Differences in distress: Variance and production of American Crocodile (Crocodylus acutus) distress calls in Belize

Abstract Acoustic communication of American Crocodiles (Crocodylus acutus) is relatively understudied. Our overall aim was to determine the acoustic structure of wild American Crocodile distress calls, distinguish call differences among size classes (hatchling, juvenile, sub‐adult, and adult), and investigate call production on a gradient of human disturbance. American Crocodile distress calls have strong frequency modulation and are comprised of multiple harmonics in a downsweeping pattern. Measured parameters (total duration, first quartile duration, maximal frequency, first quartile frequency, end frequency, slope of first quartile, slope of last quartiles) differed significantly among size classes (p < .05). Hatchling distress calls are higher in frequency and strongly modulated, whereas calls produced by sub‐adults and adults showed little modulation, are lower in frequency, and have greater overall duration. Proportion of crocodiles that produced distress calls during capture differed by size class and sampling location, particularly adult distress calls which are reported here to be produced with undocumented frequency. We determined that American Crocodiles of all size classes produce distress calls at varying rates among study sites. Our results demonstrate that American crocodiles produce distress call more frequently at sites with higher anthropogenic activity. Measured call parameters of juveniles and hatchling American crocodiles also varied among sites in relation to human disturbance. Calls recorded at sites of high anthropogenic impact have increased duration and less modulation which may adversely affect response to emitted distress calls. Proportional and call parameter variances suggest anthropogenic activity as a driver for increased call production and alteration of call parameters at high human‐impacted sites.


| INTRODUC TI ON
Social interactions between crocodylians are facilitated through a variety of acoustic and physical signals. Of the acoustic signals produced, distress calls are perhaps the most frequently recorded call. Distress calls are repetitive chirps with multiple harmonics and frequency modulation (Garrick & Lang, 1977;Vergne, Pritz, & Mathevon, 2009).
The distress call is commonly produced by juveniles as a warning to other crocodylians and also elicits a defense response from conspecifics (Britton, 2001;Campbell, 1973;Chabert et al., 2015;Herzog & Burghardt, 1977;Vergne, Aubin, Taylor, & Mathevon, 2011;Vergne et al., 2009). Production of this behaviorally significant call decreases as the size of the crocodylian increases in conjunction with reduced risk of predation (Staton, 1978). Although reported, production of distress calls by sub-adult and adult crocodylians is rare (Staton, 1978).
There is a paucity of data regarding the structure, production, and significance of distress calls among crocodylians, such as the American Crocodile (Crocodylus acutus), and the exploration is still in its infancy.
American Crocodiles are the most widely distributed crocodylian in the New World, inhabiting coastal and lowland wetlands from southern Florida, USA, to northern South America (Ernst, Ross, & Ross, 1999;Platt & Thorbjarnarson, 2000). Although now a protected species in all of its range, previous exploitation in the mid-1900s led to extirpation and declines of the American Crocodile in many areas. Populations drastically decreased in Belize between 1930 and 1970 (Platt & Thorbjarnarson, 2000;Rainwater & Platt, 2009). Although the commercial hunting that caused population declines has ceased, new threats have emerged putting recovering populations at risk. Degradation and diminishment of habitat as a result of widespread pollution and coastal development are significant threats to the American Crocodile in Belize (Platt, Rainwater, & Nichols, 2004;Platt & Thorbjarnarson, 2000;Rainwater, 2008). Ongoing research in various crocodilyan habitats of Belize is providing increasing evidence of the connection between environmental toxicity and mortality and/or morbidity currently observed in crocodiles (Tellez, unpubl. data). It is likely, however, that there are also ecological consequences of human influence on American Crocodiles in Belize ( Figure 1). Additionally, there has been little dedicated study of American Crocodile bioacoustics anywhere in its range. Increasingly, anthropogenic noise and impact in wildlife habitat is demonstrated to have deleterious effects on wildlife acoustics and ecology (Blickley & Patricelli, 2010;Hildebrand, 2005;Laiolo, 2010). As an acoustically communicative species, it is feasible that American Crocodile sound production may be affected by anthropogenic disturbance. Herein we describe the acoustic structure of American Crocodile distress calls and explore a possible link between distress call production and anthropogenic impact.

