Scincidae

4.3. On Scincidae View in CoL phylogeny and hemipenial morphology

At present, hemipenial morphology data representing all main phylogenetic clades (sensu Pyron et al., 2013) is available for approximately 3 % of all scincid species. These data help us outline a preliminary picture of the main trends in hemipenial evolution in this poorly studied lizard family.

Many studies, dealing with hemipenes, simply describe the morphology of these structures without providing figures or often present the structures only schematically. This is likely the main reason for generating a plethora of terms for homologous structures. Sometimes a complete eversion of the hemipenis is not easily achieved, which has resulted in an additional source of confusion in some works (e.g., Zhang, 1986).

Greer (1979) recognised two basic types of hemipenes (illustrated schematically on figs. 11–15) in Australian skinks, and this discrimination can be accepted for the whole family: 1) with relatively short columnar base and slightly bulbous or bilobed cap; and 2) with long and narrow base and two equally long bifurcations. The “short columnar base” is typical for the Egernia group which includes the genera Egernia , Tiliqua and Tribolonotus . The genus Cyclodomorphus and Corucia zebrata Gray, 1855 can be assigned to the same type (Ziegler and Bohme ¨, 2004). The close relationships between the members of this group are supported by molecular phylogenetic analyses ( Gardner et al., 2008; Pyron et al., 2013). The “short columnar base” type also includes the Eugongylus group (sensu Greer, 1979) and its two subgroups, i.e., the

Eugongylus View in CoL subgroup with the genera Anotis, Cryptoablepharus , Emoia View in CoL , Eugongylus View in CoL , Leiolopisma, Morethis View in CoL , and Proablepharus View in CoL , and the Lampropholis View in CoL subgroup with the genera Carlia View in CoL , Lampropholis View in CoL , and Menetia View in CoL . S´anchez-Martínez et al. (2020) studied the hemipenes of a number of Mabuya View in CoL and Trachylepis View in CoL taxa. Seven of the Mabuya species have thick hemipenes with a very short bifurcation: M. agilis , M. altamazonica View in CoL , M. caissara View in CoL , M. bistriata , M. dorsivittata , M. frenata View in CoL , M. macrorhyncha View in CoL . Similar short bifurcation is observed in Trachylepis atlantica (Schmidt, 1945) View in CoL , and Tiliqua rugosa (Gray 1825) View in CoL . Noble and Bradley (1933) provide drawings of the hemipenes of Plestiodon fasciatus View in CoL L. which appear rather similar to those of T. capensis View in CoL (the present study). The hemipenes are characterised by wide lobes, numerous folds on the lobes, short bifurcation and base, and the presence of an asulcal protrusion (lateral aspect according to fig. 12 caption in Noble and Bradley, 1933) similar to that observed in Ablepharus budaki View in CoL ( Fig. 3F View Fig ). Scincella baraensis Nguyen, Nguyen, Nguyen & Murphy, 2020 View in CoL has a very simple, balloon-shaped, non-bifurcated hemipenis without folds and protrusions ( Nguyen et al., 2020). Riopa guentheri (Peters, 1879) View in CoL also possesses a non-bifurcated hemipenis, but it is possible that the hemipenes were not completely everted and that the tip is actually not visible ( Fig. 3 View Fig in Bhilala et al., 2021). Bifurcation is lacking in Copeoglossum nigropunctatum View in CoL (Spix, 1825; S´anchez-Martínez et al., 2020).

