Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Jul 15;247(6):2616–2629. doi: 10.1111/nph.70382

Altered nutrient cycling functionality in seagrass meadows under a simulated future marine heatwave event

Alissa V Bass 1,✉, Ziyan Wang 1, Nga Man Chung 2, Mandy Wing Kwan So 2, Laura J Falkenberg 1,3, Benoit Thibodeau 1,2,✉
PMCID: PMC12371151  PMID: 40665617

Summary

  • Seagrasses are important contributors to environmental nutrient cycling in marine ecosystems and can improve water quality by absorbing excess nitrogen (N). However, these ecosystems are vulnerable to human‐mediated pressures, including marine heatwaves (MHWs), particularly those of longer duration.

  • We performed an experiment simulating a 30‐d, +5°C intensity MHW to examine the effects on maximum potential N cycling and transformation rates in different system compartments associated with the tropical seagrass Halophila ovalis.

  • Under the MHW, the seagrass exhibited higher ammonium assimilation rates in the leaves, increased respiration rates, and highly variable survival and growth than those at ambient temperature. Contrary to expectations, sediment denitrification rates were lowered under the MHW, reflecting the loss of microbial functions and therefore signifying reduced N removal benefits. Moreover, the lowered seawater total alkalinity under the MHW suggests that the habitat would provide less ocean acidification buffer under future climate change. However, cycling rates in vegetated and unvegetated sediments were not significantly different, showing that the seagrass does not strongly influence the sediment N cycling capacities at this time scale.

  • Our study demonstrates that future MHWs may alter nutrient cycling rates across ecosystem compartments in seagrass meadows, potentially leading to reductions in ecosystem function and services.

Keywords: biogeochemistry, climate change, ecosystem services, Halophila ovalis, isotopes, nitrogen

Introduction

Coastal ecosystems are some of the most important ecosystems globally due to the functions and services they underpin (Koch et al., 2009; Barbier et al., 2011). Found at the interface between the open ocean, land, and the atmosphere, coastal zones and ecosystems have immense nutrient cycling abilities, making them some of the most valuable ecosystems globally (Costanza et al., 1997). Coastal ecosystems are key nitrogen (N) sinks (Ogilve et al., 1997; Voss et al., 2011), with foundation species, such as seagrasses, providing significant benefits to the environment through N cycling (Moulton et al., 2016; Aoki et al., 2020; Olivé et al., 2022). Not only is N uptake necessary and crucial for seagrass growth, survival, and nutrient transport to higher trophic levels, but seagrasses can also provide substantial ecosystem services, such as removing anthropogenic N pollution from the environment (Gaylard et al., 2023). Indeed, seagrasses naturally incorporate N from the water and cycle it through different compartments of the ecosystem (Touchette & Burkholder, 2000; Aoki et al., 2020), with N cycling rates observed to be much higher in vegetated sediments than in bare sediments (Aoki & McGlathery, 2018; Chen et al., 2021). Moreover, the N cycle is closely intertwined with other biogeochemical cycles, such as the carbon (C) and sulfur cycles (Reay et al., 2008; Pachiadaki et al., 2017), and therefore shifts in one cycle due to external forcings can affect multiple biogeochemical cycles and the ecosystem services they provide.

Within the seagrass habitat compartments (i.e. the plant: above‐ and belowground compartments, the sediment, and the surrounding waters), reactive N species are continuously transformed, transferred, and exchanged via seagrass metabolism and microbial activity (Fig. 1). Specific N cycle transformations that are driven by microbial processes, such as dissimilatory nitrate reduction to ammonium (DNRA), help to retain N in the sediment and, following which, it can be transferred to plants, microbes, or transformed back into nitrate by nitrification (Giblin et al., 2013). Loss of N from the sediment also occurs through other transformation processes, including denitrification and anammox (anaerobic ammonium oxidation). Denitrification removes reactive N (i.e. nitrate) from the ecosystem by consuming organic C and preventing its storage, while anammox is the transformation of ammonium with nitrite as an electron donor, with both processes leading to the formation of N2, which can be lost back into the environment (i.e. the surrounding air and water column). Some previous studies show that DNRA can be much higher than denitrification in seagrass meadows (Rysgaard et al., 1996; An & Gardner, 2002; Gardner et al., 2006). By contrast, other studies report relatively low rates of DNRA, accounting for no > 5% of the total nitrate reduction into the sediment (Smyth et al., 2013) and also high anammox rates in seagrass sediments (more than denitrification in terms of N loss; Salk et al., 2017). As a result, identifying the balance between different N transformations that result in N retention or N loss is pivotal in determining the overall benefits of the seagrass ecosystem functions and services.

Fig. 1.

Fig. 1

Open in a new tab

Overview of the N cycle relating to the different system compartments of seagrass meadows (Inspired by Camillini, 2020; Herbert, 1999). It includes N uptake processes into the seagrass through assimilation of free amino acids (FAAs), urea, ammonium, and nitrate from the overlying water, along with sediment processes, anammox, nitrification, DNRA, and nitrification in oxic and anoxic depths of the sediment.

Anthropogenically induced changes to the abiotic environment are causing both direct and indirect impacts to N cycling in marine habitats (Hutchins & Fu, 2017; Wrightson & Tagliabue, 2020; Berlinghof et al., 2024). One such climate change‐related stressor, which has only recently been discovered to affect N cycling, is marine heatwaves (MHWs) (Seidel et al., 2023). MHWs are discrete anomalous warming events of water temperature, and are increasing in duration, intensity, and frequency with climate change (Hobday et al., 2016; Oliver et al., 2018). Unlike a gradual increase in temperature, associated with ocean warming, MHWs involve accelerated temperature increases that rapidly exceed thresholds of adaptive or acclimation capacities of species and therefore can lead to more adverse responses compared to ocean warming (Holbrook et al., 2020). A few studies are beginning to reveal the effect of elevated temperature on N cycling in marine systems. That is, long‐term gradual warming has been shown to have varying degrees of severity on microbes, from altering the sediment microbial structure (Seidel et al., 2023) to inhibiting the microbes involved in N processes (Hutchins & Capone, 2022). In addition, higher temperatures at lower latitudes have been found to increase the N demand of seagrass species, such as increased ammonium and nitrate uptake (Alexandre et al., 2020). Furthermore, changes in temperature from MHWs that affect seagrass metabolism will also likely temporarily increase C exudation and, consequently, seagrass carbon metabolism (Egea et al., 2023). As a result, this change can cascade to affect C availability for N cycling (Hardison et al., 2015). Previous research also shows increased N loss from sediments and limited N fixation as temperatures increase seasonally (Garcias‐Bonet et al., 2018). However, N retention through DNRA also increases seasonally due to temperature increase (Ferrón et al., 2009; Smyth et al., 2013), and DNRA can be favored over denitrification at higher temperatures (Giblin et al., 2013). Beyond these results, however, the extent to which MHWs impact seagrass N cycling has been very little studied.

Here, we addressed this knowledge gap by performing a laboratory experiment to examine how a simulated extreme and extended MHW can influence N cycling in seagrass meadows (i.e. seagrass compartments, the sediment, and the overlying water) using 15N tracers and physiological measurements. We further explored how the abundance of microbes with genes involved in N cycling is impacted by the MHW, and whether vegetated or unvegetated sediment exhibits different rates of N cycling. We adopted the tropical seagrass Halophila ovalis, the most predominant species of seagrass in Hong Kong (AFCD, 2023), as the study model. Under extreme temperature increase, we hypothesized that seagrass N uptake would increase, including N assimilation and fixation, due to increased N demand (Alexandre et al., 2020). Moreover, we anticipated increased bacterial metabolic rates with elevated temperature, and therefore that sediment N cycling rates, such as DNRA and denitrification rates, would increase, and these changes may lead to alterations in net N removal (similar to Garcias‐Bonet et al., 2018). We also hypothesized there would be differences in the amount of reactive N available in the overlying water and the total alkalinity (i.e. the amount of the proton acceptors minus the amount of proton donors; Dickson, 1981; Wolf‐Gladrow et al., 2007), due to alterations in the water‐sediment nutrient flux with temperature (Cerco, 1989) as well as carbon respiration processes (Wolf‐Gladrow et al., 2007), and that there would be higher N loss in vegetated sediments due to benthic respiration (Eyre et al., 2011). Evidence supporting these hypotheses would imply that MHWs can reduce the beneficial effects of seagrass meadow contribution to N cycling and net removal of human‐derived N inputs.

