Download Chlorophyll a specific absorption and fluorescence excitation

Deep-Sea Research, Vol. 35, No. 5, pp. 639-663, 1988.
0198-0149/88 $3.00 + 0.00
~ 1988 Pergamon Press plc.
Printed in Great Britain.
Chlorophyll a specific absorption and fluorescence excitation spectra for
fight-limited phytoplankton
B. GREG MrrcrmLL* and DALEA. KmFERt
(Received 10 Ju/y 1987; in revised form 12 November 1987; accepted 19 November 1987)
Abstract--Methods are described for the measurement of spectral absorption coefficients,
fluorescence excitation, and fluorescence yields for pigmented particles retained on filters. The
corrections required for absorption coefficients include determining increased optical pathlength
while corrections for fluorescence include determining system spectral variability, mean fight
level and reabsorption. The empirical technique is consistent with and validated by theoretical
relationships for light transmission and fluorescence of absorbing particulate material embedded
in a medium with intense scattering.
These methods were applied to a study of photoadaptation in several phytoplankton species
and revealed variations in the blue for chlorophyll a specific absorption [a~h(~.)]and fluorescence
excitation [F*(k)] of greater than 3- and 10-fold, respectively. Variations in the spectral shapes
and the magnitude of a~h(Z) and F*(Z) with photoadaptation are determined largely by the effect
of pigment absorption in discrete particles, sometimes referred to as the sieve or package effect.
A model is presented expressing F*(Z) in terms of a*(Z) which predicts large variability in F*(k)
due to cell size and cellular pigmentation and which may help reconcile the previously reported,
but unexplained variations in F*(~,). Spectral variations in the fluorescence yield appear to be
caused by variations in the fraction of light absorbed by photosystem II which fluoresces as
compared to photosystem I or photoprotective pigments which do not fluoresce. The techniques
presented provide a rapid, reproducible, and simple approach for routine analysis, particularly
for field applicationswhere particle densities are too low for direct analysis of absorption spectra.
INTRODUCTION
ANALYSIS of optical properties of particles retained on filters has proven to be a
convenient, though generally qualitative, approach for the study of dilute suspensions
found in aquatic environments. Improved understanding of primary production and light
absorption in aquatic environments requires quantitative techniques. For example, an
improved understanding of the utilization of radiant energy by the phytoplankton crop
could be expected if quantitative techniques were available for studying the relationship
between light absorption, photochemistry and fluorescence. Aquatic ecologists have
relied on simple methods for fluorescence and photosynthetic measurements in studying
natural populations. The radiocarbon tracer technique (STEEMASNNmLSEN, 1952) has
allowed description of the temporal, regional and vertical distribution of primary
production. Fluorescence methods applied to the assessment of chlorophyll concentrations/n vitro (Y~N'rscI-I and MENZEL, 1963; HOLM-HANSONet al., 1965) and in vivo
(LoRENZEN, 1966) have provided the means to easily study phytoplankton biomass. In
* Polar Research Program A-002P, Scripps Institution of Oceanography, University of California, La Jolla,
CA 92093, U.S.A.
? Department of Biological Science, University of Southern California, Los Angeles, CA 90089, U.S.A.
639
640
B . G . MITCHELL and D. A. KIEFER
addition many studies have combined these techniques together with variable light
incubations to define the photosynthesis-irradiance response of algae (RYTHER,1956;
DUNSTAN,1973; PLATr and JASSBY,1976). However, relatively few studies have dealt
with the problem of light absorption and utilization (TYLER, 1975; DUBINSKY and
BERMAN, 1976; MOREL, 1978; BANNISTER and WEIDEMANN,1984; LEWIS et al., 1985;
KaSHINO et al., 1986). This is due, in large part, to the lack of simple quantitative
techniques for directly measuring light absorption by the dilute suspension of cells in situ.
Furthermore, most studies of fluorescence or absorption of field samples have not used
techniques that provide spectral information and therefore are limited in information
content. Since the spectral composition, as well as the intensity of photosynthetically
available radiation, varies for phytoplankton in situ, techniques that provide information
on spectral absorption and fluorescence response would be valuable tools for routine
studies of phytoplankton ecology.
Development of capabilities to estimate primary production from either remote or in
situ optical sensors will require the application of physiological models of phytoplankton
growth like that proposed by KIEFER and MITCHELL (1983). An initial effort has been
presented by COLLINS et al. (1986). The lack of a method to measure directly the
absorption coefficient of marine particulates has limited the capability to validate such
models in the field.
Determinations of the absorption coefficient of laboratory cultures have been obtained
with greater accuracy and spectral resolution than field measurements. The opal glass
technique (SHIBATAet al., 1954; SHIBATA,1958) helps to minimize distortions caused by
light scattering within the sample (DuYSENS, 1956). Because cell densities are low for
field samples, the particles must be concentrated for direct absorption measurements.
KIRK (1980) first concentrated samples by filtration and then measured the absorption
coefficient of resuspended material using a spectrophotometer with an integrating sphere
attachment. WEIDEMANN and BANNISTER (1986) concentrated samples by filtration or
centrifugation and measured the re-suspensions using the Shibata technique. YENTSCH
(1957) directly measured qualitative absorption spectra for cultures and field samples
collected on filters (YENTSCH,1962). FAUSTand NORRIS (1982) have applied derivative
spectroscopy to the study of phytoplankton spectra determined on glass fiber filters. A
significant problem for analysis on filters arises from large modifications of the light
transmission by the filters: signal amplification due to increased optical pathlength
(BUTLER,1962) must be considered for quantitative determinations. KIEFER and SOOHOO
(1982) proposed a constant factor to correct for the amplification by scattering of the
glass fiber filters. MITCHELLand KILLER(1984) demonstrated that the correction factor is
not constant, but varies with sample density. A more accurate technique for correcting
the amplification by several different filter types is described here. The technique
provides a method for the determination of volume absorption coefficients (ap) for field
particles and cultures.
Qualitative measurements of chlorophyll a fluorescence excitation provide insight to
the taxonomic and photosynthetic characteristics of populations. Using uncorrected
fluorescence excitation and emission spectra, YENTSCH and YENTSCH (1979) proposed
that phytoplankton taxonomic groups could be defined according to their major accessory pigment characteristics. Recent field and laboratory studies have also demonstrated
that the shape of fluorescence excitation spectra can be interpreted as features of
photoadaptation (MITCHELLand KIEFER,1984, 1988; NEORI et al., 1984; SOOHOO et al.,
Chlorophyll a specific absorption and fluorescence
641
1986; MITCHELL,1987). Furthermore, NEORIet al. (1986) has shown that the fluorescence
excitation for Chl a in phytoplankton is representative of the action spectrum for oxygen
evolution. Methods are presented for quantitative correction of fluorescence excitation
spectra which compensate for variations in excitation energy due to optical components
of the system, sample density, and extracellular reabsorption of the fluoresced light by
the red Chl a absorption band. A simple model indicates that intraceUular reabsorption
of fluoresced light is an important cause of variations in Chl a specific fluorescence in
vivo. Both the absorption and fluorescence techniques are easily applied for either
cultures or field samples, offering aquatic ecologists an improved ability to explore the
conversion of solar energy into biomass.
MATERIALS AND METHODS
Sample preparation
Cultures of Pavlova (Monochrysis) lutheri, Dunaliella tertiolecta, and Chaetoceros
gracilis were grown in f/2 enriched sterile seawater (GuILLARD, 1975) with continuous
illumination from cool white fluorescent lamps. Variable growth irradiances for fightlimited experiments were attained with neutral density screens, or by placement of
culture vessels at different distances from the light source. All samples were filtered with
low vacuum pressure (<5 mm Hg). Filtered samples were kept dark at 0.5-3°C before
analysis of fluorescence spectra, which was done as soon as possible after filtration. Tests
of the stability of the fluorescence signal under these conditions revealed no significant
changes for samples stored up to 8 h. However, freezing of the samples results in a large
drop in the fluorescence signal, and in some cases changes in the spectral shape, even if
the samples are analysed while still frozen. Absorption spectra appear to be stable if
frozen at-20°C.
For absorption analysis, sample and blank filters were mounted on a clear plexiglass
sheet at the back of the sample compartment to minimize the distance between the
sample and detector. Before measurement the filters were moistened with filtered
seawater until saturated, because the reflectance and transmittance of the filters varies
with the degree of saturation. Saturation of both the sample and blank was verified if a
drop of water remained after lifting the filter from the mounting plexiglass plate. The
sample may be mounted between two plexiglass plates to minimize dehydration, but this
was not essential when the spectra were run within 2-3 min after saturation.
