CO2 enrichment can produce high red leaf lettuce yield while increasing most flavonoid glycoside and some caffeic acid derivative concentrations

Abstract

Carbon dioxide (CO2) enrichment is a common practice in greenhouses to increase crop yields up to 30%. Yet, reports on the effect on foliar phenolic compounds vary. We studied the effect on two red leaf lettuce cultivars, grown for 25days in growth chambers at CO2 concentrations of 200 or 1000ppm, with some plants exchanged between treatments after 11days. As expected, head mass increased with higher CO2 concentration. Regression analysis, corrected for head mass, showed increased concentrations of most flavonoid glycosides at high CO2 concentrations while only some caffeic acid derivatives were increased, and not uniformly in both cultivars. Sugar concentrations increased with CO2 concentration. Generally, conditions in the 10days before harvest determined concentrations. We suspect that phenolic compounds were mainly accumulated because plenty of precursors were available. The results indicate that CO2 enrichment can result in high yields of red leaf lettuce rich in phenolic compounds.

Full text

CO
2
enrichment can produce high red leaf lettuce yield while increasing
most flavonoid glycoside and some caffeic acid derivative concentrations
Christine Becker
⇑
, Hans-Peter Kläring
Leibniz Institute of Vegetable and Ornamental Crops, Theodor-Echtermeyer-Weg 1, 14979 Großbeeren, Germany
article info
Article history:
Received 11 September 2015
Received in revised form 10 December 2015
Accepted 11 December 2015
Available online 15 December 2015
Keywords:
CO
2
Lettuce
Lactuca sativa
Flavonoid glycosides
Anthocyanin
Caffeic acid derivatives
Sugar
abstract
Carbon dioxide (CO
2
) enrichment is a common practice in greenhouses to increase crop yields up to 30%.
Yet, reports on the effect on foliar phenolic compounds vary. We studied the effect on two red leaf lettuce
cultivars, grown for 25 days in growth chambers at CO
2
concentrations of 200 or 1000 ppm, with some
plants exchanged between treatments after 11 days. As expected, head mass increased with higher
CO
2
concentration. Regression analysis, corrected for head mass, showed increased concentrations of
most flavonoid glycosides at high CO
2
concentrations while only some caffeic acid derivatives were
increased, and not uniformly in both cultivars. Sugar concentrations increased with CO
2
concentration.
Generally, conditions in the 10 days before harvest determined concentrations. We suspect that phenolic
compounds were mainly accumulated because plenty of precursors were available. The results indicate
that CO
2
enrichment can result in high yields of red leaf lettuce rich in phenolic compounds.
Ó2015 The Authors. Published by Elsevier Ltd. This is an open accessarticle under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Carbon dioxide (CO
2
) enrichment is a commonly used method
to increase yields of greenhouse cultivated crops (Chalabi, Biro,
Bailey, Aikman, & Cockshull, 2002), featuring CO
2
concentrations
of up to 1000 ppm in the greenhouse air. Increases of up to 30%
are well possible in northern countries during autumn, winter
and spring in fruit vegetables (Willits & Peet, 1989) and also in let-
tuce (Hunt, Wilson, Hand, & Sweeney, 1984). Recently, closed
greenhouses were developed. During high solar radiation periods,
those greenhouses capture and store thermal energy to reuse it
for heating during dark and cold periods (Schmidt et al., 2011).
The closed operation mode allows for maintaining high CO
2
con-
centrations year round resulting in very high yields (De Gelder,
Dieleman, Bot, & Marcelis, 2012). However, without CO
2
supply
to a greenhouse with closed ventilation, the CO
2
concentration in
the air can decrease down to 150 ppm during the day due to the
CO
2
uptake by plants, as shown for cucumber (Kläring,
Hauschild, Heißner, & Bar-Yosef, 2007).
In contrast to the effect of the CO
2
concentration on photosyn-
thesis and yield, there are fewer reports on its effect on secondary
metabolites in vegetables. In general, the effect of CO
2
concentra-
tion on secondary metabolites seems to be low compared to other
environmental factors. Thus, Krumbein, Schwarz, and Kläring
(2006) could not find any effect on carotenoid content in tomato.
In broccoli, rising CO
2
concentration increased the total glucosino-
late concentration which however, was counteracted by a decrease
of concentration of indole glucosinolates (Schonhof, Kläring,
Krumbein, & Schreiner, 2007). The existing reports on the effect
of CO
2
enrichment on the concentration of foliar flavonoids and
phenolic acids also show mixed results, ranging from increases to
decreases or no effect.
In grapevine (Vitis vinifera L.), increased concentrations of flavo-
noids were detected (Bindi, Fibbi, & Miglietta, 2001). In strawber-
ries (Fragaria x ananassa), Wang, Bunce, and Maas (2003) found
increased anthocyanin and flavonol concentrations and
Kuokkanen, Julkunen-Tiitto, Keinänen, Niemelä, and Tahvanainen
(2001) measured increased flavonol glycoside concentrations in
birch seedlings (Betula pendula Roth).
Peñuelas et al. (1996) tested the responses of several species
and found differences: Elevated CO
2
concentrations resulted in
increased concentrations of total phenolics in wheat leaves (Triti-
cum aestivum), decreased concentrations in pine tree needles
(Pinus eldarica L.) and had no effect regarding orange tree leaves
(Citrus aurantium L.). In high CO
2
concentrations, Peltonen,
Vapaavuori, and Julkunen-Tiitto (2005) observed increased con-
centrations of phenolic acids, flavonols, flavanols and condensed
http://dx.doi.org/10.1016/j.foodchem.2015.12.059
0308-8146/Ó2015 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
⇑
Corresponding author at: French National Institute for Agricultural Research
(INRA), University of Nice Sophia Antipolis, CNRS, UMR 1355-7254, Institut Sophia
Agrobiotech, 06903 Sophia Antipolis, France. Institut de Chimie de Nice, UMR CNRS
7272, University of Nice Sophia Antipolis, Parc Valrose, 06108 Nice Cedex 2, France.
E-mail address: c_becker@gmx.org (C. Becker).
Food Chemistry 199 (2016) 736–745
Contents lists available at ScienceDirect
Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

