These methods, instrumentation and/or protocols apply to all data in this dataset:| Methods and protocols used in the collection of this data package |
|---|
| Description: |
Fluridone applications
Fluridone was applied by DBW staff at the two treated sites using standardized methods and in accordance with all regulatory permits. Treatment was organized into treatment series, each consisting of weekly applications during 11- to 16-week periods. During each year, treatment series were completed between March and November, the period in which the control program is legally permitted to apply fluridone. Weekly treatments were conducted to maintain fluridone concentration at target levels for sustained periods. Three treatment series were completed at Little Hastings: Summer 2017 (June 7 – Sept. 19; 14 applications), Spring 2018 (March 19 – July 2; 15 applications), and Fall 2018 (September 17 – November 26; 11 applications) (Table 1, Figure S1). Decker Island received two treatment series: Summer 2017 (June 6 – September 18; 15 applications) and Spring 2018 (March 26 – June 18; 11 applications). There was no treatment at Decker Island in Fall 2018 because herbicide application would have conflicted with nearby wetland construction activities.
During a treatment series, fluridone pellets were distributed throughout each site weekly using boat-mounted air blowers at a nominal rate that ranged 10 – 20 ppb (Figure S1). These application rates were based on estimated water volume of a site (area × mean depth; Decker Island: 7.2 ha x 2.1 m; Little Hastings: 57.1 ha x 1.5 m). In general, the strategy was to use a relatively quick-releasing formulation (Sonar® Q) for a few weeks at the beginning of a treatment series. Sonar® Q was followed with SonarOne®, which has a longer release time than Sonar® Q. A treatment series ended with Sonar® PR, which has the slowest release time. Pellets of all three formulations were composed of clay embedded with 5% fluridone by weight (SePRO Corporation, Carmel, IN). Water concentrations of fluridone were monitored weekly or biweekly at all sites during active treatment, in part, to inform subsequent application rates. The goal was to maintain ambient concentrations of 2-5 ppb in the water column during a treatment series.
|
| | Description: |
Fluridone concentrations in the water
Fluridone concentrations were measured immediately prior to the first applications in 2017 at Little Hastings, French Island, and Decker Island to quantify potential background levels in the environment. These sites were subsequently monitored during each treatment series over the next 18 months to determine whether concentrations were maintained within the target range of 2-5 ppb. In summer 2017, sites were sampled weekly. In spring 2018, sampling frequency was reduced to every two weeks due to limited staff resources. Because Fisherman’s Cut was added to the study in April 2018, it was only sampled for fluridone water concentrations during the last two sampling dates of the spring 2018 treatment series. In fall 2018, French Island and Little Hastings were sampled weekly. Decker Island was not treated in fall 2018, so Decker Island and Fisherman’s Cut were not sampled during this period.
Within each study site, water samples were collected from three stations at a depth of 1 m. Two stations were located at water quality monitoring stations, one in a dense SAV bed and the other in an adjacent open water area, while the third station was located in an SAV bed distant from the other stations to better capture spatial variation in fluridone concentrations. The 60-mL samples were stored in opaque polyethylene bottles. Within 24 hours of collection, samples were delivered to a University of California-Davis (UC-Davis) laboratory for fluridone analysis. This laboratory used an enzyme-linked immunosorbent assay (ELISA, also called FasTEST) to analyze all samples until May 2018 (Pandey et al. 2019). From June 2018 onward, they used high-pressure liquid chromatographic (HPLC) (Wickham et al. 2019). ELISA and HPLC results compare well across a wide range of fluridone concentrations (Netherland et al. 2002).
|
| | Description: |
Fluridone concentrations in the sediment
Monthly sediment samples were collected to capture changes in sediment fluridone concentrations during treatment and non-treatment periods and to characterize sediment composition. Bed sediment samples were collected from October 2017 to June 2018, such that the first samples were collected within two weeks of the last application of the 2017 treatments series (when 2017 sediment concentrations would likely be at a maximum), most of the spring 2018 treatment series, and the intervening months with no treatment. Therefore, samples were collected on nine dates from all sites, except Fisherman’s Cut, which was only sampled in June 2018. Sediment samples were collected by boat using a ponar dredge (15 × 15 cm opening, Forestry Suppliers Inc., Mississippi, U.S.). Vegetation, invertebrates, and debris were removed from samples. No fluridone pellets were observed in samples. Half of each ~2 kg sample was used to measure the fluridone concentration and half was used for sediment physio-chemical composition analysis. The ponar dredge was thoroughly rinsed between samples. Samples were transported to the laboratory in a cooler and stored in the dark at 4°C until analyzed. All samples were analyzed within one week.
Sediment fluridone analysis was conducted by the same laboratory at UC-Davis that analyzed the water samples. Fluridone was extracted from wet sediment using the QuEChERS method (Quick, Easy, Cheap, Effective, Rugged and Safe) (Wickham et al. 2019), and concentrations were determined using high-performance liquid chromatography coupled with an ultraviolet detector (HPLC-UV). Because these methods were still under development, sediment fluridone concentration was not analyzed during the first few months of the study. Additional sediment physio-chemical characteristics were analyzed by the Bryte Soils Laboratory of the California Department of Water Resources (CDWR) in West Sacramento, CA using standard American Society for Testing and Materials methods for soil texture classification (ASTM D2487) and mechanical particle size analysis (ASTMD 422-63, 2007).
