Grosphus angulatus, Lowe & KOvAřík, 2022

Lowe, Graeme & KOvAřík, FRANtIšEk, 2022, Reanalysis of Teruelius and Grosphus (Scorpiones Buthidae with descriptions of two new species, Euscorpius 356, pp. 1-105 : 75-79

Grosphus angulatus sp. n.

( Figs. 13–15 , 20 , 23–28 , 30–35 , 43–44 , 49–50 , 54 , 71–74 , 93–94, 103–105 , 160 , 165–167 , 175–178 , 180, 182–183 , 189 , 196–209 , 285–324, 326–327 , 393, Tabs. 13–14) http://zoobank.org/urn:lsid:zoobank.org:act:8CA04BBF-

5698-459A-8E69-AA16838667E1

Grosphus sp. nr hirtus Lowe & Kovařík, 2019: 13 , 19, 21, 30–31, 41, 43–44, 54, 60; figs. 26, 36, 40, 107, 111, 158.

TYPE LOCALITY AND TYPE REPOSITORY. Madagascar: Moramanga env., Anjiro, 1995; NZAC, GLPC .

TYPE MATERIAL. Madagascar: Moramanga env., Anjiro, 10.II.1995, 1♀ (holotype), 4♀ (paratypes) NZAC; 1♀ (paratype) GLPC.

ETYMOLOGY. The species name refers to the angulate distal vertex of the modified basal pectinal tooth ( bpt ) in females.

DIAGNOSIS (ADULT FEMALES). Medium-sized member of the genus, total length of adults around 50 mm; base color dark reddish-brown, carapace with weak variegated fuscous pattern; pedipalp patella with obsolete dorsointernal carina; femoral trichobothrium e 1 level with or slightly proximal to d 5; pedipalp chela movable finger with 11–12 median denticle subrows, fixed finger with 13; female bpt with angulate distal vertex; PTC 14– 16; spiracles wide, ovoid in profile; metasoma II–IV with 3–4 robust, dentate granules on posterior dorsosubmedian carinae; telson vesicle hemielliptic in lateral profile, with small subaculear tubercle; morphometrics, L/W ratios (n = 6): metasoma I 0.86– 0.92, metasoma II 1.06–1.10, metasoma III 1.18–1.20, metasoma IV 1.42–1.45, metasoma V 1.84–1.90, pedipalp chela 3.72–3.86, pedipalp femur 2.45–2.74, pedipalp patella 2.06–2.24.

DESCRIPTION (FEMALE). Coloration ( Figs. 285–286 , 310, 312, 314–317 ). Base color dark reddish-brown; carapace with weak variegated fuscosity; coxosternal area and sternites III–VI dark orange-brown; pectines yellow; legs dark-reddish brown to orange-brown, with more pale telotarsi; chelicerae dark reddish-brown, dorsal manus with fuscous anterior margin and reticulation.

Carapace ( Figs. 285 , 287 ). Subrectangular, W/L 1.10–1.19; medial surface level along its entire length; anterior margin slightly concave with small epistomal process; preocular L/ carapace L 0.36-0.38; surface mostly bearing fine granules of moderate density, except in some bilateral smooth strips and areas around central median, posterior median, posterior transverse and posterior marginal furrows; granulation more coarse and dense on preocular triangle; superciliary carinae granulate; macrosetae absent; lateral eye groups composed of either 3 large + 2 small ocelli (8/12 groups), or 2 large + 2 small ocelli (4/12 groups), i.e., type 5 or type 4B, respectively ( Loria & Prendini, 2014); 4 carapaces with type 5/5 pattern, 2 carapaces with type 4B/5 pattern, 1 carapace with type 4B/4B pattern (left group/right group); median eyes of moderate size, eye diameter/ carapace L 0.086.

Chelicerae ( Figs. 310–313 ). Dorsal surface of manus granulate on anterior 1/5, smooth on posterior 4/5; anterior granulate area with 8–9 macrosetae, 4–5 pale, fluorescent microsetae; dorsointernal carina strong, granulate; fingers with typical buthid dentition (Vachon, 1963), movable finger dorsal margin with two large subdistal denticles and two small basal denticles, ventral margin with subdistal and basal denticles (notched basal denticle in Figs. 312–313 is atypical or worn, and not present in the holotype and other paratypes), fixed finger with large subdistal denticle and proximal bicusp, two denticles on ventral surface; dorsal surface of movable finger smooth, with dorsal row of 5–7 pale, fluorescent microsetae.

Coxosternal area ( Figs. 286 , 288 ). All coxae smooth with sparse macrosetae and fluorescent microsetae; sternum smooth, subtriangular, with narrow slit-like posteromedial depression, bearing 2 macrosetae; genital opercula smooth, divided.

Pectines ( Figs. 28 , 32 , 34–35 , 43–44 , 286 , 288 , 320–335 ). Basal piece smooth, with concave anterior margin, surface flat without groove or pit, pectines with 3 marginal lamellae, 4–6 middle lamellae, extending to distal end of coxa IV; marginal and middle lamellae with sparse cover of macrosetae and pale, fluorescent microsetae; fulcra with 2 fluorescent microsetae; bpt with angulate distal vertex.

