Bioresource Technology 98 (2007) 1489–1493

Production of biomass and nutraceutical compounds by

Spirulina platensis under diVerent temperature and nitrogen regimes
Luciane Maria Colla, Christian Oliveira Reinehr, Carolina Reichert,
Jorge Alberto Vieira Costa ¤
Laboratory of Biochemical Engineering, Department of Chemistry, Federal University Foundation of Rio Grande (FURG), P.O. Box 474,
CEP 96201-900 Rio Grande, RS, Brazil

Received 17 February 2004; received in revised form 17 March 2005; accepted 27 September 2005
Available online 27 October 2006

Abstract

The cyanobacterium Spirulina platensis has been used by humans because of its nutritional and possibly medicinal eVects. Our study
evaluated the inXuence of temperature and nitrogen concentration in the medium on the production of biomass by this cyanobacterium
and the biomass composition in protein, lipid and phenolic compounds. We found that at 35 °C there was a negative eVect on biomass
production but a positive eVect on the production of protein, lipids and phenolics, the highest levels of these compounds being obtained in
Zarrouk’s medium containing 1.875 or 2.500 g l¡1 sodium nitrate. Higher biomass densities and productivity were obtained at 30 °C than
at 35 °C, but nitrogen concentration appeared to have no eVect on the amount of protein, lipid or phenolics, indicating that at 30 °C the
concentration of sodium nitrate in Zarrouk’s medium (2.50 g l¡1) can be reduced without loss of productivity, an important cost-saving
factor in large-scale cultivation.
© 2005 Elsevier Ltd. All rights reserved.

Keywords: Nitrogen; Nutraceuticals; Spirulina platensis; Temperature

1. Introduction single cell protein (SPC) (Anupama, 2000) and because of

Spirulina platensis is a planktonic photosynthetic Wla-
mentous cyanobacterium that forms massive populations
in tropical and subtropical bodies of water which have high
levels of carbonate and bicarbonate and alkaline pH values
of up to 11. This cyanobacterium is recognizable by the
main morphological feature of the genus, i.e. the arrange-
ment of multicellular cylindrical trichomes in an open
left-hand helix along the entire length of the Wlaments
(Vonshak, 1997).
For centuries, native peoples have harvested S. platensis
from Chad Lake in Africa and Texcoco Lake in Mexico for
use as a source of food (Vonshak, 1997), a fact which means
that Spirulina deserves special attention both as a source of
*
Corresponding author. Tel.: +55 53 233 8653; fax: +55 53 233 8745.
E-mail address: dqmjorge@furg.br (J.A.V. Costa).
its nutraceutical properties. The chemical composition of
Spirulina indicates that it has high nutritional value due to
its content of a wide range of essential nutrients, such as
provitamins, minerals, proteins and polyunsaturated fatty
acids such as gamma-linolenic acid (Miranda et al., 1998).
More recently, Spirulina has been studied because of its
therapeutic properties (Belay et al., 1993) and the presence
of antioxidant compounds (Estrada et al., 2001; Miranda
et al., 1998) such as phenolics. The occurrence of phenolic
compounds in plants is well documented and these com-
pounds are known to possess antioxidant activity in bio-
logical systems but the antioxidant characteristics of algae
and cyanobacteria are less well documented, although
decreased cholesterol levels have been reported in hyper-
cholesterolemic patients fed Spirulina (Ramamoorthy and
Premakumari, 1996) and the antioxidant activity of phyco-
biliproteins extracted from S. platensis has also been dem-
onstrated (Estrada et al., 2001).

0960-8524/$ - see front matter © 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2005.09.030
1490 L.M. Colla et al. / Bioresource Technology 98 (2007) 1489–1493

