Salamandra auriculata Holbrook, 1838b:115

Pyron, R. Alexander & Beamer, David A., 2022, A nomenclatural and taxonomic review of the salamanders (Urodela) from Holbrook’s North American Herpetology, Zootaxa 5134 (2), pp. 151-196 : 163-164

Salamandra auriculata Holbrook, 1838b:115 , pl. 28

Holbrook authored the original description ( Holbrook 1838b) of the species currently recognized as Holbrook’s Southern Dusky Salamander , Desmognathus auriculatus (see Means 1999; Means et al. 2017) with a brief account and an illustration by C. Rogers. Holbrook (1842e:47, pl. 12) reprinted the account with an expanded description and the same drawing, recolored. The characters and drawing leave little doubt that this is the swamp-dwelling Southern Dusky Salamander ( Petranka 2010; Powell et al. 2016), though this name has historically been applied to at least three distinct species, at least two of which occur near the type locality in southeastern Georgia ( Beamer and Lamb 2008; Means et al. 2017; Pyron et al. 2022a,b). Most crucially for the purposes of allocation, Holbrook stated: “The specimens from which the above description was taken came from Riceborough, in Georgia, and were furnished me by my friend Dr. Harden. ” This is present-day Riceboro, in Liberty County, Georgia.

Dr. John McPherson Berrien Harden (1810–1848) was a renowned physician of the mid-19th century in Georgia who resided, after 1833, with his wife Jane LeConte at Woodmanston, the family plantation in Liberty County, Georgia, near Riceboro ( Stephens 1975). The LeConte family produced numerous well-known naturalists ( Barnhart 1917). Indeed, there are specimens of Desmognathus auriculatus collected by “LeConte” in both Philadelphia (ANSP 14005–19) and Paris (MNHN-RA 0.4665), most likely from Maj. John Eatton LeConte, Jr. Accordingly, it seems likely that any specimens of Harden’s were likely collected at or near his home ( Stephens 1997), now preserved as LeConte-Woodmanston Plantation and Botanical Gardens National Historic Site (NRHP 73000626), ~ 5km SW of present-day Riceboro. Therefore, we regard this as the type locality under Recommendation 76A.

Holbrook (1838b) clearly referred to a series of primary syntypes (“[t]he specimens from which the above description was taken...”), but their disposition is currently unknown. Hallowell (1858) stated that there were “ 3 specimens in Mus. Acad. [ANSP], presented by Dr. Jones and Prof. Holbrook.” Examination of the holdings of the ANSP reveals two specimens from “ Dr. Jones ” with the locality “ Georgia,” but none from Holbrook. Dr. William Louis Jones (1827–1914) was raised on a plantation adjacent to Woodmanston , and often collected with the older Harden ( Stephens 1997). He donated many amphibian and reptile specimens to the ANSP, MCZ, and USNM, but only beginning after Harden’s death, and was only 11 years old when Holbrook’s (1838b) description was published. Upon examination, the Jones specimens (ANSP 938–939) are clearly referable to Desmognathus auriculatus sensu lato but lack precise locality data beyond “ Georgia ” and cannot be assigned to a specific genetic lineage at present.

Subsequently, Cope (1869) referred to 10 specimens in the lot USNM 3901 labeled “Riceboro, Ga., Dr. Jones, var. auriculata ” in his account of Desmognathus fuscus . Yarrow (1882) listed the collections of the USNM, of which those received by Dr. W. L. Jones range in date from 1851 to 1861, and for which he also only records 6 specimens cataloged under USNM 3901. Dunn (1917) stated that the type is “not known to exist,” noting that he examined 12 specimens in USNM 3901 from Riceboro and recording that they were from the “(type-locality).” Dunn (1926) then stated unequivocally that USNM 3901, consisting of 12 specimens he examined, represented the type series.

It is possible that Cope’s (1869) identification of USNM 3901 as “ var. auriculata ” in a table of numerous Desmognathus fuscus specimens formed the basis of Dunn (1926) ’s identification of that material as the type series. Note that Cope (1869) did not identify them as types. We find no reference in Dunn’s writings between 1917 and 1926 to explain his sudden reversal and acceptance of USNM 3901 as the type. He may have intended a sort of “neosyntype” designation, though this would not be valid under Article 75.3, as there was no explicit statement of intent or characters. Cochran (1961) did not list USNM 3901 in her catalog of types held at the USNM. Adler (1976) states that the catalog entry for USNM 3901 lists “Riceboro, Georgia ” without a date or collector, while the species card entry lists “ Dr. W. L. Jones ” as the collector (R.L. Crom bie, pers. c omm. to Adler, 17 May 1976). Adler (1976) discounted the possibility that USNM 3901 represents the type series, because of Cochran (1961) ’s omission thereof, and the discrepancy between Holbrook’s assertion that “ Dr. Harden ” furnished him with the type series, while “ Dr. Jones ” collected the USNM material (see below).

Our examination of USNM 3901 revealed only a single specimen in the jar, clearly referable to Desmognathus auriculatus sensu lato, with the collection information as listed above: “ Dr. W.L. Jones —Riceboro, Georgia.” The original lot of 12 was cataloged 2 August 1858, and the 12 th specimen was later marked by Cope as being destroyed. The remaining ten specimens were re-catalogued as USNM 271136–45 on 30 December 1986. It seems improbable that ANSP 938–9 or USNM 3901 are Harden’s specimens on which the description is based but were instead collected and donated by Jones much later, albeit from at or near the type locality.

