Balancing the energy flow from captured light to biomass under fluctuating light conditions

Authors


Author for correspondence: T. Jakob Tel: +49 (0)341 9736873 Fax: +49 (0)341 9736899 Email: tjakob@rz.uni-leipzig.de

Summary

  • • The balance of energy flow from light absorption into biomass was investigated under simulated natural light conditions in the diatom Phaeodactylum tricornutum and the green alga Chlorella vulgaris.
  • • The energy balance was quantified by comparative analysis of carbon accumulation in the new biomass with photosynthetic electron transport rates per absorbed quantum, measured both by fluorescence quenching and oxygen production. The difference between fluorescence- and oxygen-based electron flow is defined as ‘alternative electron cycling’.
  • • The photosynthetic efficiency of biomass production was found to be identical for both algae under nonfluctuating light conditions. In a fluctuating light regime, a much higher conversion efficiency of photosynthetic energy into biomass was observed in the diatom compared with the green alga.
  • • The data clearly show that the diatom utilizes a different strategy in the dissipation of excessively absorbed energy compared with the green alga. Consequently, in a fluctuating light climate, the differences between green algae and diatoms in the efficiency of biomass production per photon absorbed are caused by the different amount of alternative electron cycling.

Introduction

In the aquatic environment, phytoplankton experiences strong fluctuations with variations of light, nutrient availability and temperature. Whereas changes in nutrients and temperature occur mostly on timescales of days to seasons, the variability of light can occur on much shorter timescales, ranging from seconds to hours or days (Litchman & Klausmeier, 2001). Even for phytoplankton populations remaining at a constant depth within the euphotic zone, the light environment is changing in response to the daily course of sunlight and seasonal variations in irradiance. Additionally, the vertical mixing of the water column exposes algae to light intensities ranging from complete darkness to full sunlight (MacIntyre et al., 2000).

Among marine planktonic microalgae, the diatoms are the most important group because of their high productivity. They contribute up to 40% of the oceanic primary production (Tréguer et al., 1995) and strongly influence the biogeochemical cycling of silicon, carbon, nitrogen, phosphorus and iron (Buesseler, 1998). Diatoms dominate in well mixed, nutrient-rich waters (Sarthou et al., 2005), where light should be the major factor controlling phytoplankton growth. The seasonal dominance of diatoms is not well understood to date. This dominance is especially interesting because planktonic diatoms are not motile, and thus cannot escape from large vertical circulations within the water body. Certainly, their flexibility in the adjustment of the photosynthetic apparatus in response to light (Ting & Owens, 1986; Lavaud et al., 2004) is a great advantage and allows them to maintain maximal growth rates at a wide range of light intensities (Falkowski & LaRoche, 1991).

Fluctuating light conditions exert a great influence on phytoplankton species composition. In order to estimate phytoplankton growth from photosynthesis rates, it is therefore important to investigate the effects of fluctuating light conditions on photosynthesis and biomass production of microalgae. There have been a number of physiological studies dealing with the effects of fluctuating light regimes on photosynthesis in algae in comparison with nonfluctuating light conditions (Marra, 1978; Kroon et al., 1992b; Fietz & Nicklisch, 2002; Havelková-Doušováet al., 2004). It is not possible to compare their approaches directly because of differences in total incident irradiation, maximal light intensities, day length and frequencies of light regimes. Therefore it is not surprising that the reported physiological adaptations of different algal species to fluctuating light under laboratory conditions vary greatly. There are only a small number of studies on the effects of fluctuations on the growth of algae (Cosper, 1982; Ibelings et al., 1994; Tichýet al., 1995; Nicklisch, 1998; Litchman, 2000; Elliott et al., 2001). Again, differences in changes of growth rates in response to light fluctuations are probably caused by species-specificity, and by the different light regimes used in their approaches. However, it is apparent from direct measurement of growth rates (Nicklisch, 1998; Litchman, 2000; Elliott et al., 2001) and photosynthetic acclimation to fluctuating light regimes (Fietz & Nicklisch, 2002) that diatoms showed less decrease in growth rates under fluctuating light than did green algae or cyanobacteria.

Litchman & Klausmeier (2001) established a model of phytoplankton growth and competition under a wide range of fluctuation frequencies and day lengths, and found that both factors have an impact on growth and therefore on community structure. However, there has been limited modelling of growth rates under fluctuating light derived from growth–irradiance curves under constant light. Litchman (2000) observed differences between predicted and measured growth rates of algae in light regimes with fluctuations ranging from limiting to saturating or inhibiting irradiances (light conditions that should be expected in nature).

Another approach is the prediction of growth rates from the measurement of photosynthetic performance (Micheletti et al., 1998; Davis & Walsby, 2002). Determination of the quantum yield of photosynthesis can be based on measurements of C fixation, oxygen evolution or variable chlorophyll fluorescence. The latter is certainly the fastest method and, unlike 14C incorporation or oxygen-production rate, displays primary photosynthetic transport rates of electrons evolved at photosystem II (PSII) (Kolber & Falkowski, 1993). Nevertheless, it is evident that photosynthetic efficiency (regardless of the method used) does not necessarily correspond to growth efficiency, for several reasons. Whereas variable chlorophyll fluorescence is the measure of the quantum yield of PSII (ΦPSII; Genty et al., 1989), and indicates all electrons released at PSII, oxygen evolution rates at PSII are biased by oxygen-consuming processes such as mitochondrial respiration, the Mehler reaction or photorespiration. Additionally, electron cycling around PSI or PSII drives electron flow detectable by chlorophyll fluorescence, but does not lead to a concomitant increase in oxygen evolution. Other problems arise from C losses caused by the efflux of glycolate in an incomplete photorespiratory metabolism (Leboulanger et al., 1998), or by excretion of carbohydrates (Hoagland et al., 1993), difficulties in the determination of mitochondrial respiration in the light and variations in the photosynthetic quotients (Falkowski et al., 1985; Bender et al., 1987; Carignan et al., 2000). Therefore it is only under ideal conditions that a strong correlation between ΦPSII and the quantum yield of C fixation (inline image) can be observed (Genty et al., 1989). Under nonideal conditions (such as in the natural environment), differences between ΦPSII and oxygen- or C-based measurements of photosynthesis rates are both expected and observed (Geel et al., 1997; Gilbert et al., 2000a; Franklin & Badger, 2001; Longstaff et al., 2002; Figueroa et al., 2003). Gilbert et al. (2000b) showed that these discrepancies seem to be species-specific, with a reliable correlation between fluorescence- and oxygen-based photosynthesis rates in diatoms and cryptophyta, and less reliable correlations in green algae and cyanobacteria. The close correlation between fluorescence and oxygen measurements of photosynthesis in diatoms was further confirmed under natural conditions (Jakob et al., 2005). The deviations between ΦPSII and inline image are the result of the activity of alternative electron pathways, as described above. In higher plants, the combinations of alternative electron pathways are assumed to consume up to 30% of photosynthetic electrons (Asada, 1999; Badger et al., 2000; Wingler et al., 2000). However, there are no quantitative studies of the amount of alternative electrons under different environmental conditions in algae.