Hemispermatophore. Unknown.

Mesosoma ( Figs. 54 , 71–74 , 93–94 , 285–288 , 393). Tergites: pretergites smooth, with microsulcate posterior margins; tergites densely, finely granulated, with narrow, smooth transverse lateral strips on tergites II–VI; tergite I without discernible carinae, tergites II–VI with single weak, granulate median carina, tergite VII with medial hump and 2 pairs of granulate carinae; all tergites lacking macrosetae. Sternites: sternites III–VI smooth, acarinate; sternite VI smooth with two pairs of weak, granulate carinae; posterior margins of all sternites smooth; spiracles broad, hemi-elliptic; sternite III–VI macrosetae: one submedian pair, one lateral pair, two posterior marginal pairs; sternite VII macrosetae: one submedian carinal pair, one lateral pair; sternites III–VI glossy, sternite VII matte.

Metasoma ( Figs. 105 , 160 , 285–286 , 301–303 ). Segments of uniform width, robust. Carination: segments I–III with 10 complete carinae, IV with 8 complete carinae (lateral median carinae indistinct posteriorly), V with 5 complete carinae; all carinae granulate; dorsosubmedian carinae on II–VI with enlarged dentate posterior granules; ventrolateral carinae on V strongly, uniformly granulate; dorsolateral carinae on V irregularly, coarsely granulate; lateral anal margin with 2 small granules, ventral anal margin with up to 18 granules. Intercarinal surfaces: moderately dense, fine granulation on lateral, ventrolateral and ventral surfaces of all segments; I–VI with dorsomedian surfaces finely granulated or shagreened, with decreasing density on more posterior segments, dorsolateral surfaces with sparse fine granules, mostly smooth; V with dorsomedian surface smooth, dorsolateral surfaces with sparse fine granules, mostly smooth. Setation: dorsal surfaces without setae; other surfaces with numerous short macrosetae and fluorescent microsetae, mostly associated with carinae.

Telson ( Figs. 285–286 , 301–303 ). Vesicle dorsal surface smooth, lateral and ventral surfaces covered with numerous coarse and fine granules, except along lateral and paramedian longitudinal strips which are smooth; vesicle hemi-elliptic in lateral profile, with distinct subaculear tubercle; numerous short macrosetae and fluorescent microsetae on lateral and ventral surfaces; dorsal surface with posterior patch of short macrosetae; aculeus shorter than vesicle.

Pedipalps ( Figs. 289–300 , 304–309 , 318–319 ). Segments robust (see DIAGNOSIS for morphometrics). Femur: dorsointernal, dorsoexternal and ventrointernal carinae distinct, granulate; other carinae indistinct; intercarinal surfaces smooth; sparse short macrosetae and fluorescent microsetae present, mostly associated with granules and carinae. Patella: all carinae obsolete, smooth except for isolated granules on internal surface; numerous short macrosetae and fluorescent microsetae present. Chela: all carinae obsolete, surfaces smooth, with dense cover of short macrosetae and fluorescent microsetae; 11–12 median denticle subrows on fixed finger, 13 on movable finger including short subdistal row (excluding malformations), all subrows except proximal flanked by one mid-row internal and two proximal external accessory denticles. Trichobothriotaxy: orthobothriotaxic, type Aα (Vachon, 1974), femur d 2 internal, e 1 level with or slightly proximal to d 5; chela fixed finger db proximal to est.

Legs ( Figs. 285–286 , 314–317 ). Femora and patellae with granulate carinae, prolateral surfaces sparsely granulate and matte, retrolateral surfaces smooth, glossy; tibial spurs present on legs III–IV; retrolateral tarsal spurs simple, prolateral tarsal spurs with very small basal bifurcation; basitarsi with 2 axial rows of macrosetae on ventral surface, irregular macrosetae on lateral and dorsal surfaces; telotarsi with 2 axial rows of up to 8 short macrosetae on ventral surface, lateral apices with 4–6 macrosetae; tarsal ungues stout.

Measurements. See Tables 13–14.

AFFINITIES. G. angulatus sp. n. is similar to three other species of the genus: G. hirtus , G. polskyi and G. voahangyae . The four species share the following characters: metasoma I stout, mean L/W ratio ♀ <0.97, ♂ <1.02 ( Fig. 168 ); dorsointernal carina of pedipalp patella with granulation sparse or absent; pedipalp chela fixed finger with trichobothrium db level with or proximal to 0.92 est ( Fig. 167 ); leg IV tibial spur L/ tibia distal D, mean ratio 0.60–0.65 ( Fig. 50 ). These four species were often recovered as a monophyletic ‘ hirtus ’ group in cladistic analyses (e.g., Figs. 196–200, 203–209 ).

