Biodiversity inventory of the grey mullets (Actinopterygii: Mugilidae) of the Indo-Australian Archipelago through the iterative use of DNA-based species delimitation and specimen assignment methods.

Erwan Delrieu‐Trottin,Sopian Sauri,Hadi Dahruddin,Mochamad Syamsul Arifin Zein,Daisy Wowor,Jean‐Dominique Durand,Robert Hanner,Kadarusman Kadarusman,Hagi Yulia Sugeha,Tedjo Sukmono,Nicolas Hubert,Laurent Pouyaud,Philippe Keith,Wei‐Jen Chen,Yuli Fitriana,Dirk Steinke,Philippe Borsa,Régis Hocdé ,F Busson ,Gino V Limmon

doi:10.1111/eva.12926

Abstract

DNA barcoding opens new perspectives on the way we document biodiversity. Initially proposed to circumvent the limits of morphological characters to assign unknown individuals to known species, DNA barcoding has been used in a wide array of studies where collecting species identity constitutes a crucial step. The assignment of unknowns to knowns assumes that species are already well identified and delineated, making the assignment performed reliable. Here, we used DNA‐based species delimitation and specimen assignment methods iteratively to tackle the inventory of the Indo‐Australian Archipelago grey mullets, a notorious case of taxonomic complexity that requires DNA‐based identification methods considering that traditional morphological identifications are usually not repeatable and sequence mislabeling is common in international sequence repositories. We first revisited a DNA barcode reference library available at the global scale for Mugilidae through different DNA‐based species delimitation methods to produce a robust consensus scheme of species delineation. We then used this curated library to assign unknown specimens collected throughout the Indo‐Australian Archipelago to known species. A second iteration of OTU delimitation and specimen assignment was then performed. We show the benefits of using species delimitation and specimen assignment methods iteratively to improve the accuracy of specimen identification and propose a workflow to do so.

Highlights

The first round of species delimitation methods using the DNA barcode reference library of Durand et al (2017) composed of 827 sequences and 102 nominal species yielded a varying number of operational taxonomic unit (OTU) according to the methods with 105 using Refined Single Linkage (RESL), 148 using Automatic Barcode Gap Discovery (ABGD), 70 using mPTP, and 120 using mGMYC (Figure 2 and Table S3) and resulting in a consensus consisting of 113 OTUs (Figure 2)
The first iteration of species delimitation applied to the reference library revealed the presence of more than 10% cryptic diversity because 113 OTUs were extracted from 102 known taxa
DNA barcoding has prompted the development of a wide range of genomic tools

Summary

| MATERIALS AND METHODS

The baseline reference library used in this study originates from Durand et al (2017). The first round of species delimitation methods using the DNA barcode reference library of Durand et al (2017) composed of 827 sequences and 102 nominal species yielded a varying number of OTUs according to the methods with 105 using RESL, 148 using ABGD, 70 using mPTP, and 120 using mGMYC (Figure 2 and Table S3) and resulting in a consensus consisting of 113 OTUs (Figure 2). The remaining 38 cases represented not yet delineated new OTUs. The ratio between the K2P genetic distances of the selected OTUs and distance to the nearest neighbor was smaller for the three methods in the second round (Figure 6). B), and OTU 121 (Planiliza sp.)), three in New Guinea (OTU 118 (Osteomugil sp.), OTU 48 (P. labiosus), and OTU 27 (Planiliza sp.)), two in Kalimantan

Findings

| DISCUSSION

| CONCLUSION