Review Species list of the European herpetofauna - 2020 update by the Taxonomic Committee of the Societas Europaea Herpetologica
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Amphibia-Reptilia 41 (2020): 139-189 brill.com/amre
Review
Species list of the European herpetofauna – 2020 update by the
Taxonomic Committee of the Societas Europaea Herpetologica
Jeroen Speybroeck1,∗ , Wouter Beukema2 , Christophe Dufresnes3 , Uwe Fritz4 , Daniel Jablonski5 ,
Petros Lymberakis6 , Iñigo Martínez-Solano7 , Edoardo Razzetti8 , Melita Vamberger4 ,
Miguel Vences9 , Judit Vörös10 , Pierre-André Crochet11
Abstract. The last species list of the European herpetofauna was published by Speybroeck, Beukema and Crochet (2010). In
the meantime, ongoing research led to numerous taxonomic changes, including the discovery of new species-level lineages as
well as reclassifications at genus level, requiring significant changes to this list. As of 2019, a new Taxonomic Committee was
established as an official entity within the European Herpetological Society, Societas Europaea Herpetologica (SEH). Twelve
members from nine European countries reviewed, discussed and voted on recent taxonomic research on a case-by-case basis.
Accepted changes led to critical compilation of a new species list, which is hereby presented and discussed. According to
our list, 301 species (95 amphibians, 15 chelonians, including six species of sea turtles, and 191 squamates) occur within our
expanded geographical definition of Europe. The list includes 14 non-native species (three amphibians, one chelonian, and
ten squamates).
Keywords: Amphibia, amphibians, Europe, reptiles, Reptilia, taxonomy, updated species list.
Introduction
1 - Research Institute for Nature and Forest, Havenlaan 88 Speybroeck, Beukema and Crochet (2010)
bus 73, 1000 Brussel, Belgium
2 - Wildlife Health Ghent, Department of Pathology,
(SBC2010, hereafter) provided an annotated
Bacteriology and Avian Diseases, Ghent University, species list for the European amphibians and
Salisburylaan 133, 9820 Merelbeke, Belgium non-avian reptiles. A decade later, a sizable
3 - LASER, College of Biology and the Environment,
amount of new research has been produced,
Nanjing Forestry University, Nanjing, China
4 - Museum of Zoology, Senckenberg Dresden, A.B. fuelling the need for a contemporary update.
Meyer Building, Königsbrücker Landstraße 159, Within the European Herpetological Society
01109 Dresden, Germany (Societas Europaea Herpetologica; SEH) and by
5 - Department of Zoology, Comenius University in
Bratislava, Ilkovičova 6, Mlynská dolina, 842 15 invitation of the SEH Council, a newly com-
Bratislava, Slovakia posed Taxonomic Committee (SEH TC, or fur-
6 - Natural History Museum of Crete, University of Crete, ther TC) was formed in early 2019, and its
Knossou Ave. 71409, Crete, Irakleio, Greece
7 - Museo Nacional de Ciencias Naturales (MNCN-
chair was approved by SEH membership during
CSIC), c/ José Gutiérrez Abascal, 2, 28006 Madrid, the Ordinary General Meeting held in Milan,
Spain
8 - Kosmos – Museo di Storia Naturale, Università di
Pavia, Piazza Botta 9, 27100 Pavia, Italy
11 - CEFE, Université Montpellier, CNRS, EPHE, IRD,
9 - Division of Evolutionary Biology, Zoological In-
stitute, Braunschweig University of Technology, Université Paul Valéry Montpellier 3, Montpellier,
Mendelssohnstr. 4, 38106 Braunschweig, Germany France
∗ Corresponding author;
10 - Department of Zoology, Hungarian Natural History
Museum, 1088 Budapest, Baross u. 13, Hungary e-mail: jeroen.speybroeck@inbo.be
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
© Speybroeck et al., 2020. via free access
DOI:10.1163/15685381-bja10010
This is an open access article distributed under the terms of the CC-BY 4.0 License.140 J. Speybroeck et al.
September 5th 2019 (SEH News, Amphibia- Venchi and Grieco (2013). Together, these four
Reptilia 40: 551-559). sources led to a starting point species list. In the
We did not define our own limits for the geo- following, we only discuss taxonomic changes
graphical area considered here, but adopted the which deviate from this species list. We decided
limits defined by previous projects. Our goal against using online databases as starting points,
was to provide a taxonomic reference for the fu- as they are changing constantly, with historical
ture mapping projects of the SEH. Therefore, versions not remaining reliably and easily avail-
we included all areas that were part of pre- able.
vious European atlas projects (cf. Gasc et al., The taxonomic decisions adopted here are
1997; Sillero et al., 2014). We also aimed at in- not necessarily supported by all authors of this
forming the taxonomic backbone of the Fauna work. According to TC guidelines, a change to
Europaea initiative (https://fauna-eu.org), and the starting list will only be adopted if widely
therefore our geographical area also includes all supported among its members, specifically by
territories that are covered by Fauna Europaea. a >75% majority. When a change is recom-
As a result, we enlarged the geographical area mended by a large majority of the TC mem-
considered by SBC2010 to encompass all areas bers, but different members favour different
covered by both Gasc et al. (1997) and Fauna outcomes, the adopted solution may be sup-
Europaea. Areas included by Gasc et al. (1997), ported by only a simple majority (>50%). Note
but not by Speybroeck, Crochet and Beukema that this process favours taxonomic stability,
(2010), include the northern versant of the Cau- with changes requiring large support among TC
casus (including north-eastern Azerbaijan), all members to become accepted.
areas west of the Ural River (including western- TC members do not necessarily adhere to
most Kazakhstan) and west of the Ural Moun- the same species concept. While many agree
tains, and the Yekaterinburg Region. Areas in- with the General Lineage Concept (GLC) of De
cluded in Fauna Europaea, but not by Gasc et Queiroz (2007), some prefer the general frame-
al. (1997) or SBC2010, are Macaronesia (with- work of the Biological Species Concept. How-
out Cape Verde), the Greek Islands off the west- ever, all agree on using reproductive isolation as
ern Anatolian shore, and Cyprus. As such, our the primary operational criterion for the delimi-
area exceeds that of the most recent European tation of species. The majority of TC members
atlas (Sillero et al., 2014) by including Mac- is of the opinion that, while every species is a
aronesia, all Greek islands, and parts of Azer- lineage, not every lineage is a species. The com-
baijan and Kazakhstan (fig. 1). A Google Earth mon approach can be defined as either following
.kml file with the limits of the area is provided the Biological Species Concept framework, or
in the supplementary material. For the rationale as applying a Biological Species Criterion under
of these limits, we refer to Gasc et al. (1997) and the GLC. More specifically, TC members ad-
https://fauna-eu.org/data-handling). here to a “soft” version of the reproductive iso-
Upon enlarging the scope area and prior to lation criterion. As such, we allow extensive in-
discussing any taxonomic changes, a broader trogression between recognised species, as long
baseline list had to be set for species occur- as there are intrinsic barriers to gene flow that
ring outside the area considered by SBC2010. prevent wide-reaching introgression beyond the
As such, in addition to SBC2010, we followed contact zones. Even in the absence of geograph-
the taxonomy of Gasc et al. (1997, 2004, includ- ical barriers, a sufficient level of reproductive
ing the changes adopted by Dubois and Cro- isolation has to exist in order to ensure long-
chet in the 2004 reprint), and for species outside term persistence of the diverged lineages. Taxa
Europe (including Macaronesia and Cyprus), connected by bimodal or trimodal hybrid zones
Sindaco and Jeremčenko (2008) and Sindaco, (Gay et al., 2008) were unanimously treated as
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna
Figure 1. Extent of geographic area (Mollweide projection). Herpetofauna species within this area are dealt with. Shading indicates areas not included by Sillero et al. (2014).
