Introduction

The Neotropical high montane includes a variety of unique ecosystems on the planet, including montane and high Andean forests and large areas dominated by native grasses (Gradstein et al., 2008; Hofstede, 2013). The tropical alpine grasslands of America are located on mountain peaks at elevations above 3,000 m. The humid and dry punas of the Andes, the high-altitude grassland in southeastern Brazil, and the equatorial Páramos are classified within these grasslands (Ruiz et al., 2008; Christmann & Oliveras, 2020). Páramos are part of the high montane ecosystems with the largest extension, being distributed in the equatorial Andes and southern Central America, above the limit of the high Andean forests and below the snow line (Lauer, 1981; Hofstede et al., 2003; Anthelme & Peyre, 2019).

In Colombia, Páramos are considered strategic ecosystems of great importance due to the ecosystem services they provide, such as carbon sequestration and the regulation of almost 70% of the country's drinking water (Rivera Ospina & Rodríguez, 2011; Hofstede, 2013; Cabrera, 2014). The Colombian Páramos have a unique diversity in terms of species richness and endemicity on the planet. This is explained mainly by the complex geological and environmental history of the Andes (Navas, 1999; Gregory-Wodzicki, 2000; Doan, 2003; Madriñán et al., 2013). The high speciation rates are associated with differences in environmental conditions within mountain systems, which are compared to continental islands (Vuilleumier, 1970; Navas, 2002; Mendoza et al., 2015).

Amphibians are one of the most diverse lineages of vertebrates in Colombian Páramos (Ardila-Robayo & Acosta-Galvis, 2000; Lynch & Suárez, 2002). Given their life history and physiological traits, both salamanders and anurans have adapted and diversified in these environments. Their low metabolisms rates, behavior, morphological traits, and thermal adaptations contributed to compensate for overexposure to wind and UV rays or heat loss in an ecosystem where the daily circadian cycle changes abruptly. (Navas, 1996a; Navas, 2006; Carvajalino-Fernández et al., 2011; Navas et al., 2013). Most studies of Páramo amphibians focus on descriptions or species lists (Rueda-Almonacid & Hoyos, 1991; Ardila-Robayo & Acosta-Galvis, 2000; Lynch, 2001; Buitrago-González et al., 2016; Henao et al., 2019), as well as some research in Ecophysiology (Carvajalino-Fernández et al., 2011; Guarnizo & Cannatella, 2014; Navas, 1996b, 1996), morphology (Hoyos et al., 2015; González-Durán et al., 2017; Mendoza - Henao et al., 2019), bioacoustics (Fandiño et al., 1997; Gutiérrez & Lüddecke 2002) and community ecology (Gutiérrez-Lamus et al., 2004; Roach et al., 2020; Duarte-Ballesteros et al., 2021). However, despite a large amount of research carried out in the Colombian Páramos, scientific publications are scattered in various sources and some of them are difficult to access.

The study of anurans and salamanders diversity in the high mountains of Colombia is of vital importance given the accelerated loss of ecosystems in the tropical Andes (Etter & van Wyngaarden, 2000; Etter et al., 2008; Etter et al., 2018). Furthermore, habitat loss in synergy with climate change will increase the risk of extinction of amphibian species (Agudelo-Hz et al., 2019). These conservation problems are even more evident in the Páramo ecosystem because its transformation by human actions has recently increased (Alarcón et al., 2002; Rivera Ospina & Rodríguez, 2011; Cadena-Vargas & Sarmiento, 2015; Sarmiento et al., 2017). It is urgent to have a database that enables the consolidation of the current knowledge of amphibian studies in the Páramo and, from a bibliometric approach quantify and describe patterns in scientific publications and identify information gaps (Aldana-Domínguez et al., 2017; Arbeláez-Cortés, 2013a; Urbina-Cardona et al., 2023). There are some bibliometric analyzes in Colombia about birds (Estela et al., 2010), turtles (Bock & Páez, 2017), herpetofauna (Urbina-Cardona et al., 2023), and tropical dry forest ecosystem (Aldana-Domínguez et al., 2017), as well as a review on national biodiversity (Arbeláez-Cortés, 2013b). Bernal & Lynch (2008) reviewed the geographic distribution and altitude ranges of amphibian species in the Colombian Andes, but the state of knowledge of the amphibians that inhabit the Páramo ecosystem is unknown to date. Systematic reviews are currently popular and their important contribution to knowledge is increasingly recognized, which is related to the high quality and transparency that these works have (Siddaway et al., 2019). The aims of this study are 1) to describe the temporal patterns in scientific productivity between scientific articles in indexed and non-indexed journals; 2) to characterize the scientific productivity according to the taxonomic groups and the main research topics over the years; 3) to identify spatial patterns in the frequency of publications by sector, department, and Páramo complex; 4) to evaluate changes in the composition and richness of amphibian species in the Páramo complexes by sector in Colombia; and 5) discuss knowledge shortfalls that become challenges for the conservation of Páramo amphibians.

