Hoggicosa, TO OTHER

Langlands, Peter R. & Framenau, Volker W., 2010, Systematic revision of Hoggicosa Roewer, 1960, the Australian ‘ bicolor’ group of wolf spiders (Araneae: Lycosidae), Zoological Journal of the Linnean Society 158 (1), pp. 83-123 : 88-89

RELATIONSHIPS OF HOGGICOSA TO OTHER

LYCOSINE GENERA

The systematic analysis resulted in an overall poor resolution of relationships amongst genera, which appears to reflect the conservative nature of wolf spider morphology (e.g. Vink, 2001; Stratton, 2005). Indeed, resolving relationships within the Lycosinae poses a number of problems. A recent molecular attempt found that Lycosinae contain paralogous copies of the 28S rRNA gene, making this gene unsuitable for analysis ( Murphy et al., 2006). A combined effort employing morphological character sets and molecular markers may solve relationships in this morphologically conservative group, as exemplified for the New Zealand Artoriinae ( Vink & Paterson, 2003) . Additional phylogenetic signal may also be provided by morphological characters that we were not able to explore as part of the current study. For example, spinneret morphology has been used in phylogenetic analyses of other entelegyne spiders (e.g. Scharff & Coddington, 1997; Griswold et al., 2005) and shows promising intergeneric differences within wolf spiders ( Townley & Tillinghast, 2003).

We did not find an unambiguous sister group to Hoggicosa based on our character and taxon sample. Based on similarities in morphology and burrowing behaviour we included Knoelle clara and a representative of an undescribed Australian genus ‘Grey Wolf Spider’ as potential sister taxa. But neither of these nor McKay’s (1973) previous suggestion of Geolycosa were supported as sister group by our phylogeny ( Fig. 2 ). Molecular studies suggest that Australian Hogna may be closely related to Hoggicosa . The maximum parsimony analysis of molecular data by Murphy et al. (2006) found Hogna crispipes as sister to two Hoggicosa representatives. Adding to these results, Gotch et al. (2008) found that Hogna kuyani was sister to Hoggicosa , although the posterior probability of this relationship was low (54%), and Hogna crispipes was basal to both. These two Hogna species ( Hogna crispipes and Hogna kuyani ) were included in our phylogeny and found to form a monophyletic group, but a sister group relationship with Hoggicosa was not supported. In addition, a third species currently listed in Hogna ( Hogna immansueta ) was found to group with the type species of Venator , suggesting additional taxonomic problems within the six Australian species currently placed in Hogna .

Our analysis clearly shows that some Australian species currently listed in Lycosa do not form a monophyletic group with the type species of the genus, L. tarantula and a congeneric L. praegrandis . Lycosa godeffroyi and L. leuckartii were placed separately in the unresolved base of the tree, contrary to the molecular results of Murphy et al. (2006) who found very strong support for the monophyly of these two species. They represent a very diverse clade of common Australian Lycosinae characterized by a typical ‘Union-Jack-colour pattern’ on the cephalothorax with alternating dark and light radial lines (character 2). Tasmanicosa tasmanica ( Hogg, 1905) , the type species of Tasmanicosa , is part of this group and the genus Tasmanicosa is currently under revision by the junior author.

The grouping of the North American representatives of Geolycosa with the Australian genera Dingosa and Mainosa was surprising and may indicate a case of convergence. This group was defined by the lack of cymbium macrosetae and the presence of burrow palisades, which is an uncommon burrow modification in lycosids. All these species live in sandy habitats and this shared burrowing behaviour and putative associated morphology may be a result of adaptations to live in this substrate. Neither Dingosa nor Mainosa were represented in Murphy et al. (2006), but Geolycosa missouriensis which was included, was found to be the basal most lycosine (parsimony analysis) or sister to Rabidosa (Bayesian analysis).

The revalidation of Hoggicosa places L. exigua ( Roewer, 1960) from Namibia back into Hoggicosa . A critical review of Roewer’s (1960: 773–774, fig. 432A, B) original description of Hoggicosa exigua clearly shows that the epigyne is not like that of Hoggicosa or Lycosa (e.g. Zyuzin & Logunov, 2000), but much more reminiscent of Hogna (e.g. Fuhn & Niculescu- Burlacu, 1971 for the type species of Hogna , Hogna radiata Latreille, 1817 and Framenau et al., 2006 for Australian Hogna ). We therefore transfer Hoggicosa exigua to Hogna , Hogna exigua ( Roewer, 1960) pending a revision of the African Lycosidae related to Hogna . Unfortunately, we were not able to include Hogna exigua in our phylogenetic analysis to confirm this transfer.

Figures

Figure 2. Strict consensus tree showing the relationships amongst Hoggicosa species and other Lycosinae (length = 104, consistency index = 35, retention index = 53). Characters are indicated by squares: character number above and character state below square with shaded squares for nonhomoplasious characters. Bremer support and relative Bremer support (0–100) values for each branch are indicated. Inset of colour dimorphism (character 4, bold) under fast optimization (ACCTRAN) and slow optimization (DELTRAN).