We performed the first cladistic analysis of Metanephrops, the most diverse extant genus of clawed lobster, using a morphology-based data matrix of all 20 species of Metanephrops and 35 cladistically informative characters, all external hard parts. Unweighted cladistic analysis corroborates previous studies that indicated that homoplasy is rampant in the evolution of clawed lobsters. Only 5 of the 68 synapomorphies identified by the analysis are unambiguous, unreversed synapomorphies. Recent species of Metanephrops have traditionally been divided (non-cladistically) into four morphology-based groups. Cladistic analyses support the traditional, arafurensis, Atlantic/binghami, and japonicus groupings; these groups are monophyletic. The thomsoni group is not supported by the cladogram. The (two) oldest known fossil Metanephrops species occur in Late Cretaceous, shallow marine rocks of the eastern/Atlantic side of the Antarctic Peninsula. The stratigraphic and geographic occurrences of these fossil species are the basis for the previously hypothesized, Late Cretaceous, southern high latitude origin of Metanephrops. Cladistic results corroborate that Metanephrops originated in the southern high latitudes. The cladistically most plesiomorphic single species, the recent M. challengeri, and the next most plesiomorphic species, the Late Cretaceous M. rossensis, are both known from the high southern latitudes. Likewise, the most plesiomorphic species group, the binghami group, is best known from the high southern latitudes.
Introduction
Metanephrops Jenkins, 1972, is an extant, clawed lobster genus (Family Nephropidae) with a very distinctive, carinate and spiny cephalothorax (Fig. 1). It is the most diverse extant lobster genus, known by 17 extant species and 3 fossil species. The recent species are mostly Indo-Pacific, deepwater (continental slope and deeper) dwellers.
The genus Metanephrops was erected to accomodate 13 recent species previously referred to Nephrops Leach, 1814, as well as a new, Pliocene fossil species, Metanephrops motunauensis Jenkins, 1972. Since that time, 4 recent and 2 Late Cretaceous fossil species have been discovered. Holthuis (1974, p. 824-826) provided a good definition of the genus, one that we follow herein.
Metanephrops is not distinguishable on the basis of any one morphologic feature. It is, however, unique in exhibiting the combination of: 1) prominent, spinose, supraorbital carinae that extend posteriorly to near the postcervical groove, 2) a prominent, narrow, antennal carina, 3) usually three pairs of thoracic carinae [except M. neptunus (Bruce, 1965) with two pairs] (Fig. 2), and 4) symmetrical chelipeds that are generally long, slender, and sometimes strongly carinated.
Metanephrops bears strong morphologic similarities to both the recent Nephrops [although a close relationship is disputed by DNA evidence of Tam and Kornfield (1998) and Tshudy et al. (2005)] and fossil Paraclytia Fritsch, 1887 (Fig. 3). Metanephrops is most easily distinguished from Nephrops by its possession of three features: 1) the branchial carina (absent on Nephrops and one species of Metanephrops, M. neptunus), 2) prominent supraorbital carina (much shorter and lower on Nephrops), and 3) prominent antennal carina (much shorter and lower on Nephrops). Paraclytia is distinguished from Metanephrops and Nephrops in exhibiting a unique sculpture of the pleon terga and pleura, and in possessing a telson with submedian carinae converging (instead of diverging) posteriorly.
Several authors (De Man, 1916; Yaldwyn, 1954; Jenkins, 1972; Chan and Yu, 1987, 1991; Holthuis, 1991; Chan, 1997) have maintained that recent species of Metanephrops can be divided into four morphologic groups, summarized below (Table 1).
Recent Metanephrops are dwellers of the outer continental shelf and, especially, continental slope. Species have been collected from depths between 50 and 994 meters, but most commonly greater than 150 m (list in Tshudy, 2003). Recent species occur at latitudes ranging from about 35°N (Japan) to almost 50°S (New Zealand), and are distributed mainly along the western edges of the ocean basins, especially the Indo-West-Pacific. Not a single occurrence is known from the central or eastern Pacific nor from the central or eastern Atlantic (Fig. 4).
Little is known about the evolutionary or biogeographic history of Metanephrops, despite its fossil record ranging back to the Late Cretaceous. Only 3 fossil species are known: M. rossensis Feldmann et al., 1993 [Late Cretaceous (Campanian)], 2) M. jenkinsi Feldmann, 1989 [Late Cretaceous (Maastrichtian — Paleocene)], and 3) M. motunauensis Jenkins, 1972 (Pliocene). Several workers (Jenkins, 1972; Feldmann and Tshudy, 1989; Feldmann et al., 1993; Chan, 1997) have presented hypotheses on the evolution and dispersal history of the genus.
Jenkins (1972) concluded that Metanephrops evolved in the Indo-West-Pacific, where it has diversified, adding that “Metanephrops possibly reached the Atlantic around southern Africa, but more probably migrated through the Tethys prior to or during the Lower Miocene. The sea connection between the Indian and Atlantic Oceans through Tethys was lost in the late Lower Miocene” (p. 175).
