FlowCAM optimization: Attaining good quality images for higher taxonomic classification resolution of natural phytoplankton samples

Abstract

AbstractTaxonomic data on phytoplankton composition is important for ecological studies, however, such information is not easy to gather. Imaging devices and image classification software have been developed in the past decades for rapid phytoplankton assessment. Taxonomic resolution output of classification software are primarily limited by the quality of images produced by these imaging instruments. FlowCAM has been utilized in several studies for this endeavor. However, the phytoplankton categories that the instrument is currently able to discriminate are still few compared to the outputs of microscopy. This study aimed to produce high resolution FlowCAM images of fixed phytoplankton samples from natural environments without compromising sample analysis time. It was also aimed to optimize the capability of FlowCAM's VisualSpreadsheet software to automatically classify phytoplankton images. The use of FOV300 flow cell and 10X objective combination has proven to be effective in producing good quality images at a faster rate. The modified hardware configuration resulted to FlowCAM counts that were comparable to that of the standard microscopy method. FlowCAM was able to automatically classify images of dominant phytoplankton groups in the two key sardine fishery areas in the Philippines with relatively high accuracy values. These phytoplankton groups are represented by genera with complex morphological structures (e.g., setae) such as Chaetoceros and Bacteriastrum as well as those genera with simple shapes such as Pseudo‐nitzschia (thin‐elongate) and Coscinodiscus (spherical).