Evaluating a Most Probable Number Method for Assessing the Viability of Great Lakes Protists

Abstract

To support type approval testing of ballast water management systems we evaluated freshwater viability assessments for protists from the Duluth-Superior harbor of western Lake Superior using the most probable number (MPN) method. Tests were performed using varying temperatures and growth media and were compared to standard microscopic methods for determining live organism densities. Tests were also performed focusing on growth series derived from harbor water, and during an actual land-based test of a treatment system being evaluated for efficacy. We determined that growth of protists during MPN experiments was especially favored under higher temperatures and a growth medium comprising a 50 % solution of Bold Modified Basal Media. This medium also supported the growth of the greatest number of protist taxa. Based on microscopic analysis of live protists use of a treatment system during land-based testing reduced protist densities from 554 – 3000 cells/mL in the untreated water to 12 – 52 cells/mL after treatment. Corresponding assessments using the MPN method estimated respective densities of 1651 – 6060 cells/mL and 0 – 2.8 cells/mL, indicating that MPN likely overestimated viable cells in ambient harbor samples while it underestimated cell densities in treated samples. As asserted in the MPN protocols we confirmed that MPN-estimated protist densities were similar to densities in the protist size class that includes only cells strictly 10 – 50 µm in minimum dimension; protist densities including cells <10 µm were much higher than MPN estimates. However, based on all evaluations of freshly acquired samples containing a wide range of starting densities there was no correlation between MPN- and microscopy-determined densities, regardless of size class. Based on all testing, certain protist taxa were poorly favored during MPN grow-out periods (e.g., the chrysophyte Mallomonas), while others (e.g., free-living centric diatoms) tended to thrive, though there was substantial variability in taxonomic selectivity among tests. These findings contribute important freshwater data to the field of efficacy testing of ballast water treatment systems.