Optimizing methods to estimate zooplankton concentration based on generalized patterns of patchiness inside ballast tanks and ballast water discharges.

Zooplankton populations are spatially heterogeneous in nature and inside ship ballast tanks. Sampling methods should take heterogeneity into account, particularly when estimating quantitative variables such as abundance or concentration. It is particularly important to generate unbiased estimates of zooplankton concentration in ballast water when assessing compliance with new international ballast water discharge standards. We measured spatial heterogeneity of zooplankton within ballast water using three sampling methodologies. In-tank pump samples were collected at fixed depths within the vertical part of the ballast tank (side tank). Vertical net-haul samples were collected from the upper portion of the tank as a depth-integrated and historically relevant method. In-line, time-integrated samples were collected during ballast discharge by an isokinetic sample probe, likely representing the double bottom part of the ballast tank. The bias and precision associated with each sampling method were evaluated in reference to the estimated average abundance of the entire ballast tank, which was modeled from the data collected by all methods. In-tank pump samples provided robust evidence for vertical stratification of zooplankton concentration in the side tank. A consistent trend was also observed for in-line discharge samples, with zooplankton concentration decreasing through time as the ballast tank is being discharged. Sample representativeness, as compared to the tank average, varied depending on the depth or tank volume discharged. In-line discharge samples provided the least biased and most precise estimate of average tank abundance (having lowest mean squared error) when collected during the time frame of 20%-60% of the tank volume being discharged. Results were consistent across five trips despite differences in ballast water source, season, and age.