Multigene Phylogenetic Analysis Redefines Dung Beetles Relationships and Classification (Coleoptera: Scarabaeidae: Scarabaeinae) Sergei Tarasov* and Dimitar Dimitrov
Total Page:16
File Type:pdf, Size:1020Kb
Tarasov and Dimitrov BMC Evolutionary Biology (2016) 16:257 DOI 10.1186/s12862-016-0822-x RESEARCHARTICLE Open Access Multigene phylogenetic analysis redefines dung beetles relationships and classification (Coleoptera: Scarabaeidae: Scarabaeinae) Sergei Tarasov* and Dimitar Dimitrov Abstract Background: Dung beetles (subfamily Scarabaeinae) are popular model organisms in ecology and developmental biology, and for the last two decades they have experienced a systematics renaissance with the adoption of modern phylogenetic approaches. Within this period 16 key phylogenies and numerous additional studies with limited scope have been published, but higher-level relationships of this pivotal group of beetles remain contentious and current classifications contain many unnatural groupings. The present study provides a robust phylogenetic framework and a revised classification of dung beetles. Results: We assembled the so far largest molecular dataset for dung beetles using sequences of 8 gene regions and 547 terminals including the outgroup taxa. This dataset was analyzed using Bayesian, maximum likelihood and parsimony approaches. In order to test the sensitivity of results to different analytical treatments, we evaluated alternative partitioning schemes based on secondary structure, domains and codon position. We assessed substitution models adequacy using Bayesian framework and used these results to exclude partitions where substitution models did not adequately depict the processes that generated the data. We show that exclusion of partitions that failed the model adequacy evaluation has a potential to improve phylogenetic inference, but efficient implementation of this approach on large datasets is problematic and awaits development of new computationally advanced software. In the class Insecta it is uncommon for the results of molecular phylogenetic analysis to lead to substantial changes in classification. However, the results presented here are congruent with recent morphological studies and support the largest change in dung beetle systematics for the last 50 years. Here we propose the revision of the concepts for the tribes Deltochilini (Canthonini), Dichotomiini and Coprini; additionally, we redefine the tribe Sisyphini. We provide and illustrate synapomorphies and diagnostic characters supporting the new concepts to facilitate diagnosability of the redefined tribes. As a result of the proposed changes a large number of genera previously assigned to these tribes are now left outside the redefined tribes and are treated as incertae sedis. Conclusions: The present study redefines dung beetles classification and gives new insight into their phylogeny. It has broad implications for the systematics as well as for various ecological and evolutionary analyses in dung beetles. Keywords: Dung beetles, Scarabaeinae, Scarabaeidae, Model adequacy, Classification, Molecular phylogeny * Correspondence: [email protected] Department of Research and Collections, Natural History Museum, University of Oslo, P.O. Box 1172, Blindern NO-0318, Oslo, Norway © The Author(s). 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Tarasov and Dimitrov BMC Evolutionary Biology (2016) 16:257 Page 2 of 19 Background resolving shallow divergences but they are less inform- With over 6200 described species [1] and a global distri- ative for recovering higher-level relationships [23] which bution, dung beetles of the subfamily Scarabaeinae calls for assembling larger datasets to improve the ro- (Coleoptera: Scarabaeidae) provide important ecosystem bustness of phylogenetic inference. services [2]. They are one of the primary utilizers of In this paper, we reconstruct the phylogeny of dung mammalian dung on Earth and are historically one of beetles using a molecular dataset that comprises 547 ter- the most recognized and best researched groups of beetles minal taxa and 8 gene regions. This is the largest dung [1, 3–5]. Recently, dung beetles have also become an estab- beetle molecular dataset assembled to date, and includes lished model group in ecology and developmental biology a large quantity of newly sequenced data. In addition, (e.g. [6, 7]). However, a robust classification and phylogen- the present dataset has a global biogeographic coverage etic hypothesis for dung beetles is not available despite and incorporates major phylogenetic lineages and enig- many recent phylogenetic