Timescales of call variability in a South American treefrog

Timescales of call variability in a South American treefrog | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Timescales of call variability in a South American treefrog Mariana Rodriguez-Santiago, Paula Pouso, Kim Hoke, Erik Zornik This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9458485/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract Communication systems shape behavioral interactions that are crucial for recognition, territoriality, and reproduction across animals. In many anuran species, acoustic communication drives reproductive and defensive behaviors as males produce advertisement calls to establish territories and attract females. Understanding the temporal dynamics in vocalizations and their variation between and within males can provide crucial insight into the mechanisms that generate call patterning and communication. Here, we examine such dynamics in the call patterning of male Boana pulchella , a South American treefrog species that produces temporally-variant call series over time. Using rhythm analysis, we quantify the rhythmic structure of these vocalizations and characterize the variability in their temporal patterning between males. We find that rhythmic isochrony is timescale-dependent - call series are produced isochronously while the patterning of calls and notes within these series is non-isochronous yet periodic. The isochronous rhythm is maintained through an inverse relationship between series duration and the silence interval that follows it, such that longer series are followed by shorter silence intervals. Temporal features of calls and notes as well as their variation vary depending on within-series patterning dynamics and across timescales. Taken together, these results suggest that B. pulchella vocal timing is generated by hierarchically organized oscillators that produce isochronous vocalizations over a calling bout through flexible modulation at finer timescales. anuran acoustics vocalizations rhythm analysis variation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 INTRODUCTION Acoustic communication systems enable individuals to convey information that is crucial for a wide range of behaviors (Gerhardt, 1992 ). From territory defense to courtship signaling to individual recognition, senders tailor their vocal cues to elicit context-appropriate responses from receivers. The production of these signals is energetically costly (Gerhardt 1994 ; Gillooly & Ophir 2010 ) and can have detrimental results such as exposing signalers to predators (Rand & Ryan 1981 ). To mitigate costs and facilitate context-specific communication, signalers often modify aspects of their vocalizations such as the amplitude or rhythmic timing of their signals (Humfeld 2013 ; Welbergen & Davies 2008 ; Bhat et al. 2022 ). Understanding the boundaries of variation in signaler vocalizations informs the physiological mechanisms and constraints that regulate call patterning and communication. Vocalizations consist of spectral and temporal properties that contain important information such as species, social context (e.g. reproductive or competitive), and signaler condition. Spectral properties, like fundamental frequency, often play a role in species recognition, can correlate with signaler size and condition, and tend to be relatively invariant over time (Marler and Peters 1977 ; Ryan 1985 ; Gerhardt 1991 ; Gerhardt and Bee 2007 ; Woolley and Moore, 2011 ; Moore and Woolley, 2019 ; Rivera et al. 2023 ). Meanwhile, temporal features can change markedly within short and long timescales (Gerhardt 1991 ; Pollack 2000 ; Sakata et al. 2008 ). Call rate, for example, varies between and within individuals over time, fluctuates in both emission rate and periodicity (Narins 1982 ; Lopez & Narins 1991 ; Klump & Gerhardt 1992 ; Tumer & Brainard 2007 ), and can depend on factors like temperature and social cues (Baraquet et al. 2013 ; Furtado et al. 2016 ). Other timing features like call duration and inter-call intervals can vary across timescales depending on the acoustic environment (Zelick and Narins, 1985 ; Greenfield 1988 ; Tobias et al., 1998 ; Brumm 2006 ). Faster call rates often correlate with greater mating success (Sullivan 1983 ; Gibson & Bradbury 1985 ; Morris & Yoon 1989 ; Sullivan & Hinshaw 1990 ; McComb 1991 ), illustrating the importance of dynamic temporal modulation in shaping communication systems. Beyond their context-dependent functions, the timescales of temporal variability can also shed light on the physiological mechanisms that drive their production. In songbirds for example, within-individual variability in the acoustic features of syllables is linked to the trial-by-trial differences in the activity of ensembles of neurons (Sober et al. 2008 ). Furthermore, variability in the pattern of neural activity is associated with variability in song structure during learning (Olveczky et al. 2005; Kao et al. 2005 ). Thus, the temporal structure of vocalizations and their variability over time can illustrate the mechanistic principles that generate them. In anurans, vocalizations are composed of acoustic units that tend to be quite stereotyped within species but vary in their rate of production, timing, and sequencing depending on the context. Since Peter Narins’ foundational work that devised rigorous methods to measure call timing and its variability (Narins & Capranica 1977 ; Narins 1982 ; Zelick & Narins 1982 ; Zelick & Narins 1983 ), few studies have rigorously characterized vocal rhythms and temporal variation across multiple timescales. The relatively recent development of generalizable rhythm analysis schema to quantify rhythmic communication across species offers new potential for insights into both evolutionary trajectories and the underlying mechanisms of vocal communication systems (Ravignani & Norton 2017 ; Burchardt & Knörnschild 2020 ; Burchardt et al. 2021 ; Anichini et al. 2023 ; Jadoul et al. 2025 ). Here, we adopt these methods and modify their application to examine the emergence of rhythmicity in anuran vocalizations. By coupling these analyses with measures of between- and within-male variation in temporal features, we aim to inform the potential mechanisms that generate rhythmic calling across multiple timescales. Boana pulchella is a South American treefrog that lives in ponds and forested streams of Uruguay, southern Brazil, southern Paraguay, and northeastern Argentina, and can be heard calling not just at marked breeding seasons but throughout the year (Canavero et al. 2008 ; Pouso et al. 2025). Males of this species produce a simple doublet call consisting of two notes with spectral features that vary depending on body size and environmental temperature (Basso & Basso, 1987 ; Ziegler et al. 2011 ; Ziegler et al. 2016 ). Although previous studies have noted that the temporal patterning males use when calling in large choruses are highly dynamic (Basso & Basso 1987 ; Canavero et al. 2008 ), these vocal patterns and their variation remain uncharacterized across time scales. Using free-field recordings of B. pulchella , we examine variation of temporal features across timescales to generate hypotheses about the underlying oscillatory and motor control mechanisms organizing acoustic signals. During calling bouts, males can be heard appending multiple calls in series, ranging in number from a single call to five or more. We examine the rhythmic structure of these vocalizations, quantifying the periodicity in their call timing across timescales. We also examine the extent of between- and within-male variation in the timing features of their calls. Ultimately, understanding the timescales of both temporal modulation and individual variation can shed light on the potential physiological mechanisms that generate and constrain flexible communication systems. METHODS Study system and acoustic recordings The calls of 22 B. pulchella males were recorded in Paraje Zanja del Tigre, Maldonado, Uruguay in March 2020. At this site, males are readily abundant in the foliage and in the water. We held unidirectional Sennheiser microphones (Sennheiser MKE 400) connected to a digital recorder (Olympus LS 11) approximately 20 cm from each calling male to record focal males for 5 minutes. All recordings were performed between the hours of 23:00–2:00. After taking each recording, we measured the snout-vent-length (SVL), femur length, and weight of each male. Humidity levels (Extech RHT510) and air temperature (HOBO UA-002-64, Onset Computer Corporation) were logged every night. Of the 22 males recorded, four recordings were excluded due to high background noise impeding our ability to differentiate the focal male from the chorus. Signals were detected using Raven Pro 1.6 (Cornell Laboratory of Ornithology, Ithaca, NY; Charif et al, 2010). We distinguished focal males from the chorus using amplitude detection on the oscillograms with parameters set by eye based on individual recording conditions. We obtained start and end times of each note and verified each automatic detector by hand to ensure correct detection. All background chorus calls were excluded during the selection process. Tabulated selections were exported as .txt files per focal male and imported into RStudio for subsequent analysis. We used the package ‘warbleR’ (Arayas-Salas 2017) to measure spectrotemporal features of each note (for a full list of features measured, see Supp. Materials). Analysis of temporal features across timescales Preliminary analyses suggested that male call rates vary across timescales – from calling bouts that range in duration anywhere from several seconds to several minutes (Fig. 1 a), to call series within these bouts where calls are separated by silences of 1-10 seconds (Fig. 1 b), to a within-series timescale where calls are separated by silences < 1 sec on average (Fig. 1 c). These timescales were initially identified through visual inspection of the distribution of inter-call onsets (ICOs), which is the length of time between the start of consecutive doublet calls. The distribution of ICOs suggested the existence of two peaks – one very sharp, short-peak (< 1 s) and another peak distributed more widely around 1.8 s. We selected the timescale cutoffs for these ICOs based on within-male distributions to determine each male’s call timing thresholds (see Supp. Fig. 1 for examples). We identified x-axis min and max numbers for each peak distribution range and extracted the x-axis max value for the first distribution, deemed the short-peak distribution trough. Although the exact ICO distributions vary by individual male, each male had a distinct trough value which we used to bin calls as belonging to long ( \(\:\stackrel{-}{X}\) > 1s) or short ( \(\:\stackrel{-}{X}\) < 1s) timescales. We capped the analysis to calling bouts that are separated by intervals of silence of 10+ seconds. Within these calling bouts, we characterized the temporal features ( duration, onset and interval ) of three vocal timescales: the notes that make up each call, doublet calls , and series composed of rapid series of calls separated by short ICOs. Table 1 . Glossary of temporal features quantified. Term Definition call (doublet) two-note sound unit consisting of a short note 1 and longer note 2 separated from subsequent call by a silence interval longer than the interval between note 1 and 2 inter-call onset (ICO) length of time between onset of note 1 in focal call and note 1 in following call inter-call interval (ICI) length of time between offset of note 2 in focal call and note 1 onset in following call note subunit of a call, varies in duration inter-note onset (INO) length of time between onset of note 1 and note 2 in a doublet call inter-note interval (INI) length of time between offset of note 1 and onset of note 2 in a doublet call (call) series a group of calls separated from other such groups by periods of silence much longer than the inter-call intervals inter-series onset (ISO) length of time between onsets of successive call series inter-series interval (ISI) length of time between offset of one call series and onset of subsequent call series To summarize, in Table 1 we provide definitions of the major temporal features of vocalizations quantified. These definitions are adapted from Kohler et al. 2017 and modified to fit the call structure of B. pulchella. The ICO as defined here is equivalent to the inter-onset interval (IOI) used broadly in other studies. The calculations for all temporal features measured – durations, onsets and intervals – was the same across all timescales, the only difference is the vocal timescale being measured (i.e. series , call , or note ). We quantified variation in these temporal features between and within males within calling bouts and within series. All analyses were performed in RStudio (v7.1; R Core Team 2021) using custom scripts - one script for series, one for calls, and one for notes - such that no timescale was duplicated in the analysis. Specifically, in the series data frame, each row represents a single call series and contains identifying information such as its begin and end time, duration, inter-series onset (ISO), inter-series interval (ISI), as well as qualitative measures such as the series length, or total number of calls within the series (solo, two, three, four, five). In the call data frame, each row represents a single call and includes columns detailing its beginning and end time, duration, ICO, inter-call interval (ICI), as well as the series length it belongs to (solo, two, etc) and the order in the series it appears in (within-series position: 1 = first call in series, 2 = second, etc). Many calls were not emitted within a call series, and these ‘solo’ calls were omitted in within-series analyses. Given that not every series length is represented equally, we binned the ‘within-series position’ into categories: first and mid. In the notes script, each row represents a note within a call and includes columns with its begin and end time, duration, inter-note onset (INO), inter-note interval (INI), mean dominant frequency, the series length it belongs to, and within-series position. Assessing rhythmic structure across timescales of vocalizations The structure of rhythmic vocalizations is often an indicator of the type of mechanism (i.e. physiological oscillators) that generates the signal. To quantify the rhythmic structure of calling, we followed a general rhythm analysis schema established by Ravignani & Norton 2017 and expanded on by Burchardt & Knörnschild 2020 (see also Anichini et al. 2023 ; Jadoul et al. 2025 ). We use one specific method, the interval ratio (IR), to determine whether the call patterning of B. pulchella is isochronous and periodic. Isochronous signals follow metronome-like beats such that the interval between consecutive signals is uniform. Periodicity, or the underlying temporal patterns within signal series, indicates potential mechanisms and physiological constraints that generate their production. Here, we quantified the IR of signals across vocal timescales (Fig. 2 ). All interval ratios were calculated such that: where X k is the inter-signal onset for the kth call. All ratios were calculated within calling bouts such that between-bout interval ratios were excluded from all analyses. We initially assessed whether each vocal timescale was produced isochronously by quantifying all observed interval ratios between males and assessing their distribution around 0.5. The shape of these distributions can identify rhythmic isochrony such that a unimodal distribution centered around 0.5 suggests isochrony. To characterize periodicity more generally, we generated a random null distribution of interval ratios by randomly sampling pairs of intervals (with replacement) from each male’s observed range, calculating the interval ratio for each pair, and repeating this process to produce approximately 6,000 simulated ratios across individuals. We used a two-sample Kolmogorov-Smirnov (KS) test to assess potential periodicity in the structure of these vocalizations by determining whether the observed interval ratio distributions differed significantly from the null distributions. Non-isochronous periodicities were identified by peak distributions in the interval ratios not centered around 0.5. To examine whether isochronous call series timing is consistent with a pacemaker-like mechanism that generates series onsets independent of series duration or with a consistent silent period between call series, we ran a linear mixed-effects model using the ‘lme4’ (Bates et al. 2015 ) package to test how the length of the series impacts the ISI with frogID as a random effect (Fig. 3 ). Given that series longer than three calls were less prevalent than solo or two-call series, we binned series of four or more calls (deemed here ‘four+’ or ‘4+) for statistical analyses. Examining how periodicity within short timescales varies with series dynamics To assess how temporal features of calls and notes vary depending on within-series dynamics, we ran a set of mixed models examining how series length and position within series affect the duration, silence interval, and onset intervals of calls and notes. We limited the analysis to series of 3–5 calls and examined how features vary depending on the duration of the series (three, four, five calls) and their position within the series (first and mid call in series). We excluded series of two calls because inherently they do not have mid-position calls, limiting our comparisons. In addition, given the nested nature of calling timescales, we excluded calls in the last position because their interval timing is part of the series timescale. Similarly, we excluded note 2 onset and interval timing from analyses because their interval timing is part of the call timescale. We initially assessed normality of residuals visually using Q-Q plots. We used the ‘lme4’ package for linear mixed models, ‘car’ for testing significance of fixed effects with type III Wald tests (Fox & Weisberg 2019), and ‘emmeans’ (Lenth et al. 2024) for contrasts. Contrasts were performed to understand significant interactions and main effects with Bonferroni correction. All results were considered significant at the p < 0.05 level. Marginal and conditional R 2 were calculated using the ‘performance’ package (Lüdecke et al. 2021 ). We detail the statistical models run for each vocal timescale below. Calls : To examine whether the temporal features of calls (response variables: duration, ICI, and ICO) vary depending on the series duration (three, four, or five calls) or sequence order within series (first and mid calls), for each response variable we initially compared two models: an unrestricted model including individual frog identity as a random intercept and the call series duration and position within series and their interaction as fixed effects, and a restricted model without the random effect. All models with the lowest AIC included frogID as random effects. Final models are detailed in Supp. Table 1. Notes We examined how note features vary depending on series duration and sequence order within series using a similar model selection process comparing restricted and unrestricted models as described for calls. In this case, since there are two note types within doublet calls (notes 1 and 2), we ran separate models per note to examine differences in note durations and dominant frequencies (DFs) depending on series duration and within-series position. To examine whether the silence interval and onset difference between notes 1 and 2 of each doublet call vary depending on series duration and sequence order, we ran separate models with INI and INO as the dependent variables. Results of the note features models can be found in Supp. Table 2. Quantifying within- and between-male variation in spectrotemporal features across timescales To quantify variation between and within males across calling timescales, we calculated the coefficient of variation for all series, call, and note features within (CV w ) and between males (CV b ). CVs were calculated such that: CV w = sd w / X̄ w CV b = sd b / X̄ b For example, a smaller CV w indicates acoustic features with low variation and high consistency within the many calls produced by an individual male. A small CV b indicates that a particular acoustic feature has low variation and high consistency between all the males. The ratio of CV b / CV w indicates to what degree callers tend to differ more from one another compared to their own intra-individual variation over time, such that a CV ratio > 1 