Multiyear monitoring reveals seasonal and short-term dynamics of ecosystem metabolism in a temperate salt marsh channel

Multiyear monitoring reveals seasonal and short-term dynamics of ecosystem metabolism in a temperate salt marsh channel | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Multiyear monitoring reveals seasonal and short-term dynamics of ecosystem metabolism in a temperate salt marsh channel Emily J. Chua, John Supino, Kristen E. Fogaren, Hilary I. Palevsky This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6759348/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Climate change is expected to significantly alter carbon dynamics in coastal ecosystems like salt marshes. However, our capacity to detect and respond to these changes is limited by insufficient knowledge of baseline ecosystem metabolism across aquatic ecosystems. Quantifying metabolic rates—gross primary production (GPP), ecosystem respiration (ER), and net ecosystem metabolism (NEM)—reveals whether an ecosystem stores (net autotrophic) or releases (net heterotrophic) carbon. Few studies have monitored water quality at high frequency over full annual cycles in marsh-dominated estuaries, limiting estimates of seasonal metabolic variation. In this study, we examined the metabolic balance of a salt marsh channel in coastal New Jersey, an area heavily impacted by human activities and experiencing rapid sea level rise. Over three years, we monitored environmental parameters—salinity, temperature, dissolved oxygen, pH, turbidity, and chlorophyll a —at three sites within the channel. Using these data, we calculated metabolic rates and assessed relationships between environmental conditions and metabolism, including impacts of storm events. Results revealed an overall net heterotrophic system (mean NEM: -29.4 mmol O 2 m − 2 d − 1 ), with high seasonal variation ranging from slightly net autotrophic in winter/early spring (February maximum: 12.0 mmol O 2 m − 2 d − 1 ) to net heterotrophic in late spring/fall (September minimum: -88.0 mmol O 2 m − 2 d − 1 ). Temperature played a dominant role in metabolic dynamics, while storms temporarily intensified heterotrophic conditions. Our findings reveal seasonal patterns in productivity and respiration that affect carbon storage capacity and document natural ecosystem variability and responses to disturbances—insights critical for understanding how these vital blue carbon systems may respond to environmental changes. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Coastal waters are a critical component of the global carbon budget, with estuaries and their surrounding vegetation processing and cycling substantial amounts of carbon. In particular, intertidal salt marshes (and their tropical counterparts, mangroves) are among the most productive ecosystems on Earth, acting as key conduits for transferring inorganic carbon from the atmosphere to the ocean (Alongi, 2020 ). Salt marshes, which are found worldwide and are the dominant intertidal habitat along the U.S. East and Gulf coasts (Pennings & Bertness, 2001 ), provide not only a significant carbon sink but also essential ecosystem services: they support biodiversity, sustain fisheries, improve water quality by filtering pollutants, and protect shorelines from sea level rise and storms (Barbier et al., 2011 ). Salt marshes are increasingly vulnerable to a range of threats, complicating mitigation and restoration efforts. Climate-driven pressures, such as sea-level rise, altered hydrology, and more extreme weather events, can destabilize sediments and cause widespread loss of vegetation, degrading the marsh (Rolando et al., 2023 ; Valiela et al., 2018 ). These climatic impacts are compounded by human-induced stressors like coastal development, pollution, and changes in sediment supply (Gedan et al., 2009 ). Migration and loss of marshes can release large amounts of stored carbon to the atmosphere, shifting salt marshes from being net carbon sinks to carbon sources (Warnell et al., 2022 ). Understanding and managing these dynamic systems requires process-based indicators that can serve as early warnings of ecosystem functional changes. Aquatic ecosystem metabolism offers such a process-based indicator. First conceptualized by Odum ( 1956 ), aquatic ecosystem metabolism integrates all biological activity—organic matter production and consumption—within a body of water. Gross primary production (GPP) represents the total rate of organic matter produced in an ecosystem, while ecosystem respiration (ER), also known as community respiration, captures the total rate of organic matter consumed in an ecosystem via aerobic respiration. Their balance, termed net ecosystem metabolism (NEM), reflects the metabolic status of an aquatic system—whether it is a net source or sink of organic carbon. These metrics provide information on the overall carbon balance of an ecosystem, and how it responds to changes that perturb this balance. Historically, assessing the response of aquatic ecosystems to environmental change has primarily occurred through discrete measurements of dissolved oxygen (DO) concentrations (Jankowski et al., 2021 ). DO dynamics are influenced by both physical processes (gas exchange, water mixing) and biological processes (primary production and respiration). Regulatory assessments often rely on comparing these discrete measurements to threshold values, such as minimum levels to sustain aquatic life. However, advances in autonomous oxygen sensors in recent decades have enabled high-frequency DO data collection, which, when combined with high-frequency temperature and salinity measurements, enable continuous estimates of metabolic rates. These high-temporal resolution metabolism data provide unprecedented insight into short-term variability and long-term trends in aquatic system functioning. Ecosystem metabolism is thought to respond to a wide range of anthropogenic disturbances, such as altered water column mixing due to climatic changes and increased suspended sediment loads—however, the ecological consequences of these perturbations are not yet well understood (Jankowski et al., 2021 ; Kemp & Testa, 2012 ). While small changes in physicochemical parameters might not be detectable on their own, when translated into metabolic rates these subtle shifts can reveal significant functional changes in the ecosystem (Jankowski et al., 2021 ). For example, metabolic rates calculated along stressor gradients—such as urbanization or eutrophication—can identify shifts in ecosystem function. Similarly, time series analysis of metabolic rates can reveal responses to disturbance events and track recovery or degradation. Long-term metabolism records are particularly valuable for distinguishing natural variability from climate-driven trends. In salt marsh systems, carbon dynamics are especially complex due to their heterogeneous features and tight coupling of terrestrial and aquatic processes. Salt marshes are dynamic, interconnected mosaics of carbon sources and sinks: while marsh vegetation sequesters atmospheric carbon dioxide (CO 2 ) and stores “blue carbon” in their sediments, the adjacent aquatic environments—tidal channels and estuaries—often act as net CO 2 sources (Cai, 2011 ; Dai et al., 2022 ; Laruelle et al., 2010 ; Rosentreter et al., 2023 ). This paradox is explained by the “marsh CO 2 pump” concept, wherein CO 2 fixed by marsh plants is later respired in adjacent waters, and either directly emitted back to the atmosphere or exported as dissolved inorganic carbon (DIC) to the coastal ocean (Z. A. Wang et al., 2016 ; Z. A. Wang & Cai, 2004 ). The strength of this pump varies seasonally, with peak CO 2 uptake by marsh plants in spring and early summer, followed by greater carbon export from the marsh system in late summer and fall. DIC exported to, and ultimately sequestered in, the deep ocean could surpass the amount stored in marsh sediments, suggesting that the marshes’ true carbon sink capacity may be underestimated (Z. A. Wang et al., 2016 ). The spatial and temporal variability of marsh carbon fluxes contributes to persistent underrepresentation of estuarine environments in Earth system models, which struggle to capture the full complexity of coastal ocean dynamics and their implications for regional and global climate projections (Bauer et al., 2013 ; Dai et al., 2022 ; Herrmann et al., 2015 ). Several synthesis studies have sought to refine carbon budgets for estuaries and the coastal ocean by integrating estimates of ecosystem metabolism. One such global synthesis by Bauer et al. ( 2013 ) identified significant uncertainty in estuarine net carbon balances due to their spatial and temporal heterogeneity in carbon processing and fluxes, as well as the difficulty in scaling up relatively few observational studies. Another synthesis by Herrmann et al. ( 2015 ), focused on the East Coast of the U.S., found that regional estuarine waters are generally net heterotrophic but identified NEM in the Mid Atlantic Bight as the most uncertain term in their budget, emphasizing the need for further data in this region. Environmental change, driven by climate or human activity, will likely shift the balance between primary production and respiration in coastal ecosystems (Kemp & Testa, 2012 ). Climate change, including regionally rising temperatures, changes in precipitation patterns, and increased storm frequency, is already underway (IPCC, 2023 ), while human-induced eutrophication (over-enrichment with nutrients or organic matter) has resulted in more frequent and intense hypoxic conditions in estuaries worldwide (Dai et al., 2023 ). For instance, warming tends to enhance respiration more than primary production in marine environments (Boscolo-Galazzo et al., 2018 ), shifting the system toward heterotrophy, which leads to greater dissolution of metabolic CO 2 and increases acidification risk (Baumann & Smith, 2018 ). Furthermore, few studies (e.g., Buelo et al., 2024 ; Howarth et al., 2000 ; Tassone & Bukaveckas, 2019 ) have captured how short-term (daily to weekly) events, such as storms or changes in freshwater discharge, affect estuarine ecosystem metabolism. Gaining a clearer understanding of the current metabolic balance in estuaries, and the factors that influence it, is essential for predicting how carbon dynamics in these systems will respond to future environmental changes. To address these gaps and improve our understanding of the metabolic balance in temperate salt marshes, we conducted a multi-year, high-frequency monitoring campaign in a tidal marsh channel in southern New Jersey. Over a three-year period, we collected continuous measurements of key physical and biogeochemical parameters—temperature, salinity, dissolved oxygen, pH, chlorophyll a , and turbidity—at three sites in the marsh channel. Our study had three primary objectives: (1) to quantify aquatic ecosystem metabolism in the channel and characterize seasonal dynamics of environmental conditions and metabolism, establishing baseline conditions; (2) to explore associations between environmental factors and metabolic rates over seasonal and annual timescales; and (3) to assess the metabolic impacts of perturbations from episodic storm events. We also discuss the sources of uncertainty in our metabolic rate estimates and compare our results to other marsh-dominated estuaries. This work contributes valuable high-resolution data from the mid-Atlantic U.S. and improves our understanding of how salt marsh systems respond to environmental change, which may inform future management strategies for marsh conservation and restoration in the face of climate change. Methods Study Site Continuous water quality monitoring was conducted in the marsh tidal channel system landward of Seven Mile Island, a populated barrier island located in Cape May County on the southern coast of New Jersey, from June 2021–June 2024 (Fig. 1 ). This expanse of salt marsh and tidal channels is part of the Seven Mile Island Innovation Laboratory (SMIIL), a designated research area that spans the coastal region from Townsends Inlet in Avalon to Hereford Inlet in Stone Harbor, bisected by the New Jersey Intracoastal Waterway. SMIIL was established in 2019 through a partnership between the U.S. Army Corps of Engineers (USACE) Philadelphia District, the USACE Engineer Research and Development Center (ERDC), the State of New Jersey, and The Wetlands Institute (TWI) to advance and improve dredging and marsh restoration techniques (Chasten et al., 2023 ; Fall et al., 2021 ). The area encompasses over 62 km 2 of state-owned marshland, including tidal marshes, coastal lagoons, shallow bays, sounds, and tidal inlets, and is part of the Cape May Wetlands Wildlife Management Area. Average low tide depths are 0.6 m (Perkey et al., 2024 ), and tidal conditions are mixed semi-diurnal with a tidal range of approximately 1–2 m (Fall et al., 2021 ). No rivers or streams drain into the region. Monitoring Data In June 2021, three observational “open-water” platforms were constructed in the main channel of SMIIL to facilitate water quality monitoring efforts (Fig. 1 ): North platform near the channel entrance (39.103 o N, 74.765 o W), Gull platform off the southern edge of Gull Island (39.072 o N, 74.778 o W), and South platform near the channel exit (39.044 o N, 74.788 o W). These platforms are henceforth referred to as “North,” “Gull,” and “South.” All three platform sites are shallow, with mean water depths of 1.26 m (North), 1.18 m (Gull), and 2.49 m (South; Table 1 ). At each platform, two Aqua TROLL 600 multiparameter sondes (In-Situ Inc.) were mounted on a pole attached to the platform, approximately 37–75 cm off the bottom. Both sondes deployed at a platform contained sensors for conductivity, temperature, dissolved oxygen, and pH; additionally, one sonde contained a chlorophyll a sensor and the other a turbidity sensor. Table 1 Summary of physical and chemical characteristics of the three channel sites over the entire time series (2021–2024). Reported errors are the measurement uncertainty (1 standard deviation). Seasonal range is defined as the 5th–95th percentile range per year, averaged across all years in the time series. Missing data points were gap-filled using a smoothed climatological day-of-year mean, derived from a 10-day moving average over the entire time series. Depth (m) Tidal range (m) Temperature ( o C) Salinity (psu) Dissolved oxygen (%) pH Mean Mean Mean Seas. range Mean Seas. range Mean Seas. range Mean Seas. range North 1.26 1.41 14.71 ± 0.04 4.84–25.23 32.0 ± 0.5 29.4–34.5 94.1 ± 1.6 74.4–104.5 8.11 ± 0.01 7.87–8.31 Gull 1.18 1.45 15.14 ± 0.03 3.96–26.69 32.4 ± 0.2 29.3–35.2 89.8 ± 1.1 65.9–104.0 8.02 ± 0.04 7.71–8.27 South 2.49 1.41 14.45 ± 0.01 3.49–24.78 32.1 ± 0.2 29.8–35.0 92.6 ± 1.1 74.6–104.1 8.11 ± 0.01 7.88–8.32 From June 2021 to June 2024, the sondes collected near-continuous data at the three open-water platforms. Data were collected at 6-minute intervals prior to August 2022; after that, recording was set to 10-minute intervals. Initially, the Aqua TROLLs were configured for remote monitoring via telemetry; however, following issues with data dropouts in the first year of deployment, the sondes were switched to internal data logging. Throughout the three-year monitoring period, maintenance visits were conducted regularly to the platforms to download data from the Aqua TROLLs and replace them with cleaned and newly calibrated sondes. Sondes were maintained every 4–12 weeks, with an average deployment length of 9 weeks. Starting in September 2022, discrete surface water samples were obtained during maintenance visits for later laboratory analysis of dissolved oxygen (DO), dissolved inorganic carbon (DIC), and total alkalinity (TA). Discrete sampling occurred near the end of a deployment, just before a sonde was recovered from a platform, and after replacement with a re-calibrated sonde. Samples for DO were collected into volume-calibrated flasks and preserved for Winkler analysis following standard protocols (Langdon 2010 ); the bottle necks were sealed with water to further permit long-term storage (Zhang et al., 2002 ). DO discrete samples were titrated within a week of collection using a custom-built Winkler titrator with automated potentiometric end point detection (control software available via Nicholson et al., 2023 ). Samples for DIC and TA were collected into borosilicate glass bottles and preserved with saturated mercuric chloride following standard protocols (Dickson et al., 2007 ). Discrete samples for DIC and TA were analyzed using an AS-C6L DIC Analyzer and an AS-ALK2 TA Analyzer (Apollo SciTech Instruments), respectively. DIC and TA were measured from the same sample bottle, with DIC measurements made on the day the bottle was first opened and TA measurements made the same week. Data Quality Control Water quality time series data—including depth, temperature, salinity, DO concentration, pH, turbidity, and chlorophyll a —underwent a multi-step quality control process adapted from the U.S. Integrated Ocean Observing System (IOOS) Quality Assurance/Quality Control of Real-Time Oceanographic Data (QARTOD) recommendations (IOOS, 2018 ; Palevsky et al., 2023 ). Initial quality control involved human-in-the-loop (HITL) inspection of depth data, with annotations cross-referenced against field logs to flag periods of known sensor malfunction or when sensors were briefly out of water between deployments. Flagged points were removed from further analysis for all parameters. Automated tests followed, including gross range checks and spike detection using parameter-specific thresholds (Table S1 in Supplemental Information). Values exceeding these thresholds were flagged and excluded. The cleaned data were then binned to 10-minute bins to resolve irregular sampling times. A moving median test was subsequently applied to all parameters—excluding depth and temperature due to their strong diel oscillations—to identify and remove additional outliers (Leys et al., 2013 ). This test used a double median absolute deviation (MAD) approach to account for asymmetrical distributions, with data points exceeding ± 3 respective MADs from the moving median flagged and removed (Tan et al., 2022 ). For parameters with duplicate measurements (salinity, temperature, DO concentration, and pH), cleaned time series were further evaluated to produce a single final dataset per parameter. Differences between duplicates were assessed using absolute difference thresholds and visual inspection. For each deployment, the final dataset was chosen from either the more reliable sensor or the mean of both sensors, based on data completeness, agreement with adjacent deployments, and consistency with supporting data. For salinity, which was particularly noisy, data were first smoothed with a 24-hour moving median before selection. Comparisons between related parameters aided this assessment—for example, low pH values were cross-checked against DO data to identify potential hypoxic or anoxic events. Discrete samples were used to validate the sonde data: Winkler titrations for DO, and DIC and TA for pH. For pH validation, sonde data were converted to hydrogen ion concentration ([H+]) space, and [H+] values from DIC and TA measurements were calculated using the CO2SYSv3 MATLAB program (Lewis & Wallace, 1998 ; Sharp, 2023 ; Van Heuven et al., 2011 ), with propagated uncertainties following Orr et al. ( 2018 ). Discrete salinity samples were occasionally available for spot-checking. To prevent error propagation from pressure, salinity, and temperature measurements, DO concentrations were recalculated using the cleaned environmental variables and salinity- and pressure-compensation equations (Aanderaa, 2017). DO % saturation was then calculated by normalizing the measured DO concentration to the saturation DO concentration (DO % saturation = DO measured / DO sat ​ ×100%), where DO sat was computed as a function of temperature and salinity (Garcia & Gordon, 1992 ). For turbidity and chlorophyll a , each measured by a single sensor per platform, a final manual inspection was conducted after the moving median test. These optical sensors are especially vulnerable to signal drift and noise from biofilm accumulation, which obstructs light transmission through the water column and was not fully remedied by earlier quality control steps. Persistent, year-round biofouling was observed at all platforms, consistent with a recent study at the SMIIL site that reported impaired turbidity sensor performance (Perkey et al., 2024 ). Time series were reviewed on a deployment-by-deployment basis across platforms, and data affected by suspected biofouling were removed. Due to the extent of data loss, turbidity and chlorophyll a records are presented only in the Supplementary Information (Fig. S1e–f). To constrain measurement error in the Aqua TROLL data, we used the mean absolute difference between duplicate sensor readings—taken during representative deployments where both sensors were functioning—as an estimate of two standard deviations (2σ). This approach was applied to parameters for which duplicate sensors were deployed at a platform (i.e., salinity, temperature, DO, and pH). Averaged across all three platforms, the resulting standard deviations were 0.28 psu for salinity, 0.025 o C for temperature, 2.88 µmol L − 1 for DO, and 0.02 for pH. Although duplicate depth sensors were also deployed, the vertical offset between the sensors varied slightly from deployment to deployment (on the order of a few cm), precluding a precise estimate of the depth measurement error; we estimate this to be < 5 cm. For DO, we also performed linear regressions comparing the best-guess Aqua TROLL values to discrete Winkler data, using the standard error of the mean as an independent error estimate for each platform. In subsequent uncertainty analyses (see “Aquatic Ecosystem Metabolism”), we used the larger of the two DO error estimates (duplicate sensor or Winkler-based) for each platform: the duplicate sensor method yielded higher errors for North and Gull, while the Winkler comparison did for South. Meteorological Data Sources A ResponseON Weather Transmitter (R.M. Young Company) was also installed on the Gull platform to measure wind speed and direction, air temperature, relative humidity, and atmospheric pressure. Measurements were made at 6-minute intervals. The weather station record ended prematurely due to data transmission issues, and spans September 2021–May 2023. Air temperature and atmospheric pressure data were also obtained from a HOBO water level logger (Onset Computer Corp.) deployed at The Wetlands Institute, which is centrally located in SMIIL (Fig. 1 A). HOBO data were collected at 10-minute intervals from August 2022–June 2024. The air temperature and atmospheric pressure data from the Gull meteorological station and HOBO sensor were combined to produce a time series that spanned the open-water platform water quality measurement record. Similarly, to fill gaps in the Gull meteorological station wind speed record, the hourly 10-m u- and v- components of wind were obtained from the fifth generation European Centre for Medium-Range Weather Forecasts atmospheric reanalysis (ERA5; Hersbach et al., 2023 ) for the grid cell nearest to 39.04 o N, 74.79 o W and used to calculate the 10-m wind speed. Given the close agreement between the Gull wind speed measurements and the ERA5 data (Fig. S2 in Supplemental Information), the latter was used in subsequent analyses across the entire water quality measurement record. To constrain measurement error for the meteorological data, we determined 2σ from comparisons between two measurement sources, similar to the approach used for the water quality monitoring data. For atmospheric pressure, we used the mean absolute difference between the Gull meteorological station and HOBO sensor data over overlapping measurement periods, resulting in an uncertainty of 2.54 hPa. For wind speed, we used the mean absolute difference between the Gull meteorological station hourly means and ERA5 wind speeds, resulting in an uncertainty of 1.53 m s − 1 . Aquatic Ecosystem Metabolism The measured DO time series data were leveraged to estimate ecosystem metabolism rates in the main marsh channel following the “open-water” or “diel oxygen” method pioneered by Odum ( 1956 ). This method is predicated on the notion that during a daily cycle, three main processes affect the dissolved oxygen concentration of a water mass: photosynthetic production, respiration, and exchange of oxygen across the air-water interface. Metabolic rates are related to rates of dissolved oxygen change via the mass balance equation, $$\:\begin{array}{c}\frac{\partial\:\text{D}\text{O}}{\partial\:t}=P-R+D\:\#\left(1\right)\end{array}$$ where ∂ DO /∂t is the rate of change of DO concentration with respect to time, P is the photosynthetic rate, R is the respiration rate, and D is the diffusive rate of oxygen exchange across the air-water interface, all expressed on a volumetric basis (mmol O 2 m − 3 h − 1 ). A main assumption of the diel oxygen method is that measured changes in dissolved oxygen arise from biological production/consumption and gas exchange, and the effects of physical transport on the DO at a fixed sampling location are negligible. However, in estuaries like SMIIL that are strongly influenced by tidal transport and mixing, the advection of water masses with different DO histories past the measurement point can influence DO measurements. To reduce the effects of physical transport on metabolism estimates, a weighted regression model developed by Beck et al. ( 2015 ) was applied to the measured DO data. This method targets the periodicity of the tidal component while preserving the biological signal and was implemented via the R package WtRegDO . The resulting “detided” oxygen data, DO detided was then used to calculate metabolic rates using Eq. 1, as described below. This approach substantially reduced the occurrence of anomalous GPP and ER results (i.e., negative GPP and positive ER) compared to calculations performed on the raw DO data. Remaining anomalous values may reflect cases where values fall below the detection limit for this method (Caffrey et al., 2014 ). Air-water oxygen exchange ( D in Eq. 1) was modeled as the product of a gas-exchange coefficient (aka volumetric aeration coefficient or piston velocity), k (h − 1 ) and the difference between the DO saturation concentration, DO sat and the measured detided DO concentration (both in mmol O 2 m − 3 ): $$\:\begin{array}{c}D=k\left({\text{D}\text{O}}_{\text{s}\text{a}\text{t}}-\:{\text{D}\text{O}}_{\text{d}\text{e}\text{t}\text{i}\text{d}\text{e}\text{d}}\right)\#\left(2\right)\end{array}$$ DO sat is the concentration of oxygen in water that is in equilibrium with the atmosphere, typically at sea level pressure, calculated as a function of water temperature and salinity (Garcia & Gordon, 1992 ). The gas-exchange coefficient is commonly estimated from statistical relationships with wind speed and/or tidal currents (Howard et al., 2018 ; Kemp & Testa, 2012 ; Needoba et al., 2012 ). Many efforts to parameterize k have been proposed (e.g., Cole and Caraco 1998 ; Emerson et al. 2019 ; Liang et al. 2013 ; Ro and Hunt 2006 ; Stanley et al. 2009 ; Vachon and Prairie 2013 ; Wanninkhof 2014 ), and the choice of an appropriate k parameterization depends on the specific environmental conditions. Howard et al. ( 2018 ) evaluated four methods within a salt-marsh environment; however, a comprehensive comparison of their impact on metabolic rate estimates does not yet exist. As such, we performed a sensitivity analysis using three parameterizations: Ro & Hunt ( 2006 ), Wanninkhof ( 2014 ), and Emerson et al. ( 2019 ). Ro & Hunt ( 2006 ) is commonly used in other studies of estuaries (Beck et al., 2015 ; Caffrey et al., 2014 ; Murrell et al., 2018 ; Roberts et al., 2022 ; Wallace et al., 2021 ). Wanninkhof ( 2014 ) was selected for its accuracy in predicting noble gas saturation in a salt marsh pond (Howard et al., 2018 ). As these first two parameterizations do not account for gas transfer through bubbles, we included Emerson et al. ( 2019 ), which explicitly accounts for bubble-mediated flux. We calculated gas-exchange coefficients and metabolic rates using these parameterizations. The mean absolute differences between the three methods were similar, with Wanninkhof ( 2014 ) yielding intermediate results, so it was chosen for our ecosystem metabolism analysis. In this parameterization, k is a function of wind speed at 10-m height, atmospheric pressure, air temperature, depth of the water column at the sampling site, and the Schmidt number; the latter is gas-specific and depends on the water temperature and to a lesser degree on the salinity. The ERA5 wind speeds and combined atmospheric pressure and air temperature data from the Gull meteorological station and HOBO sensor were used for the k calculations, which were implemented in MATLAB using the gas_toolbox package (Manning & Nicholson, 2022 ). This parameterization provides reliable estimates of the gas-exchange coefficient for intermediate wind speeds (3–15 m s − 1 ; Wanninkhof, 2014 ), with only 0.06% of our