On the identity of Leptoxis taeniata – a misapplied name for the threatened Painted Rocksnail (Cerithioidea, Pleuroceridae)

Abstract The Painted Rocksnail, currently known as Leptoxis taeniata, is a federally threatened species native to the Mobile River basin in Alabama, USA. Presently restricted to four disjunct populations, the species is at considerable risk of extinction after a range decline of over 95% in the 20th century because of habitat alteration following impoundment of the Coosa River. Here, we reassess the identity and historical range of the Painted Rocksnail to improve communication and conservation efforts for the species. We determined that L. taeniata is a synonym of L. picta and that the name L. taeniata has been misapplied to the current concept of the Painted Rocksnail for which L. coosaensis is the oldest available name. Leptoxis coosaensis and L. picta are herein redescribed. After examination of historical material, we determined that records of the Painted Rocksnail outside the Coosa River drainage were misidentifications. Thus, we redefine the historical range of the Painted Rocksnail as restricted to the Coosa River and select tributaries above the Fall Line at Wetumpka, Alabama, rather than extending into the Alabama River as previously thought. Leptoxis coosaensis is in dire need of conservation, and management plans should take into consideration the revised historical range of the species.

(MCZ 294987) bearing an original A.A. Gould label stating "Anc. taeniatus Conr. Claiborne Al. from Conrad" (Fig. 5E). We also located an uncatalogued lot of four specimens containing another possible paralectotype at the Natural History Museum in London (NHMUK; ex Cuming collection), again labeled with the locality simply as "Alabama" (Fig. 5D). The latter lot is accompanied by an original J.G. Anthony label, annotated in his hand, "the separate one is authentic, marked so by Conrad himself, are not the others mature forms of the same?" Although one specimen is no longer conspicuously separated from the rest, we have concluded that the smallest specimen within the lot may be the specimen referred to in Anthony's note. Although it was received from Conrad, given the ambiguity about the identity of the specimen Anthony was referring to, and the locality inconsistency, we consider it only a possible type.
Pleurocerid species display high levels of morphological variation in their shells, which can overlap between close relatives (Goodrich 1922;Whelan et al. 2015). Consequently, confidently identifying species using shells alone can be difficult, particularly for poorly localized historical material including types. As such, we carefully examined possible type material of L. taeniata to determine if it matches the current concept of the Painted Rocksnail. Baker's lectotype (ANSP 27620) has a less inflated body whorl and is more elongately conical than the Painted Rocksnail. The paralectotypes in lot ANSP 413583 (formerly 27620) ( Fig. 5G-I), particularly the adult specimens ( Fig. 5G, H), are more clearly not representative of the Painted Rocksnail, indicating that the lectotype is a slightly atypical shell. The specimen labelled by Anthony at NHMUK (Fig. 5D) has a more narrowly ovate aperture than that seen in Painted Rocksnails. MCZ 294987, the one lot explicitly from the type locality (Fig. 5E), has a shell morphology that clearly does not conform to the current concept of the Painted Rocksnail (Figs. 2,3). Its body whorl is more narrowly conical and the aperture is more narrowly ovate, rather than broadly ovate in large adults. Painted Rocksnails also usually have more impressed sutures. Overall, the possible type material of L. taeniata (Fig. 5) does not conform with the current concept of the Painted Rocksnail.
In addition to examining type and historical material (Fig. 5), we evaluated pleurocerid collections from the Alabama River made in the last 30 years to help determine the identity and range of the Painted Rocksnail. No modern survey has recovered the Painted Rocksnail from any location in the Alabama River (Garner et al. 2011). These surveys included over 190 hours of dive time since 1990 in the Alabama River (JT Garner, unpubl. data). Furthermore, examination of historical museum collections has failed to produce a single lot from the Alabama River that corresponds to the current concept of the Painted Rocksnail. Lots labeled as "L. taeniata" from the Alabama River are misidentified specimens, usually of L. picta (e.g. UF 82371, ANSP 65451 Fig 5L-O; also see photographs uploaded to FigShare, https://doi.org/10.6084/ m9.figshare.5084272.v1); lots identified as "L. taeniata" from the Cahaba River drainage are also misidentified, usually of specimens of L. ampla (e.g. UF 81652, USNM 519194; see photographs uploaded to FigShare, https://doi.org/10.6084/ m9.figshare.5084272.v1). Currently, the Alabama River near Claiborne hosts healthy populations of other pleurocerids including L. picta, Pleurocera prasinata (Conrad, 1834) (Conrad 1834b), and multiple Elimia species (Garner et al. 2011). Therefore, we doubt that the apparent failure to collect the Painted Rocksnail from its ostensive type locality for over 150 years reflects extirpation of the species at that site.
