Systematic Entomology (2020), DOI: 10.1111/syen.12423 Global domination by crazy ants: phylogenomics reveals biogeographical history and invasive species relationships in the genus Nylanderia (Hymenoptera: Formicidae) JASON L. WILLIAMS1 ,YUANMENGMILESZHANG1 , MICHAEL W. LLOYD2,3 ,JOHNS.LAPOLLA4, TED R. SCHULTZ2 and A N D R E A L U C K Y 1 1Entomology & Nematology Department, University of Florida, Gainesville, FL, U.S.A., 2Department of Entomology, Smithsonian Institution, Washington, DC, U.S.A., 3Laboratories of Analytical Biology, Smithsonian Institution, Washington, DC, U.S.A. and 4Department of Biological Sciences, Towson University, Towson, MD, U.S.A. Abstract. Nylanderia (Emery) is one of the world’s most diverse ant genera, with 123 described species worldwide and hundreds more undescribed. Fifteen globetrotting or invasive species have widespread distributions and are often encountered outside their native ranges. A molecular approach to understanding the evolutionary history and to revision of Nylanderia taxonomy is needed because historical efforts based on morphol- ogy have proven insuffcient to defne major lineages and delimit species boundaries, especially where adventive species are concerned. To address these problems, we gener- ated the frst genus-wide genomic dataset of Nylanderia using ultraconserved elements (UCEs) to resolve the phylogeny of major lineages, determine the age and origin of the genus, and describe global biogeographical patterns. Sampling from seven biogeo- graphical regions revealed a Southeast Asian origin of Nylanderia in the mid-Eocene and four distinct biogeographical clades in the Nearctic, the Neotropics, the Afrotrop- ics/Malagasy region, and Australasia. The Nearctic and Neotropical clades are dis- tantly related, indicating two separate dispersal events to the Americas between the late Oligocene and early Miocene. We also addressed the problem of misidentifca- tion that has characterized species-level taxonomy in Nylanderia as a result of limited morphological variation in the worker caste by evaluating the integrity of species bound- aries in six of the most widespread Nylanderia species. We sampled across ranges of species in the N. bourbonica complex (N. bourbonica (Forel) + N. vaga (Forel)), the N. fulva complex (N. fulva (Mayr) + N. pubens (Forel)), and the N. guatemalensis com- plex (N. guatemalensis (Forel) + N. steinheili (Forel)) to clarify their phylogenetic place- ment. Deep splits within these complexes suggest that some species names – specifcally N. bourbonica and N. guatemalensis –eachareappliedtomultiplecrypticspecies.In exhaustively sampling Nylanderia diversity in the West Indies, a ‘hot spot’ for invasive taxa, we found fve adventive species among 22 in the region; many remain morpho- logically indistinguishable from one another, despite being distantly related. We stress that overcoming the taxonomic impediment through the use of molecular phylogeny and revisionary study is essential for conservation and invasive species management. Correspondence: Andrea Lucky, Entomology & Nematology Department, University of Florida, 1881 Natural Area Dr., Gainesville, FL, 32611-0620, U.S.A. E-mail: [email protected] © 2020 The Royal Entomological Society 1 2 J. L. Williams et al. Introduction native species and rapidly identifed before they can expand their ranges further. The taxonomic impediment ‘The “habitus” of a species, as every taxonomist knows, is The ant genus Nylanderia something one may take in at a glance, but be quite unable to express without wearisome prolixity.’ Nylanderia and six closely related genera (collectively called –WilliamMortonWheeler(1910). the ‘Prenolepis genus-group’) have a tumultuous taxonomic The term ‘taxonomic impediment’ (Taylor, 1983) refers to history that has only recently reached some level of stability the ongoing challenge of naming and characterizing vast num- (reviewed in LaPolla et al., 2010; see also Lapolla et al., 2012). bers of species, namely those with inconspicuous diversity LaPolla et al. (2010) found support in multigene sequence owing to broad distributions or limited morphological variation. data for the monophyly of Nylanderia.Additionalmorpho- Although ants are unusually well-characterized among insects, logical characters distinguishing Nylanderia from other genera with over 13 000 species formally described, more than 7000 were discovered in a global taxonomic revision of the genus ant species are estimated to remain undescribed (Hölldobler Prenolepis Say (Williams & LaPolla, 2016), and recent molec- & Wilson, 1990; AntWeb, 2019). More than a quarter of the ular phylogenetic and phylogenomic studies have continued to ants on Earth lack formal descriptions because, in some cases, support the monophyly of