Many teleost fish generate acoustic signals for vocal communication by the synchronized, high-frequency contraction of skeletal, sonic muscles. In midshipman, eight groups of brainstem neurons were distinguished after biocytin application to the sonic nerve that, we propose, represent the entire vocal motor circuit. Biocytin-filled terminals were ubiquitous within all areas containing labeled neurons and, together with ultrastructural evidence, suggested a serial, transneuronal transport at synaptic sites between at least three neuronal groups. The most intensely labeled neurons were positioned in the caudal brainstem and included a previously characterized pacemaker-motoneuron circuit and a newly recognized ventral medullary nucleus that itself gave rise to extensive commissural and lateral brainstem bundles linking the pacemaker circuitry to the rostral brainstem. Five additional groups formed a column rostrally within the medial brainstem adjacent to eighth nerve (octaval)-recipient nuclei largely presumed to be acoustic. This column extended dorsally up to the ventricular cell layer and as far anterior as midbrain isthmal levels. The best-defined group was in the octaval efferent nucleus that directly innervates the sacculus that is considered the auditory division of the inner ear. Saccular afferents and neurons throughout the medial column were also filled after biocytin application to the saccular nerve. This vocal-acoustic network overlaps low-threshold, electrical stimulation sites in the rostral brainstem that elicit vocalizations. The medial column must therefore be the origin of the descending pathway controlling activation of the vocal pacemaker circuitry and likely forms the basis for acoustically elicited vocalizations. We suggest this network, together with input from the pacemaker circuitry, is also the origin of a vocal-related, corollary discharge to acoustic nuclei. Direct links between vocal and acoustic brain regions are thus traits common to aquatic and terrestrial vertebrates.

No abstract

While particle motion is thought to directly stimulate the inner ear of most fish species, it is difficult to measure and might not be predictable from pressure measurements in a small tank. It is therefore important to replicate experiments conducted relative to pressure measurements using stimuli of known particle motion, to ensure that unmeasured components of the stimulus field do not produce misleading frequency response profiles. The frequency sensitivity of the inner ear of the plainfin midshipman fish, Porichthys notatus, in response to isopressure stimuli has been described. This study now examines the frequency and directional response properties of midshipman saccular afferents in response to whole-body displacements simulating acoustic particle motion. Best frequencies were distributed bimodally, with peaks at 50 Hz and 100 Hz. Most units had cosinusoidally shaped directional response profiles in the horizontal and vertical planes, though some units showed slight deviations from this pattern. A few units (probably saccular efferents) had omnidirectional directional response profiles and did not phase lock to the stimulus waveform. These results are consistent with responses of the midshipman saccular nerve to isopressure stimuli, and strengthen the hypothesis that the frequency sensitivity of the midshipman ear matches the frequency content of behaviorally relevant vocalizations.

Plainfin midshipman fish (Porichthys notatus) have a remarkable capacity to generate long duration advertisement calls known as hums, each of which may last for close to two hours and be repeated throughout a night of courtship activity during the breeding season. The midshipman's striking sound production capabilities provide a unique opportunity to investigate the mechanisms that motor neurons require for withstanding high-endurance activity. The temporal properties of midshipman vocal behaviors are largely controlled by a hindbrain central pattern generator that includes vocal motor neurons (VMN) that directly determine the activity pattern of target sonic muscles and, in turn, a sound's pulse repetition rate, duration and pattern of amplitude modulation. Of the two adult midshipman male reproductive phenotypes --types I and II--only type I males acoustically court females with hums from nests that they build and guard, while type II males do not produce courtship hums but instead sneak or satellite spawn to steal fertilizations from type I males. A prior study using next generation RNA sequencing showed increased expression of a number of cellular respiration and antioxidant genes in the VMN of type I males during the breeding season, suggesting they help to combat potentially high levels of oxidative stress linked to this extreme behavior. This led to the question of whether the expression of these genes in the VMN would vary between actively humming versus non-humming states as well as between male morphs. Here, we tested the hypothesis that to combat oxidative stress, the VMN of reproductively active type I males would exhibit higher mRNA transcript levels for two superoxide dismutases (sod1, sod2) compared to the VMN of type II males and females that do not hum and in general both of which have a more limited vocal repertoire than type I males. The results showed no significant difference in sod1 transcript expression across reproductive morphs in the VMN and the surrounding hindbrain, and no difference of sod2 across the two male morphs and females in the SH. However, we observed a surprising, significantly lower expression of sod2 transcripts in the VMN of type I males as compared to type II males. We also found no significant difference in sod1 and sod2 expression between actively humming and non-humming type I males in both the VMN and surrounding hindbrain. These findings overall lead us to conclude that increased transcription of sod1 and sod2 is not necessary for combatting oxidative stress from the demands of the midshipman highendurance vocalizations, but warrant future studies to assess protein levels, enzyme activity . CC-BY-ND 4.0 International license It is made available under a (which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint . http://dx.doi.org/10.1101/354977 doi: bioRxiv preprint first posted online Jun. 26, 2018; 3 levels, as well as the expression of other antiox...

Neural selectivity to signal duration within the auditory midbrain has been observed in several species and is thought to play a role in signal recognition. Here we examine the effects of signal duration on the coding of individual and concurrent vocal signals in a teleost fish with exceptionally long duration vocalizations, the plainfin midshipman, Porichthys notatus. Nesting males produce long-duration, multi-harmonic signals known as hums to attract females to their nests; overlapping hums produce acoustic beats at the difference frequency of their spectral components. Our data show that all midbrain neurons have sustained responses to long-duration hum-like tones and beats. Overall spike counts increase linearly with signal duration, although spike rates decrease dramatically. Neurons show varying degrees of spike rate decline and hence, differential changes in spike rate across the neuron population may code signal duration. Spike synchronization to beat difference frequency progressively increases throughout long-duration beats such that significant difference frequency coding is maintained in most neurons. The significance level of difference frequency synchronization coding increases by an order of magnitude when integrated over the entirety of long-duration signals. Thus, spike synchronization remains a reliable difference frequency code and improves with integration over longer time spans.