Marine cabled video-observatories allow the non-destructive sampling of species at frequencies and durations that have never been attained before. Nevertheless, the lack of appropriate methods to automatically process video imagery limits this technology for the purposes of ecosystem monitoring. Automation is a prerequisite to deal with the huge quantities of video footage captured by cameras, which can then transform these devices into true autonomous sensors. In this study, we have developed a novel methodology that is based on genetic programming for content-based image analysis. Our aim was to capture the temporal dynamics of fish abundance. We processed more than 20,000 images that were acquired in a challenging real-world coastal scenario at the OBSEA-EMSO testing-site. The images were collected at 30-min. frequency, continuously for two years, over day and night. The highly variable environmental conditions allowed us to test the effectiveness of our approach under changing light radiation, water turbidity, background confusion, and bio-fouling growth on the camera housing. The automated recognition results were highly correlated with the manual counts and they were highly reliable when used to track fish variations at different hourly, daily, and monthly time scales. In addition, our methodology could be easily transferred to other cabled video-observatories.

We applied data mining on YouTube videos to better understand recreational fisheries targeting common dentex (Dentex dentex), an iconic species of Mediterranean fisheries. In Italy alone, from 2010 to 2016 spearfishers posted 1051 videos compared to 692 videos posted by anglers. The upload pattern of spearfishing videos followed a seasonal pattern with peaks in July, a trend not found for anglers. The average mass of the fish declared in angling videos (6.4 kg) was significantly larger than the one in spearfishing videos (4.5 kg). Videos posted by spearfishers received significantly more likes and comments than those posted by anglers. Content analysis suggested that the differences in engagement can be related to appreciation of successful spearfishers necessitating relevant personal qualities for catching D. dentex. We also found that the mass of the fish positively predicted social engagement as well as the degree of positive evaluation only in spearfishing videos. This could be caused by the generally smaller odds of catching large D. dentex by spearfishing. Our case study demonstrates that data mining on YouTube can be a powerful tool to provide complementary data on controversial and data-poor aspects of recreational fisheries and contribute to understanding the social dimensions of recreational fishers.

The inclusion of behavioural components in the analysis of a community is of key relevance in marine ecology. Diel and seasonal activity rhythms or more longlasting\ud changes in behavioural responses determine shifts in population, which in turn affect measurable abundances. Here, we review the value of cabled videoobservatories as a new and reliable technology for the remote, long-term, and highfrequency monitoring of fishes and their environment in coastal temperate areas. We provide details on the methodological requirements and constraints to appropriately measure fish behaviour at day-night and seasonal temporal scales from fixed videostations.\ud In doing so, we highlight the relevance of an accurate monitoring capacity of the surrounding environmental variability. We present examples of multiparametric video, oceanographic, and meteorological monitoring made with the western Mediterranean platform OBSEA (www.obsea.es; 20 m water depth). Results are reviewed in relation to future developments of cabled observatory science, which will greatly improve its monitoring capability due to: i. the application of Artificial Intelligence to aid in analysis of increasingly large, complex, and highly interrelated biological and environmental data, and ii. the design of future geographic\ud observational networks to allow for reliable spatial analysis of observed populationsPostprint (published version