What do we know about advertisement calls of Mexican anurans? A bibliographic review

ABSTRACT In a fast-changing environment, it is important to keep track of biodiversity. In Mexico, new herpetofauna is described every year. Vocalisations can portray ecological and evolutionary characteristics for better understanding of species. Here, we lay out the current state of knowledge on advertisement calls of Mexican anurans, as well as the publication trends and the use of metadata through an exhaustive literature review and bibliometric analysis. We obtained 174 documents with advertisement call descriptions of 58.4% of Mexican anuran species published between 1940–2021. We identified that most of the species with call descriptions are distributed in the Neotropics and belong to endemic and endangered categories. Anuran call descriptions in publications are frequently accompanied by associated data such as relative humidity, call frequency, or recording format. The mean number of recorded characteristics per description was 12.568 (of 38). Although almost half of the Mexican species have associated advertisement call descriptions, we identified several knowledge gaps, particularly in the degree of information that each description contains. Furthermore, most of the identified studies were performed outside of Mexico by foreign researchers, despite the high level of anuran endemism in the country. We highlight the need for more detailed descriptions published by Mexican researchers.


Introduction
As a tropical and megadiverse country, Mexico harbours a biodiversity that is not only understudied (Guerra et al. 2018), but also still being inventoried.As of 2021, over 250 species of anurans have been described from Mexico, with a high endemism level of more than 60% (Frost 2021).However, climate change, overexploitation, and deforestation add constant pressures to Mexican landscapes.The rapid change of ecosystems in the country makes our understanding, monitoring, and conservation of amphibians a complex task (Frías-Alvarez et al. 2010;Ochoa-Ochoa et al. 2012;Bonilla-Moheno and Aide 2020).In addition to this, the high cost of fieldwork coupled with difficult accessibility to some regions adds further complications.To confront this problem, we must implement new study methods that help deal with the rapid loss of biodiversity while remaining affordable.
Over the past decades, the field of bioacoustics has established itself as an accessible and powerful tool to monitor and study sound-producing animals and their dynamics (Laiolo 2010;Campos-Cerqueira et al. 2016;Nelson et al. 2017).In anurans, acoustic communication is present in most species, having appeared early in the group's evolutionary history (Chen and Wiens 2020).Furthermore, vocalisations are an important lifehistory trait that can provide relevant information for studies in ecology and are relatively easy to collect (Padial et al. 2008;Ohmer et al. 2009;Both and Grant 2012;Velásquez et al. 2013;Medeiros et al. 2017;Williams et al. 2021).
Acoustic signals play an important role in ecological processes such as defending territory, inter-and intraspecific recognition, and mate selection, to name a few (Jones 1997;Kelley 2004;Dreher and Pröhl 2014;Oliva et al. 2018).Because of their role in recognising conspecifics, calls can reflect evolutionary processes such as sexual selection and reproductive isolation, effectively functioning as prezygotic barriers that can lead to speciation (Goicoechea et al. 2010;Clemens et al. 2017).Furthermore, acoustic signals are subject to abiotic pressures promoting differences between individuals and populations, which can lead to the development of geographic variation (Ohmer et al. 2009;Kight and Swaddle 2015;Weaver et al. 2020).
Advertisement calls are the most conspicuous call type among anurans, belonging to the reproductive category according to Toledo et al. (2015).These calls are produced exclusively during the breeding season to attract potential mates or detract competitors (Bradbury and Vehrencamp 2011;Toledo et al. 2015).Advertisement calls have mainly been documented for males; however, females have also been reported to emit such calls in some species (Boyd and Emerson 1999;Toledo et al. 2015).While most call descriptions are limited to the audible spectrum, there is evidence of infra-and ultrasonic advertisement calls in a few species (Bradbury and Vehrencamp 2011;Toledo et al. 2015).
In a survey of Brazilian herpetologists regarding the study of anuran vocalisations, the vast majority of researchers and students reported having documented anuran calls in the field; nevertheless, less than 20% of those surveyed had deposited their audio material into sound repositories (Dena et al. 2019), despite anuran calls being a highly informative trait.Most recordings were not catalogued, with researchers considering the cataloguing process too complicated or not having enough time to do it (Dena et al. 2019).It is very likely that a similar situation is occurring in Mexico.
As Guerra et al. (2018) stated, the first step to apply bioacoustics to herpetology is to nourish the basic knowledge of anuran vocalisations by describing the calls quantitatively and qualitatively.Over the last decades, most scientific collections have been rapidly incorporating an audio-visual format into their repository infrastructures, in addition to digitising anuran audio recordings to make them available to the general public (Schindel and Cook 2018;Mendoza-Henao et al. 2021).Audio recordings can enrich the knowledge of specimens with information about biotic interactions and behavioural traits that otherwise would be dismissed (Toledo et al. 2015).
In this study, we conduct a systematic literature review and a bibliometric analysis to understand the current state of knowledge on the advertisement calls of Mexican anurans, its possible implications, and future directions.We aim to identify: i) the families with the most advertisement call descriptions, and if the number of descriptions is related to either the regional species richness or the threat category of the species; ii) the number of species that have a call deposited in an online audio-visual repository; iii) the change over time in the number and the complexity of the descriptions (i.e.what type of metadata is reported); iv) the research areas that contribute the most descriptions; and v) trends in the number of authors per publication and their affiliations.

