Abstract Many species of group-living rodents use ultrasonic vocalisations (USVs, >20 kHz) during affiliative contacts with group members. This study analysed USVs of adult Brandt’s voles Lasiopodomys brandtii on surfaces of three natural colonies. Most USVs had a flat contour of fundamental frequency (f0). The maximum f0 was 33.91 kHz and the minimum f0 28.43 kHz on average. The average duration of USVs was 27 ms and did not exceed 76 ms. Values of USV duration, fundamental and peak frequencies did not differ between colonies. Of 366 analysed USVs, 142 (38.8%) were emitted during series of audible alarm calls of other individuals; acoustic variables did not differ between USVs produced in presence or absence of alarm calls. The USVs of wild-living Brandt’s voles were very similar in the acoustic structure with affiliative USVs, previously recorded in captivity at short-term isolation or weak disturbance in social groups. We discuss that captive environment does not affect the species-specific pattern of USVs in the Brandt’s vole.

We investigated behaviour and vocalization of wild-living male and female Mongolian gazelles during the winter rutting period, occurring at low temperatures, on average − 21.7˚C. We used both manual and automated audio recordings and manual video recordings of calling animals in rutting harems. Recording vocalizations with automated devices placed within harems documented a high vocal activity during the rut in both sexes. Harem males produced their series of rutting barks day and night. Video recordings revealed larynx retraction at each bark resulting in substantial vocal tract elongation. Harem females produced various contact calls but did not produce barks. Both sexes produced nasal alarm snorts and oral running calls, with widely open mouth during galloping. Mating behaviour and fights between rival males were not accompanied by any calls. Both male rutting barks and female contact calls were unusually high-frequency for animals of such size. We integrated data on behavioural postures and acoustic variables with data on sex-specific vocal anatomy, obtained during dissections of male and female specimens. We showed that Mongolian gazelle is the only studied ruminant, in which flexible vocal folds were evolutionarily lost. For both male and female larynges, we report the lack of any thin and flexible, rostrally directed mucosal parts in the glottal region suggesting that Mongolian gazelle, contrary to most other mammals, do not produce their calls by expiratory airflow-induced vocal fold vibration. The production mechanism of the high-frequency calls in Mongolian gazelle remains unclear. We discuss the evolution of vocal rutting behaviour in polygynous Mongolian gazelles.

Ultrasonic vocalizations (USVs) were not previously described for pikas. This study describes ultrasonic calls, produced by Daurian pikas Ochotona dauurica when disturbed by a potential threat (standing researcher or a small chair with mounted recording device) in the vicinity of the burrow entrances of their native colonies. The USVs of pikas were always produced as separate calls, followed by intervals of up to a few minutes. The USVs were never produced in regular series. Frequency range of 80

Abstract Ultrasonic acoustic communication of European rabbits, Oryctolagus cuniculus, was not investigated, although domestic rabbits are widely used as farm and laboratory animals. We examined acoustic variables of ultrasonic vocalisations (USVs, above 20 kHz) of adult and infant domestic rabbits. We analysed a total of 337 USVs from domestic rabbits, 102 USVs from 66 adults and 235 USVs from infants aged between 1 and 11 days, originated from 47 litters. Adults produced USVs when approaching researchers or towards their adult conspecifics. Infants produced USVs in response to a 10-min isolation of a litter from the mother. The fundamental frequency of 80% of USVs ranged from 17 to 43 kHz and was on average 31.0 kHz in adults and 30.6 kHz in infants. The duration of 80% of USVs was short, varying from 11 to 53 ms and was on average 22 ms in adults and 31 ms in infants. The USVs did not differ in their acoustic variables or contour shapes between adults and infants. Nonlinear phenomena were rare and occurred only in infant USVs. We show that lagomorphs represent a new mammalian taxon for which USVs are documented and compare acoustic variables of USVs in European rabbits with those in rodents.