Feldmann (1989) named the then-oldest known fossil species of Metanephrops, M. jenkinsi (Maastrichtian— Paleocene of Antarctica), but he did not comment on the history of the genus. The discovery of M. jenkinsi, however, provided the first material basis for refining Jenkins' hypothesis that the genus originated in the Indo-West-Pacific. Feldmann and Tshudy (1989) suggested that Metanephrops evolved in the high southern latitudes, in shallow water, and then radiated into lower latitude, deeper settings. They also noted some underlying morphological similarities between Metanephrops jenkinsi and Hoploparia stokesi (Weller, 1903), which co-occur in the Antarctic Late Cretaceous-Paleocene, and suggested that Hoploparia stokesi may have been the root stock from which Metanephrops evolved.
Feldmann, Tshudy and Thomson (1993) described a still-older fossil Metanephrops, M. rossensis, from the Campanian of the Antarctic Peninsula. Noting the strong similarity between M. rossensis and the recent species of Metanephrops, they concluded that M. rossensis “seems certain to have been ancestral to most or all modern species” (p. 18). They further noted that M. jenkinsi is clearly less similar to recent Metanephrops than is M. rossensis, and added that M. jenkinsi “may be an offshoot off the Metanephrops main lineage” (p. 18-19). Having now two shallow-water, Cretaceous species, they reiterated the onshore-offshore, and high latitude to low latitude, dispersal theme (p. 19).
They also reiterated the hypothesis that the closest known ancestor of Metanephrops is Hoploparia and further speculated that the ancestor of Metanephrops might be, or at least be close to, the Campanian-Maastrichtian Hoploparia antarctica Wilckens, 1907, of Antarctica and South America, as suggested by both the morphology and occurrence of H. antarctica. Unlike other, typical Hoploparia, but like Metanephrops, H. antarctica has thoracic carinae and keeled chelipeds. Hoploparia antarctica co-occurs with the oldest known Metanephrops, M. rossensis, in the Campanian portion of the Lachman Crags Member of the Santa Marta Formation on James Ross Island, Antarctic Peninsula (Feldmann et al., 1993).
Chan (1997) studied recent species extensively (and provided keys and geographic and bathymetric distribution maps for modern distributions) but, by an oversight, missed the two Cretaceous fossil species. Chan hypothesized that the genus probably originated in the Indo-Malay region, since most of the recent species (11 of 17) are present there. Chan said Jenkins (1972) was likely right in suggesting that the Atlantic group migrated to the Atlantic through the Tethys, instead of reaching the Atlantic via southern Africa.
Not since Jenkins' (1972) work has there been a taxonomically comprehensive examination of both the recent and fossil species of Metanephrops. Since 1972, six new species have been described, including the only two Cretaceous fossil species. Moreover, there has never been a cladistic analysis of the genus. Thus, the current study provides us a new opportunity to attempt reconstructing the history of the genus. We want to determine if the four, traditionally recognized, groups of Metanephrops species are supported by cladistic analysis. We also want to know what cladistic analysis indicates about the evolutionary and biogeographic history of the genus.
Four Traditional Groups
Recent species have traditionally been divided into four morphologic groups based on the ornamentation of the cephalothorax, pleon terga, and form and ornamentation of the chelipeds (Table 1). Neither of the two Cretaceous fossil species, M. rossensis and M. jenkinsi, fits neatly easily into any of the four groups.
The binghami or Atlantic group, M. binghami (Boone, 1925) and M. rubellus (Moreira, 1903), is characterized by a generally smooth (non-spiny) carapace, unsculptured pleon terga, uropods spineless dorsally, and by chelipeds that exhibit six spiny, longitudinal carinae and a flat outer margin. The Atlantic species are unique among congeners in possessing a postantennal spine, and in having a supracervical spine (named herein, character 9). Additionally, both species possess cervical and postcervical spinules that are absent in most Indo-Pacific species.
The japonicus group, consisting of M. japonicus (Tapparone-Canefri, 1873), M. andamanicus (Wood-Mason, 1892), M. armatus Chan and Yu, 1991, M. formosanus Chan and Yu, 1987, M. mozambicus Macpherson, 1990, M. sagamiensis (Parisi, 1917), and M. velutinus Chan and Yu, 1991, is characterized by a generally smooth (non-spiny) carapace, complexly sculptured pleon terga, uropods spineless dorsally, and spiny, carinate chelipeds with an outer border somewhat flat but with a longitudinal, medial depression. Terga have arched, lateral “eyebrow” furrows (character 15) that are absent in the binghami and thomsoni groups and in 2 of 4 of the arafurensis group. Terga also usually (6 of 7 recent species) have a “scales” shaped facade (character 16); “scales” are unique to this group. Chan and Yu (1987, 1988) initially said M. formosanus does not fit into any of the four groups, and was morphologically intermediate between the japonicus and thomsoni groups but, in 1997, Chan referred M. formosanus to the japonicus group (also see Chan and Yu, 1991).
