Effectiveness of drone imagery in mapping shallow-water benthic assemblages

Abstract

The present study aimed to assess the efficiency of using drones to map shallow-water benthic assemblages, compared with that of manual survey techniques. It also aimed to develop an ‘optimum protocol’ to characterize shallow-water benthic assemblages. The efficiency of mapping was assessed in terms of its comparative accuracy and cost-effectiveness. The study was carried out in six coastal sites in Malta, characterised by varying benthic complexity and fragmentation. The bays were imaged using a DJI Mavic 2 Pro drone, at altitudes between 100-300m above mean sea level. Aerial images obtained were subsequently processed using k-means clustering to distinguish between different benthic assemblages, based on their colour profiles. The results of the automated drone surveys were then compared with those of the direct manual surveys to assess their relative efficiencies in terms of area mapped per unit effort. Results obtained indicated that using a drone to map benthic cover greatly reduced the time and manpower requirements relative to direct surveys. This method also generated biotope maps with greater coverage, accuracy, and efficiency relative to manual methods. Nonetheless, while a general procedure was followed, an optimum protocol could not be produced, as each study site was treated according to its own specific characteristics.