According to Greer (1979), the second type of hemipenes is deeply bifurcated with a long narrow base and two equally long lobes. This type is characteristic of the Sphenomoprhus group (after Skinner et al., 2013) including Anomalopus View in CoL , Ctenotus View in CoL , Eremiascincus View in CoL , Hemiergis View in CoL , Lerista View in CoL , Notoscincus View in CoL , Saiphos View in CoL , Sphenomorphus View in CoL , and Concinnia queenslandiae (De Vis, 1890) . This condition is considered derived, in contrast to the columnar structure condition in the genera Eumeces View in CoL and Mabuya View in CoL which is considered basal. The latter statement is not valid for P. fasciatus ( Noble and Bradley, 1933) View in CoL and Eutropis longicaudata ( Ziegler, 2002) View in CoL , which possess a weak hemipenial bifurcation (type one, according to Greer, 1979), or some species with a substantial degree of bifurcation (type two, op. cit.) such as Eutropis multifasciata ( Zhang, 1986) View in CoL (species previously attributed to genus Eumeces View in CoL and Mabuya View in CoL ). In Tropidophorus noggei Ziegler, Thanh & Thanh, 2005 View in CoL , the hemipenes are weakly bifurcated ( Ziegler et al., 2005) although the genus Tropidophorus View in CoL is shown to be closely related to Sphenomorphus species by Pyron et al. (2013). An extreme bifurcation of the hemipenial lobes is found in Scincella rufocaudata View in CoL (Darevsky & Nguyen van Sang, 1983; Darevsky and Nguen, 1983) in which the lobes are approximately three times longer than the base. However, this condition may due to an incomplete eversion of the pedicel (the base). The general hemipenial shape, i.e., having fine lobes as long as the base, of Scincella reevesii (Gray, 1839) ( Ziegler, 2002) View in CoL is similar to the hemipenes of A. kitaibelii View in CoL and A. budaki View in CoL described by Vergilov et al. (2017). In Sphenomorphus buenloicus (Darevsky & Nguyen van Sang, 1983) View in CoL the lobes are shorter than the base, whereas in S. indicus (Gray, 1853) View in CoL they are longer than the base ( Zhang, 1986; Ziegler, 2002). In S. incognitus (Thompson, 1912) ( Zhang, 1986) View in CoL , the length of the lobes cannot be estimated due to incomplete eversion. The genus Sphenomorphus View in CoL is paraphyletic, with Scincella View in CoL deeply nested within it, and the hemipenial morphology of the groups supports this point of view ( Shea, 2012; Linkem et al., 2010; Linkem, Diesmos and Brown, 2011). The species of Ablepharus View in CoL examined in this study can also be assigned to the deeply bifurcated hemipenes type.

These examples demonstrate that the degree of bifurcation of the hemipenis, although well-defined, is an extremely homoplastic character.

Scincid hemipenial morphology has received relatively little attention and has been studied rather sporadically which may explain the inconsistency in terminology. The folds of the organ caused tremendous confusion and the lack of understanding of their homology resulted in the coining of numerous terms. One of the first authors dealing with Scincidae hemipenes was Cope (1896), who noticed the presence of "longitudinal laminae" in many scincids. S. reevesii ( Ziegler, 2002) has longitudinal folds on the sides of the base, not reaching the branches. No longitudinal folds were observed in the Ablepharus species. Cope (1896) mentioned some "cross-ribbed plicae" in E. carinata and P. obsoletus . Noble and Bradley (1933) found a series of "transverse flounces" on the side opposite to the "heart-shaped pad" in two scincid lizards. These, however, may be artifacts of sample processing, i.e., injection of salt solution and fixation with formaldehyde. McCann (1949) also mentions “transverse plicae” on the apical surface of the lobes in E. carinata . S´anchez-Martínez et al. (2020) observed “horizontal ridges” in T. atlantica . These folds are apparently homologous to the transverse folds of T. capensis ( Fig. 1E View Fig ). The term "horizontal" is inappropriate from a morphological point of view. Apparently, the anatomical position of the everted hemipenes is different from those of the isolated organs (oriented to the taste of the researcher). For this reason, the terms “transverse folds” or “transverse plicae” should be considered correct.

Lateral folds ( Fig. 3 View Fig ) can be seen on the lateral part of the hemipenial base in many Scincidae species. The term “lateral” in this case is inappropriate, because of the variable orientation of the everted organ of a living animal. Dunger (1973) mentioned "flap-like lobules" on the hemipenis of Leptosiaphos kilimensis (Stejneger, 1891) from Nigeria (now attributed to Leptosiaphos dungeri Trape, 2012 ). The lateral projections observed in Hemiergis gracilipes ( Greer 1979) are likely homologous to the "bulbous lobes" of A. budaki ( Vergilov et al., 2017) , although they are substantially more pronounced in the latter. The term "bulbous lobes" is confusing, as the term "lobes" has been used for the branches of the bifurcated hemipenis by other authors. For us, due to the lack of more appropriate term for such structures, the usage of term “lateral folds” or “lateral plicae” is acceptable. Greer (1989) mentioned that the genera Carlia , Lampropholis , Lygisaurus , and Saproscincus share a unique elongated projection of the base. This character can be an autapomorphy of the group, which forms a well-supported clade in the phylogenetic tree of Pyron et al. (2013). Linkem et al. (2011) referred to the lateral region of the "main shaft" basad of the hemipenial bifurcation as "bulbous lobe", treated it as an autapomorphy of the clade, and described the new genus Pinoyscincus . These “lobes” appear homologous to the lateral folds (as those of A. kitaibelii ; Fig. 2 View Fig ) but are wider, thicker and cover the whole lateral surface of the hemipenial base.