observed wind speeds exceeding this range and 12.6% falling below it. To constrain uncertainty in k , we used the mean absolute difference between the two parameterizations that produced the largest and smallest results (i.e., Emerson et al., 2019 and Ro & Hunt, 2006 ). The change in DO due to biological processes was calculated from Eq. 1, $$\:\begin{array}{c}\frac{\partial\:{\text{D}\text{O}}_{\text{b}\text{i}\text{o}}}{\partial\:t}=P-R=\frac{\partial\:{\text{D}\text{O}}_{\text{d}\text{e}\text{t}\text{i}\text{d}\text{e}\text{d}}}{\partial\:t}-D\:\#\left(3\right)\end{array}$$ and averaged separately during daylight periods to compute hourly rates of net (aka apparent) primary production ( P ), and during night periods to compute hourly rates of nighttime respiration ( R ). To calculate daily rates of respiration and production, respiration rates were assumed to remain constant during day and night. Daily total respiration, R t and daily gross production, P g (both in units of mmol O 2 m − 3 d − 1 ) were consequently calculated as: $$\:\begin{array}{c}{R}_{t}\:=\:R\:\times\:\:24\:hours\#\left(4\right)\end{array}$$ $$\:\begin{array}{c}{P}_{g}\:=\left(P-R\right)\times\:\:daylength\#\left(5\right)\end{array}$$ Daylength (in hours) was determined based on local sunrise and sunset times. These volumetric daily rates were multiplied by the daily mean water depth at the respective open-water platform to yield areal rates of gross primary production (GPP) and ecosystem respiration (ER) in mmol O 2 m − 2 d − 1 . Net ecosystem metabolism was then calculated as $$\:\begin{array}{c}NEM\:=\:GPP\:+\:ER\#\left(6\right)\end{array}$$ NEM reflects the overall balance between the production and respiration of organic matter by all organisms within the ecosystem. A positive NEM indicates the system is net autotrophic, producing excess organic carbon, while a negative NEM suggests the system is net heterotrophic, consuming organic carbon. When NEM approaches zero, the system is considered to be in metabolic balance. The uncertainties in each measured parameter used in the diel oxygen analysis contribute to uncertainty in each term of the mass balance (Eq. 1), and therefore the overall uncertainty in the calculated values for the metabolic rates. Uncertainties from individual parameters were propagated through the diel analysis via a Monte Carlo technique (e.g., Albert, 2020 ) using the mean standard deviation for each input parameter. First, the uncertainty in the air-sea gas exchange (Eq. 2) was estimated via a Monte Carlo simulation iterated 10 4 times using the uncertainties in salinity, temperature, depth (or pressure), DO concentration, wind speed, atmospheric pressure, and choice of k parameterization. Next, the uncertainties in the hourly rates of net primary production and nighttime respiration were calculated via a Monte Carlo simulation iterated 10 4 times containing the uncertainty in D and converted to uncertainties in the daily rates, R t and P g (Equations 4 and 5). The uncertainties in R t , P g , and depth were propagated in a final Monte Carlo simulation iterated 10 4 times to calculate GPP, ER, and NEM (Eq. 6). Determining metabolic fluxes relies on accurately estimating air-water dissolved oxygen exchange driven by wind mixing, especially in shallow systems. However, gas-exchange parameterizations are site-specific and fluxes calculated using different methods can range over an order of magnitude, often making gas exchange flux the largest error source in estimating metabolic rates via diel analysis (Howard et al., 2018 ). The primary contributors to the uncertainty in our NEM estimates, determined by isolating one error source while setting all others to zero in the Monte Carlo simulation (data not shown), were the wind speed values and the choice of gas exchange parameterization. As described earlier, we found the ERA5 wind speed data overestimated the Gull meteorological station data by an average of 1.53 m s − 1 . Consequently, our metabolic rate estimates are likely slight overestimates. Lastly, while we quantified aquatic ecosystem metabolism in terms of oxygen units, these values are not readily converted to carbon units. In shallow water systems, O 2 and CO 2 dynamics can become decoupled, and standard open-ocean stoichiometric ratios may not apply. For instance, S. R. Wang et al. ( 2018 ) observed large seasonal variability in the respiration quotient (CO 2 : O 2 ) from 0.5 to 1.5 in a marsh-dominated estuary in Georgia, U.S. This variability arises in part because anaerobic respiration—prevalent in salt marsh and estuarine sediments, especially during warmer months—does not consume O 2 but still produces CO 2 . Accurately estimating metabolic fluxes in carbon units would therefore require a full characterization of the carbonate system, which is beyond the scope of this paper. Statistics To explore the relationships between environmental conditions and metabolic rates, we calculated bivariate associations between GPP, ER, and NEM and various environmental parameters using daily-averaged data pooled from all three open-water platforms. All response and explanatory variables were scaled to have a mean of zero and a standard deviation of one, i.e., ( x – x̄ )/ s . This standardization ensured that variables with different ranges of variation were on the same scale, allowing direct comparisons between them. We selected six physical and biological parameters likely associated with estuarine metabolism based on previous studies: salinity and temperature (Bas-Silvestre et al., 2020 ; Caffrey, 2004 ; Gomez-Castillo et al., 2023 ; Nelson et al., 2017 ; Russell & Montagna, 2007 ), DO % saturation (Russell & Montagna, 2007 ), pH (Baumann & Smith, 2018 ; Lowe et al., 2019 ; Russell & Montagna, 2007 ), daily tidal range (Gomez-Castillo et al., 2023 ), and wind speed (Nelson et al., 2017 ; Russell & Montagna, 2007 ). For all statistical analyses, results were considered statistically significant at α = 0.05 and highly significant at α = 0.01. Next, we assessed relative strength of associations between environmental variables and metabolic rates using multiple linear regression. To evaluate collinearity (i.e., when explanatory variables are correlated; Zuur et al., 2007 ), we calculated pairwise correlation coefficients ( r ) and variance inflation factors (VIFs). Collinearity is typically considered high when | r | > 0.7 and VIF > 5 or 10 (Dormann et al., 2013 ; Zuur et al., 2009 ); we adopted the more conservative VIF threshold of 5. For each response variable (GPP, ER, and NEM), we fitted a multiple linear regression model using the complete annual dataset. In addition to the six environmental explanatory variables, the models also controlled for site (Gull, North, and South), season (winter, spring, summer, and fall), and the interaction between site and season to account for site- and season-specific effects. Including site as a fixed effect allowed us to isolate variations in metabolic rates attributed to environmental factors, rather than site-specific characteristics (e.g., differing depths). We also fitted separate models for each season, using the same explanatory variables as the annual model, while controlling for site. All regression models were conducted in MATLAB using the fitlm function, and VIFs were calculated with the vif function (Vasilaky, 2025 ). We also examined the effect of short-term perturbations from storm events on metabolic rates. Storm events were identified as periods between December 2021–June 2024 when the daily averaged ERA5 wind speed exceeded the 99th percentile threshold (11.9 m s − 1 ). We then calculated the mean NEM values during the storm event, as well as the means for pre- and post-event periods (5 days in length each). Paired t-tests were used to compare NEM rates before, during, and after storm events using the values pooled from all three platforms using MATLAB’s ttest function. Results Environmental Parameters Seasonal temperature patterns were nearly identical across all three sites (Fig. 2 a). Mean monthly temperatures were lowest during winter months (December–February) and highest during summer months (June–August), with a mean seasonal range of 21.47 o C across all sites. The overall site-wide mean temperature was 14.77 o C. Similarly, seasonal salinity patterns were consistent between sites, with slightly higher salinities during warmer months (Fig. 2 b). The site-wide mean salinity was 31.2 psu with a mean seasonal range of 5.4 psu. A slight freshening trend was observed at all sites in the final year of measurement, reaching an average minimum value of 26.7 psu in late December 2023/early January 2024. This trend was also observed at water-quality monitoring stations (Buoy 126 and Buoy 139) in nearby Great Bay, NJ, approximately 36 mi up the coast, as reported by the Jacques Cousteau National Estuarine Research Reserve (data not shown). Dissolved oxygen % saturation exhibited strong seasonal variation that was the inverse of temperature, peaking in winter and reaching its lowest values in summer (Fig. 2 c). The mean seasonal range across sites was 32.5%. Daily variability was greatest during warmer months, when elevated biological activity led to more intense diel swings: mid- to late-summer values ranging from hypoxic (< 30%) in the early morning to 100% saturation or supersaturation by early evening, reflecting heightened photosynthesis and respiration (Fig. 3 ). In contrast, colder temperatures in winter enhanced oxygen solubility, and DO levels remained consistently near or above saturation throughout the day, with biological influences less dominant. Site means ranged from 89.7 to 94.8% (Table 1 ), with Gull showing consistently lower saturation levels than North and South, especially during summer months. Seasonal pH patterns mirrored DO saturation, with annual highs in winter and annual lows in summer (Fig. 2 d). Similar to DO, daily variability was higher in summer than in winter (Fig. 3 ). Overall mean pH ranged from 8.02 to 8.11, with a mean seasonal range of 0.47. As with DO, pH at Gull was consistently lower than at the other two sites. Daily mean tidal range was consistent across all sites, varying predictably with lower tidal ranges during neap tides (mean ~ 1.2 m) and higher ranges during spring tides (mean ~ 1.8 m) (Fig. 2 e). The overall site-wide mean tidal range was 1.42 m (Table 1 ). Periods of highest daily mean wind speeds were recorded during winter months and early spring due to more frequent winter storms and cold fronts (Fig. 2 f). During the three-year measurement period, one tropical cyclone, Hurricane Ian, occurred in late September/early October 2022. While chlorophyll a data coverage was patchy, clear periods of elevated concentrations were consistently observed in January–March of each year, indicative of the winter–spring bloom (Fig. S1e in Supplemental Information). Turbidity data were even patchier; results from available periods suggested that higher turbidity levels corresponded to larger tidal ranges during spring tides. No seasonal patterns in turbidity were evident (Fig. S1f). Metabolic Rates Metabolic rates, calculated from the monitoring data, showed similar seasonal trends across all sites (Fig. 2 g–i). GPP and ER were inversely related, with months with the highest gross production also having the highest respiration, and vice versa. Both GPP and ER increased in magnitude from colder to warmer months. NEM was most strongly heterotrophic during warmer months, shifting to slightly autotrophic values during the winter. Continuous monitoring also revealed short-term perturbations to metabolic rates (on the order of days), which were associated with periods of high wind speed (see “Impact of Storms”). Rates of GPP, ER, and NEM averaged across all years showed some variation among channel sites (Table 2 ). Both gross production and respiration areal rates were consistently higher at South compared to the other two sites, primarily due to its deeper water depth. Annual average GPP at South was ~ 1.6 times higher than at Gull and ~ 1.9 times higher than at North. South also exhibited the widest seasonal range in GPP, fluctuating between 61.4–267.7 mmol O 2 m − 2 d − 1 . Annual average ER at South was ~ 1.3 times higher than at Gull and ~ 1.7 times higher than at North. As with GPP, South exhibited the broadest seasonal range in ER, from − 373.4– -54.9 mmol O 2 m − 2 d − 1 . All three sites were net heterotrophic on an annual basis. Despite the site differences in GPP and ER, annual average NEM was very similar at North and South, with respective means of -20.1 ± 28.5 and − 26.1 ± 34.1 mmol O 2 m − 2 d − 1 . Gull was slightly more heterotrophic with a mean rate of -42.1 ± 31.9 mmol O 2 m − 2 d − 1 , however, all three sites overlapped within their uncertainty ranges. Table 2 Summary of mean and seasonal range of metabolic rates for the three channel sites over the entire time series (2021–2024). Errors reflect propagated uncertainty calculated through the Monte Carlo analysis. Seasonal range and gap-filling methods are as described in Table 1 . GPP (mmol O 2 m − 2 d − 1 ) ER (mmol O 2 m − 2 d − 1 ) NEM (mmol O 2 m − 2 d − 1 ) Mean Seas. range Mean Seas. range Mean Seas. range North 76.3 ± 19.4 23.6–164.4 -99.3 ± 20.4 -218.4– -19.1 -20.1 ± 28.5 -91.6–32.9 Gull 87.9 ± 21.7 22.3–198.3 -132.2 ± 23.4 -308.0– -22.2 -42.1 ± 31.9 -167.2–29.3 South 144.2 ± 22.5 61.4–267.7 -172.1 ± 25.3 -373.4– -54.9 -26.1 ± 34.1 -108.3–32.3 To examine the climatological means of channel-integrated metabolism over the annual cycle, metabolic rates were averaged across all three sites by day and month of year (Fig. 4 ). From fall to early spring, GPP and ER were comparable in magnitude. However, from late spring to late summer, a period of heightened metabolic activity occurred during which respiration outstripped gross production. Summer production and respiration rates were up to nearly 4 and 6 times higher, respectively, than those observed in winter. Monthly mean GPP ranged from a low of 49.3 mmol O 2 m − 2 d − 1 in December to a high of 187.3 mmol O 2 m − 2 d − 1 in July. Similarly, monthly mean ER ranged from − 44.3 mmol O 2 m − 2 d − 1 in December to -269.9 mmol O 2 m − 2 d − 1 in July. This seasonal pattern is reflected in NEM, which remained near balance or slightly autotrophic from November to April, then became strictly heterotrophic starting in late spring, peaking in late summer. The system was most autotrophic in February (NEM of 12.0 mmol O 2 m − 2 d − 1 ) and most heterotrophic in September (-88.0 mmol O 2 m − 2 d − 1 ). Discussion Comparisons to Other Systems Marsh-dominated estuaries, characterized by surrounding tidal marshes, are strongly influenced by exchanges of materials within the marsh (Cai, 2011 ). High rates of production are often observed in marsh estuaries, which are close to land margins and receive nutrient runoff that enhances local production; moreover, in shallow waters, high light levels enhance benthic production (Caffrey 2004 ). Marsh plants are a large source of allochthonous organic matter that is respired in adjacent waters, often creating highly heterotrophic conditions in the aquatic component of marsh ecosystems (Cai, 2011 ). Despite the importance of coastal vegetated systems in driving coastal carbon fluxes, there are limited estimates of metabolic rates for marsh-dominated estuaries, particularly over the course of a full annual cycle (S. R. Wang et al., 2018 ). Herrmann et al.'s ( 2015 ) synthesis of the net ecosystem metabolism of estuaries along the U.S. East Coast found that these open estuarine waters are collectively net heterotrophic, with a best estimate of -3.2 mmol O 2 m − 2 d − 1 . For New Jersey inland bays specifically, NEM was estimated as -4.5 mmol O 2 m − 2 d − 1 . Caffrey ( 2004 ) provided annual average areal metabolic rates for a variety of shallow, well-mixed estuarine sites across U.S. coastal bioregions, showing that estuaries adjacent to marshes or mangroves were generally net heterotrophic. Notably, these smaller systems with tidal creeks and marshes often exhibited NEM rates that were several times higher than those in larger estuarine system, such as those included in Herrmann et al. ( 2015 ). Of the 42 sites studied by Caffrey ( 2004 ), half were predominantly marsh-dominated habitats, mainly located in the Mid-Atlantic, Southeast, and Pacific regions. For these marsh-adjacent sites, the annual average GPP was 305 mmol O 2 m − 2 d − 1 (with site-specific values ranging from 94 to 878 mmol O 2 m − 2 d − 1 ) and annual average ER was − 384 mmol O 2 m − 2 d − 1 (ranging from − 138 to -1,023 mmol O 2 m − 2 d − 1 ). This resulted in an annual average NEM of -81 mmol O 2 m − 2 d − 1 (ranging from − 28 to -125 mmol O 2 m − 2 d − 1 ). In comparison, Hagerthey et al. ( 2010 ) measured similar magnitudes across 64 sites in the Florida Everglades peatland, with a system-wide mean GPP of 103 mmol O 2 m − 2 d − 1 , ER of -220 mmol O 2 m − 2 d − 1 , and NEM of -117 mmol O 2 m − 2 d − 1 . Our site-wide annual averages—GPP of 102.8 mmol O 2 m − 2 d − 1 , ER of -134.5 mmol O 2 m − 2 d − 1 , and NEM of -29.4 mmol O 2 m − 2 d − 1 —fall on the lower end of Caffrey's (2004) ranges for marsh estuaries across the U.S., and indicated a much less heterotrophic system compared to the Everglades. Conversely, our metabolic rates were higher than those reported by Caffrey et al. ( 2014 ) for the salt-marsh dominated Grand Bay estuary in the Gulf of Mexico, where annual average GPP was 31 mmol O 2 m − 1 d − 1 and NEM was − 15.1 mmol O 2 m − 1 d − 1 . Similarly, Raymond et al. ( 2000 ) found an even lower annual average GPP of 22.8 mmol O 2 m − 1 d − 1 for the York River Estuary in Virginia, which has large fringing marshes. Our gross production and respiration rates were approximately double those of the Duplin salt-marsh estuary in Georgia (S. R. Wang et al., 2018 ), where annual average GPP was 49 mmol O 2 m − 1 d − 1 and ER was − 82 mmol O 2 m − 1 d − 1 ; however, our NEM was comparable to their value of -33 mmol O 2 m − 1 d − 1 . On the other hand, our values are comparable to or lower than those found in marsh-surrounded coastal lagoons. For instance, for the Ria Formosa Lagoon in Portugal, which is surrounded by salt marshes, Cravo et al. ( 2020 ) found GPP ranged from 150–300 mmol O 2 m − 1 d − 1 between winter and spring/summer, and ER from − 150– -350 mmol O 2 m − 1 d − 1 . This resulted in a near-balanced annual average NEM of -0.67 mmol O 2 m − 1 d − 1 . In contrast, Bas-Silvestre et al. ( 2020 ) reported much higher GPP and ER values for two confined coastal lagoons in La Pletera salt marsh in the Mediterranean, with annual average GPP ranging from 377–531 mmol O 2 m − 1 d − 1 and ER ranging from − 401– -491 mmol O 2 m − 1 d − 1 . The La Pletera marsh lagoons were among the most productive aquatic systems in the published literature, higher than even eutrophic estuaries. Factors Associated with Metabolism To better understand which environmental factors are associated with marsh aquatic metabolism, we created simple and multiple linear regression models. These models reveal statistical associations between variables but do not imply causation. For instance, a variable correlated with a metabolic rate may be a response to, rather than a driver of, metabolic process—or both may respond to an unmeasured third factor. Indirect relationships may also occur, where a variable correlates with metabolism only because it is linked to another factor that actually influences the metabolic process. Thus, while regression results can highlight potential drivers, confirming causality requires additional approaches, such as controlled experiments or mechanistic models. Our objectives were to identify which environmental parameters were the most strongly associated with GPP, ER, and NEM; to develop predictive models for metabolism under seasonally changing environmental conditions; and to establish a baseline framework to enable future comparisons of functioning across different marsh sites. These linear regressions are a first step in linking environmental variability to marsh metabolism in the SMIIL study area, with more sophisticated approaches (e.g., generalized additive models, machine learning) needed to fully capture non-linearities and interactions. We began with simple linear regression models between daily rates of GPP, ER, and NEM and daily means of environmental parameters (salinity, temperature, DO % saturation, pH, tidal height, and wind speed), evaluated both annually and by season. These models quantified the variance in metabolic rates explained by each variable alone, helping to identify those with the strongest standalone relationships. These models therefore served as an initial exploration of individual associations and provided a basis for identifying key patterns. We then built multiple linear regression models using the same set of environmental variables to assess their combined influence on metabolic rates across the same timescales. By incorporating multiple predictors simultaneously, these models improved explanatory power and accounted for interactions among variables, enabling clearer identification of the most influential factors while controlling for others. To compare the relative importance of each predictor, we used standardized regression coefficients. We present both the simple and multiple regression results to show how relationships persist or change when accounting for other factors. This comparison can help distinguish whether a factor directly influences metabolism or primarily works through other variables, if a strong simple relationship disappears in multiple regression. Conversely, it can reveal previously obscured relationships, if a weak simple relationship becomes stronger in multiple regression after controlling for suppressor variables. This complementary approach is useful in complex ecosystems like salt marshes where numerous factors interact to influence metabolic processes. In the simple regressions, standardized estimated coefficients are equivalent to correlation coefficients (Table 3 ). Several key annual patterns emerged. Temperature showed the strongest associations with all metabolic rates: a strong positive relationship with GPP ( r = 0.605), and strong negative relationships with ER ( r = -0.700) and NEM ( r = -0.547). These results support the notion that temperature is a primary driver of metabolism that more strongly stimulates respiration than production. After temperature, DO % saturation and pH had the strongest associations with metabolic rates, and also showed consistent patterns: negative with GPP and positive with ER and NEM. These strong bivariate associations likely reflect feedback effects where metabolic activity alters water chemistry, rather than the other way around. Salinity and wind speed showed moderate relationships: salinity was positively associated with GPP ( r = 0.234) and negatively with ER ( r = -0.257) and NEM ( r = -0.188), while wind speed was negatively related to GPP ( r = -0.217) and NEM ( r = -0.155), and slightly positive for ER ( r = 0.045). Table 3 Estimated coefficients from simple linear regression models relating metabolic rates (GPP, ER, NEM) to individual environmental variables (salinity, temperature, DO percent saturation, pH, tidal range, and wind speed) across all three channel sites, annually and by season. All variables are scaled to ( x – x̄ )/ s and are therefore unitless. Metabolic rate Dataset Sal. Temp. DO %sat pH Tidal Wind GPP Annual 0.234** (0.021) 0.605** (0.017) -0.512** (0.018) -0.505** (0.019) -0.015 (0.021) -0.217** (0.021) Winter -0.077** (0.024) 0.293** (0.070) 0.260** (0.073) 0.276** (0.054) -0.027 (0.019) 0.026 (0.021) Spring -0.031 (0.040) 0.706** (0.055) -0.470** (0.069) -0.198** (0.053) -0.005 (0.033) -0.093** (0.033) Summer -0.024 (0.043) 0.058 (0.135) -0.116* (0.046) 0.060 (0.061) -0.108* (0.047) -0.202** (0.058) Fall 0.172** (0.034) 0.421** (0.029) -0.279** (0.026) -0.231** (0.028) -0.063* (0.026) -0.107** (0.024) ER Annual -0.257** (0.021) -0.700** (0.015) 0.691** (0.016) 0.659** (0.016) 0.010 (0.021) 0.045* (0.021) Winter 0.035* (0.016) 0.191** (0.047) 0.170** (0.049) -0.098** (0.037) 0.022 (0.013) -0.045** (0.014) Spring 0.070* (0.032) -0.654** (0.042) 0.744** (0.050) 0.425** (0.040) 0.043 (0.027) -0.021 (0.027) Summer 0.055 (0.041) -0.381** (0.129) 0.268** (0.043) 0.162** (0.058) 0.095* (0.045) -0.417** (0.053) Fall 0.032 (0.050) -0.616** (0.042) 0.576** (0.032) 0.505** (0.037) 0.007 (0.039) -0.193** (0.034) NEM Annual -0.188** (0.021) -0.547** (0.018) 0.633** (0.017) 0.585** (0.017) 0.001 (0.021) -0.155** (0.021) Winter -0.022 (0.022) 0.639** (0.059) 0.569** (0.063) 0.128* (0.051) 0.007 (0.018) -0.050** (0.019) Spring 0.086** (0.029) -0.361** (0.042) 0.768** (0.042) 0.515** (0.033) 0.068** (0.024) -0.136** (0.024) Summer 0.068 (0.048) -0.589** (0.151) 0.335** (0.051) 0.342** (0.068) 0.047 (0.053) -0.929** (0.052) Fall 0.240** (0.069) -0.601** (0.063) 0.686** (0.047) 0.616** (0.053) -0.056 (0.053) -0.445** (0.044) Standard errors are reported in parentheses. *, ** indicates significance at the 95% and 99% level, respectively. Seasonal trends were also apparent. Temperature had its strongest effect on GPP in spring ( r = 0.706) but became insignificant in summer, suggesting a possible threshold or saturation effect. It also showed a strong positive relationship with NEM in winter ( r = 0.639) but was negative in other seasons. Wind speed had a strong negative relationship with NEM in summer ( r = -0.929) and was consistently negative with GPP across seasons, possibly reflecting effects of turbidity or gas exchange. Tidal range showed minimal influence overall, generally not significant for most seasons and metabolic rates. Overall, the simple regressions indicate that the association between environmental factors and metabolic rates varies seasonally, as evidenced by the significant shift in coefficient values and even sign changes across seasons. Moreover, even the effect of the primary driver, temperature, is not constant throughout the year, suggesting acclimation or interaction with other seasonal factors. Before fitting multiple regressions, we assessed collinearity between explanatory factors using pairwise correlation coefficients and VIFs. Strong correlations (| r| > 0.7) were found between salinity and temperature, pH and temperature, and DO % saturation and pH (Table S2 in Supplemental Information). However, all VIFs were below 5, with temperature (4.81), pH (4.65), and DO % saturation (3.61) indicating moderate collinearity, while salinity, tidal range, and wind speed had VIFs near 1. Therefore, we retained all six predictors in the multiple regression models. All fifteen multiple regression models were highly significant and explained 38–73% of the variation in metabolic rates, with generally higher adjusted R 2 values for GPP and ER than for NEM (Table 4 ). Annual models explained 73% and 70% of the variation in GPP and ER, respectively, and 53% of the variation in NEM. Summer models for GPP and ER explained less variation than other seasons, while the winter model for NEM was weakest. Table 4 Estimated coefficients from multiple linear regression models relating metabolic rates (GPP, ER, and NEM) to environmental variables across all three channel sites, annually and by season. Explanatory variables are shown; grouping variables for the season-specific models are site, and for the annual models are site, season, and their interactions. All variables are standardized and unitless. Metabolic rate Dataset Sal. Temp. DO %sat pH Tidal Wind R 2 adj GPP Annual -0.056** (0.014) 0.642** (0.031) -0.028 (0.024) 0.090** (0.028) -0.002 (0.011) -0.021 (0.011) 0.728** Winter -0.122** (0.014) 0.234** (0.042) 0.410** (0.056) -0.030 (0.043) 0.004 (0.012) 0.069** (0.013) 0.683** Spring -0.071** (0.022) 1.008** (0.052) 0.006 (0.063) 0.095 (0.054) 0.029 (0.018) 0.055** (0.018) 0.732** Summer -0.067 (0.038) 0.435** (0.138) -0.028 (0.049) 0.059 (0.063) -0.082* (0.039) -0.192** (0.046) 0.401** Fall 0.104** (0.028) 0.345** (0.052) -0.179** (0.035) 0.138** (0.045) -0.037 (0.020) -0.052** (0.018) 0.493** ER Annual 0.081** (0.015) -0.493** (0.033) 0.266** (0.025) 0.008 (0.029) 0.035** (0.012) -0.188** (0.012) 0.700** Winter 0.035** (0.012) 0.240** (0.034) 0.182** (0.046) -0.044 (0.036) 0.039** (0.010) -0.029** (0.010) 0.527** Spring 0.076** (0.022) -0.622** (0.051) 0.343** (0.061) 0.042 (0.053 0.034* (0.017) -0.140** (0.018) 0.604** Summer 0.093* (0.037) -0.477** (0.135) 0.221** (0.048) -0.016 (0.061) 0.057 (0.038) -0.471** (0.046) 0.384** Fall 0.000 (0.038) -0.330** (0.069) 0.454** (0.047) -0.036 (0.059) 0.001 (0.026) -0.293** (0.024) 0.572** NEM Annual 0.078** (0.018) -0.154** (0.041) 0.425** (0.031) 0.109** (0.037) 0.058** (0.015) -0.344** (0.015) 0.533** Winter -0.072** (0.018) 0.662** (0.054) 0.750** (0.072) -0.107 (0.055) 0.071** (0.015) 0.024 (0.016) 0.400** Spring 0.054* (0.024) 0.016 (0.054) 0.593** (0.066) 0.174** (0.058) 0.089** (0.019) -0.180** (0.020) 0.427** Summer 0.087* (0.036) -0.348** (0.132) 0.347** (0.047) 0.035 (0.060) 0.010 (0.037) -1.010** (0.044) 0.575** Fall 0.111* (0.052) -0.195* (0.096) 0.583** (0.065) 0.085 (0.083) -0.038 (0.037) -0.557** (0.034) 0.556** Standard errors are reported in parentheses. *, ** indicates significance at the 95% and 99% level, respectively. While the simple regression results provide a foundation for understanding primary relationships, the multiple regression results reveal which factors remain significant when controlling for others, potentially showing which correlations are direct versus indirect. Temperature remained the most influential factor overall: it was still strongly positive for GPP across seasons, and maintained a negative relationship with ER in most seasons—except in winter, as in the simple regression. The scaled influence of temperature on GPP was greater than that of any other predictor in nearly all cases, sometimes by an order of magnitude. For instance, in the annual GPP model, temperature’s scaled effect ( β = 0.642) was seven times larger than that of pH ( β = 0.090) and nearly 11.5 times that of salinity ( β = -0.056). In contrast, DO % saturation and pH, which were strongly correlated with GPP in simple regression, were often non-significant or reverse in multiple regressions, supporting the interpretation that they reflect feedbacks from metabolism rather than direct drivers. Salinity and wind speed became more important in the multiple regressions. Both were significant across most seasons for all metabolic rates, with wind speed showing a particularly strong association with NEM in summer ( β = -1.010). Multiple regressions also highlighted seasonal variability (Fig. 5 and Fig. S3 in Supplemental Information). Spring GPP had the highest association with temperature ( β = 1.008), while all other coefficients were two orders of magnitude lower. Winter metabolism showed unique patterns, as it was the only season where temperature was positively correlated with ER and NEM. Similarly, salinity had a distinct effect in fall, where it was positively associated with GPP (contrary to other seasons) but not significant for ER. Altogether, these statistical patterns offer several ecological insights into the processes associated with marsh metabolism. In particular, temperature emerged as a dominant and consistent driver, although its influence varied by season and metabolic process. Warmer conditions in spring and summer promote marsh plant growth and labile organic matter availability, which in turn fuels microbial respiration; higher temperatures also directly elevate respiration rates. This dual stimulation of GPP and ER is well-supported in the literature. For example, Caffrey et al. ( 2014 ) found that the effect of temperature is stronger for respiration than primary production, resulting in more heterotrophic conditions in summer. As a result, seasonally increasing temperatures generally lead to more negative NEM rates, consistent with our findings. Salinity also played a meaningful but more variable role. Generally, lower salinity was associated with higher GPP and ER, suggesting that freshwater inputs may support increased metabolic activity. This pattern likely reflects nutrient delivery, as salinity acts as a conservative tracer for freshwater-driven nutrient inputs to the system, which is a key factor controlling primary production. For example, in the Grand Bay estuary, Caffrey et al. ( 2014 ) observed that periods of low salinity were followed by increases in primary production. Similarly, we infer that freshwater inputs stimulate productivity in the SMIIL system, which in turn leads to increased respiration from decomposition of newly produced organic matter. Patterns in pH and DO % saturation suggest that water chemistry is strongly influenced by biological processes, rather than driving them. Although both variables were consistently correlated with metabolic rates in simple regressions, their significance often disappeared in multiple regressions, suggesting they reflect feedbacks from production and respiration. This finding aligns with Baumann & Smith ( 2018 ), who found that pH patterns were closely linked to dissolved oxygen across a diversity of shallow estuarine environments across the U.S, and that these concurrent fluctuations were driven by local metabolic processes. Similarly, Lowe et al. ( 2019 ) found that local ecosystem metabolism is the dominant driver of pH variability in a range of habitats in the Northeast Pacific coastal ocean. Both of these studies also provided strong empirical evidence that physical factors indirectly influence carbonate chemistry through affecting primary production and respiration, with the implication that increased ecosystem respiration from warmer waters exacerbates coastal acidification. Our findings build on these insights by showing seasonal peaks in the pH–GPP and pH–NEM relationships in fall and spring, respectively, suggesting that metabolic control of carbonate chemistry may be strongest during transitional periods. Finally, we explored the potential influence of physical processes such as tidal and wind mixing on metabolism. Although our turbidity and chlorophyll a datasets were compromised by biofouling, limited available data showed a correlation between turbidity and tidal range, consistent with findings from Gomez-Castillo et al. ( 2023 ). They observed that blooms typically developed during neap tides when mixing was reduced and turbidity was lower, shifting estuarine metabolism to net autotrophic conditions. These blooms subsequently dissipated during the following spring tide, when stronger mixing led to increased turbidity and a return to heterotrophy. In our system, however, tidal range was not a significant predictor of metabolism, suggesting that mixing-related processes may be less influential here than in more phytoplankton-dominated estuaries. In contrast, wind speed emerged as a more important factor in multiple regressions, particularly for NEM, where it showed strong negative effects in summer and fall. This suggests that wind-driven mixing may enhance gas exchange or turbidity, thereby influencing net carbon balances. Impact of Storms Further examination of the wind speed and metabolic rate time series suggested that temporary fluctuations in NEM were linked to short periods of elevated daily wind speed. For example, Tropical Storm Ian in early October 2022 was accompanied by a large dip in NEM at all three sites (Fig. 6 ). To investigate the potential effect of short-term storm events on net ecosystem metabolism, we identified high wind-speed events in the measurement record and compared mean NEM values during the storm event to baseline periods of 5 days before and after the events. From December 2021 to June 2024, eight events were recorded during which the daily wind speed exceeded the 99th percentile (Fig. 6 ). Due to gaps in the NEM record for each site, this corresponded to six events each for North and South, and seven for Gull. The site-wide mean NEM rate during events was − 68.69 mmol O 2 m − 2 d − 1 , while the means for the pre- and post-event periods were − 8.60 and − 12.9 mmol O 2 m − 2 d − 1 , respectively (Fig. 7 ). NEM was significantly lower during the events compared to pre-event conditions (paired t - test, p = 0.0090) and also compared to post-event conditions (p = 0.013). There was no significant difference between pre-event and post-event NEM values (p = 0.69). Elevated wind speed was consistently associated with higher respiration rates and more heterotrophic conditions, in line with findings by Gomez-Castillo et al. ( 2023 ). Wind speed affects currents and mixing in estuaries, which in turn influences dissolved oxygen levels, nutrient and organic matter transport, and turbidity and light attenuation (Kemp & Testa, 2012 ). Wind speed is also a key indicator of storm conditions. On weekly to monthly time scales, ephemeral events like storms may influence metabolic balance more than seasonal changes in environmental conditions (Russell & Montagna, 2007 ). While we focused on increases in wind speed, other indicators of storms, such as heavy precipitation and drops in barometric pressure could also serve to identify storms events. Our results suggest that high wind speed events are linked to a temporary increase in heterotrophy, which quickly returns to baseline levels after the storm passes. Various alternative mechanisms—including storm surge, heavy precipitation, and increased discharge—can also influence estuarine metabolism (Buelo et al., 2024 ). For example, higher winds may reduce the intensity of upwelling-induced production in certain systems, while elevated precipitation can increase gross production through enhanced nutrient delivery and reduced estuarine residence time. The impacts of storm events are thus complex and likely system-dependent, warranting further investigation in future studies. Conclusion Future changes in estuarine systems will depend on how anthropogenic impacts and climate change alter the balance between gross primary production and ecosystem respiration. While predicting these shifts remains challenging due to observational gaps, continuous monitoring of key parameters like dissolved oxygen, temperature, and salinity is becoming essential for capturing the dynamics of aquatic metabolism in diverse coastal environments. In this study, we quantified aquatic ecosystem metabolism in the main tidal channel of the Seven Mile Island Innovation Laboratory and found that all three monitoring sites were net heterotrophic annually, with clear seasonal patterns: both GPP and ER increased substantially from colder to warmer months, leading to peak heterotrophy in late summer/fall. This pattern supports the concept of the seasonal marsh CO 2 pump (Z. A. Wang et al., 2016 ; Z. A. Wang & Cai, 2004 ), wherein marsh-derived organic matter is respired in adjacent waters, causing the aquatic environments of salt marsh systems to often function as net CO 2 sources. These results establish a robust baseline characterization of metabolic dynamics in the SMIIL salt marsh system. Our results suggest that warming temperatures, which are projected to continue to increase across the mid-Atlantic (EPA Region 3 Climate Collaborative, 2022 ), could further drive estuarine systems in the region toward greater heterotrophy, increasing CO 2 emissions and potentially intensifying coastal acidification. Episodic storm events, expected to become more frequent and severe, may periodically amplify these metabolic responses. The SMIIL research area, designed to evaluate the use of dredged sediments to enhance marsh accretion and protection, also provides a unique opportunity to study the impacts of sea level rise on coastal ecosystems. In New Jersey, relative sea level rise is occurring at more than twice the global average rate (New Jersey Climate Change Resource Center, 2020 ), which will likely reshape coastal ecosystems with significant implications for blue carbon sequestration. While rising sea levels may enhance soil carbon accumulation in marshes, they could also trigger ecosystem transitions that alter the balance between carbon sequestration and greenhouse gas emissions (Kirwan et al., 2023 ). Our three-year dataset captures natural variability in marsh channel environmental conditions and metabolism and offers a valuable reference point for evaluating future environmental changes and management interventions. Although this study focused on aquatic metabolism of the salt marsh ecosystem, integrating these data with carbon fluxes from other marsh features—including vegetated platforms, sediments, and salt ponds—would provide a more complete understanding of system-wide carbon dynamics. Such integrated approaches are essential for refining estimates of salt marsh carbon sink capacity. As coastal marshes face mounting pressures from climate change and human activities, a holistic understanding of their carbon processing is increasingly critical to inform effective conservation and restoration strategies. Declarations Ethics approval and consent to participate Not applicable Consent for publication Not applicable Availability of data and materials Datasets generated and analyzed during the current study are available in the Biological and Chemical Oceanography Data Management Office (BCO-DMO) repository. Continuous monitoring and meteorological datasets, as well as discrete sample datasets for DO, DIC, and TA, can be accessed at https://www.bco-dmo.org/project/962228. Competing interests The authors declare that they have no competing interests. Funding This work was funded by the U.S. Army Engineer Research and Development Center (W912HZ2020061–RA3). Authors' contributions EJC performed data curation, analysis, and interpretation, and wrote the manuscript. JS led the field data collection and discrete sample laboratory analysis. KEF contributed to field data collection, assisted with data curation, and aided manuscript development. HIP secured funding, supervised the project, and supported manuscript development. All authors contributed to the conception and design of the study, and reviewed and approved the final manuscript. Acknowledgements We thank David Perkey (Engineer Research and Development Center), Kelsey Fall (University of Delaware), Lenore Tedesco and Julie Blum (The Wetlands Institute), and Roland Hagen (Rutgers University Marine Field Station) for their assistance with fieldwork. Field and laboratory support was also provided by Jose Cuevas, Yasmin Hamilton, Elle Joubert, Yoleimy Lopez Barias, and Stevie Walker (Boston College). We thank Kate Willis for ArcGIS support, and Swapnil Sharma and Melissa McTernan for guidance on statistical analyses. References Aanderaa Data Instruments AS. (2017). TD 269 Operating Manual . https://www.aanderaa.com/media/pdfs/oxygen-optode-4330-4835-and-4831.pdf Albert, D. R. (2020). Monte Carlo Uncertainty Propagation with the NIST Uncertainty Machine. Journal of Chemical Education , 97 (5), 1491–1494. https://doi.org/10.1021/acs.jchemed.0c00096 Alongi, D. M. (2020). Carbon balance in salt marsh and mangrove ecosystems: A global synthesis. Journal of Marine Science and Engineering , 8 (10), 1–21. https://doi.org/10.3390/jmse8100767 Barbier, E. B., Hacker, S. D., Kennedy, C., Koch, E. W., Stier, A. C., & Silliman, B. R. (2011). The value of estuarine and coastal ecosystem services. Ecological Monographs , 81 (2), 169–193. https://doi.org/10.1890/10-1510.1 Bas-Silvestre, M., Quintana, X. D., Compte, J., Gascón, S., Boix, D., Antón-Pardo, M., & Obrador, B. (2020). Ecosystem metabolism dynamics and environmental drivers in Mediterranean confined coastal lagoons. Estuarine, Coastal and Shelf Science , 245 , 106989. https://doi.org/10.1016/j.ecss.2020.106989 Bauer, J. E., Cai, W.-J., Raymond, P. A., Bianchi, T. S., Hopkinson, C. S., & Regnier, P. A. G. (2013). The changing carbon cycle of the coastal ocean . https://doi.org/10.1038/nature12857 Baumann, H., & Smith, E. M. (2018). Quantifying Metabolically Driven pH and Oxygen Fluctuations in US Nearshore Habitats at Diel to Interannual Time Scales. Estuaries and Coasts , 41 (4), 1102–1117. https://doi.org/10.1007/s12237-017-0321-3 Beck, M. W., Hagy, J. D., & Murrell, M. C. (2015). Improving estimates of ecosystem metabolism by reducing effects of tidal advection on dissolved oxygen time series. Limnology and Oceanography: Methods , 13 (12), 731–745. https://doi.org/10.1002/lom3.10062 Boscolo-Galazzo, F., Crichton, K. A., Barker, S., & Pearson, P. N. (2018). Temperature dependency of metabolic rates in the upper ocean: A positive feedback to global climate change? Global and Planetary Change , 170 , 201–212. https://doi.org/10.1016/j.gloplacha.2018.08.017 Buelo, C. D., Besterman, A. F., Walter, J. A., Pace, M. L., Ha, D. T., & Tassone, S. J. (2024). Quantifying Disturbance and Recovery in Estuaries: Tropical Cyclones and High-Frequency Measures of Oxygen and Salinity. Estuaries and Coasts , 47 (1), 18–31. https://doi.org/10.1007/s12237-023-01255-1 Caffrey, J. M. (2004). Factors controlling net ecosystem metabolism in U.S. estuaries. Estuaries , 27 (1), 90–101. https://doi.org/10.1007/BF02803563 Caffrey, J. M., Murrell, M. C., Amacker, K. S., Harper, J. W., Phipps, S., & Woodrey, M. S. (2014). Seasonal and Inter-annual Patterns in Primary Production, Respiration, and Net Ecosystem Metabolism in Three Estuaries in the Northeast Gulf of Mexico. Estuaries and Coasts , 37 (S1), 222–241. https://doi.org/10.1007/s12237-013-9701-5 Cai, W. J. (2011). Estuarine and coastal ocean carbon paradox: CO2 sinks or sites of terrestrial carbon incineration? Annual Review of Marine Science , 3 , 123–145. https://doi.org/10.1146/annurev-marine-120709-142723 Chasten, M., Tedesco, L., & Kopkash, G. (2023). Advancing sediment solutions in the Seven Mile Island Innovation Laboratory. Coastal Engineering Proceedings , 37 , 64. https://doi.org/10.9753/icce.v37.papers.64 Cole, J. J., & Caraco, N. F. (1998). Atmospheric exchange of carbon dioxide in a low-wind oligotrophic lake measured by the addition of SF6. Limnology and Oceanography , 43 (4), 647–656. https://doi.org/10.4319/lo.1998.43.4.0647 Cravo, A., Rosa, A., Jacob, J., & Correia, C. (2020). Dissolved oxygen dynamics in Ria Formosa Lagoon (South Portugal)—A real time monitoring station observatory. Marine Chemistry , 223 , 103806. https://doi.org/10.1016/j.marchem.2020.103806 Dai, M., Su, J., Zhao, Y., Hofmann, E. E., Cao, Z., Cai, W. J., Gan, J., Lacroix, F., Laruelle, G. G., Meng, F., Mu die ller, J. D., Regnier, P. A. G., Wang, G., & Wang, Z. (2022). Carbon Fluxes in the Coastal Ocean: Synthesis, Boundary Processes, and Future Trends. Annual Review of Earth and Planetary Sciences , 50 , 593–626. https://doi.org/10.1146/annurev-earth-032320-090746 Dai, M., Zhao, Y., Chai, F., Chen, M., Chen, N., Chen, Y., Cheng, D., Gan, J., Guan, D., Hong, Y., Huang, J., Lee, Y., Leung, K. M. Y., Lim, P. E., Lin, S., Lin, X., Liu, X., Liu, Z., Luo, Y.-W., … Zhang, Z. (2023). Persistent eutrophication and hypoxia in the coastal ocean. Cambridge Prisms: Coastal Futures , 1 , e19. https://doi.org/10.1017/cft.2023.7 Dickson, A. G., Sabine, C. L., Christian, J. R., Bargeron, C. P., & North Pacific Marine Science Organization (Eds.). (2007). Guide to best practices for ocean CO2 measurements . North Pacific Marine Science Organization. Dormann, C. F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., Marquéz, J. R. G., Gruber, B., Lafourcade, B., Leitão, P. J., Münkemüller, T., McClean, C., Osborne, P. E., Reineking, B., Schröder, B., Skidmore, A. K., Zurell, D., & Lautenbach, S. (2013). Collinearity: A review of methods to deal with it and a simulation study evaluating their performance. Ecography , 36 (1), 27–46. https://doi.org/10.1111/j.1600-0587.2012.07348.x Emerson, S., Yang, B., White, M., & Cronin, M. (2019). Air-Sea Gas Transfer: Determining Bubble Fluxes With In Situ N2 Observations. Journal of Geophysical Research: Oceans , 124 (4), 2716–2727. https://doi.org/10.1029/2018JC014786 EPA Region 3 Climate Collaborative. (2022). EPA Region 3 Climate Adaptation Implementation Plan (Mid-Atlantic). US EPA. https://www.epa.gov/climate-adaptation/climate-change-epa-mid-atlantic-region Fall, K. A., Perkey, D. W., Tyler, Z. J., & Welp, T. L. (2021). Field measurement and monitoring of hydrodynamic and suspended sediment within the Seven Mile Island Innovation Laboratory, New Jersey . https://erdclibrary.on.worldcat.org/discovery. Garcia, H. E., & Gordon, L. I. (1992). Oxygen solubility in seawater: Better fitting equations. Limnology and Oceanography , 37 (6), 1307–1312. https://doi.org/10.4319/lo.1992.37.6.1307 Gedan, K., Silliman, B., & Bertness, M. (2009). Centuries of Human-Driven Change in Salt Marsh Ecosystems. Annual Review of Marine Science , 1 , 117–141. https://doi.org/10.1146/annurev.marine.010908.163930 Gomez-Castillo, A. P., Panton, A., & Purdie, D. A. (2023). Temporal variability of phytoplankton biomass and net community production in a macrotidal temperate estuary. Estuarine, Coastal and Shelf Science , 280 (August 2022), 108182. https://doi.org/10.1016/j.ecss.2022.108182 Hagerthey, S. E., Cole, J. J., & Kilbane, D. (2010). Aquatic metabolism in the Everglades: Dominance of water column heterotrophy. Limnology and Oceanography , 55 (2), 653–666. https://doi.org/10.4319/lo.2010.55.2.0653 Herrmann, M., Najjar, R. G., Kemp, W. M., Alexander, R. B., Boyer, E. W., Cai, W., Griffith, P. C., Kroeger, K. D., Mccallister, S. L., & Smith, R. A. (2015). Net ecosystem production and organic carbon balance of U.S. East Coast estuaries: A synthesis approach. Global Biogeochemical Cycles , 29 , 96–111. https://doi.org/10.1002/2013GB004736.Received Hersbach, H., Berrisford, P., Biavati, G., Horányi, A., Muñoz Sabater, J., Nicolas, J., Peubey, C., Radu, R., Radu, I., Schepers, D., Simmons, A., Soci, D., Dee, D., & Thépaut, J.-N. (2023). ERA5 hourly data on single levels from 1940 to present [Dataset]. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.adbb2d47 Howard, E. M., Forbrich, I., Giblin, A. E., Lott, D. E., Cahill, K. L., & Stanley, R. H. R. (2018). Using Noble Gases to Compare Parameterizations of Air-Water Gas Exchange and to Constrain Oxygen Losses by Ebullition in a Shallow Aquatic Environment. Journal of Geophysical Research: Biogeosciences , 123 (9), 2711–2726. https://doi.org/10.1029/2018JG004441 Howarth, R. W., Swaney, D. P., Butler, T. J., & Marino, R. (2000). Climatic Control on Eutrophication of the Hudson River Estuary. Ecosystems , 3 (2), 210–215. IOOS. (2018). Manual for Real-Time Quality Control of Dissolved Oxygen Observation: A Guide to Quality Control and Quality Assurance for Dissiolved Oxygen Observations in Coastal Oceans (No. 2.1). https://doi.org/10.25923/q0m1-d488 IPCC. (2023). Climate Change 2023: Synthesis Report (Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, H. Lee and J. Romero (Eds.)]). IPCC. 10.59327/IPCC/AR6-9789291691647 Jankowski, K. J., Mejia, F. H., Blaszczak, J. R., & Holtgrieve, G. W. (2021). Aquatic ecosystem metabolism as a tool in environmental management. WIREs Water , 8 (4), e1521. https://doi.org/10.1002/wat2.1521 Kemp, W. M., & Testa, J. M. (2012). Metabolic Balance between Ecosystem Production and Consumption. In Treatise on Estuarine and Coastal Science (Vol. 7, pp. 83–118). Elsevier Inc. https://doi.org/10.1016/B978-0-12-374711-2.00706-3 Kirwan, M. L., Megonigal, J. P., Noyce, G. L., & Smith, A. J. (2023). Geomorphic and ecological constraints on the coastal carbon sink. Nature Reviews Earth & Environment , 4 (6), 393–406. https://doi.org/10.1038/s43017-023-00429-6 Langdon, C. (2010). Determination of dissolved oxygen in seawater by Winkler titration using the amperometric technique . Laruelle, G. G., Dürr, H. H., Slomp, C. P., & Borges, A. V. (2010). Evaluation of sinks and sources of CO2 in the global coastal ocean using a spatially‐explicit typology of estuaries and continental shelves . https://doi.org/10.1029/2010GL043691 Lewis, E., & Wallace, D. (1998). Program developed for CO2 system calculations (No. ORNL/CDIAC-105). Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory. Leys, C., Ley, C., Klein, O., Bernard, P., & Licata, L. (2013). Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. Journal of Experimental Social Psychology , 49 (4), 764–766. https://doi.org/10.1016/j.jesp.2013.03.013 Liang, J. H., Deutsch, C., McWilliams, J. C., Baschek, B., Sullivan, P. P., & Chiba, D. (2013). Parameterizing bubble-mediated air-sea gas exchange and its effect on ocean ventilation. Global Biogeochemical Cycles , 27 (3), 894–905. https://doi.org/10.1002/gbc.20080 Lowe, A. T., Bos, J., & Ruesink, J. (2019). Ecosystem metabolism drives pH variability and modulates long-term ocean acidification in the Northeast Pacific coastal ocean. Scientific Reports , 9 (1), 1–11. https://doi.org/10.1038/s41598-018-37764-4 Manning, C. C. M., & Nicholson, D. P. (2022). dnicholson/gas_toolbox: MATLAB code for calculating gas fluxes . Zenodo. https://doi.org/10.5281/zenodo.6126685 Murrell, M. C., Caffrey, J. M., Marcovich, D. T., Beck, M. W., Jarvis, B. M., & Hagy, J. D. (2018). Seasonal Oxygen Dynamics in a Warm Temperate Estuary: Effects of Hydrologic Variability on Measurements of Primary Production, Respiration, and Net Metabolism. Estuaries and Coasts , 41 (3), 690–707. https://doi.org/10.1007/s12237-017-0328-9 Needoba, J. A., Peterson, T. D., & Johnson, K. S. (2012). Method for the Quantification of Aquatic Primary Production and Net Ecosystem Metabolism Using In Situ Dissolved Oxygen Sensors. In S. M. Tiquia-Arashiro (Ed.), Molecular Biological Technologies for Ocean Sensing (pp. 73–101). https://doi.org/10.1007/978-1-61779-915-0_4 Nelson, N. G., Muñoz-Carpena, R., Neale, P. J., Tzortziou, M., & Megonigal, J. P. (2017). Temporal variability in the importance of hydrologic, biotic, and climatic descriptors of dissolved oxygen dynamics in a shallow tidal-marsh creek. Water Resources Research , 53 (8), 7103–7120. https://doi.org/10.1002/2016WR020196 New Jersey Climate Change Resource Center. (2020). Sea Level Rise in New Jersey: Projections and Impacts . Rutgers. https://njclimateresourcecenter.rutgers.edu/wp-content/uploads/2020/05/Sea-Level-Rise-in-New-Jersey.pdf Nicholson, D., Barrette, J., & Hakai, Z. (2023). boom-lab/winkler-titrator: V0.1.0-alpha (v0.1.0-alpha) [Computer software]. https://doi.org/10.5281/zenodo.8048209 Odum, H. T. (1956). Primary Production in Flowing Waters. Limnology and Oceanography , 1 (2), 102–117. https://doi.org/10.4319/lo.1956.1.2.0102 Orr, J. C., Epitalon, J.-M., Dickson, A. G., & Gattuso, J.-P. (2018). Routine uncertainty propagation for the marine carbon dioxide system. Marine Chemistry , 207 , 84–107. https://doi.org/10.1016/j.marchem.2018.10.006 Palevsky, H. I., Clayton, S., Atamanchuk, D., Battisti, R., Batryn, J., Bourbonnais, A., Briggs, E. M., Carvalho, F., Chase, A. P., Eveleth, R., Fatland, R., Fogaren, K. E., Fram, J. P., Hartman, S. E., Le Bras, I., Manning, C. C. M., Needoba, J. A., Neely, M. B., Oliver, H., … Wingard, C. (2023). OOI Biogeochemical Sensor Data: Best Practices & User Guide. The Global Ocean Observing System , 1.1.1 (July), 1–135. Pennings, S., & Bertness, M. (2001). Salt marsh communities. In Sinauer Associates Inc (pp. 289–316). Perkey, D. W., Tedesco, L. P., Fall, K. A., Huff, T. P., & Chasten, M. A. (2024). Stability of marsh edge berms constructed from fine-grained dredged sediment. Frontiers in Marine Science , August , 1–14. https://doi.org/10.3389/fmars.2024.1401225 Raymond, P. A., Bauer, J. E., & Cole, J. J. (2000). Atmospheric CO2 evasion, dissolved inorganic carbon production, and net heterotrophy in the York River estuary. Limnology and Oceanography , 45 (8), 1707–1717. https://doi.org/10.4319/lo.2000.45.8.1707 Ro, K. S., & Hunt, P. G. (2006). New Unified Equation for Wind-Driven Surficial Oxygen Transfer into Stationary Water Bodies. Transactions of the ASABE , 49 (5), 1615–1622. https://doi.org/10.13031/2013.22020 Roberts, D., MacVean, L., Holleman, R., Chelsky, A., Art, K., Nidzieko, N., Sylvester, Z., & Senn, D. (2022). Connections to Tidal Marsh and Restored Salt Ponds Drive Seasonal and Spatial Variability in Ecosystem Metabolic Rates in Lower South San Francisco Bay. Estuaries and Coasts , 45 (8), 2560–2577. https://doi.org/10.1007/s12237-022-01088-4 Rolando, J. L., Hodges, M., Garcia, K. D., Krueger, G., Williams, N., Carr Jr., J., Robinson, J., George, A., Morris, J., & Kostka, J. E. (2023). Restoration and resilience to sea level rise of a salt marsh affected by dieback events. Ecosphere , 14 (4), e4467. https://doi.org/10.1002/ecs2.4467 Rosentreter, J. A., Laruelle, G. G., Bange, H. W., Bianchi, T. S., Busecke, J. J. M., Cai, W.