Cyrioctea (Araneae, Zodariidae) in Africa: temperate Gondwanaland relict, recent radiation, or both?

. Two new species of the zodariid genus Cyrioctea Simon, 1889 are described: C. sawadee sp. nov. and C. lotzi sp. nov., both only known from males. The genus now contains seven Afrotropical species and this abundance is discussed in the context of its basal situation in the family and its apparent temperate Gondwanaland distribution, which implies a much greater age of the Zodariidae than presently accepted. Unlike most taxa with a temperate Gondwana distribution, Cyrioctea boasts a high number of species with small distribution areas. This points in the direction of a recent radiation initiated after a long period of stasis.


Introduction
The spider genus Cyrioctea is remarkable in many ways.The spiders are characterized by a transverse row of, most often six, short but sturdy spines between the eye rows.These are assumed to constitute a burrowing device.Cyrioctea do indeed live under ground and females, that hardly ever emerge from their burrows, are very rarely collected.Although the position of the genus has been the subject of some debate, these spiders clearly belong to the Zodariidae (Jocqué 1991): they lack a serrula and are provided with long anterior lateral spinnerets.Since the claw teeth are positioned in the axis of the claw, they were considered as the sister group of all other Zodariidae, in which the claws are positioned on the side of the claw facing the other one.The fi rst cladistic analysis of the family on genus level therefore used Cyrioctea as the outgroup.A recent unpublished cladistic analysis (Jocqué 2006) on morphological grounds with Amaurobius as outgroup (see Jocqué & Dippenaar 2007), fi nds Cyrioctea in the same position at the root of the Zodariidae, thus sister to all other genera in the family.
Because of its ancestral character it is not surprising to fi nd the genus on three continents: Africa, South America and Australia (Platnick 1986;Platnick & Griffi n 1988;Platnick & Jocqué 1992).This is a typical temperate Gondwanaland distribution with all the localities on the southern tip of the continents.On the other hand, the complexity of the male genitalia is startling.Moreover, the species appear to have very small distribution areas, which is also puzzling as the Gondwanaland relicts have been assumed to be leftovers of old radiations with a large distribution.
The present paper describes two new species, gives an overview and a distribution map of African Cyrioctea species.A key to the males is provided.

Material and methods
All material is preserved in 70% ethanol.Specimens were observed and measured with a Leica M10 stereo microscope.Photographs were taken with a Leica MZ16 using the LAS automontage software.Drawings were carried out with a camera lucida on a WILD M5.
All measurements are in millimetres.

Remarks
The genus Cyrioctea was described by Simon (1889), on a species from South America and redescribed in detail by Platnick (1986) and Jocqué (1991).Cyrioctea now contains 13 species of which fi ve occur in South America, one in Australia (Platnick 2012)

Diagnosis
The male of C. sawadee sp.nov.can be recognized by the modifi ed third leg with dense spination and very short tibia, and by the characters of the palp, of which the tooth on the ventral margin of the RTA is the most conspicuous.That character is shared with C. griswoldorum, which has a shorter dorsal tibial apophysis, the shape of the RTA is triangular rather than rectangular and the abdominal pattern consists of one pale spot on a dark background.

Etymology
The species name is a noun in apposition taken from the type locality.Other material examined None.
COLOUR.Carapace brownish yellow with black fovea and dark margin; palp, chelicerae, mouthparts and sternum pale brown; legs yellow, femora suffused with dark grey; abdomen pale grey with dark pattern of central spot followed by four transverse bands; venter medium grey, darkened towards yellow spinnerets.
CEPHALOTHORAX.Clypeus centre devoid of setae but with dense cluster of inward curved thick setae on either side.Eye region with a row of six slightly curved spines (Figs 1B, 3A), on right side with one short thick extra seta.Chilum poorly developed, inconspicuous.Sternum subcircular, with rather long, thin, posterior extension.

Distribution
Only known from type locality in the Western Cape Province (Fig. 4).

Diagnosis
The male of C. lotzi sp.nov.can be recognized by the pale abdomen with faint dorsal stripe, the strong spines on the fi rst leg and the details of the male palp in which the RTA is adjacent to the dorsal apophysis.

Etymology
The species name is a patronym in honour of one of the collectors.Other material examined None.

Distribution
Only known from type locality in the Free State, South Africa (Fig. 4).