As Holbrook (1838b) mentions only the few specimens received from Harden and no further collections of his own in subsequent editions, discovery of any contemporaneous Desmognathus auriculatus specimens associated with Holbrook would likely represent primary syntypes. No specimens in the ANSP or MCZ collections are known to have originated with Harden or Holbrook. We previously stated ( Pyron and Beamer 2020) that a specimen of “ D. niger ” from Holbrook (ANSP 14001) was probably the specimen referred to by Hallowell (1858), but we now conclude this is incorrect. Rather, ANSP 14001 is most probably the exemplar of “ Triton niger ” illustrated in Holbrook (1842e), which represents in part what is now called D. quadramaculatus (see accounts for Salamandra maculo-quadrata and Triton niger ).

Fortuitously, a specimen in Paris is probably referable to the type series. As noted above, Hallowell (1858) listed three specimens of “ Plethodon auriculatum ” in the ANSP, presented by Dr. Jones and Prof. Holbrook. The ANSP catalog notes only two (ANSP 938–9) from Dr. Jones, in an entry completed in the 1890’s at the earliest (see Malnate 1971; Pyron and Beamer 2020). The third specimen from Holbrook is unaccounted for but would undoubtedly be considered a type if extant. Among the specimens sent to the MNHN by Hallowell in December 1856 is one recorded in the MNHN exchange ledger (“ Inventaire des Collections ”) vol. 42 (2), page 82 as “ Plethodon auriculatum ” from Philadelphia , cataloged as “ Cylindrosoma auriculatum ” in Paris , with a locality of “Georgie.” The MNHN catalog vol. 1, page 123 lists it as MNHN-RA 0.4675, a Desmognathus auriculatus from “ Georgie ” with Hallowell as the collector. As described previously, it does not appear that Hallowell himself actually collected salamanders in southern Georgia. Rather, some of the specimens he sent from the ANSP likely lacked collector data or labels, and the MNHN did not distinguish between collector and donor in such instances (A. Dubois and A. Ohler, pers. comm.).

Thus, it seems very likely that this specimen, MNHN-RA 0.4675, is the third specimen of Desmognathus auriculatus listed by Hallowell (1858), from Holbrook ( Fig. 3 ). Therefore, we conclude that it is part of the type series from Woodmanston, collected by Harden and donated by Holbrook to the ANSP, before being transferred to the MNHN by Hallowell in December 1856. The specimen is shorter (~ 3.5 inches) than the total length of 5 inches given in the description, but this was based on a series. Furthermore, we suggest that this specimen may be the one illustrated in pl. 28, given that it matches the description of color pattern, and that Holbrook apparently preferentially preserved and deposited his illustrated specimens in Philadelphia. Specifically, the diagnostic white “portholes” on the lateral surfaces of the tail and body that characterize many populations of D. auriculatus are reduced or absent in both MNHN-RA 0.4765 and the specimen figured in pl. 28. The specimen is well-preserved, to the extent that other color pattern features are visible on the dorsal and ventral surfaces, and the absence of the portholes does not appear to be a consequence of age or preservation.

Therefore, we designate MNHN-RA 0.4675 as the lectotype of Salamandra auriculata Holbrook, 1838b , with the type locality of Woodmanston Plantation near Riceboro, Liberty Co., Georgia, collected by Harden sometime between 1833 and 1838. The remaining primary syntypes of unknown number and disposition thereby become paralectotypes. Hallowell (1858) recorded the “habitat” of the species as “Riceboro’, Georgia, South Carolina ,” apparently the first reference to populations outside of Georgia. As MNHN-RA 0.4675 bears a locality of Georgia dating to at least 1856, it is possible that this “South Carolina ” reference is simply to Holbrook’s Charleston residence as with many of his specimens, that the Jones specimens are actually from South Carolina rather than Georgia as recorded in the later ANSP catalogue, or simply that Hallowell had received reports from elsewhere. However , this does not solve the problem of allocating this name to a species-level genetic lineage. Populations referred to this taxon have a complex nomenclatural history ( Grobman 1950; Rossman 1959; Valentine 1963), but it was long treated as a single, distinct species by most recent authors ( Means 1999).

Subsequent genetic analyses have shown that populations historically included in Desmognathus auriculatus comprise multiple distinct taxa ( Beamer and Lamb 2008, 2020). One species-level lineage from Louisiana and Mississippi was recently described as D. valentinei Means, Lamb, and Bernardo, 2017 . Remaining populations comprise at least two candidate species ( Pyron et al. 2020, 2022a), D. auriculatus A (Alabama, Georgia, and Florida) and D. auriculatus B/C (Georgia, North Carolina , and South Carolina ), both of which occur near Riceboro ( Pyron et al. 2022b). While D. auriculatus A has recently been treated as the nominotypical lineage ( Means et al. 2017; Beamer and Lamb 2020; Pyron et al. 2022b), specimens from the type locality have not been sampled genetically. Therefore , which lineage (or both) occurs there, and which species is represented by the lectotype, is unknown. Additional data, analyses, and nomenclatural action will be needed to allocate this name conclusively.

Figures

FIGURE 3. Illustration of a primary syntype of Salamandra auriculata Holbrook, 1838b, and a specimen (MNHN-RA 0.4675) from “Georgie [Woodmanston Plantation near Riceboro, Liberty County, Georgia],” here designated as lectotype, in dorsal (a, c) and ventral (b, d) views. The remaining primary syntypes of unknown number and disposition thereby become paralectotypes. This is a “type by association” (Bell 1996; Pyron and Beamer 2020), as we infer that it is the ANSP specimen from Holbrook described by Hallowell (1858), sent by Hallowell from Philadelphia to the MNHN in Paris in December 1856. Grid is in cm.