tannins but decreased concentrations of flavones in birch leaves
while Sallas, Luomala, Utriainen, Kainulainen, and Holopainen
(2003) detected decreased total phenolics concentration in Norway
Spruce (Picea abies) but not in Scots Pine (Pinus sylvestris L.). The
species-related heterogeneity may be increased by different time
spans the plants were exposed to elevated CO
2
concentrations, dif-
ferent developmental stages tested in different environments, i.e.
growth chambers, greenhouses, open fields, pots or field-soil, as
well as unspecific measurements, i.e. total phenolics or total
flavonoids instead of single compounds, of different plant organs
(Goufo et al., 2014). On average, however, elevated atmospheric
CO
2
concentration may stimulate the production of flavonoids
and other phenolics (DeLucia, Nabity, Zavala, & Berenbaum, 2012).
Epidemiological studies strongly link a diet rich in phenolic
compounds with a low incidence of coronary heart disease or can-
cer (Boudet, 2007). Although in vitro tests with single substances
corroborate these beneficial effects (Mulabagal et al., 2010), they
have to be considered carefully as in vivo studies do not always
reproduce these effects (Rimbach, Melchin, Moehring, & Wagner,
2009). According to in vitro tests, synergistic and additive effects
of dietary phenolic compounds are considered to play a major role
(Boudet, 2007) but high doses of flavonols may be disadvantageous
due to possible pro-oxidative effects (Rietjens et al., 2005). Hence,
it appears wise to generally enhance their accumulation in fruits
and vegetables by horticultural approaches instead of administer-
ing single substances as dietary supplement. To achieve this we
need to monitor how their concentrations change in response to
different cultivation strategies. Major phenolic compounds in red
leaf lettuce are glycosides of cyanidin, quercetin, and luteolin as
well as esters of caffeic acid (Llorach, Martínez-Sánchez, Tomás-
Barberán, Gil, & Ferreres, 2008). The cyanidin glycosides are espe-
cially important for visual quality as they are responsible for the
red color of the leaves (Gould & Lister, 2006).
CO
2
enrichment leads to increased photosynthesis rates
(Nederhoff & Vegter, 1994). Hence, there is an abundance of assim-
ilates which can be funneled into biosynthetic pathways of all sorts
(Treutter, 2010). This becomes obvious by enhanced biomass accu-
mulation and may result in higher concentrations of secondary
metabolites due to higher availability of precursor molecules. The
pathway for the biosynthesis of flavonoids and caffeic acid deriva-
tives, the phenylpropanoid pathway, is reported to be directly
linked to the carbohydrate status of plants: Monosaccharides are
in several steps transformed into phenylalanine via the shikimate
pathway and the first step of the phenylpropanoid pathway is
de-amination of phenylalanine to obtain cinnamic acid (Schopfer
& Brennicke, 2010). Additionally, sugar has been found to directly
upregulate transcription factors involved in anthocyanin biosyn-
thesis in Arabidopsis thaliana (Solfanelli, Poggi, Loreti, Alpi, &
Perata, 2006).
According to the International Panel on Climate Change (IPCC),
the global atmospheric CO
2
concentration is rising (IPCC, 2007).
However, horticultural crops are under-represented when it comes
to studying the effects of elevated atmospheric CO
2
concentration
on phenolic compounds which leaves the IPCC with little data to
predict future crop yields and quality (Goufo et al., 2014). As far
as we know, there are no studies regarding the effect of CO
2
concentration on phenolic compounds in lettuce to date.
In a study on the effect of irradiance on phenolic compounds in
lettuce, we found that their concentration depends rather on the
radiation intensity during the last days before harvest than on
the average of the total cultivation period (Becker, Klaering, Kroh,
& Krumbein, 2013).
In the presented study, we tested the following hypotheses
under controlled conditions: (I) the CO
2
concentration has a posi-
tive effect on the concentration of flavonoid glycosides and caffeic
acid derivatives in lettuce, (II) CO
2
enrichment during lettuce cul-
tivation is more efficient in the weeks before harvest than in the
weeks after planting, and (III) high CO
2
concentrations increase
the concentrations of sugars which serve as precursors for the
phenylpropanoid pathway. We used two cultivars because differ-
ent genotypes may respond differently (Jaafar, Ibrahim, & Karimi,
2012). Instead of measuring total concentrations, we measured
single phenolic compounds using HPLC-DAD-ESI-MS
3
. Based on
our own and the experience of others (Becker et al., 2013;
Caporn, 1989) we cultivated lettuce for several weeks at a light
intensity that was saturating but not stressful, at low and high
CO
2
concentration, respectively, and standard cultivation tempera-
ture, to obtain results of practical relevance.
2. Material and methods
2.1. Plant cultivation
Red Oak Leaf and red Lollo lettuce seeds (Lactuca sativa L. var.
crispa L. cv. Eventai RZ and L. sativa L. var. crispa L., cv. Satine,
respectively; RijkZwaan, De Lier, The Netherlands) were sown in
rockwool cubes, kept at 10 °C for two days for germination and
subsequently grown in a conventional greenhouse until the exper-
iment started. When plants had developed five true leaves (five
weeks old) and weighed about 1.6 g they were transferred into four
growth chambers (Weiss Gallenkamp, Loughborough, UK) where
they were grown in 2 L containers in aerated nutrient solution.
The nutrient solution was prepared according to Sonneveld and
Straver (1988). It contained the following ions in mmol L
1
19.0 NO

3
, 1.25 NH
þ
4
, 2.0 H
2
PO

4
. 1.125 SO
2
4
, 11.0 K
+
, 4.5 Ca
2+
, 1.0
Mg
2+
, and in
l
mol L
1
40.0 Fe
3+
, 1.0 Mn
2+
, 4.0 Zn
2+
, 30.0 BðOHÞ

4
,
0.75 Cu
2+
, and 0.5 MoO
2
4
, resulting in an electrical conductivity
of the nutrient solution of 2.3 dS m
1
. The pH of the nutrient solu-
tion was 5.5. Solution taken up by the plants was replenished peri-
odically. Air temperature in the chambers was kept at 20 °C during
daytime and 15 °C at night, relative humidity was approximately
80% and radiation was supplied by halogen metal vapor lamps
Osram Powerstar HQI-BT 400 W/D Pro (Osram, Munich, Germany)
and krypton lamps. The light cycle consisted of four elements: 11 h
of darkness, 0.5 h of dawn, 12 h of light and another 0.5 h of twi-
light. During the light phase, the mean photosynthetic photon flux
density (PPFD) was 260
l
mol m
2
s
1
. During dusk and dawn only
the krypton lamps were switched on, resulting in a mean PPFD of
95
l
mol m
2
s
1
. These radiation intensities were measured with a
portable light meter LI-250 (LI–COR Inc., Lincoln, Nebraska, USA).
The daily light integral was 11.6 mol m
2
.
Technical pure CO
2
was supplied to all chambers during the
light phase to compensate for the uptake by the plants. The set
points for the control of the CO
2
concentration were 1000 ppm
and 200 ppm each in two chambers. In the latter, the CO
2
concen-
tration during the dark phase increased to about 500 ppm due to
the plant’s respiration and the lack of a CO
2
absorber. Shortly after
the beginning of the light phase it arrived at 200 ppm as result of
the plants’ photosynthesis and was then kept at this level by the
CO
2
controller.
Each chamber held 12 plants of each cultivar. Per cultivar, four
plants were harvested 11 days after planting (DAP). At the same
day four plants were exchanged between the CO
2
treatments. After
14 more days, all remaining plants were harvested. For each culti-
var, 8 replicates consisting of 4 plants from two independently
controlled chambers were installed. However, regarding Oak Leaf,
treatment 1000_200 and 200_1000, only 6 and 7 plants, respec-
tively, could be considered because the others were infected with
a fungus. As the plants were cultivated in independent pots, the
infection did not spread.
C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745 737

