Methodology:

Sampling:

Sampling took place at the P3 observation site (Northeast Scotia Sea, 52.80 degrees S, 40.14 degrees W, bottom depth 3748 m), as part of the Scotia Open Ocean Observatory programme (SCOOBIES, https://www.bas.ac.uk/project/scoobies/). A sediment trap was deployed at 378 m during research cruise JR17002, RSS James Clark Ross and recovered by research cruise DY098, RRS Discovery. The sediment trap collected samples between February and December 2018.

The opening of the trap (McLane PARFLUX, 0.5 m2 capture area; McLane labs, Falmouth MA, USA) was fitted with a baffle to prevent larger animals entering the trap. 500 ml bottles were used for automated sample collection. Bottles were filled to the brim with a fomalin (filtered seawater containing 2 percent v/v formalin, mixed with sodium tetraborate (BORAX; 0.025 percent w/v), and 0. 5 percent w/v sodium chloride). The bottles were placed in a carousel pre-programmed to rotate and collect material every calendar month, although the collection periods were shorter during the productive period (but collated to 1 month for the present analysis). Once recovered, samples were stored at 4 degrees C until analysis.

Sample processing and copepod identification:

Specimens were identified, measured and counted under an Olympus SZX16 fitted with a Canon EOS 60D DSLR camera. We assumed that copepods with an intact prosome and urosome were active swimmers. Due to their small size, nauplii, Oithona, Microcalanus and Onacea spp. were excluded from any analysis. It is not always possible to discriminate between swimmers and carcasses (i.e., perfectly intact zooplankton that had died shortly before sinking into the sediment trap). However, Ivory et al., (2014) found the number of copepod swimmers to be an order of magnitude higher than sinking carcasses, indicating that carcasses are a relatively minor component. Copepod occurrence was converted to abundance, as follows (Eq. 1).

Eq.1 Copepod abundance [individuals.m-2.day-1]=Occurence/(Trap Capture Area*Days bottle is open)

Where possible, copepods were staged and identified to species and genus, however family classifications were sometimes required, particularly where younger developmental stages made precise identification difficult. Copepod length, for each individual, was calculated from images taken with a graduated Petri dish, and subsequently analysed using ImageJ (Version 1.54g).

Data collection:

The sediment trap (Mclane PARFLUX, 0.5 m2 capture area; McLane labs, Falmouth MA, USA)

Olympus SZX16 fitted with a canon EOS 60D DSLR camera

ImageJ (Version 1.54g).

Data quality:

For the copepod count data

NA under 'Taxon/genus' or 'species' = no taxon could be assigned

NA under 'Size (urosome included)' = no size recorded