DOI: https://doi.org/10.21203/rs.3.rs-2458173/v1
Ocean-related global changes have altered phytoplankton community structure, especially the diatom-dinoflagellate competition, which further influences ecosystem structure and functions. The pivotal ecological roles of diatoms and dinoflagellates are strongly related with their biochemical compositions, while quantitative comparisons of biochemical changes between diatoms and dinoflagellates under variable environments are still limited. We investigated responses of lipid biomarkers (sterols and fatty acids (FAs)) to different temperatures (12, 18 and 24℃), nitrogen and phosphorus concentrations and their molar ratios (N:P ratio) of 10:1, 24:1 and 63:1 in marine diatom Phaeodactylum tricornutum and dinoflagellate Prorocentrum minimum. Over these wide ranges of temperature and nutrient conditions, sterol and FA profiles were relatively stable in the two species. For C-normalized contents of major sterols and FAs, warming caused non-significant changes in the diatom but an increase (up to 153%) in the dinoflagellate; eutrophication caused an overall decrease (up to 53%) in the diatom but an overall increase (up to 77%) in the dinoflagellate; in contrast, imbalanced N:P ratios caused an overall increase (up to 64%) in the diatom but an overall decrease (up to 53%) in the dinoflagellate. Under future ocean warming, eutrophication and imbalanced N:P ratios, major sterol and polyunsaturated FA contents would increase (ca. 9% ~ 48%) in the dinoflagellate, while those in the diatom would change non-significantly. This study expands our knowledge on lipid-based indicators of phytoplankton under changing environments, which by systematically linking with several other aspects of food quality will help to understand the ecological role of diatom-dinoflagellate community changes.
Diatoms and dinoflagellates begun their evolutionary trajectories to ecological prominence about 200 million years ago and still occupy ecological dominance in the contemporary ocean (Falkowski and Oliver 2007; Jeong et al. 2021). Anthropogenic and natural climatic forcing has significantly altered diatom-dinoflagellate community composition, shifting typically from diatom dominance to the co-occurrence of diatoms and dinoflagellates or dinoflagellate dominance, especially in coastal oceans (Chen et al. 2022; Klais et al. 2011). This shift can greatly affect ecosystem structure and functioning, and biogeochemical cycles of elements, e.g., reducing the associated bacterial production and the down-ward transport of organic matter (Heiskanen 1998; Smetacek 1985). The ecological importance of diatoms and dinoflagellates are strongly related with their biochemical compositions. In particular, sterols and fatty acids (FAs) are useful biomarkers in studies of food webs (Ruess and Müller-Navarra 2019) and phytoplankton productivity and community structure in the modern ocean and in the past through sedimentary records (Schubert et al. 1998; Wang et al. 2022).
Sterols are tetracyclic triterpenoids indispensable for many cellular and developmental processes in all eukaryotes such as modulating membrane fluidity, permeability, and serving as precursors of steroid hormones in animals (Martin-Creuzburg and von Elert 2009; Volkman 2003). Brassicasterol/epi-brassicasterol and dinosterol are often used as diatom and dinoflagellate biomarkers, respectively, e.g., in the Chesapeake Bay (Zimmerman and Canuel 2002), the Sea of Okhotsk (Seki et al. 2004), the East China Sea (Wu et al. 2016; Xing et al. 2016), and the Kuroshio-Oyashio region of the northwest Pacific Ocean (Wang et al. 2022). In marine diatoms and dinoflagellates, the production of these two sterols varies strongly with culture conditions. For example, per-cell brassicasterol/epi-brassicasterol contents in the diatom Phaeodactylum tricornutum significantly increased (mean percent changes: up to 37%) with increasing temperature and N:P ratios (Bi et al. 2020). Dinosterol contents in three species of marine dinoflagellates varied strongly between different growth phases (up to 168% changes) as well as under high temperature and nutrient deficient conditions (up to 107% changes) (Chen et al. 2019). Although efforts have been made to compare sterol production changes between diatoms and dinoflagellates (Ding et al. 2019), further studies are needed to disentangle the effects of environmental versus physiological changes on sterols in these two phytoplankton groups.
