The effects of salt marsh restoration on nutrient retention and nitrogen cycling

Transcription

1 The effects of salt marsh restoration on nutrient retention and nitrogen cycling Katherine Klammer Skidmore College 16 Advisor: Anne E. Giblin The Ecosystems Center at the Marine Biological Laboratory Woods Hole, MA December 17, 214

2 Abstract Restoration can alter the release and storage of nutrients in marsh sediment, resulting in a potential increase of eutrophication in adjacent water bodies. In this study, we created a set of sediment core incubations to model the impact of salt marsh restoration on sediment chemistry. We chose our site in East Sandwich, Massachusetts which had a unimpacted site, a restored site, and an impacted site. The core treatments were made to observe how 1. Draining affects sediment chemistry in impacted sites. 2. Impacted sites respond restoration with salinity change with and without nitrate added. 3. Unimpacted sites compare to the restored sites with and without nitrate added. Our incubation period was 2 weeks and pore water was collected at every treatment for a total of five treatments. Cores were sacrificed at the end to determine extractable nutrients. The cores show that extractable nitrate decreased with salinity, and the core modeling restoration was most efficient at denitrifying nutrients, with and without nitrate added. Many results are different from previous studies, indicating a longer study is necessary to overcome short term effects. Additionally, total core nutrients among duplicates had poor similarity, indicating variability in the sampling site. Keywords: Salt marsh, diking, restoration, eutrophication, nitrogen, phosphorous 1

3 Introduction Salt marshes are often cited as an important ecosystem for conservation. One of their most notable ecological services is acting as a buffer between aquatic and land ecosystems, filtering out nutrients like nitrogen and phosphorous (Nelson, 212). If these nutrients reach adjacent bodies of water, they can cause eutrophication (Nelson, 29). For this reason, marshes are especially important around agricultural areas, where nutrients are leached in high concentrations. However, anthropogenic development disturbs a marsh s natural ability to filter nutrients. Roads that are built over salt marshes can isolate sections of the marsh, effectively cutting them off from tidal flow exchange. Over time, these diked marshes become fresh and, the sediment chemistry is altered (Portnoy, 1997). Decomposition slows down as methanogenesis is the dominant microbial energy pathway. This, along with the low amount of sediment exchange from tides, allows the fresh marsh sediment to have a higher concentration of organic matter. There is also less ion exchange between the pore water and the sediments, and ammonium is adsorbed onto the sediment particles (Portnoy, 1997). Additionally, phosphorous is bound to iron and aluminum, and remains immobilized in the sediment (Portnoy, 1997). Comparatively, salt water marshes have higher rates of decomposition, as there is abundant sulfate in the water and sulfate reduction is the primary bacterial metabolism pathway (Portnoy, 1997). The sea water also increases ion exchange in pore water, meaning more nitrogen and phosphorous is released into the water. As impacted marshes are restored to salt marshes, we would expect to see an increase in decomposition, and more nutrients would be released into the pore water. Previous studies have also indicated the importance of vegetation for nutrient retention, as plants can sequester nutrients (Nelson, 212). Vegetation, and therefore uptake rates can change depending on the salinity and flooding of the site, thus changing how the full scale marsh retains and releases nutrients during restoration efforts. For our project, we went to the Game Farm in East Sandwich in Cape Cod, Massachusetts where we initially looked at nutrients in surface water and sediments along a restoration gradient. The Game Farm has an unimpacted area, where the salt marsh can exchange nutrients freely with the tidal flow; a restored area where previously inhibited tidal exchange was restored in 26; and an impacted area where, a now fresh water marsh has been diked for an estimated 5 years by a railroad. The impacted area has a small culvert which allows for some tidal exchange, but remains mostly fresh. We used experimental incubation cores to imitate restoration effects on impacted sediments with and without nitrate addition, and used sediments from the unimpacted site as a comparison. 2

