opening could be found between the valves, the crab 

 would insert fingers of the major chela into the open- 

 ing and utilize the large basal tooth on the dactyl to 

 repeatedly crush one valve. Once the valves 

 remained open, the crab would pick out the flesh with 

 its chelae. 



Panopeus obesus was successful in feeding on the 

 marsh periwinkle, Littorina irrorata, but P. simpsoni 

 was unable to chip the shell aperture of the snail. 

 Panopeus obesus initially chipped at the edge of the 

 snail' s shell aperture, and soon inserted fingers of the 

 major chela into the opening which it had created. 

 Repeated crushing with the dactyl opened the shell 

 along a direct line toward and beyond the operculum. 

 Once the crab was able to remove the operculum, it 

 could feed on the flesh. Densities of L. irrorata were 

 apparently related to those of P. obesus; where large 

 individuals of the crab were abundant, few or no L. 

 irrorata could be found. 



Large P. obesus readily fed on Sesarma cinereum, S. 

 reticulatum, and species of Uca, but less actively on 

 Eurypanopeus depressus. Prey was entrapped by 

 means of ambulatory legs and chelae. The less 

 aggressive P. simpsoni fed on any of the crabs pre- 

 sented if they were sufficiently small. 



Regardless of feeding intervals or molting occur- 

 rences, cannibalism in both species always resulted 

 when medium to large individuals remained in a tank 

 for extended periods (2-3 mo). 



Only P. simpsoni was observed feeding in the field. 

 One individual was observed feeding on algae and 

 bryozoans encrusting rocks by tearing off bits of the 

 encrustation with both chelae and passing these to 

 the third maxillipeds. 



Coloration 



The two species of mud crabs from Mobile Bay can 

 be identified by differences in coloration. Panopeus 

 obesus is dark reddish brown dorsally and cream 

 colored ventrally. The chelipeds exhibit a "veined" 

 pattern extending from upper to lower margins on the 

 external surface of the palm. These pigmentation 

 sites coincide with muscle fiber attachments within 

 the chelae. Fingers are dark brown. 



Panopeus simpsoni is variably grayish brown dorsal- 

 ly, with or without a variable cream colored stripe 

 (lacking, simple, broken, or staggered) posterior to 

 the frontal margin; some individuals have a spotted 

 appearance (grayish brown varying to dark brown 

 with yellow- white areolations). Most are white to 

 cream ventrally. Palms of the chelae are mottled with 

 brown, gray, and yellow- white, fading to yellow-white 

 ventrally; fingers are dark brown. 



Males of both species of mud crabs have a large 

 proximal red spot on the inner surface of the ischium 

 of the third maxillipeds (color variably persistent 

 after long preservation). The spot is present in all 

 females of P. obesus but completely lacking in 

 females of P. simpsoni. 



Morphology 



The carapace of P. obesus appears wider in relation 

 to its length, is more convex dorsally along the 

 anterior-posterior axis, and is armed with generally 

 blunter anterolateral teeth than that of the relatively 

 less bulky P. simpsoni (Rathbun 1930; Williams 

 1983). While differences in carapace morphology 

 and color usually suffice to distinguish living or 

 freshly preserved individuals of the two species, 

 variations among individuals within the study area 

 and throughout the entire geographic range indicate 

 that these and other possible differences should be 

 rigorously evaluated (Smith 1869; Rathbun 1930; 

 Williams 1965). 



The ecological interface between the two sympatric 

 forms was considered to be the best place to test for 

 possible differences between them, since it is the 

 area in which intergradation might most likely occur. 

 Nearly all of the measured mud crabs were collected 

 at random from habitats in the intertidal zone. 

 However, by imposing this limitation on the samples, 

 the proportion of older (and larger) individuals in 

 them may not reflect true proportions in the natural 

 populations. (Larger individuals of P. simpsoni may 

 occur subtidally but normally P. obesus did not occur 

 there.) An intertidal rubble station in proximity to a 

 mud bank sheltered many P. simpsoni as well as a 

 small number of P. obesus. Intertidal mud banks 

 yielded about 22% of the P. obesus (mainly larger 

 individuals) , but the overwhelming proportion of this 

 form came from undercut marsh (mostly smaller 

 individuals). 



The means for each character except those of the 

 palm (Table 1) were significantly greater (P= 0.05) 

 in the P. obesus samples, although the smallest 

 specimen of P. simpsoni was only 1 mm shorter than 

 its P. obesus counterpart. The relationships between 

 carapace length and width, carapace length and body 

 depth, length of merus and ischium of the third max- 

 illiped, and length and height of the palm of the major 

 chela (Fig. 2) were analyzed by regression analysis 

 (Table 2) and the ANCOVA procedure (Table 3). 

 The coefficients of determination (Table 2) were 

 >0.94 in all cases except one. The ANCOVA pro- 

 cedure (Table 3) showed a highly significant statisti- 

 cal difference between the species in each of these 



888 



