Algal Research 16 (2016) 387–398

Contents lists available at ScienceDirect

Algal Research

j ourna l homepage: www.e lsev ie r .com/ locate /a lga l

Effect of light quality supplied by light emitting diodes (LEDs) on growthand biochemical profiles of Nannochloropsis oculata and Tetraselmis chuii

Peter S.C. Schulze a,b,c, Hugo G.C. Pereira a, Tamára F.C. Santos a, Lisa Schueler a, Rui Guerra b, Luísa A. Barreira a,José A. Perales c, João C.S. Varela a,⁎a CCMar — Centre of Marine Sciences, University of Algarve, Campus de Gambelas, 8005-139 Faro, Portugalb CEOT — Centro de Electrónica, Optoelectrónica e Telecomunicações, Campus de Gambelas, 8005-139 Faro, Portugalc CACYTMAR — Centro Andaluz de Ciencia y Tecnología Marinas, University of Cádiz, Campus Universitario de Puerto Real, Cádiz, Spain

⁎ Corresponding author.E-mail address: jvarela@ualg.pt (J.C.S. Varela).

http://dx.doi.org/10.1016/j.algal.2016.03.0342211-9264/© 2016 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 21 January 2016Received in revised form 24 February 2016Accepted 27 March 2016Available online xxxx

Biochemical components obtained bymicroalgal biomass can be induced by specific wavelengths and processedto high value food/feed supplements or pharma- and nutraceuticals. Two biotechnologically relevantmicroalgae,Nannochloropsis oculata and Tetraselmis chuii, were exposed to non-tailored LEDs light sources emitting eithermono- or multichromatic light with low red but significant blue (b450 nm) photon content, or tailored lightsources with high blue or high red photon emissions: fluorescent light (FL), di- or multichromatic LED mixes.Growth ofN. oculata and T. chuii under tailored light resulted in a≈ 24% increase of the average biomass produc-tivity as compared to cultures lit by non-tailored light sources. FL induced the highest C:N ratios in both algae(N. oculata: 7.91± 0.09 and T. chuii: 11.29± 0.03), highest total lipid (48.37± 1.07%) inN. oculata and carbohy-drate (55.31± 1.02%) in T. chuii biomass. Among non-tailored light sources, monochromatic LEDs with emissionpeaks 465, 630 and 660 nm induced a≈ 29% increase of carbohydrates and a≈ 20% decrease of protein levels ascompared to LEDs peaking at 405 nm and cool- and warmwhite LEDs. In conclusion, as FL have low photon con-version efficiencies (PCE), particularly within the red wavelength range, LEDs emitting at the 390–450 and 630–690 nm wavebands should be combined for optimal carbon fixation, nitrogen and phosphate uptake.

© 2016 Elsevier B.V. All rights reserved.

Keywords:Tetraselmis chuiiNannochloropsis oculataLight emitting diode (LED)Light spectraCell physiologyBiochemical composition

1. Introduction

Microalgae can be biotechnologically processed into products, suchas bulk food, feedstock for food/feed supplements, nutraceuticals andcosmetics and they have also been considered to be a promising feed-stock for biofuel production [1,2]. Microalgal biochemical composition,and thus the levels of target added-value biomolecules, can changedue to shifting environmental parameters and/or phases of the algallifecycle [3,4].

Light quality and quantity supplied by sun- or artificial light is oneof themost important parameters for phototrophic organisms, as it isrequired for photosynthesis and the regulation of several cellularprocesses [1,2,5–9]. Even though sunlight is the most cost-effectiveenergy source to produce microalgae, artificial light may becomeeconomically feasible when production of high value biomoleculesis considered [1,10]. The key advantage of using artificial light formicroalgal production relies on a stricter regulation of parametersthat significantly impact cell proliferation, such as the photosynthet-ic photon flux density (PPFD), photoperiod and light spectra. Thistighter control of light availability can maintain cell growth 24 h

per day even in outdoor facilities, because nightly biomass lossesvia respiration are prevented when dark periods are precluded[11]. In addition, the control of light quality and intensity leads tolower variability and a higher control of microalgae biomass produc-tivity and target biochemical composition.

Artificial lighting in microalgal research and production is usuallyachieved by means of fluorescence lamps (FLs) [2] or, alternatively, bylight-emitting diodes (LEDs) [1,12–14]. LEDs are mercury-free andfast-responding artificial light sources. As they can be dimmed andhave long lifetimes (~50,000 h), LEDs can be used to cut costs both interms of energy and equipment maintenance [12–14]. The use of FLsor LEDs, however, comes at a cost and their improvement in terms ofphotosynthetic and electric efficiency is essential to obtain a widerand cheaper array of products from microalgae cultivated under artifi-cial light [1]. However, FLs are energetically inefficient, as they emitwide emission spectra, including wavelengths with low photosyntheticactivity for certain phototrophs, whereas LEDs can be designed to emitonly the required wavelengths.

Hence, LEDs can provide not only a more sustainable control of sup-plemental light during microalgal growth, but also adjust the biochem-ical composition of the biomass by means of single wavelengths atdifferent light intensities and/or pulse light frequencies [15–19]. Al-though studies about the cultivation of microalgae under different

388 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

light qualities supplied by LEDs have increased in recent years, there arestill important gaps in the knowledge of how microalgae respond tolight. The combined use of LEDs for microalgal cultivation or generalmetabolic response patterns was only partly investigated (reviewedby Schulze et al. [14]).

Fig. 1. Diagram of the experimental setup: (A) Photon distribution between 380 and 750 nm(D) two-colour mix adapted spectra and (E) Multi-colour mix spectra. (2-column fitting imagthe web version of this article.)

In order to fill in the gaps, found in the current state of the art, thepresent study aims to determine the effects of light quality on thegrowth rate, biochemical composition,morphological and physiologicalproperties of Nannochloropsis oculata and Tetraselmis chuii through theapplication of different (blue, red, white) LED and FL light sources.

of the light sources under study; (B) single colour LEDs; (C) mixed spectra light sources;e). (For interpretation of the references to colour in this figure, the reader is referred to

389P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

2. Materials and methods

2.1. Microalgae

N. oculata and T. chuii cultures were provided by the ExperimentalLaboratory of Aquatic Organisms (LEOA), University of Algarve(Portugal) and the Central Research Services in Mariculture (SCI-CM),University of Cádiz (Spain), culture collections, respectively.

2.2. Growth conditions

Experimentswere carried out in 1-L borosilicate glass flasks filled upwith 700 mL of algal culture (see Fig. 1A in supplemental data). A mod-ified F/2 medium was used as culture medium using 0.1 μm filteredseawater from the Atlantic shoreline of Cádiz (SW Spain) (salinity 38‰) supplemented with 4.39 mg L−1 EDTA, 3.15 mg L−1 FeCl3,100 mg L−1 NaNO3, 5 mg L−1 NaH2PO4, 0.098 mg L−1 CuSO4,0.105 mg L−1 ZnCl2, 0.10 mg L−1 CoCl2, 1.8 mg L−1 MnCl2 and0.63 mg L−1 Na2MoO4. The climate chamber was maintained at 22 ±2 °C and a PPFD of 100 μmol s−1 m−2 was adjusted using a 24-h photo-period (continuous lighting). Cultures were aerated with 0.2 μm-filtered air enriched with 5% CO2 (flow rate of 0.5 L min−1). Cultureswere harvested by centrifugation (3000 g, t = 10 min), freeze driedand stored in a desiccator until further analysis.

2.3. Light treatment

The qualitative spectral photon distribution of the light sourcesunder studywasmeasured viaOceanoptics 4000+(Fig. 1B–E)whereasthe quantitative PPFD was determined via Apogee MQ 100 quantumsensor. The spectral response of the used quantumsensorwas correctedby applying factors calculated according to [20].

Algae were exposed to single purple (LED 405), blue (LED 465),pure-red (LED 630), deep red (LED 660) as well as cool- and warmwhite phosphor-converted LEDs (LED CW, LED WW; Fig. 1B–C). Alter-natively, each microalgae culture was exposed to dichromatic lightwith a high (HRLB) or low (HBLR; Fig. 1D) red-to-blue ratios as wellas a multichromatic emission spectrum, mixing low (HBmix) or high(HRmix; Fig. 1E) red-to-blue ratios with “accessory”wavelengths as de-scribed in Schulze et al. [14]. Because of the high spectral matching be-tween emission spectrum of HBmix and HRmix and the absorptionspectrum of both algae (absorption spectrum provided in Fig. A1 sup-plemental data), these light sources are considered as tailored lightsources.

The LEDs used (SMD 5050 LED strips) were placed on a closure headof a cable channelmounted on awooden board (see also Fig. 1A). Refer-ence cultures for each experiment and inocula were grown under FL(Sylvania GroLux) used for phototrophic growth.

2.4. Analytical methods

Biochemical and morphological analysis of algal biomass was per-formed upon algae reaching the early stationary phase. Growth wasmonitored by cell counting (Neubauer chamber), optical density(OD680), medium nutrient concentration and pH. Algal suspensionwasfiltered and ash free dryweight (AFDW)was determined accordingto Zhu and Lee [21] using 0.5 M bicarbonate washing solution.

Residualmoisture in freeze dried biomasswas determined by drying10mg of biomass at 105 °C for 24 h to obtain the dryweight (DW). Sub-sequently the dried biomass was incinerated at 560 °C for 8 h and theash content was determined. The AFDW of the samples was calculatedby subtracting the moisture weight and the ash weight (AW) from theinitial biomass weight. Upon plotting AFDW against the OD680 of differ-ent experiments, time points and light treatments (n = 64 and n = 52for N. oculata and T. chuii, respectively), a linear correlation (r ≥ 0.97;p b 0.01) was found and used to estimate AFDW data on a daily basis.

Elemental analysis of nitrogen, hydrogen and carbon was assessedusing an elemental analyser (Vario EL iii, Elementar AnalysensystemeGmbH, Germany). Protein determination was carried out according toa modified Lowry et al. [22] method as detailed in Pomory [23]. Totallipids were determined according to a modified Bligh and Dyer [24]method as previously described in Pereira et al. [25]. Briefly, 0.8 mL dis-tilled water were added to 20 mg of microalgae biomass. Afterwards,2 mL of methanol and 1 mL of chloroform were added followed by ho-mogenizationwith an IKAUltra-Turrax in an ice bath for 60 s. Thereafter1 mL of chloroform was added and homogenized for 30 s, followed bythe addition of 1 mL of distilled water and homogenisation for another30 s. Phase separation was achieved by centrifugation (2000 g, t =2 min) and a defined volume of the chloroform layer was transferredinto a clean, pre-weighed tube and evaporated at 60 °C. Carbohydrateswere determined by subtraction.