2.2. Plant growth characteristics
For all samples, above ground (lettuce heads) and below ground
(roots) organs were harvested separately. At both harvest dates,
eight plants per cultivar and treatment were weighed to obtain
the mean head and root mass. Values are given in gram fresh mat-
ter (FM). Lettuce heads were cut in pieces. Some head fresh matter
and the roots were dried in an oven at 80 °C for three days, to
obtain dry matter content. Values for dry matter content are given
in milligram dry matter per gram fresh matter.
2.3. Sample preparation
Within 30 min after harvesting, the remaining cut lettuce heads
were frozen at 20 °C until lyophilized (Christ Beta 1–16, Oster-
ode, Germany). Only limp or deteriorated outer leaves were
removed during preparation. Dried plant matter was ground with
an ultracentrifuge mill (hole size: 0.25 mm; ZM 200, Retsch, Haan,
Germany).
2.4. Analyses of phenolic compounds
The well-established HPLC-DAD-ESI-MS
3
method for the deter-
mination of flavonol glycosides and phenolic acids, reported by
Scattino et al. (2014) was optimized for lettuce. Best results were
obtained by extracting 20 mg of lyophilized, pulverized lettuce
powder three times (600, 300, and 300
l
l) with aqueous methanol
(50% MeOH; Carl Roth GmbH, Karlsruhe, Germany) for 40, 20, and
10 min in a thermomixer (1400 rpm, 20 °C; Eppendorf ther-
momixer comfort, Germany). Between the three extraction phases,
they were centrifuged at 4.500 rpm and 20 °C for 10 min (Labofuge
400R, Heraeus Instruments, Thermo Fisher Scientific, Waltham,
USA) and the supernatants were combined. To remove larger par-
ticles, the extracts were filtered through Spin-X centrifuge filters
(0.22
l
m; Geyer GmbH, Berlin, Germany) by centrifugation at
3.000 rpm and 20 °C for 5 min. Afterwards they were transferred
to glass vials and analyzed via HPLC-DAD-ESI-MS
3
.
The anthocyanin extracts were prepared similarly to the
method applied to flavonols and phenolic acids, except for a
slightly different composition of the extraction agent and a shorter
extraction time: The extraction agent was acidified aquaeous
methanol (40% MeOH, 10% acetic acid) to a pH value of 2.6. Extrac-
tion of anthocyanin glycosides took 3 10 min.
The system used for analysis consists of an Agilent HPLC series
1100 (Agilent, Waldbronn, Germany), containing of a degaser,
binary pump, autosampler, thermostat and a photodiode array
detector (DAD). The components were separated on a Ascentis
Express F5 column with a C18 security guard (150 4.6 mm,
5
l
m; 4 4.6 mm, 5
l
m; Supelco Analytical, Sigma Aldrich,
Munich, Germany) at 30 °C using a water/acetonitrile gradient.
Solvent A consisted of 99.5% water and 0.5% acetic acid (Merck,
Darmstadt, Germany) whereas solvent B was 100% acetonitrile
(ACN; J.T. Baker, Deventer, The Netherlands). Two separate gradi-
ents were used for flavonol glycosides and phenolic acids (gradient
1) and anthocyanins (gradient 2), respectively. Gradient 1 held the
following percentages of ACN: 5% (2 min), 5–12% (13 min), 12–20%
(31 min), 20–90% (3.5 min), 90% isocratic (2.5 min), 90–5%
(0.7 min), isocratic 5% (6.3 min). Gradient 2 was distinctly shorter:
5–20% B (2 min), 20% B isocratic (1 min), 20–90% B (3.5 min), 90% B
isocratic (2.5 min), 90–5% B (1 min) and 5% isocratic (3 min). Flow
rate in both gradients was 0.85 ml/ min. Flavonol glycosides and
phenolic acids were detected in the mass spectrometer as deproto-
nated molecular ions and characteristic mass fragment ions using
an Agilent series 1100 MSD (ion trap) with ESI as ion source in neg-
ative mode. Nitrogen served as dry gas (10 l/min; 350 °C) and neb-
ulizer gas (40 psi). Helium was used as collision gas in the ion trap.
Mass optimization was performed for quercetin 3-O-glucoside
[MH]

m/z. Anthocyanidin glycosides were identified using the
positive mode. Identification of the compounds was achieved by
comparing retention time, absorption maxima and mass spectra
to that of standard substances, when available, or to literature data
(DuPont, Mondin, Williamson, & Price, 2000; Llorach et al., 2008).
Standard substances were purchased at Carl Roth GmbH
(Karlsruhe, Germany; quercetin-3-O-glucoside) and Sigma–Aldrich
GmbH (Munich, Germany; di-O-caffeoyltartaric acid and
cyanidin-3-O-glucoside).
The DAD was used for quantification, using the detection wave-
lengths 330 nm (phenolic acids), 350 nm (flavonol and flavone gly-
cosides) and 520 nm (anthocyanidin glycosides). External
calibration curves were prepared in the respective relevant
concentrations, using the standard substances where available.
Cyanidin-3-O-(6
00
-O-malonyl)-glucoside was quantified as
cyanidin-3-O-glucoside. Quercetin-3-O-(6
00
-O-malonyl)-glucoside,
quercetin-3-O-glucuronide and luteolin-7-O-glucuronide were
quantified quercetin-3-O-glucoside equivalents. The caffeic acid
derivatives were quantified as di-O-caffeoyltartaric acid
equivalents.
2.5. Analysis of sugars
For sugar analysis, 10 mg of freeze-dried plant material were
extracted with 800
l
l of aqueous ethanol (80% EtOH; Carl Roth
GmbH, Karlsruhe, Germany), vortexed and incubated for 20 min
at 78 °C. After centrifugation at 14000 rpm and 4 °C for 10 min,
the supernatant was transferred into a new Eppendorf reaction
tube. The extraction was repeated twice with 400
l
l 50% EtOH.
Of the combined supernatants, 5
l
l were and analyzed via enzy-
matic essay in microplates as described by Klopotek and Kläring
(2014).
2.6. Statistical analyses
In order to detect significant differences regarding growth char-
acteristics due to the different CO
2
concentrations, one-way
ANOVA was performed (Fisher’s F-test) for each cultivar and har-
vest date separately, followed by Tukey’s Honest Significant Differ-
ence test with a significance level of
a
= 0.05. In order to detect
significant differences regarding phenolic compounds and sugars
due to the different CO
2
concentrations, regression analysis was
performed for each cultivar separately. In order to test hypothesis
(II) that conditions shortly before harvest have much greater
impact on phenolic compound concentrations than those after
planting, all data were related to the CO
2
concentration during
the last 10 days before plant sampling at 11 and 25 DAP. In addi-
tion, polyphenol concentrations are strongly influenced by plant
development (Becker, Klaering, Schreiner, Kroh, & Krumbein,
2014b). Therefore we included the plants’ head mass (logarithmic)
into this analysis. Regression coefficients were evaluated using
Student’s t-test. A significance level of
a
= 0.05 was applied. Single
plants were considered biological replicates. Calculations were
performed using STATISTICA (version 10, Statsoft Inc., Tulsa, USA).
3. Results and discussion
3.1. Growth characteristics
As expected, we observed higher head mass (FM) with plants
cultivated at 1000 ppm compared to 200 ppm CO
2
concentration,
in both cultivars and at both harvest dates (Fig. 1). Compared to
the 200 ppm treatment, plants from the 1000 ppm treatment on
average gained a 72% higher head mass. Hunt et al. (1984) reported
738 C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745

