JOURNALOF CRUSTACEANBIOLOGY, 14(2): 191-209, 1994

GENETIC DIF1FERENTIATION IN TRANS-FLORIDIAN COMPLEXES OF AND UCA (: BRACHYURA)

Darryl L. Felder and Joseph L. Staton

ABSTRACT Most nearshore brachyurancrabs of the northern Gulf of Mexico have been treated as conspecificsof those in warm-temperateCarolinian waters on the southeasternAtlantic coast of North America. However, historicalphysiographic constraints appear to have periodically restrictedgene flow betweennorthern Gulf populationsand siblingAtlantic coast populations, and contemporarydisjuncture of ranges often persists across south Florida. The present ex- aminationof intertidalcomplexes has centeredon grapsidcrabs presently assigned to Sesarma reticulatumand ocypodid assignedto Uca minax, 2 species in which we have observed markedvariations in colorationover their distributionalrange. Genetic differentiationbetween populationshas been assayed by allozyme electrophoresis,and resultantdata have been eval- uated with F-statistics and cluster analysis of genetic distance. Allozyme divergencebetween Gulf and Atlantic populationsof the S. reticulatumcomplex is at levels previously reported for speciatedpopulations, while that betweentrans-Floridian populations of U. minax is much less pronounced. In both species, minimal divergence can be measured between 2 widely separatedAtlantic coast sample localities that were compared,while more complex gradesof differentiationare evident between sample localities compared within the northernGulf of Mexico. Trans-Floridiandivergence of populationsfor both of these species is compatiblewith models for periods of contact and subsequentisolation of Gulf and Atlantic stocks duringand since peak glacial advances in North America. Less conspicuouspatterns of genetic differen- tiation between sample localities within the northernGulf of Mexico may reflecta history of glacialand postglacialalluvial events which resultedin contemporaryphysiography of northern Gulf estuaries.

Prior to recent reevaluations,the coastal cylindrica (Stimpson), Platychirograpsus warm-temperateCarolinian decapod crus- spectabilisde Man) and some possible iso- tacean assemblageof the westernNorth At- lates of tropical stocks (Uca spinicarpa lantic was thoughtto include a largenumber Rathbun, U. marguerita Thurman, Calli- of widely distributed and highly adaptable nectes rathbunaeContreras). Evidence of a nearshoreand intertidalbrachyuran species few other endemic nearshore brachyurans that ranged from the vicinity of Cape Hat- could be deduced from published ranges, teras,North Carolina,into the northernand but there was no clear case for relationship northwestern Gulf of Mexico (Hedgpeth, of these or other Gulf endemic species to 1953; Williams, 1965; Powers, 1977). Even Atlantic coast siblings and no documenta- in cases where distributionsof species were tion of trans-Floridian geminate species known to break at the southern extreme of pairs in this group (Williams, 1965; Felder, peninsular Florida, close morphological 1973; Williams, 1974; Powers, 1977). similaritybetween disjunct populations led Evidence for historical disjuncture of most investigatorsto conclude that little or coastalbrachyuran stocks acrossthe Florida no differentiationhad occurredbetween such peninsula and ultimate speciation of those populations. These breaks in distribution stocks has accumulatedin the course of sev- were usually assumed to representnothing eral case studies from the early 1970s to the more than a contemporary disjuncture of present. In his separation of the Atlantic appropriatehabitats which accounted for a xanthidcrab Dyspanopeus sayi (Smith)from short break in an otherwise continuous the closely related Gulf of Mexico endemic range. While endemism of nearshore spe- species D. texanus(Stimpson), Abele (1972: cies was documented within some brachy- 269) discussedthe probablerole of Miocene urangenera in the Gulf of Mexico, evidence throughPleistocene events which alternate- was limited primarily to certain conspicu- ly submergedand exposed the Florida pen- ous cases of apparentrelict species (Uca sub- insula and which could have reduced gene 191 192 JOURNAL OF BIOLOGY, VOL. 14, NO. 2, 1994 flow between Gulf and Atlantic coast pop- of appropriateestuarine habitats for these ulations of the Dyspanopeuscomplex (treat- species across southern Florida. There is, ed within Neopanope prior to Martin and however, some differencein early life his- Abele, 1986) and of "other species pairs tory and capacityfor dispersalbetween typ- among the Decapoda with similar distri- ical populations of the U. minax (five zoeal bution patternsalong the Gulf and Atlantic stages)and S. reticulatum(three zoeal stages) coasts." Subsequently, additional closely complexes, and some differencein habitat related species pairs have been identified preferencebetween U. minax (low salinity that may have divergedfrom stocks isolated marshesand banks)and S. reticulatum(me- to either side of historical or contemporary dium to high salinity marshes and banks) distributionalbreaks in south Florida;these complexes. include such Gulf/Atlantic pairings as Uca Much as color differencesfirst prompted panacea Novak and Salmon/U. pugilator further study and eventual recognition of (Bosc) (Novak and Salmon, 1974), Ovalipes sibling species in the Menippe mercenaria/ floridanus Hay and Shore/O. stephensoni M. adina complex (Williams and Felder, Williams (Williams, 1976), Panopeussimp- 1986), our observationof differencesin col- soni Rathbun/P. herbstii Milne Edwards oration between trans-Floridian popula- (Williams, 1984a), and Menippeadina Wil- tions of both the U. minax and the S. re- liams and Felder/M. mercenaria(Say) (Wil- ticulatum complexes has prompted us to liams and Felder, 1986). Even where con- undertakegenetic characterizations,despite temporaryranges for some of the above may our finding of otherwise striking morpho- now overlap in the northeastern Gulf, as logical similarities between Atlantic and in the case of Uca panacea/U. pugilator Gulf of Mexico populations.However, since (Novak and Salmon, 1974; Barnwell and in both complexes some color variations Thurman, 1984), or where, as in the Menip- were also observed between geographically pe complex,introgressive hybridization may separated populations within the Gulf of occur (Bert, 1986; Williams and Felder, Mexico, an efforthas been made to compare 1986; Bert and Harrison, 1988), historical populations of both species from through- events which separated stocks across the out their ranges in the Gulf of Mexico. Fi- Florida peninsula may have facilitated ge- nally, our interestin genetic examination of netic isolation and subsequent speciation. the S. reticulatumcomplex has also been Recognition of divergent features be- promptedby recent studies of reproductive tween closely relatedtrans-Floridian sibling biology (Zimmerman and Felder, 1990, species and populations of the Brachyura 1991) and osmoregulatoryability (Staton has in various instances been facilitatedby and Felder, 1992) that have uncovered ev- comparison of morphology,coloration, be- idence of functional divergence between havior, physiology, hemocyanin electro- Gulf of Mexico and Atlantic populations of morphs, allozyme frequencies,and nucleic this group. acid sequences(as referencedabove, see also Fielder et al., 1971; Selanderet al., 1971a; MATERIALSAND METHODS Salmon et al., 1979; Sullivan et al., 1983; for genetic analyses were collected over a Neigel et al., 1991). In the presentstudy, we 2-year period (16 May 1989-1 August 1991) from lo- undertake allozyme analyses in trans-Flo- calities at intervals along the southeastern Atlantic and ridian that are treated Gulf of Mexico coasts of the United States (Fig. 1). complexes presently Within this area, we have concluded that the Sesarma as two brachyurantaxa, the fiddler reticulatum complex is reliably documented to range (Ocypodidae) Uca minax (Le Conte) and in the Gulf of Mexico from Barra del Tordo, Tamau- the marsh crab ()Sesarma reti- lipas, Mexico (Rabalais et al., 1989; Zimmerman and culatum(Say). The selection of these species Felder, 1991) to Sarasota County, Gulf coast of Florida (Abele, 1973), and from Volusia County, Atlantic coast is based upon (i) their classical and contin- of Florida, to Woods Hole, Massachusetts (Abele, 1992). ued treatment as Carolinian species that We have concluded that the Uca minax complex ranges range into the Gulf of Mexico (Hedgpeth, in the Gulf of Mexico from near the Neches River at 1953; Barnwelland Thurman, 1984; Abele the eastern extreme of the Texas coast (Wurtz and and the in dis- Roback, 1955) into northwestern Florida, perhaps to Kim, 1986), (ii) disjuncture Yankeetown, Gulf coast of Florida (Salmon, 1967); tribution of both of these species across precise records for both eastern and western extremes southern Florida, and (iii) the disjuncture of the Gulf range are questionable, since the species FELDER AND STATON: BRACHYURAN GENETICS 193

Sesarma reticulatum complex

PINC - Pivers Island, North Carolina (28) 24 Oct 89 NBGA - north of Brunswick, Georgia (28) 16 May 89 SUFL - Suwannee, Florida (13) 26 Apr 90 CAFL - west of Carabelle, Florida (30) 14 Nov 90 WBFL- West Bay, Florida (30) 26 Apr 90 DIAL- Dauphin Island causeway, Alabama (30) 9 Aug 90 PINC PRMS - tributary of Pascagoula River, Mississippi (30) 27 Oct 89 COLA - Cocodrie, Louisiana (30) 24 Jul 89 CPLA - Cypremort Point, Louisiana (30) 22 Jun 89 SPTX - Sabine Pass, Texas (30) 24 Oct 89 BSTX - Bayside, Texas (30) 3 Aug 90 RGTX - bank of Rio Grande River, Texas (16) 14 Jun 89

