A comprehensive appraisal of evolutionary diversity in venomous Asian coralsnakes of the genus Sinomicrurus (Serpentes: Elapidae) using Bayesian coalescent inference and supervised machine learning. (24th November 2021)

• Record Type: Journal Article
• Authors: Smart, Utpal; Ingrasci, Matthew J.; Sarker, Goutam C.; Lalremsanga, Hmartlawmte; Murphy, Robert W.; Ota, Hidetoshi; Tu, Ming Chung; Shouche, Yogesh; Orlov, Nikolai L.; Smith, Eric N.
• Is Part Of: Journal of zoological systematics and evolutionary research. Volume 59:Number 8(2021)
• Page Start: 2212
• Page End: 2277
• Publication Date: 2021-11-24
• Subjects: Asian coralsnakes -- Bayesian species delimitation -- comparative morphology -- Elapidae -- integrative taxonomy -- machine learning -- random forest -- Sinomicrurus; Animals -- Classification -- Periodicals; Zoology -- Periodicals; Evolution -- Periodicals; 578.012
• DOI: 10.1111/jzs.12547
• Languages: English
• ISSNs: 0947-5745

Abstract

Abstract: While species boundaries between conspicuously divergent populations of the medically important snake genus Sinomicrurus are well established, instances of erratic chromatic and meristic variation continue to confound taxonomists, since the mid‐1800s. This predicament can be attributed to an inadequate molecular phylogenetic framework and the lack of a comprehensive taxonomic representation of geographic variants. Here, we revisit lineage delineation in this genus, drawing cohesive evidence from a plurality of data and analysis types, including a promising, yet taxonomically under‐utilized, supervised machine learning algorithm (random forest). Overall, this study incorporates data generated by molecular analyses as well as morphometrics and comparative anatomy based on 236 specimens from 28 different natural history collections examined by us, and an additional 161 records from 47 other sources. Our results indicate several very divergent evolutionary lineages concealed as subspecies. Thus, to better reflect this phylogenetic diversity, we raise S . macclellandi iwasakii from the southern Ryukyus and S . m . swinhoei from Taiwan to full species, and resurrect S . annularis . We highlight the need to distinguish at species level the current subspecies of S . japonicus, namely as, S . japonicus and S . boettgeri, and provide diagnostic characters to that end. On the other hand, given unpersuasive support of lineage independence, we sink Taiwanese S . hatori into S . sauteri and S . nigriventer into S . macclellandi . We also meticulously redescribe S . peinani from mainland China and Vietnam based on a substantial number of additional specimens, while synonymizing the recently described S . houi under S . kelloggi . We conclude with a discussion on the role of regional biogeography as a primary driver of cladogenesis in the genus. Abstract : Systematics and species boundaries of the elapid snake genus Sinomicrurus were investigated using mitochondrial and nuclear markers. We utilized a Bayesian framework, reinforced by comparative morpho‐anatomical analyses leveraging comprehensive geographic representation, to diagnose and delimit species. Our morphometric evaluations contrasted classical discriminant function analyses and random forest, a tree‐based supervised machine learning technique. We conclude that previous taxonomic schemes often confounded between species‐level and population‐level variation, and we update the taxonomy by recognizing a total of nine species within Sinomicrurus . 摘要: 雖然醫學上廣受重視的 Sinomicrurus 屬蛇類群其物種之間的界限已因其明顯不同的外形而確立, 但自 1800 年代中期以來, 這類群所表現出的不穩定體色和測量形質上的變異持續困惑著分類學者。這種困境可歸因於不夠完整的分子親緣譜系框架和缺乏全面性的地理變異分類學表徵資訊。本研究重新審視該屬的親緣譜系劃分, 並從多項數據和分析類型進行聚類分析, 包括一種具有研究應用潛力但在分類學應用上卻被低估的机器学习演算法(random forest)。我們總共檢視了來自28 個自然史博物館所收藏的 236件標本, 以及其他來源的161件樣本, 結合分子遺傳變異、形態測量和比較解剖學所取得的資料並進行分析, 結果顯示 Sinomicrurus 屬的物種多樣性被嚴重低估, 幾個具有明顯分化的演化親緣譜系被評估為亞種。為了妥切反映這類群在系統發育上的多樣性, 我們將琉球南部的 S . macclellandi iwasakii 和台灣的 S. m . swinhoei 兩個亞種提升到種的階層, 並再確認 S . annularis 的種的有效性。我們也認為 S. japonicus 當前的亞種需要在種的階層上進行区别, 即應提升為 S. japonicus 和 S . boettgeri, 並提供此兩個物種的檢索資訊。另一方面, 由於缺乏種的階層上的足夠證據支持, 我們將台灣的 S . hatori 併入 S . sauteri, 也將 S . nigriventer 併入了 S . macclellandi 。本研究也依據大量的標本測量資訊重新描述了中國大陸和越南的 S . peinani, 並認定最近發表的 S . houi 應為 S . kelloggi 的同物異名。本研究结语歸納并讨论了區域性生物地理分化運作為 Sinomicrurus 屬蛇類物種分化的主要驅動因素。