In summary, it can be assumed that the acclimation of phytoplankton to dynamic light conditions under nutrient repletion depends on the conversion efficiency of photosynthetic electrons into biomass production, and includes the impact of electron losses by alternative pathways on growth and biomass formation. Therefore a new approach of the current study was to observe and quantify the energy flow from photons absorbed to biomass formed in a comparison of a diatom and a green alga under simulated natural light conditions. Furthermore, in ecological aspects it is more important to investigate the contribution of phytoplankton to the effective biomass (dry matter) production than to compare photosynthesis or growth rates separately. The model presented here is essentially based on a comparison of expected (derived from photosynthesis rates) and actual measured biomass production. The derived photosynthetic efficiency of biomass formation (Φµ) can be compared with energy losses at the level of thermal dissipation and alternative electron pathways. The quantification of alternative electrons was deduced from the difference between fluorescence- and oxygen-based photosynthesis rates. Because we were interested in the effects of light fluctuations on the efficiency of biomass formation, algae were grown in a sine light climate (SL, simulating the daily course of sunlight) and in an oscillating sine light climate (FL, simulating vertical fluctuations within the water column during the day), with the same maximum light level but with different total light doses. This is in contrast to most other studies, where the same daily irradiance was used in the comparison of constant illumination with fluctuations of different degree or frequencies. However, this procedure neglects the additional effect of adaptation to different maximum light intensities. On the other hand, light climates with different daily irradiance require careful estimation of the amount of light absorbed by the algae.

Materials and Methods

Culture conditions

Experiments were performed with the diatom Phaeodactylum tricornutum (SAG 1090-1a, Göttingen, Germany) and the green alga Chlorella vulgaris (SAG 211-11b, Göttingen, Germany). Phaeodactylum tricornutum was grown in a modified artificial seawater (ASP) 2 medium (Provasoli et al., 1975) and C. vulgaris in a Bold's basal medium (BBM) (Bischoff & Bold, 1963) at 20°C.

Algal culture and light climates were performed in a cyclostat (according to Kroon et al., 1992a) with computer-controlled regulation of light supply and at a Chl a concentration of 2 mg l−1 (±0.2). In addition, with the rectangular culture vessel (237 × 450 mm, 30 mm depth) the self-shading effects were kept to a minimum. In the further calculation of photosynthetic activity, it is important that this setup allows an accurate measure of the amount of absorbed radiation (Qphar). Illumination was provided by a high-intensity light source (HQI-1000 W-Daylight, Osram, Germany). A special computer program controlled two light regimes with a day length of 12 h (Fig. 1): (i) a sine light climate (SL, daily photon insolation 19.1 mol m−2); (ii) a sine light climate with superposition of exponential fluctuations (FL, daily photon insolation 4.2 mol m−2). Exponential fluctuations simulated vertical mixing of algae according to the ratio:

Figure 1.

Variations in light conditions applied during a 12-h light period: (a) sine light climate simulating the daily course of sunlight; (b) exponential light climate simulating vertical circulation of algae within the water body.

image(Eqn 1)

where zeu is the depth of the euphotic zone and zm is the mixing depth. The length of one cycle of fluctuation was 30 min. Both light regimes were applied to algal cultures with the same maximum photon irradiance of 980 µmol m−2 s−1 (mean irradiance), and therefore allows comparison with other studies (Nicklisch, 1998). Incident light intensity (PAR, 400–700 nm) was measured at 15 points within the culture vessel using a spherical quantum sensor (Zemoko, Kouderkerke, the Netherlands) in combination with a data display (LI-250, Li-Cor, Lincoln, NE, USA). Mixing of algae within the culture vessel and CO2 supply was achieved by bubbling with air through sintered glass at the bottom of the culture vessel.

Growth rates (µ) were determined with the following equation:

image(Eqn 2)

with:

image(Eqn 3)

where D is the dilution rate, f is dilution volume, V is the volume of the culture vessel, dx is the change in biomass during time interval dt (on the basis of Chl a concentration) and x0 is the biomass before time interval dt. Biomass was always measured at the beginning of the light period.

To achieve full adaptation to the applied light climates, algae were grown for at least 1 wk under SL or FL conditions to yield constant growth rates. All measurements described below were performed on three sampling days.

Pigment analysis

The concentration of Chl a was determined spectrophotometrically in 90% acetone extracts of cells of P. tricornutum using the equations of Jeffrey & Humphrey (1975), and in 80% acetone extracts for C. vulgaris using the equations of Ziegler & Egle (1965). Algal samples (5 ml) were collected on glass-fibre filters (Schleicher and Schuell, Dassel, Germany), 2.5 ml acetone added, and cells broken in a cell homogenizer (Braun, Melsungen, Germany). After centrifugation (1 min at 13 000g, Hermle, Z231M, Gosheim, Germany), absorbance spectra were recorded with a spectrophotometer (Hitachi U2000).

HPLC pigment analysis was performed using a Gynkotek HPLC system equipped with an HPLC pump (Gynkotek M480) and a photodiode array detector (Gynkotek UVD 340S). Algal samples (10 ml) were harvested by filtration, filters were frozen in liquid nitrogen and freeze-dried. Pigments were extracted with 90% methanol/0.2 m ammonium acetate (90/10 v/v) and 10% ethyl acetate, centrifuged for 1 min at 13 000g (Hermle Z 231M) and injected into the HPLC column. Pigment separation for P. tricornutum and C. vulgaris is described by Jakob et al. (1999, 2001), respectively.

Determination of absorbed radiation

A dual beam spectrophotometer (Zeiss M500, Jena, Germany) was used to determine cellular absorption spectra of algae. Fresh medium was used as reference. Chlorophyll a-specific in vivo absorption coefficients, a* (m2 mg−1) were calculated as follows:

image(Eqn 4)

where A is the absorption of the sample, d is the path length of the cuvette (m), 2.3 is the conversion factor from log10 to ln, and Chl is the Chl a concentration (mg m−3).

The emission spectra (400–700 nm) of the light source from the culture unit and from the combined oxygen-/fluorescence-measuring device (see below) were determined using a spectroradiometer (Tristan, Hamburg, Germany).

