There is no question that large numbers 

 of fish perished in the drying pools and riffles 

 because escape was impossible. We had occa- 

 sion to sample brook trout and other species in 

 isolated pools which suddenly dried within a day 

 or two later. As the levels of the pools receded 

 and the amount of cover was reduced, the larger 

 trout attempted to move downstream. Their at- 

 tempts to cross drying or dry riffle areas were 

 observed in several instances. Between periods 

 of rest, the trapped fish charged the downstream 

 edge of their isolated pool seeking an avenue of 

 escape. The attempts were more than probes 

 since the fish often threw themselves out of the 

 water temporarily. It may be significant that 

 the larger trout, 10 or more inches long, were 

 quick to disappear from the streams when sec- 

 tions began to dry up . 



Water temperatures :-- The droughts in 

 the autumns of 1951, 1952, and 1953 were not 

 complicated by unfavorable water temperatures. 

 During the summer drought in 1954, however, 

 temperatures as high as 82° F. were measured 

 in the isolated pools . The high temperatures 

 persisted for days at a time, and the trout were 

 in distress . Adult trout rested upright on the 

 stream bottom, and made no attempt to support 

 themselves on pectoral fins. They showed no 

 movement except a rapid rate of respiration. 

 Counts of 136 gill movements per minute were 

 made on 10 -inch trout lying in 76° water. Em- 

 body (1921) observed that brook trout in stream 

 water at 83.3° exhibited great distress and a 

 refusal to eat, but survived. 



The trout also showed no fear. One 

 could wade into the pools and touch them . No 

 mortalities, however, could be charged directly 

 to the high temperatures . Rather, the warm 

 water may have contributed to an increase of 

 losses by predation. 



Predation :-- The drying of the streams 

 caused the fish to congregate in isolated pools . 

 As the levels of these pools receded, much of 

 the usual cover became unavailable, and the 

 fish were exposed to easy predation. Further- 

 more, the high temperatures of water in July 

 and August 1954 made the trout almost helpless 

 and subject to easy capture. During autumn, ac- 

 cumulations of fallen leaves served as temporary 

 cover and the fish hid in them . 



Common watersnakes (Natrix sipedon) 

 killed many trout. This species is abundant in 

 the park, and in fact, many local fishermen shun 

 the streams in summer because of their dislike 

 for the numerous "moccasins" . As many as five 

 large watersnakes were observed prowling at 

 once in one shrunken pool. Three individuals 

 were killing trout simultaneously in another small 

 pool. One of the snakes, 31.4 inches long, killed 

 and swallowed a 7.9-inch brook trout. Another 

 slightly smaller snake killed a 7-inch trout. The 

 third snake released its still living prey and es- 

 caped. 



Following the random observations on 

 predation, selected isolated pools were checked 

 daily. The watersnakes continued to wipe out a 

 large proportion of the adult trout which were 

 trapped there. The pools generally contained 

 greater numbers of blacknose and longnose dace, 

 common shiners, suckers, sculpins, and darters 

 than trout. It was surprising, therefore, to find 

 that the snakes took the adult trout in marked 

 preference to the other species. The trout may 

 have been more susceptible to capture than the 

 other fish because of their lethargy in the warm 

 water . Also, the trout were larger than the other 

 species, with the occasional exception of hog 

 suckers and white suckers. Their size may have 

 been a factor in attracting the predators since 

 Lagler and Salyer (1947) reported finding brown 

 trout up to 8.5 inches long in the stomachs of 

 watersnakes . 



American eels occur commonly in several 

 streams, particularly on the east side of the park. 

 On the possibility that they prey on trout, dozens 

 of specimens from 14 to 33 inches long were cap- 

 tured and opened. Most of the stomach contents 

 consisted of crayfish. One eel, 24 inches long, 

 had eaten an 8 -inch trout. 



There was no evidence that other preda- 

 tors were as damaging as watersnakes to trout 

 during the summer drought. Herons were ob- 

 served rarely, but kingfishers were more abun- 

 dant than usual along the streams . We concur 

 with Larimore, Childers, and Heckrotte (1959) 

 that predators can and do remove large numbers 

 of fish from shallow pools during periods of dis- 

 continuous flow. 



