Lophyrus spinosus C. Duméril & A. Duméril, 1851, a case of mistaken identity

Lophyrus spinosus Duméril & Duméril, 1851 has been considered synonymous with Bronchocela marmorata Gray, 1845 since its original description. The name-bearing type of Lophyrus spinosus is the specimen collected by Hombron and Jacquinot (MNHN 6896) by original designation and the holotype by monotypy of Bronchocela marmorata is the specimen deposited under NHMUK 1946.8.11.16. Accordingly, these two scientific names do not share name-bearing types. Prior to the original descriptions of Lophyrus spinosus and Bronchocela marmorata Hombron & Jacquinot (1843) published a plate depicting Lophyrus spinosus, but only naming the species in French. The nomenclatural implications of this publication are discussed. Our comparison of the holotypes reveals that these two species are not identical. Therefore we resurrect Lophyrus spinosus from its synonymy with Bronchocela marmorata and show that the specimen collected by Hombron and Jacquinot actually belongs to the genus Hypsilurus. Duméril and Duméril (1851) were the first to make the name Lophyrus (= Hypsilurus) spinosus available and the authorship has to be assigned to them. Based on evidence from original travel reports and biogeography we propose that the collection locality of Lophyrus spinosus, i.e. Hypsilurus spinosus Duméril & Duméril (1851), should be corrected to Triton Bay, Kaimana, West Papua, Indonesia.


Introduction
) described a new lizard Lophyrus spinosus based on a single specimen that had been collected during the expedition of the corvettes L'Astrolabe and La Zélée to Oceania and the South Pole. The expedition had lasted from 7 th September 1837 until 6 th November 1840. All specimens collected during the Astrolabe expeditions were assigned to Hombron & Jacquinot in the introduction of a treatise of the reptiles and fishes by Jacquinot and Guichenot (1853). Lophyrus spinosus had been illustrated and named in French on Plate 3 in Hombron andJacquinot (1842-1854), a collection of plates (atlas) depicting some of the animals encountered during their voyage. The plates were distributed in 28 parts (livraisons) over a period of 12 years and the plate depicting Lophyrus spinosus had been published in 1843 according to the wrappers of the livraisons held by the British Library (Clark and Crosnier 2000). Consequently in their description Duméril and Duméril (1851) did not use "nobis" behind the name (which would indicate the description is theirs) but wrote instead: "L. [Lophyre / Lophyrus] épineux Spinosus Hombron et Jacquinot (Voy. au pôle sud et dans l'Océanie sur les corvettes l'Astrolabe et la Zélée, Rept., pl. 3, sans texte)". The next line quotes the description by Gray (1845) of "Bronchocela marmorata Gray,Cat. of Liz.,p. 242". These lines appear before the description of the species and have several nomenclatural implications. Firstly, Hombron and Jacquinot were the collectors of the specimen and probably took notes collated into a manuscript during their journey because Duméril and Duméril (1851) as well as Jacquinot and Guichenot (1853) referred to these notes. As stated by Jacquinot and Guichenot (1853: 1), in their manuscript Hombron and Jacquinot already proposed ("imposé") names for the specimens collected. But the manuscript was never published and therefore no name was made available in accordance with the Code of the International Commission on Zoological Nomenclature (ICZN 1999, in the following text "the Code"). The only published document concerning Lophyrus spinosus was the plate "Sauriens pl. 3" within the atlas published by Hombron andJacquinot (1842-1854), a figure using the French name "Lophyre épineux". No Latin scientific name was given and the French name is not available under the rules of the Code. The first valid publication, making the name Lophyrus spinosus available for nomenclatural purposes, was authored by Duméril and Duméril (1851: 91). Consequently the authorship for the taxon has to be assigned to Duméril and Duméril (1851). Secondly Duméril and Duméril (1851) appear to have considered Lophyrus spinosus conspecific with Bronchocela marmorata Gray, 1845. And in fact already Gray (1845: 242) cited Plate 3 of Hombron and Jacquinot (1843) when describing Bronchocela marmorata. However, Gray (1845) had doubts -indicated by a question mark-to allocate the specimen figured by Hombron and Jacquinot (1843) to his new species. Thus the single specimen of Lophyrus spinosus does not constitute a name-bearing type of Bronchocela marmorata according to Article 72.4.1 of the Code and we can assume that Gray's name was created based on a single specimen, namely the holotype (NHMUK 1946.8.11.16) by monotypy.
