The herpetofauna of São Tomé and Príncipe consists of nine species of amphibians, all endemic, and 21 species of terrestrial reptiles, of which 17 are endemic. Our current knowledge regarding its natural history, ecology, and distribution is limited. Here two important tools are provided to support researchers, conservationists, and local authorities in the identification of the country’s herpetofauna: an illustrated key to the herpetofauna of the two islands and surroundings islets and a DNA barcode reference library. The keys allow a rapid and unambiguous morphological identification of all occurring species. The DNA barcodes for the entire herpetofauna of the country were produced from 79 specimens, all of which are deposited in museum collections. The barcodes generated are available in online repositories and can be used to provide unambiguous molecular identification of most of the species. Future applications and use of these tools are briefly discussed.

Biological surveys, conservation, Cytochrome c oxidase subunit I (COI), DNA metabarcoding, environmental DNA, Herpetofauna, Oceanic Islands

The herpetofauna of São Tomé and Príncipe, a small insular country in the Gulf of Guinea, West Africa, has been the subject of an intense taxonomic and systematic revision in the last decade (see Bell et al. 2022a and Ceríaco et al. 2022 for an overview). Hence, the country’s herpetofauna is currently one of the best known in Africa, with nine recorded species of amphibians (Bell et al. 2022a) and 21 recorded species of reptiles (Ceríaco et al. 2022). Of these, all of the amphibian species are endemic to their respective island (Bell et al. 2022a), while 17 out of 21 reptiles are also endemic (Ceríaco et al. 2022).

Recent research efforts have focused on the taxonomic revision, systematic placements, and biogeographic patterns associated with the amphibians and terrestrial reptiles of the country. However, not much attention has been given to their natural history, distribution, and ecological relationships. Besides some anecdotal data available in taxonomic papers, only a few studies provided details on the trophic ecology of São Tomé and Príncipe herpetofauna (Manaças 1958, 1973; Delêtre and Measey 2004; Jones et al. 2006; Cascio 2022; Sousa et al. 2022). Recent research has focused on the trophic ecology of the Tinhosa Grande islet Trachylepis adamastor population (Sousa et al. 2022), and of the Príncipe Island endemic Feylinia polylepis (Cascio 2022). No data is currently available on predation by either native or introduced predators on the amphibians and reptiles of São Tomé and Príncipe. Contrary to birds (Melo et al. 2022), plants (Dauby et al. 2022), and sea turtles (Ferreira-Airaud et al. 2022), data on the habitat association of São Tomé’s amphibians and reptiles is scarce and limited to Hyperolius tree frogs (Strauss et al. 2018). These data gaps preclude a more complete understanding of the ecology of these islands’ ecosystems, its contextualization into broader scenarios, as well as the implementation of data-driven conservation strategies (Bell et al. 2022b; Soares et al. 2022).

Overcoming these knowledge gaps demands additional natural history and ecological studies using a plethora of field techniques, from traditional field surveys to the use of modern techniques such as DNA barcoding and metabarcoding. To contribute to a more accurate and easy identification of the amphibians and terrestrial reptiles of São Tomé and Príncipe, here we provide an illustrated identification key as well as a DNA barcode reference library.

The illustrated keys provided here aim to facilitate a rapid, accurate, and easy identification of the amphibians and reptiles occurring in São Tomé and Príncipe, serving as a baseline for future ecological studies and surveys, as well as conservation actions. Moreover, they will serve as an important support for the work of the forthcoming generations of researchers studying the biodiversity of these islands. In most cases, morphological identification is sufficient to answer the need of researchers, conservationists, and local authorities and constitutes a rapid and inexpensive method. The species occurring in each island are taxonomically diverse, belonging to different families and genera, and even the taxa that belong to the same genus (maybe with the exception of the members of the genus Letheobia in both islands and Schistometopum in São Tomé Island) present several conspicuous morphological characters that allow a rapid identification by even a non-herpetologist in most of the situations.

