Comparative Analysis of the Outdoor Culture of Haematococcus Pluvialis in Tubular and Bubble Column...

Journal of Biotechnology 123 (2006) 329–342

Comparative analysis of the outdoor culture of Haematococcuspluvialis in tubular and bubble column photobioreactors

M.C. Garcıa-Malea Lopez a, E. Del Rıo Sanchez b, J.L. Casas Lopez a,F.G. Acien Fernandez a, J.M. Fernandez Sevilla a, J. Rivas b,

M.G. Guerrero b, E. Molina Grima a,∗

a Department of Chemical Engineering, University of Almerıa, Canada San Urbano S/N, Almerıa 04071, Spainb Inst. Bioquımica Vegetal y Fotosıntesis, University Sevilla-CSIC, E-41092-Sevilla, Spain

Received 19 July 2005; received in revised form 27 October 2005; accepted 23 November 2005

Abstract

The present paper makes a comparative analysis of the outdoor culture of H. pluvialis in a tubular photobioreactor and abubble column. Both reactors had the same volume and were operated in the same way, thus allowing the influence of the reactordesign to be analyzed. Due to the large changes in cell morphology and biochemical composition of H. pluvialis during outdoorculture, a new, faster methodology has been developed for their evaluation. Characterisation of the cultures is carried out on amacroscopic scale using a colorimetric method that allows the simultaneous determination of biomass concentration, and the
chlorophyll, carotenoid and astaxanthin content of the biomass. On the microscopic scale, a method was developed based onthe computer analysis of digital microscopic images. This method allows the quantification of cell population, average cell sizeand population homogeneity. The accuracy of the methods was verified during the operation of outdoor photobioreactors on apilot plant scale. Data from the reactors showed tubular reactors to be more suitable for the production of H. pluvialis biomassand/or astaxanthin, due to their higher light availability. In the tubular photobioreactor biomass concentrations of 7.0 g/L (d.wt.)were reached after 16 days, with an overall biomass productivity of 0.41 g/L day. In the bubble column photobioreactor, onthe other hand, the maximum biomass concentration reached was 1.4 g/L, with an overall biomass productivity of 0.06 g/L day.The maximum daily biomass productivity, 0.55 g/L day, was reached in the tubular photobioreactor for an average irradianceinside the culture of 130 �E/m2s. In addition, the carotenoid content of biomass from tubular photobioreactor increased up to2.0% d.wt., whereas that of the biomass from the bubble column remained roughly constant at values of 0.5% d.wt. It shouldbe noted that in the tubular photobioreactor under conditions of nitrate saturation, there was an accumulation of carotenoidsdue to the high irradiance in this reactor, their content in the biomass increasing from 0.5 to 1.0% d.wt. However, carotenoidaccumulation mainly took place when nitrate concentration in the medium was below 5.0 mM, conditions which were onlyobserved in the tubular photobioreactor. A similar behaviour was observed for astaxanthin, with maximum values of 1.1 and
∗ Corresponding author. Tel.: +34 950 015032; fax: +34 950 015484.E-mail address: [email protected] (E.M. Grima).

0168-1656/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.jbiotec.2005.11.010

330 M.C.G.-M. Lopez et al. / Journal of Biotechnology 123 (2006) 329–342

0.2% d.wt. measured in the tubular and bubble column photobioreactors, respectively. From these data astaxanthin productivitiesof 4.4 and 0.12 mg/L day were calculated for the tubular and the bubble column photobioreactors. Accumulation of carotenoidswas also accompanied by an increase in cell size from 20 to 35 �m, which was only observed in the tubular photobioreactors.Thus it may be concluded that the methodology developed in the present study allows the monitoring of H. pluvialis culturescharacterized by fast variations of cell morphology and biochemical composition, especially in outdoor conditions, and thattubular photobioreactors are preferable to bubble columns for the production of biomass and/or astaxanthin.© 2005 Elsevier B.V. All rights reserved.

Keywords: Haematococcus; Photobioreactors; Astaxanthin; Digital image; Colorimeter; Outdoor

1. Introduction

Microalgae are a potential source of biomass or spe-cific products such as lipids, pigments, antioxidants,etc. One of the most recent processes based on microal-gae is the production of astaxanthin from Haematococ-cus pluvialis. Astaxanthin is a high-value carotenoidpigment with important applications in the nutraceuti-cal, cosmetics, food and feed industries (Guerin et al.,2003). The major market for astaxanthin is as a pigmen-tation source in aquaculture, primarily in salmon andtrout (Guerin et al., 2003). In this sense, the microalgaeHaematococcus pluvialis is the richest source of natu-ral astaxanthin, and it is now cultivated on an industrialscale (Olaizola and Huntley, 2003).

Astaxanthin sells for US $2500 kg with an annualworldwide market estimated at US $200 million(Lorenz and Cysewski, 2000). Although 95% of thismarket consumes synthetically derived astaxanthin,

(Tredici and Materassi, 1992; Richmond et al., 1993;Molina et al., 1994; Acien et al., 1998; Garcıa et al.,1999). However, only bubble columns and tubular pho-tobioreactors have proven capable of scaling up to highvolumes. Although the outdoor production of H. plu-vialis in tubular photobioreactors has been referenced(Olaizola, 2000; Chaumont and Thepenier, 1995), nodata from bubble column has been referenced in spiteof the highly adequate light profile inside this type ofreactor (Garcıa et al., 1999). In addition, Chaumontand Thepenier (1995) have reported the rapid variationof biomass concentration and pigment content in day-light using tubular photobioreactors, with carotenoidcontent increasing from 0.6 to 1.4% d.wt. in 4 h, from7:00 to 11:00 h.