While the plate depicting L. spinosus was already published in 1843, the publication of the text volume was delayed for another ten years, and finally Jacquinot and Guichenot (1853) stated the following in their account: "mais dont plusieurs d'entre elles [species] cependant étaient entièrement nouvelles pour la science, à l'époque ou elles ont été déposées dans les collections du Muséum de Paris (fevrier 1841)" [but several of them, however, were entirely new to science, at the time when they were deposited in the collections of the Paris Museum (February 1841)]. As the plate of Lophyrus spinosus had been published in 1843 and the specimen as well as the notes of Hombron and Jacquinot were most certainly available already in 1841, Duméril and Duméril (1851) gave nomenclatural priority to Hombron and Jacquinot (1843) and treated Bronchocela marmorata Gray, 1845 as a junior subjective synonym. This nomenclatural action has been wrongly considered as a nomen substitutum pro Bronchocela marmorata by subsequent herpetologists (see for example Wermuth 1967;Hallermann 2005).
The name-bearing type of Lophyrus spinosus is the specimen collected by Hombron and Jacquinot, MNHN 6896, by original designation as "type" and the holotype of Bronchocela marmorata must be considered a paratype as Duméril and Duméril (1851) refer to this specimen when creating the name. Thus these two scientific names do not share name-bearing types. However, by considering the two holotypes as belonging to the same species (Duméril and Duméril 1851), in this case Lophyrus spinosus Duméril & Duméril, 1851 effectively becomes a junior subjective synonym of Bronchocela marmorata Gray, 1845 on grounds of priority as Gray's description was published first. This view has been the opinion of generations of herpetologists for over 150 years. The main reason for this is probably that the type specimen of Lophyrus spinosus was purportedly from Zamboanga, Mindanao, Philippines where Bronchocela marmorata was thought to occur. Already Taylor (1922) questioned whether Bronchocela marmorata actually inhabits Mindanao and he did not include L. spinosus in his synonymy of Bronchocela marmorata. No herpetologist appears to have looked at the actual specimen or the illustration on which the description of L. spinosus was based.
Only Hallermann (2005) in his review of the genus Bronchocela Kaup, 1827 doubted that Lophyrus spinosus Duméril & Duméril, 1851 is conspecific with Bronchocela marmorata Gray, 1845 and stated that Lophyrus spinosus "is in fact a Gonocephalus sophiae (Gray, 1845)". Unfortunately Hallermann (2005) did not communicate how he arrived at his conclusion or provided evidence that would corroborate his identification of the spinosus holotype as Gonocephalus sophiae. His main reasons for the identification were actually based on an examination of the holotype of L. spinosus. The specimen was clearly not a member of the genus Bronchocela but had the general appearance of an anglehead lizard of the genus Gonocephalus Kaup, 1825. Additionally he was misled by the assumption that the type locality of L. spinosus was on Mindanao, Philippines. This combined only left one conclusion, namely to consider the specimen as Gonocephalus sophiae.
Recently, one of us (WD) came across the original plate ( Figure 1) that served as the template for the colour description of Lophyrus spinosus. We re-evaluated the original description by Duméril and Duméril (1851) and compared it to the image it was referring to and conclude that the specimen described by Duméril and Duméril (1851) is neither Bronchocela marmorata nor Gonocephalus sophiae, but has rather the appearance of a New Guinean anglehead lizard, namely Hypsilurus auritus (Meyer, 1874).
In order to evaluate the taxonomic status of Lophyrus spinosus Duméril & Duméril, 1851 we studied museum specimens including the type specimens of Gonocephalus sophiae and Hypsilurus auritus and compared the respective data to those of the holotype of Lophyrus spinosus.

Material and methods
Meristic and morphometric data were recorded from type specimens and additional material mentioned in the text. Measurements were taken using a sliding calliper with a precision of 0.1 mm or using a ruler with a precision of 1 mm. Abbreviations used are as follows: SVL: snout-vent length; TL: tail length; HL: head length; HW: head width, d tymp : diameter of the tympanum, d eye : diameter of the eye.

