Strong seasonal patterns of DOC release by a temperate seaweed community: Implications for the coastal ocean carbon cycle

Release of dissolved organic carbon (DOC) by seaweed underpins the microbial food web and is crucial for the coastal ocean carbon cycle. However, we know relatively little of seasonal DOC release patterns in temperate regions of the southern hemisphere. Strong seasonal changes in inorganic nitrogen availability, irradiance, and temperature regulate the growth of seaweeds on temperate reefs and influence DOC release. We seasonally surveyed and sampled seaweed at Coal Point, Tasmania, over 1 year. Dominant species with or without carbon dioxide (CO2) concentrating mechanisms (CCMs) were collected for laboratory experiments to determine seasonal rates of DOC release. During spring and summer, substantial DOC release (10.06–33.54 μmol C · g DW−1 · h−1) was observed for all species, between 3 and 27 times greater than during autumn and winter. Our results suggest that inorganic carbon (Ci) uptake strategy does not regulate DOC release. Seasonal patterns of DOC release were likely a result of photosynthetic overflow during periods of high gross photosynthesis indicated by variations in tissue C:N ratios. For each season, we calculated a reef‐scale net DOC release for seaweed at Coal Point of 7.84–12.9 g C · m−2 · d−1 in spring and summer, which was ~16 times greater than in autumn and winter (0.2–1.0 g C · m−2 · d−1). Phyllospora comosa, which dominated the biomass, contributed the most DOC to the coastal ocean, up to ~14 times more than Ecklonia radiata and the understory assemblage combined. Reef‐scale DOC release was driven by seasonal changes in seaweed physiology rather than seaweed biomass.


