Journal of Experimental Botany, Vol. 59, No. 2, pp. 349–359, 2008
doi:10.1093/jxb/erm316 Advance Access publication 5 February, 2008
This paper is available online free of all access charges (see http://jxb.oxfordjournals.org/open_access.html for further details)
RESEARCH PAPER
Effect of anthocyanins, carotenoids, and flavonols on
chlorophyll fluorescence excitation spectra in apple fruit:
signature analysis, assessment, modelling, and relevance
to photoprotection
Mark N. Merzlyak1,*, Thor Bernt Melø2 and K. Razi Naqvi2
1
Department of Physiology of Microorganisms, Faculty of Biology, Moscow State University, 119991,
GSP-2 Moscow, Russia
2
Department of Physics, Norwegian University of Science and Technology (NTNU), Trondheim, N-7491, Norway
Received 2 October 2007; Revised 7 November 2007; Accepted 20 November 2007
Abstract
Whole apple fruit (Malus domestica Borkh.) widely
differing in pigment content and composition has been
examined by recording its chlorophyll fluorescence
excitation and diffuse reflection spectra in the visible
and near UV regions. Spectral bands sensitive to the
pigment concentration have been identified, and linear
models for non-destructive assessment of anthocyanins, carotenoids, and flavonols via chlorophyll fluorescence measurements are put forward. The adaptation
of apple fruit to high light stress involves accumulation of these protective pigments, which absorb solar
radiation in broad spectral ranges extending from UV to
the green and, in anthocyanin-containing cultivars, to
the red regions of the spectrum. In ripening apples
the protective effect in the blue region could be
attributed to extrathylakoid carotenoids. A simple
model, which allows the simulation of chlorophyll
fluorescence excitation spectra in the visible range
and a quantitative evaluation of competitive absorption by anthocyanins, carotenoids, and flavonols, is
described. Evidence is presented to support the view
that anthocyanins, carotenoids, and flavonols play, in
fruit with low-to-moderate pigment content, the role of
internal traps (insofar as they compete with chlorophylls for the absorption of incident light in specific
spectral bands), affecting thereby the shape of the
chlorophyll fluorescence excitation spectrum.
Key words: Anthocyanins, apple fruit, carotenoids, chlorophyll,
flavonols, fluorescence, photoprotection.
Introduction
As an important defence mechanism against the deleterious effects of solar radiation, long-term adaptation of
higher plants involves synthesis of relatively stable compounds capable of serving as light screens and/or internal
traps (Day et al., 1993; Bilger et al., 1997, 2001; Smith
and Markham, 1998; Cockell and Knowland, 1999; Barnes
et al., 2000; Merzlyak and Chivkunova, 2000; Cerovic
et al., 2002; Steyn et al., 2002; Merzlyak and
Solovchenko, 2002; Pfündel et al., 2006). The build-up of
such substances in specific cell and tissue structures
reduces the fraction of radiation absorbed by potent
photosensitizers, and thereby diminishes light-induced
damage. It is helpful to distinguish between screening and
competitive absorption. The term screening will be applied
to systems in which the photosynthetic pigments can only
absorb the light that emerges from a layer or compartment
containing the protective pigments (Barnes et al., 2000;
Cerovic et al., 2002); in other cases, where the protecting
pigments must compete with photosynthetically active
pigments for light absorption, the term ‘internal trapping’
is used. Several compounds absorbing in different spectral
regions could provide photoprotection; vacuolar flavonoids
(Bilger et al., 1997; Cockell and Knowland, 1999; Barnes
et al., 2000; Cerovic et al., 2002; Pfündel et al., 2006) and
* To whom correspondence should be addressed. E-mail: [email protected]
Abbreviations: AnC, anthocyanins; Car, carotenoids; CFE, chlorophyll fluorescence excitation; Chl, chlorophyll; Flv, flavonols; Pgm, pigment.
ª 2008 The Author(s).
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which
permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
350 Merzlyak et al.
anthocyanins (AnC) (Merzlyak and Chivkunova, 2000;
Steyn et al., 2002; Pfündel et al., 2006), absorbing mainly
in the near UV and green regions of the spectrum,
respectively, have been considered in the literature. The internal trapping role was also suggested for extrathylakoid
carotenoid (Car) pigments in chloroplasts of stressed/
senescing leaves and ripening fruit (Merzlyak and
Solovchenko, 2002; Han et al., 2003; Merzlyak et al.,
2005a).
Together with other approaches (Day et al., 1993;
Cockell and Knowland, 1999; Barnes et al., 2000;
Merzlyak and Chivkunova, 2000; Bilger et al., 2001;
Solovchenko and Merzlyak, 2003; Solovchenko and
Schmitz-Eiberger, 2003; Pfündel et al., 2006) the screening effect in green plant tissues was evaluated through the
analysis of the excitation spectrum of chlorophyll (Chl)
fluorescence (Bilger et al., 1997, 2001; Barnes et al.,
2000; Cerovic et al., 2002; Barthod et al., 2007), which,
at room temperature, is emitted from chlorophyll a (Chla)
(Buschmann and Lichtenthaler, 1988; Lazár, 1999; Roháček,
2002). In leaves the ratio of Chl fluorescence excited by
UV-B to that excited by blue-green light showed
a negative correlation with the concentration of wholeleaf UV-B-absorbing pigments, and a positive correlation
with the transmittance of isolated epidermal tissue, where
flavonoids accumulate (Barnes et al., 2000). Cerovic
et al., (2002) reported that UV-excitation spectra of Chl
fluorescence from the adaxial and abaxial sides of bifacial
leaves allowed identification of UV-absorbing screening
pigments in the leaf epidermis. They found that, due to the
strong absorption by Chl at wavelengths shorter than 450
nm, differences in leaf optical properties exert only
a minor influence on the shape of UV-excitation spectra,
and ‘Chl behaves as a photon counter’. In the commercial
fruit industry, studies of the basic properties of lightscreening substances and Chl fluorescence have been applied
recently for non-destructive pigment assessment; methods
allowing sensitive and rapid determination of flavonoids
(including AnC) in the skin of apples (Hagen et al., 2006),
marketable broccoli heads (Bengtsson et al., 2006), olive
fruit (Agati et al., 2005), and grape berries (Agati et al.,
2007) have been developed. In these studies, screening by
flavonoids was quantified, on the leaf/fruit level, by using
a Chl fluorescence excitation (CFE) ‘ratio’, for example, the
ratio of the Chl fluorescence yields for different excitation
wavelengths. Further progress can be made only by relating
a CFE ‘spectrum’ to specific spectral features of Chl and
individual light-screening and/or internally-trapping pigments
in the specimen under examination.
