Mollusca From Little South Fork Cumberland River 111 



tered underground to enter subterranean channels in the Ste. Genevieve 

 limestone. The streambed was dry for approximately one kilometer, 

 after which the river emerged at the base of limestone bluffs. Below the 

 point of resurgence the water was cooler and had greater clarity, and 

 there were long, deep pools. 



A sampling transect at Freedom Church Ford (Sta. 16) yielded an 

 average 2.87 unionid individual/ m 2 , compared with 7.53 and 7.17 indiv- 

 iduals/ m 2 at upstream stations 8 and 13, respectively (Table 4). 

 Decreases in abundance at Station 16 were probably related to de- 

 creases in optimal habitat. G. Schuster (pers. comm.) reported sedimen- 

 tation at Ritner Ford in spring 1981. Although these sediments were 

 gone by fall 1981, mussel populations are incapable of withstanding 

 repeated or extended siltation. Limestone outcrops and bedrock, which 

 increasingly dominated the substrate in the river below Station 13, may 

 also reduce the amount of optimal habitat. At Freedom Church (Sta. 

 16) the ford is a continuous sheet of limestone, 30 m wide and 20 m 

 long, and unionids are restricted to gravel accumulated above and 

 below the bedrock outcrops. 



Within the river, certain general habitat preferences were apparent. 

 Without heavy spring rains in 1981, deep pools were covered with 2 to 

 10 cm of organic detritus. No live unionids were collected in these areas. 

 Neither unionids nor Corbicula occurred along stream margins where 

 water willow, Justicia americana, is abundant. However, in 1980 and 

 1981 these areas were exposed or in shallow water. In pool areas, heavi- 

 est concentrations of unionids, especially Potamilus alata, occurred 

 along current-swept banks. No live Corbicula were recorded in deep, 

 sluggish pools except at inflow areas, while shallow pools contained 

 occasional live specimens. 



Unionids were not recorded in water less than 10 cm deep, but 

 Corbicula was found in water 3 cm in depth. Greatest unionid densities 

 occurred in water from 10 to 25 cm deep. With the exception of Lampsil- 

 is ovata and Pegias fabula, all unionids could be found with the anter- 

 ior end protruding slightly from the substrate. Lampsilis ovata was 

 uncovered in gravel with the anterior end approximately 5 cm below the 

 normal substrate surface. Pegias fabula, previously reported by Starnes 

 and Starnes (1980) at the interface between pool and riffle, was also 

 collected live from the shallower, current-swept areas in the riffle 

 proper. During low flow periods, Pegias was observed reclining upon 

 the substrate between gravel and cobbles, with a somewhat more hori- 

 zontal than vertical orientation. In higher spring flows, Pegias anchored 

 into the substrate with more typical unionid orientation. Ptychobran- 

 chus subtentum was ubiquitous in riffle areas in water 10 to 25 cm deep 

 and in all but the swiftest current. Medionidus conradicus was restricted 