141
via free access
Heruntergeladen von Brill.com05/14/2020 03:08:24PM142 J. Speybroeck et al.
valid species, but opinions often differed re- reptiles. For each case, we provide the rationale
garding how much introgression was “allowed” underlying the respective decision of the TC,
across unimodal hybrid zones, reflecting differ- and conclude by providing an updated species
ent opinions relative to where to cut the grey list of the European herpetofauna.
zone of speciation, how much reproductive iso-
lation is necessary and when to treat incipient
species as valid species. For allopatric taxa, or Amphibia
when contact zones were not studied, lineages
Caudata/Urodela
that had divergence levels similar to closely re-
lated, unambiguously distinct species were ac- A series of phylogenetic studies on mitochon-
cepted as species. As an auxiliary criterion, we drial DNA, allozymes, and nuclear DNA se-
sometimes use monophyly, even though we do quences of members of the family Salamandri-
not consider it as a necessary requirement for dae (Litvinchuk et al., 2005; Weisrock et al.,
species status. 2006; Zhang et al., 2008; Kieren et al., 2018;
For supraspecific classification, the TC Veith et al., 2018) confirmed that the newt genus
agreed to accept only monophyletic units. Triturus sensu lato, as traditionally recognised,
This causes issues regarding the class Rep- is not monophyletic. Litvinchuk et al. (2005)
tilia, which in its traditional definition is pa- proposed the separation of Triturus into four
raphyletic through the exclusion of birds. All genera, among which the new genus Omma-
current hypotheses on the evolution of verte- totriton contains (the former) Triturus vitta-
brates agree that a group including squamates, tus. SBC2010 accepted this new arrangement,
turtles, Sphenodon and crocodiles, but not birds but did not explicitly acknowledge the need to
is paraphyletic (see e.g. Chiari et al., 2012a; recognise Ommatotriton, as the new genus does
Hasegawa, 2017). As a consequence, most cur- not occur in the area they considered. As we
rent classifications of Vertebrata include Aves in herein consider a wider area, we formally ac-
the class Reptilia (see e.g. Modesto and Ander- cept Ommatotriton as a separate genus.
son, 2004; Ruggiero et al., 2015). To avoid con- Litvinchuk et al. (2005) also showed that
fusion, we adopt the term ‘non-avian reptiles’ to morphology (number of trunk vertebrae, colour
refer to the components of the European fauna pattern), genome size, and allozymes (Nei’s
assigned to Testudines/Chelonii and Squamata. genetic distances of 0.44-0.83) strongly dif-
For nomenclatural decisions, including fer between populations in the two widely dis-
spelling, we followed the International Code joint areas inhabited by Ommatotriton vittatus,
of Zoological Nomenclature (the Code here- namely south-eastern Anatolia and the Levant
after, ICZN (1999 and subsequent changes), (O. v. vittatus) versus the southern and eastern
see https://www.iczn.org/). Such decisions were Black Sea and western Caucasus regions (O. v.
generally not submitted to voting, but they could ophryticus). This led them to elevate ophryticus
be discussed, as many parts of the Code can be to species level. A more comprehensive study
subject to interpretation, and many actual cases also demonstrated restricted introgression be-
can be open to different decisions, even under tween an eastern and a western taxon along the
the rules of the Code. Black Sea coast of Turkey (van Riemsdijk et al.,
In the following, we review taxonomic and 2017). Consequently, we recommend to recog-
nomenclatural changes proposed since the pub- nise three species in the genus Ommatotriton:
lication of the four literature sources that we O. vittatus (extralimital) in south-eastern Anato-
used to build our starting point, as well as other lia and the Levant, O. ophryticus in Russia, the
relevant new information pertaining the taxon- Caucasus and northern Anatolia west to the re-
omy of European amphibians and non-avian gion of Samsun, and O. nesterovi (extralimital)
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna 143
in Anatolia from Samsun to the Sea of Marmara. result of ancient admixture between T. mace-
The populations inhabiting the native Omma- donicus and T. ivanbureschi. We do not entirely
totriton range in the European part of Russia be- agree with their interpretation of the Code. As
long to O. ophryticus, while the introduced (and they rightfully state, nomina based on geneti-
persisting anno 2019; Speybroeck, pers. obs.) cally admixed individuals derived through sev-
population in north-eastern Spain is of mixed eral generations of backcrossing are indeed left
ancestry, with genetic contribution from both O. in limbo in the Code. Yet, we do not agree that
ophryticus and O. nesterovi (van Riemsdijk et the Code should be interpreted as extending pro-
al., 2018). visions of Art. 23.8 to nomina based on genet-
In addition to mtDNA sequences, Vences et ically admixed individuals, at least not without
al. (2014) used sequences from thirteen nuclear clear guidelines of what constitutes admixture
loci to improve our understanding of the his- in the sense of the Code. We do, however, agree
tory of the genus Salamandra. Their nuclear with Wielstra and Arntzen (2014) that arntzeni,
data suggest that Salamandra salamandra con- based on individuals that carry less than 50%
tains several deeply divergent lineages whose of alleles derived from the karelinii-complex
monophyly relative to S. algira received weak taxon, should not be used as the valid name for
support, although this requires confirmation. that taxon. We thus accept T. ivanbureschi as
Based on mitochondrial data the subspecies lon- the valid nomen of the taxon of the karelinii-
girostris is sister to all other S. salamandra lin- complex that occurs in Europe in the Balkans.
eages. However, nuclear genes place it with the With the extension of the geographic range of
subspecies morenica. The discordance between SBC2010, T. karelinii sensu stricto also occurs
mtDNA and nuclear genes may result from past in our area in Russia, the Crimean Peninsula and
introgression and admixture processes. Further- the Caucasus region.
