Fishes formerly included in the genus Chondrostoma (Cyprinidae: Leuciscinae) are
distributed through South and Central Europe, from the Atlantic to the Caspian and from the
Mediterranean to the Baltic. They are also found in Asia Minor, the Caucasus and
Mesopotamia.
The number of species included in the group has varied according with the authors due
to the use of different diagnostic criteria (i.e. morphological and osteological or molecular).
Previous molecular phylogenetic studies of this group yielded unresolved polytomies
although some monophyletic and statistically well supported clades were identified:
toxostoma, lemmingii, polylepis, arcasii, nasus, soetta e genei.
The present thesis aimed to achieve several objectives at various levels. At a
macroevolutionary scale this study attempted to solve the phylogeny of the fish traditionally
included in Chondrostoma with a broader sample of taxa and more DNA fragments. At the
same time the study aimed to clarify the diversification of this group in the Iberian Peninsula.
The main results can be summarized as follows.
A statistically robust phylogeny of the fish formerly included in the genus
Chondrostoma was obtained. The former polytomies were solved which indicates that they
were likely due to insufficient taxon sampling or scarcity of molecular data.
Mapping of morphological characters on the inferred phylogeny showed that several
traits considered to be diagnostic of the genus were homoplasic. Thus a revison of the genus
Chondrostoma was undertaken. This revision restricted the genus Chondrostoma to the nasus
lineage and gave rise to the new genera Pseudochondrostoma, Parachondrostoma,
Iberochondrostoma, Achondrostoma and Protochondrostoma (corresponding to the lineages
polylepis, toxostoma, lemmingii, arcasii and genei, respectively).
The cladogenetic events that gave rise to these lineages seem to have occurred 11
million years ago, excluding the hypothesis of diffusion and diversification during the
oligohaline Lago Mare phase of the Mediterranean, near the end of the Messinian. Thus the
model of dispersal proposed by Bianco does not hold for this group of fish.
Summary
XIThree of the six genera defined in this work are endemic of the Iberian Peninsula and
another one has the majority of its species in the peninsula, with a little extension to France.
These data and the chronology estimated for the diversification inside these genera suggest
that most of the radiation in this group occurred in the Iberian Peninsula long before the end
of the Miocene. It seems that in this group the contacts between the peninsula and the rest of
Europe were, in recent times, very scarce, and limited to connections between northeast Spain
and France.
In the Iberian Peninsula the genera Achondrostoma, Iberochondrostoma and
Parachondrostoma have disjunct but adjacent distribution areas which suggests that vicariant
processes may have played a fundamental role in their differentiation. Pseudochondrostoma
shows several resemblances with the genera Parachondrostoma and Chondrostoma (e.g.
inferior mouth with a horny blade, comparably large size and pre-reproductive migrations)
with the genera Parachondrostoma and Chondrostoma. The hypotheses of these
resemblances being the result of convergent evolution or ancient hybridization events are
discussed.
The phylogenetic analysis of the genus Achondrostoma showed that this genus
includes two lineages separated since the Miocene. A. arcasii is polyphyletic and includes fish
from the two lineages referred above, thus a revision of its taxonomy is urgently needed. This
phylogenetic information, combined with morphological data, allowed the identification of a
new species in the southwest of the distribution area of the genus Achondrostoma,
Achondrostoma occidentale. This new species is endemic of the Lisbon district and it is
considered Critically In Danger.
As the name macrolepidotum was unavailable it was necessary to rename
Achondrostoma macrolepidotum to A. oligolepis.
The analysis of the deep phylogeography of the genus Iberochondrostoma resulted in
the proposal of a model of speciation in which a large central species, Iberochondrostoma
lemmingii, originated in its periphery and at different geological times, diverse species with
small distribution areas. This peripatric speciation model is supported by mitochondrial and
nuclear DNA and it seems consistent with the geological history of the Iberian Peninsula in
the Tertiary.
The phylogeographic analysis of the populations of Iberochondrostoma lusitanicum
allowed the identification of distinct ESUs, which in turn impose the description of a new
species, as soon as possible, as well as revealing important information to the conservation
strategies of this Critically in Danger species.
Summary XIIThe studies on the reproductive behaviour of I lusitanicum, and on the agonistic
behaviour of Pseudochondrostoma polylepis revealed some ethological patterns relevant to
the conservation of the species. It was also the case with the study on the expansion of
Alburnus alburnus in the Iberian Peninsula, a species which has spread very quickly in recent
years.
The development of primers that allowed the amplification of a fragment of more than
900 base pairs of the nuclear beta actin gene was fundamental to almost all genetic studies
referred above. Beside its contribution to these studies, it allowed the recognition of the
paternal ancestor of Squalius alburnoides an hybridogenetic species which resulted from
crossings between S. pyrenaicus females and males phylogenetically very close to but distinct
from Anaecypris hispanica.
The fact that species of the genera Alburnus and Squalius hybridize easily and the
phylogenetic proximity between Alburnus, Anaecypris and one of the lineages that integrates
S. alburnoides, brings great concerns about the expansion of Alburnus alburnus. The high
potential of hybridization recorded between this last species and fishes of the genus Squalius
is well known and may result in the genetic descaracterization of several species endemic to
the Iberian Peninsula.
Finally, a new method was developed and validated, that allows the attribution of
different bases present in the DNA chromatograms of diploid or polyploidy fishes to each one
of the constitutive chains, taking advantage of artefacts of the sequencing process induced in
the vicinity of heterozygous indels. This method is useful in the analysis of multiple SNP’s in
the same fragment, in the identification of DNA sequences present in hybrids and in the
distinction of several kinds of polyploids.