Bryanites graeffii sp. n. is described from Samoa based on a single male specimen collected between 1862–1870 that was recently discovered in the Muséum national d’Histoire naturelle, Paris. Cladistic analysis based on 127 morphological characters from 49 exemplars of the carabid beetle tribe Platynini in the Austral-Pacific region, places the new species as adelphotaxon to Bryanites samoaensis Valentine, type species of the genus Bryanites Valentine, 1987. Bryanites comprises, along with Vitagonum Moore, 1998 of Fiji and Ctenognathus Fairmaire, 1843 of New Zealand, a clade that diverged early in the evolutionary history of Pacific platynine Carabidae. Bryanites graeffii exhibits very large body size among taxa of Platynini—16.2 mm standardized body length—with the genus characterized by vestigial flight wings and metathoracic apomorphies that are associated with flight-wing loss. Along with Blackburnia Sharp, 1878 of Hawaii, the origins of Bryanites, Vitagonum, and Ctenognathus are hypothesized to date to the Miocene, with their radiations beginning long before the origins of the geographically widespread, flight-capable species of Metacolpodes Jeannel, 1948 that colonized numerous island systems across the western Pacific. Given the numerous platynine taxa collected by extensive biotic surveys of Samoa during the first quarter of the 20^th Century, the absence of any specimens of B. graeffii since the initial collection of the unique holotype prior to 1871 suggests that this species may be extinct. Such extirpation of large platynine carabid beetles has also been documented for Hawaii, where the time of extinction of seven Blackburnia species represented only by subfossil fragments coincides with the time of human colonization and attendant introduction of the Pacific rat, Rattus exulans (Peale).

anthropogenic extinction, biogeography, brachyptery, Polynesia

It is undeniable that natural history museums represent invaluable and irreplaceable archives of biological diversity on Earth. Firstly these institutions serve as repositories for many millions of studied and described specimens: i.e. type specimens and associated subsequent collections of named species. Secondly museums also hold uncounted, unstudied specimens that have been collected and processed, but never adequately examined or understood by a taxonomic specialist. Predictably, examination of unidentified material collected long ago can produce surprising results, especially when the historical museum specimens represent the only evidence we have for existence of that particular species. For example, two Samoan species of flying fox, genus Pteropus Brisson, 1762 (Chiroptera: Pteropodidae), are known only from museum specimens collected between 1839 and 1856 (Helgen et al. 2009). In this contribution, a Samoan carabid beetle specimen collected prior to 1871, and subsequently deposited in the Paris Natural History Museum, is shown to represent the only specimen known of a new species. Moreover, the genus to which the new species is assigned is known previously only from two species based on two specimens collected in 1924 (Valentine 1987). That the new species is being described only now, over 140 years after its collection from nature, attests to the contextual information that is required to interpret biodiversity, for it is not the simple naming of organisms that occupies systematists, but more importantly, the organized naming and placing in context of all organisms. A scientific name presented in isolation from any detailed taxonomic framework is nearly useless, being suspect as a natural entity, devoid of biological context provided by knowledge of its sister group (Hennig 1966), and at best representing a single datum indexing a gross estimate of biodiversity.

The specimen described in this contribution is a member of the carabid beetle tribe Platynini belonging to the genus Bryanites Valentine, 1987 (Valentine 1987). This genus was previously known only from specimens collected by E. H. Bryan, Jr., 23-v-1924, at Salaʽilua, Savaiʽi Island, Samoa. Though the holotype specimens of Bryanites samoaensis Valentine and B. barri Valentine are assignable to the Platynini, they are aberrant representatives of that tribe. In contrast to many platynine carabid beetles, which as the name suggests are flattened dorsally, the bodies of Bryanites beetles are fusiform with a broad pronotum, their bodies convex dorsally. The two Bryanites spp. also exhibit patterns of setation on the elytra and pronotum unusual among platynine taxa. H. E. Andrewes, formerly at the Natural History Museum, London and a worldwide authority on Carabidae, placed determination labels on both specimens subsequently described by Valentine that read “? Pterostichini gen. nov.” He did this presumably because the specimens appear superficially like the larger-bodied, more robust beetles of that tribe, though their lack of a critical diagnostic feature of most Pterostichini—i.e. an externally visible plica, or fold, along the apical elytral margin—fits members of the tribe Platynini. Andrewes’ tentative, incorrect guess as to tribal membership is also understandable because the two Bryanites beetles appear nothing like the other Samoan species of Platynini that he had previously described (Andrewes 1927).

