Food chain model based on field data to predict westslope cutthroat trout (Oncorhynchus clarkii lewisi) ovary selenium concentrations from water selenium concentrations in the Elk Valley, British Columbia

Abstract

Previous studies conducted in the Elk River watershed showed that selenium concentrations are higher in aquatic biota in lentic compared to lotic habitats of the system having similar water selenium concentrations. Studies have also shown that water selenium concentrations have increased over time (~10% per year) and recent annual average concentrations have ranged up to 0.044 mg/L in areas downstream from mine discharges. For the present study, trophic transfer of selenium was characterized in lotic versus lentic habitats using concentrations measured in field-collected samples and assuming a three-step food chain of water to the base of the food web (biofilm), to benthic invertebrates, and then to westslope cutthroat trout (WCT) ovaries. Food chain models were developed for each habitat type (lotic and lentic) by combining linear regression equations for the three transfer relationships, allowing for prediction of fish ovary concentrations from water concentrations. Greater accumulation of selenium in lentic areas was mostly attributable to greater uptake at the base of the food chain compared to lotic areas. Enrichment/trophic transfer factors for selenium at all levels of the lotic and lentic food chains decreased and then became near constant as exposure concentrations increased. The lotic model predicted little increase in WCT ovary selenium concentrations over an eightfold increase in water concentrations (~0.005-0.040 mg/L), accounting for the lack of observed increase in within-area fish tissue concentrations over time despite increasing trends in water concentrations.

Food chain transfer and exposure effects of selenium in salmonid fish communities in two watersheds in the Canadian Rocky Mountains

“…Higher trophic transfer from biofilm to invertebrates in McLeod River streams suggests concentration-dependent uptake at this food chain step where greater Se accumulation occurs in invertebrates exposed to lower dietary Se concentrations (DeForest et al 2007). A similar relationship was observed in an earlier lotic trophic transfer study in the Elk River watershed (Orr et al 2012) and supports a trophic transfer model where Se uptake rates in stream-dwelling invertebrates change in relation to available dietary Se concentrations.…”

“…These fish had high mean muscle Se concentrations of 19.72 mg·kg -1 dm in brook trout and 15.07 mg·kg -1 dm in rainbow trout. These fish may have spent time in the lentic-type environment of the pond, and their tissue concentrations may therefore reflect exposure to higher trophic transfer and accumulation of Se typical in lentic food chains compared with the lotic environment where they were captured (Orr et al 2012;Simmons and Wallschläger 2005).…”

“…Creek and Swift Creek, which had water Se concentrations of 0.543 and 0.508 mg·L -1 , respectively (compared with 0.0373 mg·L -1 in the Fording River, the maximum concentration identified by Orr et al 2012). These were single point measurements, while the BC MOE guideline is prescribed based on the 30-day mean of water Se; however, elevated Se concentrations in water and biota in the mine-affected study streams may pose toxicity risks to resident aquatic life and have the potential to translate into effects in aquatic communities further downstream in the watersheds (Windward Envrionmental et al 2014).…”

“…Fish tissue samples were collected between 1 and 14 days later, except at three sites (Deerlick and Wampus creeks and the Gregg River) where fish tissue samples were collected approximately 6 weeks later. Overall, the sampling period was similar in duration to other ecosystem-scale Se models where samples were collected over a period of up to 2 months (Orr et al 2012;Presser and Luoma 2010).…”

“…Therefore, Se monitoring programs have relied on the evaluation of species-specific tissue concentrations against toxicity thresholds (Casey 2005;Minnow Environmental Inc. et al 2011). Food chain trophic transfer models have been proposed to improve Se monitoring and reduce lethal fish sampling in vulnerable populations by using Se sampling in water, biofilm, or invertebrates to model fish tissue concentrations (Orr et al 2006(Orr et al , 2012Presser and Luoma 2010). However, trophic transfer relationships between dietary concentrations and most salmonid species of the Canadian Rockies have not been reliably defined.…”