Due to the fast and great variation of Haematococ-cus cells during outdoor culture, a fast methodologyto characterize the cultures at macroscopic and micro-scopic scale is necessary. Microscopic characterisation,

consumer demand for natural products makes the syn-thetic pigments much less desirable and provides anopportunity for the production of natural astaxanthinby H. pluvialis. This strain contains 1.5–3.0% astaxan-thin and has gained acceptance in aquaculture and othermarkets as a “concentrated” form of natural astaxanthin(Lorenz and Cysewski, 2000; Olaizola and Huntley,2003). In this sense, natural astaxanthin from H. plu-vialis is currently produced in a two-step process. In

i.e. cell population and size distribution, is usually per-formed either by direct observation, or using automateddevices such as haemocytometers or more sophisti-cated cell counters (Harker et al., 1996; Tripathi etal., 1999). However, apart from cell counters, the othermethods do not quantify cell size and homogeneity,and the use of cell counters for Haematococcus cells isvery problematic due to variations in cell size betweendifferent cell morphotypes, frequently more than 10-

the first step green vegetative cells are produced under fold. Macroscopic culture characterisation is usually
controlled culture conditions, frequently indoors, usingstirred tank or bubble columns. In the second step,green cells are exposed to stress conditions (high irra-diance, nitrate and/or phosphate deprivation, high tem-perature) to induce accumulation of astaxanthin, usingopen raceways or tubular photobioreactors. Flat panel,bubble columns and tubular photobioreactors havebeen extensively proposed as outdoor closed photo-bioreactors for the industrial production of microalgae
performed as the dry weight measurement of biomassconcentration and the determination of pigment contentby spectrophotometry–HPLC (Del Campo et al., 2000).Dry weight measurements are tedious and results arenot obtained until at least 12 h after sampling. Thespectrophotometric measurement of pigment contentrequires a time-consuming extraction process of thepigments using adequate solvents, which must alsoensure effective cell wall breakage. In the case of

M.C.G.-M. Lopez et al. / Journal of Biotechnology 123 (2006) 329–342 331

H. pluvialis, efficient extraction of pigments is onlyproperly achieved using mechanical procedures of dis-ruption (grounding with alumina) or chemical methods(DMS) (Del Campo et al., 2000). In addition, the astax-anthin content of the biomass must finally be quantifiedby HPLC. In short, the measurement of biomass pig-ment content is both a time and labour consumingmethod, taking a minimum of 24 h.

In the present paper a comparative analysis on theperformance of H. pluvialis cultures carried out in bub-ble columns and tubular photobioreactors is performed.The objective is to determine the best reactor to be usedfor the outdoor production of astaxanthin, as well as toidentify the main variables governing the behaviour ofthe cultures. For this, and due to fast changes in cellmorphology and biochemical composition of this strainwhen accumulation of astaxanthin takes place, two fastmethods for the microscopic and macroscopic charac-terization of H. pluvialis cultures are developed. Themicroscopic characterization is performed by a com-puter analysis of digital images of the cultures obtainedwith a microscope. The macroscopic characterizationis carried out by measuring colour changes directly onH. pluvialis cultures, and the method is capable of mea-suring biomass concentration and pigment content. Thedevelopment of fast characterization methods as wellas determining the most adequate photobioreactor andthe main variables governing the behaviour of outdoorcultures of H. pluvialis is the first step towards opti-mising the outdoor production of astaxanthin from H.p

2

2

Caog2iibmt

demand injection of CO2. Batch cultures were car-ried out at different irradiances, nitrate concentrations,and elapsed culture times, to obtain different morpho-logical types of cells. The irradiance ranged from 50to 2000 �E m−2 s−1, whereas the nitrate concentrationranged from 3.0 to 15.0 mM, and the elapsed culturetime ranged from 6 to 12 days.

Outdoor cultures were performed in two types ofreactors: an airlift tubular reactor and a bubble columnreactor (Fig. 1). In both reactors, the culture temper-ature was controlled at 20 ◦C by using heat exchang-ers. The pH was controlled by injection of dry CO2,whereas air was bubbled to circulate the liquid andremove dissolved oxygen. The reactors were operatedin batch mode. The tubular airlift reactor had a volumeof 55 L, and consisted of a 98 m polymethylmethacry-late tube of 0.03 m diameter arranged horizontallyinside a water pool, with a degasser at a height of 3.5 m.Air was supplied at the beginning of the riser. The bub-ble column had a culture volume of 55 L, and consistedof a 2.0 m high vertical tube of 0.20 m diameter. Airwas supplied at the bottom of the column. In both pho-tobioreactors temperature, pH and dissolved oxygenprobes were installed and connected to a data acqui-sition system for on-line registration of measurementand control. The instantaneous photon-flux density ofthe photosynthetically active radiation (PAR) on a hori-zontal surface and inside the water pool were measuredon-line using two quantum scalar irradiance meters(LI-190 SA, Licor Instruments, Lincoln, NE, USA)

luvialis.

. Materials and methods

.1. Micro-organism and culture conditions

The microalga Haematococcus pluvialis (strainCAP 34/8) was from the Culture Collection of Algaend Protozoa of the Centre for Hydrology and Ecol-gy (Ambleside, UK), and was grown in an inor-anic medium free of acetate (Garcia-Malea et al.,005). Cultures for the calibration of fast character-zation methods were grown in 2 L photobioreactorsn laboratory conditions. The photobioreactors wereubbled with air at 1.0 v/v/min, with the temperatureaintained at 20 ◦C by passing thermostated water

hrough a jacket, and the pH maintained at 8.0 by on

connected to a data acquisition card. Measured valueswere integrated to determine the mean solar irradianceduring the daylight period. The mean irradiance on thesurface of the bubble column was calculated by multi-plying the irradiance on the horizontal surface by 0.48.This factor was previously determined by experimen-tal measurements (Garcıa et al., 1999). The averageirradiance inside the reactors was calculated as a func-tion of irradiance on the reactor surface, Io, the tubediameter, p, the extinction coefficient of the biomass,Ka, and the biomass concentration, Cb, as (Acien et al.,1998),

Iav = Io

KapCb(1 − exp(−KapCb)) (1)

where the extinction coefficient of the biomass, Ka, wasconsidered to be constant and equal to 0.1 m2/g.


