Groundwater releases CO2 to diverse global coastal ecosystems

Abstract

Coastal ecosystems play a major role in marine carbon budgets, but substantial uncertainties remain in the sources and fluxes of coastal carbon dioxide (CO₂). Here, we assess when, where, and how submarine groundwater discharge (SGD) releases CO₂ to shallow coastal ecosystems. Time-series observations of dissolved CO₂ and radon (²²²Rn, a natural groundwater tracer) across 40 coastal systems from 14 countries revealed large SGD-derived CO₂ fluxes. The mean groundwater partial pressure of CO₂ was 35 times higher than surface seawater. The mean SGD-derived CO₂ flux was 148 ± 226 millimoles per square meter per day (mmol m^-2 day^-1), resulting in a mean water-air CO₂ flux of 80 ± 133 mmol m^-2 day^-1. Tidal rather than diel cycles drove CO₂ enrichment in most ecosystems. Tidally driven SGD was the primary CO₂ source in mangroves, salt marshes, tidal flats, estuaries, and canals. Overall, we expand current knowledge of marine carbon cycles by demonstrating SGD as an important source of CO₂ that requires inclusion in coastal carbon budgets.

Fig. 1.. Location of study sites worldwide and general summary of observations.

(A) The map depicts 40 observation sites across 5 continents and 14 countries. The symbol size represents the mean CO₂ saturation levels in surface water, and the color scale represents the ecosystem type. The inset highlights locations along the east coast of Australia. Details about specific locations and sources of data are summarized in fig. S1 and table S1. (B) The panels show the distribution of Pco₂, ²²²Rn, and DO levels in surface water categorized by ecosystem type. Boxplots reveal the median levels of the three parameters along with their interquartile ranges. Jittered dots representing individual data points enable identification of outliers falling outside whiskers. Additional details about surface water observations can be found in table S4.

Fig. 2.. Correlations between Pco₂ and the proxies for groundwater (²²²Rn) and productivity (DO).

(A) Significant positive correlations were found between Pco₂ and ²²²Rn in 25 of 40 sites (see also figs. S4 and S5). The points above the black line in the insets depict surface water Pco₂ levels exceeding atmospheric Pco₂. The small points represent individual time-series observations, whereas the larger points represent site-specific means. (B) CO₂ versus ²²²Rn, DO, and salinity r values for each site (refer to figs. S4 and S5 and table S4).

Fig. 3.. Comparative analysis of tidal and diel effects.

The effects on Pco₂, ²²²Rn, and DO in surface water, captured during peak hours (fig. S10) of diel and tidal cycles across 40 study sites. The tidal and diel effect indices (TEI and DEI, respectively) describe the relative effect of tidal and diel processes. The white area depicts the region primarily influenced by tides, whereas the gray area represents the region primarily influenced by diurnal cycles. Each point represents an individual study location. The size of the points varies with the tidal range (m) for each site. The color scheme represents the ecosystem type. Two outliers were excluded.

Fig. 4.. Summary of SGD-derived and water-air CO₂ fluxes (mmol m⁻² day⁻¹).

Mean (median) values depicted for SGD-derived and water-air CO₂ fluxes for different ecosystem types. SGD-derived CO₂ fluxes received by the surface waters are in brown circles. Water-air CO₂ fluxes are in blue circles. Arrows represent SGD-derived CO₂ fluxes (bottom) and water-air CO₂ fluxes (top). The circle and arrow sizes qualitatively represent the value of the flux (refer to tables S2 and S5). Icons from T. Saxby retrieved from https://ian.umces.edu/media-library/.