Apples possess a fully functional photosynthetic apparatus (Blanke and Lenz, 1989) containing thylakoid-bound
Chl and Car as well as vacuolar flavonoids as principal
pigments absorbing in the visible range and near UV
(Merzlyak et al., 2003, 2005b; Bengtsson et al., 2006;
Merzlyak, 2006). As a result of development under strong
solar radiation, apples accumulate phenolics with absorption maxima in the near UV, including flavonols (Flv) and
other flavonoids (Solovchenko and Merzlyak, 2003;
Solovchenko and Schmitz-Eiberger, 2003; Merzlyak
et al., 2005b; Hagen et al., 2006; Merzlyak, 2006). Fruit
of some apple cultivars under high-light conditions are
able to accumulate AnC pigments, which absorb strongly
in the green region of the spectrum and thus impart
red colour to the fruit (Saure, 1990; Merzlyak and
Chivkunova, 2000; Ma and Cheng, 2004). Ripening is
accompanied by a progressive decline in the Chl content and the build-up of xanthophylls (Merzlyak and
Solovchenko, 2002; Solovchenko et al., 2005). Due to its
relatively low pigment content, apple fruit represents
a simple natural system in which general plant ontogenetic
and/or stress-induced pigment dynamics could be followed non-destructively and quantitatively, using measurements of diffuse reflectance (see for review Merzlyak,
2006).
In this work we have quantified competitive absorption
by Chl, Flv, Car, and AnC by examining diffuse reflection
spectra and CFE spectra of apple fruit differing in pigment
content and composition, as a result of ripening and/or
adaptation to strong sunlight. It is shown that a good
approximation to CFE spectra in the visible range can be
achieved by assuming simple relations between fluorescence, absorbed radiation, and internal trapping. The
proposed model has enabled us to carry out a quantitative
evaluation of internal trapping by AnC, Car, and Flv in
apple fruit.
Materials and methods
Plant material
AnC-free Granny Smith and Golden Delicious, and AnC-accumulating
Summer Red apple (Malus domestica Borkh.) fruit were purchased
at local market of Trondheim in Autumn 2002 and Spring 2003.
The fruits were selected on the basis of their colour. Sunlit surfaces
of Granny Smith and Golden Delicious fruits were recognized by
their paler and more yellowish colouration and specific features in
the near UV part of the diffuse reflection spectrum, mainly due to
Flv accumulation (Merzlyak and Chivkunova, 2000; Merzlyak
et al., 2005b). Sunlit surfaces of Summer Red apples were pink-tored in colour as a result of AnC pigmentation.
Spectral measurements
Whole fruit diffuse reflection spectra (against barium sulphate as
a standard) were recorded by means of a 3010 Hitachi spectrophotometer equipped with a 150 mm diameter integrating sphere
attachment. The reflectance data were sampled at 1 nm intervals in
the 350–800 nm range.
Fluorescence excitation and emission spectra were measured
using Fluorolog 3 (Jobin Yvon-Spex). Segments (c. 15325 mm)
sliced from the equatorial part of an apple, were placed in a sample
holder made by removing two adjacent sides from a standard
(1 cm2) plastic cuvette, which was itself placed in the sample
compartment of the spectrofluorimeter. The excitation beam was at
an angle of approximately 45 to the surface of the sample, and
Effect of anthocyanins, carotenoids, and flavonols on chlorophyll fluorescence 351
a right-angle arrangement was used for observing the emitted light.
For all experiments, the band pass of the excitation and emission
monochromators were 3 nm and 4 nm, respectively.
To minimize the effect of chlorophyll fluorescence induction
kinetics (Buschmann and Lichtenthaler, 1988; Lazár, 1999; Roháček,
2002), the samples were pre-irradiated directly in the spectrofluorimeter for 550 s, using light of 430 nm for AnC-free apples (5.8 mW
cm2) or 680 nm for AnC-containing apples (2.5 mW cm2). In the
course of the pre-irradiation, the fluorescence intensity fell to nearly
a half of its peak value, and reached a steady level within 450 s.
Fluorescence spectra of the samples were acquired immediately
after the termination of the pre-irradiation period; in order
to prevent fluorescence recovery, an intense excitation beam was
used, which had the additional advantage of providing a large
signal-to-noise ratio. Under these conditions, two successive
scans yielded spectra that were identical (within the limit set by
noise) if the second scan was started within 1 min after the end of
the first.
Fluorescence excitation spectra were recorded between 350 nm
and 710 nm at a scan rate of 1 nm s1 and emission set at 715 nm.
To reduce the amount of exciting light scattered towards the
detection system, a long pass filter transmitting above 700 nm
(Oriel LP70) was placed before the emission monochromator;
correction for the wavelength-dependence of the intensity of the
excitation beam was made by using correction factors supplied by
the manufacturer. For recording fluorescence emission spectra, the
excitation wavelength was set at 440 nm. Since absolute measurements of fluorescence could not be performed with a standard
spectrofluorimeter (where all the emitted light is not collected), the
analysis was limited to a comparison of relative changes in the
shapes of the recorded spectra. To minimize the influence of
uncontrolled variations in the instrumental conditions (drift, possible variability of the fluorescence quantum yield), the spectra
analysed here were all normalized (at the red Chla maximum).
Pigment analysis
The pigment content in the peals of the apples examined here was
determined with the aid of recently developed algorithms, which
make use of values of Rw, the diffuse reflectance for incident light
of wavelength w nm. The following indexes were used for
quantification: R800[1/R678–1/R800] for Chl, R800[1/R520–1/R550] for
Car, and R800[1/R520–1/R700] for AnC [Merzlyak et al., 2003], and
R800[1/R410–1/R460] for Flv, whose content was expressed as that of
rutin (Merzlyak et al., 2005b). Square brackets are used to denote
the pigment concentration.
cal diffuse reflectance of apple fruit, [R(k)]1¼6(k), may be
visualized as the sum of two subspectra: that is, 6(k)¼6(S)(k)+
6(N)(k), where the selective part 6(S)(k) contains features attributable to the various pigments, and the featureless non-selective part
6(N)(k) arises from factors other than absorption (mainly scattering
within the fruit tissue). The simulated counterparts of 6(k), 6(S)(k),
~
~ ðSÞ ðkÞ, and 6
~ ðNÞ ðkÞ,
and 6(N)(k), will be denoted by 6ðkÞ,
6
ðSÞ
ðNÞ
~
~
respectively; our main task is to model 6 ðkÞ, and 6 ðkÞ so as
to make their sum
~
~ ðSÞ ðkÞ þ6
~ ðNÞ ðkÞ
6ðkÞ
¼6
ð2Þ
agree as closely as possible with 6(k).
The first term in equation (2) was expressed as the sum
~ ðSÞ ðkÞ ¼ + SI ðkÞ, where SI ðkÞ[aI PI ðkÞ are contributions by in6
I
dividual pigment pools with known fiducial spectra, PI(k), and aI
are constants. The following pigment pools present in apple fruit
were considered: (i) PT(k), chloroplast chlorophylls and Car
associated with thylakoid membranes; (ii) PX(k), extrathylakoid
Car (mainly xanthophylls), (iii) PP(k), cuticular and vacuolar
phenolics and Flv; and (iv) PA(k), vacuolar AnC. Accordingly,
~ ðSÞ ðkÞ ¼ ST ðkÞ þ SX ðkÞ þ SP ðkÞ þ SA ðkÞ
6
ð3Þ
is defined.