With knowledge of the emission spectra of the light source and the Chl a-specific absorption, it is possible to estimate the amount of absorbed PAR (Qphar) by the algal culture, based on the following equation (according to Gilbert et al., 2000b):

image(Eqn 5)

where Qphar is the photosynthetically absorbed radiation (µmol m−2 s−1), Q is the photosynthetically available (incident) radiation (µmol m−2 s−1) dependent on the light source, and d is the optical path length (m).

Measurements of oxygen evolution and variable chlorophyll fluorescence

In calculating the amount of alternative electrons as the difference between fluorescence- and oxygen-based photosynthesis rates, it is not sufficient to measure relative values, rather absolute photosynthesis rates per absorbed photons should be measured. This requires careful estimation of Qphar and the simultaneous measurement of oxygen evolution rates and fluorescence-based electron transport. We used the combination of a pulse amplitude modulation (PAM) fluorometer (unit 101/103, Walz, Effeltrich, Germany) and a light pipette equipped with a special cuvette (Topgallant LLC, Salt Lake City, UT, USA) which allows connection of the emitter and detector unit of the PAM fluorometer. The light pipette was used as actinic light source (Xenophot Longlife HLX64642, Osram) and for the measurement of oxygen evolution with a Clark-type electrode (MI730, Microelectrodes Inc., Bedford, NH, USA), whereas the PAM 103 unit in combination with a KL 1500 (Walz, Effeltrich, Germany) delivered the saturating light pulses (3500 µmol photons m−2 s−1).

During the light phase, a sample (20 ml) was taken out of the culture vessel every 60 min. Then 3 ml of this sample were placed in the cuvette of the oxygen-/fluorescence-measuring device to measure photosynthesis–irradiance (P/E) curves (Fig. 2a). Measurements of chlorophyll content, in vivo absorption and pigment analysis by HPLC were taken from the same sample.

Figure 2.

Modelling of photosynthesis and the amount of alternative electrons in the course of the light period. (a) Comparison of fluorescence-based (filled squares) and oxygen-based (open circles) photosynthesis–irradiance curves in Phaeodactylum tricornutum (sample taken at 12:00 h). Data fitting is represented by dashed (fluorescence) and solid (oxygen) lines. (b) Comparison of integrated photosynthesis rates based on fluorescence (dotted line) and oxygen evolution (solid line) data in P. tricornutum under sine light (SL) conditions. (c) Comparison of the amount of alternative electrons under SL conditions in Chlorella vulgaris (dotted line) and P. tricornutum (solid line). (d) Comparison of the amount of alternative electrons under exponentially fluctuating (FL) conditions in C. vulgaris (dotted line) and in P. tricornutum (solid line)

The P/E curves were measured over 50 min using a computer-controlled program of increasing light intensities (each light irradiance with 4 min duration) alternating with subsequent dark phases of the same length. Measurements always started with a 4-min dark period. Gross oxygen production was derived by correcting net oxygen evolution rates by the corresponding dark respiration. Electron transport rates (ETR, µmol electrons m−2 s−1) were calculated using the following equation:

image(Eqn 6)

where ΦPSII is the quantum yield of photosystem II (van Kooten & Snel, 1990), Qphar is the amount of absorbed radiation in the measuring device, and the factor 0.5 is based on the assumption that two quanta are necessary for linear electron transport. Fluorescence measurements were corrected for background signals caused by electronic artefacts.

Additionally, online fluorescence of the algal culture was measured directly at the surface of the culture vessel (according to Lippemeier et al., 2001). For this, the glass-fibre optic of the PAM was mounted behind the culture vessel. To increase the signal, a small piece (2 × 2 cm) of aluminum foil was placed at the same spot on the front side of the culture vessel. Every 60 s a light pulse (0.8 s, 4000 µmol photons m−2 s−1) was given (KL 1500, Schott, Germany) and the variable fluorescence recorded on a computer. Because of the depth of the culture vessel and the strong mixing of algae, it was not possible to achieve complete saturation of PSII reaction centres. Therefore we used the online fluorescence data to estimate relative changes in the nonphotochemical quenching (NPQ) in the course of the light period.

Modelling of integrated photosynthesis rates and biomass formation

Because of differences in the spectra of the actinic light source in the light pipette and in the culture unit, and different optical path lengths between both systems, it was necessary to plot the P/E curves against Qphar. For modelling of photosynthesis rates as a function of changing light intensities, P/E curves were fitted using the dynamic model of Eilers & Peeters (1988). With the derived fit parameters, oxygen evolution and fluorescence-based electron transport rates can be calculated for any given light intensity during the daily course of the light climate (Fig. 2a,b). The data in between the hourly measured photosynthesis rates were derived from averaging the corresponding fit parameters.

Calculation of theoretical oxygen evolution rates, derived from the measurement of variable fluorescence, was based on the assumption that four electrons are required for the evolution of one molecule of oxygen. The theoretical oxygen production (or photosynthesis rate, P) (µmol O2 per mg Chl a h−1) is given by:

image(Eqn 7)

where ETR is the electron transport rate (derived from calculation using the fit parameters), d is the optical path length of the culture vessel (m), and chlorophyll concentration is given in mg m−3.

Algal cultures in the turbidostat were kept at a constant Chl a concentration to assure comparable cellular absorption properties. Further calculations were therefore performed on a Chl a basis. This procedure can be applied because the ratio dry mass : Chl a did not change after acclimation to the different light conditions. From the integrated oxygen- and fluorescence-based photosynthesis rates (mmol O2 per mg Chl a d−1), the theoretical biomass production on a dry matter basis (BO or BF, mg dry matter per mg Chl a d−1) can be calculated:

image(Eqn 8)

where P is the integrated photosynthesis rate; PQ is the photosynthetic quotient (see below); RD and RQD are respiration (mmol O2 per mg Chl a 12 h−1) and respiratory quotient (see below), respectively, in the dark; RL and RQL are respiration (mmol O2 per mg Chl a per 12 h) and respiratory quotient, respectively, during the light period; 12 is the conversion factor from mmol to mg carbon; and (DM : C) is the dimensionless ratio dry mass per carbon content. Respiration rates in the dark were derived from oxygen evolution rates measured before the start of the light period in the culture unit.

The measured biomass production Bµ (mg dry matter per mg Chl a d−1) was determined using the following equation:

image(Eqn 9)

where DM is the dry mass (mg per Chl a) and µ is the daily growth rate (d−1).