2.5. Fatty acid methyl esters (FAME)

Fatty acids were determined according to Lepage and Roy [26] withmodifications as described in Pereira et al. [27]. Briefly, approximately30–40 mg of freeze dried algae were transferred into reaction vessels,1.5mL of amethanol/acetyl chloride (20:1)mixwas added and homog-enized with an IKA Ultra-Turrax. An aliquot of 1 mL n-hexane wasadded and vessels were incubated at 90 °C in a water bath for 1 h. Sam-ples were cooled down and sequentially extracted with n-hexane fourtimes. Afterwards, sodium sulphate was added and themixture was fil-tered. Finally, the hexane was evaporated until dryness at 55 °C under anitrogen atmosphere and resuspended in 500 μL of hexane for GC anal-ysis. FAME were analysed on an Agilent GC–MS (Agilent Technologies6890NetworkGC System, 5973 InertMass SelectiveDetector) equippedwith a DB5-MS capillary column (25 m × 0.25 mm, 0.25 μm film thick-ness, Agilent Tech). As carrier gas helium was used. For the identifica-tion and quantification of FAME the total ion mode and a Supelco® 37Component FAME Mix (Sigma-Aldrich, Sintra, Portugal) standard wasused.

2.6. Nutrients

For nutrient analysis, the culture medium was filtered (0.7 μmnominal pore glass fibre filter) and frozen (−20 °C) until analysis.Dissolved phosphate-based phosphorus (mg P-PO4

3− L−1; Merck1.17942.0001), nitrate-based nitrogen (mg N-NO3

− L−1; Merck1.14773.0001), and ammonium-based nitrogen (mg N-NH4

+ L−1;Merck 1.14752.0001) were determined by spectrophotometricalanalysis using a Spectroquant Nova 60 (Merck Chemicals).

2.7. Morphological traits

Sedimentation chambers with a volume of ~3 mL were filled withthe cell suspension and cell lengthwas determined by an inversemicro-scope connected to NIS-Element software version 4.1 (Nikon Cop.). Cellsurface area (n = 20) was analysed via ImageJ software version 1.48(Research Service Branch, NIH, Bethesda, MD). In order to determinecell length and area, pictures of a calibration master plate containing a1 mm-long bar, subdivided in 10 μm steps (Graticules Ltd.), were ap-plied to calibrate the software for each lens used.

2.8. Data treatment

To describe growth kinetics and nutrient consumption of cultures amodel proposed byRuiz et al. [28]was used. Thismodel allowed the cal-culation of growth parameters without being influenced by lag phaseduration or initiation of the stationary phase.

Statistical analyseswere performed by one-way ANOVA and Tukey'spost-hoc test using the GraphPad Prism 6 software. Linear relationshipswere assessed via a two-tailored Pearson's test (r). Qualitative data for

390 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

the determination of fatty acid profile determination were normalizedusing arcsine transformation before carrying out statistical data analy-sis. A significance level of α = 0.05 was used in all tests. Partial LeastSquares (PLS) analysis was preformed to describe relations and correla-tions between growth and morphological parameter of both algae ex-posed to different LED treatments using Addinsoft XLSTAT statisticssoftware (release 2014.5.03). PLS analysis is a generalization of a princi-ple component analysis (PCA) where a projection model is establishedto predict Y from X, deploying PCA scores for X. In the present study,the label data of the LED treatment but also respective quantitativespectral data were used as X-block whereas growth or biochemical/morphological parameters where used as the Y-block. PLS analysis re-sults in model coefficients for the variables, called PLS-weights. Theweights for the X-variables (w) specify their importance and howmuch they participate to the modelling of Y. The weights for the Y-variables (c) indicate which Y-variables are modelled in the respectivePLS model dimensions. Plotting X and Y in a wc-plot, the relationshipsbetween both variables are obtained. Through a PLS analysis, a variableimportance in projection (VIP) plot can be obtained, which is ameasureof how important each variable (e.g. wavelength) is for growth and/orbiochemical/morphological cell parameter in a PLS model. A VIP valuehigher than 0.8 indicates a statistically significant effect of the X-variable on the Y-variable.

In order to investigate the dependence of the growth and biochem-ical parameters on the relative proportions of blue and red, each lamphas been characterized by a parameter α (Eq. (1)):

α ¼ R–BRþ B

ð1Þ

where R and B represent the red and blue photon emissions (μmolphotons m−2 s−1) of the lamps in the range 630–690 nm and 350–450 nm, respectively. The R and B components were calculated as inte-grals of the spectral distributions in these ranges. The variable α takesvalues between −1 and 1, with the particular values α = −1 for blueonly, α = 1 for red only, and α = 0 for a 50/50% mix.

The relation between the growth and biochemical parameters andthe α values was sought in the form of a quadratic regression (Eq. (2)):

P ¼ α þ bα þ cα2 ð2Þ

where P is a generic parameter and a, b and c the fitting parameters. Themotivation behind this fit is to locate hypothetical optimal mixes of Rand B. In this case, the parameter would peak around α ≈ 0 andwould drop towards α = 1 and α = −1, a behaviour that could bewell modelled by a quadratic shape. On the other side, when the datadoes suggest an optimal mix, the correlation coefficient of the regres-sion will approach zero.

3. Results and discussion

3.1. Growth parameters

In all experiments, N. oculata showed no lag phase while T. chuiishowed a short lag phase within the first 24 h, followed by an exponen-tial phase (Fig. A2). Cell concentration plateaued at t = 96 h and t =120 h for N. oculata and T. chuii, respectively, indicating the beginningof the stationary phase. In order to compare growth parameters be-tween light treatments, a normalization of the quantitative data wasdone by setting all results and parameters as relative values to theones obtained with the FL control treatment (Fig. 2).

For both algae, cell count-based growth parameters (Pcell, μcell andXcell; Fig. 2A–B) tended to be higher under light sources emitting higherlevels of red photons (LED 660, FL, HRLB, HRmix HBLR or HBmix) ascompared to other light sources (LED 405, LED 465). This is probablycaused by red light-induced cell cycle acceleration as described

elsewhere for different species such as Arthrospira (Spirulina) platensis,Chlorella spp. or Galdieria sulphuraria [6,8,29–32].

Specifically designed lamps for photosynthesis with peak emissionwavelengths around 660 nm as in LED light combinations, such asHBLR, HRLB, HBmix and HRmix, provided higher biomass growth pa-rameters (PAFDW, μAFDW and XAFDW) for N. oculata and T. chuii as com-pared to non-optimal lamps such as LED CW or LED WW (Fig. 2C–D).Thus, only monochromatic LEDs or sources lacking a combination ofred and blue light were not able to promote maximum algal growthon their own, which is in agreement with data reported elsewhere[33]. Precise light tailoring (i.e., use of light sources whose emissionspectra closely match the absorption spectra of target microalgalspecies; e.g. HRmix, HBmix) may allow algae to absorb photons moreefficiently, promoting a higher μmax-AFDW as compared to most mono-chromatic and non-tailored (e.g. LED CW, LED WW) light sources test-ed. However, precise tailored lighting resulted in lower maximumbiomass concentration (XAFDW) as compared to only dichromatic tai-lored light, for example, HBLR.

Concerning nearmonochromatic light sources,N. oculata and T. chuiishowed similar biomass productivitieswhen lit by LED465 and LED 660(Fig. 2C–D). Interestingly, a similar trend has been reported forNannochloropsis oceanica [34], while Abiusi et al. [7] detected similarproductivities for Tetraselmis suecica under blue and red LEDs onlyuntil the third day of an experiment lasting 14 days. N. oculata reacheda higher XAFDW under LED 465 when compared to that of LED 660,whereas T. chuii showed the opposite effect (XAFDW,LED_660 N

XAFDW,LED_465; Fig. 2C–D). This finding is in agreement with Abiusiet al. [7], who reported that red LED-treated T. suecica cultures had ahigher maximal biomass concentration compared to those under bluelight. This suggests that algae continue to grow when exposed to theirpreferred light quality (e.g., red light for chlorophytes; [14]), whereasnon-optimal wavelengths can cause growth inhibitionwhen cell or bio-mass concentrations reach a threshold value, as observed for differentalgae [7,35,36].

Notably, both algae grown under FL grew better than those underLED light, as in general the latter displayed lower AFDW- and cellcount-based growth parameters than those of FL-lit cells (red dashedline in Fig. 2). FLs designed for growing photosynthetic organisms natu-rally emit photons with a balanced mix of blue and red light. However,FLs emit light at a grid- and ballast-dependent frequency. In the presentstudy, an electronic ballast was used causing a flashing light frequencyof 20KHz. Therefore, aflashing light effectmay also be at play, favouringgrowth by limiting non-photochemical quenching (NPQ) and self-shading [37,38].

The low amount of photons per wavelength emitted by carefully tai-lored light sources (e.g. HRmix or HBmix) might be prone to many dif-ferent optical obstacles absorbing a wide range of wavelengths (e.g.particles, organic molecules, or pigments [39]. This might lead to alower photon penetration depth in a photobioreactor (PBR), suppress-ing growth when cell concentrations increase beyond a thresholdvalue. On the other hand, a dichromatic tailored light source (e.g.HBLR), which emits the same amount of photons than a tailored lightsource, concentrates all photons on a narrow waveband, attaining bet-ter photon penetration into the algal culture. This might also explainwhy an alga like N. oculata displayed a lower maximum biomass con-centration when exposed to a monochromatic light source (e.g. LED660) emitting wavelengths matching a peak in in vivo absorption spec-tra, compared to a light source emitting at wavelengths (e.g. LED 465)that are not absorbed as readily. Although chlorophyll a, with absorp-tion peaks at 440, 625 and 680 nm [40,41], could be found in both spe-cies, the differences in the absorption spectra of both species is relatedto their dissimilar pigment compositions. T. chuii contains chlorophyllb with absorption peaks at ~650 nm and 460–480 nm, displaying,thus, a higher in vivo absorption within these ranges (Fig. A1), com-pared to N. oculata, which does not biosynthesize the latter pigment.On the other hand, the in vivo absorption peak of N. oculata from 487

Fig. 2. Normalized cell count and AFDW based growth parameter obtained by data modelling [50] for N. oculata (A, C) and T. chuii (B, D), respectively. Indices P, X, and μ representproductivity, maximum volumetric concentration and growth rate, respectively. Reference data (red dashed line) was obtained with cells growing under FL. Statistically higher orlower values as compared to those of the reference (FL) cultures are given as + and − signed letters, respectively. Original growth curves are provided in Fig. A2 (supplemental data).Statistical differences (p b 0.05) within growth parameters are indicated by different letters. (2-column fitting image).

