Bioresource Technology 100 (2009) 1847–1851

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /b ior tech

Predictive modeling of biomass production by Spirulina platensis as function ofnitrate and NaCl concentrations

Abuzer Çelekli *, Mehmet YavuzatmacaDepartment of Biology, Faculty of Art and Science, University of Gaziantep, 27310 Gaziantep, Turkey

a r t i c l e i n f o a b s t r a c t

Article history:Received 14 July 2008Received in revised form 16 September2008Accepted 22 September 2008Available online 6 November 2008

Keywords:BiomassModified modelsNitrateSaltSpirulina platensis

0960-8524/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.biortech.2008.09.042

* Corresponding author. Tel.: +90 3423171925; faxE-mail addresses: celekli.a@gmail.com (A. Çelek

hotmail.com (M. Yavuzatmaca).

Effects of nitrate (2.0, 2.5, and 3.0 g L�1) and salt (0.5, 1.0, 1.5, 2.0 g L�1) concentrations on biomass pro-duction by Spirulina platensis was examined in the Schlösser medium. The highest (p < 0.001) biomassyields and chlorophyll a content was observed at 2.5 g L�1 nitrate and 1.5 g L�1 NaCl as 3.495 g L�1 and29.92 mg L�1, respectively. Increment rate of biomass production was especially found between 72 and216 h. Modified Richards, Schnute, Logistic and Gompertz models was successfully predicted (r2 > 0.96and RSS P 0.003) biomass production by S. platensis as function of nitrate and salt concentrations. Lowresidual sum of squares (RSS) and high regression coefficients (r2) indicated that used models were wellfitted to the experiment data and it could be regarded as sufficient to describe biomass production of Spi-rulina sp. Biological variables i.e. production rate (l) and lag time (k) for S. platensis ranged 0.012–0.034 h�1 and 2.43–5.85 h, respectively from biomass production were successfully predicted by modi-fied Logistic model according to low RSS and F-testing value.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction Various environmental conditions such as nutrients, light inten-

Spirulina sp. is planktonic photosynthetic filamentous cyano-bacterium (Vonshak et al., 1982), identified by the main morpho-logical feature of the genus, i.e. the arrangement of multicellularcylindrical trichomes in a helix along the entire length of the fila-ments. Native people harvested biomass of Spirulina from ChadLake (Africa) and Texcoco Lake (Mexico) as a source of food for cen-turies (Vonshak, 1997). However, Spirulina platensis has been com-mercially cultivated due to its biotechnological importance since1970s (Vonshak, 1997). It is known that Spirulina is a versatileorganism due to its high nutritional value such as rich in proteincontent (Cohen, 1997; Colla et al., 2007), polyunsaturated fattyacids (c-linoleic acid) (Sajilata et al., 2008), pigments (Rangel-Yagui et al., 2004; Madhyastha and Vatsala, 2007), vitamin andphenolics (Colla et al., 2007; Ogbonda et al., 2007). Additionally,biomass of Spirulina has been performed to removal unwantedmaterials such as excess fertilizer, heavy metals, textile dyes andpesticides from wastewaters (Chojnacka et al., 2005; Solisioet al., 2006; Pane et al., 2008; Lodi et al., 2008). Moreover, microal-gae play a key role in natural food chain of aquatic systems, as afood source for herbivores such as larvae of many species of zoo-plankton, mollusks, crustaceans, and fishes (Lavens and Sorgeloos,1996).

ll rights reserved.

: +90 3423601032.li), mehmet_yavuzatmaca@

sity, and pH directly affect the growth of organisms (Danesi et al.,2004; Colla et al., 2007; Ogbonda et al., 2007). Modeling of organ-isms provides that understanding of its behavior under growthconditions such as temperature, pH, nutrients etc. (Zwieteringet al., 1990; Whiting, 1995; Çelekli et al., 2008). Results of modelsrevealed prediction of microbial development, optimization ofgrowth conditions, biovolume and biomass productions, and alsoestimation of microbial safety and quality in different environmen-tal conditions. Within the last decades, several growth models(Costa et al., 2002; Çelekli et al., 2008) have been used to predictbiomass and biovolume productions by microalgae during the cul-tivation. Growth curves of bacteria have significantly described bypredicting models (Zwietering et al., 1990; Bozkurt and Erkmen,2001). Several mathematical models such as Gompertz, Logistic,Richards, Schnute, and Stannard have been developed to describethe whole microbial growth curve (Zwietering et al., 1990;Whiting, 1995). Sigmoidal growth curve contain mathematicalparameters (a, b, c, . . .) rather than parameters with a biologicalmeaning (A, l, and k) are described by most of the equations. Threeparameters models such as modified Gompertz and Logisticmodels are among the widely used models which give biologicalparameters such as lag time (k), specific growth rate (l), andasymptotic value (A) (Zwietering et al., 1990; Bozkurt and Erkmen,2001; Çelekli et al., 2008).