FAs are basic lipid constituents in algae with diverse functions including cellular membrane structure and energy storage (Guschina and Harwood 2009). FAs are taxonomic indicators at the class level of phytoplankton, with diatoms containing relatively high proportions of 16:1 and EPA (20:5n-3; eicosapentaenoic acid), and dinoflagellates being rich in 18:0, 18:1, and DHA (22:6n-3; docosahexaenoic acid) (Jónasdóttir 2019; Kelly and Scheibling 2012). Thus, the ratios of specific FAs have been used to reveal phytoplankton community structure in many areas such as the West Greenland coast (Reuss and Poulsen 2002), and the Guadalquivir River Estuary, the southwest Spain (Cañavate et al. 2019) and the Barents Sea (Kohlbach et al. 2021). Phytoplankton FA composition is strongly affected by environmental changes. Data synthesis on phytoplankton FA profiles showed that warming significantly reduced EPA proportions in diatoms and DHA proportions in dinoflagellates, while the responses of major FA groups (polyunsaturated fatty acids (PUFAs) and saturated fatty acids (SFAs)) differed between diatoms and dinoflagellates (Hixson and Arts 2016). C-normalized PUFA contents was predicted to decrease (~ 10%) in the diatom P. tricornutum under warming, nutrient deficiency and enhanced pCO2 in the future, in oceans (Bi et al. 2020). However, more research is needed to quantitatively compare FA changes between diatoms and dinoflagellates in changing environments.
Here, we test how temperature, N (P) concentrations and N:P supply ratios affect sterols and FAs in marine model species P. tricornutum and Prorocentrum minimum. These two algal species have been used in genomes and genome-enabled studies (Bowler et al. 2008; Smith et al. 2019) and marine food web dynamic research (Arndt and Sommer 2014; Hassett 2004; Jónasdóttir 1994). Brassicasterol/epi-brassicasterol, dinosterol and PUFAs were considered as the most important lipid biomarkers due to their wide applications in ecology and biogeosciences. We focus on the changes of lipid biomarker contents in the two species, while previous work mainly reported the responses of population growth to temperature and nutrient changes. Using semi-continuous culture systems, we disentangled the effects of environmental variations versus growth rate changes. Our aims are to (i) identify the key environmental factors driving the changes in sterols and FAs, and (ii) quantify and compare the effects of environmental changes on sterols and FAs in the diatom and dinoflagellate.
Monospecific cultures were performed with two algal species, i.e., P. tricornutum (Bacillariophyceae; strain MACC/B254) and P. minimum (Dinophyceae; strain HYESL63). Cultures of the two species were kept in thermo constant cabinets of 12, 18 and 24℃ at a light:dark cycle of 12:12 h and a light intensity of 100 µmol m− 2 s− 1. All cultures were grown in nutrient-enriched and sterile-filtered (0.2 µm pore size, Sartobran® P 300) natural seawater prepared according to Provasoli (1963) and Ismar et al. (2008). Nitrate and phosphate were added to natural seawater to achieve different concentrations (low, normal and high) and corresponding N:P molar ratios of 10:1, 24:1 and 63:1, with a salinity of 31 psu (Fig. S1; Table S1). Silicate was also added in diatom cultures with a concentration of 880 µmol L− 1. Each treatment was replicated three times and each culture flask was kept with 200 mL culture volume. To limit sedimentation in the experiments, all flasks were regularly agitated every day. The two algal species were acclimated to the corresponding culture conditions prior to the experiments.
Each species was first grown in batch cultures under different treatments. Algal growth was monitored daily, and cell density during the exponential growth phase was used to calculate the observed maximum growth rate (µmax, d− 1) (Bi et al. 2012). When the cultures reached the early stationary phase, certain volume of the incubation water was replaced with fresh, enriched media to attain semi-continuous cultures with the gross growth rate (µ, d− 1) of 20% of µmax. The exchanged volume was determined by multiplying the daily renewal rate (D, d− 1; D = 1 – e −µ × t, where t = 1d) by 200 mL (the culture volume). The steady state was attained when µ = D and the net growth rate (r, d− 1; r = µ - D) = 0 d− 1.
Algal cells were counted daily with an improved Neubauer hemacytometer (Glaswarenfabrik Karl Hecht GmbH) under a microscope (Olympus CX41). To analyze particulate organic carbon (POC), FAs and sterols, algal cells at steady-state conditions were harvested on pre-combusted GF/F filters (Whatman) after filtering 15–30 mL of cultures depending on cell density in the culture flask and the parameters to be determined. Samples were kept at − 80℃ after filtration.
POC was determined by an elemental analyzer (Thermo Flash 2000) (Sharp 1974). Sterols and FAs were extracted from freeze-dried filter samples using the mixture of dichloromethane and methanol with a volume ratio of 3:1 according to the method in Zhao et al. (2006). In brief, C19 n-alkanol and nonadecanoic acid were added as internal standards. The polar fraction containing sterols was eluted using dichloromethane/methanol with a volume ratio of 95:5 and silylated using N, O-bis(trimethylsilyl)-trifluoroacetamide at 70℃ for 1 h. The acid fraction that contained FAs was derivatized with MeOH/HCl (95:5, volume ratio) at 70℃ for 12 h. Sterols and fatty acid methyl esters (FAMEs) were analyzed in a gas chromatograph (Agilent Technologies 8890A) equipped with a flame ionization detector, and a HP-1 column (50 m, 0.32 mm i.d., 0.17 µm film; Agilent J&W) and a SP-2560 column (100 m, 0.25 mm i.d., 0.20 µm film; Supelco) for sterol and FAME analysis, respectively. The identification of sterols was performed with reference to a lab standard of dinosterol and commercially available standards of several sterols (brassicasterol/epi-brassicasterol, Merck; cholesterol and campesterol, Sigma). FAs were identified with reference to the standard Supelco 37 component FAME mixture. C-normalized (µg mgC− 1) and per-cell (pg cell− 1) contents of sterols and FAs were presented, and FA proportions (% of total fatty acids (TFAs)) were also reported in our study.