more, a clade comprising the subspecies fas- Using mitochondrial and nuclear DNA se-
tuosa, bernardezi (including populations at- quences, Sotiropoulos et al. (2007) and Recuero
tributed to the subspecies alfredschmidti, whose et al. (2014) investigated the phylogenetic struc-
validity was rejected, as it is phylogenetically ture of Ichthyosaura alpestris. They identified
nested within several subclades of bernardezi; several deeply divergent lineages that largely
Beukema et al., 2016) and gigliolii is recov- correspond to currently recognised subspecies,
ered with strong support. While additional data except for the subspecies alpestris, which is
are clearly needed, this suggests a high amount further divided into a western and an eastern
of evolutionary divergence within S. salaman- main lineage. Interestingly, nuclear sequences
dra, and the existence of more than one species from two loci show a nearly complete lack of
within S. salamandra cannot yet be fully ruled allele sharing between the eastern and west-
out. ern clades, suggesting reproductive isolation be-
SBC2010 recognised two species within the tween them. However, no samples were col-
former Triturus karelinii: T. arntzeni from the lected close to their contact zones. As the east-
Balkan Peninsula, and T. karelinii. Wielstra et ern and western clades should meet in the
al. (2013) suggested that the type specimens of Balkans and the southern Carpathian Moun-
T. arntzeni are in fact T. macedonicus, which led tains, we follow the recommendation of Re-
them to place T. arntzeni in the synonymy of cuero et al. (2014) and await more data on
T. macedonicus, and to create the name ivanbu- their level of reproductive isolation before mak-
reschi for the taxon of the karelinii complex that ing any taxonomic change. Thus, we do not
occurs in the Balkans and in Western Anato- follow Raffaëlli (2018) who (based on the ge-
lia. Later, Wielstra and Arntzen (2014) demon- netic data presented in the aforementioned stud-
strated that the types of T. arntzeni are in fact the ies) elevated the subspecies apuana, reiseri and
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access144 J. Speybroeck et al.
veluchiensis to species level, and we note that schmidtleri, its range limits and contact zones
such changes would make I. alpestris poly- with other European lineages remain poorly
phyletic. We thus retain I. alpestris as a single known. Although data on the contact zone with
species for now. L. v. kosswigi in Anatolia support a species-level
Pabijan et al. (2017) reconstructed phyloge- divergence between L. v. schmidtleri and the
netic relationships within the Lissotriton vul- graecus-kosswigi clade, areas of 100-300 km
garis species complex and inferred patterns of inhabited by L. vulgaris s.l. of unknown iden-
(historical) gene flow, using 74 nuclear DNA tity separate genotyped L. v. schmidtleri popu-
markers from one individual from each of 127 lations from those of L. v. vulgaris and L. grae-
locations. Five highly divergent lineages were cus (Pabijan et al., 2017; Wielstra et al., 2018).
identified within our focal area: Lissotriton The TC thus feels that more information on con-
montandoni and four lineages corresponding tact zones between L. v. schmidtleri and L. v.
to the Lissotriton vulgaris subspecies graecus, vulgaris is warranted before the species rank of
lantzi, schmidtleri and vulgaris. In spite of clear schmidtleri can be accepted. Finally, L. v. lantzi
evidence of past historical introgression, their is endemic to the Caucasus and shows an al-
contemporary gene flow is restricted. Therefore, lopatric distribution (Wielstra et al., 2018). Mi-
the authors proposed to treat the three former tochondrial data suggests that L. v. lantzi was
subspecies as species (Pabijan et al., 2017). Be- the first to diverge from other lineages around
tween L. v. vulgaris and the morphologically 3.39 (1.42-5.37) Mya (Pabijan et al., 2015). Yet,
diverged L. v. meridionalis regular episodes of nuclear data place L. v. lantzi as sister to L. v.
gene flow were identified, thus meridionalis schmidtleri and other eastern and central Eu-
was retained at subspecies level. While L. v. ropean Lissotriton, while suggesting that diver-
graecus is easily distinguished from other Eu- gence within this group initiated with the split
ropean populations by male nuptial character- of the graecus-kosswigi clade (Pabijan et al.,
istics, both L. v. lantzi and L. v. schmidtleri 2017). In conclusion, the TC recommends to
are morphologically cryptic in respect to L. v. treat Lissotriton graecus as a valid species, but,
vulgaris (Raxworthy, 1990). Rather than be- as awarding species status to lantzi but not to
ing confined to Anatolia, Pabijan et al. (2015, schmidtleri may render L. vulgaris paraphyletic,
2017) showed that L. v. schmidtleri also occurs we prefer to maintain the other taxa, including
in Greek and Turkish Thrace and on a number L. v. schmidtleri and L. v. lantzi, as subspecies
of Greek islands. Because the sampling gaps be- of L. vulgaris for the time being.
tween several of the lineages remained wide, A phylogeographic study based on two
the TC has been reluctant to accept all syste- mtDNA markers by Martínez-Solano et al.
matic conclusions of Pabijan et al. (2017). The (2006) revealed two major mitochondrial lin-
taxon graecus comes into close contact with L. eages of Miocene origin in Lissotriton boscai.
v. vulgaris along the northern and eastern bor- One of them is restricted to central and south-
ders of its distribution. Although no dense sam- western coastal Portugal, while the other oc-
pling has been performed to delineate the con- cupies the remainder of the species range, in-
tact zone in detail, and mitochondrial introgres- cluding the type locality of L. boscai. Martínez-
sion from southern L. v. vulgaris into graecus Solano et al. (2006) acknowledged that the
occurs at its northern range border (Pabijan et two lineages might represent cryptic species,
al., 2017), nuclear gene pools of these two taxa but called for additional morphological and
appear to remain reciprocally distinct in rela- molecular studies, including data on variation
tive close geographical proximity (Pabijan et al., in nuclear DNA markers. Dubois and Raffaelli
2017; Wielstra et al., 2018). We therefore accept (2009) resurrected the nomen Triton maltzani
Lissotriton graecus as a valid species. For L. v. Boettger, 1879 for the southwestern lineage in
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna 145
the new combination Lissotriton maltzani. Ac- Vörös, Ursenbacher and Jelić (2019) used 10
cording to these authors, L. maltzani can be dis- microsatellite loci to investigate patterns of dif-
tinguished from L. boscai by its smaller size ferentiation between four Croatian cave popu-
and by its paler dorsal coloration, especially in lations of Proteus anguinus. They uncovered
females, with less distinct dark spots. Teixeira long-lasting isolation between caves belong-
et al. (2015) used DNA sequences of one nu- ing to different hydrogeographic systems, with
clear gene and found overall congruence with the most ancient divergence being older than 7
mtDNA genes in terms of sequence divergence Mya. This suggests that some of the evolution-
and geographic structure. They, however, also ary lineages within this species might constitute
revealed wide areas of admixture and evidence cryptic taxa, possibly of species rank (Vörös,
for recombination, suggesting a lack of com- Ursenbacher and Jelić, 2019).
plete reproductive isolation and the presence of
incomplete speciation. In a recent study, Se- Anura
queira et al. (2020) used ten microsatellites, one Alytes obstetricans is composed of four sub-
mtDNA gene and two single copy nuclear DNA species: A. o. obstetricans, A. o. pertinax, A. o.
