Benthic imagery from nearshore drop camera surveys along Eastern Shore (Nova Scotia, Canada) to characterize shallow subtidal habitats
Description
The Department of Fisheries and Oceans Canada (DFO) conducted nearshore drop camera surveys between 2019 and 2020 to characterize the distribution and abundance of seagrass, kelp, and other benthic macroalgae along the Eastern Shore of mainland Nova Scotia. Surveys were designed to address data gaps and support mapping work (species distribution modelling, remote sensing) to inform conservation and marine spatial planning in the Eastern Shore Islands Area of Interest and Goldboro area. Still images extracted from video transects were analyzed to evaluate the abundance (% cover) of eelgrass (<i>Zostera marina</i>), 8 species or genera of large brown macroalgae (<i>Saccharina latissima</i>, <i>Laminaria digitata</i>, <i>Agarum clathratum</i>, <i>Alaria esculenta</i>, <i>Saccorhiza dermatodea</i>, <i>Desmarestia</i> spp., <i>Fucus</i> spp., <i>Ascophyllum nodosum</i>), an invasive green alga (<i>Codium fragile</i> ssp. <i>fragile</i>), 5 functional groups of benthic macroalgae (filamentous turf algae, coarsely-branched turf algae, articulated coralline algae, encrusting algae, and foliose algae), detrital material, and 5 bare substrate cover categories (mud/sand, pebbles, cobbles, boulders, bedrock). In this Version 1, we report general geolocations and image labels describing the dominant cover component per image. Specific locations will be provided in a future update.
Files
Steps to reproduce
Images in this collection were extracted from video transects conducted at 265 stations spanning roughly 135 km (linear distance) from Jeddore Harbour to New Harbour on the Atlantic coast of Nova Scotia. We selected stations based on a stratified random design to sample a range of substrate types between -1 m − 12 m depth (Chart Datum). We deployed a drop camera system (SPOT X™ Pro Squid), comprising a watertight housing connected via an umbilical cable to a topside console, from the side of a small vessel. Lasers mounted 10 cm apart on the camera frame provided scale. No auxiliary lighting was required. Live stream video feed captured by a GoPro® HERO7 camera within the housing was viewed from an LCD screen at the surface and used to maintain the camera at a consistent height above the seabed or macrophyte canopy. Actual height of the camera off bottom varied with visibility, canopy height, and vessel motion. At each station, we allowed the camera system to drift with the passive movement of the vessel to continuously record benthic habitat (2.7k, 16:9 aspect ratio, 23.98 FPS). A GPS track delineating vessel position while the camera was on bottom was recorded with a handheld unit (Garmin GPSmap 62stc; 3-m accuracy). We assumed minimal displacement between the boat and camera positions at the depths sampled. Transects ranged from 10 m − 240 m with an average length of 33.7 m ± 22.3 m (mean ± SD). To measure cover of eelgrass and benthic macroalgae, for each transect we extracted images from the recorded video roughly every 4 m using open-source software (FFmpeg v 3.3.9). We assigned each image a unique filename comprising the station name and a sequential image number corresponding with the chronological order along the transect. Using ImageJ[1], we scaled and cropped images to a standard area (0.25 m<sup>2</sup>). We calculated percent cover of eelgrass, macroalgae, and bare substrate (21 categories listed above) using a point-count method by overlaying a 10 x 10 grid on the cropped image and determining the number of grid points intersecting each group. To align the dataset with existing benthic classification standards and facilitate inclusion in deep learning projects (e.g., automated image classification), we assigned labels to each image from 2 widely used classification schemes: the Coastal and Marine Ecological Classification Standard[2] and the CATAMI Classification Scheme[3]. The assigned label describes the biotic or substrate component dominating the seabed cover within the field of view. [1] Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671-675 [2] FGDC (2012) FGDC-STD-018-2012. Coastal and Marine Ecological Classification Standard. Reston, VA [3] CATAMI Technical Working Group (2013) CATAMI classification scheme for scoring marine biota and substrata in underwater imagery – Technical Report. Accessed [1-8-2022] http://catami.org/classifcation v 1.4