Savignia

Kropf, Holger Frick Wolfgang Nentwig Christian, 2010, Progress in erigonine spider phylogeny-the Savignia-group is not monophyletic (Araneae: Linyphiidae), Organisms Diversity & Evolution (New York, N. Y.) 10 (4), pp. 297-310 : 308

Savignia -group relatives and ancestors

Based on the conformation of the male palp, Millidge (1977: fig. 200) regarded Erigonoplus as the sister taxon of the Entelecara -group plus the Savignia -group, with both having evolved from a Lophomma -like ancestor. The close relationship between those groups is supported by the present analysis, but their relationships differ from what Millidge (1977) predicted. Lophomma is sister to most members of the Savignia -group but not to Erigonoplus and Entelecara . The relation between the Entelecara -group ( Entelecara , Hybocoptus ) and the Savignia -group differs in details between most phylogenetic analyses: the Entelecara- group emerged either as a distant relative of the Savignia -group ( Hormiga 2000: fig. 40), as a clade among the non-monophyletic Savignia -group ( Miller and Hormiga 2004: fig. 7) or among a clade basal or distal to the Savignia -group (present study: Fig. 5; Frick and Muff 2009: fig. 67; Paquin et al. 2008: fig. 20; Seyfulina and Jocqué 2009: fig. 4). Therefore, the relationships between these groups remain unclear. Clade 23 includes members of the Pelecopsis -group ( Parapelecopsis , Silometopus , Hypomma , Abacoproeces , Dismodicus ), the Tapinocyba - group ( Tapinocyba , Ceratinops ), the Entelecara -group, the Erigonoplus -group ( Erigonoplus ), as well as taxa that were not assigned to any group ( Caracladus , Monocephalus ). Consequently, the relationships among these groups are close but far from clear, and likely to change with the addition of more taxa. Especially the delimitation of the Pelecopsis - group is very ambiguous, with some more taxa emerging also distal to the Savignia -group ( Panamomops , Gonatium , Grammonota ).

Conclusions and outlook

To a large extent the genus-level phylogeny of erigonines is still unknown. However, earlier studies and especially the present one have proven that the matrix of Miller and Hormiga (2004) withstands the addition of many, even closely related, taxa. Furthermore, we now have a general picture of the (character) evolution of about a third of all described erigonine genera. In the present analysis, major changes due to taxon addition occurred among genera belonging to Millidge’ s Pelecopsis and Tapinocyba groups and also concerning the position of Venia , the only Afrotropical genus considered so far. Focusing on the addition of taxa from these genus groups and African genera would expand our knowledge of erigonine evolution considerably.

Our analysis provides yet another example of how the addition of closely related taxa can change the topology of the ingroup. But the addition of outgroup taxa is also likely to change the topology of the erigonine tree. Previously, Mynogleninae looked like the sister group of erigonines (e.g. Hormiga 2000 in part; Miller and Hormiga 2004). However, in the recent work by Arnedo et al. (2009), which for the first time also included molecular data on linyphiid spiders, Micronetinae emerged as sister to Erigoninae .

The next important step in reconstructing the phylogeny of erigonine spiders will be to separate plesiomorphic from apomorphic character states by advancing research on the genus-level phylogenies of the potential sister subfamilies.

Among the erigonines we should focus on adding representatives of the remaining 40% of non-monotypic genera. This will give us a general idea of the phylogeny of erigonine spiders and potentially serve as a tool to better understand the character evolution and biogeography of about 70% of their genera and approximately 90% of the corresponding species.

The matrix of Miller and Hormiga (2004) will have its limits as to how many more taxa can be added without consideration of further characters. The present analysis shows that we might be approaching a certain limit already. The topology concerning the newly added taxa is not well supported, as it relies mainly on highly homoplastic characters. If we intend to reconstruct the entire phylogeny of erigonine spiders we should also focus on the addition of more hypotheses of homology. The very complex genital morphology of linyphiid spiders still offers many more characters to be scored, especially if already included discrete characters are refined to account for their various forms.