The sampling strategy for the water-quality assessment included quarterly sampling at the three lagoon sites and the perennial tributary T1, plus sampling two storm events at several tributary sites. The quarterly water-quality sampling occurred on November 24, 1998, January 27, 1999, May 19, 1999, and August 26, 1999. Storm sampling occurred on February 6–7, 1999, and April 10–11, 1999. A continuous water-level recorder was installed at T1 in February 1999, just prior to the storm sampling. A streamflow rating was developed by the USGS for this tributary and was used to develop a continuous streamflow record for water years 1999 and 2000. Streamflow in T1 is mainly a function of rainfall; the stream becomes almost dry during extended periods without rain. Four samples were collected at T1 during the storm hydrograph on February 6–7, 1999, and two samples were collected during the storm hydrograph on April 10–11, 1999. Seven tributaries (T1–T7) and the middle lagoon (L2) were sampled from one to four times during the February 6–7, 1999, storm runoff. During the April 10–11, 1999, storm runoff, two samples were collected at T1, and single samples were collected at T2, T3, T5, and the middle lagoon.

During November, January, and May, the lagoon sites (L1, L2, and L3) were sampled from a canoe. During August, a larger boat was used because of the one-time collection of bed-sediment samples. The quarterly water-quality samples from the lagoon were collected with a Van Dorn sampler. Samples collected for laboratory parameters were obtained about one foot below the water surface. Field parameters were also measured near the lagoon bottom at each site. The bed-sediment samples collected at the lagoon sites in August were collected with an Ekman Dredge. Quarterly samples at T1 were collected as mid-point grabs with a 3-L Teflon bottle. Storm samples were collected as midpoint grabs directly into 1-L amber glass bottles at all sites.

The water samples collected quarterly at the lagoon sites and T1 were analyzed for field parameters, major ions, nutrients, chlorophyll, and suspended sediment. Only samples collected at L1 were analyzed for dissolved organic carbon. The bed-sediment samples collected in August at the lagoon sites were analyzed for percentage of inorganic and organic carbon, iron, and phosphorus. Field parameters measured at each site during quarterly sampling included temperature, specific conductance, pH, dissolved oxygen, alkalinity, and Secchi disk transparency (lagoon sites only). Storm samples were analyzed only for field parameters (temperature, specific conductance, and pH), nutrients, and suspended sediment. The April 1999 storm sample collected at T3 was also analyzed for a suite of 47 indicators of wastewater as part of a USGS national reconnaissance on “emerging contaminants” (Kolpin and others, 2002). Except for suspended sediment and chlorophyll, all laboratory samples were stored on ice and shipped overnight to the USGS’s National Water Quality Laboratory (NWQL) in Denver, Colorado, within two days of collection. Suspended sediment samples were sent to the USGS’s California Water Science Center Sediment Laboratory in Salinas, California. Chlorophyll samples were stored on dry ice and shipped separately to the NWQL.

Samples collected for major-ion analyses (schedule 2701 at NWQL) were filtered through a 0.45-μm capsule filter and were analyzed by the inductively coupled plasma (ICP) method at the NWQL (Fishman and Friedman, 1989). Samples collected for analyses of dissolved nutrient species (schedule 2702 at NWQL) were filtered through a 0.45-μm capsule filter. Both dissolved and total nutrient species were analyzed by the colorimetry method at the NWQL (Fishman and Friedman, 1989). Chlorophyll samples (labcode 586 at NWQL) were filtered through a 0.7-μm glass fiber filter in the field, then the filter was folded and put in a petri dish, wrapped in aluminum foil and stored on dry ice until shipped to the NWQL. At the NWQL, chlorophyll samples were ruptured mechanically by centrifuge, and the pigments were separated from each other and degradation products by high-pressure liquid chromatography (HPLC). The chlorophyll-a concentrations were then determined by fluorescence spectroscopy (Britton and Greeson, 1987). The dissolved organic carbon samples (labcode 113 at NWQL) were filtered through a 0.45-μm silver filter in the field. At the NWQL, the dissolved organic carbon was determined by ultraviolet-promoted persulfate oxidation and infrared spectrometry (Brenton and Arnett, 1993). The bed-sediment samples were sieved through a 2-mm sieve in the field. Thus, the bed sediment samples analyzed for percent carbon, total phosphorus, and iron represent material less than 2 mm in diameter. The percentage of inorganic and organic carbon in bed material (labcodes 1458 and 1459, respectively at NWQL) were determined by wet-chemical oxidation and infrared spectrometry (Burkhardt and others, 1997). The total phosphorus in bed material (labcode 515 at NWQL) was determined by colorimetry (Fishman and Friedman, 1989). The iron in bed material (labcode 2566 at NWQL) was determined by ICP (Garbarino and Struzeski, 1998). The samples for wastewater indicators (custom labcode 8033 at NWQL) were analyzed by gas chromatography/mass spectrometry on unfiltered water (Barber and others, 2000).

The quality control sampling was limited in this study because of the limited budget available for additional analytical samples. During the storm sampling, one replicate and one field blank were collected for nutrients only. The field blank collected at T4 on February 7, 1999, at 1008 had no detections of nitrogen species, and low-level detections of dissolved phosphorus species (0.006 mg/L dissolved phosphorus as P; 0.019 mg/L dissolved orthophosphate as P). The reporting levels for these constituents were 0.006 and 0.018 mg/L as P, respectively. The corresponding values in the environmental sample collected at T4 were 0.070 and 0.052 mg/L as P, respectively. Because the storm samples were collected directly in 1-L glass bottles, there was no cleaning of a sample collection bottle. Thus, phosphorus contamination of the blank was probably introduced either during filtering or in the blank water itself.

A replicate was collected at T2 on February 7, 1999, at 1441. All dissolved nutrient measurements had relative percent differences of less than 10 percent. For whole-water constituents—total Kjeldahl nitrogen and total phosphorus—the relative percent differences were 50 and 17 percent, respectively. These replicates were collected as sequential replicates into separate 1-L glass bottles. This illustrates the problem with representing (and reproducing) suspended material with sequential mid-point grabs (Martin and others, 1992). The environmental sample had a suspended sediment concentration of 88 mg/L, with 36 percent of this being material coarser than clay or silt (that is, with diameter greater than 0.062 mm). The flow at this site was mostly runoff from a dairy operation, and the proper splitting of solids between replicates greatly affects suspended nutrient levels. Sequential grab samples are not a good way to reproduce suspended-nutrient concentration results.