NBGA PRMS n WBFL %k,CsPCPLA N'.A \L PTX ICO SUFL

BSTX

RGTX

Uca minax complex

PINC - Pivers Island, North Carolina (30) 24 Oct 89 P NBGA - north of Brunswick, Georgia (30) 16 May 89 OSFL - Overstreet, Florida (30) 14 Nov 89 MBAL - upper Mobile Bay, Alabama (30) 9 Aug 89 PRMS - bank of Pascagoula River, Mississippi (30) 27 Oct 89 MRLA- mouth of Mississippi River, Louisiana (14) 24 May 1990 DULA - Dulac, Louisiana (30) 24 Jul 89 CPLA - Cypremort Point, Louisiana (30) 19 Jun 89

PRMS fMBAL NBGA CPLA' OSFL " DUL

MRLA

Fig. 1. Sample localities for specimens of the and Uca minax complexes. Locality ab- breviations listed are each followed by site definition, sample size (N), and date of collection. has often been confused with a closely related but dis- of Mexico were selected to encompass as much of the tinct Gulf endemic species, Uca longisignalis Salmon distributional ranges as possible, and to represent sub- and Atsaides (see Salmon and Atsaides, 1968; von populations that were known to vary in color or sub- Hagen, 1976; Heard, 1977; Williams, 1984b). On the populations that appeared to be segregated by features Atlantic coast, U. minax appears to range from north- of coastal physiography. Successful sampling for both east Florida to Buzzards Bay, Massachusetts (Williams, complexes within their Gulf ranges was limited to those 1984b). localities from which adequate sample sizes could be Our collection localities along coastlines of the Gulf obtained and, therefore, did not represent extreme lim- 194 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Table 1. Enzyme systems and optimal buffer systems used in the analysis of population samples of Sesarma (S) and Uca (U). E.C. No. refers to index numbers recommended by the Nomenclature Committee of the International Union of Biochemistry (1984). Buffers used in this study were Tris-citrate pH 8.0 (TC 8) and Poulik pH 8.7 (P 8.7).

Enzyme system (locus abbreviation) E.C. No. Buffer system Acid phosphatase(Acp) 3.1.3.2 TC 8 (S) Alanopinedehydrogenase (Aladh) 1.5.1.17 TC 8 (S, U) Argininekinase (Ark) 2.7.3.3 P 8.7 (S, U) Enolase(Eno) 4.2.1.11 TC 8 (U) Glucose phosphateisomerase (Gpi) 5.3.1.9 P 8.7 (S, U) Glutamic-oxaloacetictransaminase 1 and 2 (Got) 2.6.1.1 TC 8 (2-S; 1-U) Hexokinase(Hk) 2.7.1.1 P 8.7 (S, U) Isocitricdehydrogenase 1 and 2 (Idh) 1.1.1.42 TC 8 (S, U) Lactic dehydrogenase(Ldh) 1.1.1.27 TC 8 (S, U) Leucyl-alanylpeptidase (PepLa) 3.4.13.- TC 8 (S, U) Leucyl-glycyl-glycylpeptidase (PepLg) 3.4.13.- TC 8 (S, U) Malic dehydrogenase1 and 2 (Mdh) 1.1.1.37 TC 8 (S, U) Mannosephosphate isomerase (Mpi) 5.3.1.8 P 8.7 (S, U) 6-phosphogluconicdehydrogenase (Pgdh) 1.1.1.49 TC 8 (S, U) Phosphoglucomutase(Pgm) 2.7.5.1 P 8.7 (S, U) Sorbitoldehydrogenase 1 and 2 (Sdh) 1.1.1.14 P 8.7 (2-S, 1-U) Superoxidedismutase (Sod) 1.15.1.1 TC 8 (S, U) its of the reportedranges. Collection localities for both zymes produced20 presumptiveloci for Sesarma and complexeson the Atlanticcoast werelimited to 2 wide- 18 presumptiveloci for Uca that producedconsistently ly separatedsites, and animals of respective species interpretableresults (Table 1).Allelic forms of enzymes from both sites were similarin coloration.Crabs were were designatedby letters, and differentloci for the capturedwhile they were active outside their burrows same enzyme system were designatedby Arabic nu- or were excavatedat low tide. An effort was made to merals. collectat least 30 individualsfrom each localityfor the Allelic frequencies and observed heterozygosities analysis, and additional intact specimens along with (HOBS)for each locus weredetermined by directcensus dissected remains were usually preservedas morpho- of the population data. Mean heterozygositywithin logical vouchers. Collection localities are identified populationswas estimatedfrom averagevalues across throughoutthe remainderof this paper by the abbre- all loci. HierarchicalF-statistics (Weir and Cockerham, viations listed in Fig. 1. 1984) were estimatedfor 3 levels of populationstruc- Tissue sampleswere either frozenin liquid nitrogen ture:individuals within localities (FIL), localities within in the field,or weredissected from animalstransported 2 regions,the Gulf coast and the Atlantic coast (FLR), live to the laboratory;all tissue sampleswere storedat and the 2 regions within the total area sampled (FRT, -70?C until they were processedprior to electropho- see Table 2). Theoretically,F values range from 0-1 resis. Prior to homogenization,subsamples of tissue with F = 1 being groups fixed for alternate alleles. (excluding hepatopancreas)were placed in 1.5-ml Preliminaryruns of the data producedsmall negative microcentrifugetubes and refrozenwith an equal vol- values, and these were set to zero in subsequentanal- ume of distilled water.These frozenpreparations were yses as suggestedby Slatkinand Barton(1989). Nei's homogenizedand then centrifugedfor 20 min at 15,000 (1978) unbiased genetic distance (DN) was calculated g. Samples were electrophoresed in 12.5% starch and then clustered using the unweighted pair-group (StarchartCorp., Smithville, Texas, U.S.A.) underthe algorithm(UPGMA, Unweighted Pair-Group Method conditions of Selanderet al. (1971b) and Murphy et Using ArithmeticAverages; Sneath and Sokal, 1973) al. (1990). An initial screeningof more than 35 en- and the neighbor-joiningalgorithm (Saitou and Nei,

Table 2. HierarchicalF-statistics adapted from Weir and Cockerham(1984) design that utilizes a random- effectsnested ANOVA model (afterKwast et al. 1990). FRTis genetic differentiationamong regionswithin the total area sampled,whereas FLR and FILare the levels of genetic differentiationof localities within regionsand individualswithin localities, respectively.

Level of variation Variance component F-statistics = Regions (R) within total (T) a2RT FRT RT/CGT = Localities (L) within regions a2LR FLR LR/(OGTT - RT) = - Individuals (I) within localities Cr2lL FIL 2IL/(72GT -- RT a2LR) Genes (G) within individuals a2G, Total 02GT FELDER AND STATON: BRACHYURAN GENETICS19 195

Table 3. Allele frequenciesand heterozygositiesfor 13 loci in 12 collectionsof the Sesarmareticulatum complex. Alleles are identifiedby lower case letter;multiple loci are identifiedby Arabicnumeral. HOB,s and HExp represent observedand expected(unbiased) heterozygote frequencies, respectively. " -" indicatesabsence of heterozygotes for localities with only homozygotes.Locality abbreviations and sample sizes are given in Fig. 1.

