On the genus Haploginglymusnext section
Our phylogenetic analysis corroborates the monophyly of the family Niphargidae but shows Niphargus to be paraphyletic as it currently stands (Figs 2; 3). The paraphyly of Niphargus was already suggested by Sket (1981) and implicitly recognized by Cene Fišer and co-workers, who in different molecular phylogenetic papers (Fišer et al., 2008; Hekmatara et al., 2013; Švara et al., 2015; Esmaeili-Rineh et al., 2015; Brad et al., 2015) showed Carinurella, Pontoniphargus, Niphargobates and Niphargellus nested within that genus. In our analysis, Haploginglymus appears also nested within Niphargus conforming a robustly supported monophyletic group. This clade is sister to another one conformed by the English Niphargellus glenniei (Spooner, 1952) and the Irish Niphargus irlandicus Schellenberg, 1932. These two taxa were already recovered as the most basal niphargids in the analyses performed by McInerney et al. (2014) and Brad et al. (2015). The retention of Haploginglymus as a valid genus in this context is thus implausible, especially taking into consideration that its presumed more remarkable diagnostic trait (the unsegmented U3 exopod) is shared also by the monotypic genera Niphargobates and Carinurella. Nevertheless, the former two genera differ remarkably from Haploginglymus in many relevant features, in accord with their peculiar life-styles.
In any event, our analysis recovers all Haploginglymus species conforming a clade geographically restricted to the Iberian Peninsula, where it replaces the rest of niphargids except for two small disjunct areas on both edges of the Pyrenees (Notenboom, 1990). As in Niphargus s.l., this monophylum includes some morphologically aberrant members such as H. morenoi that, if it were not because it displays an unsegmented U3, it would deserve that a new genus was erected to accommodate it. The potential for morphological disparity in Haploginglymus is thus as high as in Niphargus s. l., and would suggest a long, separate evolutionary history between both lineages, enabling Haploginglymus to colonise and adapt to even the most demanding subterranean niches in Iberia. Considering all these pieces of evidence and that Niphargus already accounts with more than 400 valid species, we opt here for the retention of Haploginglymus as a valid genus. As regard Haploginglymus morenoi and its aberrant morphology, our results fully agree with Iannilli et al. (2009) in considering it as a highly modified member of the genus with a peculiar, interstitial lifestyle.
Origin, intra-generic relationships and biogeography
Haploginglymus is endemic to the Iberian Peninsula, where it displays an almost generalised distribution in fresh subterranean waters and is the only niphargid known unless for two disjunct areas on both edges of the Pyrenees, where Niphargus also occurs (Notenboom, 1990). In those areas, the orography is less pronounced than at the central Pyrenean sector, to the point of enabling the local trans-Pyrenean colonization by Niphargus. The latter genus is broadly distributed across the rest of Europe and the Middle East (Esmaeili-Rineh et al., 2015), and seems to have colonised the two foregoing Spanish regions quite recently, apparently after the establishment of Haploginglymus in Iberia. Thus, at least three of the four Niphargus species recorded in Catalonia display a trans-Pyrenean distribution (Karaman, 2015a; b), whereas those from the Basque Country conform a derived monophyletic group within the niphargid tree that appears to be unrelated to Haploginglymus (v. Fig. 2). Thus, the rise of the Pyrenees with its current structural organisation already at the Paleocene/Eocene boundary, 55-47 Myr ago (Vergés et al., 2002) might have played a major role in the differentiation of Haploginglymus from the rest of niphargids, with this age representing a terminus post quem for its origin (Fig. 1).
The nine Haploginglymus samples included in our analysis conform two main monophyletic groups (Figs 2; 3; nodes B and C). One (node B) appears associated to the River Ebro and its tributaries, and comprises the three species recorded at River Matarranya, of which only H. morenoi has been formally described thus far. The second group (node C) is distributed across the southern half of the Iberian Peninsula and comprises the rest of taxa. The genetic divergence among these Haploginglymus lineages is high, as deduced from the observed values of uncorrected pairwise cox1 p-distances (Table 4). With the caveat that DNA sequences from representatives of other parts of the northern sector of the Peninsula are lacking, it is feasible to relate this primary subdivision of the genus to the uplift during the Tertiary Alpine orogeny of two intra-Peninsular mountain ranges: The Central System (that separates the Duero and Tajo river basins) and the Iberian System (that separates the Ebro river basin from the rest of the Peninsula) (v. Fig. 1).