4 Methods All samples were taken from the Game Farm in East Sandwich, Massachusetts, a salt marsh that has an impounded site, an impounded restored site, and no impoundment (Figure 1). Sampling occurred in November, after plant senescence occurred. Initial At each site, we measured the salinity of water in addition to taking triplicate surface water samples and sediment cores to determine concentration of ammonium, nitrate, and phosphate. The sediment cores were dissected into 5 cm increments and homogenized, and ammonium and nitrate were extracted. 1 ml of 1 N KCl was added to 15 grams of wet sediment, which was shaken for 2 hours. I then let the sediments settle overnight and I filtered the samples using 25 mm GF/F filters. The remaining sections of the cores were dried in a drying oven and weighed to determine wet/ dry ratio. I then extracted phosphate and total phosphorous from the dried sediments using the method from (Harwood,1969). Tidal water measurements were taken after one hour and again after six and a half hours from the beginning at the entrances to each site. Sediments were then submitted for δ 13 C abundance and δ 15 N abundance. Experimental We took twelve cores from the impacted site and four from the unimpacted site for our manipulation experiment. These cores were then drained of pore water and capped within 24 hours of sampling. Each treatment was given to duplicate cores bi weekly for two weeks (Table 1). The fresh water addition was deionized water, and we filtered sea water to use in the salinity treatments. I made a 1 µm solution of NaNO 3 in both 3 ppt salinity and DI for the nitrate additions. We stored the treatment water in a cold room through the duration of the experiment, and tested them for nutrients every time we added them to a core. On the second addition, I added 1 µmoles of δ 15 N to each of the nitrate addition cores, instead of the premade nitrate solution. At this time, we also added additional water to some cores to ensure that the sediments were completely hydrated. After a treatment was added to a core, 1 ml of pore water was drained from the bottom, filtered, and analyzed for nutrient concentration. Excess pore water from the nitrate addition cores were frozen for later isotopic analysis. At the end of the incubation, we drained the cores, dissected the cores into 5 cm increments, and homogenized the sediment. Nutrients in the sediments were extracted in the way initial cores nutrients were. Nutrient Concentration and Stable Isotopes Ammonium: After filtering the samples, I acidified them with 1 µl of 5 N HCl and refrigerated until analysis. Samples were run on the Shimadzu 161 spectrophotometer following the methods from (Solarzano, 1969) 3

5 Nitrate: After filtering the samples, I froze them for later analysis. Surface water samples, KCl extractions and treatment water was analyzed for nitrate on the Lachat (Lachat Instruments). Pore water samples were analyzed using a Teledyne T2 series Nitrous Oxide analyzer (EPA, 29). Phosphorous and Phosphate: After filtering the samples, I acidified them with 1 µl of 5 N HCl and refrigerated until analysis. Total phosphorous and phosphate were analyzed on the spectrophotometer according to Harwood. (Harwood, 1969) I took 1 ml of pore water from each of the treatments for isotope diffusion (a total of 5 ml for each core). Each sample was spiked with 5 µmoles of δ 15 N and diluted to have 1 ml total sample. 6 ml of the 3 parts per thousand salinity with 1 µm NaNO 3 was used as the control. Diffusion methods were modified from Holmes et al (Homes, 1998). Sediments from these cores were homogenized, and dried. I then sampled from the cores and ground the sediment to be analyzed for δ 15 N abundance. Results Initial Surface Water and Tidal Fluxes Initial surface water samples had high concentrations of nitrate for all sites; however, the highest concentration of nitrate was at the impacted site, and the average concentration decreased with distance from the impacted site. Ammonium concentrations increased with distance from the impacted site. Phosphate remained low among all sites, but was highest at the unimpacted site (figure 2). Nye Pond, the fresh water input to the impacted site, had a very high concentration of nitrate, but had very low concentrations of ammonium and phosphate. The impacted site had a slightly lower concentration of nitrate over the outflowing tide, but decreased over time (figure 3). The cranberry bog had a lower concentration of nitrate, but the restored site had about the same concentration of nitrate. Phosphate was high in the cranberry bog, but there were no similarly high concentrations of phosphate in the surface water of the restored site (figure 4). Nitrate was low in the unimpacted site, but generally increased as the tide went out. As the tide started to come back in, there was a decrease in nitrate and ammonium concentrations, but phosphate concentration increases (figure 5). Initial Cores Initial cores taken from the impacted site had a high percent of nitrogen and carbon in the surface sediment (figure 6). 4