INTRODUCTION
Seaweeds (marine macroalgae) are key habitat-forming organisms on temperate rocky reefs, providing essential inorganic and organic nutrient cycling services, as well as a source of food to numerous taxonomic groups (Graham, 2004;Roleda & Hurd, 2019;Steneck et al., 2002). Most of the seaweed primary production (~43%, see Krause-Jensen & Duarte, 2016 and references therein) enters the microbial food chain through the detrital pathway via the release of particulate-and dissolved organic carbon (POC and DOC; Abdullah & Fredriksen, 2004;Duarte & Cebrián, 1996;Duggins et al., 1989). Through photosynthesis, carbon dioxide (CO 2 ) not only is converted to photosynthate for use in seaweed growth or cellular maintenance but is also released by seaweed into the surrounding seawater (as DOC) through multiple passive and active mechanisms (Fogg et al., 1965;Newell et al., 1980;Paine et al., 2021;Sieburth, 1969). These mechanisms include not only the leakage of small molecules through the cell membrane, tissue breakage, and decay but also active exudation to maintain intracellular stoichiometric ratios (Gordillo et al., 2001;Moebus et al., 1974;Pregnall, 1983;Weigel & Pfister, 2021). These various mechanisms of DOC release can operate simultaneously, and release rates are influenced by environmental drivers such as nutrient availability, irradiance, and water temperature (Gordillo et al., 2001;Hall et al., 2022;Iñiguez et al., 2016;Mueller et al., 2016;Paine et al., 2021;Wada et al., 2007).
In cool temperate regions, strong seasonal fluctuations of inorganic nitrogen availability (primarily NO 3 − ), irradiance, and temperature modulate the annual growth and net primary production (NPP) of seaweeds (Chapman & Craigie, 1977;Davison, 1991;Gagné et al., 1982;Phillips & Hurd, 2003;Schaffelke & Luning, 1994). As a consequence of the closely linked seaweed metabolisms of nitrogen and carbon, seasonal changes in environmental drivers are also thought to regulate the magnitude of DOC release (Paine et al., 2021;Weigel & Pfister, 2021). This is likely due to the "overflow hypothesis," originally described for phytoplankton that posits DOC is released when photosynthesis fixes carbon at a rate that is faster than that required for growth (Fogg, 1984;Thornton, 2014). This DOC release mechanism likely occurs due to either (1) nitrogen limitation or (2) nitrogen sufficiency. Nitrogen limitation (1) suggests that when environmental conditions are unfavorable for algal growth (i.e., low NO 3 − availability), overflow of DOC may be an energy dissipation strategy to ensure that RUBISCO is able to efficiently respond and assimilate CO 2 when suitable environmental conditions become available (Abdullah & Fredriksen, 2004;Nagata, 2000;Wada et al., 2007;Wood & Van Valen, 1990). Conversely, nitrogen sufficiency (2) can prevent the "switching off" of photosynthesis in algae causing DOC to be actively released as the carbon storage capacity of the cell is exceeded (Thornton, 2014). This DOC overflow mechanism can be identified by DOC release during periods of high algal productivity and exponential growth, i.e., spring (Søndergaard et al., 2000). Nitrogen-limited and/ or -sufficient DOC release mechanisms are dependent on gross photosynthesis (defined as all photosynthetic carbon fixation, regardless of whether photosynthate is used for growth, used for cell maintenance, or released as DOC; Falkowski & Raven, 2007), which is regulated by environmental variables such as temperature and irradiance Paine et al., 2021).
Photosynthesis requires the assimilation of dissolved inorganic carbon (C i ), which is available in seawater for seaweed in two major forms, carbon dioxide (CO 2 ) and bicarbonate (HCO 3 − ). Carbon dioxide is taken up across the cellular plasma membrane via passive diffusion requiring no energy; however, HCO 3 − uptake requires active transport via the use of energy-costly CO 2 concentrating mechanisms (CCMs; Beardall & Giordano, 2002;Giordano et al., 2005;Raven et al., 2005Raven et al., , 2014Raven & De Michelis, 1980;Raven et al., 2008). CCMs in their various forms take up and convert HCO 3 − to CO 2 for assimilation by RUBISCO (Fernández et al., 2014;Giordano et al., 2005;Raven, 1991). Approximately 35% of all studied red seaweeds do not possess CCMs and rely solely on CO 2 (termed non-CCM) for use in photosynthesis (Raven et al., 2002. Non-CCM seaweeds can be identified using isotopic analysis; δ 13 C isotopes below −30‰ indicate the lack of CCM and sole passive use of CO 2 (Raven et al., 2002). In Tasmania, Australia, the seaweed flora has a globally unique assemblage with up to 90% of the red seaweeds on some reefs consisting of non-CCM species (Cornwall et al., 2015). These non-CCM seaweeds are increasingly prevalent at depth, living in low-light environments, suggesting a physiological mechanism for "preserving energy" at low-light levels (Cornwall et al., 2015;Raven et al., 2014;Raven & Hurd, 2012). The seaweed communities in Tasmania, therefore, provide a unique opportunity to study the physiological mechanisms of CO 2 use and how they may or may not underpin rates of DOC release in relation to seasonal environmental drivers.
Our study site was Coal Point, a cool temperate reef on the east coast of Bruny Island, Tasmania, Australia ( Figure 1). We hypothesized that (1) DOC release rates by all seaweed species would be higher in spring and summer than in autumn and winter and (2) non-CCM seaweeds would have reduced DOC release rates compared to CCM seaweeds as they are predicted to be C i limited under current ocean CO 2 concentrations. We seasonally measured the rates of DOC release for dominant seaweed species and the biomass of seaweed at Coal Point to elucidate the seasonal reef-scale drivers of DOC production. Combining DOC release rates measured in the laboratory (μmol C g · DW −1 · h −1 ) with biomass of seaweed at the site (g DW · m −2 ), we estimated the net release of DOC from the temperate rocky reef (Gg C · m −2 ), along with in situ seawater DOC concentrations (μmol · L −1 ) at Coal Point.