G. hirtus differs from G. angulatus sp. n. as follows: lighter base color of yellowish to reddish-yellow, with more distinct variegated fuscous patterns on body, pedipalps and legs; pedipalp patella with weakly granulate dorsointernal carina; femoral trichobothrium e 1 distal to d 5; female bpt rounded distally ( Fig. 325 vs. Figs. 320–324 ), z-scores of Fourier harmonics clustered separately ( Figs. 34–35 ), nonoverlapping in PC1-PC2 plane ( Fig. 326 ), disjunct in PC3- PC4 plane ( Fig. 327 ); metasoma II–IV with smaller dentate granules on posterior dorsosubmedian carinae; more elongate pedipalp patella, L/W ♀ 2.21–2.42. G. voahangyae differs from G. angulatus sp. n. as follows: smaller size, adult female total length around 40 mm; more distinct variegated fuscous patterns on body, pedipalps and legs; femoral trichobothrium e 1 distal to d 5; female bpt ovoid, rounded distally, z-scores of Fourier harmonics clustered separately ( Figs. 34–35 ); more elongate pedipalp patella, L/W ♀ 2.22–2.30. Only the male is known for G. polskyi , but the following differences from G. angulatus sp. n. do not exhibit strong sexual dimorphism in species of Grosphus for which both sexes are known: smaller size, adult male total length around 33 mm; lighter base color of reddish-yellow, with more distinct variegated fuscous patterns on body; femoral trichobothrium e 1 distal to d 5; metasoma II–IV without enlarged dentate granules on posterior dorsosubmedian carinae; telson with larger subaculear tubercle. The species G. tavaratra is also similar, but only the male is known. However, G. tavaratra differs in having a granulate dorsointernal carina on the pedipalp patella, femoral trichobothrium e 1 distal to d 5, and narrower spiracles, traits that do not exhibit strong sexual dimorphism in species of Grosphus for which both sexes are known.

Figures

Figure 34: Phenetic analysis of shapes of female basal pectinal teeth (bpt). Ultrametric tree obtained from hierarchical cluster analysis by UPGMA of the Euclidean distance matrix of z-scores of 32 Fourier coefficients. Font colors: Grosphus, blue; Teruelius flavopiceus, orange; T. ankarana, magenta; T. grandidieri, red; other Teruelius spp., dark yellow. Pectine images: Grosphus angulatus sp. n. (upper), Teruelius grandidieri (lower).

Figure 35: Phylogram of shapes of female basal pectinal teeth (bpt). Tree obtained from neighbor-joining cluster analysis of the Euclidean distance matrix of z-scores of 32 Fourier coefficients. Outgroup taxon: Pseudolychas ochraceus. Font colors: as in Fig. 34, with Pseudolychas black. Pectine images: Teruelius grandidieri (upper), Grosphus angulatus sp. n. (lower).

Figure 285. Grosphus angulatus sp. n., habitus. Holotype female, dorsal view. Scale bar: 5 mm.

Figure 286. Grosphus angulatus sp. n., habitus. Holotype female, ventral view. Scale bar: 5 mm.

Figures 287–288. Grosphus angulatus sp. n., holotype female. Carapace and tergites (287) and coxosternal area and sternites (288). Scale bar: 2 mm. UV fluorescence.

Figures 289–303. Grosphus angulatus sp. n. Figures 289–300. Left pedipalp chela (289, 292, 295, 298), patella (290, 293, 296, 299) and femur (291, 294, 297, 300) in dorsal (289–291), external (292–294), ventral (295–297) and internal (298–300) views. All are holotype female except external patella (293) from a paratype female, as the left trichobothrium esb is missing in the holotype. Scale bar: 2 mm. UV fluorescence. 1 Figures 301–303. Metasoma and telson of paratype female in ventral (301), left lateral (302) and dorsal (303) views. Scale bar: 4 mm. UV fluorescence.

Figures 304–309. Grosphus angulatus sp. n., pedipalp segments of female holotype (304–306, 308–309) and female paratype (307) with trichobothrial pattern indicated. Chela in external (304) and ventral (305) views. Patella in dorsal (306) and external (307) views. Femur in dorsal (308) and internal (309) views.

Figures 310–319. Grosphus angulatus sp. n. Figures 310– 313. Paratype female, right chelicera in dorsal (310–311) and ventral (312–313) views, under white light (310, 312) and UV fluorescence (311, 313). Scale bar: 1 mm. Figures 314– 317. Paratype female, right basitarsi and telotarsi in retrolateral views, legs I (314), II (315), III (316) and IV (317). Scale bar: 1 mm. Figures 318– 319. Holotype female, right chela dentition, fixed finger (318), movable finger (319). UV fluorescence. Scale bar: 1 mm.

Figures 320–327. Grosphus angulatus sp. n., variation in femalebpt. Figures 320– 324. Basal pectines showing bpt shapesof holotype(322)and four paratypes (320–321, 323– 324). White arrow (320) indicates partially modified intermediate tooth bearing a sensorial area (cf. Figs. 43–44). Figure 325. Basal pectines of female G. hirtus showing typical bpt shape.UV fluorescence (320–325).Figures 326– 327. Bivariate scatter plots of bpt scores of G. angulatus sp. n. (black circles) and G. hirtus (gray circles) for principal components PC2 vs. PC1 (326) and PC4 vs. PC3 (327) obtained from PCA of 32 Fourier coefficients from up to eighth order harmonic terms in EFA of bpt shapes of Grosphus and Teruelius (cf. Figs. 29– 33).

