Research Rapidly Increasing Polybrominated Diphenyl Ether Concentrations in the Columbia River System from 1992 to 2000 SIERRA RAYNE Department of Chemistry, Box 3065, University of Victoria, Victoria, British Columbia, Canada, V8W 3V6 MICHAEL G. IKONOMOU* Contaminants Science Section, Institute of Ocean Sciences, Fisheries and Oceans Canada, 9860 West Saanich Road, Sidney, British Columbia, Canada, V8L 4B2 BONNIE ANTCLIFFE Habitat and Enhancement Branch, Fisheries and Oceans Canada, Suite 200, 400 Burrard Street, Vancouver, British Columbia, Canada V6C 3S4
Concentrations and congener patterns of 32 individual PBDE congeners from mono- through hexa-brominated were investigated in two fish species occupying similar habitatssbut having different diets and trophic levelss and surficial sediments from several locations on the major river system of western North America, the Columbia River, in southeastern British Columbia, Canada. Total PBDE concentrations have increased by up to 12-fold over the period from 1992 to 2000 in mountain whitefish from the Columbia River, with a doubling period of 1.6 years. The rate at which PBDE concentrations are increasing in whitefish is greater than has been previously reported worldwide. At the current rate of increase, ΣPBDE will surpass those of ΣPCB by 2003 to become the most prevalent organohalogen contaminant in this region. ΣPBDE in whitefish from the mainstem of the Columbia River range up to 72 ng/g wet weight, concentrations that are 20-50-fold higher than in a nearby pristine watershed affected only by atmospheric contaminant transport. Conversely, ΣPBDE in largescale suckers were approximately an order of magnitude lower than in whitefish, demonstrating the influence of biomagnification and feeding habits. Congener patterns in whitefish from the Columbia River directly correlated with the two major commercial penta-BDE mixtures in use and represent the first time free-swimming aquatic biota such as fish have been found to contain PBDE congener patterns so similar to commercial mixtures. PBDE concentrations in sediments were not linked to a variety of investigated point sources but were instead inversely correlated with the ratio of organic carbon:organic nitrogen in surficial sediments with a pattern suggesting the dominant influence of septic field inputs from the primarily rural population.
* Corresponding author phone: (250)363-6804; fax: (250)363-6807; e-mail:
[email protected]. 10.1021/es0340073 CCC: $25.00 Published on Web 05/17/2003
2003 American Chemical Society
Introduction The Columbia River is the fifth largest river system in North America ranked by flow (6650 m3/s), draining a total basin area of 673 000 km2 in the northwestern United States and Canada before discharging into the northeast Pacific Ocean at the Washington State/Oregon border, some 1900 km from its source in two mountain lakes of the Selkirk Mountain range in British Columbia, Canada. Human settlement along the river has been present for over 10 000 years, but the 20th century saw large scale engineering projects (mostly hydroelectric developments) rapidly change the river’s hydrology. Anthropogenic activities resulting from this development have had numerous effects on the physical, chemical, and biological characteristics of the region, especially declining salmonid fish populations over the past century (1). At present, more than 400 dams exist on the Columbia River system, making it the most hydroelectrically developed river system in the world (2). However, while much is known regarding the effects of human development on the physical and biological state of the Columbia River, relatively little is known regarding organic contaminants in this system, especially in the Canadian portion. Among the organic contaminants of current interest and concern, recent evidence has confirmed that the brominated flame retardants, polybrominated diphenyl ethers (PBDEs), are ubiquitous contaminants present in all environmental compartments at concentrations up to the parts-per-million range (3, 4). In industrialized and pristine regions of North America, PBDE concentrations are increasing rapidly (3, 5) at rates which may be more rapid than was ever observed for PCBs (6) and in contrast to declining concentrations of PCBs and PCDD/Fs in the present study area (7). Indeed, the 1999 worldwide production volume of PBDEs (sum of the commercial penta-, octa-, and deca-BDE mixtures) at 67 000 tonnes is near the PCB production maximum of 100 000 tonnes in 1970 (8). At this rate of PBDE production, which does not appear to be either stabilizing or declining (3), only 14 years would be required to surpass the estimated worldwide PCB production of ∼1 000 000 tonnes which took place over the five decades from the 1930s until the late 1970s (9). Unfortunately, there are few temporal studies by which to rigorously examine how PBDE concentrations have changed in different regions during the latter 20th and early 21st centuries. Furthermore, little is known of PBDE concentrations and congener patterns in the major North American rivers (i.e. Mississippi-Missouri-Red Rock, St. Lawrence, Mackenzie-Peace, Yukon, Columbia). In the Columbia River, the large number of hydroelectric facilities have, in effect, disconnected various units of the river from each other, thereby preventing unrestricted movement of fish throughout the ecosystem. This disconnection is thought to have greatly constrained the access of anadromous fish to upstream spawning sites, possibly helping to explain large declines in regional fisheries stocks since the early 1900s (10-12). In addition, dams largely prevent the downstream movement of suspended and bedload sediments, thereby encouraging scouring of the river bed downstream. These sediments are known to be important vectors for hydrophobic contaminants such as PBDEs, which would generally prefer to reside on organic substrates than be dissolved in the water column (13-16). This compartmentalization of fish populations and sediments makes the Columbia River system a particularly intriguing one in which to examine PBDE VOL. 37, NO. 13, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2847