Field biomass surveys, ambient levels of DOC, and seaweed collection for laboratory experiments
We surveyed the seaweed assemblage of Coal Point, Bruny Island (43°19.848' S, 147°19.62′ E; Figure 1) using SCUBA over four seasons across 2020 and 2021. Surveys were in spring 2020 (September), summer 2021 (January), autumn 2021 (April), and winter 2021 (June). In each season, three transect lines (10 m long) were randomly laid out along the bottom of the rocky reef between 8 and 10 m depth, and six quadrats (0.25 m 2 ) were randomly positioned along each transect line.
The dominant, habitat-forming seaweeds at this depth were Ecklonia radiata (phylum Ochrophyta, order Laminariales) and Phyllospora comosa (phylum Ochrophyta, order Fucales), and the number of stipes of each species was tallied in each quadrat to determine their abundance. To estimate the biomass of E. radiata and P. comosa at Coal Point, four individual adult sporophytes were collected each season from 8 to 10 m depth and taken to the laboratory; each was placed in an aluminum tray and dried in an oven at 60°C until constant dry weight was achieved (~3 nights). Biomass (g DW · m −2 ) of E. radiata and P. comosa was estimated for each season by multiplying the number of stipes per m with mean dry weight of each species (g DW).
The understory at Coal Point was defined by seaweeds that were <30 cm tall. Each season, each quadrat was cleared of all understory seaweeds, which were collected in clear polyethylene zip-lock bags. In the laboratory, each quadrat of understory algae was analyzed separately and sorted by phyla (Rhodophyta, Chlorophyta, or Ochrophyta) before being placed in labeled aluminum trays and dried for constant dry weight (~3 nights). Understory biomass was calculated as the average dry weight of seaweed (g DW) per m 2 (±SE) for each season.
In situ seawater samples were taken during the seasonal biomass dive surveys throughout the dive site at the surface, above the seaweed canopy (~5 m depth), and below the canopy (8-10 m depth) for average concentrations of DOC and inorganic nitrogen (NO 3 − and NH 4 + ). All seawater samples were filtered through precombusted 0.45-μM glass filter paper (Whatman GF/F). DOC seawater samples were poisoned with 100 μL of orthophosphoric acid (85%) and stored at 4°C in glass TOC vials (40 mL, Shimadzu) until further analysis. Seawater samples (12 mL) for inorganic nitrogen were stored in 12-mL polyethylene tubes at −20°C until further analysis.

Experiments to determine the rate of DOC release
We conducted laboratory experiments to measure the rates of DOC release each season by dominant seaweeds at Coal Point. Five individuals of Ecklonia radiata (juvenile sporophytes) and Phyllospora comosa (one lateral blade from individual adult sporophytes) as well as each of the dominant red seaweeds (based on visual observation at the time of collection each season and collected as whole individuals), Lenormandia marginata, Plocamium cirrhosum, Hemineura frondosa, Delisea plumosa, and Schottera nicaeensis were collected from Coal Point at 8-10 m depth during dive biomass surveys. Juveniles of E. radiata were used because previous research revealed no difference in the DOC release rates by whole mature and whole juvenile sporophytes (see Paine et al., 2023, Figure S1). Seaweeds were all ~25 cm 2 and were transported to the laboratory, 1.5 h away, in a dark insulated container.
In the laboratory, seaweeds collected for the timecourse DOC release experiment were cleaned of epiphytes using tweezers and Kimwipes™ (KIMTECH® Science™) and placed into 1-L glass beakers with filtered seawater (0.22 μm, cartridge filters) overnight. The following day, we filled conical flasks (250 mL) with filtered seawater and took initial seawater samples at 9:00 a.m. using syringes (12 mL) to draw and filter water from each flask. Syringe and filter were rinsed twice in distilled water between each flask sampling. Thirty-six milliliters of seawater were sampled from each flask, stored in 40-mL glass TOC vials, poisoned with 100 μL of orthophosphoric acid (85% v/v), and stored at 4°C. After initial seawater samples were taken (time 0 h), replicate seaweeds were each introduced to their individual 250-mL conical flasks for the duration of the incubation. Conical flasks were left open to the atmosphere to avoid unwanted changes to the seawater carbonate chemistry, and water motion was provided to the seaweeds using a shaker table to avoid boundary layer buildup during the incubation period. There were five replicates for each species and two controls with no seaweed to ensure DOC change in seawater was facilitated by presence of seaweed. Ambient seawater temperatures measured using our dive computers were used for the experimental incubations (13°C spring, 17°C summer, 15°C autumn, and 11°C winter) and irradiance was kept constant at 150 μmol photons · m −2 · s −1 on a 12:12 light/dark cycle for all seasons. Seaweeds were removed after 24 h and 36 mL of the seawater was sampled for final DOC content. Once removed from flasks, we placed the seaweeds in an oven at 60°C for three nights to determine dry weight (g DW).

