Well-Aerated Southern Appalachian Forest Soils Demonstrate Significant Potential for Gaseous Nitrogen Loss

Understanding the dominant soil nitrogen (N) cycling processes in southern Appalachian forests is crucial for predicting ecosystem responses to changing N deposition and climate. The role of anaerobic nitrogen cycling processes in well-aerated soils has long been questioned, and recent N cycling research suggests it needs to be re-evaluated. We assessed gross and potential rates of soil N cycling processes, including mineralization, nitrification, denitrification, nitrifier denitrification, and dissimilatory nitrate reduction to ammonium (DNRA) in sites representing a vegetation and elevation gradient in the U.S. Department of Agriculture (USDA) Forest Service Experimental Forest, Coweeta Hydrologic Laboratory in southwestern North Carolina, USA. N cycling processes varied among sites, with gross mineralization and nitrification being greatest in high-elevation northern hardwood forests. Gaseous N losses via nitrifier denitrification were common in all ecosystems but were greatest in northern hardwood. Ecosystem N retention via DNRA (nitrification-produced NO3 reduced to NH4) ranged from 2% to 20% of the total nitrification and was highest in the mixed-oak forest. Our results suggest the potential for gaseous N losses through anaerobic processes (nitrifier denitrification) are prevalent in well-aerated forest soils and may play a key role in ecosystem N cycling.


Introduction
Changing patterns of nitrogen (N) deposition and climate are expected to dramatically alter ecosystem N and carbon (C) cycling. Nitrogen deposition [1][2][3], soil-warming-enhanced soil N mineralization [4,5], and increased N deposition in the southern Appalachian Mountains may potentially increase N concentrations in streams and rivers [4]. Increased N availability and higher temperatures could also result in greater forest productivity and increase the storage of atmospheric C in soil [6,7]. Additionally, greater reductions in moisture conditions in this ecosystem could affect forest productivity [8]. However, these gains could be reversed if the addition of labile C as leaf and root material results in enhanced respiration [9] and higher emissions of nitrous oxide, a potent greenhouse gas that is almost 300 times more effective than carbon dioxide at heating the atmosphere [10]. Recent studies have demonstrated the importance of anaerobic processes (e.g., denitrification and dissimilatory nitrate reduction to ammonium (DNRA)) in generally well-aerated soils [11], which are soils that are low in soil moisture and have no evidence of anaerobic conditions such as sulfate reduction, suggesting that to fully elucidate the effects of changing N deposition and temperature

Site Description
This study was conducted at Coweeta Hydrologic Laboratory, a U.S. Department of Agriculture (USDA) Forest Service experimental forest and a National-Science-Foundation-funded Long-Term Our objective was to determine the relative importance of specific N cycling processes in four major forest types of the southern Appalachian Mountains using a combination of potential and gross N cycling assays. We hypothesized (1) that sites with high N availability (cove hardwood and northern hardwood forests) would exhibit the greatest rates of gaseous N losses and that these rates explain a substantial proportion of currently unaccounted for N losses, and (2) that the importance of DNRA is greater in sites with high C availability (low C:N; northern hardwood) and lower in sites with low C availability (high C:N; mixed-oak-pine and mixed-oak sites).

Site Description
This study was conducted at Coweeta Hydrologic Laboratory, a U.S. Department of Agriculture (USDA) Forest Service experimental forest and a National-Science-Foundation-funded Long-Term Ecological Research (LTER) site in the southern Appalachian Mountains of western North Carolina. The average temperatures in Coweeta range from 20 • C midsummer (June-August) to 5 • C midwinter (December-January), with an annual average precipitation of 1900 mm. The growing season extends from May through September and receives an average precipitation of 800 mm of precipitation [35,36].
The study was conducted in four large 80 × 80 m plots established along the elevation and vegetation gradient within the Coweeta basin in 1991. All sites are located in reference watersheds (undisturbed since 1927) and include the four major vegetation types present in the southern Appalachians, including xeric mixed-oak-pine, mesic cove hardwood, mesic mixed-oak, and mesic northern hardwood forests [37,38]. Soils are predominately in the Ultisols and Inceptisols orders and have profiles characterized by a well-developed mull-type O-horizon (Oa-horizon of 2-4 cm), A horizon (7-22 cm), transitional horizon (AB or BA; 10-20 cm), and a clay accumulation layer (Bt or Bw; 30-40 cm) with total soil pedon depth ranging from 60 to 100 cm [39]. Table 1 contains more detailed climatic, vegetation and soils information.  [15], and mean annual temperature and precipitation data are from [16].

