Abstract
Movements of 20 radio-tagged wild Atlantic salmon (Salmo salar) parr were studied 1.3 km downstream of a hydropower station. Ten parr were free-ranging and ten were restricted to an area near the river bank by a 75 m(2) rectangular containment pen that was flooded at high flow, but dry at low flow except for a small refuge at the low end. The positions of contained parr were continuously recorded within a four-cell uniform grid by continuous data logging, while free-ranging fish were positioned by manual tracking four times per day. River discharge was manipulated to run five pseudo-replicated series of three step changes from 30 to 70 to 110 m(3) s(-1), with slow (2 h) and rapid (0 h) increase and decrease in discharge, as well as during day and night. No differences in movements or home ranges (mean 2770 m(2) +/- 1759 SD) were detected related to manipulated discharge variations or time of day or night for the free-ranging parr. No free-ranging parr were stranded during rapid flow reductions. In the containment pen, the parr distributed themselves relatively evenly among the cells. They moved significantly more at changing than at stable flows. In the containment pen, most of the fish that were stranded were observed during rapid flow reductions at night. Rapid reductions in water flow may cause increased mortality in large parr in shallow habitats if movements are restricted; however, rapid fluctuations did not appear to influence movements or cause stranding by free-ranging Atlantic salmon parr, which moved considerable distances regardless of discharge. Copyright (C) 2004 John Wiley Sons, Ltd.