Introduction

Cultivation of microalgae has the potential to provide critical ecosystem services through bioremediation of atmospheric industrial pollution (e.g. CO2 and NO) (Brune et al. 2009; Ho et al. 2011) and nutrient-rich waters from agriculture, aquaculture or urban sewage (Chan et al. 2014). In parallel, the resulting biomass can be used for production of commodities and high-value compounds such as protein or fatty acids (Mata et al. 2010; Pulz and Gross 2004; Stephens et al. 2010).

An important consideration for the feasibility of large-scale microalgae cultivation is water availability and salinity (Borowitzka and Moheimani 2013). Industrial sites, such as coal-fired power stations or sewage plants, may provide low-salinity wastewaters; however, in most cases, groundwater is predominantly available, which, in many parts of Australia, is often saline (<5 ppt) (Hart et al. 1991; Peck and Hatton 2003). This is of particular concern in tropical areas, where high evaporation rates year-round may lead to problems of increased salinity. For example, in a 100,000-L culture pond with a 5 ppt starting salinity and a ~5 % day−1 evaporation rate (North East Queensland, Australia, December) (BOM 2006), salinity would increase to ~7.5 ppt in 10 days and ~10 ppt in 20 days despite daily replacement. Consequently, when screening microalgae for biotechnological applications, it is important to determine species-specific halotolerance to identify species with broader salinity tolerance ranges providing greater flexibility in water requirements and applicability across different cultivation sites (Borowitzka and Moheimani 2013).

Many microalgae have the ability to tolerate fluctuations in salinity (Brown 1982; Kirst 1989; von Alvensleben et al. 2013) through the K+/Na+ pump as well as osmolyte production (e.g. glycine betaine, proline, sucrose, glycerol), which contributes to the osmotic potential for cell turgor and volume control and the latter protecting and restoring damaged proteins, nucleic acids and membrane lipids (Erdmann and Hagemann 2001). Despite this, salinity stress often leads to decreased biomass productivity due to the high energy cost of osmoregulation (Oren 1999) and is often associated with an over-production of reactive oxygen species (ROS) (Erdmann and Hagemann 2001; Mahajan and Tuteja 2005; Sudhir and Murthy 2004). Importantly, this natural response to salinity stress can also be exploited in order to manipulate the biochemical composition of microalgae, as evidenced in, e.g. increased fatty acid content with increasing salinity in the marine microalgae Isochrysis sp. and Nannochloropsis oculata (Renaud and Parry 1994) and in the freshwater microalga Chlamydomonas mexicana (Salama et al. 2014).

Microalgal nutrient uptake (e.g. nitrogen and phosphate) varies widely between species (Aravantinou et al. 2013; Dortch et al. 1984) which has multiple implications for large-scale production depending on whether these nutrients need to be purchased or whether they are freely available in nutrient-rich wastewater, requiring remediation. If nutrient-rich wastewater is available, species selection should identify species with high nutrient consumption and tolerance to eutrophic conditions for timely wastewater remediation (Mata et al. 2010). In contrast, if nutrient-rich wastewater is not available, nutrient provision will incur substantial costs, and species with lower nutrient consumption would be advantageous for biomass production. This is particularly important with the observed global peak of phosphate production and fertilizers, in general, becoming increasingly expensive if effective recycling methods are not adopted (Cordell et al. 2009; Dawson and Hilton 2011). As microalgal growth is positively correlated with nutrient availability (MacIntyre and Cullen 2005), nutrient provision at a cost will particularly affect economics of large-volume bio-products. Similarly to salinity, nutrient condition manipulations are commonly used to favourably alter the biochemical composition of microalgal biomass, for example, to induce the rapid accumulation of triacylglycerols (TAG) (Gao et al. 2013; Olofsson et al. 2014; Rodolfi et al. 2009) or pigments (Imamoglu et al. 2009) in a number of algal species in commercial production. Whilst nutrient limitation leads to cessation of active biomass production, benefits are incurred if it leads to substantially higher accumulation of target compounds (e.g. β-carotene in Dunaliella salina, astaxanthin in Haematococcus pluvialis and lipids in Nannochloropsis spp. (Richardson 2011)) through diverting carbon usage for growth to carbon storage in biomolecules.

Considering the enormous diversity of algal species (Guiry 2012) and the common stress response to up-regulate the content of cellular components that are desirable in commercial production of algae, bioprospecting for new microalgal species amenable to cultivation and environmental tolerance trials connected to biochemical plasticity remain important tasks. Research to date for Desmodesmus armatus, Mesotaenium sp. and Tetraedron sp. is limited, in particular, concerning their potential biotechnological applications (Pulz and Gross 2004). Scenedesmus spp., however, have been extensively investigated with established potential for wastewater remediation and in biotechnological applications, e.g. pigment and biofuel production (Martı́nez et al. 2000; Muller et al. 2005; Guedes et al. 2011; Garcia-Moscoso et al. 2013). Therefore, this study investigated nutrient requirements and responses of four freshwater microalgal species (D. armatus, Mesotaenium sp., Tetraedron sp. and Scenedesmus quadricauda) isolated from ash dam water of a Queensland power station to changes in salinity and flow-on effects on biochemical compositions. The second aim was to identify if nutrient limitation could be used to favourably alter the biochemical profiles and productivity of the same species and if this effect was linked to the level of salinity stress.

Materials and methods

Freshwater microalgae were isolated from Tarong power station (Stanwell Corp.) in Queensland, Australia, and maintained at the North Queensland Algal Identification/Culturing Facility (NQAIF) culture collection (James Cook University, Townsville, Australia). Of the 13 successfully isolated species, Desmodesmus armatus (Trebouxiophyceae) (culture accession # NQAIF301), Mesotaenium sp. (Conjugatophyceae) (NQAIF303), Scenedesmus quadricauda (Chlorophyceae) (NQAIF304) and Tetraedron sp. (Chlorophyceae) (NQAIF295) were chosen for the experiments, not only because they were the dominant and most abundant species but also because of their cosmopolitan distribution which minimizes biosecurity concerns. The fact that they were isolated from ash dam waters from a coal-fired power plant points to their potential resilience in demanding environments and likely suitability for remediation purposes. Individually aerated (0.45 μm filtered air; Durapore; Millipore) batch cultures were maintained at 24 °C, with a 12:12-h photoperiod and light intensity of 42 μmol photons m−2 s−1. Monoclonal cultures with low bacterial numbers (<1 mL−1) were established in a total culture volume of 2 L in modified L1 culture medium (Andersen et al. 2005), with 5 mg L−1 instead of 3 mg L−1phosphate. Preliminary studies identified high phosphate uptake rates by Mesotaenium sp. and S. quadricauda; consequently, medium phosphate concentrations were increased to reduce phosphate replenishment requirements and avoid medium phosphate depletion. To account for differences in cell size, all species were inoculated at 100 mg dry weight L−1. To maintain nutrient-replete biomass, cultures were re-fertilized with nitrate and phosphate when near depletion, until there was sufficient N replete biomass for harvest for biochemical analyses. Modified L1 culture medium was prepared at four salinities: 2, 8, 11 and 18 parts per thousand (ppt) (0.03, 0.14, 0.19 and 0.31 M, respectively) NaCl in filtered seawater (FSW) (pre-filtration Whatman GF/C, followed by 0.45 μm Durapore, Millipore). All materials were sterilized by autoclaving, and cultures were handled and inoculated aseptically in a laminar flow cabinet. Replicate cultures (2 L, n = 3) of all species were inoculated at a dry weight (DW) of ~100 mg L−1 for each salinity. Inoculation was carried out from salinity-specific, 6-month-acclimated mother cultures.

