FRESHWATER VERIFICATION OF THE MICROWISE BALLASTWISE COMPLIANCE MONITORING DEVICE

Abstract

This technical report presents findings from freshwater verification tests evaluating the performance of the MicroWISE BallastWISE compliance monitoring device, hereafter BallastWISE. BallastWISE was developed by MicroWISE, located in Ebeltoft, Denmark.

The compliance monitoring device evaluation began in August 2020 and ended in December 2020, at the Lake Superior Research Institute (LSRI) of the University of Wisconsin-Superior (UWS) in Superior, Wisconsin, USA. BallastWISE utilizes separate chambers to enumerate organisms in each of two regulated size classes, ≥10 and <50 µm (nominally protists) and ≥50 µm (nominally zooplankton). Cameras and optical chambers capture video and track motility through software analysis for the zooplankton size class. Fluorescence microscopy evaluates chlorophyll containing organisms in addition to motility tracking in the protist size class.

The verification testing was composed of three phases. Phase I testing was completed in two water types with laboratory-cultured organisms in the two regulated size classes, utilizing the single-celled protist Haematococcus pluvialis and colonial protist Scenedesmus quadricauda, and the zooplankton Eucyclops spp. and Daphnia magna. Phase II testing was completed using naturally occurring Great Lakes organisms in the Duluth-Superior Harbor of Lake Superior in the two regulated size classes. Phase III testing was completed using Duluth-Superior Harbor water and ambient organisms before and after treatment with a ballast water treatment technology (BWT) during three land-based trials. Data from all phases were analyzed for precision, accuracy, and reliability. Quantification/detection limits were also calculated from Phase I data . Phase I testing showed BallastWISE was effective at quantifying single-celled protists to within about 20% of the microscopic counts, but undercounted colonial protists. Colonial protist entity counts were close to microscopic entity counts suggesting that individuals within the colonies were not resolved. High total suspended solids (TSS) and (DOC) may slightly reduce BallastWISE sensitivity to protists. BallastWISE overcounted zooplankton in both species tested in both high and low TSS/DOC by between 150% and 420%. Phase II testing from the Duluth-Superior Harbor showed BallastWISE counts of natural assemblages of protists strictly in the ≥10 and <50 μm size class to be slightly below microscopic counts by about 35% and with high precision. Zooplankton were overestimated by BallastWISE by roughly 40% and with considerably more variation compared to microscopic counts. Phase III testing showed low BallastWISE accuracy and precision in untreated protist and zooplankton samples. This may have been caused by organism densities higher than the device’s effective upper limit of detection in the zooplankton samples, but further investigation would be needed to determine the cause of low accuracy and precision in protist analysis. BallastWISE accurately measured treated protist samples as 0 cells/mL in agreement with strict microscopic counts, but overcounted treated zooplankton samples in 2 out of 3 tests, possibly due to the method of treatment. A number of operational issues made enumeration of zooplankton unreliable, but improvements (e.g., software updates, guidance on device operation) from the developer over the period of this assessment have already improved performance. BallastWISE shows promise as a useful device for detecting and measuring protists and zooplankton in the Great Lakes as additional improvements are made.