We investigated the ontogenetic development of vocal repertoire in cubs of the Eurasian lynx Lynx lynx from birth until 3 months of age. Calls of four litters of captive lynxes kept in large outdoor enclosures were recorded within natal dens and during activities near the dens automatically and manually. We confirmed early findings that Eurasian lynx cubs are capable of producing biphonic calls with two independent fundamental frequencies in call spectra, the low (f0) and the high (g0), potentially created with two different sound sources. In the vocal repertoire of cub Eurasian lynxes, we subdivided 3 structural classes, which included in total 10 call types. The structural classes were classified based on presence/absence of one or two fundamental frequencies: two low-frequency call types with f0-singly, four high-frequency call types with g0-singly, and four biphonic call types with both f0 and g0. The f0 ranged from 0.17 kHz to 1.19 kHz, the g0 ranged from 0.21 kHz to 7.19 kHz among different call types. All the 10 calls types were present in Eurasian lynx cubs starting from the first week of age; no one of these call types disappeared to 3 months of age and no additional call types appeared to this age. Assignment the calls to the predicted call type with discriminant analysis was 94.5% for the low-frequency calls, 89.3% for the high-frequency calls, and 67.8% for the biphonic calls, thus suggesting robust classification. Acoustic parameters of both f0 and g0 frequencies decreased with age. We discuss the ontogeny of g0 in biphonic calls of mammals and the high acoustic complexity of vocal repertoire of Eurasian lynx cubs in relation to evolution of vocal traits across the studied Felidae species.

In mammals, subterranean lifestyle is thought to be associated with limited dispersal ability and low effective population size. These characteristics are expected to increase both inter- and intrapopulation differentiation while reducing intra-population genetic diversity. However, the available empirical data indicate this pattern is not universally applicable. Expanding the range of studied taxa and comparing between populations of the same species can contribute to our understanding of the factors responsible for the genetic dynamics. Previously unexamined in this context, the northern mole vole ( Ellobius talpinus ) shows promise as a model species for population genetics due to its broad geographic range and habitat plasticity, and variation in breeding system. Here, we present the first assessment of the genetic patterns in E. talpinus and compare the levels of genetic diversity of two distant populations. As expected, the highly fragmented Novosibirsk population from the extreme northeast of the species’ range showed low microsatellite diversity (observed heterozygosity, Ho = 0.34 and unbiased expected heterozygosity, uHe = 0.55). The mitochondrial D-loop was nearly monomorphic with a haplotype diversity ( Hd ) of 0.14. In contrast, the subperipheric Saratov population displayed moderate nuclear ( Ho = 0.64; uHe = 0.76) and high mitochondrial variation ( Hd = 0.83), compared to the surface-dwelling voles. In a heterogeneous landscape, significant genetic differentiation was revealed at both intermediate (dozens of kilometers) and small (several kilometers) scales. However, within continuous suitable habitat, no fine-scale (within 1000 m) spatial structure was detected, except for that resulting from kin clustering.

Environmental and social factors are among the most important drivers of evolution of the acoustic structure of calls in mammals. For nine vole species, we compared the values of acoustic variables (duration, fundamental frequency and peak frequency) and discretely varying categorical traits of ultrasonic calls (10–15 calls per pup, 905 in total) recorded from 2–5-day-old pups (7–10 pups per species, 87 in total) isolated from the nest for 2 min and then weighed and measured for body variables. The species were different in body size, in the acoustic traits of their calls, and in preferential call contour shapes, note compositions and nonlinear vocal phenomena. The acoustic structure of the ultrasonic calls was species-specific, enabling 67.8

This study investigated vocal characteristics of calls, emitted by 10 male and 17 female domestic cats Felis catus in the mating season. We analyzed the relationship between vocal traits of "meows" and "rutting cries" and cat individual identity, sex, age, sex hormones, immunoglobulins and body mass. Compared with meows, cries were longer in duration, and lower in fundamental frequency, peak frequency and power quartiles. Cries were produced in series with regular intervals between calls and contained the articulation effect flutter substantially more often than meows. The occurrence of nonlinear vocal phenomena did not differ between cries and meows. Male meows and cries had significantly lower fundamental frequency than female meows and cries; durations, peak frequency and quartiles within meows and cries did not differ between sexes. Comparison of meows and cries of 11 individuals which provided both meows and rutting cries showed that individuality was stronger expressed in cries than in meows. We did not find significant correlations of body mass, age and blood serum parameters with vocal traits. All variables of the fundamental frequency of female cries did not differ from male meows. We hypothesize that during periods of estrus, female domestic cats decrease the fundamental frequency of their far-distant cries and that they mimic meows of males.