Seawater analysis for DOC and inorganic nitrogen
DOC samples were analyzed at the Queensland Chemistry Centre using an automated total organic carbon analyzer (Analytik Jena Multi N/C 3100) via combustion at 720°C over a platinum catalyst in accordance with method 5310 D (Standard Methods Committee of the American Public Health Association, 2017). Net DOC release rates (μmol C · g DW -1 · h −1 ) were determined using the following equation: where C 0 is the DOC concentration (μmol · L −1 ) in the initial (time = 0 h) seawater sample; C f is the DOC concentration (μmol · L −1 ) in the final (24 h) seawater sample; V is the initial seawater volume (L) at time = 0 h, i.e., 0.214 L as 36 mL was sampled at time 0; T is the incubation duration (24 h); and W is the seaweed biomass (in g of dry weight). Positive rates of net DOC production indicate DOC release, while negative values indicate net DOC uptake in the system. We used a QuickChem® 8000 Automated Ion Analyzer (LaChat Instruments) to determine concentrations of NO 3 − and NH 4 + following methods outlined in Diamond (2008) and Liao (2008). δ 13 C, percent C and N, and C:N ratios Seaweeds from the DOC incubation experiments at each seasonal sampling were analyzed for percent tissue carbon, nitrogen, C:N ratios, and δ 13 C isotopes ratios. Percent tissue carbon and nitrogen contents and stable isotope ratios were analyzed using a Fisons NA1500 elemental analyzer coupled to a Thermo Scientific Delta V Plus via a Conflo IV. Sample combustion and reduction were attained at 1020°C and 650°C, respectively. Carbon and nitrogen were calculated using peak areas of standards with known concentrations. Percent carbon and nitrogen were expressed as a percentage of seaweed dry weight, and C:N ratios are given as mol:mol. Carbon isotope values were reported as δ-values (‰) relative to Vienna Peedee Belemnite (VPDB) and were corrected using 3-point calibration using certified reference standards. Internal reference materials were used to monitor reproducibility. Both precision and accuracy were 1‰ (1 SD).

DOC release by the Coal Point reef assemblage
To estimate the DOC release by the Coal Point seaweed assemblage, DOC release rates for each dominant group (Ecklonia radiata, Phyllospora comosa, and the combined understory) were scaled up using the seaweed biomass and repeated for each transect (n = 3) using the following equation: where Transect DOC is the DOC produced per m 2 (in g C · m −2 · h −1 ) in each transect; Biomass is the average seaweed dry weight per m 2 (g DW · m −2 ); DOC is the average DOC produced by the seaweed (g C · g DW -1 · h −1 ); and Time is the number of hours in a day (i.e., 24). Reef DOC was then calculated as the mean of each Transect DOC (n = 3) for each dominant group and repeated for each season. Reef DOC of each dominant seaweed group was plotted as a bar chart, and standard deviation was propagated using the following equation: where is standard deviation, R is Reef DOC , a is DOC (as above), and b is Biomass (as above). To further scale up and approximate the DOC released by the total Coal Point reef assemblage, the Reef DOC of each dominant group was summed together for each season, and standard deviation was propagated using the following equation: where, Q is the calculated standard deviation, and x, y, and z represent the Reef DOC of each of the dominant seaweed groups (E. radiata, P. comosa, and the combined understory), respectively.

Statistical analysis
Statistical analyses were performed in the software program R, version 4.2 (R Core Team, 2022), and plots were made using SigmaPlot. Inorganic seawater nitrogen concentrations and seaweed biomass were examined using a one-way ANOVA with season as the independent factor. For DOC release by individual seaweed, percent tissue carbon and nitrogen content, and reef-scale DOC release, two-way ANOVAs were conducted with the factors season × species.
Assumptions of normality and variance were assessed using diagnostic plots of model residuals, and response variables (biomass, DOC release, and nutrients) were transformed where necessary using the package bestNormalize (Peterson, 2021). For exact transformations, see Table S1 in the Supporting Information. For the one-way ANOVAs, significance was marked on our plots using letters assigned by a multiple comparison boxplot.