markers in a cline analysis framework to inves- boscai and A. o. almogavarii. The latter taxon
tigate one of the hybrid zones between L. boscai is endemic to Catalonia and adjacent areas in
and L. maltzani. The results show evidence for north-eastern Spain and southern France, and
partial reproductive isolation between L. boscai it is highly differentiated in allozyme (Arntzen
and L. maltzani, with narrow clines (3-28 km) and García-París, 1995; García-París, 1995),
consistent with selection against hybrids. We mitochondrial (Gonçalves et al., 2007, 2015)
thus recognise L. maltzani as a separate species. and microsatellite markers (Maia-Carvalho et
Wake (2012) addressed the taxonomy of al., 2018), and features peculiar osteological
the Plethodontidae, advocating to treat Aty- (Martínez-Solano et al., 2004) and bioacous-
lodes (comprising Speleomantes genei) and tics characters (Márquez and Bosch, 1995). Pos-
Speleomantes (comprising the other European sibly due to ancestral hybridisation or incom-
plethodontid species) as subgenera of a single, plete lineage sorting (Gonçalves et al., 2007;
cross-Atlantic genus Hydromantes. Addressing Maia-Carvalho et al., 2014), the nuclear phy-
the genus name confusion, Wake (2013) com- logeny, based on intron sequences, is not well
pared five potential arrangements, missing how- resolved. Using microsatellites, Maia-Carvalho
ever the arrangement of SBC2010 (i.e. two et al. (2018) identified a distinct cluster cor-
genera without subgenera). Comparing with responding to all A. o. almogavarii popula-
other plethodontid genera, he argued in favour tions, suggesting that the lack of monophyly in
of an arrangement of a single genus with mtDNA data is due to cyto-nuclear discordance.
three subgenera. However, the arguments of This cluster extends as far west as the southern
SBC2010 still stand: the position of Atylodes slopes of the Pyrenees in the north-westernmost
remains unresolved, and the European species parts of Aragon. These authors also found re-
form a well-defined monophyletic group with a stricted genetic admixture between A. o. almo-
large genetic distance from the five Californian gavarii and neighbouring subspecies (A. o. per-
species Hydromantes brunus, H. platycephalus, tinax, A. o. obstetricans). Following up on this
H. samweli, H. shastae, and H. wintu (Nascetti study, Dufresnes and Martínez-Solano (2020)
et al., 1996; Pyron and Wiens, 2011; Chiari targeted the hybrid zone between A. o. almo-
et al., 2012b; Bingham, Papenfuss and Wake, gavarii and A. o. pertinax with genomic analy-
2018). Thus, no change seems in order, and we ses using RADseq-derived markers along a fine-
maintain the European species for the time be- scale transect in Catalonia. They documented
ing in a single taxon, the genus Speleomantes. a very narrow mitochondrial (cline width ca.
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access146 J. Speybroeck et al. 13 km) and nuclear (cline width ca. 16 km) et al., 2019a, b), we accept the specific rank of transition, and detected portions of the genome Pelobates vespertinus. that were completely impermeable to gene flow. Dufresnes et al. (2019a) also demonstrated Given the absence of barriers to dispersal in the complete reproductive isolation between Pelo- contact zone, the authors concluded that these bates syriacus syriacus and P. s. balcanicus, lineages exhibit substantial (even if incomplete) warranting elevation of European populations, reproductive isolation. We adopt their recom- except those from south-eastern Bulgaria, parts mendation and treat Alytes almogavarii as a dis- of European Turkey and a number of eastern tinct species. Greek islands (including Limnos and Lesbos), Pabijan et al. (2012) found a lack of recipro- to species level as Pelobates balcanicus, with an cal monophyly in nuclear data of Iberian painted estimated Mio-Pliocene divergence (>5 Mya) frogs (Discoglossus spp.), while Dufresnes et between both taxa. Deep intraspecific diver- al. (2020a) found numerous introgressed indi- gence within each species further leads to recog- viduals in their RAD datasets and a broad hy- nise the subspecies P. balcanicus chloeae (Pelo- brid zone (average cline width of nuclear mark- ponnese) and P. syriacus boettgeri (all European ers >136 km). These results indicate weak (or parts of the distribution of Pelobates syriacus). no) restriction to gene flow and confirm that The complex pattern of genetic variation D. g. galganoi and D. g. jeanneae are better within Iberian Pelodytes has been known for treated as conspecific, as previously advocated several years (e.g., van de Vliet et al., 2012). It by SBC2010. was taxonomically formalised by Díaz- Borkin et al. (2001) reported differences in Rodríguez et al. (2017), who described two genome size between the morphologically sim- new species: Pelodytes atlanticus from Por- ilar eastern and western populations of Pelo- tugal and P. hespericus from central eastern bates fuscus, with a transition between both Spain, in addition to the previously established groups in north-eastern Ukraine and adjacent species, P. ibericus and P. punctatus. The two parts of Russia. The name Rana vespertina Pal- new lineages show no consistent morphological las, 1771 is available for the eastern taxon. Sub- or bioacoustics differences, are only weakly dif- sequently, analyses using mtDNA data across ferentiated in mtDNA (
Species list of European herpetofauna 147
from hespericus and punctatus is quite large. and most French populations of Bufo bufo (for
We thus follow the suggestion of Dufresnes which the name Bufo spinosus is available),
et al. (2020a) to accept P. atlanticus as sepa- and a clade composed of two lineages, repre-
rate species, while retaining hespericus at sub- senting ii) verrucosissimus from the Caucasus
species level as P. punctatus hespericus. and iii) bufo from northern France to Russia,
Despite growing support for recognition of while populations from Greece, southern Italy
various clades in the family Bufonidae as sepa- and Sicily and most of Anatolia carried bufo
rate genera (Stöck et al., 2006; Van Bocxlaer et mtDNA but grouped with verrucosissimus in
al., 2009), several authors advocated retaining nuclear DNA. Estimations of divergence times
Bufo for all species from the Western Palaearc-
indicated a long evolutionary history of the
tic and Central Asia, at least as an ad interim
group, starting with the split from eichwaldi at
solution (Dubois and Bour, 2010; SBC2010).
about 12 Mya, and the divergence of spinosus
Formal taxonomic action was further hampered
taking place around 6 Mya. The deep level of
by nomenclatural confusion, such as the use
genetic divergence observed between the west-
of either Pseudepidalea or Bufotes as the valid
generic nomen for the green toad group (Frost ern and eastern groups of common toads indi-
et al., 2006; Dubois and Bour, 2010). Dubois cated that these groups may be different species.
and Bour (2010) demonstrated that Pseudep- However, Garcia-Porta et al. (2012) found ex-
idalea is a junior objective synonym of Bu- tensive admixture of mitochondrial lineages be-
fotes, thereby giving priority to the latter. How- tween the eastern and western clades in the
ever, based on the presence of hybridisation Languedoc area of southern France (as also re-
between representatives of these clades, and ported by Arntzen et al., 2017 for the Provence
the restricted sampling size of previous stud- area of southern France), and detected signs of
ies, Dubois and Bour (2010) listed these nom- ancient introgression of bufo allozyme alleles
ina as subgenera, rather than genera. Accord- into spinosus. They thus suggested to treat B. b.
ing to Van Bocxlaer et al. (2010), Pyron and spinosus and B. b. bufo as conspecific pending
Wiens (2011), and Beukema et al. (2013), which studies of the contact zones. Detailed analysis of
together provide a dense sampling of bufonid the amount of reproductive isolation in two geo-
species, most genera currently recognised by graphically distant contact zones on the basis of
Frost (2019) represent well-supported mono- mtDNA, morphology and nuclear markers have
phyletic units. Moreover, time-calibrated analy- subsequently been published by Arntzen et al.
ses showed Bufo, Bufotes and Epidalea to be of
(2016, 2017). Both studies documented an ex-
similar age as or older than most other recog-
tensive amount of hybridisation and introgres-
nised bufonid genera (Beukema et al., 2013).
sion in the contact zone, but also the presence
Considering these results, we accept Bufo, Bu-
of narrow and concordant clines for most nu-
fotes and Epidalea at genus level.
clear markers, resulting in a unimodal, yet nar-
Recuero et al. (2012) published a multilocus
mitochondrial and nuclear DNA sequence data row hybrid zone of ca. 30 km wide, and indi-
set for Bufo bufo and associated species, cover- cating intrinsic barriers to gene flow in spite of
ing the entire documented range and providing incomplete reproductive isolation. Karyological
extensive genetic data. The study yields a fully analysis (albeit based on only four males and
resolved phylogeny, with the recently described a single female) identified heteromorphic sex
Bufo eichwaldi from the Talysh Mountains of chromosomes in spinosus, but not in bufo (Sko-
southern Azerbaijan and northern Iran as the rinov et al., 2018). The combined available ev-
sister taxon of a clade including three deeply idence justifies treating Bufo spinosus as a dis-
diverged lineages: i) north African, Iberian, tinct species.