Algorithms have been developed for three Whatman glass filter types (GF/F, GF/C,
934AH) and Millipore cellulose acetate type HA. For the analysis of photoadaptive
responses described here, the cultures were collected on GF/C filters.
Instrumentation
All analog optical signals were digitized, stored, and processed with a Digital
Equipment corporation PDP 1123 computer. Spectra of diffuse transmission of sample
filters relative to a blank filter were measured with a Bausch and Lomb Spectronic 2000
scanning, dual-beam spectrophotometer. The filter medium itself acted as a diffuser of
the measurement beam providing the benefits of the SHIBATA (1958) technique. No
additional diffuser was used. We wish to stress that the corrections we present are
specific to analyses made with the Spectronic 2000 spectrophotometer with sample
positioning as described. Acceptance angles and measurement geometry of different
642
B.G. MrrCHELLand D. A. KIEFER
instruments may be important. Some instruments we have tested do not have sufficient
dynamic range to resolve the absorbance since, for our configuration, the transmittances
of a blank filter relative to air was only a few percent. Recent analysis indicates that the
correction factors for pathlength amplification will depend on sample position, possibly
due to changes in the acceptance angle (A. MOREL, personal communication).
In order to derive empirical corrections of absorption measurements for filters,
experiments were conducted comparing the absorption for cells in suspension to the
same material on filters. Absorption coefficients of the suspensions for these experiments
were measured with a vertical path (Butler) transmissometer (BUTLER, 1962; KEFER and
SOOHOO, 1982; MITCHELLand KIEFER,1984), which consisted of a Bausch and Lomb high
intensity monochromator (33-86-76) coupled to a Bausch and Lomb 45 W tungstenhalide light source. Transmitted light intensities of suspensions and blanks were measured with a Gamma 20-20-10 photomultiplier and 2900 autophotometer. For measurements of suspensions with the Butler spectrophotometer, a diffuser was placed between
the sample and detector in accordance with the Shibata technique. The instrument
optical geometry has the design advantage of maximizing the angle of acceptance of the
measurement beam and minimizing effects due to settling of the cells in the suspension.
Measurements of fluorescence excitation spectra were made with the same instrument
used for absorption spectra of suspensions but reconfigured in the following way. The
mirror used to reflect the monochromatic excitation beam onto the sample was replaced
by a long wavelength transmitting dichroic filter (OCLI CSF-A). This filter reflected
wavelengths <550 nm onto the sample but allowed the Chl a fluorescence to pass to the
photomultiplier which was mounted above the dichroic mirror. A 680 nm bandpass filter
(Melles-Griot 03-FIV-065) placed in front of the photomultiplier prevented excitation
light reflected off the sample from reaching the photomultiplier while allowing the Chl a
fluorescence to pass (Fig. 1). This optical configuration facilitated calibration of the
relative spectral excitation quantum flux at the sample using the fluorescent dye,
rhodamine-B, which was dissolved in ethylene glycol at a concentration of 3 g 1-1. Under
these conditions, rhodamine-B has been found to be a quantum counter which absorbs all
photons and emits them with a constant fluorescence quantum yield in a broad band
above 600 nm (1VIELHUISH,1962). A smoothed average of five rhodamine-B spectra was
normalized to unity at the maximum (545 nm for this system) and used in the algorithm
to correct for the system spectral response. Because of spectral limitations of the dichroic
filter, spectra presented here are restricted to excitation wavelengths <525 nm. Signal
stability was tested with a fluorescent glass (Coming no. 3486) which has excitation and
PHO]'PO-LIER
MULTI
MONO
CHROMATOR
-
PHOTOMETERS]COMPUTER
I
Fig. 1. Schematicdiagram of the vertical path spectrophotometerconfiguredfor fluorescence
excitation measurements.
Chlorophylla specificabsorption and fluorescence
643
emission characteristics similar to Chl a and is stable through time (PARKERand
VAUDmSY, 1980). The signal at 512 nm was tested before and after sample measurements, and system drift was compensated by varying the high voltage to the photomultiplier. Spectral variations due to lamp aging were tested by comparing the shape of the
rhodamine-B spectra through time. If signals for the fluorescent glass or system spectral
shapes deviated significantly from nominal conditions, the lamp was replaced and the
system recalibrated.
CORRECTION ALGORITHMS
Volume absorption coefficient algorithm
To maximize the spectral absorption signals for field samples when water sample
volume is limited, measurements are sometimes made on double layers of a sample filter
which has been folded or cut in half and stacked. Therefore, a two-step procedure has
been developed to derive the phytoplankton volume absorption coefficient, aph(~, ) (m-l),
from these measurements. Beer-Lambert law for absorbing materials in pure solution
predicts a doubling of absorption when the geometric pathlength is doubled. Since the
filters diffuse the light in the measurement beam making the actual optical pathlength of
the sample greater than the geometric pathlength, theories for non-scattering solutions
are not applicable (SmSATA, 1958). BtrrLER (1962) has defined the term 13as the ratio of
the optical to geometric pathlength.
The problem of assessing 13has been approached as follows. For two-layer samples one
must first correct for 13"(~,2) for a second layer relative to a single layer and then for
13(~.,1) for a single layer relative to a non-scattering suspension. The definition of 13,
according to BtrrLER (1962) specifically referred to the absorption amplification induced
by diffusing media relative to non-diffusing media. The parameter 13" is defined as the
increased optical pathlength observed for multiple filter layers relative to a single layer,
where both are the same diffusing medium.
The optical density, the signal generated by most commercial spectrophotometers, is
defined as
OD = -Log10 (Ts/Tb),
(1)
where Ts and Tb are the transmitted irradiance of the sample and blank, respectively.
Defining ODs to be the sample optical density, X to be the ratio of the volume of sample
filtered to the clearance area of the filter, ~. to be the wavelength, and n to be the number
of stacks, the phytoplankton volume absorption coefficient can then be determined by
equation (2):
aph(~,) --
2.30Ds(~.,n)
X13(g,n)
(2)
ODs(X,n) in all eases has been corrected by subtracting the measured value at 750 nm.
We thus assume that phytoplankton absorption at 750 nm is negligible.
When measurements are made on two filters, 13"(X,2) required to convert the value to
an appropriate value for a single layer is defined by the OD for two layers divided by two
times the OD for one layer:
ODs(X,2)
13"(E,2) - 200s(X,1)"
(3)
644
B.G. MrrCHELLand D. A. KIEFER
This relationship is a function of the sample OD and the specific diffusing medium
containing the sample. Empirical studies were conducted to develop algorithms for
estimates of {)*(~.,2). The results of this analysis for Whatman GF/C filters are illustrated
in Fig. 2, which is a composite of measurements on two different volumes of culture
filtered so that there was a broad range of ODs. In principle ~l* cannot be less than 1. The
data show the dependence of ~* on ODs and agree well with the prediction of DUNTLEY'S
(1942) tWO flOW radiative transfer model.
The parameter ~ in equation (2) is defined to be the ratio of the absorption of cells
measured on a single filter layer to the same equivalent pathlength of cells suspended in a
solution of bovine serum albumin (BSA) which closely matched the refractive index of
the cells. By matching the refractive index of the medium to the cells, scattering was
mimimized but not eliminated. The BSA technique, combined with the favorable
geometry (wide acceptance angle) of our Butler vertical path spectrophotometer and use
of the Shibata technique allowed accurate determination of the volume absorption
coefficient for phytoplankton suspensions used to develop the correction algorithm for
filters. Inspection of Figs 2 and 3 reveals that both f) and ~)* show strong dependence on
the sample OD. Thus, these two parameters of optical pathlength will vary with the
amount of material filtered, and will vary spectrally for individual samples as the ODs
varies. We assume that scattering within the sample is dominated by the filter itself, and
that this scattering has no spectral dependence.
The detailed study of ~ and 13" for GF/C filters provides a convenient algorithm, which
is easily implemented on a digital computer. However, much evidence now indicates that
1.75,
~ ~
x~
"< 1.5-
-^ X
1
,
.2
i
.4
i
.6
X X [3
1
.8
ODs(X,2)
Fig. 2. The relationshipbetweenIB*(k,2)definedas ODs(k,2)/2ODs(k,1)and ODs(~.,2)for GF/C
filters. The (Vl)symbolsare experimentalobservations,and the (X) symbolsare the least-squares
best fit equation used in the correction algorithm. The solid line is the predicted relationship
according to DUNTLEY'S(1942)two-flowmodel of the optical properties of diffusingmaterial.
Chlorophyll a specific absorption and fluorescence
645
2.5
2.25
[3
×0%
a x~x u
x
13
[3
,~< 1.75
x
1.5
[:]
1.25
':
'3
',
s
OD s (X,1)
Fig. 3. The relationship between [1(~,,1) for a single GF/C filter and the observed ODs(k,1).
