Methodology:

Deployments of the Nansen net were carried out through luffing and hauling a rope attached to the net. A measurement of 50 m was marked onto the rope to determine the correct deployment length. The ship was stationary during deployment and local time, latitude and longitude of the sampling station was noted. Net was luffed and hauled by hand, with hauling commencing immediately after reaching the maximum depth. Total deployments were approximately 3 to 5 minutes in duration.

On retrieval of the net, the cod end was unclipped from the net and the sample gently decanted into a bucket. The cod-end was further rinsed with seawater taken from the same location to ensure the whole sample had been retrieved from the cod-end. This sample was then concentrated through passing through a 200 µm sieve. The sample concentrate was transferred to a Nalgene bottle of appropriate size so that it comprised no more than 10% of the total volume of the bottle. The bottle was then topped up with 4% buffered formaldehyde and a waterproof label inserted before sealing.

Sample analysis was carried out approximately 6 months after sample capture. Samples were transferred out of formalin to water and sorted under a light microscope. All individuals were identified to the lowest possible taxonomic level to generate an abundance value for each individual taxon. This value was converted to a concentration (individuals m-3) through dividing by the volume of water sampled. Volume of water sampled was determined through multiplying the opening area of the net (diameter 0.7 m, sampling area 0.385 m2) by the length of rope deployed (mostly 50 m). Post-analysis, samples were returned to the original formalin for long term curation at the British Antarctic Survey.

Data collection:

(i) Nansen 70 net (vertically hauled)

The specification for the construction of the original N70 net is detailed in Ward et al (2012) following the designed detailed Kemp et al. (1929). The N70 was a 70-cm diameter net used to sample mesozooplankton and constructed with coarse netting and two grades of silk mesh. The upper part of the net was 21 inches (53 cm) of inch (6 mm) knotted mesh followed by a canvas band and then two finer mesh sections both made from Messrs Staniers' Quadruple Extra Heavy Quality Double Twist Swiss Silk bolting cloth, the first 3 feet 200 (96.5 cm) of 40 threads per inch (TPI) and the bottom section 4 feet 500 (136 m) of 74 TPI.

(ii) Binocular light microscope

Data quality:

Deployment depth was estimated according to the length of rope luffed into the water and the angle of the rope relative to vertical. Although 50 m was the nominal maximum depth, maximum depth attained is likely to be less in most instances.

A flow meter was not deployed with the net so volume of water sampled was estimated as the product of the opening area (diameter of 0.7 m) of the net and the amount of rope deployed below the sea surface. This will incur an error resulting from any turbulence, upwelling or downwelling currents present during the time of sampling and any clogging and inefficient filtering through the net.

Sample analysis was carried out to the lowest possible taxonomic level practicable, which was not always to the species level but sometimes went further also to include developmental stages.