Fig. 1. Scheme of outdoor photobioreactors used: (A) airlift tubular photobioreactor and (B) bubble column.

2.2. Analytical methods

Classic analytical methods were used for reference.The number of cells in the cultures was determined ina Neubauer counting chamber by direct observation onan Olympus CH20 microscope. The biomass concen-tration was determined by dry weight measurements.For this purpose, 50 mL of culture was filtered througha 0.45 �m Whatmann filter, washed with 0.1N HCl andoven-dried at 80 ◦C for 24 h. Chlorophyll content wasdetermined by the method of Mackinney (1941), afterdisruption of cells by sonication for 5 min and extrac-tion with methanol at 40 ◦C for 1 h. Total carotenoidcontent was determined spectrophotometrically by themethod of Davies (1976), after disruption of cells bysonication for 5 min and extraction with acetone for3 h.

For pigment analysis by HPLC cells were ground ina mortar with alumina and the extracts prepared in pureacetone. To quantify astaxanthin, saponification wasperformed, as follows. Acetone extracts were evapo-rated under nitrogen gas and the residue re-dissolvedin pure ethyl ether. The same volume of KOH in MeOH(2%, w/v) was added. The stirred mixture was allowedto react for 15 min at 0 ◦C in darkness under nitrogengas. To stop the reaction and to remove excess alkali,

10% NaCl was added. After separation of phases, etherwas evaporated under nitrogen gas and the remainingpigment re-dissolved in pure acetone. Pigments wereanalysed by the chromatographic method of Mınguez-Mosquera et al. (1992), as modified by Del Campoet al. (2000). Standards of astaxanthin, cantaxanthinand �-carotene were kindly provided by Hoffman-LaRoche, Switzerland. Violaxanthin was a gift from Dr.L. Lubian, Instituto de Ciencias Marinas, CSIC, Cadiz,Spain. Lutein was from Sigma Chemical Co., St Louis,MO, USA.

2.3. Fast characterization methods

For microscopic characterization of the cultures dig-ital images of the samples were obtained by usinga CMOS camera (Evolution LC Color from MediaCybernetics) mounted on the same microscope (Olym-pus CH20). Images were analyzed by software usingImage-Pro Plus 4.5.1 from Media Cybernetics. Thistype of software allows us to identify each cell in thephoto and to quantify the number of events, their meansize and the standard deviation of sizes (Sabri et al.,1997).

For macroscopic characterization of the culturesa colorimetric method was used. The colour of the


Table 1Cell density, biomass concentration and pigments content of biomass from cultures used for the calibration of the fast characterization methodsproposed

Culture Cell number (cell/mL) Cb (g/L) Chlorophyll (%d.wt.) Carotenoid (%d.wt.) Astaxanthin (%d.wt.) Cell type

1 459000 0.25 5.00 1.00 0.20 Flagellated2 819000 2.50 1.50 2.00 1.80 Cysts3 1138500 0.63 4.52 0.24 0.00 Flagellated4 3937500 1.79 5.90 0.50 0.13 Flagellated5 2142000 1.88 3.84 0.72 0.43 Palmeloids6 1206000 3.33 1.12 1.79 1.64 Cysts7 3244500 1.79 5.90 0.50 0.13 Flagellated8 1521000 1.33 4.00 1.02 0.61 Palmeloids

Determinations were performed by classical methods.

samples was determined using a CM-3500d Minoltaspectrophotometer–colorimeter with Spectramagic 3.6software (Minolta, Germany). For these measurementsa special glass cuvette (3 cm wide, 5 cm high, 1 cmdeep) was filled with 12 mL of sample and colourparameters were immediately obtained, in addition totransmittance at wavelengths from 400 to 700 nm. Inthe agricultural and food industries the most popularnumerical colour-space system is the L*a*b*, whichis also referred to as the CIE-LAB system, originallydefined by the CIE in 1976 (CIE, 1986). In this colourspace, L* indicates the clearness, whereas a* and b*are the chromaticity coordinates. The parameter a*measures the red and green characteristics, whereas b*measures the yellow and blue characteristics. Anotherfrequently used colour-space system is the L*C*h*.Measurements were carried out with both colour-spacesystems.

2.4. Statistical analysis

Statistical analysis of data was carried out usingStatgraphics version 7.0 software. Analysis of varianceand correlations between the variables were also cal-culated using this software.

3. Results

Before carrying out the experiments in outdoor pho-tiwfd

under different growth conditions were used. Data fromthese cultures, measured by classical methods, are sum-marized in Table 1. The biomass concentration of thecultures ranged from 0.25 to 3.33 g/L, whereas thechlorophyll and carotenoid content ranged from 1.12to 5.90% d.wt., and from 0.24 to 2.00% d.wt., respec-tively. The cell density of the cultures ranged from0.46 × 106 to 3.94 × 106 cell/mL, and cell morphol-ogy varied from flagellated cells to palmeloid cells,and cysts.

For the calibration of the microscopic fast character-ization method proposed, samples of cultures three tosix were used. From each culture, aliquots were dilutedat five levels, and four repetitions of each dilution weremeasured, a total of 80 samples being quantified byboth the classical Neubauer counting chamber methodand the proposed digital image-analysis method. Anal-ysis of the digital images of the samples allowed us todetermine not only the cell number, but also the aver-age cell size as diameter and standard deviation of cellsize. Analysis of variance (Table 2) showed that theculture sample was the major significant variable, theresults yielded by the different images taken for eachsample showing no significant difference. On the otherhand, the dilution ratio influenced the results only athigh values of this variable. The analysis of the resultscarried out at different dilution ratios showed that toavoid the influence of this variable the measurementsby digital image analysis must be performed at celldensities higher than 105 cell/mL. Fig. 2 shows theamimc

obioreactors, the development of necessary character-zation methods was needed. In this sense, two methodsere developed for the microscopic and macroscopic

ast characterization of the cultures. For this, eightifferent batch cultures performed at laboratory scale

greement of the classical Neubauer chamber countingethod and microscopic fast characterization by digital

mage analysis, verifying the adequacy of the proposedethodology. Thus, characterization of overall batch

ultures by the proposed digital image analysis method


Table 2Statistical analysis of data obtained from microscopic characterization of the cultures (n = 80)

Cell number (cell/mL) Cell size (�m) Standard deviation size (�m)

F-ratio Significant level F-ratio Significant level F-ratio Significant level

Culture 64.732 0.0000 53.132 0.0000 7.257 0.0003

Dilution 3.226 0.0174 3.669 0.0091 0.572 0.6838Replica 0.107 0.9558 0.360 0.7817 0.012 0.9981

Influence of independent variables (culture, dilution and replica) on the measured values of microscopic characterization. Significant variablesare shadowed.

