Bioacoustics of Savignia

Savignia with uncertain species group assignments The cave dwelling S. rostellatra cannot be assigned to any species group based on morphology alone due to its very derived male copulatory organs. It clearly belongs to Savignia having the unique tooth-like basal radical process (character 159; Fig. 21G ) but its inner suprategular apophysis is fused to the suprategulum (distal part), covered with papillae and bearing a tooth ( Song & Li 2009 : fig. 28). Its cephalic lobe, however, is a simple PME lobe that would argue for the mirabilis -group. Savignia ericola as mentioned by the author ( Tanasevitch 2023a ) is conditionally placed in Savignia until a female is found, based on the different shape of the distal suprategular apophysis, lacking the broad inner suprategular apophysis and the distinct shape of the male palpal tibia bearing three long strong setae. We agree with the genus assignment, as S. ericola has the radical conformations, including the unique tooth-like basal radical process (character 159; Fig. 21G ). However, the species group assignments are more difficult. Interpretating the large sclerotised part of the suprategulum as inner suprategular apophysis rather than a distal one, would argue for the mirabilis -group. This would also be supported by the presence of an AME lobe like in S. frontata and lacking any other cephalic lobes. Misplaced species of Savignia Paik (1978) assigned S. pseudofrontata Paik, 1978 to Savignia based on its close resemblance to S. frontata . We do not agree with that: the vulva is spiral and much more complex than the globular spermathecae found in S. frontata and the radix lacks the typical processes. The radical processes and the form of the embolus ( Paik 1978 : figs 48, 49, 51) are somewhat similar to S. harmsi ( Wunderlich 1980a : figs 48–49) and D. lusiscus ( Wiehle 1963 : fig. 30). The cephalic lobe looks like one found in Araeoncus (e.g., Tanasevitch 1987 : fig. 59) and the prolateral tibial apophysis resembles what is found in some species of Diplocephalus (e.g., Isaia 2005 : fig. 4). Based only on these observations and without phylogenetic analyses, it is not possible to assign S. pseudofrontata to any other genus, but it clearly belongs to the Savignia -group. The position of S. harmsi is ambiguous in the phylogenetic analysis but always outside Diplocephalus . It does not have the diagnostic characters of males or females of Savignia . We agree with Wunderlich (1980a) , who mentioned the close resemblance of the male cephalic lobe and the male secondary genital organs to S. fronticornis . Both seem to be intermediate between Araeoncus and Diplocephalus while S. harmsi seems to be closer to Diplocephalus and S. fronticornis closer to Araeoncus . Based on figures (e.g., in Bosmans 1996 : fig. 38), S. fronticornis seems to have a lobe emerging from the retrolateral side of the male palpal tibia similar to Araeoncus . Since it is not possible to solve the appropriate generic assignment without a phylogeny that includes both taxa plus some more potential relatives, we prefer not to draw systematic conclusions for these two species here. One of these potential close relatives is S. naniplopi as discussed in Bosselaers & Henderickx (2002). This species is probably very close to S. harmsi and S. fronticornis based on the radix and its appendices. The cephalic lobe however bears the PME rather than the AME. Also, the epigyne and vulva of S. naniplopi (Bosselaers & Henderickx 2002: fig. 8) are similar to S. fronticornis ( Bosmans 1996 : fig. 40). Simon (1908) described S. erythrocephalus from Western Australia as similar to the Mediterranean S. fronticornis and Coreorgonal monoceros (Keyserling, 1884) from North America ( Millidge 1981 : fig. 147), based on the eyes and the palp but differing from them by the simpler clypeus that is neither elongated nor a small bump. No native erigonines are known from Australia and this species does not belong to the Savignia- group nor to Savignia . Thaler (1984) underlined the problematic position of S. superstes Thaler, 1984 within the erigonines but placed it in the Savignia- group based on the form of the suprategulum and the chaetotaxy despite the lack of a bisected epigyne. The cephalic lobe resembles some species of Diplocephalus (e.g., E. latifrons comb. rev.; Fig. 15F ) but differs in the position of the AME ( Thaler 1984 ) and the epigyne is close to Saloca . Due to the problematic relationships within the Savignia- group he preferred to place in its “typical” genus rather than in another monotypic genus. The current phylogenetic analysis showed that the bisected epigyne (character 178) is synapomorphic for the Savignia genus group (node 21). Savignia is deeply nested in this group so that S. superstes does not belong to Savignia . However, a pointed radical tailpiece (character 163; Fig. 8B ) is found in S. superstes ( Thaler 1984 : fig. 12) that was only observed in three taxa in the outgroup ( Walckenaeria and “ Saloca ” in node 10) pointing to its exclusion from the Savignia- group also. Oi (1960) placed S. kawachiensis Oi, 1960 in Savignia due to the close resemblance of the palp with S. frontata but also mentions the similarity of the cephalic lobe with certain species of Araeoncus . Unfortunately, the figures and text do not allow discussing this assignment. The genus assignment of S. kartalensis is also difficult. It does not belong to Savignia . Based on the literature data only, it is not possible to discuss its assignment, however, we agree with Jocqué (1985) that it belongs to the Savignia-group .

