Mapping dataset of nutrients in the Bohai and Yellow seas in June and November 2011, January 2016, September 2017, April and July 2018

Six field
surveys were conducted spanning both cold seasons (21 November–6 December 2011,
25–31 January 2016, and 28 March–16 April 2018) and warm seasons (13–28 June
2011, 29 August–16 September 2017, and 24 July–8 August 2018). Some data
obtained in the Bohai Sea during our 2011 and 2016 cruises have been releaseded
earlier at figshare.com via https://doi.org/10.6084/m9.figshare.10292384. At
26–79 grid stations, depth profiles of temperature and salinity (practical
salinity scale of 1978) were determined with a calibrated CTD recorder
(SBE911+, Sea-Bird Co., Bellevue, WA, USA) onboard R/V Dongfanghong 2. Water samples for nutrient
analyses were collected at three or four depths using a rosette sampler fitted
with 8 L or 12 L Niskin bottles, which were mounted with the CTD unit. Sea
surface water samples were collected from a depth of 2–3 m, and bottom water
samples were collected from 2–3 m above the sea bed. Water samples were
filtered on board using clean 0.45 μm cellulose acetate membranes. Each filtered
sample was divided into three subsamples. One of the subsamples was frozen at
−20 °C and then analyzed for nitrate, nitrite, and soluble reactive phosphorus
(SRP) in the laboratory. Trichloromethane (0.1% v/v) was added to the other two
subsamples, which were stored at 4 °C and then analyzed for ammonium and DSi in
the laboratory.

Nitrate
and nitrite (N+N) were measured by reducing nitrate to nitrite with a Cd
column, and then determining nitrite using the standard pink azo dye
spectrophotometric method. SRP, DSi and ammonium were measured based on
standard phosphor-molybdenum blue (for SRP), silicon molybdenum blue (for DSi),
and indophenol blue (for ammonium) spectrophotometric procedures. The samples
of nitrate, nitrite, SRP, and DSi obtained during our 2011 and 2016 cruises
were determined using a SEAL AA3 Auto-Analyzer (Bran+Luebbe GmbH, Hamburg,
Germany); and the samples of nitrate, nitrite, SRP, and DSi obtained during our
2017 and 2018 cruises were determined using a reorganized Flowsys III analyzer
(Systea S.P.A. Co., Anagni, Italy). Adjustments to the original Flowsys method
included (1) enlarging injection volumes to increase the concentration of
chromogenic reagents, (2) increasing the injection time from 60 s to 90 s, (3)
increasing the flushing time from 60 s to 90 s, and (4) replacing the cadmium
ring with a manually filled cadmium column to ensure complete and stable
nitrate reduction in the sample determination process. Ammonium was measured
using a Tri-223 continuous Auto-Analyzer. The detection limits were 0.05 μmol kg–1 for nitrate, 0.02 μmol
kg–1 for nitrite, 0.25
μmol kg–1 for
ammonium, and 0.02 μmol kg–1 for
SRP, and 0.08 μmol kg–1 for DSi.
For
data quality assurance, Reference Materials for Nutrients in Seawater from the
General Environmental Technos Co., Ltd. (Lot No. CB-1507, Kanso Co., LTD.,
Osaka, Japan) were used during analyses, and we achieved a precision level of ±
0.1 μmol kg−1 for
all nutrients other than ammonium.

Considering
non-linear variations of molar ratios with water mass mixing and the analytical
error, N* was used in this study to characterize the nutrient structure and
tracing N sources. Among the various definitions of N*, we adopted the most
common one: N* = DIN – 16 × SRP, that is, the excess DIN relative to SRP for
the growth of average phytoplankton. According to the analytical precision in
the determinations of DIN and SRP, the estimated uncertainty of N* was ± 0.5
μmol kg–1. During our
November–December 2011 cruise, the ammonium measurements failed and we used
oxidized nitrogen (the sum of nitrate and nitrite concentrations) to describe
DIN; oxidized nitrogen usually accounts for over 90% of DIN in cold seasons.

To
do the N-source apportionment work in the Bohai and Yellow seas, a budget of N*
was conducted in the two shelf seas. Finally, marine N in water samples can be
calculated as:

DINSW = (1 −
γ) × 16 × SRPfield-measured + N*KW ×
Salinity ÷ 34.7,

where
γ (%) = 59.89 × depth−0.32 %,
indicating the percentage of sediment oxygen demand in the total system aerobic
respiration. N*KW is the
Kuroshio wintertime N* at 0–200 m; salinity is the sample salinity; and 34.7 is
the Kuroshio salinity at 0–200 m. In 2011−2019, the average Kuroshio wintertime
N* at 0–200 m is −0.6 ± 0.3 μmol kg–1.