-J., Eyre, B. D., Forbrich, I., Kwon, E. Y., Maavara, T., Moosdorf, N., Najjar, R. G., Sarma, V. V. S. S., Van Dam, B., & Regnier, P. (2023). Coastal vegetation and estuaries are collectively a greenhouse gas sink. Nature Climate Change , 13 (6), 579–587. https://doi.org/10.1038/s41558-023-01682-9 Russell, M. J., & Montagna, P. A. (2007). Spatial and temporal variability and drivers of net ecosystem metabolism in Western Gulf of Mexico estuaries. Estuaries and Coasts , 30 (1), 137–153. https://doi.org/10.1007/BF02782974 Sharp, J. (2023). CO2SYSv3 for MATLAB (Version v3.2.1) [MATLAB]. https://github.com/jonathansharp/CO2-System-Extd Stanley, R. H. R., Jenkins, W. J., Lott, D. E., & Doney, S. C. (2009). Noble gas constraints on air-sea gas exchange and bubble fluxes. Journal of Geophysical Research: Oceans , 114 (11), 1–14. https://doi.org/10.1029/2009JC005396 Tan, R. Z., Markus, C., Vasikaran, S., & Loh, T. P. (2022). Comparison of 8 methods for univariate statistical exclusion of pathological subpopulations for indirect reference intervals and biological variation studies. Clinical Biochemistry , 103 , 16–24. https://doi.org/10.1016/j.clinbiochem.2022.02.006 Tassone, S. J., & Bukaveckas, P. A. (2019). Seasonal, Interannual, and Longitudinal Patterns in Estuarine Metabolism Derived from Diel Oxygen Data Using Multiple Computational Approaches. Estuaries and Coasts , 42 (4), 1032–1051. https://doi.org/10.1007/s12237-019-00526-0 Vachon, D., & Prairie, Y. T. (2013). The ecosystem size and shape dependence of gas transfer velocity versus wind speed relationships in lakes. Canadian Journal of Fisheries and Aquatic Sciences , 70 (12), 1757–1764. https://doi.org/10.1139/cjfas-2013-0241 Valiela, I., Lloret, J., Bowyer, T., Miner, S., Remsen, D., Elmstrom, E., Cogswell, C., & Robert Thieler, E. (2018). Transient coastal landscapes: Rising sea level threatens salt marshes. Science of The Total Environment , 640–641 , 1148–1156. https://doi.org/10.1016/j.scitotenv.2018.05.235 Van Heuven, S., Pierrot, D., Rae, J., Lewis, E., & Wallace, D. (2011). MATLAB Program Developed for CO2 System Calculations. ORNL/CDIAC-105b. University of St Andrews Research Portal. https://research-portal.st-andrews.ac.uk/en/datasets/matlab-program-developed-for-co2-system-calculations-ornlcdiac-10 Vasilaky, D. (2025). Vif(X) [MATLAB]. https://www.mathworks.com/matlabcentral/fileexchange/60551-vif-x Wallace, R. B., Peterson, B. J., & Gobler, C. J. (2021). Ecosystem Metabolism Modulates the Dynamics of Hypoxia and Acidification Across Temperate Coastal Habitat Types. Frontiers in Marine Science , 8 (December), 1–20. https://doi.org/10.3389/fmars.2021.611781 Wang, S. R., Di Iorio, D., Cai, W. J., & Hopkinson, C. S. (2018). Inorganic carbon and oxygen dynamics in a marsh-dominated estuary. Limnology and Oceanography , 63 (1), 47–71. https://doi.org/10.1002/lno.10614 Wang, Z. A., & Cai, W. J. (2004). Carbon dioxide degassing and inorganic carbon export from a marsh-dominated estuary (the Duplin River): A marsh CO2 pump. Limnology and Oceanography , 49 (2), 341–354. https://doi.org/10.4319/lo.2004.49.2.0341 Wang, Z. A., Kroeger, K. D., Ganju, N. K., Gonneea, M. E., & Chu, S. N. (2016). Intertidal salt marshes as an important source of inorganic carbon to the coastal ocean. Limnology and Oceanography , 61 (5), 1916–1931. https://doi.org/10.1002/lno.10347 Wanninkhof, R. (2014). Relationship between wind speed and gas exchange over the ocean revisited. Limnology and Oceanography: Methods , 12 (JUN), 351–362. https://doi.org/10.4319/lom.2014.12.351 Warnell, K., Olander, L., & Currin, C. (2022). Sea level rise drives carbon and habitat loss in the U.S. mid-Atlantic coastal zone. PLOS Climate , 1 (6), e0000044. https://doi.org/10.1371/journal.pclm.0000044 Zhang, J.-Z., Berberian, G., & Wanninkhof, R. (2002). Long-term storage of natural water samples for dissolved oxygen determination. Water Research , 36 (16), 4165–4168. https://doi.org/10.1016/S0043-1354(02)00093-3 Zuur, A. F., Ieno, E. N., & Smith, G. M. (2007). Linear regression. In Analysing ecological data . Springer. Zuur, A. F., Ieno, E. N., Walker, N., Saveliev, A. A., & Smith, G. M. (2009). Mixed effects models and extensions in ecology with R . Springer. https://doi.org/10.1007/978-0-387-87458-6 Additional Declarations The authors declare no competing interests. Supplementary Files SupplementalInformation.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6759348","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":462498236,"identity":"476c9415-dfd4-42ff-b9b3-05ae9c0ada95","order_by":0,"name":"Emily J. Chua","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxUlEQVRIiWNgGAWjYFACxsYDQDKBgYH5GJjbQISWBqgWtjRitTAwQLXwmBGnhV/scMNhHga7PIPjZ749+MFgI7vhAAEtkrMTQVqSiw3O5G437GFIMyaoxeA2WMuBxA0HcrdJMzAcTiSoxR6u5fybZ0At/wlrMZCGabmRwwbUcoCwFgmgLQfnGCQnzrzxzNywxyDZeCYhLfyz0x8+eFNhl9h3PvnZgx8VdrJ9hLSAABOPAdydRCgHAcYfRCocBaNgFIyCEQoAY9RKLC+/4JwAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-1605-5730","institution":"Boston College","correspondingAuthor":true,"prefix":"","firstName":"Emily","middleName":"J.","lastName":"Chua","suffix":""},{"id":462498238,"identity":"ac7934f1-c664-4a1a-aa69-9c2b573df2b0","order_by":1,"name":"John Supino","email":"","orcid":"https://orcid.org/0009-0006-1252-4815","institution":"Boston College","correspondingAuthor":false,"prefix":"","firstName":"John","middleName":"","lastName":"Supino","suffix":""},{"id":462498244,"identity":"704f588d-1b76-4018-b9ce-b9544ba96898","order_by":2,"name":"Kristen E. Fogaren","email":"","orcid":"https://orcid.org/0000-0002-6980-7668","institution":"Boston College","correspondingAuthor":false,"prefix":"","firstName":"Kristen","middleName":"E.","lastName":"Fogaren","suffix":""},{"id":462498246,"identity":"130054e1-f71f-4e84-8ec9-c8aadfee0076","order_by":3,"name":"Hilary I. Palevsky","email":"","orcid":"https://orcid.org/0000-0002-0488-4531","institution":"Boston College","correspondingAuthor":false,"prefix":"","firstName":"Hilary","middleName":"I.","lastName":"Palevsky","suffix":""}],"badges":[],"createdAt":"2025-05-27 12:16:56","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-6759348/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6759348/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83545526,"identity":"54c35429-f0ee-4727-9a49-20dd3aa77b40","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":2077479,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea)\u003c/strong\u003e Map of the Seven Mile Island Innovation Laboratory (SMIIL) study area in southern New Jersey, indicating locations of the three marsh channel observatories (“open-water platforms”) at which water quality monitoring sondes were deployed, and The Wetlands Institute (TWI) where ancillary weather data were obtained. \u003cstrong\u003eb)\u003c/strong\u003ePhoto depicting the open-water platform setup. The Aqua TROLL sonde is mounted on the end of a pole, which is submerged in the water during deployments and can be lifted to enable maintenance between deployments.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/455f998f930ef98b7d318dc4.png"},{"id":83545527,"identity":"e26b3eb6-4367-45ff-9a3e-a5c4378675e5","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1625349,"visible":true,"origin":"","legend":"\u003cp\u003eDaily means of environmental parameters at three open-water platforms in the Seven Mile Island Innovation Laboratory (July 2021–June 2024): a) temperature, \u003cstrong\u003eb) \u003c/strong\u003esalinity, \u003cstrong\u003ec)\u003c/strong\u003e DO %saturation (with 100% saturation shown as a horizontal black line), and \u003cstrong\u003ed)\u003c/strong\u003e pH, \u003cstrong\u003ee)\u003c/strong\u003etidal range, \u003cstrong\u003ef)\u003c/strong\u003e wind speed, along with daily rates of water-column metabolism: \u003cstrong\u003eg)\u003c/strong\u003e gross primary production (GPP), \u003cstrong\u003eh)\u003c/strong\u003e ecosystem respiration (ER), and \u003cstrong\u003ei) \u003c/strong\u003enet ecosystem metabolism (NEM). Gaps in data are due to sensor malfunction, biofouling, or sediment coverage. Wind speed data were obtained from the ERA5 atmospheric reanalysis.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/d969d1b60ea33416a56b90ac.png"},{"id":83546622,"identity":"e09b0daf-729b-47be-9372-b596f62bf87c","added_by":"auto","created_at":"2025-05-28 09:09:08","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":249224,"visible":true,"origin":"","legend":"\u003cp\u003eDiel fluctuations in pH and DO % saturation over week-long periods in \u003cstrong\u003ea)\u003c/strong\u003e January 2023 and \u003cstrong\u003eb)\u003c/strong\u003e July 2023 at Gull platform. Shading represents nighttime, and the horizontal black line denotes 100% DO saturation.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/871c7fc3e5950412a11db314.png"},{"id":83545534,"identity":"078bcc42-e525-4c49-89bf-3fc4369f5166","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":895605,"visible":true,"origin":"","legend":"\u003cp\u003eDaily and monthly climatological mean metabolic rates calculated from the entire time series across all three channel sites. Site means were taken by day of year for \u003cstrong\u003ea) \u003c/strong\u003egross primary production (GPP) and ecosystem respiration (ER) and \u003cstrong\u003ec)\u003c/strong\u003e net ecosystem metabolism (NEM), and by month of year for \u003cstrong\u003eb)\u003c/strong\u003e GPP and ER and \u003cstrong\u003ed)\u003c/strong\u003eNEM. The mean Monte Carlo propagated uncertainty is represented as shading in a) and c) and error bars in b) and d).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/a361a40725f46ce09cd0030c.png"},{"id":83545533,"identity":"f7d12c50-8c11-4080-b569-c0e9c168f65e","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1292100,"visible":true,"origin":"","legend":"\u003cp\u003eMultiple linear regressions of NEM vs. six environmental parameters, combining data from all three sites. Black points represent the season-specific values across all three years of sampling, while light grey points represent year-round values. Red lines represent the season-specific multiple regression relationships, varying the parameter indicated while holding the other parameters at their mean values. Light grey lines represent the year-round multiple regression relationships, calculated similarly. Response and predictor variables are scaled to (\u003cem\u003ex – x̄\u003c/em\u003e)/\u003cem\u003es\u003c/em\u003eand are therefore unitless.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/a2c1aea29e501e0a4953f84f.png"},{"id":83545529,"identity":"ec9ef7be-56e1-4aae-81e1-d3a8661cf12c","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":985043,"visible":true,"origin":"","legend":"\u003cp\u003eDynamic response of net ecosystem metabolism at all three channels sites to short-term perturbations induced by storms, as indicated by elevated daily mean wind speeds. Storm periods are indicated by red shading. Inset: Zoom in to Tropical Storm Ian in early October 2022; the pre- and post-storm reference periods are indicated by gray shading.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/a01886d77860617f57e6c350.png"},{"id":83545531,"identity":"efe7e113-1bc9-443f-be0e-ab27d63a6ec5","added_by":"auto","created_at":"2025-05-28 09:01:08","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":106787,"visible":true,"origin":"","legend":"\u003cp\u003eInfluence of storm events on net ecosystem metabolism (NEM), as compared to 5-day reference periods before and after each event. Data from all three sites (North, Gull, and South) were combined for a total of 19 storm events over the entire study period. On each box, the central mark indicates the median and the bottom and top edges indicate the 25th and 75th percentiles, respectively. The whisker endpoints correspond to the maximum and minimum data points that are not considered outliers, while the open circles denote outliers.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/87cc55650dbb48b343d027c0.png"},{"id":83546876,"identity":"51e03296-b657-4f5c-be03-591f6219abb5","added_by":"auto","created_at":"2025-05-28 09:17:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10251040,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/dfe2200c-b638-4eb2-bfa1-4d70ca43fde5.pdf"},{"id":83545525,"identity":"afeb3422-f31c-4965-a7ed-cd28bf692505","added_by":"auto","created_at":"2025-05-28 09:01:07","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1423688,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-6759348/v1/6108ce8b546a9f9bcc7573f6.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eMultiyear monitoring reveals seasonal and short-term dynamics of ecosystem metabolism in a temperate salt marsh channel\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCoastal waters are a critical component of the global carbon budget, with estuaries and their surrounding vegetation processing and cycling substantial amounts of carbon. In particular, intertidal salt marshes (and their tropical counterparts, mangroves) are among the most productive ecosystems on Earth, acting as key conduits for transferring inorganic carbon from the atmosphere to the ocean (Alongi, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Salt marshes, which are found worldwide and are the dominant intertidal habitat along the U.S. East and Gulf coasts (Pennings \u0026amp; Bertness, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), provide not only a significant carbon sink but also essential ecosystem services: they support biodiversity, sustain fisheries, improve water quality by filtering pollutants, and protect shorelines from sea level rise and storms (Barbier et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSalt marshes are increasingly vulnerable to a range of threats, complicating mitigation and restoration efforts. Climate-driven pressures, such as sea-level rise, altered hydrology, and more extreme weather events, can destabilize sediments and cause widespread loss of vegetation, degrading the marsh (Rolando et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Valiela et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). These climatic impacts are compounded by human-induced stressors like coastal development, pollution, and changes in sediment supply (Gedan et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Migration and loss of marshes can release large amounts of stored carbon to the atmosphere, shifting salt marshes from being net carbon sinks to carbon sources (Warnell et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Understanding and managing these dynamic systems requires process-based indicators that can serve as early warnings of ecosystem functional changes.\u003c/p\u003e \u003cp\u003eAquatic ecosystem metabolism offers such a process-based indicator. First conceptualized by Odum (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e1956\u003c/span\u003e), aquatic ecosystem metabolism integrates all biological activity\u0026mdash;organic matter production and consumption\u0026mdash;within a body of water. Gross primary production (GPP) represents the total rate of organic matter produced in an ecosystem, while ecosystem respiration (ER), also known as community respiration, captures the total rate of organic matter consumed in an ecosystem via aerobic respiration. Their balance, termed net ecosystem metabolism (NEM), reflects the metabolic status of an aquatic system\u0026mdash;whether it is a net source or sink of organic carbon. These metrics provide information on the overall carbon balance of an ecosystem, and how it responds to changes that perturb this balance.\u003c/p\u003e \u003cp\u003eHistorically, assessing the response of aquatic ecosystems to environmental change has primarily occurred through discrete measurements of dissolved oxygen (DO) concentrations (Jankowski et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). DO dynamics are influenced by both physical processes (gas exchange, water mixing) and biological processes (primary production and respiration). Regulatory assessments often rely on comparing these discrete measurements to threshold values, such as minimum levels to sustain aquatic life. However, advances in autonomous oxygen sensors in recent decades have enabled high-frequency DO data collection, which, when combined with high-frequency temperature and salinity measurements, enable continuous estimates of metabolic rates.\u003c/p\u003e \u003cp\u003eThese high-temporal resolution metabolism data provide unprecedented insight into short-term variability and long-term trends in aquatic system functioning. Ecosystem metabolism is thought to respond to a wide range of anthropogenic disturbances, such as altered water column mixing due to climatic changes and increased suspended sediment loads\u0026mdash;however, the ecological consequences of these perturbations are not yet well understood (Jankowski et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Kemp \u0026amp; Testa, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). While small changes in physicochemical parameters might not be detectable on their own, when translated into metabolic rates these subtle shifts can reveal significant functional changes in the ecosystem (Jankowski et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). For example, metabolic rates calculated along stressor gradients\u0026mdash;such as urbanization or eutrophication\u0026mdash;can identify shifts in ecosystem function. Similarly, time series analysis of metabolic rates can reveal responses to disturbance events and track recovery or degradation. Long-term metabolism records are particularly valuable for distinguishing natural variability from climate-driven trends.\u003c/p\u003e \u003cp\u003eIn salt marsh systems, carbon dynamics are especially complex due to their heterogeneous features and tight coupling of terrestrial and aquatic processes. Salt marshes are dynamic, interconnected mosaics of carbon sources and sinks: while marsh vegetation sequesters atmospheric carbon dioxide (CO\u003csub\u003e2\u003c/sub\u003e) and stores \u0026ldquo;blue carbon\u0026rdquo; in their sediments, the adjacent aquatic environments\u0026mdash;tidal channels and estuaries\u0026mdash;often act as net CO\u003csub\u003e2\u003c/sub\u003e sources (Cai, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Dai et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Laruelle et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Rosentreter et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This paradox is explained by the \u0026ldquo;marsh CO\u003csub\u003e2\u003c/sub\u003e pump\u0026rdquo; concept, wherein CO\u003csub\u003e2\u003c/sub\u003e fixed by marsh plants is later respired in adjacent waters, and either directly emitted back to the atmosphere or exported as dissolved inorganic carbon (DIC) to the coastal ocean (Z. A. Wang et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Z. A. Wang \u0026amp; Cai, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). The strength of this pump varies seasonally, with peak CO\u003csub\u003e2\u003c/sub\u003e uptake by marsh plants in spring and early summer, followed by greater carbon export from the marsh system in late summer and fall. DIC exported to, and ultimately sequestered in, the deep ocean could surpass the amount stored in marsh sediments, suggesting that the marshes\u0026rsquo; true carbon sink capacity may be underestimated (Z. A. Wang et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe spatial and temporal variability of marsh carbon fluxes contributes to persistent underrepresentation of estuarine environments in Earth system models, which struggle to capture the full complexity of coastal ocean dynamics and their implications for regional and global climate projections (Bauer et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Dai et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Herrmann et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Several synthesis studies have sought to refine carbon budgets for estuaries and the coastal ocean by integrating estimates of ecosystem metabolism. One such global synthesis by Bauer et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) identified significant uncertainty in estuarine net carbon balances due to their spatial and temporal heterogeneity in carbon processing and fluxes, as well as the difficulty in scaling up relatively few observational studies. Another synthesis by Herrmann et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), focused on the East Coast of the U.S., found that regional estuarine waters are generally net heterotrophic but identified NEM in the Mid Atlantic Bight as the most uncertain term in their budget, emphasizing the need for further data in this region.\u003c/p\u003e \u003cp\u003eEnvironmental change, driven by climate or human activity, will likely shift the balance between primary production and respiration in coastal ecosystems (Kemp \u0026amp; Testa, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Climate change, including regionally rising temperatures, changes in precipitation patterns, and increased storm frequency, is already underway (IPCC, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), while human-induced eutrophication (over-enrichment with nutrients or organic matter) has resulted in more frequent and intense hypoxic conditions in estuaries worldwide (Dai et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For instance, warming tends to enhance respiration more than primary production in marine environments (Boscolo-Galazzo et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), shifting the system toward heterotrophy, which leads to greater dissolution of metabolic CO\u003csub\u003e2\u003c/sub\u003e and increases acidification risk (Baumann \u0026amp; Smith, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Furthermore, few studies (e.g., Buelo et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Howarth et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Tassone \u0026amp; Bukaveckas, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) have captured how short-term (daily to weekly) events, such as storms or changes in freshwater discharge, affect estuarine ecosystem metabolism. Gaining a clearer understanding of the current metabolic balance in estuaries, and the factors that influence it, is essential for predicting how carbon dynamics in these systems will respond to future environmental changes.\u003c/p\u003e \u003cp\u003eTo address these gaps and improve our understanding of the metabolic balance in temperate salt marshes, we conducted a multi-year, high-frequency monitoring campaign in a tidal marsh channel in southern New Jersey. Over a three-year period, we collected continuous measurements of key physical and biogeochemical parameters\u0026mdash;temperature, salinity, dissolved oxygen, pH, chlorophyll \u003cem\u003ea\u003c/em\u003e, and turbidity\u0026mdash;at three sites in the marsh channel. Our study had three primary objectives: (1) to quantify aquatic ecosystem metabolism in the channel and characterize seasonal dynamics of environmental conditions and metabolism, establishing baseline conditions; (2) to explore associations between environmental factors and metabolic rates over seasonal and annual timescales; and (3) to assess the metabolic impacts of perturbations from episodic storm events. We also discuss the sources of uncertainty in our metabolic rate estimates and compare our results to other marsh-dominated estuaries. This work contributes valuable high-resolution data from the mid-Atlantic U.S. and improves our understanding of how salt marsh systems respond to environmental change, which may inform future management strategies for marsh conservation and restoration in the face of climate change.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Site\u003c/h2\u003e \u003cp\u003eContinuous water quality monitoring was conducted in the marsh tidal channel system landward of Seven Mile Island, a populated barrier island located in Cape May County on the southern coast of New Jersey, from June 2021\u0026ndash;June 2024 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). This expanse of salt marsh and tidal channels is part of the Seven Mile Island Innovation Laboratory (SMIIL), a designated research area that spans the coastal region from Townsends Inlet in Avalon to Hereford Inlet in Stone Harbor, bisected by the New Jersey Intracoastal Waterway. SMIIL was established in 2019 through a partnership between the U.S. Army Corps of Engineers (USACE) Philadelphia District, the USACE Engineer Research and Development Center (ERDC), the State of New Jersey, and The Wetlands Institute (TWI) to advance and improve dredging and marsh restoration techniques (Chasten et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Fall et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The area encompasses over 62 km\u003csup\u003e2\u003c/sup\u003e of state-owned marshland, including tidal marshes, coastal lagoons, shallow bays, sounds, and tidal inlets, and is part of the Cape May Wetlands Wildlife Management Area. Average low tide depths are 0.6 m (Perkey et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and tidal conditions are mixed semi-diurnal with a tidal range of approximately 1\u0026ndash;2 m (Fall et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). No rivers or streams drain into the region.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMonitoring Data\u003c/h3\u003e\n\u003cp\u003eIn June 2021, three observational \u0026ldquo;open-water\u0026rdquo; platforms were constructed in the main channel of SMIIL to facilitate water quality monitoring efforts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e): North platform near the channel entrance (39.103\u003csup\u003eo\u003c/sup\u003eN, 74.765\u003csup\u003eo\u003c/sup\u003eW), Gull platform off the southern edge of Gull Island (39.072\u003csup\u003eo\u003c/sup\u003eN, 74.778\u003csup\u003eo\u003c/sup\u003eW), and South platform near the channel exit (39.044\u003csup\u003eo\u003c/sup\u003eN, 74.788\u003csup\u003eo\u003c/sup\u003eW). These platforms are henceforth referred to as \u0026ldquo;North,\u0026rdquo; \u0026ldquo;Gull,\u0026rdquo; and \u0026ldquo;South.\u0026rdquo; All three platform sites are shallow, with mean water depths of 1.26 m (North), 1.18 m (Gull), and 2.49 m (South; Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). At each platform, two Aqua TROLL 600 multiparameter sondes (In-Situ Inc.) were mounted on a pole attached to the platform, approximately 37\u0026ndash;75 cm off the bottom. Both sondes deployed at a platform contained sensors for conductivity, temperature, dissolved oxygen, and pH; additionally, one sonde contained a chlorophyll \u003cem\u003ea\u003c/em\u003e sensor and the other a turbidity sensor.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of physical and chemical characteristics of the three channel sites over the entire time series (2021\u0026ndash;2024). Reported errors are the measurement uncertainty (1 standard deviation). Seasonal range is defined as the 5th\u0026ndash;95th percentile range per year, averaged across all years in the time series. Missing data points were gap-filled using a smoothed climatological day-of-year mean, derived from a 10-day moving average over the entire time series.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"11\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDepth (m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTidal range\u003c/p\u003e \u003cp\u003e(m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eTemperature\u003c/p\u003e \u003cp\u003e(\u003csup\u003eo\u003c/sup\u003eC)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eSalinity\u003c/p\u003e \u003cp\u003e(psu)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e \u003cp\u003eDissolved oxygen\u003c/p\u003e \u003cp\u003e(%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c11\" namest=\"c10\"\u003e \u003cp\u003epH\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.71\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.84\u0026ndash;25.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e32.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e29.4\u0026ndash;34.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e94.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e74.4\u0026ndash;104.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e8.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e7.87\u0026ndash;8.31\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGull\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15.14\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.96\u0026ndash;26.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e32.4\u0026thinsp;\u0026plusmn;\u0026thinsp;0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e29.3\u0026ndash;35.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e89.8\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e65.9\u0026ndash;104.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e8.02\u0026nbsp;\u0026plusmn; 0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e7.71\u0026ndash;8.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSouth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.45\u0026thinsp;\u0026plusmn;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.49\u0026ndash;24.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e32.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e29.8\u0026ndash;35.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e92.6\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e74.6\u0026ndash;104.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e8.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c11\"\u003e \u003cp\u003e7.88\u0026ndash;8.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFrom June 2021 to June 2024, the sondes collected near-continuous data at the three open-water platforms. Data were collected at 6-minute intervals prior to August 2022; after that, recording was set to 10-minute intervals. Initially, the Aqua TROLLs were configured for remote monitoring via telemetry; however, following issues with data dropouts in the first year of deployment, the sondes were switched to internal data logging. Throughout the three-year monitoring period, maintenance visits were conducted regularly to the platforms to download data from the Aqua TROLLs and replace them with cleaned and newly calibrated sondes. Sondes were maintained every 4\u0026ndash;12 weeks, with an average deployment length of 9 weeks.