6 Experimental In all the impacted cores, the amount of total nitrogen released through pore water over the course of the manipulation experiment was about an order of magnitude less than the total amount of extractable nitrogen left in the cores. The unimpacted 3 ppt (control core) had about half as much total nitrogen in the cores as was extracted from the sediment. In the comparison between drained and un-drained cores, there was no significant difference in the nitrogen exported in the pore water through the experiment. However, drained cores had slightly more nitrate in the efflux (figure 7). The amount of nitrogen in the efflux of the fresh impacted core was not significantly different from the amount of efflux nitrogen in the impacted 3 ppt core. However, on average, the impacted 3 ppt treatment had less nitrogen released in the efflux as the fresher water manipulations. The impacted 3 ppt treatment also had less extractable nitrate in the sediments than the impacted fresh (figure 8). The impacted 3 ppt treatment had much less total efflux and extractable nitrogen than the unimpacted 3 ppt. Additionally, extractable nitrate in the impacted 3 ppt was less than the extractable nitrate in the unimpacted 3 ppt (figure 9). In the impacted fresh water cores, the cores that had nitrate added had about 2 times more nitrogen being exported from the cores than the cores without nitrate. However, this is only slightly more than the nitrogen being exported from the cores without nitrate added plus the amount of nitrate added to the cores (figure 11). In the impacted 3 ppt salinity cores, the amount of nitrogen being exported from the nitrate addition core is slightly higher than the nitrogen being exported from the cores without nitrate added. The amount of nitrate exported from the nitrate addition cores was higher than the cores without nitrate added. Both these treatments in the impacted 3 ppt cores had lower nitrogen being exported than the unimpacted 3 ppt or the impacted fresh cores (figure 12). The amount of nitrogen released from the unimpacted 3 ppt salinity treatment with nitrate added has less nitrogen being exported than the cores without nitrate added (figure 13). Stable Isotope Analysis The sediments from the impacted fresh and unimpacted 3 ppt treatments retained most of the δ 15 N. The impacted 3 ppt had less than 1 µmoles of δ 15 N recovered, and had more of the δ 15 N come out in the efflux than the other two treatments (figure 1). 5

7 Discussion Initial Surface Water and Tidal Fluxes The tidal fluxes had concentrations of nutrients that were expected, with the exception of nitrate. However, the inflow from both Nye Pond and the cranberry bog that had high nitrate concentrations can explain the higher concentrations of nitrate in our sites. During the time of sampling, the cranberry bog was being drained, potentially increasing the concentration of nitrate in the restored area higher than it would be during other periods of the year. The decrease in nitrate concentration from the impacted site down to the unimpacted site is most likely because of dilution, or uptake from the surrounding marsh. Initial Cores The initial impacted cores had a high C:N ratio, as well as higher % N and % C than the restored or unimpacted sites. This is most likely due to the amount of organic matter that is deposited in the peat that does not decompose into inorganic nutrients. Experimental In the drained cores, we expected to see a higher concentration of ammonium and nitrate being released from the cores in efflux (Portnoy, 1997). The aeration that occurs in the cores from draining often expedites decomposition, and also increases the rate of nitrification. However, drained cores can often become acidic, suppressing nitrification (Portnoy, 1997). In our cores, the drained and the un-drained total nitrogen effluxes were similar, which could be due to the cores not being immediately processed after sampling, thus all the sediments experienced initial draining. The drained cores did become more acidic towards the end of the experiment, suggesting that the nitrate that we observed coming out of the drained cores probably occurred when the sediment cores had a higher ph, but were still being oxidized (McCarthy, 214). If the experiment had a longer duration, we might have seen the expected increase in ammonium efflux. While these drained cores are relevant to many other diked salt marshes, our impacted marsh had the fresh water input that kept the impacted marsh completely flooded. The salinity treatments on the impacted sites showed that there was little difference in the amount of nitrogen in the efflux and extractable in the fresh and 3 ppt. This is most likely because the salinity of the cores taken from the impacted site was started at about 7 ppt, and gradually dropped to about 5 ppt for both of the treatments. The impacted 3 ppt on average had less ammonium come out of the cores compared to the impacted fresh cores, which was unexpected. Portnoy et al. saw an increase in pore water ammonium when salt water was added to diked salt marsh sediment (Portnoy, 1997). This could be due to the fact that overall, there was less nitrogen in these sediment cores at the beginning of the experiment, but we were unable to test this and only have initial core data for a rough comparison. It is also possible that there 6