Materials and Methods

Sample collection and experimental preparation

In March 2024, we collected the seagrass Halophila ovalis (R. Br.) Hook. f. and sediment from Tung Chung Bay, Hong Kong (22°17.0934′N, 113°55.5946′E), where the seagrass meadow can be accessed at extreme low tides on foot. The seagrass–and the sediment in which it was rooted–was collected in trays using a shovel, and gently placed into cool boxes. Bare sediment directly adjacent to the seagrass patches was also collected to be used in the experiment for the unvegetated tanks, as well as to plant the seagrass in the vegetated treatment tanks. The seagrass and sediment were then transported to outdoor tanks at the Marine Science Laboratory at the Chinese University of Hong Kong, where they were acclimated to ex situ conditions for 2 d. During this time, the bare sediment was filtered with a metal sieve (mesh size 4 × 4 mm) to remove large biotic material and was mixed to homogenize the chemical, physical, and microbial properties. The seagrass was then removed from the original sediment and prepared for the experiment, including cutting the seagrass into ramets of three pairs of leaves, rhizome, roots, and a growing tip, so that they were all approximately the same length (Supporting Information Fig. S1). For the vegetated treatment, five individual ramets were planted into each 1.5 l tank (13 cm l × 13 cm w × 9 cm h) in 3.5‐cm deep filtered sediment, with five replicate tanks per temperature treatment. In order to examine the differences in sediment N cycling when seagrass was present vs absent, three additional tanks per temperature treatment were set up without any seagrass (i.e. unvegetated treatment), which, together with the vegetated treatment, resulted in a total of eight tanks per temperature (16 total experimental units). As there was the potential for seagrass mortality to affect replicate numbers, and to avoid overcrowding of seagrass in each tank, we chose to have an unbalanced design with more vegetated tanks compared to unvegetated tanks.

Treatment manipulation

After planting the seagrass into experimental tanks at the Marine Science Laboratory, the seagrass tanks were transported to an indoor laboratory at the Chinese University of Hong Kong to run the month‐long experiment. Four tanks were randomly placed into a single water bath, with four water baths in total; two at control temperature and two that would be ramped to MHW treatment temperature (shown in Fig. S2). These water baths were used to maintain treatment temperatures, and there was no mixing of water between any experimental tanks within a water bath. Therefore, vegetated and unvegetated tanks were haphazardly allocated to the two water baths for each treatment temperature in order to minimize bias. Seagrass tanks were then filled with artificial seawater (containing trace amounts of N and a salinity of 30 psu – approximately the optimal salinity for Halophila ovalis), with 50% of the water (c. 500 ml) manually replaced three times a week, and transparent lids placed onto each seagrass tank to prevent evaporative loss and salinity fluctuations. Full‐spectrum light was provided with LED light panels suspended above the tanks (Povi®, Guangzhou, China), providing an average of 100 μmol photons m−2 s−1 daytime light with a 12 h : 12 h, light : dark cycle. Air was gently bubbled into each tank using air stones to keep the water well mixed and aerated, and seagrass tanks were left to acclimate to indoor lab conditions for 4 d further at ambient temperature (25°C). The water temperature of the experimental tanks was controlled by two aquarium bar heaters placed into each water bath to maintain control temperature treatment: 25°C (ambient March–April water temperature in Hong Kong) and a +5°C MHW temperature of 30°C for the MHW treatment (equivalent to an ‘extreme’ category – 4× the value between the local climatological mean and the climatological 90th percentile – MHW event; Bass & Falkenberg, 2024). While this intensity and duration of MHW have not been experienced under natural conditions in Hong Kong, they have occurred independently (Bass et al., 2024) and are therefore a predicted future MHW scenario. To achieve the MHW, in the appropriate tanks, the temperature was increased by 1°C d−1 after the 4‐d laboratory acclimation period. The temperature treatments were then maintained for 30 d after ramping until end measurements and incubations were performed, allowing us to consider a long MHW event for Hong Kong (Bass & Falkenberg, 2024). The average temperature treatment for the control tanks was 25.09 (± 0.01 SE), while the MHW treatment was 30.49°C (± < 0.01 SE; Fig. S3).

Measurements

Overlying water nutrient analysis

Three days before the experiment ended, the water in the experimental tanks was replenished for the last time. At the end of the experiment, before all other measurements, overlying water was sampled from each tank by syringe filtering (0.2 μm pore‐size filter and a 50‐ml syringe), and used for measuring total alkalinity, nitrate, nitrite, and ammonium concentrations in the water. For total alkalinity and nutrient analysis, 50 and 15 ml Falcon™ tubes, respectively, were filled, leaving no headspace, with two replicate samples per tank. These were frozen at −20°C for 1 wk until analysis. Total alkalinity was measured with a titrator connected to an autosampler (888 Titrando and Robotic Analyzer, Metrohm). Ammonium, nitrate, and nitrite were measured spectrophotometrically, with ammonium concentration analyzed using a salicylate method (Giner‐Sanz et al., 2021), and nitrate and nitrite were analyzed with a modified method using vanadium (III) for small volumes (García‐Robledo et al., 2014). The detection limits were c. 1 μmol l−1 for ammonium and 0.1 μmol l−1 for both nitrate (NO3 −) and nitrite (NO2 −).

Seagrass survival, growth, and measurement allocation

Once the overlying water was sampled, seagrass ramets were removed from their temperature treatment water baths and photographed at room temperature. We quantified the number of seagrass ramets in each tank to assess mortality rates as a percentage of ramets per tank. Surviving seagrass ramets (i.e. intact rhizomes with at least one leaf attached) were then removed from the sediment, gently rinsed with seawater to remove loose sediment, and photographed from above against a white background for leaf number calculation. These photographs were then analyzed in the fiji extension of imagej (Schindelin et al., 2012) and compared against the starting leaf number of six per rhizome to calculate leaf number changes per ramet, and consequently leaf density (i.e. leaves from all ramets totaled per tank) change per tank over the 30 d. After the photographs were taken (c. 10‐min duration), seagrass ramets were sorted for short‐term incubation experiments at their treatment temperatures (schematically shown in Fig. S4).

Seagrass N fixation, assimilation, and productivity measurements

One ramet per experimental tank (i.e. five replicates in total per temperature treatment) was used for the N fixation incubation, and then another ramet per tank was used for N (ammonium) assimilation incubation and productivity measurements, also equaling five replicates per treatment. Of these 10 total ramets used for N uptake measurements, one from each temperature treatment was used as a standard to assess natural δ15N values and was therefore not exposed to enriched N. Consequently, there was one fewer replicate for N fixation for each temperature treatment, leaving four replicates instead of five. There were no visible epiphytes attached to the seagrass at the end of the experiment. To assess N fixation rates into the seagrass leaves, 5 ml 98 atom% 15N2 gas (Sigma‐Aldrich) was bubbled into 500 ml autoclaved and degassed artificial seawater to create a deoxygenated stock mixture (Mohr et al., 2010; Wannicke et al., 2018). This mixture was continuously stirred with a magnetic stir bar for 24 h, and subsequently, we dispensed 4 ml of the enriched stock and 36 ml of the degassed, artificial seawater into 40 ml glass incubation bottles, along with a single intact seagrass ramet. After 24 h, the seagrass was removed, dried at 60°C for 48 h, split into aboveground (leaves) and belowground (roots and rhizome) compartments, and encapsulated for bulk isotope analysis.

For the maximum potential ammonium assimilation rates and oxygen measurements, 100 μmol 15NH4 + enriched artificial seawater (originally N‐free) and a single seagrass ramet were added to each 250 ml clear acrylic container. High enrichment in 15NH4 + has been shown to produce the highest uptake rates and therefore represents the maximum potential assimilation rates (Alexandre et al., 2020). The containers had an oxi‐spot attached to the inside (a planar sensor spot with an oxygen‐sensitive coating; SP‐PSt3‐NAU‐D5‐YOP, PreSens, Germany), which allowed oxygen levels to be measured using a fiber optic oxygen sensor (Fibox4 trace, PreSens, Germany). Containers filled with enriched artificial seawater without seagrass were used to account for background respiration rates, which were deducted from the respiration and net primary productivity rates of seagrass before rate calculations. Containers were first sealed tightly and left to equilibrate for 15 min in the water baths at experimental temperatures under dark conditions. Afterward, oxygen levels were measured every 15 min for 45 min each for respiration (under dark conditions) and then, after another 15‐min equilibration under light conditions, for photosynthesis (also under light conditions). We then calculated the respiration rates of each ramet using the following equation:

Rate=O2start−O2end/w×t

where O2 start and O2 end are the oxygen concentrations at the beginning and end of the 45‐min period, respectively, while w is seagrass ramet dry weight biomass and t is time. Therefore, a higher value would translate into a higher respiration rate. The reverse equation was then used for net productivity calculations (O2 end−O2 start).

After all the oxygen measurements were finished (c. 90 min), the seagrass ramets were left uncovered to finish the rest of the 4‐h incubation in 15NH4 + water, and then were taken out, dried at 60°C for 48 h, split into above‐ and belowground compartments, and prepared for bulk stable isotope analysis. Isotope values are expressed as per mil relative to air and VPDB for N and C, respectively, and rates of N uptake are expressed in μmol g−1 DW h−1 and standardized to account for the natural δ15N levels for each treatment (information provided in Notes S1).