The inXuence of growth conditions on the chemical com- 2.2. Cultivation

position of Spirulina has been studied by many researchers
with the purpose of optimizing the production of economi-
cally and nutritionally interesting compounds, especially
gamma-linolenic acid (GLA) and phycocyanin (Tanticha-
roen et al., 1994; Cohen et al., 1993). Although the use of
Spirulina for the production of phycocyanin for use as a
natural pigment seems commercially eYcient, the produc-
tion of GLA from this cyanobacterium presents higher
costs when compared to production using other sources
(Cohen et al., 1993). The extraction of nutritionally active
compounds in pure form is expensive, but the direct con-
sumption of Spirulina as a nutraceutical food is a viable
alternative. Growth conditions optimized for biomass pro-
duction and productivity are usually used in the commer-
cial production of Spirulina without considering chemical
composition, but higher concentrations of potentially
useful compounds such as polyunsaturated fatty acids,
proteins and phenolics can be obtained by manipulating
growth conditions.
The objective of the work presented in this paper was
to evaluate the inXuence of nitrogen concentration and
temperature on the protein, lipid and phenolic compound
content, maximum speciWc growth rate and productivity of
S. platensis.
2. Methods
2.1. Microorganism and culture medium
The cyanobacterium S. platensis strain LEB-52 (Costa
et al., 2002, 2003) was used in this study. To prepare and
maintain the inoculum we used Zarrouk’s medium, a stan-
dard synthetic medium containing 2.50 g l¡1 sodium nitrate
as nitrogen source (Zarrouk, 1966), this medium also being
used to study the batch growth of S. platensis but in this
case the concentration of sodium nitrate was modiWed as
explained below in section Experimental design. All the
reagents used were of analytical grade, obtained from the
Merck Chemical Co. (Darmstadt, Germany) or the Synth
Chemical Co. (São Paulo, Brazil).
S. platensis was cultivated in 20 l photo-bioreactors with
an initial volume of 14 l and an initial biomass concentra-
tion of 0.15 g l¡1. Two glass tubes were passed through the
photo-bioreactor’s stopper, one for sample collection and
the other for continuous aeration. The cultures were mixed
and aerated using Wltered air at a Xux of 170 l h¡1 supplied
by diaphragm pumps. The cultures were illuminated with
daylight-type 40 W Xuorescent lights (Osram, Brazil) which
provided an irradiance of 31.35 molphotons m¡2 s¡1 (Costa
et al., 2000). Cultures were maintained in a greenhouse
under a 12 h light/12 h dark photoperiod at 30 °C and 35 °C.
To monitor biomass changes in cultures, samples were
taken every 24 h using aseptic technique. At the end of each
batch run, replicate cultures of each experiment were
pooled, Wltered, washed with distilled water to remove solu-
ble salts, centrifuged at 15,000 rpm, lyophilized and stored
at ¡20 °C for protein, lipid and phenolic compound deter-
minations.
2.3. Experimental design
For this study we used a multilevel factorial design
(MFD) in which the concentration of sodium nitrate in
Zarrouk’s medium was set at four diVerent levels (0.625,
1.250, 1.875 and 2.500 g l¡1) and the temperature at 30 and
35 °C (Table 1), all experiments being triplicated. Analysis
of variance (ANOVA) was used to compare the data and
calculate p-values.
2.4. Analytical methods
Biomass concentration (g l¡1) was calculated by measur-
ing optical density at 670 nm to produce a standard curve
relating dry weight of S. platensis biomass (g l¡1) to optical
density, this standard curve was subsequently used to calcu-
late the biomass of individual samples based on their optical
density. The calculated biomass (the average of three
experiments) were used to construct growth curves from
which we obtained maximum speciWc growth rates (max)

