Author
 |
Shappell, Nancy |
|
Submitted to: Integrated Environmental Assessment and Management
Publication Type: Other Publication Acceptance Date: 7/7/2016 Publication Date: 9/18/2016 Citation: Shappell, N.W. 2016. A transformational approach to improve interpretation of environmental data. Integrated Environmental Assessment and Management. 12(4):613-614. https://doi.org/10.1002/ieam.1810. DOI: https://doi.org/10.1002/ieam.1810 Interpretive Summary: Technical Abstract: In “iStreem® : an approach for broad-scale in-stream exposure assessment of “down the drain chemicals,” “ Kapo et al. (2015) modeled stream concentrations of two chemicals (a fragrance product – HHCB, and the insect repellant - DEET) from USGS data obtained across a large contiguous portion of the United States. In Figure 3 of the paper, the stream data were presented graphically as the cumulative percent rank (0-100%) on the Y axis and the stream concentration (µg/L) on the X axis using log scale (see Figure 1 below). Included on the graphs were the Predicted No Effect Concentration (PNEC) for each compound and the iStreem predicted concentrations. In my opinion, this traditional representation of the data poses two limitations. 1) The graphical representation using the traditional log scale from low to high concentration on the X axis left to right results in an expanded scale for the lowest of concentrations, and a nearly illegible scale for the highest concentrations. By reversing the X axis to run from highest to lowest concentrations (left to right) this situation is ameliorated (Figure 2). 2) Figure 1 as presented by Kapo et al. (2015)fails to clearly distinguish which data actually reflect samples that were at, or below, the detection limits for those locations (stated as either 0.05 or 0.5 µg/L in the supporting information, but not in the text or figure legend). I have taken the liberty of graphing crude estimates of the X and Y values for HHCB reported by USGS and used by Kapo et al. (2015) (original Figure 3A) The resulting figure provides for 1) ease of estimation of percentage of streams with highest HHCB concentrations, and their proximity to the established PNEC; 2) a clear indication of the disparity between detection limits across locations, and therefore actual concentration uncertainties reflected by those data points; and 3) discernment of potential need for lower detection limits to obtain a margin of safety greater than 20 below the PNEC. Such a margin of safety would not be possible with a PNEC of10 µg/L and a detection limit of 0.5 µg/L at ~50% of the locations. This small formatting change in data presentation would lead to more accurate assessment of the critical data provided on chemicals of environmental concern. Nancy W. Shappell, Biosciences Research Laboratory Agricultural Research Service, USDA Fargo, ND, USA |