Anaerobic potential rate incubations

Once all the seagrass was removed, sediment from each tank was collected for the enriched isotope incubation experiments. From each experimental tank, a slurry was made containing 80 g of sediment and 500 ml autoclaved, artificial N‐free seawater (30 psu) in a 500 ml Duran bottle. To make the slurry anoxic, the bottles were purged with ultra‐high purity helium (99.999%) for 2 h and oxygen concentrations were measured using an O2 sensor spot (TROXSP5, Pyroscience sensor technology) placed inside the 500 ml Duran bottle with a fiber optic O2 sensor to ensure that oxygen levels were below 1 μmol l−1. The bottles were then transferred into an anoxic glove bag, which was also flushed and filled with helium (oxygen level was checked with a PDO‐519 Oxygen Meter; Lutron, Taiwan, china), and from each 500‐ml bottle, 16 gas‐tight 12‐ml evacuated vials (Exetainer®; Labco, Lampeter, Wales, UK) were filled with the slurry using a 50‐ml syringe, which would be used for the isotope tracer incubations. The 18 vials consisted of four different reactions and different tracers (Table 1): DNRA, anammox, denitrification, and a control, with the control measuring background nitrate and nitrite during the preincubation, which would then be subtracted from anammox rates. There were also two time points (T0 – to have the beginning concentrations of each N species, and T1 – after 8 h of injecting the tracer) and two replicates each for the tracers and time point, with an extra two T1 samples for DNRA for each 500 ml bottle. The vials with the slurry were preincubated for 36 h at their treatment temperatures to equilibrate, and then subsequently the T0 samples were injected with 200 μl of 50% ZnCl2 to stop the microbial reactions, while the T1 samples were injected with their designated enriched N tracers (Table 1). Briefly, 120 μl of 10 mmol 15NH4 +, 15NO3 −, and/or 14NO3 − stock tracer solution was injected into each of the vials to measure anammox, DNRA, and denitrification, creating a final concentration of 100 μmol of each enriched solution in the 12‐ml vials. High concentrations of labeled tracers have often been used to produce maximum potential rates in slurry experiments (e.g. Dalsgaard & Thamdrup, 2002; Song et al., 2016; Yin et al., 2017). Control samples were injected with 15NH4 + in order to measure background nitrate and nitrite during the preincubation. After injecting the enriched N solutions, T1 samples were incubated at their experimental temperatures in water baths for 8 h with continuous movement on a shaker, and then stopped with 50% ZnCl2 after 8 h. Afterward, the vials measuring DNRA were flushed with helium for 20 min to remove any 15N2 gas. All vials were then sent for Membrane Inlet Mass Spectroscopy (MIMS; Lin et al., 2021), and then anammox and denitrification rates were calculated using the equations in Thamdrup & Dalsgaard (2002) and DNRA using equations in Song et al. (2016) (Notes S1), with rates of enriched N usage being calculated. Due to the high concentrations of 15N tracers, the rates calculated are the maximum potential rates, assuming there is no N limitation. For each temperature treatment, the end data points consisted of five tank replicates for the vegetated sediments and three replicates for the unvegetated sediments (i.e. one replicate per tank), each with duplicate measurements per tank replicate. Data points from the MIMS that indicated errors in the results, such as negative values, were excluded from the analysis.

Table 1.

Summary of the enriched nitrogen (N) incubation experiment procedures.

Process Reaction Experiment procedure
N fixation 15N2 to 15N seagrass 4 ml 15N2 seawater + 36 ml seawater in 40 ml vials
Anammox 15NH4 + + 14NO3 − to N2 120 μL 15NH4 + + 120 μl 14NO3 − + 12 ml slurry vials
DNRA 15NO3 − to 15NH4 + 120 μL 15NO3 − + 12 ml slurry vials
Denitrification 15NO3 − to 30N2 120 μL 15NO3 − + 12 ml slurry vials
Control 15NH4 + to 15NO3 − + 15NO2 − 120 μl 15NH4 + + 12 ml slurry vials
N assimilation 15NH4 + to 15N seagrass 100 μmol 15NH4 + in 250 ml container
Nitrification 15NH4 + to 15NO3 − 10 ml 100 μmol 15NH4 + seawater + 2 g sediment
Open in a new tab

The incubations performed in anoxic conditions (first five rows) and in an oxygen‐rich environment (last two rows).

Aerobic potential rate incubations

Analysis of nitrification rates was performed under aerobic conditions. In each 12 ml exetainer vial, 2 ± 0.1 g of wet sediment from each tank was mixed with 10 ml of oxygenated and autoclaved artificial seawater. Immediately, 50% ZnCl2 was added to half the samples at T0 (those without 15NH4 + addition), while the other half of the samples were injected with 120 μl 15NH4 + (a final concentration of 100 μmol). T1 vials were incubated at experimental treatment temperatures for 8 h and then stopped with 50% ZnCl2 (Ottosen et al., 1999), and subsequently, 10 ml of filtered water was taken from the vials after the sediment had settled. These water samples were frozen and stored at −20°C until further analysis 4 months later. Once ready for analysis, the samples were thawed and analyzed by the bacterial denitrifier method (Sigman et al., 2001). In short, 15NO3 − and 15NO2 − in the samples, produced from the nitrification of 15NH4 +, were reduced via denitrification to 15N2O gas by the nitrifying bacterium Pseudomonas aureofaciens, without being completely reduced due to the lack of N2O reductase. Before the addition of the bacteria, NO2 − was removed with sulfamic acid, with only NO3 − remaining (Granger & Sigman, 2009). After adding the bacteria and converting to gas, the gas samples were then analyzed on a GasBench II coupled to a mass spectrometer (IRMS; Thermo Delta V Advantage; Thermo Fisher Scientific, Waltham, MA, USA), with the international standards USG34 (δ15N = −1.8‰), USG35 (δ15N = 2.7‰), and the enriched standard USG32 (δ15N = 180‰) (Böhlke et al., 2003). However, because of the extremely strong enrichment of the nitrification samples, we observed a memory effect on the standards, and thus we used the raw data to perform the calculations. Therefore, the nitrification rates are considered maximum values and are interpreted as the relative comparison between treatment groups rather than absolute values. Vials with only the bacteria were used as the blanks. Under the assumption that all 15NO3 − is converted into 15N2O, the isotopic abundance of 15N2O gas is therefore the isotopic abundance of 15NO3 − from the nitrification of 15NH4 +. The concentration of 15NO3 − (μmol l−1) was calculated using the equations provided in Notes S1. Afterward, the T0 sample results were subtracted from the T1 samples to account for the background 15NO3 levels in the preincubation stage, and the resulting sediment N transformation rates are expressed in μmol kg−1 h−1.

Analysis of N cycling‐related gene abundance

DNA was extracted from remaining sediment samples (stored at −80°C before analysis). From each tank sample, DNA was extracted from c. 1 g of sediment, following the manufacturer's guidelines (Powersoil Pro Kit; Qiagen), and DNA was quantified on a NanoDrop One Microvolume UV–Vis Spectrophotometer (Thermo Fisher Scientific). For denitrification, where rates were statistically different from the anoxic slurry incubations and were the highest (see Sediment potential N transformation rates in the Results section), two marker genes for the enzymes involved in the specific N transformations were amplified, which were nitrous oxide reductase and nitrite reductase in denitrification (nosZ and nirS, respectively). We then performed quantitative PCR (qPCR) on these genes' copy number in the samples to assess any differences in gene copy numbers (μl kg−1 of sediment) for the different experimental treatments (CFX Opus 96; Bio‐Rad). More information can be found in Notes S2.

Statistical analysis

All statistical analyses were conducted in R (v.4.3.2). Before analyses, data were first checked to ensure normality and equal variances, extreme outliers were identified and removed, and then linear mixed effects models were performed in the package lme4 (Bates et al., 2015) and then an analysis of variance (ANOVA) was applied in package car (Fox & Weisberg, 2011) to generate P‐values. One‐way ANOVAs were conducted on the seagrass mortality and productivity measurements with the fixed factor ‘treatment’ (with two levels: control or MHW, n = 5 per treatment), while two‐way ANOVAs were performed for all other measurements. For example, the seagrass isotope and N and C content responses had the fixed factors ‘treatment’ and ‘compartment’ (with two levels: above‐ or belowground seagrass compartment, n = 5 for each level) while the water nutrients and sediment measurements had the fixed factors ‘treatment’ and ‘vegetation’ (i.e. vegetated vs unvegetated sediment tanks, n = 5 vegetated, n = 3 unvegetated). Multiple measurements within tanks (such as multiple seagrass ramets within a single tank, multiple slurry vials per tank, and multiple water samples per tank) were treated as ‘random effects’ in the linear mixed effects model before the ANOVA. For the measurements where there were no random effects, such as just one replicate per tank (e.g. seagrass physiology and N uptake rates), a linear model, rather than the linear mixed effects model, was performed before the ANOVA. When there were significant effects of the fixed factor interactions from the ANOVA test (e.g. vegetation × treatment), Tukey's post hoc pairwise comparisons were carried out on the fixed factors in the lsmeans package (Lenth, 2016). All data points, including within‐tank measurements, are presented on the result figures.

Results

Overlying water characteristics

The final total alkalinity of overlying water in the experimental tanks was significantly lower in the elevated temperature treatment than the control (2.26 mmol l−1 ± 0.23 SE vs 1.71 mmol l−1 ± 0.1 SE; Fig. 2a; Table S1). Under the control temperature, nitrate concentration averaged 120 μmol l−1 ± 17.2 SE, but was considerably lower with the MHW, down to 20.1 μmol l−1 ± 6.0 SE in both the absence and presence of vegetation (Fig. 2b; Table S1). Nitrite and ammonium were extremely low in our samples – below the detection limit – and are therefore not reported here.

Fig. 2.

Fig. 2

Open in a new tab

Overlying water analysis from experimental tanks, including (a) total alkalinity and (b) nitrate concentration under different temperature treatments and vegetation presence and absence. Boxplots represent median, upper, and lower quartiles, with the diamonds showing the mean values and individual points showing each measurement. Different letters indicate statistical differences from ANOVA main effects tests (‘Treatment’ and ‘Vegetation’). MHW, marine heatwave.