Fig. 2. Correlation between cell number measured by classicalNeubauer chamber counting method and digital image analysismethod, both in cells/mL.

is shown in Fig. 3. It can be observed that besides celldensity, the proposed methodology also allows us todetermine mean cell size and standard deviation of cellsize inside the culture. Data showed that both the cell

Fig. 3. Measurements obtained from digital image analysis of sam-ples from batch cultures performed on different culture conditions.

size and standard deviation of cell size were lowest forflagellated cells (cultures 1, 3, 4 and 7) and highestfor cysts (cultures 2 and 6), cell size ranging from 19 to29 �m and standard deviation ranging from 5 to 12 �m,respectively.

Regarding the macroscopic fast characterization ofthe cultures a colorimetric method was developed.Colour measurements allow one to determine thecolour coordinates (L*, a*, b*, C*, h*) of each sam-ple instantaneously, directly over the fresh culture. Forthe evaluation of the colorimetric method samples ofall the cultures were used. Aliquots from each culturewere diluted at five levels, and three repetitions of eachdilution were measured, making a total of 120 samplesquantified. Analysis of variance (Table 3) showed thatthe culture was the only significant variable; neitherculture dilution nor replica were significant. In addi-tion, colour coordinates L*, a* and C* showed highersignificance levels, and were therefore selected as cor-relation variables. In this sense, correlation analysisof data (Table 4) showed how biomass concentrationcorrelated with the clearness index L*, whereas thered coordinate a* correlated with the carotenoid andastaxanthin concentration in the culture, and greencoordinate C* correlated with the chlorophyll con-centration in the culture. Mathematical relationshipand fitting between experimental and calculated values(Fig. 4) proved that the macroscopic fast characteri-zation method allows a fast quantification of culturebiomass and pigment concentrations in only a few sec-o

oHivc

nds, and directly using the fresh culture.Once the fast characterisation methods were devel-

ped and verified, the behaviour of outdoor cultures of. pluvialis in pilot scale photobioreactors was stud-

ed. The photobioreactors used had the same cultureolume, 55 L, but very different designs. The bubbleolumn photobioreactor is characterised by a tube of


Table 3Statistical analysis of data obtained from macroscopic characterization of the cultures (n = 120)

L a* b* C h

F-ratio Significantlevel





Culture 105 0.0003 68.2 0.0356 32.2 0.0553 78.6 0.0235 16.3 0.0665Dilution 6.53 0.0173 3.21 0.1456 1.66 0.2658 4.65 0.2256 1.23 0.2685Replica 0.10 0.9668 0.23 0.9865 0.35 0.9859 0.16 0.9987 0.46 0.9866

Influence of independent variables (culture, dilution and replica) on the measured values of macroscopic characterization. Significant variablesare shadowed.

Table 4Statistical analysis of data obtained from macroscopic characterization of the cultures

Parameter Cb (g/L) Chlorophylls (mg/L) Carotenoids (mg/L) Astaxanthin (mg/L)

Coefficientcorrelation

F-ratio Coefficientcorrelation



F-ratio

L* 0.9345 227 0.7185 35 0.7120 2 0.6257 15

a* 0.7319 38 0.0935 1 0.9646 321 0.9618 295

C* 0.4881 10 0.9472 288 0.0087 1 0.1260 1

Correlation between spectrophotometric-colour parameters and measured variables of the samples. Shadow data corresponded to the bestcorrelation factor determined.

Fig. 4. Correlation between selected parameters and measured variables of the samples: (A) biomass concentration vs. L*; (B) chlorophyllsconcentration vs. C*; (C) carotenoids concentration vs. a*; (D) astaxanthin concentration vs. a*.


Fig. 5. Experimental data from the discontinuous culture of H. pluvialis in the airlift tubular photobioreactor. The biomass concentration,pigments content, cell density, cell size and deviation of cell size were determined according to fast characterization methods proposed.

large diameter (0.20 m) arranged vertically and with alow surface/volume ratio (22 m2/m3). In contrast, theairlift tubular photobioreactor is characterised by a tubeof small diameter (0.03 m) placed horizontally and witha high surface/volume ratio (188 m2/m3). Both pho-tobioreactors were inoculated with 10% of the sameinoculum, and the cultures were then operated in batchmode for 16 days. Data from the airlift tubular photo-bioreactor and the bubble column photobioreactor areshown in Figs. 5 and 6, respectively.

Fig. 5 shows how, due to the immersion of the exter-nal loop of the photobioreactor in a 15 cm deep waterpool panted white, the irradiance impinging on thereactor surface was higher than that measured on a hor-izontal plane, with mean values in the daylight periodof 1600 �E/m2s. In spite of the high irradiance, theculture was not photoinhibited, as confirmed by val-ues of chlorophyll fluorescence above 0.6 regardlessof the mean daily irradiance on the reactor surface.