\u003c/p\u003e \u003cp\u003eStarting in September 2022, discrete surface water samples were obtained during maintenance visits for later laboratory analysis of dissolved oxygen (DO), dissolved inorganic carbon (DIC), and total alkalinity (TA). Discrete sampling occurred near the end of a deployment, just before a sonde was recovered from a platform, and after replacement with a re-calibrated sonde. Samples for DO were collected into volume-calibrated flasks and preserved for Winkler analysis following standard protocols (Langdon \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2010\u003c/span\u003e); the bottle necks were sealed with water to further permit long-term storage (Zhang et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). DO discrete samples were titrated within a week of collection using a custom-built Winkler titrator with automated potentiometric end point detection (control software available via Nicholson et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Samples for DIC and TA were collected into borosilicate glass bottles and preserved with saturated mercuric chloride following standard protocols (Dickson et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Discrete samples for DIC and TA were analyzed using an AS-C6L DIC Analyzer and an AS-ALK2 TA Analyzer (Apollo SciTech Instruments), respectively. DIC and TA were measured from the same sample bottle, with DIC measurements made on the day the bottle was first opened and TA measurements made the same week.\u003c/p\u003e\n\u003ch3\u003eData Quality Control\u003c/h3\u003e\n\u003cp\u003eWater quality time series data\u0026mdash;including depth, temperature, salinity, DO concentration, pH, turbidity, and chlorophyll \u003cem\u003ea\u003c/em\u003e\u0026mdash;underwent a multi-step quality control process adapted from the U.S. Integrated Ocean Observing System (IOOS) Quality Assurance/Quality Control of Real-Time Oceanographic Data (QARTOD) recommendations (IOOS, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Palevsky et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Initial quality control involved human-in-the-loop (HITL) inspection of depth data, with annotations cross-referenced against field logs to flag periods of known sensor malfunction or when sensors were briefly out of water between deployments. Flagged points were removed from further analysis for all parameters. Automated tests followed, including gross range checks and spike detection using parameter-specific thresholds (Table S1 in Supplemental Information). Values exceeding these thresholds were flagged and excluded. The cleaned data were then binned to 10-minute bins to resolve irregular sampling times. A moving median test was subsequently applied to all parameters\u0026mdash;excluding depth and temperature due to their strong diel oscillations\u0026mdash;to identify and remove additional outliers (Leys et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). This test used a double median absolute deviation (MAD) approach to account for asymmetrical distributions, with data points exceeding\u0026thinsp;\u0026plusmn;\u0026thinsp;3 respective MADs from the moving median flagged and removed (Tan et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor parameters with duplicate measurements (salinity, temperature, DO concentration, and pH), cleaned time series were further evaluated to produce a single final dataset per parameter. Differences between duplicates were assessed using absolute difference thresholds and visual inspection. For each deployment, the final dataset was chosen from either the more reliable sensor or the mean of both sensors, based on data completeness, agreement with adjacent deployments, and consistency with supporting data. For salinity, which was particularly noisy, data were first smoothed with a 24-hour moving median before selection. Comparisons between related parameters aided this assessment\u0026mdash;for example, low pH values were cross-checked against DO data to identify potential hypoxic or anoxic events. Discrete samples were used to validate the sonde data: Winkler titrations for DO, and DIC and TA for pH. For pH validation, sonde data were converted to hydrogen ion concentration ([H+]) space, and [H+] values from DIC and TA measurements were calculated using the CO2SYSv3 MATLAB program (Lewis \u0026amp; Wallace, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Sharp, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Van Heuven et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), with propagated uncertainties following Orr et al. (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Discrete salinity samples were occasionally available for spot-checking. To prevent error propagation from pressure, salinity, and temperature measurements, DO concentrations were recalculated using the cleaned environmental variables and salinity- and pressure-compensation equations (Aanderaa, 2017). DO % saturation was then calculated by normalizing the measured DO concentration to the saturation DO concentration (DO % saturation\u0026thinsp;=\u0026thinsp;DO\u003csub\u003emeasured\u003c/sub\u003e / DO\u003csub\u003esat\u003c/sub\u003e​ \u0026times;100%), where DO\u003csub\u003esat\u003c/sub\u003e was computed as a function of temperature and salinity (Garcia \u0026amp; Gordon, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1992\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor turbidity and chlorophyll \u003cem\u003ea\u003c/em\u003e, each measured by a single sensor per platform, a final manual inspection was conducted after the moving median test. These optical sensors are especially vulnerable to signal drift and noise from biofilm accumulation, which obstructs light transmission through the water column and was not fully remedied by earlier quality control steps. Persistent, year-round biofouling was observed at all platforms, consistent with a recent study at the SMIIL site that reported impaired turbidity sensor performance (Perkey et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Time series were reviewed on a deployment-by-deployment basis across platforms, and data affected by suspected biofouling were removed. Due to the extent of data loss, turbidity and chlorophyll \u003cem\u003ea\u003c/em\u003e records are presented only in the Supplementary Information (Fig. S1e\u0026ndash;f).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo constrain measurement error in the Aqua TROLL data, we used the mean absolute difference between duplicate sensor readings\u0026mdash;taken during representative deployments where both sensors were functioning\u0026mdash;as an estimate of two standard deviations (2σ). This approach was applied to parameters for which duplicate sensors were deployed at a platform (i.e., salinity, temperature, DO, and pH). Averaged across all three platforms, the resulting standard deviations were 0.28 psu for salinity, 0.025 \u003csup\u003eo\u003c/sup\u003eC for temperature, 2.88 \u0026micro;mol L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for DO, and 0.02 for pH. Although duplicate depth sensors were also deployed, the vertical offset between the sensors varied slightly from deployment to deployment (on the order of a few cm), precluding a precise estimate of the depth measurement error; we estimate this to be \u0026lt;\u0026thinsp;5 cm. For DO, we also performed linear regressions comparing the best-guess Aqua TROLL values to discrete Winkler data, using the standard error of the mean as an independent error estimate for each platform. In subsequent uncertainty analyses (see \u0026ldquo;Aquatic Ecosystem Metabolism\u0026rdquo;), we used the larger of the two DO error estimates (duplicate sensor or Winkler-based) for each platform: the duplicate sensor method yielded higher errors for North and Gull, while the Winkler comparison did for South.\u003c/p\u003e\n\u003ch3\u003eMeteorological Data Sources\u003c/h3\u003e\n\u003cp\u003eA ResponseON Weather Transmitter (R.M. Young Company) was also installed on the Gull platform to measure wind speed and direction, air temperature, relative humidity, and atmospheric pressure. Measurements were made at 6-minute intervals. The weather station record ended prematurely due to data transmission issues, and spans September 2021\u0026ndash;May 2023. Air temperature and atmospheric pressure data were also obtained from a HOBO water level logger (Onset Computer Corp.) deployed at The Wetlands Institute, which is centrally located in SMIIL (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). HOBO data were collected at 10-minute intervals from August 2022\u0026ndash;June 2024.\u003c/p\u003e \u003cp\u003eThe air temperature and atmospheric pressure data from the Gull meteorological station and HOBO sensor were combined to produce a time series that spanned the open-water platform water quality measurement record. Similarly, to fill gaps in the Gull meteorological station wind speed record, the hourly 10-m \u003cem\u003eu-\u003c/em\u003e and \u003cem\u003ev-\u003c/em\u003ecomponents of wind were obtained from the fifth generation European Centre for Medium-Range Weather Forecasts atmospheric reanalysis (ERA5; Hersbach et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) for the grid cell nearest to 39.04\u003csup\u003eo\u003c/sup\u003eN, 74.79\u003csup\u003eo\u003c/sup\u003eW and used to calculate the 10-m wind speed. Given the close agreement between the Gull wind speed measurements and the ERA5 data (Fig. S2 in Supplemental Information), the latter was used in subsequent analyses across the entire water quality measurement record.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo constrain measurement error for the meteorological data, we determined 2σ from comparisons between two measurement sources, similar to the approach used for the water quality monitoring data. For atmospheric pressure, we used the mean absolute difference between the Gull meteorological station and HOBO sensor data over overlapping measurement periods, resulting in an uncertainty of 2.54 hPa. For wind speed, we used the mean absolute difference between the Gull meteorological station hourly means and ERA5 wind speeds, resulting in an uncertainty of 1.53 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eAquatic Ecosystem Metabolism\u003c/h3\u003e\n\u003cp\u003eThe measured DO time series data were leveraged to estimate ecosystem metabolism rates in the main marsh channel following the \u0026ldquo;open-water\u0026rdquo; or \u0026ldquo;diel oxygen\u0026rdquo; method pioneered by Odum (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e1956\u003c/span\u003e). This method is predicated on the notion that during a daily cycle, three main processes affect the dissolved oxygen concentration of a water mass: photosynthetic production, respiration, and exchange of oxygen across the air-water interface. Metabolic rates are related to rates of dissolved oxygen change via the mass balance equation,\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}\\frac{\\partial\\:\\text{D}\\text{O}}{\\partial\\:t}=P-R+D\\:\\#\\left(1\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cem\u003e\u0026part;\u003c/em\u003eDO\u003cem\u003e/\u0026part;t\u003c/em\u003e is the rate of change of DO concentration with respect to time, \u003cem\u003eP\u003c/em\u003e is the photosynthetic rate, \u003cem\u003eR\u003c/em\u003e is the respiration rate, and \u003cem\u003eD\u003c/em\u003e is the diffusive rate of oxygen exchange across the air-water interface, all expressed on a volumetric basis (mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003eA main assumption of the diel oxygen method is that measured changes in dissolved oxygen arise from biological production/consumption and gas exchange, and the effects of physical transport on the DO at a fixed sampling location are negligible. However, in estuaries like SMIIL that are strongly influenced by tidal transport and mixing, the advection of water masses with different DO histories past the measurement point can influence DO measurements. To reduce the effects of physical transport on metabolism estimates, a weighted regression model developed by Beck et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) was applied to the measured DO data. This method targets the periodicity of the tidal component while preserving the biological signal and was implemented via the R package \u003cem\u003eWtRegDO\u003c/em\u003e. The resulting \u0026ldquo;detided\u0026rdquo; oxygen data, DO\u003csub\u003edetided\u003c/sub\u003e was then used to calculate metabolic rates using Eq.\u0026nbsp;1, as described below. This approach substantially reduced the occurrence of anomalous GPP and ER results (i.e., negative GPP and positive ER) compared to calculations performed on the raw DO data. Remaining anomalous values may reflect cases where values fall below the detection limit for this method (Caffrey et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAir-water oxygen exchange (\u003cem\u003eD\u003c/em\u003e in Eq.\u0026nbsp;1) was modeled as the product of a gas-exchange coefficient (aka volumetric aeration coefficient or piston velocity), \u003cem\u003ek\u003c/em\u003e (h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and the difference between the DO saturation concentration, DO\u003csub\u003esat\u003c/sub\u003e and the measured detided DO concentration (both in mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e):\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}D=k\\left({\\text{D}\\text{O}}_{\\text{s}\\text{a}\\text{t}}-\\:{\\text{D}\\text{O}}_{\\text{d}\\text{e}\\text{t}\\text{i}\\text{d}\\text{e}\\text{d}}\\right)\\#\\left(2\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eDO\u003csub\u003esat\u003c/sub\u003e is the concentration of oxygen in water that is in equilibrium with the atmosphere, typically at sea level pressure, calculated as a function of water temperature and salinity (Garcia \u0026amp; Gordon, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1992\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe gas-exchange coefficient is commonly estimated from statistical relationships with wind speed and/or tidal currents (Howard et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Kemp \u0026amp; Testa, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Needoba et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Many efforts to parameterize \u003cem\u003ek\u003c/em\u003e have been proposed (e.g., Cole and Caraco \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Emerson et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Liang et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Ro and Hunt \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Stanley et al. \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Vachon and Prairie \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Wanninkhof \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), and the choice of an appropriate \u003cem\u003ek\u003c/em\u003e parameterization depends on the specific environmental conditions. Howard et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) evaluated four methods within a salt-marsh environment; however, a comprehensive comparison of their impact on metabolic rate estimates does not yet exist. As such, we performed a sensitivity analysis using three parameterizations: Ro \u0026amp; Hunt (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), Wanninkhof (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), and Emerson et al. (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Ro \u0026amp; Hunt (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) is commonly used in other studies of estuaries (Beck et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Caffrey et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Murrell et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Roberts et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Wallace et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Wanninkhof (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) was selected for its accuracy in predicting noble gas saturation in a salt marsh pond (Howard et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). As these first two parameterizations do not account for gas transfer through bubbles, we included Emerson et al. (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), which explicitly accounts for bubble-mediated flux.\u003c/p\u003e \u003cp\u003eWe calculated gas-exchange coefficients and metabolic rates using these parameterizations. The mean absolute differences between the three methods were similar, with Wanninkhof (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) yielding intermediate results, so it was chosen for our ecosystem metabolism analysis. In this parameterization, \u003cem\u003ek\u003c/em\u003e is a function of wind speed at 10-m height, atmospheric pressure, air temperature, depth of the water column at the sampling site, and the Schmidt number; the latter is gas-specific and depends on the water temperature and to a lesser degree on the salinity. The ERA5 wind speeds and combined atmospheric pressure and air temperature data from the Gull meteorological station and HOBO sensor were used for the \u003cem\u003ek\u003c/em\u003e calculations, which were implemented in MATLAB using the \u003cem\u003egas_toolbox\u003c/em\u003e package (Manning \u0026amp; Nicholson, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This parameterization provides reliable estimates of the gas-exchange coefficient for intermediate wind speeds (3\u0026ndash;15 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e; Wanninkhof, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), with only 0.06% of our observed wind speeds exceeding this range and 12.6% falling below it. To constrain uncertainty in \u003cem\u003ek\u003c/em\u003e, we used the mean absolute difference between the two parameterizations that produced the largest and smallest results (i.e., Emerson et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e and Ro \u0026amp; Hunt, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe change in DO due to biological processes was calculated from Eq.\u0026nbsp;1,\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}\\frac{\\partial\\:{\\text{D}\\text{O}}_{\\text{b}\\text{i}\\text{o}}}{\\partial\\:t}=P-R=\\frac{\\partial\\:{\\text{D}\\text{O}}_{\\text{d}\\text{e}\\text{t}\\text{i}\\text{d}\\text{e}\\text{d}}}{\\partial\\:t}-D\\:\\#\\left(3\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eand averaged separately during daylight periods to compute hourly rates of net (aka apparent) primary production (\u003cem\u003eP\u003c/em\u003e), and during night periods to compute hourly rates of nighttime respiration (\u003cem\u003eR\u003c/em\u003e). To calculate daily rates of respiration and production, respiration rates were assumed to remain constant during day and night. Daily total respiration, \u003cem\u003eR\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e and daily gross production, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e (both in units of mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) were consequently calculated as:\u003cdiv id=\"Equd\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}{R}_{t}\\:=\\:R\\:\\times\\:\\:24\\:hours\\#\\left(4\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Eque\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Eque\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}{P}_{g}\\:=\\left(P-R\\right)\\times\\:\\:daylength\\#\\left(5\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eDaylength (in hours) was determined based on local sunrise and sunset times.\u003c/p\u003e \u003cp\u003eThese volumetric daily rates were multiplied by the daily mean water depth at the respective open-water platform to yield areal rates of gross primary production (GPP) and ecosystem respiration (ER) in mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Net ecosystem metabolism was then calculated as\u003cdiv id=\"Equf\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equf\" name=\"EquationSource\"\u003e\n$$\\:\\begin{array}{c}NEM\\:=\\:GPP\\:+\\:ER\\#\\left(6\\right)\\end{array}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eNEM reflects the overall balance between the production and respiration of organic matter by all organisms within the ecosystem. A positive NEM indicates the system is net autotrophic, producing excess organic carbon, while a negative NEM suggests the system is net heterotrophic, consuming organic carbon. When NEM approaches zero, the system is considered to be in metabolic balance.\u003c/p\u003e \u003cp\u003eThe uncertainties in each measured parameter used in the diel oxygen analysis contribute to uncertainty in each term of the mass balance (Eq.\u0026nbsp;1), and therefore the overall uncertainty in the calculated values for the metabolic rates. Uncertainties from individual parameters were propagated through the diel analysis via a Monte Carlo technique (e.g., Albert, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) using the mean standard deviation for each input parameter. First, the uncertainty in the air-sea gas exchange (Eq.\u0026nbsp;2) was estimated via a Monte Carlo simulation iterated 10\u003csup\u003e4\u003c/sup\u003e times using the uncertainties in salinity, temperature, depth (or pressure), DO concentration, wind speed, atmospheric pressure, and choice of \u003cem\u003ek\u003c/em\u003e parameterization. Next, the uncertainties in the hourly rates of net primary production and nighttime respiration were calculated via a Monte Carlo simulation iterated 10\u003csup\u003e4\u003c/sup\u003e times containing the uncertainty in \u003cem\u003eD\u003c/em\u003e and converted to uncertainties in the daily rates, \u003cem\u003eR\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e (Equations 4 and 5). The uncertainties in \u003cem\u003eR\u003c/em\u003e\u003csub\u003e\u003cem\u003et\u003c/em\u003e\u003c/sub\u003e, \u003cem\u003eP\u003c/em\u003e\u003csub\u003e\u003cem\u003eg\u003c/em\u003e\u003c/sub\u003e, and depth were propagated in a final Monte Carlo simulation iterated 10\u003csup\u003e4\u003c/sup\u003e times to calculate GPP, ER, and NEM (Eq.\u0026nbsp;6).\u003c/p\u003e \u003cp\u003eDetermining metabolic fluxes relies on accurately estimating air-water dissolved oxygen exchange driven by wind mixing, especially in shallow systems. However, gas-exchange parameterizations are site-specific and fluxes calculated using different methods can range over an order of magnitude, often making gas exchange flux the largest error source in estimating metabolic rates via diel analysis (Howard et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The primary contributors to the uncertainty in our NEM estimates, determined by isolating one error source while setting all others to zero in the Monte Carlo simulation (data not shown), were the wind speed values and the choice of gas exchange parameterization. As described earlier, we found the ERA5 wind speed data overestimated the Gull meteorological station data by an average of 1.53 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Consequently, our metabolic rate estimates are likely slight overestimates.\u003c/p\u003e \u003cp\u003eLastly, while we quantified aquatic ecosystem metabolism in terms of oxygen units, these values are not readily converted to carbon units. In shallow water systems, O\u003csub\u003e2\u003c/sub\u003e and CO\u003csub\u003e2\u003c/sub\u003e dynamics can become decoupled, and standard open-ocean stoichiometric ratios may not apply. For instance, S. R. Wang et al. (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) observed large seasonal variability in the respiration quotient (CO\u003csub\u003e2\u003c/sub\u003e : O\u003csub\u003e2\u003c/sub\u003e) from 0.5 to 1.5 in a marsh-dominated estuary in Georgia, U.S. This variability arises in part because anaerobic respiration\u0026mdash;prevalent in salt marsh and estuarine sediments, especially during warmer months\u0026mdash;does not consume O\u003csub\u003e2\u003c/sub\u003e but still produces CO\u003csub\u003e2\u003c/sub\u003e. Accurately estimating metabolic fluxes in carbon units would therefore require a full characterization of the carbonate system, which is beyond the scope of this paper.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistics\u003c/h2\u003e \u003cp\u003eTo explore the relationships between environmental conditions and metabolic rates, we calculated bivariate associations between GPP, ER, and NEM and various environmental parameters using daily-averaged data pooled from all three open-water platforms. All response and explanatory variables were scaled to have a mean of zero and a standard deviation of one, i.e., (\u003cem\u003ex \u0026ndash; x̄\u003c/em\u003e)/\u003cem\u003es\u003c/em\u003e. This standardization ensured that variables with different ranges of variation were on the same scale, allowing direct comparisons between them.\u003c/p\u003e \u003cp\u003eWe selected six physical and biological parameters likely associated with estuarine metabolism based on previous studies: salinity and temperature (Bas-Silvestre et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Caffrey, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Gomez-Castillo et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Nelson et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Russell \u0026amp; Montagna, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), DO % saturation (Russell \u0026amp; Montagna, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), pH (Baumann \u0026amp; Smith, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Lowe et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Russell \u0026amp; Montagna, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), daily tidal range (Gomez-Castillo et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and wind speed (Nelson et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Russell \u0026amp; Montagna, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). For all statistical analyses, results were considered statistically significant at α\u0026thinsp;=\u0026thinsp;0.05 and highly significant at α\u0026thinsp;=\u0026thinsp;0.01.\u003c/p\u003e \u003cp\u003eNext, we assessed relative strength of associations between environmental variables and metabolic rates using multiple linear regression. To evaluate collinearity (i.e., when explanatory variables are correlated; Zuur et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), we calculated pairwise correlation coefficients (\u003cem\u003er\u003c/em\u003e) and variance inflation factors (VIFs). Collinearity is typically considered high when |\u003cem\u003er\u003c/em\u003e| \u0026gt; 0.7 and VIF\u0026thinsp;\u0026gt;\u0026thinsp;5 or 10 (Dormann et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Zuur et al., \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2009\u003c/span\u003e); we adopted the more conservative VIF threshold of 5. For each response variable (GPP, ER, and NEM), we fitted a multiple linear regression model using the complete annual dataset. In addition to the six environmental explanatory variables, the models also controlled for site (Gull, North, and South), season (winter, spring, summer, and fall), and the interaction between site and season to account for site- and season-specific effects. Including site as a fixed effect allowed us to isolate variations in metabolic rates attributed to environmental factors, rather than site-specific characteristics (e.g., differing depths). We also fitted separate models for each season, using the same explanatory variables as the annual model, while controlling for site. All regression models were conducted in MATLAB using the \u003cem\u003efitlm\u003c/em\u003e function, and VIFs were calculated with the \u003cem\u003evif\u003c/em\u003e function (Vasilaky, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe also examined the effect of short-term perturbations from storm events on metabolic rates. Storm events were identified as periods between December 2021\u0026ndash;June 2024 when the daily averaged ERA5 wind speed exceeded the 99th percentile threshold (11.9 m s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). We then calculated the mean NEM values during the storm event, as well as the means for pre- and post-event periods (5 days in length each). Paired t-tests were used to compare NEM rates before, during, and after storm events using the values pooled from all three platforms using MATLAB\u0026rsquo;s \u003cem\u003ettest\u003c/em\u003e function.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eEnvironmental Parameters\u003c/h2\u003e \u003cp\u003eSeasonal temperature patterns were nearly identical across all three sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Mean monthly temperatures were lowest during winter months (December\u0026ndash;February) and highest during summer months (June\u0026ndash;August), with a mean seasonal range of 21.47\u003csup\u003eo\u003c/sup\u003eC across all sites. The overall site-wide mean temperature was 14.77\u003csup\u003eo\u003c/sup\u003eC. Similarly, seasonal salinity patterns were consistent between sites, with slightly higher salinities during warmer months (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). The site-wide mean salinity was 31.2 psu with a mean seasonal range of 5.4 psu. A slight freshening trend was observed at all sites in the final year of measurement, reaching an average minimum value of 26.7 psu in late December 2023/early January 2024. This trend was also observed at water-quality monitoring stations (Buoy 126 and Buoy 139) in nearby Great Bay, NJ, approximately 36 mi up the coast, as reported by the Jacques Cousteau National Estuarine Research Reserve (data not shown).