a 30% higher absolute crop growth rate for lettuce under elevated
CO
2
concentration. We observed a larger gain in head mass. This is
partly due to the fact that our low CO
2
concentration of 200 ppm
lies way below the commonly used ‘‘ambient” CO
2
concentration
of about 335 ppm in these experiments. The main explanation,
however, is probably that the CO
2
concentration of 1000 ppm in
early growth stages resulted in increased leaf area due to increased
photosynthesis. This increased leaf area in turn potentisized the
effect of the elevated CO
2
concentrations in later growth stages.
As an approximation, at 11 DAP the mean head diameters were
on average over both cultivars 18.3 and 19.0 cm at 200 and
1000 ppm CO
2
concentration, respectively. At 25 DAP, head diam-
eters were significantly different between plants cultivated at
CO
2
concentrations of 200 ppm all the time or only the 14 days
prior to harvest (27.1 and 26.2 cm) and plants cultivated at
1000 ppm all the time or only the 14 days prior to harvest (29.7
and 29.1 cm; data not shown). Head mass in red Lollo of both
exchange treatments (1000_200 ppm and 200_1000 ppm) did not
differ from each other and lay between those of plants perma-
nently cultivated at 200 ppm and at 1000 ppm CO
2
concentration
(Fig. 1). Regarding Oak Leaf, head mass of plants from the
1000_200 ppm and 200_1000 ppm treatments, respectively, did
not differ from plants cultivated all the time at 200 and
1000 ppm CO
2
concentration, respectively (Fig. 1).
Shoot dry matter content of neither Oak Leaf nor Lollo was sig-
nificantly influenced by CO
2
concentration, neither at 11 nor at
25 DAP. Mean shoot dry matter at 11 DAP was 66 mg g
1
(Oak
Leaf) and 62 mg g
1
(Lollo) while it was 56 mg g
1
(Oak Leaf) and
50 mg g
1
(Lollo) at 25 DAP (data not shown).
The shoot/root-ratio on a dry matter basis of neither Oak Leaf
nor Lollo was significantly influenced by CO
2
concentration, nei-
ther at 11 nor at 25 DAP. Mean shoot/root-ratio at 11 DAP was
7.7 g g
1
(Oak Leaf) and 7.9 g g
1
(Lollo) while it was 9.9 g g
1
(Oak Leaf) and 11.7 g g
1
(Lollo) at 25 DAP (data not shown).
Because of this, root fresh and dry matter content are not shown
here.
3.2. Phenolic compounds
In our HPLC-DAD-ESI-MS
3
analyses of phenolic compounds in
red leaf lettuce, we identified two quercetin glycosides, one lute-
olin glycoside, one cyanidin glycoside, and several caffeic acid
derivatives. The main phenolic compound was chicoric acid
(di-O-caffeoyltartaric acid), followed by chlorogenic acid (5-O-
caffeoylquinic acid), quercetin-3-O-(6
00
-O-malonyl)-glucoside,
caffeoylmalic acid, cyanidin-3-O-(6
00
-O-malonyl)-glucoside, caf-
taric acid (caffeoyltartaric acid), quercetin-3-O-glucuronide,
isochlorogenic acid (di-O-caffeoylquinic acid) and luteolin-7-O-
glucuronide. These compounds were previously reported for red
leaf lettuce (Llorach et al., 2008).
3.2.1. Flavonoid glycosides
Results of the regression analysis are depicted in Fig. 2. Equa-
tions of the regression curves and their coefficients of determina-
tion (R
2
) as well as the p-values of the two involved factors are
given in Table 1. Two regression curves are depicted in each graph:
one for 200 and one for 1000 ppm. In both Oak Leaf and Lollo let-
tuce, cyanidin-3-O-(6
00
-O-malonyl)-glucoside was higher in plants
grown at 1000 compared to 200 ppm CO
2
concentration and
decreased with increasing physiological plant age (Table 1). The
same is true for both quercetin glycosides in both cultivars.
Luteolin-7-O-glucuronide concentration likewise decreased with
increasing physiological plant age in both cultivars. However, it
was only positively influenced by CO
2
concentration in Oak Leaf
lettuce which also had higher concentrations than Lollo. Our
results on glycosides of the flavonol quercetin and the anthocyani-
din cyanidin are in line with results reported on birch leaves and
strawberries (Kuokkanen et al., 2001; Peltonen et al., 2005; Wang
et al., 2003). While flavones decreased in birch leaves with increas-
ing CO
2
concentration (Peltonen et al., 2005), in our experiment the
luteolin glycoside was only significantly affected by CO
2
concentra-
tion in one of two cultivars.
The concentration of flavonoid glycosides was mostly influ-
enced by the CO
2
concentration in the 10 days prior to harvest. This
is demonstrated by the position of the open symbols (Fig. 2). This is
underlined by the fact that regression analyses with the mean CO
2
concentration over the growing period resulted in lower coeffi-
cients of determination (data not shown). This pattern corresponds
with results from a previous experiment where flavonoid glycoside
concentrations were determined by light intensity in the 10 days
directly before harvest rather than by previous conditions
(Becker et al., 2013).
Decreasing flavonoid concentrations with increasing physiolog-
ical plant age (Fig. 2) have been observed in lettuce previously
(Becker et al., 2014b).
Our results show a general enhancement of flavonoid biosyn-
thesis in plants grown at high CO
2
concentrations which is in line
with DeLucia et al. (2012). This effect appears to be especially
strong regarding the cyanidin glycoside (Fig. 2). The results indi-
cate that this is due to the aglycone rather than the glycosidic
Fig. 1. Effect of CO
2
concentration (ppm) on mean head mass in gram fresh matter (g FM) of two red leaf lettuce cultivars (Oak Leaf and Lollo). The grey line represents plants
cultivated at CO
2
concentrations of 200 ppm, the black line represents plants cultivated at 1000 ppm all the time. Broken lines represent plants that have been exchanged
between treatments and thus been cultivated first at 200 then at 1000 ppm (broken black line) or first at 1000 then at 200 ppm CO
2
concentration (broken grey line). Identical
letters indicate that no significant differences were detected between these treatments. Both harvest dates were evaluated separately (Tukey-test,
a
= 0.05).
C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745 739

