Towards automated long-term acoustic monitoring of endangered river dolphins: a case study in the Brazilian Amazon floodplains

Using passive acoustic monitoring (PAM) and convolutional neural networks (CNN), we monitored the movements of the two endangered Amazon River dolphin species, the boto (Inia geoffrensis) and the tucuxi (Sotalia fluviatilis) from main rivers to floodplain habitats (várzea) in the Mamirauá Reserve (Amazonas, Brazil). We detected dolphin presence in four main areas based on the classification of their echolocation clicks. Using the same method, we automatically detected boat passages to estimate a possible interaction between boat and dolphin presence. Performance of the CNN classifier was high with an average precision of 0.95 and 0.92 for echolocation clicks and boats, respectively. Peaks of acoustic activity were detected synchronously at the river entrance and channel, corresponding to dolphins seasonally entering the várzea. Additionally, the river dolphins were regularly detected inside the flooded forest, suggesting a wide dispersion of their populations inside this large area, traditionally understudied and particularly important for boto females and calves. Boats overlapped with dolphin presence 9% of the time. PAM and recent advances in classification methods bring a new insight of the river dolphins' use of várzea habitats, which will contribute to conservation strategies of these species.

The biosonar of the boto: evidence of differences among species of river dolphins (Inia spp.) from the Amazon

Echolocation clicks can reflect the anatomy of the vocalizing animal, enabling the distinction of species. River dolphins from the family Iniidae are formally represented by one species and two subspecies (Inia geoffrensis geoffrensis and I. g. humboldtiana). Additionally, two other species have been proposed (I. boliviensis and I. araguaiaensis) regarding its level of restricted distribution and morph-genetics differences. For the Committee on Taxonomy of the Society for Marine Mammalogy, the specific status of the proposed species relies on further knowledge on morphology, ecology, and genetics. Given that species-specific status is required for conservation efforts, we described and compared the echolocation clicks of Inia spp., searching for specific differences on their vocalizations. The sounds were captured with a Cetacean Research ™ C54XRS (+3/−20 dB, −185 dB re: 1V/μPa) in Guaviare River (Orinoco basin), Madeira River (Madeira basin), Xingu River (Amazon Basin), and Araguaia River (Tocantins-Araguaia basin). We found significant differences in all analyzed parameters (peak frequency, 3 dB bandwidth, 10 dB bandwidth and inter-click interval) for all species and subspecies. Differences in acoustical parameters of clicks are mainly related to the animal’s internal morphology, thus this study may potentially support with information for the species-level classification mostly of I. araguaiaensis (the Araguaian boto). Classifying the Araguaian boto separately from I. geoffrensis has important implications for the species in terms of conservation status, since it is restricted to a highly impacted river system.

Repeated downsweep vocalizations of the Araguaian river dolphin, Inia araguaiaensis

Araguaian botos (Inia araguaiaensis) are known to produce pulsed as well as tonal sounds. This study documents the first evidence for repetitive sequences of downsweep whistles in botos that appear to be shared between individuals, and the context of their occurrence is investigated. Boat surveys were conducted along the Tocantins River located in the Eastern Amazon over a period of 42 days between 2012 and 2018. Eighty-two groups of Araguaian botos were observed, and 43 h of sound recordings were acquired. 632 downsweep whistles were recorded in 10 encounters. Four of these encounters contained downsweep bouts (21 bouts with ≥2 whistles) with short inter-call intervals (bout criterion 50 s) and up to 161 whistles. A statistical relationship was not found between downsweep occurrence and any of the contextual parameters that were investigated, including socializing, travelling, feeding, group size, presence of calves, and socio-sexual displays. The rarity of these signals makes them unlikely candidates for individual or group identification. It is more likely that they are associated with very specific contexts, such as nursing or mating, both of which were rarely observed in this study. Further studies are required to investigate context specificity and elucidate the function of these signals.