Environmental factors closely affect algal growth due to theirphysiological requirements. This reason it is important to determinethe optimum culture conditions for the achievement of high yieldsof microalgae in standard media (Voltolina et al., 2005; Ogbonda

Table 1Experimental design and growth conditions for salt and nitrate concentrations

Runs Salt (g L�1) Nitrate (g L�1)

S1N1 0.5 2.0S1N2 0.5 2.5S1N3 0.5 3.0S2N1 1.0 2.0S2N2 1.0 2.5S2N3 1.0 3.0S3N1 1.5 2.0S3N2 1.5 2.5S3N3 1.5 3.0S4N1 2.0 2.0S4N2 2.0 2.5S4N3 2.0 3.0

S1N1 indicates 0.5 and 2.0 (g L�1) salt and nitrate concentrations, respectively.

1848 A. Çelekli, M. Yavuzatmaca / Bioresource Technology 100 (2009) 1847–1851

et al., 2007). Biomass production by Spirulina species are changed byvariation in the culture factors such as nitrogen source (Danesi et al.,2002; Colla et al., 2007), phosphate concentrations (Çelekli et al.,2008), initial biomass concentration (Costa et al., 2002), pH and lightintensity (Vonshak et al., 1982; Ogbonda et al., 2007). Nitrogen andNaCl are among the environmental factors that play a significantrole on the biomass production of Spirulina species (Vonshaket al., 1982; Richmond, 1992). It is known that change in environ-mental factors such as nitrate, phosphate, salt concentrations, tem-perature, etc. affected biomass production (Zeng and Vonshak,1998). Presence of high salt in the medium affected photosystem Iand II of Spirulina species due to its damage effect on protein degra-dation (Shipton and Barber, 1994). Source of nitrate is needed toproduce organic molecules such as protein and carbohydrate. Con-sequently, nitrate and sodium chloride are chosen to evaluate theeffect of these factors on biomass production by S. platensis. Theobjective of the study was (i) to examine the effect of nitrate and saltconcentrations on the biomass production by S. platensis, (ii) to pre-dict the biomass production by using modified equations ofGompertz, Logistic, Richards, and Schnute; (iii) to determine thebest model which describes the curve of biomass production.Besides, relationship between predict variable (algal biomass andbiological parameters) and response factors (nitrate and salt con-centrations) was evaluated and understand in batch culture.

2. Methods

2.1. Microorganism and growth conditions

The cyanobacterium used in the study, S. platensis obtainedfrom University of Ege, EBILTEM Culture Collection, was inoculatedon the Schlösser’s medium (Schlösser, 1982). Cells were main-tained in the culture medium of Schlösser, having the followingcomposition (per liter): 13.61 g NaHCO3, 4.03 g Na2CO3, 0.50 gK2HPO4, 2.50 g NaNO3, 1.00 g K2SO4, 1.00 g NaCl, 0.20 gMgSO4 � 7H2O, 0.04 g CaCl2 � 2H2O. All nutrients were dissolved indistilled water containing (per liter): 6 mL of metal solution(97 mg FeCl3 � 6H2O, 41 mg MnCl2 � 4H2O, 5 mg ZnCl2, 2 mgCoCl2 � 6H2O, 4 mg Na2MoO4 � 2H2O), 1 mL of micronutrient solu-tion (50.0 mg Na2EDTA, 618 mg H3BO3, 19.6 mg CuSO4 � 5H2O,44.0 mg ZnSO4 � 7H2O, 20.0 mg CoCl2 � 6 H2O, 12.6 mgMnCl2 � 4H2O, 12.6 mg Na2MoO4 � 2H2O) and 0.15 mg of B12 vita-min. The culture was incubated under 2.0 klux with measuringlight meter (Lutron Lx-130 model) continuous illumination usingcool, white fluorescent lamps.