Ultrasonic calls (USVs) produced by rodents during contacts with group members are poorly studied compared to USVs elicited by isolation, handling, mate exposition or dyad same-sex interactions on a neutral territory. In this study, we apply a procedure for eliciting the USVs accompanying contacts between adult voles living in permanent social groups for two Lasiopodomys vole species. The values of all variables of fundamental and peak frequencies were significantly higher in L. brandtii than in L. mandarinus, the duration of USVs was two times longer in L. brandtii than in L. mandarinus. At the same time, the USVs did not differ between species in the occurrence of different call contours, nonlinear phenomena and note compositions. The USVs emitted during affiliative contacts were acoustically nearly identical to the low-frequency USVs described previously for these two species in the situation of short-term isolation from conspecifics. We discuss that voles can use the same type USVs in two different situations: of short-term isolation from conspecifics and during affiliative contacts in social groups. Call-eliciting procedure applied in this study is easy and potentially appropriate for pilot and comparative studies of USVs across species, including wild-type rodents without long experience of living under laboratory conditions.

This study compares population differences of stag rutting calls (bugles) in Manchurian wapiti Cervus canadensis xanthopygus. The bugles were recorded automatically in three native wild-living populations (Ussury, Khabar and Anyi) of Far East of Russia. In the bugles, we measured 13 acoustic variables, estimated percentages of different call contours (trapeze, descending, saddle) and noted a presence of nonlinear vocal phenomena (biphonation and deterministic chaos). Between-population acoustic differences were found in bugle duration, in the high fundamental frequency (g0), in peak frequency and in percentages of the bugles with trapeze and descending contours. Overall, the bugles from the populations Khabar and Anyi were less different in their acoustic structure from each other than from the bugles of the Ussury population. Between-population differences in the bugles were related to the geographical distance between three of Manchurian wapiti populations. We discuss that all subspecies of wapiti (Manchurian, Siberian C. c. sibiricus and North American C. c. nelsoni) displayed the similar pattern of male rutting bugles with a prominent long plateau of the high fundamental frequency. Another common trait of rutting bugles in male wapiti is the facultative appearance of the second low fundamental frequency (f0). In addition, we discuss that acoustic differences between the bugles of the Manchurian wapiti and the Siberian wapiti may serve as vocal indicators discriminating males of these subspecies during passive acoustic monitoring in zones of sympatry.

Among rodents, male ultrasonic courtship songs were investigated regarding their acoustical structure only in laboratory mice (Mus musculus) and used as bioindicators for pharmacological research. This study describes the complex acoustic structure of male ultrasonic courtship songs of another rodent species, the Harting's vole (Microtus hartingi), at the levels of phrases, syllables and notes. Male songs were recorded from 13 of 59 males paired for short-term with unfamiliar female on the neutral arena during sexual behaviour including copulations or attempts of copulations. We predicted that songs of males in trials with successful copulations will be more complex than in the trials including only attempts of copulations. Against prediction, the songs of males which only displayed attempts of copulations were more complex, with longer phrases, more numbers of syllables per phrase and with shorter intervals between syllables within phrases at the same number of notes per syllable. We propose that male vole ultrasonic songs associated with mating may provide stimuli for inducing ovulation in female voles. We discuss the similarities and differences of male songs of Harting's voles with literature data on male song in domestic mice, which produce fewer complex songs when courtship ended with ejaculations than when ejaculations did not occur.

Vocal repertoires of wild-living cub and adult cheetahs (Acinonyx jubatus) in Kenya were described for the first time. Previously, cheetah vocal repertoire was only described in captivity. Acoustic parameters and contexts of production of 12 call types (7 tonal and 5 non-tonal) were investigated in adults older 4 years and cubs under 6 months of age. We demonstrated in the wild cheetahs the same basic set of 8 call types which was found previously in captive cheetahs and compared acoustic variables and behavioural contexts in the wild and in captivity. In addition, we described 4 call types that occurred in nature in the contexts which were either rare or did not occur in captivity. Along ontogeny from cubs to adults the fundamental and peak frequencies of 7 tonal call types decreased, duration of 7 tonal and 3 non-tonal call types either did not change or increased. Pulse rate of chirrs, growls and purrs did not change with age. Structure of calls in adult cheetahs slightly differed between males and females. Only in 3 high-frequency tonal call types (chirp, meow and squeal), fundamental and peak frequencies of males were lower than in females. Call duration did not differ between sexes in all call types. A distinctive trait of cheetah vocalizations in comparison with other species of felids is an unusually high fundamental frequency of some tonal call types in both cubs and adults.

