Salamandra maculo-quadrata Holbrook, 1840:121

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 : 167-173

Salamandra maculo-quadrata Holbrook, 1840:121 , pl. 26

[= S. fusca Green, 1818 ]

Black-bellied Salamanders are a southern Appalachian complex of large, heavy-bodied, semi-aquatic salamanders with dark, patternless venters ( Petranka 2010; Powell et al. 2016). For 119 years, they have usually been referred to as Desmognathus quadramaculatus after the “rediscovery of one of Holbrook’s salamanders” by Stejneger (1903). This group is polyphyletic and comprises as many as 6 distantly related species with apparently parallel or convergent morphologies ( Jackson 2005; Kozak et al. 2005; Jones and Weisrock 2018; Beamer and Lamb 2020; Pyron et al. 2020). However, this is only part of the nomenclatural difficulty associated with these taxa. As long hinted at by previous authors (e.g., Fouquette and Dubois 2014), Holbrook’s name was obviously intended to apply to populations of Salamandra fusca Green, 1818 , and it is indisputably a junior subjective synonym thereof. Astonishingly, no valid name has ever been applied to Black-bellied Salamanders at the species level.

As noted by previous authors ( Pyron and Beamer 2020), Salamandra nigra Green, 1818 was the first name applied to the Black-bellied Salamanders following Holbrook (1842e) ’s account of Triton niger ( Green, 1818) , until Stejneger (1903). In contrast, S. quadramaculata Holbrook , “1842” was generally regarded as a junior subjective synonym of Triturus fuscus Rafinesque, 1820 following its description, designated as such by Baird (1850). Cope (1859) wrote: “ D. fusca (Green) . Salamandra quadrimaculata (Holbrook) is undoubtedly the young of this species.” Stejneger (1903) even stated in his opening sentence “Among the many synonyms usually cited under Desmognathus fusca (Rafinesque) , Salamandra quadrimaculata of Holbrook a has been considered as belonging to this species without a shadow of a doubt. [a North Am. Herpet., 2 ed., V, 1842, p. 49, pl. xiii.]” Stejneger’s reasons for then declaring it distinct are unclear. Though the history of this confusion is complex, a re-discovery of some of Holbrook’s type material provides a simple, albeit disruptive, resolution of this nomenclatural ambiguity.

The description of Salamandra maculo-quadrata Holbrook (1840:121 , pl. 26) actually appeared in the fourth volume of the first edition with an illustration by J. Queen and was reprinted in the fifth volume of the second edition with an expanded description and a new engraving by J. H. Richard ( Holbrook 1842e:49, pl. 13). The fourth volume of the first edition was essentially unknown until Gill (1903a) reported its existence (see Worthington and Worthington 1976). Most workers to that point, including Stejneger (1903), instead cited Holbrook (1842e) as the taxonomic authority for this name. The later account contains a few differences from the original description, but on balance clearly refers to the same taxon, contrary to the suggestions of Adler (1976). A few points are relevant here.

First, in the first edition, Holbrook (1840) labeled the illustration (pl. 26 sequentially, misnumbered 27 or 29 in some copies; see Shimek 1924) “ Salamandra quadramaculata ,” although the text in the Table of Contents and on pages 121–122 reads “ Salamandra maculo-quadrata .” In the second edition, Holbrook (1842e) labeled pl. 13 “ Salamandra quadramaculata ,” while the Table of Contents and text on pages 49–50 uses “ Salamandra quadrimaculata .” Later authors such as Stejneger (1903) generally used the spelling “ quadrimaculatus .” In contrast, Valentine (1974) regarded Holbrook (1842e, pl. 13) as the First Reviser, stating that he had thereby fixed “ Salamandra quadramaculata ” as the correct name on the plate, albeit while using an incorrect subsequent spelling on pages 49–50.

Second, there are two different illustrations in the different volumes ( Fig. 4 ); pl. 26 by J. Queen in Holbrook (1840) and pl. 13 by J.H. Richard in Holbrook (1842e). Based on the color pattern, position of the limbs, and differing measurements given, these may represent two different specimens, neither of which are Desmognathus quadramaculatus as currently recognized; large-bodied, black-bellied salamanders from the southern Appalachians ( Stejneger 1903; Valentine 1974). Instead, they are quite clearly D. fuscus , moderately sized dusky salamanders from the northeastern United States ( Petranka 2010; Powell et al. 2016).

