Phylogenetic Relationships of Clawed Lobster Genera (Decapoda : Nephropidae) Based on Mitochondrial 16S Rrna Gene Sequences Y

Home , Cape lobster, Hoploparia, Metanephrops, Nephropsis

The University of Maine DigitalCommons@UMaine

Marine Sciences Faculty Scholarship School of Marine Sciences

2-1-1998 Phylogenetic Relationships of Clawed Lobster Genera (Decapoda : Nephropidae) Based on Mitochondrial 16S rRNA Gene Sequences Y. K. Tam

Irv Kornfield University of Maine - Main, [email protected]

Follow this and additional works at: https://digitalcommons.library.umaine.edu/sms_facpub

Repository Citation Tam, Y. K. and Kornfield, Irv, "Phylogenetic Relationships of Clawed Lobster Genera (Decapoda : Nephropidae) Based on Mitochondrial 16S rRNA Gene Sequences" (1998). Marine Sciences Faculty Scholarship. 89. https://digitalcommons.library.umaine.edu/sms_facpub/89

This Article is brought to you for free and open access by DigitalCommons@UMaine. It has been accepted for inclusion in Marine Sciences Faculty Scholarship by an authorized administrator of DigitalCommons@UMaine. For more information, please contact [email protected]. JOURNALOF CRUSTACEANBIOLOGY, 18(1): 138-146, 1998

PHYLOGENETIC RELATIONSHIPSOF CLAWED LOBSTER GENERA (DECAPODA: NEPHROPIDAE) BASED ON MITOCHONDRIAL 16S rRNA GENE SEQUENCES

YanKit Tamand Irv Kornfield

ABSTRACT Approximately350 base pairs(bp) of the mitochondrial16S rRNAgene were used to studythe phylogeneticrelationships among 5 generaof the clawed lobsterfamily Nephropidae(infraorder Astacidea), including Homarus, Homarinus, Metanephrops, Nephrops, and Nephropsis. Maximum- parsimonyanalysis, using a hermitcrab, Pagurus pollicaris (infraorder Anomura), as an outgroup, produceda tree topologyin whichHomarus and Nephrops formed a well-supportedclade thatex- cludedHomarinus. The sametree topology was obtainedfrom both neighbor-joining and maximum- likelihoodanalyses. Some morphologicalcharacters that appear synapomorphic for Nephropsand Metanephropsmay be due to convergencerather than symplesiomorphy.The currenttaxonomy, therefore,does not reflectthe phylogenyof this groupas suggestedby the moleculardata. More moleculardata and studiesusing homologousmorphological characters are neededto reacha bet- ter understandingof the phylogenetichistory of clawedlobsters.

Clawed lobsters are marinedecapods be- ical charactersmay provide valuable infor- longing to the superfamily Nephropoidea mationfor taxonomicclassification, the char- (Decapoda: Astacidea). The Nephropidae acters used in distinguishingdifferent taxa Dana, 1852, comprisingthree subfamilies and may not be homologous structuresand thus eleven genera, contains most of the clawed may not containphylogenetic signals. Thus, lobstersin this superfamily(Holthuis, 1974). the taxonomic classification of clawed lob- Amongthem are the commerciallyimportant sters may not reflect the phylogenetic rela- genera Homarus Weber, 1795, Nephrops tionships of the group. Homologous struc- Leach,1814, and Metanephrops Jenkins, 1972. turesneed to be definedand appliedin mak- Holthuis(1991) presenteda comprehensivere- ing phylogeneticinferences (Tshudy, 1993); view of clawed lobsters, emphasizingthose homoplasyof characters,such as convergence that are of interestto global fisheries. and parallelism,produce noise and mislead The Nephropidaeis an old family which data analysis. Dependingon the degree that has a fossil recordextending from the Mid- some morphologicalcharacters are conver- dle Jurassicto the Recent (Glaessner,1969). gent, phylogenetic inferences may be com- Ornamentationson the carapaceof fossils are promised.This concernapplies to inferences well preserved;these patternsof grooves and aboutfossil nephropidsas well. eminencesmay provideclues to the evolution Molecularcharacters, particularly DNA se- of lobster lineages. Based on carapacemor- quencedata, provide an independentdata set phology,Glaessner (1969) proposeda hypo- with which to construct phylogenetic hy- theticalphylogeny of astacideans.In this, the potheses(Hillis et al., 1996). Molecularstud- fossil genusPalaeophoberus Glaessner, 1932, ies of phylogenetic relationships among gave rise to the genus Hoploparia McCoy, clawed lobstersare limited.Chu et al. (1990) 1849, one line of which developedinto cray- studiedenzyme polymorphismin three spe- fish (Astacus), and anotherline into Nephrops- cies of Metanephropsin Taiwan.Hedgecock like andHomarus-like lobsters. It hadbeen pro- et al. (1977) examinedthe genetic variation posed thatfossil and Recentnephropids were between Homarus americanus H. Milne Ed- composedof two subfamilies(Mertin, 1941), wards, 1837, and H. gammarus(Linnaeus, the Nephropinaeand the Homarinae.How- 1758) by using allozyme data. The present ever, based on the morphologyof living lob- study,using molecularcharacters as an alter- sters,Holthuis (1974) did not acceptthe idea. native approachto morphology,provides in- The currenttaxonomy of clawedlobsters is dependentclues about the phylogeny of the based on morphologicalfeatures (Glaessner, clawed lobsters. By using universalprimers 1969; Holthuis,1991). Althoughmorpholog- within a conservative region in the mito-

