2.1.1 Sampling strategies
In accordance with the NSSP, NCDMF routinely samples all shellfish growing areas on a regular basis (6 times annually) using the systematic random sampling strategy, meaning samples for each station are collected at randomly scheduled timepoints throughout the year; however, there are some constraints as to when sample collection is permissible. Specifically, samples are only collected during conditions when the SGAs are open for harvest or assumed to be unimpacted by unsafe FC levels, resulting in the subsequent data not capturing peak FC concentrations. Because freshwater input and runoff are tied to increased FC concentrations in estuarine waters, precipitation intensity is used as a management indicator for closures of conditionally approved waterways (NSSP, 2019; Leight et al., 2016). In North Carolina, rainfall closure thresholds in conditionally approved waters range from 1-3 inches of rainfall within a 24-hour period (Figure 1c) and dictate if a managed shellfish area will be closed for harvesting after a meteorological event. The “emergency closure” of additional areas can occur after higher rainfall amounts are noted and, for the purposes of this analysis, those areas will be represented by a 4 inch threshold. Rainfall thresholds are assigned to conditionally managed areas. To reopen closed areas that have been temporarily closed following exceedances of those rainfall thresholds, NCDMF tests the water, and reopening will only occur after samples confirm safe harvest conditions, which is defined by waters that do not exceed a FC density of 14 MPN per 100 mL of sample (NSSP, 2019); sampling for reopening is hereafter referred to as “conditional sampling”. NCDMF balances its limited resources with speed of reopening by only conditionally sampling when the organization suspects the FC concentrations will be low enough to support reopening. SGAs A, C, D, E, and the northern half of B are characterized by moderate to low rainfall thresholds (i.e., 1-3 inches) whereas SGAs F, G, H, I, and the southern half of B have high rainfall thresholds (i.e., 4 inches; Figure 1c).
Under systematic random sampling, routine sampling in conditionally approved waters must occur when the SGA is open and not during a temporary closure. Accordingly, this stipulation creates bias in the routine monitoring data. As such, areas with a 1 inch rainfall threshold will close more often than areas with 4 inch thresholds, resulting in more restrictive sampling conditions and more homogeneous water quality samples in the long-term. Furthermore, NCDMF will change rainfall thresholds or exclude stations for areas with persistently poor water quality, resulting in the longest-standing monitoring stations typically being located in growing areas that are known to be productive with relatively good water quality, contributing another source of bias to the data.