«Taxonomic Revision, Molecular Phylogeny and Zoogeography of the huntsman spider genus Eusparassus (Araneae: Sparassidae) Dissertation for attaining ...»
Eusparassus is different from the ―African clade‖ genera by: anterior median eyes (AME) always larger or equal to anterior lateral eyes (ALE) (the ―African clade‖ members with AME smaller than ALE) and legs ventrally with two pairs of tibial spines (three pairs in the ―African clade‖). Indeed, some Eusparassus species have intermarginal denticles in their chelicerae which are absent in the genera of the African clade. Although Eusparassus resemble the ―African clade‖ member particularly by the shape of copulatory structures but it is commonly accepted that genital structures evolve and change more rapidly compared to somatic ones (Eberhard 2010). Consequently, Eusparassus might not have any close relationships to African clade genera and the similarity in the noted traits might be the result of convergent evolution. If this hypothesis is further supported, Eusparassinae has to be considered monotypic. At this point, the results are not sufficient to propose a new subfamily for the ―African clade‖ genera or exclude them from Eusparassinae but revisions and more robust phylogenetic results are required before proposing a new taxonomic rank. The exact position of Eusparassus within Sparassidae was not resolved. However, the relationship of Eusparassus with the rest of Eusparassinae genera might be recovered by adding more data supporting the backbone of the current phylogeny.
The Asian genus Rhitymna which was proposed originally to be classified in Eusparassinae (Järvi 1912) was placed separately in a clade containing genera Staianus, cf. Remmius, and Micrommata indicating no relationships to Eusparassinae. This affirmed the note by Jäger (2003) that the genus belongs to a lineage different from Eusparassinae.
4.2.3. Sparassidae and the rest of subfamilies
This study was the first comprehensive phylogenetic research on the family Sparassidae treating the majority of morphological variations and classical subfamilies. The inferred phylogeny supported Sparassidae as a monophyletic group with all available subfamilies nested within as well as the type genus for the family, Micrommata. Sparassidae split into two clades, a basal clade containing individuals of the subfamily Sparianthinae (represented by the genus Thelcticopis which features all of the typical characteristics of Sparianthinae) and a larger clade composed of the rest of Sparassidae (termed non-Sparianthinae). This latter result is concordant with morphological evidences since Sparianthinae have a kind of autapomorphic characters when compared to the rest of Sparassidae. It also indicated that Sparianthinae is likely a very early diverging group and is sister to all other Sparassidae.
Within non-Sparianthinae clade the monophyly of the subfamilies Heteropodinae sensu stricto (Asia, Africa), Palystinae (Africa) and Deleninae (Australia) were supported. Heteropodinae with representatives of six genera were recovered in a basal position within the non-Sparianthinae clade. However the backbone of the non-Sparianthinae clade was not resolved. The monophyly of Heteropodinae sensu stricto is also supported by synapomorphies for this subfamily including the trilobate membrane with well-developed lateral projections and median hook, chelicerae with three anterior and four to six posterior teeth intermarginally covered with denticles (Jäger 1998). Through the investigated taxa, just members of Heteropodinae and two species-groups of Eusparassus have intermarginal denticles in their chelicerae. Whether denticles are a plesiomorphic character state in Sparassidae or not could not be tested with the results since the backbone of the tree was not resolved. But it should be noted that the form of cheliceral denticles is different in Heteropodinae sensu stricto from those of Eusparassus. Mostly a patch of denticles is present close to the three anterior teeth in Heteropodinae (Jäger 2002;
Moradmand and Jäger 2011) while in Eusparassus, there is a line or scattered denticles throughout cheliceral furrow (Moradmand and Jäger 2012a; Moradmand 2013).
Sparassinae and the genus Olios (as the most similar genus to Eusparassus in somatic character) appeared to be polyphyletic but further studies require additional taxon sampling. Previously, the polyphyly of Olios was suggested by Rheims (2010) and noted by Jäger and Kunz (2005) according to morphological evidences. The genus Olios which is distinguished from Eusparassus by its morphology (Moradmand and Jäger 2012a, Moradmand 2013) confirmed to be a separate clade by molecular data as well.