To investigate the influence of different environmental factors on algal growth and the contents of sterols and FAs, generalized linear mixed models (GLMMs) were conducted by setting temperature, nitrate concentrations and N:P ratios as fixed effects, and µmax, cell density, per-cell POC contents, per-cell and C-normalized contents of sterols and FAs as dependent variables. The Akaike Information Criterion corrected (AICc) was used to select the best candidate among models that contained first-order effects, second-order interactions, and third-order interactions of the three environmental factors. The model with the minimum AICc value was chosen as the best candidate. The changes in AICc values ≥ 10 units were considered as a reasonable improvement in the models, and the simpler model was selected with comparable AICc values (< 10 units), unless second- or third-order interactions were significant. In our study, the changes in AICc values between different models were less than 10 units for all dependent variables, and models that contained only first-order effects of the three environmental factors were selected (Table S2).
Principal component analysis (PCA) was used to explore the differences in FA compositions (as % of TFAs and on a per carbon basis) between the two species across all treatments. The same analysis was conducted for both FA and sterol compositions (on a per carbon basis), where the variables included FA individuals accounting over 0.5% of TFAs in both species (Table S3), brassicasterol/epi-brassicasterol in the diatom, and brassicasterol/epi-brassicasterol and dinosterol in the dinoflagellate.
All statistical analyses were performed with a significance level of p = 0.05. GLMMs were conducted in SPSS 25 (IBM Corporation). PCA was carried out in R, version 3.5.1 (R Development Core Team 2010) using the packages factoextra (Kassambara and Mundt 2017) and FactoMineR (Le et al. 2008).
3.1 Population growth
The observed maximal growth rate (μmax) changed significantly with temperature (GLMMs, p ≤ 0.020, bold letters in Table S4), but not to that in nutrient conditions in most cases in the two species. The rising temperature led to a significant increase in μmax in the two species (Fig. S2a, b). Cell density at steady state changed significantly with nutrient conditions in the two species (p ≤ 0.028), while temperature showed significant effects only in the dinoflagellate (p < 0.001). Cell density at steady state showed a hump-shaped response to increasing N:P ratios in most cases, and a clear increase with increasing nutrient concentrations in the two species and with increasing temperature in the dinoflagellate (Fig. S2c-f).
3.2 Sterols
There was one major sterol (brassicasterol/epi-brassicasterol) detected in the diatom P. tricornutum in our study (Table S3). In the dinoflagellate P. minimum, four sterols were identified, i.e., dinosterol (21% ~ 63%, mean percentage of the total sterols), cholesterol (22% ~ 40%), brassicasterol/epi-brassicasterol (9% ~ 34%) and campesterol/5-ergostenol (4% ~ 18%). Due to their high abundance and wide applications, brassicasterol/epi-brassicasterol and dinosterol were considered as the most important sterols, and their responses to culture condition changes will be discussed in detail in the following sections.
GLMMs results showed that C-normalized brassicasterol/epi-brassicasterol contents in the diatom changed non-significantly in response to the three environmental factors, while brassicasterol/epi-brassicasterol in the dinoflagellate responded significantly to temperature and dinosterol responded significantly to all the three factors (p ≤ 0.011, bold letters in Table 1, Figs. 1-2). In the dinoflagellate, C-normalized brassicasterol/epi-brassicasterol contents were generally higher at warmer conditions (Fig. 2a-i), and dinosterol also increased with increasing temperature and nutrient concentrations (Fig. 2j-r) and showed a hump-shaped response to increasing N:P ratios at most cases (Fig. S3).
3.3 Fatty acids
The two species had distinct FA profiles across all treatments (Fig. 3, Figs. S4-S6, Table S3). The diatom P. tricornutum contained clearly high MUFAs (mean percentage of TFAs: 22% ~ 51%) than the dinoflagellate P. minimum (10% ~ 36%), while the proportions of other FA groups were comparable between the two species. The results of PCA showed that the two species were clearly separated along the first two dimensions for both FA proportions and C-normalized contents (Figs. S5-S6, Tables S5-S6). In the PCA plot for FA proportions, 16:1n-7, 18:2n-6c, EPA and 24:0 (indicative of the diatom), and in the opposite direction 16:0, 18:0, 20:2n-6 and DHA (indicative of the dinoflagellate) explained most of the PC1 pattern (Fig. S5).
