Revisiting the systematics of the Leptodactylus melanonotus group (Anura: Leptodactylidae): Redescription of L. petersii and revalidation of its junior synonyms

Abstract

The Leptodactylus melanonotus species group is widely distributed across the Neotropics and contains 17 valid species. Here we reassess the taxonomic status of several taxa of the group, focusing mainly on the Amazonian L. petersii and both its junior synonyms (L. brevipes and L. intermedius), based on morphological, acoustic, cytogenetic, and DNA sequence data. The phylogenetic analyses recovered a monophyletic L. melanonotus group. However, L. petersii in the broad sense was found to be polyphyletic by containing the nominal species and two other lineages more closely related to L. podicipinus than to L. petersii. Color patterns and calls also support the distinctiveness among these lineages. Moreover, one of the lineages (L. brevipes) has a unique karyotype in the genus (2n = 20 vs. 2n = 22 in the other species). The holotype of L. petersii is presumably lost and a previous designation of a neotype was later considered as invalid. Therefore, we designate a new neotype for L. petersii and redescribe the species from the upper Negro River drainage. Also, we revalidate and characterize L. brevipes and L. intermedius from their type localities. Our study contributes to the systematics, cytogenetic and phenotypic variation, and distribution patterns of leptodactylid frogs in South America.

Introduction

The Neotropics harbors the highest species richness of amphibians on Earth, even though many taxa still remain unknown or unnamed (Vasconcelos et al. 2019). The high rate of species discoveries and taxonomic reassessments in this biogeographical region over the past decades appears to mostly reflect two phenomena: the intensified effort in biodiversity surveys and inventories (Köhler et al. 2005; Fouquet et al. 2007), and the application of a variety of techniques besides morphological data, the so-called integrative taxonomy (Dayrat 2005; Bickford et al. 2007; Pante et al. 2015). Data sharing and collaborative work also have contributed to boost scientific research and to increase the current knowledge on South American herpetofauna (Vasconcelos et al. 2019).

Contemporary and more inclusive approaches have been of high importance in uncovering cryptic diversity and addressing taxonomic problems based on multiple sources of information (e.g. bioacoustics, molecular genetics, and cytogenetics). An integrated framework of taxonomy has been incorporated into more recent studies on leptodactylid frogs of the genus Leptodactylus Fitzinger, 1826 addressing species complexes previously overlooked due to varying degrees of phenotypic conservatism (Heyer and de Sá 2011; Carvalho et al. 2013; Gazoni et al. 2018; Silva et al. 2020; Magalhães et al., 2020).

The genus Leptodactylus comprises 81 valid taxa widely distributed throughout the Neotropics, from southern North America, across Central America, over the Antilles, to southern South America on both sides of the Andes (de Sá et al. 2014;Carvalho et al., 2020Magalhães et al., 2020). Heyer (1969) firstly allocated members of the genus to five species groups (L. fuscus, L. latrans, L. marmoratus, L. melanonotus, and L. pentadactylus), based mainly on morphological, behavioral and ecological traits—one of them, the L. marmoratus group, is currently equivalent to the genus Adenomera Steindachner, 1867 (de Sá et al. 2014). A total-evidence phylogeny of the genus supported the monophyly of the other four Leptodactylus species groups (de Sá et al. 2014). However, the molecular dataset alone in this same study did not recover most of the species groups; neither did it recover a monophyletic Leptodactylus (but see Fouquet et al. 2013 for an alternative hypothesis of Leptodactylus species groups using a smaller taxon sampling). This might have been partly because of the accidental use of contaminated and chimeric sequences or misidentifications, as already pointed by Schneider et al. (2019) and Silva et al. (2020).

The L. melanonotus group contains 17 valid taxa that are mostly distributed over Amazonia. The six lowland Amazonian endemic species (sensu Heyer 1994; present study for L. petersii) are L. diedrus Heyer, 1994, L. discodactylus Boulenger, 1884, L. leptodactyloides, (Andersson 1945), L. petersii, (Steindachner 1864), L. riveroi Heyer & Pyburn, 1983, and L. wagneri, (Peters 1862). Four species occur within Amazonia, but are not endemic to lowland Amazonian forests (sensu Heyer 1994; Jairam & Fouquet 2018): L. griseigularis, (Henle 1981), L. nesiotus Heyer, 1994, L. pascoensis Heyer, 1994, and L. validus Garman, 1888. The remaining seven species are not associated with Amazonia (sensu Heyer 1994; Mijares-Urrutia 1997; Galvis and de Sá 2018; present study for L. podicipinus): L. colombiensis Heyer, 1994, L. magistris Mijares-Urrutia, 1997, L. melanonotus, (Hallowell 1861), L. natalensis Lutz, 1930, L. podicipinus, (Cope 1862), L. pustulatus, (Peters 1870), and L. sabanensis Heyer, 1994.

Some species of the Leptodactylus melanonotus group have widespread distribution in South America east of the Andes, e.g. L. petersii and L. podicipinus. The varying degrees of morphological variation, allied to the complex nomenclatural history and lack of associated acoustic and DNA sequence data hinder accurate identification of populations across their geographic ranges (e.g. Heyer 1994; Fouquet et al. 2012, 2019). Leptodactylus petersii is of particular relevance because of its intricate nomenclatural history. This species was described from the upper Negro River in the Brazilian state of Amazonas, close to the country’s tripoint border with Colombia and Venezuela. The original description was based on a single specimen (the holotype), which was reported as “most likely lost” by Heyer (1970), who designated a Venezuelan specimen as a neotype. In a subsequent review, Heyer (1994) invalidated his previous designation once he noticed that the neotype corresponded to another species (L. validus; reported therein as L. pallidirostris, Lutz 1930). In the same study, he examined hundreds of specimens assigned to L. petersii distributed throughout Amazonia and some localities from the Brazilian Cerrado, and transferred two species into its synonymy: Leptodactylus brevipes Cope, 1887 from the Cerrado of western Brazil and Leptodactylus intermedius Lutz, 1930 from central Brazilian Amazonia.

In this paper, we combine mtDNA genetic diversity, DNA-based phylogenetic analyses, cytogenetic (karyotype and Ag-NOR location) and phenotypic (morphology and calls) data in order to assess species diversity and relationships among members of the L. melanonotus group, based on a comprehensive sampling across South America. Our main focus is to address the taxonomic identity of L. petersii and both its junior synonyms by gathering data from their type localities and from most of their geographic ranges.