Defining links between plant traits and ecosystem processes is key to understanding how shifts in vegetation structure and composition relating to pollution, species invasions or climate change may impact the development of ecosystems and the services they provide. In coastal marsh ecosystems, removal of nitrate by microbial denitrification is a key ecosystem service that improves coastal water quality. By reducing sediment nitrate concentrations, it could also promote the resilience of the marsh platform by inducing plants to produce more roots. Denitrification requires nitrate, organic carbon and anaerobic conditions. Correlations with plant traits that affect these variables may predict spatial and temporal variation in denitrification, and elucidate the mechanisms governing variability in this process.
We quantified changes in denitrification enzyme activity (DEA), ammonium, nitrate, ammonification and nitrification over the growing season in three coastal Spartina alterniflora wetlands on Long Island, NY. Intra-marsh variation at each site was characterized by sampling unvegetated mud and stands of short- and tall-form Spartina. DEA was regressed on nitrate and various above-ground plant traits. We used these data to test whether plants affected DEA by 1) promoting immobilization of ammonia by microbes, 2) competing with denitrifying microbes for nitrate, or 3) influencing sediment oxygenation.
DEA peaked at all sites within a two week period in late June when plant biomass was increasing rapidly. More than 50% of the cumulative denitrification for the entire season usually occurred during this short period. DEA peaks coincided with declines in exchangeable ammonium and increases in nitrate concentrations within sediments. This pattern reversed during late summer, after plant growth had stopped. Competition for nitrate by plants did not depress microbial denitrification seasonally as hypothesized. Likewise, although immobilization of ammonium by sediment nitrogen-starved microbes was universal, competition for nitrogen among microbes did not appear to prevent peaks in DEA. Instead, DEA varied with nitrate supplied either by nitrification or advection. Plants also play a key role, as nitrate predicted DEA in vegetated plots only. A combination of nitrate, specific leaf area and stem height explained almost all the spatial variation in DEA among vegetated plots. We hypothesize that more plant biomass results in greater sediment oxygenation, leading to more nitrification and denitrification. At very high biomass, by contrast, plants permanently flood the sediment with oxygen, thereby preventing denitrification. The resulting unimodal relationship between plant biomass and denitrification is consistent with a previous meta-analysis of denitrification measurements.