The evolution of sociality is puzzling and determining factors that promote the transition towards a social lifestyle is a major challenge in evolutionary biology. Animals living in social groups benefit from cooperation in foraging, brood care and protection from predators (Brockmann 1997; Brown 1983; Choe and Crespi 1997; Creel 2001; Dechmann et al. 2010; Unglaub et al. 2013), but group living also entails costs such as competition for mating partners (Huchard and Cowlishaw 2011). In the last 20 years, the social lifestyle of “non-traditional” social taxa such as clonal aphids (Abbot 2009) or spiders (Avilés 1997; Lubin and Bilde 2007) has become of increasing interest. Spiders are recognized as important model organism to study the evolution of sociality (Agnarsson 2012; Agnarsson et al. 2006; Avilés 1997; Evans 1998a; Johannesen et al. 2005; Lubin and Bilde 2007; Ruch et al. 2009; Schneider and Bilde 2008; Yip and Rayor 2013). They are generally very aggressive and sociality in spiders is extremely rare (Agnarsson et al. 2006; Bilde and Lubin 2011). Nevertheless, sociality has evolved several times independently across eight families (Agnarsson et al. 2006), suggesting strong selective benefits from living in groups. However, identification of the selective agents is difficult due to a lack of detailed natural history and taxonomy of solitary, subsocial and social species (Agnarsson 2012). Such knowledge facilitates comparisons of factors promoting social behavior in general, for instance ecological factors (Avilés and Harwood 2012; Corcobado et al. 2012) and/or kin selection (Schneider and Bilde 2008).

The generally accepted hypothesis is that sociality in spiders evolved via the ‘subsocial route’, meaning that permanent sociality derived from ancestors with extended maternal care (Lubin and Bilde 2007; Wickler and Seibt 1993). This hypothesis is corroborated by the phylogenetic reconstruction of social spider lineages (Agnarsson et al. 2006; Johannesen et al. 2007). Subsocial spiders differ from permanently social spiders in that they disperse prior to mating and thus have an outbred mating system (Agnarsson et al. 2006; Avilés 1997; Lubin and Bilde 2007). In both, subsocial and social spiders, females care intensively for offspring and the latter cooperate, for instance, in hunting, foraging, webbuilding and predator defence (Avilés 1997; Lubin and Bilde 2007; Ruch et al. 2014a; Yip and Rayor 2013). A major characteristic explaining the evolution and maintenance of sociality in spiders is the construction of a communal capture web, which allows capturing large prey items (Avilés 1997; Lubin and Bilde 2007). Non-webbuilding subsocial and social lineages are documented in only two families, huntsman spiders (Sparassidae Bertkau 1872) as well as crab spiders (Thomisidae Sundevall 1833) The social lineages of both taxa can be exclusively found in Australia (Agnarsson and Rayor 2013; Evans 1995).

To date, all subsocial and social crab spiders are described in the genus Diaea Thorell, 1869 (Evans 1995). Three species are known to be subsocial or social: Diaea socialis Main, 1988 from Western Australia, Diaea ergandros Evans, 1995 and Diaea megagyna Evans, 1995 (= Diaea inornata (Szymkowiak and Dymek 2012)) from southeastern Australia (Evans 1997). Subsocial/social Diaea mainly build nests in small-leaved Eucalyptus trees. The climatic conditions in their habitats seem to be relatively similar across the range of their distribution from southern Queensland to Tasmania as well as in Western Australia (Evans 1997). Nest inhabitants are usually related, however, groups accept immigrating spiders from other nests in Diaea ergandros (Evans 1998a; Evans and Goodisman 2002). The presence of immigrating spiderlings seems to affect group dynamics and female care in Diaea ergandros (Ruch et al. 2014b) and female care is very important for offspring survival (Evans 1998a, b; Unglaub et al. 2013).

We have recently identified another case of subsociality in crab spiders: Xysticus bimaculatus L. Koch, 1867. The discovery of social behavior in a species outside the Diaea genus suggests a possible independent evolutionary event and thus the potential to identify common drivers in the evolution of sociality in spiders. Here, we describe the natural history and subsocial lifestyle for the first time (Koch 1867, 1876) and present a re-description of the species.

