Coastal wetlands, including tidal marshes, are particularly efficient at sequestering Carbon (C) and offsetting the effects of anthropogenic C emissions. However, approximately 25% to 50% of the world’s vegetated coastal habitat has been destroyed in the past 50 years, representing a massive loss of C-sink capacity. Wetland restoration presents an opportunity to recover this ecosystem function, but significant temporal lags in the restoration of C-sink capacity have been recorded for many constructed marshes. Carbon sequestration is regulated, in part, by organic matter (OM) decomposition, a microbially-mediated process, yet microbial biomass and community composition often differ between natural and constructed marshes. Decomposition is also controlled by plant litter quality, but little is known about how litter quality differs between constructed and natural marshes. In this study, we asked: 1) do rates of litter decomposition differ between constructed and natural marshes?; and if so, 2) what factors drive these differences? To answer these questions, we quantified decomposition rates of Juncus roemerianus shoot tissue using litterbags installed in one natural and two constructed marshes along the Alabama Gulf Coast. In addition to mass loss, we measured litter quality (C:N:P, lignin), bacterial biomass, and fungal biomass in all three marshes.
Results/Conclusions
Initial litter quality of J. roemerianus shoots did not differ in lignin, carbon, nitrogen and phosphorous content between marshes. However, decomposition rates of J. roemerianus shoots were significantly higher in the natural marsh (k (instantaneous decay constant (per day)) = 0.0042 ± 0.001) than in the two constructed marshes (k = 0.0019 ± 0.0001 and k = 0.0015 ± 0.0001, respectively; ANOVA, p < 0.001). This difference in organic matter decay may have been driven by differences in fungal biomass between the marshes, as ergosterol content on shoot litter was significantly greater in the natural marsh than the constructed marshes on day 0 (ANOVA, p = 5.66e-07), day 47 (ANOVA, p = 0.007) and day 98 (ANOVA, p = 0.004). These results suggest that rates of OM processing in these constructed marshes have not reached functional equivalency with the natural marsh 33 years after their creation.