| Study areas
Belize is a small Mesoamerican nation (22,965 km 2 ) paralleled by the Mesoamerican Barrier Reef and located south of Mexico on the eastern seaboard ( Figure 2). We collected acoustic recordings in the coastal zone of Belize at Ambergris Caye, Caye Caulker, and Belize Aquaculture Limited west of the Placencia Peninsula.
Ambergris Caye is the southern extension of the Mexican Yucatan Peninsula separated from Mexico by a small channel. We collected American Crocodile recordings throughout the southern portion of the island in proximity to the highly developed San Pedro town center. We recorded captured crocodiles throughout the study area and in particular from the Coco Beach Resort lagoon, San Pedro sewage treatment ponds, and the Mahogany Bay Village development site.
Our second study area, Caye Caulker, is an offshore island located 8 km south of Ambergris Caye. American Crocodiles are distributed along the entire island and we attained recordings opportunistically throughout the study area during concurrent population surveys.
Belize Aquaculture Limited (BAL), our final study area, is a large commercial shrimp farm (3,642 ha) located approximately 8.3 km east of the northern tip of the Placencia peninsula. We recorded crocodiles captured in the effluent polishing lagoon and canal system located on the margins of the farming operation.

| Wetland impact assessment
Concurrent to the collection of behavioral observations, we assessed study sites using a 6-point scale (Table 1) to characterize human disturbance in the wetlands (Maine Department of Environmental Protection, 2013). Our assessments scored presence and severity of hydrologic modifications to the wetland, vegetative modifications to the wetland, evidence of pollutants, and acoustic disturbance and human contact. We used rankings of each category of stressor to determine an overall level of human disturbance for each study site.

| Capture techniques
As distress calls are produced when crocodiles are under duress or perceived threat, we collected recordings during capture of wild F I G U R E 1 An adult American Crocodile photographed at one of the highly human-impacted study sites sampled for acoustic recording on Ambergris Caye, Belize F I G U R E 2 Belize, Central America, located on the eastern Caribbean seaboard. Letters dictate study area location; (A) Ambergris Caye, (B) Caye Caulker, and (C) Belize Aquaculture Limited American Crocodiles. Crocodiles were captured as part of ongoing behavioral research and population surveys. The West Virginia University Animal Care and Use Committee  and the Belize Forest Department (BFD) approved the capture protocol. The lead author acquired research permits from the BFD prior to initiating any capture or recording (Ref. No. CD/60/3/15(45)). We captured and restrained crocodiles using conventional techniques detailed by Webb and Messel (1977). We captured smaller crocodiles by hand and larger animals by noosing or a treble hook. We did not secure the snout of crocodiles prior to recording to avoid altering distress call emission. Following the collection of recordings, or if the crocodile ceased to vocalize, we secured the jaws using electrical tape and applied eye coverings to reduce capture stress. We collected morphometric measurements from all captured crocodiles following the protocol of Webb and Messel (1978). We used the total length (TL) to classify captured crocodiles as hatchlings (TL < 35 cm), juveniles (TL = 36-90 cm), sub-adults (TL = 91-180 cm), or adults (TL > 181 cm). Following the completion of health assessment and morphometric measurements, we released crocodiles at site of capture.

| Sound analysis
We performed acoustic analysis to determine the structure of distress calls for each size class. We analyzed five calls per individual, in one case only 3 calls were analyzed due to heavy background noise, and measured seven acoustic variables, two temporal and five spectral, using Raven Pro 1.5 acoustic analysis software (Bioacoustics Research Program, 2014). We used spectrographic analysis (window size 1,024, overlap 80%) of the fundamental frequency to determine maximal frequency (F max , Hz), frequency at end of first quartile (F 1/4 , Hz), and final frequency (F end , Hz; Figure 3a). Using call oscillograms, we measured temporal properties for total duration (DT, s) and duration of the first quartile (D 1/4 , s; Figure 3b). Measurement windows were drawn around the fundamental frequency and the maximal frequency values at the beginning, end, and first quartile were recorded. We used frequency and temporal measurements to calculate call modulation of the first temporal quartile slope (Slope 1, Hz/s, calculated as (F 1/4 − F max )/D 1/4 ), and the slope of the remaining three temporal quartiles (Slope 2, Hz/s, calculated as (F end − F 1/4 )/ (Vergne, Aubin, Martin, & Mathevon, 2012). Concurrent to call measurements, we recorded number of calls produced by each individual for 10-, 20-, and 30-s intervals as total recording time varied between individual crocodiles. We began call counts at the first recorded call for each individual. We used size designation to organize and analyze distress call recordings by overall size class.
We performed statistical analyses using RStudio version 0.99.902 (RStudio Team, 2015). Our call parameter data did not meet the assumptions of normality or homogeneity of variance (p < .05) (Boucher, 2017). Thus, we analyzed call parameter means (DT, D 1/4 ,