Discussion
Taxa with a temperate Gondwana distribution, also called an Austral distribution (Humphreys & Parenti 1999), found on the southern tip of the continents Australia, Africa and South America, sometimes including New Zealand and New Caledonia, have been considered to be relicts (e.g.Crisci et al. 1991;Robertson & Holzenthal 2005).This type of distribution is often synonymous with the presence of ancestral characters, which is also the case for Cyrioctea.Other examples are the ancestral Ratites, fl ightless birds (Bourdon et al. 2009), Pleurodira, primitive turtles (Sterli 2010) and beetles that are herbivorous on Araucaria (Sequeira & Farrel 2001).Some of these have been qualifi ed as living fossils as they are at the base of old, large clades.Such groups, albeit rarely exhibiting a Gondwana-distribution, are most often very poor in species, e.g., Tuatara (Hay et al. 2010), Platypus (Groves 2005), Coelacanths (Forey 1998) etc.However, this phenomenon cannot be qualifi ed as a rule since there are remarkable exceptions.The "primitive" spiders with segmented abdomen belonging to the family Liphistiidae are an excellent example of such an exception: the family is composed of 5 genera and contains not less than 85 species (Platnick 2012).And in spider families for which a detailed phylogenetic analysis is available, it is not the rule that basal genera are species poor.It might be so for Lordhowea Griswold, 2001 in Cyatholipidae (Griswold 2001) or Pararchaea Forster, 1955 in Pararchaeidae (Rix 2006), but in Theridiidae, for instance, the most basal genus is the speciose Dipoena Thorell, 1869 (Agnarsson 2004).Also in families like Linyphiidae (Miller & Hormiga 2004), Ctenidae (Silva 2003) and Thomisidae (Benjamin et al. 2008), the most basal genera are rich in species.Thus it is certainly not a general phenomenon that basal taxa of a large evolutionary line tend to be relicts in the sense that they are poor in species.
But also from the morphological point of view, certain characters and more precisely the genitalia of Cyrioctea, are puzzling.In the Zodariidae, the root of the large clades is very often occupied by species with very simple genitalia.A recurring pattern is the presence of a short, straight, spine-like embolus accompanied by a short and simple median apophysis and a simple RTA in the species near the root, evolving to a long embolus accompanied by intricate supporting structures and the addition of apophyses on the palpal patella and even the femur.Examples are Diores namibia Jocqué, 1991 for Diores with more than 50 species (Jocqué 1991), Tenedos infrarmatus Jocqué & Baert, 2002 at the base of a large clade containing the genera Tenedos, Ishania, Epicratinus and Colima (Jocqué & Baert 2005), Procydrela procursor Jocqué, 1999 at the base of the Cydrelinae (Jocqué 1999), and Pentasteron simplex Baehr & Jocqué, 2001 at the base of the very large Asteron complex in which more than 100 species have already been described (Baehr & Jocqué 2001).
In other families where a similar phenomenon is assumed to occur, the trend is often obscured by the creation of separate genera exactly on the basis of these additional structures on palpal patella and femur (e.g.Wang 2002).Data on the Zodariidae, one of the few families for which there is a combination of generic and species phylogenies, show that a somatic radiation precedes the genital one.But Cyrioctea, does not qualify.All the species have a complex bulbus with several tegular appendages.In several species, however, the palpal tibia shows a typical ancestral character with the presence of a dorsal apophysis, which may be the precursor of the retrolateral tibial apophysis (RTA).But the complexity of the bulbus is incoherent with the idea that Cyrioctea might be considered a living fossil dating back from the split-off of Gondwanaland.It might therefore be argued that a recent radiation has occurred, explaining the existence of closely related species with small distribution areas.Such a recent radiation is known for the plants belonging to the Cycadales, considered a group of "living fossils" until recently (Nagalingum et al. 2011).Neither the Liphistiidae nor Cyrioctea have been subject to such a detailed analysis, but may well represent other examples of old taxa with a recent radiation.However, the most important consequence of Cyrioctea's distribution is that the family is much older than what is proposed in Penney & Selden (2011).Following Miller et al. (2010aMiller et al. ( , 2010b)), they consider Zodariidae the sister family of Penestomidae and conclude that the clade is not older than 30 ma.Since it must date back from before the split up of western Gondwanaland, its age must be pushed back to at least 100 ma (Pitman et al. 1993).
It is evident that only a detailed cladistic analysis including molecular data can corroborate the hypothesis that has been formulated here.