River above the fall line has a historical range that extends into the Alabama River or Coastal Plain physiographic region. Leptoxis ampla, a Cahaba River endemic and the sister species to the Painted Rocksnail (Whelan et al. 2015), is also found only above the Fall Line. Therefore, if the Painted Rocksnail was historically found in both the Alabama River and the Coosa River above Wetumpka, the species would represent a significant departure from distribution patterns seen among Mobile River basin Leptoxis species.
Despite examining records at seven major natural history collections [ANSP, MCZ, NHMUK, University of Michigan Museum of Zoology (UMMZ), National Museum of Natural History (USNM), North Carolina Museum of Natural Sciences (NCMNS), and Florida Museum of Natural History (UF)], we have not located a single lot from the Alabama River that could conclusively be identified as the Painted Rocksnail. After careful examination of historical collections and consideration of both contemporary surveys and broad biogeographic patterns, we conclude that all possible type material of L. taeniata more closely resembles L. picta than the current concept of the Painted Rocksnail.
Leptoxis picta was described from the same location as L. taeniata. The original description of L. picta is sufficiently vague that it could be applied to multiple Mobile River drainage Leptoxis species. Leptoxis picta was described as, "Shell sub-oval, shoulder obtusely rounded; aperture ovate, large; columella callous above; epidermis olive, with numerous quadranglular small spots disposed in revolving lines, strongly marked in the aperture" (Conrad 1834a: 342-343). To differentiate the two species, Conrad (1834b) noted that L. picta often had pigmentation spots and that L. taeniata had dark green bands, but both patterns have been documented in both species (Figs. 2,3,5). Further, Leptoxis picta was described as inhabiting pebble bars, whereas L. taeniata was observed to inhabit friable calcareous banks and siliceous breccias (Conrad 1834b). We question whether any pleurocerid species could be reliably distinguished based on minor differences in habitat preference as we have often observed individuals of the same Leptoxis species to inhabit many different microhabitats (i.e. near the banks and in the main current, both pebble and bedrock substrates; Whelan et al. pers. obs.).

Results
Taking all the above into consideration, we have concluded that the type material of L. taeniata and L. picta represents the same taxonomical species and that the two are synonyms. Leptoxis picta was described four months prior to L. taeniata and thus has priority under Article 23 of the International Code of Zoological Nomenclature (ICZN). Nevertheless, the current concept of the Painted Rocksnail represents a unique monophyletic clade in phylogenetic analyses of Leptoxis species, and possesses body coloration patterns (Fig. 4) and egg laying behavior different from that of L. picta (Whelan et al. 2015). Misapplication of the name Leptoxis taeniata to specimens from the Coosa River apparently became widespread after the publication of Tryon (1873), when L. coosaensis (Lea, 1861) was synonymized with L. taeniata. The name taeniata does not meet requirements of Art. 23.9.1 of the ICZN that prevailing usage must be maintained because the name has not been used as valid in 25 or more publications in the last 50 years (Burch 1982;Burch and Tottenham 1980;Dillon and Lydeard 1998;Holznagel and Lydeard 2000;Johnson et al. 2013;Lee et al. 2006;Lydeard et al. 1997;Lydeard et al. 1998;Strong and Köhler 2009;Tolley-Jordan et al. 2015;Whelan 2016;Whelan et al. 2015). In such instances, one possible course of action would be conservation of prevailing usage by designation of a neotype under Art. 75.6, which would require a request to the Commission to use its plenary powers to set aside any existing name-bearing types and select a neotype. However, the name L. taeniata has long been associated with an incorrect historical distribution for the Painted Rocksnail and further use of the name could perpetuate this error and create confusion for future management plans. Thus, rather than maintain prevailing usage, we here prefer to recognize the oldest available name for the Painted Rocksnail, which we have determined to be Leptoxis coosaensis (Lea, 1861).