This paper first traces the provenance of the newly described Samoan beetle specimen from its collection in nature to the present day. We cannot be certain why the specimen avoided description until now. But the documented timing of expeditions and subsequent taxonomic publications coupled with the temporal connections of the various taxonomists, and how the beetle could have passed from hand to hand, make a strong case for how a large carabid beetle from a tropical island could be ignored for well over a century to be rediscovered within a natural history museum. In order to be certain how best to classify the new species, cladistic analysis using morphological characters is used to phylogenetically place the species. After adding the new species to a little-known clade of Samoan Carabidae, the question regarding the fate of this lineage in nature is investigated, with parallels from the platynine carabid beetle fauna of the Hawaiian Islands suggesting a likely scenario for this species leading from the 19th to the 21st Century.

Taxonomic material. Specimens treated in this contribution were borrowed from the Muséum national d’Histoire naturelle, Paris (MNHN), Mdm. Helène Perrin and Dr. Thierry Deuve curators, and the B. P. Bishop Museum, Honolulu (BPBM), James Boone collection manager. Placement of the new species within the context of platynine phylogeny (Suppl. material 1) utilized an updated version of the character matrix previously presented in Liebherr (2005), with that paper citing institutional sources that provided specimens for initial matrix development. Taxa included in the analysis are distributed in the southwest Pacific region, including Australia and Asia, northeastward through Melanesia and Polynesia to the Hawaiian Islands.

Several taxa were added to complement the previous analysis. Colpodes kanak Fauvel (Fauvel 1903) of New Caledonia was added to the analysis based on examination of a female (MNHN) determined by Fauvel in 1906.Vitagonum apterum Moore of Fiji was added because it, like Bryanites, represents an aberrant, taxonomically isolated platynine lineage in the Pacific (Moore 1998). The male holotype of Bryanites samoaensis serves to represent the genus in the analysis. The female holotype of B. barri Valentine was also studied, and external anatomy of prothorax and elytra is consistent with Valentine’s (1987) taxonomic decision placing both species together in the genus. However the type specimen of B. barri is heavily damaged. The beetle was extensively disarticulated during description, demonstrated by Valentine’s illustrations of an isolated mentum, maxilla, and labium (Valentine 1987: Fig. 3). The resulting disarticulated structures were then mounted on two acetate sheets using Canada balsam as a fixative. Several isolated flecks of mounting medium remain on the sheet without any associated sclerite, suggesting strongly that flexure of the acetate caused the balsam to pop off the acetate. This interpretation explains the absence of the head capsule, mandibles, and antennae from the mounting cards. The female ovipositor is also damaged and the internal reproductive tract not present. Given this destruction of the specimen, 26 characters of the female reproductive tract and 11 characters of the head cannot be scored. Coupling that loss with the inability to score the 21 characters of the male reproductive system from the female holotype leads to at least 58 of the 127 characters necessarily being coded as missing. The poor mounting condition of the remaining sclerites cemented to the specimen cards limits confidence in assessing microsculpture covered by the Balsam, adding further uncertainty to the assessment of character states. This lack of information precludes a precise placement of this taxon in the analysis, at the same time seriously reducing resolution in a strict consensus cladogram when the taxon is included in the analysis. As B. samoaensis is the type species of Bryanites, it was decided to limit the focus of the analysis to the question of where the taxon herein classified as B. graeffii fits phylogenetically relative to an array of fully informative exemplar specimens. As we have no evidence to overturn Valentine’s (1987) decision to place B. barri in Bryanites, that decision is allowed to stand.

Laboratory Techniques. Dissection protocols used throughout development of the present character matrix are detailed in Liebherr (2015: 18–20). An ocular ratio is used to quantify eye size: the maximum width of head across eyes divided by the minimum breadth of frons between eyes. Male genitalia are preserved in polyethylene vials placed on the specimen pin. Standardized body length comprises the sum of three measurements: 1, head length measured from the medioanterior margin of the labrum to the cervical ridge; 2, median pronotal length; and 3, elytral length measured from base of elevated scutellar apex to apex of longer elytron adjacent to the suture.

Character data. The cladistic analysis was based on 127 characters, 26 scored from the female reproductive tract and gonocoxae (i.e. ovipositor), 21 scored from male genitalic structures, and 80 derived from external anatomical structures (Suppl. material 2). Six of the characters are autapomorphic for a single terminal, however these were retained as other taxa could not be scored for these characters, due either to missing female or damaged specimens. Given additional material these characters could become potentially synapomorphous.