Biomass of seaweed assemblage at Coal Point
The biomasses of Phyllospora comosa, Ecklonia radiata, and understory seaweeds were similar for each season. P. comosa was the dominant seaweed of the Coal Point Reef with ~18 times greater biomass than E. radiata on average across all sampling seasons F I G U R E 2 Biomass (g DW · m −2 ) of (a) Ecklonia radiata and Phyllospora comosa, and (b) understory seaweeds (sorted by phyla) surveyed at Coal Point, Bruny Island, during each season. The plot depicts the 95% confidence intervals (error bars), the first quartile and third quartile (box edges), and median (bold black line). Circles indicate points of biomass above or below the 95% confidence intervals (n = 18). For two-way ANOVA test results, see Table S1.
( Figure 2a). The understory seaweed assemblage was dominated by the phyla Rhodophyta and Ochrophyta (Figure 2b), with the Ochrophyta understory primarily consisting of juvenile P. comosa and E. radiata.

Seasonal concentrations of seawater DOC and inorganic nitrogen
In situ DOC concentrations at Coal Point were 72% higher in summer, at 153 μM (±9.7 SE), than in autumn, at 89 μM (±6.4 SE; p = 0.04; F value = 4.475; Figure 3a). Note, however, the large variability between seasons. NO 3 − varied seasonally, with the greatest concentrations of 11.2 μM measured during winter and the lowest (~1.64 μM) in spring and summer (p < 0.001; F value = 92.14; Figure 3b); however, concentrations of NH 4 + were unaffected by season and ranged from 0.18 to 1.44 μM (p = 0.076; F value = 2.641; Figure 3b). The seawater temperature measured at Coal Point for each sampling period was 13°C in spring, 17°C in summer, 15°C in autumn, and 11°C in winter.

DOC release by dominant seaweed species
Ecklonia radiata and Phyllospora comosa were the dominant canopy-forming seaweeds, while Lenormandia marginata and Plocamium cirrhosum were the dominant understory reef seaweeds yearround. Hemineura frondosa, Schottera nicaeensis, and Delisea plumosa were seasonally dominant understory seaweeds. For all species at Coal Point, the rate of DOC release was higher during the spring and summer compared to the autumn and winter (p = 0.014; F value = 2.435; Figure 4). Ecklonia radiata had ~320%, P. comosa ~ 1500%, L. marginata ~ 2400%, and P. cirrhosum ~ 470% greater release rates in spring and summer (Figure 4) compared to in autumn and winter, respectively (Figure 4).

Percent dry tissue carbon and nitrogen
The dry tissue C:N ratio for each seaweed species (except for Hemineura frondosa, Delisea plumosa, and Schottera nicaeensis, which were only seasonally dominant) was ~50% greater in winter compared to in spring (p < 0.001; F value = 64.4; Figure 5a). The high C:N ratio in winter was driven by the low percent dry tissue nitrogen content (~34% lower in early winter compared to in spring; Figure 5c). Isotope results confirm that Plocamium cirrhosum, H. frondosa, D. plumosa, and S. nicaeensis are non-CCM species with δ 13 C isotopes below −30‰ year-round and that Ecklonia radiata, Phyllospora comosa, and Lenormandia marginata used CCMs with δ 13 C isotopes between −15 and − 30‰ (see Figure S1).

Reef-scale DOC release
When scaled up to the entire reef at Coal Point, Phyllospora comosa released substantially (~14 times) more DOC into the system during spring and summer than Ecklonia radiata and the understory seaweed assemblage (p < 0.001; F value = 52.710; Figure 6a). When the rates of species were combined, DOC release was influenced by season with ~16 times greater release in spring and summer than in autumn and winter (p < 0.001; F value = 31.255; Figure 6b).

F I G U R E 3 Concentrations of (a) DOC (μM) and (b) inorganic nitrogen (μM of NO 3
− and NH 4 + ) measured in situ at Coal Point for each sampling season. Bars represent mean ± SE (n = 5). For significant results, seasons displaying the same letter were not significantly different in post hoc tests. For one-way ANOVA test results, see Table S1.

