Surveying scallop populations with artificial intelligence

Scallops are a shellfish delicacy precious to many New Zealanders. They are an important kaimoana species for tangata whenua and have long been a sought-after species for commercial and recreational fishers.

Scallops are a shellfish delicacy precious to many New Zealanders. They are an important kaimoana species for tangata whenua and have long been a sought-after species for commercial and recreational fishers

Scallops are suspension-feeding bivalve molluscs that play a key role in the functioning of soft-sediment marine communities. They reproduce by ‘broadcast-spawning’ their eggs and sperm into the seawater. Spawning scallops need to be closely spaced in high density aggregations to ensure fertilisation success.But recent fisheries research surveys show scallop numbers have declined and are currently very low in many areas.

There are concerns about overfishing and habitat degradation from sedimentation and bottom-contact fishing methods.The impacts of scallop dredging on the Aotearoa New Zealand marine ecosystem in vulnerable areas like the Hauraki Gulf is a major concern.

To ensure scallop fisheries research surveys have the least impact possible, NIWA has been working with the University of Canterbury and Fisheries New Zealand to develop a non-invasive automated method of counting and measuring scallops to monitor their populations. The technology allows data to be collected without disturbing the seafloor or the animals living on or in it. This is all made possible with the use of underwater drones with cameras and very clever computer learning.

Currently under development, the technique involves using artificial intelligence to remove any need to use dredging in the future; divers validate how many scallops are on the seabed and what size they are, whilst the computer looks at imagery of the same sites.

The collection and analysis of imagery, using an underwater remotely operated vehicle, gives us the same level of precision as our traditional surveys, plus provides a wealth of new information on supporting benthic habitats, says NIWA fisheries scientist James Williams. 

Developing this technology allows researchers to gather data over larger areas and improve their understanding of scallop populations, associated marine life and supporting habitats without impacting on the marine environment. 

Surveying scallops populations with artificial intelligence