of parallel information processing within the tectofugal visual system of the pigeon. J Comp Neurol 429:94–112. https://doi.org/10.1002/1096-9861(20000101)429:1<94::AID-CNE8>3.0.CO;2-5
    OpenUrlCrossRefPubMed
  13. ↵
    1. Hellmann B,
    2. Güntürkün O,
    3. Manns M
    (2004) Tectal mosaic: organization of the descending tectal projections in comparison to the ascending tectofugal pathway in the pigeon. J Comp Neurol 472:395–410. https://doi.org/10.1002/cne.20056
    OpenUrlCrossRefPubMed
  14. ↵
    1. Hu M,
    2. Naito J,
    3. Chen Y,
    4. Ohmori Y,
    5. Fukuta K
    (2003) Afferent and efferent connections of the nucleus rotundus demonstrated by WGA-HRP in the chick. Anat Histol Embryol 32:335–340. https://doi.org/10.1111/j.1439-0264.2003.00488.x
    OpenUrlCrossRefPubMed
  15. ↵
    1. Jia X,
    2. Tanabe S,
    3. Kohn A
    (2013) Γ and the coordination of spiking activity in early visual cortex. Neuron 77:762–774. https://doi.org/10.1016/j.neuron.2012.12.036 pmid:23439127
    OpenUrlCrossRefPubMed
  16. ↵
    1. Karten HJ,
    2. Hodos W
    (1970) Telencephalic projections of the nucleus rotundus in the pigeon (Columba livia). J Comp Neurol 140:35–51. https://doi.org/10.1002/cne.901400103
    OpenUrlCrossRefPubMed
  17. ↵
    1. Keller CH,
    2. Hartung K,
    3. Takahashi TT
    (1998) Head-related transfer functions of the barn owl: measurement and neural responses. Hear Res 118:13–34. https://doi.org/10.1016/S0378-5955(98)00014-8
    OpenUrlCrossRefPubMed
  18. ↵
    1. Keller CH,
    2. Takahashi TT
    (2005) Localization and identification of concurrent sounds in the owl's auditory space map. J Neurosci 25:10446–10461. https://doi.org/10.1523/JNEUROSCI.2093-05.2005 pmid:16280583
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Knudsen EI
    (1982) Auditory and visual maps of space in the optic tectum of the owl. J Neurosci 2:1177–1194. https://doi.org/10.1523/JNEUROSCI.02-09-01177.1982 pmid:7119872
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Knudsen EI
    (1984) Auditory properties of space-tuned units in owl's optic tectum. J Neurophysiol 52:709–723. https://doi.org/10.1152/jn.1984.52.4.709
    OpenUrlCrossRefPubMed
  21. ↵
    1. Knudsen EI,
    2. Konishi M
    (1978) A neural map of auditory space in the owl. Science 200:795–797. https://doi.org/10.1126/science.644324
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Knudsen EI,
    2. Konishi M
    (1980) Monaural occlusion shifts receptive-field locations of auditory midbrain units in the owl. J Neurophysiol 44:687–695. https://doi.org/10.1152/jn.1980.44.4.687
    OpenUrlCrossRefPubMed
  23. ↵
    1. Kohn A,
    2. Smith MA
    (2005) Stimulus dependence of neuronal correlation in primary visual cortex of the macaque. J Neurosci 25:3661–3673. https://doi.org/10.1523/JNEUROSCI.5106-04.2005 pmid:15814797
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Lepage KQ,
    2. Kramer MA,
    3. Eden UT
    (2011) The dependence of spike field coherence on expected intensity. Neural Comput 23:2209–2241. https://doi.org/10.1162/NECO_a_00169
    OpenUrlCrossRefPubMed
  25. ↵
    1. Marín GJ,
    2. Durán E,
    3. Morales C,
    4. González-Cabrera C,
    5. Sentis E,
    6. Mpodozis J,
    7. Letelier JC
