The new synthesis of cabled observatory science: Technology meets deep-sea ecology

Constraints in sampling repeatability at statistically relevant frequencies are limiting the progress of marine ecology, especially in the vast and still largely unexplored deep sea. Within the framework of novel cabled observatory science, it is now possible to study and monitor the fauna of geologically different ecosystems at any depth of the continental margin and abyssal plains, at sampling frequencies and over temporal durations never attained before (i.e. from seconds to decades). New multiparametric platforms endowed with video cameras that are being progressively installed in different oceans, can be used not only for a broad faunal characterization but also to quantify the massive three-dimensional displacements of marine populations in response to cyclic oceanographic, chemical, and geologic fluctuations (also measured in a multiparametric fashion). Here, we will review how automated video-imaging protocols for animal classification and counting could be implemented to transform the video-camera into one of the first intelligent marine sensors for remote, autonomous and continuous monitoring of communities in relation to their diel (i.e. inertial, internal-tidal or day-night), seasonal, and inter-annual cycles of functioning. We will also discuss the possibility to study the responses of benthic species to other more stochastic habitat changes (e.g. those induced in the water column by the meteorology), through the measurement of modifications in water column properties by observatory vertical elongations. Studies of this kind may allow an efficient modeling of marine community modifications in view of future climate change scenarios, based on alterations of the benthopelagic coupling equilibrium. A special emphasis will also be given to the faunal monitoring prior and after catastrophic events (e.g. seismic activity or tsunamis), to initiate a critical discussion on the reliability of biologically-based early warning systems based on the continuous monitoring of the deep sea. The reliability of these systems should be evaluated by considering whether significant changes in automatically video-counted benthos occur prior to, or only after, the incoming catastrophe.