Nylanderia (Blaimer et al.,2015; particular groups remain unstudied, whereas other groups are Matos-Maraví et al.,2018).Emery(1906)originallydescribed impeded by morphological crypsis; historical taxonomic con- Nylanderia over a century ago and a genus-wide taxonomic revi- fusion can compound both of these problems. Many species sion is needed because taxonomic confusion frequently leads to descriptions are more than a century old, are based only on the misidentifed or unnamed specimens. Regional revisions have worker caste, and have not been revisited in higher-level revi- included taxonomic treatment of species in some areas where sionary study. Surprisingly, one of the most species-rich ant gen- Nylanderia species richness is relatively low, including the era, Nylanderia (Emery, 1906) – which currently includes 123 Afrotropics (LaPolla et al., 2011), the Nearctic (Kallal, 2012) extant species and 25 subspecies worldwide (Bolton, 2019) – is and the West Indies (LaPolla & Kallal, 2019). Taxonomic revi- severely underdescribed, with an estimate of hundreds more sion is still needed for Nylanderia in regions where the genus species awaiting description (LaPolla et al., 2011). is most diverse: Mesoamerica, South America, Southeast Asia Phylogeny, taxonomy and species-level identifcation have and Australasia. These regions are major centres of Nylanderia long been characterized by confusion in this genus as a result species diversity, and Southeast Asia likely represents the bio- of widespread morphological convergence and a high number geographical origin of the genus (Matos-Maraví et al., 2018). In of widely distributed adventive species. As in many genera order to stabilize the chaotic taxonomy of this genus, nomencla- that are species-rich and hard to identify, obstacles to resolving ture needs to refect actual species boundaries and relationships the systematics and clarifying the nomenclature of this group among lineages. For this, a phylogeny with representative sam- include: pling from all major lineages worldwide is needed. Nylanderia is a near-globally distributed genus, but it is not 1. High species richness and abundance world- clear whether regional faunas represent monophyletic lineages wide – Nylanderia is one of the most commonly collected and it is impossible to understand how geography has shaped ant genera worldwide and includes hundreds of species them without frst knowing the major clades. LaPolla et al. (Ward, 2000; LaPolla et al., 2011). (2010) suggested that Nylanderia includes fve biogeographi- 2. Limited morphological variation – there are few eas- cally distinct clades: (i) Nearctic; (ii) Neotropics; (iii) Afrotrop- ily discretizable worker morphological characters between ics; (iv) Indomalaya; and (v) Australasia, but this study was species, and variation within a species often overlaps con- focused specifcally on genus-level relationships and only sam- siderably with that of other, similar species (Trager, 1984). pled c. 24 of the estimated hundreds of Nylanderia species. The 3. Globetrotting species – at least 15 species have been trans- hypothesis that there are fve major lineages corresponding to ported across the globe by humans and are easily mistaken these regions must be tested in order to advance global revision- for native species (Williams and Lucky, in press). ary study. Without a grasp on the global phylogeny it is diffcult to place diagnostic morphological characters in a global context. These challenges have long confounded attempts to clarify Taxonomic revisions and phylogenies over the past decade the biology and natural history of the genus at the species have brought a measure of much-needed clarity to the system- level. Recognition of the need to manage emerging invasive atics of Nylanderia at the regional level, but more questions threats such as the tawny crazy ant, N. fulva (Mayr), has added than answers persist about the origins and identities of some new urgency to the need for establishing a solid taxonomic of the most widespread globetrotting species in the genus, such foundation in this lineage (Gotzek et al., 2012). Nylanderia fulva as: (i) how many times ‘invasiveness’ has arisen in Nylande- is just one of several taxonomically diffcult species that are ria, and to which major clade or clades these invasive lineages accidentally transported along human trade routes. Resolving belong; (ii) whether the globetrotting species are monophyletic; the systematics and clarifying the nomenclature of the genus is (iii) whether species boundaries are reasonably delimited or if necessary so that non-native species can be distinguished from some species have multiple names applied to them based on © 2020 The Royal Entomological Society, Systematic Entomology,doi:10.1111/syen.12423