Sampling and analytical procedures:
Methods are described in detail in Gosnell et al. (2017).

Samples were collected to represent three separate seasons during 2014: Spring (May 20th-22nd), Summer (August 2nd-5th), and Fall (September 9th-11th). Plankton and water samples were collected from Western Long Island Sound (WLIS) and Eastern Long Island Sound (ELIS) stations in the spring and summer, while the fall collection included stations in Central Long Island Sound (CLIS), as well as along the shelf break (SB) and mid-shelf (MS) region.

At each station, triplicate substations were sampled approximately one mile apart at each site during the spring and summer seasons, while duplicate substations were sampled during the fall. Water samples were collected using a trace-metal clean GoFlo bottle attached to a Kevlar line, or with a trace-metal clean rosette system. Water was kept cold and dark until processing. Water samples were filtered for particulate methylmercury (MeHg) and total mercury (HgT), total suspended solids (TSS) and chlorophyll a (Chl a). Filtrate was saved for dissolved mercury and methylmercury, nutrients, and dissolved organic carbon (DOC) measurements. A CTD was deployed concurrent with water collections, and recorded the vertical profiles of fluorescence, oxygen, temperature, and salinity. Separate phytoplankton (seston) fractions were obtained by sequentially filtering the seawater through polycarbonate filters of sizes 0.2 µm, 5 µm and 20 µm. Larger particulate material was excluded by initially passing water through a 200 µm mesh. Zooplankton were collected by attaching an opening/closing 200 µm net (Seatec) to the Kevlar line. Deployments were from 2 meters above the bottom up to 1 meter from the surface. Depths ranged from 11 meter in Long Island Sound, to 100 meter at the Mid-Shelf and Shelf Break stations. Zooplankton were separated into size fractions of 0.2–0.5 millimeters, 0.5–1 mm, 1–2 mm, and >2 mm via mesh screens.

Each sample was digested in 4.5 M HNO3 acid solution at 60 degrees C for at least 12 hours prior to analysis. For MeHg analysis, a subsample was neutralized with KOH, diluted to a final volume of 30 mililiters, buffered with acetate buffer (pH of 4.9), and ethylated with sodium tetraethylborate (Hammerschmidt et al. 2013). Water samples were analyzed without acid digestion (Munson et al., 2014). Samples were then analyzed using a Tekran 2700 methylmercury analyzer with cold vapor atomic fluorescence (CVAFS) detection, and an external calibration curve (r2 >0.998). For total Hg analysis (HgT), the remaining digest was diluted, bromine monochloride (BrCl) was added for a minimum 24 hours preceding analysis to decompose organic matter and convert all Hg into the ionic form. Hydroxylamine hydrochloride was added to degrade excess BrCl oxidant prior to analysis using a manual system with tin chloride (SnCl2) reduction, purging with N2 onto a trap containing gold-coated beads for quantification using a Tekran 2500 CVAFS detector, and an external calibration curve (r2 >0.999). For water samples, the acidified water was treated with BrCl and then processed as for biota digests. Chlorophyll and phaeopigment were analyzed using a Trilogy fluorometer, and DOC was analyzed using a Shimazdu TOC/TN instrument.

Additional information:
QA/QC information is given in Table 1. (see Supplemental Files section below)

Results of these data are shown in Figs 2-5 and Table 1 in Gosnell et al. (2017). Additional calculated information (%MeHg, bioconcentration factors) is included in the same paper in additional tables. Stable isotope data for C, N and S for zooplankton is also reported.

Problem report:
Not all size fractions of samples were collected at all locations because of biota variability. Missing information is indicated in the dataset as 'nd'. Additionally, Spring and Summer samples were collected only at two sites with additional Fall sampling at three other sites.