Individual whole plants of Ecklonia radiata and Sargassum linearifolium (hereafter ‘Ecklonia’ and ‘Sargassum’), were collected by SCUBA in Sydney, Australia (34°03’59.0”S 151°09’44.3”E) under NSW DPI permits (P01/0059(A)-3.0 and OUT17/7525). Carefully keeping the holdfasts intact, 150 whole juvenile plants (< 30cm total length) were collected per species from a shallow subtidal rocky reef between 3 to 6m depth. Within 4h of collection, the plants were transported (by air) to the National Marine Science Centre in Coffs Harbour, Australia and placed in 1000 L mesocosms with flowing seawater, weighed down by the holdfast. Following the methods described by Straub et al. (2021) plants were acclimated for 6d before the application of experimental temperature treatments. Four ocean temperature treatments were applied over 48d. The four treatments were i) ambient (22°C), ii) warming (24°C), iii) heatwave non-variable (HWnon) – remaining at 22°C for 20d before moving to 27°C for 20d and back to 22°C for 8d, and iv) heatwave variable (HWvar) – exposed to rapid fluctuations between 22 and 24°C for 8d (temperature variability), remaining at 22°C for 6d moving to a heatwave at 27°C for 20d and back to 22°C for 8d ( Figure 2 ). The magnitude of cumulative intensity was similar among treatments throughout the experiment (warming = 96 degree-days, HWnon = 85 degree-days, and HWvar = 97 degree-days cumulative intensity above the ambient treatment; Hobday et al., 2016).

Twenty aerated mesocosms (80cm diameter × 45cm high, 230L water volume) received a continuous inflow (flowrate ≈ 2L min^-1) of filtered seawater (50μm) from Charlesworth Bay (30°16’3.78”S, 153° 8’25.60”E). Three header tanks with heater/chillers units (Aquahort Ltd.) controlled experimental temperatures at 22, 24, and 27°C, respectively. At the start of the experimental period, five tanks were assigned per treatment, and seven individual Ecklonia and Sargassum plants were placed within each tank (see Supplementary Material ). Seaweeds acclimated at 22°C were randomly allocated to ambient, HWnon, and HWvar treatments (which start at 22°C) and seaweeds acclimated at 24°C were randomly allocated to a warming treatment tank (n = 14 plants/tank).

Tissue samples were collected on day 20 (after temperature variation, before marine heatwave), day 33 (12d into the marine heatwave), and day 48 (one-week recovery post-heatwave). At each sampling, a 9 mm hole punch was used to collect samples from the meristem of Ecklonia and tweezers were used to remove approximately 20 fronds of Sargassum. Each time, three independent plants, per species (n = 3) were sampled (see Supplementary Material ). Samples were collected from three of the five independent tanks to generate experimental replicates and no plant was sampled twice, allowing the inclusion of ‘exposure time’ as an independent experimental factor. Following best practice, tissue samples were immediately snap-frozen in liquid nitrogen and frozen at – 80°C until analysis (Kumari, 2017).

Lipids were extracted from seaweed samples using a scaled-down approach modified from Folch et al. (1957) standard procedures, which accounted for the small sample weight obtained from the Ecklonia and Sargassum (0.07 ± 0.01 and 0.13 ± 0.07 g wet weight, respectively). Using analytical grade solvents (Sigma-Aldrich), samples were soaked in 2:1 chloroform:methanol, then polarized using 0.9% NaCl solution to obtain the lipophilic (chloroform) layer. The lipophilic layer was, transferred to a pre-weighed vial and dried using high-purity nitrogen gas to obtain the lipid extract weight. Lipid extracts were resuspended at 1mg mL^-1 in dichloromethane:methanol (50/50 v/v). Total lipid concentration was expressed as a percentage of wet tissue weight for each extracted sample.

