Sample collection
Copepods were collected on four seasonal cruises from the central Barents Sea to the Arctic Nansen basin north-east of Svalbard, Norway (Table S2). Cruises occurred during Autumn (5 – 27 August 2019), early winter (28 November – 17 December 2019), late winter (2 – 24 March 2021) and early spring (27 April – 20 May 2021), and visited three stations on the Barents Sea central shelf (76.00N, 31.22E), northern shelf (79.72N, 34.32E) and Nansen basin (81.83-82.16N, 28.15-29.84E, positions varied due to sea-ice drift, Fig. S1). Small-sized mesozooplankton (<1 mm) were collected in vertical 64 µm Bongo-net hauls (to full depth or max. 1000 m, ascent = 0.3 m s-1, descent = 0.5 m s-1, 60 cm mouth diameter). All large and/or gelatinous animals (1-10 cm) were removed, and the remaining suspension sieved (64 µm) to discard seawater. Ice-cold ethanol (96%, -20 °C) was then used to rinse retained mesozooplankton, before transfer into a sample bottle. The container was topped up with ice-cold ethanol and stored at -20°C.
Microsetella norvegica(Boeck, 1865), Microcalanus spp. (M. pygmaeus or M. pusillus , Sars G. O., 1900-1903) and Oithona similis (Claus, 1866) were morphologically identified under a stereomicroscope (Table S2). Up to 14 individuals per species and station were picked where available. Each specimen was thoroughly rinsed individually three times in Milli-Q water, transferred to tissue lysis buffer (E.Z.N.A Tissue DNA kit, Omega Bio-tek). Surface sterilization with bleaching was excluded, since the minute body size (< 1 mm) of the copepods analyzed herein raised concerns regarding how the treatment could penetrate and potentially alter the dietary signal. Also, existing literature, mostly based on arthropods, are in disagreement regarding the efficacy of bleaching, with one study indicating little effect on the overall dietary signal (Miller-ter Kuile et al., 2021). DNA extraction was performed per manufacturer’s protocol (“Tissue Spin Protocol”, E.Z.N.A® Tissue DNA kit, Omega Bio-Tek), with a lowered elution volume of 2 x 50 µL elution buffer. One negative without material was included with every round of extraction.
We checked all copepod and negative DNA extracts by PCR amplification of a 18S V7 fragment (~240 bp) using the universal eukaryotic primers 960F (5’-GGCTYAATTTGACTCAACRCG-3’) and modified 1200R (5’-GGGCATCACAGACCTG-3’) (Cleary & Durbin, 2016; Gast et al., 2004, Table S1). The amplifications were performed in 10 µL reaction volumes (4.8 µL MQ water, 1.0 µL DreamTaq buffer (10X), 0.2 µL DreamTaq Polymerase (5 U µL-1), 1.0 µL dNTP (2.5 mM each), 0.5 µL forward-primer (960F, 10 µM), 0.5 µL reverse-primer (1200R, 10 µM) and 2.0 µL template). PCR negatives had 2 µL MQ instead of template. Thermal cycling consisted of an initial denaturation (2 min, 95°C), and 35 cycles of denaturation (30 s, 94°C), annealing (30 s, 54°C), and elongation (30 s, 72°C), with a final elongation step (15 min, 72°C). The success of the amplifications was inspected by 1% agarose gel electrophoresis, confirming single band PCR products in all samples and no amplification in all extraction negatives.