391P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

to 490 nm (Fig. A1) can be related to antheraxanthin andviolaxanthin pigments [41,42]. These photoprotective pigmentsmay help Nannochloropsis spp. to protect the photosynthetic appara-tus against excess high energetic blue photons and thus utilize themmore efficiently [43]. Indeed, energies of blue photons are higherthan required for photosynthesis causing NPQ, which leads to pig-ment accumulation and energy quenching via the xanthophyll cycles(e.g. violaxanthin–astaxanthin–zeaxanthin cycle in chlorophytesand the diadinoxanthin–diatoxanthin cycle in diatoms) [44,45].

As a conclusion, it is likely that a low photon diversity might pro-mote maximum photon penetration depth. However, a certain photon

Table 1Nitrate-based nitrogen (RN; mg N-NO3

−; n = 6) and phosphate-based phosphorous (RP; mg P-experimental run. Nitrogen content in themedium at the harvesting time point is given as S(t).for both algae at time point of harvest (n = 3). Different letters indicate statistical differences

Treatment N. oculata

RNmg N-NO3

− L−1 d−1RPmg P-PO4

3− L−1 d−1S (t = 120 h)mg N-NO3

− L−1C:N ratio

LED 405 1.77 0.71 3.13 ± 0.13a 5.79 ± 0.10c

LED 465 1.56 0.39 0.30 ± 0.20d 6.89 ± 0.03b

LED 630 0.99 0.45 2.93 ± 0.25a 7.87 ± 0.17a

LED 660 1.19 0.45 1.75 ± 0.05b 7.53 ± 0.13a

LED CW 1.51 0.37 0.88 ± 0.38c 6.81 ± 0.18b

LED WW 1.52 0.41 0.18 ± 0.04d 6.68 ± 0.10b

FL 1.51 0.43 0.28 ± 0.04d 7.91 ± 0.09a

diversity is necessary to stimulate cell metabolic processes that are de-pendent on particular wavelengths, such as red light. In the presentstudy, but also in others [17,19,36,46], such condition was achievedthrough dichromatic rather than mono- or multichromatic tailoredlight sources. The need for suitable blue-to-red ratios is most probablyregulated via blue and red photoreceptors, such as phototropins andcryptochrome-like proteins, which appear to be important for a rapidresponse at the gene expression level in several species to changes inlight quality [14,47–49]. This in turn will influence the metabolism ofthe microalgal cultures under study and explain the differences ob-served in terms of their biochemical profile (see below).

PO43−; n = 4) consumption of N. oculata and nitrogen consumption of T. chuii during the

Intracellular carbon:nitrogen ratio (C:N ratio) and nitrogen-protein (N-prot) factors givenamong treatments within each parameter.

T. chuii

N-prot factor RNmg N-NO3

− L−1 d−1S (t = 96 h)mg N-NO3

− L−1C:N ratio N-prot factor

7.94 ± 0.56a 2.35 ≤0.1 ± 0.00a 8.30 ± 0.12c 7.81 ± 0.36a

6.00 ± 0.32b,c 2.61 ≤0.1 ± 0.00a 9.90 ± 0.27b 7.63 ± 0.22a

5.76 ± 0.09c,d 1.81 ≤0.1 ± 0.00a 8.80 ± 0.25c,d 7.36 ± 0.13b

6.14 ± 0.29b,c 2.09 ≤0.1 ± 0.00a 9.89 ± 0.26b 7.19 ± 0.17b

6.52 ± 0.27b 1.95 ≤0.1 ± 0.00a 7.72 ± 0.10c,d 7.71 ± 0.14a

6.53 ± 0.29b 2.15 ≤0.1 ± 0.00a 9.36 ± 0.12b,c 7.94 ± 0.03a

5.48 ± 0.41d 2.30 ≤0.1 ± 0.00a 11.29 ± 0.03a 7.47 ± 0.29b

392 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

3.2. Nutrient uptake

Levels of dissolved N-NO2− and N-NH4

+ throughout the experimentwere below detection limit, as almost all nitrogen was in the form ofN-NO3

−. A consistently higher N-NO3− utilization was found in both

algae grown under LED 405 and LED 465, when compared to thoseunder LED 630 and LED 660 light (Table 1). In addition, higher P-PO4

3−

utilization was measured in N. oculata cultures exposed to LED 405 ascompared to cells under red LEDs. Taken together, these results suggestthat nutrient utilization is stimulated by blue light, which in turnmaybeexplained by the activation of transporters/enzymes responsible forNO3

−, NO2− and PO4

3− uptake/assimilation [46,51]. However, the highgrowth rates of all T. chuii cultures (Fig. A2) triggered a very fast nutrientconsumption, consuming all detectable P-PO4

3− and N-NO3− within the

first 2 and 72 h of the experiment, respectively. Therefore, the differ-ences among light treatments of nutrient consumption by T. chuii cellswere less distinct than those concerning N. oculata cultures.

Indeed, algae follow different strategies to assimilate N-NO3−. For in-

stance, an obligatory photoautotroph, such as N. oculata, requires lightfor effective growth and nutrient utilization, whereas some greenalgae (e.g. chlorophytes such as T. chuii) can also assimilate NO3

−

under dark conditions [52–54]. This suggests that nutrient uptake ofalgae being capable to grow under heterotrophic conditions is lesslight quality-dependent compared to obligatory autotrophs. The slowergrowth of N. oculata caused a consistently lower C:N ratio among alltreatments compared to T. chuii (Table 1). Interestingly, a significantcorrelation between N-NO3

− uptake rates and C:N ratios among differ-ent light treatments was obtained for N. oculata (r = −0.755,p b 0.05), but not for T. chuii. This observation might be caused by amore pronounced competition between nutrient assimilation and car-bon fixation in obligatory photoautotrophs (N. oculata) when comparedto non-obligatory autotrophs, such as T. chuii [54]. In fact, a photosyn-thetically generated excited electron can only be used to provide energyfor one metabolic pathway (e.g. either nitrogen or carbon fixation;[55]). Therefore, light quality-induced higher nutrient uptake/assimila-tion (e.g. showing a high NO3

− consumption) would likely result in de-creased carbon fixation, via the Calvin cycle, and low C:N ratios inalgal cultures (e.g. LED 405 treatments; Table 1).

In conclusion, these results suggest that carbon fixation competesmore readily with nitrogen uptake/assimilation in N. oculata than inT. chuii under blue light. Conversely, N. oculata cultivated under LED630 and LED 660 lighting did not use all nitrogen available in the medi-um. Therefore, the absence of blue light may have prevented maximalgrowth due to partial inhibition of nutrient uptake [56].

3.3. Cell morphology

The morphology of microalgal cultures was significantly affected bylight quality. The mean cell weight of N. oculata was highest under FL,LED 465 and LED CW and in average ~25% lower under LED 405 andLED 660 treatments (Table 2). Cell length was also highest under FLtreatment and lowest under LED 405, LED 630 and LED 660 treatments.

Table 2Morphological cell parameters for different light treatments. Different letters indicate statistica

LED 405 LED 465 LED 630

N. oculataCell surface area (μm2) (n = 20) 7.52 ± 3.09c 8.46 ± 3.27b,c 8.06 ± 3.07Cell weight (pg cell−1) (n = 6) 6.13 ± 0.15c 8.37 ± 0.29a 6.54 ± 0.29Cell length (μm) (n = 200) 3.18 ± 0.38d 3.48 ± 0.44c 3.36 ± 0.47

T. chuiiCell surface area (μm2) (n = 20) 79.66 ± 3.84b 70.73 ± 6.02c 76.24 ± 5.0Cell weight (pg cell−1) (n = 6) 444.1 ± 25.8a 439.5 ± 6.4a 328.7 ± 12.Cell length (μm) (n = 200) 12.04 ± 1.04b 12.56 ± 1.01a 12.49 ± 1.2

Measurements of cellular length of T. chuii could not clearly distin-guish cultures with larger or smaller cells among treatments, since theshape of cells changed significantly, while cell lengths changed onlyslightly (≤10%). Similar to N. oculata, T. chuii cells exposed to FL showedalso the largest cell surface area (102.57 ± 4.24 μm2), whereas the re-maining treatments showed cells on average 30% smaller than the lat-ter. LED 405- and LED 630-treated T. chuii cells showed larger cellsurface areas compared to LED 660. Cells exposed to LED 405, LED 465and FLwere heaviest (444.1± 25.8, 439.5± 6.4 pg cell−1, respectively)as compared to those under LED 630, LED 660, LED CW and LED WWlight.