The thomsoni group, consisting of M. thomsoni (Bate, 1888), M. boschmai (Holthuis, 1964), M. challengeri (Balss, 1914), M. sibogae (De Man, 1916) and M. sinensis (Bruce, 1966b), is characterized by a smooth (non-spiny) carapace, terga that are either completely unsculptured or (on M. sinensis and M. thomsoni) traversed by a transverse furrow interrupted medially, uropods spineless dorsally and also by weakly carinate, nearly smooth, e.g., M. boschmai and M. sinensis, to finely granulate chelipeds with outer margin always angular. This combination of features is unique to this group, although each of these features is exhibited independently in other Metanephrops.
The arafurensis group, consisting of M. arafurensis (De Man, 1905), M. australiensis (Bruce, 1966a), M. neptunus (Bruce, 1965), and the fossil M. motunauensis is characterized by a generally spiny carapace, pleon terga that are deeply marked by at least one or two transverse furrows, by uropods spinulose on the dorsal surface, and by weakly carinate, finely granulose to spiny (in M. neptunus) chelipeds with the outer margin always angular. These species are unique among congeners in having a generally spiny carapace, and in having spiny uropods.
The geographic ranges of the arafurensis, japonicus, and thompsoni groups largely overlap in the Indo-West Pacific. The japonicus group ranges (via two species) much farther across the Indian Ocean, all the away to Africa.
Materials and Methods
Character Selection
We tried to make the matrix as big as possible, and used all known, cladistically informative characters (n = 35, App. I, II). Many other characters were considered but turned out to be invariant or autapomorphous, and therefore useless in determining phylogenetic relationships between Metanephrops species. Of the autapomorphous characters, several were those that made individual species distinctive and, thus, could be useful in species taxonomy, dichotomous keys, etc. All 35 characters describe external hardparts, including aspects of the cephalothorax (#1-13; Fig. 2), pleon and telson (#14-23, 35; Fig. 5), appendages (#24-34), and color pattern (#34-35). All but two characters (#3, #19) are coded as binary, with most character states being “absent” or “present”.
Taxon Selection
All species of Metanephrops were used in the analysis. All 17 recent taxa were included in the matrix, and all of these were examined firsthand by DT, T-YC, or both. All three recognized fossil taxa were included. The two Cretaceous species, M. jenkinsi and M. rossensis, were examined directly, but the Pliocene M. motunaeunsis could not be obtained for direct examination. Material used for the present analysis is listed in Appendix III. Topotypic material was examined for all species except M. andamanicus, M. sibogae and M. sinensis (though topotypic material of these three species had been examined by the second author in Chan and Yu, 1991; Chan, 1997; and at the Fisheries Research Station in Hong Kong). Specimens of each species of Metanephrops from different localities are generally similar. However, marked variations complicating the coding of one or two characters are found in three species: M. australiensis [#15 (see Chan, 1997), #34], M. thomsoni (#34), and M. velutinus (#34). In these cases, only topotypic material was used for coding the characters.
Metanephrops thomsoni from its type-locality in the Philippines differs from those found elsewhere in lacking the distinctive red bands on the large chelipeds (Chan, 1998). Although it seems that the topotypic material generally has the surface of the pleon less pitted, and the spines better developed on the thoracic median carina but less developed on the inner margin of palm, there is overlap in these characters which makes the separation rather difficult.
A very young (upper Miocene) fossil species, Wongastacia taiwanica Hu, 1983 had been described from Taiwan. A search of the holotype and paratypes of W. taiwanica deposited at the National Taiwan University, Taipei did not locate any specimens. Only one paratype, NMNS 007606 (also see Hu and Tao, 1996, Pl. 2-Fig. 6) deposited at the National Museum of Natural History, Taichung, was found, but it contains too little information to determine if it really belongs to the same species as the holotype (Hu and Tao, 1996 listed other paratypes NMNS 007618, 007618a-d at the National Museum of Natural History but they are fragments and were collected in 1994, long after the original description). Nevertheless, the general outline of the carapace and the sculpture of the pleon, particularly the characteristic pits on the surfaces of the pleon terga, in the photographs of the types show that W. taiwanica is almost identical with those recent M. thomsoni from Taiwan (see photos in Hu, 1983; Hu and Tao, 1996—fossil form; Chan and Yu, 1988; 1993—extant form). W. taiwanica is treated as a synonym of M. thomsoni in the present study.
We included two species of Hoploparia, one (H. albertaensis) as the out-group and the other (H. antarctica) because it has been hypothesized to be ancestral to Metanephrops. Hoploparia, the most diverse genus of fossil lobster, is well known from the Cretaceous (ranging from Valanginian— Lower Miocene), and pre-dates by over 50 million years the oldest known Metanephrops (Campanian — recent). Hoploparia, with its long-ranging and abundant fossil record, has been interpreted in both non-cladistic and cladistic studies as ancestral to modern lobster genera (Tshudy and Babcock, 1997), and seems a good, safe, reliable choice for determining character state polarity. Tshudy and Sorhanuus (2003) performed a cladistic analysis of Hoploparia species and concluded that Hoploparia is wastebasket genus. We stand by that conclusion but, nonetheless, believe it is possible to select a mainstream/typical-looking Hoploparia for out-group purposes. We selected the species H. albertaensis as out-group because it is an older species (Early Cretaceous: Coniacian) with a morphology typical of the genus. Hoploparia antarctica was also included in the matrix because, as discussed above, it has been hypothesized, on the basis of its morphology and co-occurrence with the oldest known fossil species of Metanephrops, to be ancestral or near to ancestral to Metanephrops (Feldmann et al., 1993).
