Abstract
Geographic variation in the vocal behavior of manatees has been reported but is largely unexplored. Vocalizations of wild West Indian manatees (Trichechus manatus) were recorded with hydrophones in Florida from Florida manatees (Trichechus manatus latirostris), and in Belize and Panama from Antillean manatees (Trichechus manatus manatus) to determine if calls varied between subspecies and geographic regions. Calls were visually classified into five categories: squeaks, high squeaks, squeals, squeak-squeals, and chirps. From these five categories, only three call types (squeaks, high squeaks and squeals) were observed in all three populations. Six parameters from the temporal and frequency domains were measured from the fundamental frequency of 2878 manatee vocalizations. A repeated measures PERMANOVA found significant differences for squeaks and high squeaks between each geographic location and for squeals between Belize and Florida. Almost all measured frequency and temporal parameters of manatee vocalizations differed between and within subspecies. Variables that may have influenced the variation observed may be related to sex, body size, habitat and/or other factors. Our findings provide critical information of manatee calls for wildlife monitoring and highlight the need for further study of the vocal behavior of manatees throughout their range.
Similar content being viewed by others
Introduction
Geographic separation can result in acoustic differences in animal vocalizations. Macrogeographic variation is observed between isolated populations, whereas microgeographic variations are associated with closely neighboring species that have the potential to interbreed1. Intraspecific variation in vocalizations in animals may occur due to factors such as genetic differences (birds2); properties of the habitat3; and species dispersion (birds4,5) among others. There have been many studies on geographic variation in multiple marine mammal species including harbor seals (Phoca vitulina)6, bearded seals (Erignathus barbatus)7, blue whales (Balaenoptera musculus)8, and bottlenose dolphins (Tursiops truncatus)9. Comparative studies of geographic differences in vocalizations and their acoustic parameters within species can improve our understanding of the evolution of the vocal repertoire and factors that influence the variation in vocalizations. While our understanding of the acoustic characteristics of West Indian manatee vocalizations has increased10,11,12,13, how the vocalizations vary by region or subspecies is still poorly understood.
The West Indian manatee is distributed in riparian and coastal systems from the Western Atlantic, Caribbean Sea, Gulf of Mexico, to the southeast of Brazil14,15. The two recognized subspecies of the West Indian manatee—the Florida manatee (T. m. latirostris) and the Antillean manatee (T. m. manatus)—are listed as threatened by the IUCN14,16,17. Florida manatees range from the Florida peninsula to the north of the Gulf of Mexico18,19 and migrate seasonally in the winter to warm water refuges20. The overall population size of Florida manatees is relatively well estimated (6350 manatees, 95% CI: 5310–739021) compared to the less studied Antillean subspecies22. The Antillean manatee inhabits shallow and warm waters in the Bahamas, Gulf coast of Mexico, the Greater Antilles, and Atlantic coasts of Central and South America14. Their populations are small23 and genetic analysis has separated them into three biogeographically distinct populations: (1) Florida, Central America, Antilles, and the Caribbean coast of South America; (2) Mexico, Central America, and the Caribbean coast of South America; and (3) the northeastern coast of South America including Guyana and Brazil24,25,26,27.
Across their range, Antillean manatee populations vary dramatically in their size and abundance. Belize has the largest population of Antillean manatees (approximately 1000 individuals) due to its richness of suitable habitats28. The distribution and population status for other regions is unknown or declining22 and connectivity between different Antillean manatee populations is poorly understood. Panama is thought to contain approximately 150 individuals and its population trend is currently unknown29. Examination of the genetics of Panamanian manatees determined they cluster with Mexico/Belize manatees30, which supports the hypothesis that manatees in Panama originated from the Mexico/Belize populations during historical colonization events and/or recent migrations24.
Tagging and photo-identification studies have revealed how far West Indian manatees migrate and potentially mix within their geographic distribution (See Deutsch et al. 202231 for a review). Several individual Florida manatees have been documented migrating to Cuba32 and the Caribbean coast of Mexico33 by matching images of scarring on these animals in both locations. Radio-tagged Antillean manatees in Mexico sometimes migrated over 100 km from Chetumal Bay, Mexico to Belize34 while most of the individuals tagged in Belize remained within 25 km of where they were captured35. In Panama, a lone female was tagged in the San San Pond Sak region and remained in this region throughout the study36. The authors cautiously suggest that the San San Pond Sak system contains habitat suitable to high site fidelity in manatees. However, more research is needed to see if Panama manatees migrate to neighboring countries. Based on these migratory patterns, it is possible that manatee subpopulations in Florida, Belize, and Panama may have differences in their vocal behavior.
Manatees are known to produce a variety of vocalizations that play a fundamental role in their communication, such as during social interactions and maintaining contact between cows and their calves37,38,39,40. Compared to other marine mammals, manatees have a relatively small vocal repertoire composed of five or six distinct call types11,37,41,42, depending on the author, that have fundamental frequency values between 0.5 and 5.0 kHz and duration between 0.2 and 0.5 s38,43,44. These graded call types contain narrow and broadband vocalizations11,40 and have been described as squeaks, high squeaks, squeak-squeals, squeals and chirps11. Squeaks, high squeaks and squeals are more commonly observed in reports of Florida manatee communication37,40,45. Recent studies found the high squeak is a stereotypical call produced by both Florida45,46 and Antillean calves46. As the animal grows, the hill-shaped contour of the high squeak appears to flatten and become a more linear call46. In addition, results showed few differences between calls produced between captive Florida and Antillean calves46.
The Florida and the Antillean manatee have similar vocal tract anatomies47,48 and cranial morphology49 which are directly related to vocal range capabilities47,48. Yet, few studies have focused on geographic variability in wild manatee vocalizations41,44. Analyses reported similarities and overlapping distributions in frequency and duration parameters measured between Antillean and Florida manatees41,44. However, these studies were based on individuals or small groups, and limited to two geographic locations. Studies addressing geographic variability could provide insight if similarities or differences in genetic variation is a factor in vocal variation in manatees. The aim of this study was to test the hypothesis that vocal characteristics of manatee populations exhibit geographic variability. Specifically, we predict divergent vocalizations between Florida and Antillean subspecies while expecting similarity within Antillean populations in accordance with their geographic proximity. To do this, we tested for differences in the acoustic parameters of manatee calls between three different geographic locations which include two subspecies of West Indian manatees (Florida and Antillean) and two regional populations of Antillean manatees.
