A multi-excitation fluorometer (MFL, JFE Advantech Co., Ltd.), originally designed to discriminate between phytoplankton species present within a population, has been redirected for use in fluorescence quantum yield (FQY) determination. While this calibration for apparent FQY requires no modification of the MFL, it is necessary to have an independent measurement of the spectral absorption coefficient of the subject fluid. Two different approaches to calibration were implemented. The primary method made use of reference fluorescent dye solutions of known quantum yield. The second method made use of acrylic fluorescent plaques and films. The two methods yielded consistent results, except in the 570 and 590 nm LED channels of the MFL. Application of the MFL in FQY determination is illustrated with an in situ Southern Ocean sample.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction and background
The light reactions of photosynthesis take place in two characteristic pigment-protein/ electron-carrier systems, known as photosystem (PS) I and II. These photosystems are equipped with a network of light-harvesting antennae pigments, which function solely to extend the range of light absorption. The absorption spectrum of a specific phytoplankton species is representative of the total absorption by all pigment-protein complexes within the cell. This includes photoprotective pigments, present within phytoplankton to aid in the dissipation of excess photon energy under high light conditions.
Cell size and pigment composition and concentration, are known to affect the absorption efficiency of phytoplankton . An increase in cell volume or intracellular pigment concentration leads to a reduction in the phytoplankton absorption efficiency due to the arrangement of pigments within chloroplasts, and chloroplasts within the cells. The phenomenon is known as the “package effect” , and is responsible for lowering the absorption coefficient of phytoplankton per unit of chlorophyll-a (chl-a) . Pigment composition and packaging are the main drivers of variation in phytoplankton-specific coefficients [4–6].
Chlorophyll-a absorbs light in the blue and red parts of the visible electromagnetic spectrum (400–700 nm), which elevates the molecule to an excited vibrational state. The return of excited chl-a to the ground state occurs either through photochemistry, non-radiative decay or re-emission as light, i.e. fluorescence . At room temperature, chl-a fluorescence is predominantly emitted by PSII and associated light harvesting complexes , and is observed as a typically conserved distribution around 685 nm . Phytoplankton fluorescence excitation spectra are derived exclusively from the re-emission of light reaching chl-a of PSII. The fixed waveband of fluorescence varies with incident light as a function of wavelength, as not all intracellular pigments contribute to PSII fluorescence, including the antennae complex of PS I and the presence of photoprotective pigments .
The natural fluorescence of chl-a is important to the non-invasive study of phytoplankton dynamics in their native environment. The quantum yield of chl-a fluorescence (ΦF), i.e. the number ratio of photons emitted as fluorescence to those absorbed by the cell, provides a first order measure of photosynthetic efficiency. The derivation of ΦF requires, aside from precise quantification of the fluorescence signal, that the spectral characteristics of the excitation energy and the spectral absorption capacity of phytoplankton be known. The ΦF in natural phytoplankton assemblages is considered an apparent ΦF, largely due to the contribution of PSI antennae pigments and photoprotective carotenoids towards phytoplankton-specific absorption and not PSII fluorescence. Chlorophyll-a fluorescence quantum yield may be derived from in situ measurements [11,12] or remotely through satellite ocean colour sensors [13,14].
Multi-excitation fluorometers, typically used to discriminate between phytoplankton taxa within a community [15–18], have been identified as a possible option for use in apparent ΦF determination . The present study describes the characterization and absolute radiometric calibration of a JFE Advantech Co., Ltd. Multi-Exciter Fluorometer (MFL, model MFL10W-CAD, SN. 0013). The MFL consists of nine excitation light-emitting diodes (LEDs, 375, 400, 420, 435, 470, 505, 525, 570 and 590 nm) and detects emitted light between 640–1000 nm. Characterized excitation LEDs and the resultant fluorescence emission, together with in situ phytoplankton-specific absorption data, allow for the derivation of the ΦF. The nine excitation wavelengths allows for unique taxonomic resolution, providing insight into the influence of phytoplankton community structure on ΦF. An improved understanding of the drivers of ΦF variability will shape the development of novel sensors and advance existing ocean colour algorithms.