Notwithstanding, DNA barcodes may play an important role in the identification of juveniles lacking good diagnosable traits, amphibian eggs, and larval individuals, and of species with very cryptic morphological variation (e.g., members of the genus Letheobia in both islands and Schistometopum in São Tomé Island). Moreover, barcodes can be useful to identify poorly preserved and/or partly digested specimens originated from stomach contents or scats of other animals, or to identify animals’ parts and/or animal products being trafficked. However, attention is always needed, and results may sometimes require additional evidence to confidently link a given BIN to a taxon. Single gene methods for species delimitation, such as the use of a single mitochondrial gene as COI, presents some caveats that need to be considered (Dufresne and Jablonski 2022). While BOLD BINs approaches are originally designed for specimen identification, not species delimitation, some abuses and misinterpretations have occurred, leading users to consider BINs as surrogates for taxa (Meier et al. 2021). In our results, most of the BINs were in accordance with the previous taxonomic identification of the respective specimen, but they disagreed in two cases, Hemidactylus longicephalus and Letheobia newtoni, in which the BOLD BIN system provided two different BINs for each taxon. This is mostly explained by the existence of intraspecific diversity within the São Tomé population of these species, which can be a result of some degree of geographic isolation between the sequenced specimens. Subsequent morphological analysis of the barcoded specimens of these two taxa, as well as sequencing of additional mitochondrial and nuclear genes which were run against existing phylogenies of the respective groups (Hedges et al. 2014; Ceríaco et al. 2020b) confirmed that the barcoded specimens represent only two taxa, H. longicephalus and L. newtoni, and no cryptic diversity exists within each taxon. Also, the use of a single mitochondrial gene makes an unambiguous identification impossible in the case of hybrid populations, such as those reported for the species of Hyperolius and Schistometopum on São Tomé Island (Bell et al. 2015; Bell and Irian 2019; O’Connell et al. 2021). For such cases, nuclear markers are needed to confidently assess their identification. When considering the previously existing DNA sequences, we also found that the specimen identified as Schistometopum thomense in Zhang and Wake (2009) groups with our single specimen of Schistometopum ephele sharing the BIN BOLD:AAN0016, both showing a divergence above 3% from all the five specimens of Schistometopum thomense in our dataset. This inconsistency roots in an understandable misidentification by Zhang and Wake (2009), as S. ephele was at that time still considered as a synonym of S. thomense (see O’Connell et al. 2021).

When a solid, complete, and taxonomically well-curated DNA barcode reference library exists, DNA metabarcoding analyses will allow a more detailed and complete glimpse to the understanding of prey patterns in both native and invasive predators (Pompanon et al. 2012; Forin-Wiart et al. 2018; Sousa et al. 2019; Mata et al. 2021; Silva et al. 2021). This is critical for our knowledge and conservation of São Tomé and Príncipe herpetofauna, as the ecological role of the amphibians and reptiles in the local food chain is mostly unknown. This is currently a major information gap because some species may be negatively affected by invasive predators (Bell et al. 2022a; Ceríaco et al. 2022), while others may be feeding on the invasive mammal populations (Ceríaco et al. 2017). DNA barcoding is a relevant method in forensics and to monitor illegal trafficking and has been successfully applied in many regions of the world, for both fauna and flora (Li et al. 2017; Smart et al. 2021; Gostel and Kress 2022). This approach can be of special relevance for the case of the Endangered São Tomé Cobra-Preta, Naja peroescobari, for which reports indicate that certain body parts (fat and meat) are being nationally commercialized and internationally trafficked for their assumed benefits for traditional medicines (Ceríaco et al. 2017, 2022).

More recently, environmental DNA approaches have been employed to contribute to the survey of vertebrate species, including amphibians and reptiles (Ficetola et al. 2019; Buxton et al. 2022; Moss et al. 2022; Nordstrom et al. 2022). These approaches can, theoretically, be faster and less dependent on taxonomic expertise (Ruppert et al. 2019) and have been used to try to document the presence of rare and ecologically cryptic and difficult to observe taxa (Rojahn et al. 2021), as well as invasive taxa (Mahon and Jerde 2016). The effectiveness of environmental DNA approaches to survey amphibians and reptiles in the wild is not generalized across the different taxonomic and functional groups, being much more effective for the case of strictly aquatic amphibians (Buxton et al. 2022; Moss et al. 2022) but tends to be less complete for the case of more terrestrial amphibians and reptiles (Kyle et al. 2022; Nordstrom et al. 2022; pers. obs.). Despite its current caveats, environmental DNA is being perceived as an important tool for future studies on African biodiversity (Heyden 2022), and could act as an important component for surveys in usually logistically difficult areas in the islands of São Tomé and Príncipe, where the traditional surveys are highly impacted by the terrain and climatic harsh conditions.

