Systematics, biogeography and evolutionary patterns of the Hydrobiidae family (Mollusca: Caenogastropoda)

• Autores: Félix Jonathan Pereira Miller
• Directores de la Tesis: María Ángeles Ramos Sánchez (dir. tes.), Diana Delicado Iglesias (codir. tes.)
• Lectura: En la Universidad Autónoma de Madrid ( España ) en 2021
• Idioma: español
• Tribunal Calificador de la Tesis: José Templado González (presid.), Juan Moreira (secret.), Marco Oliverio (voc.)
• Programa de doctorado: Programa de Doctorado en Biología por la Universidad Autónoma de Madrid
• Materias:
  • Ciencias de la vida
    • Biología animal (zoología)
• Enlaces
  • Tesis en acceso abierto en: Biblos-e Archivo
• Resumen

  • Within Mollusca, freshwater gastropods are one of the most diverse groups. Previous studies have reported that this group is comprised of approximately 4,000 valid nominal species, but they have also concluded that this estimate may represent only 25% of its actual diversity. Freshwater snails are found on every continent, except Antarctica and in nearly every aquatic habitat (lotic and lentic), including lakes, swamps, rivers, streams, springs, underground aquifers and interstitial waters, among others. The truncatelloidean family Hydrobiidae Stimpson, 1865 is the most specious assemblage of freshwater gastropods, with ca. 1,000 recognised species. However, until the application of molecular tools, most brackish or freshwater snails that were small (between 0.5 and 8 mm) and had featureless shells were typically classified as hydrobioids (= Hydrobiidae sensu lato [s. l.]). This likely resulted in an overestimation of the family’s actual species richness. In fact, the modern sensu stricto (s. str.) definition of the family, based on morphological and molecular evidence, recognised some former hydrobiid subfamilies as independent families. Taken altogether, these points highlight our incomplete understanding of the species richness and the global geographic distribution of Hydrobiidae s. str. New investigations of global-scale patterns of hydrobiid species richness and endemicity are needed to provide a baseline from which to infer the factors influencing the evolutionary processes and geographic distribution of Hydrobiidae.

    Historical biogeography is of special importance for understanding the evolutionary processes that have led to the geographic distribution and species richness of extant clades. The scarcity of these types of studies is due, in part, to the fact that the great majority of the hydrobiid genera inhabiting Mediterranean continental waters consist of narrow-range endemics in species-poor assemblages. Only a few Mediterranean genera are diverse and geographically distributed enough to be good candidates for studies aiming to reconstruct biogeographic histories and infer speciation processes. Suitable taxa for such studies include Corrosella Boeters, 1970, a genus comprised of 17 known species distributed mainly in mountain springs in the Iberian Peninsula and the lowland genera Pseudamnicola Paulucci, 1878, Islamia Radoman, 1973 and Mercuria Boeters, 1971, with 70, 45 and 26 currently recognised extant species, respectively. The latter three genera occur in a wide range of freshwater ecosystems throughout the Mediterranean basin in both Europe and North Africa. Pseudamnicola has also been reported for the Ponto–Caspian region and Mercuria for Atlantic coastal regions. The four studied genera are notable as they occupy different habitats along an elevational gradient (e.g., from the highland species of Corrosella to the saline species of Mercuria). Among these genera, Mercuria is the assemblage with the largest number of species delineated on the basis of conchological features. Given the questionable use of shell features to identify hydrobiid species, the taxonomy of Mercuria populations needs to be re-evaluated within an integrative framework.

    Although there are numerous taxonomic works that described the morphology of hydrobiids, an exhaustive review of the total number of species and an explorative analysis inferring which evolutionary factors have determined their present-day richness have not yet been performed. Furthermore, although there are various phylogenetic works describing the relationships of genera inhabiting mountainous regions, coastal streams and lakes or brackish environments, there are currently no studies comparing the evolutionary patterns (e.g., divergence times, rates of genetic substitution or diversification and biogeographic patterns) of these genera. Such studies would provide a better understanding of how the biota found in these habitats evolved. To this end, the major goal of this doctoral thesis was to test which ecological and geographic factors may have influenced (i) the disparity in species richness among regions and taxonomic groups and (ii) the evolutionary patterns among hydrobiid genera.

    To achieve this goal, we first conducted a comprehensive literature and biodiversity database review in order to determine the number of total, endemic and threatened species per freshwater ecoregion and compile information on them. We classified ecoregions as hotspots if each biodiversity index was in the top 25% of its range and assessed the effect of environmental and evolutionary factors on species richness using generalized linear models. We identified 906 extant species and 157 genera of Hydrobiidae s. str. that show a mainly Nearctic–Palearctic distribution and 11 biodiversity hotspots, most located across the Mediterranean basin. According to our dataset, 83% of the species are endemic to a single ecoregion. Out of the 43% of non-data deficient species, nearly three times more were classified as threatened than non-threatened and extinction risk peaked at 1,500 m.a.s.l. Species richness was unequally distributed over biogeographic realms, increasing with greater connectivity among ecoregions and was negatively related to annual temperature range. Latitude and precipitation seasonality partially explained the variation in richness by a non-linear relationship. The identified hotspots correspond to those of other freshwater taxa.

    Multilocus phylogenies were used to compare the evolutionary patterns and historical biogeography of the selected hydrobiid genera. Combined data from mitochondrial (mtCOI and 16S rRNA) and nuclear ribosomal (28S rRNA) gene fragments were analysed for 183 individuals of Mercuria, 116 of Pseudamnicola, 34 of Islamia and 102 of Corrosella. To obtain sequences for Mercuria, 129 locations were sampled. For Islamia, many of the sequences were generated de novo using samples deposited in the University of Giessen Systematics and Biodiversity (UGSB) collection. Sequences of Corrosella and Pseudamnicola were retrieved mainly from GenBank (NCBI).