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access148 J. Speybroeck et al.
Based on differences in morphology and Stöck et al., 2006) and demonstrated strong re-
karyotype and a study suggesting lowered fertil- productive isolation between them. As these
ity of F1 hybrids (summary in Kuzmin, 1999), two taxa belong to the two most divergent lin-
the Caucasian populations of the complex are eages (North African and Eurasian), a two-
sometimes also treated as a distinct species as species split of the green toad complex is clearly
Bufo verrucosissimus. Recent genetic studies warranted. The Sicilian taxon siculus is closely
paint a more complex picture. Firstly, based on related to the North African taxon boulengeri,
nuclear markers, the Anatolian populations, tra- which has priority over siculus. Subsequently,
ditionally excluded from B. b. verrucosissimus nuclear data were added to the picture (mi-
on the basis of morphology, are in fact geneti- crosatellites – Dufresnes et al., 2018a; Gerchen,
cally closer to this group than to B. b. bufo or B. Dufresnes and Stöck, 2018, and RADseq mark-
spinosus, even if they carry B. b. bufo mtDNA ers – Dufresnes et al., 2019c). Although cyto-
(Garcia-Porta et al., 2012; Arntzen et al., 2013). nuclear discordance was present in large areas
Secondly, mtDNA divergence between Cau- of Europe where populations of viridis carry
casian (B. b. verrucosissimus) and Eastern Eu- the variabilis mtDNA lineage, these studies
ropean (B. b. bufo) populations is much lower confirmed that overall the previously identified
than between B. b. bufo and B. spinosus. In- mtDNA lineages correspond to distinct evolu-
tionary units. Dufresnes et al. (2019c) also dis-
deed, divergence in mtDNA and allozymes be-
covered that the type locality of variabilis is in-
tween the Caucasian and Eastern European lin-
habited by the western lineage, rendering vari-
eages is even lower than between the subclades
abilis a junior synonym of viridis, and that the
of B. spinosus (Garcia-Porta et al., 2012). Most
valid name for the Anatolian lineage is sitibun-
importantly, based on frequencies of allozyme
dus. These two lineages (B. v. sitibundus and B.
alleles, populations of B. b. bufo from Greece
v. viridis) widely admix over a very large geo-
and north-western Turkey appear intermediate
graphic area in Anatolia and Russia (Dufresnes
between Caucasian and European populations
et al., 2019c), suggesting that they are best
(Garcia-Porta et al., 2012) or group with Cau-
treated as subspecies. The contact zone between
casian populations in nuclear trees (Recuero et
B. v. viridis and B. v. balearicus is narrower
al., 2012), suggesting extensive introgression
(63 km), but still extensive if compared with
between B. b. verrucosissimus and B. b. bufo in other contact zones between taxa that are here
north-western Turkey and the southern Balkans treated as different species (Gerchen, Dufresnes
(but see Arntzen et al., 2013). In our opinion, and Stöck, 2018). In addition, raising B. v.
the specific status of B. b. verrucosissimus is balearicus but not B. v. sitibundus to species
currently insufficiently supported by the avail- rank would make B. v. viridis paraphyletic (see
able evidence, and we maintain it as conspecific Dufresnes et al., 2019c). The TC therefore pre-
with Bufo bufo for the time being. ferred to treat B. v. sitibundus and B. v. baleari-
Stöck et al. (2006) elevated a number of mito- cus (and the extralimital B. v. perrini) as sub-
chondrial lineages of green toads (now Bufotes) species of B. viridis.
to species level (for Europe: B. variabilis and Dufresnes et al.’s (2019c) range-wide sam-
B. balearicus, in addition to B. viridis) and sub- pling of all known taxa of the Bufotes viridis
sequently described a new species from Sicily complex identified a total of eight diploid ge-
based on the same lines of evidence. SBC2010 netic clusters, including a new lineage endemic
argued against accepting these species on the to Cyprus, which they describe as Bufotes cy-
basis of mitochondrial DNA alone. Colliard et priensis. Estimated to be of Messinian origin
al. (2010) documented a contact zone in Sicily (5.3 Mya), this insular taxon is as old as or
between the taxa siculus and balearicus (sensu older than many species of anuran amphibians
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna 149
from the Western Palearctic. Yet, all ten indi- of other supporting characters (weak acous-
viduals from four localities carry mtDNA hap- tic divergence: see Schneider, 1974; no well-
lotypes of B. v. sitibundus. Secondary contact supported morphological characters), SBC2010
and hybridisation with the Anatolian mainland refrained from formally accepting this syste-
lineage during the Pleistocene would explain matic treatment, as it would have rested solely
this mitochondrial capture, as well as the rem- on mitochondrial DNA data. Several more re-
nants of nuclear introgression that were detected cent studies offer a better understanding of
as well. In addition to this deep genomic di- the phylogeny of this group and of the pat-
vergence, the genome of B. cypriensis is also terns of gene flow across contact zones. Firstly,
significantly larger than that of all other West- Dufresnes et al. (2015, 2016) investigated the
ern Palearctic Bufotes taxa. Morphologically, arborea-orientalis contact zones in Poland and
the Cyprus green toads are smaller than those in the Balkans with microsatellite markers. In
from mainland populations, as previously re- Poland, they found evidence of admixture over a
ported by Stugren and Tassoula (1987). The 200 km wide zone, with mosaic contacts and in-
combination of an old divergence and a larger terspersed hybrid populations, but with strongly
genome size led the TC to accept the species sta- restricted introgression at sex-linked loci and
tus of B. cypriensis. Adriatic populations could many populations of seemingly pure ancestry
in close contact to each other (Dufresnes et al.,
represent a valid subspecies of B. viridis (for
2016). In the Balkans, Dufresnes et al. (2015)
which the nomen longipes Fitzinger in Bona-
found narrow clines (30 and 32 km of aver-
parte, 1840 might be available), while green
age cline width in Serbia and Greece respec-
toads from Naxos (Cyclades) and Crete dif-
tively). In line with our treatment of taxa con-
fer genetically from other populations and de-
nected by unimodal hybrid zones with narrow
serve genomic investigations (Dufresnes et al.,
clines, we recommend affording species status
2019c). In summary, for Europe, we accept (1)
to Hyla orientalis. Secondly, the phylogenomic
Bufotes viridis, including the widespread B. v.
tree in Dufresnes et al. (2018b) unambiguously
viridis, B. v. balearicus (Italian Peninsula, Cor-
groups orientalis and molleri as sister taxa, with
sica, Sardinia, Sicily, Balearic islands) and B. v.
arborea splitting from a node basal to the mol-
sitibundus (within Europe restricted to a num- leri – orientalis divergence. Therefore, we ac-
ber of eastern Greek islands), (2) the Sicilian cept molleri at species rank as well. Although
Bufotes boulengeri siculus and (3) Bufotes cy- the close relationships between molleri and ori-
priensis. We note that the availability of the entalis could justify to treat them as conspecific,
nomen siculus remains unclear, given that the we refrain for the time being from this arrange-
corresponding description (Stöck et al., 2008) ment. We thus treat Hyla molleri and Hyla ori-
was only issued electronically and prior to 2011 entalis as valid species.