Figures 13–16. Morphometric analyses of carapace and hemispermatophores. Figure 13. Horizontal bar plot of mean carapace length (mm) (character 0) of Grosphus (n = 46, 14 spp.), Teruelius (n = 70, 21 spp.), Pseudolychas (n = 8, 3 spp.), and other outgroup taxa (n = 36, 9 spp.). Data from both sexes pooled. Error bars are standard errors. Discretization thresholds at step changes in ranked length. Figure 14. Horizontal bar plot of mean concavity angle (°) (character 4) of Grosphus (n = 49, 14 spp.), Teruelius (n = 72, 19 spp.), Pseudolychas (n = 8, 3 spp.) and other outgroup taxa (n = 22, 9 spp.). Data from both sexes were pooled. Error bars indicate standard errors. Discretization threshold at step transition in rank slope. Inset: angle defined by tangent line at midpoint between anterior-most lateral eye and carapace center. Figure 15. Horizontal bar plot of mean ratio of carapace preocular L/ carapace L (character 5) of Grosphus (n = 39, 14 spp.), Teruelius (n = 71, 21 spp.), Pseudolychas (n = 8, 3 spp.) and other outgroup taxa (n = 23, 9 spp.). Data from both sexes pooled.Error bars are standard errors.Discretization threshold at a minor step transition in rank slope. Figure 16. Bivariate logarithmic scatter plot of hemispermatophore posterior lobe width/ length ratio vs. hemispermatophore capsule length/ posterior lobe length for Grosphus (n = 13, 6 spp.), Teruelius (n = 10, 9 spp.) and outgroup Pseudolychas (n = 1, 1 sp.). Color codes of symbols or bars as indicated in Fig. 16 legend: Grosphus ‘hirtus’ group (G. angulatus sp. n., G. hirtus, G. polskyi, G. voahangyae), blue; other Grosphus spp., cyan; Teruelius flavopiceus, orange; T. ankarana, magenta; T. grandidieri, red; other Teruelius spp., yellow; Pseudolychas spp., black; other outgroups, gray.

Figures 19–22. Morphometric analyses of regular pectine teeth. Figures 19–20. Horizontal logarithmic bar plots comparing length/ width (L/W) ratios of regular pectine teeth in males (19) and females (20) of Grosphus, Teruelius and outgroup taxa. Bars are rank ordered means, error bars are standard errors; ♂ n = 52, 30 spp.; ♀ n = 66, 31 spp. Figure 21. Bivariate logarithmic scatter plot comparing pectinal tooth L/W ratios of males (ordinate) vs. females (abscissa) (21 spp.). Gray line is diagonal. Plotted values and error bars as in Figs. 19–20. Fig. 22. Regular pectine teeth (♀) of Grosphus voahangyae (left) and Teruelius ankarafantsika (right), showing measurements of length (L) and width (W). W is equal to inter-fulcral spacing. UV fluorescence. Measurements were taken at> 3 teeth away from most proximal or most distal teeth. Color codes of symbols and bars as in Fig. 16–17 legends.

Figures 23–28: Ratiometric analysis of shapes of female basal pectinal teeth (bpt). Figures 23–26. Bivariate logarithmic scatter plots of six ratiometric shape variables: roundness vs. solidity (23), circularity vs. roundness (24), maximum/ minimum caliper diameter vs. solidity (25), basal tooth width/ regular tooth width vs. perimeter attachment ratio (26). Figures 27–28. Bivariate scatter plots of first two principal components (PC2 vs. PC1) obtained from PCA of standardized logarithms of all six ratiometric variables, accounting for 85.31% and 8.43% of variance, respectively. Individual cases plotted in Fig. 27, means and standard errors for each species in Fig. 28. Profile silhouette examples are shown for analyzed species in Fig. 28. Data from 106 bpt from Grosphus (n = 31, 8 spp.), Teruelius (n = 70, 18 spp.) and Pseudolychas (n = 5, 2 spp.). Symbol colors indicated in legend of Fig. 23.

Figures 29–33: Elliptic Fourier analysis (EFA) of shapes of female basal pectinal teeth (bpt). Figure 29. Logarithmic plots of harmonic power (sums of squares of Fourier coefficients) vs. harmonic order for fits to bpt outlines of Grosphus (blue symbols) and Teruelius (yellow symbols). Error bars indicate ranges (minimum to maximum). Upper inset: examples of EFA fits to bpt from Grosphus (G. voahangyae) (left) and Teruelius (T. olgae) (right) by Fourier series with cumulative terms up to and including second (blue), fourth (green) and eighth (red) order harmonics. Contours of bpt oriented with perimeter attachment horizontal, start point at proximal vertex, area normalized.Figures 30–31. Bivariate scatter plots of bpt scores for first three principal components,PC2 vs. PC1 (Fig.30) and PC3 vs. PC2 (Fig.31), obtained from PCA of 32standardized Fourier coefficients from up to eighth order harmonic terms, accounting for 31.02%, 19.14% and 8.20% of variance, respectively (total variance 58.36%). Lower inset in Fig. 31: scree plot of eigenvalue vs. PC number.Figure 32. Bivariate scatter plot of means and standard errors of bpt scores of first two principal components, PC2 vs. PC1, for each species in Fig. 30.Figure 33. Trivariate scatter plot of bpt scores of first three principal components (PC1, PC2, PC3) rendered as 3D cross stereoscopic pair. Symbol colors as in legend of Fig. 30.Analyzed data set as in Figs. 23–28.