Methods

Study area: Páramos are present in Colombia, Costa Rica, Ecuador, Perú, and Venezuela (Hofstede et al., 2003; Cortés-Duque et al., 2013). In the case of Colombia, this ecosystem covers a total of 2.906.137 ha, which is close to 2.5% of the country's mainland and 50% of the world's Páramos (Morales et al., 2007). Páramos have political and administrative presence in 22 departments, and they are geographically present in the Central, Eastern, and Western Ranges, as well as in the Sierra Nevada de Santa Marta Mountain range (SNSM). As an environmental protection policy for the country, the government created a delimitation process that groups most of the Colombian Páramos under the figure “Páramo complexes” (Morales et al., 2007). This process resulted in a delimitation of 36 complexes grouped into five sectors (Rivera Ospina & Rodríguez, 2011; Sarmiento & Ungar, 2014; Sarmiento & León, 2015; Figure 1).

Figure 1.

Study area. (A) Distribution of Páramos in the neotropics. (B) Distribution of the 5 sectors within which the 36 Páramo complexes are grouped, according to the delimitation process of Páramo complexes for Colombia (IAvH, 2012).

The Páramo ecosystem is part of the Colombian high mountains, the geographic space corresponding to the altitudinal culminations of the mountainous formations that occur from 2700 ± 100 m (Sarmiento et al., 2013). Until now, there is no consensus about the altitudinal limits between the ecosystems of the Andean, high Andean, and Páramo orobiomes since the species replacement between these ecosystems varies greatly throughout their distribution according to their location and climatic characteristics (Cuatrecasas, 1958; Lauer, 1981; Rangel & Rangel -CH, 2000; Rodríguez et al., 2006; Cortés-Duque et al., 2013; Hofstede et al., 2014).

Literature search and databases: Our systematic literature survey included seven different search strategies to detect the largest number of documents published in indexed journals and local sources with low accessibility, between the years 1863 and 2021: 1) A comprehensive Title-Abstract-Keywords search in Web of Science (WoS including all databases: the Core collection, Korean journals and Scielo citation index) and SCOPUS on the 30^th of November 2021, between the years 1927 and 2021 and using the following search syntax: (Páramo* OR "andean forest*" OR "Tropical High mountain" ) AND (anura* OR amphibian* OR tadpole* OR frog* OR toad* OR salamander OR caudata OR gymnophion* OR caecil *). 2) A search in all the published volumes and issues of all Colombian journals indexed by MinCiencias (Ministerio de Ciencia, Tecnología e Innovación de Colombia, 2018) with the research topics of natural sciences or biological sciences ( online Appendix 1). 3) The same search syntax from search strategy 1 was used together with the term “Colombia” in the Biodiversity Heritage library (BHL) (Biodiversity Heritage, 2021) until 2021. 4) A deep search on the digital repositories of the Instituto de Investigación de Recursos Biológicos Alexander von Humboldt - IAvH (2019) (until 2021) and the book series “Colombia Diversidad Biótica” (Rangel-Ch, 2019). 5) We consolidate a taxonomic list of Páramo species reported in the scientific literature (Acosta-Galvis, 2000, Ardila-Robayo & Acosta-Galvis, 2000; Lynch & Suárez, 2002; Bernal & Lynch, 2008; Acosta-Galvis, 2015; Buitrago-González et al., 2016; Meza-Joya & Torres, 2016). 6) We compilated the available records of amphibians present in the Biodiversity Information System [by its Spanish acronym SiB], and the Global Biodiversity Information Facility (GBIF), from the delimitation of Páramos of Colombia project and the biological collections of the University of Antioquia (MHUA), National University of Colombia (ICN), Universidad del Valle (UVC), Universidad del Quindío (ARUQ) and the Alexander von Humboldt Research Institute (IAVH) to complement the geographic and taxonomic information of the species. And, finally, 7) We searched the bibliographic references for each species (until May 2021) reported in the International Union for Conservation of Nature (IUCN) red list assessments (IUCN, 2021) and the list of amphibians of Colombia by Acosta-Galvis (2021).

