Separating overlapping bat calls with a bi-directional long short-term memory network.

Acquiring clear acoustic signals is critical for the analysis of animal vocalizations. Bioacoustics studies commonly face the problem of overlapping signals, which can impede the structural identification of vocal units, but there is currently no satisfactory solution. This study presents a bi-directional long short-term memory network to separate overlapping echolocation-communication calls of 6 different bat species and reconstruct waveforms. The separation quality was evaluated using 7 temporal-spectrum parameters. All the echolocation pulses and syllables of communication calls in the overlapping signals were separated and parameter comparisons showed no significant difference and negligible deviation between the extracted and original calls. Clustering analysis was conducted with separated echolocation calls from each bat species to provide an example of practical application of the separated and reconstructed calls. The result of clustering analysis showed high corrected rand index (82.79%), suggesting the reconstructed waveforms could be reliably used for species classification. These results demonstrate a convenient and automated approach for separating overlapping calls. The study extends the application of deep neural networks to separate overlapping animal sounds.

Comparing context-dependent call sequences employing machine learning methods: an indication of syntactic structure of greater horseshoe bats.

For analysis of vocal syntax, accurate classification of call sequence structures in different behavioural contexts is essential. However, an effective, intelligent program for classifying call sequences from numerous recorded sound files is still lacking. Here, we employed three machine learning algorithms (logistic regression, support vector machine and decision trees) to classify call sequences of social vocalizations of greater horseshoe bats (Rhinolophus ferrumequinum) in aggressive and distress contexts. The three machine learning algorithms obtained highly accurate classification rates (logistic regression 98%, support vector machine 97% and decision trees 96%). The algorithms also extracted three of the most important features for the classification: the transition between two adjacent syllables, the probability of occurrences of syllables in each position of a sequence, and the characteristics of a sequence. The results of statistical analysis also supported the classification of the algorithms. The study provides the first efficient method for data mining of call sequences and the possibility of linguistic parameters in animal communication. It suggests the presence of song-like syntax in the social vocalizations emitted within a non-breeding context in a bat species.

A magnetic compass guides the direction of foraging in a bat.

Previously, two studies have provided evidence that bats can use magnetic field cues for homing or roosting. For insectivorous bats, it is well established that foraging represents one of the most fundamental behaviors in animals relies on their ability to echolocate. Whether echolocating bats can also use magnetic cues during foraging remains unknown, however. Here, we tested the orientation behavior of Chinese noctules (Nyctalus plancyi) during foraging in a plus-shaped, 4-channel apparatus under different magnetic field conditions. To minimize the effects of spatial memory on orientation from repeated experiments, naïve bats were tested only once in each experimental condition. As expected, under geomagnetic field and a food resource offered conditions, the bats significantly preferred to enter the channel containing food, indicating that they primarily relied on direct sensory signals unrelated to magnetic cues. In contrast, when we offered food simultaneously in all four channels and minimized any differences in all other sensory signals available, the bats exhibited a clear directional preference to forage along the magnetic field direction under either geomagnetic field or a magnetic field in which the horizontal component was rotated by 90°. Our study offers a novel evidence for the importance of a geomagnetic field during foraging.

Bats adjust temporal parameters of echolocation pulses but not those of communication calls in response to traffic noise.

Many studies based on acute short-term noise exposure have demonstrated that animals can adjust their vocalizations in response to ambient noise. However, the effects of chronic noise over a relatively long time scale of multiple days remain largely unclear. Bats rely mainly on acoustic signals for perception of environmental and social communication. Nearly all previous studies on noise-induced vocal adjustments have focused on echolocation pulse sounds. Relatively little is known regarding the effects of noise on social communication calls. Here, we examined the dynamic changes in the temporal parameters of echolocation and communication vocalizations of Vespertilio sinensis when exposed to traffic noise over multiple days. We found that the bats started to modify their echolocation vocalizations on the fourth day of noise exposure, with an increase of 42-91% in the total number of pulse sequences per day. Under noisy conditions, the number of pulses within a pulse sequence decreased by an average of 17.2%, resulting in a significantly slower number of pulses/sequence (P < 0.001). However, there was little change in the duration of a pulse sequence. These parameters were not significantly adjusted in most communication vocalizations under the noise condition (all P > 0.05), except that the duration decreased and the number of syllables/sequences increased in 1 type of communicative vocalization (P < 0.05). This study suggests that bats routinely adjust temporal parameters of echolocation but rarely of communication vocalizations in response to noise condition.

