You are here:
Retrospective analyses of aquatic freshwater species sensitivity using reference toxicant testing data
Citation:
Lazorchak, Jim, R. Martin, H. Haring, W. Thoeny, AND Mohamed Ghorab. Retrospective analyses of aquatic freshwater species sensitivity using reference toxicant testing data. 2020 SETAC SciCon2, Virtual, November 15 - 19, 2020.
Impact/Purpose:
Retrospective look at laboratory cultures sensitivity to reference toxicants.
Description:
KCl is a reference toxicant studied in our laboratory in relation to its effects on seven species: Pimephales promelas, Daphnia magna, Ceriodaphnia dubia, Neocloeon triangulifer, Hyalella azteca, Chironomus dilutus and Lumbriculus variegatus. In a retrospective analysis of tests for these species dating back to 1998, we assessed variation in sensitivities using control charts, simple calculations of the coefficient of variation (CV) metric, and a Bayesian hierarchical model. For each species, we calculated the CV for control charts organized for the first 20, middle 20, and last 20 tests, as well as over the entire period of record. The Bayesian hierarchical model was then fit to all data to estimate means and variation within and between species and over the period of record. CV calculations for the 96-hr acute P. promelas tests, based on 128 tests in eight years, ranged from 12% to 19% . CVs for the 48-hr acute D. magna tests, utilizing results from 81 tests conducted over nine years, ranged from 3% to 11%. CVs for the the D. magna 4-day growth tests ( N = 41) ranged from 14.9% to 15.6%. CVs for the 48-hr acute C. dubia tests ranged from 17.3% to 24.4 % across four test sets. Thus, control charts CV calculations for all tests were near or well below 20%, suggesting that the toxicity information generated in those studies was reasonably reliable and reproducible. Estimates from the Bayesian model indicated that variability in (log-scale) mean acute test LC50 (K mg/L) was much larger between species (sd = 0.77) compared to tests within (sd = 0.19). Additionally, year-to-year variation in overall mean LC50 was estimated to be quite low (sd = 0.12); and year-to-year variation in mean LC50 within species even lower (sd = 0.07). Finally, mean LC50 estimates, averaging over the period of records, were highest for C. dilutus (96-hr; 95% credible interval (CI) for mean = 4335 – 5771 mg/L), followed by P. promelas (48-hr; CI = 803 – 928 mg/L), D. pulex (48-hr; CI = 641 – 779 mg/L), D. magna (48-hr; CI = 612 – 731 mg/L), L. variegatus (96-hr; CI = 559 – 686 mg/L), C. dubia (48-hr; CI = 412 – 478 mg/L), D. pulex (96-hr; CI = 364 – 511 mg/L), and H. azteca (48-hr; CI = 306 – 374 mg/L).