Parabuthus Pocock, 1890

Kovařík, František, Lowe, Graeme, Plíšková, Jana & Šťáhlavský, František, 2016, Scorpions of the Horn of Africa (Arachnida: Scorpiones). Part VII. Parabuthus Pocock, 1890 (Buthidae) with description of P. hamar sp. n. and P. kajibu sp. n. from Ethiopia, Euscorpius 228, pp. 1-58 : 2-10

Parabuthus Pocock, 1890

( Figs. 1–204, Tables 1–2)

Buthus (Parabuthus) : Pocock, 1890: 124–125.

Parabuthus: Pocock, 1895: 309–314 , plate IX, figs. 4a–d; Fet & Lowe, 2000: 200–211 (complete reference list until 2000); Kovařík, 2009: 22, 31; Prendini & Esposito, 2010: 673–710, figs. 1–17.

= Heterobuthus Kraepelin, 1891: 205–211 (63–69) (syn. by Kraepelin, 1895: 79 (7))

= Riftobuthus Lourenço et al., 2010: 281 , figs. 1 and 2. Syn. n.

TYPE SPECIES. Androctonus (Prionurus) liosoma Ehrenberg in Hemprich et Ehrenberg, 1828

DIAGNOSIS. Total length 35–180 mm. Dorsal trichobothria of pedipalp femur arranged in α- configuration. Trichobothrium d 2 located external to dorsomedian carina of patella when carina is present. Trichobothrium eb located on fixed finger of chela. Pectines with fulcra. Pectinal teeth number 18–62. Sternum subtriangular. Tibial spurs present on third and fourth legs. Cheliceral fixed finger with two ventral denticles. Carapace without distinct carinae. Carapace in lateral view with entire dorsal surface horizontal or nearly so. Dentate margin of pedipalp chela movable finger with distinct granules divided into 9–14 rows, 3 terminal granules and one basal terminal granule. Tergites I–VI of mesosoma bear one carina. Telson without subaculear tubercle. Dorsal surfaces of first and second metasomal segments with stridulatory areas.

REMARKS ON THE HEMISPERMATOPHORES. The hemispermatophores of the two Parabuthus species cited here below are quite similar to each other. One possible minor difference is that the flagellum of P. kajibu sp. n. is shorter and thicker than that of P.abyssinicus , although we did not examine more than one sample and cannot specify the intraspecific variation. The hemispermatophores we found are also similar to published rough drawings or photographs of hemispermatophores of P. villosus (Peters, 1862) in Lamoral (1979: fig. 28), P. mossambicensis (Peters, 1861) (Fitzpatrick, 1994: fig. 2f), P. glabrimanus Prendini et Esposito, 2010 , and P. setiventer Prendini et Esposito, 2010 (their figs. 12, 17). Common features include a distally dilated pars reflecta, and the presence of a broad median lobe, hook, and pointed internal lobe. Our findings agree with previous observations that Parabuthus hemispermatophores have rather uniform structures that lack clear diagnostic differences at the species level (Fitzpatrick, 1994; Lamoral, 1979; Prendini, 2004).

We provide here descriptions and illustrations of the capsule and lobe structures of the genus Parabuthus that are more detailed than previously reported. The 2 + 1 configuration of lobes (median, internal + basal), with the flagellum fused to a broad, carinated median lobe, is consistent with other non- Buthus group members of the family (Kovařík et al., 2016). A feature that has not been documented in other buthids is the distinctly thickened ridge or basal lobe carina (blc) on the dorso-internal surface of the capsule.

REMARKS ON THE KARYOTYPES. We analyzed male karyotypes of three different Parabuthus species from the Horn of Africa ( Table 1). We used standard cytogenetic methods (e.g. Kovařík et al., 2009). The chromosome slides were stained with 5% Giemsa and the relative length of the chromosomes of the diploid set was measured for each specimen using the software Image J 1.45r (http://rsbweb.nih.gov/ij) with the plugin Levan (Sakamoto & Zacaro, 2009) based on 10 postpachytene spermatocyte nuclei.