Literature search and retrieval
We conducted a literature review to understand the state of knowledge of advertisement calls of anurans distributed within Mexico.We also performed a bibliometric analysis to determine the publication trends for this topic in Mexico.For this, we followed the methods presented by Guerra et al. (2018) with some modifications.We considered all species distributed in Mexico, exclusively or not, according to the list and taxonomy of Mexican amphibians proposed by Frost (2021; see Appendix I, Table S1).This included 262 described species as of June 2021, of which only Trachycephalus 'vermiculatus' was considered as incertae sedis.
For the literature review, we did a Boolean search for descriptions of advertisement calls of Mexican anurans through some of the main scientific literature databases: Google Scholar (www.scholar.google.com),Scopus (https://www.scopus.com/),Web of Science (https://www.webofknowledge.com),and Scielo (https://scielo.org/es/).The literature search consisted of two phases: 1) a general search using the genera plus a word combination (e.g., 'Craugastor' AND Mexic* AND call*); and 2) after a selection of key search words, a secondary search combining the species name and valid synonyms with the selected words (vocalisation, call, advertisement call) in Spanish, English, and Portuguese (e.g., 'Smilisca baudini' AND call* OR advertisement call OR chamado OR canto).We chose these languages because several species distributed in Mexico also have part of their range in Latin America, including Brazil, and North America.
The distribution and threatened status of each species were obtained from the IUCN Red List of Threatened Species (https://www.iucnredlist.org/)and the Mexican Official Norm, NOM-059 (SEMARNAT 2010).All documents obtained from the second phase were subjected to a quick inspection for at least one of the key search words.If any were found, the document was selected and added to the database.We selected: scientific papers from indexed journal (research, reviews, short communications, and scientific notes), books, catalogue notes, reports, and field guides published in the scientific literature libraries mention above.Those documents that met the requirements were identified with a unique ID number with the initials of the first author (e.g.SOF001).We only considered publications that had an original description of an advertisement call or included a novel feature of it.The documents that did not meet the requirements were deposited in separate folders for future consulting.We included molecular and physiological studies, as well as behavioural experiments in which individuals were manipulated to change their advertisement calls or studies on phonotaxis using synthetic calls (labelled 'Experimental').The literature search was conducted from February 2021 to January 2022.
We complemented the bibliographic search with a rapid survey of the availability of audio recordings for each species in some of the main digital collections: Macaulay Library of the Cornell Lab of Ornithology, USA (https://www.macaulaylibrary.org);Fonoteca Zoológica of the Museo Nacional de Ciencias Naturales of Madrid, Spain (http://www.fonozoo.com);Fonoteca de Anfibios of the Zoology Museum of the Faculty of Science of the National Autonomous University of Mexico (http://cantosa nuros.fciencias.unam.mx/),AmphibiaWeb (http://www.amphibiaweb.org),and iNaturalist (https://www.inaturalist.org/).We also searched the Fonoteca Neotropical Jacques Vielliard (FNJV-http://www2.ib.unicamp.br/fnjv/)from Brazil, as it houses the largest repository in Latin America; however, only seven records were found for Mexican species on this database: Dendropsophus ebraccatus, D. microcephalus, Engystomops pustulosus, Leptodactylus melanonotus, Rhinella horribilis, Trachycephalus 'vermiculatus', and Xenopus laevis.We decided not to include the FNJV in our results because all these species have at least one record in the other analysed digital collections.