(see Article 8.5 of the Code): whether “numer- Supported by mtDNA and genomic data, two
ous identical and durable copies” (Article 8.1) major cryptic lineages reside within Hyla inter-
were registered by the authors in parallel is yet media (Canestrelli, Verardi and Nascetti, 2007;
to be addressed (see also Dubois et al., 2013). Stöck et al., 2008; Dufresnes et al., 2018b).
Based on a high level of divergence in mi- They are not known to differ in any diagnos-
tochondrial DNA sequences of Hyla tree frogs, tic morphological or acoustic character, even
Stöck et al. (2008) have suggested to recog- if they differ in the averages of one acoustic
nise the Iberian taxon molleri and the eastern and some morphometric traits. The northern lin-
European taxon orientalis as species distinct eage occupies the Po Plain and adjacent re-
from Hyla arborea. However, given the lack gions (including Ticino in Switzerland), with
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access150 J. Speybroeck et al.
the northern Apennines acting as the biogeo- 2013b, 2017). In a phylogeographic survey us-
graphical barrier separating it from the south- ing genome-wide data, Dufresnes et al. (2020b)
ern lineage. As the name intermedia applies to found one of these mitochondrial groups to be a
the southern lineage, Dufresnes et al. (2018b, “ghost lineage”, not differentiated in the nuclear
c) coined a new name, Hyla perrini, for the genome. The two main lineages are strongly dif-
northern lineage. Note that, as its registration ferentiated and are estimated to have diverged
in the Official Register of Zoological Nomen- around 4 Mya. Despite the absence of geo-
clature (ZooBank) occurred later than the pub- graphic or ecological barriers to dispersal, they
lication of Dufresnes et al. (2018b), the name
form a narrow hybrid zone (25 km) in the east-
perrini is not made available by Dufresnes et al.
ern Cantabrian Mountains. Because partial re-
(2018b), but by Dufresnes et al. (2018c), who
productive isolation is thus likely to prevent
met all conditions of nomenclatural availabil-
these two taxa from merging, we follow the
ity. Detailed analyses of the contact zone based
recommendation of Dufresnes et al. (2020b) to
on genomic (RADseq) data by Dufresnes et al.
treat R. parvipalmata as a distinct species.
(2018b) revealed broad clines (96 km of average
cline width) for nuclear markers and detectable Using allozymes and mtDNA, a detailed
admixture over approximately 130 km. Despite analysis of gene flow patterns across the con-
a relatively high cyt b distance of around 9%, tact zone of the northern Anatolian lineage and
and even though Dufresnes et al. (2018b) ar- the Balkan lineage of the Pelophylax ridibun-
gued that the extent of the contact zone would dus – bedriagae complex in northern Greece re-
be even larger without some form of selection vealed the existence of a wide hybrid zone, with
against hybrids, the TC felt that the observed ex- introgression detectable over more than 200 km
tent of introgression and thus the lack of strong (Hotz et al., 2013). This suggests that the Eu-
reproductive barriers did not unambiguously al- ropean and Anatolian lineages are conspecific
low treatment of perrini at species rank. We thus (contra SBC2010). Since the European lineage
recommend to treat the northern lineage as Hyla splits from a node basal to the diversification of
intermedia perrini. Anatolian lineages with respect to mtDNA ac-
The authorship of the family Ranidae has cording to Plötner et al. (2012), it further sug-
been clarified by Dubois and Bour (2011), at- gests that most lineages of this complex are
tributing it to Batsch, 1796. conspecific, with the possible exception of the
Based on allozyme data (Arano, Esteban and
populations from Syria and Jordan (possibly the
Herrero, 1993; Veith et al., 2002, 2012), two ge-
‘true’ bedriagae, see Plötner et al., 2012). On
netically distinct groups have been recognised
the other hand, Plötner et al. (2010) suggested
within Rana temporaria populations of northern
hybrid breakdown between some of these lin-
Spain. The first was assigned to the subspecies
eages, indicating incipient speciation. While un-
R. t. parvipalmata and is restricted to the west-
doubtedly the last word on this subject has not
ernmost edges of the Iberian distribution (Gali-
cia and Asturias). Frogs from western Galicia been written, the TC felt that it was prema-
feature reduced feet webbing, smaller size and ture to suggest taxonomic changes, especially
lower number of pulses per call (Vences, 1992). as a large taxonomic paper based on genomic
The second group corresponds to the nominal variation patterns is in preparation (G. Mazepa,
taxon R. t. temporaria, which extends through- pers. comm.). The same applies to the newly
out Europe. Mitochondrial data suggested a proposed P. cypriensis, which was found to be
more complex picture, with four deeply differ- placed between the European and the Anato-
entiated lineages in the same area (Vences et al., lian lineages in mtDNA, but grouped with P.
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna 151
cretensis (albeit with very low support) in nu- cannot be explained by a geographic barrier ef-
clear DNA (Plötner et al., 2012). For the mo- fect of this narrow sea strait, because in E. or-
ment, we thus suggest no change to the taxon- bicularis recent gene flow occurred across the
omy adopted in SBC2010 for European Pelo- Strait of Gibraltar (Velo-Antón et al., 2015),
phylax. and possibly across the Adriatic Sea between
southern Italy and the Balkans (Vamberger et
al., 2015). In contrast, Pereira, Teixeira and
Reptilia Velo-Antón (2018), using seven microsatellite
loci, concluded that the Strait of Gibraltar cur-
Testudines/Chelonii rently impedes gene flow between Iberian and
North African pond turtles. However, they show
Based on nuclear genomic ISSR fingerprints that the break between the two clusters corre-
and mtDNA sequences, Fritz et al. (2005) de- sponds to the Central Iberian Mountains and
scribed Emys trinacris as a distinct species en- not the sea strait. Clines of microsatellites for
demic to Sicily and Calabria, albeit the record the trinacris – orbicularis contact zone across
from Calabria was later questioned (Vamberger the Strait of Messina were inferred to be very
et al., 2015). A phylogenetic study using seven wide (247 km) in Vamberger et al. (2015), al-
nuclear genes (Spinks and Shaffer, 2009) found though the significance of this estimate is ham-
E. orbicularis to be paraphyletic with respect to pered by a substantial (150 km) sampling gap
trinacris: E. o. hellenica was sister to a clade between the two taxa. The TC takes therefore
containing trinacris and the remaining lineages a conservative stance and treats trinacris as a
of E. orbicularis. This led SBC2010 to conclude subspecies of E. orbicularis, waiting for further
that trinacris should not be recognised as a studies resolving the complicated relationships
species, which has been criticised by Vamberger of this complex. In this context, we note that
and Fritz (2018). While trinacris is the most Pöschel et al. (2018) reported a sharp transition
basal lineage in mtDNA trees (Fritz et al., 2005, between occidentalis and orbicularis + gal-
2007), this position is weakly supported, and the loitalica in north-eastern Spain, matching that
level of mtDNA divergence between trinacris of distinct species recognised here, suggesting
and the other lineages is not clearly larger than that the E. orbicularis complex could comprise
that between some lineages of E. orbicularis. several species. In the meantime, we recom-
A first combined analysis of genetic differen- mend to maintain the Sicilian lineage as Emys
tiation at eight microsatellite loci and mtDNA orbicularis trinacris.