Locality Locus PINC NBGA SUFL ARFL WBFL DIAL PRMS LCLA CPLA SPTX BSTX RGTX Acp-2 a 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 b 1.000 0.982 1.000 1.000 1.000 1.000 1.000 0.983 1.000 1.000 1.000 1.000 C 0.000 0.018 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 HOBS - 0.036 0.033 - HEXP - 0.036 0.033 - Aladh a 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.937 b 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.063 HOB 0.000 HExP 0.121 Got-I a 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.0 17 0.000 0.000 0.017 0.000 b 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.983 1.000 1.000 0.983 1.000 HuBS 0.033 - - 0.033 HEXP 0.033 0.033 Got-2 a 0.000 0.000 0.000 0.000 0.000 0.033 0.000 0.017 0.033 0.000 0.017 0.031 b 0.018 0.000 0.577 0.552 0.565 0.934 0.850 0.866 0.850 0.850 0.950 0.875 C 0.000 0.000 0.000 0.000 0.000 0.033 0.000 0.000 0.000 0.000 0.033 0.000 d 0.964 0.982 0.423 0.43 1 0.435 0.000 0.150 0.117 0.117 0.150 0.000 0.094 e 0.018 0.018 0.000 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 HuBS 0.07 1 0.036 0.385 0.552 0.419 0.067 0.300 0.267 0.267 0.233 0.100 0.250 HEXP 0.071 0.036 0.508 0.518 0.500 0.129 0.259 0.238 0.267 0.259 0.098 0.232 Gpi a 0.018 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 b 0.000 0.000 0.000 0.000 0.016 0.017 0.000 0.017 0.017 0.017 0.017 0.000 C 0.000 0.000 0.654 0.793 0.758 0.800 0.883 0.850 0.733 0.900 0.900 0.719 d 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 e 0.804 0.857 0.346 0.207 0.226 0.183 0.117 0.116 0.250 0.083 0.083 0.281 f 0.178 0.143 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 HuBS 0.250 0.214 0.538 0.414 0.484 0.400 0.167 0.267 0.533 0.133 0.200 0.438 HExP 0.328 0.249 0.47 1 0.334 0.380 0.332 0.210 0.268 0.406 0.186 0.186 0.417 idh-J a 0.000 0.000 0.000 0.103 0.032 0.000 0.033 0.000 0.117 0.167 0.100 0.000 b 1.000 1.000 1.000 0.896 0.968 1.000 0.967 1.000 0.883 0.816 0.900 1.000 C 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 HuBS 0.206 0.000 - 0.067 - 0.167 0.300 0.067 HExP 0.189 0.063 - 0.066 - 0.210 0.310 0.183 idh-2 a 1.000 0.964 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 b 0.000 0.036 0.000 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 C 0.000 0.000 1.000 0.983 0.984 1.000 1.000 1.000 1.000 1.000 0.983 1.000 d 0.000 0.000 0.000 0.000 0.016 0.000 0.000 0.000 0.000 0.000 0.017 0.000 HuBS - 0.071 0.034 0.032 - 0.033 HExP - 0.070 0.034 0.032 - 0.033 mdh-2 a 0.000 0.018 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 b 1.000 0.982 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.983 1.000 HOBS - 0.036 - 0.033 HEXP - 0.036 - 0.033 196 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Table 3. Continued.

Locality Locus PINC NBGA SUFL ARFL WBFL DIAL PRMS LCLA CPLA SPTX BSTX RGTX

PepLa-2 a 0.000 0.036 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 b 1.000 0.964 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 c 0.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 HOBS 0.000 HEXP 0.070 PepLg-I a 1.000 1.000 0.077 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 b 0.000 0.000 0.923 0.000 0.000 0.033 0.000 0.000 0.000 0.000 0.000 0.000 c 0.000 0.000 0.000 1.000 1.000 0.967 1.000 1.000 1.000 1.000 1.000 1.000 HOBS 0.000 0.000 HEXP 0.148 0.066 Pgdh a 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.063 b 0.982 0.929 1.000 1.000 1.000 1.000 1.000 0.967 0.967 0.967 1.000 0.937 c 0.018 0.071 0.000 0.000 0.000 0.000 0.000 0.033 0.033 0.033 0.000 0.000 HOBS 0.036 0.071 0.000 0.000 0.067 0.000 HEXP 0.036 0.135 0.126 0.065 0.065 0.121 Pgm a 0.000 0.000 0.000 0.017 0.000 0.017 0.017 0.017 0.000 0.000 0.000 0.031 b 0.018 0.000 0.308 0.310 0.516 0.717 0.783 0.833 0.733 0.750 0.867 0.813 c 0.982 1.000 0.692 0.638 0.452 0.266 0.150 0.133 0.200 0.217 0.133 0.156 d 0.000 0.000 0.000 0.035 0.032 0.000 0.050 0.017 0.067 0.033 0.000 0.000 HOBs 0.036 0.461 0.483 0.613 0.367 0.333 0.300 0.433 0.367 0.200 0.250 HEXP 0.036 0.443 0.504 0.537 0.422 0.367 0.292 0.425 0.396 0.235 0.324 Sdh-2 a 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 b 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 c 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.949 1.000 1.000 1.000 1.000 d 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 0.000 0.000 0.000 HOBS 0.100 HEXP 0.098

1987). Standarderrors were calculated as describedby or nearly fixed differences in alleles for Idh- Nei et al. (1985) and displayed on UPGMA cluster 2, PepLa-2, and PepLg-1. For other loci, (+ 1 diagrams SE). the dominant allele in Atlantic samples was present in northwestern Florida samples, but RESULTS largely absent from western Gulf samples The Sesarma reticulatum complex was (e.g., for Got-2, Atlantic mean = 0.973, monomorphic across the entire range sam- Florida mean = 0.430, and western Gulf pled for seven of the 20 presumptive loci mean = 0.084). This pattern is repeated in examined (Ark, Hk, Ldh, Mdh-1, Mpi, Sdh- Got-2, Gpi, and Pgm loci. In the Uca minax 1, and Sod). In the Uca minax complex, complex, a similar pattern is distinguishable five of the 18 loci (Ark, Eno, PepLg, Mdh- for the Idh-2 and Sdh-l loci, although the 1, and Sod) tested were monomorphic for Gulf localities do not show a distinct east the entire range sampled. Many loci pos- to west separation. However, the extreme sessed local rare alleles with frequencies western Gulf populations from Louisiana <5% (e.g., Acp-1, Got-i, Mdh-2, and Sdh- do exhibit a distinct allelic shift from Idh- 2 for the S. reticulatum complex (Table 3); 2d to Idh-2c. Aladh, Hk, Idh-1, Mdh-2, Mpi, PepLa, and While genetic separation in the S. reti- Pgdh for the U. minax complex (Table 4)). culatum complex is greater between Gulf Atlantic and Gulf of Mexico samples of the and Atlantic regions than in the Uca minax S. reticulatum complex demonstrated fixed complex, U. minax possessed a greater av- FELDER AND STATON: BRACHYURAN GENETICS 197

Table 4. Allele frequenciesand heterozygositiesfor 13 loci in eight collections of the Uca minax complex. Alleles are identified by lower case letter; multiple loci are identified by Arabic numeral. HOBSand HEXprepresent observed and expected (unbiased) heterozygote frequencies, respectively. "-" indicates absence of heterozygotes for localities with only homozygotes.Locality abbreviationsand sample sizes are given in Fig. 1.

Locality Locus PINC NBGA OSFL MBAL PRMS SPMR DULA CPLA Aladh a 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.933 b 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.017 c 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.050 HOBS - - - 0.033 HEXP - - - - 0.128 Got-2 a 1.000 1.000 0.983 1.000 0.967 1.000 0.933 0.983 b 0.000 0.000 0.017 0.000 0.033 0.000 0.067 0.017 - HOBS 0.033 - 0.067 - 0.133 0.033 HEXP - - 0.033 - 0.065 - 0.126 0.033 Gpi a 0.000 0.000 0.051 0.000 0.017 0.036 0.000 0.000 b 0.167 0.167 0.283 0.150 0.200 0.321 0.117 0.267 c 0.650 0.716 0.433 0.567 0.400 0.357 0.567 0.567 d 0.183 0.117 0.233 0.283 0.383 0.286 0.316 0.166 HOBs 0.600 0.533 0.800 0.667 0.600 0.786 0.567 0.567 HEXP 0.524 0.453 0.686 0.586 0.664 0.712 0.574 0.590 Hk a 1.000 0.967 1.000 1.000 1.000 1.000 1.000 1.000 b 0.000 0.033 0.000 0.000 0.000 0.000 0.000 0.000 HOBS - 0.000 HEXP - 0.065 Idh-I a 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 b 1.000 1.000 0.950 0.983 1.000 1.000 1.000 0.983 c 0.000 0.000 0.033 0.017 0.000 0.000 0.000 0.017 HOBS - - 0.100 0.033 - - - 0.033 HEXP - - 0.098 0.033 - - - 0.033 Idh-2 a 0.000 0.000 0.000 0.000 0.000 0.000 0.033 0.050 b 0.000 0.000 0.000 0.000 0.000 0.000 0.033 0.000 c 0.000 0.000 0.083 0.033 0.050 0.036 0.350 0.783 d 0.033 0.000 0.467 0.684 0.567 0.643 0.334 0.167 e 0.950 1.000 0.450 0.283 0.383 0.321 0.250 0.000 f 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 HOBS 0.100 - 0.400 0.300 0.667 0.428 0.333 0.267 HEXP 0.098 - 0.582 0.459 0.538 0.500 0.713 0.362 Ldh a 1.000 1.000 1.000 1.000 1.000 0.929 1.000 1.000 b 0.000 0.000 0.000 0.000 0.000 0.071 0.000 0.000 HOBS - 0.000 - - HEXP - 0.137 - - Mpi a 1.000 1.000 0.983 1.000 1.000 1.000 1.000 1.000 b 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 HOBS 0.033 -- HEXP 0.033 - 198 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Table 4. Continued.