Figures 36–44. Basal pectinal teeth and basal middle lamellae of female scorpions.Figures 36–38. Basal pectinal structures of female Grosphus voahangyae (36) and Teruelius ankarafantsika (37–38), showing basal pectinal tooth (bpt) and basal middle lamella (bml). Terminology of Lowe & Kovařík (2019) (37) is contrasted with that of Lourenço et al. (2020) (38): middle lamellae: m1, m2, m3, m4, ....; pectine teeth: t1, t2, t3, t 4, .... Figures 39–40. Basal pectinal structures of female Uroplectes vittatus (39) and U. planimanus (40), showing basal pectinal tooth (bpt) and basal middle lamella (bml). Figures 41–42. Basal pectinal structures of female Stahnkeus subtilimanus (41) and Serradigitus wupatkiensis (42), showing basal middle lamella (bml) and multiple modified basal pectinal teeth (t, t, t, t), one of which bears a sensorial area (s). Figures 1 2 3 4 43–44. Basal pectinal structures of a female Grosphus angulatus sp. n., showing basal middle lamella (bml), modified basal pectinal tooth (bpt = t1) regular pectinal teeth (t3, t4, etc.), and a partially modified intermediate tooth (t2) bearing a sensorial area (s). Scale bars: 500 μm (36, 44), 1 mm (37–43). UV fluorescence, excitation by 395 nm LED (380–410 nm), emission filter 475 nm long pass (Edmund Optics 64633) (36–43),

Figures 45–52. Tibial spurs and tarsal setation. Figures 45–46. Tibial and tarsal segments of leg III (45) and leg IV (46) of Grosphus madagascariensis in retrolateral view. Figures 47–48. Tibial and tarsal segments of leg III (47) and leg IV (48) of Teruelius limbatus in retrolateral view.Abbreviations:Tb, tibia; Bt, basitarsus (tarsomere I); Tt, telotarsus (tarsomere II); ts, tibial spur; vs, ventral setae of telotarsus. Figures 49–50. Horizontal bar plots of the mean ratio of tibial spur L/ tibia distal D for leg III (49) and leg IV (50) (characters 10 and 11, respectively) of Grosphus (n = 34, 9 spp.), Teruelius (n = 66, 21 spp.), Pseudolychas (n = 2, 2 spp.) and other outgroup taxa (n = 15, 8 spp.). Data from both sexes pooled. Error bars are standard errors. Discretization thresholds shown at step changes in ranked length. Color codes of bars as indicated in Fig. 16 legend. Figures 51–52. Ventral telotarsal setation. Figure 51. Logarithmic distribution of number of macrosetae on ventral telotarsus III (character 12) in Grosphus (cyan bars and symbols, n = 36, 8 spp.) and Teruelius (yellow bars and symbols, n = 12, 10 spp.). Macrosetal counts on abscissa plotted on logarithmic scale. Insets: UV fluorescence photomicrographs and maps of macroseta sockets (red dots) of ventral telotarsus III in Grosphus simoni (left) and Teruelius flavopiceus (right). Figure 52. Horizontal bar plot of ratios of mean blue channel intensities of dorsal vs. ventral regions-of-interest (ROI) (white boxes) in telotarsal images of Grosphus (cyan bars) and Teruelius (yellow bars). Insets: control images (*) of T. haeckeli sp. n. (a, leg III) and G. madagascariensis (c, leg III), original (upper) and resampled (lower); test images of T. sabineae (b, leg IV) and G. ambre (d, leg IV) (resampled from: Lourenço & Wilmé, 2016; Lourenço et al., 2018). Error bars: standard deviations of ROI pixels.

Figures 53–66. Ventral views of sternites IV–VII and metasoma I. Figures 53–57. Grosphus: G. madagascariensis, ♂ (53), G. angulatus sp. n., ♀ (54), G. simoni, ♀ (55), G. voahangyae, ♀ (56), and G. hirtus, ♀ (57). Figure 58. Pseudolychas transvaalicus, ♀. Figure 59. Lychas mucronatus, ♀. Figures 60–66. Teruelius: T. ankarafantsika, ♀ (60), T. ankarana, ♂ (61), T. grandidieri, ♀ (62), T. limbatus, ♂ (63), T. flavopiceus, ♂ (64), T. mahafaliensis, ♀ (65), and T. rossii, ♂ (66).