Materials and methods
Sampling and study sites
Wild manatees were recorded in Florida, Belize, and Panama (Fig. 1). Florida manatees were recorded at Blue Springs, Florida (hereafter “Florida”), a warm, nearly crystal-clear freshwater refuge utilized by manatees when ambient water temperature drops below 20 °C in the winter. Recordings were made on 03 and 04 January 2010 and were obtained using an omnidirectional SQ26-08 hydrophone (linear frequency range: 0.02–50 kHz; sensitivity: − 169 dB re 1 V/µPa, 48 kHz) attached to a M–Audio MicroTrack 24/96 recorder (48 kHz sample rate; 16 bit). The hydrophone was suspended at a depth of 1 m below the surface in water that was 1.5 m deep.
Maps of the Florida and Antillean species range and locations where vocalizations were recorded. Florida manatees (T. m. latirostris) were recorded with a drop-in hydrophone at Blue Springs in Florida. In Belize, Antillean manatees (T. m. manatus) were recorded with stationary recorders deployed in two different resting holes near St. George’s Caye. Similarly, Antillean manatees in Panama were recorded with stationary recorders deployed in four locations throughout the Río Changuinola. Maps were generated in QGIS v. 3.28.1 using Google Satellite Imagery (URL: https://mt1.google.com/vt/lyrs=s&x={x}&y={y}&z={z}).
Vocalizations of Antillean manatees in Belize were recorded at St. George’s Caye, a small crescent-shaped island located 9.5 km east of mainland Belize and 2.5 km west of the Belize Barrier Reef (hereafter “Belize”). The site is surrounded by expansive seagrass flats and sand patches with an average depth of 2 m, deep channels (2–3 m deep), and deep depressions (2–4 m) used by manatees as resting holes (i.e., depressions of high use for Belizean manatees50). Outside of the deep channels and resting holes, water visibility is clear from the surface to the seabed. The area is regularly inhabited by manatees of all ages and sex classes51. Recordings were made during one week in July 2017 and 2.5 weeks in January 2018 using a SoundTrap 300 HF (Ocean Instruments, New Zealand). The recorder sampled at a frequency of 288 kHz (16-bit, flat frequency response: 0.02–150 kHz ± 2 dB, clip level: 172 dB re: 1 µPa) with the preamplifier gain on. The device was anchored to the seafloor in a seagrass bed adjacent to a resting hole with a cinderblock and suspended in the water column at a depth of 1 m above the seafloor in water 1.5 m deep.
In Panama, Antillean manatees were recorded in the Changuinola River (sites S2 and S4, see more information in Merchan et al., 201952), which consists of sinuous, narrow (< 20 m) brackish channels with low visibility and abundant surface and subaquatic vegetation (hereafter “Panama”). Recordings were made from April to May each year from 2015 to 2018 using Wildlife Acoustics Song Meter SM3 Marine bioacoustics programmable hydrophones (Maynard, MA, USA). Hydrophones were installed on PVC placed 1 m above the river floor at 2–3 m depth. The recorder was set up to record 2 min clips every 10 min with sampling frequency set at 96 kHz (32 bits), with the gain set at 35 dB and the sensitivity at 12 dB re: 1 V/µPa (frequency response: 0.02–192 kHz ± 5 dB; dynamic range: 81–165 db SPL).
The numbers of manatees during recording periods at each location was estimated. In Florida, manatees were counted (n = 120 animals) during a visual survey at the beginning of the recording session. The number of wild Antillean manatees in recordings from Belize ranged from 17–60 individuals, which were observed during drone flights at St. George’s Caye from 2017 to 201853,54. Between 1 and 15 manatees were present at a time in the recording area. Wild manatees in Panama were estimated at about 45–48 individuals, according to the vocalization clustering method used in Merchan et al. (2019)52.
Recording procedures for Florida manatees were approved by the US Fish and Wildlife Service LOA #63658B. Data collection in Belize was conducted under permits granted to E.A.R. from the Belize Fisheries Department (Ref. No. 000031-17, 000010-15). For Panama, the Government of Panama and the Ministerio de Ambiente provided research permits for accessing the river and the protected area. The Animal Care and Use Committee of the Smithsonian Tropical Research Institute (STRI) approved all procedures used in the work. All methods were performed in accordance with relevant guidelines and regulations as suggested in the ARRIVE guidelines.
Acoustic call selection
Analysis included a total of 15 h recordings from Florida, 116 h from Belize, and 400 h from Panama. Recordings from Florida and Belize were visually inspected in spectrograms, and all calls were manually selected using Raven 1.5 software55. Calls from Panama were obtained through automated detection (for further details see Merchan et al. 201952). This resulted in an initial pool of 11,328 vocalizations from Florida, 3262 vocalizations from Belize, and 4819 vocalizations from Panama. Underwater sounds produced close to the surface are subject to the physical limitations of the Lloyd Mirror Effect37. Interference patterns can result in a loss or attenuation of lower frequencies56,57 which may result in the loss of the fundamental frequency and the second harmonic being incorrectly distinguished as the fundamental. For this reason, we measured the harmonic interval of calls to ensure they corresponded with the measured fundamental frequency, and excluded calls that did not have a fundamental frequency. From this set, non-overlapping calls with a signal-to-noise ratio (SNR) ≥ 6 dB, with clear and identifiable parameters were selected. This left us with a pool of 1,691 vocalizations from Florida, 377 vocalizations from Belize, and 810 vocalizations from Panama for further analysis.
Remaining calls were visually categorized into five categories: squeaks, high squeaks, squeals, squeak-squeals and chirps based on the classification scheme of Brady et al. 202011 (Fig. 2).
Spectrograms of representative samples of visually classified call types as defined by Brady et al. 202042.
Squeaks and high squeaks were differentiated based on the shape of the call contour. Squeals had deterministic chaos throughout the call. Squeak-squeals contained elements of squeaks and squeals and chirps contained frequency jumps. Calls were visually classified by two independent reviewers. If they disagreed on classification, a third skilled reviewer made the final decision of call type. Classification of call types by location is in Table 1.
