Fishes of Buck Creek, Kentucky 151 



al. 1980; Gilbert 1980a; Branson and Batch 1981) and could have 

 actively or passively moved via subsurface channels into Buck Creek 

 and from there into Pitman Creek. Dispersal through Cumberland 

 River Reservoir constitutes the second mechanism. Guillory (1978) dis- 

 cussed active and passive dispersal via the main channel of the lower 

 Mississippi River by Notropis longirostris (Hay), a small-stream species. 

 Burr et al. (1980) interpreted the presence of E. buccata in the Green 

 River main channel as evidence of direct dispersal to tributary streams. 

 The final mechanism, stream capture, is plausible, but potential sites of 

 piracy with adjacent drainages could not be identified. Whether E. buc- 

 cata is expanding its range or is limited to the Cumberland River drain- 

 age on the Cumberland Plateau and Pitman Creek can only be deter- 

 mined by periodic surveys of the Fishing Creek and Big South Fork 

 Cumberland River fish faunas, which apparently do not currently 

 include this species (Comiskey and Etnier 1972; Harker et al. 1980). 



Two species assigned protection status in Kentucky by the Ken- 

 tucky Academy of Science (Branson et al. 1981) are known from the 

 Buck Creek drainage. Notropis ariommus was listed as of undetermined 

 status but must be considered rare in Buck Creek, from which only 5 

 specimens are known. Etheostoma cinereum was listed as endangered 

 and is known from only four drainages within the Cumberland river 

 system of Kentucky, including Buck Creek (Burr 1980; Warren 1981). 

 This large-stream and river darter prefers" cover such as boulders, 

 undercut banks, and rubble-gravel substrate mixed with detritus and /or 

 Justicia americana in sluggish current adjacent to swift shoals (Saylor 

 1980; Warren 1981). Although the ashy darter has not been collected 

 from Buck Creek since 1955, it may persist in suitable habitat along the 

 mainstem between KY 80 and KY 192. 



Three faunal units were discerned when the fish faunas of the 21 

 mainstem collecting sites were analyzed to determine average faunal 

 resemblance (Table 3). The units were comprised by sites 1-27, 31-33, 

 and 35-39, respectively, (hereinafter referred to as Units 1, 2, and 3) 

 based on greater than 50% shared fauna. Divergence from this standard 

 within each unit is attributed to sampling artifact. 



The fauna of Unit 3 (Table 4) was characteristic of low-gradient 

 habitats such as lakes, impoundments, and medium-to-large rivers 

 (Pflieger 1975; Smith 1979; Trautman 1981) and was markedly different 

 from that of the other units. Fifteen of the twenty-nine species collected 

 from Unit 3 were limited in distribution to this section of Buck Creek. 



Faunal differences were less pronounced between the two remaining 

 units (Table 3). Sites 1-14 were faunistically similar to 21-27 but gener- 

 ally shared 40% or less of the fauna with Unit 2 (sites 31-33). Sites 21-27 