Age and origin of the Niphargidae
Niphargids are almost completely bound to subterranean waters except for a few species recorded from surface habitats or the bottom of deep lakes (Karaman and Ruffo, 1986). The family occurs only in the Palaearctic west of the Caspian Sea and currently comprises ten valid genera, of which Niphargus, with 408 species, is by far the most species-rich. The rest embraces only 16 species (Horton and Lowry, 2013). Only Haploginglymus is allopatric with respect to the rest of niphargids, although it overlaps with Niphargus (exceptionally coexisting with it) in two small disjunct areas on both edges of the Pyrenees (Pretus and Sabater, 1990; Notenboom, 1991; Karaman, 2015a; b).
Karaman and Ruffo (1986) suggested that the diversification of the family began in the basins of the Paratethys Sea during the Tertiary Period, from which European fresh waters were subsequently colonised. But the discovery of casts of niphargids in Late Eocene Baltic amber (ca. 45-50 Myr old; Coleman and Myers, 2001; Coleman and Ruffo, 2002; Jażdżewski and Kupryjanowicz, 2010) rules out this hypothesis since the birth of Parathethys as an enclosed basin with reduced salinity and endemic faunas took place only afterwards, at the early Oligocene at most (Rögl, 1997). McInerney et al. (2014) have attributed a much older, late Cretaceous age to the family, and an origin in NW Europe rather than in the Balkan area followed by a gradual range expansion across central Europe to reach its current distribution range. A third, alternative scenario where niphargids colonized freshwaters directly from the sea multiple times independently during their evolutionary history, and where the lineage of Haploginglymus settled in Iberia when it was an island during the late Cretaceous is not favored here since, if this were the case: (1) it is hardly conceivable that Niphargus was not established throughout the entire Peninsula when it is present in French territories adjacent to the Atlantic and the Mediterranean, and vice-versa, that Haploginglymus is not present in France; (2) there are no niphargids on North Atlantic oceanic islands, nor on Mediterranean islands such as the Balearics, disconnected from the continents at least since the end of the Oligocene, but where ulterior marine transgressive pulses have constrained severely the extension of emerged land; (3) the distribution of the family is not limited to areas covered by epicontinental seas in the geological past (Notenboom, 1991); and (4) brackish water species are exceptional among members of the family (Notenboom, 1991).
Our study reveals a strongly supported sister relationship between niphargids and pseudoniphargids that deserves a comment since it is relevant for the origin and biogeography of the Niphargidae. Contrary to niphargids, which are mainly limnic and are present on territories not formerly occupied by the sea, the Pseudoniphargidae are undeniably thalassoid, being present even on oceanic islands that have never been connected to the continents (Bermuda, Canaries, Madeira, Azores; Stock et al., 1986; Stock, 1988; Stock and Abreu, 1992; Stock, 1980). The study of the phylogenetic relationships among members of the family based on both morphological (Notenboom, 1988) and molecular features (analyses currently under way in our lab) has lead to the identification of a monophyletic cluster of species of Pseudoniphargus Chevreux, 1901 cantoned on the western edge of the Pyrennees – Basque Country, bordering the Gulf of Biscay – as the most primitive lineage within the family. Remarkably, the most primitive niphargids are also found in an area adjacent to the Gulf of Biscay (Great Britain; Mcinerney et al., 2014). Accordingly, we suggest that niphargids – contrary to pseudoniphargids – have colonized continental waters only once, and from a marine ancestor common to both families, and that this common ancestor most probably lived on the NE Atlantic coasts at the end of the Cretaceous. Once established in continental waters, niphargids proceeded to spread across Europe, with the colonisation of Iberia by the ancestor of Haploginglymus taking place before the rise of the Pyrenees 55-47 Myr ago. Niphargids should continue to be considered as a primary limnic group for biogeographic purposes, despite its presumed relatively recent (late Cretaceous) marine origin and sister relationship with the (thalassoid) Pseudoniphargidae.