2017/10/16 by Andrew E. Z. Short, ANDREW EDWARD Z. SHORT · 6 citations
Agricultural and Biological Sciences · Environmental Science · #Coleoptera Taxonomy and Distribution #Freshwater macroinvertebrate diversity and ecology #Forest Ecology and Biodiversity Studies
paper · pdf · doi:10.1111/syen.12270
With more than 13 000 described species (Fig. 1, Table 1), water beetles are one of the most globally abundant groups of aquatic insects. Among insect orders, only Diptera has more aquatic taxa (though just as larvae) than Coleoptera, and the two largest water beetle families, Dytiscidae and Hydrophilidae, each have more species than either Ephemeroptera or Plecoptera. This rich diversity is not the result of a single aquatic invasion, but rather of repeated macroecological shifts from terrestrial habitats throughout more than 300 Ma of evolutionary history (Toussaint et al., 2017b). Consequently, water beetles do not form a single clade but are better described as an ecological guild distributed across at least 30 families in three of the four coleopteran suborders. Water beetles have been a popular study group among professional and amateur entomologists alike, even boasting their own international society, the UK-based Balfour-Browne Club. Because of their ecological sympatry, for centuries water beetle specialists have often collected and/or studied multiple families in this guild rather than limiting themselves to one particular lineage. The reasons for the relative popularity of aquatic beetles include: (i) their sheer abundance and broad distribution – they are easy to collect almost anywhere; (ii) their occasionally large size, with some adult species exceeding 5 cm in length, makes some species highly visible and some of these are even kept as pets; (iii) their diversity and often beautiful colour patterns, particularly evident in some diving beetles; and (iv) the breadth of their ecologies and behaviours – from incompressible plastrons in Elmidae to repeated aquatic/terrestrial habitat shifts in Hydrophilidae – ensure water beetles are tractable for a range of biological questions. Here, I attempt to summarize the current state-of-the-art of water beetle systematics, including a synthesis of recent literature on how well water beetle taxa are described, the degree to which we understand their evolutionary relationships, and in what ways they are being used as model systems in systematics. I also provide a ‘horizon scan’ of where water beetle systematics is going, and how its prominence and role in shaping our understanding of evolutionary biology will continue to increase. Due to being defined by ecology rather than a single clade, which groups fall under the aquatic beetle umbrella differ slightly depending on the specialist. For simplicity and the purposes of this review, I limit discussion to families within the five major groups in which the vast majority of species have adults and/or larvae occupying aquatic habitats: Myxophaga, ‘Hydradephaga’, Hydraenidae, Hydrophiloidea and the core aquatic Byrrhoidea (Dryopidae, Psephenidae, Lutrochidae, Elmidae). I have omitted a few groups that some workers might generally consider aquatic, in particular the Limnichidae and Scirtidae. Many additional terrestrial beetle families have taxa with significant aquatic modifications (e.g. Curculionidae, Chrysomelidae and Staphylinidae among others), but these are also not considered here. Similarly, a number of attempts have been made to partition water beetles into various categories based on the particulars of their life history (e.g. aquatic adults vs. aquatic larvae vs. aquatic adults and larvae). For a more detailed summary of how various water beetle groups are defined, and for a limited review of aquatic taxa in otherwise dominant terrestrial families, see Jäch (1998) and Jäch & Balke (2008). The taxonomy of aquatic beetles has long benefited from a dedicated and relatively large community of both professional and amateur entomologists. Consequently, although there is no question that thousands of new species of aquatic beetles remain to be discovered or described, they are among the better-known and best-catalogued groups of beetles. There are modern world catalogues for a large majority of aquatic beetle families, many of which are updated regularly (Table 1), notable exceptions being Gyrinidae and the smaller aquatic byrrhoid groups (Psephenidae, Dryopidae). Estimating the number of described species is a frequent thought experiment for taxonomists. In one recent attempt to estimate the actual diversity, Jäch & Balke (2008) suggested that water