Social vocalizations of big-footed myotis (Myotis macrodactylus) during foraging.

Acoustic signals play a crucial role in transmitting information and maintaining social stability in gregarious animals, especially in echolocating bats, which rely primarily on biological sonar for navigating in the dark. In the context of foraging without relying on tactile, visual or olfactory cues, acoustic signals convey information not only on food but also on ownership and defense of resources. However, studies on such information remain fragmentary. In the present study, we aim to document the social vocal repertoire of Myotis macrodactylus at natural foraging sites. Multiple acoustic analyses and spectrographic classification revealed a rich foraging vocal repertoire comprising 6 simple syllables and 2 composites. Discriminant function analyses associated with a subset-validation procedure provided an optimal method to spectrographically classify all recorded sounds into different syllable types. Multidimensional scaling of median values of multiple parameters further confirmed notable differences among these syllables in a 3-D space. In addition, Euclidean distance analysis showed that there were some spectral similarities between specific social vocal syllables and feeding buzzes, which implied a potential jamming role. Altogether, the data indicate that bats at foraging sites under natural conditions used variant social vocalizations with different functions in addition to echolocation calls, providing supporting evidence for further work on the function and vocal mechanisms of acoustic communication in mammals.

Fermi Surface Reconstruction and Drop in the Hall Number due to Spiral Antiferromagnetism in High-T_{c} Cuprates.

We show that a Fermi surface reconstruction due to spiral antiferromagnetic order may explain the rapid change in the Hall number as recently observed near optimal doping in cuprate superconductors [Badoux et al., Nature (London) 531, 210 (2016)]. The single-particle spectral function in the spiral state exhibits hole pockets which look like Fermi arcs due to a strong momentum dependence of the spectral weight. Adding charge-density wave order further reduces the Fermi surface to a single electron pocket. We propose quantum oscillation measurements to distinguish between commensurate and spiral antiferromagnetic order. Similar results apply to certain metals in which topological order replaces antiferromagnetic order.

Coexistence of Incommensurate Magnetism and Superconductivity in the Two-Dimensional Hubbard Model.

We analyze the competition of magnetism and superconductivity in the two-dimensional Hubbard model with a moderate interaction strength, including the possibility of incommensurate spiral magnetic order. Using an unbiased renormalization group approach, we compute magnetic and superconducting order parameters in the ground state. In addition to previously established regions of Néel order coexisting with d-wave superconductivity, the calculations reveal further coexistence regions where superconductivity is accompanied by incommensurate magnetic order.

Bats respond to very weak magnetic fields.

How animals, including mammals, can respond to and utilize the direction and intensity of the Earth's magnetic field for orientation and navigation is contentious. In this study, we experimentally tested whether the Chinese Noctule, Nyctalus plancyi (Vespertilionidae) can sense magnetic field strengths that were even lower than those of the present-day geomagnetic field. Such field strengths occurred during geomagnetic excursions or polarity reversals and thus may have played an important role in the evolution of a magnetic sense. We found that in a present-day local geomagnetic field, the bats showed a clear preference for positioning themselves at the magnetic north. As the field intensity decreased to only 1/5th of the natural intensity (i.e., 10 µT; the lowest field strength tested here), the bats still responded by positioning themselves at the magnetic north. When the field polarity was artificially reversed, the bats still preferred the new magnetic north, even at the lowest field strength tested (10 µT), despite the fact that the artificial field orientation was opposite to the natural geomagnetic field (P<0.05). Hence, N. plancyi is able to detect the direction of a magnetic field even at 1/5th of the present-day field strength. This high sensitivity to magnetic fields may explain how magnetic orientation could have evolved in bats even as the Earth's magnetic field strength varied and the polarity reversed tens of times over the past fifty million years.

Ambient noise causes independent changes in distinct spectro-temporal features of echolocation calls in horseshoe bats.

One of the most efficient mechanisms to optimize signal-to-noise ratios is the Lombard effect - an involuntary rise in call amplitude due to ambient noise. It is often accompanied by changes in the spectro-temporal composition of calls. We examined the effects of broadband-filtered noise on the spectro-temporal composition of horseshoe bat echolocation calls, which consist of a constant-frequency component and initial and terminal frequency-modulated components. We found that the frequency-modulated components became larger for almost all noise conditions, whereas the bandwidth of the constant-frequency component increased only when broadband-filtered noise was centered on or above the calls' dominant or fundamental frequency. This indicates that ambient noise independently modifies the associated acoustic parameters of the Lombard effect, such as spectro-temporal features, and could significantly affect the bat's ability to detect and locate targets. Our findings may be of significance in evaluating the impact of environmental noise on echolocation behavior in bats.

