Kinetic parameters of Scenedesmus obliquus (Chlorophyceae) growth and yield under different cultivation conditions in vitro

Scenedesmus obliquus isolation and culture activation

Scenedesmus obliquus isolation was performed based on both capillarity and successive dilution techniques (Lourenço, 2006). The experiment was carried out at the Algae Cultivation and Bioprospecting Laboratory (LACAL) of Federal University of Amapá (UNIFAP), after the species was subjected to taxonomic identification through optical microscopy. Phytoplankton samples collected in a natural aquatic ecosystem (Lagoa dos Índios, Amapá, Brazil, 0°01′54.260″N 51°06′09.665″W) were used in the study. Isolated cells were kept in test tubes filled with liquid synthetic culture medium (NPK 15:05:05) (Table 1). Cells were incubated at 23 °C, under continuous photoperiod. Isolates (11.7 ± 0.3 × 10⁴ cells mL⁻¹) were transferred to activation crops in Erlenmeyer (200 mL) flasks until the tests were run.

Experimental design and S. obliquus cultivation in photobioreactor

A three-level three-variable Box-Behnken factorial design was developed to determine the temperature (21 °C, 24 °C and 28 °C), pH (5.0, 7.0 and 9.0) and NPK dilution (12 mg/L, 16 mg/L and 20 mg/L) influence on S. obliquus cultivation (Table 2). The pH range was defined according to conditions naturally observed in Amazonian ecosystems (Barros et al., 2021). STATISTICA^® software (version 10, Statesoft–Inc., Tulsa, USA, trial version, 2011) was used in the experimental design for matrix elaboration. Independent variables were divided into X1, X2 and X3.

Inoculates (200 mL) were transferred to bioreactors—2L Erlenmeyer vials (in vitro) filled with NPK culture medium at different dilutions (12, 16 and 20 mg/L distilled water)—at the beginning of the experiments (Table 2). Initial axenic inoculum consisted of 11.7 ± 0.3 × 10⁴ cells mL⁻¹ in all assays. All experiments were simultaneously conducted in triplicate.

Inoculates were grown and kept under continuous photoperiod (28.5 µmol m⁻²s⁻¹), as already reported for S. obliquus (Breuer et al., 2013; Girard et al., 2014). Constant aeration was carried out by CO₂ (14 L min⁻¹) air diffusion through pumping (1/2 HP). Temperature in the growing room was controlled and monitored with the aid of digital thermometers (21 ± 0.4 °C, 24 ± 0.6 °C, 28 ± °0.9 °C). Different pH values were determined by adding HCl 1N and NaOH 1N (5 ± 0.7, 7 ± 0.5 and 9 ± 0.3) solutions. The pH in the culture medium was measured every 48 h with the aid of a pH-meter.

Cell density and yield

Cell density in the culture (average recorded for the three replicates) was measured on a daily basis in Neubauer hemocytometer by counting microalgal cells in optical microscope (Lourenço, 2006). Density values were used to calculate specific growth rate (µ), maximum cell concentration (N_max) and maximum yield (P_max).

Specific growth rate (µ, d⁻¹) was calculated based on the logarithmic phase regression of the cell growth curve. Maximum cell concentration (N_max, cells mL⁻¹) was represented by the maximum concentration value recorded in the logarithmic phase. Maximum yield (P_max, cel mL⁻¹ d⁻¹) was calculated through Eq. (1): (1)${\displaystyle p_{max}=\frac{\left(X_1-X_0\right)}{\left(t_1-t_0\right)}}$

wherein, X₁ is cellular concentration (cells mL⁻¹) at time t₁ (d) and X₀ (cells mL⁻¹) is cellular concentration at time t₀ (d) (Schmidell et al., 2001). Doubling time per day (T₂) was calculated through Eq. (2): (2)${\displaystyle T_2=\frac{ln2}{\mathrm{\mu }}.}$

Cellular density logistic model adjustment

Microalgae growth kinetics can often be modeled through the Verhulst logistic equation (Mazumdar et al., 2018; Yang et al., 2011), which is a model based on three important parameters: initial cell density at T₀ (cells mL⁻¹), theoretically reached maximum cell density (cells mL⁻¹) and specific growth rate (µ, day⁻¹).