Table 1
Cultivation of Spirulina platensis in media with diVerent sodium nitrate concentrations and at diVerent temperatures
Run NaNO3 (g l¡1) max (day¡1) t (h) P450 (mg l¡1 day¡1) Protein (%) Lipid (%) Phenolics (mg g¡1)
T D 30 °C
1 0.625 0.073 § 0.002 24–468 34.0 § 0.1 59.76 § 2.07 6.73 § 0.40 3.09 § 0.26
2 1.250 0.073 § 0.002 24–468 30.0 § 1.0 57.36 § 1.13 6.69 § 0.27 3.66 § 0.27
3 1.875 0.073 § 0.001 48–612 34.0 § 3.7 60.82 § 1.88 7.61 § 0.43 3.27 § 0.39
4 2.500 0.074 § 0.005 24–468 30.2 § 0.7 57.61 § 1.16 8.16 § 0.23 3.78 § 0.30
T D 35 °C
5 0.625 0.048 § 0.001 96–588 23.9 § 3.9 58.92 § 0.96 7.49 § 1.10 2.46 § 0.22
6 1.250 0.050 § 0.006 72–588 24.8 § 4.0 56.73 § 0.79 7.95 § 1.42 2.42 § 0.21
7 1.875 0.054 § 0.002 24–564 26.4 § 3.9 70.15 § 0.82 10.37 § 0.63 4.99 § 0.37
8 2.500 0.054 § 0.003 24–468 24.8 § 4.0 65.47 § 2.19 10.03 § 0.63 4.92 § 0.29
Except for t, all values show means § standard deviation.
max D maximum speciWc growth rate; t D start–end of the exponential growth phase; P450 D productivity at 450 h.
 L.M. Colla et al. / Bioresource Technology 98 (2007) 1489–1493
from the log phase of the growth curves by exponential
regression. Productivities was calculated from the equa-
tion P D (Xi ¡ X0)/ti, where P Dproductivity (mg l¡1 day¡1),
X0 D initial biomass density (mg l¡1), Xi D biomass density
at time i (mg l¡1) and ti D time interval (h) between X0 and Xi.
Protein was determined by the micro-Kjeldahl method
according to AOAC standard methods (AOAC, 1995) in
which protein is assumed to contain 16% nitrogen. Four
replicates were analyzed for each lyophilized biomass sam-
ple.
Lipid content was evaluated using Folch’s method
(Folch and Lees, 1957) by extracting lipids in a 2:1 chloro-
form/methanol mixture and determining lipid content
gravimetrically. Three replicates were used for each lyophi-
lized biomass sample.
Phenolic compounds were determined by extraction
with methanol, followed by partitioning with hexane and
precipitation of non-phenolics with Ba(OH)2 and ZnSO4.
Total phenolics were determined spectrophotometrically by
the Folin–Ciocalteau method using tyrosine as standard
(Singleton and Rossi, 1965). Five replicates were used for
each lyophilized biomass sample.
3. Results and discussion
The growth curves for 30 °C are shown in Fig. 1a while
Fig. 1b shows the curves for 35 °C. Because the S. platensis
cells had previously been adapted to the medium there was
no lag phase. The biomass concentrations achieved in the
30 °C runs (0.82–0.92 g l¡1) were higher than in the 35 °C
runs (0.59–0.65 g l¡1). It has been shown by previous work-
ers (Danesi et al., 2001; Vonshak, 1997) that the optimal
growth temperature for S. platensis is between 30 and
35 °C, with 40 °C deWnitely being deleterious to this cyano-
bacterium. In respect to increase in biomass, the best
responses were obtained at 30 °C, which agrees with the
studies by Danesi et al. (2001).
The fact that the highest biomass values occurred at
30 °C may have been due to the fact that the partial pres-
1.00
0.80
Biomass (g/l)
0.60
0.40
0.20
0.00
0 200 400
Time (h)
(a)
Fig. 1. Biomass density growth curves of Spirulina platensis growing in media containing diVerent concentrations of sodium nitrate: (a) 30 °C (Run 1 䊉,
Run 2 䊏, Run 3 䊊, Run 4 䊐); (b) 35 °C (Run 5 䊉, Run 6 䊏, Run 7 䊊, Run 8 䊐). Sodium nitrate concentration was 0.625, 1.250, 1.875 and 2.500 g l¡1
respectively for Runs 1–4, the same concentrations being used in the same order for Runs 5–8 respectively.
1491
sure of CO2 in the medium is higher at 30 °C than at 35 °C,
leading to a higher concentration of bicarbonate and conse-
quently an increased rate of photosynthesis. Another factor
that should be considered is that at higher temperatures (i.e.
35 °C) there is an increase in dark cycle respiratory activity
in which the cells use reserve material (e.g. carbohydrates)
for respiration and a concomitant decrease in cell weight
(Vonshak et al., 1982). According to Torzillo and Bernar-
dini (1991), for outdoor cultures of Spirulina up to 34% of
the biomass produced during the daylight period may be
lost through respiration at night.
Values for maximum speciWc growth rate (max), dura-
tion of the exponential phase (t), productivity at 450 h
(P450), and protein, lipid and phenolic compound content
are shown in Table 1.
In general the biomass and max values were higher at
30 °C (0.073–0.074 day¡1) than they were at 35 °C (0.048–
0.054 day¡1). In Spirulina, the lower the population density,
the higher the speciWc growth rate. This is to be expected
for a system that is primarily light-limited because reducing
the population density increases the availability of light to
each cell. However, the eVect of decreasing the population
density in a light-limited system is most pronounced at high
temperatures and much less so at low temperatures (Von-
shak et al., 1982). This explains why the highest max values
were observed at 30 °C and not at 35 °C. The max values at
30 °C appear not to have been aVected by sodium nitrate
concentration. The highest productivity (P450 D 30–
34 mg l¡1 day¡1) was obtained at 30 °C, and it thus appears
that the concentration of sodium nitrate in Zarrouk’s
medium (2.500 g l¡1) could be reduced to 0.625 g l¡1 without
loss of productivity, decrease production costs in large-
scale cultivation.
The p-values (Table 2) show that the concentration of
sodium nitrate in Zarrouk’s medium had a signiWcant eVect
on the production of protein, lipid and phenolics, while cul-
tivation temperature had a signiWcant eVect on all variables.
There were signiWcant interactions between protein, lipid
and phenolics contents at the 99% conWdence interval,
1.00
0.80
Biomass (g/l)
0.60
0.40
0.20
0.00
600 800 0 200 400 600 800
Time (h)
(b)
1492 L.M. Colla et al. / Bioresource Technology 98 (2007) 1489–1493

