Multiparametric monitoring of fish activity rhythms in an Atlantic coastal cabled observatory

Cabled video-observatories offer new opportunities to monitor fish species at frequencies and durations never attained before, quantifying the behavioural activities of their individuals, and providing ancillary data to inform stock assessment (in a fishery-independent manner). In this context, our objective was to improve the ecological monitoring capability of SmartBay observatory (20 m depth, Galway Bay, Ireland), through a pilot study dedicated to tracking of fish counts (as a proxy of populations activity rhythms), in a context where species behaviour and consequent community turnover may occur at different temporal cycles (i.e. tidal versus day-night). In order to understand how animals can regulate their behavioural activity upon those cycles, we enforced a time-lapse (1 h interval) image collection and concomitant multiparametric oceanographic plus acoustic data acquisition continuously during 24 h, over 30 days in August 2018 (when turbidity is at minimum). For each image, we classified and then counted all visible fish and derived count time series. Periodogram and waveform analyses were used to calculate their fluctuations' periodicity (i.e. the ruling cycle) and phase (i.e. peak timing in relation to the cycle). A total of 12 marine teleost species were pictured with Trisopterus minutus, Trachurus trachurus and Chelidonichthys lucerna characterized by day-night related rhythms, while others, such as Trisopterus luscus and Gadus morhua, were influenced by the tidal cycle. 24 h count patterns were compared together and investigated for time-based ecological niche-partitioning in a wave and current-affected soundscape. These findings were discussed in relation to the ecology of species and the feasibility of promising observatory-based monitoring applications in fishery assessment practices, when targeted species have commercial value.