Sample Collection
We collected organic and mineral soil layers for use in laboratory determination of potential N cycling processes five times over 16 months (November 2010, March 2011, June 2011, November 2011, and March 2012). Soils (O-horizon and 0-15 cm mineral soil) were collected using a stainless-steel (2.5 cm inside diameter) corer (n = 6; except for November 2012, when n = 3) sieved to <2 mm and subdivided for laboratory assays. We determined potential nitrification and denitrification rates to assess temporal dynamics. We also collected soils from the most active part of the soil horizon (0-10 cm mineral soil) in May 2012 to measure gross N cycling rates using 15 N isotopically labeled NO 3 and NH 4 . Samples used for potential nitrification assays were stored at room temperature for <72 h and denitrification samples were stored at 4 • C for <2 weeks before conducting the assays. Samples collected for 15 N gross assays were stored at room-temperature conditions and assays were conducted within 48 h of collection.

Nitrogen Cycling: Potential Rates
Potential nitrification and denitrification rates were determined using a modified slurry approach [32,41]. Nitrification was determined using 5 g of organic or mineral soil, mixed with 15 mL of media (2.5 mM (NH 4 ) 2 SO 4 , 0.80 mM K 2 HPO 4 , and 0.20 mM KH 2 PO 4 ) and incubated on a shaker at 150 rpm and 25 • C to maintain oxic conditions. Subsamples (1 mL) of the soil/media slurry were collected at 0.5, 2.0, 6.0-8.0, and 24.0 h for NO x analysis using colorimetric methods [42]. Potential denitrification rates were determined using the acetylene block method [43]. Five grams of soil (2 g for O-horizon) were mixed with 5 mL of media (1 mM dextrose and 1 mM nitrate), purged with He for 30 min, and placed in a 30 mL serum vial. Ten percent of the headspace volume was displaced with acetylene. The soil/media slurry was shaken at 150 rpm and 25 • C, and the head space (1 mL) was subsampled at 0, 2.0, and 6.0 h for N 2 O analysis using a GC-ECD (Shimadzu Inc., Tokyo, Japan).

Nitrogen Cycling: Gross Rates
We measured gross rates of N mineralization (i.e., ammonification), nitrification, nitrifier denitrification (reduction of ammonia to N 2 ), and DNRA using laboratory 15 N tracer approaches and analysis of the 15 N content of the microbial biomass [44,45]. We composited soil samples collected in May 2013 (n = 4) by site into a single sample for gross N analysis, and samples were sieved together to ensure complete mixing. Thirty grams of each composited soil sample were extracted using 100 mL of 2 M KCl to determine initial NO 3 and NH 4 -N content and natural 15 N abundance. To determine initial microbial biomass N and 15 N, 10 g of soil was extracted with 80 mL of 0.5 M K 2 SO 4 and a second 10 g sample was fumigated with ethanol-free chloroform, incubated for 5 days, then extracted with 0.5 M K 2 SO 4 [45,46]. The remaining composited sample was divided in two (approximately 300 g each) for use in 15 NO 3 and 15 NH 4 labeling experiments. We calculated site-specific label enrichment additions equal to 20% of extractable NO 3 and NH 4 to minimize fertilization artifacts [11]. One sample was labeled with NO 3 (3.8% ± 2.3% K 15 NO 3 (98%)) in the amounts of mixed oak-pine, 4 Following thorough mixing, 50 g of each isotopically labeled soil sample was placed in 1 L Mason jars (six replicate jars per composited soil sample). The jars were sealed with lids adapted with a septum for gas sampling purposes and incubated under field moisture conditions (ranging from 31% to 68% gravimetric soil moisture) at 20 • C. At each sample collection, we randomly selected three replicate sample jars to subsample. From the 15 NH 4 -labeled jars, we collected two 9 mL gas samples for N 2 and N 2 O concentrations and isotopic analysis ( 15 N 2 O and 15 N 2 ) after 4 and 24 h of incubation. Gas samples were stored in Labco Exetainers (Labco Limited Inc., Ceredigion, UK) at room temperature until analysis. After the gas sample collection, 20 g of soil from each incubated jar (NO 3 -and NH 4 -labeled) was extracted with 100 mL of 2 M KCl to determine NH 4 and NO 3 concentrations and their 15 N relative abundance in these specific inorganic nitrogen species. To estimate microbial biomass nitrogen (MBN) and 15 N, 10 g of soil was extracted with 80 mL of 0.5 M K 2 SO 4 , and another 10 g was fumigated with ethanol-free chloroform for 5 days and subsequently extracted with 0.5 M K 2 SO 4 according to Cabrera and Beare [46] and Silver et al. [45]. We used the difference between the nonfumigated and fumigated extractions to calculate microbial biomass, total microbial N, and microbial 15 N uptake.

