8281 - Laserfiche WebLink

Home Browse Search

230 COMMENTARY project area. These investigations were not designed to evaluate biological effects from exposure to contaminants. The NIWQP mission does not include identifying or quan- tifying biological effects (Feltz and Engberg, 1994), and by themselves, cannot be used to evaluate biological effects associated with selenium or other trace elements. However, comparing concentrations of contaminants measured in NIWQP studies with concentrations associated with ad- verse effects derived in laboratory and field studies designed to measure biological effects is a legitimate way of determin- ing whether adverse e'ffects might be expected. One critical flaw in Canton and Van Derveer (1997) is that they inter- preted the two NIWQP studies as exposure-response stud- ies instead of exposure surveys. Moreover, Canton and Van Derveer (1997) cited Lemly et al. (1993) to indicate that biological effects were not ob- served in the Gunnison River studies, yet Lemly et al. (1993) did not discuss or cite specific results from Gunnison River reports. To the contrary, Lemly et al. (1993) contains a fig- ure that identifies the Gunnison River as a site where sub- surface drainage from federal irrigation projects may be causing toxicity to fish and wildlife, based on a comparison to biological effects studies, and a table identifies the Gun- nison River as a study area where toxicity is also predicted on the basis of concentrations of selenium found in fish and bird tissues. Furthermore, a recent review of NIWQP stud- ies for land susceptible to selenium contamination predicted that a selenium problem was likely in the Gunnison River Basin-Grand Valley Project area (Seiler, 1998). This predic- tion was validated by the fact that the Gunnison-Grand Valley study area was one of only four NIWQP study areas where Kesterson-type deformities of bird embryos were found (Seiler, 1998). 4.3. Effects in Western Streams Canton and Van Derveer (1997) state that "Despite these elevated [selenium] concentrations [in streams in south- eastern Colorado], no biological impacts, such as reduced fish diversity or abundance, has been observed." This asser- tion is the foundation of their agrument for sediment-based criteria, yet none of the empirical data essential for discern- ing biological effects were given to substantiate the state- ment, i.e., dietary studies; growth, survival, and reproductive viability studies; teratogenic deformity assessment. Canton and Van Derveer (1997) base their argument on the presence of fish species and conclude that water-based selenium cri- teria were weak, based on a small dataset from a single watershed in southeastern Colorado. Apparently no re- search or literature search was done to determine whether sensitive species were present historically in southeastern Colorado streams. Faunal surveys, in the absence of historical information, are not sufficient to detect contaminant impacts, or lack thereof, in an open river system. Adverse effects on a portion of a demographically open fish population in a section of the river with contaminant impacts would be very difficult to detect and must be confirmed with detailed biological studies because of immigration of individuals from the por- tion of the population in nonaffected river reaches or tribu- tary streams. The review by Skorupa (1998) addresses this concern succinctly and states "It is common for instream studies to report the counterintuitive combination of abnor- mally elevated levels of selenium in fish tissue associated with what is viewed as a normally abundant and diverse fish fauna." The end result is that Canton and Van Derveer (1997) erroneously conclude that the toxic thresholds for selenium derived from laboratory and field studies in closed basins, i.e., demographically closed populations, do not ap- ply to stream studies. Effects of selenium on species or populations of fish in the lake and reservoir studies cited by Canton and Van Derveer (1997) were substantiated with appropriate biological tests, whereas the stream/river in- vestigation by Van Derveer and Canton (1997) was not. Canton and Van Derveer (1997) and Van Derveer and Canton (1997) repeatedly use this incorrect reasoning that the presence of fish in Colorado streams with elevated selenium concentrations indicates the absence of impacts. Their claim of no biological effects in Colorado streams cannot be confirmed from information given or referenced in their papers. On the contrary, monitoring of fish popula- tions in rivers is an insensitive measure of contaminant effects unless substantial effort is made to assess the health of the fish community. This assertion was addressed by the USEPA in their guidelines for deriving water quality cri- teria. Stephan et al. (1984) stated that "The insensitivity of most monitoring programs [for number of taxa or indi- viduals] greatly limits their usefulness for studying the validity of [water quality] criteria because unacceptable changes can occur and not be detected. Therefore, although limited field studies can sometimes demonstrate that criteria are under protective, only high quality field studies can reliably demonstrate that criteria are not under protective [i.e., overprotective]." No high quality field studies directly examining the relationship between selenium exposure and a primary response variable, such as egg viability or body condition, were cited by Canton and Van Derveer (1997) .to support their conclusion of no adverse effects in Colorado streams. Based on the authors' measure of response, the selenium poisoning of western mosquitofish (Gambusia af- finis) at Kesterson Reservoir (Saiki and Ogle, 1995) would have been a no-response case. Fausch et al. (1990) reviewed the most common ap- proaches to assessment of environmental degradation using fish communities in lotic systems, which included (1) indi- cator taxa or guilds; (2) indices of species richness. diversity. and evenness; (3) multivariate methods; and (4) the index of biotic integrity (IBI). Each of the approaches had specific