Gulf of Alaska | Lower Trophic Levels
High frequency observations of plankton distributions from the Prince William Sound plankton camera Presenter: Rob Campbell , rcampbell@pwssc.org, Prince William Sound Science Center A novel plankton imager was developed and deployed aboard a profiling mooring in Prince William Sound in 2016 to 2022. The imager consists of a 12-MP camera and a 0.137x telecentric lens, along with darkfield illumination produced by an in-line ring/condenser lens system. The profiler travels from ~60 m to the surface twice daily from March to November, producing a high resolution record of the plankton and biogeochemistry of the surface ocean during the growing season. Almost 4x106 images of individual plankters and particulates have been collected since 2016. A subset of ~2x104 images was manually identified into 43 unique classes, and a hybrid convolutional neural network classifier was developed and trained to identify the images. Classification accuracy varied among the different classes, and applying thresholds to the output of the neural network (interpretable as probabilities or classifier confidence), improved classification accuracy in non-ambiguous groups to between 80% and 100%. The identified images provide observations of individual plankters, with millisecond timestamps which may be used to infer the depth of each plankter to within ~10 cm from the pressure-time record logged by the profiler. Visualizations of those individual plankters shows high frequency variability in several groups, including short blooms (1-2 weeks) of Limacina pteropods and Oikopleura larvaceans , and variations in depth distributions over the growing season. A pronounced vertical migration was observed in Metridia copepods, and a high frequency (hourly) cycle of profiles was able to resolve the daily migration. Image based observations show promise for learning about the fine-scale dynamics and habits of numerous plankton taxa.
Alaska Marine Science Symposium 2023 132
Made with FlippingBook flipbook maker