Cell density (N_t), at any time (t), based on N_o as the initial cell density is given by Eq. (3) in this model: (3)${\displaystyle N_t=\frac{N_0.N_{Max}}{N_0+\left(N_{Max}-N_0\right)e^{-\mathrm{\mu }\mathrm{t}}}.}$

A transformed version of this Eq. (3) was used to estimate the growth parameters. It was based on adjusting the experimental raw data in the model of a nonlinear function (logistics) through the least square’s method (Mazumdar et al., 2018; Weisberg, 2005), which is expressed in Eq. (4): (4)${\displaystyle N_t=\frac{N_{Max}}{1+e^{\left(\mathrm{\theta }+\mathrm{\mu }\mathrm{t}\right)}}}$

wherein, N_t is cell count at time t, t is time after experiment has started, N_Max isthe maximum estimated cell count, N₀ is the initial cell count, µis maximum specific growth rate - θ was defined through Eq. (5): (5)${\displaystyle \theta =\ln \left(\frac{N_0}{N_{Max}-N_0}\right).}$

Statistical treatment

Model significance was tested in the R Statistic (version R Core Team, 2016) software. Logistic regression analysis was applied to assess the explainability and significance of models adjusted based on Eqs. (2) and (3).

Shapiro–Wilk normality test, followed by Levene/Bartlett tests, was carried out to assess experimental data homogeneity and homoscedasticity and to model the assessed kinetic parameters. Analysis of variance applied to repeated measurements was performed through Kruskal–Wallis test, followed by post-hoc dunn test, to find significant differences in growth parameters, in different experiments, since data distribution rejected normality. All tests were considered significant at α < 0.05.

STATISTICA software^® (version 10, Statesoft - Inc., Tulsa, USA, test version, 2011) was used for data analysis. Analysis of variance (ANOVA) was applied to transformed data to assess influence significance and interactions between independent variables, namely: temperature, pH and NPK dilution. Pareto graphs were plotted to find the significance of tested variables and to accomplish modeling through response surfaces.

Cell density and microalgal yield

The effect of different culture medium dilution, temperature and pH levels on S. obliquus cultivation was tested based on logistic models defined through Eqs. (2) and (3). Significant differences were observed in all assessed microalgal growth measurements between experiments. Maximum cell density (N_max) ranged from 32.6 to 101.5 × 10⁵ cells mL⁻¹; median was 42.1 × 10⁵ cells mL⁻¹. This variation led to significantly higher N_max in experiments 9 and 6, which were respectively defined as E9pH₉T₂₄C₂₀ (101.5 × 10⁵ cells mL⁻¹) and E6pH₅T₂₁C₂₀ (56.2 × 10⁵ cells mL⁻¹) (p = 0.04 and p = 0.004) (Table 3). Similarly, specific growth rates (µ) and maximum yield (Pmax) recorded medians of 0.24 d⁻¹ and 5.46 cells mL⁻¹ d⁻¹, respectively. Significant and higher values were also observed in experiments 9 and 6, respectively: E9pH₉T₂₄C₂₀ and E6pH₅T₂₁C₂₀ (p = 0.002 and p = 0.005).

Different experiments presented median cell-doubling time of 2.81 days. The longest doubling time (T₂) was recorded for experiment 5, i.e., E5pH₉T₂₁C₁₆ (5.2 days) (p = 0.005) - values were similar to those observed for assays 3 and 8, respectively: E3pH₅T₂₄C₁₂ (3.07 days) and E8pH₅T₂₈C₂₀ (3.8 days) (Table 3). The shortest cell doubling time was observed for experiment 9, i.e., E9pH₉T₂₄C₂₀ (1.4 ± 0.17 days) (p = 0.002) - values were similar to those recorded for experiments 1 and 6, respectively: E1pH₇T₂₁C₁₂ (2.0 ± 0.14 days) and E6pH₅T₂₁C₂₀ (1.6 ± 0.19 days).