Calculation of the photosynthetic quantum efficiency of biomass production Φµ (mol mol−1) represents the number of quanta converted into biomass in relation to the amount of photosynthetically active quanta, and was based on the assumption that the fluorescence-based photosynthesis rate is the maximum amount of electrons transported through the electron transport chain. Φµ was derived from the following equation:

image(Eqn 10)

where PF is the integrated fluorescence-based photosynthesis rate (expressed as oxygen evolution) per day, 8 is the minimum quantum requirement for the fixation of one molecule of CO2, (BF : Bµ) is the ratio of the maximum (theoretical) to the measured biomass production, and Qphar is the amount of absorbed quanta per day.

Measurements of gas exchange and elemental composition

To calculate the photosynthetic and respiratory quotient, simultaneous measurements of CO2/O2 gas exchange were performed using a respirometer (Biometric Systems, Weiterstadt, Germany). Five samples of algal suspension (20 ml each) were taken from the culture vessel at the beginning of the light phase and collected on glass-fibre filters (54 mm diameter), taking care that the filter did not become dry. The filters were placed in glass bottles (100 ml, Duran, Schott, Germany) and connected to the respirometer system. To prevent heating of the samples, a water jacket was placed between lamps and bottles and adjusted to 20°C. One bottle was kept in darkness, and the others were illuminated from beneath (halogen lamp, Kandolite 35 W, Osram) with photon irradiances ranging from 50 to 600 µmol m−2 s−1. Mean values of gas exchange were calculated from the data for five measuring cycles. After switching off the light again, five measuring cycles were performed to measure the respiratory gas exchange directly after light treatment. PQ and RQ were calculated using the following equations:

image(Eqn 11)
image(Eqn 12)

where [OER] is the absolute O2-exchange rate (µl h−1) and CER is the absolute CO2-exchange rate (µl h−1). The respiratory quotient in the dark (RQD) was derived from the sample kept in darkness, and RQL (RQ in the light) was derived from the darkened samples after illumination.

The elemental composition was measured with a CHNSO analyser (vario EL, Analytik Jena GmbH, Germany). Cell suspension (75–130 ml, corresponding to 8–10 mg organic dry matter) were harvested by centrifugation, washed three times with distilled water and freeze-dried. To trace changes in elemental composition throughout a daily course, algal samples were taken before, in the middle and at the end of the light periods.

Results and Discussion

Physiological parameters of adaptation to fluctuating light conditions

In this first section, we demonstrate that the two model organisms used in this investigation, the diatom P. tricornutum and the green alga C. vulgaris, showed typical physiological adaptations to fluctuating light conditions, as reported previously in other studies. This forms the basis for a wider application of the conclusions derived from the analysis of growth balance.

Daily growth rates of both algae were lower under FL than under SL conditions. This is mainly because of the 4.5 times higher daily light dose in SL compared with FL. Surprisingly, the growth rates in SL are only twice as high as in FL. This can be ascribed to the average PAR. The daily incident PAR corresponds to an average irradiance of 443 µmol m−2 s−1 in SL, and to 96 µmol m−2 s−1 in FL. However, growth rates of green algae and diatoms are usually saturated at photon irradiances of 200–300 µmol m−2 s−1 (Geider et al., 1985; Gilbert et al., 2000b; Litchman, 2000; Litchman et al., 2003). Therefore irradiances above the given value do not increase growth rates further. The decrease in growth rates of C. vulgaris in response to lower daily photon insolation in FL compared with SL is comparable with Dunaniella tertiolecta under dynamic light conditions (Havelková-Doušováet al., 2004). Wilson & Huner (2000) found only a small increase of growth rates of C. vulgaris from c. 0.73 to 0.87 d−1 in response to constant illumination, with PAR increasing from 100 to 400 µmol m−2 s−1, respectively. The growth rate at 400 µmol photons m−2 s−1 is almost identical to the growth rate of C. vulgaris under SL observed in the present study (Table 1). However, growth rate in FL achieved only 50% of the values given by Wilson & Huner (2000). The same ratio holds true for the comparison with data derived from a study on Sphaerocystis schroeteri (Litchman, 2000). Therefore in C. vulgaris, as in other green algae, fluctuating light climates tend to decrease growth rates in comparison with nonfluctuating light conditions.

Table 1.  Physiological parameters measured in Phaeodactylum tricornutum and Chlorella vulgaris suspended in exponentially fluctuating (FL) or sine light (SL) conditions
ParameterP. tricornutumC. vulgaris
FLSLFLSL
  1. Mean values ± SD (of at least three independent replicates) of Chl a specific absorption a*; amount of absorbed light quanta Qphar; integrated photosynthesis rate based on PAM fluorescence (PF) or on oxygen evolution (PO); amounts of alternative electrons (e); growth rate µ (on a Chl a basis); photosynthetic quotient PQ, respiratory quotient RQ measured during illumination and in darkness; integrated respiration R during light and dark period; and ratios dry mass : Chl a, carbon : dry mass, and carbon : nitrogen.

a* (m2 per g Chl a)10.96 ± 0.0511.43 ± 0.21 9.80 ± 0.0710.51 ± 0.34
Qphar (mmol per mg Chl a d−1) 26.7 ± 1.2124.3 ± 4.2 25.0 ± 0.9112.3 ± 3.4
PF (mmol O2 per mg Chl a d−1) 1.67 ± 0.07 6.14 ± 0.19 1.89 ± 0.03 5.66 ± 0.09
PO (mmol O2 per mg Chl a d−1) 1.48 ± 0.05 3.90 ± 0.09 0.93 ± 0.01 2.69 ± 0.03
Alternative e (mmol per mg Chl a d−1) 0.89 ± 0.09 8.69 ± 0.69 3.77 ± 0.0912.40 ± 0.37
Growth rate, µ (d−1) 0.28 ± 0.01 0.59 ± 0.02 0.35 ± 0.02 0.81 ± 0.04
PQ 1.52 ± 0.05 1.56 ± 0.11 1.33 ± 0.00 1.31 ± 0.01
RQ
 light 0.48 ± 0.04 0.56 ± 0.08 0.66 ± 0.01 0.74 ± 0.01
 dark 0.57 ± 0.01 0.65 ± 0.03 0.54 ± 0.12 0.49 ± 0.02
R (mmol O2 per mg Chl a per 12 h)
 light 0.66 ± 0.04 0.76 ± 0.19 0.33 ± 0.01 0.55 ± 0.02
 dark 0.46 ± 0.03 0.45 ± 0.04 0.23 ± 0.02 0.40 ± 0.03
Dry mass : Chl a38.91 ± 2.2649.71 ± 2.3624.35 ± 2.0335.97 ± 3.11
C : dry mass 0.47 ± 0.00 0.47 ± 0.00 0.49 ± 0.01 0.51 ± 0.01
C : N 5.74 ± 0.16 5.64 ± 0.36 5.86 ± 0.43 5.82 ± 0.46

Given the cell size of P. tricornutum, this is a relatively slow-growing alga compared with other small diatoms (Geider et al., 1985; Nicklisch, 1998; Litchman, 2000). Nevertheless, the growth rates under dynamic light conditions derived in this work are similar to other studies with the same organism under constant illumination with corresponding light doses (Osborne & Geider, 1987; Laws et al., 1997). The same ratio of growth rates as in P. tricornutum under FL and SL (Table 1) was found in Thalassiosira weissflogii in comparison of two light intensities in constant illumination (Clark et al., 2002), which corresponds to the light doses applied in FL and SL conditions, respectively, in the present work. This is a further confirmation of the finding that growth of diatoms under fluctuating light conditions either is not, or is only slightly decreased compared with constant illumination (Cosper, 1982; Nicklisch, 1998; Litchman, 2000).