indicates an acoustic feature that varies more between than within males while a CV ratio < 1 indicates relatively higher within-male variation and greater overlap between males. We compare how the temporal features of series, calls, and notes vary within calling bouts, and how call and note features vary within series. For within-series comparisons, we also examine how variation depends on the sequence position (see Supp. Table 3 for a comprehensive table detailing all timescale CV results). Specifically, for the within-series comparison, we compare the CV ratios of duration, interval, and onsets of calls and notes for first and mid calls in a series that contained at least three calls. For note features, we also examined how spectrotemporal features vary within a series. We highlight variation in the DF of notes 1 and 2 across timescales throughout the results. Although quantifying full spectral variation was outside of the scope of this paper, we provide CV summaries for all spectrotemporal features measured (listed in the Supp. Materials) in Supp. Table 4. Characterizing novel note types Using nomenclature originally established by Basso & Basso 1987 and expanded on by Baraquet et al. 2007 and Ziegler et al. 2011 , we initially identified calls based on the temporal sequence of notes 1 and 2 within doublet calls. We found that some notes fell outside of previously established note parameters – namely (1) notes that are not part of two-note calls but their duration and DF ranges overlap with either note 1 and note 2 (deemed ‘solos’), and (2) a novel single-note call type with a substantially longer duration than notes 1 and 2 (deemed ‘note 3’ and named “squeaks”). Exact ranges of features vary by individual male thus we automated note waveform selection based on individual male cutoffs. In Supp. Table 5, we summarize the mean ± sd and ranges of all identified note features. RESULTS Evidence for isochronous and non-isochronous periodicity depends on timescales We initially set out to examine the rhythmic patterning of B. pulchella vocalizations by quantifying the interval ratio (IR) distributions between the onsets of identified vocal timescales – series, calls, and notes – within calling bouts. We found striking IR results across timescales. Specifically, the IR distribution of call series was unimodal and centered around 0.5 (x̄ = 0.5, sd = 0.08), suggesting an isochronous call series signal within bouts. The observed distribution did not differ from a randomly generated null distribution (Fig. 2 b; null x̄ = 0.5, sd = 0.09, two-sample KS test: D = 0.04, p = 0.13). The ISO distribution showed one peak between 1–2 sec (Fig. 2 e), suggesting this is the typical series timing. We found that the call IR had multiple peaks, inconsistent with an isochronous rhythmic structure (Fig. 2 c). The observed call IR distribution differed significantly from the null distribution (D = 0.09, p < 0.001; null: x̄ = 0.49, sd = 0.04), suggesting non-isochronous periodicity. The ICO distribution had two primary peaks between (Fig. 2 f). The longer ICO peak centered at 1.5 sec (Fig. 2 f) represents the inter-call-onsets between a solo doublet or the last doublet in a call series and the next call series. The sharp ICO peak at a shorter timescale (ICO < 1 sec; Fig. 2 f inset) represents the regular timing between doublets within call series. The note IR was also not isochronous based on the presence of multiple IR peaks and the lack of any peak around 0.5 (Fig. 2 d). This distribution differed significantly from the randomly generated null distribution, which had additional peaks including one centered around 0.5 (D = 0.2, p < 0.001; null: x̄ = 0.5, sd = 0.03). The INO distribution had multiple peaks in less than 1 second (Fig. 2 g inset). Given that series are isochronous despite variation in series durations, we examined whether the isochronous series timing is maintained by changes in the silence interval after call series. We found that longer call series were followed by significantly shorter ISIs (Fig. 3 ). Pairwise contrasts show that solo calls have significantly longer ISIs than those of longer series (solo vs. two: t = 9.43, p < 0.0001; solo vs. three: t = 8.18, p < 0.0001; solo vs. four+: t = 3.74, p < 0.001); ISIs of two-call series were significantly longer than three-call series (t = 3.61, p < 0.001); and series of three calls do not differ significantly from longer call series (p = 0.81). These results show an inverse relationship between series length and the interval of silence that follows it. Timing of calls and notes varies depending on within-series dynamics To understand within-series temporal dynamics, we examined the variation in the onset intervals, silence intervals, and durations of calls and notes depending on series length and their relative position within the series. We restricted these analyses to series containing at least three calls to directly compare not just how series length affects temporal features but also the position within series (e.g., series of two calls do not have mid-position calls which would limit our comparisons). Calls We examined whether series length and sequence order within series influenced the ICO, call duration, and ICI (Fig. 4 a, b & c). We found that calls in series containing a total of four calls have a significantly longer ICO than those in three-call series (Fig. 4 a; estimate = 0.073, p < 0.001) and that calls in mid-series positions have longer ICOs than those first in series (estimate = 0.026, p = 0.005). Call durations get shorter in longer series (estimate = -0.016, p < 0.001) and are consistently and significantly shorter in mid-series positions compared to first calls (estimate = -0.016; p < 0.001). ICI of calls in four-call series are slightly longer than those in three-call series (estimate = 0.073, p < 0.001) and for calls in mid-series compared to first (estimate = 0.042, p < 0.001). Random effects explained a large portion of the variance in call duration (conditional R² = 0.76) but less so for ICI (conditional R² = 0.48) and ICO (conditional R² = 0.51). See Supp. Table 1 for a summary of all model comparisons. Notes : Across the board, temporal features of notes (durations, INI, and INO) vary distinctly depending on the within-series position regardless of the series duration (Fig. 4 d, e & f). The INO between notes 1 and 2 was significantly lower in mid-series calls than those first in series (Fig. 4 d; estimate = -0.008, p < 0.001). Duration of both notes 1 and 2 was significantly shorter in mid-series calls than those first in series (Fig. 4 e; note 1: estimate = -0.002, p < 0.001, note 2: estimate = -0.008, p < 0.001). The INI was also significantly shorter in mid-position calls compared to those first in series (Fig. 4 f; estimate = -0.006, p < 0.001). For all models, random effects explained a substantial portion of the variance (conditional R² ranges between 0.775–0.853 for note 2 duration, INI, and INO). See Supp. Table 2 for a summary of all model comparisons. Within- and between-male variation in spectrotemporal features depends on the timescale To understand how the above between- and within-series dynamics contribute to the overall patterns of variation across timescales, we initially quantified the CV w, CV b, and the CV b / CV w ratio of series, calls and notes (Fig. 5 ), and then examined the extent to which position within series and series duration accounts for the patterns of overall variation. We report the patterns of variation below, presented in order by the timescale of analysis. A table detailing all the CV w, CV b, and CV b / CV w ratio values for series, calls, and notes can be found in Supp. Table 3. Series : We find that all series-level temporal features (series duration, ISI, and ISO) vary more within males over time than between males (Fig. 5 a, d, g; CV b /CV w : series duration = 0.48; ISI = 0.34; ISO = 0.37). The degree of CV w is nearly double that of CV b for all features (CV w : duration = 68.8, ISI = 42.7, ISO = 36.6; CV b : duration = 33.22, ISI = 14.69, ISO = 13.39). Calls : Within calling bouts, we find that the ICI and ICO have a greater degree of within-male variation than between-male variation compared to call duration (CV b /CV w : ICI = 0.34, ICO = 0.34; CV b /CV w : duration = 1.06; Fig. 5 a, d and g). When we examined these patterns of variation within series, we found that the degree of between-male variation compared to within-male variation is much more similar among these call features (CV b /CV w : duration = 1.21, ICI = 0.96, ICO = 1.15; Fig. 5 b, e, and h) and the patterns of variation depend on the within-series position order (Fig. 5 c, f and i). Specifically, calls in the middle of series have a greater degree of overall variation (both within- and between-males) than the first call in a series (CV w : call duration − 1st = 5.92, mid = 7.86; ICI − 1st = 13.48, mid = 27.76; ICO − 1st = 9.04, mid = 20.59; CV b : call duration − 1st = 15.04, mid = 13.47; ICI − 1st = 16.9, mid = 34.56; ICO − 1st = 14.69, mid = 26.94). Notes Within bouts of calling, the duration of note 1 is the only note-level feature that has a higher level of within- than between-male variation (Fig. 5 g). Within series, the degree of between- to within-male variation in the INI and INO increases when taking the within-series position into account. Specifically, the degree of between-male to within-male variation in INI and INO increases in mid-position calls compared to first calls (1st CV b /CV w : INI = 2.84, INO = 4.36; mid CV b /CV w : INI = 2.91, INO = 4.43). The position-dependent increase in CV b /CV w is driven by a decrease in CV w for mid-position calls (INI: 1st CV w =7.05, mid CV w = 4.61; INO: 1st CV w = 5.36, mid CV w = 3.42). Variation in the individual note durations also depends on within-series patterning such that both note durations vary more between males than within, and the degree of between-male variation depends on within-series position (note 1 duration: 1st CV b /CV w = 1.36, mid = 1.78; note 2 duration: 1st CV b /CV w = 2.88, mid = 1.73). Overall, the CV w for the two note DFs is among the lowest CV w overall. We find that note DFs have a higher degree of between-male than within-male variation within series based on the ratio of CV b /CV w (note 1 DF: CV b /CV w = 1.12, note 2 DF: CV b /CV w = 1.13). CV w is higher in mid-position notes than first notes (note 1 DF: 1st CV w = 1.8, mid CV w = 5.37; note 2 DF: 1st CV w = 3.57, mid CV w = 6.51). Overall, note DFs retain nearly double the degree of between-male than within-male variation for first position within series (note 1 DF: 1st CV b /CV w =3.26, mid CV b /CV w 1.4; note 2 DF: 1st CV b /CV w 2.28, mid CV b /CV w = 1.2), and this is likely because the within-male variation in note DF is lower for the 1st call compared to mid calls. Detailed note analysis reveals novel and distinct note types We initially identified calls based on the temporal sequence of notes 1 and 2 and found that there were at least three distinct note types that differ in spectral and temporal properties (Fig. 6 ). Out of a total of 5701 individual notes across 18 males, 17 were ‘solos’ and 48 were note 3. Given their low occurrence rate in the dataset, we did not analyze features of solo notes and merely note their existence. We then compared the spectrotemporal features of notes 1, 2, and 3. We found that all three notes vary significantly in duration and dominant frequency (duration: X 2 =4181.5, df = 2, p < 0.0001, dom frequency: X 2 =346.83, df = 2, p < 0.0001; Supp. Table 5). While note 1 has a significantly shorter inter-note interval (INI) than notes 2 and 3 ( X 2 =4169.8, df = 2, p < 0.0001), notes 2 and 3 do not differ in INI. In addition, note 1 has a significantly shorter INO than notes 2 and 3 while these notes are not significantly different from each other ( X 2 =4170.3, df = 2, p < 0.0001; pairwise 1–2 and 1–3 p < 0.001, pairwise 2–3 p = 0.52). DISCUSSION We initially set out to characterize and quantify the temporal dynamics of B. pulchella vocalizations by quantifying the distributions of call IOIs, which is an established temporal feature measured in anuran communication studies more broadly. Consistent with previous reports in B. pulchella , we found that the vast majority of vocalizations consist of note pairs comprising a doublet call (Fig. 6 ; Basso & Basso, 1987 ; Canavero et al. 2008 ; Ziegler et al. 2011 ; Ziegler et al. 2016 ). Where we expand on the existing literature is that, in quantifying the call onset distribution, we found a series of nested vocal features. By using a rhythm-analysis approach, we found that male calling patterns occur on multiple vocal timescales, and the degree of between- and within-male variability in temporal features depends on timescale dynamics. While call series (also referred to as “trains” or “bouts”) have been described in several frog species, the rhythmic nature of these vocal patterns have not been quantified. Our results suggest that long timescale rhythmicity is not limited to elaborate or learned vocal systems, and the neural underpinnings of vocal patterning across a wide range of timescales may be ancient among tetrapods. While we highlight the dynamic vocal behavior of B. pulchella specifically, we encourage similar multi-timescale analyses across other frog species to gain an evolutionary perspective on rhythm generation complexity and its role in communication systems. Different rhythmic patterns emerge across nested vocal timescales Our findings show that B. pulchella calls were clustered into repeating series of variable length, typically between 1–4 calls (Fig. 1 ). While the number of calls in each series varied both within and between males, the series themselves were regularly timed and isochronous (Fig. 2 b). This regular series timing appears to be driven by the relationship between the total number of calls (series duration) and the silence interval that follows, such that longer series are followed by shorter silence intervals (Fig. 3 ). Unlike the longer inter-series timing, interval ratios at shorter timescales— between calls and between notes— exhibited non-isochronous patterns (Fig. 2 c and d). While the call interval ratio distribution did have a peak centered at 0.5 (Fig. 2 c), there were two other peaks around 0.2 and 0.8. The peak distributions around 0.2 and 0.8 likely result from consecutive calls with very different ICOs (e.g. a two-doublet call series with an ICO of 0.3 s for the first call and 1.5 s for the second call). The peak distribution around 0.5 could result from consecutive calls within a series (each with an ICO of around 0.3 seconds; Fig. 2 f inset) or from consecutive solo calls that are not part of a series (each with an ICO around 1.5 seconds, Fig. 2 f). The note interval ratio distribution has four peaks, two below 0.5 and two above 0.5 (Fig. 2 d). The existence of four peaks is a result of the regular repetition of two-notes within doublet calls that have two distinct inter-onset-intervals. For example, the note interval ratios below 0.5 belong to doublet calls within a series while the longer note interval ratios above 0.5 belong to call intervals between series. The note-level interval ratios lack any peak centered around 0.5, although such a peak is prominent in the null distribution generated by selecting intervals in random orders. The lack of isochrony in the observed note data is driven by the fact that, consistent with current findings, the majority of vocalizations occur as two-note doublets separated by longer silence periods from the next call (Fig. 6 c). For a visual example of these non-0.5 call and note IR distributions, see Supp. Figure 2 . The long-timescale series rhythms we found share features with behaviors of vocal systems with greater vocal repertoires and semantic complexity such those in birds, marine mammals, and primates (Weiss et al. 2014 ; Levinson 2016 ; Anichini et al. 2023 ). The distinct patterns of interval ratio distributions between timescales mirror findings in other taxa where rhythmic regularity at one level of organization does not predict regularity at nested levels (Xing et al. 2022 ; Ma et al. 2024 ). Taken together, our results suggest an analogous hierarchical organization in anuran vocal production in which the temporal structure of vocalizations is timescale-dependent. Patterns of variation across timescales Despite their isochrony, the temporal features of call series exhibited a greater degree of within- than between-male variation (Fig. 5 ). This is likely due to the fact that, while males maintain isochronous calling series across a calling bout, they can vary the total number of calls produced in quick succession, and this variation in total calls produced could be influenced by social or environmental feedback. Such rhythmic but variable vocal timing is reminiscent of chorus dynamics in E. coqui , in which a “sloppy oscillator” is proposed to generate a flexible call period that, while variable in isolation, can be entrained to a range of call intervals of neighboring frogs (Zelick and Narins, 1985 ). Such flexibility in the temporal generation of calls may be a necessity of chorusing systems, where the social context drives call patterning within bouts and modifies short timescale temporal features (Larter et al. 2022; Larter et al. 2026 ). Although outside the scope of the current study, future work can investigate the degree to which inter-series rhythms in Boana vary depending on behavioral patterns of chorusing neighbors. Given the existence of these call clusters we deem here as ‘series’, we compared the temporal properties of calls within and across series (i.e. across shorter versus longer timescales). We found that the temporal features of calls vary dynamically depending on series duration and within-series position (Fig. 4 ). Within series, mid-position calls have significantly longer inter-call onsets (ICOs) compared to the first call in series, and we conclude that these longer ICOs are driven by longer inter-call silence intervals (ICIs) because call durations of these mid-position calls are actually shorter (Fig. 4 ). The dependence of call timing on series position is reflected in distinct variation across timescales. First, within-male variation in ICO and ICI was high within bouts and significantly decreased within series (Fig. 5 ). Likewise, within-male variation in call duration was modest when looking across or within series but decreased when accounting for position within series. Unlike call timing properties, note features do not significantly vary with series duration but rather depend exclusively on the sequence order (Fig. 4 ). As noted above, call durations are shorter in the middle of calls, and we can attribute that change in call properties to both decreased durations of notes and silence within calls. Specifically, the inter-note onset and interval (INO and INI) between notes 1 and 2 get shorter in mid-position calls (Fig. 4 d and f) and the durations of both notes 1 and 2 also get shorter in mid-position calls (Fig. 4 e). The relative degree of between- male variation is higher for INI and INO in mid-position calls compared to those that come first in the series (Fig. 5 c). In fact, the mid-position INI and INO have the highest CV ratio in our study, meaning they are very consistent within males yet differ among males. These CV ratio results are an interesting comparison to the linear model results showing that individual identity explained a large and consistent portion of the variance in call duration (conditional R² = 0.76) and note durations (note 1: conditional R² = 0.76; note 2: conditional R² = 0.78), but substantially less variance in inter-call intervals (conditional R² = 0.48). The CV ratio for call duration further confirms that between-male differences in call duration exceed within-male variation, and this depends on the within-series position of calls with the first call in a call series having especially low within-male variation and thus readily measurable differences between males. For all note features except the duration of note 1, individual identity explained a large and consistent portion of the variance (Supp. Table 2). The pronounced reduction of within-male variation when accounting for series positions highlights the value of defining distinct timescales of vocal features since failing to do so results in an overestimation of within-male variation. Additionally, we want to highlight here that we found no evidence for within-series position effects on spectral features, such as the dominant frequency, of notes 1 and 2 (Supp. Table 2) and low CVw for note spectral features, suggesting these remain static within-individuals and do not vary depending on temporal dynamics. These findings are consistent with previous works showing that spectral and duration features of vocalizations are more individually stereotyped than temporal spacing features (Gerhardt 1991 ; Gerhardt and Bee 2007 ). Call duration, like spectral properties, may be constrained by individual morphology— particularly body size and the physical properties of the vocal apparatus—rendering it a relatively fixed signature of the individual. Inter-call intervals, by contrast, are more responsive to immediate context, suggesting they are regulated by mechanisms that are less tightly coupled to fixed individual traits. This functional dissociation has implications for signal evolution: if receivers use call duration to identify or assess individual callers and use inter-call timing to track calling effort or competitive state, selection could act independently on these features. Potential neural mechanisms that drive vocalization patterning across timescales Central pattern generators (CPGs) are circuits capable of generating and modulating rhythmic motor output in the absence of rhythmic input (Marder and Calabrese, 1996; Marder and Bucher, 2007). The isochronous call series rhythm seen here is consistent with a “series CPG” in which neurons exhibit consistent periods of activation. The regular call timing within each series could be driven by mechanisms nested within the series rhythm generator itself; alternatively, call timing within series could be regulated by a distinct but coupled “call CPG.” The duration of each series could, in turn, be driven by mechanisms located in either the series CPG, the call CPG, or both. While the temporal patterning of call emission within a series—including the observed shortening of call and note elements depending on sequence order—may simply reflect the dynamics of the call CPG, we propose that the cycle-by-cycle series durations are actively modified by social features such as chorus amplitude or the call properties of neighboring males. Such social signals could lead to sensory-based excitatory, inhibitory and/or modulatory inputs onto the circuit elements regulating both series onset and duration patterns. Thus, the mechanisms controlling the timing of nested rhythms within series could operate under different constraints—including ones that permit, or even require, flexibility in response to immediate contextual demands. CPGs in the anuran brainstem (Barkan et al. 2018 ; Kelley et al. 2020 ) are strong candidates for the neural substrates mediating vocal dynamics in B. pulchella . In distantly related aquatic frog species in the genus Xenopus , two reciprocally connected hindbrain nuclei – nucleus ambiguus (NA) and the parabrachial nucleus (PB) – comprise the vocal CPG (Kelley et al., 2020 ; Rhodes et al., 2007 ). While PB is able to generate neural correlates of call duration and period independent of NA, feedback from NA to PB is needed to generate the faster inter-pulse intervals (Zornik et al., 2010 ). Although the timescales involved in the Xenopus vocal CPG are faster than Boana vocal patterns, one possibility is that homologous coupled circuit components generate the shorter timescale call periods and the longer duration series timing. PB has been shown to participate in vocal production in both Xenopus (family Pipidae) and Rana (family Ranidae), suggesting that its involvement in vocal patterning is ancient and conserved across anurans (Schmidt, 1992 ; Barkan et al., 2017 ). Whole-cell physiology recordings in PB have revealed neurons that intrinsically encode species-specific call periods in X. laevis and X. petersii (Barkan et al., 2018 ), but longer timescales have not been investigated in these species. Future physiological experiments in Boana (and across anurans) could allow researchers to test hypotheses about underlying circuits that generate vocal patterns across timescales, providing potentially exciting insights into the evolutionary trajectories of nested vocal rhythms in frogs. Behavioral implications At the within-series level, the ability to append calls in series—and to modulate the timing of those calls—may allow males to signal competitive arousal or motivation. Males producing longer series may convey greater effort or urgency to both females and rival males. Faster call rates within bouts have been associated with greater mating success across a range of anuran species (Sullivan 1983 ; Morris & Yoon 1989 ; Lopez & Narins 1991 ), and the progressive temporal adjustments we describe in males quickly appending calls within series could be this type of competitive signal. At the between-series level, the isochrony we observe in series spacing may serve a different social signal. Regular, predictable calling cadences could allow male receivers to anticipate the timing of future calls by neighbors. Being able to predict when silent periods in chorus signaling may arise can help males time their next call and facilitate signal detection in noisy chorus environments (Klump & Gerhardt 1992 ; Larter & Ryan 2024 ). Regular calling cadences could similarly help female receivers localize and assess multiple callers simultaneously, consistent with preferences by female treefrogs for high and regular call rates (Tanner & Bee 2019 ; Tanner & Bee 2020 ; Tanner et al. 2025 ). These nested timescales of periodicity could provide sufficient structure for receivers to assess short timescale features that reflect male motivation, physical condition, or individual identity nested within a more predictable signal that aids detection of and selective attention toward one caller in a complex environment. Characterization of a novel call type Previous studies that examined the calling behavior of this species restricted their analyses to characterizing changes in doublet calls across contexts. While some studies highlighted the variability of male call patterning over time, we found no published descriptions of a a single-note call with a long duration and upward-trending harmonics that we deem a ‘squeak’ call. We report that males produce two main types of calls – doublets and squeaks – which are distinguished by their durations and the sequence of note types used. Across our dataset, we found a total of 48 squeaks used between 18 males. Of these 18 males, only 11 produced squeaks and the total number of squeaks per male ranged from 1 to 17. Given the limited number of total squeak calls in our dataset, we are unable to report the contextual basis for its use (e.g. whether it is used to attract females or repel nearby males). We encourage future studies to characterize squeak properties and context-dependent use to enhance our understanding of their function. Conclusion The vocalization patterning of B. pulchella reflects a hierarchical structure in which rhythmic regularity emerges at longer timescales yet maintains highly variable patterning, and temporal features at shorter-timescale retain high flexibility between and within males. The dissociation between individually stereotyped features—call duration, note duration—and more variable timing features, such as the inter-call interval, suggests that distinct mechanisms regulate different aspects of vocal patterning. Such distinct vocal timescales could be generated by CPG activity that establishes call patterns at multiple timescales (such as ISO and ICO). Oscillatory or neuromodulatory mechanisms could regulate the number of calls within each series in response to external (e.g. social) or internal (e.g. energetic capacity) feedback. This temporal flexibility has plausible consequences for both mate assessment and acoustic competition in chorus environments. Future work directly linking neural activity to the patterns described here and examining how chorus dynamics shape within-bout calling decisions will be essential for understanding the physiological and ecological forces that structure vocal communication across species. Declarations Competing Interests: The authors have no relevant competing interests to declare. Funding Declaration: This work was supported by the National Science Foundation IOS-2154203 and Universidad de la República, Programa de Desarrollo de las Ciencias Básicas (PEDECIBA). MRS was supported by a National Science Foundation Postdoctoral Research Fellowship in Biology (award no. 2109884). COMPLIANCE WITH ETHICAL STANDARDS Research Involving Animals : This study was reviewed and approved by an ethics committee (Comisión Nacional de Experimentación Animal, Universidad de la República, Protocol Number 1061). Author Contribution P.P., K.H., and E.Z. conceptualized the project to examine call variability in Boana and conducted the field recordings. M.R.S. conceptualized the addition of rhythm analysis to examine vocal timescales. M.R.S. conducted all the acoustic and statistical analyses. M.R.S. wrote the original draft. All authors commented on previous versions and approved the final manuscript. All authors contributed equally to raising philosophical conundrums and leading the team into stimulating and productive intellectual spirals. Acknowledgement We wish to thank Esteban Russi for assisting with data collection and Alvaro Mazzilli Millan and his family, Mercedes Hernández, and Isidro Rodríguez for access to the field site in Uruguay. We also thank Ana Silva, Carlos Colacce, and La Comarca for providing housing. We are grateful to the members of the Hoke lab for helpful feedback and discussion on figure clarity. Data Availability All data and scripts used for analysis will be made publicly available on github upon publication. References Anichini M, De Reus K, Hersh TA, Valente D, Salazar-Casals A, Berry C, Ravignani A (2023) Measuring rhythms of vocal interactions: a proof of principle in harbour seal pups. Philosophical Trans Royal Soc B: Biol Sci, 378 (1875) Araya-Salas M, Smith-Vidaurre G (2017) warbleR: An R package to streamline analysis of animal acoustic signals. Methods Ecol Evol 8(2):184–119 Baraquet M, Salas NE, Ludovico AM (2013) Advertisement calls and interspecific variation in Hypsiboas cordobae and Hypsiboas pulchellus (Anura, Hylidae) from central Argentina. Terr Ecol Behav 65(4):479–486 Barkan CL, Kelley DB, Zornik E (2018) Premotor neuron divergence reflects vocal evolution. J Neurosci 38(23):5325–5337 Barkan CL, Zornik E, Kelley DB (2017) Evolution of vocal patterns: tuning hindbrain circuits during species divergence. J Exp Biol 220(5):856–867 Basso NG, Basso G (1987) Analisis acustico del canto nupcial de Hyla pulchella pulchella . Mus Hist Nat Valparaiso 18:109–114 Bates D, Maechler M, Bolker B, Walker S, Christensen RHB, Singmann H, Bolker MB (2015) Package ‘lme4’. convergence , 12 (1), 2 Bhat AS, Sane VA, Seshadri KS, Krishnan A (2022) Behavioural context shapes vocal sequences in two anuran species with different repertoire sizes. Anim Behav 184:111–129 Brown CH, Beecher MD, Moody DB, Stebbins WC (1978) Localization of primate calls by old world monkeys. Science 201(4357):753–754 Brumm H (2006) Signalling through acoustic windows: nightingales avoid interspecific competition by short-term adjustment of song timing. J Comp Physiol A 192(12):1279–1285 Brush JS, Narins PM (1989) Chorus dynamics of a neotropical amphibian assemblage: comparison of computer simulation and natural behaviour. Anim Behav 37:33–44 Burchardt LS, Knörnschild M (2020) Comparison of methods for rhythm analysis of complex animals’ acoustic signals. PLoS Comput Biol 16:e1007755. 10.1371/journal.pcbi.1007755 Burchardt LS, Picciulin M, Parmentier E, Bolgan M (2021) A primer on rhythm quantification for fish sounds: a Mediterranean case study. Royal Soc open Sci, 8 (9) Canavero A, Arim M, Naya Monteverde DE (2008) Calling activity patterns in an anuran assemblage: the role of seasonal trends and weather determinants. North-Western J Zool 4(1):29–41 Chagnaud BP, Baker R, Bass AH (2011) Vocalization frequency and duration are coded in separate hindbrain nuclei. Nat Commun 2(1):346 Forrest T, Ariaratnam J, Strogatz S (1998) Synchrony in cricket calling songs: models of coupled biological oscillators. J Acoust Soc Am 103(5Supplement):2827–2827 Fox J, Weisberg S, Adler D, Bates D, Baud-Bovy G, Ellison S, Heiberger R (2012) Package ‘car’. Vienna: R Foundation Stat Comput 16(332):333 Furtado R, Santos SP, Dias TM, Bastos RP, Nomura F (2016) Vocal repertoire during reproductive and aggressive contexts of three Brazilian tree frogs: Bokermannohyla sapiranga, Hypsiboas albopunctatus and H. goianus (Anura: Hylidae). South Am J Herpetology 11(2):136–147 Gerhardt HC (1991) Female mate choice in treefrogs: static and dynamic acoustic criteria. Anim Behav 42(4):615–635 Gerhardt HC (1992) December). Multiple messages in acoustic signals. Seminars in Neuroscience, vol 4. Academic, pp 391–400. 6 Gerhardt HC (1994) The evolution of vocalization in frogs and toads. Annual review of ecology and systematics , 25 (1), 293–324 Gerhardt HC, Bee MA (2007) Recognition and localization of acoustic signals. Hearing and sound communication in amphibians. Springer New York, New York, NY, pp 113–146 Gibson RM, Bradbury JW (1985) Sexual selection in lekking sage grouse: phenotypic correlates of male mating success. Behav Ecol Sociobiol 18:117–123 Gillooly JF, Ophir AG (2010) The energetic basis of acoustic communication. Proceedings of the Royal Society B: Biological Sciences , 277 (1686), 1325–1331 Greenfield MD (1988) Interspecific acoustic interactions among katydids Neoconocephalus: inhibition-induced shifts in diel periodicity. Anim Behav 36(3):684–695 Hartbauer M, Kratzer S, Steiner K, Römer H (2005) Mechanisms for synchrony and alternation in song interactions of the bushcricket Mecopoda elongata (Tettigoniidae: Orthoptera). J Comp Physiol A 191(2):175–188 Humfeld SC (2013) Condition-dependent signaling and adoption of mating tactics in an amphibian with energetic displays. Behav Ecol 24(4):859–870 Jadoul Y, D’Orazio F, Eleuteri V, van der Werff J, Tufarelli T, Gamba M, Ravignani A (2025) Rhythmic analysis in animal communication, speech, and music: the normalized pairwise variability index is a summary statistic of rhythm ratios. Vibration 8(2):12 Kao MH, Doupe AJ, Brainard MS (2005) Contributions of an avian basal ganglia–forebrain circuit to real-time modulation of song. Nature 433(7026):638–643 K. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology (2026) Raven Pro: Interactive Sound Analysis Software (Version 1.6.5) [Computer software]. Ithaca, NY: The Cornell Lab of Ornithology. Available from https://www.ravensoundsoftware.com Kelley DB, Ballagh IH, Barkan CL, Bendesky A, Elliott TM, Evans BJ, Zornik E (2020) Generation, coordination, and evolution of neural circuits for vocal communication. J Neurosci 40(1):22–36 Klump GM, Gerhardt HC (1992) Mechanisms and function of call-timing in male-male interactions in frogs. Playback and studies of animal communication. Springer US, Boston, MA, pp 153–174 Köhler J, Jansen M, Rodriguez A, Kok PJ, Toledo LF, Emmrich M, Vences M (2017) The use of bioacoustics in anuran taxonomy: theory, terminology, methods and recommendations for best practice. Zootaxa 4251(1):1–124 Larter LC, Bernal XE, Page RA, Ryan MJ (2023) Local competitive environment and male condition influence within-bout calling patterns in túngara frogs. Bioacoustics 32(2):121–142 Larter LC, Cushing CW, Ryan MJ (2026) Cadences of the collective: conspecific stimulation patterns interact with endogenous rhythms to cue socially mediated response shifts. J Exp Biol 229(1):jeb250982 Larter LC, Ryan MJ (2024) Túngara frog call-timing decisions arise as internal rhythms interact with fluctuating chorus noise. Behav Ecol 35(4):arae034 Lenth R, Singmann H, Love J, Buerkner P, Herve M (2019) Package ‘emmeans’. R package version, 1 (3.2) Levinson SC (2016) Turn-taking in human communication–origins and implications for language processing. Trends Cogn Sci 20(1):6–14 Lopez PT, Narins PM (1991) Mate choice in the neotropical frog, Eleutherodactylus coqui. Anim Behav 41(5):757–772 Lüdecke et al (2021) performance: An R Package for Assessment, Comparison and Testing of Statistical Models. J Open Source Softw 6(60):3139. https://doi.org/10.21105/joss.03139 Ma H, Wang Z, Han P, Fan P, Chapman CA, Garber PA, Fan P (2024) Small apes adjust rhythms to facilitate song coordination. Curr Biol 34(5):935–945 Marler P, Peters S (1977) Selective vocal learning in a sparrow. Science 198(4316):519–521 McComb KE (1991) Female choice for high roaring rates in red deer, Cervus elaphus. Anim Behav 41(1):79–88 Moore JM, Woolley SM (2019) Emergent tuning for learned vocalizations in auditory cortex. Nat Neurosci 22(9):1469–1476 Morris MR, Yoon SL (1989) A mechanism for female choice of large males in the treefrog Hyla chrysoscelis. Behav Ecol Sociobiol 25:65–71 Narins PM (1982) Behavioral refractory period in neotropical treefrogs. J Comp Physiol 148(3):337–344 Narins PM, Capranica RR (1977) An automated technique for analysis of temporal features in animal vocalizations. Anim Behav 25:615–621 Ölveczky BP, Andalman AS, Fee MS (2005) Vocal experimentation in the juvenile songbird requires a basal ganglia circuit. PLoS Biol, 3(5), e153 Pollack G (2000) Who, what, where? Recognition and localization of acoustic signals by insects. Curr Opin Neurobiol 10(6):763–767 Pouso P, Cabana Á, Nieto Methol C (2026) Environmental drivers of calling activity in a southern subtropical anuran assemblage: insights from passive acoustic monitoring. Bioacoustics 35(1):75–89 Rand AS, Ryan MJ (1981) The adaptive significance of a complex vocal repertoire in a neotropical frog. Z für Tierpsychologie 57(3–4):209–214 Ravignani A, Norton P (2017) Measuring rhythmic complexity: a primer to quantify and compare temporal structure in speech, movement, and animal vocalizations. J Lang Evol 2(1):4–19 Rhodes HJ, Yu HJ, Yamaguchi A (2007) Xenopus vocalizations are controlled by a sexually differentiated hindbrain central pattern generator. J Neurosci 27(6):1485–1497 Rivera M, Edwards JA, Hauber ME, Woolley SM (2023) Machine learning and statistical classification of birdsong link vocal acoustic features with phylogeny. Sci Rep 13(1):7076 Ryan MJ (1985) The túngara frog: a study in sexual selection and communication. University of Chicago Press Sakata JT, Hampton CM, Brainard MS (2008) Social modulation of sequence and syllable variability in adult birdsong. J Neurophysiol 99(4):1700–1711 Schmidt RS (1992) Neural correlates of frog calling: production by two semi-independent generators. Behav Brain Res 50(1–2):17–30 Schwartz JJ (1993) Male calling behavior, female discrimination and acoustic interference in the Neotropical treefrog Hyla microcephala under realistic acoustic conditions. Behav Ecol Sociobiol 32(6):401–414 Sismondo E (1990) Synchronous, alternating, and phase-locked stridulation by a tropical katydid. Science 249(4964):55–58 Sober SJ, Wohlgemuth MJ, Brainard MS (2008) Central contributions to acoustic variation in birdsong. J Neurosci 28(41):10370–10379 Sullivan BK (1983) Sexual selection in Woodhouse's toad (Bufo woodhousei). II. Female choice. Anim Behav 31(4):1011–1017 Sullivan BK, Hinshaw SH (1990) Variation in advertisement calls and male calling behavior in the spring peeper (Pseudacris crucifer). Copeia 1990(4):1146–1150 Takahashi DY, Narayanan DZ, Ghazanfar AA (2013) Coupled oscillator dynamics of vocal turn-taking in monkeys. Curr Biol 23(21):2162–2168 Tanner JC, Bee MA (2019) Within-individual variation in sexual displays: signal or noise? Behav Ecol 30(1):80–91 Tanner JC, Bee MA (2020) Inconsistent sexual signaling degrades optimal mating decisions in animals. Sci Adv 6(20):eaax3957 Tanner JC, Shein-Idelson M, Hoke KL, Bee MA (2025) Complex choice environments shelter unattractive signallers from sexual selection. Proceedings of the Royal Society B: Biological Sciences , 292 (2054) Tobias ML, Viswanathan SS, Kelley DB (1998) Rapping, a female receptive call, initiates male–female duets in the South African clawed frog. Proceedings of the National Academy of Sciences , 95 (4), 1870–1875 Tomasello M, Zuberbühler K (2002) Primate vocal and gestural communication. In: Bekoff M, Allen C, Burghardt GM (eds) The cognitive animal: Empirical and theoretical perspectives on animal cognition. MIT Press, pp 293–299 Tumer EC, Brainard MS (2007) Performance variability enables adaptive plasticity of ‘crystallized’adult birdsong. Nature 450(7173):1240–1244 Walker TJ (1969) Acoustic synchrony: two mechanisms in the snowy tree cricket. Science 166(3907):891–894 Waser PM, Waser MS (1977) Experimental studies of primate vocalization: Specializations for long-distance propagation. Z für Tierpsychologie 43(3):239–263 Weiss M, Hultsch H, Adam I, Scharff C, Kipper S (2014) The use of network analysis to study complex animal communication systems: a study on nightingale song. Proceedings of the Royal Society B: Biological Sciences , 281 (1785) Welbergen JA, Davies NB (2008) Reed warblers discriminate cuckoos from sparrowhawks with graded alarm signals that attract mates and neighbours. Anim Behav 76(3):811–822 Woolley SM, Moore JM (2011) Coevolution in communication senders and receivers: vocal behavior and auditory processing in multiple songbird species. Ann N Y Acad Sci 1225(1):155–165 Xing J, Sainburg T, Taylor H, Gentner TQ (2022) Syntactic modulation of rhythm in Australian pied butcherbird song. Royal Soc open Sci, 9 (9) Zelick RD, Narins PM (1982) Analysis of acoustically evoked call suppression behaviour in a neotropical treefrog. Anim Behav 30(3):728–733 Zelick RD, Narins PM (1983) Intensity discrimination and the precision of call timing in two species of neotropical treefrogs. J Comp Physiol 153(3):403–412 Zelick R, Narins PM (1985) Characterization of the advertisement call oscillator in the frog Eleutherodactylus coqui. J Comp Physiol A 156(2):223–229 Ziegler L, Arim M, Bozinovic F (2016) Intraspecific scaling in frog calls: the interplay of temperature, body size and metabolic condition. Oecologia 181:673–681 Ziegler L, Arim M, Narins PM (2011) Linking amphibian call structure to the environment: the interplay between phenotypic flexibility and individual attributes. Behav Ecol 22(3):520–526 Zornik E, Katzen AW, Rhodes HJ, Yamaguchi A (2010) NMDAR-dependent control of call duration in Xenopus laevis. J Neurophysiol 103(6):3501–3515 Additional Declarations No competing interests reported. Supplementary Files boanatimescalessuppsub1.pdf Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 13 May, 2026 Reviews received at journal 13 May, 2026 Reviews received at journal 28 Apr, 2026 Reviewers agreed at journal 23 Apr, 2026 Reviewers agreed at journal 22 Apr, 2026 Reviewers invited by journal 22 Apr, 2026 Editor assigned by journal 21 Apr, 2026 Submission checks completed at journal 21 Apr, 2026 First submitted to journal 18 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9458485","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":628278403,"identity":"1b9146d3-ab59-4842-a67d-6fa1a7883ab7","order_by":0,"name":"Mariana Rodriguez-Santiago","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYJCCAwwMzHJA3AAXkSBGizEDAyNQSwJENUEtQMCc2EC0Ft32sw8PV1RYp/e3NzY+/PnDro6fgfngbR48WszOpBscPHMmPXfGmYPNxjwJyRKSDWzJ1ni1HEhjONjYdjh3g0RimzRDArOEwQEeM2m8Ws4/A2tJNwBqkfyRUC9hf4D/G34tNyC2JIC0SPAkHJYwYOBhI6AFaEvDmXRDiF/SjkvOOMxmbDkHr8PSmD82VFjL87c3H3z4w6aaH8h4eOMNHi1YADNpykfBKBgFo2AUYAEA6iZMAloNelEAAAAASUVORK5CYII=","orcid":"","institution":"Colorado State University","correspondingAuthor":true,"prefix":"","firstName":"Mariana","middleName":"","lastName":"Rodriguez-Santiago","suffix":""},{"id":628278404,"identity":"72f778d6-9c48-49f3-9fb0-173db71b1e54","order_by":1,"name":"Paula Pouso","email":"","orcid":"","institution":"Universidad de la República","correspondingAuthor":false,"prefix":"","firstName":"Paula","middleName":"","lastName":"Pouso","suffix":""},{"id":628278405,"identity":"a01cd791-f192-437b-a75b-5f4738e88d2c","order_by":2,"name":"Kim Hoke","email":"","orcid":"","institution":"Colorado State University","correspondingAuthor":false,"prefix":"","firstName":"Kim","middleName":"","lastName":"Hoke","suffix":""},{"id":628278406,"identity":"3fddf6a0-49d6-40aa-aa15-77d7f4956a27","order_by":3,"name":"Erik Zornik","email":"","orcid":"","institution":"Reed College","correspondingAuthor":false,"prefix":"","firstName":"Erik","middleName":"","lastName":"Zornik","suffix":""}],"badges":[],"createdAt":"2026-04-19 00:38:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9458485/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9458485/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107890985,"identity":"caca86b8-37a1-4235-b8b0-e2a185483fe0","added_by":"auto","created_at":"2026-04-27 10:03:16","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2416118,"visible":true,"origin":"","legend":"\u003cp\u003eNested timescales of vocalizations. \u003cstrong\u003ea\u003c/strong\u003e Example oscillogram of a free-field recording with two calling bouts, denoted by gray bars. \u003cstrong\u003eb \u003c/strong\u003eWithin a calling bout, males produce call series (denoted in blue) that are separated by an inter-series interval (ISI) greater than one second on average. We quantify the inter-series onset (ISO) as the length of time between the start of two consecutive series. \u003cstrong\u003ec \u003c/strong\u003eWithin a series, calls (denoted in purple) are separated by an inter-call interval (ICI) of less than one second and the inter-call onset (ICO) is the length of time between the start of consecutive calls. Calls are composed of notes which are separated by inter-note intervals (INI) of less than 0.5 s. The inter-note onset (INO) is the length of time between note onsets. Asterisks denote an example of a background chorus call that was excluded during automated signal detection.\u003c/p\u003e","description":"","filename":"fig1final174mm.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/bd6000774b01e2b04cd357fa.jpg"},{"id":108181166,"identity":"ae9c840d-9fbc-4a1d-bf5e-ea39f4a9f433","added_by":"auto","created_at":"2026-04-30 08:58:06","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3656344,"visible":true,"origin":"","legend":"\u003cp\u003eRhythmic isochrony of acoustic features depends on the vocal timescales. \u003cstrong\u003ea\u003c/strong\u003e Example oscillogram of multiple call series. Within calling bouts, the inter-series onset (ISO) is the length of time between start times of successive call series (blue). The inter-call onset (ICO) is the length of time between call onsets within a series (purple). The inter-note onset (INO) is the length of time between note onsets within calls (green). \u003cstrong\u003eb-d \u003c/strong\u003eInterval ratio distributions of series (\u003cstrong\u003eb\u003c/strong\u003e), calls (\u003cstrong\u003ec\u003c/strong\u003e), and notes (\u003cstrong\u003ed\u003c/strong\u003e) compared to randomly generated null distributions (in gray). \u003cstrong\u003ee-g\u003c/strong\u003eHistograms of the inter-onset timing distributions of series (\u003cstrong\u003ee\u003c/strong\u003e), calls (\u003cstrong\u003ef\u003c/strong\u003e), and notes (\u003cstrong\u003eg\u003c/strong\u003e). Insets in \u003cstrong\u003ef \u003c/strong\u003eand \u003cstrong\u003eg\u003c/strong\u003e are the zoomed-in distributions around 0-1 seconds. Histogram bin size = 100 ms.\u003c/p\u003e","description":"","filename":"fig2final174mm.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/e741eb51fef791ed13f59166.jpg"},{"id":107890987,"identity":"94076a53-b4f5-4d91-b278-1eb27a5c2881","added_by":"auto","created_at":"2026-04-27 10:03:16","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":84914,"visible":true,"origin":"","legend":"\u003cp\u003eShorter call series are followed by longer inter-series intervals (ISI). ISI decreases significantly with increasing series length (p \u0026lt; 0.01 for pairwise comparison of solo vs. two, three, and 4+ series duration). Plot displays ISI ≤ 5 seconds for clarity; statistical comparisons were performed on ISI \u0026lt; 10 s.\u003c/p\u003e","description":"","filename":"fig3final80mm.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/6103cedce76eff3ccc84b60b.jpg"},{"id":108006161,"identity":"ee30226a-3dc6-416b-a9cf-4b44ad8dd518","added_by":"auto","created_at":"2026-04-28 12:54:09","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1225577,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal features of calls and notes vary dynamically depending on series length (i.e. total number of calls in the series) and within-series position. For all plots, color gradient represents the within-series position of the feature. \u003cstrong\u003ea\u003c/strong\u003e Inter-call onsets (ICO) of mid-calls in three-call series are significantly longer than the first call. Calls within four-call series have longer ICOs than those in three-call series. \u003cstrong\u003eb \u003c/strong\u003eCall duration is shorter in longer series and in mid-positions within three and four-call series. \u003cstrong\u003ec \u003c/strong\u003eCalls in four-call series are longer than those in three-call series. Mid-position calls are longer in series of three and four calls. \u003cstrong\u003ed-f \u003c/strong\u003eAll note features vary significantly depending on within-series position but not depending on the length of the series. Inter-note onset (INO; \u003cstrong\u003ed\u003c/strong\u003e) is shorter in mid-position calls, both note 1 and 2 durations get shorter in mid-position calls, and the inter-note interval (INI) is shorter between mid-position calls (\u003cstrong\u003ef\u003c/strong\u003e). p\u0026lt;0.001***, p\u0026lt;0.01**\u003c/p\u003e","description":"","filename":"fig4final600dpi.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/f7f3085562d06eb6ab19c7e4.jpg"},{"id":109202785,"identity":"4600a945-3b7f-4623-89f1-6626ba51a846","added_by":"auto","created_at":"2026-05-13 14:17:13","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":2926927,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\n\u003c/p\u003e\n\u003cp\u003eVariation in temporal features depends on the timescale. \u003cstrong\u003ea-c \u003c/strong\u003eCV\u003csub\u003ew\u003c/sub\u003e of series, calls, and notes within bouts (\u003cstrong\u003ea\u003c/strong\u003e), within series (\u003cstrong\u003eb\u003c/strong\u003e), and by position in series (\u003cstrong\u003ec\u003c/strong\u003e). \u003cstrong\u003ed-f \u003c/strong\u003e\u0026nbsp;CV\u003csub\u003eb \u003c/sub\u003eof series, calls, and notes within bouts (\u003cstrong\u003ed\u003c/strong\u003e), within series (\u003cstrong\u003ee\u003c/strong\u003e), and by position in series (\u003cstrong\u003ef\u003c/strong\u003e). \u003cstrong\u003eg-i \u003c/strong\u003e\u0026nbsp;CV\u003csub\u003eb\u003c/sub\u003e:CV\u003csub\u003ew \u003c/sub\u003eratio of series, calls, and notes within bouts (\u003cstrong\u003eg\u003c/strong\u003e), within series (\u003cstrong\u003eh\u003c/strong\u003e), and by position in series (\u003cstrong\u003ei\u003c/strong\u003e). CV\u003csub\u003eb\u003c/sub\u003e:CV\u003csub\u003ew\u003c/sub\u003e \u0026gt; 1 means there is more between-male variation than within-male variation. Within bouts: Temporal features of series (duration, ISI, and ISO) all vary more within- than between males and this is driven by high CV\u003csub\u003ew\u003c/sub\u003e values \u0026gt; 30%. ICI and ICO retain more within-male variation while call duration has near-equal between-male to within-male variation. INI, INO and the DF of note 1 have more relative between-male variation although the total CV\u003csub\u003ew\u003c/sub\u003e and CV\u003csub\u003eb\u003c/sub\u003e are nearly identical for all features represented. Within series: The high CV\u003csub\u003ew\u003c/sub\u003e of call features seen at longer timescales drops significantly within-series and is driven by within-series position dynamics. While overall CV\u003csub\u003ew\u003c/sub\u003e and CV\u003csub\u003eb \u003c/sub\u003evalues stay the same for all note features within bouts and further, within series, there is marked between- male variation depending on within-series position for all note features.\u003c/p\u003e","description":"","filename":"fig5final600dpi.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/1028b85e94356a9ce901ccf5.jpg"},{"id":108006819,"identity":"35816975-6b80-4a28-a88e-e31cc185852f","added_by":"auto","created_at":"2026-04-28 12:57:29","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":613353,"visible":true,"origin":"","legend":"\u003cp\u003eAutomated detection of signals reveals novel note types. \u003cstrong\u003ea \u003c/strong\u003eOscillograms and spectrograms of note types identified in this dataset. \u003cstrong\u003eb \u003c/strong\u003eNotes were classified as four types: note 1, 2, note 3 which had a significantly longer duration than notes 1 or 2, and solos that were similar to notes 1 and 2 but occurred in isolation. \u003cstrong\u003ec \u003c/strong\u003e99% of all notes across males were either note 1 or 2, and the newly identified call (‘note 3’) made up 0.8% of all notes. Solo notes have durations and DFs in the range of notes 1 and 2 but appear outside of a two-note pattern.\u003c/p\u003e","description":"","filename":"fig6final400dpi.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/064851068cc871a8fbd3819e.jpg"},{"id":109252241,"identity":"fdfc8092-b851-4cf5-bc1d-4b414a3e1c17","added_by":"auto","created_at":"2026-05-14 09:23:54","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":11270617,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/4d6ce5bd-f5cf-4ada-962f-dfca3f0849f8.pdf"},{"id":108006389,"identity":"7d165016-ef02-454b-86a0-c985dfbad4a0","added_by":"auto","created_at":"2026-04-28 12:55:19","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":430468,"visible":true,"origin":"","legend":"","description":"","filename":"boanatimescalessuppsub1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9458485/v1/0a28da343342c8b946f67cbc.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Timescales of call variability in a South American treefrog","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eAcoustic communication systems enable individuals to convey information that is crucial for a wide range of behaviors (Gerhardt, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). From territory defense to courtship signaling to individual recognition, senders tailor their vocal cues to elicit context-appropriate responses from receivers. The production of these signals is energetically costly (Gerhardt \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1994\u003c/span\u003e; Gillooly \u0026amp; Ophir \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) and can have detrimental results such as exposing signalers to predators (Rand \u0026amp; Ryan \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e1981\u003c/span\u003e). To mitigate costs and facilitate context-specific communication, signalers often modify aspects of their vocalizations such as the amplitude or rhythmic timing of their signals (Humfeld \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Welbergen \u0026amp; Davies \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Bhat et al. \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Understanding the boundaries of variation in signaler vocalizations informs the physiological mechanisms and constraints that regulate call patterning and communication.\u003c/p\u003e \u003cp\u003eVocalizations consist of spectral and temporal properties that contain important information such as species, social context (e.g. reproductive or competitive), and signaler condition. Spectral properties, like fundamental frequency, often play a role in species recognition, can correlate with signaler size and condition, and tend to be relatively invariant over time (Marler and Peters \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e1977\u003c/span\u003e; Ryan \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e1985\u003c/span\u003e; Gerhardt \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Gerhardt and Bee \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Woolley and Moore, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Moore and Woolley, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Rivera et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Meanwhile, temporal features can change markedly within short and long timescales (Gerhardt \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Pollack \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Sakata et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Call rate, for example, varies between and within individuals over time, fluctuates in both emission rate and periodicity (Narins \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1982\u003c/span\u003e; Lopez \u0026amp; Narins \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Klump \u0026amp; Gerhardt \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Tumer \u0026amp; Brainard \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), and can depend on factors like temperature and social cues (Baraquet et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Furtado et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Other timing features like call duration and inter-call intervals can vary across timescales depending on the acoustic environment (Zelick and Narins, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e1985\u003c/span\u003e; Greenfield \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e1988\u003c/span\u003e; Tobias et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Brumm \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Faster call rates often correlate with greater mating success (Sullivan \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1983\u003c/span\u003e; Gibson \u0026amp; Bradbury \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e1985\u003c/span\u003e; Morris \u0026amp; Yoon \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e1989\u003c/span\u003e; Sullivan \u0026amp; Hinshaw \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e1990\u003c/span\u003e; McComb \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1991\u003c/span\u003e), illustrating the importance of dynamic temporal modulation in shaping communication systems.\u003c/p\u003e \u003cp\u003eBeyond their context-dependent functions, the timescales of temporal variability can also shed light on the physiological mechanisms that drive their production. In songbirds for example, within-individual variability in the acoustic features of syllables is linked to the trial-by-trial differences in the activity of ensembles of neurons (Sober et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). Furthermore, variability in the pattern of neural activity is associated with variability in song structure during learning (Olveczky et al. 2005; Kao et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Thus, the temporal structure of vocalizations and their variability over time can illustrate the mechanistic principles that generate them.\u003c/p\u003e \u003cp\u003eIn anurans, vocalizations are composed of acoustic units that tend to be quite stereotyped within species but vary in their rate of production, timing, and sequencing depending on the context. Since Peter Narins\u0026rsquo; foundational work that devised rigorous methods to measure call timing and its variability (Narins \u0026amp; Capranica \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e1977\u003c/span\u003e; Narins \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1982\u003c/span\u003e; Zelick \u0026amp; Narins \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e1982\u003c/span\u003e; Zelick \u0026amp; Narins \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e1983\u003c/span\u003e), few studies have rigorously characterized vocal rhythms and temporal variation across multiple timescales. The relatively recent development of generalizable rhythm analysis schema to quantify rhythmic communication across species offers new potential for insights into both evolutionary trajectories and the underlying mechanisms of vocal communication systems (Ravignani \u0026amp; Norton \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Burchardt \u0026amp; Kn\u0026ouml;rnschild \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Burchardt et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Anichini et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jadoul et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Here, we adopt these methods and modify their application to examine the emergence of rhythmicity in anuran vocalizations. By coupling these analyses with measures of between- and within-male variation in temporal features, we aim to inform the potential mechanisms that generate rhythmic calling across multiple timescales.