While PT(k), PX(k), and PP(k) have been determined previously
(Merzlyak, 2006), PA(k) was equated to the difference 6(S)(k)–
6(S,0)(k), where 6(S)(k) and 6(S,0)(k) now refer to Summer Red
fruit with a moderate amount of AnC and with no AnC,
respectively. A linear form a1k+a0, where a1 and a0 are constants,
was used for representing 6(N)(k) (Merzlyak, 2006).
Modelling chlorophyll fluorescence excitation spectra
In an optically clear system containing only a single fluorophore,
the fluorescence intensity is proportional to the quantity of absorbed
radiation, 1–T(k), where T(k) is the transmittance at wavelength k.
Quantitatively, the relation between I0(k), the intensity of the
collimated incident beam, and IF(k#;k), the intensity of fluorescence
emitted (in all directions) at wavelength k#, can be expressed as:
IF ðk#; kÞ ¼ cuI0 ðkÞ½1 TðkÞ
ð4Þ
ð1Þ
where c is an instrumental factor (that may depend on, among other
things, k and k#), and u is the fluorescence quantum yield; in most
one-component systems, u is independent of the wavelength of
excitation (Bursteyn, 1968; Parker, 1968). If c and u are constants
and T vanishingly small (total absorption), IF becomes proportional
to I0(k), and the sample can be used as a ‘relative quantum counter’
(Bowen, 1936).
In order to apply the foregoing reasoning to apple fruit, the
following assumptions were made: (i) only the PT(k) pigment pool
contributes to the fluorescence emission spectrum in the red
(Buschmann and Lichtenthaler, 1988; Gitelson et al., 1998; Lazár,
1999; Roháček, 2002; Ramos and Lagorio, 2006), and (ii) the
efficiency of energy transfer to Chla is high for Chlb and the
majority of Car which are in the thylakoids (Gradinaru et al., 2000;
Croce et al., 2001) and zero for those Car which lie outside. Let
C(k)¼6T(k)/6(S)(k) denote the fraction of light absorbed by the
thylakoid-bound pigments, and let ua denote the fluorescence
quantum yield of Chla; the measured intensity of fluorescence
L(k#;k) can then be expressed as
h
i
Lðk#; kÞ ¼ cua I0 ðkÞ 1 6 ðNÞ ðkÞ=6ðkÞ CðkÞ
ð5Þ
where R (k) is the diffuse reflectance of the specimen in the
absence of the pigments. On the basis of experience gained from
light-induced bleaching experiments (Merzlyak, 2006), the recipro-
When we refer to the case C¼1 (a fruit that is devoid of extrathylakoid pigments), an asterisk will be added, as shown below,
where the form taken by equation (5) for C¼1 is shown:
Modelling of pigment contribution into spectral reflection
The modelling of diffuse reflectance in the visible range was
performed in accordance with a theory developed by Atherton
(1955) for dealing with a dyed fabric; though similar to the betterknown Kubelka–Munk approach, his analysis is more pertinent for
spectral measurements, since it assumes the incident beam to be
collimated (as in a spectrophotometer). One of his principal results,
which forms the basis of our analysis, may be stated as follows: In
a specimen containing many pigments (in concentrations c1, c2, etc.,
and with absorption coefficients, e1(k), e2(k), etc., where k denotes
the wavelength of the incident light), the diffuse reflectance R(k)
satisfies the following relation (Atherton, 1955; Alderson et al.,
1961):
1=RðkÞ 1=Rð0Þ ðkÞ ¼ c1 e1 ðkÞ þ c2 e2 ðkÞ þ .
(0)
352 Merzlyak et al.
L ðk#; kÞ ¼ c ua I0 ðkÞ
6 T ðkÞ
6 T ðkÞ þ 6
ðNÞ
ð6Þ
ðkÞ
In accordance with equations (5) and (6), the light-entrapment
factor due to the extra-thylakoid pigment pools in apple fruit was
calculated as
Wðk#; kÞ[
L ðk#; kÞ
6ðkÞ
¼
Lðk#; kÞ
6 T ðkÞ þ 6 ðNÞ ðkÞ
ð7Þ
Fluorescence excitation spectra F(k#;k)¼L(k#:k)/I0(k) were modelled by using fruit with low-to-moderate pigment content, thereby
~
ensuring that the model, given 6ðkÞ
in equations (2) and (3), fitted
the reciprocal reflection spectrum with a relative error less than 10%
throughout the visible range. Since the same emission wavelength
(710 nm) is used for all excitation spectra shown here, the notation will
be simplified by suppressing k# and writing F(k) instead of F(k#;k);
furthermore, Fw [Fðk ¼ w nmÞ and FðkÞ[FðkÞ=F
678 will be set.
Results
Pigment content
As may be seen from Table 1, the specimens of apple fruit
examined here differed considerably in pigment content
and composition. Granny Smith and Golden Delicious
fruit, which contained only minor amounts of skin AnC
([AnC] <1 nmol cm2), were regarded as AnC-free. In
Golden Delicious apples, [Flv] was high in sunlit surfaces
(up to 100 nmol cm2), and much lower in the shaded
surfaces. Apples of both AnC-free cultivars displayed
a significant variation in the [Chl]/[Car] ratio. The Chl
Table 1. Pigment content (nmol cm2) in apple fruit
Chlorophylls Carotenoids Flavonoids
Flavonols Anthocyanins
Granny Smith
(n¼16)
Average
Standard
deviation
Minimum
Maximum
Median
Gold Delicious
(n¼16)
Average
Standard
deviation
Minimum
Maximum
Median
Summer Red
(n¼12)
Average
Standard
deviation
Minimum
Maximum
Median
a
2.32
2.0
2.15
0.71
20.1
15.7
–a
–
0.77
8.1
1.6
1.22
3.86
1.99
4.9
59.8
14.1
–
–
–
1.43
0.94
1.44
0.35
46.4
40.7
–
–
0.53
2.90
0.97
0.90
2.16
1.43
0.4
102.9
28.3
–
–
–
1.43
0.80
–
–
–
–
15.5
13.9
0.48
3.16
1.16
–
–
–
–
–
–
0.4
42.1
13.3
Not detemined.
content of mature specimens was higher than their Car
content, but ripening made Car the dominant pigment
(Solovchenko et al., 2005); a reduction in [Chl] was
accompanied by a remarkable increase in [Car]/[Chl] ratio
reaching 2.4–2.5 in fruit with [Chl]-<1 nmol cm2. Sunlit
surfaces of the fruit also exhibited, in comparison with the
shaded surfaces, a preferential increase in [Car] over [Chl]
(see also Merzlyak et al., 2002; Ma and Cheng, 2004;
Solovchenko et al., 2006) especially in Golden Delicious
apples (not shown). The [Chl]-[Car] correlation was
relatively high in Granny Smith apples (r2¼0.80), and much
lower in Golden Delicious fruits (r2¼0.49). No [Chl]-[Flv]
or [Car]-[Flv] correlation was found in AnC-free fruit; in
Summer Red apples, [Chl] and [AnC] were uncorrelated.
