Description of the sound diversity of two species of tuco-tucos (Ctenomys torquatus and Ctenomys lami) in natural environment

Sound signals can travel through long distances, becoming an important communication channel between animals that need to establish contact beyond the visual form. They can then be considered a relevant form of communication between species living in the underground environment. Ctenomys torquatus and Ctenomys lami are solitary subterranean rodents, thus demanding an improvement of the communicative channels, especially in territorial defense actions and meeting potential reproductive partners. This work was aimed to describe the variability of acoustic signals emitted by C. torquatus and C. lami by analyzing the physical-morphological characteristics of the signals. The study was carried out in two populations, one of each species and was selected 14 individuals of C. torquatus and 15 C. lami. The acoustic signals were recorded in a natural environment, obtaining the sounds straight from the animal tunnels. A total of 1,380 signals were captured and analyzed, 786 from C. torquatus and 594 from C. lami. It was possible to characterize 5 different types of signals, emitted by both species. Most of the analyzed sequences presented low frequency, and many of these calls exhibited characteristics of long-range signals. It was verified a sharing of sound signals in both species, as regarding the acoustic parameters as the morphology of the analyzed spectrograms. For the first time, it was possible to have access to sound data emitted by direct subterranean rodents from their tunnels in the natural environment.


Introduction 48
After capture, the individuals were weighed and visually sexed [37,[39][40][41][42], after 150 performing these procedures, the animals were returned to the burrow from where they were 151 captured. When the animal was not captured, but the burrow got closed, indicating that there 152 was at least one individual inside the tunnel gallery, this individual was considered as 153 unidentified. All animals had the position of their burrows registered using a GPS (Garmin 154 Vista®). 155 Twenty-four C. torquatus individuals were selected: 3 males, 8 females, and 13 156 unidentified. Of these 24, it was possible to obtain vocalization from 14: 6 females, 2 males, 157 and 6 unidentified. In C. lami 15 animals were selected, 3 females, 1 pup, which was not able 158 to identify the sex, and 11 unidentified individuals. It was possible to obtain sound signals from 159 all individuals.  The insertion of the microphone into the burrows occurred first with the removal of the 171 sand deposited by the animal at its entrance, then was located in the main tunnel where the 172 microphone was positioned. A minimum period of 24 hours was considered when the animals were captured and returned to their burrows so that the individuals could acclimate before the 174 recording began. 175 A sampling effort of 65 hours of sound data collection for both species was made. For 176 C. torquatus, 39 hours were recorded, and for C. lami were 26 hours. The individual recording 177 time was at least 40 minutes.

184
The terminologies adopted in this paper to describe and identify the sound signals were 185 note and phrase [44]. The note is a single sound unit with no interval, although the set of notes 186 with silent intervals between them was called phrase (Fig 1). The acoustic parameters analyzed 187 were: phrase size or note size, number of notes in the signal, dominant frequency, maximum 188 frequency, and energy in the established sound sequences.

201
From an auditory and visual investigation, it was possible to identify five different types 202 of signals in C. torquatus and C. lami: mono, drummed, tuc, sequential and squeal (Fig 2), and 203 a new nomenclature was made for the signals that not yet had been identified due to the absence 204 of sound data for the species studied. A total of 1,380 signals were recorded and analyzed, 786 205 for C. torquatus, and 594 for C. lami. In C. torquatus the vocalizations were emitted by females, 206 males, and unidentified individuals (Table 1), while in C. lami, the signals were emitted by 207 females and unidentified individuals (Table 2). For all detected signals were measured the mean,  (Table 3).  It is a pulsed signal formed by a single low frequency note (Fig 3).  Table 3.

237
It is a pulsed signal, analogous to a non-vocal sound. Formed by serial notes, with a 238 silence interval between the notes, constituting the drummed phrases (Fig 4). The phrases ranged from 2 to 8 notes, emitted by females, males, and unidentified (Table 1). In  Table 3. They are low frequency signals, formed by a set of notes with a short interval of silence 249 separating one note from another, thus forming the tuc phrases (Fig 5). The sign tuc is an 250 onomatopoeia that originated the popular name of the animal: tuco-tuco.

253
In C. torquatus, it was the signal that showed the most incidence, being recognized 474 254 tucs phrases (Fig 2). In only 1 of 14 recorded individuals, it was not possible to identify this 255 signal. The phrases' length ranged from 2 to 12 notes, emitted by females, males, and 256 unidentified individuals (Table 1). For C. lami, 119 tuc phrases were identified (Fig 2). The  It is a rhythmic signal, with similar notes in continuity separated by a short silence, 263 constituting the phrases (Fig 6).

266
For C. torquatus, 60 sequenced phrases (Fig 2) were recognized, emitted by 6 267 individuals, females, males, and unidentified (Table 1). In C. lami, it was the signal that showed  Table 3. It is a harmonic signal and can be emitted in single note or in serial form. In serial form 273 the notes are separated by a silence interval, generating the squeal phrases (Fig 7). In C. torquatus, the squeal phrases ranged from 1 to 3 notes, totaling 15 phrases (Fig 2) 277 registered in 7 individuals: females, males, and unidentified (Table 1). For C. lami, 109 squeal 278 phrases were identified (Fig 2), ranging from 1 to 3 notes, emitted by 12 female and unidentified 279 individuals ( Table 2). The mean values of the acoustic parameters investigated in the squeal 280 signal of both species are shown in Table 3. A squeal subtype was identified in C. lami (Fig 7   281 C), a signal with a more stressed harmonic. 15 phrases were recognized, emitted by 3 282 individuals: female, and unidentified (Table 2), the phrases presented from 1 to 7 notes. The 283 mean values of the acoustic parameters investigated in the squeal subtype are shown in Table   284 3.

286
It was possible to verify, in both species, a sharing of the investigated sound signals, 287 expressing similarity in both acoustic parameters analyzed and in the spectrographic design 288 (Table 3) (Fig 3, 4, 5, 6, and 7). These analogies may be related to the similarity of environments 289 that these species occupy [34,35] or to the fact that they belong to the same phylogenetic group,

291
According to the mean data obtained from the frequency bands used (  (Table 3).

355
The tuc type vocalizations are mostly emitted in isolation and with variable repetition 356 in the number of notes (Table 3). The tuc signal has already been characterized for C. design. However, the low sound frequency range is shared between species (Table 4).

365
The tuc-equivalent signals in C. talarum, C. pearsoni, and Anillaco Ctenomys sp. are 366 recognized as long-distance signals, which reiterates the character of the tuc as a long-range 367 signal.

368
The sequential type signal may also be characterized as long-distance because it is 369 analogous to the physical description made for long-range signals [33]. The same way it 370 resembles the mating signal emitted by male individuals of C. talarum [20], however in C. 371 torquatus, it was emitted by both sexes (Table 1), and in C. lami it was possible to confirm only 372 the emission by females since all animals captured and identified were females (Table 2). In addition to morphological similarity, they also use an approximate frequency range during 374 emissions (Table 5).

387
In contrast, the squeal signal presented a relatively high-frequency range compared to 388 the values of the other signal types described (