sugar moiety: The two quercetin glycosides (glucuronide and
malonylglucoside) responded similarly to each other while the
response of the quercetin and cyanidin malonylglucosides differed.
Anthocyanin concentrations are high when radiation intensity
is high, temperature is low or nutrients are deficient (Becker,
Klaering, Kroh, & Krumbein, 2014a; Becker et al., 2014b; Peng
et al., 2008). Neither condition applies to our experiment. Accumu-
lation of flavonoids often coincides with conditions that enhance
the formation of reactive oxygen species in plants (Agati et al.,
2013). However, high CO
2
concentration is said to decrease the
oxidative load in plants through unknown mechanisms (Farfan-
Vignolo & Asard, 2012). The photoprotection theory suggests
anthocyanins to be accumulated by plants to protect photosyn-
thetic tissue by absorbing photosynthetically active radiation
and/or scavenging reactive oxygen species or other radicals and
it often serves as an explanation for anthocyanin accumulation
Fig. 2. Effect of CO
2
concentration (ppm) and head mass (g fresh matter) on the concentration of flavonoid glycosides per gram dry matter (g DM) in two red leaf lettuce
cultivars (Oak Leaf and Lollo). Grey symbols represent plants cultivated at CO
2
concentrations of 200 ppm, black symbols represent plants cultivated at 1000 ppm all the time.
White symbols represent plants that have been exchanged between treatments and thus been cultivated first at 1000 then at 200 ppm (diamonds) or first at 200 then at
1000 ppm (circles). The effect of CO
2
concentration and head mass on flavonoid glycoside concentrations was evaluated via multiple regression analysis (equations in
Table 1). The solid and broken lines depict the flavonoid glycoside concentrations calculated with the regression equation for 1000 and 200 ppm CO
2
concentration,
respectively. Asterisks mark significant differences between the CO
2
treatments Cy3MG: cyanidin-3-O-(6
00
-O-malonyl)-glucoside, Q3MG: quercetin-3-O-(6
00
-O-malonyl)-
glucoside, Q3Gc: quercetin-3-O-glucuronide, L7Gc: luteolin-7-O-glucuronide.
740 C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745

(Carpenter, Keidel, Pihl, & Hughes, 2014; Gould, 2004). However, it
does not seem to apply here. The observed increase in anthocyanin
concentration in our experiment, is more likely to be related to
high sugar concentrations as described in Section 3.3.
3.2.2. Caffeic acid derivatives
Results of the regression analysis are depicted in Fig. 3 with one
curve per CO
2
concentration applied. Equations of the regression
curves and their coefficients of determination as well as the
p-values of the two involved factors are given in Table 1.
The concentration of chicoric acid was significantly higher in
Oak Leaf lettuce cultivated at 1000 compared to 200 ppm CO
2
concentration. In Lollo, its concentration was not significantly
influenced by the CO
2
concentration. In both cultivars, the
concentration significantly decreased with increasing head mass.
Chlorogenic acid concentrations were higher in 1000 compared
to 200 ppm CO
2
concentration and decreased with increasing head
mass, in both cultivars. Caffeoylmalic acid concentration in Oak
Leaf was neither influenced by the CO
2
concentration nor by head
mass. In Lollo, caffeoylmalic acid concentration was higher in
plants cultivated at 1000 compared to 200 ppm CO
2
concentration
and, remarkably, increased with increasing head mass. In both cul-
tivars, the concentration of caftaric acid was not influenced by CO
2
concentration but decreased with increasing head mass. Isochloro-
genic acid concentrations were higher in plants cultivated at 1000
compared to 200 ppm CO
2
concentration and decreased with
increasing head mass. Hence, in both cultivars 3 out of 5 caffeic
acid derivatives were positively influenced by increasing CO
2
con-
centration which is in line with the results of Peltonen et al. (2005)
obtained on birch. To our knowledge, the response of single foliar
phenolic acids to elevated CO
2
concentrations has not been studied
before. However, they are contributing to the total phenolics con-
centration which was also observed to increases in response to ele-
vated CO
2
concentration in wheat leaves but not in orange or pine
trees (Peñuelas et al., 1996). The variance among the single caffeic
acid derivatives we detected in our experiment underlines the sig-
nificance of detailed measurements, preferably using HPLC, instead
of total phenolics.
The decreasing concentrations of most caffeic acid derivatives
with increasing head mass are in line with previous results
(Becker et al., 2014b).
Chicoric, chlorogenic, caffeoylmalic, caftaric and isochlorogenic
acid resemble each other in their structure. Caftaric acid is an ester
of caffeic acid and tartaric acid while chicoric acid comprises one
more caffeoyl moiety. Chlorogenic acid is an ester of caffeic acid
and quinic acid while isochlorogenic acid comprises one more caf-
feoyl moiety. Caffeoylmalic acid is an ester of one caffeic acid and
one malic acid. In Lollo, chicoric and caftaric acid showed a uniform
response to CO
2
concentration and plant age, as did chlorogenic
and isochlorogenic acid. In Oak Leaf, only the latter showed a uni-
form response. Nevertheless, the acid bound to the caffeic acid
moiety seems to be more influential on the response regarding
CO
2
concentration and head mass than the number of caffeic acid
moieties in the molecule: We detected three caffeic acid deriva-
tives comprising only one caffeic acid moiety but their responses
to the studied factors resembled each other less than that of those
with either tartaric or quinic acid.
The response of caffeoylmalic acid is unlike the other caffeic
acid derivatives. In Oak Leaf, the factors we studied were obviously
not of influence. In Lollo, its concentration increased with plant
age. We did not observe a similar response with any of the other
phenolic compounds.
Except for their ability to absorb ultraviolet radiation and their
antioxidant activity, not much is known about the function of caf-
feic acid derivatives and neither of these characteristics is likely to
account for the observed effects in our experiment.
3.3. Sugar concentrations
Results of the regression analysis are depicted in Fig. 4, with one
curve per CO
2
concentration applied. Equations of the regression
curves and their coefficients of determination as well as the
p-values of the two involved factors are given in Table 2. In both
cultivars, sugar concentrations were significantly higher in plants
cultivated at 1000 compared to 200 ppm CO
2
concentration. The
response of sucrose, glucose, and fructose concentrations was
uniform within each cultivar (Fig. 4).
Head mass had a significant influence on sugar concentration
regarding Lollo but not regarding Oak Leaf lettuce (Table 2). In
our experiment, high flavonoid concentrations coincide with high
sugar concentrations. This is in line with the hypothesis that high
CO
2
concentrations increase precursor availability for flavonoid
Table 1
Effects of CO
2
concentration during the 10 days before harvest (CO
2
, ppm) and head mass at harvest (M
head
, g) on the concentration of flavonoid glycosides and caffeic acid
derivatives (mg (g DM)
1
). Coefficients were estimated using quasilinear regression analysis separately for each cultivar based on all samples harvested 11 and 25 DAP. Cy3MG:
cyanidin-3-O-(600-O-malonyl)-glucoside, Q3MG: quercetin-3-O-(600 -O-malonyl)-glucoside, Q3Gc: quercetin-3-O-glucuronide, L7Gc: luteolin-7-O-glucuronide. R
2
denotes the
coefficient of determination.
Cultivar Compound Equation of regression curve R
2
p-Value for
CO
2
M
head
Oak Leaf Cy3MG =1.56 + 0.000402 * CO
2
0.242 * ln(M
head
+ 1) 0.72 <0.00001 <0.00001
Q3MG =4.19 + 0.000405 * CO
2
0.772 * ln(M
head
+ 1) 0.88 0.00003 <0.00001
Q3Gc =0.71 + 0.000065 * CO
2
0.135 * ln(M
head
+ 1) 0.85 0.0005 <0.00001
L7Gc =0.37 + 0.000049 * CO
2
0.069 * ln(M
head
+ 1) 0.82 0.00001 <0.00001
Chicoric acid =22.34 + 0.00266 * CO
2
3.942 * ln(M
head
+ 1) 0.77 0.0001 <0.00001
Chlorogenic acid =4.46 + 0.000625 * CO
2
0.596 * ln(M
head
+ 1) 0.32 0.018 0.00004
Caffeoylmalic acid =0.38 + 0.000027 * CO
2
+ 0.082 * ln(M
head
+ 1) 0.06 0.745 0.060
Caftaric acid =0.62 + 0.000016 * CO
2
0.068 * ln(M
head
+ 1) 0.50 0.422 0.00000
Isochlorogenic acid =1.36 + 0.000199 * CO
2
0.282 * ln(M
head
+ 1) 0.91 <0.00001 <0.00001
Lollo Cy3MG =0.46 + 0.000159 * CO
2
0.053 * ln(M
head
+ 1) 0.43 <0.00001 0.0004
Q3MG =5.92 + 0.000529 * CO
2
0.979 * ln(M
head
+ 1) 0.80 0.001 <0.00001
Q3Gc =1.91 + 0.000169 * CO
2
0.345 * ln(M
head
+ 1) 0.82 0.001 <0.00001
L7Gc =0.20 + 0.000005 * CO
2
0.037 * ln(M
head
+ 1) 0.84 0.332 <0.00001
Chicoric acid =23.62 + 0.00049 * CO
2
3.476 * ln(M
head
+ 1) 0.72 0.459 <0.00001
Chlorogenic acid =3.63 + 0.000771 * CO
2
0.439 * ln(M
head
+ 1) 0.51 0.00001 <0.00001
Caffeoylmalic acid =0.21 + 0.000218 * CO
2
+ 0.060 * ln(M
head
+ 1) 0.53 <0.00001 0.003
Caftaric acid =0.82–0.000029 * CO
2
0.091 * ln(M
head
+ 1) 0.58 0.252 <0.00001
Isochlorogenic acid =0.42 + 0.000102 * CO
2
0.070 * ln(M
head
+ 1) 0.56 0.00001 <0.00001
C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745 741