In this way, the biomass concentration increased upto 7.0 g/L after 16 days, including the lag phase. Themaximum growth rate during the exponential phasewas 0.040 h−1, whereas the overall biomass produc-tivity was 0.41 g/L day. Concerning the pigment con-tent, accumulation of carotenoids took place duringthe first 9 days, the carotenoid content increasing from0.5 to 1.0% d.wt., due to high irradiance on the reac-tor surface. However, the major accumulation tookplace when the nitrate concentration decreased below5 mM, when maximum carotenoid values of 2.0% d.wt.were recorded. In contrast, the chlorophyll contentwas high under nitrate saturation conditions, with amean value of 1.8% d.wt. but decreasing to valuesof 0.5% d.wt. when accumulation of carotenoids tookplace due to nitrate limitation (after day 9). Concerningthe microscopic characterisation of the cultures, datashowed that the cell density increased in similar fash-ion to biomass concentration up to 4.5 × 106 cell/mL


Fig. 6. Experimental data from the discontinuous culture of H. pluvialis in the bubble column photobioreactor. The biomass concentration,pigments content, cell density, cell size and deviation of cell size were determined according to fast characterization methods proposed.

on day 9, whereas the cell size and standard devi-ation of cell size remained constant with values of21 and 5 �m, respectively. However, under conditionsof nitrate limitation, cell division diminished and celldensity remained constant in spite of higher biomassconcentration. This was due to encystment of the cells,which became red, the cell size and standard devia-tion of cell size increasing to values of 32 and 9 �m,respectively.

A different behaviour was observed from data ofthe bubble column photobioreactor (Fig. 6). Due to thevertical arrangement the irradiance on the reactor sur-face was lower than measured on a horizontal surface;maximum mean values of 425 �E/m2s were recordedin daylight. The low irradiance protected the culture bypreventing photoinhibition, with values of fluorescenceof chlorophylls above 0.6 whatever the irradiance, butit also reduced the growth and biomass concentration

of the culture. Thus, a maximum biomass concentra-tion of 1.4 g/L day and a maximum growth rate onthe exponential phase of 0.021 h−1 were recorded.The overall biomass productivity of the culture waslow, 0.06 g/L day. This imposed a low nitrate uptakerate, and so the nitrate concentration of the cultureslowly decreased from the initial value of 20.0 mMto a final value of 16 mM, meaning that the culturewas always nitrate saturated. This fact, along withthe low irradiance on the reactor surface explains thatthe chlorophyll and carotenoid content of the biomassremained roughly constant, with mean values of 1.7and 0.5% d.wt., respectively. Cell size and standarddeviation of cell size remained constant, with val-ues of 18 and 4 �m. The cell density increased from0.5 × 106 to 2.5 × 106 cells/mL as the biomass con-centration increased, but cells remained as small greenflagellated vegetative cells.


Fig. 7. Variation of astaxanthin content of the biomass during thediscontinuous culture of H. pluvialis on tubular photobioreactor andbubble column used. Data measured according to fast characteriza-tion method proposed.

Differences between the reactors also determinedifferences in the astaxanthin accumulation of thebiomass. Reactor shape is the only factor that explainswhy the astaxanthin content of the biomass from thebubble column photobioreactor only increases slightlyfrom 0.10 to 0.25% d.wt., as opposed to an increasefrom 0.10 to 1.1% d.wt. in the airlift tubular photo-bioreactor (Fig. 7). The higher astaxanthin content inthe biomass from the tubular photobioreactor is alsorelated, as in the case of total carotenoids, to the higherirradiance and the depletion of nitrate in this reactor. Itis also important to note that the major accumulationof astaxanthin took place in the tubular photobioreactorwhen nitrate was depleted.

To quantify the influence of design of the reactor inthe light availability and the behaviour of the cultures,the average irradiance to which the cells were exposedinside the culture was calculated. Fig. 8 shows howthe biomass productivity increased with the averageirradiance up to 130 �E/m2s, and then decreased dueto the low biomass concentration inside the cultures.However, the average irradiance was 10 times higher inthe tubular photobioreactor than in the bubble column,due to the different arrangement and tube diameter ofboth photobioreactors. In this sense, the low averageirradiance inside the bubble column, with maximumvalues of 70 �E/m2s, determines that the growth ratewas much lower than the specific maximum growthrate, and that no accumulation of astaxanthin takesplace in this reactor. In contrast, the high average irra-

Fig. 8. Influence of the average irradiance inside the culture on thedaily biomass productivity of H. pluvialis cultures on discontinuousmode, in both tubular and bubble column photobioreactors.

diance inside the tubular photobioreactor, with valuesof up to 600 �E/m2s, indicates that this reactor can beused to maximise the production of both biomass andastaxanthin.

Finally, in order to verify the data obtained by usingthe fast characterisation methods, a comparison wasperformed with data measured using classic methods.Fast microscopic characterisation was demonstratedto reproduce the experimental data measured usingthe Neubauer chamber counting method, as confirmedwith a statistical procedure analogous to Fig. 2 (notshown). Concerning the macroscopic characterisation,Fig. 9 summarises the agreement of measurementsof biomass concentration and pigment content carried

Fig. 9. Fitting between experimental values of biomass concentra-tf
ion and pigments content measured by classical methods and usingast method proposed.


out using classical methods and using the proposedfast method. It is clear that the proposed methodologyallows us to quantify the biomass concentration andpigment content of the biomass, and although someerrors were over 15%, residues are unscrewed, and themethodology is, therefore, useful for determining thebehaviour of the cultures.

4. Discussion

In order to produce natural astaxanthin from H.pluvialis at competitive cost the production musttake place in outdoor conditions and using optimallydesigned processes. The core of the production processis the photobioreactor to be used, and it is thereforenecessary to determine the influence of the reactorgeometry on the yield of the process. In addition,outdoor cultures of H. pluvialis are characterisedby fast variations of biochemical composition andcell morphology in response to variations of cultureconditions, such as irradiance or nitrate concentration.The availability of fast characterisation methodscan therefore contribute to the optimisation of thisprocess.