2. Calibration methods
The MFL was characterized in as much detail as possible, including measurement of LED spectra (Fig. 1), spatial distribution of illumination, LED temporal cycles, detector field of view (FOV) and zero bias over temperature. The instrument was then calibrated for quantum yield using two distinct methodologies, where the primary and most direct method made use of fluorescent dye solutions of known quantum yield and spectral characteristics. The second method utilized fluorescent acrylic plaques and films to determine the response of the MFL as a function of distance. The MFL was originally calibrated by the manufacturer against a solution of Rhodamine WT dye, with the raw output signal R expressed as the equivalent signal from Rhodamine WT dye of specific dilution in parts-per-billion (ppb in water). Since we were unable to obtain sufficient quantitative information on the fluorescence of Rhodamine WT, we did not attempt to exploit this calibration. The MFL is assumed to have a linear radiometric response that is also stable with temperature. All calibrations described here were executed at an ambient temperature of 20 ± 2 °C. All radiometric quantities are photon-based. The symbols and associated units of measurement for this work are provided in Table 1.
2.1. Fluorescent dye method
Water-soluble fluorescent dyes of known quantum yield and high photostability were sought, with an emission spectrum similar to that of chlorophyll and excitation spectrum covering the full stimulation range of the MFL (350 to 610 nm). Large Stokes shift (minimal overlap of the excitation and emission spectra of the dye) was considered advantageous in order to reduce second and higher order excitation light field effects. Ideal characteristics for quantum yield standard dyes are reviewed by Demas and Crosby .
The carboxy derivatives of two suitable ATTO-TEC dyes with different spectral characteristics were selected, namely ATTO655 and ATTO490LS, the latter having an exceptionally large Stokes shift. These dyes both have a specified quantum yield of 0.3, but full uncertainty data was not available from ATTO-TEC. The absorption spectra of these dyes, together with measured LED photon spectra are shown in Fig. 1.
2.1.1. Linear calibration model
The linear calibration model for the MFL output signal response to a fluorescent solution with index i was written as
- Rij is the MFL signal response (in “equivalent ppb Rhodamine”, but we assumed the response unitless, written “MFLu”), to the fluorescent solution i exposed to LED wavelength (spectral distribution) j,
- aij is the absorption coefficient of solution i, a mean value weighted by the spectral photon distribution of LED j,
- kij is a calibration constant for solution i, LED j, expected to be independent of solution i, leaving kj,
- pi is the MFL-relative partial quantum yield factor for the fluorophore in solution i (described in Section 2.1.3) and
- Φi is the total FQY for the fluorophore in solution i.
The linear calibration process establishes the calibration constants kij determined as
A stepwise description of the linear calibration is as follows: (1) The spectral molar absorptivity εi(λ) of the working dye stock solutions was determined using a Varian (now Agilent) Cary 500 spectrophotometer and compared to the data supplied by the manufacturer. (2) The MFL was immersed in solutions of the working stock at known dilution ratios under the assumption that the absorption coefficients were linearly related to concentration. The Rij values were established by immersing the MFL in a fluorescent solution index i (the dyes ATTO655 or ATTO490LS dissolved in deionized water (dH2O)), where the equivalent absorption coefficient aij for each LED spectral photon distribution had been established. A time-averaged response Rij of the MFL was measured for each solution. (3) The LED-specific absorption coefficients of the calibration dye solutions were calculated using spectral weighted averaging. To get good signal on all channels of the MFL, solutions of the dyes at various concentrations were used. A stock dye solution of known spectral absorption coefficient was made up using dH2O, from which we could calculate aij for any diluted solution of the stock (see Section 2.1.2). (4) The MFL partial quantum yield was determined (see Section 2.1.3). This was necessary as part of the dye emission spectrum is filtered out by the MFL cut-on filter. (5) Finally, the calibration coefficient for each LED wavelength was computed using Eq. (2).
2.1.2. Dye characterization
A Cary 500 spectrophotometer, with long path (ℓ = 10 cm) cuvettes, was used to determine the spectral molar absorptivity of the dye, and hence the spectral molar absorption coefficient of any dye solutions used for calibration purposes. The FQY Φi of the dyes was supplied by the manufacturer. The spectral molar absorptivity, ε(λ) is an inherent characteristic of the dye and does not depend on concentration. The transmission 𝒯 of the solution in the cuvette was measured to determine the absorbance A (also known as optical density), related asEq. (3) it follows that Eq. (4) was exploited by fitting a straight line to the measured absorbance of solutions at known working stock dilution ratios. The working stock absorbance was determined by extrapolating the fitted line to .