Seagrass mortality and physiology

At the end of the experiment, seagrass ramet mortality at control temperature treatment was 40% in all tanks (i.e. three out of five ramets survived) (Fig. 3a). Under the MHW temperature treatment, mortality increased to 66% (± 7.48 SE), although there was no significant difference between the temperature treatments (Table S1). Despite the relatively high mortality for the control temperature treatment, seagrass leaf density change after the 30 d was approximately zero (±2.68 SE; Fig. 3b), indicating that the new leaf production of the surviving ramets compensated for the loss of the degraded ramets. However, for the MHW treatment, leaf density change was distinctly variable, with an average loss of 3.6 leaves per tank and notably ±7.19 SE, indicating that where seagrass survived under the MHW, individuals' growth was boosted significantly or drastically reduced. Respiration rate (i.e. the oxygen consumed during dark conditions) was pronouncedly higher under the MHW treatment (Fig. 3c, Table S1) with an average of 4.55 mgO2 L−1 DW−1 h−1 (± 1.72 SE) than under the control temperature treatment, which had an average respiration rate of −1.66 (± 1.93 SE, due to respiration rate of the seagrass being lower than that of the blank control units without the seagrass). Net primary productivity (oxygen production under light conditions) was low for both temperature treatments, and both slightly negative (for the control and the MHW treatments respectively −0.29 mgO2 l−1 DW−1 h−1 ± 0.003 SE and −0.195 mgO2 l−1 DW−1 h−1 ± 0.012 SE; Fig. 3d). This indicates that there was still high respiration in each of the treatments, potentially from the high microbial respiration, which reduces the accuracy of these results. Nevertheless, under the MHW treatment rates were significantly higher than the control treatment (Fig. 3d; Table S1).

Fig. 3.

Fig. 3

Open in a new tab

Halophila ovalis seagrass growth and metabolism measurements at the end of the experiment, including (a) seagrass ramet mortality percentage per tank, (b) change in leaf density per tank, (c) respiration rate, and (d) net primary productivity rate, calculated from changes in dissolved oxygen concentrations. Boxplots represent median, upper, and lower quartiles, with the diamonds showing the mean values and individual points showing each measurement. Different letters indicate statistical differences from ANOVA main effects tests (‘Treatment’). Graphs without letters revealed no statistical differences between any treatments. MHW, marine heatwave.

Seagrass isotope enrichment and bulk nutrients

Assimilation of 15N‐enriched ammonium was significantly impacted by the interaction between temperature and seagrass compartment (Fig. 4a; Table S1). The MHW had a different effect on the different seagrass compartments, with the aboveground assimilation rate significantly increasing by an average of c. 20 μmol 15N g−1 DW h−1 compared to the control, but the belowground compartments not changing its 15N significantly (an average increase of c. 2 μmol g−1 DW h−1). By contrast, N fixation of dissolved 15N2 into the seagrass was variable and low, at an average of 0.09 μmol 15N g−1 DW h−1 across the temperature treatments (Fig. 4b; Table S1), and was not statistically different between seagrass compartments. The δ13C was higher in the above than in the belowground compartments of the seagrass and was significantly lowered under the MHW for both compartments (Fig. 4c; Table S1). Seagrass N% and C% were not affected by the MHW (although there was a trend toward a significant effect for N%; P = 0.054), but they were different between the compartments, with lower N% and C% in the belowground than in the aboveground sections (Fig. 4d,e; Table S1).

Fig. 4.

Fig. 4

Open in a new tab

Halophila ovalis seagrass nitrogen uptake rates of (a) NH4 + through assimilation and (b) N2 through fixation, along with the naturally occurring abundance of (c) δ13C, (d) N%, and (e) C% in the above (leaves) and belowground (roots and rhizome) compartments. Boxplots represent median, upper, and lower quartiles, with the diamonds showing the mean values and individual points showing all measurements. Different letters indicate statistical differences from ANOVA main effects tests (‘Treatment’ and ‘Compartment’, but a post hoc Tukey test for (a) since there was a ‘Treatment’ × ‘Compartment’ interaction). Graphs without letters revealed no statistical differences between any treatments. MHW, marine heatwave.

Sediment potential N transformation rates

The maximum potential rate of DNRA was significantly reduced by the MHW (Fig. 5a; Table S1), averaging rates of 1.28 μmol N kg−1 h−1 (± 0.4 SE) compared to 2.98 (± 0.5 SE) across vegetation treatments. Vegetated sediments showed trends toward higher potential DNRA rates (averaging twice as high) than in unvegetated sediments across both temperature treatments, although it was not statistically significant. Denitrification rates were also significantly lower, almost half the rate, under the MHW treatment than under the control (Fig. 5b; Table S1). Anammox potential rates were not altered by the MHW (Fig. 5c; Table S1), which averaged 0.44 μmol N kg−1 h−1 (± 0.03 SE) across all treatments. Furthermore, the potential rate of anammox appeared to be more reduced by the MHW in unvegetated sediments (0.48 μmol N kg−1 h−1 ± 0.07 SE down to 0.28 ± 0.1 SE) compared to vegetated sediments, which had a more stable anammox rate across temperature treatments, although these results were not statistically significant. The total sediment potential N loss (the sum of denitrification and anammox) was significantly impacted by the MHW, with the highest rates of N loss at control temperature 10.9 μmol N kg−1 h−1 (± 1.0 SE) and lower at the MHW treatment 6.21 (± 0.6 SE) (Table S1). By contrast, nitrification rates were significantly increased by the MHW temperature treatment (Fig. 5d; Table S1), with rates double that of the control (2.93 μmol N kg−1 h−1 ± 0.3 SE vs 1.46 ± 0.3 SE, respectively).

Fig. 5.

Fig. 5

Open in a new tab

Sediment maximum potential N utilization rates in (a) DNRA, (b) denitrification, (c) anammox, as well as rates of (d) nitrification. Boxplots represent median, upper, and lower quartiles, with the diamonds showing the mean values and individual points showing all measurements. Different letters indicate statistical differences from ANOVA main effects tests (‘Treatment’ and ‘Vegetation’). Graphs without letters revealed no statistical differences between any treatments. MHW, marine heatwave.

Microbial gene copies involved in sediment N cycling

Both denitrification genes examined showed a significant reduction in copy number under the MHW treatment (Fig. 6a; Table S1). The gene nirS, which encodes the enzyme nitrite reductase, was significantly lowered by the MHW, with gene copies averaging from c. 7 million copies (± 1 million SE) down to 125 000 copies (± 37 000 SE) kg−1 sediment. Additionally, vegetated sediments had double the amount, c. 4.2 million (± 1 million SE) of nirS gene copies compared to unvegetated sediment (c. 2.7 million ±1 million SE; Fig. 6a; Table S1). The gene encoding nitrous oxide reductase (nosZ) was also highly reduced under the MHW treatment, which halved in abundance (Fig. 6b; Table S1).

Fig. 6.

Fig. 6

Open in a new tab

Gene copies in the sediment encoding enzymes involved in denitrification, specifically (a) nirS in nitrite reduction and (b) nosZ in nitrous oxide reduction. Boxplots represent median, upper, and lower quartiles, with the diamonds showing the mean values and individual points showing all measurements. Different letters indicate statistical differences from ANOVA main effects tests (‘Treatment’ and ‘Vegetation’). MHW, marine heatwave.

Discussion

In this study, we assessed how ecosystem functions, specifically N cycling in seagrass meadows, can be affected by an extreme and prolonged MHW event. Our experiment revealed that the future‐scenario MHW impacted maximum potential N transformation rates in each system compartment of the tropical seagrass habitat, with variable seagrass growth and metabolism, as well as higher ammonium assimilation into the seagrass, in accordance with our hypothesis that N uptake would increase. Contrary to our hypotheses, the simulated MHW diminished maximum potential N cycling rates in the sediment, specifically denitrification, which is likely due to a reduction of these N cycling functional gene counts, resulting in a reduction in the net removal of reactive N. However, we also observed that the changes to sediment N cycling processes from the MHW do not occur in the same direction, such as increased nitrification rates (although this could not be measured accurately) in contrast to the decreased denitrification and DNRA potential rates. Given the importance of seagrass meadows in terms of their potential functions and services, alterations to N cycling functionality, including anthropogenic N net removal, can have considerable ramifications for entire ecosystems.

Seagrass physiology and nutrient responses to MHWs

The responses of seagrass physiological performance to MHWs have been widely explored in the past several years (e.g. Saha et al., 2020; Deguette et al., 2022; Bass & Falkenberg, 2023), with changes in its survival, growth, and performance expected to have impacts on biogeochemical cycling. It has been shown that MHW impacts on the growth and performance of individuals are dependent on a multitude of variables, such as past exposure, geographical range, and internal energy reserves (Smith et al., 2023). As we considered an extreme MHW treatment of a temperature considerably higher than the average temperatures experienced during this time of year (c. 30°C), the 20% reduction in survival (albeit nonsignificant) was expected – a result that has been recorded many times from MHWs (Smale et al., 2019). However, considering the seagrass tank mortality was 40% in control tanks, we expect that the naturally variable growth and turnover rates of H. ovalis may have contributed to these losses (Rasheed, 2004; Rasheed et al., 2014). We also observed that the surviving seagrass' response to the MHW in terms of growth and physiology was highly variable (Fig. 3), with some ramets faring poorly while others grew very well. This result potentially suggests some level of thermal phenotypic plasticity. From these physiological results, it is expected that if this magnitude and duration of MHW is experienced in the natural environment, it could lead to very different responses among the individual ramets, potentially leading to patchy seagrass meadows.