Fig. 3. Effect of CO
2
concentration (ppm) and head mass (g fresh matter) on the concentration of caffeic acid derivatives per gram dry matter (g DM) in two red leaf lettuce
cultivars (Oak Leaf and Lollo). Grey symbols represent plants cultivated at CO
2
concentrations of 200 ppm, black symbols represent plants cultivated at 1000 ppm all the time.
White symbols represent plants that have been exchanged between treatments and thus been cultivated first at 1000 then at 200 ppm (diamonds) or first at 200 then at
1000 ppm (circles). The effect of CO
2
concentration and head mass on caffeic acid derivative concentrations was evaluated via multiple regression analysis (equations in
Table 1). The solid and broken lines depict the caffeic acid derivative concentrations calculated with the regression equation for 1000 and 200 ppm CO
2
concentration,
respectively. Asterisks mark significant differences between the CO
2
treatments.
742 C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745

biosynthesis which results in high flavonoid concentrations: Zhang
et al. (2013) proposed that carbohydrates trigger anthocyanin
biosynthesis in Begonia semperflorens. They suggested excess car-
bohydrates to be the proximate trigger of leaves reddening in
autumn when carbohydrates are accumulated for storage. Sucrose
is reported to directly induce anthocyanin biosynthesis in A. thali-
ana (Solfanelli et al., 2006). Glucose and fructose can feed the pen-
tose phosphate pathway which provides erythrose-4-phosphate, a
precursor for the phenylalanine producing shikimate pathway,
eventually leading to flavonoid aglycone biosynthesis (Jaafar
et al., 2012; Schopfer & Brennicke, 2010). Additionally, another
glucose molecule is consumed to form glycosides.
Flavonols, flavones, caffeic acid derivatives share various
biosynthetic steps with anthocyanins. Precursor abundancy may
therefore well have had the same effect on their biosynthesis like
hypothesized for anthocyanins.
Fig. 4. Effect of CO
2
concentration (ppm) and head mass (g fresh matter) on the concentration of sugars per gram dry matter (g DM) in two red leaf lettuce cultivars (Oak Leaf
and Lollo). Grey symbols represent plants cultivated at CO
2
concentrations of 200 ppm, black symbols represent plants cultivated at 1000 ppm all the time. White symbols
represent plants that have been exchanged between treatments and thus been cultivated first at 1000 then at 200 ppm (diamonds) or first at 200 then at 1000 ppm (circles).
The effect of CO
2
concentration and head mass on sugar concentrations was evaluated via multiple regression analysis. The solid and broken lines depict the sugar
concentrations calculated with the regression equation for 1000 and 200 ppm CO
2
concentration, respectively. Asterisks mark significant differences between the CO
2
treatments.
C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745 743