Cladistic methods. The character matrix was developed in WinClada (Nixon 2002) (Suppl. material 2), and analyzed by Nona (Goloboff 1999) using the ratchet (Nixon 1999). Results from an initial analysis using 200 iterations of the ratchet were compared to subsequent analyses using 1000 ratchet iterations, and then 10,000 ratchet iterations. After each ratchet run all trees were hard collapsed—i.e. collapsing nodes on trees when they are not supported under all optimizations—and then non-optimum (longer) trees were deleted. The strict consensus was then generated from those remaining hard-collapsed trees. Identical results at the 200, 1000 and 10,000 ratchet levels permitted the conclusion that all most parsimonious trees had been discovered.

Provenance. The specimen rediscovered in the Paris Museum (Fig. 1) bears three labels (Fig. 2). The collector/locality label specifies that the beetle was collected in Samoa by “Dr. Graffe.” A second label reads “8284”, and a third label notes accession by the Paris Museum. “Dr. Graffe” is no doubt Eduard Gräffe, or Graeffe, a Swiss zoologist and naturalist who lived in Samoa from 18621870 (Graeffe 1917, Clunie and Snow 1986). Graeffe was employed by Johann Cesar VI Godeffroy, a wealthy shipping magnate in Hamburg who founded the Museum Godeffroy. The Museum Godeffroy published a journal, within which Graeffe published treatises on Samoan geography and birds (Graeffe 1873a, 1873b). He also published on the invertebrates including the insects of Fiji (Graeffe 1866). He maintained contact with scientists in Europe, including the ornithologists Otto Finsch and Gustav Hartlaub, to whom he sent a manuscript describing the birds of Tonga (Graeffe 1870). After Graeffe’s return to Hamburg, Leon Fairmaire described new species and redescribed numerous previously described beetle species from Australia, Fiji, Tonga, and Samoa (Fairmaire 1879a), many deposited in Museum Godeffroy. Fairmaire’s (1879a) paper documents a connection between Graeffe and Fairmaire via the Museum Godeffroy, and supports the transfer of the specimen described herein to Fairmaire, either by loan or by purchase, after 1870 but prior to 1879. Fairmaire described or revised other carabid beetles from Fiji and Tonga (Fairmaire 1878, 1879b, 1881a, 1881b), though no more from Samoa. When his collection passed to the Paris Museum upon his death in 1906, the “MUSEUM PARIS” accession label was placed on the specimens (e.g. Fig. 2), with all of his carabid specimens maintained together in boxes as the “Fairmaire Collection.” When Erwin and Erwin (1971) assessed the condition of the Paris Museum carabid beetle holdings, they cited shelving unit B13, shelves 12; “23 boxes of the Fairmaire Collection, misc. carabids in very poor condition with no or poor labels, ‘a mess’! (p. 10).” It is within this set of boxes that the specimen described below was found.

Figure 1. 

Male holotype, Bryanites graeffii, dorsal view.

Figure 2. 

Specimen labels associated with holotype specimen of Bryanites graeffii.

Cladistic Analysis. Subjecting the taxon-character matrix to Nona (Goloboff 1999) within the WinClada shell (Nixon 2002) resulted in 47 hard-collapsed, multiple equally parsimonious trees (MEPT) of 566 step-length under 200, 1000, or 10,000 iterations of the ratchet (Nixon 1999). These trees were constrained to be fully resolved under all character optimizations; i.e. fast or slow optimization. The strict consensus cladogram of 607 steps (Fig. 3) was rooted so that the species of Lorostema Motschulsky, 1865 comprise outgroups to the other Pacific platynine taxa. This decision was based on plesiomorphic presence of two testes in male beetles of Lorostema spp. versus the derived monorchid, or single-testis configuration (character 46, Suppl. material 2) in males of species of Blackburnia Sharp, 1878, Notagonum Darlington, 1952, Colpodes W. S. MacLeay, 1825, and Metacolpodes Jeannel, 1948 (Will et al. 2005). The cladistic relationships indicate that the newly described species should be combined as Bryanites graeffii sp. n. (Fig. 3). The two Bryanites examples are members of a clade, under all possible MEPTs, that includes Vitagonum apterum Moore of Fiji as adelphotaxon. These two genera are in turn sister group to Ctenognathus Fairmaire, 1843 of New Zealand.

Figure 3. 