Data analysis
The data analysis focused on two distinct types of metadata: bioacoustics metadata and publication metadata.We identified the bioacoustics metadata associated with each description to understand the variables and characteristics that are recorded and collected when recording a vocalisation.For this type of metadata, we retrieved the following information: collection information (e.g.humidity, temperature, coordinates, date of collection), the type of metadata collected (e.g.spectral and temporal characteristics), as well as if any recordings were made available in an audio-visual repository.
To understand the publication trends in Mexican anuran bioacoustics, we recorded the following document metadata from each publication,: 1) scientific name, 2) publication year, 3) name of the first author, 4) name and email of the correspondence author, 5) affiliation of the first author, 6) number of authors, 7) name of co-authors, 8) journal name, 9) the type of research (Descriptive/Experimental and in situ/ex situ), 9) gender of the first author, and 10) area of research.We also searched for the bioacoustics metadata following the protocol of Guerra et al. (2018) with some modifications (see Appendix I Table S2 for more details).
We categorised each document according to the type of research method used: Descriptive/Experimental and in situ/ex-situ.We defined Descriptive as those works that focused on describing a characteristic of the advertisement call without interfering with the natural vocal activity, while we considered Experimental studies those that included a manipulation of the vocal activity, whether through acoustic, physical, or chemical stimulation methods (e.g., playbacks, presentation of females, or hormone injections).On the other hand, we considered those studies done in natural or outdoor conditions as in situ and those that took place in laboratory/controlled settings as ex-situ.

Results
From the literature search, we retrieved a total of 428 documents from the analysed repositories.After filtering, we identified 173 of these to contain an original description of the advertisement call or a novel feature of it.A complete list of the references obtained from this search can be found in the Supplementary material (see Appendix II, Additional Results, Text S3 for more details).A total of 153 species of anurans occurring in Mexico have descriptions of their advertisement call published between 1940-2021, representing 58.4% of the species distributed in the country.As we considered the description of a new parameter or characteristic of an advertisement call as an original description, we found several species with multiple descriptions, with Engystomops pustulosus having the most at 27 descriptions (see S1 for more information).A total of 385 advertisement call descriptions have been published to date.The first description was for Anaxyrus cognatus in 1940 published in The American Naturalist (Bragg 1940).The average number of publications per year is 2.9, and the years with the most publications were 1984 (8 publications), and 1990, 1991, 2012, and 2014 (6 publications each; Figure 1).
In some cases, publications described advertisement calls of more than one species, as in Duellman (1970) who described 84 advertisement calls or advertisement call characteristics.The families that have the highest proportions of species with advertisement call descriptions are Rhinophrynidae (100%, albeit it is monospecific and widely distributed), Leptodactylidae (100%, but in Mexico only two species are distributed), Pipidae, and Centrolenidae, followed by Microhylidae with 83.3%.The least represented family is Eleutherodactylidae (40.5%; Figure 2).