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDissolved oxygen % saturation exhibited strong seasonal variation that was the inverse of temperature, peaking in winter and reaching its lowest values in summer (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). The mean seasonal range across sites was 32.5%. Daily variability was greatest during warmer months, when elevated biological activity led to more intense diel swings: mid- to late-summer values ranging from hypoxic (\u0026lt;\u0026thinsp;30%) in the early morning to 100% saturation or supersaturation by early evening, reflecting heightened photosynthesis and respiration (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In contrast, colder temperatures in winter enhanced oxygen solubility, and DO levels remained consistently near or above saturation throughout the day, with biological influences less dominant. Site means ranged from 89.7 to 94.8% (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), with Gull showing consistently lower saturation levels than North and South, especially during summer months. Seasonal pH patterns mirrored DO saturation, with annual highs in winter and annual lows in summer (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Similar to DO, daily variability was higher in summer than in winter (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Overall mean pH ranged from 8.02 to 8.11, with a mean seasonal range of 0.47. As with DO, pH at Gull was consistently lower than at the other two sites.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDaily mean tidal range was consistent across all sites, varying predictably with lower tidal ranges during neap tides (mean\u0026thinsp;~\u0026thinsp;1.2 m) and higher ranges during spring tides (mean\u0026thinsp;~\u0026thinsp;1.8 m) (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). The overall site-wide mean tidal range was 1.42 m (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Periods of highest daily mean wind speeds were recorded during winter months and early spring due to more frequent winter storms and cold fronts (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003ef). During the three-year measurement period, one tropical cyclone, Hurricane Ian, occurred in late September/early October 2022.\u003c/p\u003e \u003cp\u003eWhile chlorophyll \u003cem\u003ea\u003c/em\u003e data coverage was patchy, clear periods of elevated concentrations were consistently observed in January\u0026ndash;March of each year, indicative of the winter\u0026ndash;spring bloom (Fig. S1e in Supplemental Information). Turbidity data were even patchier; results from available periods suggested that higher turbidity levels corresponded to larger tidal ranges during spring tides. No seasonal patterns in turbidity were evident (Fig. S1f).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eMetabolic Rates\u003c/h2\u003e \u003cp\u003eMetabolic rates, calculated from the monitoring data, showed similar seasonal trends across all sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003eg\u0026ndash;i). GPP and ER were inversely related, with months with the highest gross production also having the highest respiration, and vice versa. Both GPP and ER increased in magnitude from colder to warmer months. NEM was most strongly heterotrophic during warmer months, shifting to slightly autotrophic values during the winter. Continuous monitoring also revealed short-term perturbations to metabolic rates (on the order of days), which were associated with periods of high wind speed (see \u0026ldquo;Impact of Storms\u0026rdquo;).\u003c/p\u003e \u003cp\u003eRates of GPP, ER, and NEM averaged across all years showed some variation among channel sites (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Both gross production and respiration areal rates were consistently higher at South compared to the other two sites, primarily due to its deeper water depth. Annual average GPP at South was ~\u0026thinsp;1.6 times higher than at Gull and ~\u0026thinsp;1.9 times higher than at North. South also exhibited the widest seasonal range in GPP, fluctuating between 61.4\u0026ndash;267.7 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Annual average ER at South was ~\u0026thinsp;1.3 times higher than at Gull and ~\u0026thinsp;1.7 times higher than at North. As with GPP, South exhibited the broadest seasonal range in ER, from \u0026minus;\u0026thinsp;373.4\u0026ndash; -54.9 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. All three sites were net heterotrophic on an annual basis. Despite the site differences in GPP and ER, annual average NEM was very similar at North and South, with respective means of -20.1\u0026thinsp;\u0026plusmn;\u0026thinsp;28.5 and \u0026minus;\u0026thinsp;26.1\u0026thinsp;\u0026plusmn;\u0026thinsp;34.1 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Gull was slightly more heterotrophic with a mean rate of -42.1\u0026thinsp;\u0026plusmn;\u0026thinsp;31.9 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, however, all three sites overlapped within their uncertainty ranges.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of mean and seasonal range of metabolic rates for the three channel sites over the entire time series (2021\u0026ndash;2024). Errors reflect propagated uncertainty calculated through the Monte Carlo analysis. Seasonal range and gap-filling methods are as described in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eGPP\u003c/p\u003e \u003cp\u003e(mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eER\u003c/p\u003e \u003cp\u003e(mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eNEM\u003c/p\u003e \u003cp\u003e(mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSeas. range\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNorth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e76.3\u0026thinsp;\u0026plusmn;\u0026thinsp;19.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23.6\u0026ndash;164.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-99.3\u0026thinsp;\u0026plusmn;\u0026thinsp;20.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-218.4\u0026ndash; -19.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-20.1\u0026thinsp;\u0026plusmn;\u0026thinsp;28.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-91.6\u0026ndash;32.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGull\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87.9\u0026thinsp;\u0026plusmn;\u0026thinsp;21.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22.3\u0026ndash;198.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-132.2\u0026thinsp;\u0026plusmn;\u0026thinsp;23.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-308.0\u0026ndash; -22.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-42.1\u0026thinsp;\u0026plusmn;\u0026thinsp;31.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-167.2\u0026ndash;29.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSouth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e144.2\u0026thinsp;\u0026plusmn;\u0026thinsp;22.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61.4\u0026ndash;267.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-172.1\u0026thinsp;\u0026plusmn;\u0026thinsp;25.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-373.4\u0026ndash; -54.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-26.1\u0026thinsp;\u0026plusmn;\u0026thinsp;34.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-108.3\u0026ndash;32.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTo examine the climatological means of channel-integrated metabolism over the annual cycle, metabolic rates were averaged across all three sites by day and month of year (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e4\u003c/span\u003e). From fall to early spring, GPP and ER were comparable in magnitude. However, from late spring to late summer, a period of heightened metabolic activity occurred during which respiration outstripped gross production. Summer production and respiration rates were up to nearly 4 and 6 times higher, respectively, than those observed in winter. Monthly mean GPP ranged from a low of 49.3 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in December to a high of 187.3 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in July. Similarly, monthly mean ER ranged from \u0026minus;\u0026thinsp;44.3 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in December to -269.9 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in July. This seasonal pattern is reflected in NEM, which remained near balance or slightly autotrophic from November to April, then became strictly heterotrophic starting in late spring, peaking in late summer. The system was most autotrophic in February (NEM of 12.0 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and most heterotrophic in September (-88.0 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eComparisons to Other Systems\u003c/h2\u003e \u003cp\u003eMarsh-dominated estuaries, characterized by surrounding tidal marshes, are strongly influenced by exchanges of materials within the marsh (Cai, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). High rates of production are often observed in marsh estuaries, which are close to land margins and receive nutrient runoff that enhances local production; moreover, in shallow waters, high light levels enhance benthic production (Caffrey \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). Marsh plants are a large source of allochthonous organic matter that is respired in adjacent waters, often creating highly heterotrophic conditions in the aquatic component of marsh ecosystems (Cai, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Despite the importance of coastal vegetated systems in driving coastal carbon fluxes, there are limited estimates of metabolic rates for marsh-dominated estuaries, particularly over the course of a full annual cycle (S. R. Wang et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Herrmann et al.'s (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) synthesis of the net ecosystem metabolism of estuaries along the U.S. East Coast found that these open estuarine waters are collectively net heterotrophic, with a best estimate of -3.2 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. For New Jersey inland bays specifically, NEM was estimated as -4.5 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCaffrey (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) provided annual average areal metabolic rates for a variety of shallow, well-mixed estuarine sites across U.S. coastal bioregions, showing that estuaries adjacent to marshes or mangroves were generally net heterotrophic. Notably, these smaller systems with tidal creeks and marshes often exhibited NEM rates that were several times higher than those in larger estuarine system, such as those included in Herrmann et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Of the 42 sites studied by Caffrey (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), half were predominantly marsh-dominated habitats, mainly located in the Mid-Atlantic, Southeast, and Pacific regions. For these marsh-adjacent sites, the annual average GPP was 305 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (with site-specific values ranging from 94 to 878 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and annual average ER was \u0026minus;\u0026thinsp;384 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (ranging from \u0026minus;\u0026thinsp;138 to -1,023 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). This resulted in an annual average NEM of -81 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (ranging from \u0026minus;\u0026thinsp;28 to -125 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). In comparison, Hagerthey et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) measured similar magnitudes across 64 sites in the Florida Everglades peatland, with a system-wide mean GPP of 103 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, ER of -220 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and NEM of -117 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Our site-wide annual averages\u0026mdash;GPP of 102.8 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, ER of -134.5 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and NEM of -29.4 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e\u0026mdash;fall on the lower end of Caffrey's (2004) ranges for marsh estuaries across the U.S., and indicated a much less heterotrophic system compared to the Everglades.\u003c/p\u003e \u003cp\u003eConversely, our metabolic rates were higher than those reported by Caffrey et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) for the salt-marsh dominated Grand Bay estuary in the Gulf of Mexico, where annual average GPP was 31 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and NEM was \u0026minus;\u0026thinsp;15.1 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Similarly, Raymond et al. (\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) found an even lower annual average GPP of 22.8 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for the York River Estuary in Virginia, which has large fringing marshes. Our gross production and respiration rates were approximately double those of the Duplin salt-marsh estuary in Georgia (S. R. Wang et al., \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), where annual average GPP was 49 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and ER was \u0026minus;\u0026thinsp;82 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e; however, our NEM was comparable to their value of -33 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eOn the other hand, our values are comparable to or lower than those found in marsh-surrounded coastal lagoons. For instance, for the Ria Formosa Lagoon in Portugal, which is surrounded by salt marshes, Cravo et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) found GPP ranged from 150\u0026ndash;300 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e between winter and spring/summer, and ER from \u0026minus;\u0026thinsp;150\u0026ndash; -350 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. This resulted in a near-balanced annual average NEM of -0.67 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. In contrast, Bas-Silvestre et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) reported much higher GPP and ER values for two confined coastal lagoons in La Pletera salt marsh in the Mediterranean, with annual average GPP ranging from 377\u0026ndash;531 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and ER ranging from \u0026minus;\u0026thinsp;401\u0026ndash; -491 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. The La Pletera marsh lagoons were among the most productive aquatic systems in the published literature, higher than even eutrophic estuaries.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eFactors Associated with Metabolism\u003c/h2\u003e \u003cp\u003eTo better understand which environmental factors are associated with marsh aquatic metabolism, we created simple and multiple linear regression models. These models reveal statistical associations between variables but do not imply causation. For instance, a variable correlated with a metabolic rate may be a response to, rather than a driver of, metabolic process\u0026mdash;or both may respond to an unmeasured third factor. Indirect relationships may also occur, where a variable correlates with metabolism only because it is linked to another factor that actually influences the metabolic process. Thus, while regression results can highlight potential drivers, confirming causality requires additional approaches, such as controlled experiments or mechanistic models.\u003c/p\u003e \u003cp\u003eOur objectives were to identify which environmental parameters were the most strongly associated with GPP, ER, and NEM; to develop predictive models for metabolism under seasonally changing environmental conditions; and to establish a baseline framework to enable future comparisons of functioning across different marsh sites. These linear regressions are a first step in linking environmental variability to marsh metabolism in the SMIIL study area, with more sophisticated approaches (e.g., generalized additive models, machine learning) needed to fully capture non-linearities and interactions.\u003c/p\u003e \u003cp\u003eWe began with simple linear regression models between daily rates of GPP, ER, and NEM and daily means of environmental parameters (salinity, temperature, DO % saturation, pH, tidal height, and wind speed), evaluated both annually and by season. These models quantified the variance in metabolic rates explained by each variable alone, helping to identify those with the strongest standalone relationships. These models therefore served as an initial exploration of individual associations and provided a basis for identifying key patterns.\u003c/p\u003e \u003cp\u003eWe then built multiple linear regression models using the same set of environmental variables to assess their combined influence on metabolic rates across the same timescales. By incorporating multiple predictors simultaneously, these models improved explanatory power and accounted for interactions among variables, enabling clearer identification of the most influential factors while controlling for others. To compare the relative importance of each predictor, we used standardized regression coefficients.\u003c/p\u003e \u003cp\u003eWe present both the simple and multiple regression results to show how relationships persist or change when accounting for other factors. This comparison can help distinguish whether a factor directly influences metabolism or primarily works through other variables, if a strong simple relationship disappears in multiple regression. Conversely, it can reveal previously obscured relationships, if a weak simple relationship becomes stronger in multiple regression after controlling for suppressor variables. This complementary approach is useful in complex ecosystems like salt marshes where numerous factors interact to influence metabolic processes.\u003c/p\u003e \u003cp\u003eIn the simple regressions, standardized estimated coefficients are equivalent to correlation coefficients (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Several key annual patterns emerged. Temperature showed the strongest associations with all metabolic rates: a strong positive relationship with GPP (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.605), and strong negative relationships with ER (\u003cem\u003er\u003c/em\u003e = -0.700) and NEM (\u003cem\u003er\u003c/em\u003e = -0.547). These results support the notion that temperature is a primary driver of metabolism that more strongly stimulates respiration than production. After temperature, DO % saturation and pH had the strongest associations with metabolic rates, and also showed consistent patterns: negative with GPP and positive with ER and NEM. These strong bivariate associations likely reflect feedback effects where metabolic activity alters water chemistry, rather than the other way around. Salinity and wind speed showed moderate relationships: salinity was positively associated with GPP (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.234) and negatively with ER (\u003cem\u003er\u003c/em\u003e = -0.257) and NEM (\u003cem\u003er\u003c/em\u003e = -0.188), while wind speed was negatively related to GPP (\u003cem\u003er\u003c/em\u003e = -0.217) and NEM (\u003cem\u003er\u003c/em\u003e = -0.155), and slightly positive for ER (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.045).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eEstimated coefficients from simple linear regression models relating metabolic rates (GPP, ER, NEM) to individual environmental variables (salinity, temperature, DO percent saturation, pH, tidal range, and wind speed) across all three channel sites, annually and by season. All variables are scaled to (\u003cem\u003ex\u003c/em\u003e \u0026ndash; \u003cem\u003ex̄\u003c/em\u003e)/\u003cem\u003es\u003c/em\u003e and are therefore unitless.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetabolic rate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDataset\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSal.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTemp.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDO %sat\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003epH\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTidal\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eWind\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGPP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAnnual\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.234**\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.605**\u003c/p\u003e \u003cp\u003e(0.017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.512**\u003c/p\u003e \u003cp\u003e(0.018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.505**\u003c/p\u003e \u003cp\u003e(0.019)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.015\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.217**\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWinter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.077**\u003c/p\u003e \u003cp\u003e(0.024)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.293**\u003c/p\u003e \u003cp\u003e(0.070)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.260**\u003c/p\u003e \u003cp\u003e(0.073)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.276**\u003c/p\u003e \u003cp\u003e(0.054)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.027\u003c/p\u003e \u003cp\u003e(0.019)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpring\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.031\u003c/p\u003e \u003cp\u003e(0.040)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.706**\u003c/p\u003e \u003cp\u003e(0.055)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.470**\u003c/p\u003e \u003cp\u003e(0.069)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.198**\u003c/p\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.005\u003c/p\u003e \u003cp\u003e(0.033)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.093**\u003c/p\u003e \u003cp\u003e(0.033)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSummer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.024\u003c/p\u003e \u003cp\u003e(0.043)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.058\u003c/p\u003e \u003cp\u003e(0.135)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.116*\u003c/p\u003e \u003cp\u003e(0.046)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.060\u003c/p\u003e \u003cp\u003e(0.061)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.108*\u003c/p\u003e \u003cp\u003e(0.047)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.202**\u003c/p\u003e \u003cp\u003e(0.058)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.172**\u003c/p\u003e \u003cp\u003e(0.034)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.421**\u003c/p\u003e \u003cp\u003e(0.029)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-0.279**\u003c/p\u003e \u003cp\u003e(0.026)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.231**\u003c/p\u003e \u003cp\u003e(0.028)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.063*\u003c/p\u003e \u003cp\u003e(0.026)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.107**\u003c/p\u003e \u003cp\u003e(0.024)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eER\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAnnual\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.257**\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.700**\u003c/p\u003e \u003cp\u003e(0.015)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.691**\u003c/p\u003e \u003cp\u003e(0.016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.659**\u003c/p\u003e \u003cp\u003e(0.016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.010\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.045*\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWinter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.035*\u003c/p\u003e \u003cp\u003e(0.016)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.191**\u003c/p\u003e \u003cp\u003e(0.047)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.170**\u003c/p\u003e \u003cp\u003e(0.049)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.098**\u003c/p\u003e \u003cp\u003e(0.037)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.022\u003c/p\u003e \u003cp\u003e(0.013)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.045**\u003c/p\u003e \u003cp\u003e(0.014)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpring\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.070*\u003c/p\u003e \u003cp\u003e(0.032)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.654**\u003c/p\u003e \u003cp\u003e(0.042)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.744**\u003c/p\u003e \u003cp\u003e(0.050)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.425**\u003c/p\u003e \u003cp\u003e(0.040)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.043\u003c/p\u003e \u003cp\u003e(0.027)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.021\u003c/p\u003e \u003cp\u003e(0.027)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSummer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.055\u003c/p\u003e \u003cp\u003e(0.041)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.381**\u003c/p\u003e \u003cp\u003e(0.129)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.268**\u003c/p\u003e \u003cp\u003e(0.043)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.162**\u003c/p\u003e \u003cp\u003e(0.058)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.095*\u003c/p\u003e \u003cp\u003e(0.045)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.417**\u003c/p\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003cp\u003e(0.050)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.616**\u003c/p\u003e \u003cp\u003e(0.042)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.576**\u003c/p\u003e \u003cp\u003e(0.032)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.505**\u003c/p\u003e \u003cp\u003e(0.037)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003cp\u003e(0.039)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.193**\u003c/p\u003e \u003cp\u003e(0.034)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNEM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAnnual\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.188**\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.547**\u003c/p\u003e \u003cp\u003e(0.018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.633**\u003c/p\u003e \u003cp\u003e(0.017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.585**\u003c/p\u003e \u003cp\u003e(0.017)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.155**\u003c/p\u003e \u003cp\u003e(0.021)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eWinter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.022\u003c/p\u003e \u003cp\u003e(0.022)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.639**\u003c/p\u003e \u003cp\u003e(0.059)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.569**\u003c/p\u003e \u003cp\u003e(0.063)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.128*\u003c/p\u003e \u003cp\u003e(0.051)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003cp\u003e(0.018)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.050**\u003c/p\u003e \u003cp\u003e(0.019)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpring\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.086**\u003c/p\u003e \u003cp\u003e(0.029)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.361**\u003c/p\u003e \u003cp\u003e(0.042)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.768**\u003c/p\u003e \u003cp\u003e(0.042)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.515**\u003c/p\u003e \u003cp\u003e(0.033)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.068**\u003c/p\u003e \u003cp\u003e(0.024)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.136**\u003c/p\u003e \u003cp\u003e(0.024)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSummer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.068\u003c/p\u003e \u003cp\u003e(0.048)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.589**\u003c/p\u003e \u003cp\u003e(0.151)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.335**\u003c/p\u003e \u003cp\u003e(0.051)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.342**\u003c/p\u003e \u003cp\u003e(0.068)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.047\u003c/p\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.929**\u003c/p\u003e \u003cp\u003e(0.052)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.240**\u003c/p\u003e \u003cp\u003e(0.069)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.601**\u003c/p\u003e \u003cp\u003e(0.063)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.686**\u003c/p\u003e \u003cp\u003e(0.047)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.616**\u003c/p\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0.056\u003c/p\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0.445**\u003c/p\u003e \u003cp\u003e(0.044)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"8\"\u003eStandard errors are reported in parentheses.