Table 2 Miranda et al. (1998), the main phenolic compounds found

Analysis of variance for the multilevel factorial design described in the in Spirulina are salicylic, trans-cinnamic, synaptic, chloro-
text
genic, quimic and caVeic acids. However, the metabolic
Factors p-value pathways for the formation of these compounds in cyano-
max Productivity Protein Lipid Phenolics bacteria and their importance are still unknown. For green
(P450) plants, Rechner et al. (2001) reported that the amino acid
X1 0.1788 0.3889 <0.0001a <0.0001a <0.0001a phenylalanine could be converted by an ammonia lyase to
X2 <0.0001a <0.0001a <0.0001a 0.0002a 0.0004a trans-cinnamic acid, which is then converted to cumaric
X1 · X2 0.3637 0.5180 <0.0001a 0.0039a <0.0001a
acid that in turn is converted to caVeic acid, these com-
a
SigniWcant at the 99% conWdence interval. X1 D nitrogen concentra- pounds being converted through several chemical reactions
tion; X2 D temperature; X1·X2 D interaction of factors.
to Xavonoids, natural pigments with antioxidant activity.
According to Duval and Shetty (2001), biosynthetic path-
indicating that the interaction between these variables
should be investigated in more detail.
The lower sodium nitrate concentrations (0.625 and
1.250 g l¡1) gave lower values for cellular proteins and lipids
(Fig. 2a and b). Nitrogen is required for synthesis of the
amino acids, which make up proteins and other cellular
components such as phycocyanin. However, at 30 °C, nitro-
gen uptake seems to be limited because the experiments
with higher concentrations of sodium nitrate (1.875 and
2.500 g l¡1) showed no increase in the level of protein, while
at 35 °C an increase was observed (Runs 7 and 8 in Table 1).
With regard to lipids, higher concentrations of sodium
nitrate resulted in an increase in lipids, similar to that which
was obtained by Manabe et al. (1992), who demonstrated
higher total lipid in Spirulina grown in media containing up
to 25 mM of ammonium chloride. However, Piorreck et al.
(1984) found that the concentration of nitrogen had little
inXuence on the total lipid and fatty acid composition of
some cyanobacteria, the inXuence of nitrogen being more
marked in eukaryotic algae. Piorreck et al. (1984) also
found that for cyanobacteria, total lipid content remained
constant at all nitrogen concentrations studied (0.001–0.1%
of potassium nitrate), with only a slight increase occurring
at the highest nitrogen concentration tested (0.1%). Olguín
et al. (2001) observed a higher content of total lipids in
Spirulina growing in Zarrouk’s medium as compared to
Spirulina cultivated under conditions of nitrogen starva-
tion. In our experiments, higher protein and lipid content
may have been associated with nitrogen uptake because
Runs 7 and 8 presented lower biomass densities which
would have meant that more light was available to individ-
ual cells, light being the only limiting factor in self-shading
cells at high densities (Qiang et al., 1996). Because there was
less biomass in the 35 °C experiments more light may have
been available to individual cells and light ceased to be a
limiting factor, allowing better use of available nitrogen
and, consequently, higher production of protein and lipids
at the higher nitrogen content of Runs 7 and 8. According
to Walach et al. (1987), higher quantities of carbohydrates
are synthesized when nitrogen availability is decreased but
carbon availability is constant. These two factors together
may explain the higher production of protein, lipids and
phenolics in Runs 7 and 8.