    Our integrative study revealed changes in the taxonomy and geographic distribution of the species of Mercuria studied. Based on the applied species delimitation methods and morphological characterisations, 14 putative species were identified in the dataset: nine correspond to species recognised through traditional taxonomic classifications and five are new to science. Among the applied molecular delimitation methods, automatic barcode gap discovery (ABGD) was the most efficient at recognising Mercuria species (match ratio value of 0.85), confirming 12 of the 14 species delimited in the study. By contrast, the tree-based methods [i.e., generalised mixed Yule coalescent (GMYC) and Poisson tree processes (PTP)] presented a relatively low match ratio value, coinciding with an overestimation of the number of species. Several authors have recognised that these methods should be interpreted with caution for groups with low vagility and poor dispersal capacities such as gastropods, contrasting the validity of the delimited species with morphological data under an integrative taxonomic approach. Morphological similarity among Mercuria species was scored on the basis of a geometric morphometric analysis of shell shape using a total of 1,148 shells from 46 populations and 21 landmarks and semilandmarks. In the PCA, the first two components accounted for 75.3% of the variation in shell shape; however, only 50% of the species were successfully delimited on the basis of the shell morphometry. Generally, the shell, operculum and genitalia of Mercuria present wide intra- and interspecific variability; therefore, the use of these characters should be used cautiously when assessing the taxonomic status of populations of this genus. Instead, we propose the use of integrative taxonomy, adding genetic information to morphological data. Under this approach, various misidentifications of previous records of Mercuria were revealed in our study, representing a significant advancement in our knowledge of this genus.

    Bayesian inference and maximum likelihood analyses recovered similar topologies (in terms of the degree of geographic restriction of clades and divergence times) for the genera Mercuria and Pseudamnicola and for Islamia and Corrosella. By applying an external calibration rate, our dated species trees showed a similar age for Corrosella and Islamia (ca. 11 Mya) and also for Mercuria and Pseudamnicola (ca. 7 Mya), with a notably younger age estimated for the latter pair. Taking these estimated ages into account, we placed the origin of the diversification of the four genera in the Miocene (from 23 to 5.3 Mya). According to the diversification principle, older groups should present more species as they have had a longer time to diversify; however, when we contrasted the ages of these genera with the number of species in each, we found discrepancies demonstrating that they do not follow this principle. We found a notable difference between the number of extant species presented by Corrosella (17) and Islamia (45) and a greater difference between that by Mercuria (26) and Pseudamnicola (70). The disparity in the observed species richness could be related to differences in environmental preferences such as in habitat type, elevational range or water parameters and also in dispersal strategies.

    To test these hypotheses, the ancestral area of species was inferred in the program BioGeoBEARS using the calibrated species trees [i.e., maximum clade credibility (MCC) trees] and the geographic distribution of each species coded by freshwater ecoregion. This analysis revealed high speciation within peninsular regions, which supports previous assumptions on the role of the Mediterranean peninsulas as evolutionary centres for hydrobiids. The inferred biogeographic history of Corrosella suggests a pattern of isolation by distance produced by vicariance processes. Cladogenesis within Islamia involved a combination of founder events from the Apennine Peninsula to the east and the west and narrow (within-ecoregion) sympatry, whereas in Pseudamnicola and Mercuria, it involved a series of dispersal and colonisation events followed by subsequent isolation.

    The BioGeoBEARS analyses showed that rates of anagenetic range expansion (d) were higher in species that occur at high (Corrosella) and low (Mercuria) elevations, whereas those that occur at medium elevations (Pseudamnicola and Islamia) tended to have low dispersal rates. Also, species that occur at high elevation (i.e., Corrosella) presented the lowest founder event (jump dispersal) speciation parameter (j) at cladogenesis, while species that occur at medium elevations (i.e., Pseudamnicola and Islamia) showed a propensity for jump dispersals. In order to assess the potential relationship between altitudinal range and dispersal, we calculated, for each genus, the dispersal ratio (dratio = (Nd + Nj) / Nlin) from the number of dispersal or range expansion events (Nd), jump-dispersal or founder events (Nj) and total lineages (Nlin) and plotted this against elevation occurrences. We observed that the dispersal ratio decreased with elevation, from 0.42 for the lowland genus Mercuria to 0.14 for the headwater genus Corrosella. Our data, unlike for other organisms, did not evidence a direct relationship between dispersal and body size.

    Overall, this study confirms that the family Hydrobiidae is a highly diverse group with hotspots of richness, endemism and threat, mainly in freshwater ecoregions of the Mediterranean basin. This distribution of the species richness was determined by geographic and climatic factors such as latitude, watershed connectivity, precipitation, seasonality and annual temperature range, but not by elevation and geological heterogeneity (probably because the spatial aggregation of these two factors concealed their true variability). Therefore, we conclude that global hotspots of Hydrobiidae richness represent areas of climatic stability with moderate precipitation and temperature seasonality that are well connected to other hydrological basins. Our study also indicates that the current patterns of species richness have been affected by the dispersal strategy of groups, influenced by the elevation range in which they occur, rather than by age or habitat type. Better access for dispersal vectors and greater habitat connectivity at the lower elevations may have facilitated long-distance dispersal, resulting in founder event (cladogenetic) speciation. On the contrary, highly specialised and isolated groups, such as the Corrosella species inhabiting mountain-top springs, are less likely to disperse and speciate (and are probably more prone to extinction) compared with those inhabiting lower elevations. Taken altogether, our results illustrate that both evolutionary and environmental factors determine global species patterns and that future changes of the latter factors may affect the species richness of hydrobiids.