Locality Locus PINC NBGA OSFL MBAL PRMS SPMR DULA CPLA

Mdh-2 a 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 b 0.967 0.967 0.983 0.983 1.000 1.000 1.000 1.000 c 0.000 0.033 0.000 0.017 0.000 0.000 0.000 0.000 d 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 e 0.017 0.000 0.000 0.000 0.000 0.000 0.000 0.000 HOBS 0.067 0.067 0.033 0.033 HEXP 0.066 0.065 0.033 0.033 PepLa a 1.000 1.000 1.000 1.000 1.000 1.000 0.983 1.000 b 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.000 HOBS - - 0.033 - HEXP - - 0.033 - Pgdh a 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 b 0.983 1.000 0.966 1.000 0.983 1.000 1.000 1.000 c 0.017 0.000 0.000 0.000 0.017 0.000 0.000 0.000 d 0.000 0.000 0.017 0.000 0.000 0.000 0.000 0.000 HOBS 0.033 - 0.067 - 0.033 - - - HEXP 0.033 - 0.066 - 0.033 - - - Pgm a 0.033 0.000 0.000 0.017 0.017 0.036 0.000 0.033 b 0.000 0.000 0.000 0.000 0.000 0.000 0.067 0.000 c 0.834 0.900 0.966 0.833 0.867 0.893 0.717 0.900 d 0.100 0.100 0.000 0.083 0.000 0.000 0.133 0.067 e 0.033 0.000 0.017 0.050 0.066 0.071 0.083 0.000 f 0.000 0.000 0.017 0.017 0.050 0.000 0.000 0.000 HOBS 0.333 0.133 0.067 0.333 0.267 0.214 0.333 0.200 HEXP 0.298 0.183 0.066 0.300 0.246 0.204 0.465 0.187 Sdh-1 a 0.000 0.000 0.000 0.000 0.000 0.000 0.017 0.017 b 0.000 0.050 0.450 0.450 0.350 0.500 0.400 0.517 c 1.000 0.933 0.533 0.533 0.650 0.500 0.566 0.433 d 0.000 0.017 0.017 0.017 0.000 0.000 0.017 0.017 HOBS 0.133 0.500 0.633 0.500 0.571 0.700 0.567 HEXP 0.128 0.521 0.521 0.463 0.518 0.527 0.553 erage number of alleles (1.64 ? 0.08) and allele at that locus. The overall jackknife a greatermean frequencyof polymorphism means (Weir and Cockerham,1984) for FIL (0.215) within populations than did the S. were similar for the two species. The U. reticulatumcomplex (.38 ? 0.05 and 0.167, minax complex possessed a significantly respectively, Table 5). Observed and ex- higherjackknife mean FLRthan did the S. pected heterozygosities were in relative reticulatum complex (a = 0.05, Student's agreementfor both complexes and suggest t-test). The major component of variation no heterozygote deficiencies within popu- in S. reticulatum was that between regions lations for both species. Mean observedhet- (FRT); that of U. minax was not statistically erozygosities (HOBS? SE) summed across different between regions (FRT)or among all individuals and loci were 0.130 ? 0.003 populations (FIL) according to jackknife for the S. reticulatum complex and 0.134 + means. The S. reticulatum complex pos- 0.004 for the U. minax complex. sessed much higher genetic differentiation Individual loci demonstrated high het- on either side of the Floridapeninsula (jack- erozygote deficiencies (large FIL values, Ta- knife mean FRT = 0.795 ? 0.090) than did ble 6) which were attributableto single in- the U. minax complex (jackknifemean FRT dividuals that were homozygous for a rare = 0.118 + 0.062). FELDER AND STATON: BRACHYURAN GENETICS 199

Table 5. Estimation of genetic variability for 20 loci within 12 demes of the Sesarma reticulatum complex and 18 loci within eight demes of the Uca minax complex. Na = mean number alleles per locus + standard error; frequency of polymorphism determined by the 95% criterion (locus is polymorphic if frequency of most common allele does not exceed 0.95). Observed (HOBS)and expected (HEXp)heterozygote frequencies are means summed across all loci + standard error.

Frequency Collection Na polymorphism HOBS HEXP

Sesarma reticulatum complex PINC 1.30 + 0.17 0.050 0.020 ? 0.005 0.023 ? 0.015 NBGA 1.35 + 0.11 0.100 0.023 ? 0.006 0.032 ? 0.020 SUFL 1.20 + 0.09 0.200 0.069 ? 0.016 0.078 ? 0.021 ARFL 1.40 ? 0.18 0.200 0.084 + 0.011 0.079 + 0.017 WBFL 1.35 + 0.15 0.150 0.077 + 0.011 0.076 ? 0.013 DIAL 1.35 ? 0.17 0.150 0.042 ? 0.008 0.047 ? 0.018 PRMS 1.30 ? 0.16 0.150 0.043 ? 0.008 0.045 ? 0.019 COLA 1.70 ? 0.25 0.200 0.050 ? 0.009 0.054 ? 0.023 CPLA 1.40 ? 0.17 0.200 0.070 ? 0.010 0.069 ? 0.021 SPTX 1.40 ? 0.17 0.200 0.055 ? 0.009 0.061 ? 0.022 BSTX 1.45 ? 0.15 0.150 0.033 ? 0.007 0.040 ? 0.021 RGTX 1.35 ? 0.15 0.250 0.047 ? 0.012 0.061 ? 0.029 Uca minax complex PINC 1.55 ? 0.23 0.111 0.063 + 0.010 0.057 ? 0.020 NBGA 1.39 ? 0.16 0.167 0.048 ? 0.009 0.050 ? 0.020 OSFL 1.89 ? 0.24 0.167 0.113 ? 0.014 0.118 ? 0.018 MBAL 1.67 ? 0.27 0.222 0.111 ? 0.013 0.107 ? 0.019 PRMS 1.61 + 0.24 0.222 0.118 ? 0.014 0.112 ? 0.018 MRLA 1.50 ? 0.22 0.278 0.111 ? 0.020 0.115 ? 0.025 DULA 1.78 ? 0.31 0.278 0.117 ? 0.014 0.136 ? 0.019 CPLA 1.72 + 0.24 0.278 0.094 ? 0.012 0.105 ? 0.022

Table 6. F-statistics for sampled populations of the Sesarma reticulatum and Uca minax complexes. FL = level of genetic differentiation among individuals within localities (positive values indicate a reduced number of heterozygous individuals in a locality); FLR= level of genetic differentiation among localities within regions (Atlantic and Gulf of Mexico); FRT = level of genetic differentiation among regions within total range sampled. Estimates for each locus are summed across all alleles for that locus. Means over all loci are arithmetic means across all alleles and loci. * signifies locus not scored for that species. - signifies that locus is monomorphic for that species; F-statistics are undefined for monomorphic loci.

S. reticulatum complex U. minax complex Locus F,L FLR FRT FIL FL-R FRT Acp 0.001 0.000 0.008 * * * Aladh 1.000 0.035 0.000 0.744 0.031 0.000 Got-1 0.002 0.000 0.000 * * * Got-2 0.042 0.147 0.674 0.000 0.017 0.000 Gpi 0.016 0.020 0.623 0.013 0.019 0.031 Hk - - - 1.000 0.000 0.112 Idh-1 0.217 0.051 0.017 0.000 0.008 0.000 Idh-2 0.001 0.001 0.984 0.245 0.364 0.224 Ldh - - - 1.000 0.051 0.000 Mpi - - - 0.001 0.001 0.000 Mdh-2 0.001 0.000 0.004 0.000 0.000 0.023 Pgdh 0.658 0.006 0.008 0.000 0.001 0.000 PepLa 1.000 0.000 0.994 0.001 0.000 0.000 PepLg 1.000 0.052 0.994 - - - Pgm 0.033 0.160 0.534 0.047 0.027 0.002 Sdh- - - - 0.000 0.126 0.112 Sdh-2 0.001 0.009 0.000 * * * Mean 0.086 0.097 0.788 0.080 0.144 0.109 Jackknife mean 0.071 0.101 0.795 0.072 0.150 0.118 SE 0.046 0.043 0.090 0.060 0.101 0.062 200 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Average Distance Between Clusters 0.3 0.25 0.2 0.15 0.1 0.05 0 I I 1- I I ~~~~~~~~I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I I I , PINC NBGA SUFL CAFL WBFL DIAL PRMS COLA CPLA SPTX BSTX RGTX Fig. 2. Dendrogram of UPGMA cluster analysis for sample localities of Sesarma reticulatum complex based on DN derived from 20 presumptive loci. Locality abbreviations identified in Fig. 1. Error bars represent + one standard error. Long bars are truncated at 0 or left margin.

