Field Deployed Ni-Amended Sediments Shows Varying Effects in Two Central MIchigan Streams

Abstract

Most research on Ni contaminated sediments has occurred under laboratory conditions, which can lead to exposure to formulated water that is not representative of environmentally relevant conditions. For example, laboratory conditions don’t accurately represent natural fluctuations in temperature, light, or flow. The purpose of our study was to 1) assess the bioavailability and toxicity of Ni contaminated sediments to stream macroinvertebrate communities under in situ conditions and 2) compare these effects across two similar streams in close proximity of each other. Sediments were removed from two, second order streams in central Michigan (Black Creek (BC) and Little Molasses (LM)) that were similar in hardness, pH, dissolved oxygen (DO), dissolved organic carbon (DOC), and alkalinity. The sediments were directly spiked with Ni at two treatment levels, low nickel (210 mg/kg dry weight) and high nickel (1889 mg/kg dry weight), equilibrated for 10 days, and re-deployed in macroinvertebrate colonization baskets within the respective streams. Sediment geochemistry (e.g. AVS, SEMNi, total Ni, total Fe, and total Mn) and water quality measurements (e.g. hardness, alkalinity, pH, and dissolved organic carbon (DOC)) were determined at deployment and three successive sampling periods (14, 28, and 56 days post-deployment). In situ chambers with Hyallela azteca were used to conduct 96-h acute toxicity tests at day 0 and 14. In addition, we deployed a microbial decomposition test from day 0 to 14 to assess Ni effect on the microbial community. We found varying responses to Ni treatments in each stream. The 96-h acute toxicity tests only showed significant toxicity in the sediment-water interface exposure at day 0 for BC. Both Ni treatments adversely impacted the microbial community in BC, but not in LM. The benthic community recolonization response was limited with only the family Gammaridae showing a response to Ni treatment in both streams on day 14. Stepwise stream-dependent regression of the measured

variables showed that the SEMNi/AVS model of potential toxicity significantly predicted Gammaridae abundance in BC, but not LM. BC showed a trend of toxicity from Ni-amended sediments with increased mortality to H. azteca (acute exposure), the microbial community, and Gammaridae.