The elemental composition (C : dry mass; C : N) of both algae did not change in response to the applied light conditions (Table 1). This is in agreement with findings under fluctuating light by Fietz & Nicklisch (2002), and with studies under constant illumination with different light irradiances (Berges & Falkowski, 1998; Bartual & Gálvez, 2002; Sarthou et al., 2005). As depicted from the standard deviations of the mean values of C : N ratios (Table 1), we observed only small changes in the C : N ratio in P. tricornutum during light or dark periods with lowest values at the beginning, and highest values at the end of light periods. A strong decrease in C : N ratio after a light period would indicate dark nitrate assimilation, as observed by Clark et al. (2002). The assimilation of nitrate in darkness was shown to be induced by high light irradiances or N limitation when N assimilation and transport could not keep up with C fixation (Clark et al., 2002). Thus we would expect this effect only under SL conditions. However, the relatively low growth rates of P. tricornutum may prevent the necessity of additional nitrate assimilation in the dark. In this context, the excretion of C (see below) could also help avoid periods of unbalanced N and C assimilation during high irradiances.

Large changes in response to light conditions have been found in the pigmentation of cells. The differences between algae grown under FL or SL are mainly caused by the increase in total light dose and average irradiance in SL compared with FL. With increasing average light irradiances and daily insolation, the chlorophyll content per cell tends to decrease (indicated by the ratio dry mass per Chl a, Table 1), which is consistent with other studies (Osborne & Geider, 1987; Kroon et al., 1992b; Janssen et al., 2001; Fietz & Nicklisch, 2002; Havelková-Doušováet al., 2004). Decreasing chlorophyll content is accompanied by a slight decrease in accessory carotenoids (fucoxanthin in P. tricornutum and lutein in C. vulgaris). In accordance with Friedman & Alberte (1986) and Ley & Mauzerall (1982), it can be assumed that this is the result of a decrease in the number of light-harvesting units in comparison with reaction centres of both photosystems. However, the reduction in pigment content did not lead to a significant decrease in the optical cross-section of cells (a*) grown under SL climates (Table 1). This is probably because of a decreased package effect within the thylakoid membranes (Falkowski et al., 1985; Janssen et al., 2001). The diatom and the green alga responded with an almost twofold increase in the amount of xanthophyll cycle pigments on the higher irradiance under SL conditions, although P. tricornutum contained a much larger xanthophyll cycle pool size than C. vulgaris already under FL conditions (Table 2). As the xanthophyll cycle (and corresponding heat emission) is one of the major processes to dissipate excessively absorbed light energy, the differences between P. tricornutum and C. vulgaris are expected to influence the quantum efficiency of photosynthesis (see below).

Table 2.  Changes in pigment composition in Phaeodactylum tricornutum and Chlorella vulgaris suspended in exponentially fluctuating or sine light conditions
ParameterP. tricornutum (mmol pigment per mol Chl a)
Chl cFxXC-poolβ-car
Sine light climate 74 ± 6670 ± 42293 ± 1239 ± 6
Fluctuating light climate118 ± 6706 ± 24185 ± 830 ± 3
 C. vulgaris (mmol pigment per mol Chl a)
Chl bLXC-poolNxβ-car
  1. Chl c, Chlorophyll c; Fx, fucoxanthin; XC-pool, xanthophyll cycle pool; β-car, β-carotin; Chl b, chlorophyll b; L, lutein; Nx, neoxanthin.

Sine light climate231 ± 9167 ± 20127 ± 940 ± 283 ± 10
Fluctuating light climate252 ± 5177 ± 3 67 ± 341 ± 176 ± 5

The major objective of the present study was to recover the new aspect of the electron balance from absorbed light to biomass. We wanted to study, in quantitative terms, how different growth rates result from physiological adaptation to the applied light conditions. This approach requires quantitative measurement of photosynthetic efficiency, the amount of alternative electrons and the efficiency of biomass production per absorbed photon, which has apparently not yet been done.

Photosynthetic quantum efficiency

The calculation of photosynthetic efficiency, the amount of alternative electrons and modelled biomass production was based essentially on photosynthesis rates derived from the measurement of variable chlorophyll fluorescence of PSII. The calculation of fluorescence-based electron transport therefore could be influenced by the limits of the fluorescence method itself, such as the possible contribution of PSI to the measurement of variable fluorescence at PSII (Pfündel, 1998) or reabsorption of emitted fluorescence. Under the conditions applied in this study, these limitations should have no impact. Furthermore, this approach requires several assumptions, including the excitation of both photosystems for linear electron transport and an even distribution of absorbed light between both photosystems. The impact of the latter aspects is discussed below in more detail.

The photosynthetic reaction is competing for absorbed light energy with the emission of fluorescence and thermal dissipation. Under low irradiances, charge separation in the reaction centre is very efficient. We observed almost the same maximum quantum efficiency for both algae at the beginning of the light period of FL and SL: c. 0.09 mol O2 per mol photons (±0.007), which equals the reciprocal value of a quantum demand of 11 quanta per mol O2 evolved (data not shown). However, under high irradiance excessive light absorption can endanger the photosynthetic apparatus, which requires the activation of regulatory reactions. Therefore the photosynthetic efficiency must be lowered to balance light absorption and the capacity of photosynthetic electron transport. Under the conditions in the present study, the effective quantum yield of photosynthesis, ΦPSII, dropped and thus increased the quantum demand to 16 and 13 mol quanta per mol O2 in FL for P. tricornutum and C. vulgaris, respectively. Under SL conditions in both algae, the effective quantum requirement was further increased to 20 mol quanta per mol O2 (Table 3).