beetles were approximately 70% described (although their circumscription of water beetles differs slightly from that of this review). Bloom et al. (2014) used the taxonomic expertise of the authors to estimate the actual diversity of each tribe of Hydrophilidae, coming to a rough calculation of 4183 species for the family; coincidentally also equating to about 70% of the currently described diversity. Using taxonomic revision data and controlling for several variables such as body size and geography, Nilsson-Örtman & Nilsson (2010) used modelling methods to estimate the actual number of species of diving beetles to be 5405, which would mean that Dytiscidae is presently ∼80% described (Nilsson, 2016). Meanwhile, some smaller families such as Epimetopidae and Lutrochidae have more than doubled in size in the last few years alone (Perkins, 2012; Maier & Short, 2014) and it is likely that they remain substantially underdescribed. The families treated in this review have an approximate total diversity of ∼13 000 described species. Assuming that this represents in aggregate between 65 and 75% of the actual diversity, there are roughly between 17 000 and 20 000 water beetle species. This large amount of remaining unknown water beetle biodiversity is, however, far from evenly distributed among biogeographical regions. The Nearctic, accounting for just ∼8% of water beetle diversity (Jäch & Balke, 2008), is both the best known and also the most species-poor region. Although new species are still to be found, they are often either isolated/unusual endemics (e.g. stygobionts; Miller et al., 2009) or ‘known unknowns’ from previously unpublished theses (e.g. Hydrochus falsus Hellman in Worthington et al., 2016). At the opposite end of the spectrum is the Neotropics, particularly tropical South America; the region is both the most species-rich (Jäch & Balke, 2008) and likely the most undescribed. More than 300 new species have been described from the Neotropics in the last 10 years alone. The Palaearctic already is relatively high in species richness but much better described than the tropical regions. The Afrotropics and particularly the Oriental region are likely to have many new species awaiting discovery. I proffer that no group of Coleoptera has received as robust and sustained study into their phylogenetic relationships as water beetles. In the last quarter of a century, more than 200 studies have generated new hypotheses of relationships based on a wide range of morphological and molecular data (Table 2). Far from the taxonomic chaos that renders some beetle groups unapproachable, the higher-level classifications for most major lineages of water beetles have been established and tested with multiple lines of evidence. Few phylogenetic studies have focused on Myxophaga. Phylogenies examining interfamilial relationships using adult (Beutel, 1999) and larval characters (Beutel et al., 1999, but excluding Lepiceridae) have found a generally consistent branching pattern of (Lepiceridae + (Torridincollidae + (Hydroscaphidae + Sphaeriusiidae))). With the exception of a recent comprehensive molecular phylogeny of Hydroscaphidae (Short et al., 2015), no phylogenies have been published that substantively examine relationships within any family of Myxophaga. Various taxa of the suborder have been the subject of detailed morphological studies (e.g. Beutel, 1998; Anton & Beutel, 2006) that provide additional discussion of phylogenetic affinities of the group examined. The branching pattern of the hydradephagan families has received significant attention, although no strong consensus on interfamilial relationships has yet emerged. Most morphology-based studies in the last 10 years suggest Gryinidae is the earliest diverging adephagan lineage and would render Hydradephaga nonmonophyletic. Various molecular phylogenies ranging from single-gene studies using 18S rDNA to more comprehensive multigene analyses have found support for a monophyletic Hydradephaga (e.g. Shull et al., 2001; Ribera et al., 2002; McKenna et al., 2015, but see Maddison et al., 2009). Using mitochondrial genomes, López-López & Vogler (2017) did recover a monophyletic Hydradephaga, but with weak support. However, in the first use of phylogenomic analysis in beetles using Ultraconserved Elements (UCE), Baca et al. (2017a) recovered a paraphyletic Hydradephaga. Beutel et al. (2008), Maddison et al. (2009) and Baca et al. (2017b) provided more detailed reviews on hypotheses of relationships among adephagan families. Most species-rich families of Hydradephaga have been the subject of phylogenetic studies in the