Yet, incorporating carbon into phenolic compounds in times of
plenty appears to be a one-way-street. There is no mechanism
known how plants could retrieve these carbon atoms in meager
times.
4. Summary and conclusions
The results partly confirm our first hypothesis that high CO
2
concentration has a positive effect on the concentration of flavo-
noid glycosides, except for luteolin-7-O-glucuronide in red Lollo,
and on 3 out of 5 caffeic acid derivatives, but different ones in each
cultivar. We could fully confirm our second hypothesis: the CO
2
concentration shortly before harvest determines the concentration
of flavonoid glycosides and caffeic acid derivatives. Additionally,
our results support the third hypothesis that high CO
2
concentra-
tions increase the availability of precursors for the biosynthesis
of phenolic compounds. From a practical viewpoint, this shows
the relevance of CO
2
enrichment of the atmosphere in greenhouses
and plant factories in order to increase the concentration of health
promoting phenolic compounds in lettuce. Ecologically inter-
preted, however, we were not able to explain our results with
existing theories on the antioxidative or photoprotective mode of
action of flavonoids and caffeic acid derivatives in plants. They
may rather act as sinks for copious amounts of photosynthates.
More detailed research into lettuce physiology is necessary here.
Acknowledgments
This research was supported by the Federal Ministry for
Environment, Nature Conservation and Nuclear Safety and the
Rentenbank managed by the Federal Ministry of Food, Agriculture
and Consumer Protection with the assistance of the Federal Agency
for Agriculture and Food. We want to thank Dr. Angelika Krumbein
very much for all her valuable input and the good cooperation. We
are furthermore very grateful for the help of Angela Schmidt,
Manuel Eduardo Porras Sanchez, Archontia Karachasani, and Ingo
Hauschild.
References
Agati, G., Brunetti, C., Di Ferdinando, M., Ferrini, F., Pollastri, S., & Tattini, M. (2013).
Functional roles of flavonoids in photoprotection: New evidence, lessons from
the past. Plant Physiology and Biochemistry, 72, 35–45.
Becker, C., Klaering, H.-P., Kroh, L. W., & Krumbein, A. (2013). Temporary reduction
of radiation does not permanently reduce flavonoid glycosides and phenolic
acids in red lettuce. Plant Physiology and Biochemistry, 72, 154–160.
Becker, C., Klaering, H.-P., Kroh, L. W., & Krumbein, A. (2014a). Cool-cultivated red
leaf lettuce accumulates cyanidin-3-O-(6
00
-O-malonyl)-glucoside and
caffeoylmalic acid. Food Chemistry, 146, 404–411.
Becker, C., Klaering, H.-P., Schreiner, M., Kroh, L. W., & Krumbein, A. (2014b). Unlike
quercetin glycosides, cyanidin glycoside in red leaf lettuce responds more
sensitively to increasing low radiation intensity before than after head
formation has started. Journal of Agricultural and Food Chemistry, 62, 6911–6917.
Bindi, M., Fibbi, L., & Miglietta, F. (2001). Free air CO
2
enrichment (FACE) of
grapevine (Vitis vinifera L.): II. Growth and quality of grape and wine in response
to elevated CO
2
concentrations. European Journal of Agronomy, 14, 145–155.
Boudet, A. M. (2007). Evolution and current status of research in phenolic
compounds. Phytochemistry, 68, 2722–2735.
Caporn, S. J. M. (1989). The effects of oxides of nitrogen and carbon dioxide
enrichment on photosynthesis and growth of lettuce (Lactuca sativa L.). New
Phytologist, 111, 473–481.
Carpenter, K. L., Keidel, T. S., Pihl, M. C., & Hughes, N. M. (2014). Support for a
photoprotective function of winter leaf reddening in nitrogen-deficient
individuals of Lonicera japonica.Molecules, 19, 17810–17828.
Chalabi, Z. S., Biro, A., Bailey, B. J., Aikman, D. P., & Cockshull, K. E. (2002). Optimal
control strategies for carbon dioxide enrichment in greenhouse tomato crops –
Part 1: Using pure carbon dioxide. Biosystems Engineering, 81, 421–431.
De Gelder, A., Dieleman, J. A., Bot, G. P. A., & Marcelis, L. F. M. (2012). An overview of
climate and crop yield in closed greenhouses. Journal of Horticultural Science and
Biotechnology, 87, 193–202.
DeLucia, E. H., Nabity, P. D., Zavala, J. A., & Berenbaum, M. R. (2012). Climate change:
Resetting plant–insect interactions. Plant Physiology, 160, 1677–1685.
DuPont, S. M., Mondin, Z., Williamson, G., & Price, K. R. (2000). Effect of variety,
processing and storage on the flavonoid glycoside content and composition of
lettuce and endive. Journal of Agricultural and Food Chemistry, 48, 3957–3964.
Farfan-Vignolo, E. R., & Asard, H. (2012). Effect of elevated CO
2
and temperature on
the oxidative stress response to drought in Lolium perenne L. and Medicago
sativa L. Plant Physiology and Biochemistry, 59, 55–62.
Goufo, P., Pereira, J., Moutinho-Pereira, J., Correia, C. M., Figueiredo, N., Carranca, C.,
... Trindade, H. (2014). Rice (Oryza sativa L.) phenolic compounds under elevated
carbon dioxide (CO
2
) concentration. Environmental and Experimental Botany, 99,
28–37.
Gould, K. S. (2004). Nature’s swiss army knife: The diverse protective roles of
anthocyanins in leaves. Journal of Biomedicine and Biotechnology, 5, 314–320.
Gould, K.S., & Lister, C. (2006). Flavonoid functions in plants. In: £.M. Anderson & K.
R. Markham (Eds.), Flavonoids: Chemistry, biochemistry and applications (pp.
397–441).
Hunt, R., Wilson, J. W., Hand, D. W., & Sweeney, D. G. (1984). Integrated analysis of
growth and light interception in winter lettuce I. Analytical methods and
environmental influences. Annals of Botany, 54, 743–757.
IPCC (2007). Climate change 2007: The physical science basis. Contribution of
working group I. In S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. B.
Averyt, M. Tignor, & H. L. Miller (Eds.). Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge
University Press.
Jaafar, H. Z. E., Ibrahim, M. H., & Karimi, E. (2012). Phenolics and flavonoids
compounds, phenylanine ammonia lyase and antioxidant activity responses to
elevated CO
2
in Labisia pumila (Myrisinaceae). Molecules, 17, 6331–6347.
Klopotek, Y., & Kläring, H. P. (2014). Accumulation and remobilisation of sugar and
starch in the leaves of young tomato plants in response to temperature. Scientia
Horticulturae, 180, 262–267.
Kläring, H. P., Hauschild, C., Heißner, A., & Bar-Yosef, B. (2007). Model-based control
of CO
2
concentration in greenhouses at ambient levels increases cucumber
yield. Agricultural and Forest Meteorology, 143, 208–216.
Krumbein, A., Schwarz, D., & Kläring, H.-P. (2006). Effects of environmental factors
on carotenoid content in tomato (Lycopersicon esculentum L. Mill.) grown in a
greenhouse. Journal of Applied Botany and Food Quality, 80, 160–164.
Kuokkanen, K., Julkunen-Tiitto, R., Keinänen, M., Niemelä, P., & Tahvanainen, J.
(2001). The effect of elevated CO
2
and temperature on the secondary chemistry
of Betula pendula seedlings. Trees, 15, 378–384.
Llorach, R., Martínez-Sánchez, A., Tomás-Barberán, F. A., Gil, M. I., & Ferreres, F.
(2008). Characterisation of polyphenols and antioxidant properties of five
lettuce varieties and escarole. Food Chemistry, 108, 1028–1038.
Mulabagal, V., Ngouajio, M., Nair, A., Zhang, Y., Gottumukkala, A. L., & Nair, M. G.
(2010). In vitro evaluation of red and green lettuce (Lactuca sativa) for functional
food properties. Food Chemistry, 118, 300–306.
Nederhoff, E. M., & Vegter, J. G. (1994). Photosynthesis of stands of tomato,
cucumber and sweet pepper measured in greenhouses under various CO
2
concentrations. Annals of Botany, 73, 353–361.
Table 2
Effects of CO
2
concentration during the 10 days before harvest (CO
2
, ppm) and head mass at harvest (M
head
, g) on the concentration of sugars (mg (g DM)
1
). Coefficients were
estimated using quasilinear regression analysis separately for each cultivar based on all samples harvested 11 and 25 DAP. R
2
denotes the coefficient of determination.
Cultivar Compound Equation of regression curve R
2
p-Value for
CO
2
M
head
Oak Leaf Total sugars =76.1 + 0.0312 * CO
2
1.74 * ln(M
head
+ 1) 0.49 <0.00001 0.500
Glucose =13.2 + 0.00695 * CO
2
+ 0.83 * ln(M
head
+ 1) 0.40 0.00005 0.283
Fructose =32.6 + 0.0138 * CO
2
1.36 * ln(M
head
+ 1) 0.36 0.00002 0.357
Sucrose =30.3 + 0.0105 * CO
2
1.22 * ln(M
head
+ 1) 0.30 0.00016 0.344
Lollo Total sugars =23.8 + 0.0346 * CO
2
+ 20.5 * ln(M
head
+ 1) 0.66 <0.00001 <0.00001
Glucose =10.4 + 0.0112 * CO
2
+ 10.5 * ln(M
head
+ 1) 0.40 0.00028 <0.00001
Fructose =31.2 + 0.0147 * CO
2
+ 4.93 * ln(M
head
+ 1) 0.31 0.00232 0.0326
Sucrose =2.98 + 0.00873 * CO
2
+ 5.15 * ln(M
head
+ 1) 0.32 0.0140 0.00361
744 C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745

