Metabolic responses of phytoplankton to combined global change drivers: temperature and resource availability

Climate change is altering local temperatures and resource availability of many ecosystems. We explore the impacts of these changes on the metabolism of phytoplankton, organisms that are crucial as the base of aquatic food webs and engines of aquatic biogeochemical cycling. Specifically, we investigate the metabolic responses of six freshwater phytoplankton species---representing green algae, diatoms, and cyanobacteria---to warming in combination with light, phosphorus, and nitrogen, the three most critical resources for phytoplankton. Using direct-infusion mass spectrometry, a high-throughput metabolomics approach, we identify interactive effects of temperature and resource availability on phytoplankton metabolism. We detect thousands of metabolites involved in key pathways, including amino acid, carbohydrate, and lipid metabolism. Our results show that resource limitation had a stronger effect on metabolism than temperature across all species. While each resource induced distinct metabolic changes, nitrogen and phosphorus limitation triggered more similar responses, whereas light limitation resulted in a unique metabolic profile. We also identified a core set of metabolic pathways involved in all responses, alongside resource-specific pathways. These findings provide mechanistic insights into how phytoplankton metabolically respond to key environmental drivers, enhancing our understanding of their responses to future climate conditions. Title Metabolic responses of phytoplankton to combined global change drivers: temperature and resource availability Running head (45 words) Temperature and resource-dependent metabolism Authors Vanessa Weber de Melo 1, Patrick Thomas 1, Marc J-F Suter 2, Anita Narwani 1 Affiliations 1 Eawag, Swiss Federal Institute of Aquatic Science and Technology, Department of Aquatic Ecology, Dübendorf, Switzerland 2 Eawag, Swiss Federal Institute of Aquatic Science and Technology, Department of Environmental Toxicology, Dübendorf, Switzerland E-mails Vanessa Weber de Melo: [email protected] Patrick Thomas: [email protected] Marc J-F Suter: [email protected] Anita Narwani: [email protected] Corresponding author Vanessa Weber de Melo Data availability statement All the metabolomics data and related metadata will be available at ERIC open data repository. During revision, all the data and code to reproduce the analyses can be anonymously accessed here: https://drive.switch.ch/index.php/s/cLIj4CKkpgPRSjT Funding statement This study was funded with a Seed grant to Marc J.-F. Suter and Anita Narwani from Eawag, an SNF Project Grant (310030_197812) to Anita Narwani, and an SNF Flexibility Grant (310030_197812/2) to Vanessa Weber de Melo. Author contributions AN and MJFS formulated the research question and obtained funding. VWM and AN designed the experiments. VWM and PT conducted the laboratory experiments. VWM performed the data analyses, with input from AN and PT. VWM and AN wrote the paper with feedback from all authors. Conflict of interest disclosure All authors declare that they have no conflicts of interest. Climate change is altering local temperatures and resource availability of many ecosystems. We explore the impacts of these changes on the metabolism of phytoplankton, organisms that are crucial as the base of aquatic food webs and engines of aquatic biogeochemical cycling. Specifically, we investigate the metabolic responses of six freshwater phytoplankton species—representing green algae, diatoms, and cyanobacteria—to warming in combination with light, phosphorus, and nitrogen, the three most critical resources for phytoplankton. Using direct-infusion mass spectrometry, a high-throughput metabolomics approach, we identify interactive effects of temperature and resource availability on phytoplankton metabolism. We detect thousands of metabolites involved in key pathways, including amino acid, carbohydrate, and lipid metabolism. Our results show that resource limitation had a stronger effect on metabolism than temperature across all species. While each resource induced distinct metabolic changes, nitrogen and phosphorus limitation triggered more similar responses, whereas light limitation resulted in a unique metabolic profile. We also identified a core set of metabolic pathways involved in all responses, alongside resource-specific pathways. These findings provide mechanistic insights into how phytoplankton metabolically respond to key environmental drivers, enhancing our understanding of their responses to future climate conditions.

metabolic responses, resource limitation, temperature-nutrient interactions, freshwater phytoplankton, metabolomics, warming

Phytoplankton play critical roles in all pelagic aquatic ecosystems, being responsible for a large portion of the global primary productivity and for essential ecosystem services such as geochemical cycling and oxygen production (Naselli-Flores & Padisák, 2023). Aquatic ecosystems are increasingly affected by multiple abiotic changes driven by climate change, which extend beyond rising surface water temperatures to include alterations in stratification, nutrient and light availability, and water quality (Sallée et al., 2021; Woolway et al., 2022). Understanding how phytoplankton respond to these combined changes is essential for predicting how aquatic ecosystems will adapt to future climate conditions. Interactive effects of temperature and resource availability on growth rates have been documented (Allen & Gillooly, 2009; Cross et al., 2015; Thomas et al., 2017; Weber de Melo et al., 2025), but the underlying metabolic changes responsible for these interactions at higher levels of biological organization remain largely unexplored (Boyd et al., 2015). Numerous studies have examined how resources influence phytoplankton metabolism. Phosphorus limitation in diatoms causes fundamental changes in their carbon metabolism, leading to the storage of excess carbon as fatty acids and the degradation of phospholipids to scavenge phosphorus (Brembu et al., 2017). Nitrogen limitation can reduce photosynthetic capacity, decrease amino acid synthesis and induce lipid accumulation in multiple phytoplankton species (Alipanah et al., 2015; Bölling & Fiehn, 2005; Brembu et al., 2017). Studies on the coccolithophore Emiliana huxleyi revealed that phosphorus starvation induces general cell-cycling arrest due to reduced nucleic acid synthesis and energy metabolism, while nitrogen limitation led to impairment of the enzymatic machinery, causing a more widespread disruption of cell functioning (Rokitta et al., 2016; Wördenweber et al., 2018). Light limitation also affects carbon metabolism and increases pigment concentration (Fisher et al., 2023). The effects of temperature on phytoplankton metabolism have also been documented. Higher temperatures require resource allocation toward carbohydrates, pigments and enzymes necessary for sustaining elevated metabolic rates in dinoflagellates (H. Zhang et al., 2022), and similar increases in pigment content, along with fatty acid accumulation, were observed in green algae (Trentin et al., 2024). The interactive effects of resource limitation and