Algal growth

Algal growth was determined by measuring turbidity (OD750 nm), correlated to cells L−1and dry weights (DW; g L−1) following previously described methods (von Alvensleben et al. 2013). Turbidity calibration curves were medium-blanked for each salinity, while DW samples were corrected for salt content using salinity-specific blanks. Results were correlated to generate linear equations (R 2 > 0.95) used to determine cell numbers and respective DWs of each species at each salinity from turbidity measurements. Growth of all species was then determined every second day using turbidity for 20 days, and obtained data were transformed to cell numbers and DWs as described in von Alvensleben et al. (2013). Volumetric biomass productivities (Eq. 1) were determined using a method modified from Su et al. (2011).

$$ \mathrm{Volumetric}\ \mathrm{Biomass}\ \mathrm{Productivity}\ \left[\mathrm{mg}\ {\mathrm{DW}\mathrm{L}}^{\hbox{-} 1}{\mathrm{day}}^{\hbox{-} 1}\right]=\frac{{\mathrm{DW}}_2\hbox{-} {\mathrm{DW}}_1}{{\mathrm{t}}_2\hbox{-} {\mathrm{t}}_1} $$
(1)

where DW 1 and DW 2 = initial and final DW (g L−1), respectively, t 1 and t 2 = initial and final culture time points (days) per identified growth period, respectively.

Nutrient analyses

Medium nitrate (NO3 −), nitrite (NO2 −) and phosphate (PO4 3−) concentrations were determined every second day. For this, 5 mL culture samples were centrifuged at 3000 × g at 20 °C for 20 min, followed by 0.45-μm filtration (Durapore; Millipore Kilsyth) of the supernatant to remove any suspended material. From this, 1.25 mL was used for nitrate (NO3 −), 1.25 mL for nitrite (NO2 −) and 200 μL for phosphate (PO4 3−) determination. The phosphate assay required dilution with deionized water (Elix 5, Millipore) when PO4 3− concentrations exceeded 2.5 mg L−1 to ensure accuracy of the assay.

Nitrate and nitrite were determined spectrophotometrically in triplicate using a method adapted from Carvalho et al. (1998) to suit a 96-well plate format as described in detail in the supplementary material. Nitrite (λ = 540 nm) and nitrate + nitrite (λ = 220 nm) concentrations were calculated by comparison of absorbences with serial dilution linear regression calibration curves (R 2 > 0.95), generated previously from known nitrate and nitrite concentrations (0–32.21 and 0–43.38 μM sodium nitrate and sodium nitrite, respectively). Nitrate concentrations were calculated by subtracting the nitrite concentration from the total nitrate + nitrite concentration.

Similarly, phosphate was determined spectrophotometrically (λ = 610 nm) in triplicate following standard methods adapted from Van Veldhoven and Mannaerts (1987) and R&D Systems, Malachite Green Phosphate Detection Kit (Cat. Number: DY996, www.RnDSystems.com) to suit a 96-well plate format, as described in detail in the supplementary material. Phosphate concentrations were then calculated from serial dilution linear regression calibration curves (R 2 > 0.95), generated previously from known phosphate concentrations (0–50 μM).

Biochemical analyses

Biomass samples for biochemical analyses were harvested from 500 mL samples by centrifugation as described previously (von Alvensleben et al. 2013) when nitrate-replete during late logarithmic growth (day 14 for D. armatus, Mesotaenium sp. and Tetraedron sp. and day 12 for S. quadricauda) and 4 days after total nitrogen (N) depletion during the initial stationary phase (days 22–28 in D. armatus, days 20–24 in Mesotaenium sp., day 24 in S. quadricauda and days 24–30 in Tetraedron sp.). Cultures were classified as nutrient-replete and nutrient-deplete based on increasing and decreasing (NO2 −) secretion patterns, and nutrient depletion was assured by harvesting 4 days after medium nutrient depletion (Malerba et al. 2012). Biomass pellets were freeze-dried and stored in air-tight vials under nitrogen at 4 °C until further analysis.

Total lipids were determined gravimetrically following a direct extraction and transesterification method as described in detail in von Alvensleben et al. (2013).

Fatty acid extraction, transesterification and analysis

Fatty acids were extracted from freeze-dried samples in a single-step extraction and transesterification procedure modified from Gosch et al. (2012) followed by GC-MS analysis. Briefly, 2 mL acetylchloride,/acetylchloride, 95:5 (v/v)) and 300 μL internal standard (nonadecanoic acid, cat# 72332‐1G‐F, Sigma-Aldrich, Australia), 0.2 mg L−1 in methanol) were added to approximately 30 mg (±0.1 mg) dry biomass in Teflon-capped glass vials. Samples were heated at 100 °C for 1 h and allowed to cool, after which 1 mL hexane was added. To ensure complete partitioning of the formed fatty acid methyl esters (FAMEs) into the hexane layer, the samples were heated again at 100 °C for 1 min to form a single methanol/hexane phase. One-millilitre de-ionized water was then added to the cooled sample to facilitate phase separation. The hexane phase containing the FAMEs was collected and filtered through a 0.2 μm PTFE syringe filter prior to injection on the GC column. All solvents were of HPLC grade. Butylated hydroxy-toluene (BHT) (0.01 %) was added as an antioxidant during the extraction.

Fatty acid analysis was carried out in scan mode on an Agilent 7890 GC (DB-23 capillary column with a 0.15 μm cyanopropyl stationary phase, 60 m × 0.25 mm inner diameter) equipped with a flame ionization detector (FID) and connected to an Agilent 5975C electron ionization (EI) turbo mass spectrometer (Agilent Technologies, Australia), for identification of FAMEs as previously described (Gosch et al. 2012; von Alvensleben et al. 2013).

