Sediment-adsorbed total mercury flux through Yolo Bypass, the primary floodway and wetland in the Sacramento Valley, California

Abstract

The fate and transport of mercury are of critical concern in lowland floodplains and wetlands worldwide, especially those with a history of upstream mining that increases the mobility of both dissolved and sediment-bound Hg in watersheds. A mass budget of total mercury (THg) quantifies sources and storage for particular areas — knowledge that is required for understanding of management options in lowland floodplains. In order to assess contaminant risk in the largest flood-control bypass, prime wetland, and restoration target in the Sacramento River basin, we estimated empirical relationships between THg, suspended sediment concentration (SSC), and streamflow (Q) for each of the major inputs and outputs using data from various publicly available sources. These relationships were improved by incorporating statistical representations of the dynamics of seasonal and intra-flood exhaustion (hysteresis) of sediment and mercury. Using continuous records of Q to estimate SSC suspended sediment flux and SSC to estimate THg flux, we computed the net transfer of sediment-adsorbed mercury through the Yolo Bypass over a decade, 1993–2003. Flood control weirs spilling Sacramento River floodwaters into the bypass deliver ~ 75% of the water and ~ 50% of the river's suspended sediment load, while one Coast Range tributary of the bypass, Cache Creek, contributes twice the THg load of the mainstem Sacramento. Although estimated sediment flux entering Yolo Bypass is balanced by efflux to the Sacramento/San Francisco Bay-Delta, there is much evidence of deposition and remobilization of sediment in Yolo Bypass during flooding. These factors point to the importance of the bypass as sedimentary reservoir and as an evolving substrate for biogeochemical processing of heavy metals. The estimates of mercury flux suggest net deposition of ~ 500 kg in the 24,000 ha floodway over a decade, dominated by two large floods, representing a storage reservoir for this important contaminant.

Introduction

Much attention has been paid to mercury sources and sinks in the Sacramento River basin of California, due to the legacy of hydraulic gold mining in the Sierra Nevada and Hg mining in the Coast Ranges, which has created persistent contamination of lowland sediments and food webs (Bouse et al., 2010, Conaway et al., 2007, David et al., 2009, Davis et al., 2008, Domagalski, 2001, Eagles-Smith et al., 2009, Gehrke et al., 2011a, Gehrke et al., 2009, Greenfield et al., 2005, Greenfield and Jahn, 2010, Marvin-DiPasquale et al., 2009, Marvin-DiPasquale et al., 2003, Roth et al., 2001, Rytuba, 2000). However, there is still great uncertainty in the mass balance of total mercury (THg) delivered to and stored in lowland floodplain environments from different source areas. Such estimates are particularly important given that lowland floodplains integrate basinwide delivery of chemical constituents and engender conditions favorable for mercury methylation (Compeau and Bartha, 1985, Gilmour et al., 1992, Marvin-DiPasquale et al., 2003). In this paper, we develop empirical relationships from hydrology, sediment, and mercury data collected by various regional and state programs to evaluate the delivery of THg to a large, engineered floodplain in the Lower Sacramento River basin and its storage over a decade. The work is relevant to contamination of food webs within a biologically productive lowland floodplain ecosystem (Sommer et al., 2001b), as well as to past and future data collection efforts in this region. These will enable better understanding of prevailing and potential contamination risks to Yolo Bypass and Sacramento/San Francisco Bay-Delta food webs. We intend the results of this analysis to augment current understanding of sediment and mercury storage in the Bypass, to clarify the relative importance of the contributing watercourses, and to inform the targeting of efforts to stabilize mercury sources.

Although the Sacramento River watershed contributes an estimated 80% of the THg moving through the Sacramento-San Joaquin Bay-Delta (Foe, 2003), a region now classified as impaired under the Clean Water Act for excessive levels of mercury and other contaminants (Larry Walker & Associates, 2002), Hg concentrations in water and suspended sediments have been assessed inconsistently in space and time. Data collection campaigns have generally been conducted within particular regions of the basin over short periods. Consequently, generalizations from such data to broader spatial and time scales are ill-defined.

Most previous studies report on a single set of event and/or annual sample data from a particular monitoring program. Estimates from Larry Walker & Associates (LWA) (2002) took advantage of several monitoring programs spanning the years 1992–2000. Mercury concentration estimates from that study were derived from univariate regressions with discharge data, where available. Where no discharge data were available, they used daily values or monthly means from the nearest discharge gauging stations to estimate regressions. There are several potential problems with such an approach. First, THg concentrations can vary significantly for a given flow rate (see LWA (2002), Fig. 2-2b). Second, since the majority of mercury tends to be adsorbed to fine sediment particles (e.g. (Domagalski, 2001, Maurice-Bourgoin et al., 2002)), a direct relationship between flow and mercury would tend to overestimate concentrations in water of the falling limb of the hydrograph if sediment hysteresis occurs (illustrated in Fig. 2-2a from LWA (2002)). Third, monthly mean discharge data dampen flood peaks that are generally responsible for the majority of the sediment/mercury flux (Singer and Aalto, 2009, Singer and Dunne, 2001).

The LWA study (and other shorter term studies) did not estimate Hg concentrations in flow overtopping Fremont Weir. This is an important omission because it is the largest source of flow, sediment, and THg derived Sacramento Basin (Domagalski, 2001, Singer and Aalto, 2009). Overall, there is a lack of long-term analyses on mercury mass balance within the Yolo Bypass, where higher rates of methylation are expected because of the organic carbon-rich, stagnant wetland environments that occur there (Rudd, 1995, Zilloux et al., 1993).

Here we will quantify the relative contributions of each major THg source to the bypass and flux out to the Bay-Delta. The study is intended to build on existing analyses that have assessed mercury in and around Yolo Bypass by providing a longer view based on mass fluxes of water, sediment and THg.