Carbon Processing
Labile carbon (LC) and potential denitrification were determined for the soil samples composited as previously described. LC was assessed using potential C mineralization incubations under anoxic (LC anoxic ) and oxic (LC oxic ) conditions via a modified potential soil respiration method [47]. Five grams of field moist composited soil were weighed into six 50 mL centrifuge tubes fitted with septa for gas sampling. Soils were adjusted to 50% gravimetric soil moisture with deionized water. Half of the tubes (n = 3) were purged with He for 1 min to assure anoxic conditions and half (n = 3) were purged with CO 2 -free air for 1 min to represent ambient oxygen conditions. Soils were incubated at 25 • C and 1 mL gas samples were taken 0.5, 4, and 24 h after capping the centrifuge tubes. Gas samples were immediately analyzed for CO 2 content using a LICOR LI-7000 setup for small volume injections. Potential respiration rates in µg C g soil −1 d −1 were determined by regression analysis. Potential denitrification rates were assessed as described earlier.

Soil Chemical, N, and 15 N Analyses
Concentrations of NH 4 , NO 3 , and NO 2 were determined colorimetrically on KCl extracts [42,48] using an AlpKem model 3590 Autoanalyzer (OI Analytical, College Station, TX, USA). Soil total carbon was determined using near-infrared reflectance spectroscopy (NIRS) techniques as described in Baas et al. [32]. Microbial biomass was determined as the difference between pre-and postfumigated K 2 SO 4 extracts analyzed for dissolved organic carbon (DOC) using a total carbon analyzer (Shimadzu TOC-V CPH TNM). MBN was determined by persulfate digestion and colorimetric analysis of NO 3 of the K 2 SO 4 extract [46]. Soil extracts were prepared for isotopic analysis using the diffusion method to capture both ammonium and nitrate (after reduction by Devarda's alloy), in which solutions were alkalized and the volatilized ammonia was trapped on an acidified filter. Nitrogen isotopic ratios of the trapped N were determined using a PDZ Europa ANCA-GSL elemental analyzer interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK) at the University of California, Davis Isotope Facility. N 2 O concentrations were determined using a gas chromatograph coupled to an electron capture detector ( 63 Ni) (Shimadzu GC-14A). Gaseous isotopic ratios (N 2 O and N 2 ) were determined using a ThermoFinnigan GasBench with a PreCon trace gas concentration system interfaced to a ThermoScientific Delta V Plus isotope-ratio mass spectrometer (Precon-IRMS, Bremen, Germany) at the University of California, Davis Isotope Facility.
Nitrogen cycling rates calculated for the different potential and gross processes were calculated as described below and are summarized in Table 2. Potential net nitrification and denitrification rates were calculated using t = 0 and t = 24 h measurements of nitrate and N 2 O, respectively, and are reported as µg kg soil −1 h −1 and in mg N m −2 d −1 using bulk densities previously determined for each site [32].
Gross nitrogen cycling N mineralization, nitrification, NH 4 + , and NO 3 − consumption rates were calculated based on isotope pool dilution and are reported as µg N g soil −1 d −1 [44,49]. Gross DNRA was calculated from the 15 NO 3 − -labeled soil incubations as the difference in 2 M KCl extractable NH 4 between t = 0 and t = 24 h, corrected for the mean residence time (MRT) and divided by the average NO 3 pool atom % 15 N excess [27,45]. We used the 15 NH 4 -labeled soils (nitrifier denitrification) to measure N 2 and N 2 O production during the 24 h incubation period, calculated as 15 N 2 and 15 N 2 O production divided by the average NH 4 atom % 15 N excess [27]. Microbial biomass was calculated as fumigated minus nonfumigated DOC from the K 2 SO 4 extracts with the efficiency correction factor of 0.45 [50].
Microbial biomass N was calculated as the difference between fumigated and nonfumigated total N in the K 2 SO 4 extract with the efficiency correction factor of 0.54 [51]. The shade indicates that this type of data is needed for estimating the specific process.