Algae undergo different morphologic changes during their cell andlife cycles, being affected by light quality and nutrient availability. Inthis study, a relationship between light quality, absorption spectra,nutrient consumption, cell size, and predominant growth stage-dependent cell morphology within a microalgal culture (here definedas “culture maturity”) was noticeable for both species. FL-treatedN. oculata cultures showed a pale-yellow colour at the end of the exper-iment while LED 465-, LED 630-, LED 660-, LED CW- and LED WW-treated cultures showed the same shift but at a slower pace. LED 405-treated cultures remained green, however (Fig. A3 in supplementarydata). Ageing N. oculata cultures have been described to become paleyellow-brown, indicating cyst formation [57]. The colour change isapparently related to an increasing ratio of yellowish carotenoids (e.g.zeaxanthin and violaxanthin) to greenish chlorophyll [5,43,58,59].Progression through the growth cycle and induction of specificcarotenoid:chlorophyll (car:chl) ratios were found to be dependentnot only on nutrient depletion from themedium, but also on light qual-ity and quantity [5,60,61]. In turn, nutrient uptake by algae is known tobe cell-size dependent. Small cells with low volume to surface ratiosusually show higher consumption rates than larger cells with highvolume to surface ratios [62]. Hence, algae tend to form naturallysmall cells in nutrient-rich environments and larger cells in nutrient-depleted growth media as described for Nannochloropsis sp. [57,62].However, in batch cultures, nutrient levels in the medium dependon uptake rates by the algae, which in turn are often controlled bythe wavelengths of incident light [46,51]. Thus, a light-dependentnutrient depletion or availability in the medium triggers algae toform either large or small cells, respectively. Therefore, this effectmight be more distinct in obligatory autotrophs, as their nutrient up-take is probably light quality-triggered than in algae growing alsounder heterotrophic conditions [54]. On the other hand, cell sizefluctuations also indicate changes in growth stage-dependent cellmorphology for some species, including N. oculata, as ageing culturestend to be dominated by larger cells, whereas a larger proportion ofsmaller cells in such cultures indicate growth limitation/arrest [8,9,63]. Interestingly, the greenish appearance of LED 405-treatedN. oculata cells in the present study (Fig. A3) and the low amountof cell aggregation (data not shown) indicate that no cysts devel-oped. Such result suggests that these cells underwent growth limita-tion or even growth arrest. This possibility is supported by the factthat the nitrogen content of the medium at the time of harvest wasthe highest value among all treatments (Table 1).

l differences among treatments within each parameter.

LED 660 CW LED WW LED FL

b,c 9.01 ± 3.51b 9.12 ± 3.53b 9.25 ± 3.43b 10.50 ± 3.91ab,c 5.31 ± 0.12d 7.56 ± 0.49a,b 6.97 ± 0.52b 8.00 ± 0.22a,bd 3.46 ± 0.44b 3.64 ± 0.40b 3.50 ± 0.37b 3.91 ± 0.41a

7b 67.35 ± 4.58c,d 66.94 ± 3.25c,d 65.58 ± 5.17d 102.57 ± 4.24a

6b 298.8 ± 29.1b 326.7 ± 19.5b 340.9 ± 25.0b 409.4 ± 28.1a

0a 12.69 ± 1.25a 11.36 ± 1.03c 12.39 ± 1.12a 12.60 ± 1.15a

393P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

3.4. Biochemical composition

For N. oculata and T. chuii the AW content averaged among all treat-ments at 5.59 ± 2.69 and 4.98 ± 1.75% of DW, respectively. N. oculatashowed similar lipid contents under all LED treatments, but significantlyhigher contents under FL (Fig. 3). Appropriate light quality can induce ahigh proportion of large, “mature” cells in the culture, which could ex-plain the significantly higher levels of total lipids in cultures grownunder FL as compared to LED-lit cells [63,64]. Nitrogen and phosphorouslimitation is known to increase cellular lipid and carbohydrate contents inNannochloropsis spp. [54,65–67]. In the present study, cultures exposed toLED 630 and LED 660 treatments showed a higher N-NO3

−t = 120 h levels

and also a higher biomass carbohydrate content. This suggests that lightsources with considerable amounts of photons with wavelengths ~400–450 nm (LED 405, LED CW and LED WW) have a negative effect oncarbohydrate:protein ratios. The high C:N ratio of FL-, LED 630- and LED660-lit N. oculata cells compared to other treatments, as well as theirhigher lipid or carbohydrate contents, could indicate nitrogen limitation[58]. However, the high N-NO3

−t = 120 h content in the medium suggests

that such nitrogen limitation is possibly due to constraints in N-uptake[51], rather than the availability of this nutrient, when cells are lit bysources containing only red photons. Although LED 465-treatedN. oculata cells showed also an elevated carbohydrate content, their C:Nratio and N-NO3

−t = 120 h was significant lower compared to LED 630

and LED 660. In this case, a light-induced N-uptake limitation cannot beassumed, but rather a limitation of the availability of this nutrient in thegrowth medium.

For T. chuii, all cultures showed anN-NO3−depletion at t=72h (data

not shown). Also P-PO43− was taken up by T. chuiiwithin 2 h. Therefore,

both cultures might have been potentially subject to N or P limitation.Under FL treatment, a high carbohydrate content and a significantlyhigher C:N ratio compared to algae illuminatedwith LEDswas detected,indicating the onset of nutrient limitation [68], a situation similar to FL-treated N. oculata. In fact, algae under N-deficiency tend to have higherC:N ratios and higher levels of carbohydrates or lipids instead of protein[69,70].

The high protein content, accompanied by low C:N ratios (Table 1),found in N. oculata and T. chuii cultures exposed to LED 405 and CWtreatments might also be attributed to high N-NO3

− consumption andlow growth rates induced by these light sources. Previously, bluelight-induced low carbohydrate:protein ratios was reported to becaused by an endogenous breakdown of carbohydrate reserves [51].Studies usingdischarge lampswithfilters letting through a considerableproportion of wavelengths below 450 nm [51,71,72] found also highprotein levels among different species. However, other studies using

Fig. 3. Biochemical composition in % of AFDW for N. oculata (A) and T. chuii (B). Statistical difcarbohydrates (dark grey bar) are indicated by different letters. (2-column fitting image).

LEDs with an emission peak at λ≈ 470 nm did not find any significantchanges in the carbohydrate:protein ratio [7,73].

Taken together, this suggests that wavelengths around ~400–450 nm promote high nitrogen uptake in microalgae, which is coupledwith high intracellular protein and low carbohydrate content. However,further research about different blue wavelengths and their effects onnutrient utilization, growth and biochemical composition onmicroalgaeis necessary to draw a final conclusion.

3.4.1. N-protein factorsThe observed nitrogen-protein (N-prot) factors in N. oculata and

T. chuii varied between 5.48–7.94 and 7.2–7.8, respectively. The N-protfactors of N. oculata tended to be higher than what was previously re-ported [58]. For T. chuii, to the authors' knowledge, the current studyis the first to report N-prot factors for this microalga. However, a re-markably lower N-prot factor of 4.37–5.08 has been described forTetraselmis gracilis [58]. N-prot factors are considered to be growthstage dependent: ageing cultures display higher N-prot factors (e.g. FLtreatment in the present study) than cultures in exponential growthphases. This effect is probably caused by decreasing nitrogen concentra-tions in themedium and increasing C:N ratios of aged cultures [58]. Thewide range of N-prot factors and C:N ratios among N. oculata cultures islikely caused by the ability of this genus to accumulate high intracellularnon-proteinaceous nitrogen (NPN) contents, which apparently changeswith growth conditions and growth phases [58].

In the present study, N. oculata and T. chuii showed the highest N-prot factors when treatedwith LED 405, LED CWand LEDWW,whereasthe lowest factors were found under FL treatment (Table 1). N-prot fac-tors and C:N ratios ofN. oculata cultures exposed to all tested treatmentsshowed a linear positive relationship (r=0.93; n=7, p b 0.05). There-by, cultures having high C:N ratio displayed a lowN-prot factor and lowN-NO3

− uptake (e.g. FL, LED 630, LED 660) and vice versa (e.g. LED 405,LED 465, LED CW, LEDWW). On the other hand, T. chuii showed no spe-cific trend between N-prot factors, C:N ratios and N-NO3

− consumptionratios among treatments (r=0.39; n=7, p N 0.05). In summary, theseresults suggest that LED light sources with considerable blue content(LED 405, 465, LED CW, WW) tend to promote high N-prot factors incontrast to light sources with high red content (LED 630, LED 660, FL),particularly in specieswith a highNPN content such asN. oculata. There-fore, the C:N ratio needs to be taken carefully into account when usingN-prot factors for total protein estimation in light quality experiments.

3.4.2. Fatty acid methyl esters (FAME) profileFor N. oculata, the major saturated fatty acids (SFA) detected were

myristic (C14:0) and palmitic (C16:0) acids, while palmitoleic acid

ferences (p b 0.05) within contents of total lipid (black bar), protein (light grey bar) and

394 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

(C16:1) was the main monounsaturated fatty acid (MUFA) observed.Regarding polyunsaturated fatty acids (PUFA), arachidonic (AA,C20:4n−6) and eicosapentaenoic (EPA, C20:5n−3) acids were pre-dominant. These results are in accordance with previous reports [16,74–76] (Table 3).

The FAME profile of T. chuii also matches what is reported in the lit-erature for the genus Tetraselmis [7,77]. Its fatty acid profile is dominat-ed by palmitic (SFA), palmitoleic and oleic (C18:1) acids whereas PUFAare mainly composed of linoleic (LA, C18:2) and α-linolenic (ALA,C18:3n−3) acids.

Concerning light treatment effects inN. oculata, the highest content ofSFA was found under LED CW and LED WW treatments, whereas LED405-treated cells showed the lowest levels (p b 0.05). Myristic acid wasfound to be highest in cultures under LED 405, LED 465 and FL. Exposureto LED 405 induced the highest contents of MUFA and PUFA. Moreover,EPA, was found to be highest in cells under LED 405, resulting in a lown−6 to n−3 (Σn−6/Σn−3) ratio (0.54 ± 0.12) whereas red LED lighttreatments showed the highest Σn−6/Σn−3 ratios. Worth to mentionis also the low MUFA content in the LED CW and LED WW treatments.The effects of different light qualities on the fatty acid composition inNannochloropsis sp. were previously investigated [16]. Although somefatty acids showed trends similar to the present study (i.e. no significantchange of C16:0 between LED 465 and 660), Das et al. [16] found largedifferences in EPA content among treatments, with white light inducingthe highest contents, a result that the present study did not confirm.These discrepant observations might be due to the use of a differentNannochloropsis species or different PPFDs employed by Das et al. [16].

T. chuii showed highest SFA content under LED 405 and FL treatments(p b 0.05; average: 37.65±0.81% of TFA) and significantly less under LEDCW (32.96 ± 0.86% of TFA). Similar to those under LED 465, FL-treatedT. chuii cultures showed also elevated content of MUFA (27.6 ± 0.65

Table 3Fatty acid profile of N. oculata and T. chuii exposed to different light qualities. Values are giventistical differences.