Cladistic Analysis
The unweighted cladistic analysis discussed herein was performed with PAUP* 4.0 (written by David L. Swofford, 2002). The search was done using the branch-and-bound algorithm, which guarantees that the most parsimonious tree(s) is found. Character state optimization in this study was implemented using the accelerated transformation approach (ACCTRAN option in PAUP). This favors an early acquisition of a new character state with a subsequent homoplasy accounted for by a reversal.
TreeView (Page, 1996) was used to draw the phylogenetic trees. Unweighted Bremer support values were calculated following the procedure outlined in Bremer (1994). Bremer support values indicate the number of extra evolutionary steps needed to collapse a node in the strict consensus tree; thus, the higher the value, the more stable the node. Bremer support values were computed, using the branch-and-bound algorithm, in the program PAUP (Version 4.0; Swofford, 2002) by successively including progressively longer trees in the strict consensus analysis; that is, until the tree was fully collapsed.
Results and Discussion
General
Unweighted analysis produced a cladogram (Fig. 6) that represents the strict consensus of 24 most parsimonious trees, and has a length = 97, retention index = 0.6907, consistency index = 0.3814, and rescaled consistency index = 0.2635. Unweighted Bremer support values are shown for resolved clades. Bremer support values for the more basal clades are low, mostly “1”'s, indicating that the more basal nodes are not particularly stable. More derived clades are better supported, with Bremer support values of 2, 3, or 4.
This cladistic analysis corroborates a conclusion of previous cladistic analyses, that homoplasy is rampant in the evolution of clawed lobsters. Abundant homoplasy in the evolution of clawed lobsters was previously reported at the generic level (Tshudy and Babcock, 1997) and Tshudy and Sorhannus (2000) reported on convergence in cheliped form at high taxonomic levels within Crustacea. In the current, species-level analysis, only 5 of the 68 synapomorphies identified by the analysis were found to be unambiguous and unreversed synapomorphies (Fig. 6). These rare, unambiguous synaporphies include: character state 24(0), scaphocerite longest medially/sub-medially, instead of externally; 9(1), supracervical spine present; 1(1), velvet-like pubescence on carapace; 2(1), generally spiny cephalothorax; and 23(1), uropodal endopod dorsal surface with spines.
Testing the Four Traditional Groupings
The four traditional groupings of the species of Metanephrops were tested cladistically herein for the first time. Cladistic analyses support the traditional, non-cladistic, morphology-based arafurensis, binghami and japonicus groups made by Jenkins (1972) and maintained/emended by Chan (1997) and others (Fig. 6). Each of these three groups is monophyletic. Additionally, Chan's (1997) referral of M. motunauensis to the arafurensis group is supported by the current cladistic results. The thomsoni group is not well supported by the cladogram, although 3 of its 5 species plot out as part of a monophlyletic group.
Given the overall morphologic similarity between M. sinensis and M. thomsoni (two members of the traditional thomsoni group), it may be surprising that they would plot so far apart on the cladogram. As coded, the species differ by 7 character states (#4, #5, #10, #13, #28, #32, #33), although close inspection shows that there were some tough calls involved in coding (#28, #32), and that one or the other species is somewhat variable (character usually present, rarely absent) in 4 of the 7 characters (M. sinensis for #4, #5, #10; M. thomsoni for #33).
The two oldest fossil species, M. rossensis and M. jenkinsi, were previously concluded, by non-cladistic means, (Tshudy, 1993, p. 214-215) to be not accommodated by any of the four traditional groups. In Tshudy's (1993) judgement, Metanephrops rossensis and M. jenkinsi are most similar, but not identical, to the binghami group (Table 1). Cladistic analysis shows the fossil species forming a monophyletic group with the binghami group.
Phylogeny and Evolutionary History
Based on fossil occurrences, the evolutionary history of Metanephrops has been hypothesized to include origin from Hoploparia in the shallow-water, Late Cretaceous seas of the southern high latitudes and subsequent dispersal into deepwater settings worldwide (Feldmann and Tshudy, 1989). This is based on the fact that the oldest known fossil species, the Late Cretaceous M. rossensis (Campanian) and M. jenkinsi (Maastrichtian—Paleocene), are known from continental shelf deposits now exposed on the Antarctic Peninsula. We can further hypothesize here that Metanephrops originated specifically in the South Atlantic Ocean, this because the fossil species are from the eastern (Atlantic) side of the Antarctic Peninsula. In the Late Cretaceous, the Atlantic Ocean was narrower, and South America and Antarctica were contiguous [(until about the Eocene) (Scher and Martin, 2006)]. This would have made M. rossensis and M. jenkinsi Atlantic species living on the shallow shelf around the eastern (Atlantic) margin of the Antarctic Peninsula.