The chromosomes of all analyzed species ( Figs. 28– 35) correspond to the cytogenetic characteristic typical for the family Buthidae : holocentric chromosomes, achiasmatic meiosis in males, and lower number of chromosomes (e.g. Mattos et al., 2013). The diploid set of P. abyssinicus consists of 16 chromosomes in both observed males from two distant localities ( Figs. 28, 30). The chromosomes gradually decrease in length from 7.51 % to 4.84 % of the diploid set in male from Ethiopia and from 7.34 % to 5.04 % of the diploid set in male from Eritrea ( Table 1). During meiosis we found only bivalents in all observed postpachytene in both males ( Figs. 29, 31). The diploid set of P. kajibu sp. n. consists of 18 chromosomes ( Fig. 32). The chromosomes gradually decrease in length from 7.17 % to 3.81 % of the diploid set ( Table 1). During meiosis we found only bivalents in all observed postpachytene in male of this species ( Fig. 33). The diploid set of P. pallidus consists of 20 chromosomes ( Fig. 34). The chromosomes gradually decrease in length from 7.11 % to 3.31 % of the diploid set ( Table 1). During meiosis we found one quadrivalent in all observed postpachytene in analyzed male ( Fig. 35).

The buthids represent the best explored scorpion family from the cytogenetic point of view. More than half of the cytogenetically analyzed scorpions so far belong to this family (in total 63 species from 22 genera) (see Kovařík et al., 2016; Schneider et al., 2016). Nevertheless, we have still limited knowledge about the karyotype variability of this family. We have available information on the karyotypes of more than three species in the genera Androctonus Ehrenberg, 1828 , Reddyanus Vachon, 1972 , Tityus C. L. Koch, 1836 and Uroplectes Peters, 1861 (Kovařík et al., 2016; Schneider et al., 2016). These genera represent three different patterns of intra- or interspecific karyotype variability. The intraspecific variability is documented in Reddyanus from Sri Lanka and Tityus from Brazil with the highest known range in T. bahiensis (Perty, 1833) (2n=5–19) ( Piza, 1944; 1949). Androctonus displays uniform karyotypes with 2n= 24 in all seven analyzed species from Northern Africa and Western Asia ( Moustafa et al., 2005; Sadílek et al., 2015). Uroplectes shows distinct interspecific variability in Africa (Kovařík et al., 2016; Newlands & Martindale, 1980). Parabuthus may be a genus with distinctive interspecific variability. Our results support this hypothesis. This fact was documented also during previous cytogenetic analysis of six species from Zimbabwe (Newlands & Martindale, 1980). It is evident that karyotyped material is not substantial enough to rule out the existence of intraspecific variability and more investigation is needed to accurately determine karyotype variability in Parabuthus species.

TAXONOMIC REMARKS. The monotypic genus Riftobuthus , with single species R. inexpectatus Lourenço et al., 2010 , is represented by a juvenile female from " Kenya, region of Turkana, N. Lokitaung", which the authors mistook for an adult female. The juvenile status of this specimen is indicated by an extraordinarily high pectinal tooth count (PTC = 36) for a buthid scorpion in its small size range, which if adult would violate the empirical scaling relation between PTC and body size of buthids ( Figs. 202–203; see discussion below). In all likelihood, Riftobuthus is a juvenile of one of the local species, Parabuthus pallidus . During our studies of Parabuthus juveniles, we were struck by the fact that in this genus there is a great deal of morphological, morphometric, and color variation ( Figs. 19–23). Lourenço et al. (2010: 281) cited a "unique combination characters" to diagnose the genus Riftobuthus , but almost all of these characters are exhibited by Parabuthus juveniles, and others are unreliable: 1) "carapace and tergites acarinated and smooth"; however, on the same page the authors state "Tergites smooth with one vestigial median carinae"; the carapace is acarinated and granulated in adult Parabuthus specimens but is often smooth in juveniles, and all Parabuthus species have tergites with one median carina; 2) "pectines extremely long with an unusually large number of teeth (36-36), dilated basal middle lamellae and weakly marked fulcra"; these are normal features for all females/juveniles of Parabuthus from the Horn of Africa (see Fig. 111 and fig. 1b in Lourenço et al., 2010: 282); 3) "the dentate margins of pedipalp chela fingers composed of linear rows of granules, forming almost a single row"; however, on the same page the authors state "outer and inner accessory granules absent from both fingers; two very small granules located proximally to the terminal granule on the movable finger", which does not correspond to their fig. 2f which is poor but indicates some accessory granules. In other papers, Lourenço has published repeated errors about "the dentate margins of pedipalp chela fingers" (see for example comments in Kovařík et Ojanguren, 2013: 209), and the granulation is sometimes developed anomalously in some specimens (see Fig. 71 versus Fig. 73); the number of these rows and their linear orientation cited by the authors for Riftobuthus correspond to Parabuthus pallidus ( Fig. 87); 4) the cited trichobothrial pattern, metasomal segment V and telson also correspond exactly to Parabuthus pallidus ; 5) the poor drawing of chelicerae (fig. 1c in Lourenço et al., 2010: 282) and the text need to be checked. From the original description, it is evident that Riftobuthus is a junior synonym of Parabuthus , and probably Riftobuthus inexpectatus is a junior synonym of P. pallidus .