efforts [1, 8–13]. As a result matic taxa. To infer the phylogeny we employed a wide interpretation of their evolutionary, ecological and devel- range of analytical approaches including direct optimization opmental features is often limited to select species and (POY), maximum likelihood (ML) and Bayesian inference large scale comparative analyses are practically impossible. (BI). The traditional substitution models used in model- The currently accepted classification splits the based methods (ML and BI) have been frequently Scarabaeinae dung beetles into 12 tribes which, over shown to poorly reflect the reality of the evolutionary the last two decades, have been the subject of 16 molecu- process [24, 25]; thus, their application can be inad- lar- and morphology-based phylogenetic studies [1, 8–22]. equate for some molecular datasets. In this study, we These studies were reviewed in detail by [1, 7]. explicitly test for model adequacy using Bayesian pos- The results of most of these studies can be character- terior assessment [25, 26] and perform partition selec- ized by three common trends. 1) They resolve early tion based on the adequacy of the selected models. branching clades or shallow nodes well, but intermediate Although data selection guided by Bayesian posterior nodes remain poorly resolved and/or weakly supported. assessment allows inferring some meaningful relation- 2) Seven tribes are always recovered as monophyletic or shipsabsentindatasetswhereitwasnotused,the nearly monophyletic (e.g. Onthophagini, Oniticellini), results of both were generally similar. The efficient while three tribes (Deltochilini, Ateuchini, and Coprini) application of data selection using model adequacy as- are always polyphyletic. The polyphyletic tribes Deltochilini sessment to large datasets, as the one used herein, is (=Canthonini) and Ateuchini (=Dichotomiini) together presently difficult due to the lack of computationally comprise ca. 55% of the total generic diversity in this advanced software. We conclude that the development group. Their highly polyphyletic concepts render the tribal of such software can, in future, boost progress of Bayesian classification in the entire subfamily extremely artificial. 3) posterior assessment methods in phylogenetics. The results of these key studies often propose conflicting Our results identify new lineages and corroborate some hypotheses [1] leading to a lack of consensus on dung bee- relationships inferred by earlier studies [1, 10–13, 19, 21]. tle evolutionary history. The consistency of clades between the molecular phyl- One morphological [1] and two molecular phylogenies ogeny presented here and the most recent morphology [11, 22] can be singled out due to their large taxon sam- based analyses [1] enables us to define new systematic ple size and global biogeographic coverage; the rest of concepts for the highly polyphyletic tribes Dichotomiini, the studies are usually limited in these respects. The glo- Deltochilini and Coprini. Over the last half-century the bal morphological phylogeny of [1] comprises all main concepts of these tribes have been constantly changing biogeographic and taxonomic lineages and provides an because clear synapomorphies which could ensure their integrative pattern of phylogenetic relationships in dung unequivocal identification have always been missing. beetles largely supported by previous publications. How- Given the principle of monophyly, we limit these tribes ever, that study also stresses the need for more data, substantially to accommodate only those genera which are primarily molecular, to corroborate its findings. closely related to their respective type genera. We use the The two available global molecular phylogenies [11] synapomorphies identified by the global morphological and [22] are similar in composition of genetic markers phylogeny of [1] to provide an effective identification of (COI, 16S, 28S and COI, 16S, 28S, 12S respectively) as these tribes within their new definitions. Many genera well as species used. mtDNA markers are known to be hitherto considered members of these tribes are now ex- saturated by fast evolution and not very informative cluded from them. We treat those genera as incertae sedis about relationships above the species level, while the and discuss the necessary steps towards their phylogeny- 16S, 28S and 12S rDNA markers are challenging to align based classification. We also expand the concept of the and analyze with traditional substitution models. These tribe Sisyphini by adding the genus Epirinus that was pre- mitochondrial and rDNA genes are