N. oculataFatty acid (%)

LED 405 LED 465 LED 630

C14:0 9.07 ± 0.38a 9.23 ± 0.44a 5.38 ± 0.73c

C16:0 33.82 ± 0.23c 41.86 ± 0.31c 38.92 ± 0.01b,c

C18:0 0.38 ± 0.05b 0.62 ± 0.12a,b 0.90 ± 0.29a

C16:1 33.82 ± 0.36a 33.54 ± 1.03a 33.99 ± 0.43a

C18:1 1.90 ± 0.11b 1.90 ± 0.05b 3.12 ± 0.23a

C18:2 (n−6) 0.72 ± 0.01a,b 0.38 ± 0.32b 1.49 ± 0.61a,b

C20:4 (n−6) 6.60 ± 0.81a 3.34 ± 0.32c 6.42 ± 0.36a

C20:5 (n−3) 13.59 ± 1.43a 9.10 ± 0.31b,c 7.82 ± 0.31b,c

∑ SFA 43.26 ± 0.56c 51.71 ± 0.25a,b 45.19 ± 1.00b,c

∑ MUFA 35.72 ± 0.24a 35.45 ± 0.98a,b 37.11 ± 0.66a

∑ PUFA 20.91 ± 0.61a,b 12.82 ± 0.95c 15.74 ± 1.29b,c

∑n−3 13.59 ± 1.43a 9.10 ± 0.31b,c 7.82 ± 0.31b,c

∑n−6 7.32 ± 0.82a,b 3.72 ± 0.64c 7.91 ± 0.98a

∑n−6/∑n−3 0.54 ± 0.12b 0.41 ± 0.06c 1.01 ± 0.08a

PUFA/SFA 0.48 ± 0.02a 0.25 ± 0.02b 0.35 ± 0.04b

T. chuiiFatty acid (%)

LED 405 LED 465 LED 630

C14:0 0.84 ± 0.24a 0.63 ± 0.03a 0.57 ± 0.02a

C16:0 36.54 ± 1.13a 33.22 ± 0.83b.c 33.39 ± 0.11b,c

C18:0 0.29 ± 0.08a,b 0.35 ± 0.04a,b 0.38 ± 0.04a,bC16:1 2.31 ± 0.25a,b,c 5.05 ± 0.18a,b 1.83 ± 0.26c

C18:1 15.05 ± 0.36b 20.01 ± 0.28a 13.24 ± 0.31c

C20:1 2.68 ± 0.35b,c 5.10 ± 0.36a 4.45 ± 0.27a

C18:2 (n−6) 14.16 ± 1.23a 6.81 ± 0.42b 13.69 ± 0.34a

C18:3 (n−3) 24.56 ± 1.62b,c 22.06 ± 0.54c 26.35 ± 1.55a,b

C20:5 (n−3) 3.45 ± 0.46d 6.55 ± 0.60a 5.84 ± 0.44a,b,c

∑ SFA 37.67 ± 1.12a 34.20 ± 0.81c 34.34 ± 0.06c

∑ MUFA 20.05 ± 0.43c 30.16 ± 0.77a 19.52 ± 0.84c,d

∑ PUFA 42.16 ± 1.17c 35.42 ± 1.39d 45.88 ± 0.77b

∑n−3 28.01 ± 1.19c 28.61 ± 1.04b,c 32.19 ± 1.11a

∑n−6 14.16 ± 1.23a 6.81 ± 0.42b 13.69 ± 0.34a

∑n−6/∑n−3 0.51 ± 0.06a,b 0.24 ± 0.01c 0.43 ± 0.03b

PUFA/SFA 1.12 ± 0.06b 1.04 ± 0.06c 1.34 ± 0.02a,b

and 30.16 ± 0.77% of TFA, respectively). LED CW-treated culturesdisplayed both the lowest MUFA and the highest PUFA contents(49.46 ± 1.07% of TFA), whereas the lowest PUFA levels were obtainedfor cells under LED 465 and FL (average: 35.06± 1.18 of TFA. The highestALA levels were found in cells under LED CW and LED WW (average:28.95 ± 1.83% of TFA) and lowest under LED 465 and LED 660(22.07 ± 0.51% of TFA). The ratio of Σn−6/Σn−3 was highest in LED660- and lowest under FL-treated T. chuii cultures (0.94 ± 0.04 and0.34±0.01%of TFA, respectively). LEDCW-lit cultures yielded thehighestPUFA/SFA ratio (1.5± 0.06), whereas cells under FL and LED 465 showedthe lowest ratio (0.92±0.05 and 1.04±0.06, respectively). Interestingly,for T. chuii, the effects of light quality on fatty acid composition have notbeen reported previously. However, Abiusi et al. [7] observed a similar re-sponse pattern for T. suecica between blue and red LED lights, whichwould be equivalent to LED 465 and 630 light sources in the presentstudy, respectively. For example, higher oleic acid content, accompaniedby higher MUFA, lower total PUFA and similar SFA contents, wasfound under blue light (λ ≈ 465 nm) compared to red LED light(λ ≈ 630 nm). These similar findings suggest an analogous response offatty acids on light qualities in the genus Tetraselmis. In fact, some authorsmentioned higher contents of PUFAswhen algae are growing actively [63,78,79]. Considering again that the C:N ratio reflects culture maturity, itcomes as no surprise that the PUFA content was lower under FL andLED 465 treatments compared to other LED light treatment. Since the re-sponse pattern of FAME profiles in microalgae under different light qual-ities is not completely understood, further research is required.

3.5. Multivariate analysis of the effect of light quality

For a global perspective about the effect of the light treatments onthe most important growth and morphological parameters of both

as percentage of total FAME (n = 3). Different letters within each fatty acid indicates sta-

LED 665 LED CW LED WW FL

6.42 ± 0.55b,c 7.09 ± 0.35b 5.15 ± 0.26c 9.44 ± 1.47a

41.19 ± 2.75a,b,c 45.60 ± 1.37a,b 48.56 ± 4.99a,b 38.65 ± 3.24c

0.81 ± 0.02a,b 0.55 ± 0.09a,b 0.65 ± 0.12a,b 0.79 ± 0.06a,b

30.49 ± 1.15a,b 28.63 ± 0.71b 26.06 ± 2.88b 28.56 ± 0.33b

3.93 ± 0.48a 2.07 ± 0.31b 1.94 ± 0.21b 4.18 ± 0.38a

1.80 ± 0.26a 0.86 ± 0.25a,b 0.89 ± 0.08a,b 1.46 ± 0.44a,b

6.19 ± 0.23a 4.58 ± 0.44b,c 6.04 ± 0.82a 6.27 ± 0.10a

7.74 ± 0.37c 9.51 ± 0.58b,c 8.64 ± 0.98b,c 10.73 ± 1.20b

48.42 ± 2.23a,b,c 53.25 ± 0.97a 54.35 ± 4.84a 48.88 ± 1.72a,b,c

34.41 ± 1.64a,b 30.69 ± 0.53b,c 28.00 ± 3.09c 32.74 ± 0.71a,b,c

15.73 ± 0.87b,c 14.96 ± 1.22b,c 15.57 ± 1.88b,c 18.46 ± 1.74b

7.74 ± 0.37c 9.51 ± 0.58b,c 8.64 ± 0.98b,c 10.73 ± 1.20b

7.99 ± 0.49a,b 5.45 ± 0.68a,b,c 6.93 ± 0.90a,b 7.73 ± 0.54a,b

1.03 ± 0.01a 0.57 ± 0.05b,c 0.80 ± 0.01a,b 0.72 ± 0.03b

0.33 ± 0.03b 0.28 ± 0.03b 0.29 ± 0.06b 0.38 ± 0.05a,b

LED 660 LED CW LED WW FL

0.60 ± 0.01a,b 0.76 ± 0.31a 0.69 ± 0.20a 0.91 ± 0.06a

34.04 ± 0.41b,c 31.85 ± 0.58c 34.54 ± 1.08a,b 36.48 ± 0.93a

0.42 ± 0.02a 0.35 ± 0.03a,b 0.37 ± 0.09a 0.23 ± 0.01b

2.62 ± 0.15a,b,c 2.37 ± 0.40a,b,c 3.18 ± 0.40a,b,c 5.12 ± 0.96a,b

15.77 ± 0.20b 12.49 ± 0.62c 16.03 ± 0.71b 19.78 ± 0.32a

3.40 ± 0.38b 2.57 ± 0.30c 2.57 ± 0.09c 2.71 ± 0.22c

14.91 ± 0.21a 15.71 ± 0.68a 7.59 ± 0.55b 4.42 ± 0.11c

22.08 ± 0.59c 27.82 ± 1.75a,b 30.08 ± 1.19a 25.91 ± 0.79b

5.93 ± 0.14a,b 5.93 ± 0.72a,b 4.76 ± 0.48b,c 4.36 ± 0.23c,d

35.05 ± 0.44b,c 32.96 ± 0.86c 35.61 ± 1.20b 37.62 ± 0.98a,b

21.79 ± 0.31c,d 17.42 ± 1.15d 21.77 ± 1.16c 27.60 ± 0.65b

42.92 ± 0.36b,c 49.46 ± 1.07a 42.43 ± 0.31c 34.69 ± 1.07d

28.01 ± 0.51c 33.75 ± 1.65a 34.84 ± 0.72a 30.27 ± 0.96b

14.91 ± 0.21a 15.71 ± 0.68a 7.59 ± 0.55b 4.42 ± 0.11c

0.53 ± 0.02a 0.47 ± 0.04b 0.22 ± 0.02d,c 0.15 ± 0.00d

1.22 ± 0.02b 1.50 ± 0.06a 1.19 ± 0.04b 0.92 ± 0.05c

395P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

algae under study a PLS analysis was conducted (Fig. 4A, B). In this anal-ysis, the X-variables (light treatment labels) that are most distancedfrom the origin (highest loading factors) have the greatest influenceon the Y-variable (growth, morphologic and biochemical parameters)located within the same or distinct different location. More variation isexplained by wc1 meaning that the loading factors on wc1 are morediscriminative than those of wc2. Fig. 4A shows the effect of the lighttreatments (X-variables) over cell count and AFDW-based growth pa-rameters (Y-variables). The plotted components wc1 and wc2 explain45.9% and 23.4% of the total variation, respectively (total: 69.3%). Fromthese results, it is apparent that tailored light combinations, such as FL,HBLR, HRLB, HBmix and HRmix provide the best biomass (PAFDW) andcell (PCell) productivities (PBN, PBC and PCN, PCT, respectively) sincethese parameters are all located in the right quadrant of the plot. Con-versely, variables corresponding to the non-optimal mono- andmultichromatic lamps treatments (LED 405, LED 465, LED 630, LED660, LED CW or LEDWW) are all located in the left side of the plot, op-posite to the productivity parameters indicating that these light sourcesare not as efficient in promoting growth. An exception was noted forT. chuii, since light sources containing considerable blue light emissions(λ N 430 nm) but none or insignificant atλ≈ 660 nm (e.g. LED 465, LEDCW, LED WW) appear to stimulate μcell of T. chuii, whereas the samegrowth parameter of N. oculata appears to be stimulated by lights emit-ting red photons (e.g. LED 660, HRLB).