8 was more microbial activity in these cores. In the study on the same sediment cores by Cocallas, she found that the impacted 3 ppt cores had more carbon dioxide production (Cocallas, 214). Because there is more carbon and nitrogen being decomposed in these cores, it is possible that bacteria are using it to build biomass. The impacted 3 ppt also had less extractable nitrate compared to the impacted fresh treatment, which is probably due to sulfur reduction preventing nitrification in these cores (Portnoy, 1997). The impacted 3 ppt treatment also had a lower amount of extractable and efflux nitrogen compared to the unimpacted 3 ppt treatment. Again, this could be due to bacteria in the core, or the amount of nitrogen that was in the cores to start with. We would expect these cores to have a higher amount of nitrogen in the efflux when compared to the unimpacted 3 ppt treatment, but over time decrease. If we lengthened the core incubation time, we might have started to see these trends. It is also possible that expected trends were altered by large rhizomes that were in some of our cores. Nelson studied salt marshes ability to sequester nitrogen, and found that mudflats removed less nitrate (-4%) than sediments with vegetated sediments (1-1%) (Nelson, 29). If we had completely included or excluded vegetation from the cores, differences between the treatments would become more apparent. Among the nitrate addition cores, the impacted 3 ppt cores had the least amount of nitrogen being released in the efflux compared to the other two treatments. This would indicate that the site undergoing restoration is much better at retaining nutrients than the impacted or the unimpacted sites. There have been few other nitrate additions with marsh restoration experiments, however, Thompson s study with restored marshes indicate that natural salt marshes have higher rates of denitrification than restored peat marshes. When nitrate is added to the restored marshes, they had lower levels of denitrification than the restored marsh without nitrate added, but had temporal variation (Thompson, 1995). When nitrate was added to the unimpacted 3 ppt cores, the unimpacted cores did a better job of retaining the added nitrate than the cores without nitrate. However, when nitrate was added to the impacted 3 ppt cores, there was little difference in nitrogen retention in the cores with nitrate and without nitrate added. When nitrate was added to the impacted fresh water cores, there was about 5 µmol more of nitrogen coming out than the cores without nitrate, indicating that the cores with added nitrate were converting the nitrate to ammonium, but not retaining it. In order to get a more accurate portrayal of the trends occurring in these cores, an experiment of longer duration would be necessary. Stable Isotopes Another explanation for the impacted 3 ppt cores having less nitrogen in the efflux and extractable compared to the unimpacted 3 ppt and impacted fresh cores could be denitrification. Even though the isotopes were only added to the nitrate addition cores, this could also explain why there was less extractable nitrate in the impacted 3 ppt cores without nitrate. These cores 7