The effective molar absorptivity for the dye in solution i for LED spectral channel j was computed by taking an integral, weighted by the LED photon spectrum ςj(λ) as
2.1.3. MFL-relative partial quantum yield
The MFL-relative partial quantum yield factor lies in the range 0 to 1 and takes into account the spectral response of the MFL detector channel. Suppose the relative spectral quantum response of the MFL is denoted ℜ(λ) and the emission spectrum of fluorescent solution i is denoted 𝒮i(λ), then the MFL-relative partial quantum yield factor is defined asFig. 2. The MFL-relative partial quantum yield factors calculated in this way and used in this calibration are provided in Table 2. For chlorophyll, the MFL-relative partial quantum yield factor was taken as the mean of the values for solution in ether and MeOH, that is, pc = 0.851.
2.2. Linear calibration application
In field usage, Eq. (2) is solved for Φ and a set of field-station, time-averaged MFL measurements Rj are input (data were continuously recorded for 10 sec every 1 min, over a 20 min period). The calibration factors kj have been provided by the dye calibration, and the LED channel phytoplankton-specific absorption coefficients have come from independent field measurements of the sample spectral absorption, determined spectrophotometrically as per the quantitative filterpad method . The partial yield factor for chlorophyll pc is used from Section 2.1.3. Since we no longer expect the FQY to be wavelength-independent, the field application of the calibration is written as
2.3. MFL distance calibration
The distance calibration adds one dimension of complexity to the simple linear calibration model described above. Here we assume the observed fluid medium to be isotropic and homogeneous but now consider the primary stimulating radiation from the MFL arriving at and returned fluorescence signal from elementary plane-parallel layers of fluorescent medium as illustrated in Fig. 3.
The differential response dR of the MFL to a plane parallel layer of thickness dz at distance z was written as the product
As in the case of the simple calibration model above, we introduced an indexing scheme to deal with different fluorophores with index i and different LED illumination spectra with index j. The differential MFL response was then written asEq. (5). The MFL-relative partial quantum yield piΦi is used to take the actual MFL spectral response into account.
In practice, attenuation of the stimulating radiation in the layers preceding the layer under consideration must be considered, as well as the attenuation of the emitted fluorescence radiation which is returned to the detector. If scattering is neglected, on the outward path, the transmittance of the excitation radiation through the prior layers isFig. 3 was hence estimated to be 11.4°.
2.3.1. Acrylic fluorescent calibration plaque
A red-orange fluorescent plastic acrylic (Perspex 4T56) plaque, given index i = 0 and of physical thickness t = 3 mm, was used as a calibration artifact. The spectral absorption coefficient of the plaque was measured on the Cary 500 spectrophotometer. The emission spectrum and quantum yield Φ0 of the plaque material were measured using a special experimental arrangement discussed in Section 2.3.5, after which the MFL-relative partial quantum yield factor p0 was calculated. This plaque was found to saturate one or more LED channels of the MFL even at quite large distance. It was therefore necessary to attenuate the excitation radiation as well as the returned radiation using a neutral (gray) absorption filter, free of fluorescence, of thickness tf and transmission 𝒯fj, respectively 𝒯′f placed over the fluorescent acrylic plaque (Fig. 4). The outward signal is reduced by and the return signal by 𝒯′f, providingFig. 4. The only unknowns in Eq. (15) are the functions gj(z).
2.3.2. Determination of gj(z)
The calibration functions gj(z) were established to within a scaling constant di by using a thin film of fluorescent material (plastic in this case) bonded onto a flat, opaque, black, solid substrate. Performing the experiment in dH2O, the response of the MFL was recorded as a function of the distance z between the fluorescent film and the front face of the MFL using a z-axis mechanical motion stage. The film was sufficiently thin that the total fluorescent signal could be regarded as arising at a single distance z of the film from the MFL. Two different films were used to verify independence of gj(z) from the fluorophore. The first film was given index i = 1, the second film was given i = 2 so thatFig. 3 may suggest that the gj(z) should peak at z ≈ 24 mm where the LED axes intersect the detector axis. However, the distance of peak response depends on this geometry as well as potential inverse-square decline in LED irradiance.
2.3.3. MFL calibration with fluorescent plaque
By measuring the response of the MFL to the gray filter with fluorescent plaque at different distances z0, it is possible to obtain multiple simultaneous equations from Eq. (15), and rearranging to obtain the unknown scaling constant di. Substituting an option from Eq. (18) into Eq. (15) and solving for d providesEq. (19) is known or measured, allowing calculation of d1j and similarly d2j or for further thin fluorescent films. It was found advisable to use larger z0 as the calibration at close range became problematic, possibly due to detector nonlinearities and saturation effects as well as geometry errors. Since the detector is common to all LED channels, it is possible that if any LED produced measurements near or beyond saturation, this could impact the quality of other LED channel measurements. Any such MFL data were therefore discarded.