Corroborating this is the statement in both editions that ( Holbrook 1840): “The range of the Salamandra maculoquadrata [ Holbrook 1842e: quadrimaculata] is much more extended in the Atlantic states than I had at first believed; it is common in Georgia and Carolina , and is an inhabitant of Pennsylvania, from which state I have received living specimens [emphasis added]; and no doubt it may be found in all the intermediate country.” Holbrook (1842e:50) clarified that: “This is a land animal, but I am ignorant of its habits, never having seen but two living specimens,” the primary syntypes which we must assume are those illustrated. Furthermore, Holbrook (1842e:81) was clearly also aware of the large-bodied, black-bellied salamanders of the southern Appalachian Mountains , having collected several and included them in his account of Triton niger ( Green, 1818) . These are MCZ A-183 and ANSP 14001, neither of which have accurate locality data, but are presumably those referenced by Holbrook (1842e:82) in his note “I have seen it in Carolina and Georgia,” and one of which is presumably that illustrated by J. Queen in pl. 27 (see Pyron and Beamer 2020).

Following this account and the revision of Baird (1850), most subsequent workers used “ Desmognathus niger ” or “ D. nigra ” to refer to the black-bellied populations. However, while the name Salamandra nigra Green, 1818 is a junior subjective synonym of S. fusca Green, 1818 (see Pyron and Beamer 2020), this is not the justification given by Stejneger (1903) for resurrecting S. quadrimaculata Holbrook, 1842e . Instead, Stejneger continued to recognize “ D. nigra ” as well as “ D. fusca ,” but did not detail his conception of those taxa. It is thus perplexing as to why he then declared that a dozen specimens from Grandfather Mountain (USNM 30891–30902) “seem to indicate that Holbrook’s name belongs to a species well separated from Desmognathus fusca and in some respect approaching D. nigra . ” He concluded by stating: “That the specimens here mentioned really belong to Holbrook’s D. quadrimaculata seems certain. So far as proportions of body and limb are concerned, Holbrook’s plate agrees perfectly with our specimens.The squarish spots in the figure appear considerably more regular than in the specimens before us, among which there is a very great individual variation in this respect, but the dorsal pattern of a couple of the medium-sized specimens is sufficiently close to show that they belong to Holbrook’s species.” He made no mention of the lack of a black belly in Holbrook’s description, or the much smaller size of Holbrook’s specimens.

Having examined this series (USNM 30891–30902), they are all metamorphosed, Black-bellied Salamanders, ranging from young to very large adults (~ 8–10cm SVL), with the typical darkened and indistinct or obscured dorsal color-pattern (more evident in smaller specimens) and patternless, darkened or blackish ventral coloration. In contrast, Holbrook’s account of the species in both volumes (1840 and 1842e) describes a species having a total length of “ 3 inches 8 lines [3.75 inches or 9.5cm],” being “a land animal,” and having a “two rows of small elongated, subquadrate red spots,” the tail having “a red mesial line,” and “the inferior surface of the whole animal is pepper-and-salt grey, with a strong purple tinge.” This is evidently not a description of Black-bellied Salamanders from the Appalachians, but clearly of the two living Pennsylvania specimens he received, obviously Desmognathus fuscus , as illustrated in pls. 26 and 13 in the first and second editions, respectively ( Holbrook 1840, 1842e).

This yields a rather unusual situation. The first name widely applied to large-bodied, semi-aquatic Appalachian populations of black-bellied salamanders was based on Holbrook (1842e) ’s erroneous expansion of Salamandra nigra Green, 1818 in his account of Triton niger ( Green, 1818) . As noted, S. nigra Green, 1818 is a junior subjective synonym of S. fusca Green, 1818 , and is therefore invalid for Black-bellied Salamanders. However, this was unknown to Stejneger (1903), who instead resurrected S. quadrimaculata Holbrook, 1842e [= S. maculo-quadrata Holbrook, 1840 ] for some populations of Black-bellied Salamanders, while also still recognizing S. nigra Green, 1818 for others. However, S. maculo-quadrata Holbrook, 1840 is also a junior subjective synonym of S. fusca Green, 1818 . Thus, both names widely applied to Black-bellied Salamanders in the past (“ Desmognathus niger ” and “ D. quadramaculatus ”) are, in fact, junior subjective synonyms of Northern Dusky Salamanders, D. fuscus .

