A popular trout fishery in the Clinch River below Norris Dam, Tennessee, is maintained by an extensive stocking program. However, survival and return rates of rainbow trout Oncorhynchus mykiss stocked as catchables are low. Twenty rainbow trout (mean total length [TL] ϭ 307 mm) that had resided in the tailwater at least 5 months were collected from the river and implanted with radio transmitters in June 1998. Similarly sized rainbow trout were implanted with radio transmitters at a hatchery and stocked into the Clinch River on 8 July 1998 (N ϭ 19; mean TL ϭ 304 mm) and 16 September 1998 (N ϭ 11; mean TL ϭ 311 mm). The stocked rainbow trout dispersed rapidly and nearly all (93%) of those fish died quickly or emigrated from the tailrace. Resident fish were significantly less active than stocked fish, and they persisted significantly longer (Kruskal-Wallis tests, P ϭ 0.0001). Poor return rates and survival of rainbow trout stocked as catchables were attributed to their rapid, long-range movements and high levels of activity. Such behaviors are energetically inefficient and probably rendered them more vulnerable to predation.

The incidence of predation by largemouth bass Micropterus salmoides on fish in Lake Conroe, Texas, was examined over 7 years at two disparate levels of habitat complexity. When areal coverage of submersed vegetation ranged from 39 to 44% of the reservoir's 8,100 hectares, largemouth bass 100 mm and smaller in total length consumed fish infrequently; most did not consume fish until they reached lengths of 140 mm and more, Following the elimination of all submersed vegetation by grass carp Ctenopharyngodon idella, fish were consumed by most largemouth bass 60 mm or longer. The initiation of piscivory at smaller sizes resulted in significantly faster first‐year growth for all largemouth bass year‐classes produced after vegetation removal. Although shifts in the structure of the forage fish community occurred, ample fish prey existed for largemouth bass before and after vegetation removal. The onset of piscivory remained similar for largemouth bass collected along the dam riprap, where vegetation was absent throughout the study. These observations support the hypothesis that habitat complexity, as mediated by vegetation abundance, was the principal factor regulating piscivory by largemouth bass in the littoral zone of Lake Conroe.

Variation in recruitment of largemouth bass Micropterus salmoides was quantified in four mainstream impoundments along an 800‐km segment of the Tennessee River and compared with various types of hydrologic and aquatic plant abundance data in an attempt to explain factors related to the formation of strong and weak year‐classes. Residuals from four catch‐curve regressions for age‐2–11 fish collected in 1993 or 1994 were used to describe year‐class strength. The formation of strong and weak year‐classes was generally synchronous in these four reservoirs. Year‐class strength was inversely related to average June–July discharge and positively associated with retention (reservoir volume/discharge) for data pooled from all reservoirs. Thus, weak year‐classes were produced during wet early‐summer conditions after largemouth bass hatched, whereas stronger year‐classes were produced during dryer early‐summer conditions. Late‐summer aquatic plant abundance and water level fluctuations during April–May while spawning was occurring were not related to largemouth bass recruitment in these four reservoirs. We speculated that higher discharges and faster flushing rates were associated with reduced production at lower trophic levels and poorer survival of young largemouth bass that ultimately affected recruitment to adult size.