In this sense, the production of astaxanthin fromH. pluvialis cultures is currently based on a two-stepprocess (Olaizola and Huntley, 2003), in which thebiochemical composition, morphology and wall of thecells undergo considerable changes. Thus, the determi-nptemcmtsacas(T1ai

of cell status and composition are very convenient. Inthis sense, the use of digital image software and col-orimeters for the microscopic and macroscopic char-acterisation of the cultures has been proved adequatein the present work. Digital image analysis softwareis nowadays more accessible and commonplace in thelaboratory, though more scarce on an industrial scale.Colorimeters are usually used in the quality control ofprocesses related with paints, ink, etc., but they havealso been used extensively in food technology for meat(Wulf and Wise, 1999), coffee (Ortola et al., 1998),tomato (Arias et al., 2000), etc.

Regarding the microscopic characterisation, thecycle and reason why the different cell types mod-ify the others is not clear (Kobayashi et al., 1997).A simplified representation of the population consist-ing of three basic cell types (flagellated, palmeloid andcyst) is usually accepted, although transitions and vari-ations of these types are also referenced (Kobayashiet al., 1997). However, most researchers only make aqualitative characterisation of the cultures indicatingonly the cell number and predominant cell morphol-ogy. The microscopic method developed in this workmakes this characterisation quantitative, measuring thecell number and also the average cell size and its stan-dard deviation, providing a more accurate descriptionof homogeneity of the cell population. This is pre-sented in Fig. 3, where the flagellated cells are shownto have a mean diameter of 20 �m, while palmeloidcells are larger, with a diameter of 25 �m, and cystsatlascdtwfi1fnb

ato

ation of the culture status is not complete if the usualractice of a single microscopic or macroscopic quan-ification is carried out (Harker et al., 1996; Tripathit al., 1999; Gong and Chen, 1998). Moreover, theethods for the macroscopic characterisation of the

ultures vary greatly from one cell type to another,ainly because the strengthened cell wall membrane

hat appears in certain cell forms requires the use ofpecial cell wall disruption systems such as groundinglumina (Del Campo et al., 2000). To further compli-ate matters, the changes in cell morphology are fastnd can take place all of a sudden induced by differenttress factors such as temperature, light, nutrients, etcBoussiba and Vonshak, 1991; Borowitza et al., 1991;jahjono et al., 1994; Harker et al., 1996; Cordero et al.,996; Sarada et al., 2002). Thus, for the optimal man-gement of these processes and for the enhancement ormprovement of new processes, a fast analysis method

re larger than 30 �m. Regarding the homogeneity ofhe population, it can also be observed that the flagel-ated cells corresponding to fast growing young cellsre more homogeneous, whereas when the culture istressed cells of slow growth (palmeloids cells andysts) appear and the homogeneity of the populationecreases because some flagellated and transition cellypes remain. From these data the great variation of celleight between different cell types can also be quanti-ed. They vary from 500 pg/cell for flagellated cells, to000 pg/cell for palmeloid cells, and even 3000 pg/cellor cysts; a direct general relationship between cellumber and dry weight measurements cannot thereforee established.

The development of fast characterisation methodsllows us to analyse the behaviour of H. pluvialis cul-ures in outdoor conditions. The experiments carriedut highlight the importance of irradiance and nitrate


concentration on the growth and astaxanthin accumula-tion of H. pluvialis. The irradiance or light availabilityis a function of location and design of the reactor. Theuse of small tube diameter horizontal photobioreac-tors enhances the irradiance on the reactor surface,thereby increasing the growth rate and the biomassand astaxanthin productivity of the reactor. Thus, themean daily irradiance on the tubular photobioreactorwas 2.8 times higher than in the bubble column, andthe maximum daily biomass productivity in the tubu-lar photobioreactor was 0.55 g/L day, as opposed to0.12 g/L day measured in the bubble column. The bub-ble column was strongly light limited, the maximumbiomass productivity in this reactor was measured atthe maximum average irradiance of 70 �E/m2s. Onthe other hand, the tubular photobioreactor was lightsaturated, and the maximum biomass productivity of0.55 g/L day was measured at an optimal average irra-diance of 130 �E/m2s. Thus, the biomass yield inlight was higher in the bubble column, 2.15 g/E, thanin the tubular photobioreactor, 0.61 g/E, although thebiomass productivity was higher in the latter. Theproductivity measured was higher than those refer-enced for the production of biomass under autotrophicconditions, 0.065 g/L day (Fabregas et al., 2001), het-erotrophic conditions, 0.17 g/L day (Hata et al., 2001),mixotrophic conditions, 0.20 g/L day (Zhang et al.,1999), and also higher than had been reached for thisstep on an industrial scale, 0.052 g/L day (Olaizola,2000). Similar biomass productivity of 0.58 g/L day hasodrtra(tclcpcotbar

Boussiba et al., 1999; Boussiba and Vonshak, 1991;Fabregas et al., 2001; Lee and Soh, 1991; Orosa etal., 2005). Moreover, some studies on the influenceof other stress factors such as light, acetate, dissolvedoxygen, pH, salinity, etc. (Kobayashi et al., 1992;Orosa et al., 2001) concluded that they might indi-rectly affect nitrogen consumption by the cells. Thedata reported in the present work support the cumu-lative character of both factors in the induction ofastaxanthin accumulation as has recently been reported(Del Rıo et al., 2005). Moreover, synthesis of astax-anthin in Haematococcus is accompanied by morpho-logical and biochemical changes (Boussiba, 2000).Our results show that cell weight increases six-foldfrom flagellated cells to astaxanthin-rich cyst cells.This value is similar to previously reported four-foldincreases in cell size from small green motile cellsto large red resting haematocysts, induced by severestress conditions (Boussiba and Vonshak, 1991), andlarger than the mere 40% increase reported undercontinuous operation (Del Rıo et al., 2005). Theselarge differences between batch and continuous oper-ation demonstrate the influence of culture conditionsand the importance of a complete characterisation ofthe cultures on both microscopic and macroscopicscale.