temperature on phytoplankton metabolism remain poorly documented; however, a few studies have described the combined effects of light and temperature on cyanobacteria (Mohanty et al., 2022) and coccolithophores (Aguilera-Sáez et al., 2019). While a tradition in the Metabolic Theory of Ecology (MTE) (Brown et al., 2004) has focused on the way in which temperature impacts respiration and photosynthesis as key metabolic pathways driving various biological processes (Bernacchi et al., 2001; Gillooly et al., 2001; Yvon-Durocher et al., 2010), metabolism comprises a complex network of interconnected reactions (Kochhar & Gujral, 2020). Previous studies have shown that changes in photosynthesis components in response to temperature depend on resource availability (Fernández-González et al., 2020), and have predicted complex effects of temperature and resource availability on macromolecular pools, which are directly linked to cell metabolism (Armin et al., 2023; Armin & Inomura, 2021). A more comprehensive understanding of other metabolic components is critical to elucidate how phytoplankton respond to complex environmental changes and could be important to explain deviations from the expectations of the MTE. To address this gap, we investigate the combined effects of temperature and resource limitation on the metabolism of freshwater phytoplankton. Recent advances in metabolomics methods now allow the fast and precise measurement of a large number of metabolites. These methods are broadly categorized into targeted and untargeted approaches, each with distinct strengths and limitations (Ribbenstedt et al., 2018). Targeted approaches provide accurate quantification of metabolite concentrations and focus on a predefined subset of metabolites. While advancements have reduced sample processing times, these methods often still require long runtimes. In contrast, untargeted approaches offer a comprehensive overview of metabolism without prior knowledge, making them valuable for hypothesis generation. They generally also have shorter sample processing times (especially when dropping chromatographic separation in favor of direct injection), enabling a greater throughput of samples. However, they are often only semi-quantitative (Schrimpe-Rutledge et al., 2016). High-throughput metabolomics methods are essential for addressing ecological questions, such as understanding organisms’ responses to multiple environmental factors, which demand the analysis of large sample sets (Walker et al., 2022). Specifically, many studies have documented plasticity in eco-physiological traits in response to changing temperature and resources; however, integrating our understanding of trait-based ecology with multivariate metabolomics data is needed to understand the mechanisms driving observed changes in organismal responses (Walker et al., 2022). In this study, we investigate the interactive effects of temperature and resource limitation on the metabolism of six freshwater phytoplankton species. We employ an untargeted metabolomics method to analyze metabolic changes in response to interactions between temperature and three key resources for phytoplankton: light, nitrogen and phosphorus. In a previous study, we investigated the interaction of these abiotic factors on population growth rate and observed changes in the temperature sensitivity of growth rate that are dependent on the type of resource limitation and also on the species (Weber de Melo et al., 2025). Here, we aim to explore the molecular mechanisms that underlie these changes in growth rate sensitivity to temperature by measuring phytoplankton metabolism. Based on previous findings, we expected to observe specific sets of pathways that respond to the different resource limitations, and also general stress-responses that are independent of the limiting resource. We expected that light limitation would affect pathways related to light harvesting and also energy metabolism, as reported in diatoms (Fisher et al., 2023). We expected that nitrogen limitation would lead to a general repression of amino acid biosynthesis, as previously demonstrated for green algae and coccolithophores (Bölling & Fiehn, 2005; Wördenweber et al., 2018). Under phosphorus limitation, we expected alterations in membrane phospholipids and in nucleic acids, molecules that require large amounts of phosphorus, and also a general reduction in photosynthetic activity and the related carbon fixation pathways (Grossman & Aksoy, 2015; S. Lin et al., 2016). Warming also greatly affects phytoplankton metabolism and growth, and we expected warming to affect central carbon metabolism required for higher growth rates but also trigger specific heat stress responses (Dedman et al., 2023; Hong et al., 2023). We measured metabolites across the most important classes, including amino acids, carbohydrates and lipids. We found that resources had a stronger effect on the metabolism of phytoplankton than temperature, and nitrogen and phosphorus limitations caused similar alterations in the metabolism of the studied species, while light limitation had unique metabolic responses.

Growth experiments We studied six freshwater phytoplankton species: Chlamydomonas reinhardtii, Scenedesmus acuminatus, Pediastrum boryanum, Cyclotella meneghiniana, Asterionella formosa and Synechoccocus sp. We will hereafter refer to the species by their genus name. All species were obtained from culture collections, except for Synechococcus, which has been isolated in our lab. Cultures were acclimated to our laboratory conditions, growing on COMBO medium (Kilham et al., 1998) at 20°C. Each species was subjected to a growth experiment, in which we manipulated temperature and three of the most important abiotic resources of phytoplankton, nitrogen, phosphorus and light. We combined four temperatures (15, 20, 25 and 30°C) with seven resource levels (a control level where resources were abundant and two levels of resource limitation for each abiotic resource), which resulted in 28 unique treatments (see Table S1 for details on the treatments). Nitrogen and phosphorus were manipulated through the medium, while light was manipulated with the use of neutral density filters. After an acclimation phase of three days to the resource and temperature level, five replicate populations of each treatment were grown in 150 ml of COMBO medium in cell culture flasks. Populations started at very low densities and were monitored until carrying capacity was reached. The experiments lasted between 9 and 16 days. For more details of the experimental design and procedures, see Weber de Melo et al., 2025. Metabolomic profiling Metabolome samples were collected at the end of the growth experiments. We sampled the control treatments during the exponential phase, to ensure populations were not strongly limited by any of the resources, and resource limitation treatments were sampled once carrying capacity was reached. The cultures of Asterionella did not grow at sufficiently high densities for metabolites sampling at 30°C, therefore this species only contains samples from between 15 and 25 °C. For each sampled population, we centrifuged 100 ml of the culture at 3200 RPM for 10 minutes, the medium was discarded, and the remaining cell pellets