Where variation between regions was ers Island, North Carolina, and Suwannee, large, it could have served to inflate the Florida, populations and a maximum DN overall FIT, which in turn could have led to of 0.323 between Brunswick, Georgia, and overestimation of other F statistics because Bayside, Texas, populations (Figs. 2, 3). - of its relationship with FIL (FRT = (FIT However, allelic composition comparisons - FIL)/(1 IL); where R refers to regional between Atlantic populations (Pivers Island level). As only Sesarma demonstrated a large and Brunswick) were essentially the same FRT, the analysis of FLR for Sesarma was (DN = 0) over a straight line distance of repeated while artificially restricting the data greater than 1,100 km. Within the Gulf of set to only the Gulf of Mexico samples. The Mexico, smaller clusters of sample localities resultant FLR (jackknife mean 0.081 + were distinguishable within eastern and 0.013) was not markedly different from the western Gulf extremes of the population. FLRdetermined in analysis of the combined Populations of the Uca minax complex Gulf and Atlantic samples. showed some evidence of this same pattern While UPGMA and neighbor-joining of divergence between the Gulf and Atlantic analyses demonstrated similar trends in the but at much lower levels of significance and data set, some authors have argued recently to a much lesser degree (mean DN = 0.0394 for the superiority of the neighbor-joining + 0.004 for all pairwise comparisons of method (Nei, 1991). By either analysis, a trans-Florida sites) than in Sesarma, and large discontinuity was shown to occur be- they exhibited no distinct subclusters of tween Atlantic and Gulf populations of the sample localities within eastern and western Sesarma reticulatum complex (Figs. 2, 3). Gulf populations (Figs. 4, 5). The allelic composition of Atlantic and Gulf populations was found to be diverged at an DISCUSSION average DN of 0.276 ? 0.009 with a range Isolation of Atlantic populations from from a minimum DN of 0.206 between Piv- Gulf of Mexico populations of the Sesarma FELDERAND STATON:BRACHYURAN GENETICS 201

limitations of our sample size, precludeour 0.237 assigning statistical significanceto a trans- Florida separation in this species, at least 0.023 on the basis of our presentgenetic approach. However, we note from ongoing genetic 0.012 r analyses of other species of the subgenus SUFL Minuca in our laboratory(Uca longisignalis and Uca rapax (Smith); unpublished data) that low heterozygosity and small genetic ' WBFL distances may be characteristicof allozyme analyses for species in this subgenus. Ge- CPLA netic distance separating regional popula- tions of Uca minax, as in other populations SPTX of Minuca that are under study, is substan- tially less than that reported in the inter- PRMS specific comparison of two members of the subgenus Celuca, Uca speciosa (Ives) and COLA Ucaspinicarpa Rathbun (D = 0.70, I = 0.50), by Salmon et al. (1979). However, those higher values for genetic distance are also not for the U. BSTX perhaps surprising speciosa/ U. spinicarpa comparison, given striking DIAL morphological and ecological characteris- tics that also separate those two well-di- verged species (Heard, 1977; Barnwelland Thurman, 1984; DLF, unpublished data). RGTX This is not to say that correspondenceis to be found between anal- Fig. 3. Dendrogramderived for sample localities of always allozymic the Sesarma reticulatumcomplex by the neighbor- ysis and morphologyor behavior in species joining algorithmbased on 20 presumptiveloci. Lo- of Uca, especially in the subgenusMinuca. calityabbreviations identified in Fig. 1. Numbersabove Salmon and Kettler (1987) argued in par- transverse bars indicate genetic distances; distances of <0.01 are not shown. ticular for the resurrection "Uca virens" from the Uca (Minuca) rapax complex on the basis of behavioralevidence, despite ap- reticulatumcomplex is evident in our anal- parentisozymic and morphologicalidentity yses, in terms of both differential fixation of the populations in question (though we for diagnostic alleles in the separate popu- conclude that design limitations of their be- lations and in the high mean FRTcalculated havioral study leave that issue yet to be re- for all polymorphic loci. Genetic distances solved). measured between Atlantic and Gulf pop- In both the S. reticulatumand U. minax ulations of Sesarma exceed those reported complexes, Atlantic populations are nearly by Bert (1986) to separate Menippe mer- identical genetically between two widely cenaria from M. adina, and fall within the separatedlocalities, while Gulf populations range of values previously reported in in- appearto be more complexlystructured over terspecificcomparisons of many other deca- such geographicdistances. In Gulf popula- pod (see Hedgecocket al., 1982; tions of the S. reticulatum complex, this Lavery and Fielder, 1993). While our ge- structureappears limited to groupingof lo- netic analysis of Atlantic and Gulf popu- calities from the extreme eastern Gulf of lations of the Uca minax complex do not Mexico (all samples from northwest Flori- exhibit a strongdifferentiation, such as that da) separatelyfrom those rangingover the seen in populations of Sesarma, a similar remainderof the northernand westernGulf, genetic pattern is inferred where subpopu- albeit at a relativelylow level of separation. lations are separatedacross the Floridapen- Similarly, but at much lower levels of sig- insula. Low numbers of alleles and small nificance,sampled localities of the U. minax genetic distances in this complex, given the complex from the eastern to midnorthern 202 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Average Distance Between Clusters

0 .04 0.03 0.02 0.01 0I I I I I I I I I I PINC

NBGA

OSFL

PRMS

MRLA

MBAL

DULA

CPLA Fig. 4. Dendrogram of UPGMA cluster analysis for sample localities of the Uca minax complex based on DN derived from 18 presumptive loci. Locality abbreviations identified in Fig. 1. Error bars represent + one standard error. Long bars are truncated at 0 or left margin.

Gulf (western Florida to the Mississippi ground color of the carapace in almost all River) group together, though rare alleles animals taken in the Gulf of Mexico ranges from sample localities west of the Missis- from reddish brown to rust, or occasionally sippi River may indicate more isolation of dark salmon; anteroventrally on the cara- that subpopulation, either historically or by pace, on anterior and posterior surfaces at contemporary barriers to gene flow. the walking leg meri, and on much of the The pronounced genetic separation be- chelipeds, lighter shades of this ground col- tween Atlantic and Gulf populations of the or may vary between vermilion red, orange, Sesarma reticulatum complex correlates well salmon, yellow orange, or yellow. Routine with a striking difference that we observed occurrences of anomalously colored speci- in their coloration. Among several common mens in the Gulf population were limited names that have been applied to the Atlan- in our collections almost exclusively to the tic coast population, the often used "purple three Florida sites, which also tended to marsh crab" is particularly appropriate. group genetically. These anomalous color Dorsally, ground color of the carapace in morphs (alluded to by personal communi- this group ranges from deep purple to blu- cation from DLF as one of two color morphs ish-black or dark slate gray, usually with "in the Gulf of Mexico" by Abele, 1992: some spots or reticulations of white or pale 1,7) typically had a broken coverage of yellow; anteroventrally on the carapace, on ground color, with light and dark patches anterior and posterior surfaces of the walk- dorsally, and occasionally had anteroven- ing leg meri, and on much of the chelipeds, tral areas of the carapace of a deep coral, or lighter shades of this ground color may range reddish-violet coloration; however, overall to violet or wine colors. By contrast, dorsal ground color still tended to be much nearer FELDER AND STATON: BRACHYURAN GENETICS 203

