The sea squirt Didemnum vexillum is a highly invasive colonial tunicate species found in the waters off the coast of New England and Canada. Although Didemnum is thought to impact fisheries, foul aquaculture facilities and alter native communities, the extent of its role as an ecological engineer remains largely unknown. More information regarding its spatial coverage is necessary in order to evaluate the ecological impacts and guide management decisions regarding this species.
Because digital cameras and video systems cover limited areas during cruises and require significant manual post-cruise data analysis, we propose to develop and test an optical sensor for detection of Didemnum via an autonomous underwater vehicle (AUV). The sensors and digital cameras on the AUV will help us map the spatial coverage of the tunicate in selected areas of the Gulf of Maine and Georges Bank, and allow us to examine benthic species diversity in the presence and absence of large mats of Didemnum. Results from this project will be used by MIT Sea Grant to develop web-based outreach materials to raise awareness of these tunicate infestations.
The first task will test the efficacy of identifying Didemnum using a system employing a hyperspectral spectroradiometer combined with an active light source. Preliminary data suggest that Didemnum has a unique spectral signature that will permit identification of areas of Didemnum mat infestation using intelligent classification algorithms. The radiometer-based approach has definite advantages in autonomous vehicle deployments, including low power consumption, high spatial resolution, and rapid data analysis. Development of the novel method will focus on laboratory analysis of field samples and field tests of a prototype system with divers in both nearshore and subtidal areas.
The first year effort will include: (1) examination of spectral reflectance of Didemnum and other common bottom materials (from different regions and seasons) from the Gulf of Maine in a laboratory setting, (2) development of feature extraction and classification algorithms to distinguish Didemnum from other bottom materials, and (3) design of a field-deployable prototype system with spectroradiometer and an active light source.
The second year objectives include: (1) thorough testing of the prototype radiometer system and classification algorithms in different regions and seasons; and (2) adaptation of the system to AUV missions.