2.1. Geographical isolation
The study of biogeography deeply shaped our understanding of speciation processes. It is of no surprise that the very broad pattern of geographic distribution of Daphnia appears to be influenced by vicariant processes such as the fragmentation of ancient continental landmasses or glaciation cycles. The genus Daphnia displays a geographic dichotomy, where species within the subgenusCtenodaphnia occur in the southern hemisphere while the subgenusDaphnia contains species commonly found in the northern hemisphere and rarely in the tropics (Hebert, 1978; Benzie, 2005; Popova & Kotov, 2013). Fossils from the Jurassic-Cretaceous boundary indicate that daphniids originated more than 145 Mya ago (Kotov & Taylor, 2011), consistent with estimates of molecular dating (Lehman, Pfrender, Morin, Crease, & Lynch, 1995; Colbourne & Hebert, 1996; Adamowicz, Petrusek, Colbourne, Hebert, & Witt, 2009). This line of evidence suggests that the two subgenera (Daphnia and Ctenodaphnia ) likely originated during the breakup of Pangea into Laurasia and Gondwana (Kotov & Taylor, 2011). Based on phylogenetic evidence, it was estimated that allopatric speciation accounts for ~42% of the speciation processes in Daphnia (Adamowicz et al., 2009; Figure 2), of which the majority (30%) represent intercontinental splits, while 12% represent intracontinental isolation. Species counterparts of D. pulex , D. pulicaria , D. obtusa ,D. curvirostris , and D. magna in European and North American displayed high sequence divergence indicating splits between 2–3 Mya (Colbourne et al., 1998; Cerny & Hebert, 1999; Hobæk & Weider, 1999; Weider et al., 1999a; De Gelas & De Meester, 2005; Markova, Dufresne, Rees, Cerny, & Kotlik, 2007). Intercontinental divergence between North and South America is also observed between daphniid species (Adamowicz, Hebert, & Marinone, 2004). As an example, South American and North American populations of the broadly distributedDaphnia ambigua show a split from a common ancestor of roughly 2 Mya (Hebert, Witt, & Adamowicz, 2003).
At the intracontinental level, lineages within species complexes often show disjointed distributions, suggestive of allopatric speciation. A phylogeographic survey of D. ambigua in North America showed regional divergence with distinct central and eastern lineages that coincide with the Appalachian mountain range barrier (Hebert et al., 2003). Genetic differentiation has also been found to occur between populations of D. pulex and Daphnia arenata on either side of the Rocky Mountains (Crease et al., 1997). The Daphnia carinata species complex, thought to have diversified roughly 70 Mya during the Mesozoic (Benzie, 1987), includes species that are endemic to various habitat-specific regions in southern Australia on either side of a montane barrier, with some of the more broadly distributed species showing regional divergence (Hebert & Wilson, 1994).
The repeated glacial cycles of the Pleistocene have been viewed as a ‘speciation pump’ (sensu Haffer, 1969) that has shaped current geographic distributions of Daphnia . Glaciation events have likely shaped the distribution of the D. pulex species complex in North America (Weider & Hobæk, 1994; Crease et al., 1997; Haileselasie et al., 2016), with many lineages originating during the Pleistocene (Colbourne et al., 1998; Crease et al., 2012). Arctic lineages in theD. pulex species complex display a biogeographic pattern consistent with a shift to a high-Arctic refuge during the last glaciation event, likely due to northward movement via passive dispersal. This dispersal colonized parts of Greenland, Iceland, Svalbard, and northern Europe (Van Raay & Crease, 1995; Weider, Hobæk, Crease, & Stibor 1996; Weider & Hobæk, 1997; Colbourne et al., 1998; Weider et al., 1999a; Weider & Hobæk, 2003). Populations from glaciated regions typically show decreased genetic divergence compared to populations from unglaciated regions (Ishida & Taylor, 2007a). Glaciation events have also shaped the distribution of closely related species Daphnia parvula and Daphnia retrocurva , withD. retrocurva distributed in northern regions of North America in large post-glacial regions (Costanzo & Taylor, 2010). Despite its broad distribution, D. obtusa shows cryptic allopatric lineages in North America (Hebert & Finston, 1996), which diverged roughly <1 Mya, suggesting recent range expansion from refugia following Pleistocene glaciation (Penton, Hebert, & Crease, 2004). In North America, molecular analyses revealed multiple cryptic lineages within the Daphnia laevis species complex, correlated with the emergence of glaciation events, followed by geographic dispersal along major migratory flyways (Taylor et al., 1998). In Europe, theDaphnia longispina species complex was found to originate roughly 5-7 Mya (Schwenk, 1993; Taylor, Hebert, & Colbourne, 1996), due to range expansion from multiple glacial refugia (Petrusek, Seda, Machacek, Ruthova, & Smilauer, 2008; Zuykova, Simonov, Bochkarev, Taylor, & Kotov, 2018a; Zuykova et al., 2018b). However, the majority of the lineages within this species complex are restricted to one continent, which distributions largely shaped by geographic barriers (Schwenk, Posada, & Hebert, 2000), although there are some exceptions (Ishida & Taylor, 2007a; Ishida & Taylor, 2007b). Recent phylogeographic surveys in Asia have uncovered a new cryptic species complex, Daphnia mitsukuri , with multiple lineages that span a wide altitude range due to multiple colonization events stemming from an eastern Palearctic refugia (Ma, Petrusek, Wolinska, Hu, & Yin, 2019). In the Daphnia carinata species complex, the restricted distribution of Daphnia nivalis in alpine glacial lakes in the Snowy Mountains of Australia was attributed to post-Pleistocene glaciation (Benzie, 1987).
Despite the overwhelming evidence that geography played an important role in driving speciation in Daphnia , particularly when looking at deep phylogenetic splits within daphniids, geographic processes cannot account for the great biodiversity of this genus. Recent studies on unglaciated regions show endemism, uncovering multiple cryptic lineages and relict groups (Benzie, 1986; Benzie & Bayly, 1996; Ishida & Taylor, 2007a; Kotov & Taylor, 2019) and these regions are considered ‘hotspots’ for biodiversity. Thus, geographic barriers could not completely account for current biogeographic patterns in certain species complexes (Kotov & Taylor, 2010), suggesting that other speciation processes, such as ecological forces might have a strong influence. Unfortunately, delineation between geographic and ecological processes is notoriously difficult to make (Taylor et al., 1996; Weider et al., 1999b; Schwenk et al., 2000; Taylor, Sprenger, & Ishida, 2005), and it has been suggested that multiple processes often shape the current biodiversity of this genus.