In short, this work demonstrates that the two anal-ysis methods proposed allow in-depth characterisationof Haematococcus pluvialis cultures whatever the mor-phology of the cells. This characterisation is fast andptsotpdk2atsiccbpt

nly been referenced in the laboratory under high irra-iance conditions (Garcıa-Malea et al., in press). Asegards astaxanthin, productivity in the tubular pho-obioreactor was 4.5 mg/L day, i.e. higher than thoseeported by Lee and Ding (1995) (0.1 mg/L day); Leend Soh (1991) (3.5 mg/L day); and Olaizola (2000)2.2 mg/L day), obtained under different stress condi-ions, but lower than the 5.6 mg/L day achieved underontinuous culture (Del Rıo et al., 2005). The higheright availability also favoured the accumulation ofarotenoids/astaxanthin in the biomass from the tubularhotobioreactor. Under nitrate saturation conditions,arotenoids were accumulated due to high irradiancen the reactor surface, but the major accumulationook place when the nitrate concentration decreasedelow 5 mM. Stimulation by nitrogen starvation ofstaxanthin accumulation in Haematococcus cells haseceived considerable attention (Borowitza et al., 1991;

rovides both the macroscopic and microscopic data ofhe culture. Fast characterisation methods allow analy-is of the behaviour of fast changing outdoor culturesf H. pluvialis. Tubular photobioreactors are showno be preferable to bubble columns for the outdoorroduction of biomass and/or astaxanthin. For the pro-uction of biomass the nitrate concentration must beept higher than 5 mM, producing flagellated cells of0 �m diameter. For the massive accumulation ofstaxanthin the nitrate concentration must be lowerhan 5 mM. Under nitrate depletion conditions cell divi-ion stops and astaxanthin content increases, and rest-ng cells as large as 30 �m diameter are obtained. Theontinuous production of astaxanthin under low nitrateoncentrations, ranging from 2 to 5 mM has recentlyeen referenced (Del Rıo et al., 2005), and the tubularhotobioreactor is the most adequate photobioreactoro perform this operation in outdoor conditions.


Acknowledgements

This research was supported by the Ministeriode Ciencia y Tecnologıa (PPQ 2001-3822-C02-02;PPQ 2001-3822-C02-01) and Junta de Andalucıa, PlanAndaluz de Investigacion III (CVI 173, CVI 263).

References

Acien, F.G., Garcıa, F., Sanchez, J.A., Fernandez, J.M., Molina,E., 1998. Modelling of biomass productivity in tubular photo-bioreactors for microalgal cultures: effects of dilution rate, tubediameter and solar irradiance. Biotech. Bioeng. 58 (6), 605–616.

Arias, R., Lee, T.C., Logendra, L., Janes, H., 2000. Correlation oflycopene measured by HPLC with the L*a*b* color readings ofa hydroponic tomato and the relationship of maturity with colorand lycopene content. J. Agric. Food Chem. 48, 1697–1702.

Borowitza, M.A., Huisman, J.M., Orborn, A., 1991. Culture ofastaxanthin-producing green alga Haematococcus pluvialis. I.Effects of nutrients on growth and cell type. J. Appl. Phycol. 3,295–304.

Boussiba, S., 2000. Carotenogenesis in the green alga Haematococ-cus pluvialis: cellular physiology and stress response. PhysiolPlanta 108, 111–117.

Boussiba, S., Vonshak, A., 1991. Astaxanthin accumulation in thegreen alga Haematococcus pluvialis. Plant Cell Physiol. 32,1077–1082.

Boussiba, S., Bing, W., Yuan, J.P., Zarka, A., Chen, F., 1999. Changesin pigments profile in the green alga Haematococcus pluvi-alis exposed to environmental stresses. Biotechnol. Lett. 21,601–604.

Chaumont, D., Thepenier, C., 1995. Carotenoid content in growingcells of Haematococcus pluvialis during a sunlight cycle. J. Appl.

C

C

D

D

D

F

Garcıa, F., Contreras, A., Acien, F.G., Fernandez, J.M., Molina,E., 1999. Use of concentric tube airlift photobioreactors formicroalgal outdoor mass cultures. Enzyme Microb. Technol. 24,164–172.

Garcia-Malea, M.C., Acien, F.G., Brindley, C., Del Rıo, E.,Fernandez, J.M., Molina, E., 2005. Modelling of growth andaccumulation of carotenoids in Haematococcus pluvialis as afunction of irradiance and nutrients supply. Biochem. Eng. J. 26,107–114.

Garcıa-Malea, M.C., Acien, F.G., Fernandez, J.M., Ceron, M.C.,Molina, E., in press. Continuous production of green cells ofHaematococcus pluvialis: modeling of the irradiance effect.Enzyme Microb. Technol.

Gong, X., Chen, F., 1998. Influence of medium components on astax-anthin content and production of Haematococcus pluvialis. Proc.Biochem. 33 (4), 385–391.

Guerin, M., Huntley, M.E., Olaizola, M., 2003. Haematococcusastaxanthin: applications for human health and nutrition. TrendsBiotechnol. 21 (5), 210–216.

Harker, M., Tsavalos, A.J., Young, A.J., 1996. Factors responsible forastaxanthin formation in the chlorophyte Haematococcus pluvi-alis. Biores. Technol. 55, 207–214.

Hata, N., Ogbonna, J.C., Hasegawa, Y., Taroda, H., Tanaka, H.,2001. Production of astaxanthin by Haematococcus pluvialis in asequential heterotrophic–photoautotrophic culture. J. Appl. Phy-col. 13, 395–402.

Kobayashi, M., Kakizono, T., Nagai, S., 1992. Effect of car-bon/nitrogen (C/N) ratio on encystment accompanied with astax-anthin formation in a green alga, Haematococcus pluvialis. J.Ferment. Bioeng. 74, 403–405.

Kobayashi, M., Kurimura, Y., Kakizono, T., Nishio, N, Tsuji, Y,1997. Morphological changes in the life cycle of the green algaHaematococcus pluvialis. J. Ferment. Bioeng. 84 (1), 94–97.