Batch cultivations were carried out in 250 mL erlenmayer flaskscontaining 100 mL the medium, placed on an orbital shaker at90 rpm for 240 h. Algal developments were postulated by changesin the concentrations of NaNO3 (N = 2.0, 2.5, and 3.0 g L�1) andNaCl (0.5, 1.0, 1.5 and 2.0 g L�1) (Table 1) at 30 ± 2 �C. Each batchculture was inoculated with an initial Spirulina biomass concentra-tion (g L�1 dry weight) of 0.2 g L�1, previously adapted to concen-trations of nitrate and sodium chloride. Control medium wasprepared without cyanobacterium in Schlösser’s medium. Experi-ments were carried out triplicate.

As reported in previous studies (Zeng and Vonshak, 1998; Costaet al., 2002; Colla et al., 2007), Spirulina biomass value was calcu-lated through Optical Density (OD) measurements using a spectro-photometer (UV/VIS Jenway 6305) and a calibration curve of ODagainst dry weight (g L�1) of Spirulina biomass. During OD mea-surement, the tendency of clumping was prevented via using dilu-tion technique for dense culture. During incubation, biomassvalues were measured for 0.0, 0.5, 1.0, 24, 48, 96, 120, 144, 168,192, 216, and 240 h. Besides, the effluent was separated from thebiomass with acetate membrane filters (0.45 lm pore size,

Sartorious, Germany) and the filtrate was waited at 80 �C for over-night. Amount of chlorophyll a was determined by spectrophotom-eter at 665 nm and 750 nm with using methanol method(Youngman, 1978).

2.2. Statistical analyses

The non-linear modified Gompertz, Logistic, Richards, andSchnute equations (Table 2) were fitted to experimental data forbiomass production by S. platensis as described by Çelekli et al.(submitted for publication). The fitting procedure was performedusing commercial computer software SigmaPlot version 10.0.1 (Sy-stat Sofware, Inc., California, USA) via the Marquardt–Levenbergalgorithm. This logarithm is used to minimize the sum of squareof differences between experimental and predict data. Besides,the logarithm calculates biological parameters, and provides resid-ual data with Residuals sum of square (RSS) values.

F-test was used to compare predicted data obtained by usingmodels (Zwietering et al., 1990; Bozkurt and Erkmen, 2001). Inall cases, RSS values from Richards and Schnute models were same.Thus, in this test, RSS obtained from Richards model was taken asan estimate of the measuring error due to the lowest RSS values. F-values were calculated as

F ¼ ðRSS2 � RSS1Þ=ðDF2 � DF1ÞRSS1=DF1

Tested against FDF2�DF1DF1

where RSS1 is the RSS from modified Richards model, RSS2 is the RSSfrom the three parameter model (modified model of Logistic orGompertz), DF1 is the degrees of freedom from modified Richardsmodel and equals n points-4, and DF2 is the degrees of freedomfrom three parameter model and equals n points-3.

Calculated biological parameters [l is growth rate (h�1), k is lagtime (h), and A is asymptote value (biomass g L�1)] among factorswere also compared using ANOVA using the SPSS version 15.0(SPSS Inc., Chicago, IL, USA). Tukey’s honestly significant difference(HSD) multiple range test was also carried out to distinguish exam-ined groups. In order to evaluate the goodness of fitting, the pre-dicted data obtained from using equations (Table 2) was plottedagainst the experimental data and regression coefficients (r2) andresiduals sum of square (RSS) values between predicted and exper-imental data were calculated.

3. Results and discussion

The cyanobacterium, S. platensis obtained from University ofEge EBILTEM Culture Collection, was used in the study. The speciesshowed slightly spiral, left direction of helix, 7–10 lm width ofcylindrical trichome, and 33–48 lm diameter of spiral with pH tol-

Table 2Equation of models

Models Modified equation References

Gompertz y ¼ A � exp � exp l�cA

� �ðk� tÞ þ 1

� �� �Zwietering et al. (1990)

Logistic y ¼ A1þexp 4l

A ðk�tÞþ2½ �gf Zwietering et al. (1990)

Richards y ¼ A 1þ v � expð1þ vÞf� exp l

A ð1þ vÞ 1� 1v

� �� ðk� tÞ

� �g:ð�1=vÞ

Zwietering et al. (1990)

Schnute y ¼ l ð1�bÞA

� 1�b�expðA�kþ1�b�atÞ

1�b

h i1=bZwietering et al. (1990)

y, biomass; A, asymptote value; l, production rate; k, the lag time; t, time; and b, c,and t are constants.

