Fish counters

: what are they counting and what does it tell us?

Abstract

The monitoring of the migratory fish, Atlantic salmon (Salmo salar), is of global importance. Within Scotland, many rivers use a resistivity counter to detect the presence of Atlantic Salmon. The data produced from these counters is currently poorly described, with no uniform method of data quality assurance. This thesis aims to describe the signals produced and ascertain the quality and comparability of data between sites and years and to further use this to monitor trends in the timing of movement and numbers of in-river migration of large salmonids. This thesis developed a more extensive list of the objects observed by resistivity counters and the resultant signals and images used in data quality assurance. The river systems used as study sites were the Awe, Beauly, Conon, Ness, Cassley and Shin, and the Tummel. Further to this, the effect of quality assurance on the accuracy of the data were tested and resistivity counters in 6 Scottish hydroelectric dams were found to be consistently accurate with under 20% error for the majority of counts. Having analysed data from instances where multiple resistivity counters were installed in hydroelectric dams throughout the system, consistent trends were found in the movement of salmonids through the system, and in response to releases of water from the dams within the Beauly and Conon systems. The trends demonstrated statistically significant relationships, having used time series linear modelling, between the numbers in passage at each counter which may be inferred as monitoring the passage of sub-stocks. Significant relationships were found between numbers in passage and water flows from dams within the system, though there were low r2
values associated with this.
In conclusion, data produced by resistivity counters across several sites in Scotland are reliable and provide accurate information on the movements of Atlantic Salmon through a river system and their responses to point source water releases during migration. The study was limited by focussing only on the Scottish and Southern Energy plc (SSE) resistivity counter within hydroelectric dams within Scotland and could be extended to cover a wider range of resistivity
counter network. Future work should focus on developing a machine learning method to fully verify and validate the data produced. This automated data could then be used to model year on year Atlantic Salmon return providing crucial information for the efficient management of fisheries.

Date of Award	28 May 2021
Original language	English
Awarding Institution	University of the Highlands and Islands
Supervisor	Eric Verspoor (Supervisor) & Mark Coulson (Supervisor)