Table 3.  Biomass production and quantum efficiency measured in Phaeodactylum tricornutum and Chlorella vulgaris suspended in exponentially fluctuating (FL) or sine light (SL) conditions
ParameterP. tricornutumC. vulgaris
FLSLFLSL
  1. Measured biomass production Bµ; ratio of modelled biomass production based on oxygen evolution (BO) to measured biomass (Bµ), ratio of modelled biomass production based on PAM fluorescence (BF) to measured biomass (Bµ); effective quantum efficiency of photosynthetic electron transport (ΦPSII); amount of photosynthetic electrons found in measured biomass (Φµ). All data derived from mean values in Table 1.

Bµ[mg dry matter per mg Chl a d−1]10.8929.33 8.5229.16
BO/Bµ 0.95 1.56 1.04 1.17
BF/Bµ 1.24 2.82 3.10 3.01
ΦPSII[mol mol−1] 0.063 0.049 0.076 0.05
Φµ (%)40.414.420.714.0

Essentially, the quantum efficiency of the photosynthetic reaction can be lowered by two photoprotective mechanisms which compete with energy trapping in the reaction center at PSII: (i) downregulation mechanisms in the PSII reaction centre (Schreiber & Neubauer, 1990); (ii) thermal energy dissipation via the fast interconversion of xanthophyll cycle (Gilmore, 1997; Niyogi, 1999). Consequently, Arbones et al. (2000) were able to demonstrate an indirect correlation between the concentration of photoprotective pigments and the quantum yield of C fixation. Therefore, in the analysis of the quantum efficiency of photosynthesis, it is important to consider the thermal dissipation of absorbed energy. The nonphotochemical quenching of chlorophyll fluorescence is a relative measure of heat dissipation. Whereas in FL the light periods were too short to induce large amounts of de-epoxidized xanthophylls and a concomitant NPQ under SL conditions, the NPQ in P. tricornutum was much higher compared with C. vulgaris (Fig. 3). In the comparison of the integrated values of NPQ in the course of the light period, the heat dissipation differed by a factor of 2 between both algae (data not shown). This was accompanied by a very strong de-epoxidation of diadinoxanthin to diatoxanthin in P. tricornutum, whereas the accumulation of antheraxanthin + zeaxanthin in C. vulgaris reached only 40% of the amount of de-epoxidized xanthophylls in the diatom (Fig. 3). In view of these results, it is surprising that the effective quantum efficiency of photosynthesis was found to be the same for both algae under nonfluctuating light conditions (Table 3). It is important to note that in P. tricornutum the increase in NPQ is strictly coupled to the activity of the xanthophyll cycle under both light conditions, whereas in C. vulgaris a significant portion of NPQ is independent of xanthophyll de-epoxidation. In SL the NPQ in the time period from 8:30 to 11:00 was not accomplished by anthera/zeaxanthin formation, and in FL we observed NPQ completely without concomitant xanthophyll cycle activity in C. vulgaris (Fig. 3). It is remarkable that the highest deviations between fluorescence- and oxygen-based photosynthesis in C. vulgaris were detected in those time periods of illumination, where NPQ was not coupled to the activity of xanthophyll cycle (cf. Figs 2c,d and 3). This strongly indicates that NPQ in C. vulgaris is affected by additional mechanisms to xanthophyll cycle-dependent quenching alone. One possible mechanism could be the redistribution of excitation energy from PSII to PSI via state transitions (for review see Larkum, 2003). Although the exact mechanism of state transition in algae is not yet clarified, it seems that the dissociation of light-harvesting complex II (LHCII) from PSII leads to a decrease in fluorescence and to a reduced absorption cross-section of PSII. This effect was observed in C. vulgaris by a concomitant decrease in the apparent and maximum fluorescence (Fig. 4) during the time when NPQ was not coupled to the xanthophyll cycle activity (Fig. 3). With further increasing irradiance in SL climate, this effect might have been overcome by a stronger reduction in the PQ pool, as indicated by the increase in apparent fluorescence (Fig. 4). State transitions can have two impacts on electron transport rates: (i) the efficiency of noncyclic electron transport between PSII and PSI is increased, which would also increase the efficiency of C fixation; and/or (ii) there is an enhanced activity of cyclic electron transport around PSI. As there was a large deviation between oxygen- and fluorescence-based photosynthesis rates in C. vulgaris, we suggest that state transitions under the conditions applied in this study led to an increased activity of cyclic PSI electron transport, and that a large proportion of the alternative electrons can be attributed to this mechanism. In specialized photosynthetic cells, this cycle around PSI is providing extra ATP in some cellular processes, such as C4 photosynthesis in higher plants (Leegood et al., 1981). Nevertheless, the activity of PSI-driven cyclic electron flow was also demonstrated in algae (Ravenel et al., 1994) and might be related to the adjustment of the NADPH : ATP ratio (Bendall & Manasse, 1995).

Figure 3.

Comparison of the daily course of nonphotochemical quenching (NPQ) in Phaeodactylum tricornutum and Chlorella vulgaris under sine light (SL, bold line) and fluctuating (FL, thin line) conditions. Comparison of the amount of de-epoxidized xanthophylls diatoxanthin (Dt; in P. tricornutum) and antheraxanthin + zeaxanthin ((Ax + Zx); in C. vulgaris) under SL (filled squares) and FL (open circles) conditions in a daily course.

Figure 4.

Changes in apparent (bold) and maximum (thin line) fluorescence derived from online PAM fluorescence measurements during cultivation of Chlorella vulgaris under sine light conditions.

To explain the discrepancy of the same quantum efficiency of photosynthesis in both algae on the one hand, and a much higher NPQ in P. tricornutum on the other, it appears likely that the amount of light absorbed by PSII is overestimated in the case of C. vulgaris. Thus, with the assumption of the occurrence of state transitions, the quantum efficiency of oxygen evolution in C. vulgaris can be expected to be higher, if calculated on the basis of the effective PSII-absorption cross-section in comparison with P. tricornutum.

