[M-IS20] Ocean Plastics, an earth science perspective

[MIS20-P01] Automatic detection for Nile Red-stained microplastic particles

*Masashi Tsuchiya¹, Tomo Kitahashi¹, Yosuke Taira², Hitoshi Saito², Kazumasa Oguri^1,3, Ryota Nakajima¹, Dhugal J. Lindsay⁴, Katsunori FUJIKURA¹, Tomohiko Fukushima^1,5 (1.Research Institute for Global Change, Japan Agency for Marine-Earth Science and Technology, 2.NEC Corporation, 3.Danish Center for Hadal Research, University of Southern Denmark, 4.Institute for Extra-cutting-edge Science and Technology Avant-garde Research, Japan Agency for Marine-Earth Science and Technology, 5.Deep Ocean Resources Development Co., Ltd.)

Keywords:Microplastics, Nile-Red staining, machine learning, automatic detection

Marine microplastics (MPs) pollution has become a major social problem. MPs have been found to be widespread not only in coastal areas but also in the open ocean, including in the polar regions and in the Mariana Trench. Due to the effects of waves, UV and physical weathering, plastics become finer during and after discharged into the ocean. However, transporting processes into the deep sea have not yet been clarified. Although seafloor is considered to be one of the sinks of MP, there is less information about them than about the floating MPs at the surface of the ocean. In this study, we developed a flow system (flow cell) in which NR-stained MPs were continuously monitored, and the fluorescent particles that passed through the flow cell were visualized under a fluorescent stereomicroscope. In addition, we constructed an automatic detection system, with machine learning, to measure the shape and number of MPs from the captured images. As a result of image acquisition and automatic discrimination using this system, we were able to discriminate MPs consisting of particles and fibers smaller than 330 µm. This system enables discrimination at more than 60 MPs per minute.