To be comparable with previous studies on measured characteristics of West Indian manatee vocalizations11,41,43,46, six parameters were characterized from the fundamental frequency (the first harmonic) of each vocalization: Minimum and Maximum Frequency (Hz) (measured from the power spectrum with a 10 dB threshold); Duration (s) (measured from the waveform); Peak Frequency (Hz) (frequency at which peak power occurs within the selection58); Bandwidth (Hz) (value of the maximum frequency minus the minimum frequency); and Center Frequency (Hz) (the frequency that divides the sound into two frequency intervals of equal energy58). Minimum and maximum frequency were calculated by generating the average power spectrum of the entire measurement box around each call with a 10 dB threshold: minimum was the point of intersection on the left of the plots and maximum was the point of intersection on the right (see Fig. 1 in Brady et al.46). Spectrograms for recordings from Florida were calculated with a time resolution of 5.33 ms and a frequency resolution of 46.8 Hz (DFT: 1024; Hanning window; 50% overlap). Belize and Panama recordings were downsampled to 48 kHz, then calculated with a time resolution of 5.33 ms and a frequency resolution of 46.8 Hz (DFT: 1024; Hanning window; 50% overlap).
Statistical analysis
PERMANOVA and SIMPER analysis
Given the results of the classification analyses, our statistical analysis focused on comparing call types with sufficient sample sizes in more than one location. Squeaks and high squeaks were compared across all three locations and squeals were compared between Belize and Florida. To identify differences in call types across populations (Florida, Belize, and Panama), the permutational analysis of variance (PERMANOVA) was run. PERMANOVA is a non-parametric, multivariate test commonly used to compare groups59,60. PERMANOVA has been applied in comparisons of vocal characteristics between groups of dolphins61 and dolphin vocalizations in different locations62. Since one animal may have contributed more than one call to the sample in each location, we ran a repeated measures PERMANOVA. To have equal sample sizes within each call category, calls were randomly selected from larger sample sizes to equal the lowest sample size in the group. For example, we randomly selected 203 squeaks from Florida and 203 squeaks from Panama to equal the lowest sample size of squeaks from Belize (N = 203). For high squeaks, we measured 81 calls from each location and 87 squeals were measured from Belize and Florida. The PERMANOVA assumption of homogeneity was assessed for each call category. The only call category that violated the assumption was squeals. However, PERMANOVA is robust to heterogeneity for balanced designs (i.e. equal sample size)63 so we continued the analysis with squeals. For each call category (squeaks, high squeaks and squeals), the PERMANOVA was performed with 999 permutations with a Bray–Curtis dissimilarity. A Dunn’s post hoc for each measured variable was conducted for groups that showed significant differences from the results of the PERMANOVA. Lastly, after z transforming the data, we applied a SIMPER analysis technique, which provides the percentage contribution of each variable that differentiates between groups64. Statistical analyses were performed using the software PAST 4.1265.
Stepwise discriminant function analysis
In addition to the SIMPER analysis, stepwise discriminant function analyses (SDFA) were used to determine which parameters were the most important in determining differences between geographic locations66. Although most frequency variables were highly correlated (Supplementary Table 1), all variables were retained to minimize loss of information67. A leave-one-out cross-validation discriminant analysis was used to predict how the model would perform on new observations. Prior probabilities were obtained from the number of vocalizations used in analysis for each call type. Statistical analyses were performed in SPSS version 2768.
Results
All vocalizations recorded from the three populations shared fundamental frequency ranges between 0.5–5.0 kHz and 0.05–0.6 s in duration. Representative samples of calls from each of the three populations are shown in Fig. 3. Means and standard deviation for all variables measured for squeaks, high squeaks, and squeals are in Supplementary Table 2. Violin plots for duration, center and peak frequency were generated in R using ggplot269 and are shown in Fig. 4.
Spectrograms of the vocalizations of the two subspecies of the West Indian manatee recorded in Belize, Panama, and Florida. Recordings of the Antillean manatee were made in Belize and Panama and recordings of the Florida manatee were gathered in Florida. Spectrogram parameters: DFT: 1024; Hanning window; 50% overlap; time resolution: 5.33 ms; frequency resolution: 46.8 Hz.
Violin plots and boxplots illustrating the values of three different acoustic parameters (center frequency, maximum frequency, and duration), compared across the three geographically distinct manatee populations. The boxplot represents the upper and lower quartile values. The black line represents the median and the red dot represents the mean. Plots for Antillean manatees are in blue and Florida manatees in red.
PERMANOVA and SIMPER analysis
The repeated measures PERMANOVA analysis indicated there were significant differences between characteristics of high squeaks (Pseudo F 252 = 43.93) and squeaks (Pseudo F 606 = 13.73) among the three populations (p < 0.05). Dunn's post hoc test (Supplementary Table 3) for squeaks showed significant differences for all measured variables (p < 0.05) between each location. For high squeaks, Dunn's post hoc test revealed no significant differences for center frequency (p = 0.127), peak frequency (p = 0.166) and minimum frequency (p = 0.520) between Florida and Panama. Belize squeaks and high squeaks were shorter in duration and higher in frequency than those from Florida and Panama. Panama squeaks and high squeaks were longer in duration than Belize and Florida. For squeals, the PERMANOVA results (Pseudo-F172 = 26.75) indicated statistical differences (p < 0.01) between Belize and Florida. Dunn’s post hoc test results indicated all measured acoustic parameters of squeals were statistically different (p < 0.05) between Belize and Florida. Belize squeals were lower in frequency and shorter in duration than squeals from Florida.
The SIMPER analysis indicated almost all frequency and temporal parameters contributed to the differences observed between and within populations. Duration contributed the most to the dissimilarity between Belize-Panama for squeaks and high squeaks as well as squeaks from Panama-Florida. For high squeaks, bandwidth, maximum and peak frequency contributed the most to the dissimilarity observed between the Florida and Antillean subspecies. Center frequency followed by peak and maximum frequency contributed the most to the dissimilarity observed between Belize and Florida for squeals (Table 2).