Lipid extracts were analyzed using direct infusion mass spectrometry (DIMS) using a high-resolution SCIEX 5600 Triple-TOF (Simons et al., 2012). Samples were run in both positive and negative modes with a flow rate of 200µL min^-1. MS/MS reactions were obtained for peaks in the range of 500–1250 Da. Obtained spectra were analyzed using Lipidview™ 1.3 software (SCIEX) for lipid class and fatty acid composition (Ejsing et al., 2006). Thirty-three lipid classes and 43 fatty acids were detected using both positive and negative detection modes. Because the negative mode detected more lipid classes and fatty acids, the subsequent analysis used negative mode data only to avoid overrepresenting lipids detected in both modes. Lipid classes and fatty acids less than 0.01% were excluded from statistical analyses, and the remaining relative proportion of 13 lipid classes and 19 fatty acids were analyzed separately. The relative compositions (total percent) summed for saturated, monounsaturated and polyunsaturated fatty acids, the ω-3 index and ω-6/ω-3 ratio were also analyzed. The fatty acid double bond isomers cannot be identified using this method of mass spectrometry which limited our ability to distinguish between some fatty acid families (e.g. C20:3n-3/6/9 and C22:5n-3/6). Consequently, the ω-3 index was calculated as the proportion of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) expressed as a percent of total fatty acids. The ω-6/ω-3 PUFA ratio was calculated as arachidonic acid (ARA) divided by the sum of EPA and DHA.

Tissue subsamples were freeze-dried and derivatized for metabolite analysis using a modified, scaled-down approach based on methods described by Mabuchi et al. (2018; adapted from Lisec et al., 2006; Pongsuwan et al., 2007). A mixed solution of 2.5:1 methanol/ultrapure water/chloroform was used to extract hydrophilic primary metabolites in 15.73 (± 2.00) and 13.19 (± 1.30) mg of freeze-dried and macerated Ecklonia and Sargassum, respectively, with 0.2 µg mg^-1 sample (dry weight) of ribitol (Sigma-Aldrich) added as an internal standard. The mixture was stirred for 5 min and centrifuged for a further 5 min (16,000xg, 4°C) and freeze-dried (~16h). Oxime formation was carried out by adding methoxyamine hydrochloride solubilized with pyridine (20mg/mL, 50µL) to the freeze-dried sample (35°C for 80 min). A trimethylsilylation reaction at 40°C for 40 min was induced using N-methyl-N-(trimethylsilyl) trifluoroacetamide (100µL).

Derivatized metabolite samples were analyzed by gas chromatography-mass spectrometry (GC-MS) using an Agilent 6890A Gas Chromatograph fitted with an Agilent 5973N Mass Selective Detector. The samples were injected with a split ratio of 50:1 using helium carrier gas at 32cm sec^–1 constant flow and an inlet temperature of 280°C. Separating volatile compounds by using a Phenomenex ZB-5 column (60m x 0.25mm x 250µm) starting at 60°C with a 3 min hold, then ramped at 8°C min^–1 to 330°C (44.75min) then held for 8min. The temperature of the mass spectrometer transfer line was 280°C. The mass spectra were recorded at 70 eV ionisation voltage over the mass range of 45 to 600amu. Peaks were identified by comparison of spectra with MS databases (WILEY 275 and NIST14), along with retention times and elution order. In instances where the compounds could not be positively identified, the compounds were assigned to a metabolite class when possible, based on examination of similarities in the major fragment ions corresponding to functional groups (see Supplementary Material ). The relative abundance of each metabolite was quantified by TIC peak area integration from the GC and expressed as a percentage of the total in each run.

Overall, the ocean warming and marine heatwave treatments showed minimal impacts on the nutrient content of Ecklonia and Sargassum. However, univariate analysis of the lipid classes showed that the interaction between species and treatment were statistically significant for 1,2-Dimyristoyl-sn-glycero-3-phosphorylethanolamine (DMPE), cardiolipin (CL) and phosphatidic acid (PA; Table 1 ). Post hoc pairwise comparisons showed that the interactive effect of species and treatment on DMPE was driven by significantly lower levels of DMPE in Sargassum from the ambient treatment relative to the HWnon treatment (p < 0.05). This difference was not observed for Ecklonia. Pair-wise tests did not detect any significant differences in CL composition among treatments for either species (p > 0.05). A significant interaction was observed for PA between species, treatment and exposure time. The post-hoc pairwise comparisons showed that after 33 days (12 days into the heatwaves) phosphatidic acid was significantly lower in Ecklonia from the HWnon temperature treatment compared to the ambient treatment (1.8%, p < 0.05); this trend was not observed for Sargassum or after a period of recovery (at 48 days). The trends in HWnon phosphatidic acid concentration throughout the experiment aligned with the HWnon maximum quantum yield of the Ecklonia meristem (as reported in Straub et al., 2021, Figure 3 ). The concentration of phosphatidic acid was higher under ambient conditions at this sampling time relative to both heatwave treatments.

