Methodology:

Samples were collected during a series of Discovery Investigations cruises to South Georgia during the late 1920s and 1930s. Two ships were involved either Discovery II or William Scoresby.

N70V plankton nets (Kemp et al. 1929) were deployed and hauled vertically through the following depth horizons 1000-750m, 750-500m, 500-250m, 250-100m 100-50m and 50-0m, where water depth was>1000m. At stations where water depth was <1000m, profiles were adjusted accordingly.

Stations of interest were identified from Discovery report Station lists (Anonymous 1931, 1932, 1941, 1942 and 1943). Samples and sample fractions had been preserved in seawater formalin solution and were located in the stacks of the Natural History Museum. My interest was in samples taken within the top 250m during the summer period spanning December to early March. Searching through the samples indicated that, for a proportion of stations, samples were entirely absent or the depth series incomplete. Nonetheless, samples from a total of 34 stations taken by Discovery II and 18 stations by William Scoresby were identified and subsequently analysed.

Data collection:

N70 plankton net

'Folsom' plankton splitter

Data quality:

Overall, condition of the samples was moderate. Some were in excellent condition whereas others were clearly showing the ravages of time and exhibited signs of decay. Nonetheless, the species of interest were readily identifiable. Many samples had been preserved in their entirety, whereas others were represented by preserved fractions or residues. In these cases, the samples had clearly been subjected to a preliminary analysis. Whether the sample examined was a residue, a fraction or an entire sample (whole) is indicated in column E of the spreadsheet. In the column sample fraction analysed, the term L&M refers to copepods large and medium as signified on the sample labels.

Visits to NOC Southampton allowed inspection of archived Discovery logbooks and sampling sheets. These indicated that preliminary analysis was restricted to macroplankton (amphipods, pteropods, euphausiids etc). Where copepods had been counted, they were not identified to species level.

Samples were assessed for copepod abundance and then split using a 'Folsom' plankton splitter (Reid and Parker, 1984) until countable aliquots were achieved. Copepodite life stages CI-CVI (see Blaxter et al. 1998) of the four species only were enumerated under a binocular microscope.

If the sample analysed was a fraction or residue, it was assumed to contain a representative number of the species in question and counts were therefore multiplied accordingly to represent totals in the entire sample. Density (ind. m-2) was determined by using Pir^2 xh where the net radius (r) was 0.35m and h is the height (depth interval) of the water column through which the net was hauled. Data presented are ind. m -2 within the depth horizon sampled by the net. To determine a value for ind. m -2 for the entire water column, it will be necessary to sum each respective depth horizon.

Discovery sample analysts were very punctilious about documenting any removals of species/stages from samples and previous experience has indicated that if such occurred sample labels were annotated accordingly.

Subsequent to analysis, all samples were transferred to PTFE bottles in Steedmans solution (as the seals on the original 'Kilner jars' were breached and could not be replaced).