An axenic strain of C. vulgaris was supplied by the Culture Collection of Algae (SAG) at Göttingen University, Germany. The permanent cultures were kept on agar in glass tubes under day light at room temperature. Prior to starting the experiments, algae stock cultures were taken from the permanent culture into Nunc EASY 75 cm² cell culture flasks (from VWR, Hannover, Germany) under sterile conditions (Clean Air CA/REV four, Labexchange, Germany) to synchronize growth for 11 days. Every 4 days, the cultures were transferred and checked for cell counts to keep them in exponential growth. For this, 750 μL of “old” culture was transferred with sterile pipettes into a new flask of fresh medium, to achieve an initial concentration of about a factor 100 smaller than in the old culture. The cultures kept in the exponential growth phase were maintained on a shaker (IKA®MTS 2/4 digital, Staufen, Germany) with 150 rpm in a climate chamber (neoLab-Thermostatschrank ET 618-4/135, Germany) at 22 ± 1°C with a light–dark cycle of 14:10 h. Four days before the tests the required amount of algae to reach an initial concentration of 2.5 × 10⁴ cells/mL was taken from the synchronized cultures and added to 50 mL green algae medium (GAM; see Table A1). This pre-culture was established to render an algae concentration suitable for the inoculation of test cultures. Cultures were incubated on a shaker at continuous illumination with 6600 lux (Sylvania Luxline plus F 15w/865 Daylight deluxe, Australia).

The oxygen optical microsensors OXR50 (optical isolation) were purchased from PyroScience (Aachen, Germany). The principle of the measurement is based on the quenching of the red-flash dye luminescence caused by collision between oxygen molecules and the red-flash dye immobilized on the sensor tip. The red-flash dye is excitable with red light at a wavelength of 610–630 nm and shows an oxygen-dependent luminescence in the near infrared (NIR, 760–790 nm) (PyroScience Manual). Details and technical specifications provided by the manufacturer (Table A2) are given in the Supplementary Material. Two-point calibrations were performed with four sensors connected to the device. Pure nitrogen gas and ambient air with known humidity (55%) were set as 0 and 100% air saturation. These two-point calibration procedures were repeated prior to each experiment. In order to exclude ambient oxygen, sample vials are well-sealed with rubber lids. The oxygen optical microsensor produces a change of luminescence in response to the depletion of oxygen in the vial due to the respiration of test organisms (Kühl, 2005). When the organisms' respiratory activity increases, the oxygen concentration decreases. The principle is the same when photosynthesis is used to measure cell activity. The sensor signals are recorded automatically by a computer. The profiles of the sensor luminescence signals directly illustrate the oxygen assimilation rate.

Algae were counted with an automatic cell counter (CASY® Model TTC, Roche Diagnostics GmbH, Mannheim, Germany) which is widely used to record healthy cells in ecotoxicological tests. It obtains the exact number of algae cells which are suspended in a special solvent (CASY® ton) inside the capillary. When the healthy cells go through the measuring pore, an enlarged level of electrical resistance is detected at once. Within a selected diameter range (to distinguish the target cells from other larger or smaller particles), dead cells and dust which do not act as an electrical barrier are not recorded.

Photosynthesis and respiration of C. vulgaris at different concentrations were measured by microsensors. Calibrations between O₂ evolution/assimilation rate and C. vulgaris cell counts were done. Experiments were performed in a large water bath with a constant temperature at 22°C (±0.1°C) in an air-conditioned room with a constant temperature of 22°C. 30 mL glass flasks closed with air-tight rubber lids (silicone septum, 3.2 mm VWR, Germany) were used. This allow the microsensor syringe needles to penetrate (Figure 1). The required concentrations of C. vulgaris (Figures 2, 3) were diluted with green algae medium in 50 mL Erlenmeyer flasks and then 10 mL test samples were transferred from the Erlenmeyer flasks into 30 mL glass flasks (the exact space volume is 31 mL) for the microsensor measurement. The volume ratio of head space and suspension is 21:10.

To assess the relationship between algae cell concentrations and O₂ evolution and assimilation rates we used the counts recorded by the algae cell counter CASY®. Each concentration was replicated four times. A temperature control flask was used in parallel to each test. Measurements were run immediately after adjusting the position of the sensor tip 1 cm above the liquid surface. The respiration tests were run in the dark, and for the photosynthesis tests ambient light was provided by three white (OSRAML 36W/865, Germany) and two warm color fluorescent bulbs (OSRAML 36W/11, Germany) with 182 μmol m⁻² s⁻¹ continuous illumination. Luminescence readings in each sensor were taken automatically every 15 s over the course of 30 min. The calibration curve was obtained within 12 h. Three independent repeats of both tests were run in winter, spring and summer.