2.3.4. Calibration application
The distance calibration process was used to establish data for gj(z) through application of Eq. (19) and then returning to Eq. (18). Now, given a medium of unknown total FQY, but known MFL-relative partial yield factor, for example chlorophyll (pc), and measured absorption coefficient aj it becomes possible to apply the calibration to a set of field measurements Rj as follows. First a set of calibration factors Gj are computed using the field sample measured absorption data as
Finally, the unknown FQY was computed essentially by solving Eq. (14) and arriving atEq. (21) must be the phytoplankton-specific absorption.
2.3.5. Plaque emission spectrum and quantum yield
The spectral absorption coefficient of the fluorescent calibration plaque was determined using the Cary 500 spectrophotometer. The emission spectrum and quantum yield Φ0 were measured by irradiating the plaque with quasi-monochromatic light in the excitation wavelength region and then measuring the apparent spectral radiance of the plaque using a calibrated spectroradiometer. The first light source used was a high intensity blue LED (Osram OSTAR Projection LE B Q8WP) having typical peak emission at 455 nm. A second high intensity green LED (Osram OSLON LT CP7P) having typical peak emission at 521 nm was also used.
Analysis began with the spectral irradiance Eλ from the Osram LED on the plaque as illustrated in Fig. 5. This was measured with a calibrated B&W Tek SpectraRad spectral irradiance meter and verified with an ASD Inc. FieldSpec 3 spectroradiometer with a Spectralon white reference panel. All spectral radiant fluxes here are in photons per second per wavelength interval. There is a wavelength-dependent fresnel reflection providing a transmittance of 𝒯F(λ) = 1 − ρF(λ) into the plaque. The plaque was assumed to have no internal scattering and the downward (as in Fig. 5) LED excitation photon irradiance within the bulk of the plaque would then decline exponentially asFig. 2.
The photon path radiance within the plaque, for modest internal incidence angles φ from the z-axis is (since the incremental path length in direction φ is dz sec φ)Eq. (29) together with Eq. (30) and Eq. (31) provides Eq. (32) are known or measured and the value of Φ0 follows. Using this method, the FQY of the fluorescent acrylic plaque (Perspex 4T56) was determined to be 0.50 for the blue LED and 0.54 for the green LED. An average value of Φ0 = 0.52 was used. It is possible that there are multiple fluorophores in this acrylic material, which could make the FQY dependent on wavelength, but the blue/green discrepancy could also be due to measurement error.
3. Calibration results and discussion
3.1. Method comparison
The field-station FQY result presented in Section 4 computed from the three calibrations shows that the calibration using ATTO490LS typically produced the lowest estimates, with the distance/plaque calibration intermediate and the ATTO655 calibration producing the highest estimates, especially in the 570 and 590 nm channels.
Due to the large Stokes shift, the ATTO490LS calibration is thought to be more reliable. It is also more consistent with the distance/plaque calibration. While the acrylic plaque and the ATTO490LS dye solutions are essentially transparent in the MFL detection spectral region, ATTO655 still has substantial absorption in the 640 to 700 nm region, which will inhibit response signal in the calibration scenario and cause artificially elevated FQY estimates in field application. This provides an argument that the calibration fluorophores should exhibit low absorption in the MFL detection spectral range, a condition not satisfied by ATTO655.
3.2. Calibration uncertainty
A full uncertainty assessment for both types of calibration is yet to be undertaken. Uncertainty in the quantum yield of the reference dyes and acrylic plaque feed directly into field sample FQY uncertainty. Besides other direct inputs to the calibration, a number of other factors could contribute to the uncertainty. The acrylic plaque has both large optical depth and strong variation of absorption coefficient with wavelength. This raises the possibility that there is significant failure of Beer’s Law for quasi-monochromatic sources such as the LEDs in the MFL. This is most relevant when the acrylic plaque is placed particularly close to the MFL. This is a potential contributor to the discrepancies in results for the 570 and 590 nm channels of the MFL.
Both forms of calibration described above are potentially compromised if used on samples having high optical attenuation, or with substantial scattering. The distance-dependent calibration method may be better suited to samples of higher attenuation or significant scattering. If the total spectral attenuation coefficient of the field sample is known, it could be used instead of absorption in Eq. (20). Adopting this approach would be equivalent to assuming that scattered photons are lost. Elevated values in the MFL backscatter channel at 880 nm could be used to flag samples with higher scattering.