    (2012) Attentional capture? Synchronized feedback signals from the isthmi boost retinal signals to higher visual areas. J Neurosci 32:1110–1122. https://doi.org/10.1523/JNEUROSCI.4151-11.2012 pmid:22262908
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Marín G,
    2. Letelier JC,
    3. Henny P,
    4. Sentis E,
    5. Farfán G,
    6. Fredes F,
    7. Pohl N,
    8. Karten H,
    9. Mpodozis J
    (2003) Spatial organization of the pigeon tectorotundal pathway: an interdigitating topographic arrangement. J Comp Neurol 458:361–380. https://doi.org/10.1002/cne.10591
    OpenUrlCrossRefPubMed
  27. ↵
    1. Marín G,
    2. Salas C,
    3. Sentis E,
    4. Rojas X,
    5. Letelier JC,
    6. Mpodozis J
    (2007) A cholinergic gating mechanism controlled by competitive interactions in the optic tectum of the pigeon. J Neurosci 27:8112–8121. https://doi.org/10.1523/JNEUROSCI.1420-07.2007 pmid:17652602
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Masino T,
    2. Knudsen EI
    (1992) Anatomical pathways from the optic tectum to the spinal cord subserving orienting movements in the barn owl. Exp Brain Res 92:194–208. https://doi.org/10.1007/BF00227965
    OpenUrlCrossRefPubMed
  29. ↵
    1. McDermott JH,
    2. Simoncelli EP
    (2011) Sound texture perception via statistics of the auditory periphery: evidence from sound synthesis. Neuron 71:926–940. https://doi.org/10.1016/j.neuron.2011.06.032 pmid:21903084
    OpenUrlCrossRefPubMed
  30. ↵
    1. Moiseff A
    (1989) Bi-coordinate sound localization by the barn owl. J Comp Physiol A 164:637–644. https://doi.org/10.1007/BF00614506
    OpenUrlCrossRefPubMed
  31. ↵
    1. Moiseff A,
    2. Konishi M
    (1981) Neuronal and behavioral sensitivity to binaural time differences in the owl. J Neurosci 1:40–48. https://doi.org/10.1523/JNEUROSCI.01-01-00040.1981 pmid:7346557
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Mpodozis J,
    2. Cox K,
    3. Shimizu T,
    4. Bischof HJ,
    5. Woodson W,
    6. Karten HJ
    (1996) GABAergic inputs to the nucleus rotundus (pulvinar inferior) of the pigeon (Columba livia). J Comp Neurol 374:204–222. https://doi.org/10.1002/(SICI)1096-9861(19961014)374:2<204::AID-CNE4>3.0.CO;2-6
    OpenUrlCrossRefPubMed
  33. ↵
    1. Mysore SP,
    2. Asadollahi A,
    3. Knudsen EI
    (2010) Global inhibition and stimulus competition in the owl optic tectum. J Neurosci 30:1727–1738. https://doi.org/10.1523/JNEUROSCI.3740-09.2010 pmid:20130182
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Mysore SP,
    2. Asadollahi A,
    3. Knudsen EI
    (2011) Signaling of the strongest stimulus in the owl optic tectum. J Neurosci 31:5186–5196. https://doi.org/10.1523/JNEUROSCI.4592-10.2011 pmid:21471353
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Mysore SP,
    2. Knudsen EI
    (2011) Flexible categorization of relative stimulus strength by the optic tectum. J Neurosci 31:7745–7752. https://doi.org/10.1523/JNEUROSCI.5425-10.2011 pmid:21613487
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Payne RS
    (1971) Acoustic location of prey by barn owls (Tyto alba). J Exp Biol 54:535–573. https://doi.org/10.1242/jeb.54.3.535
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Reches A,
    2. Gutfreund Y
    (2008) Stimulus-specific adaptations in the gaze control system of the barn owl. J Neurosci 28:1523–1533. https://doi.org/10.1523/JNEUROSCI.3785-07.2008 pmid:18256273
