Revisiting the Avian Eco‐SSL for Lead: Recommendations for Revision

Abstract

The avian ecological soil screening level (Eco‐SSL) for Pb (11 mg/kg) is within soil background concentrations for >90% of the United States. Consequently, its utility as a soil screening level is limited. Site‐specific ecological risk–based remedial goals for Pb are frequently many times greater. Toxicity reference values (TRVs) play a major role in defining Eco‐SSLs. The Pb Eco‐SSL TRV is driven by reduced egg production in Japanese quail (Coturnix japonica), which displays effects at doses both substantially lower and greater than other tested species. High variability in egg production in Japanese quail has also been observed for other contaminants. Japanese quail egg production may therefore be too variable and unreliable an effect endpoint upon which to base regulatory screening criteria. Toxicity data supporting the Eco‐SSL were reevaluated and only studies reporting both no and lowest observed adverse effect levels (NOAELs and LOAELs) for reproduction, growth, or survival were considered. Dose–response data were extracted from 10 studies both as concentrations and doses. Dose–response relationships were developed using the US Environmental Protection Agency (USEPA) Benchmark Dose Software for dietary concentrations and doses for egg production in Japanese quail and chickens. Effect levels (of 10%, 20%, and 50%) were extracted from the dose–response analyses. Species sensitivity distributions and dose–response data indicated reproduction was most sensitive to Pb and survival was least sensitive, with growth intermediate. Limited data for ringed turtle doves (Streptopelia risoria) and American kestrels (Falco sparverius) suggest lower sensitivity than chickens to Pb. The ED10 and ED20 thresholds for chickens were 4.4 and 9.8 mg·kg–1·d–1, respectively. Avian Pb Eco‐SSLs were recalculated based on the chicken ED10 and ED20, with and without a bioavailability adjustment. Revised avian Pb Eco‐SSLs for the most highly exposed species (American woodcock), based on the ED10 and assuming 100% and 50% bioavailability, were 36.3 mg/kg and 43.7 mg/kg, respectively. Integr Environ Assess Manag 2019;15:739–749. © 2019 SETAC.

Wildlife ecological risk assessment in the 21st century: Promising technologies to assess toxicological effects

“…There may be trade-offs between optimizing the analysis for statistical considerations versus incorporating relevant biological information. For example, normalizing observed biological responses to control results has often been viewed as a convenient way of harmonizing data from multiple studies and facilitating comparability among species (e.g., Fuchsman et al, 2008;Sample et al, 2019), but this approach can lose valuable information on normal variation in biological performance that may reflect data quality and animal husbandry (Blankenship et al, 2008). Normalizing responses to the control violates the assumption of independence that is central in parametric analysis and, depending on the analysis method, has the potential to introduce quantitative distortion of results (Green, 2014(Green, , 2016.…”