The confusion stems primarily from Holbrook (1840) ’s imprecise description of Salamandra maculo-quadrata and Holbrook (1842e) ’s overly broad account of Triton niger ( Green, 1818) , both of which could potentially include populations of Black-bellied Salamanders in their geographic distribution, and the latter of which is based on at least two specimens thereof. This confusion did not escape the notice of later authors such as Dunn (1926), Valentine (1974), and Adler (1976). Indeed, Valentine (1974) noted that: “Holbrook’s descriptions of S. quadra-maculata and T. niger appear to be composites of modern D. fuscus , D. quadramaculatus , and perhaps D. auriculatus . ” However, Holbrook’s account of Triton niger ( Green, 1818) carries no nomenclatural force, and S. nigra Green, 1818 was recently designated a junior subjective synonym of S. fusca Green, 1818 (see Pyron and Beamer 2020).

On the other hand, allocation of Salamandra maculo-quadrata Holbrook, 1840 still depends on the type specimens, which are not MCZ A-183 or ANSP 14001 and are not known to exist according to Dunn (1926) and Valentine (1974), who noted the need for a lectotype. Brame (1972) suggested that ANSP 490 was the holotype, but this was an error; it is actually a primary syntype of S. quadridigitata Holbrook, 1842e (see below; Malnate 1971; Wray et al. 2017). Stejneger (1903) ’s “rediscovery” of “ Desmognathus quadrimaculata (Holbrook) ” does not qualify as a valid neotype designation for the series USNM 30891–30902, as it does not contain an express statement of intent (Article 75.3). Relatedly, Schmidt (1953) restricted the type locality to the Great Smoky Mountains. Valentine (1974) suggested this refer only to the North Carolina portion due to Stejneger (1903) ’s “rediscovery” being based on specimens from Grandfather Mountain , an entirely different range in the Blue Ridge from the Smokies .

Much of the above information was previously summarized in the unpublished PhD thesis of Wooten (2008). She subsequently submitted an application to the ICZN entitled “ CASE 3427. Desmognathus quadramaculatus Holbrook, 1840 ( Amphibia: Caudata : PLETHODONTIDAE ): proposed conservation of usage of the specific name by the designation of a neotype. J.A. Wooten & L.J. Rissler ” ( Anonymous 2007). As detailed in her thesis, Wooten proposed designating UAHC 15652, collected from Noland Creek, Swain County, North Carolina (35.55575, - 83.48084) by J.A. Wooten and S.A. Eagle on 23 March 2006, as the neotype to replace an unidentifiable namebearing type under Article 75.5. They chose this location as it represents the North Carolina side of the Great Smoky Mountains , as recommended by Schmidt (1953) and Valentine (1974). This would represent the quadramaculatus F lineage of Beamer and Lamb (2020) and Pyron et al. (2020, 2022b). However, this case was never published in full before being closed ( Anonymous 2008). Thus, the type status is still undetermined, the type locality has been dubiously restricted by later authors, and the two primary syntypes may nonetheless still be extant among the known collections of Holbrook.

Indeed, a series of Salamandra fusca Green, 1818 (ANSP 814–825) is recorded in vol. 1 of the ANSP catalog as having been donated by Dr. Holbrook, with the locality of “Penns.” All are clearly Desmognathus fuscus in the modern sense ( Petranka 2010; Powell et al. 2016). All but one is in moderate but recognizable condition. They are all very soft, likely never having been fixed in formalin, probably only ethanol or a similar spirit (see Simmons 2014). One, however, is of markedly different size, coloration, shape, and condition. This specimen, ANSP 821 ( Fig. 5 ), is slightly smaller, darker, and more rigid and partially desiccated in preservative. It also matches almost exactly, in color-pattern and general description, the illustration of Salamandra maculo-quadrata Holbrook, 1840 :pl. 26, specifically in having a dark dorsum with 8 pairs of subquadrate marks on the back and a solid stripe running the length of the tail. The tail is truncated in this specimen (apparently having broken off some time after preservation), so only the beginning of this line is visible. Holbrook notes: “Length of head to anterior extremities, 7 lines [15mm]; length of body to vent, 1 inch 4 lines [34mm] …”, and our measurements of ANSP 821 match these figures when accounting for curvature.

As noted above, there are different illustrations of “ Salamandra quadramaculata ” in Holbrook (1840, pl. 26) and Holbrook (1842e, pl. 13). Holbrook (1842e:50) stated that he has received two living specimens from Pennsylvania; whether one each is illustrated in the different volumes, or the difference is attributable to the change in illustrators between volumes is unclear (see Adler 1976). In any event, the available evidence points strongly to ANSP 821 being the primary syntype described and illustrated for S. maculo-quadrata Holbrook (1840:121 , pl. 26). We hereby designate it as the lectotype, rendering the type locality “Penns[ylvania].” The second primary syntype thereby becomes a paralectotype; its whereabouts are unknown (but see below). This represents a rare addition to the type collection of both Holbrook and the ANSP (see Malnate 1971; Adler 1976). Given that the illustrations, description, and locality stated by Holbrook in the account, and the locality and apparent identity of the lectotype are all congruent with the Northern Dusky Salamander ( Desmognathus fuscus ), we reiterate earlier authors such as Baird (1850) and Cope (1859) in recognizing S. maculo-quadrata Holbrook, 1840 as a junior subjective synonym of S. fusca Green, 1818 .