We wanted to test this hypothesized evolutionary history cladistically. That is, we wanted to determine which clades and which species are most plesiomorphic and most derived and, further, wanted to test the hypothesized origin of Metanephrops. The hypothesized history includes the following four testable components.
Hoploparia is morphologically/cladistically more plesiomorphic than Metanephrops.
The Late Cretaceous fossil species M. rossensis and M. jenkinsi are morphologically/cladistically the most plesiomorphic of all known Metanephrops.
Metanephrops originated in the southern high latitudes.
Metanephrops originated specifically the South Atlantic Ocean.
First Component
The first component is not rigorously tested in this analysis, since a Hoploparia (H. albertaensis) is used as the out-group. For what it is worth, a second Hoploparia species, H. antarctica, is indeed shown to be as plesiomorphic as any Metanephrops, and more plesiomorphic than most. Given the heterogeneity of the genus Hoploparia (currently a “wastebasket genus” according to Tshudy and Sorhannus, 2003), this result was not as automatic as it might seem.
Second Component
Cladistic results oppose that the oldest fossil species, the late Cretaceous M. rossensis, is morphologically most plesiomorphic. The recent species, M. challengeri, shows only 4 apomorphies, whereas M. rossensis has 10 apomorphies. The other Metanephrops species have more apomorphies than M. rossensis and M. challengeri. Neither M. challengeri nor M. rossensis possesses any unambiguous synamporphies (all of their apomorphies/synapomorphies are duplicated elsewhere on the cladogram).
The other Late Cretaceous fossil species, M. jenkinsi, was determined cladistically to be the next most derived member (after M. rossensis) of the Atlantic/binghami clade. Metanephrops jenkinsi shows 15 synapomorphies, 5 more than does M. rossensis.
The binghami group is most plesiomorphic, and the arafurensis group most derived, according to numbers of apomorphies. The stems of the binghami, japonicus, and arafurensis groups show 8, 12, and 26 apomorphies, respectively, with terminal taxa accumulating additional apomorphies. Metanephrops neptunus, the most derived species, has 38 apomorphies. The basal stem of the arafurensis and japonicus group clades is marked by an unambiguous, unreversed synapomorphy [24(0), scaphocerite longest medially/submedially (instead of exteriorly, as in the binghami Group)]. The arafurensis group is defined by two additional unambiguous synapomorphies [2(1), spiny cephalothorax, and 23(1), uropodal endopod with spines on dorsal surface]. The only other two unambiguous, unreversed synapomorphies on the cladogram occur within the binghami clade [9(1), supracervical spine present] and within the japonicus clade [1(1) velvet-like pubescence on cephalothorax].
Third Component
Cladistic results corroborate the existing hypothesis that Metanephrops evolved in the southern high latitudes. The cladistically most plesiomorphic single species, the recent M. challengeri, and the next most plesiomorphic species, the Late Cretaceous M. rossensis, are both known from the high southern latitudes. Likewise, the most plesiomorphic species group, the binghami group, is best known from the high southern latitudes.
Fourth Component
While origination of Metanephrops in the southern high latitudes seems likely, origination specifically in the South Atlantic is highly equivocal. The two Late Cretaceous fossil species reveal at least a long history in the south Atlantic. Moreover, these fossil species are arranged cladistically as forming a monophyletic group with the extant Atlantic/binghami group, the most plesiomorphic of the traditional species groups. Taken together, these findings at least suggest an origin in the South Atlantic. On the other hand, the recent M. challengeri, the most plesiomorphic species (not species group), is known today from around New Zealand in the South Pacific. Given preservational and sampling biases, it is obviously possible that Metanephrops may have evolved in the South Pacific or southern Indian Ocean, but that any fossil record has, thus far, gone unnoticed. If origination occurred in the south Atlantic, species of Metanephrops could have dispersed outward, in the Late Cretaceous, by way of the Tethys Sea and/or around southern Africa. Later, in the Eocene, Metanephrops could have also dispersed from the Atlantic and into the Indian Ocean via the Drake Passage, newly opened between Antarctica and South America in about the Eocene (Scher and Martin, 2006).
Conclusion
Cladistic analysis herein corroborates previous studies which indicated that homoplasy is rampant in the evolution of clawed lobsters. Only 5 of the 68 synapomorphies identified by the analysis are unambiguous, unreversed synapomorphies.
Cladistic analysis supports the traditional, non-cladistic, morphology-based arafurensis, binghami and japonicus groups made by Jenkins (1972) and maintained/emended by Chan (1997) and others. These three groups are monophyletic. The thomsoni group is not well supported by the cladogram, although 3 of its 5 species plot out as part of a monophyletic group. The two oldest fossil species, M. rossensis and M. jenkinsi, were previously concluded, by non-cladistic means, to be not accommodated by any of the four traditional groups, but most similar to the binghami group. Cladistic analysis shows the fossil species forming a monophyletic group with the binghami group.
Cladistic results oppose that the oldest fossil species, M. rossensis, is morphologically most plesiomorphic. Metanephrops challengeri shows the fewest number of apomorphies of all Metanephrops species, followed by M. rossensis. Of the three traditionally recognized species groups supported herein, the binghami group is most plesiomorphic, and the arafurensis is most derived, based on numbers of synapomorphies.