The characters of the copulatory organs, trilobite membrane, eye arrangement and cheliceral dentition are currently the available morphological characters for the classification of Sparassidae. The inferred phylogeny revealed that these characters have phylogenetic values at least in the subfamilies Sparianthinae, Heteropodinae sensu stricto, Palystinae and Deleninae. However these characters should be applied in combination since possibility of convergence is important to consider.
Very recently Agnarsson and Rayor (2013) tested the phylogenetic relationships of Australian Deleninae. They focused exclusively on Deleninae and the evolution of sociality in its members. Despite they recovered Deleninae monophyletic, but they did not include Sparianthinae, Staianinae and even putative close relatives to Deleninae such as Sparassinae in their analyses.
4.2.4. ‘Laterigradae’, Dionycha and RTA-clade Further studies focused on the phylogenetic position of Sparassidae within the order Araneae and test its relationships to assumed closely related families. Sparassidae share the character of crab-like posture (laterigrade position of legs) with the entelegyne families Selenopidae (wall crab spiders), Philodromidae (running crab spiders) and Thomisidae (―true‖ crab spiders). These families together with Sparassidae were grouped under the name ‗Laterigradae‘ by Latreille (1802) and to date they are all unplaced families within the Dionycha of the RTA-clade. It was assumed that these four crab-like families may eventually be clustered near each other (Coddington and Levi 1991). All representatives of the ‗Laterigradae‘ families and a majority of the RTA-clade families with focus on Dionycha spiders were included in the phylogenetic analyses.
The ‗Laterigradae‘ was not recovered as a monophyletic group thus the similar crab-like leg posture in all four families (Sparassidae, Thomisidae, Philodromidae and Selenopidae) is better explained by evolutionary convergence. As noted above, Sparassidae was placed in an outgroup position towards the rest of the RTA-clade members. Thomisidae was firmly nested within Lycosoidea. However, Philodromidae and Selenopidae were found to be more closely related in a clade together with two family members of contrasting body organisation Salticidae and Corinnidae.
‗Laterigradae‘ families showed no further relationship towards each other which suggest that the term should be abandoned.
The monophyly of Dionycha was not supported in the current phylogeny and this result is in accordance with Agnarsson et al. (2013). The Dionycha was solely erected on the simple character of the two-clawed legs and composed of several heterogeneous spider families (Coddington 2005). The RTA-clade recovered fully monophyletic with Sparassidae nested within the clade using spiders of the non-RTAclade (Orbiculariae, Eresoidea and Palpimanoidea) as outgroups. This result was also retained by the analyses of Spagna and Gillespie (2008) and Miller et al. (2010), while these authors did not include Sparassidae in their studies.
4.3. Origination and zoogeography of Eusparassus Here I combined the results from molecular dating and phylogeny as well as current knowledge on Eusparassus distribution range to discuss about the zoogeography of the genus members.
Extant Eusparassus species are distributed across the Old World from southern Africa to Mediterranean region, and from Arabia and Middle East toward Central and South Asia. The previous records of the genus from Madagascar (―Eusparassus‖ laterifuscus), Americas (―Eusparassus‖ shefteli) and SE Asia (―Eusparassus‖ lilus) were proven to be misidentifications (Moradmand and Jäger 2012a; Moradmand 2013). The genus represents one of the evolutionary successful groups of spiders since it exists at least from early Tertiary till present (~50 million years), according to the Eusparassus amber fossil record in northern Europe (Dunlop et al. 2011). To trace and discuss the findings of this study on the origin of Eusparassus, I should give some notes on the origination of Sparassidae and its comprising subfamilies (see chapter 3.3).
The origin of Sparassidae was dated back to early Jurassic (186 MA) by the molecular dating analyses, when they diverged from the rest of RTA-clade families.
