Abstract
The FluoroProbe (FP; bbe Moldaenke, GmbH) is increasingly deployed as a means to measure in situ abundance and composition of phytoplankton communities, yet few rigorous evaluations have been made of its performance. In this study, phytoplankton strains were grown under standardized conditions to test FP performance across a range of biomass concentrations (0.2 to 20 mu g chlorophyll a (chl a) L-1). FP estimates of in vivo chl a were compared with chl a values from acetone extraction analysis. Overall, in vivo chl a detected by the FP was well correlated with extracted values (r(2) = 0.85, n = 409, p < 0.0001). However, the difference between in vivo FP measures and the acetone extracted method was high in some cases. Accuracy did not change over a range of chromophoric dissolved organic matter concentrations. Calibration with each laboratory strain improved FP accuracy over the standard instrument factory settings. The performance of two different instruments was not significantly different (p = 0.38), showing good data intercompatibility. In the majority of experiments involving multiple phytoplankton groups at various proportions and concentrations, the FP resolved all groups. The FP is a cost-effective and practical method for identifying trends in phytoplankton dynamics but has limitations with respect to accurate total chl a measurement.