Expected and observed conditions during the SAGE iron addition experiment in Subantarctic waters

The SAGE iron addition experiment was conducted from R.V. Tangaroa east of South Island, New Zealand, in late March–early April 2004. A desktop survey of climatological data was completed before the experiment, providing information to inform site selection and experiment design. The desktop survey is presented here in updated and enhanced form in order to explain the site selection and describe the conditions expected at the site during the experiment in comparison with those actually encountered.The experiment site was in Subantarctic waters between the Subtropical and Subantarctic Fronts. These waters are characterised by high surface macronutrient concentration, low iron concentration and low chlorophyll. The preferred site based on the desktop survey was in the vicinity of 173.5°E, 47.5°S, in Southern Bounty Trough. The actual release location was chosen immediately before the release and was 112 km to the northwest of this at 172°32′E, 46°44′S. The surface water here has typically come from the southwest (over the northern Campbell Plateau) or the southeast (through Pukaki Gap) and the mean current is directed towards ENE at ∼0.1 m s−1. The release location is well removed from regions of high eddy kinetic energy to the east (where the Subantarctic Front reaches its northern limit) and the west (where fine-scale instabilities develop on the Southland Front, which flows along the continental shelf). Typical conditions at the release site at the end of March are: surface temperature 12 °C; mixed layer depth 40 m; surface chlorophyll concentration ∼0.3 mg m−3; surface photosynthetically active radiation (PAR) 23 E m−2 d−1; surface nutrient concentrations 8–10 mmol m−3 (nitrate), 0.5–0.8 mmol m−3 (phosphate), 1–2 mmol m−3 (silicate) and 0.1–0.5 nM (iron); 99th percentile wind speed 19–21 m s−1. At this time of year, surface PAR is well below its summer maximum, the mixed layer is beginning its seasonal deepening and the silicate concentration is at its seasonal minimum. These factors may have limited the phytoplankton response to iron addition and were compounded in March–April 2004 by strong winds early in the experiment (substantially exceeding the 99th percentile in speed), lower than the average SST, larger than the average mixed layer depth, silicate concentration at the bottom end of the expected range and initially low PAR.