Fig. 4. PLS plot of the light treatments (X-variables) and (A) the cell count and AFDWbased grofromN. oculata (blue crosses) and T. chuii (green crosses).Weights of X and Y variables indicatedand both algae on growth andbiochemical andmorphological parameter is given by a Variable IN-NO3

− uptake rate; PCell: Cells productivity; PAWFD: Biomass productivity; C:N: Carbon:nitrogento colour in this figure legend, the reader is referred to the web version of this article.)

In Fig. 4B, 71.5% of the total variability between the studied lightsources (LED 405, LED 465, LED 630, LED 660, LED CW, LED WW andFL; X-variable) and major morphological and biochemical effects (Y-variable) of both algae are explained through the components wc1(52.6%) and wc2 (18.9%). This plot evidences that the most importantlight treatments for high protein contents and N-prot factors in bothalgae and high total lipid content in T. chuiiwere those having consider-able photon emissions at 400–450 nm but none at ~660 nm (e.g. LED405, LED CW and LED WW; upper and lower left quadrant). On theother hand, LEDs with only red light (LED 630 and LED 660) were gen-erally associated with high carbohydrate content and C:N ratio inN. oculata (lower right quadrant). Cultures exposed to FL, which con-tains considerable amounts of blue light (λ ≈ 440 nm) and red light(λ≈ 660 nm) appears to exertmost influence on C:N ratios, cell surfacearea (T Area) and carbohydrates (CBH) in T. chuii but also total lipidsand cell surface in N. oculata (N Area).

The evaluation of VIP plots from additional PLS analyses with spec-tral data of each light treatment as X variables revealed that particularblue and red wavelengths between ~390–450 nm and ~630–690 nmare themost important ranges influencingmicroalgal growth, biochem-istry and morphology (Fig. 4C). Based on these findings, the optimalratio between both wavelengths ranges for biomass productivity(PBN, PBT), maximum biomass concentration (XBN, XBT) and the bio-chemical components proteins, carbohydrates and total lipids was

wth parameters or (B) morphological and biochemical parameters (Y-variables) obtainedarew and c, respectively (seeMaterials andmethods). The importance of eachwavelengthmportance Plot (VIP-) (C). Abbreviations: N:N. oculata; T: T. chuii; CBH: carbohydrates; RN:ratio; Area: Cell surface area. (2-column fitting image). (For interpretation of the references

Fig. 5. Alpha plot showing ratios between the most important red and blue wavelengths ranges on biomass productivity PBN; PBT (A), maximum biomass concentration XBN, XBT (B),proteins (C), carbohydrates (D) and total lipids (E). Abbreviations: N: N. oculata; T: T. chuii. (2-column fitting image). (For interpretation of the references to colour in this figure legend,the reader is referred to the web version of this article.)

396 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

analysed through an alpha plot (Fig. 5A, B, C, D and E, respectively).Briefly, for highest biomass productivity, a relative proportion ofα ≈ 0.6 can be observed for both species. For highest biomass concen-tration in the medium, N. oculata and T. chuii require a high blue andred light content, respectively. Both algae tend to synthesize proteinand carbohydrate under high blue and red light levels, respectively.Total lipids are low in both algae when exposed to high red light levels(α ≥ 0.9). However, T. chuii accumulates lipids more efficiently whenexposed to light with red light levels (α ≤ -0.8) as compared toN. oculata with highest productivity at higher red light levels (α ≈ 0.5).

4. Conclusions

The present study has indicated that the tested algae require mainlydichromatic light composed of blue (~390–450 nm) and red (~660 nm)photons, although other ranges have also contributed to growth as well.These ranges can be emitted by LEDs with significant efficiency (PCE~50%) and precision (full width at half-maximum; FWHM: ~20 nm),making them a promising alternative light source formicroalgal cultiva-tion compared to broad band emitting FLs (PCE ~30%,). Overall, the datahere presented suggest that red and blue photons stimulate carbon fix-ation andnutrient uptake, respectively. As a consequence, culturematu-rity and growth are affected, which, in turn, influence the morphologyand biochemical composition of the microalgal cell.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.algal.2016.03.034.

Contributions

Peter Schulze participated in the design of the experiments and wasthemain contributor for laboratory work, analysis, interpretation of thedata and drafting of the manuscript (Email: peter.schulze86@gmx.de).Hugo Pereira contributed to the analysis of biochemical components

in samples, interpretation of data, drafting the article, including artworkand revision of themanuscript (Email: hgpereira@ualg.pt). Tamára San-tos and Lisa Schueler helped in sample analysis aswell as in the draftingof the article, including artwork, technical support and revision of themanuscript (Email: tamarafilipasantos@gmail.com; lisam.schueler@yahoo.com). Rui Guerra contributed with his statistical expertise forproper interpretation of the data, He also participated in the draftingand critical revision of the manuscript and data processing (Email:rmfguerra@gmail.com). Luísa Barreira contributed to the present man-uscript with her knowledge of analytical chemistry and helped to thedesign this study, analyse and interpret data as well as to raise the nec-essary funds for the work (Email: lbarreir@ualg.pt). José Perales andJoão Varela contributed to the conception and design of the study, anal-ysis and interpretation of the data, critical revision of the article, provi-sion of study materials and funding as well as administrative,technical, and logistic support (Email: joseantonio.perales@uca.es).

Acknowledgements

Wewish to thankMargarida Castro fromCCMar, University of Algar-ve and Edward P.Morris from the department structure and functioningof aquatic ecosystems (EDEA, RMN214), University of Cádiz for theirsupport in statistical analyses and light measurements, respectively.This study was funded by the Foundation for Science and Technology(FCT, Portugal; CCMAR/Multi/04326/2013). Hugo Pereira is a PhD stu-dent funded by the Portuguese Foundation for Science and Technology(SFRH/BD/105541/2014).

References

[1] W. Blanken, M. Cuaresma, R.H. Wijffels, M. Janssen, Cultivation of microalgae on ar-tificial light comes at a cost, Algal Res. 2 (2013) 333–340, http://dx.doi.org/10.1016/j.algal.2013.09.004.

397P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

[2] A. Carvalho, S. Silva, J. Baptista, F.X. Malcata, Light requirements in microalgalphotobioreactors: an overview of biophotonic aspects, Appl. Microbiol. Biotechnol.89 (2011) 1275–1288, http://dx.doi.org/10.1007/s00253-010-3047-8.

[3] A. Otero, D. García, J. Fábregas, Factors controlling eicosapentaenoic acid productionin semicontinuous cultures of marine microalgae, J. Appl. Phycol. 9 (1997) 465–469,http://dx.doi.org/10.1023/A:1007930804367.

[4] S. Castro Araújo, V.M.T. Garcia, Growth and biochemical composition of the diatomChaetoceros cf.wighamii brightwell under different temperature, salinity and carbondioxide levels. I. Protein, carbohydrates and lipids, Aquaculture 246 (2005)405–412, http://dx.doi.org/10.1016/j.aquaculture.2005.02.051.

[5] A. Solovchenko, I. Khozin-Goldberg, L. Recht, S. Boussiba, Stress-induced changes inoptical properties, pigment and fatty acid content of Nannochloropsis sp.: implica-tions for non-destructive assay of total fatty acids, Mar. Biotechnol. 13 (2011)527–535, http://dx.doi.org/10.1007/s10126-010-9338-3.

[6] C. Koc, G.A. Anderson, A. Kommareddy, Use of red and blue light-emitting diodes(LED) and fluorescent lamps to grow microalgae in a photobioreactor, Isr. J.Aquacult. 65 (2013) (IJA_65.2013.2797, http://hdl.handle.net/10524/32474).

[7] F. Abiusi, G. Sampietro, G. Marturano, N. Biondi, L. Rodolfi, M. D'Ottavio, M.R. Tredici,Growth, photosynthetic efficiency, and biochemical composition of Tetraselmissuecica F&M-M33 grown with LEDs of different colors, Biotechnol. Bioeng. 111(2013) 956–964, http://dx.doi.org/10.1002/bit.25014.

[8] C.-G. Lee, B.Ø. Palsson, Photoacclimation of Chlorella vulgaris to red light from light-emitting diodes leads to autospore release following each cellular division,Biotechnol. Prog. 12 (1996) 249–256, http://dx.doi.org/10.1021/bp950084t.

[9] H. Oldenhof, V. Zachleder, H. Van Den Ende, Blue-and red-light regulation of the cellcycle in Chlamydomonas reinhardtii (Chlorophyta), Eur. J. Phycol. 41 (2006)313–320, http://dx.doi.org/10.1080/09670260600699920.

[10] S. Leu, S. Boussiba, Advances in the production of high-value products bymicroalgae, Ind. Biotechnol. 10 (2014) 169–183, http://dx.doi.org/10.1089/ind.2014.1518.

[11] D. Zou, K. Gao, J. Xia, Dark respiration in the light and in darkness of three marinemacroalgal species grown under ambient and elevated CO2 concentrations, ActaOceanol. Sin. 30 (2011) 106–112, http://dx.doi.org/10.1007/s13131-011-0096-5.

[12] M. Olle, A. Viršile, The effects of light-emitting diode lighting on greenhouse plantgrowth and quality, Agric. Food Sci. 22 (2013) 223–234 (http://www.ojs.tsv.fi/index.php/AFS/article/view/7897).

[13] US–DOE, Solid-State Lighting Research and Development, Multi-Year Program Plan(DOE/EERE-0961), Office of Energy Efficiency & Renewable Energy, United StatesDepartment of Energy, 2013 (http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/ssl_mypp2013_web.pdf).