4. Application example
The application of the two calibration methods to in situ oceanic data is presented in Fig. 6. The sample was collected from the Southern Ocean during the South African National Antarctic Expedition (SANAE 53) research cruise, onboard the R.V. S.A. Agulhas II, during the austral summer of 2013/2014. The MFL was continually exposed to surface water from the ship’s underway seawater supply, as part of an onboard optics suite. Data were continuously recorded for 10 sec every 1 min, and averaged for 20 min over the time of the station. The corresponding total particulate absorbance was measured spectrophotometrically (Shimadzu UV-2501 PC, ISR-2200 internal integrating sphere), as per the quantitative filterpad method . The non-algal particulate absorbance was measured , and absorption coefficients were determined . The non-algal absorption was removed from the total particulate absorption to yield the phytoplankton-specific absorption .
The two different calibration methods show similarity, notably the ATTO490LS dye and the distance calibration, providing some evidence that the calibration procedure is robust. The relationship between calibration methods for 168 stations proved consistent (data not shown). The FQY describes the ratio of light fluoresced from PSII to that absorbed by both photosynthetic and non-photosynthetic pigments within the cell. The derivation of this apparent FQY for phytoplankton in their natural environment exploits the variable nature of phytoplankton-specific absorption and PS II fluorescence to potentially provide insight to the surrounding environment. Nutrient availability , light conditions  and taxonomy  all have a profound influence on the FQY of chl-a. Complementary nutrient, photophysiological, pigment and size data exists for all stations sampled on the SANAE53 cruise, however, interpretation of FQY variability is beyond the scope of this paper. These calibrations are designed to serve as a reproducible platform for the optical manipulation of similar instruments. The unique aspect of wavelength-specific FQYs provides insight into the dynamics of individual phytoplankton species, their photosynthetic efficiencies and thus their respective contributions to carbon export in our changing climate.
A multi-excitation fluorometer has been calibrated for apparent fluorescence quantum yield using two methods and a number of fluorescent reference materials. The first method of calibration utilised fluorescent dye solutions of known FQY and the second made use of fluorescent plastic plaques and films. These methods are thought to be valid in the regime of low sample attenuation (absorption plus scattering), where the plaque method may be more applicable in higher attenuation situations. The derived FQY results from field measurements in the Southern Ocean are encouraging and largely consistent across calibration methods and materials. The inherent uncertainty of these methods and results is still to be quantified.
An instrument-type comparison would allow for development of a standardized calibration protocol for utilising multi-excitation fluorometers in FQY derivation. This improved instrument functionality will expand on the range of measurements captured by a single instrument. The thermostable MFL can be used to make individual acquisitions or record continuously over entire transects, allowing for routine measurements of FQY, concomitantly with species composition. The more field observations recorded, the more we can improve existing ocean colour FQY algorithms, allowing for continuous monitoring of phytoplankton dynamics in the global oceans.
Council for Scientific and Industrial Research (CSIR) Parliamentary grant (SNA2011112600001); Applied Centre for Climate and Earth Systems Science (ACCESS) National Research Foundation (NRF) grant (69461); South African National Antarctic Program (SANAP) NRF grant (SNA14073184298); University of Cape Town’s Marine Research Institute (Ma-Re); Nansen-Tutu Centre for Marine Environmental Research, Ma-Re.
We thank JFE Advantech, Co. Ltd., for providing technical data on the MFL. In addition, we are grateful to Mr. Bertus Theron of CSIR for extensive help with many of the laboratory experiments, and to Drs. Thomas J. Ryan-Keogh and Marié E. Smith for their valuable input.