The respiration rate of surviving seagrasses was significantly higher under the MHW treatment (Fig. 3). Such increases indicate elevated metabolism, which would enhance the N requirements of the plants (Alexandre et al., 2020). We observed a significant boost in the seagrass aboveground compartment's assimilation rate of ammonium and, although not statistically significant, a slight natural increase in seagrass N% in above and belowground compartments under the MHW (Fig. 4). These results together signify that more N is retained in the seagrass, which may become more evident over longer exposure periods. These results show, therefore, that the MHW increases the N demands of seagrass yet simultaneously reduces their net metabolism. This lowered metabolic activity was also revealed by the lowered seagrass δ13C with the MHW for both the above‐ and belowground compartments, which can happen when carbon requirements decrease and there is higher discrimination against 13C under stressful conditions (Vizzini et al., 2010). Consequently, lowered net metabolism can reduce the availability of organic C to support N transformations (Egea et al., 2023). MHWs appear to be becoming more frequent around Hong Kong coastal waters and are having deleterious effects on marine organisms (Oh et al., 2023; Zhao et al., 2023; Bass & Falkenberg, 2024). While there have currently been no recorded in situ observations of MHW impacts on the seagrass meadows in Hong Kong, these results indicate that future scenario MHWs could restrict the productivity of seagrasses that do persist.

MHW impacts on N transfer between seagrass meadow system compartments

Nutrient fluxes within seagrass meadows are inherently dynamic, with continuous losses and gains between compartments (Hemminga et al., 1991). In this study, we observed that N transformations between some seagrass meadow compartments (i.e. the water, seagrass, and sediment) were significantly lower under the MHW treatment (summarized in Fig. 7), which contrasts with our hypothesis that these processes would be enhanced under this condition. With the MHW, maximum potential N cycling within the sediment was lowered for both DNRA and denitrification. However, while the maximum potential rates of denitrification decreased the most under MHW (down from c. 12 to 6 μmol kg−1 h−1), this reduced rate of denitrification remained slightly higher than DNRA rates overall (Fig. 5). These rates are comparable to N cycling rates under lower enrichment concentrations, with net denitrification and DNRA in tropical seagrass meadows estimated at 8 and 1 μmol N m−2 h−1, respectively (Chen et al., 2021). Moreover, the ratio of denitrification to DNRA was elevated under the MHW treatment (from c. 3.5 : 1 to 5 : 1). Therefore, while the absolute value of these fluxes was diminished under the MHW, the sediment was characterized by a stronger dominance of denitrification than of DNRA during the MHW. As a result, the net N removal service provided by the sediment still exists under the MHW, albeit significantly reduced.

Fig. 7.

Fig. 7

Open in a new tab

Summary of the average flux of maximum potential rates of seagrass leaf N uptake (ammonium assimilation and N2 fixation), sediment N loss (denitrification and anammox), and N retention (DNRA and nitrification) in the experimental tanks. Arrows show the direction of the N movement in the different seagrass meadow system compartments, with the size of the arrows indicating the magnitude of the flux. Rates are in μmol N h−1, standardized by the approximate wet weight of the sediment and seagrass in each tank (1.15 kg of sediment and 0.112 g seagrass, from unpublished data). Error values indicate the standard deviation between treatment replicates. MHW, marine heatwave.

The abundance of microbes with genes encoding enzymes that perform N transformations can greatly influence rates of N cycling (Wallenstein & Vilgalys, 2005; Mattoo & Suman, 2023). Therefore, the reduction in sediment N cycle functioning under the MHW we observed may have occurred, in part, due to the reduced microbial abundance, or reduced gene count, from the MHW. Indeed, two of the genes involved in denitrification measured were in agreement that microbial abundance is correlated to N cycling rates (Fig. 6). While temperature would usually increase metabolic transformations, such as those performed by microbes (shown previously in Lai et al., 2021), other studies on ocean warming effects have also shown that DNRA and denitrification rates were markedly lowered due to C losses through increased benthic respiration (Brin et al., 2015). We believe, therefore, that the higher respiration rates from the increased temperature may be marginally contributing to the lowered DNRA rates, as well as lowered microbial processes. We show here that MHWs can alter the abundance of important sediment N cycling genes, which could influence ecosystem biogeochemical cycling.

The overlying water nutrient levels can be influenced directly by bacteria in the water column, as well as indirectly through various processes in the sediment and in the vegetation (Bianucci et al., 2012; Herbert, 1999). While not directly measured in this experiment, the considerably lower nitrate concentrations during MHW (Fig. 2) may have arisen from higher N uptake by microphytobenthos with increased temperature from the MHW treatment (Vieira et al., 2013) and, for the vegetated treatments, the higher N uptake by seagrass under the MHW. The higher ratio of denitrification to DNRA may also have favored the loss of N from the system rather than internal recycling. This lowered water column nitrate may indicate an increase in the N uptake potential of the seagrass during the MHW. The markedly lower total alkalinity (i.e. more H+ or less HCO3 −; Dickson, 1981; Wolf‐Gladrow et al., 2007), particularly under the MHW, is likely to have arisen from the sediment nitrification (H+ release), as well as organic matter degradation, both of which are expected to be enhanced with warming (von Lützow & Kögel‐Knabner, 2009). As total alkalinity is related to the amount of protons that can be accepted (Middelburg et al., 2020), the lowered total alkalinity in the MHW treatment therefore suggests that there is a reduced buffering capacity of the seagrass meadow overlying water to ocean acidification.

MHW impacts on seagrass meadow net N removal capacity

Quantifying nitrogen dynamics in ecosystem‐associated coastal sediments can reveal the potential ecosystem services provided, including net N removal capabilities, and whether it is a net source or sink of N (McGlathery et al., 2004). In general, studies show that seagrasses are beneficial for the environment, in that they can facilitate N removal from the water by storing it as plant biomass (McGlathery et al., 2007; Reynolds et al., 2016). In our study, the rates of N fixation in seagrass were low (c. 0.09 μmol g−1 DW h−1), with other records of N fixation in H. ovalis and other species of the same genus recorded to have natural rates of 12–18 μmol N2 m−2 h−1 (Cardini et al., 2018; Carlson‐Perret et al., 2019). However, similarly low rates have been recorded in other species, such as Zostera beds (Cook et al., 2015). Along with the low rates of fixation, our results could not confirm that the highest rates of fixation occur in the belowground compartments of the seagrass. Higher rates of N fixation in belowground compartments are primarily due to the presence of diazotroph N‐fixing bacteria associated with the seagrass roots (Welsh, 2000). Moreover, the lack of epiphytic algae (which also perform N fixation; Marino et al., 2023) could have also contributed to the lowered N fixation rate. While the experimental design included homogenizing the sediment before the 4‐wk experiment, and could have therefore influenced the root‐associated microbiome and therefore diazotroph activity, previous studies have shown that recovery of the seagrass root and sediment microbiome can be observed after 2 or 3 wk (Galand et al., 2016; Wang et al., 2021). Nevertheless, we did confirm that the seagrass leaves are the main site for assimilation, with significantly higher rates of ammonium assimilation than the belowground compartments. These rates are similar to those of other marine macrophytes (Smart et al., 2022) and also agree with previous studies showing that NH4 + uptake continues to increase with temperature (Alexandre et al., 2020), although this was only applicable for the aboveground compartments. By contrast, previous studies have revealed that MHWs reduce NO3 −‐assimilation rates (Bass et al., 2024). We potentially see that the effects of the MHW on uptake rates are different with ammonium compared to nitrate because seagrasses have a higher affinity for ammonium (Touchette & Burkholder, 2000; Nayar et al., 2018). Overall, although we do see diminished long‐term N removal of the overall seagrass meadow along with the slightly increased mortality rates, the significantly higher assimilation of ammonium shows that those that do persist can retain their beneficial function in net N removal.

The presence of seagrass appeared to have little impact on the net N removal capacity of the underlying sediment, as there was very little difference between the vegetated and unvegetated sediment N transformation rates (Fig. 5), contrary to our hypothesis that there would be differences in N loss due to benthic respiration. This may be due to the homogenization of the sediment before the experiment, by which we lost the structural and organic matter differences between vegetated and unvegetated sediment that may mediate some of these processes, which may not have been recovered in the 30‐d experiment. However, our results might also be typical of pioneer or quickly colonizing species (Olesen et al., 2004), such as the seagrass used here – Halophila ovalis – which has a simplified root system and only penetrates a few centimeters into the sediment. Therefore, benthic respiration may not be very different between vegetated and unvegetated sediments in these circumstances. Despite this lack of significance, there did appear to be a slight buffering capacity of the vegetated sediment on N cycling rates, such that the average potential transformation rates, particularly denitrification and anammox (the two processes contributing to total N loss), had a larger rate difference between the control and the MHW treatments in unvegetated sediments, while a smaller rate change was observed in vegetated sediments with the MHW. These results suggest that the seagrass meadow, in comparison to bare sediment, may partially retain contemporary N cycling rates in adverse climatic conditions, such as MHWs.