Amino acid analyses

Amino acid profiling of biomass from nutrient-replete and nutrient-deplete cultures at 2 and 11 ppt salinity was carried out by the Instrument Analysis Center of Shanghai Jiao Tong University using a Hitachi L8900 Amino Acid Analyzer (Tokyo, Japan) in accordance with the Chinese national standard protocol for amino acid determination (GB/T 5009.124-2003). Briefly, 10 mg samples were hydrolyzed for 22 h at 110 °C in a vacuum glass tube, using 6 M HCl. Samples were dried in a vacuum, and the residue was diluted with 0.02 M HCl prior to analysis on the amino acid analyzer, together with amino acid standards. Alanine, arginine, aspartic acid, cysteine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tyrosine and valine were quantified. The sum of these amino acids is considered total protein throughout this study.

Carbohydrate determination

Total carbohydrate content was determined by difference, containing also dietary fibre and some organic acids, following the method by Sims (1978) using Eq. 2. For this, ash (dry inorganic) content (mg g−1 DW) was determined by combustion at 500 °C for 24 h in a muffle furnace (Yokogawa-UP 150, AS1044).

$$ \mathrm{Carbohydrate}\ \left(\%\ \mathrm{wt}\right)=100\%\hbox{-} \left(\%\mathrm{Ash}+\%\ \mathrm{Total}\ \mathrm{lipids} + \%\mathrm{Protein}\right) $$
(2)

Total lipid and FA productivities

Total fatty acid (TFA) productivities were determined using Eq. 3, where TFA1 was determined in nutrient-replete conditions and TFA 2 in nutrient-deplete conditions, and t 1 and t 2 represent the replete and deplete harvest time points for FA determination, respectively.

$$ \mathrm{F}\mathrm{A}\ \mathrm{Productivity}\ \left[\mathrm{mg}\ {\mathrm{L}}^{\hbox{-} 1}{\mathrm{day}}^{\hbox{-} 1}\right]=\frac{{\mathrm{TFA}}_2\hbox{-} {\mathrm{TFA}}_1}{{\mathrm{t}}_2\hbox{-} {\mathrm{t}}_1} $$
(3)

Statistical analyses

All statistical analyses were carried out in Statistica 12 (StatSoft Pty Ltd.). Repeated measures ANOVAs were used to determine the effects of salinity on culture growth and biomass productivity and the effects of both salinity and nutrient status on biochemical profiles within and between species. One-way ANOVAs were used to clarify single variable effects generally within species across salinities. Tukey’s post hoc tests were used to determine significant differences assigned at P < 0.05. Homogeneity of variances and normality assumptions were verified using Levene’s tests. Non-metric multidimensional scaling (nMDS, Primer 6) (Clarke 1993) was used to evaluate patterns in fatty acid and amino acid quality (% of total FA or amino acid (AA)) relative to salinity and nutrient status. Increasing distances between points in the ordination plot represent decreasing similarity between those points. The Bray-Curtis similarity coefficient was used as a distance measure, derived from square-root-transformed data of individual fatty acid proportions (mean of replicates, n = 3). Pearson product correlations for the relative proportions of the individual FAs or groups of FAs (>0.8) are shown in the vector loading plot. The length and direction of the vectors indicate the strength of the correlation and direction of change between the two axes, i.e. the relative differences in specific fatty acids between the demarcated groups.

Results

Effect of salinity on growth

Lower salinities (2, 8 and 11 ppt) had no effect on growth of D. armatus, S. quadricauda or Tetraedron sp.; however, 18 ppt significantly reduced or completely inhibited growth in all three species (Fig. 1a, c, d) (repeated measures ANOVA: D. armatus; F (1,3) = 95.5, P < 0.05, S. quadricauda; F (1,3) = 179.5, P < 0.05 Tetraedron sp. F (1,3) = 44.3, P < 0.05). In Mesotaenium sp., cultures at 2 and 8 ppt reached similar final biomass content (~1.1 g L−1), but there was a significant decrease in biomass production with increasing salinity at 11 (0.8 g L−1) and 18 ppt (0.2 g L−1) (repeated measures ANOVA, F (1,3) = 1549.4, P < 0.05) (Fig. 1b).

Fig. 1
figure 1

Mean biomass growth (mg DW L−1) of Desmodesmus armatus (a), Mesotaenium sp. (b), Scenedesmus quadricauda (c) and Tetraedron sp. (d) at 2, 8, 11 and 18 ppt determined using % transmittance at 750 nm. n = 3. Standard error is shown. DW dry weight

Full size image

At 2 and 8 ppt, highest final biomass density was achieved by Mesotaenium sp., >1000 mg L−1, whereas all other species reached between 650 and 950 mg L−1 (Fig. 1). At 11 ppt, biomass density was highest for D. armatus and Mesotaenium sp. (~930 and 850 mg L−1, respectively). With similar growth patterns from 2 to 11 ppt, S. quadricauda and Tetraedron sp. reached the lowest biomass density (730–770 and 710–760 mg L−1, respectively) (Fig. 1). At 18 ppt, D. armatus exhibited the highest growth reaching ~700 mg L−1 (Fig. 1), whereas Mesotaenium sp. and Tetraedron sp. cultures showed minimal growth (260–300 mg L−1) and S. quadricauda growth was completely inhibited, producing insufficient biomass for biochemical composition analysis at this salinity.

With the exception of Mesotaenium sp., salinity had no significant effect on biomass productivity (mg L−1 day−1) from 2 to 11 ppt for each species (repeated measures ANOVA, F (1,2) = 0.4, P = 0.6) (Fig. 2) but was significantly lower at 18 ppt in all species (repeated measures ANOVA, F (1,3) = 90.2, P < 0.05) (Fig. 2). Mesotaenium sp. showed similar productivities at 2 and 8 ppt, but a significant decrease (one-way ANOVA, F = (1,3) = 88.8, P < 0.05) at both 11 and 18 ppt. Between species, Mesotaenium sp. cultured at 2 and 8 ppt had the highest biomass productivity (52–54 mg L−1 day−1) (repeated measures ANOVA, F (1,9) = 9, P < 0.05) (Fig. 2). Productivities were similar (31–43 mg L−1 day−1) in D. armatus and S. quadricauda cultures from 2 to 11 ppt, as well as Mesotaenium sp. at 11 ppt and Tetraedron sp. at 2 ppt. Overall, Tetraedron sp. had the lowest biomass productivity at salinities above 2 ppt (10–31 mg L−1 day−1). At 18 ppt, D. armatus showed the highest productivity (17 mg L−1 day−1) followed by Tetraedron sp. and Mesotaenium sp. with minimal productivity (9–10 mg L−1 day−1) (Fig. 2).