In order to test the accuracy and efficiency of the method we compared O₂ assimilation rates measured by oxygen optical microsensors with a traditional manometer. It was designed by Otto Heinrich Warburg (Warburg, 1956; Drobnik, 1960) and has been used widely in plant physiology and soil microbiology (Syrett, 1951; Kleinzeller, 1965; Schaefer et al., 2005). The Warburg manometer is an analytical instrument for measuring the pressure of gases and vapors from biochemical reactions It was originally used to study respiration and uptake of gaseous O₂ and the production of CO₂ by various cells and tissues (Warburg, 1956; Kohler, 1973).

In this study, measurements were performed with soils from a sandy research site at Bremen University (ReviTec®; see Table 1 for soil properties). Samples were taken from three different existing experimental plots (Kf, Wild, Mix) at the ReviTec® site which had been established by Prof. Hartmut Koehler and PhD student Thomas Buse. Field plots had been inoculated with the cultures in April 2011, and samples were taken in April 2012. Sample Control was pure washed river sand. Sample Kf was taken from plots inoculated with laboratory cultured Klebsormidium flaccidium as representative filamentous terrestrial algae. Sample Wild had been inoculated with a community of wild algae consisting of different single cell and filamentous algae species from the ReviTec® site, and sample Mix was from a plot inoculated with cultures of K. flaccidium and the wild population.

Substrate-induced respiration (SIR) rate, a method for calculating soil microbial biomass (Anderson and Domsch, 1978; Sparling et al., 1981; Kaiser et al., 1992; Ananyeva et al., 2011) was measured by both Warburg and optical microsensors. Soil from the four different samples (10 g each) was separately mixed with 0.5 mL 10% glucose and distilled water to reach 50% water holding capacity. Alkaline absorption (KOH) was used in Warburg bottles to absorb CO₂. The volume of CO₂ was recorded every 30 min for 6 h at 22°C.

Sample preparation for the oxygen optical microsensors measurement was similar to the procedure described for the aquatic algae cultures. Instead of algae medium, the prepared soil samples (10 g soil mixed with 0.5 mL 10% glucose and distilled water) were transferred into 30 mL glass flasks and closed with air tight rubber lids. The experiments were performed in a water bath with a constant temperature of 22°C and measurements were run immediately after adjusting the position of the sensor tip 1 cm above the soil. In the end, microbial biomass carbon of the samples was calculated based on the maximum initial respiratory response at 22°C according to Kaiser et al. (1992). The following equation was used: x = 30y, where x is total microbial carbon biomass and y is maximum initial rate of CO₂ respiration (mL CO₂ kg⁻¹ soil h⁻¹) or O₂ consumption (mL O₂ kg⁻¹ soil h⁻¹).

Statistical analyses were performed with SPSS 17.0. Regression analyses and Pearson correlations were conducted to analyze the data on oxygen evolution and assimilation rate. For comparison between the Warburg manometer and optical microsensor in differently treated field soils, a GLM was applied to analyse the main effects and interactions between treatment and method (hypothesizing no difference between the methods and no interactions between both factors). Then a one-way ANOVA was run, followed by a post-hoc Dunnett-test to compare each treatment to the respective control of two methods.

This experiment was undertaken to create calibration curves between C. vulgaris cell counts and oxygen evolution/assimilation rate in algae medium during the process of algae photosynthesis/respiration. Pearson correlations (p < 0.001) between algae cell counts and O₂ evolution/assimilation over a wide cell number range were observed in all experiments.

No significant difference was observed between the three repeated tests. Regression analyses indicated linear relationships between algae cell counts and both O₂ evolution (r² = 0.894, r² = 0.899, r² = 0.888 in three repeated tests; see Figure 2 and Table 2) and assimilation (r² = 0.939, r² = 0.956, r² = 0.978 in three repeated tests; see Figure 3 and Table 2) although constants and slope differed between the three experiments (Table 2). The results show that algae cell counts are more closely related to O₂ assimilation rate than to O₂ evolution rate. Figure 2 suggests that the slope of O₂ evolution at lower algae cell counts is steeper than at high cell concentration, which was observed in all the three repeats.

O₂ assimilation rates in the soil samples from the field experiment were analyzed by both oxygen optical microsensors and traditional Warburg (Figure 4). No significant difference was observed between the results measured by these two methods (Table A3). Both methods showed that samples Control and sample Kf had the lowest O₂ assimilation rates and did not differ significantly. Sample Wild with wild algae had the highest O₂ assimilation rate and sample Mix with mixed culture of wild algae and laboratory cultured algae was intermediate (not significant in the group of microsensor, see Table A4). This indicates that the laboratory cultured filamentous algae K. flaccidium did not grow as well as the wild algae species in the field test.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.