9 also had a lot more nitrogen being lost in pore water than the other cores, indicating that the sea water was releasing more nitrogen in the efflux compared to the soil in these cores. Conclusions It is important to understand how impacted, restored, and unimpacted marshes retain and release nutrients. The restoration process itself can also release nutrients from sediments causing eutrophication in adjacent water bodies. However, the results from our cores often differed from findings in other, longer term experiments. We found that impacted 3 ppt salinity cores were initially best at retaining nitrogen, with and without nitrate additions. However, we also found that the impacted 3 ppt salinity was better at retaining nitrogen than the unimpacted 3 ppt cores. Additionally, the unimpacted 3 ppt was superior at retaining added nitrate compared to the cores that did not have nitrate added and to the impacted fresh water cores. Finally, efflux and extractable nitrate decreased in the impacted sites as salinity increased, indicating less nitrification was occurring in these cores. In order to apply our findings to the East Sandwich Game Farm salt marshes, as well as future salt marsh restoration projects, an experiment of longer duration would be necessary. This will hopefully reduce the variability between duplicate cores and show long term trends with each core treatment. Acknowledgements I would like to thank Anne E. Giblin for her guidance throughout this project, as well as my project partners Arianna Cocallas, and Christine McCarthy. I would also like to thank the Thornton W. Burgess Society for allowing us to use the East Sandwich Game Farm as a sampling site. I would like to thank Marshall Otter for running isotope samples for me. Finally, I would like to thank Sam Kelsey and Rich McHorney for helping me run the Nox Box and Lachat. 8

10 References Cocallas, Arianna Response of green house gasses (CO2, CH3, N2O) emissions to the increased salinity and nitrogen inputs along a salt marsh restoration gradient. Woods Hole: Semester in Environmental Science. EPA. 29. Teledyn- Advanced Pollution Instrumentation Model 2EU Ultra Sensitive NO/NO2/ NOy Analyzer with Model 51Y converter. Harwood, J.E., R.A. van Steenderen and A.L. Kuhn A rapid method for orthophosphate analysis at high concentrations in water. Water Research 3: Holmes, R. M., et al Measuring 15N- NH4 in marine, estuarine, and fresh water: an adaption of the ammonia diffusion method for samples with low ammonium concentrations."marine Chemistry 6: J.W. Portnoy, A.E. Giblin Biological Effects of Seawater Restoration to Diked Salt Marshes. Ecological Application 7(3): Joanna L. Nelson, Erika S. Zavaleta Salt Marsh as a Coastal Filter for the Oceans: Changes in Function with Exerimental Increases in Nitrogen Loading and Sea-Level Rise. Public Library of Science. Jones, Keith Nitrogen Fixation in a Salt Marsh. Journal of Ecology: Solarzano, L. 1969Determination of ammonium in natrual waters by phenol hypochlorite method. Limnol. Oceangr. 14: Lachat Instruments. Measuring Nitrate/ Nitrite Using the Lachat Flow Injection Analyzer. Standard Operating Procedure. McCarthy, Christine The effects of restoring tidal circulation to diked salt marshes on alkalinity generation and sediment profiles of metals and sulfur. Semester in Environmental Science. Nelson, Joanna L. 29 Does salt marsh function as a coastal filter for nutrient additions from land? Santa Cruz: University of California. Sigman, D. M., M. A. Altabet, R. Michener, D. C. McCorkle, B. Fry, and R. M. Holmes Natural abundance-level measurement of nitrogen isotopic composition of oceanic nitrate: an adaption of the ammonia diffusion method. Marine Chemistry 57: Thompson, Suzanne P Seasonal Patterns of Nitrification and Denitrification in a Natural and a Restored Salt Marsh. Estuaries :

