Titolo della tesi: Genetic and Morphological diversity in Sarmarutilus rubilio (Bonaparte, 1837), an Italian endemic freshwater fish
EXTENDED ABSTRACT
Intraspecific diversity, i.e. the genomic and phenotypic diversity present within and among populations, is the basis for species adaptation and diversification, thus understanding its magnitude and the mechanisms that originate and maintain it, is fundamental for the conservation of biodiversity.
Such diversity is particularly high in freshwater fishes, as severe limitations to populations’ gene flow, between different hydrographic networks or even within the same basin, are responsible for high intraspecific genetic differentiation. In addition in these taxa, several mechanisms of morphological adaptations are known to generate phenotypic variations across populations inhabiting different environments within species’ ranges.
Considering that the Mediterranean area is a hotspot of freshwater fish diversity with many threatened species, I decided to investigate the genetic and morphological diversity within the South European roach Sarmarutilus rubilio, an endemic Italian species that is currently classified as Vulnerable by the International Union for the Conservation of Nature. Due to its wide ecological niche, this species is present in several basins, which are characterized by different environmental features and past hydrogeological history; so S. rubilio was considered a good model for testing the existence of still undetected intraspecific diversity, the knowledge of which will help future conservation interventions for this species.
Chapter 1 - Phylogeography of Sarmarutilus rubilio (Cypriniformes: Leuciscidae): Complex Genetic Structure, Clues to a New Cryptic Species and Further Insights into Roaches Phylogeny
Italy hosts 27 endemic and sub-endemic freshwater fish species and their origin and distribution is strongly related to its complex paleogeographic and paleoclimatic history: tectonic movement and sea-level variations during Miocene and Pleistocene promoted genetic diversification and allopatric speciation, and mountainous ridges isolated native species from the postglacial expansions of Eastern-European ones. Geological events determined different fish species assemblages, geographically localized within the Italian peninsula, so that three distinct ichthyogeographic districts were recognized, named Padany-Venetian (PV),
Tuscany-Latium (TL), and Apulia-Campania (AC), but their borders are still not clearly defined. The South European roach Sarmarutilus rubilio inhabits various freshwater environments across the three Italian districts, and sightings of this species were reported also in Albanian lakes. Thus its distribution could likely disguise a strong genetic structure, which was also suggested by preliminary data that evidenced high intraspecific clusters divergence.
Starting with these assumptions, I investigated both the complete mitochondrial control region (CR) in almost two hundred specimens from the whole species range and the barcoding portion of the cytochrome oxydase I (COI) in a subset of Italian and Albanian Sarmarutilus-like individuals, to test the presence of different South European roach genetic clusters in Italy and assess the relationship between Italian and Albanian fishes. For the latter purpose comparisons with COI sequences from almost all other roach species (which inhabit the basins from Central Europe to Ponto-Caspian region) were also performed to better assess their taxonomical status. In addition, I analyzed two nuclear markers (Cyprinid formerly unknown nuclear Polymorphism – Cyfun P, and 1° intron of the ribosomal protein S7 – RpS7) in a few Italian specimens showing a highly divergent COI profile to obtain further insights about S. rubilio phylogeography.
The results highlighted a marked genetic divergence between S. rubilio and all other roach species, including Albanian individuals, and among Italian samples, they revealed the existence of three deeply divergent geographic haplogroups (lineages), named A, B and C, whose distribution only partially matched the extension of Italian ichthyogeographic districts. Haplogroup C likely corresponds to a new putative cryptic species (its status is supported by divergence in both mitochondrial and nuclear markers despite the lack of morphological differences) and is located exclusively at the northern border of the South European roach range (Magra-Vara basin); haplogroup B is restricted to South Italy and fixed in the Fondi plain; haplogroup A is widespread across the entire range, both in Tyrrhenian and Adriatic slope of Italy, and in some sites, it is in co-occurrence with C or B. These lineages likely originated in allopatry, as a consequence of the tectonic uplifting of the Apuan Alps in the north and of the Colli Albani Volcano in the south during the Pleistocene, which promoted isolation and vicariance. Moreover, subsequent secondary contacts through sea level variations and even river captures along the Apennine Mountains were responsible for the observed lineage coexistence in some of the investigated sites. Finally, considering the overall complex
genetic structure of S. rubilio and the genetic uniqueness of populations in Magra Vara basin (where lineage C was found) and Fondi plain in Southern Lazio (where lineage B was fixed), I suggested focusing conservation interventions on these two geographic areas and to avoid translocation of Sarmarutilus between different basins for sustaining the declining ones.