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"8\"\u003e*, ** indicates significance at the 95% and 99% level, respectively.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eSeasonal trends were also apparent. Temperature had its strongest effect on GPP in spring (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.706) but became insignificant in summer, suggesting a possible threshold or saturation effect. It also showed a strong positive relationship with NEM in winter (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.639) but was negative in other seasons. Wind speed had a strong negative relationship with NEM in summer (\u003cem\u003er\u003c/em\u003e = -0.929) and was consistently negative with GPP across seasons, possibly reflecting effects of turbidity or gas exchange. Tidal range showed minimal influence overall, generally not significant for most seasons and metabolic rates. Overall, the simple regressions indicate that the association between environmental factors and metabolic rates varies seasonally, as evidenced by the significant shift in coefficient values and even sign changes across seasons. Moreover, even the effect of the primary driver, temperature, is not constant throughout the year, suggesting acclimation or interaction with other seasonal factors.\u003c/p\u003e \u003cp\u003eBefore fitting multiple regressions, we assessed collinearity between explanatory factors using pairwise correlation coefficients and VIFs. Strong correlations (|\u003cem\u003er|\u003c/em\u003e \u0026gt; 0.7) were found between salinity and temperature, pH and temperature, and DO % saturation and pH (Table S2 in Supplemental Information). However, all VIFs were below 5, with temperature (4.81), pH (4.65), and DO % saturation (3.61) indicating moderate collinearity, while salinity, tidal range, and wind speed had VIFs near 1. Therefore, we retained all six predictors in the multiple regression models.\u003c/p\u003e \u003cp\u003eAll fifteen multiple regression models were highly significant and explained 38\u0026ndash;73% of the variation in metabolic rates, with generally higher adjusted \u003cem\u003eR\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e values for GPP and ER than for NEM (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Annual models explained 73% and 70% of the variation in GPP and ER, respectively, and 53% of the variation in NEM. Summer models for GPP and ER explained less variation than other seasons, while the winter model for NEM was weakest.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u0026nbsp;\u003ctable id=\"Tab6\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eEstimated coefficients from multiple linear regression models relating metabolic rates (GPP, ER, and NEM) to environmental variables across all three channel sites, annually and by season. Explanatory variables are shown; grouping variables for the season-specific models are site, and for the annual models are site, season, and their interactions. All variables are standardized and unitless.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMetabolic rate\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDataset\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSal.\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTemp.\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDO %sat\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003epH\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eTidal\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eWind\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003csub\u003eadj\u003c/sub\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGPP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAnnual\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.056**\u003c/p\u003e\n \u003cp\u003e(0.014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.642**\u003c/p\u003e\n \u003cp\u003e(0.031)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.028\u003c/p\u003e\n \u003cp\u003e(0.024)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.090**\u003c/p\u003e\n \u003cp\u003e(0.028)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.002\u003c/p\u003e\n \u003cp\u003e(0.011)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.021\u003c/p\u003e\n \u003cp\u003e(0.011)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.728**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWinter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.122**\u003c/p\u003e\n \u003cp\u003e(0.014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.234**\u003c/p\u003e\n \u003cp\u003e(0.042)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.410**\u003c/p\u003e\n \u003cp\u003e(0.056)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.030\u003c/p\u003e\n \u003cp\u003e(0.043)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003cp\u003e(0.012)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.069**\u003c/p\u003e\n \u003cp\u003e(0.013)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.683**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpring\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.071**\u003c/p\u003e\n \u003cp\u003e(0.022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.008**\u003c/p\u003e\n \u003cp\u003e(0.052)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.006\u003c/p\u003e\n \u003cp\u003e(0.063)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.095\u003c/p\u003e\n \u003cp\u003e(0.054)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.029\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.055**\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.732**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSummer\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.067\u003c/p\u003e\n \u003cp\u003e(0.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.435**\u003c/p\u003e\n \u003cp\u003e(0.138)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.028\u003c/p\u003e\n \u003cp\u003e(0.049)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.059\u003c/p\u003e\n \u003cp\u003e(0.063)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.082*\u003c/p\u003e\n \u003cp\u003e(0.039)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.192**\u003c/p\u003e\n \u003cp\u003e(0.046)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.401**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.104**\u003c/p\u003e\n \u003cp\u003e(0.028)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.345**\u003c/p\u003e\n \u003cp\u003e(0.052)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.179**\u003c/p\u003e\n \u003cp\u003e(0.035)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.138**\u003c/p\u003e\n \u003cp\u003e(0.045)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.037\u003c/p\u003e\n \u003cp\u003e(0.020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.052**\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.493**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAnnual\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.081**\u003c/p\u003e\n \u003cp\u003e(0.015)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.493**\u003c/p\u003e\n \u003cp\u003e(0.033)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.266**\u003c/p\u003e\n \u003cp\u003e(0.025)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003cp\u003e(0.029)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.035**\u003c/p\u003e\n \u003cp\u003e(0.012)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.188**\u003c/p\u003e\n \u003cp\u003e(0.012)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.700**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWinter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.035**\u003c/p\u003e\n \u003cp\u003e(0.012)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.240**\u003c/p\u003e\n \u003cp\u003e(0.034)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.182**\u003c/p\u003e\n \u003cp\u003e(0.046)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.044\u003c/p\u003e\n \u003cp\u003e(0.036)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.039**\u003c/p\u003e\n \u003cp\u003e(0.010)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.029**\u003c/p\u003e\n \u003cp\u003e(0.010)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.527**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpring\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.076**\u003c/p\u003e\n \u003cp\u003e(0.022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.622**\u003c/p\u003e\n \u003cp\u003e(0.051)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.343**\u003c/p\u003e\n \u003cp\u003e(0.061)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.042\u003c/p\u003e\n \u003cp\u003e(0.053\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.034*\u003c/p\u003e\n \u003cp\u003e(0.017)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.140**\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.604**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSummer\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.093*\u003c/p\u003e\n \u003cp\u003e(0.037)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.477**\u003c/p\u003e\n \u003cp\u003e(0.135)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.221**\u003c/p\u003e\n \u003cp\u003e(0.048)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.016\u003c/p\u003e\n \u003cp\u003e(0.061)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.057\u003c/p\u003e\n \u003cp\u003e(0.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.471**\u003c/p\u003e\n \u003cp\u003e(0.046)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.384**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003cp\u003e(0.038)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.330**\u003c/p\u003e\n \u003cp\u003e(0.069)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.454**\u003c/p\u003e\n \u003cp\u003e(0.047)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.036\u003c/p\u003e\n \u003cp\u003e(0.059)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003cp\u003e(0.026)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.293**\u003c/p\u003e\n \u003cp\u003e(0.024)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.572**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNEM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAnnual\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.078**\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.154**\u003c/p\u003e\n \u003cp\u003e(0.041)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.425**\u003c/p\u003e\n \u003cp\u003e(0.031)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.109**\u003c/p\u003e\n \u003cp\u003e(0.037)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.058**\u003c/p\u003e\n \u003cp\u003e(0.015)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.344**\u003c/p\u003e\n \u003cp\u003e(0.015)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.533**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWinter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.072**\u003c/p\u003e\n \u003cp\u003e(0.018)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.662**\u003c/p\u003e\n \u003cp\u003e(0.054)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.750**\u003c/p\u003e\n \u003cp\u003e(0.072)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.107\u003c/p\u003e\n \u003cp\u003e(0.055)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.071**\u003c/p\u003e\n \u003cp\u003e(0.015)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.024\u003c/p\u003e\n \u003cp\u003e(0.016)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.400**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSpring\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.054*\u003c/p\u003e\n \u003cp\u003e(0.024)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.016\u003c/p\u003e\n \u003cp\u003e(0.054)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.593**\u003c/p\u003e\n \u003cp\u003e(0.066)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.174**\u003c/p\u003e\n \u003cp\u003e(0.058)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.089**\u003c/p\u003e\n \u003cp\u003e(0.019)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.180**\u003c/p\u003e\n \u003cp\u003e(0.020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.427**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSummer\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.087*\u003c/p\u003e\n \u003cp\u003e(0.036)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.348**\u003c/p\u003e\n \u003cp\u003e(0.132)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.347**\u003c/p\u003e\n \u003cp\u003e(0.047)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.035\u003c/p\u003e\n \u003cp\u003e(0.060)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.010\u003c/p\u003e\n \u003cp\u003e(0.037)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-1.010**\u003c/p\u003e\n \u003cp\u003e(0.044)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.575**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.111*\u003c/p\u003e\n \u003cp\u003e(0.052)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.195*\u003c/p\u003e\n \u003cp\u003e(0.096)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.583**\u003c/p\u003e\n \u003cp\u003e(0.065)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003cp\u003e(0.083)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.038\u003c/p\u003e\n \u003cp\u003e(0.037)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.557**\u003c/p\u003e\n \u003cp\u003e(0.034)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.556**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"9\"\u003eStandard errors are reported in parentheses.\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"9\"\u003e*, ** indicates significance at the 95% and 99% level, respectively.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003c/p\u003e\n\u003cp\u003eWhile the simple regression results provide a foundation for understanding primary relationships, the multiple regression results reveal which factors remain significant when controlling for others, potentially showing which correlations are direct versus indirect. Temperature remained the most influential factor overall: it was still strongly positive for GPP across seasons, and maintained a negative relationship with ER in most seasons\u0026mdash;except in winter, as in the simple regression. The scaled influence of temperature on GPP was greater than that of any other predictor in nearly all cases, sometimes by an order of magnitude. For instance, in the annual GPP model, temperature\u0026rsquo;s scaled effect (\u003cem\u003e\u0026beta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.642) was seven times larger than that of pH (\u003cem\u003e\u0026beta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.090) and nearly 11.5 times that of salinity (\u003cem\u003e\u0026beta;\u003c/em\u003e = -0.056). In contrast, DO % saturation and pH, which were strongly correlated with GPP in simple regression, were often non-significant or reverse in multiple regressions, supporting the interpretation that they reflect feedbacks from metabolism rather than direct drivers. Salinity and wind speed became more important in the multiple regressions. Both were significant across most seasons for all metabolic rates, with wind speed showing a particularly strong association with NEM in summer (\u003cem\u003e\u0026beta;\u003c/em\u003e = -1.010).\u003c/p\u003e\n\u003cp\u003eMultiple regressions also highlighted seasonal variability (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e and Fig. S3 in Supplemental Information). Spring GPP had the highest association with temperature (\u003cem\u003e\u0026beta;\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.008), while all other coefficients were two orders of magnitude lower. Winter metabolism showed unique patterns, as it was the only season where temperature was positively correlated with ER and NEM. Similarly, salinity had a distinct effect in fall, where it was positively associated with GPP (contrary to other seasons) but not significant for ER.\u003c/p\u003e\n\u003cp\u003eAltogether, these statistical patterns offer several ecological insights into the processes associated with marsh metabolism. In particular, temperature emerged as a dominant and consistent driver, although its influence varied by season and metabolic process. Warmer conditions in spring and summer promote marsh plant growth and labile organic matter availability, which in turn fuels microbial respiration; higher temperatures also directly elevate respiration rates. This dual stimulation of GPP and ER is well-supported in the literature. For example, Caffrey et al. (\u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e) found that the effect of temperature is stronger for respiration than primary production, resulting in more heterotrophic conditions in summer. As a result, seasonally increasing temperatures generally lead to more negative NEM rates, consistent with our findings.\u003c/p\u003e\n\u003cp\u003eSalinity also played a meaningful but more variable role. Generally, lower salinity was associated with higher GPP and ER, suggesting that freshwater inputs may support increased metabolic activity. This pattern likely reflects nutrient delivery, as salinity acts as a conservative tracer for freshwater-driven nutrient inputs to the system, which is a key factor controlling primary production. For example, in the Grand Bay estuary, Caffrey et al. (\u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e) observed that periods of low salinity were followed by increases in primary production. Similarly, we infer that freshwater inputs stimulate productivity in the SMIIL system, which in turn leads to increased respiration from decomposition of newly produced organic matter.\u003c/p\u003e\n\u003cp\u003ePatterns in pH and DO % saturation suggest that water chemistry is strongly influenced by biological processes, rather than driving them. Although both variables were consistently correlated with metabolic rates in simple regressions, their significance often disappeared in multiple regressions, suggesting they reflect feedbacks from production and respiration. This finding aligns with Baumann \u0026amp; Smith (\u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e), who found that pH patterns were closely linked to dissolved oxygen across a diversity of shallow estuarine environments across the U.S, and that these concurrent fluctuations were driven by local metabolic processes. Similarly, Lowe et al. (\u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e) found that local ecosystem metabolism is the dominant driver of pH variability in a range of habitats in the Northeast Pacific coastal ocean. Both of these studies also provided strong empirical evidence that physical factors indirectly influence carbonate chemistry through affecting primary production and respiration, with the implication that increased ecosystem respiration from warmer waters exacerbates coastal acidification. Our findings build on these insights by showing seasonal peaks in the pH\u0026ndash;GPP and pH\u0026ndash;NEM relationships in fall and spring, respectively, suggesting that metabolic control of carbonate chemistry may be strongest during transitional periods.\u003c/p\u003e\n\u003cp\u003eFinally, we explored the potential influence of physical processes such as tidal and wind mixing on metabolism. Although our turbidity and chlorophyll \u003cem\u003ea\u003c/em\u003e datasets were compromised by biofouling, limited available data showed a correlation between turbidity and tidal range, consistent with findings from Gomez-Castillo et al. (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). They observed that blooms typically developed during neap tides when mixing was reduced and turbidity was lower, shifting estuarine metabolism to net autotrophic conditions. These blooms subsequently dissipated during the following spring tide, when stronger mixing led to increased turbidity and a return to heterotrophy. In our system, however, tidal range was not a significant predictor of metabolism, suggesting that mixing-related processes may be less influential here than in more phytoplankton-dominated estuaries. In contrast, wind speed emerged as a more important factor in multiple regressions, particularly for NEM, where it showed strong negative effects in summer and fall. This suggests that wind-driven mixing may enhance gas exchange or turbidity, thereby influencing net carbon balances.\u003c/p\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eImpact of Storms\u003c/h2\u003e\n \u003cp\u003eFurther examination of the wind speed and metabolic rate time series suggested that temporary fluctuations in NEM were linked to short periods of elevated daily wind speed. For example, Tropical Storm Ian in early October 2022 was accompanied by a large dip in NEM at all three sites (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e). To investigate the potential effect of short-term storm events on net ecosystem metabolism, we identified high wind-speed events in the measurement record and compared mean NEM values during the storm event to baseline periods of 5 days before and after the events.\u003c/p\u003e\n \u003cp\u003eFrom December 2021 to June 2024, eight events were recorded during which the daily wind speed exceeded the 99th percentile (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e). Due to gaps in the NEM record for each site, this corresponded to six events each for North and South, and seven for Gull. The site-wide mean NEM rate during events was \u0026minus;\u0026thinsp;68.69 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, while the means for the pre- and post-event periods were \u0026minus;\u0026thinsp;8.60 and \u0026minus;\u0026thinsp;12.9 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, respectively (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e). NEM was significantly lower during the events compared to pre-event conditions (paired t\u003cem\u003e-\u003c/em\u003etest, p\u0026thinsp;=\u0026thinsp;0.0090) and also compared to post-event conditions (p\u0026thinsp;=\u0026thinsp;0.013). There was no significant difference between pre-event and post-event NEM values (p\u0026thinsp;=\u0026thinsp;0.69).\u003c/p\u003e\n \u003cp\u003eElevated wind speed was consistently associated with higher respiration rates and more heterotrophic conditions, in line with findings by Gomez-Castillo et al. (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Wind speed affects currents and mixing in estuaries, which in turn influences dissolved oxygen levels, nutrient and organic matter transport, and turbidity and light attenuation (Kemp \u0026amp; Testa, \u003cspan class=\"CitationRef\"\u003e2012\u003c/span\u003e). Wind speed is also a key indicator of storm conditions. On weekly to monthly time scales, ephemeral events like storms may influence metabolic balance more than seasonal changes in environmental conditions (Russell \u0026amp; Montagna, \u003cspan class=\"CitationRef\"\u003e2007\u003c/span\u003e). While we focused on increases in wind speed, other indicators of storms, such as heavy precipitation and drops in barometric pressure could also serve to identify storms events. Our results suggest that high wind speed events are linked to a temporary increase in heterotrophy, which quickly returns to baseline levels after the storm passes. Various alternative mechanisms\u0026mdash;including storm surge, heavy precipitation, and increased discharge\u0026mdash;can also influence estuarine metabolism (Buelo et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). For example, higher winds may reduce the intensity of upwelling-induced production in certain systems, while elevated precipitation can increase gross production through enhanced nutrient delivery and reduced estuarine residence time. The impacts of storm events are thus complex and likely system-dependent, warranting further investigation in future studies.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eFuture changes in estuarine systems will depend on how anthropogenic impacts and climate change alter the balance between gross primary production and ecosystem respiration. While predicting these shifts remains challenging due to observational gaps, continuous monitoring of key parameters like dissolved oxygen, temperature, and salinity is becoming essential for capturing the dynamics of aquatic metabolism in diverse coastal environments.\u003c/p\u003e \u003cp\u003eIn this study, we quantified aquatic ecosystem metabolism in the main tidal channel of the Seven Mile Island Innovation Laboratory and found that all three monitoring sites were net heterotrophic annually, with clear seasonal patterns: both GPP and ER increased substantially from colder to warmer months, leading to peak heterotrophy in late summer/fall. This pattern supports the concept of the seasonal marsh CO\u003csub\u003e2\u003c/sub\u003e pump (Z. A. Wang et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Z. A. Wang \u0026amp; Cai, \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), wherein marsh-derived organic matter is respired in adjacent waters, causing the aquatic environments of salt marsh systems to often function as net CO\u003csub\u003e2\u003c/sub\u003e sources. These results establish a robust baseline characterization of metabolic dynamics in the SMIIL salt marsh system.\u003c/p\u003e \u003cp\u003eOur results suggest that warming temperatures, which are projected to continue to increase across the mid-Atlantic (EPA Region 3 Climate Collaborative, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), could further drive estuarine systems in the region toward greater heterotrophy, increasing CO\u003csub\u003e2\u003c/sub\u003e emissions and potentially intensifying coastal acidification. Episodic storm events, expected to become more frequent and severe, may periodically amplify these metabolic responses.\u003c/p\u003e \u003cp\u003eThe SMIIL research area, designed to evaluate the use of dredged sediments to enhance marsh accretion and protection, also provides a unique opportunity to study the impacts of sea level rise on coastal ecosystems. In New Jersey, relative sea level rise is occurring at more than twice the global average rate (New Jersey Climate Change Resource Center, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), which will likely reshape coastal ecosystems with significant implications for blue carbon sequestration. While rising sea levels may enhance soil carbon accumulation in marshes, they could also trigger ecosystem transitions that alter the balance between carbon sequestration and greenhouse gas emissions (Kirwan et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Our three-year dataset captures natural variability in marsh channel environmental conditions and metabolism and offers a valuable reference point for evaluating future environmental changes and management interventions.\u003c/p\u003e \u003cp\u003eAlthough this study focused on aquatic metabolism of the salt marsh ecosystem, integrating these data with carbon fluxes from other marsh features\u0026mdash;including vegetated platforms, sediments, and salt ponds\u0026mdash;would provide a more complete understanding of system-wide carbon dynamics. Such integrated approaches are essential for refining estimates of salt marsh carbon sink capacity. As coastal marshes face mounting pressures from climate change and human activities, a holistic understanding of their carbon processing is increasingly critical to inform effective conservation and restoration strategies.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eEthics approval and consent to participate\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\n\u003ch2\u003eConsent for publication\u003c/h2\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\n\u003ch2\u003eAvailability of data and materials\u003c/h2\u003e\n\u003cp\u003eDatasets generated and analyzed during the current study are available in the Biological and Chemical Oceanography Data Management Office (BCO-DMO) repository. Continuous monitoring and meteorological datasets, as well as discrete sample datasets for DO, DIC, and TA, can be accessed at https://www.bco-dmo.org/project/962228.\u003c/p\u003e\n\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThis work was funded by the U.S. Army Engineer Research and Development Center (W912HZ2020061\u0026ndash;RA3).