Based on geographic occurrences of fossil and recent species, and, now, corroborative cladistic results, it seems likely that Metanephrops originated in the southern high latitudes. A more precise location of origin remains highly equivocal. We look forward to new information, especially from new fossil occurrences and from molecular studies, for testing and refining hypotheses on the origin and evolution of Metanephrops.
Acknowledgments
We wish to thank S. Ahyong, A. Crosnier, D. Evans, D. A. Lee, B. Richer de Forges, M. Takeda, M. Tavares, and H. Watabe for obtaining and exchanging Metanephrops material with TYC and therefore making the present work possible. We thank M. Tabecki, British Antarctic Survey, and W. Blow, R. Lemaitre, and K. Reed, Smithsonian Institution (USNM), for providing access to, and loans of, specimens. Some specimens used in the present work were collected by the “Panglao 2004 and 2005” expeditions. The “Panglao 2004” Marine Biodiversity Project was a collaboration between Muséum national d'Histoire naturelle, Paris (MNHN, Principal Investigator, P. Bouchet) and University of San Carlos, Cebu City (Principal Investigator, D. Largo). The “Panglao 2004” was supported by the Total Foundation for Biodiversity and the Sea, the French Ministry of Foreign Affairs, and the ASEAN Regional Center for Biodiversity Conservation (ARCBC). The Philippines Bureau of Fisheries and Aquatic Resources (BFAR) is acknowledged for issuing a research permit on the material collected by the “Panglao 2004 and 2005”. The “Panglao 2005” deep-sea cruise on board M/V D/A/ BRAR was a collaboration between the MNHN (Principal Investigator, P. Bouchet) and BFAR (Principal Investigator, L. Labe), and support by the Total Foundation for Biodiversity and the Sea, the French Ministry of Foreign Affairs, the National University of Singapore (PKL Ng) and the University of San Carlos, Cebu City (D. Largo, N. Saguil). This work was partially supported by research grants from the National Science Council, Taiwan, R.O.C. and Center for Marine Bioscience and Biotechnology of the National Taiwan Ocean University to TYC, and a grant from the Edinboro University Faculty Senate to Tshudy. We thank R. M. Feldmann and an anonymous reviewer for their helpful criticisms of an earlier version.
References
Appendices
Appendix II
Explanation of Characters Analyzed in Cladistic Analysis
All 35 characters describe external hardparts, including aspects of the cephalothorax (#1-13), pleon and telson (#14-23, 35), appendages (#24-34), and color pattern (#34-35). Details on the coding of any characters are available from the first author.
Cephalothorax
1. Velvet-like pubescence on cephalothorax (>80% coverage above branchial carina).
2. Cephalothorax cuticle: general surface ornamentation
3. Supraorbital spines (“SO”): 3 or more than 3?
4. Postorbital spine (“PO”): positioned behind orbit (not more than half-way to cervical groove) and at about mid-orbital height (see M. andamanicus for example). In recent Metanephrops, this spine is situated above and anterior to the larger gastrolateral (GL) spine. It is often smaller than the metorbital spine (“MO”), which is located behind the base of the orbit.
5. Extra, smaller postorbital spine(s) (“XPO”) around PO (see M. boschmai for example).
6. Postantennal spine(s) (“PA”), a spine posterior to the keel-like antennal spine (see M. rubellus for example).
7. Cervical spinules (“C”, along posterior edge of cervical groove and directed anteriorly; see M. rubellus for example).
8. Postcervical spinules (“PC”, along posterior edge of postcervical groove and directed anteriorly; see M. rubellus for example).
9. Supracervical spine (“SC”, named herein) above cervical groove (located midway to dorsal median and posterior to gastrolateral spine; see M. rubellus for example).
10. Spinules above the gastrolateral spine (“GL”).
11. Flank spines (“FL”, named herein) on cephalic region, just anterior to inflection of postcervical groove.
12. Lateral carina (“L”) with spines or subspiniform projections (coarser than common granulation; see M. arafurensis for example).
13. Relative spacing of the two rows of thoracic median dorsal spines (less than or equal to 0.1, or greater than 0.1) determined by dividing their spacing by the length of thoracic region (measured medially). Spacing measured at 2nd spine pair posterior to the much larger, anteriormost pair (unless few pairs, then roughly 1/3 to posterior margin).
Pleon
14. Transverse furrow along posterior margin of pleon terga 2-4 (Fig. 5a, b). There is interspecific variation in the width of the furrow, the nature of its dorso-medial termination (simple or in combination with other furrows), and the degree to which the furrow parallels the posterior margin.
15. “Eyebrow” furrows (Fig. 5a). These are crescentic, concave-down, longitudinal furrows that may or may not be contiguous with transverse furrow. Several species lacking this furrow do exhibit an apunctate (and often darker colored) crescentic region in the same position. Photos can be misleading here; need to examine specimens in oblique light. Furrow is usually of a width as seen on M. mozambicus, but is unusually wide in M. japonicus.