Historical distribution. Coosa River above the Fall Line from Wetumpka, Alabama, upstream to the confluence of Terrapin Creek and the Coosa River in Cherokee County, Alabama. Some large Coosa River tributaries including Choccolocco, Buxahatchee, Talladega, and Terrapin creeks.
Remarks. USNM 121295 (Fig. 3A) originates from the Lea collection, is from the published L. coosaensis type locality, leg. Showalter, and is the shell figured by Lea (1863), with the number 65 inked onto the apertural aspect of the body whorl corresponding to the figure number. Lea (1863) indicated that he had six specimens, yet there are seven specimens distributed among the three simultaneously accessioned lots now registered as USNM 121294 (Fig. 3F-I), USNM 1456804 (see photographs on FigShare, https://doi.org/10.6084/m9.figshare.5084272.v1), and USNM 121296 (Fig. 3J). Consequently, either the specimen count in Lea (1863) was in error, or one of the included shells was acquired subsequently and has no type status. Lea (1861) described Anculosa coosaensis from the Coosa River and provided a brief description in Latin. In a subsequent extended description in English, Lea (1863) characterized the species as, "smooth, obtusely conical, thick, dark horn-color, very much banded; spire elevated, obtuse at the apex; sutures very much impressed; whorls four, very much constricted below the sutures, the last large; aperture rounded, white, much banded within; columella thickened, incurved, dark purple; outer lip acute and expanded." In the remarks, he commented that the aperture is more than half the length of the shell and that the bands may be interrupted. This description, as well as the type material ( Fig. 3F-J) match the current concept of the Painted Rocksnail (Fig. 2). Although formerly considered a synonym of L. taeniata, L. coosaensis is the oldest available name for the Painted Rocksnail.
We have been unable to locate type material of Anculosa taeniata lucida Goodrich, 1944. No holotype was designated, nor was a figure provided, but based on the original description and the type locality of tributaries of the Coosa River, we conclude that this entity does not merit recognition at the subspecies level and synonymize it with L. coosaensis.

Leptoxis picta (Conrad, 1834)
Anculosa picta Conrad, 1834a: 343, pl. 1, fig. 16. Possible syntype MCZ 294989 (4 spms); possible syntypes USNM 12074 (2 spms). "Alabama River" [near Claiborne]. Anculosa taeniata Conrad, 1834b: 63. Lectotype ANSP 27620 (Baker, 1964;as "27620a"); paralectotypes ANSP 413583 (3 spms; formerly 27620); paralectotypes MCZ 294987 (4 spms); possible paralectotype NHMUK uncatalogued (1 spm). "Alabama River at Claiborne." Other references: Leptoxis picta-Haldeman 1848: 3, figs. 74-80; Burch and Tottenham 1980: 154, fig. 476;Burch 1982: 42, fig. 476;Lydeard et al. 1997: 117-128;Dillon and Lydeard 1998: 113-121, fig. 2 hence this does not constitute a valid lectotype designation. Furthermore, there is no evidence that the lot has any type status; it originated from the Wheatley collection via the University of Pennsylvania and there is no evidence on the labels or in the original ANSP ledger that the lot was obtained from Conrad. It is possible that Baker knew that Wheatley received material from Conrad, but the original label is not in Conrad's handwriting (G. Rosenberg, pers. comm.). Despite Conrad stating that the type material had been deposited in the ANSP, we have been unable to locate any other possible type material during several searches of the collections. USNM 12074 (Fig. 5A, B) and MCZ 294989 (Fig. 5C) both resemble the shell figured by Conrad and have labels indicating they were received from Conrad. However, both lots are accompanied by labels bearing the less-specific locality Alabama, rather than Alabama River or Alabama River at Claiborne. Moreover, as mentioned, the possible syntypes were not found in ANSP, the stated repository of the types. Consequently, it is possible that neither MCZ 294989 nor USNM 12074 are syntypical and so we refrain from designating a lectotype.
