190
S. Bernard et al.
3. Ultrastructure. The presence of intracellular organelles of appreciably different refractive index to the surrounding cell, e.g. gas vacuoles in cyanobacterial species,
or calcium carbonate platelets associated with coccolithophores, can have considerable impact on cellular scattering properties (Volten et al. 1998; Whitmire et al.
2010) and thus ocean colour (Bracher et al. 2009; McKinna et al. 2011). Blooms
dominated by vacuolate species are not of ecological consequence in the southern
Benguela and will not be considered further here. Whilst coccolithophore blooms
occur in the Benguela, and can be discriminated using ocean colour radiometry
(Henderiks et al. 2012; Weeks et al. 2004), such blooms are not associated with
harmful impacts and are not considered further here.
4. Fluorescence line height and quantum yield. Both MERIS and the Moderate
Resolution Imaging Spectometer (MODIS) have a set of bands, centred at the
chlorophyll a fluorescence emission peak of ± 683 nm, to measure the sun induced fluorescence of the phytoplankton assemblage e.g. Gower et al. 1999. This
signal has a complex dependency on phytoplankton biomass, taxonomy and physiological status (Behrenfeld et al. 2009; MacIntyre et al. 2010). There is evidence
indicating that diatoms may exhibit elevated fluorescence quantum yields due
to photo-machinery able to rapidly respond to short term variability in photoacclimation in a high-mixing environment (Lavaud et al. 2002; Nymark et al.
2009). There is thus potential to use fluorescence quantum yield to discriminate
between diatom- and dinoflagellate-dominated blooms, of considerable value to
HAB application in the Benguela, and such an approach is explored further below.
An example of the large differences in the spectral reflectance, with associated and
causal differences in the measured phytoplankton absorption (as per Bernard et al.
2006) and modelled backscattering coefficients (as per Bernard et al. 2009)—as the
principle IOPs determining ocean colour—can be seen in Fig. 10.1. Data from a
diverse range of bloom types, dominated by a range of assemblages, are shown: a
mono-specific bloom (Chl a = 12.8 mg m
−3 ) of the very small-celled pelagophyte
Aureococcus anophagefferens (Probyn et al. 2010); a very high biomass bloom
(Chl a = 309.0 mg m
−3 ) of the PSP-toxic, large-celled, chain-forming dinoflagellate Alexandrium catenella (Bernard et al. 2009); a high biomass bloom (Chl
a = 172.5 mg m
−3 ) of the relatively small celled dinoflagellate Prorocentrum triestinum (ibid.); a moderate biomass bloom (Chl a = 18.0 mg m
−3 ) of the autotrophic,
large celled, phycoerythrin containing ciliate M. rubra (unpublished); and a moderate
biomass bloom (Chl a = 20.7 mg m
−3 ) dominated by a variety of diatom species.
The Aureococcus bloom (with cell counts of ± 1.5 × 10
9 cells l
−1 ) is notable
for very high reflectance values (red spectra, Fig. 10.1a), caused by the enhanced
backscattering resulting from the ± 2 μm effective diameter cell size (Fig. 10.1c),
which also results in the notably high Chl-specific phytoplankton absorption values (Fig. 10.1d) (Quirantes and Bernard 2006). This can be contrasted with the
Alexandrium bloom, where the large effective diameter of ± 30 μm results in
Chl a values > 20 times larger than the Aureococcus bloom, despite cell counts
(± 9.8 × 10
6 cells l
−1 ) ± two orders of magnitude smaller. The Alexandrium bloom
S. Bernard et al.
3. Ultrastructure. The presence of intracellular organelles of appreciably different refractive index to the surrounding cell, e.g. gas vacuoles in cyanobacterial species,
or calcium carbonate platelets associated with coccolithophores, can have considerable impact on cellular scattering properties (Volten et al. 1998; Whitmire et al.
2010) and thus ocean colour (Bracher et al. 2009; McKinna et al. 2011). Blooms
dominated by vacuolate species are not of ecological consequence in the southern
Benguela and will not be considered further here. Whilst coccolithophore blooms
occur in the Benguela, and can be discriminated using ocean colour radiometry
(Henderiks et al. 2012; Weeks et al. 2004), such blooms are not associated with
harmful impacts and are not considered further here.
4. Fluorescence line height and quantum yield. Both MERIS and the Moderate
Resolution Imaging Spectometer (MODIS) have a set of bands, centred at the
chlorophyll a fluorescence emission peak of ± 683 nm, to measure the sun induced fluorescence of the phytoplankton assemblage e.g. Gower et al. 1999. This
signal has a complex dependency on phytoplankton biomass, taxonomy and physiological status (Behrenfeld et al. 2009; MacIntyre et al. 2010). There is evidence
indicating that diatoms may exhibit elevated fluorescence quantum yields due
to photo-machinery able to rapidly respond to short term variability in photoacclimation in a high-mixing environment (Lavaud et al. 2002; Nymark et al.
2009). There is thus potential to use fluorescence quantum yield to discriminate
between diatom- and dinoflagellate-dominated blooms, of considerable value to
HAB application in the Benguela, and such an approach is explored further below.
An example of the large differences in the spectral reflectance, with associated and
causal differences in the measured phytoplankton absorption (as per Bernard et al.
2006) and modelled backscattering coefficients (as per Bernard et al. 2009)—as the
principle IOPs determining ocean colour—can be seen in Fig. 10.1. Data from a
diverse range of bloom types, dominated by a range of assemblages, are shown: a
mono-specific bloom (Chl a = 12.8 mg m
−3 ) of the very small-celled pelagophyte
Aureococcus anophagefferens (Probyn et al. 2010); a very high biomass bloom
(Chl a = 309.0 mg m
−3 ) of the PSP-toxic, large-celled, chain-forming dinoflagellate Alexandrium catenella (Bernard et al. 2009); a high biomass bloom (Chl
a = 172.5 mg m
−3 ) of the relatively small celled dinoflagellate Prorocentrum triestinum (ibid.); a moderate biomass bloom (Chl a = 18.0 mg m
−3 ) of the autotrophic,
large celled, phycoerythrin containing ciliate M. rubra (unpublished); and a moderate
biomass bloom (Chl a = 20.7 mg m
−3 ) dominated by a variety of diatom species.
The Aureococcus bloom (with cell counts of ± 1.5 × 10
9 cells l
−1 ) is notable
for very high reflectance values (red spectra, Fig. 10.1a), caused by the enhanced
backscattering resulting from the ± 2 μm effective diameter cell size (Fig. 10.1c),
which also results in the notably high Chl-specific phytoplankton absorption values (Fig. 10.1d) (Quirantes and Bernard 2006). This can be contrasted with the
Alexandrium bloom, where the large effective diameter of ± 30 μm results in
Chl a values > 20 times larger than the Aureococcus bloom, despite cell counts
(± 9.8 × 10
6 cells l
−1 ) ± two orders of magnitude smaller. The Alexandrium bloom