Scottish Marine and Freshwater Science Volume 5 Number 18: Depth use and movements of homing Atlantic salmon (Salmo salar) in Scottish coastal waters in relation to marine renewable energy development

Methods

The methods used in the study are presented in Godfrey et al. (2014) and are summarised here, highlighting differences between the years. During the months of May and June, 2013 and May, June and July, 2014, salmon were caught in bag nets at Armadale (Lat. 58.556; Long. -4.093), near Bettyhill, on the north coast of Scotland. To minimise the impact of capture or tagging on subsequent behaviour, only fish in good condition and those >70 cm in length were retained: the mean length of fish used in the study was 78.2 cm in 2013 and 77.1 cm in 2014. Scale reading indicated that all the tagged fish were multi-sea-winter (predominantly two-sea winter (2 SW)) maiden salmon, with the exception of four fish (three from 2013) that were identified as repeat spawners.

In 2013, a total of 50 PSATs, comprising 30 High Rate X-Tags (Microwave Telemetry, Inc.) and 20 MiniPATs (Wildlife Computers, Inc.) were fitted, and programmed to release from fish after 1 to 10 days of data collection. The MiniPAT release mechanism, though designed for use in sea water, was found to be functional in freshwater during the first year of the study. Thus longer deployments were possible in 2014, when a further 85 PSATS (all MiniPATs) were deployed (22 x 5 days, 21 x 10 days, 21 x 15 days and 21 x 20 days), without the risk of fish which entered rivers being lost from the sample. Fish handling and release protocols followed those described by Godfrey et al. (2014), except that the tag attachment system in 2014 was a modified version of that described by Lacroix (2013) because of the longer deployment periods used.

PSATs record and archive data during deployment, relaying the stored information, together with location, to the ARGOS satellite network (www.argos-system.org/) when the pop-up mechanism is activated. The tags were programmed to sample depth and temperature at intervals of 75 s (MiniPATs) and 10-95 s (X-Tags) depending on deployment duration. After release the tags had approximately 14 days of battery life to communicate the stored data. The extent of data recovered from tags via satellite is influenced by the sky view of the tag and the alignment of the aerial during the transmission phase, and may range from 0-100% depending on where the tag pops up. If tags can be physically recovered then it is possible for the complete dataset to be extracted from X-Tags, while recovered MiniPATs additionally allow the user to access archived data at very high resolution (1 or 3 s intervals). To date, 11 and 34 tags have been recovered from the 2013 and 2014 field seasons respectively.

Swimming depths are reported for salmon from the moment of release until the moment of tag pop-up, unless river entry (n=26) or predation events (n=4) were detected from the temperature data, or unless tags were released from fish early (Godfrey et al., 2014). In these cases the tail of the depth records was trimmed so that the resulting data set represents the behaviour of free-swimming salmon at sea. Depth data were analysed as cumulative frequency, using the full resolution of the tags (0.5 m for MiniPATs, 1.35 m for X-Tags). In each year individuals with fewer than 200 depth records were excluded from the analysis (Godfrey et al., 2014).

The error associated with pop-up position was estimated based on the measured drift of tags immediately following first location by the satellite network, for the time period equivalent to the elapsed time between tag detachment from the fish and first location by the satellite network (Godfrey et al., 2014).