Results
In his review of the genus Bronchocela Hallermann (2005) stated that Lophyrus spinosus Duméril & Duméril, 1851 does not represent Bronchocela marmorata Gray, 1845, but that it is conspecific with Gonocephalus sophiae (Gray, 1845). However, a comparison of the morphometric data reported by Boulenger (1885) for the syntypes of Gonocephalus sophiae (NHMUK 1946.8.27.10-15, incl. sophiae and syntypes of its synonym petersi) and the holotype of Lophyrus spinosus (MNHN 6896, data from Brygoo (1988) and our measurements) already rules out this identification. G. sophiae hardly ever grows larger than a total length of 350 mm. Boulenger (1885) gives measurements for the largest specimen as follows: total length 326 mm, head 33 mm, body 78 mm and tail 215 mm. Not only is G. sophiae clearly smaller and differing from the measurements of the holotype of L. spinosus (total length 565 mm, head 35 mm, body 105 mm and tail 425 mm [our measurements]; Brygoo (1988) reports 575 mm total length and 420 as the tail length), the species also differ greatly in the ratio of tail length to snoutvent length, i.e. G. sophiae TL/SVL = 1.94 (largest male specimen of the syntype series) and L. spinosus TL/SVL = 3.04 (male holotype).
In their original description of Lophyrus spinosus Duméril and Duméril (1851) stated the following (our translation) "On the temple, before the enlarged scales situated behind the eye, there exists an elliptical plane, ... of which the anterior part is [surrounds] the tympanum. It [the ellipse] is bordered by a double row of scales that are considerably larger [than the surrounding scales] and approximately the same size as the postocular scales... The general colour is grey on the underside and reddish on the back and the flanks; the sides of the head and neck are near brick-red, as well as the broad rings on the tail, in alternation with irregular grey rings; the fingers and the edge of the throat fan are green-yellowish". The colour description clearly refers to Plate 3 in Hombron and Jacquinot (1843) [see Figure 1].
More importantly there is one specific character given in the description of Duméril and Duméril (1851) that does neither fit Bronchocela marmorata nor Gonocephalus sophiae, i. e. the elliptical ring surrounding the tympanic area. As far as we are aware there is only one species of agamid lizard that shows this character, namely Hypsilurus auritus (Meyer, 1874). Meyer (1874) described Gonyocephalus (Hypsilurus) auritus based on five syntypes, that were deposited in the collection of the Dresden Museum (MTKD 398, 400-402) and in Berlin (ZMB 8782). Three of the original specimens (MTKD 398, 401-2) were destroyed during World War II. One specimen (MTKD 400) had been exchanged with a museum that could not be identified as the handwriting on the note was illegible and the specimen was therefore presumed lost by Manthey and Denzer (2006). Consequently these authors considered the only remaining specimen (ZMB 8782) as the holotype. However, this is not a valid nomenclatural act according to the Code as the name was based on a series of syntypes and this fact cannot be changed after the original publication unless a lectotype is designated. The label of the specimen (ZMB 8782) stated that it had been collected on Jobi [Yapen Island, West Papua, Indonesia] which was accordingly cited as the type locality by Manthey and Denzer (2006). Later Franzen and Glaw (2007) discovered a specimen (ZSM 187/1913) in the Munich collection that had been collected by Meyer in "Doré, westl. Neuguinea" [Manokwari, Indonesia] and was still bearing the original MTKD number "400". A re-inspection of the handwritten catalogue of the MTKD revealed that the note reads: "abgegeben an Zool. Mus. des bay. Staates im Tausch" (exchanged with Zool. Mus. of the Bavarian State), nowadays Zoologische Staatssammlung München. Under these circumstances the potential "holotype" status of the Berlin specimen cannot be held up and ZMB 8782 as well as ZSM 187/1913 have to be considered as syntypes. Additionally to the type locality "Yapen Island" given by Manthey and Denzer (2006) must be added the origin of the Munich specimen collected on mainland New Guinea. Both Yapen Island and Manokwari lie within the Geelvink Bay (Cenderawasih Bay) area and consequently constitute the type localities (symprotonymotope) for Hypsilurus auritus. The type locality encompasses both places, Yapen Island and Manokwari in Cenderawasih Bay as long as no lectotype is designated (see Frétey et al. 2018 for different categories of type localities, their precision and restriction).