A specimen in Paris (MNHN-RA 0.4671) may also bear on this matter ( Fig. 6 ). The December 1856 donation from Hallowell contained a specimen that the “Inventaire” records as “ Plethodon quadrimaculatum ” from the ANSP, received as “ Plethodon (Cylindrosoma) quadri-maculatum ” in the MNHN, from “Philadelphie.” The MNHN catalog labels it “ Desmognathus quadrimaculatus ,” with different, later handwriting crossing out the species and writing in “ auriculatus ,” with the locality “Philadelphie” and donor as “Hallowell.” A much later margin entry in red pen notes “= D. fuscus ( Green, 1818) .” Hallowell (1856) recorded this species as dubious (“[P.] quadrimaculatus?”), and this the only specimen we have encountered that was accessioned in a museum under this name. While it is possible that Hallowell collected this specimen himself in Philadelphia and used a disfavored name to catalog it, we suggest it is equally likely that this specimen was in the ANSP under that name from Holbrook. Therefore, it may represent Holbrook’ second primary syntype from Pennsylvania, perhaps even that illustrated by J. H. Richard in Holbrook (1842e, pl. 13). Lacking a firm connection between this specimen and Holbrook, we do not go so far as to declare this formally, but the specimen itself appears to be a Northern Dusky Salamander ( D. fuscus ), likely from Pennsylvania as labeled, reinforcing the junior subjective synonymy of S. maculo-quadrata Holbrook, 1840 therewith.

While this turn of events (rediscovery of a type and relegation to synonymy) is both simple and straightforward, it is nonetheless disruptive of 119 years of literature on Black-bellied Salamanders. One solution would be conservation of prevailing usage by a neotype (Article 75.6), whereby the present authors maintain prevailing usage (Article 82) and request the Commission use its plenary power (Article 81) to set aside the lectotype and designate a neotype. This is undesirable for three primary reasons. The first is that there exists a valid, available name for some Black-bellied Salamanders, Desmognathus quadramaculatus amphileucus Bishop, 1941 . The Commission could use UMMZ 89767 (the holotype of D. q. amphileucus from “Demorest, Habersham County, Georgia ”) as the neotype. This would thereby render it a junior objective synonym of Salamandra quadramaculata Holbrook, 1840 and allow D. quadramaculatus to be maintained. However, this would only provide a valid name for one of the six candidate species, quadramaculatus A ( Beamer and Lamb 2020; Pyron et al. 2022b).

Second and more importantly, doing so would not alleviate the taxonomic uncertainty, instability, and lack of universality induced by creating a clearly erroneous linkage between a specimen of one species (a Black-bellied Salamander) and the description of another (the Northern Dusky Salamander). While 119 years of prevailing usage linkgs Salamandra quadramaculata Holbrook, 1840 from Stejneger (1903) to the present day, there is also literature spanning a period of at least 53 years from Baird (1850) to Stejneger (1903) during which numerous workers (e.g., Gray 1850; Cope 1869) recognized the clear and obvious junior subjective synonymy of S. quadramaculata Holbrook, 1840 with S. fusca Green, 1818 .

We recognize that more workers applied Salamandra quadramaculata Holbrook, 1840 to Black-bellied Salamanders between 1903 and 2022 than applied it to Northern Dusky Salamanders between 1840 and 1903. Regardless, designating a specimen of a Black-bellied Salamander as the neotype of a taxon with a history of recognition as a junior subjective synonym of another widespread species would only increase confusion, as the type specimen would then be clearly at odds with the diagnosis in the original description, and the subsequent usage of that name for 63 years. This situation would only be compounded if MCZ A-183 or ANSP 14001 were chosen as the neotype, given their origin with Holbrook (1842e) ’s account of Triton niger ( Green, 1818) , itself also a junior subjective synonym of S. fusca Green, 1818 . Both specimens are also in extremely poor condition, bear few identifiable characteristics, and lack accurate localities to tie them to any of the candidate species.