Lee, Y.K., Ding, S.Y., 1995. Effect of dissolved oxygen partial pres-sure on the accumulation of astaxanthin in chemostat cultures

L

L

M

M

M

O

Phycol. 7, 529–537.IE Central Bureau, 1986. Colorimetry, second ed. CIE Publica-

tion no. 15.2, Wien, Kegelgasse 27 A-1030, CIE Central Bureau,Austria.

ordero, B., Otero, A., Patino, M., Arredondo, B.O., Fabregas, J.,1996. Astaxanthin production from the green alga Haematococ-cus pluvialis with different stress conditions. Biotechnol. Lett.18, 213–218.

avies, D.H., 1976. Carotenoids. In: Goodwin, T.W. (Ed.), Chem-istry and Biochemistry of Plant Pigments, vol. 2. Academic Press,New York, pp. 38–165.

el Campo, J.A., Moreno, J., Rodrıguez, H., Vargas, M.A., Rivas,J., Guerrero, M.G., 2000. Carotenoid content of chlorophyceanmicroalgae: factors determining lutein accumulation in Muriel-lopsis sp. (Chlorophyta). J. Biotech. 76, 51–59.

el Rıo, E., Acien, F.G., Garcıa-Malea, M.C., Rivas, J., Molina-Grima, E., Guerrero, M.G., 2005. Efficient one-step productionof astaxanthin by the microalga Haematococcus pluvialis in con-tinuous culture. Biotech. Bioeng. 91 (7), 808–815.

abregas, J., Otero, A., Maseda, A., Domınguez, A., 2001. Two-stagecultures for the production of astaxanthin from Haematococcuspluvialis. J. Biotech. 89, 65–71.

of Haematococcus lacustris (Chlorophyta). J. Phycol. 31, 922–924.

ee, Y.K., Soh, C.W., 1991. Accumulation of astaxanthin in Haema-tococcus lacustris (Chlorophyta). J. Phycol. 27, 575–577.

orenz, R.T., Cysewski, G.R., 2000. Commercial potential forHaematococcus microalgae as a natural source of astaxanthin.TIBTECH 18, 160–167.

ackinney, J., 1941. Absorption of light by chlorophyll solutions. J.Biol. Chem. 140, 315–322.

ınguez-Mosquera, M.I., Gandul-Rojas, B., Gallardo-Guerrero,M.L., 1992. Rapid method of quantification of chlorophylls andcarotenoids in virgin olive oil by high-performance liquid chro-matography. J. Agric. Food Chem. 40, 60–63.

olina, E., Garcıa, F., Sanchez, J.A., Urda, J., Acien, F.G.,Fernandez, J.M., 1994. Outdoor chemostat culture of Phaeo-dactylum tricornutum UTEX 640 in a tubular photobioreactorfor the production of eicosapentaenoic acid. Biotechnol. Appl.Biochem. 20, 279–290.

laizola, M., Huntley, M.E., 2003. Recent advances in commercialproduction of astaxanthin from microalgae. In: Fingerman, M.,Nagabhushanam, R. (Eds.), Biomaterials and Bioprocessing. Sci-ence Publishers.


Olaizola, M., 2000. Commercial production of astaxanthin fromHaematococcus pluvialis using 25000-liter outdoor photobiore-actors. J. Appl. Phycol. 12, 499–506.

Orosa, M., Franqueira, D., Cid, A., Abalde, J., 2001. Carotenoidaccumulation in Haematococcus pluvialis in mixotrophicgrowth. Biotechnol. Lett. 23, 373–378.

Orosa, M., Franqueira, D., Cid, A., Abalde, J., 2005. Analysis andenhancement of astaxanthin accumulation in Haematococcuspluvialis. Biores. Technol. 96, 373–378.

Ortola, M.D., Londono, L., Gutierrez, C.L., Chiralt, A., 1998. Influ-ence of roasting temperature on physiochemical properties ofdifferent coffees. Food Sci. Technol. Int. 4, 59–66.

Richmond, A., Boussiba, S., Vonshak, A., Kopel, R., 1993. A newtubular reactor for mass production of microalgae outdoors. J.Appl. Phycol. 5, 327–332.

Sabri, S., Richelme, F., Pierres, A., Benoliel, A.M., Bongrand, P.,1997. Interest of image processing in cell biology and immunol-ogy. J. Inmunol. Methods 208, 1–27.

Sarada, R., Tripathi, U., Ravishankar, G.A., 2002. Influence ofstress on astaxanthin production in Haematococcus pluvialis

grown under different culture conditions. Process Biochem. 37,623–627.

Tjahjono, A.E., Hayama, Y., Kakizono, T., Terada, Y., Nishio, N.,Nagai, S., 1994. Hyper-accumulation of astaxanthin in a greenalga Haematococcus pluvialis at elevated temperatures. Biotech-nol. Lett. 16, 133–138.

Tredici, M.R., Materassi, R., 1992. From open ponds to alveolarpanels: the Italian experience in the development of reactors forthe mass cultivation of phototrophic microorganisms. J. Appl.Phycol. 4, 221–231.

Tripathi, U., Sarada, R., Ramachandra, S., Ravishankar, G.A., 1999.Production of astaxanthin in Haematococcus pluvialis culturedin various media. Biores. Technol. 68, 197–199.

Wulf, D.M., Wise, J.W., 1999. Measuring muscle color on beefcarcasses using the L*a*b* color space. J. Anim. Sci. 77,2418–2427.

Zhang, X.W., gong, Z.D., Chen, F., 1999. Kinetic models for astax-anthin production by high cell density mixotrophic culture ofthe microalga Haematococcus pluvialis. J. Indust. Microbiol.Biotechnol. 23, 691–696.

Comparative Analysis of the Outdoor Culture of Haematococcus Pluvialis in Tubular and Bubble Column...

Documents

Transcript of Comparative Analysis of the Outdoor Culture of Haematococcus Pluvialis in Tubular and Bubble Column...