The Chl contents of the shaded and sunlit surfaces of
Summer Red apples did not differ significantly, and on
average equalled that in Golden Delicious apples. The AnC
content of Summer Red apples grown under the shade was
low ([AnC] <1.3 nmol cm2), but considerably higher in
sunlit surfaces, with [AnC] reaching 40 nmol cm2.
Anthocyanin-free fruit
Effect of chlorophyll content: Representative diffuse reflection and CFE spectra in the green-red region, where
only Chl contributes to absorption, are presented in Fig. 1
for six samples of AnC-free apple fruit differing in Chl
content. In this spectral range reflectance possessed
features belonging to Chla (minimum at 678 nm) and
Chlb (shoulder around 650 nm), and showed a monotonic
decrease with increasing Chl content, without noticeable
changes in the positions of the minima (Fig. 1A). The
CFE spectra were more resolved and showed qualitative
changes dependent on the Chl content. In fruit with low
[Chl], the fluorescence excitation maximum was located
around 678 nm and the fluorescence intensity near 650 nm
was relatively small. An increase in [Chl] brought about
pronounced changes in the shape, a considerable flattening of the spectra and a progressive shift of the maximum
to shorter wavelengths. Simultaneously, a pronounced
enhancement of the band 650 nm and intensification of
fluorescence emission in the green and far-red ranges were
observed. In a fruit with high [Chl] (8.06 nmol cm2) the
amplitude of FðkÞ
spectrum exceeded 0.5 in the green
region, and the maximum was situated at 653 nm (Fig. 1B).
The shapes of Chl fluorescence emission spectra of apple
fruit also showed a strong dependence on [Chl]. The fluorescence maxima in these spectra were located at 682 nm
and 685 nm, in fruits with low (c. 1 nmol cm2) and high
(c. 8.1 nmol cm2) Chl content, respectively. The latter
spectrum was much broader with an intense band 700–740
nm (not shown) suggesting a strong effect of fluorescence reabsorption (Gitelson et al., 1998; Ramos and Lagorio, 2006).
Sunlit versus shaded fruit: To display the differences
between the sunlit and shaded surfaces of AnC-free fruit,
Effect of anthocyanins, carotenoids, and flavonols on chlorophyll fluorescence 353
apples were selected with a Chl content low enough to
avoid distortion of the main Chl peak in the CFE spectra
(Fig. 1). Averaged spectra for non-reflected radiation,
QðkÞ[1 RðkÞ, which represents light absorption by (the
pigments) and scattering (by the tissue), are plotted in
Fig. 1. (A) Reflection spectra, R(k), and (B) normalized (at 678 nm)
chlorophyll fluorescence excitation spectra, FðkÞ,
of Golden Delicious
apples. The value of [Chl]/(nmol cm2) for each specimen is given
in parentheses following the identification label: 1 (0.58), 2 (1.23),
3 (1.92), 4 (2.43), 5 (2.87), 6 (8.06).
Fig. 2A for six shaded surfaces (curve 1) and six sunlit
surfaces (curve 2); the corresponding plots of normalized
CFE spectra appear in Fig. 2B. Compared with shaded
surfaces, sunlit fruit surfaces were characterized by larger
values of [Flv] and [Car], and lower values of Q in the far
red region. Typically for sunlit fruit surfaces (Merzlyak and
Chivkunova, 2000; Merzlyak et al., 2005b), reflectance
below 500 nm decreased and Q(k) increased, reaching
a high and nearly constant level in the near UV (Fig. 2A).
Compared with their reflection counterparts, the CFE
spectra of apple fruit (Fig. 2B) displayed narrower bands,
associated with Chla (main maxima at 678 nm and 440
nm), Chlb (shoulder at 650 nm, a band near 460 nm) and
Fig. 2. (A, B) Curves 1 and 2 show, respectively, the average spectrum
for non-reflected radiation, QðkÞ[1 RðkÞ, and the average normal
ized chlorophyll fluorescence excitation spectrum, FðkÞ,
of six AnCfree apples with low Chl content; error bars (6 standard error) are also
shown for each curve. Labels 1 and 2 identify shaded and sunlit
surfaces, respectively. Key for the corresponding pigment contents
(average 6SE, nmol cm2) (1): [Chl]¼0.8160.11, [Car]¼1.2760.31,
[Flv]¼10.965.4.
(2):
[Chl]¼0.8260.14,
[Car]¼1.5460.57,
[Flv]¼85.7620.3. The right vertical scale in (B) refers to curve 3,
2 ðkÞ, where the subscripts 1 and 2
1 ðkÞ=F
which is a plot of the ratio F
indicate shaded and sunlit surfaces, respectively.
354 Merzlyak et al.
Car (peaks around 465–470 nm and 485–490 nm). In
shaded fruit the ratio of Chla fluorescence maximum in
the Soret band to that in the red was, on average, c. 0.7,
approximately 2-fold higher than that in sunlit fruit
surfaces. In sunlit fruit surfaces below 520 nm, fluorescence intensity decreased considerably and two Chla
bands near 420 nm and 383 nm evident in the spectra of
shaded fruits were reduced and absent, respectively. In
these samples, fluorescence contained no spectral details
between 350 nm and 400 nm.
Signature analysis: Correlation analysis was applied for
investigating the influence of the principal pigments on
the absorption and fluorescence spectra of apple fruit in
the visible and near-UV range. In Fig. 3, which refers to
AnC-free fruit, the solid curves in each panel show the
spectra of correlation coefficient, r, of the linear relationship between the concentration of a pigment [Pgm], where
Pgm stands for Chl, Car or Flv, and a k-dependent
quantity that is Q(k) for the upper row and FðkÞ
for the
lower row; the dotted curves in Fig. 3C and D display the
FðkÞ-QðkÞ
correlation. High correlation between
QðkÞ-½Chl was observed in a narrow band around 700
nm and in a wider band, situated at 580660 nm for
Granny Smith and 610630 nm for Golden Delicious
apples. At shorter wavelengths the correlation for [Chl]
Fig. 3. Plots of the k-dependence of the correlation coefficient, r, for the Q-[Pgm] correlation (A, B), and the F-½Pgm
correlation (C, D), where Pgm
correlation.
is Chl (curve 1), or Car (curve 2), or Flv (cure 3). Curve 4 (broken line) shows the k-dependence of the correlation coefficient of the F-Q
(A, C) Granny Smith; (B, D) Golden Delicious apple fruit.
