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

Highlights

• Cabled video-observatories allow new opportunities for fish communities monitoring.

• SmartBay coastal platform (Galway, Ireland) was used to survey fish activity rhythms.

• 1-h time-lapse image and oceanographic data acquisition occurred over 30 days.

• Most of fishes showed day-night but not tidal activity patterns.

• A time-based ecological niche-partitioning was found in nocturnal and diurnal species.

Abstract

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.

Introduction

The exploration of the benthic oceanic realms is rapidly changing, thanks to advances in imaging and sensor (i.e. oceanographic and geochemical) technologies, installed on cabled observatories (e.g. Widder et al., 2005; Schoening et al., 2012; Vardaro et al., 2013; Osterloff et al., 2016; De Leo et al., 2018). The image and contemporary environmental data collection these platforms provide, provide ecologists with high-value ecological data which experimentally approximate species Hutchinson niche (Aguzzi et al., 2020). In fact, for each species in each frame, the number of individuals can be directly related with concomitant environmental characteristics, which define the “eco-field” of those individuals (Farina and Belgrano, 2004). When multivariate time series analysis approaches are used on those bio- and environmental data sets, new indications of species life traits can be derived.

A potential of this monitoring approach is the temporal fluctuations in counted individuals can be considered as a proxy of variations in local abundance. Individuals within a population responds to habitat cycles in photophase and tides, consequently modulating their behavioural activity. In the case of fishes which show high motility, changes in behavioural swimming rate directly influence video-counting levels (e.g. Condal et al., 2012; Aguzzi et al., 2010, Aguzzi et al., 2012, Aguzzi et al., 2015a; Chauvet et al., 2018; Rountree et al., 2020). At the same time, recurring synchronous increments in counts for different species can be linked to their trophic status, deriving information on interspecific relationships and thus allowing the food-web architecture to be depicted (Aguzzi et al., 2020). Finally, when this highly integrated biological and environmental monitoring is executed over consecutive years, it can, in principle, also provide clues on the overall performance of benthic ecosystems' productivity (e.g. Thomsen et al., 2017), being derived information of relevance for management policies (Danovaro et al., 2017).

Time-lapse image and coupled multiparametric video monitoring could also be used to explore communities' response to stochastic environmental events e.g. seismic activity or meteorology (e.g. Aguzzi et al., 2012; Matabos et al., 2014). To date, the response of animal behaviour to those effects is still poorly characterized. The analyses of the soundscape (i.e. the sounds in an environment from geophysical, oceanographic or biological source, including maritime activities) (Pijanowski et al., 2011) could provide an additional layer of information since its quality might condition fish behaviour (e.g. predators' presence; Rountree et al., 2006; Slabbekoorn et al., 2010; Wall et al., 2014). Moreover, the simple passive acoustic recognition of fish vocalizations could complement traditional survey methods (Staaterman et al., 2017), allowing species traceability beyond imaging methods (Rountree et al., 2020).

Time-lapse multiparametric synchronous video, oceanographic and sound monitoring by cabled observatories may provide a roadmap to deliver ancillary behavioural information, to explain the temporal variability in catches of commercial species. As per image spotting probability, activity rhythms influence species availability to fishing at all depths of the continental margin and in the deep sea (Aguzzi and Company, 2010; Aguzzi et al., 2015b). In the case of fish, important ancillary data could be produced by cabled observatory video-monitoring, by adding new behavioural observations for capture management models (e.g. Follana-Berná et al., 2019; Rountree et al., 2020). In fact, the comparison of video and trawl data has an established research record using movable lander assets (Priede and Merrett, 1996; Bailey et al., 2007) and an important advancement has been recently conducted with Baited Remote Underwater Video Systems (e.g. BRUVS; Bicknell et al., 2016; Langlois et al., 2018).

To the best of our knowledge, there are no previous studies on the behaviour of coastal Atlantic fish communities by cabled observatories, combining time-lapse imagery with concomitant oceanographic monitoring including sound. Therefore, the modulatory effect on the behaviour of coastal fishes (and resulting species composition) of combined tidal and day-night cycles in a variable soundscape, is still poorly known. Here, we used such a multiparametric video and environmental monitoring approach to track several fish species of commercial value in Galway Bay. Species activity rhythms response (in terms of video-capture availability) was evaluated into an area where behavioural modulation must occur upon tidal and light mixed regimes, and potentially, upon a non-previously quantified environmental noise.