D ISCUSSION
Our results show clear seasonal trends in rates of DOC release from the seaweed assemblage at Coal Point, Tasmania and support our first hypothesis that DOC release would be higher in spring and summer than in autumn and winter. Our second hypothesis was not supported by our results as all species released DOC within the same range. Note that the use of juveniles may have underestimated the DOC release by reproductive Ecklonia radiata; however, all seaweeds used in this experiment were non-reproductive for consistency. Reproductive tissue would have likely driven higher DOC release (Brawley et al., 1999;Muhlin et al., 2011;Swanson & Druehl, 2002) but was not considered in this study. The calculated rate of DOC release by E. radiata at 8 m depth in spring (1.04 g C · m −2 · d −1 ) was in the same range as previously estimated over a year at 10 m depth at West Island South Australia (537 g C · m −2 · y −1 ; Fairhead, 2001). The combined release rates of the reef were ~ 16 times greater in spring and summer than in autumn and winter, likely driven by seasonal patterns in seaweed gross photosynthesis and growth, regulated by the variations in NO 3 − and seawater temperature (varying from 11°C in winter to 17°C in summer; López-Sandoval et al., 2011;Mueller et al., 2014). Ambient NO 3 − concentrations were lowest in the water column in spring and summer and may be associated with the period of maximal rates of DOC release for all dominant reef species. This is only the second study to compare DOC release rates between seaweed species with varied C i uptake strategies, the first being, Paine et al. (2023). It has been hypothesized that non-CCM seaweeds are C i -limited under current ocean CO 2 concentrations (Cornwall et al., 2017;Cornwall & Hurd, 2019); we predicted, due to this limitation of photosynthesis, that DOC release rates would be lower compared to seaweeds with CCMs given the close links between photosynthesis and DOC release (Barrón et al., 2014;Fogg, 1984;Marañón et al., 2004;Paine et al., 2021). Instead, our results show that yearround, seaweeds that solely use CO 2 as their source of C i (Plocamium cirrhosum, Hemineura frondosa, Delisea plumosa, and Schottera nicaeensis) release DOC within the same range as those that use a CCM (Ecklonia radiata, Phyllospora comosa, and Lenormandia marginata). These results suggest that C i uptake strategy is not a key factor in regulating the mechanisms of DOC release.
Due to the intricately linked carbon and nitrogen metabolisms, seasonal nitrogen physiology is thought to be one of the primary drivers of DOC release in seaweeds. However, this is the first study to compare seasonal C:N ratios with rates of DOC release. In many temperate marine systems, nitrogen uptake and storage as well as the biochemical composition of seaweeds vary in response to changes in water temperature, nitrogen concentrations, wave motion, and light availability (Flukes, 2015;Kain, 1989;Lüning, 1993;Phillips & Hurd, 2004;Roleda & Hurd, 2019). The seasonal nutrient status of the seaweeds can be inferred from their C:N ratios with values <10 mol:mol indicating nitrogen sufficiency and >20 mol:mol suggesting nitrogen limitation (Harrison & Druehl, 1982;Hurd et al., 1996Hurd et al., , 2014Shepard et al., 2023). Maximal growth of Ecklonia radiata and Phyllospo comosa during spring (Sanderson, 1990;Wernberg et al., 2019) coincided with the lowest seasonal C:N values for each species in this study, suggesting that nitrogen was sufficient for seaweed growth and that DOC release was, therefore, likely due to photosynthetic overflow caused by nitrogen sufficiency. C:N ratios increased into summer, suggesting that stored nitrogen was used for growth when NO 3 − concentrations were low in the water column and that DOC was released as an overflow in response to nitrogen limitation.
Minimal growth periods of Ecklonia radiata and Phyllospora comosa occurred in winter, likely due F I G U R E 4 Rates of DOC released (μmol C · g DW −1 · h −1 ) by dominant seaweed collected at Coal Point at each sampling season, spring 2020 (September, white bars), summer 2021 (January, light gray bars), autumn 2021 (April, gray bars), and winter 2021 (June, dark gray bars). Ecklonia radiata and Phyllospora comosa were the dominant brown seaweeds year-round, Lenormandia marginata and Plocamium cirrhosum were the dominant red seaweeds year-round, and Hemineura frondosa, Schottera nicaeensis, and Delisea plumosa were seasonally dominant red seaweeds. Bars represent mean ± SE (n = 5). For two-way ANOVA results, see Table S1. to lower temperatures and suboptimal light levels (Sanderson, 1990;Wernberg et al., 2019), and coincided with the highest C:N values (20-30 mol:mol) observed in this study. This suggests that even though seawater NO 3 − concentrations were highest in winter, there was limited nitrogen uptake due to the low requirements of nitrogen for seaweed growth at this time of year. Previous studies observed the highest seasonal C:N values for E. radiata during autumn at between 30 and 50 mol:mol (Miller, 2004;Wernberg et al., 2019), suggesting that nitrogen stores were drawn down throughout the previous spring and summer. E. radiata and, particularly, P. comosa had relatively high C:N values year-round (especially in summer, autumn, and winter), indicating severe nitrogen limitation for growth (~30 mol:mol) and supporting previous seasonal C:N values for both species (Miller, 2004;Smart et al., 2022;Wernberg et al., 2019).
Seasonal growth patterns of the red seaweeds Lenormandia marginata, Plocamium cirrhosum, Hemineura frondosa, Delisea plumosa, and Schottera nicaeensis have not been previously studied . Therefore, the drivers of DOC release are considered based solely on their nitrogen status. Compared to Ecklonia radiata and Phyllospora comosa, the red seaweeds had a higher nitrogen content year-round and lower C:N ratios of between 10 and 15 mol:mol, likely due to the accumulation of their nitrogen-rich accessory pigments, phycobiliproteins (Barufi et al., 2011;Figueroa et al., 1997;Shepard et al., 2023). To sustain growth, these nitrogen stores F I G U R E 5 (a) C:N ratios (mol:mol), (b) percent dry tissue carbon, and (c) percent dry tissue nitrogen of dominant seaweed species used for DOC incubation experiments during spring 2020 (September, white bars), summer 2021 (January, light gray bars), autumn 2021 (April, gray bars), and winter 2021 (June, dark gray bars). Dashed line represents the critical threshold for nitrogen sufficiency at >20 mol:mol with seaweeds experiencing nitrogen limitation (Harrison & Druehl, 1982;Hurd et al., 1996Hurd et al., , 2014. Bars represent mean ± SE (n = 5). For two-way ANOVA test results, see Table S1. can be assimilated during periods of nitrogen or light limitation (Lapointe, 1981;Marinho-Soriano, 2012;Paine et al., 2020). During spring and summer, C:N ratios of L. marginata indicated sufficient nitrogen for growth (~10 mol:mol), suggesting high growth rates of the species coinciding with the period of greatest DOC release. H. frondosa was a dominant red species during spring and summer but was not observed at the study site during autumn and winter. D. plumosa and S. nicaeensis were the dominant red seaweeds occupying the reef when H. frondosa was not observed.
Phyllospora comosa was the greatest contributor of DOC to the coastal oceanic carbon pool of Coal Point, releasing 11.75 g C · m −2 · d −1 in summer, equating to ~14 times more DOC than Ecklonia radiata and the understory assemblage. This substantial DOC release was attributed to the large biomass of P. comosa on the reef compared to other species, a finding supported by numerous studies identifying the species as one of the most abundant and widespread habitat-forming seaweeds in eastern Australia (Coleman & Wernberg, 2017;Flukes, 2015;Sanderson, 1990). DOC released by this dominant reef species must therefore be considered an important contributor to the coastal ocean carbon cycle sustaining the microbial food web and associated higher trophic levels (Duggins et al., 1989;Elliott Smith & Fox, 2021). Previous research on the nitrogen physiology of P. comosa identified that the species takes up nitrogen using biphasic mechanisms, using active uptake mechanisms (requiring energy) when nitrogen concentrations are low and passive uptake mechanisms (diffusion, no energy required) when nitrogen concentrations are high (Buchanan et al., 2000;Smart et al., 2022). This biphasic uptake is considered a physiological adaption to survive in areas with large variations in nitrogen concentrations, such as temperate regions, and may explain the species high growth rates and comparably nitrogen-limited tissue (C:N ratios of 25-40 mol:mol) with substantial DOC release as the species undergoes a period of imbalance between gross photosynthesis and growth-facilitating DOC release.