| Wetland impact assessment
Ambergris Caye had the highest degree of human disturbance (rank = 3-4; x̄ = 3.7; SE = 0.19) due to extensive habitat modification, pollution, and contact with humans. Moreover, all Ambergris sites had nearly constant anthropogenic sound present from development, and road, boat, and air traffic. Anthropogenic impact on Caye Caulker varied in the study area and ranged from high (rank = 4) to no disturbance (rank = 0). We determined Caye Caulker to be a site of low-moderate (rank = 2-3; x̄ = 2.5; SE = 0.30) disturbance with few sources of constant sound production. Our wetland assessments of the BAL effluent lagoon and surrounding areas determined the study site to have low overall human impact (rank = 1-2; x̄ = 1.8; SE = 0.15). Despite being a commercial aquaculture facility, the effluent lagoon is located on the margins of the farm and receives little chronic anthropogenic sound production, has minimal pollution, and the crocodiles rarely come into direct contact with people.

| Call production
We captured 89 American Crocodiles during the entirety of this class call production varied between sites ( Figure 5). However, only hatchlings (p = .001) and adults (p = .03) called more often at Ambergris Caye than at the combined Caye Caulker and BAL sites (Table 4).

| Call structure
American Crocodile calls are acoustically complex with multiple harmonics. General call structure is the same for all size classes, but call parameters differed among size classes and allowed for successful differentiation of individual's calls to their respective size class. Individually, the measured parameters achieved successful differentiation of larger size classes (sub-adult and adult) from the smaller size classes (juveniles and hatchlings). When applied together, measured call parameters separated recorded crocodiles into homogeneous groups of their respective size classes. Hatchlings and juveniles emitted calls with higher overall frequencies and call modulation. Adult and sub-adult American Crocodiles produced distress calls at significantly lower frequencies and have reduced call modulation. There is also a positive correlation between increasing body size and call duration.
Total call duration increased incrementally with body size. This increase in call duration and decrease in call frequency has also been noted for Indian Gharial (Gavialis gangeticus) (Bonke, Whitaker, Rodder, & Bohme, 2015). Animal sound production changes as body size increases (Chabert et al., 2015). Lower frequencies are generated by larger crocodiles as the call production mechanisms are larger and produce longer wavelengths (Britton, 2001;Chabert et al., 2015).
Crocodylians use low-frequency sound production to communicate long-distances, particularly during courtship (Dinets, 2013;Vergne et al., 2009;Vliet, 1989). Hatchling distress calls are higher in frequency and will have limited range as high frequencies do not travel long-distances as effectively. Post-hatching American Crocodile hatchlings will remain in cohesive groups monitored closely by the maternal female (Thorbjarnarson, 1989). Call production likely does not need to travel long-distance to elicit appropriate defense response. However, microhabitat use by juveniles and sub-adult and adult American Crocodiles differs from hatchlings and yearlings (Thorbjarnarson, 1989). The production of juvenile distress calls at lower frequencies may be beneficial in garnering response from conspecifics that have greater dispersion within a given habitat. Also, larger American Crocodiles occupy more open microhabitats and long-distance signals can travel unimpeded by the dense shoreline cover that hatchlings prefer.
Differences in call parameters among size classes may reflect differences in behavioral responses. Investigation of distress call information coding demonstrated that crocodylian distress calls produced at higher pitch, with higher frequencies eliciting greater behavioral response from other crocodylians (Staton, 1978;Vergne et al., 2011).
Hatchling American Crocodiles produce distress calls at significantly higher frequencies and have greater modulation. During collection of distress calls, we noted that nearby conspecifics reacted more intensely to distress calls from hatchlings compared to the other size classes. Response to distress calls of captured hatchlings was primarily from other hatchlings in the same pod. Size of hatchlings also differed between nest sites and capture of smaller hatchlings from other pods elicited greater response from nonrelated females.
Hatchlings also produced distress calls more frequently than the other size classes, which may be another reflection of the significant behavioral response the call elicits. It is likely that the increased behavioral response to hatchling distress calls aids in juvenile survivorship by decreasing predation risk. However, if response to distress calls decreases as frequency and signal modulation decreases it is counterintuitive that sub-adult and adult American Crocodiles produce distress calls with frequency. Distress calls do not appear to have intrinsic value to individual survivorship as sub-adult and adult crocodiles have few natural predators (Pooley & Ross, 1989;Staton, 1978). We did not detect any call response from conspecifics when recording distress calls of sub-adult and adult crocodiles.
As such, it makes little sense for crocodiles to produce energetically costly calls that have no perceived benefit. Adult crocodile distress calls, to our best knowledge, have never been reported in such high incidence rates. It may be that the production of distress calls by larger American Crocodiles is indeed the result of anthropogenic activity such as the hydrologic and vegetative modifications we observed in the wetlands, but more likely due to enhanced human presence as evidenced by increased garbage dumping and artificial lighting at night. We further surmise this may be due in part to past or present illegal hunting pressure or harassment in the most modified wetlands.