16. Scales-shaped facade (Fig. 5b), a region of positive relief in a scales-like configuration; i.e., two posterior elements “suspended” from a transverse anterior element [“main facade” of Chan and Yu (1991, p. 22)]. Best displayed by M. japonicus (note: significant variation in details of the scales, e.g.,USNM 104182, n = 4). Metanephrops sagamiensis and M. andamanicus also good examples but scales shape is less detailed.
17. Longitudinal furrow above (and parallel to) tergum-pleuron boundary (Fig. 5a, b). Well developed on M. andamanicus. Minimum standard is M. formosanus.
18. 5th somite: spine(s) near posterior end of (longitudinal) lateral ridge at tergum-pleuron boundary (Fig. 5c) (see M. armatus for example); i.e., on about posterolateral corner of tergum.
19. 6th tergum: median anterior spine or spine pair (Fig. 5c). See M. japonicus for example of spine pair.
20. 6th tergum: median, mid-length (approximately) spine pair (Fig. 5c) See M. japonicus for example.
21. 6th pleomere tergum-pleuron boundary w/lateral spines (other than large one at posterior termination) (Fig. 5c). See M. arafurensis for example.
22. Telson: spines on rays of V-shaped carina (Fig. 5d). See M. boschmai for example.
23. Uropodal endopod—dorsal surface spinules (Fig. 5d). See M. arafurensis for example.
Appendages
24. Scaphocerite—longest medially/sub-medially, e.g., M. japonicus, or exteriorly, e.g., M. binghami?
25. Merus of cheliped: ventral (sharp) edge with many sharp spinules, i.e. serrated, e.g., M. japonicus, or large spines, e.g., M. binghami?
26. Cheliped upper surface: median/submedian keel (well developed on M. japonicus).
27. Cheliped upper surface: spines on (longitudinal) median/submedian (spines well developed on M. japonicus).
28. Cheliped lower surface: median/submedian keel (for reference, it is well developed on M. japonicus; present but least developed on M. sinensis).
29. Cheliped lower surface: spines on (longitudinal) median/submedian (spines well developed on M. japonicus).
30. Cheliped outer margin: double-ridged/squared-off keel (for reference, M. andamanicus).
31. Cheliped inner margin: pronounced keel (for reference, M. andamanicus).
32. Cheliped inner margin: spines? (for reference, M. andamanicus).
33. Cheliped inner margin with some large spines? [reference for “large” is M. japonicus; surprisingly, there are no “gray areas” (intermediate conditions); coding is easy].
Appendix III
Material Examined
“TYC”—National Taiwan Ocean University collection except otherwise stated, examined by Chan, “DT”—examined by Tshudy.
For TYC, “(F)” indicates direct observation on coloration of fresh material, “(P)” of color photograph on fresh material. For DT, all recent specimens preserved in alcohol unless otherwise indicated.
M. andamanicus—TYC: 16 specimens (F) from Pratas, South China Sea. DT: one male, one female on loan from Chan (collected 1989 by Taiwan Fisheries Research Institute; Pratas, S. China Sea, identified by Chan).
M. arafurensis—TYC: 2 specimens (P) from Indonesia (Zoologisch Museum, University of Amsterdam and Muséum national d'Histoire naturelle, Paris, examined by T-Y Chan in 1997 and with many photographs taken for the present analysis). DT: used photos in Chan, 1997 (figs. 1, 2, 5).
M. armatus—TYC: 47 specimens (F) from Taiwan; 2 specimens (P) from East China Sea. DT: one male, two females from Chan (collected 1998 by commercial trawler, Dasi, NE Taiwan).
M. australiensis—TYC: 9 specimens (P) from Australia; 21 specimens (F) from the Philippines; 1 specimen from Indonesia (on loan by Chan from the Muséum national d”Histoire naturelle, Paris) . DT: one male from Chan (collected 1985 by CSIRO, NW Australia, 18°19′S, 117°49′E, 414 m).
M. binghami—TYC: photo in Sea Frontiers 21(3): cover page. DT: many specimens (9 jars) from Caribbean (United States National Museum numbers USNM 11429, 98656, 136712, 136720, 136724, 136716, 170695, 170697, 170698).
M. boschmai—TYC: 16 specimens (P) from Australia. DT: two ovigerous females on loan from Chan (collected CSIRO, W. Australia, 31°29′S, 114°54′E).
M. challengeri—TYC: 6 specimens (P) from markets in Paris and Sydney, supposedly imported from New Zealand. DT: one male on loan from Chan (supermarket, Paris).
M. formosanus—TYC: 44 specimens (F) from Taiwan DT: two males, one female on loan from Chan (collected 1998 by commercial trawler, Dasi, NE Taiwan, 5/7/98).
M. japonicus—TYC : 10 specimens (P) from Japan. DT: one male, one ovigerous female from Chan (collected in lobster pot in Tokyo Submarine Canyon, 180-210 m, mud and fine sand).
M. jenkinsi—DT: borrowed several specimens from Smithsonian (holotype USNM 424598; paratypes USNM 424599-424603, 424605, 424613-424614).