sequences found concordant patterns for both
markers: no evidence for admixture between Squamata
Sicilian pond terrapins and the remaining lin- The family Agamidae is sometimes credited to
eages was found, while extensive admixture was Fitzinger, 1826 or Gray, 1827 (see e.g. Melville
found between some other lineages of E. orbic- and Smith, 1987). In fact the nomen was first
ularis (Pedall et al., 2011). Using 15 microsatel- published by Spix in 1825 as “Familia Aga-
lite loci and mtDNA sequences, Vamberger et mae”. Even if Spix did not use the correct suf-
al. (2015) found evidence for limited admixture fix ‘-idae’, Article 11.7.1.3 of the Code states
between the two taxa in one Sicilian population. that “a family-group name of which the family-
In addition, Vamberger et al. (2015) reported group name suffix is incorrect is available with
concordant clines for both markers, with cline its original authorship and date, but with a cor-
centres matching the Strait of Messina. Vam- rected suffix”. Other conditions of availability
berger et al. (2015) argue that the coincidence all seem to be fulfilled in Spix (1825), who is
of the cline centres with the Strait of Messina thus the author of the nomen.
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access152 J. Speybroeck et al.
Macey et al. (2000a) sequenced a 1685-1778 spelling is not an available name. If one fol-
bp long segment of mtDNA (including the ND1, lows this and Art. 11.7.1.1 (“a family-group
ND2 and COI genes) to assess phylogenetic name when first published must [. . . ] be a noun
relationships of acrodont lizards. Their results in the nominative plural formed from the stem
suggested that the agamid genus Laudakia is of an available generic name”), a family-group
paraphyletic, yet, low bootstrap support pre- name based on an incorrect subsequent spelling
vented definite conclusions. Subsequent stud- of an available genus name is not available.
ies based on mitochondrial and nuclear genes Yet, the Code also states that “a family-group
(ND2, RAG1) by Melville et al. (2009) and name is an incorrect original spelling and must
Edwards and Melville (2011) recovered Lau- be corrected if it is formed from an incorrect
dakia as monophyletic with high support. Baig subsequent spelling of a generic name” (Art.
et al. (2012) summarised the results of the afore- 32.5.3.3) and that “a family-group name based
mentioned studies in a morphology-based revi- upon [. . . ] an incorrect spelling of the name of
sion of Laudakia. Despite failing to find distinct the type genus must be corrected” (Art. 35.4.1).
morphological variation within the genus, and It is thus clear that the Code has no inten-
acknowledging that Melville et al. (2009) and tion to make a family-group name unavailable
Edwards and Melville (2011) recovered Lau- based on an incorrect spelling of the name of
the type genus (see also Dubois, 2010). In con-
dakia as monophyletic, Baig et al. (2012) par-
clusion, the family-group name Camaelonia is
titioned Laudakia into three genera acknowl-
available as published by Rafinesque (1815), but
edging its potential paraphyly (Macey et al.,
its spelling needs to be corrected. The nowadays
2000a). This taxonomic act was subsequently
prevailing spelling Chamaeleonidae should be
criticised by Pyron, Burbrink and Wiens (2013),
preserved (Art. 29.5), with its authorship at-
who confirmed the monophyly of Laudakia us-
tributed to Rafinesque (1815) instead of Gray
ing a supermatrix approach. Within our focal
(1825).
area, the classification proposed by Baig et al.
SBC2010 adopted the inclusion of Cyrtopo-
(2012) would affect Laudakia stellio and L. cau-
dion kotschyi in the well-supported mono-
casia, as these authors placed the former species
phyletic genus Mediodactylus, as proposed by
in the newly erected genus Stellagama, and the
Macey et al. (2000b), Červenka, Frynta and
latter into Paralaudakia. While these genera Kratochvíl (2008) and Bauer et al. (2013).
were rapidly adopted by the wider herpetolog- By enlarging the focal area of SBC2010, we
ical community, we do not follow the split of hereby also include the former C. russowi,
Laudakia, pending substantial evidence to reject now Mediodactylus russowii. Another species
its monophyly, and therefore retain L. stellio and occurring in the area considered by Gasc
L. caucasia in Laudakia. et al. (1997) but not in that of SBC2010
The family Chamaeleonidae is sometimes is the former Cyrtopodion caspium. We fol-
credited to Gray, 1825 (e.g. Glaw, 2015). How- low Bauer et al. (2013) in recognising the
ever, Rafinesque (1815) published the name well-resolved genus Tenuidactylus and list this
“Famille CAMÆLONIA”, based on the genus species as Tenuidactylus caspius. Furthermore,
“Camæleo Daud.”, which is an incorrect subse- an introduced, well-established population of
quent spelling of the available genus Chamaeleo Tenuidactylus fedtschenkoi has recently been re-
Laurenti, 1768. The Code is somewhat am- ported from the city of Odessa (Ukraine). The
biguous regarding the availability of family- native range of T. fedtschenkoi lies in central
group nomina based on an incorrect subsequent Asia (western Pamiro-Altay mountains), and it
spelling of an available genus. Article 19.1 of was probably transported passively to Odessa
the Code states that an incorrect subsequent (Duz’, Kukushkin and Nazarov, 2012). Not long
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessSpecies list of European herpetofauna 153
after the record from Odessa, a revision of delimitation of the ranges and contact zones of
the genus Tenuidactylus led to the description these new species requires further investigation
of a new species from Uzbekistan and Turk- (e.g. M. danilewskii and M. orientalis), we ac-
menistan, which was split from T. fedtschenkoi cept these species here. The following species
and named T. bogdanovi (Nazarov and Po- are thus recognised in our area: M. kotschyi
yarkov, 2013). Krasylenko and Kukushkin (mainland Balkans, most of Aegean islands, and
(2017) provided an update on the status of the Italy (Apulia)), M. bartoni (Crete and nearby
non-native Odessa population and assigned it to islets), M. danilewskii (Black Sea region and
T. bogdanovi. south-western Anatolia), M. oertzeni (southern
Within Mediodactylus kotschyi, many sub- Dodecanese Islands), and M. orientalis (Levant,
species have traditionally been recognised. As Cyprus, southern Anatolia, and south-eastern
indicated by Kasapidis et al. (2005), the ge- Aegean islands).
