SUBTERRANEAN AMPHIPOD STYGONECTES 163 



absent on both the pereonites and pleonites, but coxal gills are present 

 on the seventh pereopods. The range of S. bifurcatus is predominately 

 restricted to caves just north and east of the Llano uplift area in 

 central Texas, although two populations, known only from one 

 specimen each, occur south of the uplift area in Kendall County. 

 Theoretically, according to geological structure, the populations in 

 Kendall County caves should be extrinsicaUy isolated by strati- 

 graphic differentiation from populations fm-ther north, but appreciable 

 phenotypic differences are not discernible. Probably the best interpre- 

 tation of this range is to suppose a recent, total isolation to cave 

 waters of popidations that also formerly inhabited shallow ground 

 waters near the surface when the central Texas climate was wetter 

 during the early Quaternary. At that time, gene flow could have 

 been maintained over a wider area by mechanisms already suggested 

 to account for the dispersal of interstitial amphipods. 



The northern part of the range of S. bifurcatus is presently restricted 

 to four caves in San Saba, Lampasas, and Coryell Counties. Tippits 

 Cave in Coryell County is developed in Edwards limestone and ac- 

 cording to Reddell (in litt.), this cave is situated near the top of a 

 hill and appears to be largely isolated from the surrounding cavernous 

 limestone. The two caves in San Saba County (Harrells and Gor- 

 mans) and the one in Lampasas County (Sullivan Knob) lie some 

 45 to 50 miles west of Tippits Cave and are developed in Ordovician 

 limestone. Tippits and Sullivan Knob Caves are further separated 

 by the Lampasas River, whereas the latter cave is in turn separated 

 from Gormans and Harrells Caves by the Colorado River. There 

 is no evidence of morphological variation between any of these four 

 populations, and one is forced to reckon that limited dispersal may 

 be possible betw een these localities. This deduction is further sub- 

 stantiated if one also considers that three of these fom* caves are 

 similarly inhabited by populations of S. russelli (discussed in some 

 detail belowO and that nowhere in the range of this highly variable 

 species are populations more homogeneous than in these caves. 

 Morphological evidence strongly indicates that at least limited 

 dispersal by subterranean routes may occur along a narrow corridor 

 between southern Coryell County and eastern San Saba County. 



The genetic affinities of S. reddelli are partially obscured by the 

 lack of know-n males and the rather aberrant pereopods, which are 

 more slender and proportionately more elongate (especially pereopod 

 5) than in any other species of the tenuis group. Discounting the 

 pereopods, how^ever, this species appears to share a number of morpho- 

 logical affinities with S. bifurcatas, and both species are possibly 

 products of a relatively close common ancestry. In an earlier paper, 

 I suggested that the origin of S. reddelli might have come about by 