FIGURE 1. Map showing topographic features of the study region discussed in the text and sampling sites for mountain whitefish and surficial sediments in the Columbia and Kootenay River systems in southeastern British Columbia, Canada. Sample numbers correspond to data provided for individual fish and sediments in the Supporting Information. concentrations and patterns. In addition, increasing concentrations of potentially toxic organic contaminants such as PBDEs may be an additional stressor (17-19), that when combined with other human influences such as hydroelectric development, may further hinder recovery of some fish populations in the Columbia River. In the study region of the Columbia River and a major tributary, the Kootenay River, in southeastern British Columbia, Canada (Figure 1), thalwegs are generally shallow (2-12 m) during normal discharge events and river velocities are moderate to high (75 to >150 cm/s), favoring a riverbed composed of medium to large cobbles and small boulders and/or parent bedrock material with a limited littoral area (20, 21). Smaller numbers of depositional environments are found where gravels, cobbles, sand, and silt materials collect (20, 22). The Kootenay River joins the Columbia River ∼10 km downstream from Hugh Keenleyside Dam and 46 km upstream from the Canada-United States border, after which the Columbia River continues its course 1100 km through the United States before discharging into the Pacific Ocean. The average flow of the Columbia River at the Canada-United States border is ∼2630 m3/s, of which 1150 m3/s is from the Arrow Lakes reservoir system controlled by the Hugh Keenleyside Dam, 790 m3/s from the Kootenay River system, and 690 m3/s from the Pend d’Oreille River which enters the Columbia River 3 km before the international border (22). In the study region, 25 species of fish have been recorded (22), of which mountain whitefish and largescale suckers are considered resident (21) and thus good indicators of local contaminant inputs and sources. Of these two species, mountain whitefish are the most abundant sportfish in the Columbia and lower Kootenay Rivers with a population that has remained relatively stable at 35 000-40 000 individuals since the early 1980s (20, 22). Whitefish tend to occupy high 2848
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 13, 2003
velocity habitats with cobble or boulder substrates and riffle and run areas adjacent to cobble and boulder deposition areas. Their diet is mainly bottom feeding of aquatic insect larvae (e.g. caddisflies), small mollusks, and occasionally fish (23). Most adult whitefish in this area occupy deepwater habitats in the day and relocate to shallow-water habitats for feeding at night (24). Juveniles typically reside in shallow nearshore habitats for rearing, as these areas provide the greatest quantity of food and cover (20, 21). Largescale suckers, in comparison, use a wider variety of habitats than whitefish with adults residing in both deepwater and nearshore habitats and juveniles occupying shallow stream margins. The diet of immature suckers is mostly planktonic, consisting mainly of water fleas and copepods, and shifts to bottom feeding of invertebrates for adult individuals (23). Largescale suckers generally spawn in May and June, while whitefish spawn during the period from November through February (21). The population of largescale suckers is estimated to range between 21 000-57 000 and is also thought to be relatively stable (22). Together with surficial sediment samples, the resident nature of these two fish species and their different feeding habits offer a means of investigating the concentrations and patterns of PBDEs in this section of the Columbia River system. To facilitate an understanding of the temporal and spatial trends of PBDEs in a major North American river, we determined the concentrations of 32 individual PBDE congeners from mono- through hexabrominated in mountain whitefish, largescale suckers, and surficial sediments from several locations on the Columbia and Kootenay River systems in southeastern British Columbia, Canada. Whitefish samples obtained over the period from 1992 to 2000 shed insights into how PBDE concentrations have changed in this aquatic system over the past decade. Largescale sucker
samples were taken to help determine how PBDE concentrations and congener patterns differ for fish occupying the same region but with different feeding habits. Surficial sediments were also collected to assist in identifying potential PBDE sources and to examine whether congener patterns and concentrations differed between sediments and fish. Together, these samples form one of the first comprehensive data sets on PBDE concentrations and congener patterns in a major North American river.
Experimental Section Sample Collection and Preparation. Sampling for mountain whitefish (Prosopium williamsoni) in the Columbia and Kootenay River systems took place during the first 2 weeks of July in 1992, 1994/1995, and 2000. Whitefish samples from the Slocan River reference site (WF39-41) were obtained by angling methods in an attempt to avoid tissue damage induced by electroshocking. Angling methods were unsuccessful at the Genelle (WF15-33) and Beaver Creek (WF1-14) sites on the Columbia River, and whitefish were collected using electroshocking. Whitefish samples collected in Kootenay Lake (WF34-38) in early July of 1998 near the outflow to the Kootenay River as well as largescale suckers (Catostomus macrocheilus; LS1-6) collected in early July of 2000 downstream of the city of Nelson on the Kootenay River were also obtained by electroshocking. Details on the methods of electroshocking and age determination are provided elsewhere (25, 26). Sampling sites are shown in Figure 1; details regarding each location are discussed throughout the manuscript. Surficial sediments were collected from 11 depositional locations (S1-11) on the Columbia and Kootenay River systems on August 18 and 19, 2001. Samples were collected from the top 2-3 cm of fine-grained silt, clay, and organic materials within 2 m of the shoreline. All materials for collecting sediments were stainless steel or amber glass and were washed with successive rinses of tap water, distilled water, 95% ethanol, and acetone in the field prior to collection. Sediments collected at each site were placed in solvent rinsed amber glass jars and stored on dry ice during sampling and transport and at -20 °C in the laboratory prior to processing and analysis (