Fig. 2. Changes in the biomass of Spirulina platensis at 1.0 g L�1 NaCl as function ofnitrate concentrations. Predicted values are obtained from modified Logistic model.

A. Çelekli, M. Yavuzatmaca / Bioresource Technology 100 (2009) 1847–1851 1849

erance range 9–11, also reported by Vonshak et al. (1982) andKomárek and Anagnostidis (1989).

Effects of cultivation time on biomass productions by S. platensisare showed in Figs. 1–4. Statistical analysis revealed that there wasno significant differences (p > 0.05) observed in the lag phase forbiomass production. Short lag time was consequence of acclimi-nated cell transferred to fresh medium with same conditions. Thisresult was also observed by Kim et al. (2007) who found that bio-mass production by S. platensis had a lag time. However, someresearchers reported that S. platensis had no lag phase (Collaet al., 2007; Radmann et al., 2007). After this phase, this biomassproduction showed tremendous increment (p < 0.05). Figs. 1–4indicated that this rate of biomass production was especially foundbetween 72 and 216 h. This result was in agreement with findingof Soletto et al. (2005) for S. platensis. On the other hand, longerproductivity rates for biomass between 100 and about 400 h werefound for Spirulina species (Danesi et al., 2002; Colla et al., 2007).

Impacts of nitrate concentrations, on the algal biomass through-out the cultivation are showed on Figs. 1–4. Statistical analysis(two-way ANOVA) indicated that biomass production by the cya-nobacterium increased sharply (p < 0.001) from lag time about0.220 g L�1 to about 1.435, 2.626, and 2.326 g L�1 at 2.0, 2.5, and3.0 g L�1 nitrate (N) values, respectively. Both 2.5 and 3.0 g L�1 Nfor all salt concentrations (Figs. 1–4) stimulated to produce higherbiomass values than at 2.0 g L�1 N. Biomass production at 2.5 g L�1

N for all salt concentrations, except 2.0 g L�1 NaCl, were found tobe higher than those nitrite concentrations (Figs. 1–4). Increasingnitrate concentration caused to increase biomass production at2.0 g L�1 NaCl, given on Fig. 4. Consequently, the highest(p < 0.001) biomass yields as 3.495 g L�1 was observed at2.5 g L�1 nitrate (Fig. 3). S. platensis used in the present study

Fig. 1. Effect of nitrate concentration on the biomass of Spirulina platensis at0.5 g L�1 NaCl. Predicted values are obtained from Modified Logistic model at% 95confidence interval.

achieved higher biomass concentration comparing to previousstudies concerning on S. platensis (Lodi et al., 2005; Narayanet al., 2005; Radmann et al., 2007). This optimum nitrate concen-tration as 2.5 g L�1 could be due to N:P weight ratio of 5 favoredimprovement of biomass production by S. platensis verified bysome researchers (Bulgakov and Levich, 1999; Radmann et al.,2007; Madhyastha and Vatsala, 2007). As reported previously,changes in the availability of nutrients affected biomass produc-tion by Spirulina species (Narayan et al., 2005; Colla et al., 2007;Yang et al., 2008). S. platensis produced considerable biomass val-ues in the present study with NaNO3 as nitrate source, comparedwith utilization of urea as a nitrate source (Yang et al., 2008),ammonium sulphate (Soletto et al., 2005), glycerol (Narayanet al., 2005), mixotrophic with glucose as carbon source (Lodiet al., 2005) and nitrogen regimes (Colla et al., 2007).

Effect of NaCl on the biomass values are plotted on Figs. 1–4.Statistical analysis (three-way ANOVA) indicated that optimumbiomass production was observed (p < 0.001) at 1.0 g L�1 NaCl.These results are in agreement with previous studies (Andradeand Costa, 2007; Colla et al., 2007; Madhyastha and Vatsala,2007). In the present study, S. platensis preferred higher nitrate

Fig. 3. Variation in the biomass of Spirulina platensis at 1.5 g L�1 NaCl as function ofnitrate concentrations. Predicted values are obtained from modified Logistic model.

Fig. 4. Influence of nitrate concentration on the biomass production by Spirulinaplatensis at 2.0 g L�1 NaCl. Predicted values are obtained from modified Logisticmodel.