We generated a database and manually removed duplicates for a total of 882 documents. For each of these documents, the abstract, methods, study area, results, and supplementary material were read following the PRISMA protocol (Liberati et al., 2009; Shamseer et al., 2015), to ensure that all eligible documents met at least one of the following inclusion criteria: 1) It includes studies in ecosystems above 2700 m in Colombia; 2) It presents data on species reported in the taxonomic list compiled from the scientific literature (search strategy 5), and available records (search strategy 6); and 3) Study areas coincide spatially with some Páramo complex. After a complete review and exclusion of documents that did not meet the selection criteria (n = 477 documents), we consolidated a final database (Figure 2) comprising 405 eligible documents ( online Appendix 2) to search the following information:

Figure 2.

PRISMA methodology (Liberati et al., 2009) used in the present study. Acronyms: BHL = Biodiversity Heritage library; IAVH = Alexander von Humboldt Research Institute; IUCN = International Union for Conservation of Nature; GBIF = Global Biodiversity Information Facility; SIB = Biodiversity Information System of Colombia

Data analysis: Knowledge regarding Páramos amphibians was explored through descriptive graphs to visualize temporal patterns in publications over the years and by type of document. We used bar plots to contrast the number of papers in scientific journals within the Journal Citation Reports (JCR) (Web of Science Group - Clarivate) and Scimago (SJR) (Scopus) with publications in non-indexed sources to contrast in which source of information (indexed or non-indexed) the literature of amphibians Páramos is being generated and their temporal trends. In this way, it can be evidenced if the literature produced on this topic in the country is remaining in easily accessible repositories and if they are published in high-impact scientific journals. Documents were characterized according to the 21 research topics for each taxonomic order and visualized through a Foam tree (Carrot Search FoamTree, 2019). In the Foam tree, each category is divided into polygons of different colors and sizes according to their class and frequency, respectively.

We extracted the information on “Research topics” from  online Appendix 2 to describe the number of publications by research topic and year range and we transformed this data into a square root to calculate the association index of Whittaker (1952). With the Whittaker matrix, we created a heat map to show the temporal patterns in the 21 research topics. In the heat map, the frequency of research topics is represented by a color gradient that goes from low association (cold tones) to a high degree of association (warm tones) (Somerfield & Clarke 2013). Additionally, based on the similarity profiles (SIMPROF routine; Clarke et al., 2008), the grouping between 21 research topics was statistically validated to identify sets of topics that had similar productivity patterns over time.

To identify the spatial patterns in productivity, we generated a database from 1) Each record by species and geographical unit (department, Páramo complex, sector), reported in the documents of  online Appendix 2; 2) The information on the Extent of Occurrence (EOO) from each species of the taxonomic list (search strategy 5) and available in the last Red list assessments (IUCN, 2021); and 3) The records downloaded from the GBIF and the SIB. For this procedure, we used the “intersection” geoprocessing tool in QGis 3.24.1 ( Qgis.org, 2022) and both the occurrences (GBIF and SIB) and the EOO in the Redlist (IUCN, 2021) were extracted and overlapped with the cartographic information on the Páramo complexes delimitation of Colombia at a scale of 1:100,000 (IAvH 2012). We excluded from this analysis all the papers related to taxonomic lists that only presented general information at the department level without reporting records in the Páramo (n = 8). We also excluded 372 records from the literature with uncertain geographic information about the association of the amphibian species to the Páramo ecosystem and 775 GBIF records that did not intersect with Páramo complexes.

