RECOMMENDED ASSESSMENT PROTOCOLS FOR USE WITH AMBIENT TOXICITY TESTS IN THE 305(B) REPORTING AND 303(D) LISTING PROCESS

All states have at least narrative statements concerning the preclusion of toxics in toxic amounts in their water quality standards. However, most states do not include numeric water quality standards for toxicity of ambient water or sediment. This is further complicated by the lack of regulations addressing sediments. Currently there is very little guidance available to states on evaluating both lethal and sublethal ambient toxicity data for waters and sediments for use in developing the 305(b) and 303(d) lists. Parsons is currently conducting an assessment of ambient toxicity in Texas on water bodies based on 6 stream segments for sediment toxicity and 4 stream segments for water toxicity. This assessment will provide a means to address observed impairments and to make decisions regarding an appropriate strategy to restore water quality.This paper describes an assessment methodology for ambient water and sediment toxicity testing data in Texas, which will be used to identify water quality impairments in the TMDL area. These assessment criteria have been recommended to the Texas Natural Resource Conservation Commission (TNRCC) for use in determining attainment of aquatic life uses in surface waters. States may chose to use the criteria when developing their 305(b) Report and 303(d) List required by the Clean Water Act. Whole sediment toxicity tests are used in the assessment to establish a lower limit for the minimum significant difference for statistical interpretation of data and alpha levels based on number of replicates. The whole sediment toxicity tests differ from the use of the elutriate test which is recommended to be used only for testing dredge material or when an activity, such as remediation, is expected to cause excessive re-suspension of sediment.The proposed ambient water toxicity testing criteria contained herein, takes into account new EPA WET guidance, the uncertainty associated with designating a stream as not meeting aquatic life uses based solely on sublethal toxicity testing, recommends using an alpha level equal to 0.05 and the Fisher's Exact statistical test on lethality. In addition, it recommends using an alpha level equal to 0.01 and the statistical t-test on sublethal data.Furthermore, it is recommended that sublethal toxicity test failures of less than and equal to 30% are, in and by themselves, inconclusive for regulatory purposes; and, the waterbody should continue to be judged as fully supporting aquatic life uses. Sublethal toxicity test failure rates greater than 30% but less than 50%, by themselves, provide inconclusive evidence that the waterbody is not supporting aquatic life uses. Nevertheless, test failures in the above range do indicate that the water may be partially supporting the designated use, but additional testing is warranted. Sublethal toxicity test failure rates greater than 50%, by themselves, provide evidence that toxicity exists and the water should be designated as not supporting aquatic life uses and additional testing and potential toxicant identification are warranted.This paper provides supporting evidence that the above criteria are reasonable, considering that toxicity test results are used to make regulatory “yes or no” decisions concerning ambient 303(d) listing. The supporting evidence includes a summary of the controversy concerning ambient toxicity test variability and Type I errors.