138 TAMAND KORNFIELD:PHYLOGENETICS OF CLAWEDLOBSTER GENERA 139

Table 1. Species studied and the sampling localities.

Species Infraorder Abbreviations Sampling locality

Homarus americanus H. Milne Edwards, 1837 Astacidea HA Gulf of Maine, U.S.A. Homarus gammarus (Linnaeus, 1758) Astacidea HG Guernsey, U.K. Homarinus capensis (Herbst, 1792) Astacidea HC Cape Province, South Africa Metanephrops mozambicus Macpherson, 1990 Astacidea MM Natal, South Africa Nephrops norvegicus (Linnaeus, 1758) Astacidea NN Celtic Sea, Ireland Nephropsis aculeata Smith, 1881 Astacidea NA Massachusetts, U.S.A. Nephropsis stewarti Wood-Mason, 1872 Astacidea NS Natal, South Africa Panulirus longipes (A. Milne Edwards, 1868) Palinuridea PL South China Sea, Hong Kong Scyllarides nodifer (Stimpson, 1866) Palinuridea SN Gulf of Mexico, U.S.A. Pagurus pollicaris Say, 1817 Anomura PP Massachusetts, U.S.A.

chondrial genome, a homologous DNA seg- Pagurus pollicaris Say, 1817 (infraorderAnomura), was ment from all the taxa can be amplified and used as an outgroup for parsimony analysis. DNA was prepared from samples of muscle from the compared. abdomen or pereiopods. MtDNA was prepared by phe- A now conventional source for DNA se- nol/chloroform extraction of proteinase-K-digested tis- quences is the mitochondrion (Avise, 1994), sues (Ausubel et al., 1989). The DNA templates were then an organelle that contains multiple copies of subjected to polymerase chain reaction (PCR) (Saiki et standard et a small, maternally inherited genome, mito- al., 1988), using protocols (Palumbi al., 1991). A 570-base-pair (bp) region within the mtl6S rRNA chondrial DNA (mtdna). MtDNA sequences (16S) gene was amplified using primers 16SA (5'CGC- were used to define taxonomic relationships CTGTTTATCAAAAACAT3') and 16SB (5'CTCCG- among species of Homarus and distinguish GTTTGAACTCAGATC3') (Xiong and Kocher, 1991). from them the genus Homarinus Komfield, PCR amplification was performed using 30 cycles of 30 s s / the lobster of South Africa 94?C, / 50?C, 60 72?C, 90 s; the initial denatu- 1995, Cape ration step at 94?C lasted 5 min and the final extension (Kornfield et al., 1995). The 16S ribosomal step at 72?C lasted 10 min. Double-stranded PCR prod- RNA (rRNA) gene in the mitochondrial ucts were subjected to asymmetric PCR from both 5'-di- genome contains conservative regions which rection to generate templates for dideoxy sequencing et have been used in many phylogenetic stud- (Sanger al., 1977), using 35 cycles of the same PCR conditions. Prior to sequencing, PCR products were pu- ies at generic and higher taxonomic levels for rified by filtration through Millipore Ultrafree-MC re- a wide variety of organisms (Hillis et al., generated cellulose membrane filters of 30,000 nominal 1996). Recently, Crandall and Fitzpatrick molecular weight limit (NMWL) (Millipore Corporation). (1996) used a mitochondrial 16S rRNA gene Sequences were aligned using ESEE (Cabot and Beck- to the of enbach, 1989). Secondary structures of the partial mtl6S sequence study relationships cray- rRNA gene sequences were inferredusing Mulford (Jaeger fishes. The present study examined relation- et al., 1989) to assure homologous sequence alignment. ships among five of the 11 extant genera of Sequences without 2 highly variable regions (Parker and the family Nephropidae by comparing par- Kornfield, 1996) were subjected to all data analyses (Fig. A matrix of tial sequences of the mtl6S rRNA gene. Our 1). sequence divergence using Kimura's two- parametermethod (Kimura, 1980) was generated and sub- objective was to compare phylogenetic hy- jected to neighbor-joining analysis using MEGA (Kumar potheses based on this molecular data set to et al., 1993). Five hundred bootstrap replicates were per- the current taxonomy based on morphologi- formed to access the confidence level at each branch cal features. (Felsenstein, 1985). Maximum-likelihood trees were con- structed using the program DNAML in PHYLIP, Version 3.5 was con- MATERIALSAND METHODS (Felsenstein, 1993). Parsimony analysis ducted using the exhaustive search option in PAUP Ver- Seven species in 5 genera of the clawed lobster fam- sion 3.1.1 (Swofford, 1993); alignment gaps were in- ily Nephropidae were studied: Homarus americanus, cluded as characters and only phylogenetically informa- Homarus gammarus, Homarinus capensis (Herbst, 1792), tive characters (Hillis et al., 1996) were used. To assess Metanephrops mozambicus Macpherson, 1990, Nephrops the heuristic confidence in the parsimony trees generated norvegicus (Linnaeus, 1758), Nephropsis aculeata Smith, by PAUP, 2,000 bootstrap replicates were performed. In 1881, and Nephropsis stewarti Wood-Mason, 1872 (Table order to compare the effects of including and excluding 1). One individual from each species was used for analy- the 2 highly variable regions, all available data (i.e., 474 sis. A spiny lobster, Panulirus longipes (A. Milne Ed- bp) were also subjected to maximum-parsimony analysis wards, 1868), and a slipper lobster, Scyllarides nodifer using PAUP as described above. Different tree topologies (Stimpson, 1866), both belonging to the infraorderPalin- were compared by using the user-defined tree function in uridea, were also included in the analysis. The hermit crab MacClade, Version 3 (Maddison and Maddison, 1992). 140 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 18, NO. 1, 1998

