Influence of toxic equivalency factor scheme and method for treating non‐detect values on soil dioxin levels

Abstract

The purpose of this study was to evaluate the influence of four toxic equivalency factor (TEF) schemes (i.e., human/mammalian [Tl, T2], fish [T3], bird [T4]) and three substitution methods for treating non‐detect (ND) data (i.e., ND = 0 [Nl], ND = 0.5 DL [N2], ND = DL [N3] where DL = detection limit) on polychlorinated dibenzo‐p‐dioxin (PCDD) and dibenzofuran (PCDF) toxic equivalent (TEQ) concentrations in agricultural soils in Washington state (USA). Soil TEQs (n = 54) were determined for 12 combinations of TEF and ND levels and log‐transformed for statistical analysis. Significant bivariate correlations (Bonferroni p < .05) were observed for TEQ among all TEF levels within each ND level and between N2 and N3 within and among TEF levels. A two‐factor ANOVA demonstrated significant effects (p < .0001) for TEF, ND, and their interaction. Mean TEQ at T3 was significantly lower (p < .05) than other TEF levels at Nl, whereas mean TEQ at T4 was significantly higher (p < .05) than Tl both at N2 and N3. Mean TEQs at all ND levels differed significantly (p < .05) at each TEF level with Nl < N2 < N3. These results illustrate that TEF and ND methods can notably impact calculations for low concentrations of soil dioxin.

Key words:

• Risk assessment
• dioxin
• toxic equivalency factor (TEF)
• toxic equivalent (TEQ)
• non‐detect data
• censored data

Notes

Corresponding author. Tel.: [phone]; Fax: [phone]; E‐mail: [email protected].