M. motunauensis—None; examined only from the original description and figures in Jenkins (1972).
M. mozambicus—TYC: 1 specimen (P) from Madagascar, 2 specimens from fish markets in Japan with places of origin unknown. DT: Paris Museum paratype (one ovigerous female) from Chan (collected Madagascar, 1972, 12°38′S, 48°15.5′E, identified by Macpherson, 1987).
M. neptunus—TYC: 1 specimen (P) from Pratas, South China Sea; 4 specimens (F) from the Philippines; 4 specimens (P) from Australia. DT: two females (one large and ovig., one small) on loan from Chan (collected 1986 by CSIRO, NW Australia, 18°19′S, 117°49′E, 414 m, identified by Chan, 1986).
M. rossensis—DT: borrowed from British Antarctic Survey, BAS IN. 2161, 2181, and two unnumbered specimens.
M. rubellus—TYC: 6 specimens specimens from Brazil. DT: one male on loan from Chan (collected 1961 by Calypso, South America, 35°5′S, 52°33′W, 11fm.)
M. sagamiensis—TYC: 26 specimens (F) from Taiwan; 1 specimen (P) from Japan.
M. sibogae—TYC: 8 specimens (P) from Australia. DT: two females from Chan (collected 1990 by CSIRO, N. Australia, 7°45′S, 130°10′E; identified by Chan).
M. sinensis—TYC: 44 specimens (F) from the Philippines. DT: one male on loan from Chan (collected Musorstom III, Philippines, identified by Chan).
M. thomsoni—TYC: 36 specimens (F) from the Philippines; 61 specimens (F) from Taiwan; 2 specimens (P) from Viet Nam. DT: one male, one ovigerous female on loan from Chan (collected 1985 by Musorstom III; Philippines, 12°6′N, 121°15′E, identified by Chan).
M. velutinus—TYC: 3 specimens (F) from the Philippines; 7 specimens (P) from Australia; 1 specimen from Salomon Island. DT: two females on loan from Chan (collected 1989 by CSIRO, W Australia. 28°9′S, 113°10′E, approx. 600 m).
Fig. 2. Schematic line drawing of Metanephrops cephalothorax in A, right lateral and B, dorsal views, showing positions of grooves (lower case), carinae and spines (upper case). a = antennal groove; A = antennal spine; B = branchial carina; bc' = ventral extension of branchiocardiac groove (see Tshudy, 1993); c = cervical groove; C = cervical spinules; CM = cephalic median carina; CS = cervical spine; FL = flank spines; GL = glastro-lateral spine; h = hepatic groove; H = hepatic spine; I = intermediate carina; ic = intercervical groove; L = lateral carina; MO = metorbital spine; pc = postcervical groove; PA = postantennal spine; PC = postcervical spinules; PM = posteromarginal carina; PO = postorbital spine; SC = supracervical spine; SO = supraorbital carina/spines; TM = thoracic median carina/spines; XPO = extra postorbital spine(s).
Fig. 3. Schematic drawings of cephalothoraxes of Paraclytia, Metanephrops, Nephrops and Hoploparia, emphasizing differences in supraorbital, antennal and thoracic ornamentation.
Fig. 4. Geographic distribution of all Metanephrops species, fossil (3) and recent (17), with symbols showing traditional groupings. The arafurensis group: 
= M. arafurensis, 
= M. motunauensis, 
= M. neptunus, 
= M. australiensis. The binghami group: ▴ = M. binghami, ▾= M. rubellus. The japonicus group: 
= M. formosanus, 
= M. japonicus, 
= M. mozambicus, 
= M. andamanicus, 
= M. armatus, 
= M. sagamiensis, 
= M. velutinus. The thompsoni group: 
= M. boschmai, 
= M. challengeri, 
= M. sinensis, 
= M. sibogae, 
= M. thompsoni. Cretaceous fossil species: 
= M. jenkinsi, ♦ = M. rossensis.
Fig. 5. Selected aspects of pleon morphology of Metanephrops used in the cladistic analysis. A, third tergum, M. mozambicus; B, third tergum, M. japonicus; C, fifth tergum, showing spine at posterior end of lateral ridge, as on M. armatus, and sixth tergum, showing positions of median anterior (MA) and median mid-length (MM) spine pairs and, also, lateral spines, as on M. arafurensis; D, telson and uropods, showing spines on rays of telson, as on M. boschmai, and spines on endopodal uropod, as in arafurensis group.
Fig. 6. Cladogram produced by unweighted analysis; represents strict consensus of 24 most parsimonious trees, and has a length = 97, retention index = 0.6907, consistency index = 0.3814, and rescaled consistency index = 0.2635. Character bars indicate unambiguous, unreversed synapomorphies. Unweighted Bremer support values are shown for resolved clades. 
referring to fossil species.
Table 1. Morphologic characteristics of the four traditionally (non-cladistically) recognized groups (modified from Chan, 1997), expanded here to include new characters (*) and the two Cretaceous fossil taxa, M. jenkinsi (J) and M. rossensis (R). Neither fossil species fits neatly into any of the four groups. **Chelipeds: binghami and japonicus groups with outer margin flat (squared off; not angular), but japonicus group also with longitudinal medial depression. Abbreviations: “ang.”, angular; “sp.”, spiny.