Stepwise discriminant function analysis
The results of the SDFA for each call type are provided in Supplementary Tables 4–6, including the standardized canonical coefficients, eigenvalues, wilks lambda, chi-square, degrees of freedom, cumulative variance, and p-values. For squeaks and high squeaks, there were two significant discriminant functions (p < 0.001). Two parameters described 100% of the variance in squeaks. Duration was the most important in determining differences between geographic locations followed by bandwidth. For high squeaks, three parameters described 100% of the variance which included (in order of importance): center frequency, maximum frequency, and duration. Squeal results indicated one significant discriminant function (p < 0.001) and the most important parameter was maximum frequency, followed by minimum frequency and duration. The leave-one-out classification analysis classified 60.9% of squeaks, 73.3% of high squeaks, and 72.4% of squeals to the correct geographic location (Table 3). All correct classification scores were greater than expected by chance (squeaks and high squeaks: 1/3 = 33%; squeals: ½ = 50%).
Discussion
Our study revealed statistically significant differences in acoustic parameters of classified call types between the Florida and Antillean subspecies at the geographic scale as well as within subspecies (Belize and Panama). Both the SIMPER analysis and SDFA suggest duration plays a significant role in distinguishing between Antillean populations for squeaks and high squeaks, as well as between Florida and Antillean subspecies for squeaks. Additionally, they concur that maximum frequency is a critical parameter in differentiating between Antillean populations for squeals and between Florida and Antillean subspecies for high squeaks. The variances in parameter importance between the two analyses can be attributed to differences in their methodologies. The SIMPER analysis computes pairwise Bray–Curtis dissimilarity within and between groups64 whereas SDFA utilizes a stepwise variable selection procedure to iteratively include or exclude one variable at a time until the optimal discriminant function is achieved66. Furthermore, previous research has also suggested differences in vocal parameters between populations of the Antillean subspecies but were not statistically analyzed. For example, Ramos et al. (2020)13 found that the mean fundamental frequency of calls produced by Antillean manatees in Belize was lower than those reported for other populations from Central and South America, but higher than the calls reported from Puerto Rico. High squeaks are produced by both Antillean and Florida manatee calves, while adults produce squeaks37,45,46. This study indicates differences in acoustic parameters between adults (squeaks) and calves (high squeaks) between geographic locations. Alternative explanations of these differences may be due to differences between subspecies, geographic separation, differences in sex, current and/or historical connections between populations, habitat, and the number of manatees recorded.
Differences in the acoustic parameters of manatee vocalizations could be due to geographic separation within and across subspecies, and restricted gene flow between manatee populations70,71,72. Geographic isolation contributes to local adaptation to different environments associated with phenotypic and genetic divergence that can influence vocal characteristics in closely related species73. Genetic differences found between the Antillean and Florida manatees support their subspecies classification19 and their different biogeographic groupings24,74. Variation in vocal characteristics between Florida and Antillean manatees are expected based on geographic distance. Yet, this study observed higher dissimilarity scores, and showed a gradient in certain acoustic parameters for high squeaks and squeaks between Antillean manatee populations. Therefore, geographic distance alone cannot explain the variation observed. One explanation could be related to body size. Body condition analysis of 380 wild caught Antillean manatees from Puerto Rico, Cuba, Mexico, Belize, Colombia, and Brazil revealed two distinct ecotypes that were differentiated based on body size75. Further, Florida manatees are generally larger than Antillean manatees76,77,78. Our study showed high squeaks and squeaks from Florida manatees were lower in frequency than Belize and Panama. Differences in body size can influence frequency and temporal parameters79 and therefore, may be a source of the variation observed in this study.
Differences in the number of manatees we recorded at each site could explain some of the variation detected in their vocalizations across populations. Recordings in Florida involved substantially more manatees at a single time (n = 120 individuals) than recordings gathered over 3.5 weeks in Belize (range: 17–60 individuals53) and Panama (range: 45–48 manatees52). The number of individuals recorded in the area can influence vocal complexity, which in turn can influence the size of the vocal repertoire or diversity of call types produced80. For example, chickadees produced a greater diversity of note types as the number of individuals in the group increased81. Brady (2020)82 observed 171 aggregations of manatees that ranged in size from 1 to 20 animals. They found the occurrence of call types (specifically chirps and squeak-squeals) increased with the number of animals. In this study, we could only estimate how many individuals were present during the recording time frame and it is unknown if all animals present were vocalizing. Although, it is probable that the fewer number of animals estimated in Panama and Belize influenced the lack of chirps and squeak-squeals recorded from these locations.
Maximum, peak and minimum frequency contributed to the variation between locations for each call type which could also suggest sex differences influence variation in calls. In captive environments, where most studies of manatee communication have been conducted, previous work has shown that temporal and frequency parameters differ with the sex of manatees42,43. Several studies have reported that females have higher values of maximum and minimum fundamental frequency, but lower mean peak frequencies with respect to males43,83. The unknown distribution of sexes in the populations recorded in this study cannot be ruled out as a source of variation found in the three sites and further research is needed to examine these factors.
Although the characteristics of each of the habitats in this study were not measured, differences between habitats could have influenced the observed variation. Studies of geographic variability in the vocalizations of some marine mammal species suggest that different environmental conditions (e.g., water depth, sediment type, pH, salinity, and temperature) can influence regional variation in vocal parameters84,85,86. While all three recording sites were shallow water habitats, Florida and Belize manatees were recorded from fresh and salt water environments respectively. Animals have also been observed to adjust vocal parameters of their calls due to environmental factors varying within their habitats. For example, Amazon river dolphins (Inia geoffrensis) altered the duration, center frequency, maximum frequency, and bandwidth of vocalizations depending on whether they were in clear or rich-sediment water87. For this study, almost all the previously mentioned parameters were different between geographic locations for squeaks and high squeaks. Panama’s riverine habitat is different from recording sites in Belize and Florida as it includes four different acoustic subhabitats88. Acoustic parameters of vocalizations in Panama may differ as an adaptation to communication in diverse habitats with different propagation characteristics.