Audio-visual repositories, biogeographic region distribution, and threat category
We identified that Mexican anurans have a large percentage of their diversity represented in online audio-visual repositories (46%, Figure 3).Of the analysed repositories, the Macaulay Library houses the largest number of Mexican species, with 80 species represented by at least one audio file on their database.The repository with the least species from Mexico represented in their database was the Fonoteca de Sonidos with only 28 species.The families with the largest number of species with deposited advertisement calls are Hylidae (44 species, 43.14%), Bufonidae (20 species, 57.14%), and Eleutherodactylidae (15 species, 40.54%).In Mexico, 157 anuran species are recognised as endemic (either endemic or micro endemic), making up 59.69% of the country's anuran biodiversity.Most of the endemic species in Mexico are distributed in the Neotropics (95 species), where Hylidae is the most represented family with 41 species.A high proportion of Neotropical species are classified under one of the threat categories or do not have enough information to be evaluated (Figure 4).On the other hand, most of the endemic Nearctic species are considered as Least Concern.This pattern is also seen in the other biogeographic regions.Of all the endemic species, only 42 have their advertisement call described (26.75%).
The family in Mexico with the highest richness in the Neotropics is Hylidae (41 species), which also contains most of the endangered species (43 across all biogeographic regions).Several Hylidae genera have a high number of endangered species: Charadrahyla (8 endangered species/10 total species), Ecnomiohyla (2/2), Exerodonta (4/5), Megastomatohyla (4/4), Plectrohyla (2/2), Quilticohyla (3/3), and Sarcohyla (22/ 26).Most species in these genera also have a very limited distribution, being considered as microendemic in this review (Supplementary material, Appendix II), with some even considered as Possibly Extinct by the IUCN, such as Sarcohyla charadricola, S. pachyderma, S. psarosema, S. sabrina, and S. siopela.Not surprisingly, none of these five species have their advertisement calls deposited in any of the online repositories analysed in this review.However, the advertisement calls of S. pachyderma, S. psarosema, S. sabrina, and S. siopela were verbally described by Duellman in 1970, although these descriptions focused on abiotic characteristics and provide only a shallow explanation of the call characteristics.

Bioacoustics metadata
We found the average number of bioacoustics metadata to be 12.568 characteristics per description (SD = 6.023), with more recent descriptions tending to include more characteristics (Figure 5).The most reported characteristics were the vegetation type or habitat, a verbal description of the advertisement call, and call frequency, while the least reported were relative humidity, number of bits, and recording format (Table 1).The most common categories to be reported alongside the descriptions were qualitative information (data used to characterise the locality and a verbal description of the call), followed by spectral parameters (data that visually and numerically characterise the frequency-related characters of the advertisement calls).The least reported category was quantitative information (data used to quantitatively characterise the locality and calling behaviour of the species).
The most basic descriptions were those by Wayne McAlister in 1961 that only included the call rate in the descriptions of the species: Anaxyrus punctatus, A. woodhousii, A. debilis, Incilius alvarius, and Rhinella horribilis.The description with the most metadata reported belongs to Leptodactylus fragilis, with 31 characteristics published by Rodríguez-Cabrera et al. (2018).

Trends in advertisement call publications
From the metadata associated with the publications, we found that most publications have been produced by foreign (i.e., non-Mexican) institutions (91.62%), and most of the research locations took place in other countries (74.3%).The number of authors per publication has changed throughout the years, with a general tendency of more authors in recent years (Figure 6).We also note that an overwhelming percentage of publications on Mexican anurans advertisement calls have male first authors (83.8%).The descriptions have been published in 93 journals, with the most popular ones being: Copeia (28 publications), Journal of Herpetology, Herpetologica (11 publications each), Behavioural Ecology, Sociobiology, and Catalogue of American Amphibians and Reptiles (10 publications each).Most descriptions of advertisement calls have been in studies classified predominantly as descriptive and in situ, followed by studies in which the authors used a stimulation method within the field sites (Table 2).The areas of research in which most advertisement call descriptions were made have been in bioacoustics, followed by behaviour (Figure 7).

