|A mesocosm tool to optically study phytoplankton dynamics|Peperzak, L.; Timmermans, K.R.; Wernand, M.R.; Oosterhuis, S.; van der Woerd, H.J. (2011). A mesocosm tool to optically study phytoplankton dynamics. Limnol. Oceanogr., Methods 9: 232-244. dx.doi.org/10.4319/lom.2011.9.232
In: Limnology and Oceanography: Methods. American Society of Limnology and Oceanography: Waco, Tex.. ISSN 1541-5856; e-ISSN 1541-5856, meer
|Auteurs|| || Top |
- Peperzak, L.
- Timmermans, K.R., meer
- Wernand, M.R., meer
- Oosterhuis, S., meer
- van der Woerd, H.J.
The accuracy of remote sensing algorithms for phytoplankton biomass and physiology is difficult to test under natural conditions due to rapid changes in physical and biological forcings and the practical inability to manipulate nutrient conditions and phytoplankton composition in the sea. Therefore, an indoor mesocosm was designed to examine the optical properties of phytoplankton under controlled and manipulated conditions of irradiance, temperature, turbulence, and nutrient availability. Equipped with hyperspectral radiometers and bottom irradiance meters, it is shown that under semi-natural environmental conditions biogeochemically relevant species as Emiliania huxleyi and Phaeocystis globosa can be grown with good precision (+/- 20%) between duplicate mesocosms and between duplicate sensors (< 5% deviation). The accuracy of chlorophyll estimates by absorption, using an Integrating Cavity Absorption Meter, and fluorescence using water-leaving radiance was 74% to 80%, respectively, as it was negatively influenced by changes in phytoplankton physiology. Biomass detection was limitedto 1 to 2 mu g chlorophyll/L with an apparent linearity to 50 mu g chlorophyll/L. Estimates of the quantum efficiency of fluorescence (phi approximate to 0.01) were comparable to real-world estimates derived from satellite observations. It is concluded that the mesocosms adequately simulate natural conditions with sufficient accuracy and precision and that they offer an important tool in validating assumptions and hypotheses underlying remote sensing algorithms and models.