Presented are the experimental observations (D) and the equation used in the correction
algorithm(X).
the filtration efficiency of GF/C filters may not be sufficient to retain the picoplankton,
including cyanobacteria which are now recognized as being important components of the
phytoplankton (LI et al., 1983; PLATr et al., 1983; ITURRIAOAand MITCHELL, 1986).
Therefore, additional techniques have been developed which allow the use of three filter
types with finer particle retention capabilities than the GF/C. These include Whatman
glass fiber types 934AH and GF/F, and Millipore cellulose acetate type HA. Empirical
relationships between [3(~.)and OD(Z) were assessed by filtering five different volumes of
a culture of D. tertiolecta onto GF/C filters and each of the other three filter types. A
simple plot of the OD(Z) observed for the finer retention filters against the OD(X) for the
GF/C revealed linear relationships with the slopes >1.0 and intercepts of 0.0. This was
true for double or single layer comparisons. Figure 4 shows the results of this analysis for
a single GF/F sample plotted against a single GF/C sample. Since the correlation
coefficients (r2) for least-squares linear fits of these data were >0.98 for 300 points in
each comparison, OD data collected for various filter types can reliably be converted to
the equivalent OD on a GF/C filter. Once this is done, the detailed algorithm developed
for GF/C filters can be used to correct the spectra. The slopes of these relationships are
higher for the finer retention filters as shown in Table 1. This is apparently due to denser
weave and thicker glass fiber mats in the 934AH and G/FF, and more refractive material
in the HA.
The agreement between the corrections for the different filter types is illustrated in
Fig. 5. The 10-fold range of maximum OD for the uncorrected spectra is a much broader
646
B . G . MrrCHELLand D. A. KIEFER
O. 2 5
O. 2 0
/'x
0.15
LL
LL
n
0
0.10
O. 05
O.
OE
0.00
,
i
0.05
0.10
i
0.15
i
0.20
0.25
ODGFC ( i )
Demonstration of the linearity between the ODsQ.) measured for various filter types and
that observed for GF/C filters. Presented here are the ODs(~.,1) for single layers of G F / F
vs G F / C . The regression equation for 3011 data points (400-700nm) is O D G F F ( ~ . , 1 )
F i g . 4.
= 0.0 + 1.30DGFC(~.,1)
r~ = 0.998.
Table 1. Regression analysis of several filter types relative to Whatman GF/C.
Optical density of Millipore HA, Whatman 934 A H and GF/F are the dependent
variables; optical density of Whatman GF/C is the independent variable. For
analyses, the intercept is 0.0
Filter type
GF/C
934 AH
GF/F
HA
Mean
slope
S.D.
Number of
volumes
1.0
1.2
1.3
1.5
0.10
0.09
0.07
5
5
5
range in sample density than one would encounter in routine analyses. The spectra are
statistically indistinguishable, except in the Soret peak where the maximum and minimum values at 435 nm (for the H A and GF/C types, respectively) are significantly
different (P < 0.005, n = 5). The flattest peaks were obtained with the GF/C and
934AH types which are the least diffusing, judging from the slopes in Table 1. If the
measurement and reference beams are not adequately diffused by light scattering within
the filter, then the benefits of the SHIBATA (1958) technique will not be fully realized
unless an additional diffuser is used.
Chlorophyll a fluorescence excitation algorithm
In order to quantitatively compare fluorescence excitation spectra between different
samples, several corrections are necessary. The minimum requirements include correction for variability of the spectral quantum flux of the excitation beam, the reflectivity
647
Chlorophyll a specific absorption and fluorescence
5
34AH
(_)
4 #f/.~ GFIC
~~
LL
LL
3
I-Z
W
U
Z
O
#
CL
O
m
1
0
400
I
i
i
I
450
500
550
(500
i
1550
700
WAVELENGTH
Fig. 5. A comparison of the correction algorithm applied to samples of D. tertiolecta filtered
onto the four filter types studied. Each curve represents the mean value of the derived volume
absorption coefficient [aph(m-1)] for five different volumes filtered representing a 10-fold range in
raw OD for each filtertype. The curvesfor differentfiltersare symbolizedas follows:Millipore
HA (
); Whatmanglassfiber types GF/F (-----); 934AH (--); and GF/C (-.-).
and absorption of excitation energy by the sample, and reabsorption of emitted light in
the fluorescent waveband. Correction for these factors provides what is defined to be the
relative fluorescence excitation spectrum. Further corrections for system geometry, and
absolute calibration would yield a measure of the absolute fluorescence excitation
spectrum. Our system geometry and calibration procedure ensures that the latter factors
remain constant for the measurements, but an absolute calibration has not been
performed, so the techniques described here provide relative fluorescence excitation
spectra. The fluorescence, F, measured by any system can be expressed in simplest terms
as
F = GR~fAeAf.
(4)
G and R represent geometric and reflectivity factors, which are assumed to be constant.
The reflectivity is defined as the fraction of quanta reflected by the sample filter in the
backward direction divided by the total incident quanta. The measuring geometry (G) is
constant, but R is probably not constant over all possible sample densities. Nonconstancy in R is due to the fact that as ODs increases, the probability for escape in the
backward direction decreases. Limitations imposed by this constraint are discussed
below. The fluorescent quantum yield is represented by ~f while Ae and Af are
parameters for the absorption of excitation quanta and reabsorption of fluoresced
quanta. General expressions which allow analysis of equation (4) have been derived by
FORST~R (1951) who presented phenomenological equations for correction of luminescence data for both turbid and strongly absorbing samples. Many authors have applied
these concepts to a variety of problems (WEBER and TEALE, 1957; MELHUIS8, 1961;
648
B . O . MITCHELL and D. A. KIEFER
DEMAS and CROSBY, 1971; PARKERand RE~S, 1962). A general summary can be found in
LIPSETr (1967). The product [Ae Ad in equation (4) has been stated more explicitly by
WINEFORDNERet al. (1972):
AeAr = [fL(3.e)(l - 10-°Ds(~°)) d~,~] [f(1-10-OD'(b)) dXf],
[fODs(~,f) dXf]
(5)
where L(~.~) is the spectral irradiance quantum flux of the excitation energy and ODs(~.~)
and ODs(~f) are the optical density for the sample at the excitation and fluorescence
wavelengths, respectively. The first integral in equation (5) is the absorbed irradiance,
integrated over the bandwidth of the excitation source. The final ratio of integrals in
equation (5) accounts for reabsorption of fluorescence within the sample and is integrated over the emission bandwidth. The values for F, ODs(~.~) and ODs(~.f) are measured
directly for each sample. L(Xe) is constant for all samples through normalization with the
rhodamime-B spectrum and the fluorescent glass standard. To compare spectra quantitatively, the measurement and correction techniques must provide linearity with variable
sample volume over a practical range expected for samples. A comparison of corrections
at 485 nm for L(485) only, and a full correction utilizing an analytical solution of
equation (5) is provided in Fig. 6. One can see that over a narrow range corresponding to
very low OD, linearity exists for a simple correction involving the excitation energy only.
However, satisfactory absorption measurements at such low OD are impractical due to
instrumentation noise, especially in the low absorbing yellow-green (550-650 nm)
spectral region for phytoplankton. Linearity can be achieved over a broad range of
1.2
l.O
w
u
z
w
u
0~
w
no
x
•
x
X
X
0.8
X
-3
1.1-
0.6
XX
w
>
b-
X
0.4
<
_J
w
rr
0.2
,/
0.0
o
10
I
i
I
I
20
30
40
50
VOLUME
60
(ml)
Fig. 6. A comparison of the CM a fluorescence excitation at 485 nm corrected for the excitation
energy only (x) and a full correction applying an analytical solution of equation (5) (O). The
experiment was conducted over a 100-fold range in sample density determined by the volume
filtered. Note the non-linearity at relatively low sample densities if the appropriate corrections for
reabsorption and mean excitation quantum flux in the sample are not considered. The corrections
presented here linearize the response over a broad range of sample densities.
649
Chlorophyll a specific absorption and fluorescence
sample density by-using the fluorescence corrections presented here. For the data
presented in Fig. 6, the ODs for the chlorophyll peaks at 435 and 675 ranged from 0.008
to 0.48 and 0.005 to 0.29, respectively. The non-linearity after correction which occurs at
higher sample volume [OD(435) > 0.3] is believed to be caused by changes in the term R
in equation (4). We did not measure R, but have used our empirical data (Fig. 6) to
estimate the ODs where it becomes a significant concern. To ensure high quality
z..•
o.o,~
to
" ~ /
o.no
0. 0 0 0
400
i
i
450
500
i
550
i
600
--
i
650
7DO
WAVELENGTH
2.0
b
Z
0
U
X
L~J,
bjLIU
Z
h,
U
Ln
LU
n,0
-7
_I
b..