We initially discovered nests inhabited by several individuals of Xysticus bimaculatus L. Koch, 1867 in July 2011 on trees along the Enoggera Reservoir, Queensland, Australia (27°26'27.69"S, 152°55'29.03"E). We later surveyed spider nests in November 2011, April 2012 and November 2012 (N = 166) at four locations around Brisbane (Brisbane Forest Park, Toohey Forest, Mt Coot-tha, Mt Tibrogargan). During these surveys, we measured the nests and identified the trees these were built in. We determined the group composition (number, developmental stage and sex) of spiders inhabiting the nests. We used these data to pinpoint the dispersal stage of spiderlings, which is an indicator of the degree of sociality (Avilés and Harwood 2012; Lubin and Bilde 2007). All immature individuals are referred to as ‘spiderlings’. We moreover recorded prey items as well as commensals and potential predators in active nests that were inhabited by at least one spider (N = 131).

For the species re-description, specimens were compared with collection material located at the Australian Museum, Sydney, the Queensland Museum, Brisbane and the Zoological Museum Hamburg and included species from the genera Cymbacha, Diaea, Tharpyna and Xysticus (see Suppl. material 1 (material examined), type Xysticus bimaculatus see Figure 1). The description of the seta pattern was performed using the format described by Ramirez (2003).

Since the type locality has not been accurately specified in the original description, the species was re-described from specimens collected in the Enoggera Reservoir in April 2012. Specimens were stored in 70% EtOH and examined using a Zeiss Discovery V20 stereo microscope and imaged with a Zeiss MCr camera and a Leica M205A with a Leica 290 camera as well as with a Keyence Digital Microscope VHX-500 F. The images were edited and plates arranged using Adobe Photoshop CS4.

Female copulatory organs were dissected and macerated using pancreatin (Alvarez-Padilla and Hormiga 2007) and imaged with a Zeiss MCr camera mounted on a Olympus BX60 light microscope.

The left male palpus (sperm transfer organ) was stained with a 1.0% iodine solution overnight and critical point dried for the micro-tomographic analyses. The dry palp was mounted onto an insect pin and scanned with an Xradia MicroXCT-200 X-ray imaging system (Carl Zeiss X-ray Microscopy Inc., Pleasanton, USA) at 30 kV and 6 W (20.0 scintillator-objective lens unit, 6 seconds exposure time, 1.18 µm pixel size). The data were processed using the 3D analysis software AMIRA v5.4.2 (Visage Imaging, Berlin, Germany). Selected parts of the palp were reconstructed by delineation of the contours in each section and surfaces were computed using the surface editor.

We report the demographics of Xysticus bimaculatus, a non-webbuilding subsocial crab spider from southern Queensland. Its lifestyle appears to be very similar to the subsocial crab spider Diaea ergandros (Evans, 1995). Like in other subsocial crab spiders, the presence of a caring female seems to be important for offspring survival in Xysticus bimaculatus. We found higher numbers of spiderlings in nests with a caring adult female present and a similar pattern was found in Diaea ergandros (Unglaub et al., 2013). The presence of an adult female is beneficial in Diaea ergandros, but also in the subsocial huntsman spider Delena cancerides, since adult spiders are able to capture prey more efficiently (Evans 1998a, b; Yip and Rayor 2011). We found that the likelihood of an adult living Xysticus bimaculatus female being present in the nest was high when spiderlings were very young but declined when spiderlings were older. In Diaea ergandros some females are consumed by their offspring (matriphagy) (Evans et al. 1995) and it remains to be studied whether matriphagy occurs in Xysticus bimaculatus as well and could explain the reported pattern.

Unlike subsocial Diaea, Xysticus bimaculatus builds its nests mostly from Acacia and not from Eucalyptus leaves. This may favor the occurrence of the species in areas that are dominated by Acacia melanoxylon, which is however widely distributed and common along the Australian east coast. We only recorded those trees that were used for nest construction and did not quantify potentially available host trees, but both Acacia and Eucalyptus trees were present in all of our study sites. We never found Diaea ergandros and Xysticus bimaculatus occurring sympatrically. Diaea ergandros seems to be absent along the northern coast of New South Wales and southern coast of Queensland (Evans 1997) and so far we did not detect Xysticus bimaculatus nests south of Queensland.