Ecklonia and Sargassum each showed minimal differences in fatty acids among treatments or among treatments over time ( Table 2 ). However, there was a significant interaction between species and treatment for the saturated fatty acids C15:0 ( Table 2 ). The post-hoc pairwise comparisons showed that the interaction between species and treatment for C15:0 was largely associated with a statistically significant (p < 0.05) increase under the warming treatment compared to the HWnon and HWvar treatments for Ecklonia which was not observed for Sargassum. Although statistically significant, the changes in C15:0 were less than 1% of the total fatty acid composition. A significant interaction between species, treatment and exposure time was detected for total saturated fatty acid (SFA) and the polyunsaturated fatty acids C20:3 and C20:4 ( Table 2 ). However, these three-way interactions were not supported by the detection of any two-way interactions or main effects and no meaningful pairwise differences were detected by the post-hoc tests.

When grouped into broad metabolite classifications, alkaloids showed a significant interaction between species and treatments. The post-hoc pairwise comparisons showed that the interaction between species and treatment was largely associated with a statistically significant (p < 0.05) decrease in alkaloids under the HWvar (p < 0.05) compared to the ambient tanks for Ecklonia which was not observed for Sargassum. No other significant interactions in broad metabolite groups, including amino acids and sugars, were found within the seaweed species in response to treatments or with treatments and exposure time ( Table 3 ).

The total extracted lipids of Ecklonia and Sargassum comprised on average 1.400 ± 0.113 and 1.100 ± 0.114% g^-1 of the seaweeds’ total wet weight respectively, with no significant differences detected between species, experimental treatments or sampling period ( Table 1 ). The percent composition of lipid classes showed that 92.3% (12 of 13) were significantly different between Ecklonia and Sargassum ( Figure 4 ; Table 1 ). The relative concentration of phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI) and monomethylphosphatidyl ethanolamine (MMPE) were significantly higher in Ecklonia compared to Sargassum ( Table 1 ). While PA, phosphatidylserine (PS), CL, diacylglycerol pyrophosphate (DGPP), DMPE, phosphatidylinositol-4-phosphate (PIP) and cytidine diphosphate diacylglycerol (CDPDAG) were significantly higher in Sargassum in comparison to Ecklonia ( Table 1 ). The percent of sulfatide galactosylceramide (SGalCer) was not significantly different between the two seaweed species ( Table 1 ).

Univariate analysis of the fatty acids showed that 83% (20 of 24) were significantly different between Ecklonia and Sargassum ( Table 2 ). Sargassum was 7.4% higher in total saturated fatty acids than Ecklonia ( Figure 4 ; Table 2 ), driven by significantly higher levels of C12:0, C16:0 and C20:0 when compared to Ecklonia ( Table 2 ). Monounsaturated fatty acids were also 3.4% higher in Sargassum compared to Ecklonia ( Figure 4 ; Table 2 ), with C16:1 and C22:1 significantly higher in Sargassum than Ecklonia ( Table 2 ). Polyunsaturated fatty acids were found to be 10.8% higher in Ecklonia when compared to Sargassum ( Figure 4 ; Table 2 ). Omega-3 (C18:3, C20:5), omega-6 (C20:4, C22:4) and C20:3 fatty acids were significantly higher in Ecklonia when compared to Sargassum ( Table 2 ). Compared to Ecklonia, Sargassum was significantly higher in C18:2 (omega-6), C20:2, C22:5 and C22:6 (omega-3; Table 2 ). Importantly, arachidonic acid (ARA, 20:4), a crucial long-chain polyunsaturated fatty acid in the omega-6 pathway, was 6.2% higher in Ecklonia than Sargassum ( Figure 4 ). α–Linolenic (ALA, 18:3), important to the omega-3 pathway ( Figure 1 ), was also 3.9% higher in Ecklonia ( Figure 4 ). Omega-3 docosahexaenoic acid and omega-6 docosapentaenoic acid showed statistically significant changes however, with only ≤ 0.05% difference between the two species ( Figure 4 ). Omega-3 eicosapentaenoic acid (EPA, 20:5) was, however, 1.6% higher in Ecklonia when compared to Sargassum ( Figure 4 ). The ω-3 index was 1.5% higher in Ecklonia which shows the seaweed species has a higher omega-3 content than Sargassum ( Figure 4 ). Sargassum had a 10.8% higher ω-6:ω-3 ratio of 16.8 compared to 6.0 in Ecklonia ( Figure 4 ).