0
0. I
400
i
i
i
i
i
425
450
475
500
525
550
WAVELENGTH
Fig. 7. Mean spectra for determinations of three different volumes bracketed by the standard
error of the mean for (a) the Chl a specific absorption a~h(k) and (b) the Chl a specific
fluorescence F*(Z). T h e spectra are for Pavlova lutheri grown at 25 ~tEin m -2 s-1.
650
B . G . MITCHELL and D. A. KIEFER
absorption spectra, and linearity in the fluorescence correction which assumes R to be
constant, a working range for ODs between 0.1 and 0.3 is recommended for the peak
phytoplankton absorption in the Soret band. This corresponds approximately to 1-3 ~tg
Chl a on a 25 mm filter or 0.3-1.0 ~tg Chl cm-2.
Figure 7a and b illustrate the mean spectra for Chl a specific absorption and
fluorescence excitation [a~h(~.) and F*(k)], respectively, bracketed by the standard error
of the mean for Pavlova lutheri grown at 25 ~tEin m -2 s-1. The data represent results
from three separate volumes filtered corresponding to a 5-fold range of sample density.
The results indicate rather good reproducibility of the absorption and fluorescence
techniques described.
RESULTS
Spectral changes: the effects o f discreteness
The techniques described here were applied to a study of the response of nutrient
saturated batch cultures to light limitation. Absorption, fluorescence excitation and
fluorescence efficiency spectra showed spectral shifts associated with increases in the
relative contribution in the accessory pigment bands compared to the Chl a peak in the
Soret at 435 nm. Spectral summaries of the data for P. lutheri and D. tertiolecta are
presented in Figs 8-10. Table 2 summarizes data for these species and for Chaeloceros
gracilis. NEORI el al. (1984) and MITCHELLand KIEFER (1984) proposed that the ratio of
the fluorescence signal in the accessory pigment bands to the signal at 435 nm can be used
as an indicator of photoadaptation for field populations. For this study the F-ratio is
defined to be [F(465)+F(550)]/F(435), where F(~.) is the corrected fluorescence excitation at wavelength ~,. An analogous parameter, the A-ratio can be defined for
absorption spectra. Inspection of Figs 8 and 9 reveals that low-light-adapted cultures
have higher A- and F-ratios which are summarized in Table 2.
Analysis of the summary data in Table 2 indicates that spectral shifts are not due
exclusively to increases in accessory photosynthetic pigments relative to Chl a or changes
in ~f. The consequences of packaging the pigment molecules into discrete particles are
hypothesized to be of equal or greater significance in determining the spectral shifts
which are observed. Comparing the P. lutheri culture growing at the highest light level to
the intermediate light level, Chl a and c per cell increase 2.5- and 6.6-fold, respectively.
The increases in the F- and A-ratios for this adaptive transition are presumed to be
caused primarily by increases in accessory pigments relative to Chl a. However, adaptation for growth at the lowest irradiance induces a further two-fold increase in Chl a per
cell, while Chl c per cell remains constant. In spite of this increase in the Chl a:Chl c
ratio, both the F- and A-ratios increase further. This result may be caused by increases in
other accessory photosynthetic pigments which do not covary with Chl c. However, we
hypothesize that the enhanced spectral ratios in this photoadaptive range, for this
species, are caused largely by the packaging or sieve effect which governs in vivo
absorption properties of discrete particles (DuYSENS, 1956; MOREL and BRICAUD, 1981;
BRICAUD el al., 1983; LATIMER, 1983; FALKOWSlOel al., 1985; COLLINS et al., 1985;
DUmNSKVel al., 1986). This effect is expected to be more pronounced in the strongly
absorbing regions of the spectrum altering the spectral ratios for fluorescence and
absorption. Our hypothesis is supported by the data of GALLAGHERel al. (1984) who
found that the ratios among Chl a, Chl c, and fucoxanthin for three clones of
651
Chlorophyll a specific absorption and fluorescence
P.
0,035
0.030
a
640
170\
Z
o
o0
O. 0 2 0
U
0.015
U
O. OJO
W
LUTHERI
/"
"
A
A
0.005[
0.000/
400
i
i
i
e
i
450
500
550
600
650
700
WAVELENGTH
O.
O. Oe;O
TERTIOLECT^
~75
Z
0
i.--4
b
/\
0.04';
/
\
n,Z
uE
U
W
(1.
~)
0.015
'\\\
/Xll
O. 000
400
450
500
550
600
650
700
WAVELENGTH
Fig. 8.
The photoadaptive response of the Chl a specific spectral absorption coefficient a~h(~, )
(m 2 mg Chl a -1) for P. lutheri (a) and D. tertiolecta (b). Curves for relative light levels for each
culture are symbolized as: high (--); moderate ( - - - ) ; low (---). Each curve is the mean of 2 or 3
separate determinations for different volumes. The significant drop in a~h and increases in the Aratio and ratios of aph(675)/aph(435) at lower growth irradiance are attributable primarily to the
effects of absorption in discrete particles.
652
B.G. MITCHELLand D. A. KIEFER
P.
LUTHERI
a
2.0
1.5
c
Oe)
,"" 25
\.~._..
U
1.0
0
k
0"5I
0.0
400
i
i
I
!
42';
450
475
500
52";
WAVELENGTH
D.
8
TERTIOLECTA
b
Z
0
\_/
O
c
4
0
U_
k
n
400
i
s
i
!
425
450
475
500
52';
WAVELENGTH
Fig. 9. The photoadaptive response of the Chl a specific fluorescence excitation spectra for
683 nm emission (F*) for the same samples presented in Fig. 8: P. lutheri (a), D. tertiolecta (b).
The fine symbolism is consistent with Fig. 8. Changes in shape and magnitude of the spectra are
attributable to the effects of discreteness and intracellular reabsorption of fluorescence.
Skeletonema costatum behave differently for identical growth conditions with no
consistent trend in ratios for high- and low-light cultures. Even though the pigment
ratios did not show a clear pattern, they did observe that both total pigment per cell
and fluorescence excitation by accessory pigments relative to Chl a increased at low light
for each clone.
653
Chlorophyll a specillc absorption and fluorescence
P.
15
>..
U
Z
bJ
U
10
h
h
Ld
. . . . .~ .25
...
LIJ
U
Z
hi
U
tn
hi
nO
"1
J
I.l.
LUTHERI
-..
X'////
-~'-'~...~-~-~.~//
/
640
a
0
I
400
425
I
I
i
450
475
500
525
WAVELENGTH
D.
35
TERTIOLECTA
30
U
Z
LU
25
u
D-q
U.
It.
L;J
bJ
u
Z
bJ
0
bo
bJ
n0
2O
\575
"--"
]5
]0
.J
I.L
b
n
i
400
425
i
i
i
450
475
SOD
52';
WAVELENGTH
Fig. 10. The fluorescence efficiency spectra for the same samples presented in Figs 8 and 9:
P. lutheri (a), D. tertiolecta (b). Line symbolism is consistent with Figs 8 and 9. The shapes and
magnitudes of the efficiency spectra are proposed to be dependent primarily on the relative
absorption by photosystem II which fluoresces as compared to photosystem I or photoprotective
pigments which do not fluoresce under the experimental conditions described in the text.
Chlorophyll a specific absorption and fluorescence excitation spectra
The hypothesized effects of discreteness are consistent with the trends in the quantitative Chl a specific absorption and fluorescence spectra in Figs 8 and 9. Fig. 8 indicates
that the Chl a specific absorption in the blue [e.g. a~h(435)m 2 mg Chl a-1] is significantly
lower at low growth irradiance compared to the higher growth irradiances. This effect is
654
B . G . MrrcrmLL and D. A. KmFER
Table 2. Pigment and optical parameters for three species during light-limited exponential growth. Units are
(m -1) for aph, (m 2 rag-1 Chl a) for a*ph and fluoresceneelChl a) for F*
Parameter
Pavlova lutheri
Chaetoceros gracilis
Dunaliella tertiolecta
ttEin m-2 s-1
Chl alc or alb
A-ratio
F-ratio
E-ratio
aph(675)laph(435)
a*ph(400-700)
a*ph(435)
a*ph(675)
F*(400-525)
F*(435)
Relative Of(435)
640
6.0
0.64
0.85
3.20
0.433
0.011
0.030
0.013
1.34
2.05
0.53
170
2.3
0.73
1.05
2.80
0.520
0.01
0.025
0.013
1.53
2.16
0.63
25
5.0
0.96
1.24
2.40
0.650
0.008
0.017
0.011
1.31
1.69
0.90
300
3.3
0.75
1.20
2.95
0.480
0.011
0.027
0.013
0.74
1.05
0.30
100
3.3
0.95
1.30
2.60
0.620
0.010
0.021
0.013
0.79
1.06
0.40
33
3.3
0.96
1.40
2.65
0.580
0.009
0.019
0.011
0.55
0.71
0.35
575
2.8
0.81
0.84
2.09
0.415
0.020
0.053
0.022
4.50
6.80
0.58
220
3.2
0.77
0.88
2.34
0.415
0.024
0.065
0.027
6.50
10.00
0.57
80
3.0
0.91
1.05
2.11
0.563
0.014
0.032
0.018
5.10
7.30
1.30
23
2.9
0.98
1.21
2.27
0.714
0.010
0.021
0.015
2.20
3.10
0.70
much more pronounced in D. tertiolecta than in P. lutheri. Changes in a~h(675) are not as
large as in the blue, which is expected since the red peak is less strongly absorbing than
the blue peak. Since the magnitude of absorption in the accessory pigment region from
460 to 530 nm is intermediate between the blue and red peaks, one would expect that this
region will not be affected as much as the 435 nm peak. Thus, without any changes in the
relative concentration of accessory pigments, increased pigment concentration per cell
will lead to spectral shifts in the absorption and fluorescence excitation spectra. Indeed,
for D. tertiolecta, no significant change in the Chl a:Chl b ratios were observed, although
spectral shifts comparable to those observed for P. lutheri were evident.