Characteristics and microgeographic variation of whistles from the vocal repertoire of beluga whales (Delphinapterus leucas) from the White Sea

The beluga whale (Delphinapterus leucas) produces a variety of sounds, including whistles as well as pulsed, noisy, and biphonic vocalizations. This study presents the fine-scale, microgeographic characteristics of beluga whistles from data collected in four locations across Onega Bay and Dvina Bay in the White Sea, Russia. Ten parameters were measured from 1232 whistles. The whistles had a fundamental frequency from 0.4 to 25.0 kHz and duration from 0.04 to 3.93 s. Although two distinct types could be recognized in the White Sea beluga's whistle repertoire, other whistles formed a graded continuum. Among them, "flat" whistle contours with no inflection points appear to be the most common (39.7%), to be followed by frequency-modulated whistles: ascending (27.1%) and descending (15.6%). Non-linear phenomena detected in the whistles included frequency jumps (23.1%), biphonations (13.2%), sidebands (5.2%), and subharmonics (0.5%). The whistles varied in frequency parameters and duration across the locations, while those recorded in the same location in different years showed minimal difference. Beluga whistles appear to be an extremely diverse class of vocalizations. This, together with the lack of clear correlations between the behavior of whales and whistle production suggests whistles may perform multiple functions within the beluga whale communication system.

The newly described Araguaian river dolphins, Inia araguaiaensis (Cetartiodactyla, Iniidae), produce a diverse repertoire of acoustic signals

The recent discovery of the Araguaian river dolphin (Inia araguaiaensis) highlights how little we know about the diversity and biology of river dolphins. In this study, we described the acoustic repertoire of this newly discovered species in concert with their behaviour. We analysed frequency contours of 727 signals (sampled at 10 ms temporal resolution). These contours were analyzed using an adaptive resonance theory neural network combined with dynamic time-warping (ARTwarp). Using a critical similarity value of 96%, frequency contours were categorized into 237 sound-types. The most common types were emitted when calves were present suggesting a key role in mother-calf communication. Our findings show that the acoustic repertoire of river dolphins is far from simple. Furthermore, the calls described here are similar in acoustic structure to those produced by social delphinids, such as orcas and pilot whales. Uncovering the context in which these signals are produced may help understand the social structure of this species and contribute to our understanding of the evolution of acoustic communication in whales.

Social sounds produced by franciscana dolphins, Pontoporia blainvillei (Cetartiodactyla, Pontoporiidae)

Franciscana dolphin (Pontoporia blainvillei) whistles were documented for the first time during 2003-2013 in Babitonga Bay estuary, South Brazil, together with burst pulses. Recordings were made from small boats under good sea conditions, and recording equipment that allowed analysis of sounds up to 96 kHz. The recordings were made in the presence of 2-31 franciscana dolphins. During 23 h and 53 min, 90 whistles and 51 burst pulse series were recorded. Although Guiana dolphins (Sotalia guianensis) inhabit nearby waters, none were observed in the area during the recordings. The authors recorded ten types of whistles. The initial frequency varied between 1.6 and 94.6 kHz, and the final frequency varied between 0.7 and 94.5 kHz; the authors were not able to determine if dolphin whistles exceeded the 96 kHz recording limit of the authors' equipment, although that is likely, especially because some whistles showed harmonics. Whistle duration varied between 0.008 and 0.361 s. Burst pulses had initial frequencies between 69 and 82.1 kHz (77 ± 3.81). These results showed that P. blainvillei produces whistles and burst pulses, although they seem to be produced infrequently.

Vocalizations of Amazon river dolphins (Inia geoffrensis): Characterization, effect of physical environment and differences between populations

The vocal repertoire of the Amazon river dolphin and its geographic variations are still poorly known, especially in relation to ecological variables. Here the acoustic characteristics of low frequency pulsed vocalizations, with single or multiple pulses, recorded in two protected areas of the Amazon were described and differences in acoustic emissions related to water properties were analyzed. Both frequency and time parameters differ relative to abiotic condition of water turbidity. Changes in the animals' acoustic behavior might be due to differences in sound propagation between rich-sediment water and clear water. Geographic variation was found in frequency and time parameters, requiring further investigation.