Conclusion

Examining the relationship between N cycling rates and extreme temperature anomalies helps us predict the impacts of future MHW events on coastal biogeochemistry. We show here that extended and extreme MHWs can impact seagrass ecosystem functions and services in terms of nutrient cycling and ocean acidification buffering capacity. Although the short‐term N sink through uptake into the seagrass, specifically ammonium assimilation, increased during the MHW due to increased N demands, we saw significant reductions in denitrification rates in the sediment (lowering permanent N lost from the ecosystem), as well as reduced water column nutrients and seagrass net metabolism. As such, MHWs can potentially alter the N cycling rates in seagrass meadows, leading to them providing lowered net N removal ecosystem services. Future research is needed to elucidate whether the cause of this N loss was directly due to microbial loss or stress from the reduction of gene copy numbers observed in this study. With the addition of field studies and laboratory experiments on other seagrass species, such as those with longer lifespans and more complex belowground root structures, a clearer picture of the future effects of MHWs on seagrass meadows can be constructed. Notably, we will be better able to understand the extent to which patterns are repeated across studies, allowing for the assessment of the generality of our conclusions. This study, therefore, contributes to the growing body of research aimed at developing our understanding of how MHWs can affect coastal ecosystems and their functions.

Competing interests

None declared.

Author contributions

AVB, ZW, LJF and BT planned and designed the research. AVB performed the experiments with help from ZW, NMC and MWKS. AVB analyzed the data and wrote the manuscript with input from all authors.

Disclaimer

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.

Supporting information

Fig. S1 Starting conditions of the seagrass ramets in the experimental tanks.

Fig. S2 The experimental setup, consisting of four water baths (blue rectangles) arranged randomly on two shelf levels.

Fig. S3 Average temperatures (n = 2 loggers per treatment) experienced in the experimental tanks during the 30‐d experiment.

Fig. S4 Schematic diagram of all the measurements taken after the 30‐d experiment.

Notes S1 Details of calculations for isotope values to rate.

Notes S2 Quantitative PCR additional information.

Table S1 ANOVA results.

Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

NPH-247-2616-s001.pdf (1,013.4KB, pdf)

Acknowledgements

This work was partially supported by the Areas of Excellence Scheme; Research Grants Council of Hong Kong Special Administrative Region, China (Project Reference No. AoE/P‐601‐23 N, Earth‐HK). AVB was financially supported by the Hong Kong PhD Fellowship Scheme from the Research Grants Council of Hong Kong, and ZW and MNC were supported by a postgraduate studentship provided by The Chinese University of Hong Kong. The authors would like to thank the Agriculture, Fisheries and Conservation Department of Hong Kong for approval of seagrass collection for this experiment, and the Stable Isotope Laboratory at Hong Kong University and Tim Sik Chan for seagrass stable isotope and nitrification isotope analysis, respectively. The authors also extend their thanks to Dr Xianbiao Lin at the Ocean University of China for the MIMS analysis. The authors also thank the three anonymous reviewers for their constructive feedback.

Contributor Information

Alissa V. Bass, Email: alissabass@link.cuhk.edu.hk.

Benoit Thibodeau, Email: benoit.thibodeau@cuhk.edu.hk.

Data availability

All data can be found on the CUHK Research Data Repository (https://doi.org/10.48668/L2X6GO).

