Human activities have decreased global salt marsh surface area with a subsequent loss in the ecosystem functions they provide. The creation of marshes in terrestrial systems has been used to mitigate this loss in marsh cover. Although these constructed marshes may rapidly recover ecosystem structure, biogeochemical processes may be slow to recover. We compared denitrification and dissimilatory nitrate reduction to ammonium (DNRA) rates between a 32‐year‐old excavation‐created salt marsh (CON‐2) and a nearby natural reference salt marsh (NAT) to assess the recovery of ecosystem function. These process rates were measured at 5 cm increments to a depth of 25 cm to assess how plant rooting depth and organic matter accumulation impact N‐cycling. We found that, for both marshes, denitrification and DNRA declined with depth with the highest rates occurring in the top 10 cm. In both systems, N‐retention by DNRA accounted for upwards of 75% of nitrate reduction, but denitrification and DNRA rates were nearly 2× and 3× higher in NAT than CON‐2, respectively. Organic matter was 6× lower in CON‐2, likely due to limited plant belowground biomass production. However, there was no response to glucose additions, suggesting that the microbial functional community, not substrate limitation, limited nitrate reduction recovery. Response ratios showed that denitrification in CON‐2 recovered in surficial sediments where belowground biomass was highest, even though biomass recovery was minimal. This indicates that although recovery of ecosystem function was constrained, it occurred on a faster trajectory than that of ecosystem structure.

Human activities have led to 1–2% of coastal wetlands lost per year globally, with subsequent losses in ecosystem services such as nutrient filtering and carbon sequestration. Wetland construction is used to mitigate losses of marsh cover and services resulting from human impacts in coastal areas. Though marsh structure can recover relatively quickly (i.e., <10 years) after construction, there are often long‐term lags in the recovery of ecosystem functions in constructed marshes. We conducted a year‐long study comparing seasonal plant productivity, ecosystem respiration (ERCO2), denitrification, and dissimilatory nitrate reduction to ammonium (DNRA) between two 33‐year‐old constructed marshes (CON‐1, CON‐2) and a nearby natural reference marsh (NAT). We found that CON‐1 and CON‐2 were structurally similar to NAT (i.e., plant aboveground and belowground biomass did not differ). Likewise, gross ecosystem productivity (GEP), ERCO2, and net ecosystem exchange (NEE) were similar across all marshes. Further, DNRA and denitrification were similar across marshes, with the exception of greater denitrification rates at CON‐2 than at the other two sites. While pore‐water ammonium concentrations were similar across all marshes, organic matter (OM) content, pore‐water phosphate, nitrate + nitrite, and hydrogen sulfide concentrations were greater in NAT than CON‐1 and CON‐2. Collectively, this work suggests that current marsh construction practices could be a suitable tool for recovering plant structure and some ecosystem functions. However, the lag in recovery of pore‐water nutrient stocks and OM content also suggests that some biogeochemical functions may take longer than a few decades to fully recover in constructed marshes.

Dissolved organic carbon (DOC) flux from rivers in the pan‐Arctic watershed represents an important connection between major terrestrial carbon stocks and the Arctic Ocean. Previous estimates of Arctic carbon flux and dissolved organic matter (DOM) seasonal dynamics have relied predominantly on measurements from the six major Arctic rivers, yet these may not be representative of northern high‐latitude constrained smaller watersheds. Here, we evaluate DOC concentration and DOM composition in the Onega River, a small Arctic watershed, using optical measurements and ultrahigh resolution mass spectrometry. Compared to the six largest Arctic rivers, DOC, absorbance at a350, and indicators of terrestrial DOM (e.g., specific UV absorbance at 254 nm, modified aromaticity index, relative abundance of condensed aromatics and polyphenolics) were elevated in the Onega throughout the year. Seasonality was also generally muted in comparison to the major Arctic rivers with relatively elevated DOC and terrestrial markers in both spring and fall seasons. The Onega exhibits a strong relationship between a350 and DOC, and its organic‐rich nature is apparent in its high DOC yield (4.85 g m2yr−1), and higher chromophoric DOM per unit DOC than the six largest Arctic rivers. As DOC yield from the Onega may be more representative of smaller northern high‐latitude rivers, we derived a new pan‐Arctic DOC flux scaling estimate which is over 50% higher than previous estimates scaled solely from the six major Arctic rivers. These observations suggest that smaller northern high‐latitude rivers may be underrepresented in Arctic carbon flux models and highlights uncertainty around constraining the export of DOC to the Arctic Ocean.

No abstract