C-normalized contents of most FA groups changed significantly with nutrient condition changes in the two species, while those in the dinoflagellate also showed significant responses to temperature (p ≤ 0.027, Table 1, Fig. 3). In the diatom, SFAs, MUFAs and TFAs generally had lower contents at higher nutrient concentrations and at N:P = 24:1 (Fig. 3a), while PUFA contents showed non-significant changes. In the dinoflagellate, the contents of SFAs, PUFAs and TFAs overall increased with increased temperature, whereas all FA groups showed low contents at the highest N:P supply ratios (P deficiency) (Fig. 3b).
For the most abundant ω3-PUFA components, C-normalized contents of EPA in the diatom overall increased with increasing nutrient concentrations (p = 0.003, Table 1, Fig. 3c), while DHA contents had non-significant changes. In the dinoflagellate, DHA was the most abundant ω3-PUFA component, showing an overall increase with increasing nutrient concentrations (p = 0.010, Fig. 3d) and a decrease with increasing N:P ratios (p < 0.001).
The effects of temperature, N (P) concentrations and N:P ratios were simultaneously investigated on the growth and lipid composition (sterols and FAs) in the diatom P. tricornutum and the dinoflagellate P. minimum. Sterols and FAs showed markedly different responses to environmental changes, and in most cases even opposite, between the two species (Table 2, Table S7). Different responses to temperature can be attributed to species-specific optimal temperature, i.e., a wider range and lower values of temperature (5 ~ 25 ℃) for P. tricornutum (William and Morris 1982) and a narrower range and higher values (18 ~ 26.5℃) for P. minimum (Grzebyk and Berland 1996). Different strategies of nutrient utilization and lipid production regulation can explain the overall opposite responses of lipids to nutrient condition changes between the two algal species (Elferink et al. 2020, Margalef 1978, Zhuang et al. 2015).
4.1 Response of sterols to the changes in temperature and nutrient regimes
We observed characteristic sterol composition in the two algal species (representing two algal classes), with brassicasterol/epi-brassicasterol and dinosterol (up to 63% of total sterols) as the major sterol in P. tricornutum and in P. minimum, respectively (Table S3). Similar results have been found in other strains of P. tricornutum (Ballantine et al. 1979, Bi et al. 2020, Rampen et al. 2010) and P. minimum (Leblond and Chapman 2002, Volkman et al. 1999).
Of the three environmental factors considered in this study, warming caused the strongest variations in C-normalized sterol contents, i.e., a significant increase (37% ~ 153%) in P. minimum but non-significant changes (a slight increase) of brassicasterol/epi-brassicasterol in P. tricornutum (Table 1, Table 2). Similarly, previous studies found that C-normalized contents of brassicasterol/epi-brassicasterol and dinosterol significantly increased with increasing temperature (12℃-25℃) in three species of dinoflagellates (P. minimum, Prorocentrum donghaiense and Karenia mikimotoi) (Chen et al. 2019, Ding et al. 2019). The increased sterol contents with warming is related to the biochemical functions of sterols in cell membrane fluidity of phytoplankton (Ford and Barber 1983, Hartmann 1998), that is, more sterols can maintain physiological functions of cell membrane at high temperatures.
High nutrient concentrations (under balanced N:P ratios) caused a significant increase (77%) in C-normalized dinosterol contents in P. minimum and non-significant changes (a 35% increase) of brassicasterol/epi-brassicasterol in P. tricornutum (Table 1, Table 2). Piepho et al. (2010) also found a significant positive correlation between phosphate concentrations and C-normalized contents of major sterols in freshwater phytoplankton (fungisterol, chondrillasterol and 22-dihydrochondrillasterol in Scenedesmus quadricauda, ergosterol and fungisterol in Chlamydomonas globose, stigmasterol in Cryptomonas ovata, 24-methylenecholesterol in Cyclotella meneghiniana) under 100 μmol m−2 s−1 (a same light intensity in our experiments). This positive correlation between nutrient concentrations and sterol contents may be explained by an increase in the concentrations of isopentenyl diphosphate, a precursor of sterol biosynthesis, as nutrient concentrations increase (Guschina and Harwood 2009, Piepho et al. 2010, Volkman 2003).