The final database ( online Appendix 3) contains a total of 2835 records with the following information: species, publication, year, Páramo complex, sector, and department. Given that, in the same study, a species could occur in different departments or Páramo complexes, the information is reported for each of the geographical units (Páramo complex within a sector and in a department). From the  online Appendix 4, the number of publications per geographic unit was estimated and maps were generated using the QGis 3.24.1 program ( Qgis.org, 2022). Similarly, heat maps were created (Somerfield & Clarke, 2013) to explore the association between 21 research topics by departments. And the grouping (SIMPROF routine) between research topics and departments was statistically validated in the PRIMER 7.0.13 and PERMANOVA add on program (Clarke & Gorlye, 2006, Anderson et al., 2008; Clarke & Gorley, 2015).

Finally, based on the specific records (museum specimen with data georeferenced or verifiable Páramo locality or EOO polygon coincidence with Páramo complex; n = 1979), a taxonomic list of amphibians was generated for the Colombian Páramo complexes. We excluded from this analysis all specimens that were reported as “sp.” “cf.” or “aff.” or within species complexes (eg Atelopus “ignescens”) (n = 1101). We run a multivariate analysis of variance based on permutations (PERMANOVA; Anderson, 2001) to determine the effect of the geographic sector of the Páramo complex on amphibian alpha and beta diversities. The Jaccard similarity index was calculated to evaluate changes in the composition of the assemblages, and for richness, the Euclidean distances between the Páramo complexes were also calculated. The analysis was performed under a partial sum of squares (type III) and 9999 permutations of the residuals under a reduced model. When differences between sectors were found, a student's t-test was calculated as a post hoc test. Finally, we explore the association of the presence of species by Páramo complex and sector with a heat map (Somerfield & Clarke, 2013). The classification of species was validated with the SIMPROF routine in PRIMER 7.0.13 and PERMANOVA add on program (Anderson et al., 2008; Clarke & Gorley, 2015).

Results

Scientific productivity on Páramo amphibians: Types of publication and temporal trends

The 405 documents were published between 1863-2021 and included: 160 research articles, 134 IUCN species Red list assessments, 35 book chapters, 29 SiB and GBIF data occurrence reports, 15 herpetological notes, 12 CARC catalog files, 8 review articles, 9 books, and 2 data papers. The number of publications increased by 77.18% since the year 2000, mainly with contributions from the IUCN Red list assessments, research articles in scientific journals (indexed or non-indexed), and herpetological notes (Figure 3).

Figure 3.

Temporal pattern of scientific productivity on amphibians from the Colombian Páramo differentiated by the type of document.

The scientific articles (data papers, research articles, reviews, and herpetological notes n = 185), were published in 73 journals. 115 of these publications (62%) are found in 42 journals indexed in the JCR of Web of Science (45.36%) or the Scopus SRJ (54.63%), of which the ones with the greatest contribution were, in descending order: Caldasia, Zootaxa, Herpetologica and Journal of Herpetology (Figure 4A). On the other hand, 285 documents are not indexed, of which 25.26% correspond to articles published in scientific journals, 46.26% to evaluations of the IUCN Red list assessments, 12.45% to book chapters, 8.54% to occurrences from SiB-GBIF, 4.27% to the catalog of the CARC catalog files, and 3.20 % to books. The 43.75% of the non-indexed publications are Colombian and within these, we found two scientific journals (Revista de la Academia Colombiana de Ciencias Exactas, Físicas y Naturales and Biota Colombiana) and two books: El Libro Rojo de Anfibios de Colombia (Rueda-Almonacid et al., 2004) and Ranas Arlequines (Rueda-Almonacid et al., 2005) as the most representative from this group (Figure 4B).

Figure 4.

The number of publications on Páramo amphibians in Colombia in (A) journals indexed in SRJ and JRC. (*) correspond to internationally indexed journals from Colombian and (B) in non-indexed publications.

Taxonomic and thematic characterization of scientific literature

A total of 142 species were recorded within the Amphibia class with distribution in Colombia's Páramo complexes: Order Anura with 134 species in six families, and Order Caudata with five species in one family (Plethodontidae) (See  online Appendix 4). 95 species of our proposed list are reported as endemic to Colombia. Most of the publications were about anurans (94.68%), while 20 documents were reported for salamanders, and no information was found for the order Caecilia in the Páramos of Colombia. At the family level, Strabomantidae presented the highest number of studies (n = 219), of which the genera Pristimantis (n = 201) and Niceforonia (n= 34) were the most representative. The second-best represented family in the documents was Bufonidae (n = 116) with the genera Atelopus (n = 89), and Osornophryne (n= 32), followed by the family Hylidae (n=79) with the genera Dendropsophus (n= 62) and Hyloscirtus (n=44). The species with the highest frequency of appearance in the database publications were Dendropsophus molitor (Schmidt, 1857) (n=59), Pristimantis bogotensis (Peters, 1863) (n=44), Hyloxalus subpunctatus (Cope, 1899) (n=29) and Atelopus carrikeri Ruthven 1916 (n=27).

