Introduction

The use of algae to temporarily sequester CO2 has been often proposed (Singh and Ahluwalia 2012). These propositions have rarely taken into account the physiological complexity of algal responses to elevated CO2, which may be especially crucial if biomass utilization is also intended. A substantial increase of [CO2] may generate an imbalance in the rate of supply of C relative to other nutrients (Beardall and Giordano 2002; Montechiaro and Giordano 2010; Raven et al. 2011, 2012). The increase in external C:N ratio due to higher external [CO2] may cause a reorganization of organic cell composition, leading to an increase of those pools with lower N content (Lynn et al. 2000; Palmucci et al. 2011). This is a well known response that is often invoked to stimulate oleogenesis (Gushina and Harwood 2006). On first principles, this strategy is more promising when the increase of the C:N ratio is afforded by an increase of C rather than by a decrease of N: a reduction of N supply sufficient to appreciably affect the C:N ratio is likely to negatively affect growth rate, which is instead often stimulated by a higher C:N associated with increased CO2 availability (and fixation). Morphological and functional constraints that orientate C allocation pathways in a species-specific mode (Palmucci et al. 2011; Palmucci and Giordano 2012) may complicate things. At elevated CO2, cells may also benefit from the downregulation of the CO2 concentrating mechanisms (CCM), which constitute one of the main features of extant aquatic photosynthetic organisms (Giordano et al. 2005). Regrettably, a quantification of the amount of energy diverted to growth after CCM downregulation is difficult, because the extent and precision of CCM modulation has been only partially studied (Beardall and Giordano 2002; Raven et al. 2011, 2012; Wu et al. 2012). Furthermore, the impact of CCM downregulation would depend on the energy (light) available to the cultures. In brief, the outcome of an increased supply of CO2 may be very different from what one may expect. In order to make a first step towards experimentally addressing this problem, we compared the biomass quality of four marine species with different CCMs and taxonomy subjected to an increase of pCO2.

Materials and methods

The impact of a pCO2 increase on biomass quality was tested on four marine microalgae species: Dunaliella salina, with a very active CCM at ambient pCO2 (Giordano and Bowes 1997); Protoceratium reticulatum, with a not very active CCM (Ratti et al. 2007); Thalassiosira weissflogii; allegedly a C4 diatom (Reinfelder et al. 2004; Roberts et al. 2007); and Phaeodactylum tricornutum, a C3 diatom (Tachibana et al. 2011). All species were grown axenically and semicontinuously in ESAW medium (Berges et al. 2001), pH 8.2, at an irradiance of 100 μmol photons m2 s−1 (PAR) and at 20 °C. Cell numbers were determined as described in (Domenighini and Giordano 2009). Cultures were acclimated to growth at different pCO2 (390 and 1,000 μL L−1) and diluted daily with medium pre-equilibrated with the desired pCO2 (Reinfelder et al. 2004) to maintain constant cell concentrations. Measurements were taken at time zero (t0), after 24 hours (t1), and 20 days (t20).

Fourier transform infrared (FTIR) spectra were acquired and analyzed as described by Domenighini and Giordano (2009), Palmucci et al. (2011), and Palmucci and Giordano (2012). The relative abundances of carbohydrates, lipids, and proteins were calculated from the ratio between the integrals of the relevant absorption bands, after deconvolution of the spectra. The differential spectra between the time points (t1-t0, t20-t0, t20-t1) were calculated, in the spectral region from 1,800 to 800 cm−1, by subtracting the absorptions at each wavelength in the spectrum collected at the earlier time point from those in the spectra collected at the later time point.

Statistics

All measurements were repeated for three independent cultures. The significance of mean differences was tested with a one-way analysis of variance, followed by Dunnett’s post hoc test, using the GraphPad Prism 4.03 software (GraphPad Software, San Diego, CA, USA). The level of significance was set at 0.05.

Results and discussion

In all four species, the overall composition changed appreciably as a consequence of exposure to elevated pCO2, both for one and 20 days (p < 0.05; Fig. 1a).

Fig. 1
figure 1

Biomass composition of cells of four algae species acclimated to 390 μL L−1 CO2 (t0) and exposed for 1 (t 1 ) and 20 days (t 20 ) to 1,000 μL L−1 CO2. In the case of diatoms, the silica pool was also considered. a Variation of the overall cellular composition, expressed as the integral of the differential spectra; b Lipid to protein ratio; c Carbohydrate to protein ratio; d Carbohydrate to lipid ratio; e Silica to protein ratio; f Silica to carbohydrate ratio; g Silica to lipid ratio

Full size image

The change in overall cell composition (as derived from the integral of the differential spectra) was typically lower between t20 and t0 than between t1 and t0. The only exception was T. weissflogii, where cell composition changed by the same magnitude in both intervals. In all species, however, the differential spectra between t1 and t20 were appreciably different from zero. This suggests that, although feedback mechanisms may mitigate the initial impact of an environmental perturbation, a drift of the biomass quality occurred over 20 days, either because of genotype selection (Collins and Bell 2004) or because of longer-term acclimatory processes. This warns against extrapolating results from short-term experiments to estimate productivity of commercial systems and points towards the need for frequent biomass monitoring. In this respect, we propose the use of FTIR spectroscopy as the optimal method to keep biomass under control, due to the ease of its use (no extraction is required and no substantial sample preparation; simply dehydration is sufficient), to the fact one method is used for all major pools (thus no artifact can be introduced as a consequence of different extractive or analytical methods), and for its relatively low cost (Domenighini and Giordano 2009; Palmucci et al. 2011; Palmucci and Giordano 2012).

More detailed analyses show that the lipid to protein and the carbohydrate to lipid ratios varied in a species-specific manner and not always in the direction of a greater relative C allocation to lipids (p < 0.05; Table 1; Fig. 1b, d).

Table 1 Variation of the main organic pools in four algal species, after 24 h (t1) and for 20 days (t20) since transfer from 390 (t0) to 1,000 μL L−1 CO2
Full size table

D. salina was the only species in which the carbohydrate to protein ratio decreased significantly, after 24 h at elevated pCO2, and continued to decrease till t2 (Fig. 1c). In P. tricornutum, no change was observed in this ratio after 24 hours, but a significant decrease was measured after an extended period at elevated pCO2. In both T. weissflogii and P. reticulatum, the carbohydrate to protein ratio increased after 24 hours and showed a reversion after 20 days at high pCO2 (in T. weissflogii, this ratio reverted to the value at t0). The ratios of silica to proteins, carbohydrates, and lipids are suggestive of an only moderate change in the silica content (if any) in T. weissflogii, with higher pCO2; a more pronounced trend towards higher silicification with higher pCO2 was observed in the typically little silicified P. tricornutum (Lewin et al. 1958; Table 1; Fig. 1e–g). The fundamental importance of species selection clearly emerges from these observations, in agreement with results obtained in other conditions (Palmucci et al. 2011). Some common assumptions about the relations between nutrition and biomass quality (e.g., high C:N = oleogenesis) should thus be reconsidered: C allocation patterns can be very different in different species and can lead to profoundly distinct biomass compositions for similar growth conditions.