last decade. Miller & Bergsten (2012) recently published a detailed total-evidence phylogeny of Gyrinidae based on morphology and five genes, and revised the classification of the family accordingly. Previous classifications of gyrinids were based principally on morphological works by Beutel (1989a,b, 1990) and Beutel & Roughley (1994). Still, additional recent studies have continued to revise this classification and illuminate our knowledge of whirligig beetle relationships (e.g. Gustafson & Miller, 2017). The phylogeny and classification of Noteridae was revised by Miller (2009) based on a cladistic analysis of morphological data. However, an expanded dataset (in both taxa and characters) using five genes negated that classification and supplanted it with a new one (Baca et al., 2017a). No molecular data has been applied to the evolution of Haliplidae, and the phylogenetic validity of some smaller genera is uncertain; phylogenies based on morphological data have been inferred for the family (Beutel & Ruhnau, 1990) as well as the genus Brychius (Mousseau & Roughley, 2007). By far the largest family, Dytiscidae has been subject to more phylogenetic studies than any other water beetle group (Table 2). Although there have been numerous and persistent efforts to resolve relationships within various genera, tribes, and subfamilies, the family as a whole has been surprisingly little studied. Using a combination of adult and immature characters and a single gene, Miller (2001a, 2003) made the first attempt to infer a family-wide phylogeny. Subsequently, Ribera et al. (2008) provided the first substantial multi-gene molecular phylogenetic estimate for the family. However, many of these other early larger-scale studies were confounded by incomplete sampling of major lineages or poor resolution/low support along the backbone of the tree. Miller & Bergsten (2014) provided a major attempt at a family-wide classification using a thoroughly sampled, six-gene dataset combined with adult morphology, which remains the most complete and robust estimate to date. Michat et al. (2017) provide an independent estimate of dytiscid phylogeny based solely on larval characters. The larval phylogeny largely agrees with Miller & Bergsten (2014) with regard to the monophyly of most tribes and subfamilies, though there are exceptions (e.g. Dytiscinae, Agabini, Hydroporini), and the relationships among subfamilies differs substantially (e.g. Laccophinae being a highly nested, late diverging lineage in Miller & Bergson but sister to the remaining Dytiscidae in Michat et al., 2017). Relationships among the families of Hydrophiloidea have been the subject of more than a dozen studies over the last 25 years (Table 1). Although a definitive resolution has yet to be reached, the monophyly of Hydrophiloidea, and that of its six constitute families is not in dispute. Short & Fikáček (2013) recently revised the classification of Hydrophilidae based on an analysis of six genes. Among the most significant changes were that the genera Horelophus and Horelophopis, previously presumed to be primitive early diverging lineages and considered their own subfamilies, actually were highly derived taxa nested within other tribes (the inclusion of the latter was supported independently by a concurrent morphological study (Minoshima et al., 2013)). Of the five smaller families, only the internal relationships of Helophoridae have been examined (Fikáček et al., 2012a,b). In a large two-gene study, McKenna et al. (2015) place the Hydrophiloidea as well as Hydraenidae in the context of staphyliniform evolution, being sister to Histeroidea and within Staphylinoidea respectively. The family Hydraenidae has never been the subject of a comprehensive cladistic analysis with the current classification (see Hansen, 1998) based largely on detailed morphological studies by Perkins (1980, 1997). The megadiverse genus Hydraena, with nearly 1000 described species, has been the subject of several recent molecular phylogenies to sort out this explosive diversity (Trizzino et al., 2011, 2013). The relationships between, and within, the families of aquatic Byrrhoidea are the least understood among water beetle lineages. How the families are positioned within Elateriformia, and even if they are themselves reciprocally monophyletic, remains unclear. In the most robust study to date, Kundrata et al. (2016) inferred the phylogeny of Byrrhoidea using four genes. In their analyses, neither Psephenidae nor the subfamilies of Elmidae (Elminae and Larainae) were monophyletic. However, their taxon sampling within each family was