General and historical overview
Mexican herpetological biodiversity is still understudied, with many new species being described in recent years (e.g. de la Torre-Loranca et al. 2020;Grünwald et al. 2021).Anurans are no exception (e.g.Canseco-Márquez et al. 2017;Grünwald et al. 2018).New   descriptions of advertisement calls or their variations are also being published every year (e.g.Grünwald et al. 2018;Hernández-Herrera and Pérez-Mendoza 2021).In Mexico, the description of advertisement calls started in the 1940s and continues to this day.Recent research has highlighted the ecological and evolutionary value of studying and describing the basic characteristics of anuran vocalisations (Toledo et al. 2015;Köhler et al. 2017;Dena et al. 2019;Augusto-Alves et al. 2021).Mexico, however, is still on a learning curve regarding anuran bioacoustics.Despite an early start in the description of advertisement calls in Mexico (1940), the average number of publications per year is low (under three).The last decade, however, has seen an increase in the publication of documents and number of descriptions related to Mexican anuran advertisement calls, a trend similarly observed in Brazil, Sri Lanka, and Colombia (Guerra et al. 2018;De Silva and Wijayathilaka 2019;Rivera-Correa et al. 2021).Interest in anuran vocalisations and their associated biotic and abiotic characteristics has become a more visible topic in recent years due to the explosion of the use of ecoacoustic methods such as Passive Acoustic Monitoring (PAM) in both academia and conservation (Ferreira et al. 2018;Ulloa et al. 2019;Brodie et al. 2021;Vega-Hidalgo et al. 2021;Znidersic and Watson 2022).Sugai et al. (2019) found anurans to be the second most PAM surveyed terrestrial vertebrate group, but most of the data analysis is done manually.Doing a manual data analysis requires experience in identifying calls and lots of hours to analyse the information.Having a robust dataset can facilitate and reduce the time spent in manually processing data, making PAM a more practical tool to use.PAM can help us understand not only individual species dynamics, but also broader ecological processes at the community and ecosystem levels.However, only a couple of studies of this type have been performed in Mexico (Sugai et al. 2019).
Access to audio recording equipment and analysing software has facilitated the approach of researchers and amateurs in the collection of anuran vocalisations (Toledo et al. 2015;Weaver et al. 2020;Mendoza-Henao et al. 2021).Nevertheless, as stated before, the use of this type of method requires a baseline knowledge of the acoustic natural history traits of the focal species to be implemented properly.The description of phylogenetic signals such as advertisement calls can strengthen the number, type, and quality of research and its applications to conservation (Schneider and Sinsch 2007;Toledo et al. 2015;Guerra et al. 2018;Mendoza-Henao et al. 2021), but the description of these and other vocalisations require standardisation to homogenise and facilitate their comparisons.
Even though 45% of Mexican species have published descriptions of their advertisement calls, we identified some biases in the descriptions related to their taxonomy, category of threat, and biogeographic distribution.Regarding taxonomy, the mostrepresented families are Centrolenidae, Leptodactylidae, Pipidae, and Rhinophrynidae with 100% of species in these families having descriptions.However, this was expected as these families are monospecific for Mexico.On the other hand, the families with the fewest proportion of described species are Craugastoridae and Eleutherodactylidae, both of which contain taxonomic uncertainty and are comprised of species with elusive and disperse habits (Ospina et al. 2013;Sá et al. 2014) that make studying their natural history difficult.

Advertisement calls in online audio-visual repositories
In our review, we found that 120 species (45.62%) have at least one advertisement call deposited in an online audio-visual repository.The online repository with the most Mexican species on their database was Macaulay Library of Cornell University, USA, with 80 species.This repository is one of the biggest and oldest audio-visual archives in the world (Budney et al. 2014;Mendoza-Henao et al. 2021), and it contains the world's largest acoustic dataset for birds.However, it does not have a specific area dedicated to anuran vocalisations.Families like Hylidae and Bufonidae have an especially high representation of their calls deposited in the Macaulay Library.
The second highest representation of Mexican species was found on iNaturalist, with 73 species in its database.Citizen science initiatives have grown significantly in recent times (Liebenger 2015;Seltzer 2019;Aristeidou et al. 2021).The families Eleutherodactylidae and Ranidae have the most advertisement calls deposited in this citizen science platform, an interesting observation given that they represent the families with the least amount of advertisement calls formally described (Figure 2).
Citizen science promotes the interest of the general public in taxonomic groups that might be overlooked by researchers because of their difficulty to study or their restricted distribution (Smith et al. 2015;Liebenger 2015).Amateur observers might stumble upon such taxa on a camping trip in the desert or a family holiday in the mountains, which can lead to novel observations (and recordings) that can be later verified by specialists.Citizen science platforms like iNaturalist can complement the scientific knowledge of natural history traits (Smith et al. 2015;Seltzer 2019;Aristeidou et al. 2021), which can then be used for everything from creating hypotheses to training artificial intelligence call detectors for monitoring and conservation (Dickinson et al. 2012;Smith et al. 2015;Forti et al. 2022).
Citizen science is an area of research that involves non-scientifically trained people in the collection and interpretation of scientific data.Unlike traditional curated collections, citizen science collections are a cost-effective way of collecting larger datasets, to create a wider research community (Callaghan et al. 2021), and to generate awareness among the general public about the environment.Despite being a powerful tool for anuran bioacoustics research (Penone et al. 2013;Weaver et al. 2020;Forti et al. 2022), citizen science has certain limitations that must be considered by scientists when managing and interpretating information (Kosmala et al. 2016;Aceves-Bueno et al. 2017).The careful planning and management of citizen science projects is important to guarantee the quality and reliability of the collected data and its interpretation, since these repositories inherently receive a high number of outlying records that must be filtered.
The Fonoteca de Sonidos de Anfibios (http://cantosanuros.fciencias.unam.mx/) in the Faculty of Sciences of the National Autonomous University of Mexico (FC-UNAM, for its acronym in Spanish) is a recently created Mexican initiative that focuses on collecting national anuran vocalisations (Zamudio-Torres et al. 2020).The effort to consolidate the acoustic knowledge of Mexican anurans and the standardisation of recordings for public access in the country is growing.Despite having the smallest number of species in its database (29), initiatives such as the Fonoteca aim to promote and reinforce amphibian research in Mexico.Making audio-visual archives accessible for all is necessary to involve local actors in science production and decision-making.
Our results agree with other studies in that pairing or complementing citizen science data collection with professional data collection can provide scientists with more robust datasets (Spear et al. 2017;Pernat et al. 2021).In the case of advertisement call descriptions in Mexico, we encourage the non-scientific community to get involved with citizen science projects such as iNaturalist as well as open-access local museum collections such as the Fonoteca de Sonidos de Anfibios (UNAM) to improve the knowledge of anuran bioacoustics in our country.

