8281 - Laserfiche WebLink

Home Browse Search

228 I c,. 2. SUPPORT FOR 211g1L WATER-BASED CRITERION COMMENTARY The current standard of 5 Ilg/L was established based almost solely on information from Belews Lake, North Carolina. USEPA (1987) states "The freshwater Criterion Continuous Concentration (CCC) should be between 10 Ilg/L and the concentrations in the unaffected portion of Belews Lake, which is near or below 5 Ilg/L. Therefore, the CCC will be set at 5 Ilg/L." Since that time, investigations have found that there were adverse effects in the "unaffec- ted" arm. the portion of the lake south of Highway 158. Holland (1979) reported that selenium concentrations were biomagnified substantially up the food chain in this arm of Belews Lake. He found about 2 Ilg/L in water, 23-25 Ilg/g in plankton, 26-31 J.lg/g in benthic invertebrates, and 18-47 Ilg/ g in three fish species. Sorensen et al. (1984) re- ported adverse effects in fish from this arm of the lake, which had 4 Ilg/L in water in 1976-1977 and 3 Ilg/L in 1980- 1981, i.e., elevated selenium residues in muscle and liver, higher condition factor, lower hematocrit, and adverse his- topathological changes in liver, kidney, and ovary com- pared to reference fish. The findings of Holland (1979) and Sorensen et al. (1984), from research at the same site used by the USEP A to establish the 1987 criterion, support the need for a national water quality criterion below 5 Ilg/L. Engberg et al. (1998) summarized federal and state per- spectives on regulation and remediation of irrigation-in- duced selenium problems and concluded that the current USEP A criterion for selenium of 5 Ilg/L was underprotec- tive. Recently, Skorupa (1998) reviewed the findings of 12 "real-world" examples of selenium poisoning and concluded that a national water-based criterion of < 5 J.lg/L was broadly justified. He reviewed information from selenium- contaminated sites at Belews Lake, North Carolina, Kester- son Reservoir, California, Hyco Reservoir, North Carolina, Martin Reservoir, Texas, Richmond (Chevron marsh), Cali- fornia, Tulare Basin, California, Salton Sea, California, Kendrick Reclamation Project, Wyoming, Red Rock Ranch (agroforestry), California, Ouray National Wildlife Refuge (NWR), Utah, and Sweitzer Lake, Colorado, which included cooling reservoirs at coal-fired power plants, wetlands for treating petroleum wastes, and argicultural irrigation sources. Moreover, several comprehensive reports from state and university technical review committees as well as research scientists have also recommended 52 J.lg/L for the protec- tion of aquatic birds and mammals (SWRCBC, 1987; UCC, 1988; DuBowy, 1989; Skorupa and Ohlendorf, 1991; CEPA, 1992; Peterson and Nebeker 1992), thus indicating wide support for this criterion concentration. In-depth reviews by Maier and Knight (1994) and Lemly (1993a, 1996) rec- ommended 2 Ilg/L as a probable safe water concentration based on a convergence of concentrations of concern, derived in several laboratory and field studies, for the protection of fish and wildlife from selenium toxicity. The states of Arizona (1992) and New Mexico (1995) ha ve estab- lished a water quality standard for selenium for the protec- tion of aquatic life at 2Ilg/L. Thus, there seems to be substantial support for a national water-based criterion of 2 Ilg/L. All of these recommendations are based on the bioaccumulation of selenium from water into the food chain and the subsequent effects of dietary exposure to higher trophic levels. 2.1. International Support From an international perspective. there is additional support for a U.S. water quality criterion of 2 Ilg, L The water quality standard for selenium in Canada is I ~lg,L (CCREM, 1987). This document has recently been revised and the Canadian water quality standard for selenium has remained at Illg/L (N. Nagpal, Inland Waters Directorate, personal communication, 1998). A study of lakes in Sweden that were treated with selenite to mitigate high concentra- tions of mercury in fish concluded that waterborne concen- trations of selenium needed to be kept below 2 Ilg/L to avoid undesirable bioaccumulation of selenium in fish and unintentional side effects such as the complete die-off of perch observed in several lakes (Paulsson and Lundbergh, 1991,1994). On behalf of the Directorate General for Envir- onmental Protection, The Netherlands. Emans et al. (1993) evaluated several extrapolation methods using multiple spe- cies (MS) toxicity data to predict the no observed adverse effect concentration (NOEC) of various environmental con- taminants on ecosystems and then validated the extrapola- tion method with results from field studies. For selenium. they calculated extrapolation values for the MS NOEC of 2.2 and 2.5 Ilg/L. 3. FOOD CHAIN TOXICITY THRESHOLD The critical link in the recommendation of 2 Ilg/L as the potentially safe waterborne selenium concentration for the protection of fish and wildlife resources is bioaccumulation and biomagnification into the food chain. Maier and Knight (1994) and Lemly (1993a, 1996) recommended 3 Ilg/g dry weight as the dietary threshold for selenium toxicity to fish and wildlife. This concentration in the food chain is often achieved at low waterborne selenium concen- trations over a long period of time or by rapid loading of an aquatic ecosystem over a short period of time. For example, Lemly (1997) reported developmental de- formities in young fish from Belews Lake 10 years after selenium inputs to the lake were stopped in 1986. He found in 1996 that waterborne selenium concentrations were < Illg/L, and even though sediment concentrations of sel- enium were 65-75% lower than in 1986, they still were sufficiently elevated (1-4 Ilg/g) to contaminate benthic