were snap frozen in liquid nitrogen and stored at -80°C until processing (Fig. S1). We focused our study on polar metabolites, since they cover a large diversity of molecules that are part of the central metabolism and are involved in key cellular functions in phytoplankton (Heal et al., 2021). Polar metabolites were extracted with a methanol-based protocol. Metabolomics data was generated following the direct-infusion mass spectrometry workflow described in Southam et al., 2016. Samples were analyzed in an Orbitrap Q Exactive Plus mass spectrometer (Thermo Fisher Scientific) with a direct infusion, chip-based nanoelectrospray ion source (Triversa Nanomate with HD ESI chip, Advion Biosciences) in positive mode (1.2 kV, resolution of 140’000 @ m/z 200). For each species, one QC sample was created by pooling all the 140 samples of the experiment. Blank, QC and experimental samples were analyzed in three technical replicates, which were divided into five equal batches containing one of the five experimental replicates of each treatment. Samples were injected in a random order within each batch. Metabolomics data analyses Raw data was processed with DIMSpy (Weber & Zhou, 2020) following the recommended replicate, blank and sample filter steps described in Southam et al., 2016. The R package structToolbox (Lloyd et al., 2021) was used for data quality control and signal correction, following these steps: signal drift and batch correction were performed with the Quality Control Robust Spline method, samples that had more than 50% of missing data were removed, a Kruskal-Wallis test was used to remove features not reliably measured in the QC samples, and features with high analytical variation were excluded. Table S2 summarizes the samples and number of features remaining after data quality control for each species. These metabolic features represent distinct ions detected by the mass spectrometer, but note that detected ions can also represent fragmentation products of original metabolites present in algae. We hereafter refer to all such detected metabolic features/ions as “metabolites”. Data normalization was performed with the R package MetaboAnalystR (Pang et al., 2020) by applying a quantile normalization to the features, a log normalization to the samples, and the data was scaled by mean centering, as suggested by previous metabolomics pipelines (Southam et al., 2016; Sun & Xia, 2024). Two related methods were used to visualize the data and identify metabolites important in each of our treatments, Partial Least-Squares Discriminant Analysis (PLS-DA) and sparse Partial Least-Squares Discriminant Analysis (sPLS-DA), which were performed with the R package mixOmics (Rohart et al., 2017). While PLS-DA performs dimensionality reduction by taking into account the treatments of the samples, sPLS-DA does the same procedure but also performs feature selection, i.e. determines the minimum number of features that are necessary to correctly identify the samples treatments. We performed three separate sPLS-DA to identify metabolites most affected by temperature and resources in isolation and combined. The first set of analyses labeled the samples according to the 7 resource levels, the second set of analyses only labeled them with the four temperatures, and the third set of analyses used the full treatment combining both resource and temperature, resulting in 28 unique levels. We used a permutational ANOVA (PERMANOVA) to test for differences between the seven resource levels in the metabolomics data. We performed this with the R package vegan (Oksanen et al., 2024). Two separate pathway enrichment analyses were performed using the mummichog algorithm (S. Li et al., 2013) implemented in MetaboAnalystR. First, we identified the overrepresented pathways when comparing the control condition to each one of the six low resource treatments, which identifies the pathways most affected by resource limitation. Subsequently, we compared the control condition at 15°C to the four temperatures in each resource limitation treatment, which identifies the interactive effect of resource limitation and warming. We also performed analyses comparing only the control at 15°C with the control condition at the three higher temperatures, to identify the pathways affected by temperature only. However, for most species there were not enough features that were different between the treatments, which prevented the use of the method to identify the enriched pathways. For all the enrichment analyses, we used all measured features after the quality control and normalization steps described above. The KEGG database for Chlamydomonas reinhardtii was used as a reference for all green algae and diatoms, while the KEGG database for Synechococcus elongatus was the reference for the Synechococcus samples.

After data filtering and quality control, we measured 6,422 metabolites across 122 samples of Chlamydomonas, 5,926 metabolites across 95 samples of Scenedesmus, 6,617 metabolites across 110 samples of Pediastrum, 6,682 metabolites across 139 samples of Synechococcus, 2,860 metabolites across 123 samples of Cyclotella, and 2,636 metabolites across 88 samples of Asterionella . All species retained at least two biological replicates in each one of the treatments, except for Asterionella in which no sample at 30°C was analyzed due to the reduced cell density during the experiments, which prevented any metabolites extraction. Resource limitation of various kinds had pronounced effects on phytoplankton metabolism (Fig. 1). The type of resource limitation had a greater influence on sample clustering than temperature (Fig. 1, Fig. S2-3). PERMANOVA tests revealed significant differences between the clusters of the seven resource levels in all six species (Fig. S2). Nitrogen and phosphorus limitation affected metabolism in a more similar manner compared to light limitation. Samples from nitrogen and phosphorus limitation showed substantial overlap, while samples under light limitation formed a distinct cluster, regardless of the species or temperature (Fig. 1, Fig. S2). The control condition, in which all resources were abundantly provided in a balanced supply, consistently clusters between the light-limited samples and the nitrogen- and phosphorus-limited samples, which provides further evidence that light limitation induces a distinct set of metabolic changes compared to nitrogen and phosphorus limitations. We identified the metabolites most affected by resource levels, temperature levels, and the combination of both using sPLS-DA (Fig. 1, Fig. S4–S5). Consistent with the PCA and PLS-DA results, the influence of resource levels on phytoplankton metabolism was stronger than that of temperature. This is evident from the smaller classification error rates in analyses based solely on resource levels compared to those based solely on temperature levels (Table S3). The only exception was Asterionella, which exhibited a lower classification error rate for the temperature-based sPLS-DA, suggesting a comparatively stronger temperature signal in its metabolism. Analyses incorporating the full combination of resource and temperature levels consistently showed higher classification error rates than those based only on resource