Effect of anthocyanins, carotenoids, and flavonols on chlorophyll fluorescence 355
decreased with a concomitant increase in the correlation
for [Car]. The maxima for [Car] were observed near 510–
515, 470, and 440 nm, whereas the minima were situated
around 480 nm and 450 nm; these spectral features were
more pronounced in Golden Delicious apples. The
QðkÞ-½Flv correlation, weak in the orange-red region,
increased significantly toward the violet, with peak
positions opposite to those for the other two pigments.
Between 360 nm and 420 nm the F-Q
correlation
considerably exceeded the F-½Flv
correlation, and showed
maxima near 360–370 nm and 400–410 nm for Granny
Smith and Golden Delicious apples, respectively. The
spectra of F-½Pgm
correlation contained fewer details
(Fig. 3C, D). For [Chl] the correlation was relatively high
near 700 nm and in the band 520–630 nm. In the green
region, higher correlation was recorded over a broader
spectral band for Granny Smith fruit (with a larger Chl
content) than for Golden Delicious. At shorter wave
lengths, the Q-[Chl] correlation as well as F-½Chl
correlation displayed a significant decline. Throughout the
whole spectral range the shape of the correlation spectra
for [Car] was similar to that for [Chl] but the correlation
was much weaker (especially, in Golden Delicious fruit).
The F-½Flv
correlation was low for k >450 nm, and the
correlation spectrum contained few spectral details. At
shorter wavelengths, the shape of the spectrum (for
Granny Smith apples) resembled that for [Chl] approaching –0.7 near 375 nm, whereas in Golden Delicious apples
the negative correlation in the broad band 380–420 nm
was higher and reached –0.88 near 420 nm.
In Granny Smith apples, the F-Q
correlation spectrum
contained the same features as the F-½Chl
correlation
spectrum. In this fruit, positive and relatively high
correlation was found near 700 nm and in the band 600
nm and 650 nm, whereas in Golden Delicious apples the
latter band was much narrower with the maximum near
630 nm. Below 530–550 nm in both cultivars, r decreased
sharply with prominent minima near 440 nm and 480 nm,
became negative at shorter wavelengths, and reached
0:7 in the 360–370 nm band.
Anthocyanin-containing fruit
The results for Summer Red apples, which are able to
accumulate AnC pigments on their sunlit surfaces, are
summarized in Fig. 4. The build-up of AnC brought about
a considerable content-dependent increase in Q(k) at
wavelengths shorter than 620 nm, along with the shift of
absorption band from 510–520 nm to c. 550 nm. At high
AnC content Q(k) exceeded 0.94 in the spectral range
350–575 nm, and Chl absorption in the red appeared on
a large ‘AnC background’ (Fig. 4A). With increasing AnC
content, FðkÞ
suffered a noticeable decrease; the effect
was particularly pronounced (relative fluorescence intensity of 4–5%) between 520 nm and 575 nm. As regards
the blue region, Chl and Car spectral features, easily
Fig. 4. Plots showing the effect of anthocyanins on the reflection and
CFE spectra of 12 Summer Red apples. (A) Q(k) and (B) FðkÞ
spectra.
The values of [Pgm]/(nmol cm2 for curves 1–4 are as follows: Chl:
0.5, 1.12, 1.07, 1.02; AnC: 0.7, 19.2, 32.0, 42.1. The plots in (C) show
spectra for the Q-[Pgm] correlation (1 and 2, left scale) and F-½Pgm
correlation (3 and 4, right scale); Pgm: Chl: (1 and 3, solid lines); AnC:
(2 and 4, broken lines).
discernible in apples with moderate AnC, were suppressed
in the sample with high [AnC] (Fig. 4B).
For the whole data set of Summer Red apples, the
Q-[Chl] correlation was relatively high (r >0.9) in 660–698
nm band (Fig. 4C); the Q-[Anc] correlation, very weak in
this region, attained its maximum near 600 nm, and
remained high down to 400 nm. In the entire spectral
range, the F-½Chl
correlation showed a weak correlation
356 Merzlyak et al.
with [Chl]; in contrast, the F-½AnC
correlation had a large
negative value, peaking near 580 nm, with additional bands
around 500, 464, and 430 nm. The F-Q
correlation
possessed weak positive bands around 645 nm and 700
nm and showed a negative correlation between 450 nm and
560 nm (r >0.9) (not shown) resembling that for the F-½Flv
correlation.
Spectral reconstruction of pigment absorption and
modelling chlorophyll fluorescence excitation spectrum
Spectral reconstruction of the reflection spectra revealed
that all samples contained a pool of extrathylakoid Car,
which is characteristic of ripening (Merzlyak, 2006).
Figure 5 demonstrates the reconstruction for fruit with the
following pigment content and composition: (A) approximately equal amounts of Car and Chl, (B) low Chl
content with elevated amounts of Flv and Car, and (C)
AnC-accumulating Summer Red apple. The modelling
provided a good simulation of CFE spectra in the spectral
range 600–700 nm (cf. curves 1 and 3 in Fig. 5D–F). In
general, in the blue-green region the modelling also
succeeded in reconstructing the shapes of the spectra, the
positions of the main maxima, and the ratios between the
ðkÞ spectra
peaks. In addition, in Fig. 5 D–F, the model F
calculated for fruit ‘devoid’ of the internal-trapping PX,
PP, and PA pigment pools (curves 2) and the spectral
dependence of the trapping factors (curves 4) are
presented.
Pigment assessment with chlorophyll fluorescence
measurements
The results of correlation analysis were employed for
assessment of Chl, Flv, and AnC content via Chl
fluorescence measurements with linear models. For the
fluorescence index in the form, [Chl]¼0.109+0.0463F700/
F678, the determination coefficient, r2, and the root mean
square error (RMSE) of Chl determination were 0.86 and
0.03 nmol cm2, respectively. For the analysis of AnC in
Summer Red apples, 580 nm was taken as the diagnostic
wavelength for AnC, since the F-½AnC
correlation
reaches its minimum here (Fig. 4, curve 4); for normalization, 700 nm was chosen, since this is the wavelength at
which the F-½Chl
correlation becomes large for AnC-free
fruit (Fig. 3C, D, curve 1) and low for Summer Red
apples (Fig. 4C, curve 3). With the fluorescence index in
the form [Anth]¼ –7.10+15.73F700/F580, r2 and RMSE
for AnC estimation turned out to be 0.88, and 5.0 nmol
cm2, respectively. For estimating [Flv], the diagnostic
wavelength was taken as 400 nm, and the normalizing
½Pig correlation
wavelength as 440 nm (where the F
was low for Chl, Car, and Flv). For AnC-free fruit with
[Flv] <100 nmol cm2, values of 0.89 nmol cm2 and 9.2
nmol cm2 were found for r2 and RMSE, respectively, by
using the fluorescence index [Flv]¼ –27.5+24.03F440/
Fig. 5. Spectral reconstruction of reciprocal reflection spectra (A–C)
and chlorophyll fluorescence excitation spectra (D–F) of apple fruit.