We examined the ontogeny of pup ultrasonic isolation-induced calls (USVs) in a highly social Arvicolinae rodent species, the Harting's vole (Microtus hartingi). We recorded, with two-days intervals, the USVs of 55 individual pups from 11 litters from 1 to 12 days of age and measured body mass and the linear body size variables of each subject. The number of pups producing USVs decreased from 100 % at 1-2 days of age to 11.5 % at 11-12 days of age. Call rate changed with age, with maximum at 3-4 d of age. It ranged from 121.8 to 94.2 USVs/min between 1-2 and 7-8 days of age, decreased to 36.4 USVs/min at 9-10 days of age and to 1.5 USVs/min at 11-12 days of age. Overall, pup USVs were becoming simpler in their contour shapes with increase of age. Pup age class significantly affected all acoustic variables for the exclusion of the start and maximum fundamental frequencies, which values did not display significant changes with age. Body mass negatively correlated with USV acoustic variables, for the exclusion of call rate. A high percent of USVs contained nonlinear phenomena. We discuss the differences and similarities of pup Harting's vole USV ontogeny with other vole species and summarize the traits which make pup USVs of the Harting' vole a convenient model for biomedical research.

We investigate the acoustic structure of alarm calls in a highly social rodent, the Harting's vole (Microtus hartingi) and describe the phenomenon of collective shouting of bursts of alarm calls which could be produced in synchronized series. The alarm calls of Harting's voles were recorded using an automatic device from 10 different family groups, containing from 4 to 15 potential callers, released to outdoor enclosures. Natural predators and humans served as call-eliciting stimuli. We reveal acoustic differences between urgent alarm calls, evoked by close-vicinity predators and identified from recordings by their calls and by rustle noise from the caller's escape to burrow, and other alarm calls. We also reveal acoustic parameters between alarm calls produced in bursts and other alarm calls (produced in individual series or by a few non-synchronized callers). We discuss why the alarm calls of Harting's voles are unusually high-frequency (about 17 kHz on average in the maximum fundamental frequency) among other vole species producing highfrequency alarms. Adaptive significance of producing the collective synchronized bursts of alarm calls by Harting's voles remains unclear without direct observations of vocal vigilance in this species under natural conditions.

Subterranean rodents represent an interesting model system for molecular ecologists. Their lifestyle is associated with fragmented environments, limited dispersal ability, and low fecundity. These characteristics all are expected to increase inter- and intrapopulation differentiation and reduce intra-population genetic diversity, yet the published empirical data revealed the lack of generality of this pattern. This emphasizes the importance of accumulating more data on individual species to understand what factors shape the genetic dynamics of populations. This study represents the first characterization of the population genetic diversity and fine-scale genetic structure of a highly specialized subterranean vole, Ellobius talpinus. We used nine microsatellite loci and a fragment of the mitochondrial D-loop to investigate genetic patterns of two distant populations: the Novosibirsk population at the extreme northeastern edge of the species range, and the sub-peripheral Saratov population. As a result, the two populations exhibit distinct patterns of genetic structure. The Novosibirsk population showed a low nuclear diversity with an observed heterozygosity (Ho) of 0.34 and an unbiased expected heterozygosity (uHe) of 0.55. Mitochondrial D-loop was nearly monomorphic with over 90% of the observed haplotypes being identical, resulting in a haplotype diversity (Hd) of 0.14 and nucleotide diversity (π) of 0.0004. In contrast, the Saratov population displayed moderate nuclear (Ho = 0.64; uHe = 0.76) and high mitochondrial variation (Hd = 0.83; π = 0.034), compared to the surface-dwelling voles. Nine haplotypes representing four well-differentiated mitochondrial clades were found in the Saratov population. In a heterogeneous landscape, significant genetic differentiation was revealed at both intermediate (dozens of kilometers) and fine (several kilometers) scales. However, within the continuous suitable habitat, no fine-scale spatial structure was observed apart from that caused by kin clustering. The spatial genetic patterns revealed in E. talpinus appear to reflect a combination of effects of strong social structure, local natural and anthropogenic barriers limiting dispersal opportunities, and occasional long-distance dispersal. ### Competing Interest Statement The authors have declared no competing interest.

We investigated acoustic variation responsible for the individuality of alarm calls produced by 50 Brandt's voles Lasiopodomys brandtii derived from 50 different wild-living colonies. For the first time, we described the calling pattern of Brandt's voles, producing a long series of short alarm calls with short inter-call intervals. The alarm calls displayed four different contours of fundamental frequency but were nevertheless strongly individually distinct within a series of 50 successive alarm calls per caller (2500 analyzed alarm calls). The average value of correct assignment of alarm calls to individuals with discriminant function analysis was 15 times higher than the value expected by chance and was robust, not decreasing with cross-validation. We discuss that the highly individualistic alarm calls provide a basis for individual recognition of callers by colony members. At the same time, heterogeneity of call contours makes the long call series less monotonous, potentially preventing habituation and promoting alertness of conspecific call recipients.