References

  1. AFCD . 2023. Seagrasses in Hong Kong: distribution . [WWW document] URL https://www.afcd.gov.hk/English/conservation/con_wet/con_wet_sea/con_wet_sea_dis/con_wet_sea_dis.html [accessed 19 June 2024].
  2. Alexandre A, Quintã R, Hill PW, Jones DL, Santos R. 2020. Ocean warming increases the nitrogen demand and the uptake of organic nitrogen of the globally distributed seagrass Zostera marina . Functional Ecology 34: 1325–1335. [Google Scholar]
  3. An S, Gardner WS. 2002. Dissimilatory nitrate reduction to ammonium (DNRA) as a nitrogen link, versus denitrification as a sink in a shallow estuary (Laguna Madre/Baffin Bay, Texas). Marine Ecology Progress Series 237: 41–50. [Google Scholar]
  4. Aoki LR, McGlathery KJ. 2018. Restoration enhances denitrification and DNRA in subsurface sediments of Zostera marina seagrass meadows. Marine Ecology Progress Series 602: 87–102. [Google Scholar]
  5. Aoki LR, McGlathery KJ, Oreska MPJ. 2020. Seagrass restoration reestablishes the coastal nitrogen filter through enhanced burial. Limnology and Oceanography 65: 1–12.32801395 [Google Scholar]
  6. Barbier EB, Hacker SD, Kennedy C, Koch EW, Stier AC, Silliman BR. 2011. The value of estuarine and coastal ecosystem services. Ecological Monographs 81: 169–193. [Google Scholar]
  7. Bass AV, Falkenberg LJ. 2023. Two tropical seagrass species show differing indicators of resistance to a marine heatwave. Ecology and Evolution 13: e10304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bass AV, Falkenberg LJ. 2024. Seasonal effects and trophic pressure shape the responses of species interactions in a tropical seagrass meadow to marine heatwaves. Oikos 2024: e10382. [Google Scholar]
  9. Bass AV, Falkenberg LJ, Thibodeau B. 2024. Seagrasses under stress: independent negative effects of elevated temperature and light reduction at multiple levels of organization. Limnology and Oceanography 2: 12759. [Google Scholar]
  10. Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting linear mixed‐effects models using lme4 . Journal of Statistical Software 67: 1–48. [Google Scholar]
  11. Berlinghof J, Montilla LM, Peiffer F, Quero GM, Marzocchi U, Meador TB, Margiotta F, Abagnale M, Wild C, Cardini U. 2024. Accelerated nitrogen cycling on mediterranean seagrass leaves at volcanic CO2 vents. Communications Biology 7: 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bianucci L, Fennel K, Denman KL. 2012. Role of sediment denitrification in water column oxygen dynamics: comparison of the North American East and West Coasts. Biogeosciences 9: 2673–2682. [Google Scholar]
  13. Böhlke JK, Mroczkowski SJ, Coplen TB. 2003. Oxygen isotopes in nitrate: new reference materials for 18O : 17O : 16O measurements and observations on nitrate‐water equilibration. Rapid Communications in Mass Spectrometry 17: 1835–1846. [DOI] [PubMed] [Google Scholar]
  14. Brin LD, Giblin AE, Rich JJ. 2015. Effects of experimental warming and carbon addition on nitrate reduction and respiration in coastal sediments. Biogeochemistry 125: 81–95. [Google Scholar]
  15. Camillini N. 2020. Carbon and nitrogen cycling in seagrass ecosystems. [Doctoral dissertation]. East Lismore, NSW, Australia: Southern Cross University. [Google Scholar]
  16. Cardini U, Van Hoytema N, Bednarz VN, Al‐Rshaidat MMD, Wild C. 2018. N2 fixation and primary productivity in a red sea Halophila stipulacea meadow exposed to seasonality. Limnology and Oceanography 63: 786–798. [Google Scholar]
  17. Carlson‐Perret NL, Erler DV, Eyre BD. 2019. Comparison of dinitrogen fixation rates in two subtropical seagrass communities. Marine Chemistry 209: 62–69. [Google Scholar]
  18. Cerco CF. 1989. Measured and modelled effects of temperature, dissolved oxygen and nutrient concentration on sediment‐water nutrient exchange. Hydrobiologia 174: 185–194. [Google Scholar]
  19. Chen J‐J, Erler DV, Wells NS, Huang J, Welsh DT, Eyre BD. 2021. Denitrification, anammox, and dissimilatory nitrate reduction to ammonium across a mosaic of estuarine benthic habitats. Limnology and Oceanography 66: 1281–1297. [Google Scholar]
  20. Cook P, Evrard V, Woodland R. 2015. Factors controlling nitrogen fixation in temperate seagrass beds. Marine Ecology Progress Series 525: 41–51. [Google Scholar]
  21. Costanza R, d'Arge R, de Groot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, O'Neill RV, Paruelo J et al. 1997. The value of the world's ecosystem services and natural capital. Nature 387: 6630. [Google Scholar]
  22. Dalsgaard T, Thamdrup B. 2002. Factors controlling anaerobic ammonium oxidation with nitrite in marine sediments. Applied and Environmental Microbiology 68: 3802–3808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Deguette A, Barrote I, Silva J. 2022. Physiological and morphological effects of a marine heatwave on the seagrass Cymodocea nodosa . Scientific Reports 12: 7950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Dickson AG. 1981. An exact definition of total alkalinity and a procedure for the estimation of alkalinity and total inorganic carbon from titration data. Deep Sea Research Part A. Oceanographic Research Papers 28: 609–623. [Google Scholar]
  25. Egea LG, Jiménez‐Ramos R, Romera‐Castillo C, Casal‐Porras I, Bonet‐Melià P, Yamuza‐Magdaleno A, Cerezo‐Sepúlveda L, Pérez‐Lloréns JL, Brun FG. 2023. Effect of marine heat waves on carbon metabolism, optical characterization, and bioavailability of dissolved organic carbon in coastal vegetated communities. Limnology and Oceanography 68: 467–482. [Google Scholar]
  26. Eyre BD, Ferguson AJP, Webb A, Maher D, Oakes JM. 2011. Denitrification, N‐fixation and nitrogen and phosphorus fluxes in different benthic habitats and their contribution to the nitrogen and phosphorus budgets of a shallow oligotrophic sub‐tropical coastal system (southern Moreton Bay, Australia). Biogeochemistry 102: 111–133. [Google Scholar]
  27. Ferrón S, Alonso‐Pérez F, Anfuso E, Murillo FJ, Ortega T, Castro CG, Forja JM. 2009. Benthic nutrient recycling on the northeastern shelf of the Gulf of Cádiz (SW Iberian Peninsula). Marine Ecology Progress Series 390: 79–95. [Google Scholar]
  28. Fox J, Weisberg S. 2011. An R companion to applied regression. Thousand Oaks, CA, USA: SAGE. [Google Scholar]
  29. Galand PE, Lucas S, Fagervold SK, Peru E, Pruski AM, Vétion G, Dupuy C, Guizien K. 2016. Disturbance increases microbial community diversity and production in marine sediments. Frontiers in Microbiology 7: 1950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. García‐Robledo E, Corzo A, Papaspyrou S. 2014. A fast and direct spectrophotometric method for the sequential determination of nitrate and nitrite at low concentrations in small volumes. Marine Chemistry 162: 30–36. [Google Scholar]
  31. Garcias‐Bonet N, Fusi M, Ali M, Shaw DR, Saikaly PE, Daffonchio D, Duarte CM. 2018. High denitrification and anaerobic ammonium oxidation contributes to net nitrogen loss in a seagrass ecosystem in the central Red Sea. Biogeosciences 15: 7333–7346. [Google Scholar]
  32. Gardner WS, McCarthy MJ, An S, Sobolev D, Sell KS, Brock D. 2006. Nitrogen fixation and dissimilatory nitrate reduction to ammonium (DNRA) support nitrogen dynamics in Texas estuaries. Limnology and Oceanography 51: 558–568. [Google Scholar]
  33. Gaylard S, Gabrynowicz S, Lavery P, Waycott M. 2023. Contribution of seagrass productivity to waste treatment in a highly oligotrophic urbanised coast. Ecosystem Services 62: 101534. [Google Scholar]
  34. Giblin AE, Tobias CR, Song B, Weston N, Banta GT, Rivera‐Monroy V. 2013. The importance of dissimilatory nitrate reduction to ammonium (DNRA) in the nitrogen cycle of coastal ecosystems. Oceanography 26: 124–131. [Google Scholar]
  35. Giner‐Sanz JJ, Leverick G, Pérez‐Herranz V, Shao‐Horn Y. 2021. Optimization of the salicylate method for ammonia quantification from nitrogen electroreduction. Journal of Electroanalytical Chemistry 896: 115250. [Google Scholar]
  36. Granger J, Sigman DM. 2009. Removal of nitrite with sulfamic acid for nitrate N and O isotope analysis with the denitrifier method. Rapid Communications in Mass Spectrometry 23: 3753–3762. [DOI] [PubMed] [Google Scholar]
  37. Hardison AK, Algar CK, Giblin AE, Rich JJ. 2015. Influence of organic carbon and nitrate loading on partitioning between dissimilatory nitrate reduction to ammonium (DNRA) and N2 production. Geochimica et Cosmochimica Acta 164: 146–160. [Google Scholar]
  38. Hemminga MA, Harrison PG, van Lent F. 1991. The balance of nutrient losses and gains in seagrass meadows. Marine Ecology Progress Series 71: 85–96. [Google Scholar]
  39. Herbert RA. 1999. Nitrogen cycling in coastal marine ecosystems. FEMS Microbiology Reviews 23: 563–590. [DOI] [PubMed] [Google Scholar]
  40. Hobday AJ, Alexander LV, Perkins SE, Smale DA, Straub SC, Oliver ECJ, Benthuysen JA, Burrows MT, Donat MG, Feng M et al. 2016. A hierarchical approach to defining marine heatwaves. Progress in Oceanography 141: 227–238. [Google Scholar]
  41. Holbrook NJ, Sen Gupta A, Oliver ECJ, Hobday AJ, Benthuysen JA, Scannell HA, Smale DA, Wernberg T. 2020. Keeping pace with marine heatwaves. Nature Reviews Earth and Environment 1: 482–493. [Google Scholar]
  42. Hutchins DA, Capone DG. 2022. The marine nitrogen cycle: new developments and global change. Nature Reviews Microbiology 20: 401–414. [DOI] [PubMed] [Google Scholar]
  43. Hutchins DA, Fu F. 2017. Microorganisms and ocean global change. Nature Microbiology 2: 1–11. [DOI] [PubMed] [Google Scholar]
  44. Koch EW, Barbier EB, Silliman BR, Reed DJ, Perillo GM, Hacker SD, Granek EF, Primavera JH, Muthiga N, Polasky S et al. 2009. Non‐linearity in ecosystem services: temporal and spatial variability in coastal protection. Frontiers in Ecology and the Environment 7: 29–37. [Google Scholar]
  45. Lai TV, Ryder MH, Rathjen JR, Bolan NS, Croxford AE, Denton MD. 2021. Dissimilatory nitrate reduction to ammonium increased with rising temperature. Biology and Fertility of Soils 57: 363–372. [Google Scholar]
  46. Lenth RV. 2016. Least‐squares means: the R package lsmeans . Journal of Statistical Software 69: 1–33. [Google Scholar]
  47. Lin X, Lu K, Hardison AK, Liu Z, Xu X, Gao D, Gong J, Gardner WS. 2021. Membrane inlet mass spectrometry method (REOX/MIMS) to measure 15N‐nitrate in isotope‐enrichment experiments. Ecological Indicators 126: 107639. [Google Scholar]
  48. von Lützow M, Kögel‐Knabner I. 2009. Temperature sensitivity of soil organic matter decomposition—what do we know? Biology and Fertility of Soils 46: 1–15. [Google Scholar]
  49. Marino R, Hayn M, Howarth RW, Giblin AE, McGlathery KJ, Berg P. 2023. Nitrogen fixation associated with epiphytes on the seagrass Zostera marina in a temperate lagoon with moderate to high nitrogen loads. Biogeochemistry 166: 211–226. [Google Scholar]
  50. Mattoo R, Suman BM. 2023. Microbial roles in the terrestrial and aquatic nitrogen cycle—implications in climate change. FEMS Microbiology Letters 370: fnad061. [DOI] [PubMed] [Google Scholar]