The changes in the fluorescence excitation spectra presented in Fig. 9 are more
complicated than the observations for absorption. For all species, the intermediate light
level has slightly enhanced Chl a specific fluorescence excitation [F*(X)] throughout the
spectral range studied as compared to the highest light level. Two possibilities must be
considered in explaining this phenomenon: increases in the ratio of PSII/PSI antennae
Chl a, or increases in Of. By increasing Chl a in PSII relative to PSI, a larger fraction of
the chlorophyll present will be fluorescent, since PSI Chl a is considered to be nonf l u o r e s c i n g (BUTLER, 1978). An increase in ~f(~.) may also play a role in increased F*(L);
however these data are not adequate to distinguish between these two hypotheses.
At the lowest growth irradiance, a significant drop in F* (~) is observed for all species.
Again, two probable explanations exist for this observation. First, a decrease in PSII/PSI
Chl a ratios could reverse the trend observed between the highest and intermediate light
levels. PSI absorption may be enhanced more as cells undergo a transition from
moderate to low irradiance in order to keep the flux of excitons balanced between the two
photosystems. A second factor leading to the observed decrease in the specific fluorescence may be attributable to increasing intracellular reabsorption of the fluoresced light.
This phenomenon, which has been shown to cause spectral shifts in fluorescence emission
spectra (CoLur~s et al., 1985) is expected to be increasingly important with increased
Chl a per cell. Although these data cannot be used to test the significance of the first
hypothesis, they are sufficient to test the second.
A model of specific fluorescence based on specific absorption
The 3-fold drop in F*(435) for D. tertiolecta and photoadaptive variability in spectral
ratios associated with minimal changes in Chl a:Chl b ratios can be reconciled by
Chlorophylla specificabsorptionand fluorescence
655
examining absorption and fluorescence at the cellular level. COLLINS et al. (1985)
examined the role of intracellular reabsorption of fluoresced light in modifying the
spectral shape of the emission spectrum, They presented a general description for the
volume integral of fluorescent flux from a cell:
F(Le) = ~f(~,e)L(~.e)Qa(ke)~r2Q*a(kf)
(6)
L(ke) is the excitation irradiance, ~f(ke) is the fluorescence efficiency for absorbed light,
and nr 2 is the cell's geometric cross-section. The efficiency factor for absorption, Qa and
intracellular reabsorption Q~* have been defined by MOREL and BRICAUD(1981):
Qa(k) = 4/3Q*(~,)aem(k)r
(7)
a*(~,) = (ci/acm(~,))a~,h(~,),
(8)
where ci is the intracellular Chl a concentration and acm(L) is the spectral absorption of
cellular material dispersed into a theoretical suspension. By substituting equations (7)
and (8) into equation (6) an expression for the chlorophyll specific fluorescence (F*) at
the cellular level can be represented by
F*(~,) = [~f(~,e)lac*(kf)]a~h(Le)a~h(kf),
(9)
where F*(ke) is defined to be the emergent cellular fluorescence corrected for excitation
energy divided by the intracellular Chl a and acre* is the Chl a specific absorption
coefficient for cellular material dispersed into a theoretical solution:
F*(~,e) = F(~,e)l[L(~,e)4/3nr3ci]
(10)
a*m(Ef) = a~(~,,)/ci.
(11)
Since absorption by cultures in the red peak centered at 675 nm is dominated by Chl a,
a ~ ( ~ ) is assumed to be a constant. Thus equation (9) can be utilized to assess the
relative importance of Of(~,e) vs the product a~h(~,e) × a~n(~) in determining the
magnitude of the chlorophyll specific cellular fluorescence. Examination of Table 2
reveals that for the species studied, the apparent variability in ~f(435) is a factor of 4
while the product a~,n(435) x a~,h(675) varies by 10-fold. Although variations in @f are
significant, in determining the magnitude of F*(~,e), one may conclude that the product
of a~,h(Z,~) X at~h(~) is of greater significance. Figure 11 is a plot of the relationship
between the product a~h(435) X a~,h(675) and F*(435) for the three species studied and
the predicted variability based on a~h(~,) estimated using the model embodied in equation
(9). The modeled predictions assume that a~m(435) is equal to 5 x 105 m -~ and that the
ratio of acm(435)/a~m(675) is 2.3 (see also BRICAUD et al., 1983). For the three species
studied under light-limited growth, a significant positive trend was observed for this
relationship and the model predictions provide a good fit to the observations. These
results support the hypothesis that the magnitude of F*().) is determined largely by the
physical consequences of absorption in discrete particles.
As predicted by MOREL and BRICAUD (1981) and BRICAUDet al. (1983), if the acm(Jq)
and ci for a theoretical cell suspension are held constant, while cellular diameters vary,
then a~h(L) will decrease for large cells as a consequence of the discreteness effect which
effectively results in self-shading of pigment molecules on the interior of plastids, thereby
decreasing pigment specific absorption efficiencies. One predicts, therefore, that F* for
656
B.G. MITCHELLand D. A. KIEFER
larger cells, for constant acm(~,) and ci, will vary inversely with the absolute size of the
cell, and the relative absorption of the excitation band. The latter prediction results from
a larger drop in a~h(~,) in the more strongly absorbing region of the spectrum (MORELand
BRICAUD, 1981).
Analysis of Chl a fluorescence for size-fractionated field populations (ALPINEand
CLOERN,1985) and variable-sized phytoplankton in culture (LOFTUSet al., 1972) agree
with these predictions. Using the nominal fraction or cell sizes reported by these authors,
and the same assumptions regarding acm(~,) used to model the response of photoadaptation in Fig. 11, the response of F* to variations in cellular diameter was investigated.
The relative predictions of this model with the relative observations of ALPINE and
CLOERN (1985) and LoF'rUs et al. (1972) are presented in Table 3. The predicted range of
variability in F* agrees well with observations. Although one must assume that acm(435)
is constant for lack of the appropriate data, it certainly is not constant. The fact that
F*(435) varies 6-fold for D. tertiolecta with presumably little change in cell diameter is
attributable directly to large variability in aem(435) and acre(675). Nevertheless, the
overall agreement is very encouraging. Based on theory, the observed variability induced
by photoadaptation, and the observed variability attributed to size, we believe that the
concepts embodied in equation (9) should be carefully considered in efforts to interpret
the wide-ranging values for F* reported in the literature (Lorrus et al., 1972; KIEFER,
1973a; ALPINEand CLOERN, 1985). More detailed laboratory studies should be conducted
in order to test the hypotheses presented here.
Fluorescence efficiency spectra
Compared to the absorption and fluorescence excitation spectra, the fluorescence
efficiency spectrashow smaller and inconsistent changes in shape depending upon
15'
10"
r~
r"l
L
5'
gl~
0
0
DD
.0;05
.0;1
a;h(435) X al;h(675)
.0;15
.002
Fig. 11. The relationship between P(435) and the product a;h(435) x a;h(675) for the three
spedes studied. The experimentalobservations are symbolizedas (D). The solid line is the
predictionof the modelembodiedin equation (9).