Low-frequency signals produced by Northeast Atlantic killer whales (Orcinus orca)

Killer whale acoustic behavior has been extensively investigated; however, most studies have focused on pulsed calls and whistles. This study reports the production of low-frequency signals by killer whales at frequencies below 300 Hz. Recordings were made in Iceland and Norway when killer whales were observed feeding on herring and no other marine mammal species were nearby. Low-frequency sounds were identified in Iceland and ranged in duration between 0.14 and 2.77 s and in frequency between 50 and 270 Hz, well below the previously reported lower limit for killer whale tonal sounds of 500 Hz. Low-frequency sounds appeared to be produced close in time to tail slaps, which are indicative of feeding attempts, suggesting that these sounds may be related to a feeding context. However, their precise function is unknown, and they could be the by-product of a non-vocal behavior rather than a vocal signal deliberately produced by the whales. Although killer whales in Norway exhibit similar feeding behavior, this sound has not been detected in recordings from Norway to date. This study suggests that, like other delphinids, killer whales produce low-frequency sounds, but further studies will be required to understand whether similar sounds exist in other killer whale populations.

High initial and minimum frequencies of Sotalia guianensis whistles in the southeast and south of Brazil

Recent studies have proposed latitudinal variations for the whistles of Sotalia guianensis, since parameters of frequency increase from the southernmost distribution of the species (State of Santa Catarina, Brazil) toward the North. The last study on the whistles of S. guianensis, conducted in Costa Rica, confirmed the geographical variation of whistles, regardless of frequency responses used in the samples. Initial and minimum frequencies of whistles of the Costa Rican populations were higher than were the frequencies of Brazilian populations. In this current study, the whistles of S. guianensis were recorded with a sampling frequency of 48 kHz, the highest frequency band used in Brazilian studies. The initial and minimum frequency of whistles recorded and analyzed were similar to parameters of Costa Rican populations and show little or no difference when compared to the average of each study. Therefore, whistle frequency variables do not increase with latitude. It is likely that variations in acoustic parameters S. guianensis are mainly linked to environmental characteristics of the habitats in which the populations occur. Thus, the repertoire sound S. guianensis appears linked to similar environments with similar physical characteristics and not linked to population characteristics (dialects).

Echolocation behavior of franciscana dolphins (Pontoporia blainvillei) in the wild

Franciscana dolphins are small odontocetes hard to study in the field. In particular, little is known on their echolocation behavior in the wild. In this study we recorded 357 min and analyzed 1019 echolocation signals in the Rio Negro Estuary, Argentina. The clicks had a peak frequency at 139 kHz, and a bandwidth of 19 kHz, ranging from 130 to 149 kHz. This is the first study describing echolocation signals of franciscana dolphins in the wild, showing the presence of narrow-band high frequency signals in these dolphins. Whether they use other vocalizations to communicate or not remains uncertain.

Killer whales (Orcinus orca) produce ultrasonic whistles

This study reports that killer whales, the largest dolphin, produce whistles with the highest fundamental frequencies ever reported in a delphinid. Using wide-band acoustic sampling from both animal-attached (Dtag) and remotely deployed hydrophone arrays, ultrasonic whistles were detected in three Northeast Atlantic populations but not in two Northeast Pacific populations. These results are inconsistent with analyses suggesting a correlation of maximum frequency of whistles with body size in delphinids, indicate substantial intraspecific variation in whistle production in killer whales, and highlight the importance of appropriate acoustic sampling techniques when conducting comparative analyses of sound repertoires.