[14] P.S.C. Schulze, L.A. Barreira, H.G.C. Pereira, J.A. Perales, J.C.S. Varela, Light emitting di-odes (LEDs) applied to microalgal production, Trends Biotechnol. 32 (2014)422–430, http://dx.doi.org/10.1016/j.tibtech.2014.06.001.

[15] Y. Zhao, J. Wang, H. Zhang, C. Yan, Y. Zhang, Effects of various LED light wavelengthsand intensities on microalgae-based simultaneous biogas upgrading and digestatenutrient reduction process, Bioresour. Technol. 136 (2013) 461–468, http://dx.doi.org/10.1016/j.biortech.2013.03.051.

[16] P. Das,W. Lei, S.S. Aziz, J.P. Obbard, Enhanced algae growth in both phototrophic andmixotrophic culture under blue light, Bioresour. Technol. 102 (2011) 3883–3887,http://dx.doi.org/10.1016/j.biortech.2010.11.102.

[17] W. Fu, Ó. Guðmundsson, G. Paglia, G. Herjólfsson, Ó. Andrésson, B. Palsson, S.Brynjólfsson, Enhancement of carotenoid biosynthesis in the green microalgaDunaliella salina with light-emitting diodes and adaptive laboratory evolution,Appl. Microbiol. Biotechnol. 97 (2013) 2395–2403, http://dx.doi.org/10.1007/s00253-012-4502-5.

[18] A. Lunka, D. Bayless, Effects of flashing light-emitting diodes on algal biomass pro-ductivity, J. Appl. Phycol. 25 (2013) 1679–1685, http://dx.doi.org/10.1007/s10811-013-0044-1.

[19] C. Yan, L. Zhang, X. Luo, Z. Zheng, Effects of various LED light wavelengths and inten-sities on the performance of purifying synthetic domestic sewage by microalgae atdifferent influent C/N ratios, Ecol. Eng. 51 (2013) 24–32, http://dx.doi.org/10.1016/j.ecoleng.2012.12.051.

[20] J. Ross, M. Sulev, Sources of errors inmeasurements of PAR, Agric. For. Meteorol. 100(2000) 103–125, http://dx.doi.org/10.1016/S0168-1923(99)00144-6.

[21] C. Zhu, Y. Lee, Determination of biomass dry weight of marine microalgae, J. Appl.Phycol. 9 (1997) 189–194, http://dx.doi.org/10.1023/A:1007914806640.

[22] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, Protein measurement with theFolin phenol reagent, J. Biol. Chem. 193 (1951) 265–275 (http://devbio.wustl.edu/InfoSource/ISPDFs/Lowry%201951.pdf).

[23] C.M. Pomory, Color development time of the Lowry protein assay, Anal. Biochem.378 (2008) 216–217, http://dx.doi.org/10.1016/j.ab.2008.04.015.

[24] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. J.Biochem. Physiol. 37 (1959) 911–917, http://dx.doi.org/10.1139/o59-099.

[25] H. Pereira, L. Barreira, A. Mozes, C. Florindo, C. Polo, C.V. Duarte, L. Custódio, J. Varela,Microplate-based high throughput screening procedure for the isolation of lipid-rich marine microalgae, Biotechnol. Biofuels 4 (2011) 1–12, http://dx.doi.org/10.1186/1754-6834-4-61.

[26] G. Lepage, C.C. Roy, Improved recovery of fatty acid through directtransesterification without prior extraction or purification, J. Lipid Res. 25 (1984)1391–1396 (http://www.jlr.org/content/25/12/1391.short).

[27] H. Pereira, L. Barreira, F. Figueiredo, L. Custódio, C. Vizetto-Duarte, C. Polo, E. Rešek,A. Engelen, J. Varela, Polyunsaturated fatty acids of marine macroalgae: potential fornutritional and pharmaceutical applications, Mar. Drugs 10 (2012) 1920–1935,http://dx.doi.org/10.3390/md10091920.

[28] J. Ruiz, Z. Arbib, P. Álvarez-Díaz, C. Garrido-Pérez, J. Barragán, J. Perales,Photobiotreatment model (PhBT): a kinetic model for microalgae biomass growth

and nutrient removal in wastewater, Environ. Technol. 34 (2013) 979–991,http://dx.doi.org/10.1080/09593330.2012.724451.

[29] H.-B. Chen, J.-Y. Wu, C.-F. Wang, C.-C. Fu, C.-J. Shieh, C. Chen, C.-Y. Wang, Y.-C. Liu,Modeling on chlorophyll a and phycocyanin production by Spirulina platensisunder various light-emitting diodes, Biochem. Eng. J. 53 (2010) 52–56, http://dx.doi.org/10.1016/j.bej.2010.09.004.

[30] C.H. Shu, C.C. Tsai, W.H. Liao, K.Y. Chen, H.C. Huang, Effects of light quality on the accu-mulation of oil in amixed culture of Chlorella sp. and Saccharomyces cerevisiae, J. Chem.Technol. Biotechnol. 87 (2012) 601–607, http://dx.doi.org/10.1002/jctb.2750.

[31] I.N. Stadnichuk, A.A. Bulychev, E.P. Lukashev, M.P. Sinetova, M.S. Khristin, M.P.Johnson, A.V. Ruban, Far-red light-regulated efficient energy transfer fromphycobilisomes to photosystem I in the red microalga Galdieria sulphuraria andphotosystems-related heterogeneity of phycobilisome population, BBA-Bioenerget-ics 1807 (2011) 227–235, http://dx.doi.org/10.1016/j.bbabio.2010.10.018.

[32] M. Sánchez-Saavedra, C. Jiménez, F. Figueroa, Far-red light inhibits growth but pro-motes carotenoid accumulation in the green microalga Dunaliella bardawil, Physiol.Plant. 98 (1996) 419–423, http://dx.doi.org/10.1034/j.1399-3054.1996.980226.x.

[33] M.D.P. Sánchez-Saavedra, D. Voltolina, The growth rate, biomass production andcomposition of Chaetoceros sp. grown with different light sources, Aquac. Eng. 35(2006) 161–165, http://dx.doi.org/10.1016/j.aquaeng.2005.12.001.

[34] C.-Y. Chen, Y.-C. Chen, H.-C. Huang, C.-C. Huang, W.-L. Lee, J.-S. Chang, Engineeringstrategies for enhancing the production of eicosapentaenoic acid (EPA) from an iso-lated microalga Nannochloropsis oceanica CY2, Bioresour. Technol. 147 (2013)160–167, http://dx.doi.org/10.1016/j.biortech.2013.08.051.

[35] C. Yan, X. Luo, Z. Zheng, Performance of purifying anaerobic fermentation slurry usingmicroalgae in response to various LED light wavelengths and intensities, J. Chem.Technol. Biotechnol. 88 (2013) 1622–1630, http://dx.doi.org/10.1002/jctb.4010.

[36] B. Choi, J.-H. Lim, J. Lee, T. Lee, Optimum conditions for cultivation of Chlorella sp. FC-21 using light emitting diodes, Korean J. Chem. Eng. 30 (2013) 1614–1619, http://dx.doi.org/10.1007/s11814-013-0081-0.

[37] R.M.R. Sastre, Kopplung physiologischer und verfahrenstechnischer Parameter beimWachstum und bei der Produktbildung der Rotalge Porphyridium purpureum, KITScientific Publishing, 2010 http://dx.doi.org/10.5445/KSP/1000015569.

[38] E. Sforza, D. Simionato, G.M. Giacometti, A. Bertucco, T. Morosinotto, Adjusted lightand dark cycles can optimize photosynthetic efficiency in algae growing inphotobioreactors, PLoS One 7 (2012), e38975 http://dx.doi.org/10.1371/journal.pone.0038975.

[39] C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Particles, JohnWiley & Sons, 2008 http://dx.doi.org/10.1002/9783527618156.

[40] R. Kandilian, E. Lee, L. Pilon, Radiation and optical properties of Nannochloropsisoculata grown under different irradiances and spectra, Bioresour. Technol. 137(2013) 63–73, http://dx.doi.org/10.1016/j.biortech.2013.03.058.

[41] T.G. Owens, J.C. Gallagher, R.S. Alberte, Photosynthetic light-harvesting function ofviolaxanthin in Nannochloropsis spp. (Eustigmatophyceae), J. Phycol. 23 (1987)79–85, http://dx.doi.org/10.1111/j.0022-3646.1987.00079.x.

[42] L. Pilon, H. Berberoğlu, R. Kandilian, Radiation transfer in photobiological carbon di-oxide fixation and fuel production by microalgae, J. Quant. Spectrosc. Radiant. 112(2011) 2639–2660, http://dx.doi.org/10.1016/j.jqsrt.2011.07.004.

[43] L.M. Lubián, O. Montero, I. Moreno-Garrido, I.E. Huertas, C. Sobrino, M. González-delValle, G. Parés, Nannochloropsis (Eustigmatophyceae) as source of commerciallyvaluable pigments, J. Appl. Phycol. 12 (2000) 249–255, http://dx.doi.org/10.1023/A:1008170915932.

[44] A.R. Kommareddy, G.A. Anderson, Study of light as a parameter in the growth ofalgae in a photo-bioreactor (PBR), ASAE International Meeting Las Vegas, USA, 2003.

[45] R. Goss, B. Lepetit, Biodiversity of NPQ, J. Plant Physiol. 172 (2015) 13–32, http://dx.doi.org/10.1016/j.jplph.2014.03.004.

[46] T.-H. Kim, Y. Lee, S.-H. Han, S.-J. Hwang, The effects of wavelength and wavelengthmixing ratios on microalgae growth and nitrogen, phosphorus removal usingScenedesmus sp. for wastewater treatment, Bioresour. Technol. 130 (2013) 75–80,http://dx.doi.org/10.1016/j.biortech.2012.11.134.

[47] A. Falciatore, C. Bowler, The evolution and function of blue and red light photorecep-tors, Curr. Top. Dev. Biol. 68 (2005) 317–350, http://dx.doi.org/10.1016/S0070-2153(05)68011-8.

[48] A. Möglich, X. Yang, R.A. Ayers, K. Moffat, Structure and function of plant photore-ceptors, Annu. Rev. Plant Biol. 61 (2010) 21–47, http://dx.doi.org/10.1146/annurev-arplant-042809-112259.