Chlorophyll a specific absorption and fluorescence
657
Table 3. Comparison of model prediction of Chl a specificfluorescence using equation (9) with the field data of
ALn'tNe and CLOE~ (1985), and the laboratory data of Loertls et al. (1972). Estimates of p' assume spherical
cells and a,:m(435) of 2 x 10-5 as approximated by MOREL and BRtCAVO(1981)
Size fractionation data of ALPINE and CLOEaN (1985)
Nominal
Assumed
size fraction
diameter
Estimated
(Ima)
(lxm)
p'(435)
<5
5-22
>22
4
14
30
1
3
6
Relative fluorescence per Chl a
Model prediction
Observed
6.0
2.0
1.0
4.4
2.0
1.0
Species studied by LOFTUSet al. (1972)
Unidentified
nanoplankter
Monochrysis sp.
Dunaliella
tertiolecta
Fragillaria sp.
Phaeodactylum
tricornutum
Gymnodynium
splendens
Gymnodynium
nelsoni
2
0.4
15.0
11.0
6
6
1.2
1.2
10.0
10.0
5.0
4.0
10
10
2.0
2.0
7.2
7.2
5.0
4.7
35
7.5
2.0
2.3
48
9.5
1.0
1.0
photoadaptation. Whereas the F- and A-ratios exhibit increases for low-light-adapted
cells, the analogous fluorescence efficiency ratio (E-ratio) does not show a similar pattern
(Fig. 10, Table 2). For the three species studied, the A- and F- ratios increase between
the highest and lowest light levels. The E-ratio decreases by 25% for P. lutheri and by
10% for C. gracilis, while no pattern was observed for D. tertiolecta. The opposing
trends between the E-ratio and both the A- and F-ratios may be attributed to the fraction
of light absorbed which is coupled to the fluorescing mechanism of photosystem II (PSII).
The observations for all ratios may be explained by a simple hypothesis where changes in
the A- and F-ratios are determined primarily by the packaging effect and spectral
augmentation of the relative magnitude at longer wavelengths by accessory pigments.
These two effects will increase these ratios for low-light-adapted cells. The E-ratio is
proposed to be determined primarily by the ratio of absorption by pigments coupled to
PSII to pigments which are either coupled to PSI or are not coupled to the photosynthetic light harvesting system. Various carotenoid pigments have been proposed as photoprotective pigments which absorb light energy but do not transfer the energy to the
photosystems (KaiNS~:',', 1968; MaTrmWS-ROa'Hand K~sKY, 1970). These pigments have
absorption maxima below 500 nm (1L~iNowrrcri and GODVn~JEE, 1969).
If changes in the A- and F-ratios are both due primarily to the package effect, and the
pigment complement remains constant while intracellular concentrations vary, then the
changes will cancel each other in the calculation of the fluorescence efficiency, resulting
in a flat efficiency spectrum. Two hypotheses, which may act simultaneously, are
proposed which would be consistent with a drop in the E-ratio at lower growth
irradiance. First, if the ratio of photosynthetic to non-photosynthetic or protective
pigments increased at lower irradiance, then one would predict that the E-ratio would
decrease, since absorption by photoprotective pigments occurs below 500 nm. Thus as
658
B . G . MITCHELL and D. A. KIEFER
the fraction of light absorbed by non-photosynthetic pigments decreases at lower growth
irradiance, one expects an increased probability that quanta absorbed in the blue will be
coupled to the fluorescence system. Since the ratios reflect blue-green and green relative
to the blue, the E-ratio would drop. Second, if low-light adaptation is achieved primarily
by augmentation of PSII, then the fraction of blue light absorbed by Chl a in PSII and
coupled to the fluorescing mechanism will increase. In either case, the increased
probability of quanta absorbed being coupled to fluorescence will be spectrally biased,
and inversely proportional to the absorption spectra of either non-photosynthetic
pigments in the case of the first hypothesis, or Chl a in the second.
These two explanations are not mutually exclusive, and may occur simultaneously, or
in stages. For example, photoprotective pigments might be lost in a transition from high
to moderate irradiance, while significant augmentation of photosynthetic light harvesting
pigments might not occur until lower growth irradiances are attained. The data of the
current study are not sufficient to resolve which alternative hypothesis is valid for the
species studied, since measurement of PSI and PSII in addition to a more detailed
pigment study coupled with a complete optical analysis would be required.
DISCUSSION
A critical question regarding these techniques is whether this empirical approach to
correction of complex optical phenomenon can truly be effective in providing quantitative results. This is of particular concern for measurement of marine particulate
absorption coefficients since no simple alternative is currently available for such measurements in field studies. Furthermore, the original technique of KIEFER and SOOHOO(1982)
has come to be applied to field studies recently (LEWISet al., 1985; CARDERand STEWARD,
1985; KIsrnNo et al., 1985, 1986), often without consideration for the non-constancy of 13
as discussed by MITCHELLand IrdEFER (1984), or for the significant effects upon 13that
result from different filter types as discussed here. The empirical corrections for
measurements made on filters have been developed by direct comparison with measurements using the SHIBATA(1958) technique on concentrated suspensions. A comparative
survey of reported literature values for a~h(~,)and the data presented here for cultures
are compiled in Table 4. Unfortunately, there is no extensive literature on a~h(k), since
careful consideration of measurement geometry and technique is required. Nevertheless,
the data of MOREL and BRICAUD(1981), BR!CAUDet al. (1983) and DUBINSKVet al. (1986)
provide the most detailed analyses using traditional techniques for comparison with the
results presented here. Cell sizes, intracellular Chl a and estimates of Qa were also
tabulated, where possible.
Of obvious significance is the fact that a~h(~.) in the Soret peak (~435-440 nm) varies
by at least a factor of 4 for cultures. Thus, it is not justified, as is common practice, to
assume a~h(~.) to be constant for such applications as calculation of photosynthetic
quantum yields (BANNISTER, 1974, 1979; TYLER, 1975; DUalNSKV and BERMAN, 1976;
MOREL, 1978). The sources of variability in a~h(L) are many, in theory. KIEFER et al.
(1979) have reported that senescence of batch cultures, or nutrient limitation in
continuous culture, are important physiological considerations. Starting with the optical
theory of absorption by particles in non-absorbing media, MOREL and BRICAUD (1981)
proposed that the effect of "discreteness" explains why the Beer-Lambert law is not
appropriate for absorption by suspensions of unicellular algae. They studied a variety of
659
Chlorophyll a specific absorption and fluorescence
Table 4. Summary of chlorophyll specific and cellular absorption for various species in unialgal culture from
data reported in the literature, and from the present report
Citation and species
Ein
m - 2 day -1
a'h(435--440)
m~ mg-t Chl a
Cellular
diameter
m x 10~
pg Chl a
cell -1
Intraeellular
mg Chl a m -3
x 106
Qa(435)
BANNISTER(1979)
Chlorella pyrenoidosa
33
0.031
.
.
.
28
28
28
28
0.03
0.04
0.045
0.085
13.5
8.5
6.7
3.4
53
6
52
2.8
52
2.8
0.062
0.028
0.025
0.014
0.048
0.03
27.0
27.0
12.0
12.0
6.0
6.0
2.8
6.6
5.1
13.7
0.12
0.25
17
-
0.04--0.10
0.04-0.07
-
0.78
.
BRICAUDet al. (1983)
Hymenomonas elongata
Tetraselmis maculata
Platymonas sp.
Coccolithus huxleyi
DUBINSKYet al. (1986)
Prorocentrum micans
Prorocentrum micans
Thalassiosira weisflogii
Thalassiosira weisflogii
lsochrysis galbana
lsochrysis galbana
KIEFER et al. (1979)
Thalassiosira pseudonana
Pavlova lutheri
KIRK (1975)
Chlorella sp.
MITCHELLet al. (1984)
Pavlova lutheri
3.8
0.52
0.27
0.045
2.94
1.62
1.87
1.14
0.79
0.36
0.37
0.22
0.27
0.64
5.6
15.1
1.1
2.2
0.3
0.32
1.1
1.7
0.21
0.26
-
0.18
0.18--0.4
2.9
0.54
-
0.035
8.0
35
0.04
.
35
35
35
0.03
0.09
0.045
6.0
3.8
26.5
0.54
0.031
1.05
4.5
1.1
0.11
0.65
0.25
0.10
55
14
5
3
0.03
0.025
0.023
0.017
6.0
6.0
6.0
6.0
0.14
0.35
0.39
0.7
1.25
3.15
3.5
6.25
0.15
0.31
0.32
0.64
21
3.5
26
9
3
0.023
0.018
0.027
0.021
0.019
5.2
5.7
.
.
.
0.63
1.7
.
.
.
.
.
.
8.5
17.5
.
.
.