Imbalanced N:P ratios (N and P deficiency) resulted in a significant decrease (11% ~ 27%) in C-normalized dinosterol contents in P. minimum and non-significant changes in brassicasterol/epi-brassicasterol in both species (Table 1, Table 2). Our results show that N and P deficiency had non-significant effects on per-cell dinosterol contents (Table S4, Table S7), but they led to a significant increase (107% ~ 157%) in per-cell POC contents in P. minimum (Table 2). Thus, the decrease in C-normalized dinosterol contents can be attributed to the increase in the per-cell POC contents under nutrient deficient conditions. Under N deficiency, the synthesis of N-containing macromolecules such as protein, nucleic acids and chlorophyll is reduced (Berdalet et al. 1994), and the carbon fixed by phytoplankton is mainly used for the synthesis of energy storage substances such as starch and triglycerides (Dunstan et al. 1993, Ran et al. 2019, Thompson 1996). Accordingly, the synthesis of membrane lipids such as sterols can be also reduced. Under P deficiency, phytoplankton also accumulate energy storage substances (mainly starch), and there is an increase in the concentrations of triose phosphate in chloroplasts, which provides more precursors for carbohydrate synthesis. Also, P deficiency results in an increase in the ratio of glycerate triphosphate to orthophosphate, which increased the activity of AGPase enzyme (Geigenberger 2011) and improved the efficiency of cellular carbohydrate synthesis. The increases in both carbohydrate synthesis precursor and synthetase activity lead to an increase in cellular starch synthesis in phytoplankton (Ran et al. 2019), which can explain the increase in per-cell POC contents and the decrease in C-normalized dinosterol contents in our study.
In summary, under warming, eutrophication and imbalanced N:P ratios in future oceans C-normalized contents of major sterols in the dinoflagellate P. minimum would potentially increase (ca. 48%), while those in the diatom P. tricornutum showed non-significant changes (Table 2). While the increase in dietary sterol contents favors consumer growth and development such as the freshwater herbivore Daphnia magna (Martin-Creuzburg et al. 2014) and the clam Corbicula fluminea (Basen et al. 2012), dinoflagellates may not be the best diets for consumers due to the toxicity in most of them (Aquino-Cruz et al. 2018, Tameishi et al. 2009). Thus, further studies are needed to understand the ecological roles of nutritional quality (sterols) versus toxicity in dinoflagellates, particularly in highly eutrophic coastal areas where dinoflagellates could become dominant.
4.2 Response of FAs to the changes in temperature and nutrient regimes
The two algal species in our study had very characteristic FA compositions, showing the diatom P. tricornutum containing high proportions of MUFAs, mainly 16:1n-7, as well as EPA, and the dinoflagellate P. minimum containing high amounts of 16:0 and DHA (Fig. S5, Table S3). Similar FA profiles have been found in several species of diatoms such as P. tricornutum, Skeletonema costatum and Chaetoceros muelleri (Bi et al. 2014, Pernet et al. 2003, Renaud et al. 1994, Zhukova and Aizdaicher 1995), as well as in the dinoflagellates Gymnodinium sp., Pyramimonas sp. and Prorocentrum sp. (Li and Zhou 1999, Liu et al. 2010, Reitan et al. 1994, Wang et al. 2008, Zhukova and Aizdaicher 1995).
Warming resulted in an overall increase (6% ~ 35%) in C-normalized contents of most FA groups in the two species, and a slight decrease in PUFA contents in P. tricornutum (Table 2). Hixson and Arts (2016) also found that the proportions of PUFAs showed a positive correlation with increasing temperature in dinoflagellates and a negative one in diatoms. High temperature usually enhances the metabolic rate and consequently FA production in phytoplankton (Yang 1988), which can explain the high FA contents at high temperature in our study. On the other hand, higher unsaturated FA contents at lower temperatures could provide homeoviscosity (Los et al. 2013, Sinensky 1974), contributing to negative correlations between PUFA contents and warming in P. tricornutum in our study and other species of marine algae in previous work (Bi et al. 2017, Jiang and Gao 2004, Tseng et al. 2021). These two processes of FA production regulation functioning differently in the two species, being partially attributed to their specific adaptation strategies to warming.
We observed that high nutrient concentrations caused a significant decrease (ca. 40% ~ 50%) in C-normalized contents of SFAs and MUFAs, but an increase (ca. 40% ~ 70%) in PUFA contents in the two species (Table 2). Similarly, the positive effects of high nitrate concentrations on PUFA contents (on a basis of carbon or dry weight) or proportions have been also found in other strains of P. tricornutum (Bi et al. 2014, Liao et al. 2000) and S. costatum (Zhang 2010), and the dinoflagellate Akashiwo sanguinea (Liu et al. 2019). Also, Fernandes et al. (2016) found that the ratio of (SFAs + MUFAs)/PUFAs decreased by ~ 80% in a high nutrient medium compared to that in a low nutrient medium in the green algae Nannochloropsis gaditana, the cryptophyte Rhodomonas marina and the prymnesiophyte Isochrysis sp. These changes in FA contents can be explained by the regulations of high nutrient concentrations on FA biosynthesis. Especially under N saturation, the carbon chain extension and desaturation pathway of FAs are activated in phytoplankton (Flynn et al. 1992). The regulations in FA biosynthesis result in the conversion of most MUFAs to PUFAs and can also explain the decrease in C-normalized contents of SFAs and MUFAs and the increase in PUFAs observed in our study.