\u003c/p\u003e \u003cp\u003e \u003cem\u003eBoana pulchella\u003c/em\u003e is a South American treefrog that lives in ponds and forested streams of Uruguay, southern Brazil, southern Paraguay, and northeastern Argentina, and can be heard calling not just at marked breeding seasons but throughout the year (Canavero et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Pouso et al. 2025). Males of this species produce a simple doublet call consisting of two notes with spectral features that vary depending on body size and environmental temperature (Basso \u0026amp; Basso, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1987\u003c/span\u003e; Ziegler et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Ziegler et al. \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Although previous studies have noted that the temporal patterning males use when calling in large choruses are highly dynamic (Basso \u0026amp; Basso \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1987\u003c/span\u003e; Canavero et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), these vocal patterns and their variation remain uncharacterized across time scales.\u003c/p\u003e \u003cp\u003eUsing free-field recordings of \u003cem\u003eB. pulchella\u003c/em\u003e, we examine variation of temporal features across timescales to generate hypotheses about the underlying oscillatory and motor control mechanisms organizing acoustic signals. During calling bouts, males can be heard appending multiple calls in series, ranging in number from a single call to five or more. We examine the rhythmic structure of these vocalizations, quantifying the periodicity in their call timing across timescales. We also examine the extent of between- and within-male variation in the timing features of their calls. Ultimately, understanding the timescales of both temporal modulation and individual variation can shed light on the potential physiological mechanisms that generate and constrain flexible communication systems.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy system and acoustic recordings\u003c/h2\u003e \u003cp\u003eThe calls of 22 \u003cem\u003eB. pulchella\u003c/em\u003e males were recorded in Paraje Zanja del Tigre, Maldonado, Uruguay in March 2020. At this site, males are readily abundant in the foliage and in the water. We held unidirectional Sennheiser microphones (Sennheiser MKE 400) connected to a digital recorder (Olympus LS 11) approximately 20 cm from each calling male to record focal males for 5 minutes. All recordings were performed between the hours of 23:00\u0026ndash;2:00. After taking each recording, we measured the snout-vent-length (SVL), femur length, and weight of each male. Humidity levels (Extech RHT510)\u003c/p\u003e \u003cp\u003eand air temperature (HOBO UA-002-64, Onset Computer Corporation) were logged every night.\u003c/p\u003e \u003cp\u003eOf the 22 males recorded, four recordings were excluded due to high background noise impeding our ability to differentiate the focal male from the chorus. Signals were detected using Raven Pro 1.6 (Cornell Laboratory of Ornithology, Ithaca, NY; Charif et al, 2010). We distinguished focal males from the chorus using amplitude detection on the oscillograms with parameters set by eye based on individual recording conditions. We obtained start and end times of each note and verified each automatic detector by hand to ensure correct detection. All background chorus calls were excluded during the selection process. Tabulated selections were exported as .txt files per focal male and imported into RStudio for subsequent analysis. We used the package \u0026lsquo;warbleR\u0026rsquo; (Arayas-Salas 2017) to measure spectrotemporal features of each note (for a full list of features measured, see Supp. Materials).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAnalysis of temporal features across timescales\u003c/h3\u003e\n\u003cp\u003ePreliminary analyses suggested that male call rates vary across timescales \u0026ndash; from calling bouts that range in duration anywhere from several seconds to several minutes (Fig. \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea), to call series within these bouts where calls are separated by silences of 1-10 seconds (Fig. \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), to a within-series timescale where calls are separated by silences \u0026lt; 1 sec on average (Fig. \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). These timescales were initially identified through visual inspection of the distribution of inter-call onsets (ICOs), which is the length of time between the start of consecutive doublet calls. The distribution of ICOs suggested the existence of two peaks \u0026ndash; one very sharp, short-peak (\u0026lt; 1 s) and another peak distributed more widely around 1.8 s. We selected the timescale cutoffs for these ICOs based on within-male distributions to determine each male\u0026rsquo;s call timing thresholds (see Supp. Fig. \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e for examples). We identified x-axis min and max numbers for each peak distribution range and extracted the x-axis max value for the first distribution, deemed the short-peak distribution trough. Although the exact ICO distributions vary by individual male, each male had a distinct trough value which we used to bin calls as belonging to long (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{X}\\)\u003c/span\u003e\u003c/span\u003e \u0026gt; 1s) or short (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\stackrel{-}{X}\\)\u003c/span\u003e\u003c/span\u003e \u0026lt; 1s) timescales. We capped the analysis to calling bouts that are separated by intervals of silence of 10+ seconds. Within these calling bouts, we characterized the temporal features (\u003cem\u003eduration, onset\u003c/em\u003e and \u003cem\u003einterval\u003c/em\u003e) of three vocal timescales: the \u003cem\u003enotes\u003c/em\u003e that make up each call, doublet \u003cem\u003ecalls\u003c/em\u003e, and \u003cem\u003eseries\u003c/em\u003e composed of rapid series of calls separated by short ICOs.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eTable 1\u003c/strong\u003e. \u003cstrong\u003eGlossary of temporal features quantified.\u003c/strong\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable border=\"1\"\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTerm\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDefinition\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ecall (doublet)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003etwo-note sound unit consisting of a short note 1 and longer note 2 separated from subsequent call by a silence interval longer than the interval between note 1 and 2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-call onset (ICO)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between onset of note 1 in focal call and note 1 in following call\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-call interval (ICI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between offset of note 2 in focal call and note 1 onset in following call\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003enote\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003esubunit of a call, varies in duration\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-note onset (INO)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between onset of note 1 and note 2 in a doublet call\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-note interval (INI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between offset of note 1 and onset of note 2 in a doublet call\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(call) series\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ea group of calls separated from other such groups by periods of silence much longer than the inter-call intervals\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-series onset (ISO)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between onsets of successive call series\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003einter-series interval (ISI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003elength of time between offset of one call series and onset of subsequent call series\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo summarize, in Table\u0026nbsp;1 we provide definitions of the major temporal features of vocalizations quantified. These definitions are adapted from Kohler et al. 2017 and modified to fit the call structure of \u003cem\u003eB. pulchella.\u003c/em\u003e The ICO as defined here is equivalent to the inter-onset interval (IOI) used broadly in other studies. The calculations for all temporal features measured \u0026ndash; \u003cem\u003edurations, onsets\u003c/em\u003e and \u003cem\u003eintervals\u003c/em\u003e \u0026ndash; was the same across all timescales, the only difference is the vocal timescale being measured (i.e. \u003cem\u003eseries\u003c/em\u003e, \u003cem\u003ecall\u003c/em\u003e, or \u003cem\u003enote\u003c/em\u003e). We quantified variation in these temporal features between and within males within calling bouts and within series. All analyses were performed in RStudio (v7.1; R Core Team 2021) using custom scripts - one script for series, one for calls, and one for notes - such that no timescale was duplicated in the analysis. Specifically, in the \u003cem\u003eseries\u003c/em\u003e data frame, each row represents a single call series and contains identifying information such as its begin and end time, duration, inter-series onset (ISO), inter-series interval (ISI), as well as qualitative measures such as the series length, or total number of calls within the series (solo, two, three, four, five). In the \u003cem\u003ecall\u003c/em\u003e data frame, each row represents a single call and includes columns detailing its beginning and end time, duration, ICO, inter-call interval (ICI), as well as the series length it belongs to (solo, two, etc) and the order in the series it appears in (within-series position: 1\u0026thinsp;=\u0026thinsp;first call in series, 2\u0026thinsp;=\u0026thinsp;second, etc). Many calls were not emitted within a call series, and these \u0026lsquo;solo\u0026rsquo; calls were omitted in within-series analyses. Given that not every series length is represented equally, we binned the \u0026lsquo;within-series position\u0026rsquo; into categories: first and mid. In the \u003cem\u003enotes\u003c/em\u003e script, each row represents a note within a call and includes columns with its begin and end time, duration, inter-note onset (INO), inter-note interval (INI), mean dominant frequency, the series length it belongs to, and within-series position.\u003c/p\u003e\n\u003ch3\u003eAssessing rhythmic structure across timescales of vocalizations\u003c/h3\u003e\n\u003cp\u003eThe structure of rhythmic vocalizations is often an indicator of the type of mechanism (i.e. physiological oscillators) that generates the signal. To quantify the rhythmic structure of calling, we followed a general rhythm analysis schema established by Ravignani \u0026amp; Norton \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2017\u003c/span\u003e and expanded on by Burchardt \u0026amp; Kn\u0026ouml;rnschild \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2020\u003c/span\u003e (see also Anichini et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jadoul et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). We use one specific method, the interval ratio (IR), to determine whether the call patterning of \u003cem\u003eB. pulchella\u003c/em\u003e is isochronous and periodic. Isochronous signals follow metronome-like beats such that the interval between consecutive signals is uniform. Periodicity, or the underlying temporal patterns within signal series, indicates potential mechanisms and physiological constraints that generate their production. Here, we quantified the IR of signals across vocal timescales (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). All interval ratios were calculated such that:\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ewhere X\u003csub\u003ek\u003c/sub\u003e is the inter-signal onset for the kth call. All ratios were calculated within calling bouts such that between-bout interval ratios were excluded from all analyses.\u003c/p\u003e \u003cp\u003eWe initially assessed whether each vocal timescale was produced isochronously by quantifying all observed interval ratios between males and assessing their distribution around 0.5. The shape of these distributions can identify rhythmic isochrony such that a unimodal distribution centered around 0.5 suggests isochrony. To characterize periodicity more generally, we generated a random null distribution of interval ratios by randomly sampling pairs of intervals (with replacement) from each male\u0026rsquo;s observed range, calculating the interval ratio for each pair, and repeating this process to produce approximately 6,000 simulated ratios across individuals. We used a two-sample Kolmogorov-Smirnov (KS) test to assess potential periodicity in the structure of these vocalizations by determining whether the observed interval ratio distributions differed significantly from the null distributions. Non-isochronous periodicities were identified by peak distributions in the interval ratios not centered around 0.5.\u003c/p\u003e \u003cp\u003eTo examine whether isochronous call series timing is consistent with a pacemaker-like mechanism that generates series onsets independent of series duration or with a consistent silent period between call series, we ran a linear mixed-effects model using the \u0026lsquo;lme4\u0026rsquo; (Bates et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) package to test how the length of the series impacts the ISI with frogID as a random effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Given that series longer than three calls were less prevalent than solo or two-call series, we binned series of four or more calls (deemed here \u0026lsquo;four+\u0026rsquo; or \u0026lsquo;4+) for statistical analyses.\u003c/p\u003e\n\u003ch3\u003eExamining how periodicity within short timescales varies with series dynamics\u003c/h3\u003e\n\u003cp\u003eTo assess how temporal features of calls and notes vary depending on within-series dynamics, we ran a set of mixed models examining how series length and position within series affect the duration, silence interval, and onset intervals of calls and notes. We limited the analysis to series of 3\u0026ndash;5 calls and examined how features vary depending on the duration of the series (three, four, five calls) and their position within the series (first and mid call in series). We excluded series of two calls because inherently they do not have mid-position calls, limiting our comparisons. In addition, given the nested nature of calling timescales, we excluded calls in the last position because their interval timing is part of the series timescale. Similarly, we excluded note 2 onset and interval timing from analyses because their interval timing is part of the call timescale.\u003c/p\u003e \u003cp\u003eWe initially assessed normality of residuals visually using Q-Q plots. We used the \u0026lsquo;lme4\u0026rsquo; package for linear mixed models, \u0026lsquo;car\u0026rsquo; for testing significance of fixed effects with type III Wald tests (Fox \u0026amp; Weisberg 2019), and \u0026lsquo;emmeans\u0026rsquo; (Lenth et al. 2024) for contrasts. Contrasts were performed to understand significant interactions and main effects with Bonferroni correction. All results were considered significant at the \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 level. Marginal and conditional R\u003csup\u003e2\u003c/sup\u003e were calculated using the \u0026lsquo;performance\u0026rsquo; package (L\u0026uuml;decke et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). We detail the statistical models run for each vocal timescale below.\u003c/p\u003e \u003cp\u003e \u003cem\u003eCalls\u003c/em\u003e: To examine whether the temporal features of calls (response variables: duration, ICI, and ICO) vary depending on the series duration (three, four, or five calls) or sequence order within series (first and mid calls), for each response variable we initially compared two models: an unrestricted model including individual frog identity as a random intercept and the call series duration and position within series and their interaction as fixed effects, and a restricted model without the random effect. All models with the lowest AIC included frogID as random effects. Final models are detailed in Supp. Table\u0026nbsp;1.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eNotes\u003c/strong\u003e \u003cp\u003eWe examined how note features vary depending on series duration and sequence order within series using a similar model selection process comparing restricted and unrestricted models as described for calls. In this case, since there are two note types within doublet calls (notes 1 and 2), we ran separate models per note to examine differences in note durations and dominant frequencies (DFs) depending on series duration and within-series position. To examine whether the silence interval and onset difference between notes 1 and 2 of each doublet call vary depending on series duration and sequence order, we ran separate models with INI and INO as the dependent variables. Results of the note features models can be found in Supp. Table\u0026nbsp;2.\u003c/p\u003e \u003c/p\u003e\n\u003ch3\u003eQuantifying within- and between-male variation in spectrotemporal features across timescales\u003c/h3\u003e\n\u003cp\u003eTo quantify variation between and within males across calling timescales, we calculated the coefficient of variation for all series, call, and note features within (CV\u003csub\u003ew\u003c/sub\u003e) and between males (CV\u003csub\u003eb\u003c/sub\u003e). CVs were calculated such that:\u003c/p\u003e \u003cp\u003eCV\u003csub\u003ew\u003c/sub\u003e = sd\u003csub\u003ew\u003c/sub\u003e / X̄\u003csub\u003ew\u003c/sub\u003e\u003c/p\u003e \u003cp\u003eCV\u003csub\u003eb\u003c/sub\u003e = sd\u003csub\u003eb\u003c/sub\u003e / X̄\u003csub\u003eb\u003c/sub\u003e\u003c/p\u003e \u003cp\u003eFor example, a smaller CV\u003csub\u003ew\u003c/sub\u003e indicates acoustic features with low variation and high consistency within the many calls produced by an individual male. A small CV\u003csub\u003eb\u003c/sub\u003e indicates that a particular acoustic feature has low variation and high consistency between all the males. The ratio of CV\u003csub\u003eb\u003c/sub\u003e / CV\u003csub\u003ew\u003c/sub\u003e indicates to what degree callers tend to differ more from one another compared to their own intra-individual variation over time, such that a CV ratio\u0026thinsp;\u0026gt;\u0026thinsp;1 indicates an acoustic feature that varies more between than within males while a CV ratio\u0026thinsp;\u0026lt;\u0026thinsp;1 indicates relatively higher within-male variation and greater overlap between males. We compare how the temporal features of series, calls, and notes vary within calling bouts, and how call and note features vary within series. For within-series comparisons, we also examine how variation depends on the sequence position (see Supp. Table\u0026nbsp;3 for a comprehensive table detailing all timescale CV results). Specifically, for the within-series comparison, we compare the CV ratios of duration, interval, and onsets of calls and notes for first and mid calls in a series that contained at least three calls. For note features, we also examined how spectrotemporal features vary within a series. We highlight variation in the DF of notes 1 and 2 across timescales throughout the results. Although quantifying full spectral variation was outside of the scope of this paper, we provide CV summaries for all spectrotemporal features measured (listed in the Supp. Materials) in Supp. Table\u0026nbsp;4.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCharacterizing novel note types\u003c/h2\u003e \u003cp\u003eUsing nomenclature originally established by Basso \u0026amp; Basso \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1987\u003c/span\u003e and expanded on by Baraquet et al. 2007 and Ziegler et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2011\u003c/span\u003e, we initially identified calls based on the temporal sequence of notes 1 and 2 within doublet calls. We found that some notes fell outside of previously established note parameters \u0026ndash; namely (1) notes that are not part of two-note calls but their duration and DF ranges overlap with either note 1 and note 2 (deemed \u0026lsquo;solos\u0026rsquo;), and (2) a novel single-note call type with a substantially longer duration than notes 1 and 2 (deemed \u0026lsquo;note 3\u0026rsquo; and named \u0026ldquo;squeaks\u0026rdquo;). Exact ranges of features vary by individual male thus we automated note waveform selection based on individual male cutoffs. In Supp. Table\u0026nbsp;5, we summarize the mean\u0026thinsp;\u003cem\u003e\u0026plusmn;\u003c/em\u003e\u0026thinsp;sd and ranges of all identified note features.