50
19
7
2
0.054
0.065
0.032
0.021
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
MOREL and BRICAUD(1981)
Platymonas suecica
Coccolithus huxleyi
Chaetoceros protuberans
Present report
Pavlova lutheri
Pavlova lutheri
Pavlova lutheri
Pavlova lutheri
Thalassiosira pseudonana
Thalassiosira pseudonana
Chaetoceros gracilis
Chaetoceros gracilis
Chaetoceros gracilis
Dunaliella tertiolecta
Dunaliella tertiolecta
Dunaliella tertiolecta
Dunaliella tertiolecta
0.67
1.2
s p e c i e s g r o w n at r e l a t i v e l y h i g h l i g h t l e v e l s a n d c o n c l u d e d t h a t i n t e r - s p e c i f i c v a r i a b l e s
i n c l u d i n g c e l l size a n d c e l l u l a r p i g m e n t a t i o n c a u s e s i g n i f i c a n t c h a n g e s in a~h(~,). T h e d a t a
o f D u a i ~ s K v et al. (1986) a n d this r e p o r t i l l u s t r a t e t h a t g r o w t h i r r a d i a n c e f o r n u t r i e n t s a t u r a t e d cells c a u s e d s i g n i f i c a n t c h a n g e s in a~h(E) w h i c h a r e m a n i f e s t a t i o n s o f d i s c r e t e n e s s c a u s e d b y v a r i a t i o n s in i n t r a c e l l u l a r p i g m e n t a t i o n .
T h i s p h e n o m e n o n is g e n e r a l s i n c e it is w e l l d o c u m e n t e d t h a t l i g h t l i m i t a t i o n r e s u l t s in
i n c r e a s e d c e l l u l a r p i g m e n t a t i o n ( L A w s a n d BANNISTER, 1980; FALKOWSra et al., 1985).
660
B . G . MrrcrmLL and D. A. KIEFER
The consequences for phytoplankton ecology, particularly in deep chlorophyll maxima,
may be significant. Since much evidence now indicates that the deep chlorophyll maxima
in the oceans results from increased cellular pigmentation rather than increased cell
concentrations (Kmv'ER et al., 1976; BEERS et al., 1982), one can expect that the actual
light harvesting capability of the cells in these maxima is not directly proportional to the
observed increases in the Chl a concentrations of vertical profiles. The data presented
here indicate that fully shade-adapted cells may have chlorophyll specific absorption
coefficients 4 times lower than their high-light-adapted counterparts. This phenomenon
will be of less significance for small cells, including picoplankton. There may be an
evolutionary selective advantage for small phytoplankton in the persistent deep chlorophyll maxima of open oceans.
The results presented for a~,h based on quantitative corrections for analysis on filters
fall in the range of values obtained by the SHmATA(1958) technique. This validation for
cultures justifies the application of these corrections to particle absorption in the field.
Although the non-constancy of]3 with sample density and filter type poses analytical
problems, the general availability of digital computers makes the additional computational needs relatively insignificant. The techniques, properly applied, still offer the easiest
and most accurate methods for direct determination of the particulate absorption
coefficient ap(~L) for dilute suspensions of marine particles. One must be aware, however,
that aph(g) will contribute a variable fraction to total particulate absorption, since
significant detrital absorption is present in field samples (KIEFER and SOOHOO, 1982;
MITCHELL, 1987; MITCHELLand KIEFER, 1988).
The increasing application of in vivo fluorescence in biological oceanography requires
a continuing effort to interpret the results. KIEFER (1973a,b) has discussed a variety of
problems related to the direct extrapolation of in vivo fluorescence to estimates of Chl a
concentrations. L o ~ u s et al. (1972) and ALPINE and CLOERN (1985) have presented
evidence that cell size may influence F* in vivo. The strong dependence of F* on
intracellular reabsorption by Chl a concentrations which is demonstrated here must be
considered as well. The extent of intracellular reabsorption is dependent on cell size
which determines the pathlength for escape of fluoresced quanta and intracellular Chl a
concentration. Use of Chl a flurorescence to study phytoplankton dynamics requires that
the data be interpreted in the context of the physical as well as the physiological factors
which control and modify the signals
Acknowledgements--We wish to express our gratitude to D. K. Clark of the National Oceanic and Atmospheric Administration and M. Blizard of the Office of Naval Research who sponsored this work.
REFERENCES
ALPINEA. E. and E. CLOERN(1985) Differences in in vivo fluorescence yield between three phytoplankton size
classes. Journal of Plankton Research, 7, 381-391.
BANNISTERT. T. (1974) Production equations in terms of chlorophyll concentration, quantum yield, and upper
limit to production.Linmology and Oceanography, 19, 1-12.
BANNISTERT. T. (1979) Quantitative description of steady state, nutrient-saturated algal growth including
adaptation. Linmology and Oceanography, 24, 76-96.
BANNISTERT. T. and A. D. WEIDEMANN(1984) The maximum quantum yield of phytoplankton photosynthesis
in situ. Journal of Plankton Research, 6, 275-294.
BEERSJ. R., F. M. H. REID and G. L. STEWART(1982) Seasonal abundance of the microplankton population in
the North Pacific central gyre. Deep-Sea Research, 29, 227-245.
Chlorophyll a specific absorption and fluorescence
661
BRICAUDA., A. MOREL and L. PRIEUR (1983) Optical efficiency factors of some phytoplankters. Limnology
and Oceanography, 28, 816--832.
BUTLER W. L. (1962) Absorbtion of light by turbid materials. Journal of the Optical Society of America, 52,
292-299.
BUTLERW. L. (1978) Energy distribution in the photochemical apparatus of photosynthesis. Annual Review of
Plant Physiology, 29, 345-378.
CARDERK. L. and R. G. STEWARD(1985) A remote sensing reflectance model of the red-tide dinoflagellate off
West Florida. Limnology and Oceanography, 30, 286-298.
COLLINS O. J., O. A. KIEFER, J. B. SOOHOO and I. S. MCDERMID (1985) The role of reabsorption in the
spectral distribution of phytoplankton fluorescence emission. Deep-Sea Research, 32, 983-1003.
COLLINS D. J., D. A. KIEFER, J. B. SOOHOO, C. STALLINGSand W. YANO (1986) A model for the use of
satellite remote sensing for the measurement of primary production in the ocean. In: Ocean Optics, Vol.
VIII, M. BLIZARD,editor, SPIE, Bellingham, Washington, pp. 1-14.
DEMAS J. N. and G. A. CROSBY(1971) The measurement of photoluminescence quantum yields. A review.
Journal of Physical Chemistry, 75, 991-1024.
DUBINSKVZ. and T. BERMAN(1976) Light utilization efficiencies of phytoplankton in Lake Kinneret (Sea of
Galilee). Limnology and Oceanography, 21,226-230.
DtmINSKY Z., P. G. FALKOWSrd and K. WYMAS (1986) Light harvesting and utilization by phytoplankton.
Plant and Cell Physiology, 27, 1335-1349.
DUNSTANW. M. (1973) A comparison of the photosynthesis-light intensity relationship in phylogenetically
different marine microalgae. Journal of Experimental Marine Biology and Ecology, 13, 181-187.
Dtn'rrLEY S. Q. (1942) The optical properties of diffusing materials. Journal of the Optical Society of America,
32, 61-70.
DUYSEr~S L. N. (1956) The flattening of the absorption spectrum of suspensions, as compared to that of
solutions. Biochimica et Biophysica Acta, 19, 1-12.
FALKOWSKIP. G., Z. DUBINSKYand K. WYMAN (1985) Growth-irradiance relationships in phytoplankton.
Limnology and Oceanography, 30, 311-321.
FAUST M. A. and K. H. NORRIS (1982) Rapid in vivo spectrophotometric analysis of chlorophyll pigments in
intact phytoplankton cultures. British Phycological Journal, 17, 351-361.
FORSTER T. (1951) Fluoreszenze organischer Verbindungen. Vandenhoek and Ruprecht, Gtttingen, 273 pp.
GALLAGHERJ. C., A. M. WOOD and R. S. ALBERTE (1984) Ecotypic differentiation in the marine diatom
Skeletonema costatum: influence of light intensity on the photosynthetic apparatus. Marine Biology, 82,
121-134.
GUILLARDR. R. L. (1975) Culture of phytoplankton for feeding marine invertebrates. In: Culture of marine
invertebrate animals. W. L. SMITHand M. H. CHANELY,editors, Plenum Press, New York, pp. 29-60.
HOLM-HANSENO., C. J. LORENZEN,R. W. HOLMESand J. D. H. STRICKLAND(1965) Fluorimetric determination of chlorophyll. Journal du Conseil. Conseil International pour l'Exploration de la Met, 30, 3-15.
ITURRIAGAR. and B. G. MITCHELL(1986) Chroococcoid cyanobacteria: a significant component in the food
web dynamics of the open ocean. Marine Ecology, Progress Series, 28, 291-297.
KIEFER D. A. (1973a) Fluorescence properties of natural phytoplankton populations. Marine Biology, 22,
263--269.