Statistical Procedures
We tested for differences in potential nitrogen cycling among sites using a mixed model approach with sample date and site as fixed effects. If a significant interaction between date and site was detected, differences between sites were analyzed at each sample date and differences between sample dates were analyzed for each site. We used a one-way ANOVA to test for differences among sites for gross N cycling rates, N 2 and N 2 O flux rates, C mineralization rates, and microbial biomass (C and N). Data were log-transformed as necessary to meet ANOVA assumptions of normal distribution, which was confirmed using the goodness-of-fit function in JMP. All variability errors are presented by the standard error of the mean unless specified. We used linear regression analysis to investigate relationships between gross nitrogen cycling processes and potential respiration and potential denitrification rates. The northern hardwood site differed significantly from other sites in process rate measurements and organic matter concentrations [16], predicting different potential process drivers in this site. Therefore, we conducted additional tests of the relationships between N cycling processes excluding this site. Statistical analyses were conducted using JMP 11.0 (SAS Institute Inc., Cary, NC, USA). Significant effects are reported at p < 0.05 unless stated otherwise.

Nitrogen Cycling
Gross nitrogen transformation rates, including N mineralization, nitrification, DNRA, nitrifier denitrification, and N 2 O production, were determined in May 2012 (Table 3), as well as potential nitrification and potential denitrification from November 2010 through March 2012 (Figure 2). Gross N mineralization rates varied among sites and were greater in northern hardwood compared with mixed oak and mixed oak-pine (F 3,8 = 7.2, p = 0.01), while gross nitrification (F 3,8 = 7.2, p = 0.001) rates were greater in northern hardwood compared with all other sites. DNRA rates were greater in northern hardwood compared with mixed oak-pine (F 3,8 = 19.2, p < 0.001). The percent of nitrate produced from nitrification that was reduced by DNRA was greater in mixed oak (20%) compared with northern hardwood (2%) (F 3,7 = 3.8, p = 0.07). Percentages were 9% for mixed oak-pine and 16% for cove hardwood. Nitrifier denitrification also differed by site, with the greatest rates occurring in northern hardwood compared with cove hardwood and mixed oak-pine; rates in mixed oak were greater than mixed oak-pine (F 3,8 = 19.2, p < 0.05). Nitrous oxide flux was greatest in soils from northern hardwood, while measurements from all other sites were below detection. Initial KCl extractable NO 3 − and NH 4 + ranged from 0.77 to 5.6 µg N g soil −1 (Table 4).

Microbial Biomass and C Lability
Microbial biomass C did not differ among sites; however, the microbial biomass N was significantly lower in mixed oak-pine compared with the other sites (F3,12 = 12.3, p < 0.001; Table 4). C mineralization rates (Table 4) differed significantly by site (F3,14 = 11.7, p < 0.001) and were greater in northern hardwood soils compared with cove hardwood and mixed oak-pine (p < 0.001). Potential C mineralization rates were greater under oxic conditions (F1,14 = 33.3, p < 0.0001) for all sites. Interaction between site and redox treatment was not significant.