In the formercase, differencesin timing and 0.031 phase-setting of reproductive events sug- gest possible adaptation to higher seasonal temperaturesof the Gulf populations, while in the lattercase osmoregulatorydifferences suggest a possible divergence in adaptive strategyto accommodate contemporaryor historical differences in the salinity struc- PRMS ture and tidal rangeof the Gulf and Atlantic habitats. In both these studies, differences between the Atlantic and Gulf populations MBAL were deemed to warranttreatment of Gulf populations under the label "Sesarma sp. (nr. reticulatum),"as opposed to S. reticu- MRLA latum which was applied to the Atlantic populations.Since, in both studies, only an- 0.017 imals from the western Gulf (Louisiana) were used for comparisonsto Atlantic pop- ulations, it is unknown whether the obser- vations well to the DULA reported apply equally eastern Gulf subpopulation(Florida). Fig. 5. Dendrogramderived for sample localities of Appropriatesalt, brackish,intermediate, the Uca minax complex by the neighbor-joiningal- and fresh marsh habitats (,Juncus, gorithm based on 18 presumptiveloci. Locality ab- Scirpus,Sagittaria, Vigna,and shrubsalong breviationsidentified in Fig. 1. Numbersabove trans- tidal stream are available verse bars indicate genetic distances;distances <0.01 banks) widely are not shown. along most of the northernand northwest- ern Gulf shoreline for colonization by the S. reticulatumcomplex. Even though larval orangeor reddishbrown, than it did shades life history is restrictedto three zoeal stages of purple or violet. While we found these in these animals (lasting about 17 days; color morphs to somewhat resemble pat- Zimmermanand Felder, 1991), close prox- terns in juvenile specimens of the tropical imity of adjacent habitats would appear to species Sesarma curacaoense de Man (as maintain gene flow between sample locali- was also reported to occur in populations ties over most of the Gulf range. Only west of S. reticulatumfrom southeasternFlorida; of Cape San Blas in northwesternFlorida, Abele, 1992) and agreethat additionalstudy and perhaps in extreme south Texas, are is needed in these rangeextremes, we do not there substantialdistances devoid of appro- find that presentlyknown differencesin col- priate habitat. The former of these breaks or, morphology, or allozymes support rec- may favor maintenance of a slight genetic ognition of eastern Gulf populations as a separationthat we observed between Flor- separatespecies from the remainderof the ida demes and the remainder of the Gulf Gulf of Mexico population. Rather, in ac- population. As for the two widely sepa- cord with the species separation suggested ratedAtlantic coast localities,both color and in our previous reports (Zimmerman and allozyme characteristicsappear to be very Felder, 1990, 1991), the color distinction similar, and this may relate to relative which supportsrecognition of separatespe- continuity of appropriatehabitat and effec- cies (Felder and Zimmerman, in prepara- tiveness of longshore dispersal within this tion) is that which separatesGulf of Mexico region. However, inclusion of additional from Atlantic populations. sample localities from the Atlantic coast In addition to divergent coloration in range of S. reticulatumwill be required in the two populations of Sesarma, possible order to more fully understand dynamics divergence at a functional level has been there. reported in reproductive biology (Zim- Relationshipof colorationto evidence for merman and Felder, 1991) and in osmo- genetic separationin the U. minax complex regulatoryability (Statonand Felder, 1992). is less apparent than in Sesarma. Typical 204 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994 coloration with strikingdeep red markings panmictic, polymorphic group distributed at articulationsbetween articles of the che- within a formerlycontinuous areaof warm- lipeds and walking legs (a pattern most de- temperatehabitat. While the contemporary veloped in males), was characteristic of expanse of appropriatehabitat is interrupt- specimens from the Atlantic population (as ed in southern Florida by calcareous sub- previously described by Williams, 1965; strates,tropical waters, and absence of typ- 1984b; Crane, 1975) and was also well de- ical warm-temperatemarsh vegetation, there veloped in animals from the three western- is evidence that salt-marshhabitats, appro- most localities (Louisiana)of the Gulf pop- priate sedimentary facies, and warm-tem- ulation (as previously noted by Felder, perateclimatic conditions formerlyextend- 1973). This articular coloration was very ed across southern reaches of the Florida weak to absent in almost all animals taken peninsula, perhaps persisting into the Ho- from Overstreet, Florida, and was weakly locene transgression(see review by Barn- developed in almost all animals comprising well and Thurman, 1984). It appears even samples from Alabama and Mississippi. In more probable that such continuous habi- comparison to our genetic grouping of the tats would have prevailedduring earlier pe- Gulf of Mexico localities, it was poorly de- riods, perhaps during a peak of glacial ad- veloped in at least three of the four localities vance, when warm-temperateenvironments that formedthe easternto midnorthernGulf were generally displaced to the south. Of group. Except for this divergence in color- particularsignificance for intertidalspecies, ation and the limited evidence for genetic winter temperaturesduring peak glacial ad- divergence that we can presently measure vance plummeted at latitudes north of 30?, between Atlantic and Gulf populations of while they fell only slightly in tropical seas the U. minax complex, little else is known (McIntyreet al., 1976), and it is thus doubt- to reflecttrans-Floridian divergence. In both ful that appropriateclimate for warm-tem- the Atlantic and Gulf populations, larval perate marsh-dwelling forms would have life history encompasses five planktonic persisted during this period either in the zoeal stages, which at least in the Gulf pop- northernGulf of Mexico or along much of ulation complete developmentwithin about the southeastern seaboard of the United 20 days (Hyman, 1920; Flynn and Felder, States.Evidence suggests that, at 18,000 ybp, in preparation).The tidal stream bank and wintersurface-water temperatures along the marshhabitats occupied are similar in most continentalmargins of the Gulf averaged4- respects between all sampled localities of 5?Clower than at present,and the 20?winter the Atlantic and Gulf populations, except isotherm was displaced into the southern that those of upper Mobile Bay and at the Gulf of Mexico (Brunner,1982). It has been two sites west of the Mississippi River ap- hypothesizedthat such changeswould have pear to be more isolated within upper es- shifted ranges of both temperate and trop- tuaries than are others east of the Missis- ical intertidal crabs to relatively lower lat- sippi River and those of the Atlantic coast. itudes, perhapsinto refugiaof the southern Such physical isolation may at least in part Gulf of Mexico (Barnwell and Thurman, account for a slight genetic separation of 1984). With sea-level and coastal sea tem- these demes from the remainderof the Gulf perature rise during the Holocene trans- population. gression, and increasein both these param- Distributionsof both the grapsidand the eters during the "post-glacialoptimum" to ocypodidcrab populations that we have here levels even greaterthan at present (Gribbin examined are conspicuouslyhabitat-depen- and Gribbin, 1990: 123), it is likely that dent. Within slight ranges of variation (as panmictic warm-temperate intertidal as- noted above), their contemporarypreferred semblagesmoved north and that gene flow habitatsare warm-temperateintertidal veg- became, to various degrees, restricted by etated banks and peaty soils of estuaries, discontinuity of warm-temperatehabitats typically within or adjacent to marshes of across southern extremes of the Florida varied salinity. As for historical origins of peninsula. With a slight subsequent fall in divergenttrans-Floridian populations in ei- both sea-level and sea temperaturesduring thergroup, divergent populations may have the last 7,000 years, there appears to have been derived from partitioning of a once been both more extensive emergenceof the FELDER AND STATON: BRACHYURAN GENETICS 205

Florida peninsula and some reextension of amino acids, and fluctuatingstochastic pop- warm-temperate habitats to the south ulation parameters which contribute syn- (Scholland Stuiver, 1967; Field et al., 1979), ergisticallyto variation among populations. which could, in the first case, have favored Although several researchershave correlat- furtherisolation of partitionedpopulations ed allozymic distances with albumin im- or, in the latter case, have facilitated some munologicaldistances (Wilson et al., 1974; increased gene flow between them. Case et al., 1975; Sarich, 1977; Maxson and Thus, a complex set of physicaldynamics, Maxson, 1979; Wyles and Gorman, 1980), both during and following peak glacial ad- the majority of these correlations involve vance, could account for the trans-Floridian parametersestimated from data on herpe- genetic partitioningwe observe in contem- tofauna and may not be readily applicable porarypopulations. A less conspicuous but to a wide rangeof phylogeneticgroups. We complex genetic structurewithin the Gulf conclude that, while allozyme patternsmay may then reflect the history of genetic seg- greatly enhance studies of biogeography,it regation in subpopulations which postgla- is not possible to derive adequateestimates cially dispersed from panmictic southern of divergence times for species or popula- (glacially displaced warm-temperate)pop- tions on the basis of an allozyme "clock" ulations northward along western Gulf as was earlier done by McCommas (1982) shorelines.Regardless of whether such sub- and others. It is also noteworthy that the populations may have originatedby such a very existence of a "molecular clock" is western Gulf tract or were later isolates of questionable(see Gillespie, 1991: 139-141). the eastern Gulf population, the history of In the course of their studies on trans- gene flow between sites in the northernGulf Floridian partitioningof Hydractiniapop- would almost certainlyhave been impacted ulations (commensal hydroids of hermit by postglacialalluvial flow and sedimentary crabs),Cunningham et al. (1991) have also events that were key determinants of con- arguedthat contactmust have predatedpeak temporaryphysiography and hydrology in glacial advances, since ranges of nearshore the northernGulf of Mexico. From changes cold-temperate Carolinian ostracods were in major river coursesto changesin the dis- not found to have extended south of north- tribution of coastal sediments and vegeta- eastern Florida duringthese glacial periods tion, an arrayof factorshas likely impacted (Cronin, 1988). However, this observation present distribution and partitioning of must be examined in the full context of gla- northernGulf populations. cial events as they affectedeastern Florida, In studies of trans-Floridiandivergence including changes in sea level during peak in the Menippecomplex, anothernearshore glaciationthat would have placedthe shore- brachyurancrab genus, much earlier his- line over 100 m lower than present eustatic torical events (12-2 mybp) have been sug- levels (Stanley, 1986). Given the narrow- gested to explain isolation of two species ness of the eastern Florida shelf, a signifi- and subsequenthybridization between them cant drop in sea level would effectivelyelim- (Bert, 1986;Bert and Harrison,1988). These inate a shallow-waterregion on the eastern events, such as periodic opening of the Su- margin of Florida, since the coast would wannee Straits during the Miocene or Plio- give way directly to the Blake Plateau. At cene, were invoked to explain genetic rela- this lower stand of sea level, depths adjacent tionships in large part because of their to shorelinehere would have been near 650 congruencewith divergence times estimat- m or greater (Sheridan and Enos, 1979). ed from allozyme data. However, it is ques- Therefore,for absence of appropriatehab- tionable whether divergence times can be itat, a shallow, cold-temperate Carolinian predicted with reliability for these stocks continental shelf ostracod fauna might not from allozyme data. While neutral theory be expected to occur off easternFlorida dur- predicts divergences between protein and ing peak glacial advance, regardless of DNA sequencesof species to occur at nearly whetheror not warm-temperatespecies were a constantrate (Wilson et al., 1977; Kimura, being displaced to the south. Even without 1983), detection of this variationat the allo- this complication of the record, we would zyme level is confounded by silent substi- argue that measurable paleotransitions of tutions, substitutions for similarly charged an offshore continental shelf ostracod as- 206 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994 semblagewould not necessarilyreflect con- certainanemones (McCommas, 1982), stone ditions directly impacting shallow coastal crabs (Bert, 1986), slipper shell gastropods or intertidalfaunal distributions during peak (Hoagland,1984), and mussels(Sarver et al., glacial advance. Strong evidence for cold, 1992). However, several other studies have dry climates in terrestrialenvironments of documented divergenceof Gulf and Atlan- the southeastern United States between tic populationsof invertebrateson the basis 23,880 and 14,600 ybp (Watts and Stuiver, of DNA analyses; these include studies of 1980) suggeststhat shallow coastal environ- horseshoe crabs, Limulus (see Saunders et ments would also have been considerably al., 1986); oysters, Crassostrea(see Reeb colder during this period than at present. and Avise, 1990); hydrozoans,Hydractinia AverageJanuary temperatures in the south- (see Cunninghamet al., 1991); and hermit eastern United States were perhaps 17?C crabs, Pagurus (see Cunningham et al., lower than at present, and "evidence sug- 1992). Reeb and Avise (1990) provided ev- gests that tropicalconditions disappearedin idence that resultsfrom allozyme and DNA seas fringingthe Florida peninsula" (Stan- investigationscan exhibit discordantand ir- ley, 1986). reconcilable patterns, and Karl and Avise It is difficult to envision any historical (1992) suggestedthat, in some instances,the scenario for evolution of intertidal and balancing selection among allozymes may nearshoretrans-Floridian species complex- preventthem from differentiatingin parallel es that would not involve major impacts of with mitochondrial DNA and single copy glacial activity, especially given extensive nuclear DNA. Yet, we observe that, where evidence for latitudinal displacements in significant patterns are demonstrated by water-surfacetemperatures, pluvial effects allozymedata in divergentAtlantic and Gulf on coastal habitats, sea level changes, populations, they are in some generalways changesin distributionsof coastal vegetated concordant with those based upon DNA. habitats, and periodic massive coastal dis- We note certain overall similarities among chargesof glacial meltwatersthat appearto patternsfrom the analysis of allozyme and have occurredin the northernGulf of Mex- mtDNA data for Crassostrea(see Karl and ico and southeasternUnited States during Avise, 1992: fig. 2B, C; note that FL2, Stu- the last 200,000 and especially the last art, Florida,is an Atlantic locality), the pat- 20,000 years (see Croninet al., 1981; Brun- tern for Menippe (see Bert, 1986: fig. 2B), ner, 1982; Barnwell and Thurman, 1984). that for Sesarma (presentstudy; Figs. 2, 3) Even assumingthat the genetic characterof and, to a lesser extent, that for Uca (present intertidal and nearshore trans-Floridian study;Figs. 4, 5). All suggesta unique west- populationswas impacted by periodic con- ern Gulf of Mexico group,differentiation of tactsand subsequentseparations dating from populations between the northeasternand the Miocene or Pliocene, it appears likely western Gulf, and relative homogeneity of that genetic signaturesof those prior events Atlantic populations. would be obfuscated by a series of much later glacially modulated population dis- ACKNOWLEDGEMENTS placements and resultant episodes of pop- Among many individuals who assisted with our field ulation contact and separation.Whether by collections, we especially thank T. L. Zimmerman, A. latitudinal displacementsof warm-temper- B. Williams, M. C. deVries, J. M. Felder, R. D. Felder, ate populations or modifi- S. C. Rabalais, J. R. Meriwether, and B. E. Felgenhauer. physiographic We also gratefully acknowledge the assistance ofD. W. cation of habitats,glaciations would appear Foltz in development of our allozyme analysis proto- to have periodicallyaltered effectivenessof cols and for his comments, along with those of T. M. the Florida Peninsula as a barrierto gene Bert, C. W. Cunningham, J. E. Neigel, B. E. Felgen- for intertidaland nearshore hauer, and J. F. Jackson, on drafts of the manuscript. flow, especially This was in marine study supported part under funding to D. populations. L. Felder through the Louisiana NSF/EPSCoR Pro- Finally, we conclude that allozyme anal- gram grant RI 1-8820219, Louisiana Education Qual- yses can contribute effectively toward re- ity Support Fund grants 86-LUM(2)-084-09 and solving genetic consequences of such iso- (1987-88)-ENH-BS-10, United States Mineral Man- lation events. Previous to this agement Service Cooperative Agreement 13-35-0001- study, 30470, and a grant from the Coypu Foundation. This allozymic differencesbetween Gulf and At- is contribution No. 35 of the USL Center for Crusta- lantic invertebrateswere addressedonly for cean Research. FELDER AND STATON: BRACHYURAN GENETICS 207

