Cricotopus

Krosch, Matt N., Cranston Fls, Peter S., Baker, Andrew M. & Vink, Sue, 2015, Molecular data extend Australian Cricotopus midge (Chironomidae) species diversity, and provide a phylogenetic hypothesis for biogeography and freshwater monitoring, Zoological Journal of the Linnean Society (Zool. J. Linn. Soc.) 175 (3), pp. 496-509 : 505-506

AUSTRALIAN CRICOTOPUS

The non-monophyly of the Australian Cricotopus fauna with regard to species from elsewhere in Asia and the austral region clearly implies a history of multiple colonization events into and possibly out of Australia. The recognized centre of extant diversity for both Cricotopus and Paratrichocladius is in the Palaearctic, and this is assumed to represent a centre of origin for both groups. Given this assumption, the nested and scattered phylogenetic positions of select specimens from Singapore, Chile, New Zealand, USA, and South Africa within clades of Australian taxa suggests that Australian Cricotopus are not likely to share a single common origin on the continent. Instead, this pattern suggests a history characterized by waves of colonization into and dispersal out of Australia. Determining the directionality for these migrations is fraught with difficulty under the current sampling strategy, requiring instead intensive and targeted sampling from non-Australian locations, and so lies outside the scope of this discussion. Nevertheless, divergence events involving other austral landmasses do not largely accord with the classic geological timeline of continental breakup, especially for African C. flavozonatus (47 Mya) and C. albitibia (18 Mya); however, the estimated divergence of the Chilean species at around 26 Mya falls at the lower end of the age estimates for trans- Antarctic connections between Australia and South America via the Drake Passage ( Barker & Burrell, 1977), and divergence of New Zealand taxa around 28 Mya accords with much recent evidence for the post- Gondwanan mid–late Palaeogene divergence of many New Zealand groups (e.g. Krosch & Cranston, 2013; Buckley, Krosch & Leschen, 2014). Discordance between austral divergence events and Gondwanan fragmentation is not surprising, given this genus is considered to have originated in the Eurasian region and is not considered a Gondwanan relic (as opposed to true austral chironomid groups that have received attention in recent years; Cranston et al., 2010; Krosch et al., 2011; Krosch & Cranston, 2013). That said, although the estimated tmrca values for New Zealand and South America may yet suggest a role for continental vicariance in the evolution of this genus, longdistance dispersal events are not unknown among the chironomids that inhabit the austral landmasses ( Krosch et al., 2011).

Within the Australian Cricotopus , our sampling strategy allowed us to estimate divergence times of taxa on certain Australian islands, namely Tasmania and Lord Howe Island. Both islands were proposed to possess endemic species not found on the mainland ( C. tasmania and C. howensis , respectively; Drayson, 1992; Cranston, 1996); however, we have shown clearly that C. tasmania occurs also on mainland Australia, albeit in apparently lower abundance. The non-monophyly of island versus mainland specimens within C. tasmania suggests that this taxon may well have diversified in isolation on the island prior to colonizing the mainland. Likewise, its sister taxon, C. annuliventris , is known mostly from the mainland but also from a small number of locations in Tasmania, and exhibits a similar phylogenetic pattern to C. tasmania , which suggests that this species may have colonized Tasmania following divergence. The divergence of the two species around 14 Mya obviously pre-dates the most recent closure of Bass Strait at the end of the last glacial cycle (∼18 000 years ago; Lambeck & Chappell, 2001), along with Pleistocene divergences reported for some other taxonomic groups (e.g. Toon et al., 2007; Gongora et al., 2012; Martin & Zuccarello, 2012), but accords with other molecular phylogenetic studies that report Miocene–Pliocene divergences between mainland and Tasmanian taxa (e.g. Waters & White, 1997; Symula, Keogh & Cannatella, 2008). Thus, we argue that these taxa diverged in allopatry, separated by the proto- Bass Strait, and have both since dispersed across an intermittent sea barrier to colonize the opposite landmass. Whether this occurred via direct transoceanic dispersal, by island-hopping across the Bass Strait, or during periods of exposure of the Bassian Isthmus landbridge remains speculative. Nevertheless, there are apparent precedents for dispersal between geographically distant austral landmasses in other chironomid groups ( Krosch et al., 2011; Krosch & Cranston, 2013). Somewhat in contrast, phylogenetic relationships within other Cricotopus species for which we have Tasmanian representatives ( C. conicornis , C. hillmani , and C. parbicinctus ) all supported mainland and Tasmanian populations as reciprocally monophyletic, with divergences of Late Miocene–Pliocene age. This suggests that populations of these three species were isolated across the proto-Bass Strait and diverged in allopatry, but have not dispersed across the strait since. Differential dispersal among congeners that were isolated and diverged in allopatry across Bass Strait is also known from terrestrial mammals ( Antechinus ; T. Mutton, pers. comm., 2014), and highlights the biogeographical complexity of the region.