netic substructure of Mediodactylus kotschyi The endemic gecko from the Selvagens Is-
shows a high degree of divergence, suggest- lands has variously been treated as the valid
ing that M. kotschyi represents a species com- species Tarentola bischoffi (e.g. Rebelo, 2008;
plex. Nearly range-wide data (with limited ar- Sindaco and Jeremčenko, 2008; Uetz, Freed and
eas in eastern and northern Anatolia excluded) Hošek, 2019), or as a subspecies of the Ca-
of three mtDNA and three nuclear DNA frag-
narian T. boettgeri as T. boettgeri bischoffi (e.g.
ments allowed unravelling the evolutionary his-
Carranza et al., 2000; Pleguezuelos, Márquez
tory of Mediodactylus kotschyi (Kotsakiozi et
and Lizana, 2002; Gübitz, Thorpe and Malho-
al., 2018). Divergence dates back to 15 Mya,
tra, 2005; Sá-Sousa et al., 2009). Its phyloge-
and several of the main lineages show overlap-
netic affinities were examined by Carranza et
ping distribution areas. Divergence in mtDNA
al. (2000, 2002) and Gübitz, Thorpe and Mal-
is known to be often quite large in geckos (e.g.
hotra (2005). The genetic data available (al-
Nagy et al., 2012). However, distances between
beit based only on mtDNA) show that the mi-
the main lineages are particularly large in this
tochondrial lineages of bischoffi are closely re-
case (>10% 16S and >15% cyt b and COI
lated to those of T. boettgeri, especially to the
p-distances). In addition, nuclear data groups
subspecies T. b. hierrensis from El Hierro, and
specimens in concordance with their mtDNA
lineage, and not with their geographical ori- are in fact nested inside the mitochondrial di-
gin, which suggests a high level of repro- versity of T. b. boettgeri from Gran Canaria,
ductive isolation between them. The proposed suggesting that the Selvagens Islands were re-
splits also largely agree with morphological cently colonised from El Hierro or Gran Canaria
data (Štěpánek, 1937, 1939; Szczerbak and Gol- (Carranza et al., 2000, 2002; Gübitz, Thorpe
ubev, 1996). The narrow contact between M. and Malhotra, 2005). Treating bischoffi as a
orientalis and M. danilewskii in the Western valid species, while retaining hierrensis as a
Taurus Mountains, identified by Kotsakiozi et subspecies of boettgeri (as done by the above
al. (2018), corresponds to a lack of morpho- authors who rank bischoffi as a species), is
logical intergrades (despite the narrow gap be- thus likely to render Tarentola boettgeri para-
tween the subspecies danilewskii and cilicien- phyletic. The amount of mitochondrial diver-
sis in the same area, see Rösler, Schmidtler and gence between hierrensis and bischoffi is also
Moravec, 2012), while these new species are smaller than within the populations of T. de-
very close to the M. kotschyi groups defined by lalandii from Tenerife or between the popula-
Beutler (1981). Five species were recognised, tions of T. angustimentalis from Fuerteventura
for all of which previously existing names are and Lanzarote. Based on the available data, we
available (Kotsakiozi et al., 2018). While the thus recommend to treat the geckos from the
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free access154 J. Speybroeck et al.
Selvagens Islands as conspecific with the popu- distribution, indicate that the three lineages cor-
lations from El Hierro and Gran Canaria, as Tar- respond to three independent species. The name
entola boettgeri bischoffi. of the eastern species is often spelled P. ed-
Sánchez-Vialas et al. (2018) re-examined the wardsianus. As shown by Crochet (2015), this
description of Algyroides hidalgoi Boscá, 1916, is an incorrect subsequent emendation that does
a nomen for which the holotype has been lost, not meet the requirements of the Code. As a
to settle its position in the synonymy of the consequence, the valid spelling of the eastern
genus Algyroides. They argued that the charac- species is Psammodromus edwarsianus. Thus,
ters of the holotype, said to originate from the we add Psammodromus edwarsianus and P. oc-
Sierra de Guadarrama and described by Boscá, cidentalis to the list of the European herpeto-
fall within the morphological variability of Al- fauna species.
gyroides marchi Valverde, 1958, and designated A 9 Mya split marks the divergence of Timon
a specimen of A. marchi as neotype of A. hi- lepidus nevadensis from the nominal subspecies
dalgoi. This would make A. marchi a junior T. l. lepidus (Miraldo et al., 2013). After studies
subjective synonym of A. hidalgoi. Because the on genetic (Paulo et al., 2008; Miraldo et al.,
conditions for automated reversal of precedence 2011, 2013) and morphological differentiation
are not met (Art. 23.9.1 of the Code), A. hi- (Mateo and Castroviejo, 1990; Mateo, López-
dalgoi would become the valid nomen of the Jurado and Guillaume, 1996), mtDNA and mi-
Spanish Algyroides. Several TC members felt, crosatellites were used to investigate gene flow
however, that the interpretation of the descrip- patterns in a zone of secondary contact (Miraldo
tion of Boscá (1916) by Sánchez-Vialas et al. et al., 2013). While hybridisation and introgres-
(2018) left room for doubt, and that Boscá may sion were observed, gene flow was shown to be
not have described a specimen of Spanish Algy- restricted. The cline width for nuclear markers
roides when he created the name Algyroides hi- was estimated at around 10 km (although with
dalgoi. If so, this could affect the validity of the a sampling gap of around 20 km, it may actu-
neotype designation, respective to Art. 75.3.5 of ally be less). Furthermore, mostly pure popu-
the Code, and result in the unnecessary change lations are present on either side of the sam-
of the well-established and widely used name A. pling gap. Considering this together with the
marchi. Although opinions in the TC were di- aforementioned old genetic divergence, we ac-
vided about this, we recommend to maintain for cept the proposal of Miraldo et al. (2013) and
the time being the use of A. marchi, in anticipa- treat T. nevadensis as a valid species.
tion of an upcoming application to the Commis- Lacking range-wide sampling and adequate
sion, which would maintain marchi in use until molecular analysis, the taxonomy of the Lac-
the Commission has ruled on the case. erta trilineata-pamphylica complex has re-
Morphological and molecular (mtDNA and mained unresolved, until mitochondrial phylo-
nuclear) data support the split of Psammod- genies showed the eastern Anatolian species L.
romus hispanicus into three distinct lineages pamphylica to be nested within trilineata (God-
(Fitze et al., 2011, 2012; Mendes et al., 2017). inho et al., 2005; Ahmadzadeh et al., 2013;
While not all areas of potential contact have Sagonas et al., 2014). Remarkably, Sagonas et
been sampled (despite a recent distribution up- al. (2014) found evidence that central Aegean
date by Molina et al., in press), the available populations (L. t. citrovittata) are closely related
data seem to warrant accepting these lineages to L. pamphylica. Yet, this was poorly supported
as distinct species. Age estimates, lineage al- across analyses and the biogeographically sur-
lopatry, the lack of mitochondrial and nuclear prising relationship between L. t. citrovittata
haplotype sharing between lineages, bioclimatic and L. pamphylica could reflect a methodologi-
niche divergence, and the current biogeographic cal artefact (long branch attraction). Thus, the
Heruntergeladen von Brill.com05/14/2020 03:08:24PM
via free accessYou can also read