Figures 67–90. Cuticular surface microstructure of medial sternites VII and VI. Figures 67–70. Grosphus madagascariensis, ♂. Sternites VII (67, 69) and VI (68, 70). Figures 71–74. Grosphus angulatus sp. n., ♀. Sternites VII (71, 73) and VI (72, 74). Figures 75–78. Lychas mucronatus, ♀. Sternites VII (75, 77) and VI (76, 78). Figures 79–82. Teruelius limbatus, ♀. Sternites VII (79, 81) and VI (80, 82). Figures 83–86. Teruelius flavopiceus, ♀. Sternites VII (83, 85) and VI (84, 86). Figures 87–90. Teruelius ankarana, ♀. Sternites VII (87, 89) and VI (88, 90). Images acquired under reflected white light epi-illumination at lower magnification (scale bar 100 μm in Fig. 68) of sternites on intact animal (67–68, 71–72, 75–76, 79–80, 83–84, 87–88), and Nomarski trans-illumination at higher magnification (scale bar 40 μm in Fig. 70) of dissected sternites after soft tissue removal (69–70, 73–74, 77–78, 81–82, 85–86, 89–90).

Figures 91–104. Light reflection properties of sternites VII and VI. Figures 91–102. Spatial distribution of intensity at 45° reflection angles of laser beam with 45° incidence to normal on sternites VII (odd numbered figures) and VI (even numbered figures) from Grosphus madagascariensis ♂ (91–92), Grosphus angulatus sp. n. ♀ (93–94), Lychas mucronatus, ♀ (95–96), Teruelius limbatus, ♀ (97–98), Teruelius flavopiceus, ♀ (99–100), and Teruelius ankarana, ♀ (101–102). Illumination source: 650 nm laser diode (650MDLC-5-1235), 5 mW, focused to 40 μm diameter spot on medial areas of sternites. Imaging device: Canon EOS 7D Mark II digital camera with 100 mm f/2.8 macro lens focused on translucent white diffuser screen intercepting reflected beams. TIFF files generated by linear RAW conversion. Scale: largest circle in each bounding box subtends 0.1022 sr. Figure 103. Horizontal bar chart comparing mean radii of dispersion of reflected beams in Figs. 91–102. Means obtained from pixel-normalized radial density functions computed by ImageJ 1.52a Radial Profile plugin, normalized to mean radius of specular reflection off a silver front surface mirror. Inset: sternites VI and VII from T. ankarana, ♀. Figure 104. Horizontal bar chart comparing mean intensities of reflected beams shown in Figs. 91–102. Mean intensities computed for pixels within central circle subtending 0.0637 sr, expressed as percentage mean intensity of specular reflection off a silver front surface mirror.

Figures 105–129. Ventral views of metasomal segment I showing variation in granulation and development of paired ventrosubmedian of carinae in Grosphus and Teruelius. Species and sex indicated in labels under each figure. White arrows identify ventromedial intercarinal surfaces between paired ventrosubmedian carinae in figures with oblique views of the segment. Reflected white light illumination: 106–107, 114, 117, 126–129; UV fluorescence: 105, 108–113, 115–116, 118–125; resampled from published images: 106, 117, 126–127 (Lourenço et al., 2009, 2018; Ref. MNHN-RS-RS1314). Material of Figs. 110–111 determined and labeled as ‘Grosphus garciai’ (FMNH 73434, FMNH 73436), currently a junior synonym of G. hirtus.

Figures 150–164. Granulation of ventrosubmedian carinae of metasomal segment I. Figures 150–157. Optical image analysis of carinal granulation in Grosphus hirtus, ♂ (150–153) and Teruelius mahafaliensis, ♂ (154–157). Granule patterns (white on black) resolved by binary thresholding of image gray level with maximum entropy algorithm (150, 154), carinae traced as piecewise linear paths following granules or ridges (151, 155, red lines, UV fluorescence), fluctuations in gray levels (8-bit) along traced carinae associated with granulation, normalized to total carinal length (152, 156), and power spectra of fluctuations after subtraction of mean gray level and linear trends (153, 157). Images resampled to 340 pixel width (bilinear down-sample, or cubic up-sample) and smoothed by Gaussian filter, radius 2 pixels (thresholding) or 1 pixel (spectral analysis). Scale bars: 400 μm (151, 155). Figure 158. Bivariate logarithmic scatter plot of integrated power of gray level fluctuations (spatial frequency range 10–26 granules/ carina) vs. mean length of granules along carinal axis resolved by binary thresholding for Grosphus (cyan symbols, n = 30, 8 spp.) and Teruelius (yellow, orange, magenta and red symbols, n = 44, 18 spp.). T. feti represented by juvenile holotype male. ‘Granule’ length = 1 if thresholded regions merge into single carina. Figures 159–164. Ventral views of metasomal segment I showing different carinal granulation in Grosphus vs. Teruelius. G. madagascariensis, ♂ (159), G. angulatus sp. n., ♀ (160), T. flavopiceus, ♀ (161), T. ankarana, ♀ (162), T. grandidieri, ♂ (163), and T. limbatus, ♀ (164). Reflected white light images. Scale bars: 2 mm.

