\u2018Barcode fishing\u2019 for archival DNA from historical type material overcomes taxonomic hurdles, enabling the description of a new frog species

Mark D Scherz,Jörn Köhler,Frank Glaw,Miguel Vences,Safidy M Rasolonjatovo,Loïs Rancilhac,Michaela Preick,Andolalao Rakotoarison,Michael Hofreiter,Annemarie Ohler,Achille P Raselimanana

doi:10.1038/s41598-020-75431-9

Abstract

Taxonomic progress is often hindered by intrinsic factors, such as morphologically cryptic species that require a broad suite of methods to distinguish, and extrinsic factors, such as uncertainties in the allocation of scientific names to species. These uncertainties can be due to a wide variety of factors, including old and poorly preserved type specimens (which contain only heavily degraded DNA or have lost important diagnostic characters), inappropriately chosen type specimens (e.g. juveniles without diagnostic characters) or poorly documented type specimens (with unprecise, incorrect, or missing locality data). Thanks to modern sequencing technologies it is now possible to overcome many such extrinsic factors by sequencing DNA from name-bearing type specimens of uncertain assignment and assigning these to known genetic lineages. Here, we apply this approach to frogs of the Mantidactylus ambreensis complex, which was recently shown to consist of two genetic lineages supported by concordant differentiation in mitochondrial and nuclear genes. These lineages co-occur on the Montagne dʼAmbre Massif in northern Madagascar but appear to have diverged in allopatry. We use a recently published bait set based on three mitochondrial markers from all known Malagasy frog lineages to capture DNA sequences from the 127-year-old holotype of Mantidactylus ambreensis Mocquard, 1895. With the obtained sequences we are able to assign the name M. ambreensis to the lowland lineage, which is rather widespread in the rainforests of northern Madagascar, leaving the microendemic high-elevation lineage on Montagne d’Ambre in north Madagascar in need of description. We describe this species as Mantidactylus ambony sp. nov., differing from M. ambreensis in call parameters and a smaller body size. Thus, using target enrichment to obtain DNA sequence data from this old specimen, we were able to resolve the extrinsic (nomenclatural) hindrances to taxonomic resolution of this complex. We discuss the broad-scale versatility of this ‘barcode fishing’ approach, which can draw on the enormous success of global DNA barcoding initiatives to quickly and efficiently assign type specimens to lineages.

Highlights

A substantial, albeit inestimable, portion of the potentially over 100 million eukaryote species awaiting taxonomic description[1] cannot be described due to some kind of intrinsic or extrinsic hindrance to taxonomy
After selecting the reads matching a library of reference sequences with a similarity threshold of 90%, a total of 17,847 and 290,739 reads were left for alignment, respectively
For the cox[1] gene, 17,828 reads were aligned to the reference sequence, with a total of 214 nucleotides recovered for the total alignment length of 442 nt, with 89 nt missing at the beginning, a larger stretch of 126 missing nucleotides in the middle, and 10 nt missing at the end of the alignment; coverage for cox 1 was > 800 for the initial part, and only 4–50 for the last parts of the reconstructed sequence

Summary

Introduction

A substantial, albeit inestimable, portion of the potentially over 100 million eukaryote species awaiting taxonomic description[1] cannot be described due to some kind of intrinsic or extrinsic hindrance to taxonomy. Some such cases can be solved by narrowing down the collection locality of the specimens, identification of the current topotypical species in that area, and careful morphological examination to assign the name to one of the possibilities This can take weeks of thorough work, can still result in an answer that is not wholly unambiguous, and is made considerably more challenging when the collected specimens were not accompanied with precise or accurate locality data[3,4,5]. We designed a set of baits that targets three mitochondrial markers (16S rRNA [16S], cytochrome oxidase I [cox1], and cytochrome b [cyt-b]) from a pooled set of Malagasy frog species[13] This method allowed us to clarify the identity of several old names in the genus Mantidactylus Boulenger, 189514 from Madagascar

Objectives

Methods

Results

Conclusion