The strong correlation between NPQ and the activity of the xanthophyll cycle in P. tricornutum suggests that there was no redistribution of excitation energy between PSII and PSI. Therefore the quantum efficiency of photosynthesis in the diatom is expected to be controlled mainly by heat dissipation. However, this does not exclude the possibility of cyclic electron transport at PSI, as an important proportion of Chl a in diatoms is already bound to PSI under high-light growth conditions (Fietz & Nicklisch, 2002). Lavaud et al. (2002) presented evidence that cyclic electron transport at PSII is highly active in diatoms under high-light conditions. The PSII cycle activity should result in an increased ratio between fluorescence- and oxygen-based electron transport because of the omitted oxygen release. Furthermore, it could also lower the reduction state of the PQ pool and, consequently, lead to an overestimation of the fluorescence-based electron transport and quantum efficiency. In higher plants the Mehler reaction and photorespiration are suggested to lead to differences between fluorescence- and oxygen-/C-based photosynthesis rates (Schreiber & Neubauer, 1990). The Mehler–ascorbate peroxidase cycle involves electron transport, but no net O2 exchange (Schreiber et al., 1994). Together with photorespiration, this alternative electron pathway has in common that it is activated when CO2 fixation becomes limited. However, diatoms are able to perform CO2-concentrating mechanisms (Rotatore et al., 1995). Furthermore, it was suggested previously that they behave like C4 plants (Reinfelder et al., 2000, 2004). Although, it has been shown that metabolic intermediates of the photorespiratory pathway were also found in diatoms (Schnitzler-Parker et al., 2004), the physiology of diatoms should rather suppress the activity of Mehler reaction or photorespiration.

In analysing the balance of energy flow from Qphar to biomass, it is necessary not only to investigate the kinds of alternative electron pathway, but also to quantify the amount of alternative electrons.

Modelling the amount of alternative electrons

To determine the amount of alternative electrons, fluorescence- and oxygen- based P/E curves were measured and the integrated photosynthesis rates compared. In this study we used a model for the conversion of variable chlorophyll fluorescence into photosynthetic oxygen evolution, was based on the currently established models of the conversion of absorbed light into photosynthetic electron transport (Gilbert et al., 2000b; Figueroa et al., 2003). In this aspect, it was advantageous that the cells of P. tricornutum and C. vulgaris exhibited comparable optical properties (indicated by a*, Table 1). Hence both algal cultures absorbed almost the same amount of incident PAR (within 10% deviation) under both fluctuating and nonfluctuating light conditions (Table 1). A major point of criticism of the models of photosynthesis is that fluorescence-based modelling of photosynthesis rates essentially depends on the amount of absorbed light. The fluorescence yield displays the energy conversion at PSII; however, calculation of fluorescence electron transport is based on cellular absorption. With the assumption that the linear transport of one electron requires excitation at PSII and PSI, it is also assumed that the amount of absorbed light is evenly distributed between both photosystems. Nevertheless, the effective absorption cross-section of PSII and PSI is highly variable, species-specific, and adapts to growth and light conditions. On the other hand, the energy conversion at PSI is very efficient. Therefore every absorbed photon which is not used in charge separation at PSII or thermally dissipated can be expected to be employed at PSI, and is detectable as a difference between fluorescence- and oxygen-based photosynthesis rates. Thus we assume that fluorescence-based photosynthesis displays true photosynthetic electron transport at either or both photosystems, and can be applied in calculating the number of alternative electrons.

From the measurements of P/E curves (Fig. 2a,b), it is obvious that in both algae a close correlation of fluorescence-based (which represents the capacity of electron release at PSII) and oxygen-based photosynthesis rates was found under low light intensities. At higher irradiances we observed increasing differences between both methods. As oxygen evolution at PSII is overlaid by other O2-consuming processes, the differences between the fluorescence and oxygen methods can be attributed to the activity of alternative electron pathways. Table 1 and Fig. 2(c,d) show that alternative electron pathways differed not only under SL and under FL, but also between P. tricornutum and C. vulgaris. Generally, in both algae the amount of alternative electrons was higher in cells grown in SL compared with FL. This was because of the higher mean incident PAR, which increases alternative electron cycling, and to the higher total daily light dose in the SL climate. However, whereas in P. tricornutum the PF/PO ratio was 1.13 and 1.57 in FL and SL, respectively, we found high PF/PO ratios in C. vulgaris in both light climates (2.03 and 2.10 in FL and SL, respectively). This resulted in much higher amounts of alternative electrons under both light conditions in C. vulgaris (50 and 55% of photosynthetic electrons were used in alternative electron pathways under FL and SL conditions, respectively) compared with P. tricornutum (13 and 35%, respectively; Table 1; Fig. 2c,d). This difference is the basis for interpretation of the photosynthetic energy conversion into biomass.

Comparison of modelled and measured biomass production

At first it seems contradictory that, considering the differences in the amount of alternative electrons, daily growth rates were found to be higher in C. vulgaris than in P. tricornutum under both light conditions (Table 1). Nevertheless, growth rates were determined on a Chl a basis and did not take into account differences in the ratio dry mass per Chl a and the resulting energy requirement for biomass formation. Therefore in ecological terms the C-related biomass production is the parameter of higher impact. Here we observed the same rates of dry matter production in SL for both algae, but 28% higher for the diatom compared with the green alga under FL (Table 3). If, on the one hand, the photosynthetic capacity (PF) and effective quantum efficiency of photosynthesis were found to be comparable in both algae under the respective light conditions, and, on the other hand, there were differences in dry matter formation, then we should expect deviations in the photosynthetic efficiency of growth. To calculate the conversion efficiency of photosynthetically absorbed energy into biomass, it was necessary to model the theoretical biomass production.

The modelling of theoretical biomass production was based on measured photosynthesis rates. For this purpose, we also accounted for respiration and the photosynthetic and respiratory quotient. In this way it was possible to include losses of photosynthetic electrons (and consequently losses in C fixation) which are not detectable by the difference between PF and PO, for example caused by nitrate reduction or photorespiration in combination with C excretion. The accuracy of this model is validated by the fact that there was a close correlation between measured and oxygen-based modelled biomass production in FL (Table 3). In SL, an overestimation of 17% of the modelled oxygen-based biomass production compared with the measured biomass was observed in C. vulgaris. Even higher (56% overestimation) was the discrepancy in P. tricornutum grown under SL conditions. This loss of oxygen-based electrons could have been caused by several processes: (i) with the measuring procedure of O2 and CO2 gas exchange and the calculation of PQ, it was not possible to account for all alternative electrons under SL conditions; (ii) underestimation of respiration rates; (iii) production of extracellular carbohydrates; and (iv) photorespiration in combination with the excretion of glycolate, because this C loss is not detectable by gas-exchange measurements. In P. tricornutum, the large overestimation of biomass production based on oxygen measurements is unlikely to be caused by the measuring procedure. Cells in the culture vessel are well supplied with CO2, and C limitation can be excluded. To compensate for the overestimated biomass formation, the respiration rate would have to be increased by more than 100%.

In benthic (Smith & Underwood, 2000; de Brouwer & Stal, 2002), but also in planktonic diatoms (Zlotnik & Dubinsky, 1989; Penna et al., 1999) the excretion of extracellular polymeric substances (EPS) has been demonstrated and could be a reason for the observed differences between measured and oxygen-based biomass formation. It has been shown that the production of EPS is light-dependent and has a function in cell motility (Lind et al., 1997), or may be the result of an unbalanced growth (de Brouwer & Stal, 2002). This means that the amount of fixed C exceeds the capacity of C storage within the cell. Hence the short light periods under FL conditions are supposed to suppress the production of EPS, whereas in SL with a much higher light dose, a high photosynthetic production might induce the excretion of EPS.