The most frequent research topics within the anuran documents were Natural history (242), Conservation (138), and Systematics and Taxonomy (109). In contrast, the order Caudata was included more often in Taxonomic lists (12), Natural history (10), and Systematics and Taxonomy (8) studies (Figure 5).

Figure 5.

Polygons´ visualization of the research topics within which the scientific literature on the amphibians from the Colombian Páramo separated by taxonomic order was classified.

The publication on Natural history and on Systematics and taxonomy has been maintained over the years, but the state of knowledge on Páramo amphibians in Colombia has been enriched since the 2000s with other 19 research topics. These topics were classified into four groups with similar patterns, of which 47,6% were classified in the same group, and the other three groups presented between 1 and 7 research topics (Figure 6). The first studies in Biogeography, Ecophysiology and Bioacoustics, as well as an increase in publications on topics such as Systematics and Taxonomy, and Natural history, can be traced back to the 70s. Later, in the 90s, publications on Ecophysiology and Bioacoustics continued to increase; and, from the 2000s onwards, new topics have appeared such as Infectious Diseases, Community ecology, Reproductive ecology, Diet, and Climate change. In contrast, research topics in Population ecology, and Ethology have shown stagnation or decline. Geographic distribution, Conservation, and Natural history are some of the most representatives of the last decade.

Figure 6.

Heat map of the study research topics on amphibians from the Colombian Páramo over time. The tree on the left shows the classification of the 21 research topics into 4 groups according to their degree of association (Whittaker index) with the year ranges.

Spatial patterns from the scientific literature

Of the 22 departments with Páramos in Colombia, Cundinamarca registered the highest number of studies (n= 93), followed by Antioquia (n=64), Nariño (64), Cauca (n=61), and Boyacá (59) (Figure 7A). The highest number of publications were reported in the Eastern ranges sector (n=159), in which 17 Páramo complexes are distributed, and the Central sector (n=129), in which eight Páramos complexes are present (Figure 7C-D). The Páramo complexes that registered the highest number of studies were Chingaza (N=70), in the Eastern range sector; Los Nevados (n=62), in the Central range sector; and La Cocha-Patascoy (n=56), in the Nariño-Putumayo sector.

Figure 7.

Geographic patterns the number of documents on amphibians of Colombian Páramos (A) by department (B) Sierra Nevada de Santa Marta (C) Eastern range sector, (D) Central range sector, (E) Nariño-Putumayo sector, (F) Western range sector.

From the number of publications by research topics, the departments were classified into nine groups according to the degree of association with research topics. Besides, the 21 research topics were classified into 4 groups with similar patterns showing great thematic diversity between departments. Three of these departments (Nariño, Cauca, and Cundinamarca) are grouped with many publications on Natural history and Conservation (Figure 8). Cundinamarca represented 85% of the research topics; followed by Antioquia, Boyacá, Tolima, Caldas, Cauca, Nariño, Quindío, and Putumayo, which presented between 10 and 16 topics. Whereas the departments with the lowest number of publications were Arauca, Caquetá, Huila, Tolima, Cauca, Norte de Santander, and Santander, which represented between 1 and 2 research topics.

Figure 8.

Heat map of research topics by Department. The tree on the left shows the classification into 4 groups of the 21 research topics by their degree of association (Whittaker index) with the departments. The classification of the 22 departments into 9 groups is also evident. Classification trees represent, with red nodes, those entities that have similar publishing patterns. The degree of association is shown as a gradient of colors ranging from warm tones (high degree of association) to cold tones (low degree of association).