Table 3Residual sum of square (RSS) and confidence interval (R2) values obtained fromcomparison of experimental data and predicted data for biomass production bySpirulina platensis

Runs Richard (4) Schnute (4) Logistic (3)a Gompertz (3)a

RSS R2 RSS R2 RSS R2 RSS R2

S1N1 0.018 0.978 0.018 0.978 0.023 0.974 0.028 0.969S1N2 0.013 0.999 0.013 0.999 0.036 0.998 0.133 0.991S1N3 0.025 0.983 0.025 0.983 0.053 0.964 0.059 0.959S2N1 0.014 0.999 0.014 0.999 0.042 0.996 0.094 0.991S2N2 0.033 0.998 0.033 0.998 0.040 0.997 0.110 0.993S2N3 0.018 0.999 0.018 0.999 0.043 0.996 0.112 0.991S3N1 0.002 0.999 0.002 0.999 0.016 0.992 0.032 0.984S3N2 0.013 0.999 0.013 0.999 0.080 0.996 0.225 0.989S3N3 0.029 0.998 0.029 0.998 0.033 0.997 0.060 0.994S4N1 0.018 0.982 0.018 0.982 0.037 0.963 0.042 0.958S4N2 0.015 0.995 0.015 0.995 0.015 0.995 0.019 0.994S4N3 0.003 0.999 0.003 0.999 0.037 0.997 0.111 0.991

Abbreviations are given at Table 1.a RSS and R2 show residuals sum of square and regression coefficient,

respectively.

Table 4F-values of modified Logistic and Gompertz models and F-table value

Runs F-values of the model

Logistic (3) Gompertz (3) F-table (F1,9) at a = 0.05

S1N1 0.65 1.67 5.12S1N2 1.92 27.69 5.12S1N3 1.59 4.08 5.12S2N1 2.00 17.14 5.12S2N2 0.53 7.00 5.12S2N3 1.74 15.67 5.12S3N1 2.63 45.00 5.12S3N2 2.51 48.92 5.12S3N3 0.36 3.21 5.12S4N1 1.54 4.00 5.12S4N2 0.00 0.80 5.12S4N3 2.76 108.00 5.12

Abbreviations are given at Table 1.

1850 A. Çelekli, M. Yavuzatmaca / Bioresource Technology 100 (2009) 1847–1851

concentration at 2.0 g L�1 NaCl than that of lower salt concentra-tions. This nitrate value might be needed to produce organic mol-ecules such as protein and carbohydrate. Previously, Anacystisnidulans enhanced activity of cytochrome oxidase at salt enrichedmedium (Molitor et al., 1986).

Statistical analyses revealed that chlorophyll a content of Spiru-lina sp was significantly affected (p < 0.001) by cultivation time,NaNO3 and NaCl concentrations. The maximum chlorophyll a valuewas found to be 29.92 ± 0.17 mg L�1 at 2.5 g L�1 nitrate and1.5 g L�1 NaCl. However, slightly higher chlorophyll a concentra-tion was found at 1.0 g L�1 NaCl. In the present study, remarkablerelationship (p < 0.01;r2 > 0.97) was observed between biomassand chlorophyll a values in S. platensis along cultivation. Similarchlorophyll a value was also found in the study of Lodi et al.(2005) for S. platensis, whereas it was lower than previous studies(Danesi et al., 2004; Rangel-Yagui et al., 2004; Madhyastha andVatsala, 2007).

The modified models, in given Table 2, were performed to de-scribe the biomass production by S. platensis in the modifiedSchlösser’s medium including four salt and three nitrate concen-trations (Table 1). The high r2 (0.959–0.999) and low RSS (0.003–0.225) (Table 3) indicated that used modified models (Table 1)were well fitted to the experiment data. Therefore, they could beregarded as sufficient to describe biomass production of S. platensisRSS and F values obtained from models for comparison of theexperimental and predicted data are given in Tables 3 and 4. Inall cases, RSS values from the Richards and Schnute models werethe same (Table 3) because these models are basically same, inagreement with results by Zwietering et al. (1990) and Bozkurtand Erkmen (2001). RSS values obtained from four parametersmodels (Richards and Schnute) were always lower (p < 0.01) thanthose from three parameter models (Gompertz and Logistic), alsomentioned by Zwietering et al. (1990) and Bozkurt and Erkmen(2001). That is, four parameter models reasonably gave the best fit-ting to describe the biomass production by S. platensis ModifiedGompertz, three parameter, model had the highest RSS values thanthose models. It was also found that there were no significant dif-ference (p > 0.05) exist between RSS values of Richards and Logisticmodels. Bozkurt and Erkmen (2001) mentioned that RSS value ob-tained from Richards was also found lower than those fromGompertz and Logistic models for prediction of Yersinia enterocoli-tica inactivation. RSS from Richards model was selected and took asthe estimate of measuring error for the calculation of F-test be-