\u003c/p\u003e\n\n\u003ch2\u003eAuthors\u0026apos; contributions\u003c/h2\u003e\n\u003cp\u003eEJC performed data curation, analysis, and interpretation, and wrote the manuscript. JS led the field data collection and discrete sample laboratory analysis. KEF contributed to field data collection, assisted with data curation, and aided manuscript development. HIP secured funding, supervised the project, and supported manuscript development. All authors contributed to the conception and design of the study, and reviewed and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eWe thank David Perkey (Engineer Research and Development Center), Kelsey Fall (University of Delaware), Lenore Tedesco and Julie Blum (The Wetlands Institute), and Roland Hagen (Rutgers University Marine Field Station) for their assistance with fieldwork. Field and laboratory support was also provided by Jose Cuevas, Yasmin Hamilton, Elle Joubert, Yoleimy Lopez Barias, and Stevie Walker (Boston College). We thank Kate Willis for ArcGIS support, and Swapnil Sharma and Melissa McTernan for guidance on statistical analyses.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAanderaa Data Instruments AS. (2017). \u003cem\u003eTD 269 Operating Manual\u003c/em\u003e. https://www.aanderaa.com/media/pdfs/oxygen-optode-4330-4835-and-4831.pdf\u003c/li\u003e\n\u003cli\u003eAlbert, D. R. (2020). Monte Carlo Uncertainty Propagation with the NIST Uncertainty Machine. \u003cem\u003eJournal of Chemical Education\u003c/em\u003e, \u003cem\u003e97\u003c/em\u003e(5), 1491\u0026ndash;1494. https://doi.org/10.1021/acs.jchemed.0c00096\u003c/li\u003e\n\u003cli\u003eAlongi, D. M. (2020). Carbon balance in salt marsh and mangrove ecosystems: A global synthesis. \u003cem\u003eJournal of Marine Science and Engineering\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(10), 1\u0026ndash;21. https://doi.org/10.3390/jmse8100767\u003c/li\u003e\n\u003cli\u003eBarbier, E. B., Hacker, S. D., Kennedy, C., Koch, E. W., Stier, A. C., \u0026amp; Silliman, B. R. (2011). The value of estuarine and coastal ecosystem services. \u003cem\u003eEcological Monographs\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e(2), 169\u0026ndash;193. https://doi.org/10.1890/10-1510.1\u003c/li\u003e\n\u003cli\u003eBas-Silvestre, M., Quintana, X. D., Compte, J., Gasc\u0026oacute;n, S., Boix, D., Ant\u0026oacute;n-Pardo, M., \u0026amp; Obrador, B. (2020). Ecosystem metabolism dynamics and environmental drivers in Mediterranean confined coastal lagoons. \u003cem\u003eEstuarine, Coastal and Shelf Science\u003c/em\u003e, \u003cem\u003e245\u003c/em\u003e, 106989. https://doi.org/10.1016/j.ecss.2020.106989\u003c/li\u003e\n\u003cli\u003eBauer, J. E., Cai, W.-J., Raymond, P. A., Bianchi, T. S., Hopkinson, C. S., \u0026amp; Regnier, P. A. G. (2013). \u003cem\u003eThe changing carbon cycle of the coastal ocean\u003c/em\u003e. https://doi.org/10.1038/nature12857\u003c/li\u003e\n\u003cli\u003eBaumann, H., \u0026amp; Smith, E. M. (2018). Quantifying Metabolically Driven pH and Oxygen Fluctuations in US Nearshore Habitats at Diel to Interannual Time Scales. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(4), 1102\u0026ndash;1117. https://doi.org/10.1007/s12237-017-0321-3\u003c/li\u003e\n\u003cli\u003eBeck, M. W., Hagy, J. D., \u0026amp; Murrell, M. C. (2015). Improving estimates of ecosystem metabolism by reducing effects of tidal advection on dissolved oxygen time series. \u003cem\u003eLimnology and Oceanography: Methods\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(12), 731\u0026ndash;745. https://doi.org/10.1002/lom3.10062\u003c/li\u003e\n\u003cli\u003eBoscolo-Galazzo, F., Crichton, K. A., Barker, S., \u0026amp; Pearson, P. N. (2018). Temperature dependency of metabolic rates in the upper ocean: A positive feedback to global climate change? \u003cem\u003eGlobal and Planetary Change\u003c/em\u003e, \u003cem\u003e170\u003c/em\u003e, 201\u0026ndash;212. https://doi.org/10.1016/j.gloplacha.2018.08.017\u003c/li\u003e\n\u003cli\u003eBuelo, C. D., Besterman, A. F., Walter, J. A., Pace, M. L., Ha, D. T., \u0026amp; Tassone, S. J. (2024). Quantifying Disturbance and Recovery in Estuaries: Tropical Cyclones and High-Frequency Measures of Oxygen and Salinity. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(1), 18\u0026ndash;31. https://doi.org/10.1007/s12237-023-01255-1\u003c/li\u003e\n\u003cli\u003eCaffrey, J. M. (2004). Factors controlling net ecosystem metabolism in U.S. estuaries. \u003cem\u003eEstuaries\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(1), 90\u0026ndash;101. https://doi.org/10.1007/BF02803563\u003c/li\u003e\n\u003cli\u003eCaffrey, J. M., Murrell, M. C., Amacker, K. S., Harper, J. W., Phipps, S., \u0026amp; Woodrey, M. S. (2014). Seasonal and Inter-annual Patterns in Primary Production, Respiration, and Net Ecosystem Metabolism in Three Estuaries in the Northeast Gulf of Mexico. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e(S1), 222\u0026ndash;241. https://doi.org/10.1007/s12237-013-9701-5\u003c/li\u003e\n\u003cli\u003eCai, W. J. (2011). Estuarine and coastal ocean carbon paradox: CO2 sinks or sites of terrestrial carbon incineration? \u003cem\u003eAnnual Review of Marine Science\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e, 123\u0026ndash;145. https://doi.org/10.1146/annurev-marine-120709-142723\u003c/li\u003e\n\u003cli\u003eChasten, M., Tedesco, L., \u0026amp; Kopkash, G. (2023). Advancing sediment solutions in the Seven Mile Island Innovation Laboratory. \u003cem\u003eCoastal Engineering Proceedings\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e, 64. https://doi.org/10.9753/icce.v37.papers.64\u003c/li\u003e\n\u003cli\u003eCole, J. J., \u0026amp; Caraco, N. F. (1998). Atmospheric exchange of carbon dioxide in a low-wind oligotrophic lake measured by the addition of SF6. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e43\u003c/em\u003e(4), 647\u0026ndash;656. https://doi.org/10.4319/lo.1998.43.4.0647\u003c/li\u003e\n\u003cli\u003eCravo, A., Rosa, A., Jacob, J., \u0026amp; Correia, C. (2020). Dissolved oxygen dynamics in Ria Formosa Lagoon (South Portugal)\u0026mdash;A real time monitoring station observatory. \u003cem\u003eMarine Chemistry\u003c/em\u003e, \u003cem\u003e223\u003c/em\u003e, 103806. https://doi.org/10.1016/j.marchem.2020.103806\u003c/li\u003e\n\u003cli\u003eDai, M., Su, J., Zhao, Y., Hofmann, E. E., Cao, Z., Cai, W. J., Gan, J., Lacroix, F., Laruelle, G. G., Meng, F., Mu die ller, J. D., Regnier, P. A. G., Wang, G., \u0026amp; Wang, Z. (2022). Carbon Fluxes in the Coastal Ocean: Synthesis, Boundary Processes, and Future Trends. \u003cem\u003eAnnual Review of Earth and Planetary Sciences\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e, 593\u0026ndash;626. https://doi.org/10.1146/annurev-earth-032320-090746\u003c/li\u003e\n\u003cli\u003eDai, M., Zhao, Y., Chai, F., Chen, M., Chen, N., Chen, Y., Cheng, D., Gan, J., Guan, D., Hong, Y., Huang, J., Lee, Y., Leung, K. M. Y., Lim, P. E., Lin, S., Lin, X., Liu, X., Liu, Z., Luo, Y.-W., \u0026hellip; Zhang, Z. (2023). Persistent eutrophication and hypoxia in the coastal ocean. \u003cem\u003eCambridge Prisms: Coastal Futures\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e, e19. https://doi.org/10.1017/cft.2023.7\u003c/li\u003e\n\u003cli\u003eDickson, A. G., Sabine, C. L., Christian, J. R., Bargeron, C. P., \u0026amp; North Pacific Marine Science Organization (Eds.). (2007). \u003cem\u003eGuide to best practices for ocean CO2 measurements\u003c/em\u003e. North Pacific Marine Science Organization.\u003c/li\u003e\n\u003cli\u003eDormann, C. F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carr\u0026eacute;, G., Marqu\u0026eacute;z, J. R. G., Gruber, B., Lafourcade, B., Leit\u0026atilde;o, P. J., M\u0026uuml;nkem\u0026uuml;ller, T., McClean, C., Osborne, P. E., Reineking, B., Schr\u0026ouml;der, B., Skidmore, A. K., Zurell, D., \u0026amp; Lautenbach, S. (2013). Collinearity: A review of methods to deal with it and a simulation study evaluating their performance. \u003cem\u003eEcography\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(1), 27\u0026ndash;46. https://doi.org/10.1111/j.1600-0587.2012.07348.x\u003c/li\u003e\n\u003cli\u003eEmerson, S., Yang, B., White, M., \u0026amp; Cronin, M. (2019). Air-Sea Gas Transfer: Determining Bubble Fluxes With In Situ N2 Observations. \u003cem\u003eJournal of Geophysical Research: Oceans\u003c/em\u003e, \u003cem\u003e124\u003c/em\u003e(4), 2716\u0026ndash;2727. https://doi.org/10.1029/2018JC014786\u003c/li\u003e\n\u003cli\u003eEPA Region 3 Climate Collaborative. (2022). \u003cem\u003eEPA Region 3 Climate Adaptation Implementation Plan\u003c/em\u003e (Mid-Atlantic). US EPA. https://www.epa.gov/climate-adaptation/climate-change-epa-mid-atlantic-region\u003c/li\u003e\n\u003cli\u003eFall, K. A., Perkey, D. W., Tyler, Z. J., \u0026amp; Welp, T. L. (2021). \u003cem\u003eField measurement and monitoring of hydrodynamic and suspended sediment within the Seven Mile Island Innovation Laboratory, New Jersey\u003c/em\u003e. https://erdclibrary.on.worldcat.org/discovery.\u003c/li\u003e\n\u003cli\u003eGarcia, H. E., \u0026amp; Gordon, L. I. (1992). Oxygen solubility in seawater: Better fitting equations. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e(6), 1307\u0026ndash;1312. https://doi.org/10.4319/lo.1992.37.6.1307\u003c/li\u003e\n\u003cli\u003eGedan, K., Silliman, B., \u0026amp; Bertness, M. (2009). Centuries of Human-Driven Change in Salt Marsh Ecosystems. \u003cem\u003eAnnual Review of Marine Science\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e, 117\u0026ndash;141. https://doi.org/10.1146/annurev.marine.010908.163930\u003c/li\u003e\n\u003cli\u003eGomez-Castillo, A. P., Panton, A., \u0026amp; Purdie, D. A. (2023). Temporal variability of phytoplankton biomass and net community production in a macrotidal temperate estuary. \u003cem\u003eEstuarine, Coastal and Shelf Science\u003c/em\u003e, \u003cem\u003e280\u003c/em\u003e(August 2022), 108182. https://doi.org/10.1016/j.ecss.2022.108182\u003c/li\u003e\n\u003cli\u003eHagerthey, S. E., Cole, J. J., \u0026amp; Kilbane, D. (2010). Aquatic metabolism in the Everglades: Dominance of water column heterotrophy. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e(2), 653\u0026ndash;666. https://doi.org/10.4319/lo.2010.55.2.0653\u003c/li\u003e\n\u003cli\u003eHerrmann, M., Najjar, R. G., Kemp, W. M., Alexander, R. B., Boyer, E. W., Cai, W., Griffith, P. C., Kroeger, K. D., Mccallister, S. L., \u0026amp; Smith, R. A. (2015). Net ecosystem production and organic carbon balance of U.S. East Coast estuaries: A synthesis approach. \u003cem\u003eGlobal Biogeochemical Cycles\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e, 96\u0026ndash;111. https://doi.org/10.1002/2013GB004736.Received\u003c/li\u003e\n\u003cli\u003eHersbach, H., Berrisford, P., Biavati, G., Hor\u0026aacute;nyi, A., Mu\u0026ntilde;oz Sabater, J., Nicolas, J., Peubey, C., Radu, R., Radu, I., Schepers, D., Simmons, A., Soci, D., Dee, D., \u0026amp; Th\u0026eacute;paut, J.-N. (2023). \u003cem\u003eERA5 hourly data on single levels from 1940 to present\u003c/em\u003e [Dataset]. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.adbb2d47\u003c/li\u003e\n\u003cli\u003eHoward, E. M., Forbrich, I., Giblin, A. E., Lott, D. E., Cahill, K. L., \u0026amp; Stanley, R. H. R. (2018). Using Noble Gases to Compare Parameterizations of Air-Water Gas Exchange and to Constrain Oxygen Losses by Ebullition in a Shallow Aquatic Environment. \u003cem\u003eJournal of Geophysical Research: Biogeosciences\u003c/em\u003e, \u003cem\u003e123\u003c/em\u003e(9), 2711\u0026ndash;2726. https://doi.org/10.1029/2018JG004441\u003c/li\u003e\n\u003cli\u003eHowarth, R. W., Swaney, D. P., Butler, T. J., \u0026amp; Marino, R. (2000). Climatic Control on Eutrophication of the Hudson River Estuary. \u003cem\u003eEcosystems\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(2), 210\u0026ndash;215.\u003c/li\u003e\n\u003cli\u003eIOOS. (2018). \u003cem\u003eManual for Real-Time Quality Control of Dissolved Oxygen Observation: A Guide to Quality Control and Quality Assurance for Dissiolved Oxygen Observations in Coastal Oceans\u003c/em\u003e (No. 2.1). https://doi.org/10.25923/q0m1-d488\u003c/li\u003e\n\u003cli\u003eIPCC. (2023). \u003cem\u003eClimate Change 2023: Synthesis Report\u003c/em\u003e (Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, H. Lee and J. Romero (Eds.)]). IPCC. 10.59327/IPCC/AR6-9789291691647\u003c/li\u003e\n\u003cli\u003eJankowski, K. J., Mejia, F. H., Blaszczak, J. R., \u0026amp; Holtgrieve, G. W. (2021). Aquatic ecosystem metabolism as a tool in environmental management. \u003cem\u003eWIREs Water\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(4), e1521. https://doi.org/10.1002/wat2.1521\u003c/li\u003e\n\u003cli\u003eKemp, W. M., \u0026amp; Testa, J. M. (2012). Metabolic Balance between Ecosystem Production and Consumption. In \u003cem\u003eTreatise on Estuarine and Coastal Science\u003c/em\u003e (Vol. 7, pp. 83\u0026ndash;118). Elsevier Inc. https://doi.org/10.1016/B978-0-12-374711-2.00706-3\u003c/li\u003e\n\u003cli\u003eKirwan, M. L., Megonigal, J. P., Noyce, G. L., \u0026amp; Smith, A. J. (2023). Geomorphic and ecological constraints on the coastal carbon sink. \u003cem\u003eNature Reviews Earth \u0026amp; Environment\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(6), 393\u0026ndash;406. https://doi.org/10.1038/s43017-023-00429-6\u003c/li\u003e\n\u003cli\u003eLangdon, C. (2010). \u003cem\u003eDetermination of dissolved oxygen in seawater by Winkler titration using the amperometric technique\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eLaruelle, G. G., D\u0026uuml;rr, H. H., Slomp, C. P., \u0026amp; Borges, A. V. (2010). \u003cem\u003eEvaluation of sinks and sources of CO2 in the global coastal ocean using a spatially‐explicit typology of estuaries and continental shelves\u003c/em\u003e. https://doi.org/10.1029/2010GL043691\u003c/li\u003e\n\u003cli\u003eLewis, E., \u0026amp; Wallace, D. (1998). \u003cem\u003eProgram developed for CO2 system calculations\u003c/em\u003e (No. ORNL/CDIAC-105). Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory.\u003c/li\u003e\n\u003cli\u003eLeys, C., Ley, C., Klein, O., Bernard, P., \u0026amp; Licata, L. (2013). Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. \u003cem\u003eJournal of Experimental Social Psychology\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(4), 764\u0026ndash;766. https://doi.org/10.1016/j.jesp.2013.03.013\u003c/li\u003e\n\u003cli\u003eLiang, J. H., Deutsch, C., McWilliams, J. C., Baschek, B., Sullivan, P. P., \u0026amp; Chiba, D. (2013). Parameterizing bubble-mediated air-sea gas exchange and its effect on ocean ventilation. \u003cem\u003eGlobal Biogeochemical Cycles\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(3), 894\u0026ndash;905. https://doi.org/10.1002/gbc.20080\u003c/li\u003e\n\u003cli\u003eLowe, A. T., Bos, J., \u0026amp; Ruesink, J. (2019). Ecosystem metabolism drives pH variability and modulates long-term ocean acidification in the Northeast Pacific coastal ocean. \u003cem\u003eScientific Reports\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(1), 1\u0026ndash;11. https://doi.org/10.1038/s41598-018-37764-4\u003c/li\u003e\n\u003cli\u003eManning, C. C. M., \u0026amp; Nicholson, D. P. (2022). \u003cem\u003ednicholson/gas_toolbox: MATLAB code for calculating gas fluxes\u003c/em\u003e. Zenodo. https://doi.org/10.5281/zenodo.6126685\u003c/li\u003e\n\u003cli\u003eMurrell, M. C., Caffrey, J. M., Marcovich, D. T., Beck, M. W., Jarvis, B. M., \u0026amp; Hagy, J. D. (2018). Seasonal Oxygen Dynamics in a Warm Temperate Estuary: Effects of Hydrologic Variability on Measurements of Primary Production, Respiration, and Net Metabolism. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(3), 690\u0026ndash;707. https://doi.org/10.1007/s12237-017-0328-9\u003c/li\u003e\n\u003cli\u003eNeedoba, J. A., Peterson, T. D., \u0026amp; Johnson, K. S. (2012). Method for the Quantification of Aquatic Primary Production and Net Ecosystem Metabolism Using In Situ Dissolved Oxygen Sensors. In S. M. Tiquia-Arashiro (Ed.), \u003cem\u003eMolecular Biological Technologies for Ocean Sensing\u003c/em\u003e (pp. 73\u0026ndash;101). https://doi.org/10.1007/978-1-61779-915-0_4\u003c/li\u003e\n\u003cli\u003eNelson, N. G., Mu\u0026ntilde;oz-Carpena, R., Neale, P. J., Tzortziou, M., \u0026amp; Megonigal, J. P. (2017). Temporal variability in the importance of hydrologic, biotic, and climatic descriptors of dissolved oxygen dynamics in a shallow tidal-marsh creek. \u003cem\u003eWater Resources Research\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(8), 7103\u0026ndash;7120. https://doi.org/10.1002/2016WR020196\u003c/li\u003e\n\u003cli\u003eNew Jersey Climate Change Resource Center. (2020). \u003cem\u003eSea Level Rise in New Jersey: Projections and Impacts\u003c/em\u003e. Rutgers. https://njclimateresourcecenter.rutgers.edu/wp-content/uploads/2020/05/Sea-Level-Rise-in-New-Jersey.pdf\u003c/li\u003e\n\u003cli\u003eNicholson, D., Barrette, J., \u0026amp; Hakai, Z. (2023). \u003cem\u003eboom-lab/winkler-titrator: V0.1.0-alpha (v0.1.0-alpha)\u003c/em\u003e [Computer software]. https://doi.org/10.5281/zenodo.8048209\u003c/li\u003e\n\u003cli\u003eOdum, H. T. (1956). Primary Production in Flowing Waters. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(2), 102\u0026ndash;117. https://doi.org/10.4319/lo.1956.1.2.0102\u003c/li\u003e\n\u003cli\u003eOrr, J. C., Epitalon, J.-M., Dickson, A. G., \u0026amp; Gattuso, J.-P. (2018). Routine uncertainty propagation for the marine carbon dioxide system. \u003cem\u003eMarine Chemistry\u003c/em\u003e, \u003cem\u003e207\u003c/em\u003e, 84\u0026ndash;107. https://doi.org/10.1016/j.marchem.2018.10.006\u003c/li\u003e\n\u003cli\u003ePalevsky, H. I., Clayton, S., Atamanchuk, D., Battisti, R., Batryn, J., Bourbonnais, A., Briggs, E. M., Carvalho, F., Chase, A. P., Eveleth, R., Fatland, R., Fogaren, K. E., Fram, J. P., Hartman, S. E., Le Bras, I., Manning, C. C. M., Needoba, J. A., Neely, M. B., Oliver, H., \u0026hellip; Wingard, C. (2023). OOI Biogeochemical Sensor Data: Best Practices \u0026amp; User Guide. \u003cem\u003eThe Global Ocean Observing System\u003c/em\u003e, \u003cem\u003e1.1.1\u003c/em\u003e(July), 1\u0026ndash;135.\u003c/li\u003e\n\u003cli\u003ePennings, S., \u0026amp; Bertness, M. (2001). Salt marsh communities. In \u003cem\u003eSinauer Associates Inc\u003c/em\u003e (pp. 289\u0026ndash;316).\u003c/li\u003e\n\u003cli\u003ePerkey, D. W., Tedesco, L. P., Fall, K. A., Huff, T. P., \u0026amp; Chasten, M. A. (2024). Stability of marsh edge berms constructed from fine-grained dredged sediment. \u003cem\u003eFrontiers in Marine Science\u003c/em\u003e, \u003cem\u003eAugust\u003c/em\u003e, 1\u0026ndash;14. https://doi.org/10.3389/fmars.2024.1401225\u003c/li\u003e\n\u003cli\u003eRaymond, P. A., Bauer, J. E., \u0026amp; Cole, J. J. (2000). Atmospheric CO2 evasion, dissolved inorganic carbon production, and net heterotrophy in the York River estuary. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(8), 1707\u0026ndash;1717. https://doi.org/10.4319/lo.2000.45.8.1707\u003c/li\u003e\n\u003cli\u003eRo, K. S., \u0026amp; Hunt, P. G. (2006). New Unified Equation for Wind-Driven Surficial Oxygen Transfer into Stationary Water Bodies. \u003cem\u003eTransactions of the ASABE\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(5), 1615\u0026ndash;1622. https://doi.org/10.13031/2013.22020\u003c/li\u003e\n\u003cli\u003eRoberts, D., MacVean, L., Holleman, R., Chelsky, A., Art, K., Nidzieko, N., Sylvester, Z., \u0026amp; Senn, D. (2022). Connections to Tidal Marsh and Restored Salt Ponds Drive Seasonal and Spatial Variability in Ecosystem Metabolic Rates in Lower South San Francisco Bay. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(8), 2560\u0026ndash;2577. https://doi.org/10.1007/s12237-022-01088-4\u003c/li\u003e\n\u003cli\u003eRolando, J. L., Hodges, M., Garcia, K. D., Krueger, G., Williams, N., Carr Jr., J., Robinson, J., George, A., Morris, J., \u0026amp; Kostka, J. E. (2023). Restoration and resilience to sea level rise of a salt marsh affected by dieback events. \u003cem\u003eEcosphere\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(4), e4467. https://doi.org/10.1002/ecs2.4467\u003c/li\u003e\n\u003cli\u003eRosentreter, J. A., Laruelle, G. G., Bange, H. W., Bianchi, T. S., Busecke, J. J. M., Cai, W.-J., Eyre, B. D., Forbrich, I., Kwon, E. Y., Maavara, T., Moosdorf, N., Najjar, R. G., Sarma, V. V. S. S., Van Dam, B., \u0026amp; Regnier, P. (2023). Coastal vegetation and estuaries are collectively a greenhouse gas sink. \u003cem\u003eNature Climate Change\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(6), 579\u0026ndash;587. https://doi.org/10.1038/s41558-023-01682-9\u003c/li\u003e\n\u003cli\u003eRussell, M. J., \u0026amp; Montagna, P. A. (2007). Spatial and temporal variability and drivers of net ecosystem metabolism in Western Gulf of Mexico estuaries. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e30\u003c/em\u003e(1), 137\u0026ndash;153. https://doi.org/10.1007/BF02782974\u003c/li\u003e\n\u003cli\u003eSharp, J. (2023). \u003cem\u003eCO2SYSv3 for MATLAB (Version v3.2.1)\u003c/em\u003e [MATLAB]. https://github.com/jonathansharp/CO2-System-Extd\u003c/li\u003e\n\u003cli\u003eStanley, R. H. R., Jenkins, W. J., Lott, D. E., \u0026amp; Doney, S. C. (2009). Noble gas constraints on air-sea gas exchange and bubble fluxes. \u003cem\u003eJournal of Geophysical Research: Oceans\u003c/em\u003e, \u003cem\u003e114\u003c/em\u003e(11), 1\u0026ndash;14. https://doi.org/10.1029/2009JC005396\u003c/li\u003e\n\u003cli\u003eTan, R. Z., Markus, C., Vasikaran, S., \u0026amp; Loh, T. P. (2022). Comparison of 8 methods for univariate statistical exclusion of pathological subpopulations for indirect reference intervals and biological variation studies. \u003cem\u003eClinical Biochemistry\u003c/em\u003e, \u003cem\u003e103\u003c/em\u003e, 16\u0026ndash;24. https://doi.org/10.1016/j.clinbiochem.2022.02.006\u003c/li\u003e\n\u003cli\u003eTassone, S. J., \u0026amp; Bukaveckas, P. A. (2019). Seasonal, Interannual, and Longitudinal Patterns in Estuarine Metabolism Derived from Diel Oxygen Data Using Multiple Computational Approaches. \u003cem\u003eEstuaries and Coasts\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(4), 1032\u0026ndash;1051. https://doi.org/10.1007/s12237-019-00526-0\u003c/li\u003e\n\u003cli\u003eVachon, D., \u0026amp; Prairie, Y. T. (2013). The ecosystem size and shape dependence of gas transfer velocity versus wind speed relationships in lakes. \u003cem\u003eCanadian Journal of Fisheries and Aquatic Sciences\u003c/em\u003e, \u003cem\u003e70\u003c/em\u003e(12), 1757\u0026ndash;1764. https://doi.org/10.1139/cjfas-2013-0241\u003c/li\u003e\n\u003cli\u003eValiela, I., Lloret, J., Bowyer, T., Miner, S., Remsen, D., Elmstrom, E., Cogswell, C., \u0026amp; Robert Thieler, E. (2018). Transient coastal landscapes: Rising sea level threatens salt marshes. \u003cem\u003eScience of The Total Environment\u003c/em\u003e, \u003cem\u003e640\u0026ndash;641\u003c/em\u003e, 1148\u0026ndash;1156. https://doi.org/10.1016/j.scitotenv.2018.05.235\u003c/li\u003e\n\u003cli\u003eVan Heuven, S., Pierrot, D., Rae, J., Lewis, E., \u0026amp; Wallace, D. (2011). \u003cem\u003eMATLAB Program Developed for CO2 System Calculations. ORNL/CDIAC-105b.\u003c/em\u003e University of St Andrews Research Portal. https://research-portal.st-andrews.ac.uk/en/datasets/matlab-program-developed-for-co2-system-calculations-ornlcdiac-10\u003c/li\u003e\n\u003cli\u003eVasilaky, D. (2025). \u003cem\u003eVif(X)\u003c/em\u003e [MATLAB]. https://www.mathworks.com/matlabcentral/fileexchange/60551-vif-x\u003c/li\u003e\n\u003cli\u003eWallace, R. B., Peterson, B. J., \u0026amp; Gobler, C. J. (2021). Ecosystem Metabolism Modulates the Dynamics of Hypoxia and Acidification Across Temperate Coastal Habitat Types. \u003cem\u003eFrontiers in Marine Science\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(December), 1\u0026ndash;20. https://doi.org/10.3389/fmars.2021.611781\u003c/li\u003e\n\u003cli\u003eWang, S. R., Di Iorio, D., Cai, W. J., \u0026amp; Hopkinson, C. S. (2018). Inorganic carbon and oxygen dynamics in a marsh-dominated estuary. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e(1), 47\u0026ndash;71. https://doi.org/10.1002/lno.10614\u003c/li\u003e\n\u003cli\u003eWang, Z. A., \u0026amp; Cai, W. J. (2004). Carbon dioxide degassing and inorganic carbon export from a marsh-dominated estuary (the Duplin River): A marsh CO2 pump. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(2), 341\u0026ndash;354. https://doi.org/10.4319/lo.2004.49.2.0341\u003c/li\u003e\n\u003cli\u003eWang, Z. A., Kroeger, K. D., Ganju, N. K., Gonneea, M. E., \u0026amp; Chu, S. N. (2016). Intertidal salt marshes as an important source of inorganic carbon to the coastal ocean. \u003cem\u003eLimnology and Oceanography\u003c/em\u003e, \u003cem\u003e61\u003c/em\u003e(5), 1916\u0026ndash;1931. https://doi.org/10.1002/lno.10347\u003c/li\u003e\n\u003cli\u003eWanninkhof, R. (2014). Relationship between wind speed and gas exchange over the ocean revisited. \u003cem\u003eLimnology and Oceanography: Methods\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(JUN), 351\u0026ndash;362. https://doi.org/10.4319/lom.2014.12.351\u003c/li\u003e\n\u003cli\u003eWarnell, K., Olander, L., \u0026amp; Currin, C. (2022). Sea level rise drives carbon and habitat loss in the U.S. mid-Atlantic coastal zone. \u003cem\u003ePLOS Climate\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(6), e0000044. https://doi.org/10.1371/journal.pclm.0000044\u003c/li\u003e\n\u003cli\u003eZhang, J.-Z., Berberian, G., \u0026amp; Wanninkhof, R. (2002). Long-term storage of natural water samples for dissolved oxygen determination. \u003cem\u003eWater Research\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e(16), 4165\u0026ndash;4168. https://doi.org/10.1016/S0043-1354(02)00093-3\u003c/li\u003e\n\u003cli\u003eZuur, A. F., Ieno, E. N., \u0026amp; Smith, G. M. (2007). Linear regression. In \u003cem\u003eAnalysing ecological data\u003c/em\u003e. Springer.\u003c/li\u003e\n\u003cli\u003eZuur, A. F., Ieno, E. N., Walker, N., Saveliev, A. A., \u0026amp; Smith, G. M. (2009). \u003cem\u003eMixed effects models and extensions in ecology with R\u003c/em\u003e. Springer. https://doi.org/10.1007/978-0-387-87458-6\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"fa49b79e-cbdc-4aa5-8b6e-93c5731dc68f","identifier":"10.13039/100006505","name":"Engineer Research and Development Center","awardNumber":"W912HZ2020061–RA3","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Boston College","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6759348/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6759348/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eClimate change is expected to significantly alter carbon dynamics in coastal ecosystems like salt marshes. However, our capacity to detect and respond to these changes is limited by insufficient knowledge of baseline ecosystem metabolism across aquatic ecosystems. Quantifying metabolic rates\u0026mdash;gross primary production (GPP), ecosystem respiration (ER), and net ecosystem metabolism (NEM)\u0026mdash;reveals whether an ecosystem stores (net autotrophic) or releases (net heterotrophic) carbon. Few studies have monitored water quality at high frequency over full annual cycles in marsh-dominated estuaries, limiting estimates of seasonal metabolic variation. In this study, we examined the metabolic balance of a salt marsh channel in coastal New Jersey, an area heavily impacted by human activities and experiencing rapid sea level rise. Over three years, we monitored environmental parameters\u0026mdash;salinity, temperature, dissolved oxygen, pH, turbidity, and chlorophyll \u003cem\u003ea\u003c/em\u003e\u0026mdash;at three sites within the channel. Using these data, we calculated metabolic rates and assessed relationships between environmental conditions and metabolism, including impacts of storm events. Results revealed an overall net heterotrophic system (mean NEM: -29.4 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), with high seasonal variation ranging from slightly net autotrophic in winter/early spring (February maximum: 12.0 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) to net heterotrophic in late spring/fall (September minimum: -88.0 mmol O\u003csub\u003e2\u003c/sub\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e d\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Temperature played a dominant role in metabolic dynamics, while storms temporarily intensified heterotrophic conditions. Our findings reveal seasonal patterns in productivity and respiration that affect carbon storage capacity and document natural ecosystem variability and responses to disturbances\u0026mdash;insights critical for understanding how these vital blue carbon systems may respond to environmental changes.\u003c/p\u003e","manuscriptTitle":"Multiyear monitoring reveals seasonal and short-term dynamics of ecosystem metabolism in a temperate salt marsh channel","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-28 09:01:03","doi":"10.21203/rs.3.rs-6759348/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b1994bd4-32d3-49e2-b9e0-3d32ef3071f0","owner":[],"postedDate":"May 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-07-01T00:27:57+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-28 09:01:03","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6759348","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6759348","identity":"rs-6759348","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}