Ruling out individual variation in the current study is difficult due to the inability to reliably localize to the calling animal in all locations. Localization to an individual can be difficult to ascertain in marine mammals due to the number of animals vocalizing (e.g., in spinner dolphins [Stenella longirostris]89), animals vocalizing in close proximity to one another (harbor seals90), and an inability to observe the animal being recorded (sperm whales [Physeter macrocephalus]91 and bearded seals85). Even with the inability to localize to individual animals, the previously mentioned studies still observed variation in vocalizations between geographic locations. The data from this study indicates that the quartile ranges for the measured variables are similar across all locations, despite differences in the number of calls or individuals. Standard errors of variables measured from Florida are lower than those in the other sites, despite the larger sample size and population. These findings suggest that other factors, such as the environment, may be playing a role in the variation of call characteristics. Further research is necessary to better understand the factors contributing to these variations.
Identifying how the acoustic characteristics of manatee vocalizations vary geographically is critical to improving our understanding of their activity in different populations and habitats. For example, their detection in large datasets gathered with stationary passive acoustic recorders can provide insights into temporal trends in their occurrence and behavior92. The range of frequencies observed in this study are well within the hearing range of manatees56,93, which suggests that differences in frequency could still be perceived by conspecifics. Recent studies have observed a hybrid species between Amazonian and Antillean manatees94 in which Amazonian vocalizations are higher in frequency than Antillean species63. This suggests that vocal differences may not be a barrier to mating, as observed, for example, in sperm whales95. Considering that Florida manatees have been more commonly observed in Antillean manatee habitats (e.g.,32,33), it is unlikely that variation in vocal characteristics would be a barrier to communication.
Conclusions
Our study provides novel insights into the differences in vocal characteristics of the West Indian manatee, which may be due to geographic separation. Variability of manatee vocalizations between isolated groups and subspecies highlights the importance of studying their acoustic behavior in different regions. Generalizing our findings to other manatee populations is limited by the ecosystem characteristics. More research efforts are needed to investigate the influence of habitat, addressing one of the main limitations of the present study. Continued research on the variation in acoustic characteristics of manatee vocalizations in Florida and the Western Caribbean will facilitate improved acoustic monitoring of manatees and understanding of how their vocal behavior is shaped by environmental and evolutionary factors.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon request.
References
Krebs, J. R. & Kroodsma, D. E. Repertoires and Geographical Variation in Bird Song. In Advances in the Study of Behavior 143–177 (Academic Press, 1980).
Marler, P. Specifc distinctiveness in the communication signals of birds. Behaviour 11, 13–39 (1957).
Ey, E. & Fisher, J. The ‘Acoustic adaptations hypothesis’ a review of the evidence from birds, anurans and mammals. Bioacoustics 19, 21–48 (2009).
Mundinger, P. C. Microgeographic and Macrogeographic Variation in the Acquired Vocalizations of Birds. In Acoustic Communication in Birds: Song Learning and Its Consequences (eds Kroodsma, D. et al.) 147–208 (Academic Press, UK, 1982).
McGregor, P. K., Peake, T. M. & Gilbert, G. Communication, Behaviour, and Conservation. In Behaviour and Conservation (eds Gosling, L. & Sutherland, W.) 261–285 (Cambridge University Press, 2000).
Van Parijs, S. M., Hastie, G. D. & Thompson, P. M. Geographical variation in temporal and spatial vocalization patterns of male harbour seals in the mating season. Anim. Behav. 58, 1231–1239 (1999).
Risch, D. et al. Vocalizations of male bearded seals, Erignathus barbatus: classification and geographical variation. Anim. Behav. 73, 747–762 (2007).
Samaran, F. et al. Seasonal and geographic variation of southern blue whale subspecies in the Indian Ocean. PLoS ONE 8, e71561 (2013).
Jones, G. J. & Sayigh, L. S. Geographic variation in rates of vocal production of free-ranging bottlenose dolphins. Mar. Mamm. Sci. 18, 374–393 (2002).
Umeed, R., Attademo, F. L. N. & Bezerra, B. The influence of age and sex on the vocal repertoire of the Antillean manatee (Trichechus manatus manatus) and their responses to call playback. Mar. Mamm. Sci. 34, 577–594 (2017).
Brady, B., Hedwig, D., Trygonis, V. & Gerstein, E. Classification of Florida manatee (Trichechus manatus latirostris) vocalizations. J. Acoust. Soc. Am. 147, 1597–1606 (2020).
Merchan, F., Guerra, A., Poveda, H., Guzmán, H. M. & Sanchez-Galan, J. E. “Bioacoustic classification of Antillean manatee vocalization spectrograms using deep convolutional neural networks. Appl. Sci. 10, 1–22 (2020).
Ramos, E. A. et al. The Antillean manatee produces broadband vocalizations with ultrasonic frequencies”. J. Acoust. Soc. Am. 147, 80–86 (2020).
Deutsch, C. J. Trichechus manatus ssp. Latirostris, North American manatee. IUCN Red List Threat. Species, pp. 1–6. (2008).
Marsh, H., O’Shea, T. J. & Reynolds, J. E. Ecology and Conservation of the Sirenia: Dugongs and Manatees 143–207 (Cambridge University Press, 2012).
Deutsch, C. J., Self-Sullivan, C. J., and Mignucci-Giannoni, A. Trichechus manatus, West Indian manatee. IUCN Red List Threat. Species, pp. 1–19. (2008).
Self-Sullivan, C., and Mignucci-Giannoni, A. Trichechus manatus ssp. Manatus, Antillean manatee. IUCN Red List Threat. Species, 1–3 pp. (2008).
Powell, J. A. & Rathbun, G. B. Distribution and abundance of manatees along the northern coast of the Gulf of Mexico. Gulf Mexico Sci. 7, 1–28 (1984).
Domning, D. P. & Hayek, L. A. Interspecific and intraspecific morphological variation in manatees (Sirenia: Trichechus). Mar. Mamm. Sci. 2, 87–144 (1986).
Laist, D. W. & Reynolds, J. E. Influence of power plants and other warm-water refuges on Florida manatees. Mar. Mamm. Sci. 21, 739–764 (2005).