HA GOCCGATTACTCAGTCAATATGCCTATGGCTTTATGTGCAAGAATTTAAAA100 HG ...... T ... HC ...... N ...... T...... G . .A .A. NN ...... C ...... T ..GT ...... GT ... NA ...... A ...... G.T...... CC...... TT.G...... AGT. NS ...... A.G..T...... G.T ...... CC ...... T.G ...... G.AGT. MM ...... C...... T...... G ...... G .....A...... G ...... GG.A... SN ...... AA.G ...... A. ...T....T...... S...G .... G.....A ....A.G..T...... T..G.T. PL ...... NNA.G...... T...... A-....T.. .C...... G.A..G...... C..C..A...... ATTG...... TGCT.T. PP A.....A ...... G..T...... A.AGA.T....T .... A...G... .A ...... AGTATC.T ....T..TT.A.T.

HA AAAATTGAATT-TGACTTTTAAGTGAAAAGGCTTAAATATTTTAAAGGGACGATAAGACCCTATAAAGCTTAATA-ATTTAGTATATAATTAGATGAGTT 200 HG .T ...... HC GT.-..C..AGT ...... T.C ...... A.. NN ...... A...... CC....A .A .. NA GG...... A...... CGG...... CG ...... -C.....T. ..A..A.T.. NS ..G...... A ...... CG...... G...... -C.AT..T. . A..A.T.. MM ..GT.....A ... A ...... G ...... G...... G ...... TA . ..G. ., A.A.. A..A ... SN ..TC.....A.. .A..A ...... GAGC0. G ...... T..T...... T.GCCT.....GGTT.AA.A.. PL ..TT ....A. . A .G ...... GAGA..GG ...... T.. .G..GGTCAA.A... .TA.T. .TT...TAG. PP ..T...... T...... AAAA ...... T...T,C.T.AA.TA. . .T.T.AATTT.A.A..

HA GAAAGTTTAATATTAT-TTATATAC-TAAATTATTTCGTTG-----GGGCGACGATGATATAATTT--GTAACTGTTT--AAATTTAAAATACAG-AGAT 300 HG A ...... G ...... C ...... HC . T ...... G..-GT...... A...... A ...... T. .A..... NN A ...... C ...... -T ...... T...... NA ATT...A ...... A...... GCC-GGC...... AA..G...--. T .....A.. NS ATT . . A.... .A...... AT.-GG ...... A...... AAG.G...-- .C ....TG. MM TT... G.T.....G .....CT ....A ...... A...... G.GG....-G..C..A..T.. SN AGGT-.C.T..T.A..C.....-.GG.C..A .....T...... A. . .AGA.G ....AAAAA ...... CC.TA. ...-.. G.TA.-..A.. PL AGTTA.AAGC.TAA...... T.TG.C.....T...... AGG.G ...... A ...... CC.T.G.G.AA.G. . ..TA.TT.TT. PP AT . -.. .T. . .AA. . -- .A.AG.., T.--G...... TGCTG ... G.ATA ...... TAAA..TA .... .C.AT .....-... .CA.T. .T.

HA -ATTT-GT-GTG--TAA-TGATCC-TT-GTTGTTGATTAAAAATTTAAGTTACTTTAGGGATAACAGCGTTATTTATTTTGAGAGTTCATATCGACAAA 400 HG ...... HC ..G ...... A .T ....G...... A ...... NN ...... T . A...... NA ....C ...... T...... G ...... C...... NS ..G.C.....A.T....G...... G .. A...... G..A ...... MM ...... T.....G...... A..AA ...... T.. SN .G.... AA... .TGT...... TAATAA.GAA.TT.GAA...... A_CC...... A... PL TT. ..GT. .T. .TT.-G.A...... CT...AA..-. .TC.GA.C...... A .C.TC...... C ...... A.GG PP ..G..G.A.TAATA...GT...... -.AAAA .. . .GA.GA.....A...... T . AT C...... C...... AG.... C.....