cause of its lowest RSS value and high r2. The calculated F-test val-ues for Logistic and Gompertz models are given in Table 4. Logisticmodel gave the better fitting for biomass production by S. platensisthan that from Gompertz model for all cases according to lowerRSS and F-testing value (Tables 3 and 4). F-test revealed that Logis-tic model was accepted in all cases, whereas Gompertz model wasaccepted in 50% of the cases. In other words, there was no signifi-cant difference observed for F values between Richards and Logisticmodel. On the other hand, Gompertz model was rejected in 50% ofthe cases within 95% confidence interval. However, Zwieteringet al. (1990) and Bozkurt and Erkmen (2001) stated that Gompertzmodel showed the better fitting than Logistic model. The Logistic,three parameter, model could be regarded as sufficient to describebiomass production by S. platensis Three parameter models arepreferred rather than four parameter models because the threeparameter model is simpler and easier to use (Whiting, 1995;Zwietering et al., 1990; Bozkurt and Erkmen, 2001). Besides, bio-logical parameters from three parameter models can be easilyfound. From that point, Logistic model can be used to evaluate bio-mass production by S. platensis as function of nitrate and saltconcentrations.

Biological variables i.e, the growth rate (l), the lag time (k), andthe asymptote value (A) were calculated by modified Logistic mod-el (Table 5). The asymptote values for the cyanobacterium variedfrom 1.29 to 17.45 g L�1 for S3N1 and S1N1, respectively. Asymp-tote value is important to harvest yield of organism for biotechno-logical purposes. Through the cultivations, growth rate, l for S.platensis ranged from 0.012 to 0.034 h for S3N1 and S1N1, respec-

Table 5Mean values and standard deviations of biological parameters obtained from Logisticmodel; A is asymptote value (mg L�1); l is growth rate (h�1); k is lag time (h)

Biological parameters

A l k

S1N1 17.453 ± 0.482 0.0338 ± 0.001 5.849 ± 0.091S1N2 3.384 ± 0.005 0.019 ± 7.1 � 10�5 3.872 ± 0.041S1N3 2.178 ± 0.064 0.0063 ± 0.0001 4.037 ± 0.080S2N1 3.520 ± 0.112 0.0166 ± 0.0004 4.091 ± 0.074S2N2 3.518 ± 0.053 0.019 ± 0.0002 3.919 ± 0.055S2N3 3.248 ± 0.026 0.017 ± 0.0001 3.991 ± 0.005S3N1 1.285 ± 0.003 0.0062 ± 0.0001 0.012 ± 0.001S3N2 3.951 ± 0.005 0.0234 ± 0.0005 3.373 ± 0.056S3N3 3.454 ± 0.013 0.0159 ± 0.000 3.103 ± 0.052S4N1 1.559 ± 0.023 0.004 ± 2.8 � 10�5 2.817 ± 0.083S4N2 1.846 ± 0.022 0.0074 ± 2.3 � 10�5 2.432 ± 0.075S4N3 3.191 ± 0.005 0.017 ± 7.1 � 10�5 3.875 ± 0.063

Abbreviations are given in Table 1.

A. Çelekli, M. Yavuzatmaca / Bioresource Technology 100 (2009) 1847–1851 1851

tively (Table 5). This species had higher l than in previous studiesfor S. platensis (Lodi et al., 2005; Andrade and Costa, 2007; Collaet al., 2007). However, lower l was determined than for Scenedes-mus obliquus (Çelekli et al., 2008). It is consequence of species hav-ing different response to different environmental conditions. Lagtime (k) varied between 2.43 and 5.85 h obtained by the modifiedLogistic model. Earlier Similar results of lag time were also found infindings of statistical analysis.

Spirulina species have high potential of biomass production forvarious biotechnological purposes. This high amount of biomasscan be used as food source and feed supplement due to contentsof high nutritional value also can be used to treatment of un-wanted pollutants from artificial wastewaters (Solisio et al.,2006; Pane et al., 2008; Lodi et al., 2008).