Subsequently around 163 MA, the hypothesised Sparassidae ancestor diverged into two clades: the Sparianthinae as basal group and the non-Sparianthinae clade. The further divergence happened rapidly within the non-Sparianthinae much later (106– 97 MA) into current diverse groups of Sparassidae [e.g. African clade (Africa), Heteropodinae (Africa, Australasia), Palystinae (Africa), Polybetinae (Americas) and Deleninae (Australia)]. Since the backbone of the phylogenetic relationships among the non-Sparianthinae clade was not resolved, the relationships between Eusparassus and the African clade genera (as proposed Eusparassinae) cannot be explicitly diagnosed.
However, a hypothesis of Eusparassus origination can be proposed based on the molecular dating, historical biogeography and geological events: Eusparassus lineage probably originated around 70 MA in SW Africa in the border of the Namib Desert. This hypothesis is supported with the following evidences. Extant and fossil Eusparassus species occurred exclusively in Africa and Eurasia meaning that the genus does not have a Gondwanan distribution range and likely originated after the breakup of the supercontinent Gondwanaland. Africa after splitting from other Gondwanaland plates around 110–95 MA remained isolated, stabilized and unchanged until collided and connected with Eurasia in the Paleocene around 60 MA (Sanmartín and Ronquist 2004). But this connection was not still complete to allow dispersal from Africa+Europe to Asia because of a great sea barrier named Turgai Strait which divided the Palearctic into an eastern and western half until 30 MA when the Strait dried up and permitted extensive biotic exchanges (Sanmartín et al. 2001).
Since Eusparassus species are representatives of arid to semiarid regions, its origination and dispersal must be connected to desertification and distribution of deserts. The most suitable place for their ancestors could be the transition zone of a desert with stable environment at the estimated divergent time (70 MA). The Namib Desert with the age of 130–55 MA (Ward et al. 1983; Ward 2009) is the best candidate for the ancestral origin of Eusparassus. The hypothesis of southern African origin of Eusparassus is further supported by the current distribution range of the members of the African clade and also two basal Eusparassus species-groups (tuckeri- and jaegeri-group) within and in the borders of the Namib Desert. The Namib Desert has provided stable environments for its unique diversity of desert dwelling organisms with high levels of endemism (Barnard et al. 1998; Ward 2009).
Other deserts of the Old World are much younger, in contrast to the Namib Desert, the Sahara as the largest desert of the world formed not older than 7 MA (Schuster et al. 2006). The Asian deserts appeared or expanded mostly during the global desertification initiating around 23 Ma, in Miocene (Potter and Szatmari 2009).
The proposed six species-groups of Eusparassus are distributed in separate geographic regions but partially overlap with neighbouring group distribution range.
The inferred phylogeny recovered a clade of Eusparassus species with promising support (outlined here as non-southern African clade) and the southern African group members in a basal position (see chapter 3.3: figs 3, 5). The southern African group including E. tuckeri and E. educatus (the tuckeri-group) and E. jaegeri (jaegeri-group) recovered in a basal position within the Eusparassus clade but their relationships were not resolved. The tuckeri-group members are restricted to SW Africa at the borders of the Namib and Kalahari Deserts (Namibia, western South Africa and southern Angola). The jaegeri-group are neighbouring them in SE Africa (eastern Namibia, South Africa, Botswana and Zimbabwe). The members of the tuckeri- and jaegeri-group might be the oldest lineages of Eusparassus.
The non-southern African clade subsequently split into two major clades, one containing the dufouri-group + Cercetius perezi and another one containing walckenaeri-group + doriae-group members. The first clade separated from the second clade around 41 MA with a common ancestor probably lived in NE Africa and Arabia. The dufouri-group members are currently distributed in NW Africa (six species in Tunisia, Algeria and Morocco) and Iberian Peninsula (two species in Spain and Portugal). Most likely the Iberian species originated from NW Africa. However, this could not be tested with the available data since only one representative of NW African (E. oraninensis) was included in the analyses. But the results showed that the Iberian species have a more recent common ancestor compared to the Moroccan species around 23 MA. Cercetius perezi nested within the non-southern African clade but its exact position was not resolved. Nevertheless, morphological evidences suggest that C. perezi could be related to dufouri- and walckenaeri-group members.
The distribution of C. perezi among these groups from the Horn of Africa to Arabian Peninsula supports this notion.