Fig. 2
figure 2

Biomass productivity (mg L−1 day−1) of Desmodesmus armatus, Mesotaenium sp., Scenedesmus quadricauda and Tetraedron sp. n = 3. Standard error is shown. Statistical relations within and between species are shown

Full size image

Nutrient dynamics

Within species, the highest total N-uptake was always at 2 ppt (1.5–1.9 mg L−1 day−1) (Table 1). Total N-uptake was lower but not significantly different in cultures at 8 and 11 ppt in D. armatus, Mesotaenium sp. and S. quadricauda (1.4–1.5 mg L−1 day−1) (repeated measures ANOVA, F (1,2) = 2.7, P = 0.1) (Table 1). Tetraedron sp. showed a significantly lower intermediate uptake at 8 and 11 ppt (one-way ANOVA, F (1,3) = 27.3, P < 0.05) (~1.1 mg L−1 day−1). Between species, the highest total N uptake was observed in S. quadricauda and Tetraedron sp. cultures at 2 ppt (~1.9 mg L−1 day−1), followed by Mesotaenium sp. (1.6 mg L−1 day−1), then D. armatus (1.5 mg L−1 day−1) (Table 1).

Table 1 Total N (nitrate uptake corrected for nitrite secretion) and phosphate uptake rates (mg L−1 day−1) of Desmodesmus armatus, Mesotaenium sp., Scenedesmus quadricauda and Tetraedron sp.
Full size table

Salinity had no significant effect on phosphate uptake in D. armatus (0.59–0.64 mg L−1 day−1) (one-way ANOVA, F (1,3) = 0.02, P = 0.9), S. quadricauda (1.5–2.1 mg L−1 day−1) (one-way ANOVA, F (1,2) = 1.7, P = 0.2) or Tetraedron sp. (0.36–0.57 mg L−1 day−1 ) (one-way ANOVA F (1,3) = 1.8, P = 0.2). In Mesotaenium sp., phosphate uptake was significantly higher at 2 ppt (~2.3 mg L−1 day−1) (one-way ANOVA, F (1,3) = 49.1, P < 0.05), but not significantly different between cultures from 8 to 18 ppt (0.34–0.75 mg L−1 day−1) (main-effects ANOVA, F (1,2) = 3.4, P = 0.06) (Table 1). Between species, phosphate uptake was highest in Mesotaenium sp. at 2 ppt (2.2 mg L−1 day−1) and S. quadricauda (1.5–2.1 mg L−1 day−1) at all salinities (repeated measures ANOVA, F (1,3) = 69.8, P < 0.05), requiring re-fertilization every second day.

Correlating biomass productivity (mg L−1 day−1) and total N and P uptake rates (mg L−1 day−1) (Table 2) showed that, despite differences in uptake rates, these were closely correlated with biomass productivity resulting in similar N-uptake per mg g−1 DW across all salinities within species. The highest total N-uptake per unit biomass was observed in S. quadricauda and Tetraedron sp. at 2 ppt (~0.06 mg N mg−1 DW L−1 day−1) (Table 2). Phosphate uptake per unit biomass within species was generally lower at 2 ppt (except in Mesotaenium sp. and S. quadricauda) and increased at higher salinities. The highest phosphate uptake per unit biomass was observed in S. quadricauda at 2 ppt (~0.06 mg N mg−1 DW L−1 day−1) (Table 2).

Table 2 Nutrient consumption per unit biomass and protein
Full size table

Biochemical composition

Total lipid and TFA content. Total lipid content was 50 % higher than TFA content, which is indicative of a large contribution of other lipid-soluble compounds, particularly pigments, to the total lipid fraction (Lim et al. 2012; von Alvensleben et al. 2013). Salinity had minor effects on total lipid content in nutrient-replete conditions, with Mesotaenium sp. showing the highest total lipid content (20–25 % of DW) and Tetraedron sp. the lowest (14–18 % of DW) (Fig. 3). Nutrient depletion induced an increase of total lipid with increasing salinity in S. quadricauda (repeated measures ANOVA, F (1,2) = 14.3, P = 0.03) and Tetraedron sp. (~5–10 % increase) (repeated measures ANOVA, F (1,3) = 0.161, P = 0.9) whereas nutrient depletion had no effect on total lipid content in D. armatus (repeated measures ANOVA, F (1,3) = 1.1, P = 0.4) and Mesotaenium sp. cultures (F (1,3) = 0.9, P = 0.5) (Fig. 3).

Fig. 3
figure 3

Total lipid and total FA contents (mg L−1) of Desmodesmus armatus (a), Mesotaenium sp. (b), Scenedesmus quadricauda (c) and Tetraedron sp. (d) in nutrient-replete and nutrient-deplete conditions. n = 3. Standard error is shown

Full size image

Salinity did, however, have a significant effect on TFA contents in all species in both nutrient-replete and nutrient-deplete conditions (species- and nutrient-specific one-way ANOVAs, P < 0.05), except for S. quadricauda in replete conditions (one-way ANOVAs, F (1,2) = 2.2, P = 0.2). These significant effects were mainly driven by the small variance between replicates. In nutrient-replete conditions, TFA content decreased with increasing salinity in D. armatus from 2 (9 %) to 18 ppt (7 %) (one-way ANOVA, F (1,3) = 71.6, P < 0.05), Mesotaenium sp. (7–8 %) (one-way ANOVA, F (1,3) = 204.9, P < 0.05) and S. quadricauda (6.7–7.2 %) (one-way ANOVA, F (1,2) = 2.2, P = 0.2) from 2 to 11 ppt. In nutrient-replete Tetraedron sp. cultures, TFA content increased at 8 and 11 ppt (~6.5 %). In all species, nutrient-replete cultures at 18 ppt contained the lowest TFA, except for Mesotaenium sp. which contained its highest TFA content (12–13 % of DW) at this salinity.

Nutrient depletion had a small but significant effect on TFA driving an increase (1–6 % increase) in TFA content in all species (repeated measures ANOVAs, D. armatus: F (1,1) = 289, P < 0.05, Mesotaenium sp.: F (1,1) = 15, P < 0.05, S. quadricauda: F (1,1) = 58.8, P < 0.05, Tetraedron sp.: F (1,3) = 323.8, P < 0.05) at all salinities, except in Mesotaenium sp. at 18 ppt where a decrease (1 %) was observed. Between species, TFA content was highest in D. armatus and Mesotaenium sp. (7–13 % of DW) and lowest in S. quadricauda and Tetraedron sp. (5–7 % of DW) (Fig. 3).