Remarks This genus was described based on male specimens of S. frontata from England according to the six eyes (overlooking the AME on the AME lobe) and other eye characters, which are not diagnostic anymore. The spelling of Savignia was a cause for debate (e.g., Bonnet 1958 ; Bosselaers & Hendrickx 2002; summary of spellings used since the original description in World Spider Catalog 2025 ). Even though Blackwall (1833) dedicated this genus name to Jules-César Savigny he spelled it as Savignia throughout the original description (two times). Bonnet (1958: 3937 , note 39) referring to the rules of nomenclature (art. 8, par. H, note b) neglected the spelling of Savignia ( Blackwall 1833 ) and Savignyia ( Simon 1926 ) and proposed Savignya as the correct spelling. However, according to the International Code of Zoological Nomenclature (1999) , Savignia is an original spelling (art. 32.1.); therefore, the correct spelling of this genus is Savignia as spelled in the original work of Blackwall (1833) . Status Species name Scored according to Savignia (Z) S. amurensis Eskov, 1991 Eskov 1991a : figs 1.1–6 Savignia (M) S. badzhalensis Eskov, 1991 Eskov 1991a : figs 2.1–6 Savignia (Z) S. basarukini Eskov, 1988 Eskov 1988 : figs 53–59 Savignia (Z)* S. birostra ( Chamberlin & Ivie, 1947 ) Marusik 1988 : figs 5–7; Lasut et al . 2009 : figs 1, 4 Savignia (Z) S. borea Eskov, 1988 Eskov 1988 : figs 60–66 Savignia (Z) S. bureensis Tanasevitch & Trilikauskas, 2006 Tanasevitch & Trilikauskas 2006: figs 1–8 Savignia (M) S. centrasiatica Eskov, 1991 Eskov 1991a : figs 3.1–4 Savignia (“M”) S. ericola Tanasevitch, 2023 Tanasevitch 2023a : figs 11–12 Savignia (Z) S. eskovi Marusik, Koponen & Danilov, 2001 Marusik et al . 2001 : figs 49–52 Savignia (M)* S. frontata Blackwall, 1833 Merrett 1963 : fig. 68a–d; Roberts 1987 : figs 36c, 39f Savignia (M)* S. mirabilis ( Eskov, 1988 ) comb. nov. Eskov 1988 : figs 18–24 Savignia (“M”) S. rostellatra Song & Li, 2009 Song & Li 2009 : figs 19–35 Savignia (Z)* S. saitoi Eskov, 1988 Eskov 1988 : figs 67–72 Savignia (M) S. ussurica Eskov, 1988 Eskov 1988 : figs 76–81 Savignia (Z) S. yasudai ( Saito, 1986 ) Saito 1986 : figs 23–24, 26 Savignia (Z)* S. zero Eskov, 1988 Eskov 1988 : figs 83–85; Lasut et al . 2009 : figs 9, 12 Misplaced Savignia species Not Savignia S. erythrocephalus ( Simon, 1908 ) Simon 1908 Not Savignia S. fronticornis ( Simon, 1884 ) Bosmans 1996 : figs 36–40 Not Savignia * S. harmsi Wunderlich, 1980 Wunderlich 1980a : figs 46–51 ??? S. kartalensis Jocqué, 1985 Jocqué 1985 : figs 1–7 ??? S. kawachiensis Oi, 1960 Oi 1960 : figs 154–157 Not Savignia S. naniplopi Bosselaers & Henderickx, 2002 Bosselaers & Henderickx 2002: figs 1–11 Not Savignia S. pseudofrontata Paik, 1978 Paik 1978 : figs 40–51 Not Savignia S. superstes Thaler, 1984 Thaler 1984 : figs 6–14 Eskov (1988) redescribed Savignia according to the slightly curved and backwards directed embolus; the curved distal suprategular apophysis and the leaf-like broadened inner suprategular apophysis; the broadly rounded, terminally unciform prolateral tibial apophysis; the distinctly emarginated anterior part of the epigyne and the pronounced dorsal plate covering half of the oval spermathecae. The current analysis agrees with most of his diagnostic characters but found some more that are discussed hereafter. The genus Savignia is not monophyletic in the current analysis due to two taxa, S. producta and S. harmsi that are both proposed to be transferred to other genera. However, many more taxa that we did not include in our analysis likely do not belong to Savignia as defined by Eskov (1988) . We agree with Eskov (1988) that Savignia has a few very derived characters that are not found in any other genus within the Savigniagroup . According to our analysis this is true for node 46 that includes all Far East Savignia species plus S. frontata and is well supported by a Bremer support of 6, a jackknife support of 99 (MP analysis; Fig. 26 ) and a bootstrap of 100 (ML analysis; Fig. 27 ). We hereby include one more species, S. mirabilis comb. nov. , that is a simpler type of Savignia (node 45) but shows most of the derived characters also. This is why Eskov (1988) described the species as a Diplocephalus rather than a Savignia . In the current analysis we included six out of 24 known species of the genus Savignia (see Table 8 ). Out of all Savignia taxa, eleven occur in Russia (one also in Alaska), four in the Far East ( Japan , China or Korea ), four are known from Europe only and two are occurring all over the Palaearctic. Another one from Australia ( S. erythrocephalus ( Simon, 1908 )) can be excluded from the Savignia- group. One from the Comoro Islands ( S. kartalensis Jocqué, 1985 ) and one from Ethiopia ( S. ericola Tanasevitch, 2023 ) both have uncertain assignments. We included the two Palaearctic species, one that is only found in Europe and three from the most Eastern regions of Russia (around Magadan ). The Russian species are underrepresented in our matrix and the three species from Japan , China and Korea are not represented. Since the Russian species are very closely related to each other it seems that we cover most of the known variety within the Russian species. Since the three species scored from Russia occur in the most Eastern regions (around Magadan ), the species from Japan , China and Korea would not have added generally new morphological features (e.g., figures in Eskov 1988 , 1991a ).

Diagnosis Males of Savignia are well recognisable by the specific embolic division, where the embolus forms a so called “transparent outgrowth” (character 116; Fig. 21G ) and also by the radical processes including an unique tooth-like basal radical process (character 159; Fig. 21G ) and the dorsal radical process having a fold (character 141; Fig. 21D ). Homoplastic but also present in all species of node 45 is the elongated tip of the embolus bearing a slim projection (character 123; Lasut et al . 2009 : figs 4, 8, 12) and the massive inner suprategular apophysis (character 85; Fig. 17G ). Additionally, males often have a triangular palpal tibia with a hook-like prolateral tibial apophysis ending in a pointed tip ( Fig. 19C ). Females are best recognised by the roundly broadened anterior end of the ventral plate (character 182), the straight parallel interior margins of the ventral plate, the triangular shape of the ventral plate and oval-shaped spermathecae, oblique to parallel to each other. However, this is modified in certain species like S. rostellatra ( Song & Li 2009 : figs 33, 35) with very derived characters. Synapomorphies Savignia (node 45; Bremer support of 2, jackknife support of 92 (MP analysis; Fig. 26 ) and a bootstrap of 100 (ML analysis; Fig. 27 )) is supported by the following synapomorphies ( Fig. 30 ): the column apophysis (character 98; Fig. 21D ); the transparent outgrowth at the embolus (character 116; Fig. 21D ); the dorsal radical fold (character 141; Fig. 21D ) and the tooth-like basal radical process (character 159; Fig. 21D ). Additionally, some homoplastic characters support Savignia . Among these are the massive, highly sclerotised inner suprategular apophysis (character 85) and the twisted embolus (character 118). Node 46 corresponds mainly to what Eskov (1988) considered as Savignia excluding S. mirabilis comb. nov. It is worth mentioning this clade since it is supported by many characters that account for most distal species of Savignia and are therefore diagnostic for most species of Savignia (i.e., in the current analysis for all but S. mirabilis comb. nov. with simpler palpal conformations). Node 46 is unambiguously supported by the roundly broadened anterior end of the ventral plate (character 182). Several homoplastic characters also support this clade. Among them are the highly sclerotised, massive, blunt distal suprategular apophysis (character 90, state 10; Fig. 17G ); the distal tooth on the radical tailpiece (character 170; Fig. 20H ) and the AME lobe (character 209; Fig. 15J ). Another character that is found in all Savignia but was not scored in the current analysis is the retrolaterally curved, usually twisted prolateral tibial apophysis of the male palp (e.g., Eskov 1988 : fig. 85).