The sulci are bordered by thick folds for which we encourage the use of the term “labia” ( Fig. 3 View Fig ) (these structures correspond to the “sulcal lips” of Klaver and Bohme ¨1986). The labia extend to the sulcal openings at the apices of the lobes where they transform into small elongated projections, i.e., the “terminal awns” (e.g., Montingelli et al., 2022). Vergilov et al. (2017) referred to these structures as the “terminal awls [sic]”. In Trachylepis and Mabuya , S´anchez-Martínez et al. (2020) referred to the labium as the “fold of the base of the sulcus spermaticus”. The figure labelling of this “fold” by S´anchez-Martínez et al. (2020) corresponds to the base of the labia of the hemipenes (where the structure is thicker) but does not include the apex of the lobes. Cope (1896) describes the labia (we assume that the author refers to this structure) as “welt” that is “opposite the sulcus spermaticus”.

The term “asulcal protrusion” was proposed by Vergilov et al. (2017) for A. budaki , it was observed in A. rueppellii (this study), and in all cases it is homologous with a similar structure in T. capensis . This structure has also been observed in other species. S´anchez-Martínez et al. (2020) referred to this structure as the “rounded fold”. Noble and Bradley (1933) presented a similar protuberance on the asulcal surface of the hemipenis of Eumeces fasciatus , calling it a “tubercule”. A similar but less protruded fold (very similar to the one in T. capensis ), is present in Pinoyscincus abdictus abdictus (Brown & Alcala 1980) ( Linkem et al., 2011) . Presumably, all these structures are homologous and apparently represent a homoplastic trait.

Thе use of inappropriate terminology is not limited to the hemipenial folds and has been applied to variety of other structures. For example, Vergilov et al. (2017) divided the hemipenis into apex, truncus, and pedicel based on the terminology of Klaver and Bohme ¨(1986) for Chamaeleonidae . This terminology, however, appears incompatible with the scincid hemipenis, which differs substantially.

There are a number of examples regarding the usefulness of hemipenial morphology in support of phylogenetically close relationships. Bohme ¨(1988) pointed out the similarity in hemipenial morphology in ecologically distinct genera, such as the specialised semiaquatic genus Cophoscincopus and the morphologically unspecialised Leptosiaphos ianthinoxantha (B¨ohme, 1975). Both taxa possess a collar-like ring, which was interpreted as a synapomorphy by the author. This finding likely indicates a close relationship between the two taxa and was partly supported by Trape et al. (2012). B¨ohme (1988) also mentioned the close similarity between L. ianthinoxantha and the species present in Dunger (1973), here interpreted as L. dungeri , both of which have bifurcated hemipenes. Bohme ¨et al. (2000) included figures of the hemipenes of the closely related Cophoscincopus greeri B¨ohme, Schmitz, & Ziegler, 2000 and C. durus (Cope, 1862) . The authors mentioned that the organs were densely covered with delicate pustules except the whole apical area, however, they did not discuss the nature of these structures.

No scincid species have been previously reported to possess calcified structures, as also shown in the present study. Their absence can, to some extent, be used in the general diagnosis of scincid lizards. Noble and Bradley (1933) noted that the hemipenes of scincids and iguanids are less complex than those of teiids and lacertids which are, in relation to mating behaviour, more generalised in the first two groups. Although this field remains largely unexplored, the driving force of sexual selection has already been partly discussed ( Eberhard, 1985; 2010).

The usage or interpretation of the specific hemipenial morphology for higher taxonomic analyses, however, is less informative, e.g., the lizard family Lacertidae (B¨ohme, 1971). This is mainly due to numerous homoplastic traits found in non-congeners (for example the asulcal protrusion in A. budaki and T. capensis ; Fig. 1E View Fig ; 2B, C View Fig ). Similarly, homoplastic characters can be pointed out for relatively closely related boid snakes (Ziegler and Bohme ¨, 1997). This leads to the conclusion that hemipenial differentiation should be used only after a stable phylogeny of a group of interest has been established. In this respect, satisfactory results for the family Gymnophtalmidae were presented by Nunes et al. (2014).