Photorespiration, which is usually activated under conditions of high light intensities and high O2/low CO2 concentration (Beardall, 1989), could also account for the observed loss of oxygen-based electrons. In higher plants, fixation of O2 results in the formation of glycolate, which is subsequently metabolized in the photorespiratory pathway. In algae, however, a substantial release of glycolate is often seen as a result of photorespiratory activity (Leboulanger et al., 1997, 1998). It was suggested that the excretion of glycolate is increased under conditions where the production of glycolate overwhelms the capability of cells to recycle it (Schnitzler-Parker et al., 2004). Under the conditions of the present study, we can definitely rule out high O2 concentrations during cultivation of algae. However, we do not want to exclude the possibility that, even with well aerated cell cultures, in periods of high light intensity the intracellular CO2 concentration might not satisfy the photosynthetic demand for C. From the measurements performed in this study, it is not possible to estimate the extent of a certain alternative electron pathway. Therefore we explicitly want to include the possible contribution of other alternative electron pathways. In future investigations we plan to analyse the metabolic products of photosynthetic C fixation, which should help to clarify this question.

The comparison of the fluorescence-based modelling of biomass formation with the measured biomass revealed an overestimation of the modelled biomass by a factor of 3 in C. vulgaris under both light conditions (Table 3). However, in P. tricornutum we found (only in SL) a large difference between fluorescence-based modelled and measured biomass production, which was comparable with C. vulgaris. Under FL only 24% of the photosynthetic electrons in the diatom were not found in the biomass. The comparison of the quantum efficiency of photosynthesis revealed that both algae performed photosynthesis at the same rate of light utilization under FL and SL, respectively. Therefore the photosynthetic quantum efficiency could not account for the differences between modelled and measured biomass production. Nevertheless, we found large deviations in the conversion efficiency of photosynthetically absorbed energy into biomass (Φµ) between the diatom and the green alga under fluctuating light conditions. In C. vulgaris, under FL conditions, 21% of the photosynthetic electrons are found in the biomass compared with 14% under SL conditions. In P. tricornutumΦµ is increased to 40% in FL compared with 14% in SL. The data appear to be correlated with the amount of alternative electrons. The highest photosynthetic efficiency was measured in P. tricornutum under FL conditions, where the amount of alternative electrons was the lowest. In contrast, cells of C. vulgaris showed the highest activity of alternative electron pathways in SL conditions, with the lowest efficiency in the conversion of photosynthetic energy into biomass. This suggests that the diatom is able to adapt the activity of alternative electron pathways to the environmental conditions, whereas the constant ratio of PF/PO in C. vulgaris under FL and SL conditions suggests that this is not the case in the green alga. In nonfluctuating light climates, the energy dissipation in both algae, irrespective of the mechanism, results in the same efficiency of biomass formation. This means that under high irradiance, alternative electron pathways can partly replace or supplement the activity of the xanthophyll cycle in the dissipation of excessive energy. Therefore the results (especially with C. vulgaris) show that alternative electron transport is not necessarily an unavoidable waste of energy, rather than a mechanism to adjust light absorption to the capacity of metabolism to prevent photoinhibitory damage.

Strategies in dissipation of excess energy

In the natural environment, light conditions require adaptation of the algal photosynthetic apparatus to periods of light limitation and also to excessive light supply. The dissipation of excessively absorbed light takes place at two levels: (i) at the level of PSII and especially LHCII via thermal dissipation and/or adjustment of the absorption cross-section; (ii) at the level of redirecting electrons that overwhelm the capacity of CO2 fixation into safe pathways, where the energy is dissipated or at least partly used in metabolism. In this study it was shown that the regulation of energy dissipation at both levels is species-specific, and was dependent on the applied light conditions.

Diatoms obviously favour energy dissipation at the level of LHCII and PSII. In consequence, the amount of alternative electrons is kept to a minimum under fluctuating light conditions. In this way, diatoms achieve a very high conversion efficiency of photosynthetic energy into biomass. Under dynamic but nonfluctuating light conditions, thermal dissipation of excessive energy is not sufficient to prevent the release of photosynthetically energized electrons that are not used in C fixation. The dissipation of electrons is realized by cycling of electrons in still unknown sinks. The increase in the amount of alternative electrons drastically decreases the efficiency of biomass production.

Green algae also dissipate a large part of excessively absorbed light at the LHCII level as heat, but the degree of thermal dissipation is less than in diatoms. Additionally, these algae seem to redistribute the absorbed light energy between PSII and PSI, especially under fluctuating light conditions and under low and medium irradiance in nonfluctuating light climates. Unlike diatoms, green algae are unable to adjust the activity of alternative electron pathways in response to irradiance. It is unclear whether this is the consequence of less effective energy dissipation at the LHCII/PSII level, or an obligatory event in the regulation of electron transport. The result, however, is a much lower conversion efficiency of photosynthetic energy into biomass under fluctuating light conditions. In a dynamic but nonfluctuating light climate, the mixture of less thermal dissipation and increased alternative electron cycling operates with the same efficiency as in diatoms.

Conclusions

Two main conclusions can be drawn from the data presented here.

  • (i) The differences found in growth efficiency between diatoms and green algae may partly explain the seasonal dominance of diatoms, especially in turbulent waters. Our results were obtained under nutrient-replete conditions, and may be very different from stages where nutrient availability is the major factor controlling phytoplankton growth.
  • (ii) The data indicate that upscaling from PSII electron transport to biomass production is critical, as it depends on both the type and physiological adaptation of phytoplankton. Therefore photosynthesis rates alone may not predict changes in the phytoplankton structure.

It will be important to determine how nutrient limitation alters the growth balance, and whether there are different types of regulation of energy conversion into biomass, such as a green and a nongreen type. Furthermore, cyanobacteria may not fit into these types, as they do not perform a nonphotochemical quenching of absorbed energy via the xanthophyll cycle. Here, electron cycling may be the prominent mechanism of energy dissipation. The application of the fluorescence-based productivity measurements to ecological studies requires empirical correction factors for each type to allow the prediction of growth and biomass production.

Acknowledgements

This study was supported by the Deutsche Forschungsgemeinschaft (DFG, WI 764/10-1). We would like to thank Günter Wünsche (Faculty of Chemistry, University Leipzig) for technical assistance in the C/H/N analysis.

Ancillary