Diagnosis

This genus can be distinguished from all other members of Buthidae by the following combination of characters: (a) pedipalp fingers with 7-10 principal rows of denticles which are only slightly overlapping, oblique and flanked by external and internal supernumerary denticles on each side (missing in first and second instar immatures); (b) absence of stridulatory organ on sternite III and pectines; adult males with metasomal segments longer and narrower than females (^{Armas, 2009}).

Centruroides ruana sp. nov.

(Morphometric data:Tables 2 and 3;Figs. 1-3 and 6-8)

Figure 1 Habitus of dorsal and ventral views of holotype male and paratype female of Centruroides ruana sp. nov. Bars=10mm. 

Figure 2 Adult male paratype of Centruroides ruana sp. nov.: (a and b) carapace; (c and d) mesosoma dorsal; (e) tergite VII; (f) sternite VII; (g) pedipalp femur dorsal; (h) pedipalp patella dorsal; (i) pedipalp chela, dorsoexternal view; (j) pedipalp chela ventrointernal view. Notes: photographs (a) and (b) were taken under UV light to show granulation. Bars=1mm. 

Figure 3 Adult male paratype of Centruroides ruana sp. nov.: (a) pigmentation leg IV; (b) pigmentation leg I; (c) basal pectinal plate; (d) metasomal segment V lateral; (e) telson lateral view; (f) telson ventral view; (g) telson dorsal view. Bars=1mm. 

Diagnosis

The species is distinguished by the combination of the following characters: Scorpions of medium to moderately large adult size (including telson) for the genus. C. ruana sp. nov. is a typical “striped scorpion”, basically yellow to ochre, paler ventrally, with 2 dorsal longitudinal blackish stripes on the mesosoma, separated by a slightly wider pale stripe. Anterior median area of the carapace strongly pigmented in the anterior half except in the anterior median furrow and 2 areas above the lateral central furrow; the rear half area is slightly dark pigmented on both sides of the posterior median furrow and on the posterior lateral carinae which is more intense; intercarinal surface with dense and well-defined granules. Pedipalps moderately elongated; hand evenly oval and moderately incrassate. Basal plate of pectines in males is rectangular with the anterior notch V-shaped, anterior margin slightly concave and lobulated, and posterior margin straight, in females the anterior margin is only slightly notched medially and with posterior margin rounded. Pectinal tooth counts 24-26 (mode 25; n = 14 in males) vs. 22-25 (mode 23; n = 14) in females. Metasoma with dorsolateral, lateral, supramedian and inframedian carinae well developed and serrated; vesicle slightly elongated. Sexual dimorphism evident, females with pedipalps shorter, less slender and more strongly carinate, bulkier hand, metasoma more strongly carinate and with shorter segment V, telson vesicle shorter and rounder, with a stronger subaculear tubercle.

Taxonomic summary

Type data. Adult male holotype (CNAN-T0874), 1 adult male and 2 adult female paratypes (CNAN-T0875), 6 adult male and 6 adult female paratypes (CAFBUM 2015-0503), all with the following data: Mexico: Michoacán: Buenavista Municipality: Felipe Carrillo Puerto (= La Ruana), Ejido “Santa Rita”, 19.1224°N, 102.7117°W; 7.III.2013; coll. L. Gaona E.

Distribution

Only known from 2 nearby localities of the lowland Balsas Basin, in west-central Michoacán (Fig. 4). Both localities are in the Apatzingán Valley (Fig. 5).

Figure 4 Sexual dimorphism in females Centruroides ruana sp. nov.: (a) tergite VII; (b) pedipalp chela dorsal view; (c) pedipalp chela ventral view; (d) basal pectinal plate; (e) metasomal segment V, lateral view; (f) telson, lateral view. Bars=1mm. 

Figure 5 Known geographical distribution of Centruroides ruana sp. nov.: type-locality (•) and Apatzingán (▵), Michoacán, Mexico. Modified from Google Maps. 

Remarks

The new species is similar in size and coloration to Centruroides infamatus (Fig. 6), and both taxa are also geographically close. The latter can be distinguished by the less slender pedipalps, with femur, patella and chela shorter; hand of pedipalp slightly less elongated; metasoma and telson longer (Table 2). Also, metasoma with dorsal lateral, lateral supramedian, and lateral inframedian carinae less developed and granular to crenulated except the dorsal laterals, which are finely serrated and even denticulate (Fig. 6.11, 6.14). The basal plate of pectines in males of C. infamatus with the anteromedian notch shallow and the anterior margin not lobulated, in females with the anterior margin concave but not notched and posterior margin clearly recurved (Fig. 7).

Figure 6 Morphological comparison between: (1), Centruroides ruana sp. nov.; (2), C. balsasensis; (3), C. infamatus; (4), C. limpidus, and (5), C. tecomanus. (a) pigmentation on the carapace; (b) pigmentation of mesosoma; (c) metasoma, lateral; (d) metasomal segment V, and (e) vesicle and subaculear tubercle. Bars=1mm. 

Figure 7 Comparison of pectinal plates of males and females of Centruroides ruana sp. nov. C. balsasensis, C. infamatus, C. limpidus and C. tecomanus. Bars=1mm. 