Interestingly, we observed that the effects of N and P deficiency on FA group contents were opposite to that of enhanced nutrient concentrations in P. tricornutum, but similar in certain cases in P. minimum (Table 2). Under N deficiency, C-normalized contents of SFAs and MUFAs significantly increased, and PUFA contents decreased in P. tricornutum, consistent with previous findings (Bi et al. 2014, Bi et al. 2017, Trivedi et al. 2022, Wang et al. 2019), indicating the accumulation of SFAs and MUFAs and potentially triacylglycerols (TAGs), which can support phytoplankton growth at favorable environmental conditions (Dunstan et al. 1993, Khozin-Goldberg et al. 2002). This mechanism, however, did not apply in P. minimum, as SFA contents showed a significant decrease and PUFAs significantly increased under N deficiency in our study, as well as in the dinoflagellate Symbiodinium spp., which showed a decrease in SFA proportions under N deficiency (Weng et al. 2014). These different responses of FAs between diatoms and dinoflagellates may be largely attributed to their differential strategies of nutrient utilization (Margalef 1978), e.g., higher carbon-specific nitrate uptake rates in diatoms than dinoflagellates (Litchman et al. 2007).
Overall, characteristic FA profiles in the two species underlined fluctuations at different temperatures and nutrient regimes. Remarkably, C-normalized contents of FA groups varied differently, in most cases were opposite, between the two species in response to imbalanced N:P ratios and warming. For the ecologically important FA group (PUFAs), its C-normalized contents in the dinoflagellate P. minimum would overall increase (ca. 9%), while those in the diatom P. tricornutum had non-significant changes in future ocean (warming, eutrophication and imbalanced N:P ratios) (Table 2). Our results are consistent with the FA responses in other species of diatoms and dinoflagellates in previous studies, suggesting differential FA production regulation in diatoms and dinoflagellates under imbalanced N:P ratios and warming conditions.
4.3 Implications for the application of lipid biomarkers in marine ecology
Our PCA results showed a clear separation between in the two species of marine phytoplankton (Fig. 4, Table S8), indicating relatively unique and stable profiles of sterols and FAs in the two species across all treatments. Indeed, such characteristic sterol and FA composition correlated positively with abundance ratios between diatoms and dinoflagellates over a wide range of combinations of temperature and nutrient regimes in laboratory bi-cultures (Bi et al. 2021), and they have been applied as useful proxies for diatom-dinoflagellate community in aquatic environments (Cañavate 2019, Dalsgaard et al. 2003, Galloway and Winder 2015, Hernandez et al. 2008, Wang et al. 2022, Wu et al. 2016). While the contents of sterols and FAs varied as culture conditions changed, our findings provide important evidence for the applicability of sterols and FAs as indicators for diatom-dinoflagellate community in highly variable environments.
Quantitatively, we observed different responses of sterol and FA contents between the two species under variable conditions. Under future ocean warming, eutrophication and imbalanced N:P ratio scenarios, C-normalized contents of major sterols and PUFAs in the diatom P. tricornutum showed non-significant changes, while those in the dinoflagellate P. minimum showed an overall increase (ca. 9% ~ 48%) (Table 2). Consumers feeding on high dietary sterols and PUFAs potentially have higher growth rates, causing higher transfer efficiencies in food webs (Martin-Creuzburg et al. 2005, McInerney et al. 2022, Müller-Navarra et al. 2004, Ruess and Müller-Navarra 2019). An example of this is that high egg production rates were observed in marine copepods feeding on dinoflagellates (Evjemo et al. 2008, Vehmaa et al. 2011). In the western Gulf of Finland, decreased dinoflagellate abundance and increased cyanobacteria abundance in response to heatwave resulted in large variability in in situ egg production of copepods in spring and summer (von Weissenberg et al. 2022). While higher nutritional food quality (e.g., sterols and FAs) in dinoflagellates may be estimated under future ocean scenarios, other factors contributing to food quality such as cell size and toxicity should also be considered in future studies on the effects of dinoflagellate predominance in marine ecosystems.
The present study evaluated the variability of two lipid classes (sterols and FAs) over a wide range of combinations of temperature, N (P) concentrations and N:P ratios in two algal species representing particular phytoplankton groups, i.e., diatoms and dinoflagellates. It provides the first empirical evidence of relatively unique and stable sterol and FA profiles in the two species across large ranges of temperature and nutrient regimes, and thus verified the application of sterols and FAs as indicators for diatom-dinoflagellate community composition. Furthermore, this study shows that warming and high nutrient concentrations had the most pronounced effects on C-normalized contents of sterols and FAs, respectively. We further quantified the effects of the three environmental factors, showing different response trends, in some cases even opposite, between the two species. Our results thus provide an important basis to quantitatively assess the changes in diatom-dinoflagellate community structures using lipid biomarkers, and to compare the ecological roles of diatoms and dinoflagellates in food web dynamics under future ocean scenarios.
