A novel use of Stereo Baited Remote Underwater Video and Drop-Down Video for biodiversity and marine landscape mapping and prediction

PLOS ONE | https://doi.org/10.1371/journal.pone.0319355 April 3, 2025 7 / 23 PLOS ONE Marine landscape and biodiversity mapping using SBRUVs and DDV per [4] (Fig 3); DDV images differed from the SBRUV images due to the line of sight being perpendicular to the seafloor. Videos were analysed using Event Measure™ software [54] with a 3 minute lead in time allowing for sediment to settle. Prior to deployment, all SBRUV cameras systems were calibrated, measurements were conducted within confined controlled water and calibrated using CAL software (version 3.30) [59]. The maximum number of individuals pres- ent on the screen at any one time (MaxN) was recorded for each species. MaxN was determined by reviewing the footage and identifying the single frame in which the highest number of indi- viduals of a given species were simultaneously visible, in order to ensure consistency and avoid overestimation. A still image was captured from each station for substratum analysis at the beginning of each deployment. Each still image from SBRUV was cropped to a set aspect ratio that included the best average coverage, calculated by taking a random selection of 50 screen- grabs and calculating the average aspect ratio that would include the aspect ratio that encom- passed the best field of view (FOV) to capture both seabed and fauna, the resulting images was overlaid with a grid of 10 x 10 cells (DDV images being perpendicular to the seafloor did not require cropping) (Figs 4 and 5). The substratum that comprised the majority within each

grid square was recorded to create a proportional coverage of each substratum type at each station. Grid squares that could not be identified due to poor image clarity or obstruction were classified as unknown and excluded. Substrata were classed as algae, seagrass, maerl, gravel, mud, sand, boulder, and pebble with size based on measurements taken using EventMeasure™ [54]. Some substratum types such as boulders, seagrass, maerl, both living and dead, were too sparse in their occurrence to successfully produce robust data sets and were combined into a single category (e.g., maerl into the ‘gravel’ substratum type (Table 3)). Dead maerl, however, may also play an important role in the provision of interstitial spaces for a variety of species. Species of red, brown, and green algae were merged into a single designation to account for all algal types encountered. Still images were captured from all DDV deployments, two still images were used from each deployment, one at the drop and one from the haul, as these points could Fig 4. Grid overlay (10 × 10) of stereo baited remote underwater video (SBRUV) image seen from right hand camera. Screen grab of right hand SBRUV camera showing 10 × 10 grid overlay of image depicting substrata composition of South Arran survey point SA_201, each grid cell was classified by the substrata which com- prised > 50% of each grid cell. SBRUV bait arm is present on the left of the image. https://doi.org/10.1371/journal.pone.0319355.g004