Drivers of reef-scale DOC production
Large seasonal DOC fluxes from seaweed (up to 50-fold per unit biomass or surface area) have been reported in northern hemisphere temperate and subtropical regions with pronounced seasonality, supporting the results of this study (Haas et al., 2010;Wada et al., 2007). Generally, concentrations of DOC increase in surface waters in temperate regions during spring are linked to a nutrient-facilitated increase in NPP (increase of 5-10 μmol C · kg −1 ) and decrease in summer and autumn due to micro-heterotrophic (i.e., bacterial) consumption of DOC and nutrient limitation of NPP (Carlson et al., 1994;Hansell et al., 2012). During temperate winters, seaweed gross photosynthesis and growth are usually limited by low temperatures and reduced light intensity, which subsequently regulates DOC release (Johnston et al., 1977). Light intensity also likely played a role in regulating the DOC release with season in this study (Mueller et al., 2014(Mueller et al., , 2016Reed et al., 2015). Supporting this, in California, DOC release rates by the dominant reef species, Macrocystis pyrifera, varied unpredictably with season and instead correlated with irradiance (Reed et al., 2015).
The release of DOC by the reef at Coal Point was likely driven by change in the balance between seaweed gross photosynthesis and growth associated with seasonal environmental drivers (Fairhead & Cheshire, 2004;Miller et al., 2011). Biomass of the seaweed assemblage at Coal Point did not change significantly with the season and was unlikely to be a strong F I G U R E 6 Seasonal DOC release by the seaweed assemblage (in g C · m −2 · d −1 ) at Coal Point, Bruny Island, (a) release separated by dominant seaweed groups and (b) combined DOC release by the total seaweed assemblage. Bars represent mean ± SD (n = 3). For significant results, seasons displaying the same letter were not significantly different in post hoc tests. For ANOVA test results, see Table S1. driver of the changes in reef-scale DOC release. The observed difference in DOC release with season is instead probably associated with the seasonal fluctuations in NO 3 − and changes in seawater temperature (measured at 13°C in spring, 17°C in summer, 15°C in autumn, and 11°C in winter). NO 3 − concentrations decreased in the water column from winter to spring due to the increase in autotrophic growth and remained constant during summer before beginning to increase again in autumn with the water column overturn. It is also interesting to note that the partial pressure of carbon dioxide (pCO 2 ) in water over the continental shelf south of Tasmania follows a similar seasonal trend with partial pressure increasing between austral autumnwinter, associated with mixing of surface waters with DIC-rich deeper waters, before decreasing during austral spring-summer as biological uptake removes DIC from the surface waters (Borges et al., 2008). Due to the combination of these environmental factors, the seaweeds were likely undergoing a period of high gross photosynthesis during spring and early summer. DOC was, therefore, released at higher rates compared to winter and autumn due to imbalances with growth and gross photosynthesis. The significant DOC release by the seaweed assemblage was not observed in the in situ DOC seawater samples, and although difficult to underpin, this was possibly due to consumption by heterotrophs in spring.