Meyer's (1874) original description was rather short and read as follows (our translation): "Related to the preceding species [Gonyocephalus (Hypsilurus) binotatus], but much smaller. Crest hardly interrupted, low. A black mark on the sides of the head that encircles the ear. No large plates below the tympanum. Gular pouch large, covered with small keeled scales." The reference to Hypsilurus binotatus was presumably made because of the dark coloured mark on ei-ther side of the neck. Otherwise the similarity between these two species is rather low.
Further and more detailed descriptions of Hypsilurus auritus can be found in Peters and Doria (1878, in Italian, as Gonyocephalus [Arua] auritus) and in Denzer (2006, 2016). Manthey and Denzer (2006) described the elliptical ring around the tympanum as consisting of slightly enlarged scales and used this character to differentiate between Hypsilurus auritus and all other species of the genus. The aforementioned descriptions by Peters and Doria (1878) and Denzer (2006, 2016) agree well with the description of Lophyrus spinosus given by Duméril and Duméril (1851) and could be referred equally to both Lophyrus spinosus and Gonyocephalus (Arua) auritus.
Consequently, we compared the description and illustration as well as the holotype of Lophyrus spinosus (MNHN 6896, male) (Figure 2) to the syntypes of H. auritus (ZMB 8782, male; ZSM 187/1913, sex undetermined) ( Figure 3) in order to clarify the taxonomic status of Lophyrus spinosus. Although all three specimens are similar in their general appearance and in having an ellipsoid of enlarged scales around the tympanum we found differences with respect to several characters (see Table 1).
The morphometric data of L. spinosus also differ from those that have been reported for H. auritus specimens. Urban (1977, unpubl. PhD thesis) examined 45 auritus specimens (27 males, 18 females) and recorded a maximum SVL of 130 mm (vs. 140 mm in spinosus) and a maximum TL of 390 mm (vs. 425 mm in spinosus). The tympanum / eye ratio in auritus was given as 1 (vs. smaller than 1 [0.6-0.76] in spinosus: right side d tymp 5.0 mm, d eye 6.6 mm / ratio ~ 0.76, left side d tymp 4.5 mm, d eye 7.5 mm / ratio = 0.6). With respect to auritus Urban (1977) characterized the shape of nuchal crest scales as triangular and equal in size to that of the diameter of the eye (vs. lanceolate and larger than the diameter of the eye in spinosus).
Furthermore the colouration of the L. spinosus holotype is different from that of the two known type specimens of H. auritus (see Figures 2, 3). This may be partly a result of preservation, but -if considered in conjunction with the colour description by Duméril and Duméril (1851) and the accompanying illustration by Hombron and Jacquinot (1843) -the type specimens of H. auritus will probably not have displayed a colouration similar to that of L. spinosus. All live specimens of H. auritus known so far have been described as "dorsally predominantly green with a brown or turquoise shadings" (see Manthey and Denzer 2016). This shadings or mottling can still be seen in both syntypes of H. auritus (Figure 3) but is not observed in   the holotype of L. spinosus nor shown on the original illustration of Hombron and Jacquinot (1843). Although the illustration of L. spinosus is pretty detailed and clearly depicting the actual type there is one character that could not be verified and may be down to the artist's impression; the illustrated specimen has nuchal and dorsal crest continuous, however the type does not. We examined the type for a possible loss of crest scales on the neck and concluded that no scales are missing and therefore nuchal and dorsal crest have to be considered as interrupted.
As a result of our research we remove Lophyrus spinosus Duméril & Duméril, 1851 from its synonymy with Bronchocela marmorata Gray, 1845. Furthermore we resurrect L. spinosus to full species status to which the name Hypsilurus spinosus (Duméril & Duméril, 1851) should be applied. Based on our comparisons with respect to the type material we conclude that Hypsilurus spinosus and Hypsilurus auritus (Meyer, 1874) are not conspecific and should be considered as distinct species Given the "nomenclatural rollercoaster" Lophyrus spinosus went through, we present a short list of synonyms and chresonyms as follows:

Discussion
Hypsilurus spinosus belongs to the H. nigrigularis species group of as defined by Manthey and Denzer (2006). All members of this group (H. nigrigularis [Meyer, 1874], H. geelvinkianus [Peters & Doria, 1878], H. auritus and H. spinosus) are only known to be distributed in the western part of New Guinea as well as adjacent islands (and some questionable localities on the eastern part of New Guinea; see below). The type locality of Lophyrus (now Hypsilurus) spinosus (Zamboanga, Mindanao, Philippines) reported by Duméril and Duméril (1851) is clearly erroneous as the genus Hypsilurus is not distributed further west than some offshore islands of New Guinea and Waigeo. The type specimen of H. spinosus therefore must have been collected on New Guinea or an adjacent island.