4. Conclusions

Both four and three parameters models give good fits data forbiomass production by S. platensis as function of nitrate and saltconcentrations. Modified Richards, Schnute and Logistic modelswere significantly better than modified Gompertz model to predictbiomass production of the species. The predicted data obtainedfrom the model expressed optimum biomass value at 2.5 g L�1 ni-trate and 1.5 g L�1 salt concentrations. From low RSS and F-value,Logistic model can be used to evaluate biomass production by S.platensis under nitrate and salt concentrations. It is indicated thatSpirulina species achieved considerable high biomass value at thestudied environmental conditions.

Acknowledgements

We are grateful to Dr. Hüseyin BOZKURT for his help and con-structive comments for paper. The project (FEF.07.03) was sup-ported by Scientific Research Projects Executive Council ofUniversity of Gaziantep.

References

Andrade, M.R., Costa, J.A.V., 2007. Mixotrophic cultivation of microalga Spirulinaplatensis using molasses as organic substrate. Aquaculture 264, 130–134.

Bozkurt, H., Erkmen, O., 2001. Predictive modeling of Yersina enterocolitica inTurkish Feta cheese during storage. J. Food Eng. 47, 81–87.

Bulgakov, N.G., Levich, A.P., 1999. The nitrogen:phosphorus ratio as a factorregulating phytoplankton community structure. Arch. Hydrobiol. 146, 3–22.

Çelekli, A., Balcı, M., Bozkurt, H., 2008. Modelling of Scenedesmus obliquus; functionof nutrients with modified Gompertz model. Bioresour. Technol. 99, 8742–8747.

Çelekli, A., Yavuzatmaca, M., Bozkurt, H., submitted for publication. Effects ofphosphate concentrations and pH regimes on biomass production by Spirulinaplatensis. Biomass Bioenergy.

Chojnacka, K., Chojnacka, A., Górecka, H., 2005. Biosorption of Cr3+, Cd2+, and Cu2+

ions by blue-green algae Spirulina sp.: kinetics, equilibrium and the mechanismof the species. Chemosphere 59, 75–84.

Cohen, Z., 1997. The chemical of Spirulina. In: Vonshak, A. (Ed.), Spirulina platensis(Arthrospira), Physiology, Cell-Biology and Biotechnology. Taylor and Francis,London.

Colla, L.M., Reinehr, C.O., Reichert, C., Costa, J.A.V., 2007. Production of biomass andnutraceutical compounds by Spirulina platensis under different temperature andnitrogen regimes. Bioresour. Technol. 98, 1489–1493.

Costa, J.A.V., Colla, L.M., Filho, P.D., Kabke, K., Weber, A., 2002. Modelling of Spirulinaplatensis growth in fresh water using response surface methodology. World J.Microbiol. Biotechnol. 18, 603–607.

Danesi, E.D.G., Rangel-Yagui, C.O., Carvalho, J.C.M., Sato, S., 2002. An investigation ofeffect of replacing nitrate by urea in the growth and production of chlorophyllby Spirulina platensis. Biomass Bioenergy 23, 261–269.

Danesi, E.D.G., Rangel-Yagui, C.O., Carvalho, J.C.M., Sato, S., 2004. Effect of reducingthe light intensity on the growth and production of chlorophyll by Spirulinaplatensis. Biomass Bioenergy 26, 329–335.

Kim, M.K., Park, J.W., Park, C.S., Kim, S.J., Jeune, K.H., Chang, M.U., Acreman, J., 2007.Enhanced production of Scenedesmus sp. (green microalgae) using a newmedium containing fermented swine wastewater. Bioresour. Technol. 98,2220–2228.

Komárek, J.A., Anagnostidis, K., 1989. Modern approach to the classification systemsof cyanophytes. 4-Nostocales. Arch. Hydrobiol. (Suppl. 82), 247–345.

Lavens, P., Sorgeloos, P., 1996. Manual on the production and use of live food foraquaculture. FAO Fish. Papers 361, 7–42.

Lodi, A., Binaghi, L., Faveri, D.D., Carvalho, J.C.M., Converti, A., 2005. Fed-batchmixotrophic cultivation of Arthrospira (Spirulina) platensis (Cyanophyceae) withcarbon source pulse feeding. Ann. Microbiol. 55 (3), 181–185.

Lodi, A., Soletto, D., Solisio, C., Converti, A., 2008. Chromium (III) removal bySpirulina platensis biomass. Chem. Eng. J. 136, 151–155.