[49] C. Brunet, R. Chandrasekaran, L. Barra, V. Giovagnetti, F. Corato, A.V. Ruban, Spectralradiation dependent photoprotective mechanism in the diatom Pseudo-nitzschiamultistriata, PLoS One 9 (2014), e87015 http://dx.doi.org/10.1371/journal.pone.0087015.

[50] J. Ruiz, P. Álvarez-Díaz, Z. Arbib, C. Garrido-Pérez, J. Barragán, J. Perales, Performanceof a flat panel reactor in the continuous culture of microalgae in urban wastewater:prediction from a batch experiment, Bioresour. Technol. 127 (2013) 456–463,http://dx.doi.org/10.1016/j.biortech.2012.09.103.

[51] A. Kamiya, T. Saitoh, Blue-light-control of the uptake of amino acids and of ammoniain Chlorellamutants, Physiol. Plant. 116 (2002) 248–254, http://dx.doi.org/10.1034/j.1399-3054.2002.1160215.x.

[52] J.G. Day, A.J. Tsavalos, An investigation of the heterotrophic culture of the green algaTetraselmis, J. Appl. Phycol. 8 (1996) 73–77, http://dx.doi.org/10.1007/BF02186225.

[53] M. Azma, M.S. Mohamed, R. Mohamad, R.A. Rahim, A.B. Ariff, Improvement of medi-um composition for heterotrophic cultivation of green microalgae, Tetraselmissuecica, using response surface methodology, Biochem. Eng. J. 53 (2011) 187–195,http://dx.doi.org/10.1016/j.bej.2010.10.010.

[54] C. Hipkin, R. Thomas, P. Syrett, Effects of nitrogen deficiency on nitrate reductase, ni-trate assimilation and photosynthesis in unicellular marine algae, Mar. Biol. 77(1983) 101–105, http://dx.doi.org/10.1007/BF00396306.

398 P.S.C. Schulze et al. / Algal Research 16 (2016) 387–398

[55] N.P. Hüner, R. Bode, K. Dahal, L. Hollis, D. Rosso, M. Krol, A.G. Ivanov, Chloroplastredox imbalance governs phenotypic plasticity: the “grand design of photosynthe-sis” revisited, Front. Plant Sci. 3 (2012) http://dx.doi.org/10.3389/fpls.2012.00255.

[56] H.K. Kwon, S.J. Oh, H.S. Yang, D.M. Kim, I.J. Kang, Y. Oshima, Laboratory study for thephytoremediation of eutrophic coastal sediment using benthic microalgae and lightemitting diode (LED), J. Fac. Agr. Kyushu Univ. 58 (2013) 417–425 (http://catalog.lib.kyushu-u.ac.jp/handle/2324/27374/p417.pdf).

[57] S. Palanichamy, V. Rani, Observations on the long term preservation and culture ofthe marine microalga, Nannochloropsis oculata, J. Mar. Biol. Assoc. India 46 (2004)98–103 (http://eprints.cmfri.org.in/id/eprint/2017).

[58] S.O. Lourenço, E. Barbarino, P.L. Lavín, U.M. Lanfer Marquez, E. Aidar, Distribution ofintracellular nitrogen in marine microalgae: calculation of new nitrogen-to-proteinconversion factors, Eur. J. Phycol. 39 (2004) 17–32, http://dx.doi.org/10.1080/0967026032000157156.

[59] M.R. Brown, C.D. Garland, S.W. Jeffrey, I.D. Jameson, J.M. Leroi, The gross and aminoacid compositions of batch and semi-continuous cultures of Isochrysis sp. (cloneT.ISO), Pavlova lutheri and Nannochloropsis oculata, J. Appl. Phycol. 5 (1993)285–296, http://dx.doi.org/10.1007/BF02186231.

[60] M. Alsull, W.M.W. Omar, Responses of Tetraselmis sp. and Nannochloropsis sp. isolat-ed from Penang National Park coastal waters, Malaysia, to the Combined Influencesof Salinity, Light and Nitrogen Limitation, ICEES 2012, pp. 142–145 (http://psrcentre.org/images/extraimages/23.%20312073.pdf).

[61] M.-Y. Lee, B.-S. Min, C.-S. Chang, E. Jin, Isolation and characterization of a xantho-phyll aberrant mutant of the green alga Nannochloropsis oculata, Mar. Biotechnol.8 (2006) 238–245, http://dx.doi.org/10.1007/s10126-006-5078-9.

[62] M. Hein, M.F. Pedersen, K. Sand-Jensen, Size-dependent nitrogen uptake in micro-and macroalgae, Mar. Ecol. Prog. Ser. 118 (1995) 247–253, http://dx.doi.org/10.3354/meps118247.

[63] P.A. Hodgson, R.J. Henderson, J.R. Sargent, J.W. Leftley, Patterns of variation in the lipidclass and fatty acid composition of Nannochloropsis oculata (Eustigmatophyceae) dur-ing batch culture, J. Appl. Phycol. 3 (1991) 169–181, http://dx.doi.org/10.1007/BF00003699.

[64] M.R. Brown, S.W. Jeffrey, J.K. Volkman, G.A. Dunstan, Nutritional properties ofmicroalgae for mariculture, Aquaculture 151 (1997) 315–331, http://dx.doi.org/10.1016/S0044-8486(96)01501-3.

[65] D. Pal, I. Khozin-Goldberg, Z. Cohen, S. Boussiba, The effect of light, salinity, and ni-trogen availability on lipid production by Nannochloropsis sp, Appl. Microbiol.Biotechnol. 90 (2011) 1429–1441, http://dx.doi.org/10.1007/s00253-011-3170-1.

[66] H. Hu, K. Gao, Response of growth and fatty acid compositions ofNannochloropsis sp.to environmental factors under elevated CO2 concentration, Biotechnol. Lett. 28(2006) 987–992, http://dx.doi.org/10.1007/s10529-006-9026-6.

[67] P. Bondioli, L. Della Bella, G. Rivolta, G. Chini Zittelli, N. Bassi, L. Rodolfi, D. Casini, M.Prussi, D. Chiaramonti, M.R. Tredici, Oil production by the marine microalgaeNannochloropsis sp. F&M-M24 and Tetraselmis suecica F&M-M33, Bioresour. Technol.114 (2012) 567–572, http://dx.doi.org/10.1016/j.biortech.2012.02.123.

[68] J. Fábregas, M. Patiño, E. Vecino, F. Cházaro, A. Otero, Productivity and biochemicalcomposition of cyclostat cultures of the marine microalga Tetraselmis suecica,Appl. Microbiol. Biotechnol. 43 (1995) 617–621, http://dx.doi.org/10.1007/BF00164763.

[69] C. Adams, B. Bugbee, Nitrogen retention and partitioning at the initiation of lipid ac-cumulation in nitrogen-deficient algae, J. Phycol. 50 (2014) 356–365, http://dx.doi.org/10.1007/BF00164763.

[70] D.H. Turpin, Effects of inorganic N availability on algal photosynthesis and carbonmetabolism, J. Phycol. 27 (1991) 14–20, http://dx.doi.org/10.1111/j.0022-3646.1991.00014.x.

[71] J. Marchetti, G. Bougaran, T. Jauffrais, S. Lefebvre, C. Rouxel, B. Saint-Jean, E.Lukomska, R. Robert, J.P. Cadoret, Effects of blue light on the biochemical composi-tion and photosynthetic activity of Isochrysis sp.(T-iso), J. Appl. Phycol. 25 (2013)109–119, http://dx.doi.org/10.1007/s10811-012-9844-y.

[72] A. Kamiya, Light-induced changes in themolecular size of starch in colorless Chlorel-la cells, Plant Cell Physiol. 26 (1985) 759–763 (http://pcp.oxfordjournals.org/con-tent/26/4/759.short).

[73] M. Baba, F. Kikuta, I. Suzuki, M.M. Watanabe, Y. Shiraiwa, Wavelength specificity ofgrowth, photosynthesis, and hydrocarbon production in the oil-producing greenalga Botryococcus braunii, Bioresour. Technol. 109 (2012) 266–270, http://dx.doi.org/10.1016/j.biortech.2011.05.059.

[74] E. Forján, I. Garbayo, M. Henriques, J. Rocha, J.M. Vega, C. Vílchez, UV-A mediatedmodulation of photosynthetic efficiency, xanthophyll cycle and fatty acid produc-tion of Nannochloropsis, Mar. Biotechnol. 13 (2011) 366–375, http://dx.doi.org/10.1007/s10126-010-9306-y.

[75] S. Renaud, D. Parry, L.-V. Thinh, C. Kuo, A. Padovan, N. Sammy, Effect of light inten-sity on the proximate biochemical and fatty acid composition of Isochrysis sp. andNannochloropsis oculata for use in tropical aquaculture, J. Appl. Phycol. 3 (1991)43–53, http://dx.doi.org/10.1007/BF00003918.

[76] J. Fábregas, A. Maseda, A. Domínguez, A. Otero, The cell composition ofNannochloropsis sp. changes under different irradiances in semicontinuous culture,World J. Microbiol. Biotechnol. 20 (2004) 31–35, http://dx.doi.org/10.1023/B:WIBI.0000013288.67536.ed.

[77] C.L. Teo, M. Atta, A. Bukhari, M. Taisir, A.M. Yusuf, A. Idris, Enhancing growth andlipid production of marine microalgae for biodiesel production via the use of differ-ent LED wavelengths, Bioresour. Technol. 162 (2014) 38–44, http://dx.doi.org/10.1016/j.biortech.2014.03.113.

[78] H.M. Guzmán, A. de la Jara Valido, L.C. Duarte, K.F. Presmanes, Estimate by means offlow cytometry of variation in composition of fatty acids from Tetraselmis suecica inresponse to culture conditions, Aquac. Int. 18 (2010) 189–199, http://dx.doi.org/10.1007/s10499-008-9235-1.

[79] G. Dunstan, J. Volkman, S. Barrett, C. Garland, Changes in the lipid composition andmaximisation of the polyunsaturated fatty acid content of three microalgae grownin mass culture, J. Appl. Phycol. 5 (1993) 71–83, http://dx.doi.org/10.1007/BF02182424.