ten overlay one another (Sullivan et al. in press). By 

 comparison, in P. herbstii, s.s., at least six loci are also 

 present with multiple alleles segregating at many of 

 these loci (unpubl. data). The anodal zone of P. 

 obesus differs from P. herbstii— most P. obesus' 

 anodal zones (35 of 37 crabs examined) usually con- 

 tain a single band, while in P. herbstii two bands are 

 present in this zone (bands 5 and 6, Fig. 1). When 

 double bands are present in P. obesus (because the 

 individual is heterozygous at locus 5), the extra band 

 does not align with band 6 but moves between bands 

 4 and 5. We are able to differentiated obesus from P. 

 herbstii in every case by color pattern and by hemocy- 

 anin pattern. The hemocyanin pattern is extremely 

 variable in both of these species, yet the differences 

 described above were present in all individuals of P. 

 obesus from Beaufort to St. Petersburg. 



Electrophoretic analyses of the hemocyanins in 

 specimens of P. lacustris (n = 23) from Big Pine Key, 

 Fla., revealed five or six banded patterns which fell 

 into seven phenotypic classes. This variable species 

 shows a single major fast band, but it aligns with band 

 6 of P. herbstii instead of band 5. The remainder of 

 the hemocyanin pattern is distinct when compared 

 with those of the other forms (Fig. 1). Electrophoretic 

 analyses of the hemocyanins in eight specimens of P. 

 simpsoni from Louisiana reveals three phenotypic 

 patterns which indicate polymorphism in this pop- 

 ulation also. Comparison of the patterns with those of 

 P. herbstii reveals certain distinctions. A single fast 

 band aligning with band 5 of P. herbstii is prominent 

 in all gels. Band 6 is absent in all of the individuals of 

 P. simpsoni which we examined. Comparison with P. 

 obesus from the Carolinas and Florida reveals a 

 similarity of electrophoretic mobilities but distinct 

 intensity changes. In P. obesus, bands 3 and 4 are 

 usually prominent; in P. simpsoni, band 5 is most 

 prominent. 



Our original samples (n = 246) of Panopeus from 

 Beaufort, N.C., contained about 2% P. obesus. At 

 Charleston, S.C., {n = 38) the frequency was 24%, 

 and at St. Petersburg, Fla., (n = 19) it was 89% P. 

 obesus. All individuals were collected in rocky areas 

 or on oyster bars. Turner (1979) and Turner and 

 Lyerla ( 1 980) indicated that P. obesus was common in 

 the upper intertidal at North Inlet, S.C., where the 

 marsh grass, Spartina, grew in abundance. At 

 Beaufort, N.C., our upper intertidal samples yielded 

 virtually 100%- P. obesus. A transect of an intertidal 

 region containing both oyster rubble and Spartino 

 revealed overlap for the two species only in the mid- 

 dle intertidal where oyster rubble and marsh grass 

 were adjacent. Panopeus obesus was associated with 

 burrows at the base of Spartina clumps; P. herbstii 



was in and under oyster shells. Panopeus herbstii 

 alone occurred in the lower intertidal where only 

 shells were present, and P. obesus alone was present 

 in the upper intertidal where Spartina dominated; 

 but in areas of lower salinity it appeared that P. 

 obesus tended to displace P. herbstii on oyster bars. 



Examination of adult females in June showed one 

 ovigerous female out of five P. obesus whereas five of 

 seven adult females of P. herbstii were ovigerous. In- 

 dividuals of P. obesus appeared to live in burrows and 

 would position themselves in the entrances to defend 

 them. In general, P. obesus appeared more aggressive 

 than P. herbstii. Stomach analyses of P. obesus 

 revealed oyster spats (the primary food source of P. 

 herbstii), shell and sea urchin fragments, and eggs 

 and walking legs of Ucapugnax (Smith) and perhaps 

 Sesarma sp., which are all primary food sources for P. 

 obesus. Although the number of crabs examined was 

 small, the ratio of the cheliped dactyl length (inner 

 length, base to tip) divided by the carapace width 

 averaged 0.25 for P. obesus and 0.22 for P. herbstii. If 

 P. obesus does prey on other crabs in considerable 

 numbers rather than mainly on oysters, one might ex- 

 pect a longer dactyl as compared with the oyster- 

 feeding P. herbstii. 



Discussion 



Genetic variability at the hemocyanin loci in all pop- 

 ulations of Panopeus which we have sampled com- 

 plicates comparisons. Such variability is char- 

 acteristic of many, but not all, temperate xanthid 

 species (unpubl. data). Additionally, there are shifts 

 in allelic frequencies in geographically separated 

 populations, and the occurrence of local alleles is not 

 unusual. However, in all areas where we have sam- 

 pled two species, hemocyanin patterns can be 

 designated as specific in spite of the "within-species" 

 variability. Repeated sampling of the same in- 

 dividual over time has always yielded identical pat- 

 terns, and the alleles at each locus are invariably in 

 Hardy- Weinberg equilibrium. 



We believe the evidence for the existence of two 

 gene pools is very strong. In addition to hemocyanin 

 data, Turner and Lyerla (1980) found unique alleles 

 at the amylase, esterase, and malate dehydrogenase 

 loci in the two South Carolina species. For instance, 

 the most abundant amylase allele in P. obesus was not 

 even present in adjacent populations of P. herbstii. 



Habitat preferences are very clear in the Carolinas, 

 at or near the northern limit of range for P. obesus. In 

 western Florida, P. obesus is more abundant than it is 

 in North Carolina and may occupy a broader range of 

 habitats. Feeding habits in the two regions appear to 



884 