In contrast, the estimated divergence time for the Lord Howe Island endemic C. howensis from its sister mainland species C. hillmani of around 6 Mya accords closely with the geological time frame for the emergence of Lord Howe Island (6.4–6.9 Mya; McDougall, Embleton & Stone, 1981) as a result of volcanism along the Lord Howe Rise. This accords with molecular phylogenetic information from other freshwater invertebrates ( Page et al., 2005), but contrasts with other much older Lord Howe Island endemic insect taxa (e.g. Buckley, Attanayake & Bradler, 2009). No information exists concerning the Cricotopus faunas of other nearby South Pacific Islands (e.g. Norfolk Island), so it remains unknown whether the arrival of Cricotopus on Lord Howe Island was facilitated by island hopping via extant or now-submerged land areas or by transoceanic dispersal, possibly mediated by West Wind Drift ( Cook & Crisp, 2005). Interestingly, there does not seem to be a close relationship between C. howensis and our New Zealand representative, given the many instances of apparent colonization of New Zealand via the Lord Howe Rise during the Eocene–Oligocene, around the time at which we estimate the New Zealand taxon to have diverged (node F); however, greater sampling of New Zealand’s Cricotopus diversity is needed to resolve this question fully.

The C. albitibia group also presents an interesting biogeographical and ecological conundrum. Cricotopus albitarsis and C. wangi are clearly close genetically (although the equivocal position of the genetically divergent form of C. albitarsis renders the actual sister grouping of C. albitarsis + C. wangi uncertain), but they exhibit divergent larval morphology and ecology ( Drayson et al., 2015). Larvae of C. wangi live in the hygropetric zone of waterfalls and shallow but rapid riffles in only a few locations in northern Australia. Cricotopus albitarsis , on the other hand, is widespread and abundant across mainland Australia, in both lotic and some lentic habitats. Moreover, C. albitibia group species from outside Australia possess broad tolerances to impact (e.g. Beneberu et al., 2014). Possibly the common ancestor of C. albitarsis + C. wangi was tolerant to ecosystem impact, and the subsequent specialization of C. wangi represents diversification into a vacant niche. Interestingly, the ‘divergent NT C. albitarsis ’ taxon possessed an almost identical 28S sequence to putative C. albitibia from Papua New Guinea ( PNG; data not shown, but available on request). The PNG specimens were not successfully sequenced for the other loci used, and thus the relationships between these three taxa remain somewhat enigmatic. Several possibilities exist: (1) all three taxa represent a single species, with C. wangi an ecologically and morphologically divergent variant and ‘divergent NT C. albitarsis ’ representing highly divergent populationlevel structure (NT + PNG versus eastern Australia); (2) C. albitarsis and ‘divergent NT C. albitarsis ’ form a single species, with genetic divergence related to population-level structure, and with C. wangi as an incipient species diverging from within C. albitarsis ; (3) the equivocal placement of the ‘divergent NT C. albitarsis ’ is misleading and this form is actually sister to eastern Australian C. albitarsis , with C. wangi as sister taxon to this clade; or (4) the ‘divergent NT C. albitarsis ’ represents a third species distributed on both sides of the Torres Strait for which morphological differences have either not developed yet, are not apparent in the life stages (pupae and larvae) or specimens sampled currently, or exist in characters other than those covered by current identification keys. The first two scenarios do not fit the data adequately given the quite distinct morphological differences between C. albitarsis and C. wangi at all life stages ( Drayson et al., 2015), and the placement of C. wangi as sister taxon to all eastern Australian C. albitarsis , rather than emerging from within the clade. Furthermore, no C. albitarsis from the NT were placed within the clade of eastern Australian specimens. The latter two scenarios emerge as the more likely; however, determining these relationships definitively will require additional sampling in northern Australia and PNG.