The L'Astrolabe and La Zélée expedition sailed along the New Guinean coast three times. The first approach was from Ambon Island towards the southwestern coast of New Guinea but they did not land and proceeded to the north-western coast of Australia. Their second approach coming from Australia past Aru Island was in April 1839 as detailed in Dumont d'Urville (1844: 108-145). The expedition visited Dobo on Aru during 12 th -21 th April 1839, subsequently proceeding to the southwestern coast of New Guinea. From 24 th April until 30 th April 1839 they anchored in Dubus Harbour (Havre Dubus à la baie de Triton). Nowadays this area is called Kaimana, West Papua.
The expedition's third and last approach was from the Louisiade Archipelago near the eastern most tip of New Guinea in late May / early June 1840 (see Dumont d'Urville 1846: 205-242). They intended to anchor at Orangerie Bay and Hood Point (both on New Guinea) but decided against it. On 31 st May they visited Darnley (Erub, Arruob), an island in the Torres Street but did not collect natural history specimens. On 1 st June 1840 L'Astrolabe and the accompanying La Zélée went aground near Tudu Island in the Torres Street. It took them 10 days to recover the ships. Visits were made to Tudu Island but the island is described as having no water, mostly sandy soil with grass and some scrubs as well as a small patch with trees and palms on the northern end of the island. The lack of water and the vegetation would most certainly not support an arboreal lizard population of a Hypsilurus species that is typically found in the rainforest. This chain of events renders it highly unlikely that the type was collected during their third approach.
Consequently, Hombron and Jacquinot could have collected the specimen only on New Guinea or perhaps on Aru Island. Aru (and Kei) have been well sampled by Beccari in 1873 (see Doria 1874;Peters and Doria 1878) and later in 1908 by Roux (1910) 1999). Triton Bay is only approx. 75 km away from the Cenderawasih Bay (the combined type locality of the auritus syntypes) but on the southern coast of New Guinea.
The Naturalis collection holds two specimens (ZMA. RENA.18893) that were collected at Etna Bay, merely 25 km away from our assumed collection locality for H. spinosus. These specimens have three scales between the nasal and first supralabial, a character found in the holotype of H. spinosus. Else they rather agree with the characters described above for H. auritus, e. g. the development of the nuchal and dorsal crest. Without additional material it is impossible to determine whether the differences we found between the type specimens of auritus and spinosus are sufficient to define each taxon or whether they represent clinal variations of phenotypic characters within the same species as the Etna Bay specimens would indicate. Until further (topotypic) specimens of spinosus become available we propose to consider both Hypsilurus spinosus and H. auritus as species in their own right.
The distributional ranges of the two species cannot be defined currently. Meyer (1874) did not cite a type locality and only gave New Guinea as the general distribution of H. auritus. In a later paper Meyer (1886) cites Doré, Passim, and Rubi as localities where he collected H. auritus. The locality for the type specimen ZMB 8782 is given in the handwritten catalogue of as Jobi [Ansus on Yapen Island]. All these localities lie in north-western New Guinea along the Geelvink (Cenderawasih) Bay that therefore could serve as the type locality as detailed above. A further restriction is, however, impossible. Manthey and Denzer (2006,2016) reported Hypsilurus auritus from Misol Island, West Papua (including Vogelkop [Birdshead] Peninsula) and Papua province, Indonesia. Records further east (in Papua New Guinea) were considered doubtful. Some of the specimens that have been reported in earlier publications as H. auritus may belong to H. spinosus instead. Therefore it will be necessary to examine collection specimens and compare them to the characters given above for each species.
Additionally, further fieldwork will help to clarify not only the distribution but also the taxonomic status of different populations. It may well be the case that there exist additional undescribed species that are morphologically similar to H. auritus and H. spinosus.