Martin, J. et al. Combining information for monitoring at large spatial scales: first statewide abundance estimate of the Florida manatee. Biol. Conserv. 186, 44–51 (2015).
Castelblanco-Martínez, D. N., Nourisson, C., Quintana-Rizzo, E., Padilla-Saldivar, J. & Schmitter-Soto, J. J. Potential effects of human pressure and habitat fragmentation on population viability of the Antillean manatee Trichechus manatus manatus: a predictive model. Endanger. Species Res. 18, 129–145 (2012).
Self-Sullivan, C. & Mignucci-Giannoni, A. A. West Indian Manatees (Trichechus manatus) in the Wider Caribbean Region Sirenian Conservation: Issues and Strategies in Developing Countries 36–46 (University Press of Florida, 2012).
Vianna, J. A. et al. Phylogeography, phylogeny and hybridization in trichechid sirenians: Implications for manatee conservation. Mol. Ecol. 15, 433–447 (2006).
Hunter, M. E. et al. Low genetic variation, and evidence of limited dispersal in the regionally important Belize manatee. Anim. Conserv 13, 592–602 (2010).
Nourisson, C. et al. Evidence of two genetic clusters of manatees with low genetic diversity in Mexico and implications for their conservation. Genetica 139, 833–842 (2011).
Barros, H. M. et al. Cranial and chromosomal geographic variation in manatees (Mammalia: Sirenia: Trichechidae) with the description of the Antillean manatee karyotype in Brazil. J. Zool. Syst. Evol. Res. 55, 73–87 (2016).
Quintana-Rizzo, E., and Reynolds, J. III Regional management plan for the West Indian manatee (Trichechus manatus). United Nations Environmental Program. CEP Techn Rep No. 48. UNEP Caribbean Environment Programme, Kingston, Jamaica. (2010).
Guzman, H. M. & Condit, R. Abundance of manatees in Panama estimated from side-scan sonar. Wild. Soc. Bull. 41, 556–565 (2017).
Díaz-Ferguson, E., Hunter, M. & Guzmán, H. M. Genetic composition and connectivity of the Antillean manatee (Trichechus manatus manatus) in Panama. Aqu. Mamm. 43, 378–386 (2017).
Deutsch, C. J., Castelblanco-Martínez, D. N., Groom, R. & Cleguer, C. Movement Behavior of Manatees and Dugongs: I Environmental Challenges Drive Diversity in Migratory Patterns and Other Large-Scale Movements. In Ethology and Behavioral Ecology of Sirenia 155–231 (Springer, 2022).
Alvarez-Alemán, A., Beck, C. A. & Powell, J. A. First report of a Florida manatee (Trichechus manatus latirostris) in Cuba. Aqu. Mamm 36, 148 (2010).
Castelblanco-Martínez, D. N. et al. First documentation of long-distance travel by a Florida manatee to the Mexican Caribbean. Ethol. Ecol. & Evol. 34, 545–556 (2021).
Castelblanco-Martínez, D. N. et al. Movement patterns of Antillean manatees in Chetumal Bay (Mexico) and coastal Belize: A challenge for regional conservation. Mar. Mamm. Sci. 29, E166–E182 (2013).
Auil, N. E, Belize conservation programme ten year summary. Report by Wildlife Trust, 29 pp. (2007).
Gonzalez-Socoloske, D., Oliveira-Gómez, L. D. & Ford, R. E. Gentle giants in dark waters: Using side-scan sonar for manatee research. Open Rem. Sens. J. 5, 1–14 (2012).
O’Shea, T. J. & Poché, L. B. Aspects of underwater sound communication in Florida manatees (Trichechus manatus latirostris). J. Mamm. 87, 1061–1071 (2006).
Steel, C. Vocalization patterns and corresponding behavior of the West Indian manatee (Trichechus manatus). Ph.D. thesis, Florida Institute of Technology, (1982).
Hartman, D. Ecology and arbourr of the manatee (Trichechus manatus) in Florida. Am. Soc. Mamm. 5, 153 (1979).
Miksis-Olds, J. L. & Tyack, P. L. Manatee (Trichechus manatus) vocalization usage in relation to environmental noise levels. J. Acoust. Soc. Am. 125, 1806–1815 (2009).
Alicea-Pou, J. A. Vocalizations and behavior of Antillean and Florida manatee (Trichechus manatus): Individual variability and geographical comparison. M.Sc. thesis, San Francisco State University. (2001).
Umeed, R., Niemeyer Attademo, F. L. & Bezerra, B. The influence of age and sex on the vocal repertoire of the Antillean manatee (Trichechus manatus manatus) and their responses to call playback. Mar Mamm Sci 34, 577–594 (2018).
Sousa-Lima, R. S., Paglia, A. P. & Da Fonseca, G. A. B. Gender, age, and identity in the isolation calls of Antillean manatees (Trichechus manatus manatus). Aquat. Mamm. 34, 109–122 (2008).
Nowacek, D. P., Casper, B. M., Wells, R. S., Nowacek, S. M. & Mann, D. A. Intraspecific and geographic variation of West Indian manatee (Trichechus manatus spp.) vocalizations (L). J. Acoust. Soc. Am. 114, 66–69 (2003).
Brady, B., Moore, J. & Love, K. Behavior related vocalizations of the Florida manatee (Trichechus manatus latirostris). Mar. Mamm. Sci. 38, 975–989 (2022).
Brady, B. et al. Manatee calf call contour and acoustic structure varies by species and body size. Sci. Rep. 12, 19597 (2022).
Grossman, C. J., Hamilton, R. E. & De Wit, M. The vocalization mechanism of the Florida manatee (Trichechus manatus latirostris). OnLine J. Biol. Sci. 14, 127–149 (2014).
Landrau-Giovannetti, N., Mignucci-Giannoni, A. A. & Reidenberg, J. S. Acoustical and anatomical determination of sound production and transmission in West Indian (Trichechus manatus) and Amazonian (T. inunguis) manatees. Anat. Rec. 297, 1896–1907 (2014).
Barros, H. M. D. D. R. et al. Cranial and chromosomal geographic variation in manatees (Mammalia: Sirenia: Trichechidae) with the description of the Antillean manatee karyotype in Brazil. J. Zool. Syst. Evol. Res. 55, 73–87 (2016).