HA AAAGTTTGCGACCTCGATGTTGAATTAAAAATTNNCCATGGCGTAGGAGTTGTGGAGGTAG- -GTCTGTTCGAC 474 HG ...... C...... HC ...... TGT...... C ....A..AT ...... NN ...... NA G ...... T..TAT.G...... C....AA.ATA...... NS G...... T. .TAT..C ...... C.....A.ATA ...... MM ...... T..CT..G ., .T ....A ....AC..T-....AG ...... SN GG..A ...... G.-CCTTT.C..T.C..A ...... AA.A.G.A...... PL .GG.G ...... G.ACCTTTG.. .T.N ..C..C.ACTAGA.G...... PP ...... GT..CTTT.C..T.C..C..C C...AT.A.AGA......

Fig. 1. Sequence data for the partial mitochondrial 16S rRNA gene for all taxa in the study. Abbreviations for taxa are defined in Table 1. Sequences over-shadowed by "=" represent highly variable regions with ambiguous homologous sequence alignment which were excluded from all data analyses. "." indicates identity with the reference sequence of Homarus americanus; "-" indicates gap; N indicates an undetermined nucleotide.

RESULTS these variable regions indicated ambiguous sequence alignment, these two regions were Approximately 450 (bp) of the mtl6S excluded from all subsequent data analyses. rRNA gene were sequenced from all individu- In all, about 350 nucleotides were used for data als (Fig. 1). All sequences have been deposited analysis. Table 2 presents estimates of pairwise in GenBank (accession numbers U11238, sequence divergence of all taxa using the two- Ul 1247, U55843, U96083-U96089). In Fig. 1, parameter method of Kimura (1980). Among the over-scored regions represent two highly the clawed lobsters, Nephropsis and Metane- variable segments within the amplified se- phrops were quite divergent genetically (>7%) quences. Since the secondary structures of from the other taxa. However, the extent of TAMAND KORNFIELD:PHYLOGENETICS OF CLAWEDLOBSTER GENERA 141

Table 2. Two-parameter estimates of sequence divergence (Kimura, 1980) based on partial mitochondrial 16S rRNA gene sequences. Estimates are expressed as percentages. Abbreviations refer to Table 1.

HA HG HC NN NA NS MM SN PL PP

HA - 0.88 4.54 1.78 10.76 9.38 7.44 22.40 26.00 23.98 HG - 4.54 2.08 11.09 10.05 8.10 22.78 25.95 24.37 HC- 6.13 9.38 9.04 7.75 23.95 26.72 25.97 NN- 11.12 10.41 9.14 24.50 26.46 24.40 NA - 4.21 12.54 29.08 27.39 30.33 NS - 10.40 25.88 26.93 26.71 MM - 21.97 22.77 26.84 SN 16.34 21.46 PL - 27.49 PP

divergenceof Nephropsnorvegicus was un- ridea, had an average genetic divergenceof expected.This speciesexhibited much less di- 25.2%? 0.60%from the clawed lobsters(in- vergence (1.93% ? 0.2%) from the two spe- fraorderAstacidea), while the hermit crab cies of Homarusthan was observedbetween Paguruspollicaris (infraorderAnomura) had species of Homarus and Homarinuscapen- an average genetic divergence of 26.1% ? sis (4.54%? 0%), a taxon until recentlycon- 0.9% from the clawed lobsters.This similar generic with Homarus. The slipper lobster level of geneticdivergence among infraorders Scyllaridesnodifer and the spiny lobsterPan- indicatesthat mutationsin the mtl6S rRNA uliruslongipes, both in the infraorderPalinu- gene are saturatedat this taxonomic level.

= 5.33x y f A m m a aB

A A

B A o A AA A 0rw 0 A 30 o o 343 40 o 0 0 0 0 ,.Q 20 0 0 0 0 k 0

0 10- Between lobster taxa at genus and family level 0 0 Between lobster taxa above family but below infraorder level 0 0 Between Ilonarius and other clawed lobsters Between lobster taxa at infraorder level A Between each studied lobster and hermit crab

. . . . 20 30 40 50

Number of Transversions Fig. 2. Number of transitions versus number of transversions in all pairwise comparisons of partial mitochondrial 16S rRNA gene sequences. This figure gives an indication of the extent of transitional bias and the extent of saturation in substitutions. 142 JOURNALOF CRUSTACEANBIOLOGY, VOL. 18, NO. 1, 1998

Fig. 3. 50%majority-rule consensus of ninemost-parsimonious trees based on maximum-parsimonyanalysis of partial mitochondrial16S rRNAgene sequencesfor lobsters.Tree was rootedusing Paguruspollicaris. Bootstrap val- ues (2,000 replicates)are shownon the branches.Each of the nine most-parsimonioustrees required164 steps and had a consistencyindex of 0.707.