Drivers of Gross Nitrogen Cycling
We used potential denitrification and carbon mineralization rates to explain the variance in the gross nitrogen cycling processes we measured (Table 5). Potential denitrification rates were positively related to gross N mineralization, nitrification, nitrifier denitrification, and DNRA rates (Table 5). Oxic carbon mineralization rates were positively correlated with DNRA and nitrifier denitrification. Anoxic carbon lability estimates were correlated with nitrifier denitrification. The northern hardwood system was an outlier in many analyses and excluding this site from statistical analyses revealed that labile carbon was negatively correlated with N2O production (Table 5). N mineralization was positively correlated to gross nitrification rates, nitrifier denitrification, and DNRA (Table 6). Gross nitrification rates were positively correlated with nitrifier denitrification ( Table 6). All sites included in analyses "All" and northern hardwood site is excluded from analysis in "Excl NH". DNRA = dissimilatory nitrate reduction to ammonium; Cmin = carbon mineralization; pDNF = potential denitrification; N2O-NH4 + = N2O fluxes in NH4 + -labeled soils; N2O-NH4 + = N2O fluxes in   Potential nitrification and denitrification rates varied from 0 to 0.46 and 0 to 307 ng g soil −1 h −1 , respectively ( Figure 2). Mean potential nitrification rates did not differ among sites, while mean potential denitrification rates were significantly greater in northern hardwood than in all other sites, with no significant differences among the other sites ( Table 4). The averages along all sampling times (Table 4) showed that the northern hardwood site exhibited greater rates than any of the other sites.

Microbial Biomass and C Lability
Microbial biomass C did not differ among sites; however, the microbial biomass N was significantly lower in mixed oak-pine compared with the other sites (F 3,12 = 12.3, p < 0.001; Table 4). C mineralization rates (Table 4) differed significantly by site (F 3,14 = 11.7, p < 0.001) and were greater in northern hardwood soils compared with cove hardwood and mixed oak-pine (p < 0.001). Potential C mineralization rates were greater under oxic conditions (F 1,14 = 33.3, p < 0.0001) for all sites. Interaction between site and redox treatment was not significant.

Drivers of Gross Nitrogen Cycling
We used potential denitrification and carbon mineralization rates to explain the variance in the gross nitrogen cycling processes we measured (Table 5). Potential denitrification rates were positively related to gross N mineralization, nitrification, nitrifier denitrification, and DNRA rates (Table 5). Oxic carbon mineralization rates were positively correlated with DNRA and nitrifier denitrification. Anoxic carbon lability estimates were correlated with nitrifier denitrification. The northern hardwood system was an outlier in many analyses and excluding this site from statistical analyses revealed that labile carbon was negatively correlated with N 2 O production ( Table 5). N mineralization was positively correlated to gross nitrification rates, nitrifier denitrification, and DNRA (Table 6). Gross nitrification rates were positively correlated with nitrifier denitrification (Table 6). Table 5. Spearman's correlation coefficients between gross nitrogen cycling rates and edaphic characteristics (carbon mineralization rates (C min ) incubated under oxic or anoxic conditions and potential denitrification). All sites included in analyses "All" and northern hardwood site is excluded from analysis in "Excl NH". DNRA = dissimilatory nitrate reduction to ammonium; C min = carbon mineralization; pDNF = potential denitrification; N 2 O-NH 4 + = N 2 O fluxes in NH 4 + -labeled soils; N 2 O-NH 4 + = N 2 O fluxes in NH 4 + -labeled soils. Sample number was between 9 and 12 for the All analysis and 7 and 9 for the Excl NH analysis. Correlations in bold highlight significant relationships. ** p < 0.01, * p < 0.05, † p < 0.1. Table 6. Pearson's correlation coefficients between gross nitrogen cycling processes.

Discussion
This study focused on elucidating how potential and gross rates of soil N transformations differ among representative sites in the southern Appalachian Mountains and with microbial biomass C and N and soil C availability. We measured high potential denitrification rates in these well-aerated soils along with significant rates of nitrifier denitrification, both of which result in soil N 2 gas emissions. We also found that while DNRA rates were low, this pathway may result in significant nitrogen retention in the mixed-oak site.