Acknowledgments
We would like to thank Hailong Zhang, Gue’e Jin, Yu Zhan and Xiaoke Xin for technical support. Yaoyao Wang, Peng Peng, Jiawei Gao and Xiaohan Bao are acknowledged for their assistance in the experiments. This study was supported by the National Natural Science Foundation of China (grant numbers 41876118, 41630966). This is MCTL (Key Laboratory of Marine Chemistry Theory and Technology) contribution #299.
Author contributions
RB and MZ contributed to the plan of experiments. Experiment conduction and sample analysis were performed by CZ (Cao Zhong), with assistance of CZ (Chuanli Zhang), JC, LL and DY. Data analysis were performed by CZ (Cao Zhong). The first draft of the manuscript was written by CZ (Cao Zhong) and RB, and MZ commented on previous versions of the manuscript. All authors have read and approved the final manuscript.
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval
This article does not contain work that required ethical approval as no animals were approached.
Data Availability
The datasets generated during and/or analysed during the current study are available in the supplementary materials.
Table 1 Summary of the selected GLMM results showing the effects of temperature (T), N (P) concentrations (Nut. conc.) and N:P ratios (N:P) on per-cell POC contents and C-normalized lipid contents in Phaeodactylum tricornutum and Prorocentrum minimum. Significant results (p < 0.05) are shown in bold. TFAs: total fatty acids, SFAs: saturated fatty acids, MUFAs: monounsaturated fatty acids, PUFAs: polyunsaturated fatty acids, EPA: eicosapentaenoic acid (20:5n-3), DHA: docosahexaenoic acid (22:6n-3).
Species |
Variable |
Factor |
Coefficient |
SE |
t |
p |
n |
P. tricornutum |
POC/cell |
Intercept |
1.150 |
0.130 |
8.827 |
<0.001 |
80 |
|
|
T |
0.005 |
0.006 |
0.712 |
0.479 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
0.636 |
0.527 |
|
|
|
N:P |
<0.001 |
0.001 |
−0.249 |
0.804 |
|
|
Brassicasterol/POC |
Intercept |
0.476 |
0.138 |
3.454 |
0.001 |
79 |
|
|
T |
−0.003 |
0.007 |
−0.435 |
0.665 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
1.504 |
0.137 |
|
|
|
N:P |
−0.001 |
0.002 |
−0.422 |
0.674 |
|
|
TFA/POC |
Intercept |
2.263 |
0.129 |
17.488 |
<0.001 |
74 |
|
|
T |
−0.003 |
0.006 |
−0.402 |
0.689 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−6.258 |
<0.001 |
|
|
|
N:P |
0.004 |
0.001 |
2.531 |
0.014 |
|
|
SFA/POC |
Intercept |
1.830 |
0.139 |
13.149 |
<0.001 |
74 |
|
|
T |
−0.001 |
0.007 |
−0.124 |
0.901 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−7.726 |
<0.001 |
|
|
|
N:P |
0.005 |
0.002 |
3.169 |
0.002 |
|
|
MUFA/POC |
Intercept |
1.827 |
0.137 |
13.337 |
<0.001 |
74 |
|
|
T |
<0.001 |
0.007 |
0.032 |
0.975 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−7.411 |
<0.001 |
|
|
|
N:P |
0.005 |
0.001 |
3.168 |
0.002 |
|
|
PUFA/POC |
Intercept |
1.444 |
0.163 |
8.833 |
<0.001 |
74 |
|
|
T |
−0.006 |
0.008 |
−0.743 |
0.460 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
1.773 |
0.081 |
|
|
|
N:P |
−0.001 |
0.002 |
−0.483 |
0.630 |
|
|
EPA/POC |
Intercept |
2.715 |
0.368 |
7.370 |
<0.001 |
74 |
|
|
T |
−0.014 |
0.018 |
−0.737 |
0.463 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
3.056 |
0.003 |
|
|
|
N:P |
−0.003 |
0.004 |
−0.829 |
0.410 |
|
|
DHA/POC |
Intercept |
0.349 |
0.188 |
1.857 |
0.069 |
74 |
|
|
T |
−0.013 |
0.010 |
−1.351 |
