The toxicity of neodymium substances to soil invertebrates in a boreal soil, and the impacts of soil type, aging, and leaching on metal bioavailability and toxicity.

Abstract

A clear gap in ecotoxicological data was identified regarding the impacts of the technologically critical lanthanide neodymium (Nd) on soil invertebrates in boreal soil. To address this gap, two inorganic metal salts (NdCl3 and Nd2(SO4)3), an organometallic compound Nd-2-ethylhexanoate (Nd-2-EHA) and its organic component 2-ethylhexanoic acid (2-EHA) were examined. Toxicity exposures were conducted using the earthworm Dendrodrilus rubidus, the collembolan Proisotoma minuta, and the oribatid mite Oppia nitens, in two Canadian boreal soil horizons (i.e., organic and mineral), using standardized soil toxicity test methods; effects included survival and reproduction. Different exposure scenarios were investigated, including aging of test soil for 6 months, and the impact of leaching on toxicity test endpoints. Differences in species sensitivity were observed, with D. rubidus identified as the most sensitive test species. The study highlighted the importance of reporting chemical parameters within toxicity tests, such as electrical conductivity and pH, to better understand test system influences and factors affecting toxicity. Toxicity estimates from leached and aged NdCl3 exposures provided the least confounded assessment of Nd toxicity (reproductive median inhibitory concentration [IC50] for D. rubidus: 890 [767-950] mg Nd kg-1 dry soil), as leaching reduced salt-related effects. Incorporating measures of extractable Nd proved valuable for comparing toxicity across the substances. Despite the lowest measured extractable and total Nd, the organometallic, Nd-2-EHA, was the most toxic compound tested (reproductive IC50 for D. rubidus: 17 [10-30] mg Nd kg-1 dry soil), largely influenced by the organic component (2-EHA) when tested at equivalent concentrations. Exposure to 2-EHA and Nd-2-EHA demonstrated a significant decrease in toxicity with aging. These findings aligned with existing frameworks such that a metal-moiety approach may be insufficient for organometallics, wherein inorganic and organic influences must be considered.