Metabolites were grouped into broad classes to further assess functional differences between species (see Supplementary Material ). We found 91.7% (11 of 12) of metabolite classes significantly differed in content between species (see Supplementary Material ). Ecklonia and Sargassum metabolites consist largely of simple sugars (93.3 and 96.5%, respectively; Figure 5A ), of which Sargassum had significantly higher levels. The remaining metabolite composition of Sargassum comprised mostly of complex sugars (2.5%) and alcohols (0.5%; Figure 5B ). Ecklonia metabolite composition was more diverse containing complex sugar (2.4%), and significantly higher levels of heterocycles (1.3%), aromatic alcohols (1.2%), alcohols (0.9%) and alkaloids (0.5%) than Sargassum ( Figure 5B ; Table 3 ).

Projections of Ecklonia lipid availability off eastern Australia (28.0 – 37.5°S) revealed lipid reductions in warmer, lower latitudes where Ecklonia will range contract and increases within cooler, poleward and deeper regions of the study extent by 2100 where Ecklonia biomass will increase under all climate scenarios ( Figure 6 ). Overall, total net lipid gains of 18.5% (4.9 g.m^-2; Figure 6A ) and 29.4% (7.7 g.m^-2; Figure 6B ) are projected under RCPs 2.6 and 6.0, respectively. In contrast, a 3.5% total net loss in Ecklonia derived lipids are projected under RCP 8.5 (-0.9 g.m^-2; Figure 6C ). As the majority of current Ecklonia biomass and subsequent lipid availability occurs at shallow to intermediate reef depths (i.e. <20 m), projected losses are associated with these depths and areas of the coast (primarily lower latitudes) where seabed temperatures are above ~26°C ( Figure 6 ). Projected increases in lipid availability at high latitudes occur across the entire depth range under RCP 2.6, but are mostly restricted to deeper reefs under RCP 8.5 (>20 m; Figure 6 ).

Despite both ocean warming and MHWs having caused range contractions, local losses and genetic and physiological changes in Ecklonia and Sargassum in Australia (Vergés et al., 2016; Wernberg et al., 2016a; Straub et al., 2019; Coleman et al., 2020a; Gurgel et al., 2020), we observed minimal changes in the lipid and metabolite composition of these species in response to these ocean stressors. Our results were unexpected and contrary to other studies that observed changes in the fatty acid and metabolite composition of seaweeds in response to temperature changes in both field (Schmid et al., 2017; Gaubert et al., 2019) and laboratory settings (Renaud et al., 2002; Gosch et al., 2015). No biologically significant changes were detected in the fatty acids of Ecklonia or Sargassum in response to the warming and marine heatwave profiles. Increases in fatty acid saturation with increasing temperatures have been widely observed in marine algae (Renaud et al., 2002; Barbosa et al., 2020) and other marine plant species (Beca-Carretero et al., 2018). The unsaturation of fatty acids at low temperatures and saturation at high temperatures is an adaptative strategy algae use to maintain cell membrane fluidity and photosynthetic capacity with changing temperatures (Los and Murata, 2004; Schmid et al., 2021). Although this mechanism has commonly been observed in response to seasonal changes in temperature (Schmid et al., 2014; Schmid et al., 2017), this thermal adaptation may also assist seaweeds in coping with predicted ocean warming and MHWs (Britton et al., 2020). This response was not observed in our experiment for either Ecklonia or Sargassum, which could indicate that these species cannot adjust their fatty acid biosynthesis rapidly in response to unpredictable extreme temperature events. However, Britton et al. (2021) suggest that species that can also occupy intertidal or shallower areas such as Ecklonia and Sargassum may be better adapted to temperature fluctuations and potentially do not need to adjust fatty acid composition to maintain optimum membrane fluidity, or that the adjustments are small and hard to identify. Alternatively, our results could indicate a level of resilience, as the individuals sampled at each time point were survivors, which may have been better able to cope with warmer temperatures (e.g. thermally tolerant genotypes, Vranken et al., 2021; see Straub et al., 2021 for mortality throughout the experiment).

