KOSMOS: An Open Source Underwater Video Lander for Monitoring Coastal Fishes and Habitats

Abstract

Background: Monitoring the ecological status of coastal ecosystems is essential to track the consequences of anthropogenic pressures and assess conservation actions. Monitoring requires periodic measurements collected in situ, replicated over large areas and able to capture their spatial distribution over time. This means developing tools and protocols that are cost-effective and provide consistent and high-quality data, which is a major challenge. A new tool and protocol with these capabilities for non-extractively assessing the status of fishes and benthic habitats is presented here: the KOSMOS 3.0 underwater video system. Methods: The KOSMOS 3.0 was conceived based on the pre-existing and successful STAVIRO lander, and developed within a digital fabrication laboratory where collective intelligence was contributed mostly voluntarily within a managed project. Our suite of mechanical, electrical, and software engineering skills were combined with ecological knowledge and field work experience. Results: Pool and aquarium tests of the KOSMOS 3.0 satisfied all the required technical specifications and operational testing. The prototype demonstrated high optical performance and high consistency with image data from the STAVIRO. The project’s outcomes are shared under a Creative Commons Attribution CC-BY-SA license. The low cost of a KOSMOS unit (~1400 €) makes multiple units affordable to modest research or monitoring budgets.

Identifying fit-for purpose methods for monitoring fish communities

“…However, to our knowledge no studies focusing on fish using V-RUV have been conducted. Some studies also focus on a 360°view, which can be achieved either by using multiple cameras in an array, 360°cameras or rotating systems (Wells et al, 2008;Aguzzi et al, 2011;Pelletier et al, 2012;Schobernd et al, 2014;Starr et al, 2016;Mallet et al, 2021;Pelletier et al, 2021a;Pelletier et al, 2021b). Many of these techniques can be used not only with a single camera but also with a stereo camera, allowing more accurate length and field of view calculations (Harvey et al, 2002).…”

“…H-RUV (Fedra and Machan, 1979;Jan et al, 2007) RUV Rotating (Aguzzi et al, 2011;Pelletier et al, 2012;Pelletier et al, 2021a;Pelletier et al, 2021b) RUV 360° (Mallet et al, 2021) BRUV H-BRUV (Ellis and DeMartini, 1995) BRUV V-BRUV (Priede and Merrett, 1996; BRUV 360° (Wells et al, 2008;Schobernd et al, 2014) BRUV Stereo (e.g. Harvey et al, 2012;Langlois et al, 2015;Letessier et al, 2015) BRUV Rotating stereo (Starr et al, 2016) Hammerl et al 10.3389/fmars.2023.1322367 extracted from the filter, sequenced, detected and then analyzed (see Garlapati et al, 2019).…”