Phytoplankton biomass is a crucial measure of the health of ocean ecosystems. An impressive synthesis of the relevant data, stretching back to more than 100 years ago, provides a connection with climate change.
In 1865, Father Pietro Angelo Secchi was asked to map the clarity of the Mediterranean Sea for the Papal navy. He invented the simplest of oceanographic instruments: a 20-centimetre-wide white disk that is lowered until the observer loses sight of it, and for nearly 100 years determinations of Secchi depth were a routine part of oceanographic observations1,2 (Fig. 1, overleaf). Secchi-depth determinations assess light penetration in the upper ocean, and can be related to phytoplankton abundance. Along with measurements of the upper-ocean concentration of chlorophyll, which is found in all phytoplankton, Secchi-disk depths provide the only data available for assessing changes in the global ocean biosphere over the past century.
Boyce et al.3 (page 591 of this issue) have revisited those data, and have synthesized all available information to assess changes in phytoplankton biomass on decadal to centennial timescales, and over regional to global spatial scales. Taking great care, they created time series of phytoplankton biomass in the pelagic ocean, quantified as surface chlorophyll concentrations. They find a strong correspondence between this chlorophyll record and changes in both leading climate indices and ocean thermal conditions. They also show statistically significant long-term decreases in chlorophyll concentrations for eight of the ten ocean basins, and for the global aggregate.
Boyce and colleagues' findings are consistent with analyses of satellite observations of ocean colour, in which decreases in indices of phytoplankton productivity are mirrored by increases in ocean warming4,5,6. Satellite ocean-colour observations sample the entire globe within two days. In fact, in less than 30 seconds, the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) makes nearly half a million independent observations — equivalent to the entire historical record synthesized by Boyce and colleagues. But high-quality, global satellite observations of the ocean biosphere have been available for little more than a decade — too short a time to quantify and understand the causes of long-term trends7.
The analyses of Boyce et al. document the historical record. Looking into the future, however, satellite measurements will be the main source of data for assessing change in pelagic ecosystems. The principle is simple — the colour of the 'water-leaving' sunlight is used to determine chlorophyll concentrations. Turning that principle into practice is not simple.
First, satellites measure the reflected sunlight at the top of the atmosphere, and, typically, fewer than 10% of the photons detected relate to the oceans' water-leaving signal. Hence an atmospheric correction is required to quantify a much smaller ocean-colour signal8. Furthermore, the measurements must be accurate and stable enough to assess change over inter-annual timescales9. This requires both the on-orbit assessment of alterations in sensor characteristics over time, and a procedure to provide absolute sensor calibration8,9,10. Finally, a bio-optical model is needed to convert the remote assessments of ocean colour to oceanographically relevant quantities11, along with field observations to validate the satellite results. Thus, many interdependent components are required to create satellite observations of ocean colour that will be useful in assessing the response of ocean ecosystems to climate change.
Another consideration is that satellite missions planned at present have lifespans of only about five years, so establishing a multi-decadal time series of observations requires data from several missions. Unlike the Secchi disk, the performance of spaceborne radiometers degrades with time, and the sensors often differ in design and performance from mission to mission. To ensure that geophysical changes are not confounded by instrumental changes and/or mission transitions, ocean-colour time series must be continuously monitored and periodically updated through reprocessing of the entire data record.
A prime example is the most recent reprocessing of the ocean-colour data sets provided by SeaWiFS and by the Moderate Resolution Imaging Spectroradiometer on Aqua (MODIS-Aqua). Advances in instrument calibration and improved atmospheric correction and bio-optical models, coupled with an unprecedented attention to processing consistency, reduced the discrepancy between the two missions for deep-water, global mean chlorophyll concentrations from 12% to less than 2% over their common mission times12.
Characterization of the changes and functions of pelagic ecosystems, however, requires more than just a measurement of changes in the chlorophyll concentration. An advantage of multispectral satellite ocean-colour measurements is that distributions of other optically active constituents can all be determined9,13,14,15,16. Examples are the concentrations of coloured dissolved and particulate organic matter, particle-size spectra, and phytoplankton physiological status based on remote sensing of ocean-fluorescence properties. A better understanding of the nature of climate-change impacts on the ocean biosphere will result from an assessment based on this broader suite of parameters.
Boyce et al.3 make a sorely needed contribution to our knowledge of historical changes in the ocean biosphere. Their identification of a connection between long-term global declines in phytoplankton biomass and increasing ocean temperatures does not portend well for pelagic ecosystems in a world that is likely to be warmer — phytoplankton productivity is the base of the food web, and all life in the sea depends on it.
Unfortunately, owing to the costs and complexity of satellite ocean-colour systems and competing priorities for science funding, our future ability to assess these changes is also in jeopardy. In the United States, a National Research Council study is under way to assess issues concerning sustained satellite ocean-colour observations17. Improving the long-term understanding of changes in pelagic ecosystems, initiated by Boyce et al.3, will depend on the resolution of these issues.
Tyler, J. E. Limnol. Oceanogr. 13, 1–6 (1968).
Preisendorfer, R. W. Limnol. Oceanogr. 31, 909–926 (1986).
Boyce, D. G., Lewis, M. R. & Worm, B. Nature 466, 591–596 (2010).
Behrenfeld, M. J. et al. Nature 444, 752–755 (2006).
Polovina, J. J., Howell, E. A. & Abecassis, M. Geophys. Res. Lett. doi:10.1029/2007GL031745 (2008).
Martinez, E., Antoine, D., D'Ortenzio, F. & Gentili, B. Science 326, 1253–1256 (2009).
Henson, S. A. et al. Biogeosciences 7, 621–640 (2010).
Gordon, H. R. J. Geophys. Res. doi:10.1029/96JD02443 (1997).
McClain, C. R. Annu. Rev. Mar. Sci. 1, 19–42 (2009).
Franz, B. A., Bailey, S. W., Werdell, P. J. & McClain, C. R. Appl. Opt. 46, 5068–5082 (2007).
O'Reilly, J. E. et al. J. Geophys. Res. doi:10.1029/98JC02160 (1998).
Loisel, H., Nicolas, J.-M., Deschamps, P.-Y. & Frouin, R. Geophys. Res. Lett. doi:10.1029/2002GL015948 (2002).
Siegel, D. A., Maritorena, S., Nelson, N. B. & Behrenfeld, M. J. J. Geophys. Res. doi:10.1029/2004JC002527 (2005).
Behrenfeld, M. J. et al. Biogeosciences 6, 779–794 (2009).
Kostadinov, T. S., Siegel, D. A. & Maritorena, S. J. Geophys. Res. doi:10.1029/2009JC005303 (2009).
About this article
Journal of Natural Products (2020)
IOP Conference Series: Earth and Environmental Science (2020)
Underway spectrophotometry along the Atlantic Meridional Transect reveals high performance in satellite chlorophyll retrievals
Remote Sensing of Environment (2016)
Marine Pollution Bulletin (2016)
A Multidisciplinary Remote Sensing Ocean Color Sensor: Analysis of User Needs and Recommendations for Future Developments
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (2016)