Mangrove forests are hot spots in the global carbon cycle, yet the fate for a majority of mangrove net primary production remains unaccounted for. The relative proportions of alkalinity and dissolved CO 2 [CO 2 *] within the dissolved inorganic carbon (DIC) exported from mangroves is unknown, and therefore, the effect of mangrove DIC exports on coastal acidification remains unconstrained. Here we measured dissolved inorganic carbon parameters over complete tidal and diel cycles in six pristine mangrove tidal creeks covering a 26°latitudinal gradient in Australia and calculated the exchange of DIC, alkalinity, and [CO 2 *] between mangroves and the coastal ocean. We found a mean DIC export of 59 mmol m À2 d À1 across the six systems, ranging from import of 97 mmol m À2 d À1 to an export of 85 mmol m À2 d À1 . If the Australian transect is representative of global mangroves, upscaling our estimates would result in global DIC exports of 3.6 ± 1.1 Tmol C yr À1 , which accounts for approximately one third of the previously unaccounted for mangrove carbon sink. Alkalinity exchange ranged between an import of 1.2 mmol m À2 d À1 and an export of 117 mmol m À2 d À1 with an estimated global export of 4.2 ± 1.3 Tmol yr À1 . A net import of free CO 2 was estimated (À11.4 ± 14.8 mmol m À2 d À1 ) and was equivalent to approximately one third of the air-water CO 2 flux (33.1 ± 6.3 mmol m À2 d À1 ). Overall, the effect of DIC and alkalinity exports created a measurable localized increase in coastal ocean pH. Therefore, mangroves may partially counteract coastal acidification in adjacent tropical waters.

Mangrove forests produce significant amounts of organic carbon and maintain large carbon stocks in tidally inundated, anoxic soils. This work analyzes new and published data from 17 regions spanning a latitudinal gradient from 22°N to 38°S to assess some of the global drivers (temperature, tidal range, latitude, and rainfall) of mangrove carbon stocks. Mangrove forests from the tropics have larger carbon stocks (895 ± 90 t C ha−1) than the subtropics and temperate regions (547 ± 66 t C ha−1). A multiple regression model showed that 86% of the observed variability is associated with annual rainfall, which is the best predictor of mangrove ecosystem carbon stocks. Therefore, a predicted increase in rainfall along the tropical Indo‐Pacific may increase mangrove forest carbon stocks. However, there are other potentially important factors that may regulate organic matter diagenesis, such as nutrient availability and pore water salinity. Our predictive model shows that if mangrove deforestation is halted, global mangrove forest carbon stocks could increase by almost 10% by 2115 as a result of increased rainfall in the tropics.

[1] Coral reefs are characterized by high gross productivity in spite of low nutrient concentrations. This apparent paradox may be partially reconciled if seawater recirculation in permeable sediments over large (meters) and long (hours to days) scales is an important source of recycled nitrogen and phosphorus to coral reefs. In this paper we use radon ( 222 Rn, a natural tracer) to quantify tidally driven pore water (or groundwater) exchange between (1) an offshore coral cay island and its fringing reef lagoon and (2) a reef lagoon and the surrounding ocean. As seawater infiltrates Heron Island at high tide, it acquires a radon signal that can be detected when pore waters emerge from carbonate sands at low tide. A nonsteady state model indicated that vertical pore water upwelling rates (or saline submarine groundwater discharge) were >40 cm/d within the reef lagoon and >100 cm/d outside the lagoon at low tide. Within the lagoon, tidal pumping and temperature-driven convection were the main driving forces of pore water advection. At low tide, the reef lagoon level is about 1 m higher than the surrounding ocean. As a result, a steep hydraulic gradient develops at the reef edge, driving unidirectional filtration through the reef framework. Groundwaters were highly enriched in nitrate (average of 530 mmol, likely influenced by bird guano) relative to lagoon waters (1.9 mmol). Rough but conservative estimates indicated that groundwater-derived nitrate fluxes (7.9 mmol/m 2 /d) can replace the entire lagoon nitrate inventory every <19 days. We speculate that as offshore coral islands "breath" (inhale seawater at high tide and exhale groundwater at low tide), they release nutrients that lead to sustained productivity within coral reefs.

Nitrous oxide (N2O) is an important greenhouse gas, but large uncertainties remain in global budgets. Mangroves are thought to be a source of N2O to the atmosphere in spite of the limited available data. Here we report high resolution time series observations in pristine Australian mangroves along a broad latitudinal gradient to assess the potential role of mangroves in global N2O budgets. Surprisingly, five out of six creeks were under-saturated in dissolved N2O, demonstrating mangrove creek waters were a sink for atmospheric N2O. Air-water flux estimates showed an uptake of 1.52 ± 0.17 μmol m−2 d−1, while an independent mass balance revealed an average sink of 1.05 ± 0.59 μmol m−2 d−1. If these results can be upscaled to the global mangrove area, the N2O sink (~2.0 × 108 mol yr−1) would offset ~6% of the estimated global riverine N2O source. Our observations contrast previous estimates based on soil fluxes or mangrove waters influenced by upstream freshwater inputs. We suggest that the lack of available nitrogen in pristine mangroves favours N2O consumption. Widespread and growing coastal eutrophication may change mangrove waters from a sink to a source of N2O to the atmosphere, representing a positive feedback to climate change.

Wetlands represent the largest natural source of methane; however, very few studies have simultaneously quantified the three main atmospheric flux pathways (i.e., diffusive, ebullition, and plant‐mediated). Unlike better‐studied northern hemisphere systems, many Australian subtropical wetlands undergo extreme wet/dry oscillations, which may strongly impact methane dynamics. We assessed diurnal methane emissions of multiple pathways during three distinct seasonal events within an Australian freshwater wetland. Six‐fold higher methane emissions occurred during summer compared to autumn floods (which followed an extensive dry‐period) and winter/cool conditions. Over three seasons, diffusion represented the highest average areal fluxes (25.9 ± 73.2 mmol m−2 d−1) but were within range of fluxes through water lily aerenchyma (20.8 ± 41.5 mmol m−2 d−1). Average ebullition rates were 5.5 ± 9.7 mmol m−2 d−1. Water column CH4 displayed high spatiotemporal variability, ranging from 55.0 to 253.5 μmol L−1. Time series δ13C–CH4 isotope measurements revealed an oxidation fraction of ~ 15% at night‐time and ~ 36% during day‐time, and night‐time diffusive fluxes were consistently ~ three‐fold higher than day‐time fluxes. By aggregating seasons and weighting for changes in lily coverage, plant‐mediated fluxes accounted for ~ 59% of the annual methane emissions, whereas ebullition and diffusion each accounted for ~ 20%. The up‐scaled annual area‐weighted wetland methane flux (combined pathways) was 27.3 ± 36.7 mmol m−2 d−1. We contend that water lilies (Nymphaea sp.) are the significant carbon source, mediator, and conduit for methane fluxes in this system, and the extremely large seasonal variability of methane emissions reflect dynamic redox oscillations driven by oscillating wet and dry conditions.