COS 66-7 - Using plant traits to predict denitrification potential in salt marsh ecosystems

Thursday, August 11, 2016: 9:50 AM
Palm B, Ft Lauderdale Convention Center
Mary Alldred1, Stoycho Velkovsky2 and Stephen B. Baines2, (1)Natural Sciences, Baruch College CUNY, New York, NY, (2)Ecology and Evolution, Stony Brook University, Stony Brook, NY

Human activities have altered major biogeochemical cycles and the diversity and distribution of species on a global scale, yet our ability to quantify the effects that changes in ecological communities have on ecosystem processes lags far behind management needs.  Specifically, microbial denitrification represents an ecosystem service on which human societies depend for nitrogen removal and maintenance of water quality.  Denitrification is known to respond to differences in the characteristics of wetland plant communities but remains a notoriously difficult process to measure and predict.  Models that incorporate traits of wetland plant communities may provide a complementary approach to traditional hydrologically and chemically based models in predicting rates of denitrification among wetland ecosystems.  In this study, we used a combination of mesocosm experiments and field surveys to determine the influence of the dominant salt-marsh grass Spartina alterniflora on sediment oxygenation and microbial nitrogen-cycling processes.  In replicated experimental mesocosms, we examined the effects of belowground plant traits on sediment oxygen concentrations and key processes of the nitrogen cycle including mineralization, nitrification, and denitrification potential.  We also performed field surveys of plant traits, sediment characteristics, and nitrogen-cycling processes in 11 Spartina-alterniflora dominated salt marshes spanning a variety of land-use conditions on Long Island, NY.


In experimental mesocosms, denitrification potentials were found to correlate strongly to plant traits that enhance sediment aeration (i.e., total root mass and rhizome width), and thus facilitate coupled nitrification-denitrification.  Analysis of field surveys further revealed that simple linear models including plant root mass and the nitrogen content of plant leaves provided predictions of denitrification potentials among sites that were comparable to models incorporating measurements of sediment carbon and nitrogen content as predictors.  Among all sites, denitrification potentials in Spartina-dominated sediments were double those measured in adjacent non-vegetated control plots on average, with the influence of vegetation increasing with total aboveground and belowground biomass of the plant community.  Together our results support the utility of trait-based approaches in understanding the influence of plant communities on sediment conditions and important ecosystem processes.  Moreover, our results suggest that knowledge of the trait composition of plant communities may simplify the task of predicting ecosystem processes, even those as complicated and difficult to predict as denitrification.  Trait-process associations offer one promising approach to the challenge of predicting rates of ecosystem processes in the face of rapid changes in ecological communities due to pressures such as land development, climate change, sea-level rise, and species invasions.