Page 10 - Delicado_Machordom_2015

Evolutionary patterns of Pseudamnicola D. Delicado et al.

In general, P. (Corrosella) species occur in isolated habi- event that has not occurred often during the evolution of
tats, present a clear biogeographic pattern of northern and this family.
southern phylogenetic clustering (shown in Delicado et al.
2013) and genetically distinct species. On the other hand, Temporal history and biogeographic origin. All of the phylo-
the geographic patterns of P. (Pseudamnicola) are not as genetic inferences performed in this work have suggested
explicit. A possible reason could be because the number of the division of Pseudamnicola s. l. into three lineages, which
samples examined is more limited, and thus, the entire dis- may correspond to three different genera. Observing their
tribution range of this subgenus has not been covered. It patterns of diversification, each of these genera is likely to
would be interesting to first, extend the study in the sampled have experienced different evolutionary processes in space
regions, and second, to genetically examine the species and time. Based on the available taxonomic sampling per-
described from Morocco and Algeria (Ghamizi et al. 1997; formed for each lineage and the coalescence analysis
Gl€ oer et al. 2010) in order to investigate whether the (Fig. 3) using an evolutionary rate previously cited for hyd-
current distribution of the group is a result of stochastic or robiids, we estimate that the split leading to these three lin-
tectonic processes. In any case, as no clear biogeographic eages likely occurred during the Upper–Middle Miocene
pattern is present among the studied populations, an alterna- (28–17 Ma). Although not all of the described Pseudamnicola
tive explanation for their distribution pattern may be long- s. l. species have been included and not all of the areas of its
distance colonization followed by isolation. As the principal distribution range have been sampled in this work, our
aim of this work is to study the relationships between species results of the ancestral area estimation suggest that the Ibe-
and their biogeographic distribution, no population genetic rian Peninsula has played an important role in the diversifi-
level analyses have been performed, thus we cannot hypoth- cation of the group as it is, to date, the only region in
esize which dispersal mode the populations of this subgenus which all three proposed genera inhabit. Thus, given the
have followed. Further research at the population level and more restrictive distribution pattern of Corrosella and the
over a larger geographic area is required. distribution pattern of Didacus gen. n. (composed of Didacus
Nevertheless, phylogenetic analysis of the group does gasulli) with respect to that of Pseudamnicola, their evolution
reveal the existence of three well-supported lineages within may be a result of a peripatric (allopatric) speciation that
the genus. Previous anatomical studies (Delicado et al. occurred in the Iberian Peninsula (as shown in Fig. 3). Dur-
2012, 2014) reflect substantial differences between the sub- ing the Oligocene and Miocene (between 35 Ma and
genera and highlighted P. (P.) gasulli as a different entity 5.33 Ma), the Iberian Peninsula suffered a compressive per-
bearing evidences of divergence. Taking all these results iod with the loss as well as adhesion of several continental
into account, we suggest that the three lineages may corre- fragments (Hevia 2004) in which most of the Iberian moun-
spond to three different genera, raising Corrosella to the tain ranges originated, thereby affecting the region’s hydro-
category of genus once again (as in Boeters 1970) and logical system. The creation of these physical barriers may
removing P. (P.) gasulli from P. (Pseudamnicola), thus mak- have caused an isolation process followed by vicariance by
ing itself a new genus here designated as Didacus n. gen. In which the geographic range of the last common ancestor of
addition to the anatomical and morphological characteris- the three genera was fragmented in a relatively short period
tics that sufficiently distinguish them as different genera, of time, thus leading to the separate lineages. However, it
the genetic divergences that exist between them (uncor- seems that some populations not only remained isolated but
rected distances ranged between 11.1% and 14.3% for also evolved and adapted to new habitat conditions in a
COI and between 6.7% and 8.4% for 16S) are similar to mountainous environment (as in the case of the genus Cor-
those reported between other genera. For instance, rosella). It is noteworthy that our results revealed significant
between genera belonging to the sister subfamily Hydrobii- differences in habitat features (altitude and conductivity)
nae (based on phylogenies in Wilke et al. 2001, 2013; Szar- between genera (Table 1), which may be a consequence of
owska 2006), for example Adriohydrobia, Hydrobia, Peringia an adaptive process. Therefore, this may be a case in which
and Ventrosia, molecular distances range between 10.4% there is some implicit degree of natural selection in allopat-
and 14.8% for COI and 2.3% and 5.8% for 16S (uncor- ric speciation, as postulated by Wright (1931).
rected pairwise distances, Wilke 2003). Therefore, the sub- With respect to Corrosella species, the inclusion of addi-
family Pseudamnicolinae would be composed of three tional Pseudamnicola sequences to the *BEAST analyses did
genera, one strict freshwater and two euryhaline freshwa- not substantially change the divergence times previously
ter. Thus, within the current phylogeny for hydrobiids estimated for the three major cladogenetic events leading
(Wilke et al. 2013), these two subfamilies are very interest- to diversification within the Corrosella clade (see Delicado
ing from an evolutionary perspective because they may et al. 2013). In the analysis here, these events are estimated
represent a transition between two different environments, to have occurred slightly earlier (ca. 12 Ma, 6 Ma and

412 ª 2015 Royal Swedish Academy of Sciences, 44, 4, July 2015, pp 403–417