levels. This is expected, as the full analysis included 28 groups, compared to 7 groups in the resource-only analysis, dividing the statistical power among a greater number of experimental treatments to be analyzed. The number of metabolites selected as the most important for classifying the samples into their treatments also varied according to the species. For most species, a larger number of metabolites was selected for the resource-only model in comparison to the interactive temperature x resource, and these were also the analyses with the lowest classification error rates (Table S3). There was some overlap between the metabolites selected by each sPLS-DA, but for most species the largest part of the metabolites were uniquely selected for each of the analyses, indicating different sets of metabolites responding to resource, temperature and to the combination of both (Fig. 2). The metabolic pathways significantly enriched in response to resource limitation in each species were identified with the mummichog algorithm. We identified 11 pathways significantly enriched in Chlamydomonas across all six treatments , 19 pathways across four treatments in Scenedesmus, 8 pathways across four treatments in Pediastrum and three pathways in the cyanobacterium Synechococcus . In the diatoms, we identified 8 pathways significantly enriched across four treatments in Cyclotella and three pathways across two treatments in Asterionella (Fig. 3) . Two species, Asterionella and Cyclotella, had a much smaller set of metabolites measured, which reduces the power to detect enriched pathways in response to the experimental treatments. We identified a small number of metabolic pathways that appear to play a crucial role across all types of resource limitation. The pathways for arachidonic acid metabolism, tyrosine metabolism and cutin, subserine and wax biosynthesis were significantly enriched under all three resource limitations — light, nitrogen, and phosphorus — and in multiple species (Fig. 3A). Most pathways were significantly enriched in only one of the resource treatments: 10 pathways unique to light, eight pathways unique to nitrogen and six pathways unique to phosphorus limitation (Fig. 3B). Carbon fixation was uniquely enriched under light limitation, while energy metabolism was specific to phosphorus limitation. Two specific classes of metabolic pathways, nucleotide metabolism and the biosynthesis of secondary metabolites, were enriched exclusively under nitrogen limitation. In contrast, pathways associated with carbohydrate metabolism, lipid metabolism, and amino acid metabolism seem to be relevant for resource limitation in a species-dependent manner, since they were enriched in response to specific resource limitations and in different species (Fig. 3A). We also identified the pathways that showed an effect of both temperature and resource level by comparing the control condition at 15°C to each resource and temperature level (Fig. S6). In many species, the number of metabolites that changed was too small for the analyses to be performed, but for four species, we could identify significantly enriched pathways in these comparisons. Out of the 21 pathways, 12 were also present in the resource only analyses, while 9 pathways were only identified when analyzing the effect of both resource and temperature level. The KEGG class of glycan biosynthesis and metabolism (Fig. S6) was only present in such interactive analyses.

We evaluated the effects of the three most important resources for phytoplankton - light, nitrogen and phosphorus - and their interaction with temperature in the metabolism of six freshwater phytoplankton species. Comparative analyses of cell metabolism in response to multiple resources are rare in freshwater phytoplankton. As expected, each of the three resources elicited a distinct metabolic response, although nitrogen and phosphorus limitation resulted in more similar metabolic profiles compared to light limitation, as evident in all PCA and PLS-DA analysis (Fig. 3, Fig. S2-3). Previous studies in marine phytoplankton compared the effects of nitrogen and phosphorus starvation on the cell metabolism and also encountered distinct responses to these resources, both in the proteome of Thalassiosira pseudonana (Q. Lin et al., 2017) and in the metabolome of Emiliania huxleyi (Wördenweber et al., 2018), indicating that different resource stresses require specific cell responses. In our study, the limiting resource was the primary factor driving changes in the metabolism of the studied species, while temperature had a secondary effect on the metabolism of most species (Fig. 1). Hessen et al., 2017 evaluated the effects of phosphorus and temperature in the cellular stoichiometry of the green algae Chlamydomonas and also observed a stronger effect of nutrient limitation in comparison to temperature, corroborating the increased relevance of the resource status in comparison to temperature to functioning of the cells. Mohanty et al., 2022 compared the interactive effects of temperature and light in two cyanobacteria and observed a stronger effect of temperature on the metabolism of Hapalosiphon while light had a stronger influence on Planktothricoides. We studied one cyanobacterium, and interestingly, sPLS-DA analyses for temperature in this species showed a similar error rate to the analyses for the resource limitation alone (Table S3 and Fig. S5), indicating a similar influence of both resources and temperature on the measured metabolites. Cyanobacteria are among the phytoplankton groups best adapted to warmer temperatures (Carey et al., 2012) and this may suggest that their metabolism is more responsive to temperature compared to the other functional groups. One of the studied diatoms, Cyclotella, showed a stronger metabolic response to temperature in comparison to resource level (Table S3 and Fig. S5), but the limited number of metabolites measured in this species warrants cautions in drawing general conclusions about its metabolism. Unlike cyanobacteria, diatoms are usually cold-adapted species, and therefore, a stronger metabolic response to temperature might be related to their narrower thermal niche. We observed that light limitation led to changes in pathways related to carbon fixation, which has been previously seen in other phytoplankton species and has been related to a general reduction in the available energy necessary for growth (Fisher et al., 2023; Fisher & Halsey, 2016). Multiple pathways related to amino acid biosynthesis were also significantly overrepresented in the nitrogen-limited populations, such as tryptophan, tyrosine and phenylalanine, but this was only observed in the most limiting treatment level (Fig. 3). Since amino acids are one of the largest pools of nitrogen in the cells, reductions in the biosynthesis of these molecules are expected under nitrogen stress and have been previously observed in multiple functional groups (Liefer et al., 2019; Park et al., 2015). Pathways related to sulphur metabolism were enriched in response to P limitation in the green algae Scenedesmus and in response to N limitation in the cyanobacteria Synechococcus . Previous studies have observed that, under phosphorus limitation, both cyanobacteria (Martin et al., 