Fatty acid profiles

The most abundant (>5 mg g−1 DW) fatty acids present in all species were C16:0, C18:1, C16:4, C18:2 and C18:3 (α-linolenic acid), except in Mesotaenium sp., which did not contain C16:4 but had high amounts of C16:2 (S. Table 2). Excluding cultures at 18 ppt, Mesotaenium sp. contained the highest C16:0 (15–35 mg g−1 DW) and C18:2 concentrations (10–35 mg g−1 DW) and D. armatus the highest C16:4 (9–15 mg g−1 DW) and C18:3 (α) (20–29 mg g−1 DW) concentrations. MDS analyses (Fig. 4) showed distinct groupings of biomass samples based on their fatty acid profiles. In D. armatus (Fig. 4a, stress 0.01), nutrient-replete cultures from 2 to 11 ppt and nutrient-deplete cultures at 2 ppt grouped together and were characterized by high proportions of omega-3 (ω-3) (52–56 %) and polyunsaturated fatty acids (PUFA) (~70 %, mainly C16:4 and C18:3 α-linolenic) as indicated by the vector loadings to the top right of the ordination plot. Nutrient-replete cultures at 18 ppt had similar profiles to deplete cultures at 8 and 11 ppt and were characterized by higher proportions of omega-6 (ω-6) fatty acids (~18 %) (Pearson product correlation >0.5, data not shown). Actual TFA concentrations (mg g−1 DW) for all species and treatments are listed in S. Tables 1, 2, 3 and 4. Mesotaenium sp. grouped by salinity rather than nutrient status (Fig. 4b, stress 0). Cultures at 2 ppt were characterized by a higher proportion of ω-3 (~30 %, mainly C18:3, α-linolenic) whereas biomass from 8 and 11 ppt had a higher proportion of total PUFA (53–66 %, mainly C18:3) and ω-6 contents (43–53 %, mainly C20:3). The MDS of fatty acid profiles from S. quadricauda yielded groups strongly separated by culture nutrient status (Fig. 4c, stress 0), with a higher proportion of ω-3 (~54 %, mainly C16:4 and C18:3) and PUFA (~67 %) in replete cultures and an increase in proportion of ω-6 (3 % increase, mainly driven by an increase in C18:2) fatty acids in deplete cultures. In Tetraedron sp., fatty acid profiles were affected by both nutrient status and salinity. Replete cultures at 2 ppt showed distinctly different profiles to all other cultures (Fig. 4d, stress 0.01) with the highest proportions of PUFA (57–64 %), lower proportions of saturated fatty acids (SFA) (~20 %) and monounsaturated fatty acids (MUFA) (~12 %) and the highest total ω-3 content (~52 %). Cultures at 8 and 11 ppt were most affected by nutrient depletion as indicated by the large distance between the points in the ordination plot, with an increase in proportion of MUFA contents (15–20 %, mainly C18:1) and decrease in PUFA contents (15–30 %, mainly C16:4 and C18:3). Cultures at 18 ppt were least affected by nutrient status, with nutrient-replete and nutrient-deplete cultures grouping close together with similar FA proportions in both conditions.

Fig. 4
figure 4

Fatty acid quality of Desmodesmus armatus (a), Mesotaenium sp. (b), S. quadricauda (c) and Tetraedron sp. (d) at different salinities and nutrient status. Multidimensional scaling (MDS) ordinations show clusters of samples based on the similarity of their fatty acid profiles (% of total FA). The vector loadings of the specific fatty acids relate to the clusters of samples in the ordination plots (correlations of R > 0.8). The direction and size of each vector indicate the relative abundance of that variable in samples in the corresponding section of the ordination plot

Full size image

Lipid and fatty acid productivities

Due to high biomass productivity (Fig. 2), Mesotaenium sp. at 2 and 8 ppt showed the highest total lipid (11–13 mg L−1 day−1) and TFA productivities (4–4.5 mg L−1 day−1) (S. Table 5). Nutrient-replete fatty acid productivities were highest from 2 to 11 ppt in all species. Fatty acid productivities were mainly driven by biomass productivities, with the exception of high productivities of C18:1 in Mesotaenium sp. at 18 ppt and C18:2 in Tetraedon sp. at 11 ppt which were driven by a high FA content rather than biomass productivity (S. Table 6). Mesotaenium sp. at all salinities had the highest C16:0, C16:2 and C18:2 productivities, and D. armatus had the highest C16:4 productivity (S. Table 6). The highest C18:1 productivity was observed in Tetraedron sp. at 11 and Mesotaenium sp. at 18 ppt., while the highest C18:3 productivity occurred for D. armatus and Mesotaenium sp. at 2 ppt (S. Table 6).

AA contents

Total protein content (as the sum of AA) and essential AA (EAA) content were highest in D. armatus (309–450 mg g−1 DW and 150–233 mg g−1 DW, respectively) and lowest in Mesotaenium sp. (179–260 mg g−1 DW, 92–135 mg g−1 DW, respectively) (Table 3). Except for Mesotaenium sp. and Tetraedron sp. at 2 ppt and S. quadricauda at 11 ppt, nutrient depletion leads to a decrease in amino acids (AA) in all cultures. With the exception of Tetraedron sp., AA contents at 11 ppt were always lower than at 2 ppt in both nutrient-replete and nutrient-deplete conditions. Although AA content decreased with nutrient depletion, proportions of total AA content (%) remained similar, maintaining a similar profile regardless of conditions. In all species, the predominant EAAs were histidine and leucine with the highest concentrations (34–64 mg g−1 DW and 21–39 mg g−1 DW, respectively), whereas methionine and arginine contents were lowest (5–9 mg g−1 DW and 5–10 mg g−1 DW, respectively). More specifically, the highest histidine contents (64 mg g−1 DW) were observed in D. armatus and Tetraedron sp. and the highest leucine contents (39 mg g−1 DW) in D. armatus. Lysine contents were highest in D. armatus and S. quadricauda (27 mg g−1 DW) (for detailed AA profiles, see Supplementary Tables 7, 8, 9 and 10).

Table 3 Total amino acid, essential amino acid and carbohydrate contents (mg g−1 DW) of Desmodesmus armatus, Mesotaenium sp., Scenedesmus quadricauda and Tetraedron sp. at 2 and 11 ppt in nutrient-replete and nutrient-deplete conditions
Full size table

Carbohydrate contents

Carbohydrate content was highest in Tetraedron sp. (393–465 mg g−1 DW) and lowest in S. quadricauda (263–367 mg g−1 DW) (Table 3). At 2 ppt, nutrient depletion induced a carbohydrate content increase (20–50 mg g−1 DW) in all species. At 11 ppt, carbohydrate contents were higher than at 2 ppt in all species when nutrient-replete; however, nutrient depletion only induced a further increase in Tetraedron sp. (~40 mg g−1 DW).

Discussion

This study identified species-specific effects of both salinity and nutrient status on growth and biochemical profiles of D. armatus, Mesotaenium sp., S. quadricauda and Tetraedron sp.