The time frame over which seaweeds might display a temperature-driven change in fatty acid composition is poorly understood. Changes in seaweed lipids in response to elevated temperature have been observed after as little as three days in the giant kelp Macrocystis pyrifera (Schmid et al., 2020), and differences in seaweed fatty acids have been observed during and after a period of recovery from MHWs (Britton et al., 2020). However, we do not yet understand whether these temperature-driven changes in physiological performance and nutritional properties are species-specific, adaptations of distinct genetic populations or correlate to natural processes (e.g. growth and reproduction). Twelve days into the simulated non-variable heatwave we observed substantially lower levels of phosphatidic acid in Ecklonia when compared to the ambient temperature treatment. However, following a period of recovery from the heatwave, this difference was no longer detected, suggesting recovery. Levels of phosphatidic acid were initially also relatively low in the warming and variable heatwave treatments but increased over time. Phosphatidic acid is an important component of seaweed chloroplast membranes (Melo et al., 2015), which are important for photosynthesis (Wang and Benning, 2012). During the marine heatwave treatments, Ecklonia photosynthetic quantum yield was significantly lower than the warming and ambient treatments but these differences were no longer observed after the recovery period ( Figure 3 ; Straub et al., 2021). As only surviving plants were sampled at each time point, the performance improvement could be reflective of the responses of the better adapted survivors at 56 days compared to individuals sampled at 33 days. However, similar reductions in photosynthesis have been observed in other brown seaweeds (e.g. Phyllospora comosa) in response to MHWs, and following a period of recovery no lasting negative effects were detected (Britton et al., 2020). Therefore, the reduction in phosphatidic acid during the marine heatwave may be indicative of a stress response and connected to photosynthetic performance, however, further research is required to establish phosphatidic acid as a biomarker for photosynthetic stress responses in seaweeds.

Simulated ocean warming and marine heatwave treatments showed minimal effects on the metabolite content of both seaweed species. In particular, there was no evidence for impacts of temperature on complex and simple sugars which indicates there were no major effects on the potential beneficial properties (e.g. anti-inflammatory), energy content or calorific value of these species (Bayu et al., 2021). For example, fucoidan sugars like fructose have several biological activities including anti-inflammatory and anti-tumour properties which were unaffected by the temperature changes (Deniaud-Bouët et al., 2017). Amino acids function as important building blocks for protein (Maschek and Baker, 2008) and were unaffected by simulated ocean warming and marine heatwave scenarios. This suggests that temperature treatments did not trigger stress-induced protein degradation (Kumar et al., 2016), which would be expected to increase the levels of some amino acids in the tissues (Araújo et al., 2011). Additional studies are required to determine if the stability in metabolite content is a sign of resilience in the species’ or attributed to the sampling of survivors at each time point. While the application of metabolomics to marine system biology is increasing globally, the main obstacle that remains is the accurate annotation and identification of metabolites (Kumar et al., 2016; Lawson et al., 2022). A change was detected in an unidentified metabolite in response to the simulated temperatures, however, without sufficient context on the function of these compounds, the potential impacts on Ecklonia metabolism and nutritional quality could not be interpreted. For the future application of metabolomics techniques to marine seaweeds, access to advanced technology and the establishment of large mass spectral libraries centred on model organisms are needed (Kumar et al., 2016; Lawson et al., 2022).

