Laboratory application of the Aquascope approach of automated imaging and classification for long-term plankton monitoring

Abstract

Anthropogenic changes such as climate change and pollution have strong effects on plankton and its ecosystem services. In freshwater, phytoplankton forms the basis of the food web and plays a key role in water quality. It is therefore important to monitor the plankton community. In Switzerland, lake plankton has been monitored for decades by cantonal authorities. But in the future, there will be a limitation in the availability of trained taxonomists that are able to count plankton samples under the microscope as well as time and budget constraints. New technologies will be needed to fulfil the federal requirements on the monitoring of lakes in the future. Here we focus on two such instruments, both based on plankton automated imaging: the FlowCam and the Aquascope. The first is a commercially available instrument, the second is custom made. They are the only automated imaging instruments that can track freshwater plankton across the desired size range (phytoplankton and zooplankton). Their design makes them the state of the art technologies to make quantitative observations of plankton. While the FlowCam is a laboratory instrument, the Aquascope is originally designed for field deployment. In this report, we focus the empirical work on the laboratory application of the Aquascope approach of automated imaging and classification for long term plankton monitoring. We find that there are major advantages of the Aquascope approach but the application as a laboratory instrument still needs further development. The major advantages of the Aquascope approach compared to traditional, human-supervised, microscopy methods are the automation and the speed of the data processing, the availability of individual level traits and the management of the data generated. One of the major drawbacks is the lack of taxonomic resolution (mostly at the genus level). Nevertheless, we show that with a larger benchmarking project, it might be possible to continue the long term time series on a coarse taxonomic level as usually reported by cantonal authorities to the general public. We discuss advantages and limitations of Aquascope relative to traditional microscopy and FlowCam, and propose ways forward for future development and application of this approach.