Potential effects of anthropogenic noise on echolocation behavior in horseshoe bats.

We previously reported that band-pass filtered noise (BFN, bandwidth 20 kHz) affected the echolocation behavior of horseshoe bats in different ways depending on which frequencies within the bats' hearing range BFN was centered. We found that call amplitudes only increased when BFN was centered on the dominant frequency of the bats' calls. In contrast, call frequencies were shifted for all BFN stimuli centered on or below the dominant frequency of echolocation calls including when BFN was centered at 20 kHz (BFN20), which is far below the range of the bats' echolocation call frequencies. In the present study we focused on masking effects of BFN20 and used it as a model stimulus for anthropogenic noise to investigate how this noise affected call parameters of echolocation calls, such as their frequency, amplitude, duration and rate. We show that only call frequencies shifted in response to masking BFN20, whereas all other call parameters tested exhibited no noise-related changes. Our findings provide an empirical basis to quantitatively predict the impact of anthropogenic noise on echolocation behavior in bats. They also emphasize the need for a better understanding of the impact of anthropogenic noise on bioacoustic communication and orientation systems in general.

Different auditory feedback control for echolocation and communication in horseshoe bats.

Auditory feedback from the animal's own voice is essential during bat echolocation: to optimize signal detection, bats continuously adjust various call parameters in response to changing echo signals. Auditory feedback seems also necessary for controlling many bat communication calls, although it remains unclear how auditory feedback control differs in echolocation and communication. We tackled this question by analyzing echolocation and communication in greater horseshoe bats, whose echolocation pulses are dominated by a constant frequency component that matches the frequency range they hear best. To maintain echoes within this "auditory fovea", horseshoe bats constantly adjust their echolocation call frequency depending on the frequency of the returning echo signal. This Doppler-shift compensation (DSC) behavior represents one of the most precise forms of sensory-motor feedback known. We examined the variability of echolocation pulses emitted at rest (resting frequencies, RFs) and one type of communication signal which resembles an echolocation pulse but is much shorter (short constant frequency communication calls, SCFs) and produced only during social interactions. We found that while RFs varied from day to day, corroborating earlier studies in other constant frequency bats, SCF-frequencies remained unchanged. In addition, RFs overlapped for some bats whereas SCF-frequencies were always distinctly different. This indicates that auditory feedback during echolocation changed with varying RFs but remained constant or may have been absent during emission of SCF calls for communication. This fundamentally different feedback mechanism for echolocation and communication may have enabled these bats to use SCF calls for individual recognition whereas they adjusted RF calls to accommodate the daily shifts of their auditory fovea.

Ambient noise induces independent shifts in call frequency and amplitude within the Lombard effect in echolocating bats.

The Lombard effect, an involuntary rise in call amplitude in response to masking ambient noise, represents one of the most efficient mechanisms to optimize signal-to-noise ratio. The Lombard effect occurs in birds and mammals, including humans, and is often associated with several other vocal changes, such as call frequency and duration. Most studies, however, have focused on noise-dependent changes in call amplitude. It is therefore still largely unknown how the adaptive changes in call amplitude relate to associated vocal changes such as frequency shifts, how the underlying mechanisms are linked, and if auditory feedback from the changing vocal output is needed. Here, we examined the Lombard effect and the associated changes in call frequency in a highly vocal mammal, echolocating horseshoe bats. We analyzed how bandpass-filtered noise (BFN; bandwidth 20 kHz) affected their echolocation behavior when BFN was centered on different frequencies within their hearing range. Call amplitudes increased only when BFN was centered on the dominant frequency component of the bats' calls. In contrast, call frequencies increased for all but one BFN center frequency tested. Both amplitude and frequency rises were extremely fast and occurred in the first call uttered after noise onset, suggesting that no auditory feedback was required. The different effects that varying the BFN center frequency had on amplitude and frequency rises indicate different neural circuits and/or mechanisms underlying these changes.

Behavioural and neurobiological implications of linear and non-linear features in larynx phonations of horseshoe bats.

