Governments, scientists, and economists all agree that we urgently need to step up our ocean conservation game – not only to combat overwhelming biodiversity loss and climate change, but also to support over 30 million blue economy jobs. There is even widespread consensus about minimum conservation targets required to prevent irreversible damage: 72 countries have joined the 30x30 initiative, which calls for at least 30% of global oceans to be designated Marine Protected Areas (MPAs) or have similar status by 2030. With only 6% of the world’s oceans currently under any form of protection, this ambitious target underscores how dramatic a shift towards conservation is needed this decade.
Photo credit: Marine Protection Atlas/MPAtlas.org
When properly managed, MPAs are effective tools for the restoration and long-term protection of critical marine habitats and biodiversity. While there are clear targets for the square footage of MPAs to be created, the real-world impact of these zones is contingent upon effective planning, management, monitoring, and enforcement, which pose a more complex challenge. Crucially, the effectiveness of MPAs is not typically described in national and international reports, which leads to a disconnect between protected status on paper and actual ecosystem recovery. In Canada, only ten of 18 MPAs have active management and monitoring plans, records of monitoring, or enforcement activities in place.
One of the main barriers to effective MPA management is our ability to efficiently monitor them. Collecting and analysing data is central to identifying biodiversity hotspots of particular ecological importance, measuring changes in animal populations and progress towards conservation goals, and making evidence-based adjustments to policy and management strategies. Monitoring MPAs is difficult in large part because they can span large, remote areas. According to the Marine Protection Atlas, there are 36 MPAs each covering over 100 000 km2, the largest of which encompasses over 1.6 million km2 off the coast of Antarctica.
Adopting technological innovations and new scientific research techniques to monitor current and future MPAs is a critical component to feasibly scaling ocean protections. For example, eDNA metabarcoding is a novel technique for rapidly surveying general species biodiversity by amplifying trace DNA found in water samples, and in some cases can replace destructive and time-consuming trawl surveys. Another major avenue for scaling MPA management is the use of remote sensed imagery – non-invasive imagery taken from planes, drones, or satellite – to track macroscopic oceanographic trends such as productive phytoplankton blooms, or combining satellite data with ship existing ship positioning information to track commercial fishing efforts and monitor compliance at sea.
One powerful application of remote sensed imagery is using aerial images to obtain finer-scale data related to marine mammal demographics, abundance, spatial distributions, seasonal migration patterns, and critical habitats for foraging or calving. The health of large-bodied, long-lived species such as whales (considered sentinel species) is an important indicator for overall ecosystem health – an important metric for MPAs, which intend to preserve entire ecosystems.
Above: an aerial image taken in the Canadian Arctic. Whale identifications (green) made by Möbius, Whale Seeker’s automated detection system. Photo credit: Department of Fisheries and Oceans Canada
To date, aerial whale surveys have been key for assessing sensitive whale populations such as Arctic bowhead whales and belugas, and the critically endangered North Atlantic right whale. Its application as a widespread MPA monitoring tool has been limited, however, largely due to the resource and time-intensive image analysis process which requires human analysts to individually identify and count whales.
Recognising the current limitations of aerial image annotation and the urgent need for reliable, scalable solutions for monitoring vast, remote MPAs, Whale Seeker created Möbius. Our flagship technology leverages artificial intelligence to not only decrease turnaround time for aerial surveys, but also to standardise annotation protocols, allowing for more scientifically rigorous comparisons of assessments across survey years and regions. Most importantly, AI helps us take MPA monitoring to the big leagues by making whale population assessments truly manageable at large scales.
To make meaningful progress towards protecting ocean biodiversity, we must hold our growing marine protected areas to high standards, ensuring active management and monitoring. We believe that AI-assisted marine mammal monitoring has an important role to play by lowering the barriers to making dynamic, data-driven conservation policy decisions.