Adjustment to the logistic model applied to microalgae cell density

The estimated model significantly adjusted itself to the experimental data, since r² ranged from 0.65 (experiment 5) to 0.99 (experiment 6) (Fig. 1). All modeled growth curves presented noticeably short or, practically, no lag phase (growth induction or latency). All experiments presented exponential growth phase between the 2nd and 7th experiment days (Figs. 1 and 2). The end of the stationary phase and the start of the decline phase (cell death) in the crop were observed between the 10th and 16th day of microalgal culture. The maximum concentration phase was detected when N_Max was also observed.

The logistic curve adjustment showed significant differences between Nt responses recorded for the nine models tested under different pH, temperature and N:P:K concentration conditions. Experiments E9pH₉T₂₄C₂₀ (N_max = 101.5 × 10⁵ cel mL⁻¹) and E6pH₅T₂₁C₂₀ (N_max = 56.2 × 10⁵ cel.mL⁻¹) (p = 0.002 and p = 0.007) recorded the highest cell density growth expression. (Fig. 2A).

The adjusted curves showed significant differences when the different factors (temperature × pH × medium dilution) (Figs. 2B, 2C and 2D) used for factorial planning worked in the same microalgal crop growth parameters. Higher microalgaceate cell density was observed in microalgal growth at 24 °C (N_max = 49.0 × 10⁵ cells mL⁻¹) (p = 0.01), in 20 mg/L culture-medium dilution (N_max = 46.5 × 10⁵ cells mL⁻¹) (p = 0.02). Crops stored at pH 9 tended to present higher cell density (p = 0.05).

Assessing S. obliquus cultivation based on response surface modeling

Factorial planning allowed assessing the influence of different factors on response variables: doubling time (T₂), maximum cell density (N_max), maximum yield (P_max) and specific growth rate (µ, day⁻¹). Pareto graphs pointed towards the strong effect of NPK medium dilution on variables T₂ (Fig. 3A) (p = 0.02), P_max (Fig. 3B) (p = 0.03) and µ(day⁻¹) (Fig. 3C) (p = 0.01). Analysis of variance (ANOVA) applied to these variables’ model showed coefficients of determination (r²) equal to 0.83, 0.79 and 0.87, respectively. Similarly, there was strong crop pH (p = 0.03) effect on N_max response in the S. obliquus culture (Fig. 3D). ANOVA applied to the model of variable N_max presented coefficient of determination (r²) equal to 0.87.

The response surface analysis (Figs. 4 and 5) showed that pH ranging from 8.5 to 9.0 had strong influence on biomass yield based on variable P_max. Wide temperature range (21 to 28 °C) affected S. obliquus yield (Fig. 4A). The higher the NPK concentration in the crop, the higher the biomass yield in comparison to the effect of different pH values (Figs. 4B and 4C).

According to the response surfaces, the highest NPK concentrations had more influence on cell doubling time T₂ than crop pH (Fig. 4D). Temperature ranging from 24 °C to 29 °C had strong effect on S. obliquus cell doubling when it was conditioned in 24 mL/L NPK (Fig. 4E). Temperature highly interacted with T₂ when temperature ranged from 24 °C to 28 °C. The pH value had influence on T₂ when it was higher than 8.5 (Fig. 4F).

The pH > 9.0 in association with temperatures higher than 24 °C had strong influence on S. obliquus specific growth rate (µ, day⁻¹) (Fig. 5A). Growth rate was assessed based on NPK dilution and temperature (Fig. 5B), as well as on NPK and pH dilution (Fig. 5C). Higher NPK concentrations had strong influence on the specific growth rate in both response surfaces. Temperatures between 24 °C and 29 °C, in association with pH higher than 9, affected S. obliquus growth rate (Figs. 5B and 5C).

Response surface applied to maximum cell density (N_max) based on crop temperature and pH showed interaction between temperatures > 27 °C and pH > 9.0 (Fig. 5D). Higher N_max was observed when temperature ranging from 27 °C to 29 °C was associated with higher NPK concentrations in S. obliquus culture (Fig. 5E). Crops conditioned at pH ranging from 7.5 to 9.5 presented high cell density when they were conditioned under the highest NPK concentration (Fig. 5F).