Strict consensus cladogram of 49 taxa of Pacific platynine carabid beetles. Type species, where included, indicated by asterisks. See text for further explanation of cladistic analysis. Distributional areas of species include: As, Asia; Au, Australia; F, Fiji; HI, Hawaiian Islands; I, India including Sri Lanka; Jp, Japan; Jv, Java; NC, New Caledonia; NG, New Guinea; NZ, New Zealand; Ph, Philippine Islands; R, Rapa; Sa, Samoa; So, Solomon Islands; Sunda, Sunda Islands; T, Tahiti and Society Islands; V, Vanuatu.

Other relationships inherent in the tree are very similar to those discussed in Liebherr (2005, fig. 78), although the inclusion of the two Bryanites spp. plus V. apterum in this analysis results in the Ctenognathus spp. joining with those taxa much closer to the root node the cladogram (Fig. 3). Also, Notagonum kanak (Fauvel) comb. n. is placed such that it must be removed from the genus Colpodes and newly combined with Notagonum.

Taxonomic challenges remain regarding monophyletic classification of species in the genera Colpodes and Notagonum, and these must be addressed via a more comprehensive and informative phylogenetic analysis, such as those that include DNA-molecular characters (e.g. Maddison 2012). Colpodes is certainly polyphyletic. The genus is based on Colpodes brunneus (W. S. MacLeay), placed here as sister species to a second Javan species, C. latus Louwerens (Fig. 3). A third Javan species, C. brittoni Louwerens, is also closely related, with beetles of all three species exhibiting exceedingly protruded compound eyes and very broad pronotal lateral margins (Liebherr 1998). Other species combined with Colpodes in this analysis—from Fiji and Tahiti (Fig. 3)—must be recombined with different generic names to accommodate the monophyletic taxa Helluocolpodes Liebherr, 2005 and Metacolpodes. Notagonum has been a genus of convenience since Darlington (1952) proposed it. In his words, “It must be admitted that it is hard to draw a line between Colpodes in my partly restricted sense and some of the forms which I am including in Notagonum, but I am convinced that when Colpodes is broken up the various species of Notagonum will properly form at least one and perhaps more separate genera (Darlington 1952: 129).”

The results of the cladistic analysis succeed in the goal of placing the new species as a member of a Samoan lineage, Bryanites, that exhibits a close biogeographic relationship to the Fijian relict, Vitagonum (Moore 1998). This clade, and its sister group Ctenognathus of New Zealand, diverged early in the history of Pacific Platynini. Previously, Liebherr (2005) concluded that the Hawaiian genus Blackburnia colonized the Hawaiian Island chain long before the origins of the present high islands, and perhaps as long ago as 28 Mya when Kure became the first in a consistently present subaerial chain of islands generated by the Hawaiian Island volcanic hotspot (Duncan and Clague 1985). In the Samoan Island chain, the Alexa Bank seamount is hypothesized to have originated as a subaerial island 22 Ma (McDougall 2010), setting the earliest date possible for colonization of the Samoan Islands hotspot volcanic chain by a Bryanites ancestor. Subaerial origin of the Fijian archipelago is dated to a similar time—22–25 Ma (Gill and McDougall 1973, Whelan et al. 1985)—establishing a maximal time of origin for the ancestor of Vitagonum. And volcanism along the Kermadec-Lau-Tonga Ridge system would have enhanced colonization prospects for the ancestor of New Zealand Ctenognathus 10—15 Mya. Thus all area relationships among the early diverging Pacific platynine lineages, including Bryanites, occupy continental areas or island chains available for colonization in the Miocene. This supports Valentine’s (1987) hypothesis that Bryanites is a relict Samoan lineage.

The question of whether any or all of the Bryanites species are extinct is necessarily open ended. Nevertheless, examination of the circumstances of their collections taken within the context of other biological surveys of Samoa paint a grim picture as to the possibility of their continued existence in nature. The single specimen of Bryanites graeffii was collected prior to 1871 by a naturalist principally interested in birds and marine invertebrates (Graeffe 1917). Nonetheless, we know that he collected insects during his trips to Fiji (Graeffe 1868), and during 1869 near Apia (Graeffe 1917). But only the one specimen of B. graeffii described herein is known to have passed to Leon Fairmaire for eventual deposition in the Paris Museum. The British Museum survey of 1924–1925 (Kami and Miller 1998), summarized for Carabidae by Andrewes (1927), led to description of five other platynine carabids— Colpodes buxtoni Andrewes, C. pacificus Andrewes, C. hopkinsi Andrewes, C. piceus Andrewes, and C. anomalus Andrewes, the first three of which are now assigned to Metacolpodes (Fig. 3, Suppl. material 1; Liebherr 2005). Andrewes (1927) also examined previously collected taxonomic material, including that from 1905 collections for the Vienna Museum (Rechinger 1914) and 1912–1913 collections for the Berlin Museum (Friedrichs 1914).