Dolphin social intelligence: complex alliance relationships in bottlenose dolphins and a consideration of selective environments for extreme brain size evolution in mammals

Bottlenose dolphins in Shark Bay, Australia, live in a large, unbounded society with a fission-fusion grouping pattern. Potential cognitive demands include the need to develop social strategies involving the recognition of a large number of individuals and their relationships with others. Patterns of alliance affiliation among males may be more complex than are currently known for any non-human, with individuals participating in 2-3 levels of shifting alliances. Males mediate alliance relationships with gentle contact behaviours such as petting, but synchrony also plays an important role in affiliative interactions. In general, selection for social intelligence in the context of shifting alliances will depend on the extent to which there are strategic options and risk. Extreme brain size evolution may have occurred more than once in the toothed whales, reaching peaks in the dolphin family and the sperm whale. All three 'peaks' of large brain size evolution in mammals (odontocetes, humans and elephants) shared a common selective environment: extreme mutual dependence based on external threats from predators or conspecific groups. In this context, social competition, and consequently selection for greater cognitive abilities and large brain size, was intense.

Predation by killer whales (Orcinus orca) and the evolution of whistle loss and narrow-band high frequency clicks in odontocetes

A disparate selection of toothed whales (Odontoceti) share striking features of their acoustic repertoires including the absence of whistles and high frequency but weak (low peak-to-peak source level) clicks that have a relatively long duration and a narrow bandwidth. The non-whistling, high frequency click species include members of the family Phocoenidae, members of one genus of delphinids, Cephalorhynchus, the pygmy sperm whale, Kogia breviceps, and apparently the sole member of the family Pontoporiidae. Our review supports the 'acoustic crypsis' hypothesis that killer whale predation risk was the primary selective factor favouring an echolocation and communication system in cephalorhynchids, phocoenids and possibly Pontoporiidae and Kogiidae restricted to sounds that killer whales hear poorly or not at all (< 2 and > 100 kHz).

The freshwater dolphin Inia geoffrensis geoffrensis produces high frequency whistles

Because whistles are most commonly associated with social delphinids, they have been largely overlooked, ignored, or presumed absent, in solitary freshwater dolphin species. Whistle production in the freshwater dolphin, the boto (Inia geoffrensis geoffrensis), has been controversial. Because of its sympatry with tucuxi dolphins (Sotalia fluviatilis), a whistling species, some presume tucuxi whistles might have been erroneously assigned to the boto. Using a broadband recording system, we recorded over 100 whistles from boto dolphins in the Yasunf River, Ecuador, where the tucuxi dolphins are absent. Our results therefore provide conclusive evidence for whistle production in Inia geoffrensis geoffrensis. Furthermore, boto whistles are significantly different from tucuxi whistles recorded in nearby rivers. The Ecuadorian boto whistle has a significantly greater frequency range (5.30-48.10 kHz) than previously reported in other populations (Peru and Colombia) that were recorded with more bandwidth limited equipment. In addition, the top frequency and the range are greater than in any other toothed whale species recorded to date. Whistle production was higher during resting activities, alone or in the presence of other animals. The confirmation of whistles in the boto has important implications for the evolution of whistles in Cetacea and their association with sociality.

Whistles of tucuxi dolphins (Sotalia fluviatilis) in Brazil: comparisons among populations

A quantitative and comparative analysis of the acoustic whistles parameters of tucuxi dolphins (Sotalia fluviatilis) from eight areas in Brazilian waters were conducted. Tucuxi produced mainly upsweep whistles, and despite their similar characteristics to other delphinids, whistles were shorter and less complex in shape. Some signals had fundamental frequency components above 24 kHz, and a broader band system may be useful in further studies concerning tucuxi whistles. The frequency parameters of the whistles had the lowest variation coefficients, whereas the duration and number of inflections had the highest variation coefficients. About 65% of all whistles had common characteristics between study localities, but SF, MinF, MeF, F1/4, F1/2, and F3/4 increased in frequency from southern to northern populations. The magnitude of the whistle characteristics variation was comparatively smaller between adjacent sites areas than between nonadjacent ones. However, for some pairwise comparisons the whistles were more similar between nonadjacent areas than between adjacent ones, indicating that other factors besides genetic and mixing affected tucuxi whistle characteristics. Only a few numbers of whistle variables were significantly different between adjacent areas and the parameter ranges of whistles overlapped, which made it difficult to discriminate whistles between adjacent tucuxi populations.