References and links
1. A. Morel and A. Bricaud, “Theoretical results concerning light absorption in a discrete medium, and application to specific absorption of phytoplankton,” Deep. Sea Res. Part A 28, 1375–1393 (1981). [CrossRef]
2. L. N. M. Duysens, “The flattening of the absorption spectrum of suspensions, as compared to that of solutions,” Biochim.Biophys. Acta 19, 1–12 (1956). [CrossRef]
3. A. Bricaud, M. Babin, A. Morel, and H. Claustre, “Variability in the chlorophyll-specific absorption coefficients of natural phytoplankton,” J. Geophys. Res. 100, 13321–13332 (1995). [CrossRef]
4. S. Sathyendranath, L. Lazzara, and L. Prieur, “Variations in the spectral values of specific absorption of phytoplankton,” Limnol. Oceanogr. 32, 403–415 (1987). [CrossRef]
5. V. Stuart, S. Sathyendranath, T. Platt, H. Maass, and B. D. Irwin, “Pigments and species composition of natural phytoplankton populations: effect on the absorption spectra,” J. Plankton Res. 20, 187–217 (1998). [CrossRef]
6. S. E. Lohrenz, A. Weidemann, and M. Tuel, “Phytoplankton spectral absorption as influenced by community size structure and pigment composition,” J. Plankton Res. 25, 35–61 (2003). [CrossRef]
7. P. G. Falkowski and J. A. Raven, Aquatic Photosynthesis (Princeton University, 2007), chap. 2.
9. G. H. Krause and E. Weis, “Chlorophyll fluorescence and photosynthesis: The basics,” Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, 313–349 (1991). [CrossRef]
10. V. Lutz and S. Sathyendaranath, “Changes in the in vivo absorption and fluorescence excitation spectra with growth irradiance in three species of phytoplankton,” J. Plankton Res. 23, 555–569 (2001). [CrossRef]
11. S. Maritorena, A. Morel, and B. Gentili, “Determination of the fluorescence quantum yield by oceanic phytoplankton in their natural habitat,” Appl. Opt. 39, 6725–6737 (2000). [CrossRef]
12. J. R. Morrison, “In situ determination of the quantum yield of phytoplankton chlorophyll a fluorescence: A simple algorithm, observations, and a model,” Limnol. Oceanogr. 48, 618–631 (2003). [CrossRef]
13. M. Babin, A. Morel, and B. Gentili, “Remote sensing of sea surface sun-induced chlorophyll fluorescence: consequences of natural variations in the optical characteristics of phytoplankton and the quantum yield of chlorophyll a fluorescence,” Int. J. Remote. Sens. 17, 2417–2448 (1996). [CrossRef]
14. M. J. Behrenfeld, T. K. Westberry, E. S. Boss, R. T. O’Malley, D. A. Siegel, J. D. Wiggert, B. A. Franz, C. R. McClain, G. C. Feldman, S. C. Doney, J. K. Moore, G. Dall’Olmo, A. J. Milligan, I. Lima, and N. Mahowald, “Satellite-detected fluorescence reveals global physiology of ocean phytoplankton,” Biogeosciences. 6, 779–794 (2009). [CrossRef]
16. M. Beutler, K. H. Wiltshire, B. Meyer, C. Moldaenke, C. Lüring, M. Meyerhöfer, U.-P. Hansen, and H. Dau, “A fluorometric method for the differentiation of algal populations in vivo and in situ,” Photosynth. Res. 72, 39–53 (2002). [CrossRef]
17. M. Beutler, K. H. Wiltshire, M. Arp, J. Kruse, C. Reineke, C. Moldaenke, and U. P. Hansen, “A reduced model of the fluorescence from the cyanobacterial photosynthetic apparatus designed for the in situ detection of cyanobacteria,” Biochim. Biophys. Acta 1604, 33–46 (2003). [CrossRef] [PubMed]
18. M. Yoshida, T. Horiuchi, and Y. Nagasawa, “In situ multi-excitation chlorophyll fluorometer for phytoplankton measurements: Technologies and applications beyond conventional fluorometers,” in Proceedings of the OCEANS (IEEE, 2011), pp. 1–4.
19. M. Ostrowska, “Dependence between the quantum yield of chlorophyll a fluorescence in marine phytoplankton and trophicity in low irradiance level,” Opt. Appl. 41, 567–577 (2011).
20. J. N. Demas and G. A. Crosby, “Measurement of Photoluminescence Quantum Yields. A Review,” The J. Phys. Chem. 75, 991–1024 (1971). [CrossRef]
21. C. S. Yentsch, “Measurement of visible light absorption by particulate matter in the ocean,” Limnol. Oceanogr. 7, 207–217 (1962). [CrossRef]
22. M. Kishino, M. Takahashi, N. Okami, and S. Ichimura, “Estimation of the Spectral Absorption Coefficients of Phytoplankton in the Sea,” Bull. Mar. Sci. 37, 634–642 (1985).
23. C. S. Roesler, “Theoretical and experimental approaches to improve the accuracy of particulate absorption coefficients derived from the quantitative filter technique,” Limnol. Oceanogr. 43, 1649–1660 (1998). [CrossRef]