2023; Van Mooy et al., 2006) and green algae (Sato et al., 2000) decrease their phosphorus requirements by using sulphate as a substitute for phosphate in lipid biosynthesis, which is in line with our observations and might be a more general phytoplankton response to nutrient stress. The expected change in nucleic acids as a result of phosphorus starvation was not observed in our study. Pyrimidine metabolism is the only pathway belonging to the nucleotides metabolism that was significantly enriched in our study, but this only occurred in response to nitrogen limitation (Fig. 3). We found a small number of pathways that were enriched across multiple resource types and in multiple species, indicating that these pathways might be related to general stress-response in phytoplankton, independent of the factor causing the stress (Fig. 3). The pathway related to the metabolism of the amino acid tyrosine has been overrepresented in response to all three resource limitations. Previous studies have identified a role of tyrosine phosphorylation in signaling and regulating multiple cell activities such as growth, photosynthesis and responses to stressful conditions in different phytoplankton species, such as cyanobacteria (Yang et al., 2013; C.-C. Zhang et al., 2005) and in the model diatom Phaeodactylum tricornutum (Chen et al., 2014), which could also explain the patterns observed in this study. Multiple pathways related to lipid metabolism were enriched in response to all three manipulated resources. Accumulation of neutral lipids, specially triacylglycerol, due to adverse environmental conditions is observed in several phytoplankton species and is a form of carbon and energy storage, which is often explored by biofuel and biomaterial technologies (Hu et al., 2008). In our study, the biosynthesis of fatty acids, which are precursors of triacylglycerols, was enriched in Scenedesmus, Asterionella and Pediastrum, while sphingosine degradation, which leads to an increase in fatty acids, was enriched in Chlamydomonas (Fig. 3) and also in response to specific temperatures combined with light or phosphorus limitation in the three studied green algae (Fig. S6). Arachidonic acid (ARA) metabolism, also part of fatty acid metabolism, was enriched in response to all three resource limitations (Fig. 3) and also in response to the combination of warming and nitrogen or phosphorus limitation (Fig. S6). This pathway was only enriched in the three green algae of the study. We detected ARA enrichments in Chlamydomonas and Scenedesmus, while other metabolites that are also part of the ARA pathway such as prostaglandins and leukotriene were enriched in all three green algae. These metabolites are polyunsaturated fatty acids, a class of molecules that has been shown to play an important role in temperature responses in multiple phytoplankton species by regulating membrane fluidity, also known as homeoviscous adaptation (Holm et al., 2022). Nutrient limitation also impacts these metabolic pathways. Previous studies in Porphyridium cruentum, a red algae, observed accumulation of ARA in response to nitrogen limitation (T. Li et al., 2024), and prostaglandins have been found in the marine diatom Thalassiosira rotula (Di Dato et al., 2020) at the end of the exponential phase, which corresponds with the onset of nutrient limitation. Taken together, our results corroborate the crucial role of lipid metabolism, especially fatty acids, to stress responses in multiple phytoplankton groups. Untargeted metabolomics methods, such as the one used in this study, generate numerous hypotheses that should be further tested and validated using targeted approaches to precisely detect and quantify metabolites of interest. Despite measuring a large number of metabolites, many of the molecules we measured here could not be reliably annotated and their functions remain unknown. Targeted methods could help validate some of the patterns observed here and also provide a more detailed understanding of the pathways involved in these metabolic responses. Additionally, our study focused solely on polar metabolites, and future research incorporating both polar and non-polar metabolites could provide a more detailed characterization of the complex metabolic changes occurring in phytoplankton (Llewellyn et al., 2015). Under growth limitation, metabolites related to pathways directly linked to the limiting resource might be so scarce that they are immediately used by the cells or are present at extremely low abundances, potentially hindering their detection by metabolomics methods. Integrating complementary data types, such as transcriptomics, proteomics and macromolecular pools (for example, Fisher et al., 2023; Park et al., 2015), would enable a more complete characterization of the cellular rearrangements required to cope with warming and resource limitation in phytoplankton. While phytoplankton cell models often focus on the important role of photosynthesis and respiration in driving growth, our findings highlight metabolic shifts across multiple key cellular components, including energy storage and lipid metabolism, which might have significant consequences for population dynamics and primary productivity. These insights could improve the parametrization of such cell models, which could then be incorporated into ecosystem models to better predict primary production and other key ecosystem services in aquatic ecosystems under climate change (Hutchins & Tagliabue, 2024).

Our study demonstrates that light, nitrogen, and phosphorus limitation elicit distinct metabolic responses in six freshwater phytoplankton species, with nitrogen and phosphorus limitations showing greater metabolic similarity compared to light limitation. While temperature had a secondary effect on metabolism for most species, specific functional groups like diatoms and cyanobacteria seemed to be more influenced by this driver. Lipid metabolism, particularly the biosynthesis of fatty acids and arachidonic acid, plays a central role in the phytoplankton response to all resource limitations, likely reflecting adaptive mechanisms for energy storage, membrane fluidity, and stress management. These results underscore the importance of resource availability in shaping the metabolic profile of phytoplankton, with implications for understanding their ecological functioning and responses to climate change. Integrating metabolomics with other omics approaches will help to further elucidate the complex physiological responses of phytoplankton under environmental change, providing valuable insights for ecosystem modeling and predicting primary production in future aquatic ecosystems. Acknowledgments We thank Marta Reyes, Sereina Gut, Shannon Eckhardt, Lisa Marchand, Alexa Agatiello, and Max Hofland for technical assistance during the experimental work. We thank Sebastian Salzmann and Severin Ammann for technical assistance with the metabolomics methods. We acknowledge access to the Piz Daint supercomputer in the Swiss National Supercomputing Center. This study was funded with a Seed grant to Marc J.-F. Suter and Anita Narwani from Eawag, an SNF Project Grant (310030_197812) to Anita Narwani, and an SNF Flexibility Grant (310030_197812/2) to Vanessa Weber de Melo.