REMARKS ON BUTHID PECTINAL TOOTH COUNTS. Soleglad (1973: 353–361, tabs. 1–5, figs. 13–14) first reported a positive correlation between pectinal tooth count (PTC) and adult body length for certain groups of North American vaejovid and chactid scorpions. Soleglad & Fet (2003b: 61, 163–169, figs. 109–113, D1–D11) extended this observation to other chactoid families ( Euscorpiidae , Superstitioniidae ), and we recently confirmed it in more detail for the scorpiopine subfamily of Euscorpiidae using carapace length as a proxy for body size (Kovařík et al., 2015). The correlation slope varies significantly between taxonomic groups and has been used as a diagnostic character ( Soleglad, 1973). Here we show that a similar positive correlation holds for the largest scorpion family Buthidae ( Fig. 202), establishing this phylogenetic scaling law for the majority of extant scorpion taxa (herein termed ' Soleglad's Law '). In contrast, we propose that as a rule PTC remains invariant when body size varies ontogenetically rather than phylogenetically, i.e. across different instars in the same species of scorpion. In buthids, this was previously indicated by the similarity of PTC values in juvenile and adult Tityus fasciolatus Pessôa, 1935 (Lourenço, 1980: 812–813), and Centruroides gracilis (Latreille, 1804) (Francke & Jones, 1982) . We confirm it for three additional species of Old World buthids ( Fig. 203; only females plotted, but a similar correlation exists for males (data not shown)). Together with Soleglad's Law, ontogenetic invariance of PTC provides a useful test for whether an individual is juvenile or adult. Since juvenile and adult PTCs should be similar, a juvenile will deviate significantly from the scaling relation of Soleglad's Law because its PTC will be too high for its body size or carapace length. This is exemplified by the cases of Alloscorpiops troglodytes Lourenço et Pham, 2015 (c.f. Kovařík et al., 2015), and Riftobuthus inexpectatus , which we discussed above.

What is the meaning of Soleglad's Law? Scorpion pectines bear dense fields of mechanoreceptive and chemoreceptive sensillae that are thought to be sensing substrate texture and substrate-borne chemical signals (reviewed in Farley, 2011). When a scorpion sweeps pectine combs over the substrate, it may be sampling a coarse spatial map of physical and/or chemical stimuli on that substrate. The pectine teeth may be basic units of sensory input for such spatial mapping. The rows of teeth on a comb could contribute a series of ‘pixels’ for building up the map. Soleglad’s Law says that larger scorpions bear more teeth per comb, hence they would be acquiring higher resolution maps relative to their body size. The regression line in Fig. 202 yields a standard allometric scaling equation: PTC = a.(CL) b = 7.892.(CL) 0.5266 ~ √(CL), where CL = carapace length. Scaling is sublinear, with exponent b having a value between two extremes, 0 <b <1: (i) b = 0 corresponds to PTC being constant, and the size of individual teeth increasing in direct proportion to body size, i.e. the size scale of substrate stimuli resolved is proportional to the size of the scorpion; (ii) b = 1 corresponds to PTC being proportional to body size (linear scaling), and the size of individual teeth beng constant, i.e. there would an absolute size scale of substrate stimuli resolved, independent of the size of the scorpion. The intermediate value of b suggests that larger species of scorpions may have some need to resolve finer substrate features, but not as much fine grained detail as smaller species. Soleglad’s Law only applies to sexually mature scorpions, which suggests that it is relevant to the task of tracking sex pheromone trails. If trails are deposited by mature conspecifics, their spatial extent should scale upwards with body size, which could explain why larger species do not need to resolve the same level of detail as smaller species. The sublinear power law may be related to how the structure or effective width of a pheromone chemical trail scales with body size. For example, perhaps larger species lay down relatively narrower trails than smaller species. A similar analysis of male buthids (N = 745 species) yielded b = 0.4464, and the exponents for scorpiopines (Kovařík et al., 2015) were 0.3717 (males) and 0.4033 (females). We hypothesize that this is a universal law with similar scalings in other scorpion families.

Figures 24–25: Parabuthus abyssinicus, Localities 12EM (24), Ethiopia, 11°43'30"N 40°58'45"E, 404 m a.s.l. and 12EW (25), Ethiopia, Awash, 09°00'34.5"N 40°17'56.5"E, 1012 m. a.s.l.