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Reches A,
    2. Gutfreund Y
    (2009) Auditory and multisensory responses in the tectofugal pathway of the barn owl. J Neurosci 29:9602–9613. https://doi.org/10.1523/JNEUROSCI.6117-08.2009 pmid:19641123
    OpenUrlAbstract/FREE Full Text
  39. ↵
    1. Reches A,
    2. Netser S,
    3. Gutfreund Y
    (2010) Interactions between stimulus-specific adaptation and visual auditory integration in the forebrain of the barn owl. J Neurosci 30:6991–6998. https://doi.org/10.1523/JNEUROSCI.5723-09.2010 pmid:20484641
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Reynaert B,
    2. Morales C,
    3. Mpodozis J,
    4. Letelier JC,
    5. Marín GJ
    (2023) A blinking focal pattern of re-entrant activity in the avian tectum. Curr Biol 33:1–14.e4. https://doi.org/10.1016/j.cub.2022.10.070
    OpenUrlCrossRefPubMed
  41. ↵
    1. Shimizu T,
    2. Bowers AN
    (1999) Visual circuits of the avian telencephalon: evolutionary implications. Behav Brain Res 98:183–191. https://doi.org/10.1016/S0166-4328(98)00083-7
    OpenUrlCrossRefPubMed
  42. ↵
    1. Smith MA,
    2. Kohn A
    (2008) Spatial and temporal scales of neuronal correlation in primary visual cortex. J Neurosci 28:12591–12603. https://doi.org/10.1523/JNEUROSCI.2929-08.2008 pmid:19036953
    OpenUrlAbstract/FREE Full Text
  43. ↵
    1. Sridharan D,
    2. Boahen K,
    3. Knudsen EI
    (2011) Space coding by gamma oscillations in the barn owl optic tectum. J Neurophysiol 105:2005–2017. https://doi.org/10.1152/jn.00965.2010 pmid:21325681
    OpenUrlCrossRefPubMed
  44. ↵
    1. Straight PJ,
    2. Gignac PM,
    3. Kuenzel WJ
    (2024) Mapping the avian visual tectofugal pathway using 3D reconstruction. J Comp Neurol 532:e25558. https://doi.org/10.1002/cne.25558
    OpenUrlCrossRefPubMed
  45. ↵
    1. von Campenhausen M,
    2. Wagner H
    (2006) Influence of the facial ruff on the sound-receiving characteristics of the barn owl's ears. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 192:1073–1082. https://doi.org/10.1007/s00359-006-0139-0
    OpenUrlCrossRefPubMed
  46. ↵
    1. Wang YC,
    2. Jiang S,
    3. Frost BJ
    (1993) Visual processing in pigeon nucleus rotundus: luminance, color, motion, and looming subdivisions. Vis Neurosci 10:21–30. https://doi.org/10.1017/S0952523800003199
    OpenUrlCrossRefPubMed
  47. ↵
    1. Wang Y,
    2. Shanbhag SJ,
    3. Fischer BJ,
    4. Peña JL
    (2012) Population-wide bias of surround suppression in auditory spatial receptive fields of the owl's midbrain. J Neurosci 32:10470–10478. https://doi.org/10.1523/JNEUROSCI.0047-12.2012 pmid:22855796
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. Zandvakili A,
    2. Kohn A
    (2015) Coordinated neuronal activity enhances corticocortical communication. Neuron 87:827–839. https://doi.org/10.1016/j.neuron.2015.07.026 pmid:26291164
    OpenUrlCrossRefPubMed
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Auditory Competition and Stimulus Selection across Spatial Locations from Midbrain to Forebrain in Barn Owls
Andrea J. Bae, Brian J. Fischer, José L. Peña
Journal of Neuroscience 11 December 2024, 44 (50) e1298242024; DOI: 10.1523/JNEUROSCI.1298-24.2024
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.