Third and perhaps most crucially, despite the massive amount of literature utilizing the name under prevailing usage, Black-bellied Salamanders are not monophyletic; they are not even a single species, let alone one which is coherently diagnosed and consistently recognized ( Kozak et al. 2005; Wooten and Rissler 2011; Beamer and Lamb 2020). Instead, the currently known populations are arranged in at least six lineages representing candidate species in two distantly related clades (Pyron et al. 2020, 2022b). Some lineages have exchanged mitochondrial haplotypes with populations of D. marmoratus ( Moore, 1899) , presumably due to hybridization in sympatry ( Jackson 2005; Jones et al. 2006). Specimens of Black-bellied Salamander are also often involved in taxonomic mix-ups with D. marmoratus in the field (e.g., Jones and Weisrock 2018; see Pyron et al. 2020). Thus, any neotype designation (e.g., UAHC 15652) would involve arbitrarily allocating the identity of a single mitonuclear candidate species (e.g., quadramaculatus A or F), while doing nothing to alleviate the taxonomic uncertainty of the remaining lineages previously referred to as D. quadramaculatus . Nor would it provide a stable and universal basis for the recognition of the newly validated D. quadramaculatus itself, as the original description would still clearly refer to another species, D. fuscus .

As of 29 January 2022 on Google Scholar, at least 983 publications have used the name “ Desmognathus quadramaculatus ” and 22 have used “ D. quadrimaculatus ,” representing 119 years of prevailing usage since Stejneger (1903). We have discovered that this is not in taxonomic accord with the existing name-bearing type of Salamandra maculo-quadrata Holbrook, 1840 . However, we argue that stability and universality experience no additional threat from designating S. maculo-quadrata Holbrook, 1840 as a junior subjective synonym of S. fusca Green, 1818 . In contrast, taxonomic and nomenclatural uncertainty would only be increased by designation of a neotype. Thus, we do not invoke Article 75.6. Instead, we conclude that abandoning the use of D. quadramaculatus for Black-bellied Salamanders and relegating it to the synonymy of Northern Dusky Salamanders ( D. fuscus ) is the solution providing the greatest stability. Resurrecting available, valid names for some candidate species and erecting new names for others is consequently the only suitable path forward.

Therefore, future revisions should likely reinstate the name Desmognathus quadramaculatus amphileucus Bishop, 1941 for the quadramaculatus A lineage occupying the Blue Ridge Mountains south of the Great Smoky Mountains (see Beamer and Lamb 2020; Pyron et al. 2022b). The remaining populations, from the Great Smoky Mountains to the Gauley River in West Virginia ( quadramaculatus C, D, E, F, and G) consequently lack any valid, available names at present. We suggest that they continue to be referred to as D. “ quadramaculatus ” (in quotes) or D. sp. “ quadramaculatus ” by interested parties for the time being as new species descriptions are prepared for these taxa. Future workers investigating the group for the first time may initially be perplexed by the discontinuous shift in species names, but this is necessary for both nomenclatural and taxonomic reasons, to reflect the evolutionary and systematic history of the group accurately, as well as the historical publication record.

Figures

FIGURE 4. Illustrations of Salamandra maculo-quadrata Holbrook, 1840 by J. Queen (a, b) and J. H. Richard (c, d) from Holbrook (1842e), in dorsal (a, c) and ventral (b, d) views. Note the misnumbering on the first plate (b).

FIGURE 5. Lectotype (ANSP 821) of Salamandra maculo-quadrata Holbrook, 1840, type locality “Penns[ylvania].” in dorsal (a) and ventral (b) views.The locality, origin, and disposition of the second primary syntype and an unknown number of possible secondary or tertiary syntypes is unknown (see below), but they thereby become paralectotypes. Note the distinct “subquadrate” dorsal blotches (a). Ruler is in mm.

FIGURE 6. Historical specimen (MNHN-RA 0.4674) of Salamandra maculo-quadrata Holbrook, 1840 in dorsal (a) and ventral (b) views. This specimen originated in the ANSP, from the December 1856 donation to Paris from Hallowell, where it is recorded with a locality of “Philadelphia” and a donor of “Hallowell.” Given its proximity to the lectotype of S. auriculata (MNHN-RA 0.4675), the nonspecificity of “Hallowell” as donator rather than collector, the locality congruence with Holbrook’s specimens, and the unusually late date for use of the name as valid, we suggest that this may possibly represent Holbrook’s second primary syntype. Lacking any direct mention of Holbrook in connection with the specimen, we do not conclude this formally, but its existence and age is consistent with the junior subjective synonymy of S. maculo-quadrata Holbrook, 1840 with S. fusca Green, 1818. Grid is in cm.