Figures 165–168. Metasoma and pedipalp characters. Figure 165. Horizontal bar plot of mean ratio R = d(Eb2, Eb3)/ d(Eb1, Eb2), of inter123 trichobothrial distances on pedipalp chela manus (character 36) for Grosphus (n = 50, 7 spp.), Teruelius (n = 61, 13 spp.), Pseudolychas (n = 6, 3 spp.) and other outgroup taxa (n = 25, 8 spp.). Error bars are standard errors. Discretization threshold at step change in ranked ratio. Figure 166. Bivariate logarithmic scatter plot of positions of pedipalp fixed finger trichobothria db vs. est for Grosphus (n = 40, 13 spp.) and Teruelius (n = 57, 21 spp.). Distances of db and est to distal terminus of fixed finger normalized to corresponding distance of manus Et (inset diagrams). Figure 167. Horizontal logarithmic bar plot of mean ratio of distances of trichobothria db and est from tip of fixed finger (character 35). Data as in Fig. 166. Error bars are standard errors. Discretization threshold at step change in ranked ratio. Color codes of symbols in Fig. 166 and bars in Figs. 165, 167–168 as in Fig. 16 legend. Figure 168. Horizontal bar plot of mean ratio of metasoma I L/W in males (character 18) of Grosphus (n = 20, 12 spp.), Teruelius (n = 28, 18 spp.), Pseudolychas (n = 5, 2 spp.) and other outgroup taxa (n = 12, 9 spp.). Error bars are standard errors. Discretization thresholds at step changes in ranked ratio.

Figures 173–178: Analysis of position of petite ‘trichobothrium’ d on pedipalp femur of Grosphus, Teruelius and Pseudolychas. Figures 2 173–174. Cartesian x-y coordinates for digitizing positions of d and dorsointernal carina in Grosphus madagascariensis (173) and Teruelius 2 olgae (174). Coordinates of d 2 are (d 2x, d 2y) indicated by cyan (173) or yellow (174) symbol. Positions of granules of dorsointernal carina indicated by black symbols. Granules fitted by B-spline (magenta curves) or parametric function (green curves). Ymax: asymptotic value of parametric function. Figure 175. Bivariate scatter plot of minimum distances, s, of d from parametric curve vs. d from B-spline curve (both 2 2 normalized against Y) for femora of Grosphus (n = 34, 14 spp.), Teruelius (n = 49, 20 spp.) and Pseudolychas (n = 3, 3 spp.). Gray diagonal max line: least squares regression for Grosphus and Teruelius data, R = 0.8131, P <0.0001. Symbol colors indicated in legend. Black arrows mark isolated outlier points. Figure 176. Bivariate scatter plot of means of minimum distances for species of Grosphus and Teruelius, summarizing data of Fig. 175. Error bars are standard errors. Gray diagonal line: least squares regression, R = 0.9700, P <0.0001. Figure 177. Bivariate scatter plot of normalized x-y coordinates of d 2 for data in Fig. 175. Abscissas normalized against femur length, L femur, ordinates against Y max. Figure 178. Bivariate scatter plot of means of normalized x-y coordinates for species of Grosphus and Teruelius, summarizing data of Fig. 177. Error bars are standard errors. Data extracted from images of specimens and published figures of femur in dorsal aspect showing granules and trichobothria, in which d 2 could be identified.

Figures 179–182. Morphometrics of telson lateral profiles. Figure 179. Elliptic Fourier analysis (EFA) of lateral profiles. Upper panel: EFA curve fits to profiles of Grosphus hirtus, G. mandena, Teruelius ankarafantsika and T. annulatus by Fourier series with cumulative terms up to and including second (blue), fourth (green), eighth (red) and sixteenth (orange) order harmonics. Telson profiles oriented with dorsal surface horizontal, start point at anterior limit of vesicle, peduncle truncated, area normalized. Lower panel: Histogram plots of harmonic loadings of first two principal components, PC1 and PC2, obtained from PCA of 64 standardized Fourier coefficients from up to sixteenth order harmonic terms, accounting for 30.33% and 19.54% of variance, respectively (total variance 49.87%). Bars are heat map coded by loading values, harmonics with highest positive loadings labelled (red bars). Figure 180. Bivariate scatter plot of subspace of first two principal components rotated (PC1*, PC2*) to minimize variance of Teruelius spp. (yellow symbols) along vertical axis. Data from 117 telson profiles from Grosphus (n = 35, 14 spp.), Teruelius (n = 65, 21 spp.), Pseudolychas (n = 6, 3 spp.) and other outgroup taxa (n = 11, 4 spp.), both males and females. Upper inset: scree plot of eigenvalue vs. PC number. Lower inset: legend for symbol colors: Grosphus ‘hirtus’ group (= G. angulatus sp. n., G. hirtus, G. polskyi, G. tavaratra, G. voahangyae), blue; other Grosphus spp., cyan; Teruelius flavopiceus, orange; T. ankarana, magenta; T. grandidieri, red; other Teruelius spp., yellow; Pseudolychas spp., black; other outgroup taxa, gray. Figure 181. Bivariate scatter plot of mean values of PC2* vs. PC1* for male telson profiles of Grosphus (12 spp.), Teruelius (16 spp.), Pseudolychas (3 spp.) and other outgroup taxa (4 spp.). Figure 182. Bivariate scatter plot of mean values of PC2* vs. PC1* for female telson profiles of Grosphus (7 spp.), Teruelius (15 spp.), Pseudolychas (3 spp.) and other outgroup taxa (4 spp.). Profile silhouette examples shown for analyzed species in Figs. 181–182. Error bars in Fig. 182 are standard errors. Symbol colors in Figs. 181–182 as in legend of Fig. 180.