11 Figures and Tables Table 1: Experimental Treatments [page 11] Table 2: Phosphate and Phosphorous in Initial Cores (supplemental) [page 12] Table 3: Phosphate and Phosphorous in Experimental Cores (supplemental) [page 13] Figure 1: Site Map [page 14] Figure 2: Surface Water Nutrients (page 18) [page 15] Figure 3: Impacted Tidal Fluxes [page 16] Figure 4: Restored Tidal Fluxes [page 17] Figure 5: Unimpacted Tidal Fluxes [page 18] Figure 6: Initial Core % C and N [page 19] Figure 7: Nitrogen in Experimental Drainage Core Comparison [page 2] Figure 8: Nitrogen in Experimental Salinity Addition Core Comparison [page 21] Figure 9: Nitrogen in Experimental Salinity Addition Vs. Control [page 22] Figure 1: δ 15 N in Nitrate Addition Cores [page 23] Figure 11: Nitrate Addition Vs. Non- Nitrate: Impacted Fresh [page 24] Figure 12: Nitrate Addition Vs. Non- Nitrate: Impacted 3 ppt salinity [page 25] Figure 13: Nitrate Addition Vs. Non- Nitrate: Unimpacted 3 ppt salinity [page 26] 1

12 Impacted Drained Impacted Control Impacted 3 ppt Salinity Impacted Restored Unimpacted Control Cores Taken From: Impacted Impacted Impacted Impacted Unimpacted Experimental Deionized Deionized Treatment Water Water 3 ppt Water 3 ppt Water 3 ppt Water Drained Yes No No No No Nitrate Added (1 µm) No Yes No Yes Yes Table 1: The experimental treatment plan for the core manipulation. Cores were run in duplicate. Additionally, cores that are indicated with nitrate added also had comparison cores that underwent the same treatment with no nitrate added. 11

13 Initial Core Extractable Phosphate (µmoles) Extractable Phosphorous (µmoles) Restored Restored Impacted Impacted Unimpacted Unimpacted Table 2: The sum of extractable phosphorous and phosphate in the initial cores, in µmoles shown in duplicate. 12

14 Treatment Core Impacted Fresh Drained (1) Impacted Fresh Drained (2) Impacted Fresh (1) Impacted Fresh (2) Impacted Fresh +Nitrate (1) Impacted Fresh + Nitrate (2) Impacted 3 ppt (1) Impacted 3 ppt (2) Impacted 3 ppt (1) Impacted 3 ppt (2) Efflux of phosphate (µmol) Extractable Phosphate (µmol) Extractable Phosphorous (µmol) N/A N/A N/A N/A Impacted 3 ppt + Nitrate (1) N/A N/A Impacted 3 ppt + Nitrate (2) N/A N/A Unimpacted 3 ppt (1) Unimpacted 3 ppt (2) Unimpacted 3 ppt + Nitrate (1) N/A N/A Unimpacted 3 ppt + Nitrate (2) N/A N/A Table 3: The sum of efflux and extractable phosphate and phosphorous in each treatment. Sums of efflux were taken over the addition period; sums of extractable were taken with depth of each core at the end of the experiment. 13

15 Figure 1: A map of the East Sandwich Game Farm with surface water salinities shown at each site at low tide. The orange arrows indicate the areas where each site was or is blocked from tidal flow. The yellow star is the location of sampling for Nye Pond, and the blue star is the sampling site for the cranberry bog. 14

16 Concentration (µm) Avg PO4 Avg NO3 Avg NH4 1 5 Impacted Restored Unimpacted Figure 2: The average of three surface samples taken from each site. 15

17 Concentration (µm) Phosphate Nitrate Ammonium Nye Pond Nitrate Input. 9: 12: 15: Time of day Figure 3: Outgoing tide nutrient concentrations for the impacted sites. All samples were taken from culvert that allowed some tidal flux interchange. 16

18 Concentration (µm) Phosphate Nitrate Ammonium Cranberry Bog Nitrate Cranberry Bog Phosphate. 9: 12: 15: Time of day Figure 4: Outgoing tide nutrient concentrations for the restored site. All samples were taken from the North side of the bridge that originally blocked the restored area from tidal flow. 17

19 Cocnentration (µm) Phosphate Ammonium Nitrate 1.. 9: 12: 15: Time of day Figure 5: Outgoing tide nutrient concentrations at the unimpacted site. All samples were taken at the entrance to the unimpacted marsh (not shown on map). The last points in the time series are the incoming tide nutrient concentrations. 18