The nutritional composition between Ecklonia and Sargassum varied considerably. All 68 nutritional properties detected were present in both species, however, 82% differed substantially in content between species. The magnitude of difference between species was larger among lipid classes and fatty acids than metabolites. The lipid content of both Ecklonia and Sargassum was low (1.4 and 1.1%, respectively), with no difference detected between species and were similar to levels previously measured (Verma et al., 2017; Skrzypczyk et al., 2019). Ecklonia and Sargassum had lower total lipid levels than other Australian seaweed species such as Cystophora torulosa and C. polycystidea (8.4 and 5.7%, respectively), but were higher than the commercial species Sargassum fusiforme (0.7%, Skrzypczyk et al., 2019). Macroalgae are typically low in total fatty acids (in comparison to other algae e.g. microalgae) but have a high proportion of valuable polyunsaturated fatty acids (Schmid et al., 2014). This was supported by our results, as Ecklonia and Sargassum fatty acid composition was comprised of 39.3 and 28.6% polyunsaturated fatty acids, respectively. Although both species contained relatively high proportions of polyunsaturated fatty acids, higher levels were found in Ecklonia, while Sargassum contained substantially higher levels of saturated and monounsaturated fatty acids. Concentrations of polyunsaturated fatty acids tend to be higher in species from cold water and saturated fatty acids higher in species from warmer water, which could explain why temperate-tropical adapted species Sargassum displays higher levels of saturation (Choudhary et al., 2021).

The higher levels of polyunsaturated fatty acids in Ecklonia were driven by higher levels of omega-3 fatty acids (alpha-linolenic and docosapentaenoic) and arachidonic acid (an omega-6 fatty acid) when compared to Sargassum. Despite omega-3 fatty acids being primarily known for their benefits to human nutrition, these valuable fatty acids are also important to marine systems as an essential component of the cells of marine plants, animals and bacteria (Valentine and Valentine, 2004; Gladyshev et al., 2018). For example, numerous fish species have high eicosapentaenoic and docosahexaenoic acid content that is believed to support fast continuous swimming through its role in muscle cell metabolism (Gladyshev et al., 2018). Ecklonia displayed a higher omega-3 index than Sargassum and could be a better source of these fatty acids in temperate ecosystems in comparison. By contrast, Sargassum had a higher omega-6 to omega-3 ratio. Due to the opposing effects of omega-6 and omega-3 fatty acids (i.e. pro- and anti-inflammatory), a balanced consumption is recommended in human diets (Simopoulos, 2002). However, research into the importance of omega fatty acids for marine food chains is limited and the importance of this ratio for marine consumers is unclear. The species differences in nutritional composition are important in the context of nutritional seascapes because these two species are projected to change significantly and contrastingly under climate change.

Seaweed-herbivore interactions are complex (Hay and Fenical, 1988; Poore et al., 2012; Wernberg et al., 2013b). Consequently, structural shifts in seaweed availability may not equate to a suitable dietary replacement for consumers. Palatability is a key factor influencing herbivory and is influenced by macronutrient composition, for example, high protein and fat content often lead to higher palatability. Several other palatability factors including secondary metabolites, morphology and physical stress can also influence the herbivory of seaweeds (Cheng et al., 2022). Research has shown abalone species prefer a diet of brown algae over red or green algae which may be driven by high levels of polyunsaturated acids, in particular, eicosapentaenoic and arachidonic fatty acids (Nelson et al., 2002; Guest et al., 2008; Holland et al., 2021). Secondary metabolites are typically active in low concentrations and can impact the palatability for herbivores (Hay, 1996). These compounds were found in higher content in Ecklonia, including aromatic alcohols, heterocycles and alkaloids and could be used in defence against predators and pathogens (Hay, 1996; Engel et al., 2002). Some seaweeds also actively deter herbivory through chemical defences, for example, brown seaweeds contain phlorotannins which can discourage consumption by herbivores (Toth and Pavia, 2007). Simple sugars can affect the sweetness of seaweeds (Patrick et al., 2013), which may influence the consumption preferences of herbivores. These sugars were higher in Sargassum than Ecklonia, however, both species contained high quantities. These palatability factors may impact the uptake of nutrients from seaweeds despite their nutritional quality, providing a novel insight into potential nutrient transfer in marine food webs through metabolomics.