KIEFER D. A. (1973b) Chlorophyll a fluorescence in marine centric diatoms: Responses of chloroplasts to light
and nutrient stress. Marine Biology, 23, 39-46.
KIEFER D. A. and J. B. SOOHOO (1982) Spectral absorption by marine particles of coastal waters of Baja
California. Limnology and Oceanography, 27, 492-499.
KIEFERD. A. and B. G. MITCHELL(1983) A simple, steady state description of phytoplankton growth based on
absorption cross section and quantum efficiency. Limnology and Oceanography, 28, 770-776.
KIEFER D. A., R. J. OLSONand O. HOLM-HANSEN(1976) Another look at the nitrite and chlorophyll maxima
in the central North Pacific. Deep-Sea Research, 23, 1199-1208.
KIEFERD. A., R. J. OLSOr~and W. H. WILSON(1979) Reflectance spectroscopy of marine phytoplankton. Part
1. Optical properties as related to age and growth rate. Limnology and Oceanography, 24, 664--672.
KIRK J. T. O. (1980) Spectral properties of natural waters: Contribution of the soluble and particulate fractions
to light absorbtion in some inland waters of southeastern Australia. Australian Journal of Marine and
Freshwater Research, 31,287-296.
KISHINO M., N. TAKAHASHI, N. OKAMI and S. ICmMURA (1985) Estimation of the spectral absorption
coefficients of phytoplankton in the sea. Bulletin of Marine Science, 37, 634--642.
KISHINOM., N. OKAMI, N. TAKAHASrtl,and S. ICI-RMURA(1986) Light utilization efficiency and quantum yield
of phytoplankton in a thermally stratified sea. Limnology and Oceanography, 31,557-566.
KRINSKYN. I. (1968) The protective function of carotenoid pigments. In: Photophysiology, Vol. III, A. C.
GIESE, editor, Academic Press, New York, pp. 123-195.
LATIMER P. (1983) The deconvolution of absorption spectra of green plant materials--Improved corrections
for the sieve effect. Photochemistry and Photobiology, 38, 731-734.
662
B . G . MITCHELLand D. A. KIEFER
LAWS E. A. and T. T. BANNISTER(1980) Nutrient and light-limited growth of Thalassiosira fluviatilis in
continuous culture, with implications for phytoplankton growth in the ocean. Limnology and Oceanography, 25, 457-473.
LEWISM. R., R. E. WARNOCKand T. PLATr (1985) Absorption and photosynthetic action spectra for natural
phytoplankton populations: Implications for production in the open ocean. Limnology and Oceanogra-
phy, 30, 794-806.
LI W. K. W., D. V. StraaA RAO, W. G. HARRISOS,J. J. ~ S ,
B. IRWINand T. PLATr (1983) Autotrophic
picoplankton in the tropical ocean. Science, 219, 292-295.
LIPSL'TrF. R. (1967) The quantum efficiency of luminescence. In: Progress in Dielectrics, Vol. 7, J. B. BXRKS,
editor, CRC Press, Cleveland, pp. 217-319.
L o ~ u s M. E., D. V. StmaA RAO and H. H. SEUGER (1972) Growth and dissipation of pliytoplankton in
Chesapeake Bay. I. Response to a large pulse of rainfall. Chesapeake Science, 13, 282-299.
LORENZENC. J. (1966) A method for the continuous measurement of/n vivo chlorophyll concentration. DeepSea Research, 13, 233-227.
MATHEWS-ROTHW. H. and N. I. KSJNSKY(1970) Studies on the protective function of the carotenoid pigments
of Sarcina lutea. Photochemistry and Photobiology, 11, 419-428.
MELHUISH W. H. (1961) Quantum efficiencies of fluorescence of organic substances: Effect of solvent and
concentration of the fluorescent solute. Journal of Physical Chemistry, 65, 229--235.
MELHtaSH W. H. (1962) Calibration of spectrofluorometers for measuring corrected emission spectra. Journal
of the American Optical Society, 52, 1256-1258.
MITCHELLB. G. (1987) Ecological implications of variability in marine particulate absorption and fluorescence
excitation spectra. Ph.D. dissertation. The University of Southern California, 196 pp.
MITCHELLB. G. and D. A. KIEFER(1984) Determination of absorption and fluorescence excitation spectra for
phytoplankton. In: Marine phytoplankton and productivity, O. HOLM-HANSEN,L. BOLISand R. GILLES,
editors, Springer-Verlag, Berlin, New York, pp. 157-169.
MITCHELL B. G. and D. A. KIEFER (1988) Variability in pigment specific particulate fluorescence and
absorption spectra in the northeastern Pacific Ocean. Deep-Sea Research, 35, 665--689.
MORELA. (1978) Available, usable, and stored radiant energy in relation to marine photosynthesis. Deep-Sea
Research, 25, 673-688.
MORELA. and A. BRICAUD(1981) Theoretical results concerning light absorption in a discrete medium, and
application to the specific absorption of phytoplankton. Deep-Sea Research, 28, 1375-1393.
NEORI A., O. HOLM-HANSEN, B. G. MITCHELL and D. A. tQEWR (1984) Photoadaptation in marine
phytoplankton. Changes in spectral absorption and excitation of chlorophyll a fluorescence. Plant
Physiology, 76, 518-524.
NEORI A., M. VERNET, O. HOLM-HANSENand F. T. HAXO (1986) Relationship between action spectra for
chlorophyll a fluorescence and photosynthetic 02 evolution in algae. Journal of Plankton Research, 8,
537-548.
PARKER C. A. and W. T. REES (1962) Fluorescence spectrometry. Analyst, 87, 83-111.
PARKERR. R. and D. J. VAUDRE¥(1980) A proposed reference standard for in vivo chlorophyll a fluorometry.
CSIRO Division of Fisheries and Oceanography, Report number 125.
PLA'rr T. and A. D. JASSBY(1976) The relationship between photosynthesis and light for natural assemblages
of coastal marine phytoplankton. Journal of Phycology, 12, 421-430.
PLATr T., D. V. StmnA RAO and B. IRW~ (1983) Photosynthesis of picoplankton in the oligotrophic ocean.
Nature, 28, 702-704.
RAnI~OWITCrIE. and GOXrtl~JEE (1969) Photosynthesis. John Wiley, New York, 273 pp.
RYrI~R J. H. (1956) Photosynthesis in the ocean as a function of light intensity. Limnology and Oceanography, 1, 61-70.
SHIBATAK. (1958) Spectrophotometry of intact biological materials. Absolute and relative measurements of
their transmission, reflection and absorption spectra. Journal of Biochemistry, 45, 599--623.
SHIBATAK., A. A. BENSONand M. CALVIN(1954) The absorption of suspensions of living micro-organisms.
Biochimica et Biophysica Acta, 15, 461-470.
SOOHOO J. B., D. A. KIEFER, D. J. COLLn~Sand I. S. MCDERMID(1986) In vivo fluorescence excitation and
absorption spectra of marine phytoplankton: I. Taxonomic characteristics and responses to photoadaptation. Journal of Plankton Research, 8, 197-214.
S T E E ~ NmLSE~ E. (1952) The use of radio-active carbon for measuring organic production in the sea.
Journal du Consiel. Conseil International pour l'Exploration de la Met, 18, 117-140.
TYLER J. E. (1975) The in situ quantum efficiency of natural phytoplankton populations. Limnology and
Oceanography, 20, 976-980.
WEBER G. and F. W. TEALE (1957) Determination of the absolute quantum yield of fluorescent solutions.
Transactions of the Faraday Society, 53, 646--655.
WEIDEMA~ D. and T. T. BANNISTER(1986) Absorption and scattering coefficients in Irondequoit Bay.
Limnology and Oceanography, 31,567-583.
Chlorophyll a specific absorption and fluorescence
663
WINEFORDNERJ. D., S. G. S ~
and T. C. O'HAWR (1972) Luminescence spectrometry in analytical
chemistry. Wiley Interscience, New York, 354 pp.
YEm'SCHC. S. (1957) A non-extractive method for the quantitative estimation of chlorophyll in algal cultures.
Nature, 179, 1302-1304.
YE~'SfH C. S. (1962) Measurement of visible light absorption by particulate matter in the ocean. Limnology
and Oceanography, 7, 207-217.
YE~CrS~IC. S. and D. W. MENZEI~(1963) A method for the determination of phytoplankton chlorophyll and
phaeophytin by fluorescence. Deep-Sea Research, 10, 221-231.
YE~rSCI-I C. S. and C. M. YE~Crs~ (1979) Fluorescence spectral signatures: The characterization of
phytoplankton populations by the use of excitation and emission spectra. Journal of Marine Research, 37,
471-483.