(A–C, right scale) Curve 1 (solid line) is 6(k) and curve 7 (symbols)
~
is the corresponding model 6ðkÞ,
curve 2 is PT(k), curve 3 is PX(k),
curve 4 is PP(k), curve 5 is PA(k), and curve 6 is 6 ðNÞ ðkÞ. (D–F, left
scale) Curve 1 (solid line) is FðkÞ,
and curve 2 (symbols) is the
ðkÞ, and curve 4 (right
corresponding model, curve 3 (symbols) is F
scale) is W(k). Pigment content (nmol cm2) in (A) and (D) (Granny
Smith): [Chl] 1.59, [Car] 1.64, [Flv] 11.5; (B) and (E) (Golden
Delicious): [Chl] 0.59, [Car] 1.45, [Flv] 102.9; (C) and (F) (Summer
Red): [Chl] 1.12, [AnC] 19.2.
F400. In this Flv content range with the power model the
determination coefficient of 0.91 was achieved. Attempts
to employ the fluorescence technique for the assessment
of total [Car] in the spectral band 450–500 nm were
unsuccessful.
Discussion
In line with previous investigations (Barnes et al., 2000;
Bilger et al., 2001), this study was undertaken to
consolidate the employment of Chl as an intrinsic probe
for evaluating light screening efficiency in plants; analysis
of whole apple fruit reflectance and Chl fluorescence was
applied to samples widely differing in Chl, Car, Flv, and
AnC content, due to differences in the degree of fruit
ripeness and illumination during fruit development.
Effect of anthocyanins, carotenoids, and flavonols on chlorophyll fluorescence 357
Although precise fruit prehistory was unknown, spectral
reconstruction revealed a pool of extrathylakoid Car in
each sample, characteristic of the ripening process
(Merzlyak and Solovchenko, 2002; Solovchenko et al.,
2005, 2006; Merzlyak, 2006). Furthermore, our previous
observations make it plausible that the specimens with
a high [Chl]/[Car] ratio represent mature fruit at the early
stages of ripening whereas those with lower [Chl] but with
enhanced [Car] are characteristic of advanced stages of
ripening (Merzlyak and Solovchenko, 2002; Merzlyak
et al., 2002; Solovchenko et al., 2005, 2006). A rich Flv
content and a higher ½Car=½Chl ratio is inherent to
Granny Smith and Golden Delicious apples exposed to
strong solar radiation during their growth as compared
with shaded fruit (Merzlyak et al., 2002, 2005b;
Solovchenko et al., 2006). As a response to strong sunlight
(Saure, 1990), Summer Red apples accumulated considerable amounts of AnC on their sunlit surfaces. It should be
noted that in sunlit fruit surfaces Flv (Granny Smith and
Golden Delicious) and AnC (Summer Red) content
exceeded those of Chl and Car by a factor of 10–100.
CFE spectra of apple fruit were structured and contained
details which to a lesser degree were expressed in
reflection spectra. In AnC-free apples in the green-red
region where only Chla and b participate in light
absorption, the CFE spectra were highly resolved. In fruit
with low [Chl], fluorescence excitation maximum coincided with the main Chla absorption maximum at 678 nm
and between 550 nm and 700 nm the shapes of the
measured fluorescence spectra were close to the model
spectra. The growth of Chl content was accompanied by
remarkable changes in the shape of the spectra, the
increase in fluorescence excitation ratio F650/F678 and the
shift of maximum position towards shorter wavelengths in
CFE spectrum reaching as large as 15–18 nm in apples
with high Chl. Evidently together with an increase of
normalized fluorescence in the green-orange and far-red
regions, along with an increase in Chl, the spectral
changes in the red involve re-absorption of Chl fluorescence also manifested in emission spectra of apple fruit
(Ramos and Lagorio, 2006). Below 550 nm CFE, spectra
of AnC-free apple fruit, in addition to the sharp maximum
near 440 nm, showed Chla bands around 386 nm and 418
nm as shoulders not apparent in reflection spectra.
To determine spectral bands in CFE spectra sensitive to
individual pigments, correlation analysis was carried out.
It should be noted that apart from absorption of incident
radiation by Flv and Car, the decrease in the F-½Chl
correlation in AnC-free fruit results from scattering and
could also be associated with a non-linear relationship
between light absorption and pigment content. In particular, this effect appears as the gap near 678 nm in the
correlation spectra between spectral reflection and Chl in
apple fruit (Merzlyak et al., 2003). In addition, in fruit
with high [Chl], fluorescence re-absorption brought about
an apparent shift of the fluorescence maximum. These
reasons account for the low F-½Chl
correlation observed
in the 650–690 nm band. With the exception of this band,
spectral regions correlated linearly with Chl content were
found in the narrow band near 700 nm, and in the green
up to approximately 500 nm. Below 500 nm, where the
Q-[Car] correlation increased, the Q-[Chl] correlation
underwent a considerable decline, as did the F-½Chl.
The
F-½Car
correlation was consistently low, and with the
cultivars studied here, the lower the [Chl]-[Car] correla
tion, the lower was the F-½Car
correlation. A further
decrease of the F-½Chl
correlation occurred at wavelengths at which Flv exerted a dominant contribution on
the reflection spectra and the negative F-½Flv
correlation
was strong. In the 380–420 nm band, especially high
Q-[Flv] correlation and F-½Flv
correlation, reaching the
maximum near 410 nm, was found for Golden Delicious
fruit, characteristic of sun-exposed fruit accumulating high
amounts of vacuolar Flv (mainly quercetin-3-glycosides)
(Solovchenko and Merzlyak, 2003; Merzlyak et al.,
2005b; Hagen et al., 2006).
The CFE of sunlit and shaded fruit surfaces of Golden
Delicious apples contrasting in Flv displayed a remarkable
difference below 500 nm. Notably, in these fruit the
influence of the Flv was pronounced even in the blue
range of the spectrum. In sunlit versus shaded fruit Chl
fluorescence intensity was markedly lower, especially
between 360 nm and 400 nm, resulting in obliteration of
Chla spectral features. In addition, the ‘shaded-to-sunlit’
ratio spectrum (Fig. 2B) revealed two bands near 455 nm
and 485 nm attributable to Car (Gradinaru et al., 2000;
Croce et al., 2001; Merzlyak and Solovchenko, 2002;
Merzlyak et al., 2003), whose content was reported
(Merzlyak et al., 2002; Ma and Cheng, 2004; Solovchenko
et al., 2006) to increase in sun-exposed apple fruit. The
modelling of CFE spectrum of a Golden Delicious fruit
with an increased Flv content indicates that near 400 nm
internal trapping by Flv (and, to some extent, also by
Car) are responsible for an almost 8-fold decrease of Chl
fluorescence.