0.182 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−0.811 |
0.421 |
|
|
|
N:P |
0.002 |
0.002 |
0.908 |
0.368 |
|
P. minimum |
POC/cell |
Intercept |
2.805 |
0.132 |
21.245 |
<0.001 |
78 |
|
|
T |
−0.014 |
0.007 |
−2.150 |
0.035 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−6.187 |
<0.001 |
|
|
|
N:P |
0.005 |
0.001 |
3.451 |
0.001 |
|
|
Brassicasterol/POC |
Intercept |
−0.449 |
0.124 |
−3.622 |
0.001 |
79 |
|
|
T |
0.021 |
0.006 |
3.347 |
0.001 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
0.458 |
0.648 |
|
|
|
N:P |
−0.003 |
0.001 |
−1.906 |
0.060 |
|
|
Dinosterol/POC |
Intercept |
−0.742 |
0.154 |
−4.823 |
<0.001 |
79 |
|
|
T |
0.050 |
0.008 |
6.541 |
<0.001 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
3.018 |
0.003 |
|
|
|
N:P |
−0.004 |
0.002 |
−2.609 |
0.011 |
|
|
TFA/POC |
Intercept |
1.903 |
0.119 |
16.038 |
<0.001 |
63 |
|
|
T |
0.024 |
0.006 |
3.713 |
<0.001 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−0.220 |
0.827 |
|
|
|
N:P |
−0.007 |
0.001 |
−5.178 |
<0.001 |
|
|
SFA/POC |
Intercept |
1.569 |
0.136 |
11.497 |
<0.001 |
63 |
|
|
T |
0.017 |
0.007 |
2.261 |
0.027 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−0.065 |
0.948 |
|
|
|
N:P |
−0.005 |
0.002 |
−2.956 |
0.004 |
|
|
MUFA/POC |
Intercept |
1.367 |
0.119 |
11.462 |
<0.001 |
60 |
|
|
T |
0.008 |
0.007 |
1.263 |
0.212 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
−2.307 |
0.025 |
|
|
|
N:P |
−0.003 |
0.001 |
−2.530 |
0.014 |
|
|
PUFA/POC |
Intercept |
1.381 |
0.190 |
7.258 |
<0.001 |
62 |
|
|
T |
0.024 |
0.010 |
2.362 |
0.022 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
2.734 |
0.008 |
|
|
|
N:P |
−0.011 |
0.002 |
−4.780 |
<0.001 |
|
|
DHA/POC |
Intercept |
1.395 |
0.175 |
7.975 |
<0.001 |
61 |
|
|
T |
0.010 |
0.009 |
1.034 |
0.306 |
|
|
|
Nut. conc. |
<0.001 |
<0.001 |
2.659 |
0.010 |
|
|
|
N:P |
−0.010 |
0.002 |
−4.772 |
<0.001 |
|
Table 2 The mean percent changes in population growth, per-cell POC contents and C-normalized lipid contents in response to warming, high nutrient concentrations (at balanced N:P ratios) and N- and P-deficiency (at the normal nutrient concentration) in Phaeodactylum tricornutum and Prorocentrum minimum. Significant changes are shown in color according to GLMMs. TFAs: total fatty acids, SFAs: saturated fatty acids, MUFAs: monounsaturated fatty acids, PUFAs: polyunsaturated fatty acids, EPA: eicosapentaenoic acid (20:5n-3), DHA: docosahexaenoic acid (22:6n-3).
Species |
Variables |
Effect |
||||
Warming |
High nutrient |
−N |
−P |
|||
P. tricornutum |
μmax |
14% |
4% |
19% |
18% |
|
|
Cell density |
−6% |
213% |
−24% |
−38% |
|
|
POC/cell |
20% |
8% |
−35% |
−26% |
|
|
Brassicasterol/POC |
2% |
35% |
59% |
56% |
|
|
SFA/POC |
7% |
−53% |
52% |
37% |
|
|
MUFA/POC |
10% |
−50% |
64% |
49% |
|
|
PUFA/POC |
−6% |
72% |
−33% |
−32% |
|
|
TFA/POC |
6% |
−43% |
44% |
32% |
|
|
EPA/POC |
−5% |
103% |
−17% |
−11% |
|
|
DHA/POC |
−16% |
214% |
69% |
−1% |
|
P. minimum |
μmax |
64% |
−17% |
−22% |
−46% |
|
|
Cell density |
250% |
240% |
−41% |
−50% |
|
|
POC/cell |
−15% |
−55% |
107% |
157% |
|
|
Brassicasterol/POC |
37% |
−7% |
0% |
−16% |
|
|
Dinosterol/POC |
153% |
77% |
−11% |
−27% |
|
|
SFA/POC |
35% |
−28% |
−39% |
−53% |
|
|
MUFA/POC |
22% |
−40% |
18% |
−14% |
|
|
PUFA/POC |
12% |
43% |
32% |
−52% |
|
|
TFA/POC |
40% |
−25% |
−13% |
−53% |
|
|
DHA/POC |
9% |
35% |
60% |
−42% |
Increase |
Decrease |