Bacchus, M. L. C., Dunbar, S. G. & Self-Sullivan, C. Characterization of resting holes and their use by the Antillean manatee (Trichechus manatus manatus) in the drowned cayes Belize. Aquat. Mamm. 35, 62–71 (2009).
Ramos, E. A. et al. Small drones: A tool to study, monitor, and manage free-ranging Antillean manatees in Belize and Mexico. Sirenews 67, 13–16 (2017).
Merchan, F., Echevers, G., Poveda, H., Sanchez-Galan, J. E. & Guzman, H. M. Detection and identification of manatee individual vocalizations in Panamanian wetlands using spectrogram clustering. J. Acoust. Soc. Am. 146, 1745–1757 (2019).
Landeo-Yauri, S. S., Ramos, E. A., Castelblanco-Martínez, D. N., Niño-Torres, C. A. & Searle, L. Using small drones to photo-identify Antillean manatees: A novel method for monitoring an endangered marine mammal in the Caribbean Sea. Endanger. Species Res. 41, 79–90 (2020).
Ramos, E. A., Maloney, B., Magnasco, M. O. & Reiss, D. Bottlenose dolphins and Antillean manatees respond to small multi-rotor unmanned aerial systems. Front. Mar. Sci. 5, 316 (2018).
Lisa, K. Yang Center for Conservation bioacoustics at the cornell lab of ornithology. Raven Pro: Interactive sound analysis software (Version 1.5) [Computer software]. Ithaca, NY: The Cornell lab of ornithology. Available from https://ravensoundsoftware.com/. (2022).
Gerstein, E. R., Gerstein, L., Forsythe, S. E. & Blue, J. E. The underwater audiogram of the West Indian manatee (Trichechus manatus). J. Acoust. Soc. Am. 105, 3575–3583 (1999).
Medwin, H. Sounds in the Sea (Cambridge University Press, 2005).
Charif, R. A., Waack, A. M., and Strickman, L. M. Raven Pro 1.4 User’s Manual. (The Cornell Lab of Ornithology, Ithaca, NY), 379 (2010).
Anderson, M. J. A new method for non-parametric multivariate analysis of variance. Aust. Ecol. 26, 32–46 (2001).
Anderson, M. J. Distance-based tests for homogeneity of multivariate dispersions. Biometrics 62, 245–253 (2006).
La Manna, G., Manghi, M., Pavan, G., Lo Mascolo, F. & Sarà, G. Behavioural strategy of common bottlenose dolphins (Tursiops arbourr) in response to different kinds of boats in the waters of Lampedusa Island (Italy). Aquat. Conserv. Mar. Freshw. Ecosyst. 23, 745–757 (2013).
Perez-Ortega, B., Daw, R., Paradee, B., Gimbrere, E. & May-Collado, L. J. Dolphin-watching boats affect whistle frequency modulation in bottlenose dolphins. Front. Mar. Sci. https://doi.org/10.3389/fmars.2021.618420 (2021).
Anderson, M. J. & Walsh, D. C. I. What null hypothesis are you testing? PERMANOVA, ANOSIM and the mantel test in the face of heterogeneous dispersions. Ecol. Monogr. 83, 557–574 (2013).
Clarke, K. R. Non-parametric multivariate analyses of changes in community structure. Aust. J. Ecol. 18, 117–143 (1993).
Hammer, Ø., Harper, D. A. T. & Ryan, P. D. PAST: paleontological statistics software package for education and data analysis [computer software]. Palaeontol. Electron. 4, 9 (2001).
Huberty, C. J. Discriminant analysis. Rev. Educ. Res. 45(4), 543–598 (1975).
Charlton, B. D. et al. Implications of vocalizations during giant panda breeding. J. Acoust. Soc. Am. 130, 2460–2460 (2011).
IBM Corp. Released 2020. IBM SPSS Statistics for Windows, Version 27.0. Armonk, NY: IBM Corp. (2020).
Wickham, H. ggplot2: Elegant graphics for data analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org. (2016).
Poole, J., Tyack, P., Stoeger-Horwath, A. & Watwood, S. Elephants are capable of vocal learning. Nature 434, 455–456 (2005).
Filatova, O. A. et al. Cultural evolution of killer whale calls: Background, mechanisms and consequences. Behaviour 152, 2001–2038 (2015).
Brakes, P. & Dall, S. R. X. Marine mammal arbourr: A review of conservation implications. Front. Mar. Sci. 3, 1–15 (2016).
Janik, V. M. & Slater, P. J. Traditions in Mammalian and Avian Vocal Communication. In The Biology of Traditions: Models and Evidence 213–235 (Cambridge university press, 2003).
García-Rodríguez, A. I. et al. Phylogeography of the West Indian manatee (Trichechus manatus): How many populations and how many taxa?. Mol. Ecol. 7, 1137–1149 (1998).
Castelblanco-Martínez, D. N. et al. Analysis of body condition indices reveals different ecotypes of the Antillean manatee. Sci. Rep. 11, 1–14 (2021).
Wong, A. W. et al. Monitoring oral temperature, heart rate, and respiration rate of West Indian manatees (Trichechus manatus) during capture and handling in the field. Aquat. Mamm. 38, 1–16. https://doi.org/10.1578/AM.38.1.2012.1 (2012).
Bonde, R. K. et al. Biomedical health assessments of the Florida manatee in Crystal River – providing opportunities for training during the capture, handling, and processing of this endangered aquatic mammal. J. Mar. Anim. Ecol. 5, 17–28 (2012).
Converse, L. J., Fernandes, P. J., Macwilliams, P. S. & Bossart, G. D. Hematology, serum chemistry, and morphometric reference values for Antillean Manatees (Trichechus manatus manatus). J. Zoo Wildl. Med. 25, 423–431 (1994).
Taylor, A. M., Charlton, B. D. & Reby, D. Vocal Production by Terrestrial Mammals: Source, Filter, and Function. In Vertebrate Sound Production and Acoustic Communication (ed. Suthers, R. S.) 229–259 (Springer International Publishing, 2016).