Another species that is morphologically similar and found in the same geographic region is C. balsasensis (Fig. 6), which can be differentiated from C. ruana sp. nov. by the coloration paler and less densely patterned, the color pattern on carapace and tergites (Fig. 6), and the carapace with anterior margin noticeably notched, convex, with granules that are denser on the median part. General habitus of C. balsasensis slightly less slender, including pedipalps with more robust chelae (obvious in the fingers/manus length ratio) (Tables 3 and 4); basal pectinal plate of males of C. balsasensis with anterior margin almost straight and not lobulated; in females the anterior margin is clearly concave and the posterior margin is strongly convex. Males of C. balsasensis with subaculear tubercle markedly stronger (Fig. 7).

Centruroides limpidus is more distantly related to C. ruana sp. nov. based on morphological criteria, but both species are sympatric in the Felipe Carrillo Puerto area. Although C. limpidus is present in very low abundance here, it is easy to recognized by the carapace color pattern, which has 2 pairs of thin, longitudinal, sub parallel blackish stripes: 1 lateral across the entire carapace (Fig. 6.21, 6.24), and 1 submedian that extends shortly behind the ocular tubercle (Fig. 6.1, 6.2), leaving most part of the surface without pigment. Other differences are morphometric features (for measurements see Tables 3 and 4).

Centruroides tecomanus is another species that could be confused with the new species herein described (Figs. 6 and 7), owing to its general appearance; however, in addition to the morphometric differences (Tables 3 and 4), the coastal distribution of this species; the constant color pattern of its carapace, always with 4 thin and well defined dark lines (Fig. 6.1, 6.5); and the subaculear tubercle well developed and generally with its point oriented toward the point of aculeus (^{Ponce-Saavedra et al., 2009}) (Fig. 6.21, 6.25) are enough to distinguish this species from the newly described one.

Additionally we made a comparison of granulation pattern of the carapace of C. ruana sp. nov., C. balsasensis, C. infamatus, C. limpidus, and C. tecomanus using images in UV light. Differences in the patterns of granulation of central median carinae (cmc) and central lateral carinae (clc) as well as in the granulation located above the posterior transverse furrow (ptf) can be appreciated between species (Fig. 8).

Figure 8 Granulation comparison of carapace between: (1) Centruroides ruana sp. nov.; (2) C. balsasensis; (3) C. infamatus; (4) C. limpidus, and (5) C. tecomanus. Central median carinae (cmc), central lateral carinae (clc), and posterior transverse furrow (ptf). Bars=1mm. 

Morphometric analysis

A morphometric statistical analysis was conducted in order to reinforce the diagnostic comparison between C. ruana sp. nov., C. balsasensis, C. infamatus, C. limpidus, and C. tecomanus. As a whole, 54 variables were included: 28 measurements, 25 derived proportional ratios and a meristic count. Matrices were constructed using the following labels for each taxon: C_ruana, C_balsasensis, C_infamatus, C_limpidus, and C_tecomanus. A Cluster Analysis (CA) was performed using Euclidean distances, standardized data and Ward's Minimum Variance Method followed by a Principal Component Analysis (PCA). The results revealed that specimens previously assigned to a particular taxon, consistently clustered together and isolated from others, forming groups of defined species (Fig. 9a).

Figure 9 (a) clustering plots obtained with Ward's method and standardized data; (b) ordination graphics of 2 first principal components of PCA, and (c) graphics obtained from canonical discriminant analysis. All analyses were made with 13 variables in females and 14 in males. Centruroides ruana sp. nov. (plus symbol), C. infamatus (square), C. balsasensis (triangle), C. limpidus (letter X), C. tecomanus (circle). 

In order to reduce the number of variables supporting the species distinction and to discard any potential effect of size among the 3 taxa (a clean-up process), all matrices were depurated by eliminating those variables found to be highly correlated or with little information. The former are simple redundant, while the latter represent variation which is very poorly explained through the PCA. As outcomes, a thirteen variable matrix was obtained for males and another with fourteen for females (Tables 3 and 4).

Using these variables and running separate analyses for both males and females, a second CA was calculated, and again, specimens of each previously identified taxon clustered together and were separated from all other species (Fig. 9a). A subsequent PCA applied to these matrices yielded 85.67% and 87.55% of explained variation in 3 principal components for both males and females, respectively (Fig. 9b). Further, a canonical discriminant analysis (CDA) demonstrated that the morphometric distinction of all species is adequate (Fig. 9c): tests such as Wilks’ Lambda, Pillai's Trace, Hotelling-Lawley and Roy's Max Root, all proved significant (p < 0.05) and yielded null misclassifications. In males (Table 3), thirteen variables were found to present univariate significant statistical difference (p < 0.05). In females (Table 4), the number of such variables was fourteen (p < 0.05).

The morphologic and morphometric data support C. ruana sp. nov. as a valid species, clearly distinct from the morphologically similar C. balsasensis and C. infamatus and from C. limpidus, which is the species found closest geographically. The difference with C. tecomanus was also clear. The overall resemblance in the habitus among C. limpidus and C. balsasensis and C. ruana makes it very likely that this new species is actually more widespread across the Balsas Basin, but has been misidentified in collections.