Analysis of diatoms, pollen, and the carbon-isotopic composition of a sediment core from a brackish marsh in the northern part of the San Francisco Estuary has provided a paleosalinity record that covers the past 3000 yr. Changes in marsh composition and diatom frequencies are assumed to represent variations in freshwater inflow to the estuary. Three periods of relatively high salinity (low freshwater inflow) are indicated, 3000 to 2500 cal yr B.P., 1700 to 730 cal yr B.P., and ca. A.D. 1930 to the present. The most recent period of high salinity is primarily due to upstream storage and water diversion within the Sacramento–San Joaquin watershed, although drought may also have been a factor. The two earlier high-salinity periods are likely the result of reduced precipitation. Low salinity (high freshwater flow) is indicated for the period 750 cal yr B.P. to A.D. 1930.

No abstract

Wetland rapid assessment methods (RAMs) can provide a cost effective, scientifically defensible estimate of wetland and riparian condition for use in ambient and project monitoring in resource management and regulatory programs. Those who have chosen to develop a RAM to assess wetland and riparian condition are faced with a range of issues and important choices that they must make throughout the development process. This paper is intended as a practical guide to RAM development. Six basic stages in the RAM development process are discussed: (1) organize RAM development by identifying the intended applications, assessment endpoints, and geographic scope of the RAM and forming appropriate teams to advise and review the development process and its products; (2) build a scientific foundation for method development by conducting a literature review, choosing a wetland classification system, building conceptual models, and identifying the major assumptions underlying the model; (3) assemble the method as a system of attributes and metrics that describe a full range of conditions; (4) verify the ability of the method to distinguish between wetlands along a continuum of conditions; (5) calibrate and validate the method against sets of quantitative data representing more intensive measures of wetland condition; and (6) implement the method through outreach and training of the intended users. Important considerations within each of these stages lead to choices in accuracy, precision, robustness, ease of use, and cost. These are identified and the tradeoffs of the various options discussed. Experience with the ongoing development and implementation of the California Rapid Assessment Method (CRAM) is used to illustrate these stages and associated choices in RAM development.

No abstract