limited, and many relationships, especially those among families, were not supported statistically. No comprehensive phylogeny for Dryopidae or Elmidae has yet been undertaken. The phylogeny of Psephenidae was inferred based on adult morphology (Lee et al., 2007), and a new classification proposed. Recently, Jäch et al. (2016) elevated Protelmini (a small group of about six described species from the Afrotropics and Neotropics) from a tribe of Elmidae to its own family, Protelmidae. In proposing this new family, however, no phylogeny or other evidence was offered to support this change. Because there are not yet any reviewable data for the justification of this new family, I have treated it as part of Elmidae for the purpose of this review (similarly, a figure of an elmid phylogeny depicted in Kodada et al. (2016) is based on unpublished studies for which the data, taxon sampling, and clade support cannot be evaluated). Aquatic beetles have a rich fossil record due to strongly sclerotized bodies and predilection for habitats such as and that In to on evolution, the robust fossil record for of phylogenies and aquatic beetles as a model in evolutionary Although are to the are not (e.g. 2009). The last in particular has a of new to and – of – and of (e.g. in Hydrophiloidea, Fikáček et al., see also & 2015, for a review). The majority of aquatic beetle families are known in the fossil the and suborder is in several (e.g. et al., including of from & et al., et al., – a family to the Hydradephaga, including some families (e.g. to Gyrinidae and have a fossil record of both larval and adult to at least the (Beutel et al., & of the family have been but are known as far as the & et al. beetle to have been described from and including of Balke, Miller, Miller & Balke et al., and & The families and described are to the although the family is and to be et al. (2013) and & (2014) for recent of fossil Hydradephaga. The fossil record of Hydrophiloidea has been expanded in recent years (Fikáček et al., Fikáček & and Ma to the early have been to be to the lineage or characters that would the for review of for six families Epimetopidae have been described, though the few currently to are of The family Hydraenidae is known from a of to the & Recently, et al. (2017) described a fossil from and provided a of for the family. Although most aquatic byrrhoid families are known in the fossil the known are with Elmidae and Psephenidae known from the and Dryopidae and Lutrochidae from the & elmid from was from the family et al., see also for a review of Elmidae). The of fossil Psephenidae are in et al. those in Elmidae are in Jäch et al. At taxonomic our knowledge of aquatic beetle larvae is with the larvae even of the families and been described & et al., Of the nearly families considered in this review, the or of at least one of each have been described, as well as most of the subfamilies and higher-level diversity of each lineage. The of molecular to life as has been a to specialists to or water beetle larvae would have otherwise been at best (e.g. & 2012; et al., water beetles were early of such methods et al., and remain a group for the and of (e.g. Bergsten et al., 2012; et al., 2013). The of robust et al., has made least to some of taxonomic of unknown larvae more which also has for aquatic et al., However, these often on what are already the most and and any use of the in the more tropical remains Consequently, both the a of adults remains in most The larvae of most genera are described of the larvae of the genera and remain as the et al. (2013) a from to the family larvae were unknown although with or molecular data. Few recent are and some (e.g. remain definitive for some be found in & and Beutel & Hydradephaga, most subfamilies and tribes have the immature described for at least one exceptions and among Gyrinidae and the genus Of the subfamilies and tribes currently within the larvae for only one remain unknown & 2016). Michat et al. (2017) provide a comprehensive review and discussion of dytiscid larvae in a phylogenetic as well as and to the larvae of the subfamilies and tribes of the diving beetles of the The by et al. is also a of larval and focused on Similarly, larvae have been described for of families, subfamilies and tribes of Hydrophiloidea, and continued our knowledge at the genus and species remains substantially et al., 2016). Of the four currently subfamilies of Hydraenidae, most larval are from the more species-rich and and The larvae of was described for the first by & but immature remain unknown for the The larvae of aquatic are best known due to their