At 35 °C higher nitrogen concentration (Runs 7 and 8) Fig. 2. Plots of means of two-way interaction eVects for: (a) protein con-
promoted higher levels of phenolics (Fig. 2c). According to tent, (b) lipid content and (c) phenolics.
 L.M. Colla et al. / Bioresource Technology 98 (2007) 1489–1493
ways that lead to the formation of Xavonols and phenyl-
propanols are related to the pentose-phosphate (Calvin)
cycle, and the amounts synthesized are characteristic of
each organism. These same pathways can lead to the for-
mation of such compounds in cyanobacteria because the
Calvin cycle is part of the photosynthetic carbon Wxing
mechanism of these organisms (Fay, 1993).
Although the mechanisms used by cyanobacteria to syn-
thesize phenolic compounds and their metabolic functions
in cells are still unknown, we found that at 35 °C there was
a signiWcant increase in the quantity of phenolic com-
pounds synthesized by the cells, and this increase was also
related to an increase in other cellular components such as
proteins and lipids. In spite of the fact that biomass produc-
tion at 30 °C presented better results in terms of productiv-
ity and number of cells, cultivation at 35 °C presented
advantages related to the quantities of useful compounds
such as proteins, lipids and, especially, phenolics.
4. Conclusions
In this paper we have demonstrated that temperature has
an important inXuence on the production of biomass, pro-
teins, lipids and phenolics by S. platensis (35 °C having a neg-
ative eVect on biomass production but a positive eVect on the
production of protein, lipids and phenolics, the highest levels
of these compounds being obtained at 35 °C and 1.875 g l¡1
sodium nitrate or 2.500 g l¡1 sodium nitrate). Therefore, cul-
tivation at 35 °C can be exploited when the purpose is to pro-
duce S. platensis with nutritional characteristics. Highest
biomass density and productivity were obtained at 30 °C,
although nitrogen concentration appeared to have no eVect
on the quantity of protein, lipid or phenolics produced,
which indicates that at this temperature the concentration of
sodium nitrate in Zarrouk’s medium (2.50 g l¡1) can be
reduced to 0.625 g l¡1 without loss of biomass productivity.
References
Anupama, P.R., 2000. Value-added food: single cell protein. Biotechnol-
ogy Advances 18, 459–479.
AOAC, 1995. OYcial Methods of Analysis, 16th ed. Association of OYcial
Analytical Chemists, Arlington, USA. ISBN 0-935584-54-4.
Belay, A., Ota, Y., Miyakawa, K., Shimamatsu, H., 1993. Current knowl-
edge on potential health beneWts of Spirulina. Journal of Applied Phy-
cology 5, 235–241.
Cohen, Z., Reungjitchachawali, M., Angdung, W., Tanticharoen, M., 1993.
Production and partial puriWcation of -linolenic acid and some pig-
ments from Spirulina platensis. Journal of Applied Phycology 5, 109–115.
Costa, J.A.V., Colla, L.M., Duarte Filho, P., 2003. Spirulina platensis
growth in open raceway ponds using fresh water supplemented with
carbon, nitrogen and metal ions. Zeitschrift für Naturforsch 58c, 76–80.
Costa, J.A.V., Colla, L.M., Duarte Filho, P.F., Kabke, K., Weber, A., 2002.
Modelling of Spirulina platensis growth in fresh water using response
surface methodology. World Journal of Microbiology and Biotechnol-
ogy 18, 603–607.
Costa, J.A.V., Linde, G.A., Atala, D.I.P., Mibielli, G.M., Kruger, R.T.,
2000. Modelling of growth conditions for cyanobacterium Spirulina
1493
platensis in microcosms. World Journal of Microbiology and Biotech-
nology 16, 15–18.
Danesi, E.D.G., Rangel, C.O., Pelizer, L.H., Carvalho, J.C.M., Sato, S.,
Moraes, I.O., 2001. Production of Spirulina platensis under diVerent
