Apostolepis is a diverse neotropical snake genus, which has been historically subjected to poor taxonomic descriptions, largely based on either a small type series or subjective diagnoses. We evaluate the case of Apostolepis barrioi Lema, 1978 and its intricate taxonomic history, suggesting its synonymization with Apostolepis dimidiata (Jan, 1862), and providing brief commentary on the taxonomic instability that has been plaguing the genus.

Elapomorphini, Neotropical, synonymy, taxonomy

The Neotropical dipsadid snake genus Apostolepis Cope, 1862 comprises over 30 species, with an even broader synonym list, being marked by a systemic proliferation of “poorly defined” taxa, described based on single or few individuals with poor documentation of variation (Vanzolini 1986; Ferrarezzi 1993; Harvey 1999; Ferrarezzi et al. 2005; Nogueira et al. 2012). Apostolepis dimidiata (Jan, 1862) is a small-sized fossorial snake that occurs in the Cerrado, Chaco and Atlantic Forests at Argentina, Brazil and Paraguay (Cei 1993; Giraudo and Scrocchi 1998; Harvey 1999). Jan (1862) described Elapomorphus dimidiatus based on a specimen from “Brazil”, and allocated it to the subgenus Elapomojus Jan, 1862. Later, Peters (1880) described Elapomorphus erythronotus based on a specimen from “São Paulo” in southeastern Brazil. Cope (1887) presented two brief taxon descriptions, Apostolepis erythronotus lineatus and Rhynchonyx ambiniger vittatus, both from Chapada dos Guimarães, Mato Grosso, in central-western Brazil. Boulenger (1896) elevated both of Cope’s subspecies to species level and placed Elapomorphus erythronotus in Apostolepis. Werner (1897) described Apostolepis nigriceps based on two specimens, of which only one has a known locality given as “São Paulo”, in southeastern Brazil. Lema (1978) described three new species for Paraguay: Apostolepis barrioi from the Ypané River, Cororo, Concepcíon Province; Apostolepis ventrimaculatus from “Paraguay”; and Apostolepis villaricae from Villa Rica, Concepcíon Province. Later, Lema (1986) would synonymize A. erythronota, A. nigriceps and A. ventrimaculatus with A. dimidiata, while also revalidating A. lineata. Lema (1993) presented a review on the morphological variation of A. dimidiata, while also allocating the species he previously described, A. barrioi and A. ventrimaculatus, as synonyms of the former.

Recently, Cabral et al. (2017) presented a revalidation of A. barrioi, diagnosing it from all congeners based on an immaculate white venter, narrow dorsolateral stripes not in contact with the ventrals, and a terminal black shield. There is also a wide overlap between the meristic variation of A. barrioi and A. dimidiata, such as in the number of ventral scales (222–256 in A. barrioi; 214–264 in A. dimidiata) and subcaudal scales (23–55 in A. barrioi; 22–39 in A. dimidiata), in its morphometric variation (given in mm), in snout-vent length (188–542 in A. barrioi; 180–676 in A. dimidiata) and tail length (16–45 in A. barrioi; 16–60 in A. dimidiata), and geographic variation, since both species are sympatric along their whole distribution, as reported by the authors (Cabral et al. 2017: 246). Furthermore, the authors present a comparative table of Apostolepis species in which A. barrioi is stated as having an immaculate venter and A. dimidiata, a venter heavily pigmented with black, having only the edge of the ventrals white. This is in clear conflict with the original description, considering that the holotype of A. dimidiata presented an immaculate yellow venter according to the original description “[...] parte inferiore del corpo é giallastra, meno la testa che inferiormente ha del nero sugli inframascellari e sulle squame che stanno in vicinanzi ai sottolabiali” (En: lower part of the body is yellowish, except for the head, that has black inferiorly, in the inframaxillary (region) and in the scales near the infralabials) (Jan 1862: 48). Unfortunately, this specimen (holotype of A. dimidiata) could not be examined, since it was destroyed during the Second World War. It is also relevant that Lema (1993: 47) presents a plate that encompasses all known ventral pattern variation for A. dimidiata, with a gradual change from immaculate yellow to black ventral patterns.

Considering that Apostolepis barrioi and A. dimidiata share the same morphological features and variation (Fig. 1), present virtually identical geographic distribution, and both descriptions are based on specimens that have the same ventral coloration, which was erroneously cited as “diagnostic” at the time, we argue that A. barrioi Lema, 1978 should be relegated as a junior synonym of A. dimidiata (Jan, 1862). The work of Cabral et al. (2017) seems to follow a recent, genus-wide trend, in which several species have been described based on poorly supported diagnoses (e.g. Apostolepis mariae Borges-Nojosa et al. 2017 (2016); A. roncadori Lema, 2016; A. thalesdelemai Borges-Nojosa et al. 2016 (2017); A. underwoodi Lema & Campbell, 2017). None of these studies presented descriptions of osteology or hemipenes, nor do they include molecular support for their proposed species, relying exclusively upon highly variable morphological characters such as coloration and body shape. It is also noteworthy that, A. barrioi was described, synonymized, and then revalidated by the same author over a timespan of almost 40 years.

Figure 1. 

Dorsal, lateral and ventral illustrations of previously recognized taxa, Apostolepis dimidiata (A) and A. barrioi (B), according to the diagnoses of Cabral et al. (2017). However, these represent merely phenotypic variations of A. dimidiata and, according to the descriptions of Jan (1862) and Lema (1978), both holotypes of A. barrioi and A. dimidiata present the bottom coloration.

Unfortunately, several interest conflicts among researchers in the past decades have caused strong instability in Apostolepis, as well as most Elapomorphini taxa (Fig. 2). Here we present an overview of the currently recognized species of Apostolepis as well as their known specimens and diagnoses. It is noteworthy that poor diagnoses and small type series are usually associated with previous synonymizations (Table 1). We urge our fellow authors not to commit taxonomic malpractice and to carefully generate, rethink and analyze their data, providing compelling evidence for their claims. The careless proliferation and splitting of taxa may present deleterious consequences not only to the field of taxonomy but also to directing conservation efforts. An integrative revision, preferably incorporating aspects of external and internal morphology, along with molecular data, is largely warranted in order to mitigate and reevaluate the taxonomy of Apostolepis as a whole.

Figure 2. 

The impact of the taxonomic history of Apostolepis Cope, 1862 species. The line refers to the cumulative number of species considered as valid during the time span, suffering either reductions from synonymies or additions from descriptions and revalidations. Inset picture: Apostolepis sp. from Serra do Cachimbo, Pará, Brazil.

We thank Peter Uetz (VCU) and Henrique Caldeira Costa (UFV), who kindly provided us with some rare literature. We are also deeply indebted to Tarcísio Lyra dos Santos Abreu (UnB) and Russell Gray (UF-IFAS) for helping troubleshoot our analyses, and to Editor Robert Jadin (UW-Eau Claire) and Francisco Franco (IBSP), who provided constructive criticism and comments on our manuscript. Omar M. Entiauspe-Neto and D. Loebmann thank CNPq for a PIBIC grant (136628/2016-8); D. Loebmann is a Research Fellow of CNPq (Proc. #310651/2017-4).