Figure2 depictsthe extentof transitionalbias cies of Homarus and Nephrops consistently and also the level of saturationin mutated groupedtogether. Homarinus was always ex- sites by plotting the number of transitions cluded from the Homarus-Nephrops clade. against the number of transversionsin all Anotherstrongly supported clade was formed pairwise comparisons of taxa. The slope by the two species of Nephropsis.The phy- showed in Fig. 2 gives a roughindication of logenetic positions of Homarinusor Meta- the initial transition-to-transversionratio nephropscould not be resolved with confi- amongclosely relatedspecies. A slope of 5.0 dence (but see below). When the tree topol- indicatesa 10:1transition-to-transversion bias. ogy was constrained(by using MacCladeV.3) The figureillustrates partial saturation among so that Homarus and Homarinus formed a generawithin an infraorder,while saturation clade while Nephrops and Metanephrops is close to completebetween infraorders. formedanother clade, the total tree length is Figure3 presentsa 50%majority rule con- eight steps (i.e., tree length = 172) morethan sensus of nine most parsimonioustrees re- the total tree length in Fig. 3, while the con- sultingfrom a maximum-parsimonyanalysis sistency index is smaller (=0.67). Figure 4 of the mtl6S rRNA gene sequences. The shows a neighbor-joiningtree basedon a dis- weighing of transitionto transversionin the tance matrixcalculated by using the two-pa- analysis is 1 to 5. Parsimonyanalysis based rametermethod of Kimura(1980). Five hun- on transversionsalone gave a topology con- dred bootstrapreplications of the neighbor- sistentto thatusing both transitions and trans- joining analysis gave a topology similar to versions.Maximum-parsimony, using the her- thatof Fig. 3 when brancheswith confidence mit crabas an outgroup,yielded nine shortest levels less than 50% were collapsed. The trees with tree length of 164 and consistency bootstrappedneighbor-joining analysis indi- indexof 0.707. The nextsix shortesttrees have cates strong support (BootstrapProportion a tree length of 165 and a consistencyindex [BP] = 84%) for Homarinus being a sister of 0.703. The differencesin topology among taxon to the Homarus-Nephrops clade. In the all these trees were the relative positions of maximum-parsimonyanalysis, the two spe- taxawithin a clade containingall clawed lob- cies of Nephropsisformed a sisterclade with sters. However,in this largerclade, the spe- low confidencelevel (BP = 52%, Fig. 3) to TAM AND KORNFIELD: PHYLOGENETICS OF CLAWED LOBSTE R GENERA 143

.007 gammarus

'"- ... Nephropsis aculeata .042 .080 99 .014 99 Nephropsis stewarti

Metanephrops mozambicus

.066 Scyllarides nodifer .032

Panulirus longipes .087 Pagurus pollicaris Fig. 4. Neighbor-joining tree for lobsters based on partial mitochondrial 16S rRNA gene sequences. Numbers above branches are branch lengths. Numbers below branches indicate confidence levels (%) for branch length. all other clawed lobsters. This relationship was the clade (Table 2) could not be resolved with not resolved in the neighbor-joining analysis confidence. Another strong clade was formed (BP = 47%). Maximum-likelihood analysis by the two species of Nephropsis. Homarinus also yielded a consistent tree topology as did was always excluded from the Homarus- both maximum-parsimony and neighbor-join- Nephrops clade, but its position as sister to ing analyses, after collapsing branches whose this clade, supported by neighbor-joining confidence limits overlapped zero. analysis, was ambiguous under maximum- In order to see the effect of including the parsimony. All clawed lobsters formed a sig- two highly variable regions on the outcome nificant (BP = 100, Fig. 3; BP = 99, Fig. 4) of the cladistic analysis, all available data clade relative to both spiny and slipper lob- (i.e., 474 bp of the mtDNA 16S gene) were sters, which together formed a clade with a subjected to maximum-parsimony analysis moderate bootstrap value under parsimony using PAUP under the same settings as the (BP = 74, Fig. 3), and with a higher support analysis without the highly variable regions. (BP = 95, Fig. 4) under neighbor-joining. The maximum-parsimony analysis yielded four shortest trees of tree length and consis- DISCUSSION tency index equal to 316 and 0.668, respec- Ornamentation, such as grooves, spines, tively. The next four shortest trees have tree and carinae of the carapace of both fossil and lengths of 317. The consensus tree is identi- extant clawed lobsters, has been used by car- cal to the consensus tree which resulted from cinologists as a clue to the phylogenetic re- the analysis without the highly variable relationships among lobsters. In the hypotheti- gions. However, the inclusion of these highly cal evolutionary scheme of astacideans of variable regions did not improve the resolu- Glaessner (1969), based on carapace mor- tion of the cladogram. phology, both Nephrops-like and Homarus- In summary, the various data analyses like lobsters diverged separately from the yielded identical tree topologies in which a Hoploparia lineage during the Middle or Late strong (BP = 93-99, Figs. 3, 4, respectively) Cretaceous. As described in the species cat- clade was formed by the two species of alogue of lobsters (Holthuis, 1991), Homarus Homarus and Nephrops, although suggested has a smooth abdomen lacking grooves and relationships among these three taxa within spines and smooth first chelipeds without 144 JOURNALOF CRUSTACEANBIOLOGY, VOL. 18, NO. 1, 1998