Composition of amphibians in Páramo complexes and comparison between sectors

We found differences in the species richness (Pseudo-F= 13.8; p.perm= 0.0001; size effect = 68%) and in the composition of the amphibian assemblages (Pseudo-F= 4.15 p.perm= 0.0001; size effect = 34.3%) between sectors. The general pattern in assemblage composition shows that Sierra Nevada de Santa Marta (SNSM), Cordillera Central, and Nariño-Putumayo presented noticeable differences from other sectors such as Eastern ranges and Western, which is associated with the high rate of species turnover that occurs in the high Colombian mountain. The 142 species were classified into 42 groups according to their distribution in the 35 Páramos complexes (Picachos was not included because we do not report species in this complex) and within the five sectors (Figure 9, Table 1). A high degree of endemicity can be observed in Páramos complexes such as SNSM in which all species are unique to this area including the genus Serranobatrachus (Arroyo et al., 2022). Complexes like Perijá, located in the north region of the eastern sector, also report endemic species like Pristimantis reclusus (Lynch, 2003) and Tachiramantis cuentasi (Lynch, 2003), and the Duende complex (Western Sector) presents unique fauna like Bolitoglossa hiemalis Lynch 2001, Pristimantis duende (Lynch, 2001) and Pristimantis xeniolum (Lynch, 2001). Almost all the Páramos complexes of the eastern sector share species with wide distribution in the eastern range like Pristimantis elegans (Peters, 1863), P. bogotensis, D. molitor or Bolitoglossa adspersa (Peters, 1863). On the other hand, complexes in the Central sector located to the south (Guancas-Puracé-Coconucos and Sotará) share around 12 species with the complexes Doña Juana Chimayoy and La Cocha Patascoy (Nariño-Putumayo sector).

Figure 9.

Heat map of species by Páramo complex. The tree on the left shows the classification into 42 groups of 142 species. The classification trees validated with the SIMPROF routine, represent with red nodes, those entities that have similar distribution patterns by Páramo complex (see Table 1) and by sector: .= Central; = Nariño-Putumayo; = Oriental; = Sierra Nevada de Santa Marta; and = Western.

Table 1.

Number of Studies and Species Reported by the Scientific Literature in the Páramo Complexes Within Five Sectors of Colombia.

Discussion

Based on the 405 published documents, this review is the study with the highest bibliographic, geographical, thematic, and taxonomic coverage of amphibians of the Páramo ecosystem. Generally, the literature compiling data on Páramo amphibians was limited to species lists constructed from literature reviews in scientific journals (Lynch & Suárez, 2002; Buitrago-González et al., 2016) and in some cases, it was complemented with information from biological collections (Ardila-Robayo & Acosta-Galvis, 2000; Bernal & Lynch, 2008). However, our study, revels that a low percentage of this literature is published in scientific journals indexed in the JCR of Web of Science or the SRJ of Scopus. The highest percentage of the publications in our database are unindexed and are represented mainly by the IUCN red list assessments, which are open access (IUCN, 2021). Most discontinued journals (national or foreign) like Lozania, Trianea, Life Science Occasional paper, and Miscellaneous publications (University of Michigan. Museum of Zoology) are found in repositories that are not always easy to track or access and do not have major visibility.

In the present study, we included two documents considered grey literature, specifically books and reports from government institutions. These documents were accessed through search strategies 4 and 7, followed in the methodology. Several authors recognize the importance of grey literature in creating an overview of knowledge for a particular topic, as noted by Mahood et al. (2013), Hortal et al. (2015), and Adams et al. (2017). This type of literature often contains anecdotal scientific data or data obtained through an absent or lax research design (e.g.Christie et al., 2020, 2021). Still, it complements state-of-the-art on underexplored sites or understudied species. This is especially relevant to complement conservation actions for populations of native species and their habitat, as such measures depend on baseline information about natural history, life histories, geographic distribution, and their short-, medium- and long-term spatial dynamics. For species with a distribution restricted to naturally fragmented ecosystems (e.g., páramo amphibians), grey literature repositories may represent the only information available to guide management and conservation actions.

