FN Archimer Export Format
PT J
TI High-Resolution Vertical Habitat Mapping of a Deep-Sea Cliff Offshore Greenland
BT 
AF Van Audenhaege, Loic
   Broad, Emmeline
   Hendry, Katharine R.
   Huvenne, Veerle A. I.
AS 1:1,2;2:1,3;3:4;4:1;
FF 1:PDG-REM-EEP-LEP;2:;3:;4:;
C1 National Oceanography Centre, Southampton, United Kingdom
   Ifremer, Centre de Bretagne, REM/EEP, Laboratoire Environnement Profond, Plouzané, France
   School of Ocean and Earth Science, University of Southampton, Southampton, United Kingdom
   School of Earth Sciences, University of Bristol, Bristol, United Kingdom
C2 NOC, UK
   IFREMER, FRANCE
   UNIV SOUTHAMPTON, UK
   UNIV BRISTOL, UK
SI BREST
SE PDG-REM-EEP-LEP
IN WOS Ifremer UPR
   DOAJ
   copubli-europe
IF 5.247
TC 2
UR https://archimer.ifremer.fr/doc/00700/81223/85512.pdf
   https://archimer.ifremer.fr/doc/00700/81223/85513.pdf
LA English
DT Article
DE ;marine habitat mapping;deep-water vertical cliff;ROV;multibeam echosounder;terrain point cloud;Greenland glacial trough;suspension-feeding community;underwater exploration
AB Recent advances in deep-sea exploration with underwater vehicles have led to the discovery of vertical environments inhabited by a diverse sessile fauna. However, despite their ecological importance, vertical habitats remain poorly characterized by conventional downward-looking survey techniques. Here we present a high-resolution 3-dimensional habitat map of a vertical cliff hosting a suspension-feeding community at the flank of an underwater glacial trough in the Greenland waters of the Labrador Sea. Using a forward-looking set-up on a Remotely Operated Vehicle (ROV), a high-resolution multibeam echosounder was used to map out the topography of the deep-sea terrain, including, for the first time, the backscatter intensity. Navigational accuracy was improved through a combination of the USBL and the DVL navigation of the ROV. Multi-scale terrain descriptors were derived and assigned to the 3D point cloud of the terrain. Following an unsupervised habitat mapping approach, the application of a K-means clustering revealed four potential habitat types, driven by geomorphology, backscatter and fine-scale features. Using groundtruthing seabed images, the ecological significance of the four habitat clusters was assessed in order to evaluate the benefit of unsupervised habitat mapping for further fine-scale ecological studies of vertical environments. This study demonstrates the importance of a priori knowledge of the terrain around habitats that are rarely explored for ecological investigations. It also emphasizes the importance of remote characterization of habitat distribution for assessing the representativeness of benthic faunal studies often constrained by time-limited sampling activities. This case study further identifies current limitations (e.g., navigation accuracy, irregular terrain acquisition difficulties) that can potentially limit the use of deep-sea terrain models for fine-scale investigations. 
PY 2021
PD JUL
SO Frontiers In Marine Science
SN 2296-7745
PU Frontiers Media SA
VL 8
UT 000669348600001
DI 10.3389/fmars.2021.669372
ID 81223
ER

EF