ridges, while Nephrops has a grooved ab- cal requirement.Tshudy (1993) examined40 domen and grooved first chelipeds with morphologicalcharacters for cladistic analy- spines. In addition, the first chelipeds of sis of the clawed lobster families Chileno- Homarusare wide and thick, while those of phoberidaeand Nephropidae. Twenty-nine of Nephropsare slender and much longer than the 40 characterswere informativephyloge- wide. In additionto havinga small body size, netically.Among these 29 informativechar- Homarinus capensis has a smooth body with acters,only nine were reliable indicators of phy- first chelipeds that are fully covered with logeny (i.e., were not homoplasic).Tshudy's hairs. The shape of its body and the smooth (1993) cladisticanalysis suggests that the ex- morphology of its abdomen and chelipeds isting, intuitivesuprageneric classification of have been the basis for placingit in the genus clawed lobstersis phylogeneticallyincorrect. Homarus.However, additional distinct mor- The present study, which used molecular phologicalcharacters, such as the presenceof charactersas an alternativeapproach to mor- a dense coat of setae on the outer surfaceof phological characters,provides independent the first chelipeds and scattered setae dis- clues aboutthe phylogenyof clawedlobsters. tributedover otherbody parts,and extensive The most parsimonious molecular trees genetic divergence (Kornfield et al., 1995) (Fig. 3) reveal a significantclade formedby suggest that this species constitutes a sepa- all clawed lobsters in the present study. rate genus. The exact phylogeneticposition Within this group, Nephrops and Homarus of Homarinus capensis remains enigmatic, al- forma well-supportedinternal clade (bp = 93, thoughthe distinctionof this taxonfrom spe- Fig. 3) that excludes Homarinus.The exact cies of Homarusis clear and significant.All phylogenetic position of Homarinusis am- species of Metanephropshave previously biguous, althoughthe neighbor-joiningtree been regarded as belonging to Nephrops. suggests (bp = 84%) that Homarinus is a Based on the relativesize of the left andright sister taxon to the Homarus-Nephrops clade. first chelipeds, the size and abundanceof In general habitus, Homarinus is more spines on the carapace,the numberof ridges Homarus-like rather than Nephrops-like. If on the carapace,and the marginsof the ros- the molecularphylogeny presentedhere re- trum,Jenkins (1972) removedall except the flectsthe "true"phylogeny of clawedlobsters, Europeanspecies fromNephrops to form the the similarities between Homarus and genus Metanephrops. Both the study of Homarinus,such as wide andthick claw with Homarinus and the study of Metanephrops a smoothpalm and smooth abdomen,would indicatethat taxonomymay change as more be due to convergenceor symplesiomorphy, specimensbecome availableand more acute not synapomorphy.The disparityin genetic examinationsof the comparativemorphology and morphological divergence between of clawed lobstersare undertaken.This sup- Homarus and Homarinus illustrates that taxa ports the idea that morphologicalcharacters that are genetically divergentneed not pre- used in taxonomicclassification may or may sent obvious morphologicalautapomorphies. not contain phylogenetic signals. Further- It would appearthat the groovedpalms with more, such changesin taxonomicranking do spines and ridges shared by Nephrops and not automatically support previous or Metanephropsare due to convergencerather prevailing hypotheses of phylogenetic re- than symplesiomorphy.The formerexplana- lationships, for example, that Homarinus tion requirestwo independentcharacter-state must be closely relatedto Homarus,or that changes, while the latterrequires three. This Metanephrops is a sister genus to Nephrops. hypothesis assumes that the plesiomorphic In the case of Nephrops norvegicus, the in- stateis smoothpalms lacking ornamentation. equalityin size of the firstchelipeds is a char- Ourresults supportthe decision by Holthuis acter shared also with both Homarus and (1974) to retainHomarus-like lobster taxa in Homarinus, but not with either Metanephrops the Nephropinae,instead of erectinga sepa- or Nephropsis.On the otherhand, the distinct rate subfamilyHomarinae. The phylogenetic grooves and ridges on the chelipeds, cara- affinities of Homarinus remain tentative, pace, and abdomen in Nephrops are also since the presentstudy cannot completely re- found in Metanephrops, but not in Homarus solve the phylogenetic relationshipsamong or Homarinus. Homarinus, Metanephrops, and the Homarus- In cladistic analysis, homology is a criti- Nephrops clade. While Homarinus was not TAMAND KORNFIELD:PHYLOGENETICS OF CLAWEDLOBSTER GENERA 145