Aguilera-Sáez, L. M., Abreu, A. C., Camacho-Rodríguez, J., González-López, C. V., Del Carmen Cerón-García, M., & Fernández, I. (2019). NMR metabolomics as an effective tool to unravel the effect of light intensity and temperature on the composition of the marine microalgae Isochrysis galbana. Journal of Agricultural and Food Chemistry, 67 (14), 3879–3889.Alipanah, L., Rohloff, J., Winge, P., Bones, A. M., & Brembu, T. (2015). Whole-cell response to nitrogen deprivation in the diatom Phaeodactylum tricornutum. Journal of Experimental Botany, 66 (20), 6281–6296.Allen, A. P., & Gillooly, J. F. (2009). Towards an integration of ecological stoichiometry and the metabolic theory of ecology to better understand nutrient cycling. Ecology Letters, 12 (5), 369–384.Armin, G., & Inomura, K. (2021). Modeled temperature dependencies of macromolecular allocation and elemental stoichiometry in phytoplankton. Computational and Structural Biotechnology Journal, 19, 5421–5427.Armin, G., Kim, J., & Inomura, K. (2023). Saturating growth rate against phosphorus concentration explained by macromolecular allocation. MSystems, 8 (5), e0061123.Bernacchi, C. J., Singsaas, E. L., Pimentel, C., Portis, A. R., Jr, & Long, S. P. (2001). Improved temperature response functions for models of Rubisco-limited photosynthesis. Plant, Cell & Environment, 24 (2), 253–259.Bölling, C., & Fiehn, O. (2005). Metabolite profiling of Chlamydomonas reinhardtii under nutrient deprivation. Plant Physiology, 139 (4), 1995–2005.Boyd, P. W., Dillingham, P. W., McGraw, C. M., Armstrong, E. A., Cornwall, C. E., Feng, Y.-Y., Hurd, C. L., Gault-Ringold, M., Roleda, M. Y., Timmins-Schiffman, E., & Nunn, B. L. (2015). Physiological responses of a Southern Ocean diatom to complex future ocean conditions. Nature Climate Change, 6 (2), 207–213.Brembu, T., Mühlroth, A., Alipanah, L., & Bones, A. M. (2017). The effects of phosphorus limitation on carbon metabolism in diatoms. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 372 (1728). https://doi.org/10.1098/rstb.2016.0406Brown, J. H., Gillooly, J. F., Allen, A. P., Savage, V. M., & West, G. B. (2004). Toward a metabolic theory of ecology. Ecology, 85 (7), 1771–1789.Carey, C. C., Ibelings, B. W., Hoffmann, E. P., Hamilton, D. P., & Brookes, J. D. (2012). Eco-physiological adaptations that favour freshwater cyanobacteria in a changing climate. Water Research, 46 (5), 1394–1407.Chen, Z., Yang, M.-K., Li, C.-Y., Wang, Y., Zhang, J., Wang, D.-B., Zhang, X.-E., & Ge, F. (2014). Phosphoproteomic analysis provides novel insights into stress responses in Phaeodactylum tricornutum, a model diatom. Journal of Proteome Research, 13 (5), 2511–2523.Cross, W. F., Hood, J. M., Benstead, J. P., Huryn, A. D., & Nelson, D. (2015). Interactions between temperature and nutrients across levels of ecological organization. Global Change Biology, 21 (3), 1025–1040.Dedman, C. J., Barton, S., Fournier, M., & Rickaby, R. E. M. (2023). The cellular response to ocean warming in Emiliania huxleyi . Frontiers in Microbiology, 14, 1177349.Di Dato, V., Barbarinaldi, R., Amato, A., Di Costanzo, F., Fontanarosa, C., Perna, A., Amoresano, A., Esposito, F., Cutignano, A., Ianora, A., & Romano, G. (2020). Variation in prostaglandin metabolism during growth of the diatom Thalassiosira rotula. Scientific Reports, 10 (1), 5374.Fernández-González, C., Pérez-Lorenzo, M., Pratt, N., Moore, C. M., Bibby, T. S., & Marañón, E. (2020). Effects of Temperature and Nutrient Supply on Resource Allocation, Photosynthetic Strategy, and Metabolic Rates of Synechococcus sp. Journal of Phycology, 56 (3), 818–829.Fisher, N. L., & Halsey, K. H. (2016). Mechanisms that increase the growth efficiency of diatoms in low light. Photosynthesis Research, 129 (2), 183–197.Fisher, N. L., Halsey, K. H., Suggett, D. J., Pombrol, M., Ralph, P. J., Lutz, A., Sogin, E. M., Raina, J.-B., & Matthews, J. L. (2023). Light-dependent metabolic shifts in the model diatom Thalassiosira pseudonana. Algal Research, 74 (103172), 103172.Gillooly, J. F., Brown, J. H., West, G. B., Savage, V. M., & Charnov, E. L. (2001). Effects of size and temperature on metabolic rate. Science, 293 (5538), 2248–2251.Grossman, A. R., & Aksoy, M. (2015). Algae in a phosphorus-limited landscape. In Annual Plant Reviews Volume 48 (pp. 337–374). John Wiley & Sons, Inc.Heal, K. R., Durham, B. P., Boysen, A. K., Carlson, L. T., Qin, W., Ribalet, F., White, A. E., Bundy, R. M., Armbrust, E. V., & Ingalls, A. E. (2021). Marine community metabolomes carry fingerprints of phytoplankton community composition. MSystems, 6 (3). https://doi.org/10.1128/mSystems.01334-20Hessen, D. O., Hafslund, O. T., Andersen, T., Broch, C., Shala, N. K., & Wojewodzic, M. W. (2017). Changes in Stoichiometry, Cellular RNA, and Alkaline Phosphatase Activity of Chlamydomonas in Response to Temperature and Nutrients. Frontiers in Microbiology, 8, 18.Holm, H. C., Fredricks, H. F., Bent, S. M., Lowenstein, D. P., Ossolinski, J. E., Becker, K. W., Johnson, W. M., Schrage, K., & Van Mooy, B. A. S. (2022). Global ocean lipidomes show a universal relationship between temperature and lipid unsaturation. Science (New York, N.Y.), 376 (6600), 1487–1491.Hong, T., Huang, N., Mo, J., Chen, Y., Li, T., & Du, H. (2023). Transcriptomic and physiological responses of a model diatom ( Phaeodactylum tricornutum ) to heat shock and heat selection. Ecological Indicators, 153 (110420), 110420.Hu, Q., Sommerfeld, M., Jarvis, E., Ghirardi, M., Posewitz, M., Seibert, M., & Darzins, A. (2008). Microalgal triacylglycerols as feedstocks for biofuel production: perspectives and advances. The Plant Journal: For Cell and Molecular Biology, 54 (4), 621–639.Hutchins, D. A., & Tagliabue, A. (2024). Feedbacks between phytoplankton and nutrient cycles in a warming ocean. Nature Geoscience, 17 (6), 495–502.Kilham, S. S., Kreeger, D. A., Lynn, S. G., Goulden, C. E., & Herrera, L. (1998). COMBO: a defined freshwater culture medium for algae and zooplankton. Hydrobiologia, 377 (1), 147–159.Kochhar, S. L., & Gujral, S. K. (2020). Concepts of Metabolism. In Plant Physiology: Theory and Applications (pp. 153–167). Cambridge University Press.Li, S., Park, Y., Duraisingham, S., Strobel, F. H., Khan, N., Soltow, Q. A., Jones, D. P., & Pulendran, B. (2013). Predicting network activity from high throughput metabolomics. PLoS Computational Biology, 9 (7), e1003123.Li, T., Li, C., Wang, W., Wu, H., Wu, H., Xu, J., & Xiang, W. (2024). Reconstruction of long-chain polyunsaturated acid synthesis pathways in marine red microalga Porphyridium cruentum using lipidomics and transcriptomics. Marine Drugs, 22 (2), 82.Liefer, J. D., Garg, A., Fyfe, M. H., Irwin, A. J., Benner, I., Brown, C. M., Follows, M. J., Omta, A. W., & Finkel, Z. V. (2019). The macromolecular basis of phytoplankton C:N:P under nitrogen starvation. Frontiers in Microbiology, 10, 763.Lin, Q., Liang, J.-R., Huang, Q.-Q., Luo, C.-S., Anderson, D. M., Bowler, C., Chen, C.-P., Li, X.-S., & Gao, Y.-H. (2017). Differential cellular responses associated with oxidative stress and cell fate decision under nitrate and phosphate limitations in Thalassiosira pseudonana: Comparative proteomics. PloS One, 12 (9), e0184849.Lin, S., Litaker, R. W., & Sunda, W. G. (2016). Phosphorus physiological ecology and molecular mechanisms in marine phytoplankton. Journal of Phycology, 52 (1), 10–36.Llewellyn, C. A., Sommer, U., Dupont, C. L., Allen, A. E., & Viant, M. R. (2015). Using community metabolomics as a new approach to discriminate marine microbial particulate organic matter in the western English Channel. Progress in Oceanography, 137, 421–433.Lloyd, G. R., Jankevics, A., & Weber, R. J. M. (2021). struct: an R/Bioconductor-based framework for standardized metabolomics data analysis and beyond. Bioinformatics (Oxford, England), 36 (22–23), 5551–5552.Martin, R. M., Denney, M. K., Pound, H. L., Chaffin, J. D., Bullerjahn, G. S., McKay, R. M. L., Zastepa, A., Jones, K. A., Castro, H. F., Campagna, S. R., & Wilhelm, S. W. (2023). Sulfolipid substitution ratios of Microcystis aeruginosa and planktonic communities as an indicator of phosphorus limitation in Lake Erie. Limnology and Oceanography, 68 (5), 1117–1131.Mohanty, B., Majedi, S. M., Pavagadhi, S., Te, S. H., Boo, C. Y., Gin, K. Y.-H., & Swarup, S. (2022). Effects of light and temperature on the metabolic profiling of two habitat-dependent bloom-forming Cyanobacteria. Metabolites, 12 (5), 406.Naselli-Flores, L., & Padisák, J. (2023). Ecosystem services provided by marine and freshwater phytoplankton. Hydrobiologia, 850 (12–13), 2691–2706.Oksanen, J., Simpson, G. L., Blanchet, F. G., Kindt, R., Legendre, P., Minchin, P. R., O’Hara, R. B., Solymos, P., Stevens, M. H. H., Szoecs, E., Wagner, H., Barbour, M., Bedward, M., Bolker, B., Borcard, D., Carvalho, G., Chirico, M., De Caceres, M., Durand, S., … Weedon, J. (2024). vegan: Community Ecology Package (Version version 2.6-6.1). https://CRAN.R-project.org/package=veganPang, Z., Chong, J., Li, S., & Xia, J. (2020). MetaboAnalystR 3.0: Toward an optimized workflow for global metabolomics. Metabolites, 10 (5), 186.Park, J.-J., Wang, H., Gargouri, M., Deshpande, R. R., Skepper, J. N., Holguin, F. O., Juergens, M. T., Shachar-Hill, Y., Hicks, L. M., & Gang, D. R. (2015). The response of Chlamydomonas reinhardtii to nitrogen deprivation: a systems biology analysis. The Plant Journal: For Cell and Molecular Biology, 81 (4), 611–624.Ribbenstedt, A., Ziarrusta, H., & Benskin, J. P. (2018). Development, characterization and comparisons of targeted and non-targeted metabolomics methods. PloS One, 13 (11), e0207082.Rohart, F., Gautier, B., Singh, A., & Lê Cao, K.-A. (2017). mixOmics: An R package for ’omics feature selection and multiple data integration. PLoS Computational Biology, 13 (11), e1005752.Rokitta, S. D., von Dassow, P., Rost, B., & John, U. (2016). P- and N-depletion trigger similar cellular responses to promote senescence in eukaryotic phytoplankton. Frontiers in Marine Science, 3 . https://doi.org/10.3389/fmars.2016.00109Sallée, J.-B., Pellichero, V., Akhoudas, C., Pauthenet, E., Vignes, L., Schmidtko, S., Garabato, A. N., Sutherland, P., & Kuusela, M. (2021). Summertime increases in upper-ocean stratification and mixed-layer depth. Nature, 591 (7851), 592–598.Sato, N., Hagio, M., Wada, H., & Tsuzuki, M. (2000). Environmental effects on acidic lipids of thylakoid membranes. Biochemical Society Transactions, 28 (6), 912–914.Schrimpe-Rutledge, A. C., Codreanu, S. G., Sherrod, S. D., & McLean, J. A. (2016). Untargeted metabolomics strategies-challenges and emerging directions. Journal of the American Society for Mass Spectrometry, 27 (12), 1897–1905.Southam, A. D., Weber, R. J. M., Engel, J., Jones, M. R., & Viant, M. R. (2016). A complete workflow for high-resolution spectral-stitching nanoelectrospray direct-infusion mass-spectrometry-based metabolomics and lipidomics. Nature Protocols, 12 (2), 310–328.Sun, J., & Xia, Y. (2024). Pretreating and normalizing metabolomics data for statistical analysis. Genes & Diseases, 11 (3), 100979.Thomas, M. K., Aranguren-Gassis, M., Kremer, C. T., Gould, M. R., Anderson, K., Klausmeier, C. A., & Litchman, E. (2017). Temperature-nutrient interactions exacerbate sensitivity to warming in phytoplankton. Global Change Biology, 23 (8), 3269–3280.Trentin, R., Moschin, E., Custódio, L., & Moro, I. (2024). Temperature effects on growth, metabolome, lipidic profile and photosynthetic pigment content of Microglena antarctica (chlorophyceae): A comprehensive analysis. Algal Research, 79 (103461), 103461.Van Mooy, B. A. S., Rocap, G., Fredricks, H. F., Evans, C. T., & Devol, A. H. (2006). Sulfolipids dramatically decrease phosphorus demand by picocyanobacteria in oligotrophic marine environments. Proceedings of the National Academy of Sciences of the United States of America, 103 (23), 8607–8612.Walker, T. W. N., Alexander, J. M., Allard, P.