MERCURY TMDLS FOR ARIVACA AND PEÑA BLANCA LAKES, AZ

Waterbodies in many locations throughout the United States have concentrations of mercury in fish tissue in excess of guidelines for issuing fish consumption advisories. As a result, many of these waters have been included on 303(d) lists as not supporting uses and in need of a TMDL. In many cases, fish tissue concentrations are elevated even though there is no evidence of concentrations of mercury in water in excess of ambient water quality criteria. The TMDL analysis requires a determination of assimilative capacity, which, for this type of impairment, should be the loading rate of mercury that would result in acceptable mercury concentrations in fish tissue. Unfortunately, use of the TMDL procedure to address mercury in fish tissue presents a number of difficult technical problems. Most notably, fish body burdens of mercury can be only weakly and non-linearly related to watershed loads of mercury, because bioaccumulation is determined by the rates of mercury methylation in water and sediment and rates of trophic transfer. Mercury TMDLs also often involve multi-media problems, as atmospheric deposition is often a major pathway of mercury loading. Finally, as with many other bioaccumulative pollutants, there may be significant storage of mercury in waterbody sediments which buffers the response of fish to changes in external mercury load.Along with the technical difficulties, there are important practical challenges to completing mercury TMDLs driven by fish tissue concentrations. The complexity of the problem suggests the potential need for a correspondingly complex modeling and analytical approach. Yet, EPA and the courts have set forth ambitious schedules for completing TMDLs that limit both time and resources available for analysis. How can such TMDLs be estimated in a timely manner and at a reasonable cost?We describe the development of TMDLs to address mercury in fish in Arivaca and Peña Blanca Lakes in southern Arizona. Both lakes lack permitted point sources of mercury load and appear to meet ambient water quality criteria for mercury, yet have mercury concentrations in tissue of game fish that present a risk to human health. The TMDL analyses can thus be described as technically difficult. At the same time, the practical challenges of completing TMDLs under a short, court-imposed time line and with a limited budget were present. The successful completion of the two TMDLs in the face of these technical and practical challenges demonstrates that the problem is not an insoluble one. Key aspects include the development of risk hypotheses and a targeted modeling or linkage analysis of loading and assimilative capacity. The uncertainty inherent in the rapid approach to the TMDLs provides important insights into where effort should be focused in future attempts of this sort.