LITERATURECITED on the Atlantic inner shelf of the United States.- GeologicalSociety of AmericaBulletin 90: 618-628. Abele, L. G. 1972. A reevaluationof the Neopanope Fielder, D. R., K. R. Rao, and M. Fingerman. 1971. taxana-sayi complex with notes on N. packardii A female-limited lipoprotein and diversity of he- (Crustacea:Decapoda: Xanthidae) in the north- mocyanincomponents in the dimorphicvariants of westernAtlantic.-Chesapeake Science 13: 263-271. the fiddlercrab, Uca pugilator,as revealed by disc 1973. , distributionand ecology electrophoresis.- Comparative Biochemistry and of the genus Sesarma (Crustacea,Decapoda, Grap- Physiology39B: 291-297. sidae) in eastern North America, with special ref- Gillespie, J. H. 1991. The causes of molecularevo- erenceto Florida.-AmericanMidland Naturalist 90: lution.-Oxford University Press, New York, New 375-386. York. Pp. 1-336. 1992. A review of the grapsid crab genus Gribbin,J., and M. Gribbin. 1990. Childrenof the Sesarma (Crustacea: Decapoda: Grapsidae) in ice.-B. BlackwellLtd., Oxford,England, Pp. 1-199. America, with the description of a new genus.- Hagen, H. 0. von. 1976. Review: Joycelyn Crane, SmithsonianContributions to Zoology 527: 1-60. Fiddlercrabs ofthe world.Ocypodidae: Genus Uca.- , and W. Kim. 1986. An illustratedguide to Crustaceana31: 221-224. the marinedecapod crustaceans of Florida.-Florida Heard, R. 1977. Review: Joycelyn Crane, Fiddler Departmentof EnvironmentalRegulation Technical crabs of the world. Ocypodidae: Genus Uca.- Series 8(1), Parts 1 and 2: 1-760. NortheastGulf Science 1: 52-53. Barwell, F. H., and C. L. Thurman,II. 1984. Tax- Hedgecock,D., M. L. Tracey, and K. Nelson. 1982. onomy and biogeographyof the fiddlercrabs (Ocy- Genetics.-In: D. E. Bliss, ed.-in-chief,The biology podidae:Genus Uca) of the Atlanticand Gulf coasts of Crustacea.Vol. 2. L. G. Abele, ed., Embryology, of eastern North America.-Zoological Journal of morphologyand genetics. Pp. 283-403. Academic the LinneanSociety 81: 23-87. Press, New York, New York. Bert,T. M. 1986. Speciationin westernAtlantic stone Hedgpeth,J. W. 1953. An introductionto the zoo- crabs (genus Menippe):the role of geological pro- geographyof the northwesternGulf of Mexico with cesses and climatic events in the formationand dis- referenceto the invertebratefauna.-Publications of tributionof species.-Marine Biology 93: 157-170. the Instituteof MarineScience, University of Texas , and R. G. Harrison. 1988. Hybridizationin 3: 111-224. westernAtlantic stone crabs (genus Menippe):evo- Hoagland,E. K. 1984. Use of moleculargenetics to lutionary history and ecological context influence distinguishspecies of the gastropodgenus Crepidula species interactions.-Evolution 42: 528-544. (Prosobranchia:Calyptraeidae).-Malacologia 25: Brunner,C. A. 1982. Paleoceanographyof surface 607-628. waters in the Gulf of Mexico during the late Qua- Hyman,O.W. 1920. The developmentof Gelasimus ternary.-QuaternaryResearch 17: 105-119. afterhatching.-Journal of Morphology33: 485-501. Case, S. M., P. G. Haneline, and M. F. Smith. 1975. Karl,S. A., and J. C. Avise. 1992. Balancingselection Protein variation in several species of Hyla. -Sys- at allozymeloci in oysters:implications from nuclear tematic Zoology 24: 281-295. RFLPs.-Science 256: 100-102. Crane,J. 1975. Fiddlercrabs of the world.-Prince- Kimura, M. 1983. The neutraltheory of molecular ton University Press, Princeton,New Jersey.Pp. 1- evolution. -Cambridge University Press, Cam- 736. bridge,England. Pp. 1-367. T. M. 1988. Evolutionof marineclimates of Kwast, K. E., D. W. Foltz, and W. B. Stickle. 1990. Cronin, of the the U.S. Atlantic coast the past four million Populationgenetics and systematics Leptaste- during hexactis N. J. Shackleton,R. G. West, and D. Q. rias (Echinodermata:Asteroidea) species years.-In: 105: 477-489. Bowen, eds., The three million evolution complex.-Marine Biology past years: and D. R. Fielder. 1993. Low of climatic in the North Atlantic region. Lavery, S., allozyme variability variation in the coconut crab latro.--Com- Pp. 327-356. The Royal Society, London, Great Birgus Britain. parative Biochemistryand Physiology 104B: 353- 359. ,B. J. T. A. J. E. and J. P. Szabo, Ager, Hazel, Martin,J. W., and L. G. Abele. 1986. Notes on male Owens. 1981. Quaternaryclimates and sea levels in the crab of the U.S. Atlanticcoastal 211: 233- pleopod morphology brachyuran family plain.-Science sensu 240. Panopeidae Ortmann, 1893, Guinot (1987) (Decapoda).-Crustaceana50: 182-198. Cunningham,C. W., L. W. Buss, and C. Anderson. Maxson, L. R., and R. D. Maxson. 1979. Compar- 1991. Molecularand geologicalevidence of shared ative albumin and biochemical evolution in pleth- history between hermit crabs and the symbiotic ge- odontid salamanders.-Evolution 33: 1057-1062. nus Hydractinia.-Evolution 45: 1301-1316. McCommas,S. A. 1982. Biochemicalgenetics of the , N. W. Blackstone, and L. W. Buss. 1992. sea anemone Bunodosomacavernata and the zoo- Evolution of king crabs from hermit crab ances- geographyof the Gulf of Mexico.-Marine Biology tors.-Nature 355: 539-542. 68:169-173. Felder, D. L. 1973. An annotatedkey to crabs and McIntyre,A., T. C. Moore, B. Anderson,W. Balsam, lobsters(Decapoda, Reptantia) from coastal waters A. Be, C. Brunner,J. Cooley,T. Crowley,G. Denton, of the northwesternGulf of Mexico.-Louisiana State J. Gardner,K. Geitzenauer,J. D. Hays, W. Hutson, UniversitySea GrantPublication Number LSU-SG- J. Imbrie,G. Irving,T. Kellogg,J. Kennatt,N. Kipp, 73-02: 1-103. G. Kukla,H. Kukla,J. Lozano, B. Luz, S. Mangion, Field, M. E., E. P. Meisburger,E. A. Stanley,and S. J. R. K. Matthews,P. Mayewski,B. Molfino, D. Nin- Williams. 1979. Upper Quaternarypeat deposits kovich, N. Opdyke, W. Prell, J. Robertson, W. F. 208 JOURNAL OF CRUSTACEAN BIOLOGY, VOL. 14, NO. 2, 1994