temperatures and urea feeding regimes for chlorophyll attainment. In:
Proceedings of the Eighth International Congress on Engineering and
Food, vol. 2, pp. 1978–1982.
Duval, B., Shetty, K., 2001. The stimulation of phenolics and antioxidant
activity in pea (Pisum sativum) elicited by genetically transformed anise
root extract. Journal of Food Biochemistry 25 (5), 361–377.
Estrada, J.E., Bescós, P., Villar Del Fresno, A.M., 2001. Antioxidant activ-
ity of diVerent fractions of Spirulina platensis protean extract. Il Far-
maco 56, 497–500.
Fay, P., 1993. The Blue-Greens. Edward Arnold, London. ISBN 0-7131-
2878-x.
Folch, J., Lees, M., 1957. A simple method for isolation and puriWcation of
total lipids from animal tissues. Journal of Biological Chemistry 226,
497–509.
Manabe, E., Hirano, M., Takano, H., Ishikawa-Doi, N., Sode, K., Matsu-
naga, T., 1992. InXuence of ammonium chloride on growth and fatty
acid production by Spirulina platensis. Applied Biochemistry and Bio-
technology 34/35, 273–281.
Miranda, M.S., Cintra, R.G., Barros, S.B.M., Filho, J.M., 1998. Antioxi-
dant activity of the microalga Spirulina maxima. Brazilian Journal of
Medical and Biological Research 31, 1075–1079.
Olguín, E., Galicia, S., Angulo-Guerrero, O., Hernández, E., 2001. The
eVect of low light Xux and nitrogen deWciency on the chemical compo-
sition of Spirulina sp. (Arthospira) grown on digested pig waste. Biore-
source Technology 77, 19–24.
Piorreck, M., Baasch, K.H., Pohl, P., 1984. Biomass production, total pro-
tein, chlorophylls, lipids and fatty acids of freshwater green and blue-
green algae under diVerent nitrogen regimes. Phytochemistry 23 (2),
207–216.
Qiang, H., Guterman, H., Richmond, A., 1996. Physiological characteris-
tics of Spirulina platensis (cyanobacteria) cultured at ultrahigh cell den-
sities. Journal of Phycology 32, 1066–1073.
Ramamoorthy, A., Premakumari, S., 1996. EVect of supplementation of
Spirulina on hypercholesterolemic patients. Journal of Food Science
and Technology 33 (2), 124–128.
Rechner, A.R., Spencer, J.P.E., Kuhnle, G., Hahn, U., Rice-Evans, C.A.,
2001. Novel biomarkers of the metabolism of caVeic acid derivatives
in vivo. Free Radical Biology and Medicine 30 (11), 1213–1222.
Singleton, V.L., Rossi, J.A., 1965. Colorimetry of total phenolics with
phosphomolybdic–phosphothgstic acid reagents. American Journal of
Enology and Viticulture 16, 144–158.
Tanticharoen, M., Reungjitchachawali, M., Boonag, B., Vondtaveesuk, P.,
Vonshak, A., Cohen, Z., 1994. Optimization of -linolenic acid (GLA)
production in Spirulina platensis. Journal of Applied Phycology 6,
295–300.
Torzillo, G., Bernardini, P., 1991. On-line monitoring of chlorophyll Xuo-
rescence to assess the extent of photoinhibition of photosynthesis
induced by high oxygen concentration and low temperature and its
eVect on the productivity of outdoor cultures of Spirulina platensis
(cyanobacteria). Journal of Applied Phycology 34, 504–510.
Vonshak, A., 1997. Spirulina platensis (Arthrospira). Physiology, Cell-
biology and Biotechnology. Taylor & Francis, London. ISBN 0-2035-
8670-0.
Vonshak, A., Abeliovich, A., Boussiba, S., Arad, S., Richmond, A., 1982.
Production of Spirulina biomass: eVects of environmental factors and
population density. Biomass 2, 175–185.
Walach, M.R., Bazin, M., Pirt, J., 1987. Computer control of carbon–nitro-
gen ratio in Spirulina platensis. Biotechnology and Bioengineering 29,
520–528.
Zarrouk, C., 1966. Contribution à l’étude d’une cyanophycée. InXuence de
divers facteurs physiques et chimiques sur la croissance et la photosyn-
thèse de Spirulina maxima. Ph.D. Thesis, Université de Paris, Paris.