Although the photoprotective role of AnC in apple fruit
has been questioned (Ma and Cheng, 2004), the results of
the present study show that these pigments exert an
unmistakable effect on Chl fluorescence. The increase in
[AnC] (mainly, cyanidin-3-galactoside) (Hagen et al.,
2006) in Summer Red fruit was accompanied by a content-dependent decrease of Chl fluorescence in a broad
band up to 650 nm, and at high AnC content only a weak
Chla peak at 440 nm was detected (Fig. 4A, B). The
effect of AnC was so strong that the correlation between
Chl fluorescence and its content was very weak throughout the whole visible range. Our modelling of the CFE
spectrum of a Summer Red fruit with a moderate content
of AnC has revealed that the pigment exerts a large
influence on spectral features between 450 nm and 650
358 Merzlyak et al.
nm, peaking near 530 nm where W(k) approached 8. It is
also remarkable that vacuolar Flv and AnC which at high
concentrations govern optical properties in the near UV
(Merzlyak et al., 2005b) and green (Merzlyak et al., 2003)
spectral ranges, respectively, displayed similar (but with
opposite sign) correlation with Q(k) and FðkÞ.
In ripening fruit, Car (presumably those localizing in
chloroplast plastoglobuli) (Merzlyak and Solovchenko,
2002) exerted a significant effect on CFE that was
manifested by a reduction in the F-½Car
correlation
between 400 nm and 500 nm. According to our estimates,
light trapping attributable to the extrathylakoid Car in the
band 440–490 nm in unripe apples with high [Chl] was
small (not shown), increased significantly at advanced
stages of fruit ripening (Fig. 5A, D) and in sunlit fruit
(Fig. 5B, E) when W(k) was close to 2 and 3, respectively.
It is noteworthy that model FðkÞ
spectra contained poorly
resolved and overlapping bands of Chlb and Car absorption between 450 nm and 490 nm inherent in the spectra
of higher plant light harvesting pigment–protein complexes (Croce et al., 2000; Gradinaru et al., 2000) and
PT(k) pool. By contrast, recorded CFE spectra contained
distinct maxima near 464–472 nm and 485–490 nm (Figs
2, 5). The modelling of the spectra of AnC-free fruit
strongly suggests that the fluorescence in this band
resulted mainly from the overlapping absorption by PX(k)
and PT(k) pigment pools. As could be seen from Fig. 5,
the shapes of the spectra, the ratios between maxima and
the positions of maxima in the bands of combined Chl and
Car absorption were, to a large extent, reproduced in the
model spectra.
Together with reflection spectroscopy, Chl fluorescence
is a promising tool for rapid and non-destructive analysis
of pigments able to compete with Chl for light absorption
(Cerovic et al., 2002; Agati et al., 2005; Bengtsson et al.,
2006; Hagen et al., 2006; Barthod et al., 2007). The
analysis carried out in this study showed that, in AnC-free
apples, Chl content itself could be determined from CFE
spectral measurements using a linear model and wavelengths at which absorption is relatively small. Compared
with reflection, Chl fluorescence between 500 nm and 650
nm was less sensitive to the influence of Car whose
contribution was especially expressed in ripening and
light-stressed fruit. The analysis of Chl, however, was not
applicable to fruit containing moderate-to-high amounts of
AnC pigments. In the AnC-free fruit examined, the assessment of Flv was feasible using the CFE ratio, F440/F400.
For Summer Red apples the CFE ratio, F700/F580, was
found to be effective in the assessment of AnC, even in
fruit with a high AnC content. Close precision of AnC and
total Flv analysis was reported in recent measurements
with a PAM Chl fluorometer in Aroma apples (Hagen
et al., 2006). Although quantitative analysis of total Car in
AnC-free apple fruit with reflection spectroscopy was
developed (Merzlyak et al., 2003), using the fluorescence
technique for this purpose was not successful; the failure
may be due to the involvement of different Car pools in
light harvesting and in screening, as implied by the data
presented here.
Collectively, the signature analysis, the spectra of fruit
with contrasting pigment content, quantitative relationships together with the modelling of CFE spectra, suggest
that Flv, Car, and AnC are able to exert strong photoprotective effects in specific spectral regions. It is remarkable that the adaptation of apple fruit to high light stress
involves the accumulation of the pigments with strongly
overlapped absorption that provide screening and internal
light trapping of solar radiation in broad spectral ranges
extending from UV to the green and, in AnC-accumulating cultivars, to the red regions of the visible spectrum. In
conclusion, it should also be noted that spectrophotometric analysis and the possibilities of reconstruction of
pigment contributions in total spectral absorption provide
opportunities for quantitative evaluation of efficiency of
the photoprotective pigment in intact plant tissues.
Acknowledgements
This work was supported in part by grants for MNM from the
Research Council of Norway (NFR) and the Russian Fund for Basic
Research No. 06-04-4883. The authors are grateful to Dr AE
Solovchenko for his helpful discussion.
References
Agati G, Meyer S, Matteini P, Cerovic ZG. 2007. Assessment of
anthocyanins in grape (Vitis vinifera L.) berries using a noninvasive chlorophyll fluorescence method. Journal of Agricultural
and Food Chemistry 55, 1053–1061.
Agati G, Pinelli P, Ebner SCS, Romani A, Cartelat A,
Cerovic ZG. 2005. Non-destructive evaluation of anthocyanins
in olive (Olea europaea) fruits by in situ chlorophyll fluorescence
spectroscopy. Journal of Agricultural and Food Chemistry 53,
1354–1363.
Alderson JV, Atherton A, Derbyshyre AN. 1961. Modern
physical techniques in color formulation. Journal of the Society
of Dyers and Colorists 77, 657–669.
Atherton A. 1955. The relation of the reflectance of dyed fabrics
to dye concentration and the instrumental approach to
colour matching. Journal of the Society of Dyers and Colorists
71, 389–398.
Barnes PW, Searles PS, Ballaré CL, Ryel RJ, Caldwell MM.
2000. Non-invasive measurements of leaf epidermal transmittance
of UV radiation using chlorophyll fluorescence: field and
laboratory studies. Physiologia Plantarum 109, 274–283.
Barthod S, Cerovic Z, Epron D. 2007. Can dual chlorophyll
fluorescence excitation be used to assess the variation in the
content of UV-absorbing phenolic compounds in leaves of
temperate tree species along a light gradient? Journal of
Experimental Botany 58, 1753–1760.
Bengtsson GB, Schöner R, Lombardo E, Schöner J, Borge GIA,
Bilger W. 2006. Chlorophyll fluorescence for non-destructive
measurement of flavonoids in broccoli. Postharvest Biology and
Technology 39, 291–298.
Effect of anthocyanins, carotenoids, and flavonols on chlorophyll fluorescence 359
Bilger W, Veit M, Schreiber L, Schreiber U. 1997. Measurement
of leaf epidermal transmittance of UV radiation by chlorophyll
fluorescence. Physiologia Plantarum 101, 754–763.
Bilger W, Johnsen T, Schreiber U. 2001. UV-excited chlorophyll
fluorescence as a tool for assessment of UV-protection by the
epidermis of plants. Journal of Experimental Botany 52, 2007–
2014.