The Bishop Museum’s Whitney South Seas Expedition of 1924, including E. H. Bryan, Jr. as entomologist, also visited Samoa (Evenhuis 2007). Bryan’s field notes for 23 May 1924–the date labeled on the two specimens later described as Bryanites (Valentine 1987)–included: “Put away large quantity of specimens collected yesterday (Evenhuis 2007: 125).” Some of those “yesterday’s” specimens were collected when he “Dug several things from partly rotten bank [sic bark] of large tree:- centipedes, termites, sowbugs, tree cricket and moderately large slender fuscous Carabid (Evenhuis 2007: 125).” Given that the description of a “moderately large slender fuscous Carabid” on Samoa can only fit a Bryanites beetle, it appears that Bryanites, either the B. barri or B. samoaensis individual that Bryan collected, was associated with a downed log covered with rotten bark. Bryan described the situation within which he was collecting at that camp as “Two large Ac[h]atinella [sic] landshells on Orange tree in banana patch at about 2200’. [Met] Mr. Beck and returned to hut. Mosquitoes not bad, but rats after our provisions. Very good collecting at this elevation (Evenhuis 2007: 125).” Being a field biologist based on Oahu, Bryan no doubt mistook a native Samoan Samoana or Eua land snail (Pilsbry 1909–1910, Cowie 1992) for the precinctive Oahu genus Achatinella, but his sighting of native land snails in proximity to rats in 1924 suggests that the native elements did not have long to exist at the site given the presence of invasive rats. There is both indirect and direct evidence that rats prey on native brachypterous carabid beetles. Deposits in Makauwahi Cave on Kauai include abundant subfossil fragments of extinct, brachypterous Blackburnia carabid beetles in stratigraphic layers deposited prior to human colonization and the associated introduction of the Pacific rat, Rattus exulans (Peale) (Liebherr and Porch 2015). Subfossil fragments of those presently extinct taxa are absent from the stratigraphic column in deposits laid down from shortly after the time of human colonization up to the present. More directly, stomach contents of alien black rats (Rattus rattus L.) have been shown to include fragments of a native Hawaiian Blackburnia beetle species, as well as native katydids, weevils and spiders (Shiels et al. 2013). The presence of predatory, climbing, night active rats is uniformly deleterious to native communities on tropical islands (Harper and Bunbury 2015), with Samoa not a likely exception to this rule.

Given these various lines of data, it seems unlikely that Bryanites beetles currently exist in nature. Only through the perspicuous collecting by E. H. Bryan, Jr. and the passing of his two specimens to J. Manson Valentine could the first two species of this genus be described (Valentine 1987). The subsequent discovery of a single additional specimen collected over 140 years ago that traveled a circuitous route from Samoa, to the Museum Godeffroy, to the Paris Museum adds another entry point to study of the Bryanites radiation. Are we better for knowing about this evolutionary story even after all known players have likely left the scene? To the degree that the hotspot volcanic chain of Samoa supported a very unique radiation of platynine Carabidae, as did Hawaii, we have learned that the Blackburnia radiation of Hawaii had an analog in Samoa, with Samoan species inhabiting the islands of Savaiʽi and Upolu. We have characterized a large-bodied, distinctive species that may turn up in Samoan subfossil deposits, much like the large-bodied subfossil Blackburnia spp. of Kauai (Liebherr and Porch 2015). And should these or any other Bryanites beetles have survived forest conversion, and the plagues of rats and other invasive species so that they may be collected during a future biological survey, that sample will be connected to specimens held in two other biodiversity hotspots on Earth.

The curators at the Bishop Museum, Honolulu, and the Muséum national d’Histoire naturelle, Paris have graciously let me work in their collections over the years with free access to specimens and with a minimum of distractions. I thank the following for their support of my curatorial and collection-based research efforts: James Boone, Neal Evenhuis, Scott E. Miller, Gordon Nishida, G. Allan Samuelson (Bishop Museum); Thierry Deuve, Jean Menier, Helène Perrin, Azadeh Tagavian (Paris Museum). I thank Martin Baehr, Zoologische Staatssammlung München, for sharing his findings regarding species identities in the Notagonum marginellum-submetallicum-complex. The underlying research on Hawaiian platynine beetles was supported by National Science Foundation grants DEB-9208269, DEB-9806349, and DEB-0315504.