included in the study by Tshudy (1993), need not presentextensive morphological au- Nephrops and Metanephrops were grouped tapomorphies.Taxonomic classification of together in the same clade which excluded clawed lobsters based solely on superficial Homarusin his study. Thus, there is incon- morphological characters does not reflect gruence between morphologicaland molec- phylogeneticrelationships, and studies on fos- ular data in inferring clawed-lobster phy- silized specimenscan give only hypothetical logeny, althoughboth suggestedthat the cur- resultsthat need to be testedor complemented rent taxonomy lacks phylogenetic basis. In by otherapproaches. Both morphologicaland orderto confirmand to improvethe resolution molecularapproaches complement each other of the phylogenetic relationships among and are needed to infer phylogeny. clawed-lobstergenera, further studies involv- ing othermorphological or geneticcharacters, ACKNOWLEDGEMENTS including longer sequences of the mtl6S We thank Richard Lord (Guernsey), Tow Cross (Ire- rRNA gene or othergenes, are necessary. land), Elizabeth Bourke (Ireland), Brian Flawn (South Africa), and Roy Melville Smith (South Africa) for as- Paleontologistsuse morphologicalcharac- sistance in obtaining specimens for study. We are grate- ters and stratigraphicrecords of extinct and ful to Robert Steneck for encouraging and supporting our extanttaxa to constructphylogeny. However, research on lobsters. We thank Alexander N. Parker for there are many limitationson the use of fos- providing advice on data analysis. We benefited from limited numberof early discussions with Dale Tshudy, as well as from later sils, including specimens, comments on the manuscriptby him and a colleague. This the degreeof preservationor completenessof research was supported in part by NOAA Sea Grant the fossils, and the difficulty in identifying (NA909AAD-SG499 and NA36RG0110) and NSF homologous characters for comparison. (DEB952183). Therefore,additional approaches are needed. LITERATURE While molecularapproaches may not be gen- CITED erally applicableto fossil crustaceans,both Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, and molecular J. G. Seidman, J. A. Smith, and K. Struhl (eds.). 1989. morphological approaches Short protocols in molecular biology.-John Wiley & should be applied, if possible, when con- Sons, New York, New York. Pp. 1-387. structing phylogenies. Parker (1997) gave Avise, J. C. 1994. Molecular markers,natural history and an enlighteningdiscussion of the advantages evolution.-Chapman and Hall, New York, New York. and of either Pp. 1-511. disadvantages using morpho- Boore, J. L., T. M. Collins, D. Stanton, L. L. Daehler, and logical or molecularapproaches to phyloge- W. M. Brown. 1995. Deducing the pattern of arthro- netic studies,as well as the utilityof the com- pod phylogeny from mitochondrial DNA arrange- bined approach. De Queiroz et al. (1995) ments.-Nature 376: 163-165. gave a comprehensivereview of arguments Cabot, E. L., and A. T. Beckenbach. 1989. Simultane- in favorof each of these views. The "totalev- ous editing of multiple nucleic acid and protein sequences with ESEE.-Computation and Application in idence"approach is currentlydebated, and it Biosciences: Cabios 5: 233-234. is unclear whether different sets of data Chu, K. H., S. C. Lee, and S. C. M. Tsoi. 1990. Enzyme should be analyzed separatelyor combined polymorphism in three species of Metanephrops (De- and analyzedsimultaneously. It has been ar- capoda: Nephropidae) from Taiwan.-Biochemical Systematics and Ecology 18: 555-558. gued that combining data sets can enhance Crandall, K. A., and J. F. Fitzpatrick, Jr. 1996. Crayfish detectionof real phylogeneticgroups. How- molecular systematics: using a combination of proce- ever, combinedanalyses may also give mis- dures to estimate phylogeny.-Systematic Biology 45: leading results when there is heterogeneity 1-26. data sets. de Queiroz, A., M. J. Donoghue, and J. Kim. 1995. Sep- among arate versus combined analysis of phylogenetic evi- In conclusion, it appearsthat Nephropsis dence.-Annual Review of Ecology and Systematics closely related to Homarus, while Homarinus 26: 657-681. is outsidethe clade containingthese two gen- Felsenstein, J. 1985. Confidence limits on phylogenies: era.The resemblance an approach using the bootstrap.-Evolution 39: superficialmorphological 783-791. of Homarinus to Homarus, and Nephrops to 1993. PHYLIP (Phylogenetic Inference Pack- Metanephrops,is due to convergenceor sym- age, computer software). Version 3.5.-University of plesiomorphyand does not reflectthe pattern Washington, Seattle, Washington. of relationshipsrevealed by the mtl6S rDNA Glaessner, M. F. 1969. Decapoda.-In: R. C. Moore, ed., data herein. These results also il- Treatise on Invertebrate Paleontology. Part R. Arthro- presented poda 4 (vol. 2): R399-R628. University of Kansas, lustrate that taxa which are genetically di- Lawrence, Kansas, and Geological Society of Amer- vergent, such as Homarinus and Homarus, ica, Boulder, Colorado. 146 JOURNALOF CRUSTACEANBIOLOGY, VOL. 18, NO. 1, 1998