frequent and their in Psephenidae is better known as larvae than as Dryopidae remain an of the genera, larval for only about This from most known larvae of the family being and collected by aquatic insect Water although defined by their for aquatic ways of a broad of and have to their terrestrial as larvae or on multiple Many aquatic habitats are and in and a more to in terrestrial This ecological with has positioned water beetles as a study group for to ecological and (e.g. Ribera et al., Bloom et al., The of comprehensive and also has our knowledge of water beetle ecology (e.g. for Aquatic Coleoptera Jäch & Balke (2008) provide a of water beetle habitats and the categories into which they be particularly recent to the ecology of aquatic beetles have been the of diving beetles in and the of a understood and guild of water beetles. of of new species and many new lineages how much remains unknown about water beetle aquatic habitats such as and systems have long been known to water beetle have been described from a of water beetle families including Hydrophilidae and However, the majority of these lineages are known from one or a few species. This with the of a vast community of diving beetles in in (e.g. & about species have been described in little more than a and many more with the of these was known but not the of their phylogenetic diversity or species Phylogenies many of these taxa to be derived of otherwise lineages et al., et al., 2016). This not be yet it the and morphological changes that place lineages a new Water beetles from habitats such as and generally were considered or lineages rather than major of phylogenetic in their own with the of two previously beetle families from from South and et al., Balke et al., and from & the first of Lutrochidae & Short, and Noteridae 2009) have been discovered in South genera have been described in Dytiscidae (e.g. Miller & Miller & 2011, Miller, Hydrophilidae (e.g. Short & Short & 2014) and Hydroscaphidae Short et al., In lineages known from few taxa have been found to be substantially (e.g. Perkins Short, Short & & Short, 2012; 2016). have only additional in these habitats which in have yet more previously is likely more species will be described in the coming the habitat from its to one that is more accounting of the diversity within water beetle lineages is to and both taxonomic and evolutionary within the The water beetle community is to a Most water beetle families have modern catalogues which are updated on a to taxonomic in the few remaining be and those that have already been continue to be is that of these catalogues are or for or these catalogues into a be the modern molecular phylogenies with morphological data for lineages. In group molecular support for relationships has of the morphology and classification of families considered (e.g. the of the terrestrial and aquatic into subfamilies, or the derived of within Short & Baca et al., 2017a). between morphological and molecular have not been due to a in the of but are due to and of In the evolution both within and between aquatic and terrestrial habitats in repeated and evolution of morphological characters often were used as characters to taxa various which to The families of aquatic particularly Elmidae and a comprehensive modern phylogenetic of the higher-level relationships of Hydraenidae remain and that both groups have substantial ecological major classification changes are Many that in total or in part on water beetles have been out in the last quarter of a century, including those in (Jäch & (Jäch & Balke, (e.g. and especially those in the Oriental and only to illuminate our of the biodiversity of these rather than the of their a many and tropical South and the of tropical including and by likely will the and of new taxa and lineages. on habitats such as to a the of and will likely new of their aquatic beetles used as in evolutionary biology and the to on Although and to a for some the recent of Miller & (2016) comprehensive on diving beetle biology and into the to which water beetle is Of the 10 water beetle families with more than three described genera, modern for just Dytiscidae & Noteridae Gyrinidae & Hydroscaphidae (Short et al., and Psephenidae (Lee et al., 2007). still comprehensive for the of Hydraenidae and I to the for this review as well as I to Gustafson and Maier for their data on the species richness of the Gyrinidae and Lutrochidae, respectively. Fikáček and Ribera provided on early of this I Balke Fikáček Maddison and (Dryopidae, Psephenidae, Hydraenidae, Haliplidae, for use of their beetle This was supported in part by