Seaweeds supply coastal ecosystems with nutrients directly through macroalgal detritus (Duggins et al., 1989) and herbivory (Dunton and Schell, 1987) that are transferred to higher trophic levels ( Figure 1 ; Vanderklift and Wernberg, 2010; Jin et al., 2020). Global changes in seaweed distribution and biomass have been driven by ocean warming (Krumhansl et al., 2016; Martínez et al., 2018), localized marine heatwave events (Wernberg et al., 2016a; Straub et al., 2019; Félix-Loaiza et al., 2022), and local anthropogenic stressors (Reed et al., 1994; Coleman et al., 2008; Strain et al., 2014), altering the nutritional seascapes of temperate environments. By 2100, the Australian distribution of Ecklonia is projected to contract by 49-71% under RCPs 2.6 and 6.0, respectively. Similarly, a 41-57% contraction in the distribution of Sargassum is projected under RCPs 2.6 and 6.0, respectively (Martínez et al., 2018). These estimates are based on correlative responses between each species and environmental variables (primarily SST) only and recent work incorporating key ecological interactions into future projections (e.g. urchin herbivory) of kelp biomass loss are more optimistic (Castro et al., 2020; Davis et al., 2021). For example, incorporating ecological interactions with urchins moderated projected net changes in Ecklonia biomass on the south-east coast of Australia to +18.5, +29.4% and -3.5% under RCPs 2.6, 6.0 and 8.5, respectively (Davis et al., 2021).

Projections of change in Ecklonia lipid availability across Australia’s east coast were made by combining our quantified lipid content with predicted alterations in distribution and biomass (Davis et al., 2021). While we found total net gains in Ecklonia lipid availability across south-east Australia under RCPs 2.6 and 6.0 and a small magnitude of total loss under RCP 8.5, the changes in lipid availability are associated with considerable spatial variation due to the shifting distribution of Ecklonia biomass off eastern Australia. For example, in northern areas of the study extent where it is projected that Ecklonia will undergo a 55-276 km range contraction (Davis et al., 2021), a complete loss of Ecklonia derived lipid availability is likely while increases are expected in more poleward regions as biomass is projected to increase due to weakening urchin herbivory. Increasing water temperatures are also projected to drive Ecklonia redistributions from shallow reefs to cooler, deeper reefs in mid-latitude areas (Davis et al., 2021). These changes may have major implications for marine consumers as the availability of nutrients shifts across not only a latitudinal but also a vertical dimension, particularly if consumers do not experience similar shifts in distribution. However, local drivers such as eutrophication may buffer the negative physiological effects of ocean warming in some areas (Fernández et al., 2020). The results also suggest that the redistribution of Ecklonia and Sargassum will considerably alter the availability of nutrients to marine ecosystems in areas where patterns of species co-occurrence are affected and there is a transition from one species to the other. For example, the predicted shift from Ecklonia to Sargassum dominated reefs at lower latitudes (Vergés et al., 2014; Vergés et al., 2019) would signify increasing saturation in available fatty acids and loss of metabolite diversity, while the loss of canopy cover, in general, would lead to an overall loss of valuable essential fatty acids for secondary consumers ( Figure 1 ). Changes at the primary level could scale up through multiple trophic levels (e.g. primary, secondary, tertiary and finally humans) altering the trophic transfer of nutrients throughout entire marine systems (Jin et al., 2020). Combining lipidomic and metabolomic techniques with traditionally measured carbon, nitrogen and phosphate ratios will provide a greater depth to quantifying the nutritional changes in seaweeds (Peters et al., 2005). Moreover, to fully understand potential change in nutritional seascapes under climate change, we still require better understanding of food web dynamics as they relate to consumer and producer climate-redistributions and their interactions.