| Fish type | Time | Test statistic | P-value | n |
---|
 | Mean | CI | Median | U2 |  |  |
---|
Lake | Hatchery | 12:24 | 08:43–16:45 | 12:05 | 0.26 |  < 0.03 | 75 |
Wild | Random | 05:21–19:49 | 11:09 | 0.11 |  > 0.1 | 43 |
Estuary | Hatchery | Random | 11:38–19:37 | 15:25 | 0.16 |  < 0.1 | 63 |
Wild | Random | 13:08 – 22:22 | 19:52 | 0.16 |  < 0.1 | 38 |
Marine arrays | Hatchery | 13:32 | 10:59–16:29 | 13:31 | 0.3 |  < 0.01 | 39 |
Wild | Random | 07:43–14:57 | 09:54 | 0.19 |  < 0.1 | 22 |
- Times presented in UTC + 1. Data in bold indicate departure from circular uniformity, i.e. Atlantic salmon (Salmo salar) smolt movements were more likely to occur at a specific time of day. ‘Random’ means the data follow a uniform distribution around the 24-h clock and ‘n’ is the total number of fish detected passing the receivers at each section. Mean and median times at which smolts left the lake, left the estuary and arrived at the final marine arrays are presented. Confidence intervals (CI) for the population mean time were calculated using a bootstrapping approach assuming an underlying von Mises distribution. Watson’s critical value is 0.19 for all sample sizes