Trends in advertisement call records: biogeography and threat category
Of the 262 anurans species with distributions in Mexico, 60% are endemic.According to Ochoa-Ochoa et al. ( 2014), the regions with the highest amphibian endemism are the Neotropics and the Transition Zone, especially along the Pacific Coast and the mountain ranges.Additionally, more than half of the endemic species included in this review fall under one of the threatened categories according to the IUCN Red List, and only a quarter of the endemic species have their advertisement calls described, making our knowledge of their vocal repertoire scarce.
Differences between threat categories were found among biogeographic regions.For example, Nearctic endemic species are mostly listed as Least Concern, due in part by species in the northern part of the country generally have wider distribution ranges, a highly relevant factor in establishing threat categories (https://www.iucnredlist.org/).On the other hand, most of the Neotropical endemic species are considered endangered mainly due to their restricted distributions, and in several cases are only known from their type localities (see Ochoa-Ochoa et al. 2014).
Mexico, like many other countries from the Global South, is facing rapid changes to its ecosystems (Bonilla-Moheno and Aide 2020).This represents a threat to Mexican biodiversity especially for anuran amphibians which, due to the characteristics of their life histories, are highly vulnerable to environmental disturbances (Beebee and Griffiths 2005).We need more public and private initiatives that prioritise conservation efforts for endemic and microendemic species since they correspond to more than half of all anuran species in the country.The creation of well-planned protected natural areas and the administration of those already established is urgently needed to reduce the effects of deforestation, change of land use, and pollution on Mexican amphibians (Ochoa-Ochoa et al. 2009;Juárez-Ramírez et al. 2016).Our results show that endemic species with Neotropical distributions are the least known acoustically, and we suggest acoustic data collection efforts target such areas including the Pacific coast and mountain ranges.