Mammalian vocalizations exhibit large variations in their spectrotemporal features, although it is still largely unknown which result from intrinsic biomechanical properties of the larynx and which are under direct neuromuscular control. Here we show that mere changes in laryngeal air flow yield several non-linear effects on sound production, in an isolated larynx preparation from horseshoe bats. Most notably, there are sudden jumps between two frequency bands used for either echolocation or communication in natural vocalizations. These jumps resemble changes in 'registers' as in yodelling. In contrast, simulated contractions of the main larynx muscle produce linear frequency changes, but are limited to echolocation or communication frequencies. Only by combining non-linear and linear properties can this larynx, therefore, produce sounds covering the entire frequency range of natural calls. This may give behavioural meaning to yodelling-like vocal behaviour and reshape our thinking about how the brain controls the multitude of spectral vocal features in mammals.

Fermi-surface truncation from thermal nematic fluctuations.

We analyze how thermal fluctuations near a finite temperature nematic phase transition affect the spectral function A(k,ω) for single-electron excitations in a two-dimensional metal. Perturbation theory yields a splitting of the quasiparticle peak with a d-wave form factor, reminiscent of a pseudogap. We present a resummation of contributions to all orders in the Gaussian fluctuation regime. Instead of a splitting, the resulting spectral function exhibits a pronounced broadening of the quasiparticle peak, which varies strongly around the Fermi surface and vanishes upon approaching the Brillouin-zone diagonal. The Fermi surface obtained from a Brillouin-zone plot of A(k,0) seems truncated to Fermi arcs.

Variation in the resting frequency of Rhinolophus pusillus in Mainland China: effect of climate and implications for conservation.

This study describes variation patterns in the constant frequency of echolocation calls emitted at rest and when not flying ("resting frequency" RF) of the least horseshoe bat, Rhinolophus pusillus, on a broad geographical scale and in response to local climatic variables. Significant differences in RF were observed among populations throughout the species range in Mainland China, and this variation was positively and significantly related to climate conditions, especially environmental humidity, but the variability was only weakly associated with geographical distance. Sex dimorphism in the RF of R. pusillus may imply that female and male might keep their frequencies within a narrow range for sex recognition. Moreover, bats adjusted resting frequency to humidity, which may imply partitioning diet by prey size or the influence of rainfall noise. The results indicate that bats adjust echolocation call frequency to adapt to environmental conditions. Therefore, environmental selection shape the diversity of echolocation call structure of R. pusillus in geographically separated populations, and conservation efforts should focus on changes in local climate and effects of environmental noise.

Duration-sensitive neurons in the inferior colliculus of horseshoe bats: adaptations for using CF-FM echolocation pulses.

The present study examines duration-sensitive neurons in the inferior colliculus (IC) of the least horseshoe bat, Rhinolophus pusillus, from China. In contrast to other bat species tested for duration selectivity so far, echolocation pulses emitted by horseshoe bats are generally longer and composed of a long constant-frequency (CF) component followed by a short downward frequency-modulated (FM) sweep (CF-FM pulse). We used combined CF-FM pulses to analyze the differential effects that these two pulse components had on the duration tuning in neurons of the horseshoe bat's IC. Consistent with results from other mammals, duration-sensitive neurons found in the least horseshoe bat fall into three main classes: short-pass, band-pass, and long-pass. Using a CF stimulus alone, 54% (51/95) of all IC neurons showed at least one form of duration selectivity at one or more stimulus intensities. In 65 of the 95 IC neurons tested with CF pulses, we were also able to test their duration selectivity for a combined CF-FM pulse, which increased the ratio of duration-sensitive neurons to 66% (43/65). Seven to 15 neurons that failed to show duration tuning for CF bursts became duration sensitive for CF-FM pulses, with most of them exhibiting short-pass (depending on stimulus intensity, between 4 and 8 neurons) or band-pass tuning (1-3 neurons). Increasing stimulus intensities did not affect the duration tuning in 53% (23/43) of duration-sensitive neurons for CF bursts and in about 26% (7/27) for CF-FM stimuli. In the remaining neurons, increasing sound levels generally reduced the ratio of duration-sensitive neurons to 33% for CF and 37% for CF-FM stimulation. In those that remained duration sensitive, louder CF bursts shortened best durations in band-pass neurons and cutoff durations in short- and long-pass neurons, whereas louder CF-FM stimuli reduced the cutoff durations only in short-pass neurons. Bandwidths of band-pass neurons were not significantly affected by any stimulus configuration, with only a slight trend for increasing bandwidths for louder CF bursts (but not CF-FM stimuli). Best durations and cutoff durations reached higher values than those in the other bat species examined so far and roughly match the longer durations of echolocation pulses emitted by horseshoe bats. Therefore presentation of a CF-FM stimulus improved the duration sensitivity in IC neurons by increasing the ratio of duration-tuned neurons and making them less susceptible to changes in signal intensity.

