Thesis (Ph. D.)--Joint Program in Oceanography (Massachusetts Institute of Technology, Dept. of Biology; and the Woods Hole Oceanographic Institution), February 2002. / Vita. / Includes bibliographical references (leaves 153-161). / I used direct measurements of nitrogen gas (N₂ fluxes and a ¹⁵N stable isotope tracer to determine the contribution of denitrification to salt marsh sediment N cycling. Denitrification in salt marsh tidal creekbottoms is a major sink for groundwater nitrate of terrestrial origin. I studied creekbottom denitrification by direct measurements of N₂ fluxes in closed chambers against a low-N₂ background. I undertook experiments and simulation modeling of sediment N₂ fluxes in closed chambers to optimize the key experimental parameters of this approach. Denitrification in these sediments was driven by the degradation of labile organic matter pools which are depleted during long incubations. Sediment thickness was the most important parameter controlling the required incubation time. Errors of up to 13% with gas headspaces and 80% with water headspaces resulted from headspace N2 accumulation and the resulting collapse of the sediment-water diffusion gradient. These errors could be eliminated by using headspaces of sufficient thickness. Headspace flushing to reduce ammonium accumulation did not affect denitrification rates, but caused transient disturbance of N₂ flux rates. Direct measurements of 0₂, C0₂, N₂, and inorganic N fluxes from the sediments of a salt marsh tidal creek were made in order to examine the interaction of denitrification with the carbon, oxygen, and N cycles. Organic carbon concentration and lability were the primary controls on metabolic rates. C0₂/N flux ratios averaged 6.1, indicating respiration driven by algal biomass. / (cont.) Allochthonous denitrification accounted for 39% of total sediment denitrification (2.7 mol N m⁻² yr⁻¹). 46% of remineralized ammonium was denitrified, while the contribution of autochthonous denitrification to 0₂ and C0₂ fluxes was 18% and 10%, respectively. A ¹⁵N-ammonium tracer was used to study competition between plants and nitrifying bacteria for remineralized ammonium. In undisturbed sediments of Spartina alterniflora, plant uptake out-competed nitrification-denitrification, with plant uptake accounting for 66% of remineralized ammonium during the growing season. Under N fertilization (15.5 mol m⁻² yr⁻¹), both plant N uptake and denitrification increased, but denitrification dominated, accounting for 72% of the available N. When plant uptake was hydrologically suppressed, nitrification-denitrification was stimulated by the excess N, shifting the competitive balance toward denitrification. / by Michael Robert Hamersley. / Ph.D.

Previous research into bottom-up processes on saltmarshes has mainly focused on the influence of plant succession on herbivores. This study will present original research exploring the influence of bottom-up processes in a saltmarsh ecosystem between three trophic levels: Orchestia, redshanks, and sparrowhawks. Density dependence, may be the dominant top-down effect when higher numbers of sparrowhawks and redshanks are present, and may mask top-down and bottom-up trait effects which are constant. Bottom-up effects begin to emerge when cold conditions force redshanks from muddy creeks onto the saltmarsh to forage for Orchestia, because their primary prey, Corophium become less available. Larger flocks form and feeding on Orchestia requires them to balance a need to profit from the best available feeding patches and to be vigilant to sparrowhawk attack. Redshank vulnerability is compounded, because Orchestia hide in cold temperatures, so probing in the soil with their heads down makes them more vulnerable to sparrowhawk attack. Larger flocks may be able to exploit areas closer to sparrowhawk-concealing cover at the terrestrial boundary because they feel safer in greater numbers. Warmer temperatures make Orchestia more active which attracts redshanks, which can simultaneously feed and be vigilant because they peck and catch crawling and jumping Orchestia with their heads up. Consequently, increased flock size may temporarily depress Orchestia abundance, so that redshanks become spaced, leaving isolated individuals more vulnerable to attack. Therefore, it is a temperature-dependent bottom-up process which impacts upon both Orchestia and redshank behaviour, which then may influence the hunting success of sparrowhawks. Whether the characteristics of this saltmarsh ecosystem and the trophic dynamics can be compared to other examples is questionable. Saltmarshes probably differ in their topography and the way in which environmental conditions affect them that then defines which species are present and how these species interact.

Climate change and increasing sea level elevation are predicted to increase salinity in estuarine tidal wetlands in the Southeastern United States. Since much of the ecosystem function in these areas is predicated upon salinity regimes, many fundamental changes are likely to occur as a result. The influence of salinity and SO4 2- reduction on PO4 3- release from tidal wetland soils was evaluated along a salinity gradient at three sites in The St. John’s River, Florida using both field and laboratorybased methods. Porewater was sampled over the course of 10 months to determine ambient levels of SO4 2- and PO4 3-. Lab-based experiments, soils samples were subjected to seawater and SO4 2- treatments in an attempt to induce PO4 3- release. Salinity was lowest at Sixmile Creek (0.45 ± 0.1 g kg-1) and Goodby’s Creek (2.05 ± 2.3 g kg-1) and much higher at Sister’s Creek (27.81 ± 3.1 g kg-1). The organic content of soils was highest (82.35% ± 5.11) at Sixmile Creek, intermediate at Goodby’s Creek (64.45% ± 7.02) and lowest at Sister’s Creek (32.11% ± 9.61). Total soil P was highest at the freshwater Sixmile Creek (1101.64 ± 220.2 μg g-1), intermediate at the brackish Goodby’s Creek (719.61 ± 114.3 μg g-1) and lowest at the Sister’s Creek saltmarsh (475.85 ± 110.9 μg g-1). Porewater PO4 3- was higher at Sixmile and Goodby’s Creek sites (9.44 ± 15.6, 8.99 ± 14.7 !g L-1, respectively) compared to Sister’s Creek (0.6 ± 3.1 !g L-1). Porewater SO4 2- was lower at Sixmile (70.73± 57.58 !g L-1) and Goodby’s Creeks (124.35 ± 152.5 !g L-1) compared to Sister’s Creek (1931.41 ± 557.82 !g L-1). Temporal and spatial trends indicated that SO4 2- and PO4 3- in porewater was likely due to floodwater content and that direct reaction between analytes in soils was unlikely. The addition of aerated seawater failed to cause PO4 3- release from any sites. The incubation of soils under anaerobic conditions, in the presence of Na2SO4 induced SO4 2- reduction, but inhibited PO4 3- flux from both Sixmile and Goodby’s Creek, which is attributed here to likely S- toxicity (Roychoudhury et al., 1999). PO4 3- flux from Sister’s Creek increased in association with Na2SO4 concentration, likely due to more Fe availability to mitigate Stoxicity. Ambient seawater additions to soils under anaerobic conditions followed a similar trend, but the results were not statistically conclusive. Overall, both field and labbased data indicated that Tidal wetland porewater PO4 3- likely originates from floodwaters and that increased salinity and SO4 2- reduction did not directly enhance soil PO4 3- fluxes.