20 Depth (cm) % N % C Impacted Restored Unimpacted Figure 6: The initial sediment % composition of nitrogen and carbon from the impacted, restored, and unimpacted sites. 2 19

21 µmoles Nitrogen Drained Fresh Efflux Undrained Fresh 5 Extractable Drained Fresh Figure 7: The comparison of the sum of µmoles in each of the cores for drained and un-drained cores. The cores from each treatment are shown in duplicate. The efflux, shown on the left is the sum of nitrogen found collected from the pore water throughout the two week treatment. The extractable was the sum of nitrogen in the sediment cores after the treatment period. The dashed line indicates how much nitrogen was found in the initial core for this site, ± 1.5 µmoles of total nitrogen was found in the impacted site. Nitrate in triplicate initial cores was minimal, and separation is not shown Undrained Fresh Ammonium Nitrate Extractable Nitrogen in Initials 2

22 µmoles Nitrogen 125 Efflux 7 Extractable Nitrate Ammonium Extractable.. Nitrogen in Initials Impacted Fresh Impacted 3 ppt Impacted 3 ppt Impacted Fresh Impacted 3 ppt Impacted 3 ppt Figure 8: Comparison of the sum of µmoles in each of the cores in the salinity treatment cores, shown in duplicate. The efflux is the total nitrogen collected from the pore water throughout the two week treatment. The extractable is the amount of nitrogen in the sediment cores after the treatment period. The dashed line indicates how much nitrogen was found in the initial core for this site; ± 1.5 µmoles of total nitrogen was found in the impacted site. 21

23 µmoles Nitrogen 8 7 Efflux 8 Extractable Nitrate Ammonium Extractable..Nitrogen in..initials Impacted 3 ppt Unimpacted 3 ppt Impacted 3 ppt Unimpacted 3 ppt Figure 9: The comparison of the total amount of nitrogen in the impacted 3 ppt cores (restoring treatment), and the unimpacted 3 ppt cores (natural salt marsh control). Each treatment is shown with duplicate cores. The efflux is the total nitrogen that left the core through pore water drainage during incubation. The extractable nitrogen is the ammonium and nitrate left in the cores after the incubation period. The dashed line is the average amount of nitrogen extracted from the initial cores. The impacted site had ± 1.5 µmoles of total nitrogen, and the unimpacted site had ± 65.2 µmoles nitrogen. 22

24 µmoles δ 15 N Recovered Water d15 N Soil d15 N Impacted Fresh Impacted 3 ppt Unimpacted 3 ppt Figure 1: The comparison of how much δ 15 N was recovered from the Impacted Fresh cores, the Impacted 3 ppt Cores, and the Unimpacted 3 ppt cores. The dark blue line represents the 1 µmoles of δ 15 N that was put into the nitrate addition cores at the second treatment addition. The amount of δ 15 N that came out in the efflux through the experiment is shown in blue; the amount of δ 15 N in the soil after the incubation period is shown in brown. 23

25 µmoles Nitrogen Nitrate Ammonium 4 2 Impacted Fresh Impacted Fresh + Nitrate Figure 11: Comparison between the total nitrogen collected in the efflux from Impacted Fresh cores with and without nitrate added throughout the experiment. The dashed green line indicates 5 µmoles of nitrate were added to that core throughout the experiment. 24

26 µmoles nitrogen Nitrate Ammonium Impacted 3 ppt Impacted 3 ppt + Nitrate Figure 12: Comparison between the total nitrogen collected in the efflux from Impacted 3 ppt cores with and without nitrate added throughout the experiment. The dashed green line indicates 5 µmoles of nitrate were added to that core throughout the experiment. 25

27 µmoles nitrogen Nitrate Ammonium Unimpacted 3 ppt Unimpacted 3 ppt + Nitrate Figure 13: Comparison between the total nitrogen collected in the efflux from the unimpacted 3 ppt cores with and without nitrate added throughout the experiment. The dashed green line indicates 5 µmoles of nitrate were added to that core throughout the experiment. 26