Electrical resistivity near Pomeranchuk instability in two dimensions.

We analyze the dc charge transport in the quantum critical regime near a d-wave Pomeranchuk instability in two dimensions. The transport decay rate is linear in temperature everywhere on the Fermi surface except at cold spots on the Brillouin zone diagonal. For pure systems, this leads to a dc resistivity proportional to T(3/2) in the low-temperature limit. In the presence of impurities the residual impurity resistance at T=0 is approached linearly at low temperatures.

A mechanism for vocal-respiratory coupling in the mammalian parabrachial nucleus.

Mammalian vocalizations require the precise coordination of separate laryngeal and respiratory motor pathways. Precisely how and where in the brain vocal motor patterns interact with respiratory rhythm control is unknown. The parabrachial nucleus (PB) is known to mediate key respiratory reflexes and is also considered a principle component of the mammalian vocal motor pathway, making it a likely site for vocal-respiratory interactions, yet a specific role for the PB in vocalizing has yet to be demonstrated. To investigate the role of the PB in vocal-respiratory coordination, we pharmacologically manipulated synaptic activity in the PB while spontaneously vocalizing horseshoe bats were provoked to emit either short, single syllable or long, multisyllabic vocal motor patterns. Iontophoresis of the GABAA agonist muscimol (MUS) into the lateral PB extended expiratory durations surrounding all vocalizations and increased mean call durations. Alternatively, application of the GABAA antagonist bicuculline methiodide (BIC) shortened expirations and call durations. In addition, BIC eliminated the occurrence of multisyllabic vocalizations. BIC caused a mild increase in quiet breathing rates, whereas MUS tended to slow quiet breathing. The results indicate that GABAA receptor-mediated inhibition in the lateral PB modulates the time course of respiratory phase switching during vocalizing, and is needed for proper coordination of calling and breathing in mammals. We hypothesize that vocal-respiratory rhythm entrainment is achieved at least in part via mechanisms similar to other forms of locomotor-respiratory coupling, namely somatosensory feedback influences on respiratory phase-switching in the lateral PB.

Vocal communication in adult greater horseshoe bats, Rhinolophus ferrumequinum.

Whereas echolocation in horseshoe bats is well studied, virtually nothing is known about characteristics and function of their communication calls. Therefore, the communication calls produced by a group of captive adult greater horseshoe bats were recorded during various social interactions in a free-flight facility. Analysis revealed that this species exhibited an amazingly rich repertoire of vocalizations varying in numerous spectro-temporal aspects. Calls were classified into 17 syllable types (ten simple syllables and seven composites). Syllables were combined into six types of simple phrases and four combination phrases. The majority of syllables had durations of more than 100 ms with multiple harmonics and fundamental frequencies usually above 20 kHz, although some of them were also audible to humans. Preliminary behavioral observations indicated that many calls were emitted during direct interaction with and in response to social calls from conspecifics without requiring physical contact. Some echolocation-like vocalizations also appeared to clearly serve a communication role. These results not only shed light upon a so far widely neglected aspect of horseshoe bat vocalizations, but also provide the basis for future studies on the neural control of the production of communicative vocalizations in contrast to the production of echolocation pulse sequences.

Spontaneous, generalized lipidosis in captive greater horseshoe bats (Rhinolophus ferrumequinum).

During a routine 6-month quarantine period, 3 of 34 greater horseshoe bats (Rhinolophus ferrumequinum) captured in mainland China and transported to the United States for use in echolocation studies were found dead with no prior history of illness. All animals were in good body condition at the time of death. At necropsy, a large amount of white fat was found within the subcutis, especially in the sacrolumbar region. The liver, kidneys, and heart were diffusely tan in color. Microscopic examination revealed that hepatocytes throughout the liver were filled with lipid, and in some areas, lipid granulomas were present. renal lesions included moderate amounts of lipid in the cortical tubular epithelium and large amounts of protein and lipid within Bowman's capsules in the glomeruli. In addition, one bat had large lipid vacuoles diffusely distributed throughout the myocardium. The exact pathologic mechanism inducing the hepatic, renal, and cardiac lipidosis is unknown. The horseshoe bats were captured during hibernation and immediately transported to the United States. It is possible that the large amount of fat stored coupled with changes in photoperiod, lack of exercise, and/or the stress of captivity might have contributed to altering the normal metabolic processes, leading to anorexia and consequently lipidosis in these animals.