-M., Baines, O., Baldy, V., Bardgett, R. D., Capdevila, P., Coley, P. D., David, B., Defossez, E., Endara, M.-J., Ernst, M., Fernandez, C., Forrister, D., Gargallo-Garriga, A., Jassey, V. E. J., Marr, S., Neumann, S., Pellissier, L., … Salguero-Gómez, R. (2022). Functional Traits 2.0: The power of the metabolome for ecology. The Journal of Ecology, 110 (1), 4–20.Weber de Melo, V., Suter, M. J.-F., & Narwani, A. (2025). Light and nutrients modulate the temperature‐sensitivity of growth in phytoplankton. Limnology and Oceanography Letters, 10 (1), 91–100.Weber, R. J. M., & Zhou, J. (2020). DIMSpy: Python package for processing direct-infusion mass spectrometry-based metabolomics and lipidomics data . Zenodo. https://doi.org/10.5281/ZENODO.3764169Woolway, R. I., Sharma, S., & Smol, J. P. (2022). Lakes in hot water: The impacts of a changing climate on aquatic ecosystems. Bioscience, 72 (11), 1050–1061.Wördenweber, R., Rokitta, S. D., Heidenreich, E., Corona, K., Kirschhöfer, F., Fahl, K., Klocke, J. L., Kottke, T., Brenner-Weiß, G., Rost, B., Mussgnug, J. H., & Kruse, O. (2018). Phosphorus and nitrogen starvation reveal life-cycle specific responses in the metabolome of Emiliania huxleyi (Haptophyta). Limnology and Oceanography, 63 (1), 203–226.Yang, M.-K., Qiao, Z.-X., Zhang, W.-Y., Xiong, Q., Zhang, J., Li, T., Ge, F., & Zhao, J.-D. (2013). Global phosphoproteomic analysis reveals diverse functions of serine/threonine/tyrosine phosphorylation in the model cyanobacterium Synechococcus sp. strain PCC 7002. Journal of Proteome Research, 12 (4), 1909–1923.Yvon-Durocher, G., Jones, J. I., Trimmer, M., Woodward, G., & Montoya, J. M. (2010). Warming alters the metabolic balance of ecosystems. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 365 (1549), 2117–2126.Zhang, C.-C., Jang, J., Sakr, S., & Wang, L. (2005). Protein phosphorylation on Ser, Thr and Tyr residues in cyanobacteria. Journal of Molecular Microbiology and Biotechnology, 9 (3–4), 154–166.Zhang, H., Gu, B., Zhou, Y., Ma, X., Liu, T., Xu, H., Xie, Z., Liu, K., Wang, D., & Xia, X. (2022). Multi-omics profiling reveals resource allocation and acclimation strategies to temperature changes in a marine dinoflagellate. Applied and Environmental Microbiology, 88 (17), e0121322. Data Accessibility All the metabolomics data and related metadata will be available at ERIC open data repository. During revision, all the data and code to reproduce the analyses can be anonymously accessed here: https://drive.switch.ch/index.php/s/cLIj4CKkpgPRSjT Benefit-Sharing Section Benefits Generated: Benefits from this research accrue from the sharing of our data and results on public databases as described above. Author Contributions AN and MJFS formulated the research question and obtained funding. VWM and AN designed the experiments. VWM and PT conducted the laboratory experiments. VWM performed the data analyses, with input from AN and PT. VWM and AN wrote the paper with feedback from all authors. Tables and Figures Figure 1. Sparse partial least squares discriminant analysis (sPLS-DA) with full treatment information across six phytoplankton species, Chlamydomonas reinhardtii (A), Scenedesmus acuminatus (B), Pediastrum boryanum (C), Synechococcus sp. (D), Asterionella formosa (E) and Cyclotella meneghiniana (F). Circle colors indicate resource levels, while circle size represents temperature levels. Axis labels show the percentage of variation explained by each latent variable (LV). sPLS-DA used the full treatment information, i.e. resource and temperature level, for metabolite selection and sample classification. Axis labels show the percentage of variation explained by each latent variable (LV). The number following each species name indicates the metabolites selected by the analyses. Figure 2. Venn diagrams showing the overlap of the metabolites selected by sPLS-DA analyses for six phytoplankton species, Chlamydomonas reinhardtii (A), Scenedesmus acuminatus (B), Pediastrum boryanum (C), Synechococcus sp. (D), Asterionella formosa (E) and Cyclotella meneghiniana (F). sPLS-DA was performed using three different inputs: temperature information only, resource information only, or the full treatment information, including both resource and temperature. The colors of the circles represent the number of metabolites in each category and in the overlaps between the three groups. Figure 3. Functional analysis of metabolomics data in response to resource limitation. (A) KEGG pathway enrichment analysis comparing the control condition (abundant light, nitrogen, and phosphorus) to each resource limitation treatment. Samples were grown at four different temperatures for each resource level, but temperatures were pooled within each resource condition for this analysis. Colors represent different KEGG pathway classes, and the y-axis labels indicate the species in which each pathway was enriched. (B) Upset plot showing the number of enriched pathways shared across the three resource limitation treatments. Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License.

Authors Metrics & Citations Metrics Article Usage 329views 158downloads Citations Download citation Vanessa Weber de Melo, Patrick Thomas, Marc Suter, et al. Metabolic responses of phytoplankton to combined global change drivers: temperature and resource availability. Authorea. 06 March 2025. DOI: https://doi.org/10.22541/au.174126975.57115189/v1 DOI: https://doi.org/10.22541/au.174126975.57115189/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.