Trends in advertisement call records: bioacoustics metadata
Even though the amount of advertisement call descriptions for Mexican anurans is relatively high, most of the descriptions have scarce, limited, or moderated information (Muñoz Santos 2020).In contrast with results previously reported in a bachelor's thesis by Muñoz Santos (2020) that identified 159 species with advertisement call descriptions, our review found 118 species with at least one advertisement call description.Differences in selection criteria could explain the discrepancy in results.Muñoz-Santos included theses, meeting abstracts, and audio files in his selection criteria, while we focused on scientific papers, books, catalogues, and reports.
To understand the degree of knowledge portrayed by each description, we identified the main characteristics reported for each call.Almost every description had relevant qualitative information, followed by the inclusion of spectral parameters.The least reported characteristics were temporal parameters of the calls.We found similar patterns to those reported by Guerra et al. (2018) that identified a lack of information regarding the temporal and spectral parameters of calls (e.g., note duration, inter-note duration, and visual representations of the call).We also observed a lack of information regarding collecting information, such as whether the specimen was collected with an associated voucher number, and, if so, the type (public/private) and name of the collection where it was deposited.
The shortage of reporting these metadata compromises the comparisons among studies (Guerra et al. 2018).However, we noticed that in recent years, descriptions have become more robust by incorporating more characteristics.The six categories and 38 characteristics of calls considered in this study are based on the information that can be incorporated into the descriptions (see Appendix I, Additional Methods, text S2 for a complete description of the categories and characteristics).Spectral and temporal parameters, like call duration and frequency, are the most informative categories for describing vocalisations because they provide direct information on the calls; however, environmental and recording information can make the descriptions more robust.The presence of these characteristics is an indicator of the quality of the description, as they provide information not only about the call itself but also about the biotic and abiotic circumstances under which the recording took place (Köhler et al. 2017).Knowledge of the recording, analysing, and collecting methods helps in the standardisation of techniques, while qualitative and quantitative information about the environment and social interactions reinforces the understanding of the species' natural history (Köhler et al. 2017;Mitchell et al. 2020;Yip et al. 2020).Moreover, the inclusion of all these characteristics together guarantees the reproduction and comparison between results (Köhler et al. 2017;Guerra et al. 2018).
Before Köhler et al. (2017), the use of acoustic terminology was ambiguous (Wells 1977).Some authors described call-based characteristics while others used note-based descriptions, making comparisons between studies complicated.Köhler et al. (2017) presented a classification guide for anuran vocalisations that can be easily applied to most descriptions.Shortly after, Emmrich et al. ( 2020) accentuated the importance of unifying terminology, this time proposing a guild classification for advertisement calls.These two classification guidelines complement each other, and we recommend using both for the descriptions of new advertisement calls.
Despite the importance of all the previously discussed characteristics, every anuran vocalisation description should minimally contain a detailed report of the following parameters according to Guerra et al. (2018): duration of calls and inter calls (note/ pulses too), calling rate, note/pulse number, dominant and fundamental frequency of the calling unit, bandwidth, harmonics, and difference between upper and lower frequencies.The proper description of anuran vocalisations should be standardised and encouraged among researchers and amateurs to assure high quality data.

Trends in advertisement call publication
Our results show a clear dominance of first authors affiliated with foreign institutions in the publication of Mexican anuran advertisement call descriptions, particularly from the United States (78%).This is similar to the results of Rivera-Correa et al. (2021) for Colombia.This suggests scientific colonialism, inaccessibility to bioacoustics information in Mexico, and/or a lack of interest for this topic in the country.Similarly, the number of studies with sampling populations outside of Mexico is very high.Anurans generally have species-specific calls (Blair 1958;Toledo et al. 2015), but geographic variation among populations has been widely reported (Bernal et al. 2005;Amézquita et al. 2009;Ohmer et al. 2009;Rodríguez-Tejeda et al. 2014;Zornosa-Torres and Toledo 2019;Hernández-Herrera and Pérez-Mendoza 2021).Having records of local populations can prevent biases in the interpretation of studies using bioacoustics data.
An increase in the number of authors per publication is observed in recent years, likely due to collaboration alliances that can help reduce research expenses and publication costs, test complex hypotheses, or simply build scientific networks (Baethge 2008;Gazni et al. 2012;Abramo and D'Angelo 2015;Poulin and Presswell 2016).Classical descriptions tend to have one or two authors but with limited descriptions, while recent descriptions are more robust due to collaborations between researchers of different areas (Poulin and Presswell 2016;Guerra et al. 2018) and publications appearing in higher-ranked journals (Abramo and D'Angelo 2015).We also noticed the clear dominance of males as first authors in advertisement call descriptions.This may be attributed to risky fieldwork conditions in Mexico, especially considering that most Mexican anurans have nocturnal activity periods.However, it is well documented that herpetology is a male-dominated field and that the representation of women as main authors, editors, and principal investigators is among the lowest in zoology (Salerno et al. 2019;Distel 2020;Grosso et al. 2021).Most of the descriptions have been published in indexed scientific journals with a herpetological or a behavioural scope, making the descriptions more visible to their target audiences.Lastly, we observed that most of the descriptions are characterised as in situ descriptive studies.This type of study ensures natural conditions for the call emissions but might present high levels of variation due to uncontrolled variables.The most common categories of research that produce advertisement call descriptions are Bioacoustics and Behaviour, which aligns with some of the descriptions with more detailed information.