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eEvidence for isochronous and non-isochronous periodicity depends on timescales\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWe initially set out to examine the rhythmic patterning of \u003cem\u003eB. pulchella\u003c/em\u003e vocalizations by quantifying the interval ratio (IR) distributions between the onsets of identified vocal timescales \u0026ndash; series, calls, and notes \u0026ndash; within calling bouts. We found striking IR results across timescales. Specifically, the IR distribution of call \u003cem\u003eseries\u003c/em\u003e was unimodal and centered around 0.5 (x̄ = 0.5, sd\u0026thinsp;=\u0026thinsp;0.08), suggesting an isochronous call \u003cem\u003eseries\u003c/em\u003e signal within bouts. The observed distribution did not differ from a randomly generated null distribution (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb; null x̄ = 0.5, sd\u0026thinsp;=\u0026thinsp;0.09, two-sample KS test: D\u0026thinsp;=\u0026thinsp;0.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.13). The ISO distribution showed one peak between 1\u0026ndash;2 sec (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee), suggesting this is the typical series timing.\u003c/p\u003e \u003cp\u003eWe found that the call IR had multiple peaks, inconsistent with an isochronous rhythmic structure (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). The observed call IR distribution differed significantly from the null distribution (D\u0026thinsp;=\u0026thinsp;0.09, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; null: x̄ = 0.49, sd\u0026thinsp;=\u0026thinsp;0.04), suggesting non-isochronous periodicity. The ICO distribution had two primary peaks between (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). The longer ICO peak centered at 1.5 sec (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef) represents the inter-call-onsets between a solo doublet or the last doublet in a call series and the next call series. The sharp ICO peak at a shorter timescale (ICO\u0026thinsp;\u0026lt;\u0026thinsp;1 sec; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef inset) represents the regular timing between doublets within call series.\u003c/p\u003e \u003cp\u003eThe note IR was also not isochronous based on the presence of multiple IR peaks and the lack of any peak around 0.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). This distribution differed significantly from the randomly generated null distribution, which had additional peaks including one centered around 0.5 (D\u0026thinsp;=\u0026thinsp;0.2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001; null: x̄ = 0.5, sd\u0026thinsp;=\u0026thinsp;0.03). The INO distribution had multiple peaks in less than 1 second (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg inset).\u003c/p\u003e \u003cp\u003eGiven that series are isochronous despite variation in series durations, we examined whether the isochronous series timing is maintained by changes in the silence interval after call series. We found that longer call series were followed by significantly shorter ISIs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Pairwise contrasts show that solo calls have significantly longer ISIs than those of longer series (solo vs. two: t\u0026thinsp;=\u0026thinsp;9.43, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; solo vs. three: t\u0026thinsp;=\u0026thinsp;8.18, p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; solo vs. four+: t\u0026thinsp;=\u0026thinsp;3.74, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001); ISIs of two-call series were significantly longer than three-call series (t\u0026thinsp;=\u0026thinsp;3.61, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001); and series of three calls do not differ significantly from longer call series (p\u0026thinsp;=\u0026thinsp;0.81). These results show an inverse relationship between series length and the interval of silence that follows it.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eTiming of calls and notes varies depending on within-series dynamics\u003c/h2\u003e \u003cp\u003eTo understand within-series temporal dynamics, we examined the variation in the onset intervals, silence intervals, and durations of calls and notes depending on series length and their relative position within the series. We restricted these analyses to series containing at least three calls to directly compare not just how series length affects temporal features but also the position within series (e.g., series of two calls do not have mid-position calls which would limit our comparisons).\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCalls\u003c/strong\u003e \u003cp\u003eWe examined whether series length and sequence order within series influenced the ICO, call duration, and ICI (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, b \u0026amp; c). We found that calls in series containing a total of four calls have a significantly longer ICO than those in three-call series (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea; estimate\u0026thinsp;=\u0026thinsp;0.073, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and that calls in mid-series positions have longer ICOs than those first in series (estimate\u0026thinsp;=\u0026thinsp;0.026, p\u0026thinsp;=\u0026thinsp;0.005). Call durations get shorter in longer series (estimate = -0.016, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and are consistently and significantly shorter in mid-series positions compared to first calls (estimate = -0.016; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). ICI of calls in four-call series are slightly longer than those in three-call series (estimate\u0026thinsp;=\u0026thinsp;0.073, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and for calls in mid-series compared to first (estimate\u0026thinsp;=\u0026thinsp;0.042, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Random effects explained a large portion of the variance in call duration (conditional R\u0026sup2; = 0.76) but less so for ICI (conditional R\u0026sup2; = 0.48) and ICO (conditional R\u0026sup2; = 0.51). See Supp. Table\u0026nbsp;1 for a summary of all model comparisons.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eNotes\u003c/em\u003e: Across the board, temporal features of notes (durations, INI, and INO) vary distinctly depending on the within-series position regardless of the series duration (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed, e \u0026amp; f). The INO between notes 1 and 2 was significantly lower in mid-series calls than those first in series (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed; estimate = -0.008, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Duration of both notes 1 and 2 was significantly shorter in mid-series calls than those first in series (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee; note 1: estimate = -0.002, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, note 2: estimate = -0.008, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The INI was also significantly shorter in mid-position calls compared to those first in series (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef; estimate = -0.006, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). For all models, random effects explained a substantial portion of the variance (conditional R\u0026sup2; ranges between 0.775\u0026ndash;0.853 for note 2 duration, INI, and INO). See Supp. Table\u0026nbsp;2 for a summary of all model comparisons.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eWithin- and between-male variation in spectrotemporal features depends on the timescale\u003c/h2\u003e \u003cp\u003eTo understand how the above between- and within-series dynamics contribute to the overall patterns of variation across timescales, we initially quantified the CV\u003csub\u003ew,\u003c/sub\u003e CV\u003csub\u003eb,\u003c/sub\u003e and the CV\u003csub\u003eb\u003c/sub\u003e / CV\u003csub\u003ew\u003c/sub\u003e ratio of series, calls and notes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), and then examined the extent to which position within series and series duration accounts for the patterns of overall variation. We report the patterns of variation below, presented in order by the timescale of analysis. A table detailing all the CV\u003csub\u003ew,\u003c/sub\u003e CV\u003csub\u003eb,\u003c/sub\u003e and CV\u003csub\u003eb\u003c/sub\u003e / CV\u003csub\u003ew\u003c/sub\u003e ratio values for series, calls, and notes can be found in Supp. Table\u0026nbsp;3.\u003c/p\u003e \u003cp\u003e \u003cem\u003eSeries\u003c/em\u003e: We find that all series-level temporal features (series duration, ISI, and ISO) vary more within males over time than between males (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, d, g; CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: series duration\u0026thinsp;=\u0026thinsp;0.48; ISI\u0026thinsp;=\u0026thinsp;0.34; ISO\u0026thinsp;=\u0026thinsp;0.37). The degree of CV\u003csub\u003ew\u003c/sub\u003e is nearly double that of CV\u003csub\u003eb\u003c/sub\u003e for all features (CV\u003csub\u003ew\u003c/sub\u003e: duration\u0026thinsp;=\u0026thinsp;68.8, ISI\u0026thinsp;=\u0026thinsp;42.7, ISO\u0026thinsp;=\u0026thinsp;36.6; CV\u003csub\u003eb\u003c/sub\u003e: duration\u0026thinsp;=\u0026thinsp;33.22, ISI\u0026thinsp;=\u0026thinsp;14.69, ISO\u0026thinsp;=\u0026thinsp;13.39).\u003c/p\u003e \u003cp\u003e \u003cem\u003eCalls\u003c/em\u003e: Within calling bouts, we find that the ICI and ICO have a greater degree of within-male variation than between-male variation compared to call duration (CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: ICI\u0026thinsp;=\u0026thinsp;0.34, ICO\u0026thinsp;=\u0026thinsp;0.34; CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: duration\u0026thinsp;=\u0026thinsp;1.06; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, d and g). When we examined these patterns of variation within series, we found that the degree of between-male variation compared to within-male variation is much more similar among these call features (CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: duration\u0026thinsp;=\u0026thinsp;1.21, ICI\u0026thinsp;=\u0026thinsp;0.96, ICO\u0026thinsp;=\u0026thinsp;1.15; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, e, and h) and the patterns of variation depend on the within-series position order (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, f and i). Specifically, calls in the middle of series have a greater degree of overall variation (both within- and between-males) than the first call in a series (CV\u003csub\u003ew\u003c/sub\u003e : call duration \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;5.92, mid\u0026thinsp;=\u0026thinsp;7.86; ICI \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;13.48, mid\u0026thinsp;=\u0026thinsp;27.76; ICO \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;9.04, mid\u0026thinsp;=\u0026thinsp;20.59; CV\u003csub\u003eb\u003c/sub\u003e : call duration \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;15.04, mid\u0026thinsp;=\u0026thinsp;13.47; ICI \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;16.9, mid\u0026thinsp;=\u0026thinsp;34.56; ICO \u0026minus;\u0026thinsp;1st\u0026thinsp;=\u0026thinsp;14.69, mid\u0026thinsp;=\u0026thinsp;26.94).\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eNotes\u003c/strong\u003e \u003cp\u003eWithin bouts of calling, the duration of note 1 is the only note-level feature that has a higher level of within- than between-male variation (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eg). Within series, the degree of between- to within-male variation in the INI and INO increases when taking the within-series position into account.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSpecifically, the degree of between-male to within-male variation in INI and INO increases in mid-position calls compared to first calls (1st CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: INI\u0026thinsp;=\u0026thinsp;2.84, INO\u0026thinsp;=\u0026thinsp;4.36; mid CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e: INI\u0026thinsp;=\u0026thinsp;2.91, INO\u0026thinsp;=\u0026thinsp;4.43). The position-dependent increase in CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e is driven by a decrease in CV\u003csub\u003ew\u003c/sub\u003e for mid-position calls (INI: 1st CV\u003csub\u003ew\u003c/sub\u003e =7.05, mid CV\u003csub\u003ew\u003c/sub\u003e= 4.61; INO: 1st CV\u003csub\u003ew\u003c/sub\u003e= 5.36, mid CV\u003csub\u003ew\u003c/sub\u003e = 3.42). Variation in the individual note durations also depends on within-series patterning such that both note durations vary more between males than within, and the degree of between-male variation depends on within-series position (note 1 duration: 1st CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e = 1.36, mid\u0026thinsp;=\u0026thinsp;1.78; note 2 duration: 1st CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e = 2.88, mid\u0026thinsp;=\u0026thinsp;1.73). Overall, the CV\u003csub\u003ew\u003c/sub\u003e for the two note DFs is among the lowest CV\u003csub\u003ew\u003c/sub\u003e overall. We find that note DFs have a higher degree of between-male than within-male variation within series based on the ratio of CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e (note 1 DF: CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e = 1.12, note 2 DF: CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e = 1.13). CV\u003csub\u003ew\u003c/sub\u003e is higher in mid-position notes than first notes (note 1 DF: 1st CV\u003csub\u003ew\u003c/sub\u003e = 1.8, mid CV\u003csub\u003ew\u003c/sub\u003e = 5.37; note 2 DF: 1st CV\u003csub\u003ew\u003c/sub\u003e = 3.57, mid CV\u003csub\u003ew\u003c/sub\u003e = 6.51). Overall, note DFs retain nearly double the degree of between-male than within-male variation for first position within series (note 1 DF: 1st CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e =3.26, mid CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e 1.4; note 2 DF: 1st CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e 2.28, mid CV\u003csub\u003eb\u003c/sub\u003e/CV\u003csub\u003ew\u003c/sub\u003e = 1.2), and this is likely because the within-male variation in note DF is lower for the 1st call compared to mid calls.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eDetailed note analysis reveals novel and distinct note types\u003c/h2\u003e \u003cp\u003eWe initially identified calls based on the temporal sequence of notes 1 and 2 and found that there were at least three distinct note types that differ in spectral and temporal properties (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Out of a total of 5701 individual notes across 18 males, 17 were \u0026lsquo;solos\u0026rsquo; and 48 were note 3. Given their low occurrence rate in the dataset, we did not analyze features of solo notes and merely note their existence. We then compared the spectrotemporal features of notes 1, 2, and 3. We found that all three notes vary significantly in duration and dominant frequency (duration: \u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e =4181.5, df\u0026thinsp;=\u0026thinsp;2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001, dom frequency: \u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e =346.83, df\u0026thinsp;=\u0026thinsp;2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; Supp. Table\u0026nbsp;5). While note 1 has a significantly shorter inter-note interval (INI) than notes 2 and 3 (\u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e =4169.8, df\u0026thinsp;=\u0026thinsp;2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), notes 2 and 3 do not differ in INI. In addition, note 1 has a significantly shorter INO than notes 2 and 3 while these notes are not significantly different from each other (\u003cem\u003eX\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e =4170.3, df\u0026thinsp;=\u0026thinsp;2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001; pairwise 1\u0026ndash;2 and 1\u0026ndash;3 p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, pairwise 2\u0026ndash;3 p\u0026thinsp;=\u0026thinsp;0.52).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eWe initially set out to characterize and quantify the temporal dynamics of \u003cem\u003eB. pulchella\u003c/em\u003e vocalizations by quantifying the distributions of call IOIs, which is an established temporal feature measured in anuran communication studies more broadly. Consistent with previous reports in \u003cem\u003eB. pulchella\u003c/em\u003e, we found that the vast majority of vocalizations consist of note pairs comprising a doublet call (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e; Basso \u0026amp; Basso, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1987\u003c/span\u003e; Canavero et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Ziegler et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Ziegler et al. \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Where we expand on the existing literature is that, in quantifying the call onset distribution, we found a series of nested vocal features. By using a rhythm-analysis approach, we found that male calling patterns occur on multiple vocal timescales, and the degree of between- and within-male variability in temporal features depends on timescale dynamics. While call series (also referred to as \u0026ldquo;trains\u0026rdquo; or \u0026ldquo;bouts\u0026rdquo;) have been described in several frog species, the rhythmic nature of these vocal patterns have not been quantified. Our results suggest that long timescale rhythmicity is not limited to elaborate or learned vocal systems, and the neural underpinnings of vocal patterning across a wide range of timescales may be ancient among tetrapods. While we highlight the dynamic vocal behavior of \u003cem\u003eB. pulchella\u003c/em\u003e specifically, we encourage similar multi-timescale analyses across other frog species to gain an evolutionary perspective on rhythm generation complexity and its role in communication systems.\u003c/p\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDifferent rhythmic patterns emerge across nested vocal timescales\u003c/h2\u003e \u003cp\u003eOur findings show that \u003cem\u003eB. pulchella\u003c/em\u003e calls were clustered into repeating series of variable length, typically between 1\u0026ndash;4 calls (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). While the number of calls in each series varied both within and between males, the series themselves were regularly timed and isochronous (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). This regular series timing appears to be driven by the relationship between the total number of calls (series duration) and the silence interval that follows, such that longer series are followed by shorter silence intervals (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eUnlike the longer inter-series timing, interval ratios at shorter timescales\u0026mdash; between calls and between notes\u0026mdash; exhibited non-isochronous patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec and d). While the call interval ratio distribution did have a peak centered at 0.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), there were two other peaks around 0.2 and 0.8. The peak distributions around 0.2 and 0.8 likely result from consecutive calls with very different ICOs (e.g. a two-doublet call series with an ICO of 0.3 s for the first call and 1.5 s for the second call). The peak distribution around 0.5 could result from consecutive calls within a series (each with an ICO of around 0.3 seconds; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef inset) or from consecutive solo calls that are not part of a series (each with an ICO around 1.5 seconds, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef).\u003c/p\u003e \u003cp\u003eThe note interval ratio distribution has four peaks, two below 0.5 and two above 0.5 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). The existence of four peaks is a result of the regular repetition of two-notes within doublet calls that have two distinct inter-onset-intervals. For example, the note interval ratios below 0.5 belong to doublet calls within a series while the longer note interval ratios above 0.5 belong to call intervals between series. The note-level interval ratios lack any peak centered around 0.5, although such a peak is prominent in the null distribution generated by selecting intervals in random orders. The lack of isochrony in the observed note data is driven by the fact that, consistent with current findings, the majority of vocalizations occur as two-note doublets separated by longer silence periods from the next call (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec). For a visual example of these non-0.5 call and note IR distributions, see Supp. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe long-timescale series rhythms we found share features with behaviors of vocal systems with greater vocal repertoires and semantic complexity such those in birds, marine mammals, and primates (Weiss et al. \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Levinson \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Anichini et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The distinct patterns of interval ratio distributions between timescales mirror findings in other taxa where rhythmic regularity at one level of organization does not predict regularity at nested levels (Xing et al. \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ma et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Taken together, our results suggest an analogous hierarchical organization in anuran vocal production in which the temporal structure of vocalizations is timescale-dependent.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003ePatterns of variation across timescales\u003c/h2\u003e \u003cp\u003eDespite their isochrony, the temporal features of call series exhibited a greater degree of within- than between-male variation (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This is likely due to the fact that, while males maintain isochronous calling series across a calling bout, they can vary the total number of calls produced in quick succession, and this variation in total calls produced could be influenced by social or environmental feedback. Such rhythmic but variable vocal timing is reminiscent of chorus dynamics in \u003cem\u003eE. coqui\u003c/em\u003e, in which a \u0026ldquo;sloppy oscillator\u0026rdquo; is proposed to generate a flexible call period that, while variable in isolation, can be entrained to a range of call intervals of neighboring frogs (Zelick and Narins, \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e1985\u003c/span\u003e). Such flexibility in the temporal generation of calls may be a necessity of chorusing systems, where the social context drives call patterning within bouts and modifies short timescale temporal features (Larter et al. 2022; Larter et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2026\u003c/span\u003e). Although outside the scope of the current study, future work can investigate the degree to which inter-series rhythms in \u003cem\u003eBoana\u003c/em\u003e vary depending on behavioral patterns of chorusing neighbors.\u003c/p\u003e \u003cp\u003eGiven the existence of these call clusters we deem here as \u0026lsquo;series\u0026rsquo;, we compared the temporal properties of calls within and across series (i.e. across shorter versus longer timescales). We found that the temporal features of calls vary dynamically depending on series duration and within-series position (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Within series, mid-position calls have significantly longer inter-call onsets (ICOs) compared to the first call in series, and we conclude that these longer ICOs are driven by longer inter-call silence intervals (ICIs) because call durations of these mid-position calls are actually shorter (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The dependence of call timing on series position is reflected in distinct variation across timescales. First, within-male variation in ICO and ICI was high within bouts and significantly decreased within series (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Likewise, within-male variation in call duration was modest when looking across or within series but decreased when accounting for position within series.\u003c/p\u003e \u003cp\u003eUnlike call timing properties, note features do not significantly vary with series duration but rather depend exclusively on the sequence order (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). As noted above, call durations are shorter in the middle of calls, and we can attribute that change in call properties to both decreased durations of notes and silence within calls. Specifically, the inter-note onset and interval (INO and INI) between notes 1 and 2 get shorter in mid-position calls (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed and f) and the durations of both notes 1 and 2 also get shorter in mid-position calls (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee). The relative degree of between- male variation is higher for INI and INO in mid-position calls compared to those that come first in the series (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). In fact, the mid-position INI and INO have the highest CV ratio in our study, meaning they are very consistent within males yet differ among males.\u003c/p\u003e \u003cp\u003eThese CV ratio results are an interesting comparison to the linear model results showing that individual identity explained a large and consistent portion of the variance in call duration (conditional R\u0026sup2; = 0.76) and note durations (note 1: conditional R\u0026sup2; = 0.76; note 2: conditional R\u0026sup2; = 0.78), but substantially less variance in inter-call intervals (conditional R\u0026sup2; = 0.48). The CV ratio for call duration further confirms that between-male differences in call duration exceed within-male variation, and this depends on the within-series position of calls with the first call in a call series having especially low within-male variation and thus readily measurable differences between males. For all note features except the duration of note 1, individual identity explained a large and consistent portion of the variance (Supp. Table\u0026nbsp;2). The pronounced reduction of within-male variation when accounting for series positions highlights the value of defining distinct timescales of vocal features since failing to do so results in an overestimation of within-male variation. Additionally, we want to highlight here that we found no evidence for within-series position effects on spectral features, such as the dominant frequency, of notes 1 and 2 (Supp. Table\u0026nbsp;2) and low CVw for note spectral features, suggesting these remain static within-individuals and do not vary depending on temporal dynamics.\u003c/p\u003e \u003cp\u003eThese findings are consistent with previous works showing that spectral and duration features of vocalizations are more individually stereotyped than temporal spacing features (Gerhardt \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Gerhardt and Bee \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Call duration, like spectral properties, may be constrained by individual morphology\u0026mdash; particularly body size and the physical properties of the vocal apparatus\u0026mdash;rendering it a relatively fixed signature of the individual. Inter-call intervals, by contrast, are more responsive to immediate context, suggesting they are regulated by mechanisms that are less tightly coupled to fixed individual traits. This functional dissociation has implications for signal evolution: if receivers use call duration to identify or assess individual callers and use inter-call timing to track calling effort or competitive state, selection could act independently on these features.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePotential neural mechanisms that drive vocalization patterning across timescales\u003c/h2\u003e \u003cp\u003eCentral pattern generators (CPGs) are circuits capable of generating and modulating rhythmic motor output in the absence of rhythmic input (Marder and Calabrese, 1996; Marder and Bucher, 2007). The isochronous call series rhythm seen here is consistent with a \u0026ldquo;series CPG\u0026rdquo; in which neurons exhibit consistent periods of activation. The regular call timing within each series could be driven by mechanisms nested within the series rhythm generator itself; alternatively, call timing within series could be regulated by a distinct but coupled \u0026ldquo;call CPG.\u0026rdquo; The duration of each series could, in turn, be driven by mechanisms located in either the series CPG, the call CPG, or both. While the temporal patterning of call emission within a series\u0026mdash;including the observed shortening of call and note elements depending on sequence order\u0026mdash;may simply reflect the dynamics of the call CPG, we propose that the cycle-by-cycle series durations are actively modified by social features such as chorus amplitude or the call properties of neighboring males. Such social signals could lead to sensory-based excitatory, inhibitory and/or modulatory inputs onto the circuit elements regulating both series onset and duration patterns. Thus, the mechanisms controlling the timing of nested rhythms within series could operate under different constraints\u0026mdash;including ones that permit, or even require, flexibility in response to immediate contextual demands.\u003c/p\u003e \u003cp\u003eCPGs in the anuran brainstem (Barkan et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Kelley et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) are strong candidates for the neural substrates mediating vocal dynamics in \u003cem\u003eB. pulchella\u003c/em\u003e. In distantly related aquatic frog species in the genus \u003cem\u003eXenopus\u003c/em\u003e, two reciprocally connected hindbrain nuclei \u0026ndash; nucleus ambiguus (NA) and the parabrachial nucleus (PB) \u0026ndash; comprise the vocal CPG (Kelley et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Rhodes et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). While PB is able to generate neural correlates of call duration and period independent of NA, feedback from NA to PB is needed to generate the faster inter-pulse intervals (Zornik et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Although the timescales involved in the \u003cem\u003eXenopus\u003c/em\u003e vocal CPG are faster than \u003cem\u003eBoana\u003c/em\u003e vocal patterns, one possibility is that homologous coupled circuit components generate the shorter timescale call periods and the longer duration series timing. PB has been shown to participate in vocal production in both \u003cem\u003eXenopus\u003c/em\u003e (family Pipidae) and \u003cem\u003eRana\u003c/em\u003e (family Ranidae), suggesting that its involvement in vocal patterning is ancient and conserved across anurans (Schmidt, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Barkan et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Whole-cell physiology recordings in PB have revealed neurons that intrinsically encode species-specific call periods in \u003cem\u003eX. laevis\u003c/em\u003e and \u003cem\u003eX. petersii\u003c/em\u003e (Barkan et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), but longer timescales have not been investigated in these species. Future physiological experiments in \u003cem\u003eBoana\u003c/em\u003e (and across anurans) could allow researchers to test hypotheses about underlying circuits that generate vocal patterns across timescales, providing potentially exciting insights into the evolutionary trajectories of nested vocal rhythms in frogs.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eBehavioral implications\u003c/h2\u003e \u003cp\u003eAt the within-series level, the ability to append calls in series\u0026mdash;and to modulate the timing of those calls\u0026mdash;may allow males to signal competitive arousal or motivation. Males producing longer series may convey greater effort or urgency to both females and rival males. Faster call rates within bouts have been associated with greater mating success across a range of anuran species (Sullivan \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e1983\u003c/span\u003e; Morris \u0026amp; Yoon \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e1989\u003c/span\u003e; Lopez \u0026amp; Narins \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1991\u003c/span\u003e), and the progressive temporal adjustments we describe in males quickly appending calls within series could be this type of competitive signal.\u003c/p\u003e \u003cp\u003eAt the between-series level, the isochrony we observe in series spacing may serve a different social signal. Regular, predictable calling cadences could allow male receivers to anticipate the timing of future calls by neighbors. Being able to predict when silent periods in chorus signaling may arise can help males time their next call and facilitate signal detection in noisy chorus environments (Klump \u0026amp; Gerhardt \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Larter \u0026amp; Ryan \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Regular calling cadences could similarly help female receivers localize and assess multiple callers simultaneously, consistent with preferences by female treefrogs for high and regular call rates (Tanner \u0026amp; Bee \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Tanner \u0026amp; Bee \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Tanner et al. \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These nested timescales of periodicity could provide sufficient structure for receivers to assess short timescale features that reflect male motivation, physical condition, or individual identity nested within a more predictable signal that aids detection of and selective attention toward one caller in a complex environment.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eCharacterization of a novel call type\u003c/h2\u003e \u003cp\u003ePrevious studies that examined the calling behavior of this species restricted their analyses to characterizing changes in doublet calls across contexts. While some studies highlighted the variability of male call patterning over time, we found no published descriptions of a a single-note call with a long duration and upward-trending harmonics that we deem a \u0026lsquo;squeak\u0026rsquo; call. We report that males produce two main types of calls \u0026ndash; doublets and squeaks \u0026ndash; which are distinguished by their durations and the sequence of note types used. Across our dataset, we found a total of 48 squeaks used between 18 males. Of these 18 males, only 11 produced squeaks and the total number of squeaks per male ranged from 1 to 17. Given the limited number of total squeak calls in our dataset, we are unable to report the contextual basis for its use (e.g. whether it is used to attract females or repel nearby males). We encourage future studies to characterize squeak properties and context-dependent use to enhance our understanding of their function.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe vocalization patterning of \u003cem\u003eB. pulchella\u003c/em\u003e reflects a hierarchical structure in which rhythmic regularity emerges at longer timescales yet maintains highly variable patterning, and temporal features at shorter-timescale retain high flexibility between and within males. The dissociation between individually stereotyped features\u0026mdash;call duration, note duration\u0026mdash;and more variable timing features, such as the inter-call interval, suggests that distinct mechanisms regulate different aspects of vocal patterning. Such distinct vocal timescales could be generated by CPG activity that establishes call patterns at multiple timescales (such as ISO and ICO). Oscillatory or neuromodulatory mechanisms could regulate the number of calls within each series in response to external (e.g. social) or internal (e.g. energetic capacity) feedback. This temporal flexibility has plausible consequences for both mate assessment and acoustic competition in chorus environments. Future work directly linking neural activity to the patterns described here and examining how chorus dynamics shape within-bout calling decisions will be essential for understanding the physiological and ecological forces that structure vocal communication across species.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting Interests:\u003c/h2\u003e\n\u003cp\u003eThe authors have no relevant competing interests to declare.\u003c/p\u003e\n\u003ch2\u003eFunding Declaration:\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThis work was supported by the National Science Foundation IOS-2154203 and Universidad de la Rep\u0026uacute;blica, Programa de Desarrollo de las Ciencias B\u0026aacute;sicas (PEDECIBA). MRS was supported by a National Science Foundation Postdoctoral Research Fellowship in Biology (award no. 2109884).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOMPLIANCE WITH ETHICAL STANDARDS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResearch Involving Animals\u003c/strong\u003e: This study was reviewed and approved by an ethics committee (Comisi\u0026oacute;n Nacional de Experimentaci\u0026oacute;n Animal, Universidad de la Rep\u0026uacute;blica, Protocol Number 1061).\u003c/p\u003e\n\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\n\u003cp\u003eP.P., K.H., and E.Z. conceptualized the project to examine call variability in Boana and conducted the field recordings. M.R.S. conceptualized the addition of rhythm analysis to examine vocal timescales. M.R.S. conducted all the acoustic and statistical analyses. M.R.S. wrote the original draft. All authors commented on previous versions and approved the final manuscript. All authors contributed equally to raising philosophical conundrums and leading the team into stimulating and productive intellectual spirals.\u003c/p\u003e\n\u003ch2\u003eAcknowledgement\u003c/h2\u003e\n\u003cp\u003eWe wish to thank Esteban Russi for assisting with data collection and Alvaro Mazzilli Millan and his family, Mercedes Hern\u0026aacute;ndez, and Isidro Rodr\u0026iacute;guez for access to the field site in Uruguay. We also thank Ana Silva, Carlos Colacce, and La Comarca for providing housing. We are grateful to the members of the Hoke lab for helpful feedback and discussion on figure clarity.\u003c/p\u003e\n\u003ch2\u003eData Availability\u003c/h2\u003e\n\u003cp\u003eAll data and scripts used for analysis will be made publicly available on github upon publication.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAnichini M, De Reus K, Hersh TA, Valente D, Salazar-Casals A, Berry C, Ravignani A (2023) Measuring rhythms of vocal interactions: a proof of principle in harbour seal pups. Philosophical Trans Royal Soc B: Biol Sci, \u003cem\u003e378\u003c/em\u003e(1875)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAraya-Salas M, Smith-Vidaurre G (2017) warbleR: An R package to streamline analysis of animal acoustic signals. Methods Ecol Evol 8(2):184\u0026ndash;119\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaraquet M, Salas NE, Ludovico AM (2013) Advertisement calls and interspecific variation in \u003cem\u003eHypsiboas cordobae\u003c/em\u003e and \u003cem\u003eHypsiboas pulchellus\u003c/em\u003e (Anura, Hylidae) from central Argentina. Terr Ecol Behav 65(4):479\u0026ndash;486\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarkan CL, Kelley DB, Zornik E (2018) Premotor neuron divergence reflects vocal evolution. J Neurosci 38(23):5325\u0026ndash;5337\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarkan CL, Zornik E, Kelley DB (2017) Evolution of vocal patterns: tuning hindbrain circuits during species divergence. J Exp Biol 220(5):856\u0026ndash;867\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBasso NG, Basso G (1987) Analisis acustico del canto nupcial de \u003cem\u003eHyla pulchella pulchella\u003c/em\u003e. Mus Hist Nat Valparaiso 18:109\u0026ndash;114\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBates D, Maechler M, Bolker B, Walker S, Christensen RHB, Singmann H, Bolker MB (2015) Package \u0026lsquo;lme4\u0026rsquo;. \u003cem\u003econvergence\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhat AS, Sane VA, Seshadri KS, Krishnan A (2022) Behavioural context shapes vocal sequences in two anuran species with different repertoire sizes. Anim Behav 184:111\u0026ndash;129\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown CH, Beecher MD, Moody DB, Stebbins WC (1978) Localization of primate calls by old world monkeys. Science 201(4357):753\u0026ndash;754\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrumm H (2006) Signalling through acoustic windows: nightingales avoid interspecific competition by short-term adjustment of song timing. J Comp Physiol A 192(12):1279\u0026ndash;1285\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrush JS, Narins PM (1989) Chorus dynamics of a neotropical amphibian assemblage: comparison of computer simulation and natural behaviour. Anim Behav 37:33\u0026ndash;44\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurchardt LS, Kn\u0026ouml;rnschild M (2020) Comparison of methods for rhythm analysis of complex animals\u0026rsquo; acoustic signals. PLoS Comput Biol 16:e1007755. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pcbi.1007755\u003c/span\u003e\u003cspan address=\"10.1371/journal.pcbi.1007755\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurchardt LS, Picciulin M, Parmentier E, Bolgan M (2021) A primer on rhythm quantification for fish sounds: a Mediterranean case study. Royal Soc open Sci, \u003cem\u003e8\u003c/em\u003e(9)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCanavero A, Arim M, Naya Monteverde DE (2008) Calling activity patterns in an anuran assemblage: the role of seasonal trends and weather determinants. North-Western J Zool 4(1):29\u0026ndash;41\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChagnaud BP, Baker R, Bass AH (2011) Vocalization frequency and duration are coded in separate hindbrain nuclei. Nat Commun 2(1):346\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eForrest T, Ariaratnam J, Strogatz S (1998) Synchrony in cricket calling songs: models of coupled biological oscillators. J Acoust Soc Am 103(5Supplement):2827\u0026ndash;2827\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFox J, Weisberg S, Adler D, Bates D, Baud-Bovy G, Ellison S, Heiberger R (2012) Package \u0026lsquo;car\u0026rsquo;. Vienna: R Foundation Stat Comput 16(332):333\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFurtado R, Santos SP, Dias TM, Bastos RP, Nomura F (2016) Vocal repertoire during reproductive and aggressive contexts of three Brazilian tree frogs: Bokermannohyla sapiranga, Hypsiboas albopunctatus and H. goianus (Anura: Hylidae). South Am J Herpetology 11(2):136\u0026ndash;147\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerhardt HC (1991) Female mate choice in treefrogs: static and dynamic acoustic criteria. Anim Behav 42(4):615\u0026ndash;635\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerhardt HC (1992) December). Multiple messages in acoustic signals. Seminars in Neuroscience, vol 4. Academic, pp 391\u0026ndash;400. 6\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerhardt HC (1994) The evolution of vocalization in frogs and toads. \u003cem\u003eAnnual review of ecology and systematics\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(1), 293\u0026ndash;324\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerhardt HC, Bee MA (2007) Recognition and localization of acoustic signals. Hearing and sound communication in amphibians. Springer New York, New York, NY, pp 113\u0026ndash;146\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGibson RM, Bradbury JW (1985) Sexual selection in lekking sage grouse: phenotypic correlates of male mating success. Behav Ecol Sociobiol 18:117\u0026ndash;123\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGillooly JF, Ophir AG (2010) The energetic basis of acoustic communication. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e277\u003c/em\u003e(1686), 1325\u0026ndash;1331\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenfield MD (1988) Interspecific acoustic interactions among katydids Neoconocephalus: inhibition-induced shifts in diel periodicity. Anim Behav 36(3):684\u0026ndash;695\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHartbauer M, Kratzer S, Steiner K, R\u0026ouml;mer H (2005) Mechanisms for synchrony and alternation in song interactions of the bushcricket Mecopoda elongata (Tettigoniidae: Orthoptera). J Comp Physiol A 191(2):175\u0026ndash;188\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHumfeld SC (2013) Condition-dependent signaling and adoption of mating tactics in an amphibian with energetic displays. Behav Ecol 24(4):859\u0026ndash;870\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJadoul Y, D\u0026rsquo;Orazio F, Eleuteri V, van der Werff J, Tufarelli T, Gamba M, Ravignani A (2025) Rhythmic analysis in animal communication, speech, and music: the normalized pairwise variability index is a summary statistic of rhythm ratios. Vibration 8(2):12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKao MH, Doupe AJ, Brainard MS (2005) Contributions of an avian basal ganglia\u0026ndash;forebrain circuit to real-time modulation of song. Nature 433(7026):638\u0026ndash;643\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eK. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology (2026) Raven Pro: Interactive Sound Analysis Software (Version 1.6.5) [Computer software]. Ithaca, NY: The Cornell Lab of Ornithology. Available from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ravensoundsoftware.com\u003c/span\u003e\u003cspan address=\"https://www.ravensoundsoftware.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelley DB, Ballagh IH, Barkan CL, Bendesky A, Elliott TM, Evans BJ, Zornik E (2020) Generation, coordination, and evolution of neural circuits for vocal communication. J Neurosci 40(1):22\u0026ndash;36\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlump GM, Gerhardt HC (1992) Mechanisms and function of call-timing in male-male interactions in frogs. Playback and studies of animal communication. Springer US, Boston, MA, pp 153\u0026ndash;174\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eK\u0026ouml;hler J, Jansen M, Rodriguez A, Kok PJ, Toledo LF, Emmrich M, Vences M (2017) The use of bioacoustics in anuran taxonomy: theory, terminology, methods and recommendations for best practice. Zootaxa 4251(1):1\u0026ndash;124\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarter LC, Bernal XE, Page RA, Ryan MJ (2023) Local competitive environment and male condition influence within-bout calling patterns in t\u0026uacute;ngara frogs. Bioacoustics 32(2):121\u0026ndash;142\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarter LC, Cushing CW, Ryan MJ (2026) Cadences of the collective: conspecific stimulation patterns interact with endogenous rhythms to cue socially mediated response shifts. J Exp Biol 229(1):jeb250982\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarter LC, Ryan MJ (2024) T\u0026uacute;ngara frog call-timing decisions arise as internal rhythms interact with fluctuating chorus noise. Behav Ecol 35(4):arae034\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLenth R, Singmann H, Love J, Buerkner P, Herve M (2019) Package \u0026lsquo;emmeans\u0026rsquo;. R package version, \u003cem\u003e1\u003c/em\u003e(3.2)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLevinson SC (2016) Turn-taking in human communication\u0026ndash;origins and implications for language processing. Trends Cogn Sci 20(1):6\u0026ndash;14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLopez PT, Narins PM (1991) Mate choice in the neotropical frog, Eleutherodactylus coqui. Anim Behav 41(5):757\u0026ndash;772\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eL\u0026uuml;decke et al (2021) performance: An R Package for Assessment, Comparison and Testing of Statistical Models. J Open Source Softw 6(60):3139. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21105/joss.03139\u003c/span\u003e\u003cspan address=\"10.21105/joss.03139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa H, Wang Z, Han P, Fan P, Chapman CA, Garber PA, Fan P (2024) Small apes adjust rhythms to facilitate song coordination. Curr Biol 34(5):935\u0026ndash;945\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarler P, Peters S (1977) Selective vocal learning in a sparrow. Science 198(4316):519\u0026ndash;521\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcComb KE (1991) Female choice for high roaring rates in red deer, Cervus elaphus. Anim Behav 41(1):79\u0026ndash;88\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore JM, Woolley SM (2019) Emergent tuning for learned vocalizations in auditory cortex. Nat Neurosci 22(9):1469\u0026ndash;1476\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMorris MR, Yoon SL (1989) A mechanism for female choice of large males in the treefrog Hyla chrysoscelis. Behav Ecol Sociobiol 25:65\u0026ndash;71\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNarins PM (1982) Behavioral refractory period in neotropical treefrogs. J Comp Physiol 148(3):337\u0026ndash;344\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNarins PM, Capranica RR (1977) An automated technique for analysis of temporal features in animal vocalizations. Anim Behav 25:615\u0026ndash;621\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003e\u0026Ouml;lveczky BP, Andalman AS, Fee MS (2005) Vocal experimentation in the juvenile songbird requires a basal ganglia circuit. PLoS Biol, 3(5), e153\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePollack G (2000) Who, what, where? Recognition and localization of acoustic signals by insects. Curr Opin Neurobiol 10(6):763\u0026ndash;767\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePouso P, Cabana \u0026Aacute;, Nieto Methol C (2026) Environmental drivers of calling activity in a southern subtropical anuran assemblage: insights from passive acoustic monitoring. Bioacoustics 35(1):75\u0026ndash;89\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRand AS, Ryan MJ (1981) The adaptive significance of a complex vocal repertoire in a neotropical frog. Z f\u0026uuml;r Tierpsychologie 57(3\u0026ndash;4):209\u0026ndash;214\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRavignani A, Norton P (2017) Measuring rhythmic complexity: a primer to quantify and compare temporal structure in speech, movement, and animal vocalizations. J Lang Evol 2(1):4\u0026ndash;19\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRhodes HJ, Yu HJ, Yamaguchi A (2007) Xenopus vocalizations are controlled by a sexually differentiated hindbrain central pattern generator. J Neurosci 27(6):1485\u0026ndash;1497\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRivera M, Edwards JA, Hauber ME, Woolley SM (2023) Machine learning and statistical classification of birdsong link vocal acoustic features with phylogeny. Sci Rep 13(1):7076\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRyan MJ (1985) The t\u0026uacute;ngara frog: a study in sexual selection and communication. University of Chicago Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSakata JT, Hampton CM, Brainard MS (2008) Social modulation of sequence and syllable variability in adult birdsong. J Neurophysiol 99(4):1700\u0026ndash;1711\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchmidt RS (1992) Neural correlates of frog calling: production by two semi-independent generators. Behav Brain Res 50(1\u0026ndash;2):17\u0026ndash;30\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwartz JJ (1993) Male calling behavior, female discrimination and acoustic interference in the Neotropical treefrog Hyla microcephala under realistic acoustic conditions. Behav Ecol Sociobiol 32(6):401\u0026ndash;414\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSismondo E (1990) Synchronous, alternating, and phase-locked stridulation by a tropical katydid. Science 249(4964):55\u0026ndash;58\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSober SJ, Wohlgemuth MJ, Brainard MS (2008) Central contributions to acoustic variation in birdsong. J Neurosci 28(41):10370\u0026ndash;10379\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan BK (1983) Sexual selection in Woodhouse's toad (Bufo woodhousei). II. Female choice. Anim Behav 31(4):1011\u0026ndash;1017\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan BK, Hinshaw SH (1990) Variation in advertisement calls and male calling behavior in the spring peeper (Pseudacris crucifer). Copeia 1990(4):1146\u0026ndash;1150\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakahashi DY, Narayanan DZ, Ghazanfar AA (2013) Coupled oscillator dynamics of vocal turn-taking in monkeys. Curr Biol 23(21):2162\u0026ndash;2168\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTanner JC, Bee MA (2019) Within-individual variation in sexual displays: signal or noise? Behav Ecol 30(1):80\u0026ndash;91\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTanner JC, Bee MA (2020) Inconsistent sexual signaling degrades optimal mating decisions in animals. Sci Adv 6(20):eaax3957\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTanner JC, Shein-Idelson M, Hoke KL, Bee MA (2025) Complex choice environments shelter unattractive signallers from sexual selection. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e292\u003c/em\u003e(2054)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTobias ML, Viswanathan SS, Kelley DB (1998) Rapping, a female receptive call, initiates male\u0026ndash;female duets in the South African clawed frog. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e95\u003c/em\u003e(4), 1870\u0026ndash;1875\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTomasello M, Zuberb\u0026uuml;hler K (2002) Primate vocal and gestural communication. In: Bekoff M, Allen C, Burghardt GM (eds) The cognitive animal: Empirical and theoretical perspectives on animal cognition. MIT Press, pp 293\u0026ndash;299\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTumer EC, Brainard MS (2007) Performance variability enables adaptive plasticity of \u0026lsquo;crystallized\u0026rsquo;adult birdsong. Nature 450(7173):1240\u0026ndash;1244\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalker TJ (1969) Acoustic synchrony: two mechanisms in the snowy tree cricket. Science 166(3907):891\u0026ndash;894\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWaser PM, Waser MS (1977) Experimental studies of primate vocalization: Specializations for long-distance propagation. Z f\u0026uuml;r Tierpsychologie 43(3):239\u0026ndash;263\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeiss M, Hultsch H, Adam I, Scharff C, Kipper S (2014) The use of network analysis to study complex animal communication systems: a study on nightingale song. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e281\u003c/em\u003e(1785)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWelbergen JA, Davies NB (2008) Reed warblers discriminate cuckoos from sparrowhawks with graded alarm signals that attract mates and neighbours. Anim Behav 76(3):811\u0026ndash;822\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoolley SM, Moore JM (2011) Coevolution in communication senders and receivers: vocal behavior and auditory processing in multiple songbird species. Ann N Y Acad Sci 1225(1):155\u0026ndash;165\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXing J, Sainburg T, Taylor H, Gentner TQ (2022) Syntactic modulation of rhythm in Australian pied butcherbird song. Royal Soc open Sci, \u003cem\u003e9\u003c/em\u003e(9)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZelick RD, Narins PM (1982) Analysis of acoustically evoked call suppression behaviour in a neotropical treefrog. Anim Behav 30(3):728\u0026ndash;733\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZelick RD, Narins PM (1983) Intensity discrimination and the precision of call timing in two species of neotropical treefrogs. J Comp Physiol 153(3):403\u0026ndash;412\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZelick R, Narins PM (1985) Characterization of the advertisement call oscillator in the frog Eleutherodactylus coqui. J Comp Physiol A 156(2):223\u0026ndash;229\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiegler L, Arim M, Bozinovic F (2016) Intraspecific scaling in frog calls: the interplay of temperature, body size and metabolic condition. Oecologia 181:673\u0026ndash;681\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiegler L, Arim M, Narins PM (2011) Linking amphibian call structure to the environment: the interplay between phenotypic flexibility and individual attributes. Behav Ecol 22(3):520\u0026ndash;526\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZornik E, Katzen AW, Rhodes HJ, Yamaguchi A (2010) NMDAR-dependent control of call duration in Xenopus laevis. J Neurophysiol 103(6):3501\u0026ndash;3515\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":true,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-comparative-physiology-a","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jcpa","sideBox":"Learn more about [Journal of Comparative Physiology A](http://link.springer.com/journal/359)","snPcode":"359","submissionUrl":"https://submission.nature.com/new-submission/359/3","title":"Journal of Comparative Physiology A","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"anuran, acoustics, vocalizations, rhythm analysis, variation","lastPublishedDoi":"10.21203/rs.3.rs-9458485/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9458485/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCommunication systems shape behavioral interactions that are crucial for recognition, territoriality, and reproduction across animals. In many anuran species, acoustic communication drives reproductive and defensive behaviors as males produce advertisement calls to establish territories and attract females. Understanding the temporal dynamics in vocalizations and their variation between and within males can provide crucial insight into the mechanisms that generate call patterning and communication. Here, we examine such dynamics in the call patterning of male \u003cem\u003eBoana pulchella\u003c/em\u003e, a South American treefrog species that produces temporally-variant call series over time. Using rhythm analysis, we quantify the rhythmic structure of these vocalizations and characterize the variability in their temporal patterning between males. We find that rhythmic isochrony is timescale-dependent - call series are produced isochronously while the patterning of calls and notes within these series is non-isochronous yet periodic. The isochronous rhythm is maintained through an inverse relationship between series duration and the silence interval that follows it, such that longer series are followed by shorter silence intervals. Temporal features of calls and notes as well as their variation vary depending on within-series patterning dynamics and across timescales. Taken together, these results suggest that \u003cem\u003eB. pulchella\u003c/em\u003e vocal timing is generated by hierarchically organized oscillators that produce isochronous vocalizations over a calling bout through flexible modulation at finer timescales.\u003c/p\u003e","manuscriptTitle":"Timescales of call variability in a South American treefrog","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-27 10:03:11","doi":"10.21203/rs.3.rs-9458485/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-13T15:40:25+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-13T15:04:30+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-28T22:28:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"177549762611373206388720512742075245985","date":"2026-04-23T09:26:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"81827840228026011627911420282324254053","date":"2026-04-22T21:39:50+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-22T12:15:31+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-21T07:47:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-21T07:46:52+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of Comparative Physiology A","date":"2026-04-19T00:27:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-comparative-physiology-a","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jcpa","sideBox":"Learn more about [Journal of Comparative Physiology A](http://link.springer.com/journal/359)","snPcode":"359","submissionUrl":"https://submission.nature.com/new-submission/359/3","title":"Journal of Comparative Physiology A","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"0dfcafe9-212e-4299-b6dd-db60a135bb32","owner":[],"postedDate":"April 27th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-13T15:40:25+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-13T15:04:30+00:00","index":18,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-05-13T15:54:27+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-27 10:03:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9458485","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9458485","identity":"rs-9458485","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}