126 J. F. Fitzpatrick, Jr. 



broad areola; the shape of the rostrum, chela, carapace and pereiopodal 

 coxae; and color pattern. In the subgenus there are two principal 

 assemblages, not formally recognized by Hobbs (1972). In one group, a 

 full complement of terminal elements is present on the gonopod; in the 

 other {gibbus, raneyi, spiculifer) the cephalic process is absent; in P. 

 (Pe.) ouachitae Penn the cephalic process is also sometimes absent. 



I am not sure that Hobbs would still believe that P. (Pe.) vioscai 

 Penn is the most "primitive" extant species of the subgenus, but there is 

 no doubt that a reduction in terminal elements is apomorphic. Of the 

 three species so disposed, all are found in the most eastern part of the 

 range of the subgenus, while species with a full complement of terminal 

 elements also found in that part of the range have quite specialized 

 pleopods and annuli ventrales (Fig. 1). The more generalized species are 

 found from Mississippi westward. 



Populations of P. (Pe.) ouachitae (or siblings) occur allopatrically 

 in Arkansas and Mississippi. This species seems to be morphologically 

 intermediate between species with the full-complement of terminal ele- 

 ments and those with a short-complement. Further, the populations of 

 P. (Pe.) vioscai that occur east of the Mississippi River have a much 

 more modified cephalic process than those west of the river. They are 

 sufficiently different that work I have in progress will probably result in 

 my proposing subspecies categories for the two forms. The siblings, P. 

 (Pe.) penni Hobbs and P. (Pe.) clemmeri Hobbs, are so distributed that 

 the more eastern form is also the more remote (from the ancestral type) 

 form (Fitzpatrick 1977a). The entire picture suggests an invasion of the 

 lower Gulf Coastal Plain by an early offshoot of Procambarus stock, 

 and subsequent reinvasion of the southeastern United States along cor- 

 ridors located near the present coastline (Fig. 2). 



Except for the nearly unique subgenus Lacunicambarus , which 

 Hobbs (1969:163) believed to have been "one of the earliest branching 

 stocks," Cambarus is represented in the central Gulf area only by C. 

 (Depressicambarus) striatus Hay. Bouchard (1978) assigned this species 

 to the superspecific assemblage he considered the more advanced, yet 

 one must remember that Hobbs (1969) believed Depressicambarus to 

 represent a moderately early digression in cambarid evolution. Hobbs 

 (1969:169) conceded that his proposed dispersal corridors to this region, 

 especially for Lacunicambarus, are tenuous. 



The representatives of Orconectes in the area are all members of 

 specialized and advanced Virilis and Palmeri Groups. Except for Falli- 

 cambarus fodiens (Cottle) and F. uhleri (Faxon), all members of that 

 genus occur on the Gulf Coastal Plain or in reasonable proximity to its 

 central and western parts. Further, the most primitive species lie in 

 southwestern Louisiana and southwestern Arkansas, "probably not far 

 from the ancestral home of the genus" (Hobbs 1969:124). The most 

 "primitive" Faxonella, Fx. creaseri Walls, is found in northcentral Loui- 

 siana, while Hobbseus is confined to the middle and upper Tombigbee 