Madhyastha, H.K., Vatsala, T.M., 2007. Pigment production in Spirulina fussiformis indifferent photophysical conditions. Biomol. Eng. 24, 301–305.

Molitor, V., Erber, W.W.A., Peschek, G.A., 1986. Increased levels of cytochromeoxidase and sodium–proton antiporter in the plasma membrane of Anacystisnidulans after growth in sodium enriched media. FEBS Lett. 204, 251–256.

Narayan, M.S., Manoj, G.P., Vatchravelu, K., Bhagyalakshmi, N., Mahadevaswamy,M., 2005. Utilization of glycerol as carbon source on the growth, pigment andlipid production in Spirulina platensis. Int. J. Food Sci. Nutr. 56, 521–528.

Ogbonda, K.H., Aminigo, R.E., Abu, G.O., 2007. Influence of temperature and pH onbiomass production and protein biosynthesis in a putative Spirulina sp..Bioresour. Technol. 98, 2207–2211.

Pane, L., Solisio, C., Lodi, A., Mariottini, G.L., Converti, A., 2008. Effect of extracts fromSpirulina platensis bioaccumulating cadmium and zinc on L929 cells. Ecotox.Environ. Safe. 70 (1), 121–126.

Radmann, E.M., Reinehr, C.O., Costa, J.A.V., 2007. Optimization of the repeated batchcultivation of microalga Spirulina platensis in open raceway ponds. Aquaculture265 (1–4), 118–126.

Rangel-Yagui, C.O., Danesi, E.D.G., Carvalho, J.C.M., Sato, S., 2004. Chlorophyllproduction from Spirulina platensis: cultivation with urea addition by fed-batchprocess. Bioresour. Technol. 92, 133–141.

Richmond, A., 1992. Spirulina. In: Borowitzka, M.A., Borowitzka, L.J. (Eds.),Microalgal Biotechnology. Cambridge University Press, London.

Sajilata, M.G., Singhal, R.S., Kamat, M.Y., 2008. Fractionation of lipids andpurification of c-linolenic acid (GLA) from Spirulina platensis. Food Chem. 109,580–586.

Schlösser, U.G., 1982. Sammlung von Algenkulturen. Ber. Deutsch Bot. Ges. 95, 181–276.

Shipton, C.A., Barber, J., 1994. In vivo and in vitro photoinhibition reactions generatesimilar degradation fragments of D1 and D2 photosystem-II reaction centerproteins. Eur. J. Biochem. 220 (3), 801–808.

Soletto, D., Binaghi, L., Lodi, A., Carvalho, J.C.M., Converti, A., 2005. Batch and fed-batch cultivations of Spirulina platensis using ammonium sulphate and urea asnitrogen sources. Aquaculture 243, 217–224.

Solisio, C., Lodi, A., Torre, P., Converti, A., Del Borghi, M., 2006. Copper removal bydry and re-hydrated biomass of Spirulina platensis. Bioresour. Technol. 97,1756–1760.

Voltolina, D., Villa, H.G., Correa, G., 2005. Nitrogen removal and recycling byScenedesmus obliquus in semicontinuous culture using artificial wastewater anda simulated light and temperature cycle. Bioresour. Technol. 96, 359–362.

Vonshak, A., 1997. Spirulina platensis (Arthrospira): Physiology. Cell-Biology andBiotechnology. Taylor and Francis, London.

Vonshak, A., Abeliovich, A., Boussiba, S., Arad, S., Richmond, A., 1982. Production ofSpirulina biomass: effects of environmental factors and population density.Biomass 25, 341–349.

Whiting, R.C., 1995. Microbial modeling in foods. Crit. Rev. Food Sci. Nutr. 35, 467–494.

Yang, C., Liu, H., Li, M., Yu, C., Yu, G., 2008. Treating urine by Spirulina platensis. ActaAstronaut. 63, 1049–1054.

Youngman, R.E., 1978. Measurement of chlorophyll-a. Water Research Center, Tech.Rap. Tr-82.

Zeng, M.T., Vonshak, A., 1998. Adaptation of Spirulina platensis to salinity-stress.Comp. Biochem. Phys. A 120, 113–118.

Zwietering, M.H., Jongenburger, I., Rombouts, F.M., Van’t Riet, K., 1990. Modeling ofbacterial growth curve. Appl. Environ. Microbiol. 56, 1875–1881.