Orthochirus Karsch , 1891

Kovařík, František & Lowe, Graeme, 2022, Review of Orthochiroides Kovařík 1998 with description of a new species (Scorpiones: Buthidae), Euscorpius 349, pp. 1-42 : 35

Orthochirus Karsch, 1891

The proposed synonymy of Orthochiroides with Orthochirus by Lourenço & Ythier (2021: 339) equates to the hypothesis that these two genera together comprise a monophyletic group. To test this hypothesis, we conducted a cladistic analysis of an ingroup that included all species of Orthochiroides , plus selected exemplar species representing Orthochirus and other morphologically similar genera of small buthids ( Neobuthus Hirst, 1911 , Gint Kovařík et al, 2013, Butheolus Simon, 1882 , Xenobuthus Lowe, 2018 , Baloorthochirus Kovařík, 1996 and Fetilinia Lowe & Kovařík, 2021 ; Table 3). As outgroup taxon we selected Compsobuthus eritreaensis Kovařík, Lowe, Plíšková & Šťáhlavský, 2016 . This selection was guided by results of a molecular phylogenetic study in which this species, along with other Compsobuthus species, was placed in the immediate sister clade of a larger clade including all of our DNA-sampled ingroup taxa (Štundlová et al., 2022). We analyzed a matrix of 43 discrete morphological characters ( Tables 3, 5) These characters were selected on the basis of: (i) their relative intrageneric stability for ingroup taxa, with most having minor or no interspecific variation within each genus; and (ii) their systematic variation across genera of the ingroup, which was potentially informative about higher level relationships. They included 25 binary characters and 18 unordered multistate characters. Phylogenetically uninformative characters were excluded from the analyses. Two binary characters (characters 1 and 2: loss of anterosubmedian and central median carinae on the carapace) were tested under assumptions of both reversibility and irreversibility. Intrageneric stability ensured that results about intergeneric relationships would not be sensitive to our representation of the larger genera ( Neobuthus and Orthochirus ) by a minority of their named species. Additional exemplars of these genera would have identical, or almost identical, character scores as those already included, and are therefore expected to only introduce polytomies at their respective genus nodes.

We conducted 22 analyses, assuming either equal weights, or implied weights with varying strengths of concavity (k = 1–6, 8, 10, 15, 30). In total, nine distinct most parsimonious trees (T i, i = 0–8) were retrieved ( Table 4). A monophyletic including all species of Orthochiroides , clade 1 = (( Orthochiroides insularis , O. socotrensis ), ( O. somalilandus , O. vachoni )), was resolved in 14/22 analyses. Clade 1 was separated from, and basal to, other ingroup taxa. It received significant support in 9/11 analyses that assumed irreversibility of characters 1 and 2 (e.g., upper gray panels in Figs. 205– 206, 209 ). If reversibility was permitted, the carapace carinae were initially lost for the entire ingroup and later regained in Gint , Somalibuthus and Xenobuthus . Irreversibility assumes that once lost, a compound structure like a carina is less likely to be reconstructed de novo in these genera, which also possess other plesiomorphic characters, than it is to be lost independently in the more specialized lineages (clades 1, 2 and 3), which also possess other apomorphic characters. In most cases, imposing irreversibility had little effect on tree lengths, but increased support for clade 1 ( Orthochiroides ) which gained two synapomorphies (i.e., carinal losses).

In all of the analyses and tree topologies, either clade 2 = ( Baloorthochirus becvari , ( Fetilinia dentator , ( Orthochirus glabrifrons , ( O. afar , O. ‘ innesi ’)))), or clade 3 = ( Fetilinia dentator , ( Orthochirus glabrifrons , ( O. afar , O. ‘ innesi ’))), or both, were resolved with significant support (jackknife statistic 50% – 71%) ( Table 4, gray cells). This result was independent of reversibility of characters 1 and 2. Clades 2 and 3 excluded all species of Orthochiroides , and included Orthochirus and either one or both of two other genera, Baloorthochirus and Fetilinia . This implies that a group including only Orthochiroides and Orthochirus is polyphyletic, thereby refuting the proposed synonymy of these two genera. A genus including both Orthochiroides and Orthochirus would also need to include Baloorthochirus and Fetilinia , which have very different metasomal morphologies. Even then, it would require fusion of clades 1 and 2, which were not resolved as sister groups in any of our analyses.A monophyletic genus that includes both Orthochirus and Orthochiroides necessitates the inclusion the entire set of ingroup taxa ( Figs. 205–209 , 211 ).

A subset of our ingroup taxa was included in an extensive DNA analysis of buthid phylogeny (Štundlová et al., 2022). The relationships among these taxa reconstructed by Bayesian inference and maximum likelihood analysis applied to the entire buthid dataset are shown in Fig. 210 . The species representing Orthochiroides occupied a basal position, separated from the lineages containing Orthochirus (shaded panels in Fig. 210 ). This separation, as well as the overall topology, were consistent our morphological analyses. Taken together, these findings further confirm the status of Orthochiroides as a genus separate from Orthochirus , as originally proposed by Kovařík (1998, 2004). The synonymy of Orthochiroides with Orthochirus by Lourenço & Ythier (2021) is hereby invalidated, and we reinstate Orthochiroides Kovařík, 1998 stat. rev., with a differential diagnosis provided by the seven characters (i) – (vii).

Off the nine different trees obtained in our analyses, the most frequent was T 0 (e.g., Fig. 205 ). It was retrieved repeatedly under equal weights and under implied weights with moderate to weak concavity constants. It also enjoyed the highest support values for clades 1, 2 and 3, and was our preferred phylogenetic hypothesis. Unambiguous synapomorphies for T 0 are mapped in Fig. 211 .

Figures

Figure 211: Phylogenetic tree with topology T, retrieved by cladistic analyses of 43 morphological characters under equal weights and 0 implied weights with concavity constants k = 10 – 30 (IW), and k = 4 – 30 (IW*). Boxes indicate unambiguous synapomorphies, filled boxes unique (derived once with reversals allowed), open boxes homoplasious synapomorphies (derived more than once). Numbers above nodes are character identifiers, those below nodes character states. Gray panels 1, 2 and 3: clades of interest.

Figures 205–210: Phylogenetic analysis of Orthochiroides, and Orthochirus and related buthid genera. Figures 205–209. Examples of most parsimonious trees (T 0, T 2, T 3, T 4 and T 7) retrieved by cladistic analyses of 43 morphological characters under equal and implied weights. Numbers above indicate jackknife supports (36% resampling probability),those below relative Bremer supports, computed for implied weights(IW), concavity constantk = 1 (207),and implied weightswith characters 1 and2 irreversible (IW*), concavity constants k =10 (205),k = 6 (206),k = 1 (208) and k= 3 (209). Gray panels 1, 2 and 3: clades of interest.Figure 210. Phylogenetic trees of a subset of the exemplar species in Figs.205–209, reconstructed by Bayesian inference (BEAST) and maximum likelihood (ML) analyses of multilocus sequence data (Štundlová et al., 2022) (node supports indicated).