Open data provided by databases and reported in our study (e.g., GeneBank, IUCN, SiB-GBIF) have accelerated research in the field of Conservation biology, as well as in research topics in Evolutionary biology and Systematics (Lacher et al., 2012; Baxevanis & Bateman, 2015). However, despite being in an open science context, we found just two data papers in our database (Henao et al., 2019; Mendoza-Henao et al., 2019). Data papers are a type of recent academic publication that offers relevant information regarding the visibility and re-use of data that can be available for public use and maintains both scientific rigor and public confidence (Roche et al., 2015; Jiao & Darch, 2020; Machuca-Martínez, 2020). This is of great importance as the public availability of primary biodiversity data is essential for ecological research (Huang et al., 2013), future meta-analyses, and decision-making in critical ecosystems such as the Páramo. However, the low frequency of these types of papers may be because the authors 1) do not want to share their primary data before they are published; 2) do not want to undergo a peer review procedure again, which may place an additional burden; or 3) do not want to subject their data to a statistical reanalysis by other colleagues (Huang et al., 2013; Wicherts et al., 2011). To advance the state of knowledge of Páramo amphibians it is essential that authors have the good practice of uploading their raw data in repositories (e.g. Dryad, Mendeley Data or ResearchGate) and report the download link in their publications.

On the other hand, the biological collections provided data for 56.45% of the investigations reported in scientific articles (Data papers, research articles, reviews, and herpetological notes), mainly related to research topics such as Systematics and Taxonomy, Anatomy and Morphology, Natural history, taxonomic lists, biogeography, and geographical distribution. These data highlight the importance of collections as sources of scientific data to document the country's diversity (Medina et al., 2016). Nevertheless, the lack of availability and updating of records in several of the collections consulted in the GBIF is notorious. The largest number of specimens reported in our database are concentrated in the collections of the ICN and IAvH, but many of them lack georeferencing and altitudinal data. Other specimens are omitted in these GBIF reports and the oldest ones (mainly those of the nineteenth and twentieth centuries) present errors related to their localities that require an update (Vásquez-Restrepo, 2021).

We report a low percentage of studies whose data source is from fieldwork (24.19%) or laboratory data (2.22%). The lack of field information contributes to the underestimation of the amphibian diversity reported for the Páramos. There are recent efforts related to explorations promoted by academic or governmental institutions like Expedition BIO-MinCiencia; and Delimitation of Páramos of Colombia (Acosta-Galvis & Borja-Acosta, 2021; Henao et al., 2019; Ravelo & Martínez, 2019). These works sought to increase the knowledge of the Páramos fauna and flora. However, the information is not available for all paramos complexes and the reports accessible in the GBIF had more than 30% of taxonomic gaps most related to the Pristimantis genus. Other data from these reports just provided “Human Observation” but not a specimen deposited in a scientific collection. Some factors that could explain the low number of fieldwork are 1) the difficult access to these sites and the abrupt conditions of climate and topography, 2) political-economic deficiencies (Economic resources, research permits), and 3) public safety problems (Arbeláez-Cortés, 2013a). This represents a Linnean shortfall (Hortal et al. 2015) in which there is a large gap for the currently formally described species to approach the total number of amphibian species that inhabit the Páramos and are still unknown to us (Brito 2010).

Implications for Conservation

Colombian Páramos register an alarming degree of intervention with the replacement of natural cover and only 58.33% are within the national system of protected areas (Gómez et al., 2015, Cortés-Duque et al., 2013). According to the analyses carried out by the IAvH (Moreno et al., 2020) on a national scale, all Páramo complexes report an increase in transformed coverage from 0.6 to 0.9% due to agricultural activities, ranching, and mining (Burbano-Girón et al., 2020). On a regional scale, the Eastern, Central, and Nariño-Putumayo sectors show a pattern of increase in transformed areas up to 18.3% in the last decades, which could be associated with the “upward movement” or “upward expansion” of the agricultural frontier and the intensification of animal production in the sub-Páramo belt, which has been occurring in different parts of the equatorial Andes (Hess, 1990). These data can be worrying since these sectors account for more than 80% of the species reported as endemic to the Páramo or with populations established in these sites, and these are organisms with restricted areas of occupation compared to other biological groups. Impacts on the natural cover can lead to habitat conversion for many species of amphibians. They have a long-term negative effect on amphibian populations and increase the probability of direct mortality of species (Nowakowski et al., 2017; Catenazzi, 2015). Currently, 66.19% of Páramo amphibians are within some state of threat: Vulnerable (VU), Endangered (EN), Critically Endangered (CR); mainly due to 1) Areas of Occupation (AOO) between 10 and 2000 km2, associated with the low geographical records that exist per species and 2) the increase in threats related to categories such as “Agriculture & Aquaculture”, “Residential & commercial development”, “Invasive and other problematic species, genes & diseases” and “Energy production & mining” (Stuart et al., 2008, IUCN, 2021).