Growth

Despite being freshwater species, this study showed that D. armatus, S. quadricauda and Tetraedron sp. are relatively halotolerant, with similar growth up to 11 ppt salinity, whereas Mesotaenium sp. has a lower salinity tolerance with optimal growth up to 8 ppt. Growth responses to salinity have implications for on-site cultivation of these four microalgal species. Although these species tolerated 18-ppt salinity to varying degrees, large-scale cultivation at this salinity will not be viable. Allowing for salinity increases due to evaporation, only D. armatus can potentially be cultured up to 11 ppt. Scenedesmus quadricauda and Tetraedron sp. should ideally be grown between 2 and 8 ppt and Mesotaenium sp. at 2 to potentially 5 ppt, reducing the evaporation margin, after which growth will be compromised. For example, considering average tropical East Queensland evaporation rates (~5 % daily), these findings imply that D. armatus, S. quadricauda and Tetraedron sp. could be cultured in saline groundwater around 5 ppt with daily water replacements for up to ~24 days (~11 ppt) without adverse effects of salinity. In contrast, Mesotaenium sp. in these conditions would be affected by salinity after ~12 days (~8 ppt), consequently requiring complete water replacement every 12 days, which has serious environmental implications depending on water availability at cultivation sites.

Biomass production of D. armatus, Mesotaenium sp. and Tetraedron sp. in this study is difficult to compare due to limited or absence of reports on growth patterns for these species. In general, biomass production was lower than in previous reports for S. quadricauda and other chlorophytes being examined for biotechnological potential (Patil 1991; Tiftickjian and Rayburn 1986; Zhou et al. 2011; Dickinson et al. 2013). Lower biomass content was likely due to light limitation, as the current study was performed under controlled laboratory conditions with an average light intensity of 40 μmol photons m−2 s−1compared to 100–200 μmol photons m−2 s−1(Tiftickjian and Rayburn 1986; Dickinson et al. 2013) or natural daylight (Patil 1991). This is of little concern, as the aim was to identify suitably halotolerant freshwater species, and culture conditions were not optimized to maximize productivity. Furthermore, large-scale cultivation at high population densities is likely to reduce penetrating light intensities to similar levels.

The observed patterns of decreasing growth with increasing salinity are to be expected for freshwater microalgal species, as increasing culture salinity (mainly Na+ and Cl−) may lead to an over-production of ROS which cause not only oxidative stress, enzyme inactivation and reduction of photosynthetic rates (Sudhir and Murthy 2004; Mahajan and Tuteja 2005), but also cellular ionic imbalance and subsequent water loss (Setter and Greenway 1979; Erdmann and Hagemann 2001). Acclimation to high salinities includes three processes: (1) restoration and maintenance of cell turgor and volume, (2) changes in permeability of the cell membrane and regulated uptake (K+) and expulsion (Na+) of ions and (3) the accumulation of osmoprotectant compatible solutes and stress proteins (Brown 1976; Erdmann and Hagemann 2001). Approximately 20 different compatible solutes have been shown to occur in microalgae with variable degrees of osmoprotection and salinity compensation. This has implications for growth, which will decrease, if ATP utilization is predominantly for osmotic regulation and/or nitrogen taken up is utilized for N-based osmoregulatory solute synthesis (e.g. proline, glycine betaine) (Vanlerberghe and Brown 1987; Erdmann and Hagemann 2001).

Nutrient dynamics, specifically nitrogen fluxes, can provide insight into possible osmoregulatory mechanisms, particularly when considered in combination with changes in biochemical profiles. In this study, the similarities in N uptake per unit biomass between salinities and the decrease of AA contents, specifically proline and glycine (Suppl. Tables 7, 8, 9 and 10) with increasing salinity in all species, indicate that these species are unlikely to produce N-containing osmolytes reported in microalgae (glycine betaine and/or proline). Although not specifically studied here, we hypothesize that osmoregulation in these species could be achieved via accumulation of carbohydrates which has been reported for other chlorophyte species, e.g. Chlamydomonas sp., Chlorella emersonii, Dunaliella sp. and Stichococcus bacillaris (Ben-Amotz and Avron 1983; Erdmann and Hagemann 2001). This is supported by increased carbohydrate contents with increasing salinity under nutrient-replete conditions.

For remediation/nutrient provision purposes, S. quadricauda and Tetraedron sp. at 2 ppt had the highest total N uptake, and Mesotaenium sp. and S. quadricauda had the highest phosphate uptake. This has dual implications depending on cultivation site, where, if nutrients have to be added at a cost, strain selection should be towards low-nutrient-consumption species (e.g. D. armatus), and, if nutrient-rich wastewater is available, high-nutrient-uptake species (e.g. Mesotaenium sp. and S. quadricauda) should be selected. The faster phosphate uptake in Mesotaenium sp. and S. quadricauda compared to the other species could be due to higher biomass production, specifically in Mesotaenium sp. or an indication of storage capacity by these species. Microalgae can store phosphate as polyphosphate for later use when external phosphate becomes limiting (Powell et al. 2009). Excessive uptake and polyphosphate storage are either a consequence of nutrient starvation followed by re-exposure (Aitchison and Butt 1973), or ‘luxury uptake’, which does not require prior nutrient starvation (Eixler et al. 2006). In this instance, luxury uptake is most likely, as cultures were maintained nutrient-replete until intentional depletion.

Mesotaenium sp. at low salinities (2 to 8 ppt) had the highest biomass, lipid and FAME productivities using the least nutrients, making it the most suitable species for cultivation when fertilization incurs a cost. For wastewater remediation of nitrate and phosphate, S. quadricauda is the most suitable species showing the highest removal rates of these nutrients. These findings correlate with previous reports that have identified S. quadricauda as an effective species for nutrient-rich wastewater remediation (Martı́nez et al. 2000; Shi et al. 2007; Dickinson et al. 2013).

Biochemical profiles

Total lipid contents were at the lower end of the range reported for other green algae but correspond to previous findings for a number of chlorophyte species (Griffiths and Harrison 2009) (see comparisons in Table 4).

Table 4 Biomass productivity and biochemical content comparison between species isolated from the ash dam of a Queensland coal-fired power station (this study) and published data
Full size table

An increase in total lipid and FAs in microalgae following nutrient depletion has been shown in previous studies, e.g. Chlorella vulgaris (Converti et al. 2009) and Scenedesmus subspicatus (Dean et al. 2010), and is often used for large-scale microalgal culture manipulations (Sharma et al. 2012). However, this study shows that the degree of this effect varies between species and salinity conditions. In D. armatus and Mesotaenium sp., neither salinity nor nutrient depletion had any effects on total lipid content, whereas in S. quadricauda and Tetraedron sp., nutrient depletion only had significant effects at higher salinities from 8 and 11 ppt, respectively. This shows that nutrient depletion and/or salinity stress is not universally effective lipid induction methods, supporting previous studies using Chlorella sorokiniana (Griffiths and Harrison 2009), Chlorella sp., Scenedesmus sp., (Rodolfi et al. 2009), Tetraselmis sp. and Nannochloropsis sp. (Reitan et al. 1994).

