Experiment of light effects on phytoplankton morphometric traits and nutrient utilization

Near-surface seawater containing the natural mix of phytoplankton, protozoa and bacteria was pumped from Kiel Fjord, Baltic Sea, on 31 October 2016 and distributed through a hose system into 9 mesocosms (volume 1500 L, 1.5m in diameter and 1m high) which were evenly placed in three walk-in climate chambers. Light phase: the mesocosms were exposed to three different constant irradiance levels at a 9:15 hrs light:dark cycle. Light was supplied by computer controlled (GHL, Prometheus) LED light units. Each light treatment was triplicated and evenly distributed among the climate chambers. The three different light intensities were assumed to represent low light stress (low: 31 µmol quanta m-2 s-1), moderately limiting to saturating conditions (medium: 72 µmol quanta m-2 s-1), and saturating to slightly inhibiting conditions (high: 139 µmol quanta m-2 s-1). The experimental temperature was 11°C, representing in situ conditions. Fluorescence measurements were done every day in the lab at the same time using a subsample from the mesocosms. The different light treatments were applied until fluorescence measurements indicated stationary phase of growth (day 7 for the medium and high light treatments, day 11 for the low light treatment). An automatic propeller was used to mix the mesocosms gently, representing natural water waving to avoid sedimentation and achieve homogeneity of the plankton community. Nutrient phase: after reaching stationary phase, the light pre-conditioned communities were split into two nutrient treatments. One of them received a saturating nutrient pulse and the other was not modified as it served as control. For the new treatments 120 L-1 volume floating plastic bags were used (0.5m in diameter and 0.6m high), set in pairs (one nutrient and one control treatment) inside the bigger mesocosms and under each light unit (maximum distance from light: 1m). The plastic bags containing the new treatments were filled gradually using a 10 L bucket to achieve the same initial plankton assemblage. The content of the bags was stirred manually twice per day. Light intensity for the second experimental phase was set at the moderately limiting to saturating conditions (medium light: 72 µmol quanta m-2 s-1) for all treatments. The split into nutrient treatments and controls took place at day 11 for low light and at day 8 for medium light and high light, respectively. The nutrient pulse targeted final concentrations of 2 µmol L-1 phosphate, 32 µmol L-1 nitrate, and 32 µmol L-1 silicate in each mesocosm. Growth was followed until the nutrient augmented treatments had reached the stationary phase. Microscopic phytoplankton counts were performed according to Utermöhl's (1958) sedimentation method. Geometric proxies according to Hillebrand et al. (1999) were used to calculate cell volumes (V; in µm3). Samples for dissolved inorganic nutrients were taken together with the phytoplankton samples. The samples were filtered through pre-washed (10% HCl) cellulose acetate filters and frozen immediately until analysis. Samples for POC, PON and POP measurements were filtered onto prewashed (in 5-10% HCl) and precombusted (6 h, 550 °C) Whatman GF/F filters. POC and PON was determined by gas chromatography (Sharp, 1974) in an elemental analyser (Thermo Flash 2001, Thermo Fisher Scientific Inc., Schwerte, Germany). POP was converted to orthophosphate and determined colorimetrically (Hansen and Koroleff, 2007).