| Call production
Our study is the first to document the frequency of distress call production of American Crocodiles in response to capture. a Proportion of calls within the overall column, followed by the same uppercase letter, is not significantly different among size classes (p > .05).

F I G U R E 5 Proportion of American
b Proportion of calls within a column (location), followed by the same lowercase letter, is not significantly different among size classes (p > .05).
Furthermore, this is the first record of adult American Crocodiles producing distress calls with such regularity. Adult production of distress calls has been reported but described to be a rarity (Staton, 1978). Our results show that a high proportion (36%) of adult American Crocodiles in Belize produce distress calls. This is markedly higher than the production of distress calls for subadults (22%). In particular, we found that the highest proportion of distress calls produced by both adults and sub-adults came from crocodiles captured on Ambergris Caye study. Ambergris Caye also had the highest proportion of juvenile and hatchling call production. With the exception of hatchlings, we did not find the proportion of call production to be statistically significant for size classes between study sites, which may have been due to our small sample size. This result may be somewhat underestimated as sample size for each size class is low for all sites. However, we noted ecological significance in distress call production. Higher distress call production for hatchlings and adults occurred at the study area with high human impact (rank = 3-4). We found, for juveniles and subadults, lower distress call production at Caye Caulker (rank = 2-3) and BAL (rank = 1-2). Greater call production may be a result of anthropogenic impact in crocodile habitat. As such, production of distress calls could be a metric to determine increased stress levels in American Crocodiles. However, as indicated above, low sample sizes may contribute to this finding, and we suggest further research to investigate the relation among call structure, call production, and anthropogenic disturbance. Moreover, anthropogenic impact has been demonstrated to cause deleterious shifts in wildlife behavior and sound production (Laiolo, 2010).
Anthropogenic sound production causes temporal and frequency shifts in wildlife sound production (Barber, Crooks, & Fristrup, 2010;Laiolo, 2010 than those at the other sites but produced calls of longer duration. This contradicts results from previous study that demonstrate increased body size results in increased call duration (Bonke et al., 2015). We found maximal frequency, F 1/4 , F end , and call modulation of Slope 2 higher at Caye Caulker and BAL sites, which is also contradictory of previous research in which larger crocodiles produce sound at lower frequencies due to increased size of the sound production apparatus (Britton, 2001).
Our results for variance of call parameters between sites complement those of our analysis of distress call production. We observed that the behavioral response of American Crocodiles to distress calls decreased as frequency and modulation decreased, and duration increased. Specifically, American Crocodiles, particularly adults (most likely the mother), responded more intensely in defense of hatchlings by approaching within 7 m of the calling individual. Moreover, previous research on information encoding of juvenile crocodylian calls determined that frequency modulation is the key parameter for behavioral response (Vergne et al., 2012).

| CON CLUS IONS
Our results indicate that anthropogenic disturbance may be altering American Crocodile sound production. This could have overarching effects on the behavioral response to distress calls impacting juvenile and hatchling survivorship if defense response is lessened by altered call production. American Crocodiles also use a variety of other acoustic signals to communicate, particularly during courtship, and it is feasible that other calls produced may also be impacted by anthropogenic disturbance. There is ever increasing evidence of human activity and sound production affecting avian, terrestrial, and aquatic wildlife sound production and mitigation of bioacoustic conflict between people and wildlife is emerging. The study of crocodylian bioacoustics is still in its infancy but steps should be taken to further study crocodylian acoustic communication, particularly in response to environmental and acoustic stressors.

ACK N OWLED G M ENTS
We thank the Belize Forest Department for their support of this and WVA00812) during manuscript preparation. We thank N.
Mathevon and an anonymous reviewer for helpful comments on the manuscript.

CO N FLI C T O F I NTE R E S T
None declared.