Tryon (1873) considered both Leptoxis foremani (Lea, 1843) and L. flammata (Lea, 1843) to be synonyms of L. picta. Burch and Tottenham (1980) restored L. foremani to species status, but retained L. flammata as a synonym of L. picta. Both L. picta and L. foremani are reciprocally monophyletic and valid species (Whelan et al. 2015). However, based on shell morphology we here consider L. flammata and L. foremani to be synonyms. As both were described concurrently (Lea 1843), we here take the right of First Reviser (ICZN Art. 24.2) and establish the priority of L. foremani over L. flammata, making L. flammata a subjective junior synonym of L. foremani. Leptoxis zebra (Anthony, 1860) was also considered by Tryon (1873) to be a synonym of L. picta, but the type material (MCZ 161794, see shells photographs on FigShare, https://doi. org/10.6084/m9.figshare.5084272.v1) resembles L. foremani. Consequently, we here consider L. zebra also to be a junior synonym of L. foremani.

Discussion
Today, the Painted Rocksnail, i.e. Leptoxis coosaensis, is mostly restricted to Coosa River tributaries (Fig. 1). In tributary habitats, pleurocerids are generally smaller than main stem conspecifics. Consequently, modern specimens of wild caught individuals typically are smaller and have more eroded spires than those seen in types and some historical material. Juveniles grown in captivity with uneroded spires have four extremely compressed whorls ( Fig. 2; Whelan et al. 2015). We failed to find any specimens with costae and consider the specimen with prominent costae figured in Burch and Tottenham (1980: Fig. 486) as the Painted Rocksnail to be a probable misidentification, possibly of L. showalterii. Pleurocerids are notoriously difficult to identify, and similarities in shell morphology of L. picta and L. coosaensis, particularly of the juveniles, undoubtedly contributed to the confusion in application of an incorrect scientific name to the Painted Rocksnail for nearly 150 years (Tryon 1873).
Leptoxis coosaensis is listed as threatened under the Endangered Species Act as "L. taeniata" (Clark 1998). The species is currently restricted to four disjunct popula-tions in the Coosa River drainage in eastern Alabama. Those in Choccolocco Creek, Ohatchee Creek, and Buxahatchee and Watson creeks appear stable, while the status of the population in the Logan Martin Dam tailwaters is unclear because of low abundance and/or difficulties associated with sampling at this location (i.e. depths that require diving in high flow and poor visibility).
Listing of this species was based, in part, on the misperception that the historical distribution included a long stretch of the Alabama River from which it had been extirpated during the 20th century (Clark 1998). Furthermore, many museum lots identified as "L. taeniata" represent taxonomical species different from the Painted Rocksnail, typically of L. picta or L. ampla. The latter is endemic to the Cahaba River drainage above the Fall Line in Alabama; consequently, these misidentifications have resulted in erroneous reports that the Painted Rocksnail was historically present in the Cahaba River (Burch and Tottenham 1980;Goodrich 1922;Mirachi et al. 2004). Conversely, records of L. ampla in the Coosa River drainage are misidentifications, typically of L. coosaensis. In light of our reanalysis, the historical range of L. coosaensis is here revised to have been restricted to the Coosa River and its tributaries, which is a considerably smaller historical range for the Painted Rocksnail than previously believed (Clark 1998). Nevertheless, the current range reduction of 90% from historical occupancy given by Clark (1998) appears to have been conservative.
The historical range of L. coosaensis just in the Coosa River proper is a distance of approximately 317 km (Fig. 1). Since L. coosaensis is now known to inhabit less than 10 km of the main stem Coosa River, its range has declined over 95% in that river alone. In addition, Leptoxis coosaensis is believed to be extirpated from four of the eight Coosa River tributaries from which it was known. As such, even with a redefined and reduced historical range, this species is in obvious need of continued protection. Management efforts for pleurocerid snails in the Mobile River basin have focused on habitat improvement and captive propagation and reintroduction. Reintroduction should never occur outside a species' historical range, which is one reason why clarifying the range of L. coosaensis is so important. Recovery efforts should include a review of historic tributaries that once supported L. coosaensis to determine if any sites are appropriate for reintroduction. As with most listed mollusks, establishment of additional populations contributes to species recovery, and is necessary for the possible delisting of the species (Hartfield 2005).