Hedgecock,D., K. Nelson, J. Simons, and R. Shleser. Palumbi,S., A. Martin,S. Romano,W. 0. McMillan,L. 1977. Genicsimilarity of Americanand European spe- Stice, and G. Grabowski. 1991. The simple fool's cies of the lobsterHomarus.-Biological Bulletin 152: guideto PCR.Version 2.-Department of Zoologyand 41-50. KewaluMarine Laboratory, University of Hawaii,Hon- Hillis, D. M., C. Moritz,and B. K. Mable(eds.). 1996. olulu (specialpublication). Pp. 1-23. Molecularsystematics. Second edition.-Sinauer As- Parker,A. 1997. Combiningmolecular and morphological sociates,Inc., Sunderland,Massachusetts. Pp. 1-655. data in fish systematics:examples from the Cyprino- Holthuis,L. B. 1974. The lobstersof the superfamily dontiformes.-In:T. D. Kocherand C. A. Stepien,eds., Nephropoideaof the AtlanticOcean (Crustacea:De- Molecularsystematics of fishes. Pp. 163-188. Aca- capoda).-Bulletin of MarineScience 24: 723-884. demic Press,New York,New York. . 1991. FAO species catalogue.Vol. 13. Marine , and I. Kornfield. 1996. An improvedamplifi- lobstersof the world.An annotatedand illustrated cat- cationand sequencingstrategy for phylogeneticstud- alogue of species of interest to fisheries known to ies usingthe mitochondriallarge subunit rRNA gene.- date.-FAO FisheriesSynopsis 125 (vol. 13): 1-292. Genome39: 793-797. Jaeger,J. A., D. H. Turner,and M. Zuker. 1989. Pre- Saiki, R., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. dictingoptimal and suboptimal secondary structure for Higuchi,G. T. Horn,K. B. Mullis, and H. A. Erlich. RNA.-In: R. F. Doolittle,ed., Molecularevolution: 1988. Primer-directedenzymatic amplification of DNA computeranalysis of protein and nucleic acid se- with a thermostableDNA polymerase.-Science 239: quences.-Methods in Enzymology183: 281-306. 487-491. Jenkins,R. J. F. 1972. Metanephrops,a new genus of Sanger,F., S. Nicklen,and A. R. Coulson. 1977. DNA late Plioceneto Recentlobsters (Decapoda, Nephrop- sequencingwith chain-terminatinginhibitors.-Pro- idae).-Crustaceana22: 161-177. ceedings of the National Academy of Science 74: Kimura,M. 1980. A simplemethod for estimatingevo- 5463-5467. lutionaryrate of base substitutionsthrough compara- Swofford,D. L. 1993. Phylogeneticanalysis using par- tive studies of nucleotide sequences.-Journal of simony(PAUP). Version 3.1.1.-Illinois NaturalHis- MolecularEvolution 16: 111-120. tory Survey,Champaign, Illinois. Kornfield,I., A. B. Williams,and R. S. Steneck. 1995. Tshudy,D. M. 1993. Taxonomyand evolution of the Assignmentof Homaruscapensis (Herbst, 1792), the clawed lobster families Chilenophoberidaeand Capelobster of SouthAfrica, to the new genusHomar- Nephropidae.-Ph.D. thesis, Kent State University, inus (Decapoda: Nephropidae).-Fishery Bulletin, Kent, Ohio. DissertationAbstracts International 54: UnitedStates 93: 97-102. 4055-B. Kumar,S., K. Tamura,and M. Nei. 1993. MEGA:Mo- Xiong, B., and T. D. Kocher. 1991. Comparisonof mi- lecularEvolutionary Genetics Analysis, version 1.0.- tochondriaDNA sequencesof seven morphospecies PennsylvaniaState University, University Park, Penn- of black flies (Diptera:Simuliidae).-Genome 34: sylvania16802. 306-311. Maddison,W. P., andD. R. Maddison.1992. MacClade: RECEIVED:21 1997. analysis of phylogenyand characterevolution. Ver- January ACCEPTED:13 June 1997. sion 3.-Sinauer Associates,Inc., Sunderland,Massa- chusetts.Pp. 1-398. Address:Department of Zoology and School of Marine Mertin,H. 1941. DecapodeKrebse aus dem Subhercy- Sciences,University of Maine,Orono, Maine 04469-5751, nen und BraunschweigerEmscher und Untersenon, U.S.A. (e-mail: [email protected](YKT), sowie Bermerkungenuiber einige verwandteFormen in [email protected](IK)) derOberkreide.-Nova Acta Leopoldina, n.f., 10 (68): 152-257.