None detected: What “zero” indicates in direct counts of aquatic microorganisms in treated ballast water

Abstract

International limits on the concentrations of living organisms in ballast water are now in force for commercial ships. Microscopy-based, “direct count” assays estimate the concentrations of organisms. These assays are used in performance tests of Ballast Water Management Systems (BWMSs), which are shipboard technologies designed to reduce living organisms in ballast water to below the discharge limits. Here, we examine the factors that affect the method detection limit (MDL) of direct count assays. The MDL depends upon the volumes sampled, concentrated (or diluted), and analyzed, so the MDL will vary as these volumes vary from analysis-to-analysis. Decreasing the MDL to detect exceedingly rare individuals (e.g., ~1 individual per m3 or fewer) is possible, but problematic, given the challenges in analyzing dynamic communities of living organisms: increasing sample and analysis volumes to lower the MDL will likely accelerate the loss of individuals, as organisms are concentrated to several orders of magnitude above in situ concentrations and held for extended times. Results of direct count assays may be widely disseminated as an indicator of the performance of BWMSs, and when no organisms are detected, concentrations may be shown as “0 organisms per m3 or mL” rather than “<MDL”, as appropriate. When clearly derived and reported, a numerical MDL for the direct count assay—the lowermost value distinct from “zero”—provides context to results and transparency into the sensitivity of the assay.