Phytoplankton primary production: 14C-in situ and 14C-incubator methods compared

Abstract

Abstract We estimated phytoplankton depth-integrated primary production rates at an open sea station in the northern Baltic Proper in 2004–2009 by parallel in situ and ship-board incubations in an “ICES incubator”. Observations were strongly linearly related (r2 = 0.67, p < 0.00001) and produced similar seasonal signals, but incubator estimates were significantly lower in spring (∼5–20%, March–May) and summer (∼30%, June–August), but not in autumn (September–October). Incubator PBmax was significantly lower in all seasons (∼10, 23, and 14% in spring, summer, and autumn), and the initial slope of the PB/E-curve (αB) was two-thirds of in-situ αB in spring and autumn, and half in summer, but not significantly different in autumn, and had an order of magnitude lower standard deviation. Parallel incubations of mixed 0–10 m hose samples at discrete depths in situ and in the incubator gave similar differences in PBmax and αB. This strongly suggests that the main factor causing the lower photosynthesis rates in the incubator was light quality, since all other potential sources of variability were similar. However, we cannot rule out that the lower incubator PBmax was partly due to insufficient light compared to in situ, particularly in summer. Our results suggest that a major challenge to the harmonization of 14C in situ, and artificial light and deck incubator measurements, is to simulate spectral composition changes with depth or, alternatively, use phytoplankton action spectra or constructed action spectra that are less complicated and time-consuming to establish (Kyewalyanga et al., 1997).