CONCLUSIONS
We have shown clear seasonal trends in rates of DOC release from the temperate seaweed assemblage at Coal Point, Bruny Island, Tasmania, likely associated with changing rates of gross photosynthesis of the seaweeds. Our results add to the growing literature on environmental regulation of seaweed DOC release and highlight the significant regulatory role that nitrogen availability and temperature have on seaweed physiology and subsequent DOC release. We highlight the importance of the seaweed reef in providing DOC to the detrital food web-which, in turn, supports higher trophic levels-and how co-limiting environmental factors underpin the carbon cycle in temperate coastal oceans.

ACK NOWLEDG M ENT S
We thank Dr. Damon Britton, Dr. Cayne Layton, Jane Ruckert, Simon Talbot, and the Hurd laboratory group for their help in the laboratory, field, and interpretation of results. We acknowledge and pay our respects to the palawa people of lunawanna-allonah and nipaluna, whose water and land upon which we work. Laureate Fellowship FL160100131 to P.W.B., ARC DP160103071 funding to GD-P and CLH, and ARC DP200101467 to CLH. MS was funded consecutively through a Science Foundation Ireland grant (18/FR/6198) and a European Commission Marie-Curie Actions Postdoctoral Fellowship (Project 101066815-ASPIRE). Open access publishing facilitated by University of Tasmania, as part of the Wiley -University of Tasmania agreement via the Council of Australian University Librarians.

DATA AVAI L ABI LI T Y STATEM ENT
Available upon request.