N limitation under continued photosynthetic carbon acquisition leads to diversion of carbon from growth to storage (Becker 1994; Rodolfi et al. 2009), as production of N-containing compounds such as proteins, nucleic acids and chlorophylls is inhibited, therefore resulting in reduced growth and biomass productivity. Although nutrient depletion for 4 days increased TFA content and was statistically significant in all species, differences were small mainly driven by the small variance between replicate fatty acid samples. As such, larger-nutrient-depletion periods would need to be applied which cannot be recommended as a means to favourably alter the biochemical profiles and productivity of these species due to impacts on cultivation footprints required to sustain biomass yields.

While FA comparisons between studies show a degree of variability, generally due to differences in culture conditions, FA profiles here were similar to those of other chlorophytes, with C16:0, C16:2, C18:1, C18:2 and C18:3 being the predominant fatty acids (Dunstan et al. 1992). More specifically, D. armatus, Mesotaenium sp. and Tetraedron sp. had comparable FA profiles to other reports for these species (Lang et al. 2011) and S. quadricauda FA profiles are comparable to those described for the same species by Ahlgren and Hyenstrand (2003) (compare with Suppl. Tables 1, 2, 3 and 4).

Fatty acid profiles were affected differently between the four species: Mesotaenium sp. was affected by culture salinity, whereas in D. armatus and S. quadricauda, nutrient availability had the greatest influence and Tetraedron sp. was affected by a combination of both. In all species, nutrient depletion induced an increase of SFA and MUFA driven mainly by an increase of C16:0 and C18:1, respectively and a decrease of C18:3 (except Tetraedron sp. at 2 ppt). Similar changes have been reported for a number of green algae, e.g. Botryococcus braunii and the eustigmatophyte Nannochloropsis sp. (Reitan et al. 1994; Zhila et al. 2005; Rodolfi et al. 2009; Su et al. 2011), and are most likely due to the accumulation of neutral lipids such as triacylglycerols, which in the Chlorophyceae, which have been observed to contain mainly C16:0 and C18:1 FA (Becker 1994).

Previous reports for the effects of increasing salinities on microalgal FA profiles have shown similar patterns to nutrient depletion with increases of C18:1 contents in B. braunii (Rao et al. 2007; Zhila et al. 2011), Isochrysis sp., D. salina (Ben-Amotz et al 1985) and Desmodesmus abundans (Xia et al. 2014) and increases of C16:0 in B. braunii (Rao et al. 2007; Zhila et al. 2011), Nannochloropsis oculata and Nitzschia frustulum (Renaud and Parry 1994). However, reports on the effects of sodium chloride on microalgal fatty acids are scarce and often contradictory (Zhila et al. 2011). Furthermore, it is also unclear if fatty acid composition plays a role in microalgal osmoregulation (Renaud and Parry 1994). A primary role of fatty acids in algae is related to functions of cell membranes and metabolic processes (Guschina and Harwood 2006). The degree of membrane fatty acid unsaturation is also a significant parameter in algal adaptation to environmental conditions. Fatty acid changes in response to high salinities are required to maintain membrane fluidity and prevent destruction (Zhila et al. 2011).

The distinct effect of salinity on Mesotaenium sp FA profiles are difficult to explain and put into the context of current literature. The high concentrations of C16:0 and C18:1 at 18 ppt are potentially due to the inhibited growth at this salinity having similar effects to nutrient limitation (see above) which results in the accumulation of TAG containing C16:0 and C18:1 FA (Ben-Amotz et al. 1985).

Microalgal PUFAs have a promising biotechnological market for food, feed and material applications (Lligadas et al. 2010; Pulz and Gross 2004). Examples include the following oleic acid (C18:1, ω-9), linoleic (C18:2, ω-6), α-linolenic (C18:3, ω-3) and stearidonic acid (C18:4, ω-3). Oleic acid (C18:1) can be used to produce fatty acid-derived diols and polyols, from which polyurethanes can be synthesized through polyaddition reactions with organic isocyanates (Lligadas et al. 2010). In this study, C18:1 content was significantly increased in all species by a combination of nutrient depletion and high salinities (11–18 ppt). The highest C18:1 content was observed for Tetraedron sp. at 11 ppt (54 mg g−1 DW) and Mesotaenium sp. at 18 ppt (35 mg g−1 DW). Both could be potential candidates for bioplastic manufacturing; however, Mesotaenium sp. would require a two-step cultivation process (Su et al. 2011), with biomass production at 2 ppt followed by salinity stress (18 ppt).

For dietary applications, linoleic acid (C18:2, ω-6) and α-linolenic acid (C18:3, ω-3) are essential nutrients for immune system function and tissue regeneration processes (de Jesus Raposo et al. 2013). They are also important precursors for other ω-6 and ω-3 FAs (Guil-Guerrero 2007), with distinct cellular functions (Simopoulos 2002). An imbalance in ω-6 and ω-3 FA ratios in current ‘western diets’ has been linked to a range of diseases such as cardiovascular disorders, diabetes, obesity, inflammatory processes, increased susceptibility to viral infections, certain types of cancer, autoimmune disorders, rheumatoid arthritis, asthma and depression (Simopoulos 2002; Guil-Guerrero 2007). Consequently, a ~1:1 ω-6/ω-3 uptake ratio has been recommended to ensure good health and normal development. This is an important consideration when identifying novel feed and FA sources. Desmodesmus armatus, S. quadricauda and Tetraedron sp. have low ω-6/ω-3 (generally <0.4:1) ratios and could therefore be beneficial as ω-3 nutritional supplements. Mesotaenium sp., on the other hand, had a particularly high ω-6/ω-3 ratio at salinities above 8 ppt (3–10:1) driven by a high C18:2 content, making it a possible candidate for pharmacological applications in the topical treatment of skin hyperplasias (Proksch et al. 1993).

Stearidonic acid (SA, C18:4) has also been shown to possess a number of health benefits and bioactive properties to prevent a range of conditions including certain cancers, arthritis and thrombosis (Guil-Guerrero 2007). Microalgae have previously been suggested as a potential source of SA (Guil-Guerrero 2007). In this study, C18:4 was present in D. armatus and Tetraedron sp.; however, only D. armatus at lower salinities contained notable amounts (7 mg g−1 DW) of this FA. Depending on the viability of targeting this FA for health purposes, D. armatus is therefore a suitable candidate for further research to improve C18:4 productivity yields.