Similar to Diaea ergandros, nests of Xysticus bimaculatus serve as foraging areas and major prey types are beetles (Coleoptera), but also wasps and ants (Hymenoptera) (Evans 1998a). In contrast to Diaea ergandros nests (Unglaub et al. 2013), we only found very few potential predators inside nests of Xysticus bimaculatus however, the nest may still protect spiders from predators that we did not detect.

We found that nests contain on average 10 spiderlings in Xysticus bimaculatus, which is fewer than in Diaea ergandros, where nests contain on average 45 inhabitants (Evans 1995). However, spiderling numbers in Xysticus bimaculatus did not significantly decrease with increasing age, indicating that spiders have a relatively long period of communal activities. The finding that spiders disperse only shortly before maturation suggests a transitory stage between subsocial and permanently social (Lubin and Bilde 2007). In almost all social spiders studied to date, a transition from subsociality to sociality is accompanied by a switch from outbreeding to inbreeding, which has major consequences for speciation processes (Agnarsson 2012; Agnarsson et al. 2006; Agnarsson et al. 2013; Bilde et al. 2005; Johannesen et al. 2007). An exception can be found in social spiders of the genus Tapinillus (Oxyopidae), which is thought to be outbred because it does not have a female-biased sex ratio (Avilés 1994). It would be highly interesting to investigate the mating system and sex-ratio of Xysticus bimaculatus and to compare it with other subsocial and social crab spiders.

The taxonomy of Thomisidae is challenging and a revision of most genera is needed (Benjamin et al. 2008; Szymkowiak 2007). Similarly, a recent molecular phylogeny of Sparassidae showed that two genera with subsocial species previously described as Eodelena are synonymous with Delena and all three known group-living Delena are closely related (Agnarsson and Rayor 2013). A molecular phylogeny of the group-living Thomisidae may thus help to understand whether sociality has evolved multiple times in this family or whether the species, albeit being assigned into different genera, are closely related as well. Since thomisid genera often lack a clear definition and diagnosis, species were assigned (especially in Australia) to the most common and cosmopolitan genera Diaea, Misumena, Thomisus and Xysticus (Lethinen 2002; Szymkowiak 2007). However, the taxonomic status of these genera is highly problematic. For example, Jantscher (2002) studied various thomisid genera of central Europe with a focus on the genus Xysticus and found at least three different groups within this genus characterized by the organization of the male palp (further previous suggestions of subgroups within Xysticus s.l. are reviewed in Jantscher (2002) and not addressed here). Since Xysticus bimaculatus lacks tegular structures it does not belong to the group “Xysticus s. str.” sensu Jantscher (2002), which is characterized by a complex tegular structure and at least two distinct tegular apophyses. Based on the apomorphies proposed by Jantscher (2002), Xysticus bimaculatus might be part of the “Proxysticus” group characterized by the three distinct tibial apophyses. Nevertheless, these suggestions are only based on data of European material and comprehensive studies of Xysticus s.l., a group which is likely paraphyletic (Jantscher 2002) are still lacking.

Although crab spiders have a worldwide distribution (Platnick 2014) group-living crab spiders can be exclusively found in Australia. This continent has a history of long isolation and is renowned for its harsh environmental conditions (Herberstein et al. 2014). It has been suggested that certain evolutionary phenomena are more pervasive in Australia, such as cooperative breeding or deception (Herberstein et al. 2014). Some solitary Australian crab spiders, for example, use their body UV reflection as deceptive signal to attract and hunt naïve pollinators (Heiling et al. 2004). The harsh environmental conditions prevalent in Australia may as well have played a role in the evolution of sociality in two spider families (Thomisidae and Sparassidae). The multiple independent origins across spider families provide the opportunity for comparative investigations aiming to unravel selective forces being responsible for the evolution of this lifestyle. Since both Thomisidae and Sparassidae do not build capture webs, alternative perspectives on key factors for the evolution of sociality need to be considered (Evans 1998a). Ecologically rather similar, the subsocial Xysticus and Diaea are a very suitable model to study their behavior and its drivers on comparative grounds.