Comparisons of discrete and integrative sampling accuracy in estimating pulsed aquatic exposures. (November 2016)

• Record Type: Journal Article
• Authors: Morrison, Shane A.; Luttbeg, Barney; Belden, Jason B.
• Is Part Of: Environmental pollution. Volume 218(2016)
• Page Start: 749
• Page End: 756
• Publication Date: 2016-11
• Subjects: Environmental monitoring -- Integrative sampling -- Pulsed exposure -- Spot sampling; Pollution -- Periodicals; Pollution -- Environmental aspects -- Periodicals; Environmental Pollution -- Periodicals; Pollution -- Périodiques; Pollution -- Aspect de l'environnement -- Périodiques; Pollution -- Effets physiologiques -- Périodiques; Pollution; Pollution -- Environmental aspects; Periodicals; Electronic journals; 363.73
• DOI: 10.1016/j.envpol.2016.07.071
• Languages: English
• ISSNs: 0269-7491

Abstract

Abstract: Most current-use pesticides have short half-lives in the water column and thus the most relevant exposure scenarios for many aquatic organisms are pulsed exposures. Quantifying exposure using discrete water samples may not be accurate as few studies are able to sample frequently enough to accurately determine time-weighted average (TWA) concentrations of short aquatic exposures. Integrative sampling methods that continuously sample freely dissolved contaminants over time intervals (such as integrative passive samplers) have been demonstrated to be a promising measurement technique. We conducted several modeling scenarios to test the assumption that integrative methods may require many less samples for accurate estimation of peak 96-h TWA concentrations. We compared the accuracies of discrete point samples and integrative samples while varying sampling frequencies and a range of contaminant water half-lives (t50 = 0.5, 2, and 8 d). Differences the predictive accuracy of discrete point samples and integrative samples were greatest at low sampling frequencies. For example, when the half-life was 0.5 d, discrete point samples required 7 sampling events to ensure median values > 50% and no sampling events reporting highly inaccurate results (defined as < 10% of the true 96-h TWA). Across all water half-lives investigated, integrative sampling only required two samples to prevent highly inaccurate results and measurements resulting in median values > 50% of the true concentration. Regardless, the need for integrative sampling diminished as water half-life increased. For an 8-d water half-life, two discrete samples produced accurate estimates and median values greater than those obtained for two integrative samples. Overall, integrative methods are the more accurate method for monitoring contaminants with short water half-lives due to reduced frequency of extreme values, especially with uncertainties around the timing of pulsed events. However, the acceptability of discrete sampling methods for providing accurate concentration measurements increases with increasing aquatic half-lives. Graphical abstract: Highlights: We compared the accuracies of discrete point and integrative samples. The influence of aquatic half-lives and sample frequency was evaluated. Higher sample frequency provided the best estimate of toxicologically-relevant concentrations. Integrative methods reduced the frequency of extreme values. Abstract : Integrative methods provided better predictions of the true 96-h TWA concentration when water dissipation was rapid; however, discrete samples were equally effective for long aquatic half-lives.