While DNA barcoding is a powerful and useful tool to answer multiple ecological questions, the traditional taxonomic practice remains the fundamental part of biological research and it is impossible to be substituted by any novel technical approaches (Engel et al. 2022). This work is itself a proof of this, as the assembling of this solid and trustworthy DNA barcoding library was entirely dependent on historical and modern taxonomic works that extensively reviewed the identity of all occurring herpetological taxa (Bell et al. 2022a; Ceríaco et al. 2022) and the collection of specimens subsequently deposited in public accessible natural history collections (Rocha et al. 2014). Being one of the taxonomically most well-known and intensively reviewed herpetofauna of Africa, without many taxonomic uncertainties, a considerable number of available specimens and tissue samples and a relatively modest number of occurring taxa, the herpetofauna of São Tomé and Príncipe presents an ideal case for assembling a complete and trustworthy country-wide DNA barcoding library. For the São Tomé and Príncipe birds, a similar group in terms of stable taxonomy, available specimens and tissue samples and a relatively manageable number of taxa (Melo et al. 2022) are another perfect candidate group for the creation of such a library. For other vertebrate groups such as mammals (Rainho et al. 2022) and fishes (da Costa et al. 2022), the rapid assembly of a complete and trustworthy country-wide DNA barcoding library is made difficult by the considerable number of taxonomic uncertainties plaguing some of its species complexes, too large numbers of occurring taxa, and few readily available samples. This situation is even more striking for the case of invertebrates for which the available taxonomic data and for which more fieldwork, collection of new vouchers, and detailed taxonomic revisions are urgently needed (Bell et al. 2022b; Crews and Esposito 2022; Dijkstra et al. 2022; Mendes and Bivar-de-Sousa 2022; Nève et al. 2022; Panisi et al. 2022). As noted by Edward O. Wilson (2017) and Engel et al. (2022), “more [taxonomists’] boots on the ground” are needed to overcome this current taxonomic impediment and provide the basis for the preservation of the unique biodiversity of these islands.

The present work is the result of multiple initiatives dedicated to the study of the biodiversity of São Tomé and Príncipe. The first three authors have been conducting herpetological studies in the country since 2013 in collaboration with the local authorities, namely the General Office for the Environment from the Ministry of Environment and Natural Resources, the Regional Government of Príncipe, and Príncipe Obo National Park. Fieldwork, specimen collection, and export permits were issued by these authorities, which we acknowledge. A special thanks is owed to José Cassandra, former President of the Regional Government of Príncipe, Daniel Ramos, former director of Príncipe Obo National Park, and Arlindo Ceita Carvalho and Lourenço Monteiro de Jesus, respectively the former and present directors of the General Office for the Environment from the Ministry of Environment and Natural Resources. These activities have mostly been supported by the three first authors’ personal funds, with occasional contributions of Príncipe Trust, Escola Portuguesa de São Tomé & Príncipe, and the Museu de História Natural e da Ciência da Universidade do Porto. Rayna Bell, Lauren Scheinberg, and Erica Ely from the California Academy of Sciences are especially thanked for providing tissue samples from the collections under their care.

Finally, we thank Ostelino Rocha (Balô), Pedro Ceríaco, Sofia Perestrelo, Estrela Matilde, Manuela Costeira, Carlos Gomes (Kalu), Sandra Ferreira, Pedro Lorena, António Almeida, and Suzete Pereira for their assistance, support, and friendship during the fieldwork during the last ten years. The authors would like to thank Aaron M. Bauer, Olivier Pauwels, and a third anonymous reviewer for the comments and suggestions on the original version of this manuscript.