RESULTS
Hydrologic setting
California is characterized as having a Mediterranean climate with distinct wet and dry seasons. These seasons result in large fluctuations in freshwater inflow to the northern reach of San Francisco Bay. The magnitude of these fluctuations can vary greatly from year-to-year.

The study described here occurred over a period of extreme year-to-year differences in hydrodynamics (hydrograph). Drought conditions persisted from 1989 through 1992. In 1993 there was an end to a 7-year drought but in 1994 the drought conditions returned for one year. The years following 1994 were "wet" with several of the years being considered extreme.

The seasonal and annual hydrologic variability captured during this study exposes aquatic organisms living in the estuary to highly variable environmental conditions. These include fluctuations in salinity, contaminant inputs from both the rivers and the bay, both dissolved and sediment bound, suspended sediment loads, and carbon load to the bay. Understanding the seasonal and annual variability is critical to deciphering natural and anthropogenic effects.

Ag Concentrations
For the overall study period, the grand mean Ag concentrations in clam tissues were highest at the two mid estuary sites, CNWS and CS , and lowest at the two end-estuary sites. Annually, Ag concentrations were observed to decrease over the study period with the two mid estuary sites, CNWS and CS showing the most dramatic decrease (annual means). Variability showed was highest at the two most contaminated sites, CNWS and CS.

The monthly patterns of tissue Ag concentrations showed that variability in annual mean silver concentrations was related to inflow from the Sacramento and San Joaquin rivers. Silver was seen to accumulate in clam tissues during extended periods of low flow. Periods of increased discharge were associated with decreases in Ag concentrations in clam tissues (monthly means at Carquinez Strait ).

Reproductive Activity
With a few exceptions, the majority of clams at the estuary end-member sites were observed to be reproductive, 81% and 80% for MI and SPB, respectively. In contrast, the populations at the CNWS and CS were only reproductive 67% and 61% of the months. For these two mid-estuary sites, the proportion of clams reproductively active was seen to increase over throughout the study period. Comparison of reproductive activity with Ag concentrations showed that the increase was coincident with a decrease in metal concentration (reproductive activity and Ag concentration). When Ag concentrations were highest (annual mean >1ppb), clams were reproductively active only 20 to 60% of the year. In contrast, when Ag concentrations were lowest (annual mean <1ppb), clams were reproductively active 70 to 100% of the year. A significant correlation was observed when the reproduction data was combined with the metal data (Reproductive activity vs Ag).

Discussion
The most likely source of Ag contamination is anthropogenic input. The high concentrations observed at the two mid-estuary sites during the early part of this study suggest local sources. However, it is difficult to determine the exact source of contamination. In 1993, new restrictions were imposed to reduce the amount of Ag in industrial wastewater. For example, at the Contra Costa County sewage treatment plant, the concentration of Ag in effluent was seen to drop from 11ppb to <1ppb. The decrease in inputs coupled with the changing hydrologic environment is most likely responsible for the drop in Ag contamination.

A number of factors may affect reproductive activity in clams. A short list includes changes in salinity, organo-metal complexes, specifically organotin, and/or pesticides. The evidence indicates that these influences can be discounted. Discharge patterns which govern salinity gradients do not entirely match the observed trend of increasing reproductive activity. Contaminants, such as organotin complexes or pesticides which have sources in the Napa River and Central Valley respectively, may be discounted because their sources are external to the estuary and the contamination is clearly from internal sources. The most compelling evidence of the role of Ag is the correlation between a drop in reproductive activity and increasing Ag concentration.

Determining the effects of anthropogenic influences on ecological systems is difficult due to natural environmental variability. The case study sumarized above looked at hydrology, tissue concentrations of silver and reproductive variability of the clam Potamocorbula amurensis. Through the use of high intensity monthly sampling over a long time-period, patterns in environmental variability can be identified and cause-and-effect relationships evaluated.