Ruddiman, H. Sachs, T. Saito, N. Shackleton,H. Whitt. 1979. Behavioraland biochemicalevidence Thierstein,and P. Thompson. 1976. The surface for species distinctivenessin the fiddlercrabs, Uca of the iceage Earth.-Science 191: 1131-1144. speciosaand U. spinicarpa.-Evolution33: 182-191. Murphy, R. W., J. W. Sites, D. G. Buth, and C. H. Sarich,V. M. 1977. Rates, samplesizes, and the neu- Haufler. 1990. Proteins I: Isozyme electrophore- tralityhypothesis for electrophoresisin evolutionary sis.-In: D. M. Hillis and C. Moritz,eds., Molecular studies.-Nature 265: 24-28. systematics.Pp. 45-126. Sinauer,Sunderland, Mas- Sarver,S. K., M. C. Landrum,and D. W. Foltz. 1992. sachusetts. Geneticsand taxonomy of ribbedmussels (Geuken- Nei, M. 1978. Estimationof averageheterozygosity sia spp.).-Marine Biology 113: 385-390. and genetic distance from a small number of indi- Saunders,N. C., L. G. Kessler,and J. C. Avise. 1986. viduals.-Genetics 89: 583-590. Genetic variation and geographicdifferentiation in 1991. Relative efficienciesof differenttree- mitochondrialDNA of the horseshoecrab, Limulus makingmethods for moleculardata. -In: M. M. Mi- polyphemus.-Genetics 112: 613-627. yamoto and J. Cracraft,eds., Phylogeneticanalysis Scholl, D. W., and M. Stuiver. 1967. Recent sub- of DNA sequences.Pp. 90-128. Oxford University mergence of southern Florida: a comparison with Press, New York, New York. adjacentcoasts and other eustaticdata. -Geological J. C. Stephens,and N. Saitou. 1985. Methods Society of America Bulletin 78: 437-454. forcomputing the standarderrors of branchingpoints Selander, R. K., W. E. Johnson, and J. C. Avise. in an evolutionarytree and their applicationto mo- 1971a. Biochemicalpopulation genetics of fiddler leculardata from humansand apes.-Molecular Bi- crabs(Uca).-Biological Bulletin 141: 402. ology and Evolution 2: 66-85. ,M. H. Smith, S. Y. Yang, W. E. Johnson, and Neigel,J. E., D. L. Felder,C. A. Chlan,and R. LaPorte. J. B. Gentry. 197 b. Biochemicalpolymorphism 1991. Cloning and screeningof DNA probes for and systematicsin the genus Peromyscus.I. Varia- geneticstudies in stone crabs(Decapoda: Xanthidae: tion in the old field mouse (Peromyscuspoliono- Menippe).--Journalof CrustaceanBiology 11: 496- tus).-Studies in Genetics 6: 49-90. 505. Sheridan,R. E., and P. Enos. 1979. Stratigraphicevo- NomenclatureCommittee of the InternationalUnion lution of the BlakePlateau after a decadeof scientific of Biochemistry. 1984. Enzymenomenclature: rec- drilling.-In: M. Talwani, W. Hay, and W. B. F. ommendationsof the NomenclatureCommittee of Ryan, eds., Maurice Ewing Series 3. Pp. 109-122. the InternationalUnion of Biochemistryon the no- AmericanGeophysical Union, Washington,D.C. menclatureand classificationof enzyme-catalyzed Slatkin,M., and N. H. Barton. 1989. A comparison reactions.-Academic Press, San Diego, California. of threeindirect methods for estimatingaverage lev- Pp. 1-646. els of gene flow.-Evolution 43: 1349-1368. Novak, A., and M. Salmon. 1974. Uca panacea, a Sneath,P. H. A., and R. R. Sokal. 1973. Numerical new species of fiddlercrab from the Gulf coast of taxonomy-the principlesand practiceof numerical the United States.-Proceedings of the BiologicalSo- classification.-W. H. Freeman & Company, San ciety of Washington87: 313-326. Francisco,California. Pp. 1-573. Powers, L. W. 1977. A catalogueand bibliography Stanley, S. M. 1986. Anatomy of a regional mass of the crabs of the Gulf of Mexico.-Contributions extinction.Plio-Pleistocene decimation of the west- in MarineScience 20(Supplement):1-190. ern Atlantic bivalve fauna.-Palaios 1: 17-36. Rabalais,S. C., W. M. Pulich, N. N. Rabalais, D. L. Staton,J. L., and D. L. Felder. 1992. R. Osmoregulatory Felder, K. Tinnin, and R. D. Kalke. 1989. A capacities in disjunct western Atlantic populations biologicaland physiologicalcharacterization of the of the Sesarma reticulatumcomplex (Decapoda: Rio Carrizal Estuary,Tamaulipas, Mexico.-Con- Grapsidae).-Journalof CrustaceanBiology 12: 335- tributionsin MarineScience 31: 25-37. 341. C. and J. C. Avise. 1990. A dis- Reeb, A., genetic Sullivan,B., K. Miller,K. Singleton,A. G. Scheer,and continuityin a continuouslydistributed species: mi- A. B. Williams. 1983. of tochondrialDNA in the America Crassostrea Electrophoreticanalyses oyster, hemocyaninsfrom four species of mud crabs,genus virginica.-Genetics 124: 397-406. with observations on the of P. and M. Nei. 1987. The Panopeus, ecology Saitou, N., neighbor-joining obesus.- Fishery Bulletin, United States 81: 883- method: a new method for reconstructingphyloge- 885. netictrees. -Molecular Biologyand Evolution 4: 406- 425. Watts,W. A., and M. Stuiver. 1980. Late Wisconsin M. 1967. Coastal and climateof northernFlorida and the originof species- Salmon, distribution,display, rich deciduousforest.-Science 210: 325-327. sound production by Florida fiddler crabs (genus Uca).- Behavior 15: 449-459. Weir, B. S., and C. C. Cockerham. 1984. Estimating ,and S. P. Atsaides. 1968. Behavioral,mor- F-statistics for the analysis of population struc- phologicaland ecologicalevidence for two new spe- ture.-Evolution 38: 1358-1370. cies of fiddlercrabs (genus Uca) from the Gulf coast Williams, A. B. 1965. Marine decapod crustaceans of the United States.- Proceedingsof the Biological of the Carolinas.-Fishery Bulletin, United States Society of Washington81: 275-290. 65: 1-298. ,and M. K. Kettler. 1987. The importanceof 1974. Swimming crabs of the genus Calli- behavioraland biochemicaldifferences between fid- nectes (Decapoda: Portunidae).-Fishery Bulletin, dler crab taxa, with special referenceto Uca rapax United States 72: 685-798. (Smith)and U. virens(Salmon and Atsaides).-Con- 1976. Distinction betweena Gulf of Mexico tributionsin MarineScience 30: 63-76. and a CarolinianAtlantic species of the swimming , S. D. Ferris,D. Johnston,G. Hyatt, and G. S. crab Ovalipes (Decapoda: Portunidae).-Proceed-