Study area. The VERMIX cruise (12–31 July 2016) in the northeastern North Sea sampled along five transects (Tr.1–5) and Fv/Fm measurements were taken on Tr. 2–5 and were used for the phytoplankton analyses. Measurements of turbulence, hydrography, nutrients and chlorophyll a distributions and sampling conditions were described in a previous study[12]. Measurements of nitrate and phosphate[12] suggest that denitrification might be occurring in the shallow area ( < ~ 60 m) south of the shelf edge towards the central North Sea. More recently, the shallow and the shelf edge area were shown to be weakly connected whereas a larger connectivity was found between the shelf edge and water above the Norwegian trench[28]. Therefore, we refer to visited stations in three groups: shallow (< 60 m), shelf edge (60–120 m), and deep (> 120 m).
F v /F m corrected for photoinhibition (Fv/Fm*). At each station, water was collected from 4 selected depths for determination of chlorophyll a fluorescence kinetics, transferred directly to a 300 ml dark glass flask and stored for ~ 30 minutes until measurement of Fv/Fm. Subsequently, 100 ml of sample were transferred to a 200 ml transparent glass flask and incubated for ~ 4 hrs at 50 µmol photons m− 2 s− 1 in a temperature-controlled room, with a temperature adjusted to match the average temperature of the collected water samples. This light treatment has been shown to remove any signal in the Fv/Fm caused by photoinhibition[18]. After the incubation, Fv/Fm was again determined. Differences in post-incubation Fv/Fm were assumed to reflect differences in nutrient status in the phytoplankton communities with low Fv/Fm* values indicating nutrient stress.
Fv/Fm was measured using a FastOcean FRRf 3 sensor (Chelsea Technologies Group, UK) with a dark chamber installed, and data was acquired in the FASTpro 8 software. A single turnover protocol was used with a saturation phase consisting of 100 flashlets with a 2 µs pitch and 16 sequence repetitions. For each measurement, the sensitivity (PMT eht) was automatically optimised by the software and the LED light intensity (ELED) was manually optimised. The saturating light was composed of 450, 530, and 624 nm wavelengths in a constant ratio of 1:0.5:0.8. The detection limit of the system was Fv/Fm = 0.15. All values below the detection limit were presented as 0.15 here. For depths < 30 m, the highest Fv/Fm value (dark/light incubated) was presented to reflect the maximum PSII electron transport potential. For depths > 30 m, the Fv/Fm as a result of dark incubation only was presented as no photoinhibition was expected at these depths.
Chlorophyll. Total chlorophyll a and size fractions of chlorophyll a were determined at selected stations at 5 m, the depth of the deep chlorophyll a maximum (DCM) and at 30 m. Seawater samples tapped from Niskin bottles and filtered through Whatman GFF (0.7), 3, and 10 µm pore size Millipore filters. For each sampling depth and filter size, triplicate filtrations of 200 − 500 ml seawater for the GFF filters and 400 − 1000 ml seawater for the 3 and 10 µm filters were performed. The samples were immediately frozen and later extracted for a minimum of 6 h in 5 ml ethanol (96%) in the dark at 6 oC. Chlorophyll a concentrations were measured using a Triology Laboratory Fluorometer (Turner Designs, CA, USA), which had been calibrated using a chlorophyll a standard from the DHI Group (Hørsholm, Denmark).
Copepod egg production rate (EPR). Centropages typicus and Temora longicornis females were collected at 23 stations with a 180 µm mesh plankton net, (1 m, ring diameter) fitted with a 5 L non-filtering cod-end. The sampling depth was about 2 m under the DCM and the net was towed vertically with a towing speed of approximately 15 m per minute.
The contents of the cod-end were gently transferred into a bucket with surface (5 m) water from the respective stations. Immediately after sampling, 20 active undamaged females were selected under a stereomicroscope and introduced into 250 ml bottles (1–3 females per bottle) filled with 64 µm screened ambient water. Four to 11 replicates were incubated per station. Screening was done in order to remove all eggs and nauplii from the initial incubation water. Females were incubated in darkness for 24 h at temperatures appropriate for the ambient temperature at the station where they were collected. The incubation was terminated by filtering the content of the bottles onto a 20 µm sieve and then washing the sieve contents into a small Petri-dish. Females were then removed and measured for length and the eggs counted.
Zooplankton sampling and processing. Zooplankton samples were collected at 14 stations (latitudes ranging between 56.65–57.83 °N and longitudes between 6.28–8.25 °E) (See Supplementary Table S1 online) using a Multinet (Hydro-Bios, Kiel; mesh size 200 µm). Zooplankton were stored in 4% formalin in the freezer. Samples were size fractionated using a sieve and zooplankton split into two size groups, > 640 µm and < 640 µm. Samples were then digitally scanned on an Epson Perfection V800 photo (2400 dpi, 16 bit gray), using the software Vuescan. In total, 1340 images were produced, representing 134 samples (5 images for each sub-sample). Images were processed and statistics with the zooimage package[31] and after the creation of a classifier, individuals were automatically identified, measured and counted (see Supplementary methods online).