Peltonen, P. A., Vapaavuori, E., & Julkunen-Tiitto, R. (2005). Accumulation of
phenolic compounds in birch leaves is changed by elevated carbon dioxide and
ozone. Global Change Biology, 11, 1305–1324.
Peng, M., Hudson, D., Schofield, A., Tsao, R., Yang, R., Gu, H., ... Rothstein, S. J. (2008).
Adaptation of Arabidopsis to nitrogen limitation involves induction of
anthocyanin synthesis which is controlled by the NLA gene. Journal of
Experimental Botany, 59, 2933–2944.
Peñuelas, J., Estiarte, M., Kimball, B. A., Idso, S. B., Pinter, P. J., Wall, G. M., ... Hensrik,
D. L. (1996). Variety of responses of plant phenolic concentration to CO
2
enrichment. Journal of Experimental Botany, 47, 1463–1467.
Rietjens, I. M. C. M., Boersma, M. G., van der Woude, H., Jeurissen, S. M. F., Schutte,
M. E., & Alink, G. M. (2005). Flavonoids and alkenylbenzenes: Mechanisms of
mutagenic action and carcinogenic risk. Mutation Research/Fundamental and
Molecular Mechanisms of Mutagenesis, 574, 124–138.
Rimbach, G., Melchin, M., Moehring, J., & Wagner, A. E. (2009). Polyphenols from
cocoa and vascular health – A critical review. International Journal of Molecular
Sciences, 10, 4290–4309.
Sallas, L., Luomala, E.-M., Utriainen, J., Kainulainen, P., & Holopainen, J. K. (2003).
Contrasting effects of elevated carbon dioxide concentration and temperature
on Rubisco activity, chlorophyll fluorescence, needle ultrastructure and
secondary metabolites in conifer seedlings. Tree Physiology, 23, 97–108.
Scattino, C., Castagna, A., Neugart, S., Chan, H. M., Schreiner, M., Crisosto, C. H., ...
Ranieri, A. (2014). Post-harvest UV-B irradiation induces changes of phenol
contents and corresponding biosynthetic gene expression in peaches and
nectarines. Food Chemistry, 163, 51–60.
Schmidt, U., Huber, C., Dannehl, D., Rocksch, T., Tantau, H. J., & Meyer, J. (2011).
Effect of special climate conditions in closed greenhouses on coefficient of
performance and plant growth – Preliminary tests for optimizing closed
greenhouse control. Acta Horticulturae, 893, 429–436.
Schonhof, I., Kläring, H.-P., Krumbein, A., & Schreiner, M. (2007). Interaction
between atmospheric CO
2
and glucosinolates in broccoli. Journal of Chemical
Ecology, 33, 105–114.
Schopfer, P., & Brennicke, A. (2010). Pflanzenphysiologie (Vol. 7). Heidelberg:
Spektrum Akademischer Verlag.
Solfanelli, C., Poggi, A., Loreti, E., Alpi, A., & Perata, P. (2006). Sucrose-specific
induction of the anthocyanin biosynthetic pathway in Arabidopsis.Plant
Physiology, 140, 637–646.
Sonneveld, C., & Straver, N. (1988). Nutrient Solutions for Vegetables and Flowers
Grown in Water or Substrates (vol. 7). Naaldwijk, The Netherlands: Glasshouse
Crops Research Station, p. 16.
Treutter, D. (2010). Managing phenol contents in crop plants by phytochemical
farming and breeding-visions and constraints. International Journal of Molecular
Sciences, 11, 807–857.
Wang, S. Y., Bunce, J. A., & Maas, J. L. (2003). Elevated carbon dioxide increases
contents of antioxidant compounds in field-grown strawberries. Journal of
Agricultural and Food Chemistry, 51, 4315–4320.
Willits, D. H., & Peet, M. M. (1989). Predicting yield responses to different
greenhouse CO
2
enrichment schemes – Cucumbers and tomatoes. Agricultural
and Forest Meteorology, 44, 275–293.
Zhang, K., Li, Z., Li, Y., Li, Y., Kong, D., & Wu, R. (2013). Carbohydrate accumulation
may be the proximate trigger of anthocyanin biosynthesis under autumn
conditions in Begonia semperflorens.Plant Biology, 15, 991–1000.
C. Becker, H.-P. Kläring / Food Chemistry 199 (2016) 736–745 745