Chapter 2 - Phenotypic plasticity over genetic diversity: ecomorphological patterns revealed in the eurytopic and threatened Italian endemic freshwater fish Sarmarutilus rubilio (Bonaparte, 1837)
Correct assessment of diversity among freshwater fishes, which is crucial for their conservation, could be a challenging task. Indeed numerous cases of cryptic species are reported, i.e. species that can be recognized and discriminated only by molecular approaches, and this can lead to underestimation of the actual diversity; on the other hand, several taxa show intraspecific morphological diversity across their geographic range, as a response faced by different populations to changes in environmental features (i.e. phenotypic plasticity), and misinterpreting this phenomenon can cause species inflation. Moreover genetic and morphological diversity may be correlated, especially when differences in phenotypic traits are genetically fixed between diverging populations. Within the South European roach Sarmarutilus rubilio, a freshwater fish endemic of the Italian peninsula, three different genetic lineages (named A, B, C) with distinct geographic distributions were found; lineage C divergence from the others was high suggesting it could belong to a new species, but no morphological differences were observed. In addition, the South European roach is an omnivorous species with a broad ecological niche: it can be found in different freshwater ecosystems, from streams to lakes and reservoirs, from sea level to 1250 m above the sea; therefore any attempt to investigate morphological differences among lineages should take into account the putative effect of phenotypic plasticity.
In this study, I analysed 202 S. rubilio specimens from 13 sampling sites (12 lotic and 1 lentic freshwater environments), to assess the degree of body shape variation in this species and test the existence of morphological differences associated with different genetic backgrounds, i.e. between lineages, and/or
environmental parameters. In detail geometric morphometrics analysis was performed to evaluate morphological differences in fishes’ body shape, using seventeen landmarks identified on specimens' left side. Genetic diversity was assessed using the mitochondrial Control Region (CR), while differences among sampling sites were defined using measures of seventeen environmental parameters, such as site elevation, stream depth, percentage of pools and vegetation in the investigated watercourse.
Low statistical support was found for the relationship between morphological and genetic differences, anyway lineage B specimens showed deepest body shapes than those belonging to A and C, and no morphological differences were observed between the latter two, despite their high genetic divergence. Conversely, the correlation between morphological and overall environmental parameters was more relevant and revealed some interesting outcomes. Streamlined body shapes were observed in sites scarcely altered by human intervention and with fast water flow, and on the opposite deeper body shapes were observed in canals and one reservoir with slow/still water flow. This is consistent with the “shape-water velocity” patterns observed in other fish species: indeed streamlined body shapes are suited for swimming in high water flow, as they minimize drag; conversely, deep body shapes are fitted for complicated locomotor patterns where water flow is slower. In conclusion, the results suggested morphological diversity in the South European roach is mainly the result of phenotypic plasticity in response to hydrodynamic patterns rather than of different genetic backgrounds, and in particular that water velocity may be the major driver in determining morphological responses. Knowledge about S. rubilio intraspecific morphological variation will be useful to predict the effects of habitat alterations (e.g. water abstraction and climatic changes) on lineages’ and populations' survival and thus for their sustainable management and conservation.
Chapter 3 - Microsatellite polymorphism in Sarmarutilus rubilio: insights into the complex phylogeographic history of an Italian endemic freshwater fish
In freshwater fish species microsatellite loci are very useful in disentangling genetic structure. Due to their high mutation rate, they can provide insights about recent isolation or contemporary gene flow between populations and complement mitochondrial investigations, which generally better depict historical
processes. Phylogeographic investigations in Sarmarutilus rubilio, based on mitochondrial markers analysis, revealed that three different lineages originated in allopatry during the Pleistocene and that they currently coexist in some sites, both in northernmost (Magra-Vara basin) and southernmost (South Italy) areas of this species’ native range. Moreover, open questions remained, i.e. whether the observed coexistence was the result of a past temporary connection between basins, or the consequence of human-mediated translocation of specimens in the last centuries, and whether further genetic subdivision exists within lineages for populations belonging to different basins. Therefore I further investigated the genetic structure of S. rubilio using 9 microsatellite loci in the same populations previously characterized for the mitochondrial markers. The results revealed strong correlation with previously observed mitochondrial genetic structure, and also highlighted differentiation between populations where two lineages coexist and the others, suggesting their independent evolution after the occurred temporary connection between basins. The strongest differentiation was observed for populations from a restricted area between Central and South Italy (i.e. Fondi plain) where lineage B was fixed, likely due to their long-term isolation. Alleles strictly associated with lineage B were found, but none was conversely observed for lineage C. Thus its putative cryptic species status, highlighted by previous phylogeographic investigation, was not supported. As expected further genetic substructures were revealed also at the interbasin level within lineage A; noteworthy the lack of differentiation between the Tiber River (in the Tyrrhenian slope) and Tronto River (in the Adriatic one) isolated by the Apennine chain could be likely due to the translocations of specimens.
Finally, microsatellite investigations highlighted the genetic uniqueness of populations in the Magra-Vara basin and Fondi plain, which have limited geographic extension, and therefore they may be more susceptible to habitat alterations due to climate change or human activities, or to other threats such as the introduction of allochthonous species. It is highly recommended to consider these populations as distinct Management Units and prioritize them in conservation actions required by the European Habitats Directive for S. rubilio.