13 Western GOM Sentinel Zooplankton Indicators
Description: The data presented here is from the NERACOOS-MBON Wilkinson Basin Time Series (WBTS) station. Runge et al. 2025 (Runge et al. 2025) analysis initially included observations from 2005 up through March of 2023, this update follows the same methods and now includes observations up to November 2025.
Found in: State of the Ecosystem - Mid-Atlantic (2024+), State of the Ecosystem - Indicator Catalog (2024+)
Indicator category: Published methods
Contributor(s): Jeffrey A. Runge, Cameron R.S. Thompson, Shawn Shellito, Emma C. Dullaert, Isabel A. Honda, Douglas Vandemark, Dylan Pugh, Riley Young-Morse, Jackie Motyka, Rebecca J. Jones, Lee Karp Boss, Rubao Ji
Data steward: Cameron Thompson cameron@neracoos.org
Point of contact: Cameron Thompson cameron@neracoos.org
Public availability statement: Source data are publicly available.
13.1 Methods
13.1.1 Data sources
Calanus finmarchicus abundance and stage-composition, and total mesozooplankton biomass data are from the NERACOOS–MBON Wilkinson Basin Time Series (WBTS) station. The most recent source data are publicly available through the point of contact and github, archived data is hosted on the NERACOOS ERDDAP server and on the OBIS portal. The datasets are updated as additional observations are processed, following sample processing, QA/QC, and data management by the MBON team.
Observations 2004–2017: https://data.neracoos.org/erddap/tabledap/WBTS_CFIN_2004_2017.html
Observations up to 2020: https://data.neracoos.org/erddap/tabledap/WBTS_CFIN_start_2020.html
WBTS Calanus dataset (Darwin Core) in the OBIS portal: https://obis.org/dataset/5ef55cd8-05a1-4569-8e17-ceb224e40f59
WBTS Calanus abundance and mesozooplankton biomass dry weight: https://github.com/sea-scope/Gulf_of_Maine_Plankton_Sentinel_Indicators/blob/master/data/sentinel/prepared/WBTS_prepared.csv
This submission extends the published WBTS seasonal indices of Runge et al. (2025)(Runge et al. 2025), which covered January 2005 through March 2023, to include observations through November 2025 using identical data-processing and analysis methods.
13.1.2 Data analysis
All analyses were conducted in R (R Core Team), using the mgcv package for Generalized Additive Models (GAMs; Wood(Wood 2017)). For each time series variable, an annual climatology GAM was specified with a cyclic cubic spline for day of year and a spline for year with an autoregressive term to address autocorrelation, fit by Restricted Maximum Likelihood (REML). Using the same specification, seasonal-trend GAMs were fit to data filtered to each seasonal period (all years within a given season) to examine interannual change within a season. Where necessary, a square-root transformation was applied to achieve an approximately normal distribution for the analysis; in the depicted figures these values are displayed in their untransformed units.
Model robustness was assessed with DHARMa residual diagnostics (Hartig(Hartig 2022)): a Kolmogorov–Smirnov test on scaled residuals, plus dispersion and outlier checks. GAMs that passed the significance and diagnostic tests were visualized with ggplot2. Seasonal-trend GAMs were depicted only where the day-of-year smoother was significant, and annual-climatology GAMs only where the year smoother was significant.
The accepted GAM of each variable is used to estimate the expected average value and 95% confidence interval, either to depict climatology over an annual cycle or to depict the trend in each season across years. These estimated expected average values are the indices and can be calculated for any combination of year and day of year. For depicting the annual climatology, the year is set to 2012 while days range from 1 to 365. For depicting interannual trends within a season, a single representative day is set within each season (day of year 37, 110, 197, and 306 for winter, spring, summer, and fall, respectively) while years vary across the record (2005–2025).