|
| | Description: |
Effect of tides on fluridone concentrations
Data were collected at Little Hastings to determine the impact of tides on fluridone concentrations. These data were collected during the sixth week of the Fall 2018 treatment series. ISCO 6700 Portable Samplers (autosamplers) were positioned in three locations to capture spatial variation in fluridone concentrations. They were programmed to collect water every two hours beginning October 22, 2018 at 14:00 PST and ending October 24, 2018 at 06:00 PST, which spanned three tidal cycles. Autosamplers rinsed Teflon-lined tubing three times prior to collecting each sample into a 350-mL ISCO glass bottle in the ice filled holding chamber. There were 63 field samples (3 stations × 21 samples) plus 4 quality control samples. Quality control samples consisted of one field blank to test for fluridone contamination of the equipment (purified water; randomly assigned to one autosampler) and three reference samples to test for potential fluridone degradation (purified water spiked with 0.1500 ppb fluridone; one per autosampler). Half the samples were retrieved the morning of October 23, and the remaining samples were retrieved the morning of October 24. Samples were immediately transported on ice to the United States Geological Survey (USGS) Pesticides Fate Research Laboratory in Sacramento, CA for fluridone extraction by liquid chromatography tandem mass spectrometry. For details about the analysis methodology, see Orlando et al. (2020). Tidal stage data were downloaded from the USGS website for the nearby monitoring station (USGS 11455315; 38°14'34.84", -121°41'03.43"; 530 m from the study site).
|
| | Description: |
SAV biomass and community composition
Biomass and species composition of SAV was monitored at Little Hastings, French Island, and Decker Island during the two weeks prior to the first fluridone applications (May 25 to June 7, 2017) to determine baseline conditions of the sites. Then, sites were sampled every two months until December 2018, after the last fluridone applications were completed (10 sampling events total). Sampling at Fisherman’s Cut was initiated in April 2018 (5 sampling events).
A point intercept sampling method was used. ArcGIS was used to randomly select a set of stations at each site for SAV sampling, specifying a minimum of 30 m between any two points (except Fisherman’s Cut because of its small size). During 2017, 40 stations per site were sampled. During 2018, sampling was reduced to 20 stations per site (a random subset of the original 40 stations), based on a power analysis of the 2017 biomass data (Figure 1). A sample was generally collected within a ~20-m radius of its corresponding station. This approach prevented the removal of vegetation for one sample from affecting biomass of samples collected in subsequent months.
Sample collection was completed using the methods of Johnson and Newman (2011). Briefly, the steps were as follows: (1) lower a long-handled, double-headed thatch rake vertically to the bottom (rake head width of 35 cm with fourteen 5.5-cm-long metal teeth on either side), (2) turn the rake three times on its vertical axis while in contact with the bottom, (3) pull the rake straight up to the surface while continuing to rotate it. In high density SAV areas, the maximum capacity of the rake was sometimes exceeded, and therefore biomass may have been underestimated in some instances. Percent of total sample volume that each SAV species comprised was recorded to the nearest 10%. Samples were transported in garbage bags at ambient temperatures during the one-hour drive to the laboratory where they were stored for up to a week in a refrigerator prior to processing.
Samples were processed to estimate biomass following the methods described in Bickel and Perrett (2015). In brief, samples were rinsed to remove sediment and macroinvertebrates. Once water ran clear through the sample, it was placed into an 18.9-L salad spinner and rotated 20 times (1 revolution per sec) to remove excess water. Wet biomass was measured with a bench scale (Ohaus ES BV Bench Series, accuracy of 1.0 g). This method produces values comparable in consistency to those obtained from dry biomass (Bickel and Perrett 2015).
|
| | Description: | Phytoplankton
In May 2017, we initiated monthly phytoplankton sampling at three study sites. The sites included Little Hastings and Decker Island, which were both treated with fluridone, and French Island, which served as an untreated reference site for Little Hastings. In May 2018, we started sampling Fisherman’s Cut as a reference site for Decker Island. Within each study site, two samples were collected from high-density SAV areas, and two were collected from low-density SAV/open water areas. Half of these phytoplankton samples were collected next to continuous water quality sonde stations, as were the discrete water quality and zooplankton samples. Two full sets of samples were collected prior to initiation of the fluridone treatments (May, June 2017). The goal of this sampling design was to compare phytoplankton abundance and taxonomic composition among study sites (i.e., treated vs. untreated), between habitat types within sites (e.g., high-density SAV vs. low-density SAV/open water), and before vs. after treatment within sites. Monthly sampling continued until the end of the study in December 2018. Samples consisted of ~60 mL of surface water, which were fixed on site with Lugol’s solution and stored in amber glass bottles. All monthly samples were collected at or shortly after high slack tide. Samples were kept at room temperature and away from direct sunlight until they were analyzed.
We contracted with BSA Environmental Services, Inc. (https://www.bsaenv.com/) to analyze the samples. Phytoplankton identification and enumeration were performed using the Utermöhl microscopic method (Utermöhl 1958) and modified Standard Methods (American Public Health Association 2012). An aliquot was placed into a counting chamber and allowed to settle for a minimum of 12 hours. The aliquot volume, normally 10–20 milliliters (mL), was adjusted according to the algal population density and turbidity of the sample. Aliquots were enumerated at a magnification of 800x using a Leica DMIL inverted microscope. For each settled aliquot, phytoplankton in randomly chosen transects were counted. Taxa were enumerated as they appeared along the transects. A minimum of 400 total algal units were counted, and a minimum of 100 algal units of the dominant taxon. For taxa that were in filaments or colonies, the number of cells per filament or colony was recorded. Phytoplankton were identified to the lowest taxonomic level possible, which was most commonly genus. Organism counts for each sample were converted to organisms per milliliter. BSA estimated cell biovolume for up to ten individuals per taxon per sample. Biovolume estimates were derived by taking linear measurements of cell dimensions and calculating volumes based on the basic shapes the cells most resembled (e.g., sphere, cylinder, cone).
|
|
|
|