Blanke MM, Lenz F. 1989. Fruit photosynthesis. Plant, Cell and
Environment 12, 31–46.
Bowen EJ. 1936. Heterochromatic photometry of ultra-violet
region. Proceedings of the Royal Society of London, Series A
154, 349–353.
Bursteyn EA. 1968. Extinguishing of protein fluorescence. I.
Principles, solutions of triptophane, tyrozin, and denaturated
proteins. Biophyzika (Biophysics USSR) 13, 433–442.
Buschmann C, Lichtenthaler HK. 1988. Reflectance and chlorophyll fluorescence signatures of leaves. In: Lichtenthaler HK,
ed. Applications of chlorophyll fluorescence. Kluwer Academic,
325–332.
Cerovic ZG, Ounis A, Cartelat A, Latouche G, Goulas Y,
Meyer S, Moya I. 2002. The use of chlorophyll fluorescence
excitation spectra for the non-destructive in situ assessment of
UV-absorbing compounds in leaves. Plant, Cell and Environment
25, 1663–1676.
Cockell CS, Knowland J. 1999. Ultraviolet radiation screening
compounds. Biological Reviews 74, 311–345.
Croce R, Cinque G, Holzwarth AR, Bassi R. 2000. The
Soret absorption properties of carotenoids and chlorophylls in
antenna complexes of higher plants. Photosynthesis Research 64,
221–231.
Croce R, Müller MG, Bassi R, Holzwarth AR. 2001. Carotenoidto-chlorophyll energy transfer in recombinant major light-harvesting
complex (LHCII) of higher plants. I. Femtosecond transient absorption measurements. Biophysical Journal 80, 901–915.
Day TA, Martin G, Vogelmann TC. 1993. Penetration of UV-B
radiation in foliage: evidence that the epidermis behaves as a nonuniform filter. Plant, Cell and Environment 16, 735–741.
Gitelson AA, Buschmann C, Lichtenthaler HK. 1998. Leaf
chlorophyll fluorescence corrected for re-absorption by means of
absorption and reflectance measurements. Journal of Plant
Physiology 152, 283–296.
Gradinaru CC, van Stokkum IHM, Pascal AA, van Grondelle R,
van Amerongen H. 2000. Identifying the pathways of energy
transfer between carotenoids and chlorophylls in LHCII and CP29.
A multicolor, femtosecond pump-probe study. Journal of Physical
Chemistry (B) 104, 9330–9342.
Hagen SF, Solhaug KA, Bengtsson GB, Borge GIA, Bilger W.
2006. Chlorophyll fluorescence as a tool for non-destructive
estimation of anthocyanins and total flavonoids in apples.
Postharvest Biology and Technology 41, 156–163.
Han Q, Shinohara K, Kakubari Y, Mukai Y. 2003. Photoprotective role of rhodoxanthin during cold acclimation in
Cryptomeria japonica. Plant, Cell and Environment 26, 715–723.
Lazár D. 1999. Chlorophyll a fluorescence induction. Biochimica
et Biophysica Acta 1412, 1–28.
Ma F, Cheng L. 2004. Exposure of the shaded side of apple fruit to
full sun leads to up-regulation of both the xanthophyll cycle and
the ascorbate–glutathione cycle. Plant Science 166, 1479–1486.
Merzlyak MN. 2006. Modeling pigment contributions to spectral
reflection of apple fruit. Photochemical and Photobiological
Sciences 5, 748–754.
Merzlyak MN, Chivkunova OB. 2000. Light stress induced
pigment changes and evidence for anthocyanin photoprotection
in apple fruit. Journal of Photochemistry and Photobiology (B)
55, 154–162.
Merzlyak MN, Solovchenko AE. 2002. Photostability of pigments
in ripening apple fruit: a possible photoprotective role of
carotenoids during plant senescence. Plant Science 163, 881–888.
Merzlyak MN, Solovchenko AE, Chivkunova OB. 2002. Patterns
of pigment changes in apple fruits during adaptation to high
sunlight and sunscald development. Plant Physiology and
Biochemustry 40, 679–684.
Merzlyak MN, Solovchenko AE, Gitelson AA. 2003. Reflectance
spectral features and non-destructive estimation of chlorophyll,
carotenoid and anthocyanin content in apple fruit. Postharvest
Biology and Technology 27, 88–103.
Merzlyak M, Solovchenko A, Pogosyan S. 2005a. Optical
properties of rhodoxanthin accumulated in Aloe arborescens Mill.
leaves under high-light stress with special reference to its
photoprotective function. Photochemical and Photobiological
Sciences 4, 333–400.
Merzlyak MN, Solovchenko AE, Smagin AI, Gitelson AA.
2005b. Apple flavonols during fruit adaptation to solar radiation:
spectral features and technique for non-destructive assessment.
Journal of Plant Physiology 162, 151–160.
Parker CA. 1968. Photoluminescence of solutions with applications to photochemistry and analytical chemistry. Amsterdam,
London, New York: Elsevier Publishing Company.
Pfündel EE, Agati G, Cerovic ZG. 2006. Optical properties of
plant surfaces. In: Riederer M, Müller C, eds. Biology of the plant
cuticle. Annual Plant Reviews, Vol. 23. Oxford: Blackwell
Publishing, 216–249.
Ramos ME, Lagorio MG. 2006. A model considering light
reabsorption processes to correct in vivo chlorophyll fluorescence
spectra in apples. Photochemical and Photobiological Sciences 5,
508–512.
Roháček K. 2002. Chlorophyll fluorescence parameters: the
definitions, photosynthetic meaning, and mutual relationships.
Photosynthetica 40, 13–29.
Saure MC. 1990. External control of anthocyanin formation in
apple. Scientia Horticulture 42, 181–218.
Smith J, Markham KR. 1998. Tautomerism of flavonol glucosides: relevance to plant UV protection and ower colour. Journal
of Photochemistry and Photobiology (A) 118, 99–105.
Solovchenko A, Merzlyak M. 2003. Optical properties and
contribution of cuticle to UV protection in plants: experiments
with apple fruit. Photochemical and Photobiological Sciences 2,
861–866.
Solovchenko AE, Avertcheva OV, Merzlyak MN. 2006. Elevated
sunlight promotes ripening-associated pigment changes in apple
fruit. Postharvest Biology and Technology 40, 183–189.
Solovchenko AE, Chivkunova OB, Merzlyak MN,
Gudkovsky VA. 2005. Relationships between chlorophyll and
carotenoid pigments during on- and off-tree ripening of apple
fruit as revealed non-destructively with reflectance spectroscopy.
Postharvest Biology and Technology 38, 9–17.
Solovchenko AE, Schmitz-Eiberger M. 2003. Significance of skin
flavonoids for UV-B protection in apple fruits. Journal of
Experimental Botany 54, 1977–1984.
Steyn WJ, Wand SJE, Holcroft DM, Jacobs G. 2002. Anthocyanins in vegetative tissues: a proposed unified function in photoprotection. New Phytologist 155, 349–361.