13.1.3 Data processing
Station. The Wilkinson Basin Time Series (WBTS) Station (42°51.7′N, 69°51.8′W; previously Station WB-7) is located 60 km from Portsmouth, New Hampshire, in the northwest corner of Wilkinson Basin at an average station depth of 257 m. Since the start of the time series in December 2004 it has been accessed by day trips aboard the University of New Hampshire research vessel R/V Gulf Challenger. Contingent on funding support, the station was sampled at approximately monthly intervals between January 2005–August 2008, April 2012–May 2013, October 2015–July 2017, and at less frequent intervals in other years. The station continues to be sampled at approximately monthly intervals since 2020 as part of the U.S. MBON program.
Field sampling. Sampling at the WBTS station generally follows guidelines established by the Atlantic Zone Monitoring Program (AZMP) operated in Canadian Maritime waters by Fisheries and Oceans Canada (Mitchell et al. 2002(Mitchell et al. 2002)). All samples were taken during daytime hours, typically mid-morning to mid-afternoon. To measure zooplankton abundance and biomass, net casts were made using a SEA-GEAR 0.75 m diameter single ring or twin-ring 200 µm mesh net. The nets were towed vertically at approximately 40 m/min with the ring starting 5–7 m from the bottom. Samples were preserved in 4% buffered formaldehyde.
Laboratory processing. In the laboratory, samples were split with a Folsom Splitter, with half of the sample designated for measurement of total zooplankton biomass and the other for taxonomic enumeration.
Biomass. The sample is filtered onto one or more 47 mm diameter 200 µm mesh nitex screens preweighed in a plastic petri dish. The split sample was poured through the filter or screen mounted in a filter holder assisted with gentle vacuum pumping, rinsed with 100 ml of tap water, dried in an oven at 65 °C for 24–48 h, and then collectively weighed on a Mettler Toledo PG403-S microbalance (1 mg precision).
Enumeration. The sample is drained of formaldehyde solution on a fine mesh screen and placed in a 4 L beaker containing a known quantity of filtered seawater (typically 2500–3000 ml). Subsamples were taken while randomly stirring with either a 25 ml Hensen-Stempel pipette or a large-mouth pipette, emptied into a graduated cylinder to measure subsample volume. Multiple subsamples were taken to ensure at least 50 Calanus copepodid stages were enumerated (typically about 3% of the total half-sample) under a Leica MZ 12.5 Zoom stereo microscope. All Calanus copepodid stages in the subsamples were enumerated.
Measurement calculations. Water-column biomass (g dry weight m⁻²) and abundance (number of individuals m⁻²) were calculated by dividing the total-sample measurement (accounting for the split and aliquot of subsamples) by the net volume sampled (net depth multiplied by area of the ring net, 0.4418 m²), and then multiplying that m-3 value by the station depth. Net volume filtered was also determined by a General Oceanics flowmeter installed in the mouth of the net, but the geometrically determined volumes were chosen as the standard because reliable flowmeter data were not available for every cast.
Derived indices. Three indices are computed from the processed samples:
Calanus Abundance Index: abundance (no. m⁻²) of late copepodid stages in the water column, calculated as the total abundance of stage C3 to adult stage C6 of C. finmarchicus. The index is restricted to late copepodid stages to make it comparable with NOAA EcoMon data, which are collected with a 333 µm mesh net that allows escape of younger stages C1–C2.
Copepodid Stage Index (CSI): the weighted mean copepodid stage composition of the C. finmarchicus population, calculated as the sum of each stage number multiplied by its abundance, divided by total abundance across all stages (C1–C6). A larger value indicates a population dominated by older stages; a smaller value indicates a younger-stage population.
Total Mesozooplankton Biomass: total dry weight (g m⁻²) of mesozooplankton in the water column captured by a 200 µm mesh, an integrated measure of the entire mesozooplankton community.
Seasonal definitions. To examine subannual change in relation to the life cycle of C. finmarchicus, the year is divided into four seasons based on the population phenology:
| Season | Day-of-year window | Calendar dates | Representative DOY |
|---|---|---|---|
| Winter | 1–73 | 1 Jan – 15 Mar | 37 |
| Spring | 74–147 | 16 Mar – 27 May | 110 |
| Summer | 148–247 | 28 May – 4 Sep | 197 |
| Fall | 248–365 | 5 Sep – 31 Dec | 306 |
catalog link https://noaa-edab.github.io/catalog/wbts_zoo.html