The trend towards the transformation and replacement of natural Páramo cover by crops or grazing areas shows that habitat conservation is a high priority for the persistence of amphibian species. Studies carried out in some Páramos show the use of microhabitats associated with endemic plants from these areas, such as cushions and shrubs, tussock grasses (Calamagrostis spp.), rosette plants and stemmed rosette plants of Espeletia spp (Lynch, 1998; Lynch, 2000; Acosta-Galvis, 2015). Also, it is important to note that, at finer spatial scales, each Páramo can generate a unique variety of microclimates or microecosystems to which amphibians could be adapted (Navas, 1996a; Buitrago-González et al., 2016; Duarte-Ballesteros et al., 2021). However, the absence of population data or quantitative records of natural history data for most species is notorious, without regard to the large gaps that exist on a spatial and thematic scale and the concentration of studies on a few research topics, complexes of Páramos, and even common species. The gap in knowledge of species population dynamics is called a Prestonian shortfall and represents a challenge from long-term monitoring to assessing the risk of species extinction and implementing management and conservation actions for populations and their habitats (Hortal et al., 2015). More than 80% of the species that inhabit or transit the Páramo in Colombia only have their descriptions information or have been mentioned in the literature in species lists but little is known about their ecological requirements or data related to specific threats: Infectious diseases, invasive species, and climate change.

On the other hand, current information gaps related to the assignment of the nominal status of species and errors in geographical records can affect categorization in Red Lists, as they cause inaccuracy in the estimation of the EOO and AOO. Consequently, we could be hiding the status of some species with degrees of threat (de Magalhães et al., 2018). Currently, eight of the species with a presence in the Páramo are found with Deficient Data (DD). Two of them are uncategorized, and 80.85% of the species endemic to certain Páramos complexes are distributed only in one “Protected area” (natural parks or private ecological reserves) as a conservation figure (See  appendix 4). Although a few studies report the use of good practice protocols for amphibian sampling (e.g., using gloves to handle animals), there is a clear lack of efforts to mitigate threats to the habitat of these species or to implement concrete actions for their conservation. (i.e., amphibian translocation, citizen science, engagement of landowners and other volunteers in managing land for amphibians, or management plans with the support of environmental institutions; Sutherland et al. 2019. This lack of effort is mainly due to the absence of field data, including population size, distribution and trends, life history and ecology, population trends, and even taxonomy (IUCN, 2021). The knowledge shortfalls identified in this paper (Linnean, Wallacean, Prestonian shortfalls; Hortal et al., 2015) show thematic, spatial, and temporal biases that skew the possible evidence on the conservation (sensu Christie et al., 2020, 2021)) of amphibians inhabiting the páramos of Colombia. In a regional context, this is of great importance since the tropical Andes is the place in the world with the largest number of amphibian species with protection figures gaps in terms of protected areas (Nori et al., 2015).

These data reveal the need to increase the knowledge not only of the Páramos inventory but of the biology and distribution of the species that inhabit them, because this information contributes to the generation of monitoring programs and conservation actions focused on the attributes of the habitat of the species. Moreover, the information generated in the scientific literature and records in databases (such as SiB and GBIF) can strengthen other initiatives like the Colombian community of practice called "Biomodelos" ( http://biomodelos.humboldt.org.co/es), which seek to generate more accurate distribution models and in which only 30 species (only 21.12%) of our list are reported. Once validated by experts, the distribution of species in Biomodelos is used as input to calculate different metrics including AOO and EOO, distribution in protected areas, and changes in distribution under future scenarios of landscape transformation (Velásquez-Tibatá et al., 2019). For this reason, it becomes an indispensable tool to make the processes of amphibian extinction risk assessment in Colombia more efficient (Urbina-Cardona et al., 2023). It is important to highlight that generating data in situ involves a greater sampling effort, the use of specialized equipment, and reagents for tissue collection or animal experimentation. These are activities that require greater financial support, which is a challenge since the country is currently investing > 1% of its gross domestic product in science, technology, and innovation (The World Bank, 2018; Guevara, 2021).