Future directions
In this study, we gathered a robust database that allows for insights into the state of knowledge on advertisement call descriptions for Mexican anurans.Our results show that despite the volume of descriptions to date, the knowledge of Mexican anuran calls is far from complete.We found several shortfalls in the knowledge of the advertisement calls, specifically for those of highly threatened, tropical species (Hylidae) and taxonomically problematic groups (Eleutherodactylidae).With this work, we hope to provide useful information for the planning of future research regarding anuran bioacoustics in Mexico to increase our understanding of these understudied taxa.
Bioacoustics is a powerful tool in herpetological research that can help elucidate evolutionary relationships or prioritise conservation.The availability of bioacoustics information such as detailed call descriptions can facilitate the incorporation of bioacoustics in other herpetological studies.Like tissue and specimen collections, audiovisual repositories are important reservoirs of information.The collection of audio-visual archives in accessible local or national repositories can encourage more Mexican researchers to become interested in the vocal repertoires of anurans (Zamudio-Torres et al. 2020;Mendoza-Henao et al. 2021).We also want to emphasise that advertisement calls are not the only type of vocalisation produced by anurans and that the recording of other call types should also be encouraged.
To study anuran vocalisations outside of descriptions, we need access to more vocalisations (Guerra et al. 2018;Mendoza-Henao et al. 2021).Unity is strength, and local initiatives such as the Fonoteca de Sonidos de Anfibios (UNAM) can pair with targeted citizen science projects (e.g., iNaturalist's Bioblitz) to deposit new records in their collections.The recent incorporation of AI technologies into citizen science can decrease the number of outlying records that usually diminish the quality of these types of datasets (Kosmala et al. 2016).Depositing anuran vocalisations into audio-visual repositories, especially open and local ones, is important to facilitate the future use of this information in more specialised studies.
Even though interest in anuran vocalisations has been around for more than 50 years in Mexico, standardised terminologies and guidelines for data collection, processing, and interpretation are still lacking.This has hindered the advance of anuran bioacoustics in the country.We stress the importance of promoting studies on anuran vocalisations and the creation of an introductory manual that welcomes people to this research area.The present work provides a general overview of the problems surrounding the description of acoustic anuran signals in Mexico, and we hope it can serve as a baseline for future anuran bioacoustics research in the country.

Figure 1 .
Figure 1.Number of publications per year that describe at least one advertisement call.We considered only those documents that had an original description or a novel characteristic of the call.Illustration: Tripion petasatus by Samantha Ordoñez-Flores.

Figure 2 .
Figure 2. Percentage of species per family with at least one description of their advertisement call.The white number in the salmon bar indicates the total number of species per family.

Figure 3 .
Figure 3. Number of Mexican anurans with at least one advertisement call registered in online audiovisual repositories, grouped by family.Total species = total number of species per family; Macaulay = Macaulay Library, Cornell University, USA; Fonoteca-FC = Fonoteca de Sonidos de Anfibios de México of the Faculty of Sciences, National Autonomous University of Mexico, Mexico.

Figure 4 .
Figure 4. Number of endemic Mexican species with descriptions of their advertisement call per biogeographic region and per conservation category according to the International Union for the Conservation of Nature (IUCN) red list of threatened species.We divided the country into the three broadest biogeographic regions (Neotropics, Nearctic, and Transition Zone) and classified each endemic species' distribution to one or more biogeographic regions.Those with ranges spanning all three biogeographic regions were considered wide-ranging.Categories of threat: Data Deficient (DD), Least Concern (LC), Near Threatened (NT), Vulnerable (VU), Critically Endangered (CR).Not evaluated or missing species on the red list were considered as NE.

Figure 5 .
Figure 5. Number of bioacoustics metadata included in each description per year.

Table 1 .
Metadata information reported in 385 advertisement call descriptions of Mexican anurans.

Table 2 .
Number of publications recovered for the distinct categories of study types.
Figure 7. Percentage of studies with advertisement call descriptions per research area.