Figure 183. Morphometrics of telson lateral profiles. Trivariate scatter plot of first three unrotated principal components (PC1, PC2, PC3) of harmonics extracted by Elliptic Fourier analysis (Fig. 179), rendered as a 3D cross stereoscopic pair. Symbol colors as in legend of Fig. 180.

Figures 184–189. Variation in UV fluorescence of preserved material. Figures 184–185. Teruelius limbatus, female in dorsal (184) and ventral (185) views under uniform UV Illumination (395 nm LED array). White arrows in Fig. 185 indicate areas on right lateral metasoma IV, V and telson with strong photobleaching. Scale bar: 10 mm. Figures 186–187. Histograms showing calculated distributions of fluorescence intensity for left lateral metasoma V and telson (186), and right lateral metasoma V and telson (187). Pixel intensities calculated from JPEG images by luminance grayscale conversion and inversion of the OECF of the camera (Canon EOS 5DsR). OECF estimated by 5th-order polynomial fit to relationship between gray values of 66 linear RAW converted vs. sRGB encoded images of a uniform test target illuminated by white light LED driven by variable current source. Figure 188. Distribution of measured fluorescence intensity (photodetector current) of medial sternite VI in Grosphus (cyan bars and symbols, n = 25, 5 spp.) and Teruelius (yellow bars and symbols, n = 39, 7 spp.). Inset image: Grosphus madagascariensis (♂) (left) and Teruelius ankarana (♂) (right) under UV light. Figure 189. Raw data of Fig. 188 plotted by species. Each symbol represents a measurement from one specimen. Horizontal lines indicate ranges of variation. Numbers on the right are coefficient of variation (CV). Cyan symbols: Grosphus; yellow symbols: Teruelius; green symbols: reference data from control batches of two buthids,

Figures 196–199. Examples of MPTs retrieved by phylogenetic analyses of Grosphus and Teruelius, rooted by 4 different outgroup taxa: Charmus laneus (196), Karasbergia methueni (197), Lychas mucronatus (198) and Microcharmus variegatus (199). Weighting schemes and data matrices indicated above each tree (see Table 7). Numbers above nodes are jackknife by symmetric resampling supports, those below relative Bremer supports. Color coding of groups according to legend in Fig. 16.

Figures 200–203. Examples of MPTs retrieved by phylogenetic analyses of Grosphus and Teruelius, rooted by 4 different outgroup taxa: Pseudolychas ochraceus (200), P. pegleri (201), Tityobuthus monodi (202) and Somalicharmus whitmanae (203). Weighting schemes and data matrices indicated above each tree (see Table 7). Numbers above nodes are jackknife by symmetric resampling supports, those below relative Bremer supports. Color coding of groups according to legend in Fig. 16.

Figures 204–205. Mapping of unambiguous synapomorphies in two example MPTs retrieved by phylogenetic analyses of Grosphus and Teruelius, rooted by 2 different outgroup taxa: Charmus laneus (204) and Pseudolychas ochraceus (205). Numbers above nodes are discrete characters (Table 1), those below are discrete character states (Table 2).

Figures 206–209. Examples of MPTs retrieved by phylogenetic analyses of Grosphus and Teruelius, with multiple outgroup taxa in topologies fixed by backbone constraints. Figures 206–207. MPTs retrieved with 11 outgroup taxa in constrained topologies determined by the buthid MPT of Fig. 276, from discrete characters (206) or discrete and continuous characters (207), analyzed under implied weights. Figures 208–209. MPTs retrieved with 7 outgroup taxa in constrained topologies determined by molecular phylogenetic analyses of buthids (Fig. 190), from discrete characters (208) or discrete and continuous characters (209), analyzed under prior (209) or implied (208) weights. Numbers above nodes are jackknife by symmetric resampling supports, those below relative Bremer supports. Color coding of groups according to legend in Fig. 16.

Figures 328–331. Teruelius haeckeli sp. n., habitus. Figures 328– 329. Holotype male in dorsal (328) and ventral (329) views. Figures 330–331. Paratype female in dorsal (330) and ventral (331) views. Scale bars: 10 mm.

Figures 332–333. Teruelius haeckeli sp. n., holotype male. Carapace and tergites (332) and pectinal area and sternites (333).

Figures 334–335. Teruelius haeckeli sp. n., holotype male. Carapace and tergites (334) and coxosternal area and sternites (335). UV fluorescence.