In this study, we describe the acoustic structure of ultrasonic alarm calls of Mongolian gerbils Meriones unguiculatus in the wild and verify these calls as belonging to Mongolian gerbils by comparison of their acoustic parameters with alarm calls recorded in captivity. Both in captivity and in the wild, the alarm calls of Mongolian gerbils represented prolonged calls with an average duration of 118 ms and a flat contour and an average maximum fundamental frequency of 26.84 kHz. We found that alarm calls of captive Mongolian gerbils were shorter and higher in fundamental frequency and followed in a quicker succession than in the wild. Although the dataset size is not sufficient to determine significant acoustic variation between the populations, we discuss the potential reasons of the acoustic differences between the ultrasonic alarm calls produced in the wild and in captivity in our study and between the alarm calls reported in literature for different captive populations. We propose a method for non-invasive estimation of occupancy of the burrows by Mongolian gerbils in fragmented colonies at very low population density, by presence of the ultrasonic alarm calls.

This study investigates the effects of individuality and harem-holding status on the acoustic parameters of rutting calls (bugles) of 14 stags of Siberian wapiti Cervus canadensis sibiricus. The stags competed for mating with hinds in a large herd, free ranging inside a fenced area with complex landscape. We considered as harem-holders the stags, which kept a harem of 5 or more hinds for at least 2 days. Of the 15 days of observations enveloping the most active rutting period, during 3 days there was only one harem in the enclosure, during 9 days two harems, during 2 days three harems and during 1 day four harems. Stag bugles displayed both individual and status-related variation. Harem-holders compared with peripheral stags had shorter bugles with higher minimum fundamental frequency. After winning a harem and changing status from harem-candidate to harem-holder, stags shortened the duration of their bugles, lowered the beginning and maximum fundamental frequency and increased the minimum fundamental frequency. Significantly higher than the levels expected by chance, discriminant analysis classified 78.9% of bugles by the correct stag status and 53.2% of bugles by the correct individual callers. Different acoustic parameters encoded the status and individuality of the bugles. Status was encoded by the duration of the start and end parts and by the beginning and end fundamental frequencies. Individuality was encoded by the maximum fundamental frequency. We discuss that rutting calls of Siberian wapiti, although individualized, do not represent vocal signatures. However, these calls reliably mark stag harem-holding status.

This study investigates acoustic variation of human-audible sonic (below 20 kHz) and human-inaudible ultrasonic (above 20 kHz) calls in a wild subterranean rodent, the Northern Mole Vole (Ellobius talpinus), under 3 call-eliciting tests conducted during captures for 1 day. The Contact-in-Tunnel Test modeled contacts of 2 individuals during digging earth in a burrow tunnel. The Restraint Test modeled restraint of an animal by a surrogate predator (human hand). The Release-to-Burrow Test modeled acoustic communication of many family members returned to their home burrow after their isolation for about 8 h, from morning to evening. We described 8 call types: 3 sonic, 3 ultrasonic, and 2 expanding from sonic to ultrasonic range of frequencies; 6 call types were described for the first time for this species. No relationship was found between acoustic parameters and proxies of body size (body mass and the width of 2 upper incisors). No sex differences were found in body size or the acoustic parameters. Different call types prevailed in different tests: wheeks and upsweeps were made during peaceful interactions; squeaks and squeals were related to animal discomfort during the Restraint Test; rasps were only made in Release-to-Burrow Tests when animals were plugging the burrow entrance; and variative calls did not show any relationship with type of test. Based on presence or absence of certain call types in the tests, we evaluate their potential communicative role in comparison with published data on vocal repertoires of other subterranean rodents. Northern mole voles are truly subterranean rodents living in the permanent darkness of their burrows. Reduced vision is compensated with well-developed sonic and ultrasonic acoustic communication. Three call-eliciting test designs revealed: (1) ultrasonic calls used by animals when meeting each other; (2) calls related to discomfort; and (3) hissing rasps used during plugging the damaged burrow tunnel. This study shows for the first time that subterranean rodents use numerous ultrasonic calls during friendly communication.