Pollard, K. A. & Blumstein, D. T. Evolving communicative complexity: Insights from rodents and beyond. Philos. Trans. R. Soc. Lond. B Biol. Sci. 367, 1869–1878 (2012).
Freeberg, T. M. Social complexity can drive vocal complexity: Group size influences vocal information in Carolina Chickadees. Psychol. Sci. 17, 557–561 (2006).
Brady, B. A. Call Categorization and Vocal Behavior of the Florida Manatee (Trichechus manatus latirostris). Doctoral dissertation, Florida Atlantic University. (2020).
Sousa-Lima, R. S., Paglia, A. P. & Da Fonseca, G. A. B. Signature information and individual recognition in the isolation calls of Amazonian manatees, Trichechus inunguis (Mammalia: Sirenia). Anim. Behav. 63, 301–310 (2002).
Morisaka, T., Shinohara, M., Nakahara, F. & Akamatsu, T. Effects of ambient noise on the whistles of Indo-pacific bottlenose dolphin populations. J. Mamm. 86, 541–546 (2005).
Risch, D. et al. Vocalizations of male bearded seals, Erignathus barbatus: classification and geographical variation. Anim. Behav. 73, 747–762 (2007).
Baron, S. C., Martinez, A., Garrison, L. P. & Keith, E. Differences in acoustic signals from Delphinids in the western North Atlantic and northern Gulf of Mexico. Mar. Mammal Sci. 24, 42–56 (2007).
Amorim, T. O. S., Andriolo, A., Reis, S. S. & dos Santos, M. E. Vocalizations of amazon river dolphins (Inia geoffrensis): Characterization, effect of physical environment and differences between populations. J. Acoust. Soc. Am. 139, 1285–1293 (2016).
Rivera Chavarría, M., Castro, J. & Camacho, A. ’The relationship between acoustic habitat, hearing and tonal vocalizations in the Antillean manatee (Trichechus manatus manatus, Linnaeus, 1758). Biol. Open 4, 1237–1242 (2015).
Bazúa-Durán, C. & Au, W. W. Geographic variations in the whistles of spinner dolphins (Stenella longirostris) of the Main Hawaiian Islands. J. Acoust. Soc. Am. 116, 3757–3769 (2004).
Bjørgesæter, A., Ugland, K. I. & Bjørge, A. Geographic variation and acoustic structure of the underwater vocalization of arbour seal (Phoca vitulina) in Norway, Sweden and Scotland. J. Acoust. Soc. Am. 116, 2459–2468 (2004).
Amano, M., Kourogi, A., Aoki, K., Yoshioka, M. & Mori, K. Differences in sperm whale codas between two waters off Japan: possible geographic separation of vocal clans. J. Mamm. 95, 169–175 (2014).
Rycyk, A. M., Berchem, C. & Marques, T. A. Estimating Florida manatee (Trichechus manatus latirostris) abundance using passive acoustic methods. J. Acoust. Soc. Am. Express Lett. 2, 051202 (2022).
Gaspard, J., Bauer, G., Reep, R., Dziuk, K. & Mann, D. Audiogram and auditory critical rations of two Florida manatees Trichechus manatus latirostris. J. Exp. Biol. 215, 1442–1447 (2012).
Luna, F. O. et al. Genetic connectivity of the west Indian manatee in the southern range and limited evidence of hybridization with Amazonian manatees. Front. Mar. Sci. 7, 574455 (2021).
Weilgart, L. & Whitehead, H. Group-specific dialects and geographical variation in coda repertoire in South Pacific sperm whales. Behav. Ecol. Sociobiol. 40, 277–285 (1997).
Acknowledgements
This work was the product of a Master thesis of JDR-A who received a CONACYT scholarship. We thank Dr. Roberto Sosa-López and his students Victor Vargas and Marcos Quiroz Oliva (Laboratory of Bioacoustics and Animal Behavior of CIIDIR-Oaxaca) for support in data processing. The authors are thankful for the continued support of manatee research by the Belize Fisheries Department, Forest Department and St George’s Caye Village Council. Funding for recordings in Panama was provided by AES- Changuinola, S.R.L., the Smithsonian Tropical Research Institute (STRI), Gas Natural Fenosa Panama (now Naturgy Panama), and the Secretaria Nacional de Ciencia, Tecnología e Innovación de Panamá (SENACYT), through the contract FID18-076. F.M., H.P. and J.E.S.-G. acknowledge administrative support provided by CEMCIT-AIP and Universidad Tecnológica de Panamá. We thank AES engineers Jorge E. Da Silva, Glaister Tejada, Juan Carlos Brito, and Rodolfo Ayarza and Carlos A. Guevara of STRI and Alexis Montenegro and Alfredo Caballero of AAMVECONA for their support and cooperation during the project. The Sistema Nacional de Investigación (SNI), SENACYT-Panama supports research activities by F.M., H.P., J. E.S.-G., and H.M.G. Thanks to Joe Arena for assistance with graphics.
Author information
Authors and Affiliations
Contributions
J.D.R.-A., E.A.R., D.N.C.-M., B.B., and Y.H. conceived of the presented idea. Y.H., D.N.C.-M., G.P.-L., B.B. and E.A.R. were involved in planning and supervising the work of JDR-A. BB provided manatee recordings from Florida; E.A.R., K.A.C., M.M.-M., and L.S. from Belize; and H.G. and F.M. from Panama. J.D.R.-A. and B.B. visually inspected the recordings and developed quantitative measurements. J.D.R.-A., Y.H. and B.B. performed the numerical calculations and statistical analysis. F.M., K.C., J.S.G. and B.B. developed the machine learning analysis. J.D.R.-A., B.B., E.A.R., Y.H., D.N.C.-M. and M.M.-M. contributed to the interpretation of the results. J.D.R.-A., B.B., and E.A.R. took the lead in writing the manuscript. E.A.R. and JDR-A prepared graphic material. All authors provided critical feedback and helped shape the research, analysis and manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Reyes-Arias, J.D., Brady, B., Ramos, E.A. et al. Vocalizations of wild West Indian manatee vary across subspecies and geographic location. Sci Rep 13, 11028 (2023). https://doi.org/10.1038/s41598-023-37882-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-37882-8
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