Lipid and fatty acid productivities were generally low in this study compared to the same species in other studies (Rodolfi et al. 2009; Zhou et al. 2011), which is likely due to the low growth rates, as actual total lipid and fatty acid contents were comparable to previous studies (Dunstan et al. 1992; Ahlgren and Hyenstrand 2003; Rodolfi et al. 2009; Zhou et al. 2011). Consequently, future research should focus on increasing biomass productivity.

AA profiles in this study were similar to previous reports for Chlorella sp. and Scenedesmus sp. (Brown and Jeffrey 1992; Ahlgren and Hyenstrand 2003), except for histidine concentrations which were considerably higher in species in this study (up to 6.5 % of DW). The decrease of AA concentrations observed in all species in this study following nutrient depletion has been extensively documented and is most likely due to the diversion from protein production to carbohydrate or lipid production in the absence of N for protein synthesis (Flynn 1990; Rodolfi et al. 2009; Mata et al. 2010). As mentioned earlier and as documented in other microalgal species (Brown and Hellebust 1978; Greenway and Setter 1979; Vanlerberghe and Brown 1987), the species in this study do not use AA-based osmoregulation to combat salinity stress, as indicated by salinity-induced decreases in AA content, particularly proline and glycine.

Feed protein quality is determined by AA digestion and absorption by animals and their respective AA requirements for metabolic processes. In general, the ideal protein source for an organism contains the same AA content and AA proportions as the organism itself (Brown and Jeffrey 1992; De Silva et al. 2012). Limitation of one or more specific AAs restricts growth and results in the inability to utilize other essential AAs (De Silva et al. 2012) which becomes problematic when formulating feeds for farmed animals as certain EAAs are often limiting, e.g. methionine, lysine and threonine in fish, shrimp, cattle, swine and poultry feeds (D’Mello 1993; Kung Jr and Rode 1996; Rawles et al. 2013; Nunes et al. 2014). Currently, optimizing animal feed protein quality is not only carried out by supplementing feed with synthetic AAs but can also be achieved by AA blending from other sources with high concentrations of target AAs, e.g. plants, algae and insects (Boland et al. 2013), or as a by-product of biotechnological processes such as biofuel production (Williams and Laurens 2010). This study has shown that D. armatus had the highest AA contents and would be the most suitable species for AA production, in particular, lysine or as a feed supplement for species where lysine is often limiting, e.g. giant clam (Tridacna gigas) aquaculture (Brown 1991).

The carbohydrate content increase following N-depletion is due to the diversion of carbon from protein synthesis to carbohydrate and lipid production (see above). This increase is consistent with previous studies showing a carbohydrate increase following N depletion in Scenedesmus obliquus (Ho et al. 2012). Although this study did not specifically focus on carbohydrate production and composition in the four study species, microalgae are a potential source of sugars such as xylose, arabinose, mannose, galactose, glucose and, the less common, sugars rhamnose, fucose and uronic acids (Cheng et al. 2011; Ho et al. 2012; Krienitz et al. 1999), with an interesting potential for commercialization (Draaisma et al. 2013).

Conclusions

While all species cultured at salinities of 2–18 ppt, Mesotaenium sp. was the least salinity tolerant and D. armatus was the most halotolerant species of the dominant microalgae isolated from ash dam water of a Queensland coal-fired power station. Nitrogen uptake rates correlated with biomass irrespective of salinity, which together with decreased levels of proline and glycine at higher salinities suggest that salinity tolerance in these species is not achieved by glycine betaine or proline accumulation, as described for some other chlorophytes. Increased carbohydrate contents suggest instead that carbohydrate-based osmoregulatory mechanisms could be involved in salinity acclimation. The total lipid content data of the examined species suggest that neither increased salinity nor nitrogen depletion should be viewed as a universal mechanism to increase total lipids or fatty acids, as D. armatus and Mesotaenium sp. did not respond significantly to either treatment and S. quadricauda and Tetraedron sp. were only significantly affected by higher salinities. This is further corroborated by the finding that the FA profile was predominantly influenced by salinity in Mesotaenium sp., by nutrient status in D. armatus and S. quadricauda and by a combination of the two in Tetraedron sp. In general though, the isolated species responded to nutrient limitation with an increase in SFA and MUFA, particularly C16:0 and C18:1, which is well known from the literature. Generally, D. armatus, S. quadricauda and Tetraedron sp. were characterized by low ω-6/ω-3 ratios making them potential candidates for ω-3 supplements. In contrast, Mesotaenium sp. was characterized by a ω-6/ω-3 ratio of 2–10:1, making it unsuitable for diet supplementation with ω-3 FAs, yet it could be a pharmacological candidate for the topical treatment of skin hyperplasias.

The overall species responses from this study can now be used to produce a species selection matrix to target species for scaled production based on their salinity tolerance and plasticity in biochemical composition (Table 5). This study confirmed that S. quadricauda is an ideal candidate for environmental services, such as nitrogen and phosphate remediation, as it had the highest uptake rates. This study further identifies that the organism would be suitable across a salinity range of 2 < 11 ppt. D. armatus and Mesotaenium sp., on the other hand, stood out for biomass production under nutrient-poor conditions from 2 < 18 and 2 < 8 ppt, respectively. Such situations are typically encountered when producing carbon dioxide-supplemented biomass at coal-fired power stations in Australia where large amounts of nutrients and/or nutrient-rich water sources are generally unavailable. Both algae also had the highest FA content with a profile suitable for lipid-based biofuel production for on-site consumption. Tetraedron sp. and, Mesotaenium sp., in particular (5.4 % of DW, 45 % of TFA), excelled in accumulation of C18:1 at 8 < 11 and >11 < 18 ppt, respectively, a valuable precursor for bio-degradable plastic production. It needs to be recognized though that such production would require a two-step approach, where biomass accumulation would require cultivation at 2 ppt with subsequent salt stress used to shift the FA profile in favour of C18:1 accumulation, the feasibility of which still requires demonstration. D. armatus also has demonstrated pharmaceutical potential through accumulation of stearic acid when cultured at 2 ppt. While biomass yields and productivities are yet to be demonstrated on site, with regard to freshwater requirements, the results of this study suggest that D. armatus, S. quadricauda and Tetraedron sp. are sufficiently salinity tolerant to only require freshwater make-up water after 24-day cultivation based on East Queensland daily evaporation rates and salinity concentrations of available water sources, while Mesotaenium sp. could only be cultivated for 12 days under the same conditions. These results have major implications for cultivation site and product range selection for these new isolates.

Table 5 Decision matrix for species selection used in this study isolated from ash dam water of a Queensland power station for remediation, low-nutrient-based cultivation, high fatty acid contents and bioplastic and nutritional potential based on salinity tolerance
Full size table