  51. McGlathery KJ, Sundbäck K, Anderson IC. 2004. The importance of primary producers for benthic nitrogen and phosphorus cycling. In: Nielsen SL, Banta GT, Pedersen MF, eds. Estuarine nutrient cycling: the influence of primary producers: the fate of nutrients and biomass. Dordrecht, The Netherlands: Springer, 231–261. [Google Scholar]
  52. McGlathery KJ, Sundbäck K, Anderson IC. 2007. Eutrophication in shallow coastal bays and lagoons: the role of plants in the coastal filter. Marine Ecology Progress Series 348: 1–18. [Google Scholar]
  53. Middelburg JJ, Soetaert K, Hagens M. 2020. Ocean alkalinity, buffering and biogeochemical processes. Reviews of Geophysics 58: e0681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Mohr W, Großkopf T, Wallace DWR, LaRoche J. 2010. Methodological underestimation of oceanic nitrogen fixation rates. PLoS ONE 5: e12583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Moulton OM, Altabet MA, Beman JM, Deegan LA, Lloret J, Lyons MK, Nelson JA, Pfister CA. 2016. Microbial associations with macrobiota in coastal ecosystems: patterns and implications for nitrogen cycling. Frontiers in Ecology and the Environment 14: 200–208. [Google Scholar]
  56. Nayar S, Loo MGK, Tanner JE, Longmore AR, Jenkins GP. 2018. Nitrogen acquisition and resource allocation strategies in temperate seagrass Zostera nigricaulis: uptake, assimilation and translocation processes. Scientific Reports 8: 17151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Ogilve B, Nedwell DB, Harrison RM, Robinson A, Sage A. 1997. High nitrate, muddy estuaries as nitrogen sinks: the nitrogen budget of the River Colne estuary (United Kingdom). Marine Ecology Progress Series 150: 217–228. [Google Scholar]
  58. Oh H, Kim G‐U, Chu J‐E, Lee K, Jeong J‐Y. 2023. The record‐breaking 2022 long‐lasting marine heatwaves in the East China Sea. Environmental Research Letters 18: 064015. [Google Scholar]
  59. Olesen B, Marba N, Duarte CM, Savela RS, Fortes MD. 2004. Recolonization dynamics in a mixed seagrass meadow: the role of clonal versus sexual processes. Estuaries 27: 770–780. [Google Scholar]
  60. Olivé I, García‐Robledo E, Silva J, Pintado‐Herrera MG, Santos R, Kamenos NA, Cuet P, Frouin P. 2022. Contribution of the seagrass Syringodium isoetifolium to the metabolic functioning of a tropical reef lagoon. Frontiers in Marine Science 9: 867986. [Google Scholar]
  61. Oliver ECJ, Donat MG, Burrows MT, Moore PJ, Smale DA, Alexander LV, Benthuysen JA, Feng M, Sen Gupta A, Hobday AJ et al. 2018. Longer and more frequent marine heatwaves over the past century. Nature Communications 9: 732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Ottosen LDM, Risgaard‐Petersen N, Nielsen LP. 1999. Direct and indirect measurements of nitrification and denitrification in the rhizosphere of aquatic macrophytes. Aquatic Microbial Ecology 19: 81–91. [Google Scholar]
  63. Pachiadaki MG, Sintes E, Bergauer K, Brown JM, Record NR, Swan BK, Mathyer ME, Hallam SJ, Lopez‐Garcia P, Takaki Y et al. 2017. Major role of nitrite‐oxidizing bacteria in dark ocean carbon fixation. Science 358: 1046–1051. [DOI] [PubMed] [Google Scholar]
  64. Rasheed MA. 2004. Recovery and succession in a multi‐species tropical seagrass meadow following experimental disturbance: the role of sexual and asexual reproduction. Journal of Experimental Marine Biology and Ecology 310: 13–45. [Google Scholar]
  65. Rasheed MA, McKenna SA, Carter AB, Coles RG. 2014. Contrasting recovery of shallow and deep water seagrass communities following climate associated losses in tropical north Queensland, Australia. Marine Pollution Bulletin 83: 491–499. [DOI] [PubMed] [Google Scholar]
  66. Reay DS, Dentener F, Smith P, Grace J, Feely RA. 2008. Global nitrogen deposition and carbon sinks. Nature Geoscience 1: 430–437. [Google Scholar]
  67. Reynolds LK, Waycott M, McGlathery KJ, Orth RJ. 2016. Ecosystem services returned through seagrass restoration. Restoration Ecology 24: 583–588. [Google Scholar]
  68. Rysgaard S, Risgaard‐Petersen N, Sloth NP. 1996. Nitrification, denitrification, and nitrate ammonification in sediments of two coastal lagoons in Southern France. In: Caumette P, Castel J, Herbert R, eds. Coastal lagoon eutrophication and anaerobic processes (C.L.E.AN.): nitrogen and sulfur cycles and population dynamics in coastal lagoons a research programme of the environment programme of the EC (DG XII). Dordrecht, The Netherlands: Springer, 133–141. [Google Scholar]
  69. Saha M, Barboza FR, Somerfield PJ, Al‐Janabi B, Beck M, Brakel J, Ito M, Pansch C, Nascimento‐Schulze JC, Jakobsson Thor S et al. 2020. Response of foundation macrophytes to near‐natural simulated marine heatwaves. Global Change Biology 26: 417–430. [DOI] [PubMed] [Google Scholar]
  70. Salk KR, Erler DV, Eyre BD, Carlson‐Perret N, Ostrom NE. 2017. Unexpectedly high degree of anammox and DNRA in seagrass sediments: description and application of a revised isotope pairing technique. Geochimica et Cosmochimica Acta 211: 64–78. [Google Scholar]
  71. Schindelin J, Arganda‐Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B et al. 2012. fiji: an open‐source platform for biological‐image analysis. Nature Methods 9: e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Seidel L, Broman E, Nilsson E, Ståhle M, Ketzer M, Pérez‐Martínez C, Turner S, Hylander S, Pinhassi J, Forsman A et al. 2023. Climate change‐related warming reduces thermal sensitivity and modifies metabolic activity of coastal benthic bacterial communities. The ISME Journal 17: 855–869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Sigman DM, Casciotti KL, Andreani M, Barford C, Galanter M, Böhlke JK. 2001. A bacterial method for the nitrogen isotopic analysis of nitrate in seawater and freshwater. Analytical Chemistry 73: 4145–4153. [DOI] [PubMed] [Google Scholar]
  74. Smale DA, Wernberg T, Oliver EC, Thomsen M, Harvey BP, Straub SC, Burrows MT, Alexander LV, Benthuysen JA, Donat MG. 2019. Marine heatwaves threaten global biodiversity and the provision of ecosystem services. Nature Climate Change 9: 306–312. [Google Scholar]
  75. Smart JN, Schmid M, Paine ER, Britton D, Revill A, Hurd CL. 2022. Seasonal ammonium uptake kinetics of four brown macroalgae: implications for use in integrated multi‐trophic aquaculture. Journal of Applied Phycology 34: 1693–1708. [Google Scholar]
  76. Smith KE, Burrows MT, Hobday AJ, King NG, Moore PJ, Sen Gupta A, Thomsen MS, Wernberg T, Smale DA. 2023. Biological impacts of marine heatwaves. Annual Review of Marine Science 15: 119–145. [DOI] [PubMed] [Google Scholar]
  77. Smyth AR, Thompson SP, Siporin KN, Gardner WS, McCarthy MJ, Piehler MF. 2013. Assessing nitrogen dynamics throughout the estuarine landscape. Estuaries and Coasts 36: 44–55. [Google Scholar]
  78. Song GD, Liu SM, Kuypers MMM, Lavik G. 2016. Application of the isotope pairing technique in sediments where anammox, denitrification, and dissimilatory nitrate reduction to ammonium coexist. Limnology and Oceanography: Methods 14: 801–815. [Google Scholar]
  79. Thamdrup B, Dalsgaard T. 2002. Production of N2 through anaerobic ammonium oxidation coupled to nitrate reduction in marine sediments. Applied and Environmental Microbiology 68: 1312–1318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Touchette BW, Burkholder JM. 2000. Review of nitrogen and phosphorus metabolism in seagrasses. Journal of Experimental Marine Biology and Ecology 250: 133–167. [DOI] [PubMed] [Google Scholar]
  81. Vieira S, Ribeiro L, da Silva JM, Cartaxana P. 2013. Effects of short‐term changes in sediment temperature on the photosynthesis of two intertidal microphytobenthos communities. Estuarine, Coastal and Shelf Science 119: 112–118. [Google Scholar]
  82. Vizzini S, Tomasello A, Maida GD, Pirrotta M, Mazzola A, Calvo S. 2010. Effect of explosive shallow hydrothermal vents on δ13C and growth performance in the seagrass Posidonia oceanica . Journal of Ecology 98: 1284–1291. [Google Scholar]
  83. Voss M, Baker A, Bange HW, Conley D, Cornell S, Deutsch B, Engel A, Ganeshram R, Garnier J, Heiskanen A‐S et al. 2011. Nitrogen processes in coastal and marine ecosystems. In: Bleeker A, Grizzetti B, Howard CM, Billen G, van Grinsven H, Erisman JW, Sutton MA, Grennfelt P, eds. The European nitrogen assessment: sources, effects and policy perspectives. Cambridge, UK: Cambridge University Press, 147–176. [Google Scholar]
  84. Wallenstein MD, Vilgalys RJ. 2005. Quantitative analyses of nitrogen cycling genes in soils. Pedobiologia 49: 665–672. [Google Scholar]
  85. Wang L, English MK, Tomas F, Mueller RS. 2021. Recovery and community succession of the Zostera marina rhizobiome after transplantation. Applied and Environmental Microbiology 87: e02326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  86. Wannicke N, Benavides M, Dalsgaard T, Dippner JW, Montoya JP, Voss M. 2018. New perspectives on nitrogen fixation measurements using 15N2 gas. Frontiers in Marine Science 5: 120. [Google Scholar]
  87. Welsh DT. 2000. Nitrogen fixation in seagrass meadows: regulation, plant–bacteria interactions and significance to primary productivity. Ecology Letters 3: 58–71. [Google Scholar]
  88. Wolf‐Gladrow DA, Zeebe RE, Klaas C, Körtzinger A, Dickson AG. 2007. Total alkalinity: the explicit conservative expression and its application to biogeochemical processes. Marine Chemistry 106: 287–300. [Google Scholar]
  89. Wrightson L, Tagliabue A. 2020. Quantifying the impact of climate change on marine diazotrophy: insights from earth system models. Frontiers in Marine Science 7: 635. [Google Scholar]
  90. Yin G, Hou L, Liu M, Li X, Zheng Y, Gao J, Jiang X, Wang R, Yu C, Lin X. 2017. DNRA in intertidal sediments of the Yangtze Estuary. Journal of Geophysical Research: Biogeosciences 122: 1988–1998. [Google Scholar]
  91. Zhao Y, Chen M, Chung TH, Chan LL, Qiu J‐W. 2023. The 2022 summer marine heatwaves and coral bleaching in China's Greater Bay Area. Marine Environmental Research 189: 106044. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Fig. S1 Starting conditions of the seagrass ramets in the experimental tanks.

Fig. S2 The experimental setup, consisting of four water baths (blue rectangles) arranged randomly on two shelf levels.

Fig. S3 Average temperatures (n = 2 loggers per treatment) experienced in the experimental tanks during the 30‐d experiment.

Fig. S4 Schematic diagram of all the measurements taken after the 30‐d experiment.

Notes S1 Details of calculations for isotope values to rate.

Notes S2 Quantitative PCR additional information.

Table S1 ANOVA results.

Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

NPH-247-2616-s001.pdf (1,013.4KB, pdf)

Data Availability Statement

All data can be found on the CUHK Research Data Repository (https://doi.org/10.48668/L2X6GO).

Articles from The New Phytologist are provided here courtesy of Wiley

RESOURCES