Nitrogen Mineralization and Nitrification
The rates of gross N mineralization and nitrification in this study exhibited similar patterns to net transformation rates determined on these sites between 1992 and 2010 using the closed core in situ incubation method [14,15,52]. The proportion of mineralized N that was subsequently nitrified was also similar for both net and gross measurements in the cove hardwood (net = 11%; gross = 9%), mixed-oak (net = 14%; gross = 10%), and northern hardwood (net = 50%; gross = 50%) sites. However, the fraction of N nitrified in the low-N mixed-oak-pine site was substantially greater when measured using gross methods (20%) compared with net (9%) [15]. This is similar to the findings of Hart et al. [53] and suggests that net mineralization and nitrification rates can be poor predictors of the mechanisms regulating gross nitrogen cycling dynamics. Our data suggest we should be cautious when using net rates to parameterize mechanistic nitrogen cycling models, especially in low-N sites.

Gaseous Nitrogen Losses
The reduction of nitrate to N 2 via nitrifier denitrification proved to be a potentially important process in all sites. We measured greater nitrifier denitrification in the high-elevation sites (mixed oak and northern hardwood) than in the low-elevation sites (oak-pine and cove hardwood). Nitrifier denitrification could be responsible for the release of N 2 or it could be the product of the anammox pathway [34,54], but the current experimental design could not distinguish between the two processes. However, Davies et al. (unpublished data) measured gross anammox and denitrification in the mixed-oak site in the Coweeta basin, and their data suggested that anammox contributed <5.7% of the N 2 produced during nitrate reduction, suggesting that nitrifier denitrification is, in fact, the likely dominant pathway for N 2 formation from NH 4 . We found little accumulation of labeled nitrate during measurement of gross nitrifier denitrification, suggesting that, in situ, this process reduces NO 3 with a final product of N 2 , as posited by Wrage et al. [55]. For example, in the mixed-oak site, which had high nitrifier denitrification rates, potential denitrification rates were not high, suggesting that N 2 was produced from NH 4 and it was unlikely to be the product of a coupled nitrification-denitrification pathway.

Nitrogen Retention by DNRA
Inorganic nitrogen retention, calculated as the ratio of gross DNRA to gross nitrification, is an important process in moist tropical forests [45,56], yet it has rarely been considered in temperate ecosystems [57]. We detected DNRA in all sites, similar to a study conducted in nonsaturated temperate zone soils by [11]. Our data suggest that in mixed-oak ecosystems, often characterized by low nitrification rates, DNRA could be a crucial nitrogen retention pathway. DNRA could reduce up to 20% of the NO 3 produced during nitrification back to NH 4 , preventing loss by either denitrification or leaching. We measured significant rates of gross 15 NO 3 reduction, which could also occur via ANRA. On the other hand, we observed no change in microbial biomass 15 N during the incubation, making ANRA less likely [58].

Drivers of Soil Nitrogen Cycling
We examined the relationships between gross and net N transformation processes to discern the primary drivers of soil N cycling. This information is timely, as N deposition across the southeastern United States has shifted from being dominated by NO 3 to NH 4 forms of nitrogen [59,60]. It is likely that greater ammonium availability would enhance nitrifier denitrification N 2 O emissions in the aerated soils of the southern Appalachian Mountains [61]. Gross nitrifier denitrification was correlated with both potential denitrification and labile carbon commonly released from root exudates, shown to increase nitrifier denitrification and fluxes of N 2 and N 2 O [62].
Research conducted by Brumme et al. [63] and Groffman et al. [64] in forest ecosystems that were historically exposed to high N deposition in Germany and the northeastern United States has suggested that soils high in organic matter yield the greatest N 2 O fluxes. Their data suggest that N 2 O fluxes will increase as soil inorganic N concentrations increase. Our data from the northern hardwood forest support this finding, where we found high potential N 2 O fluxes in the soils with high total organic matter and nitrate concentrations [14]. Additionally, we found that N 2 O production was negatively correlated with labile carbon in NH 4 -amended incubations. The positive correlation between labile C and denitrification suggests that increasing carbon availability in N-rich ecosystems would increase N 2 O emissions due to nitrifier denitrification. This process is positively correlated with N 2 O fluxes in the NH 4 -labeled soils, making this the most likely process responsible for N 2 O production in the high N sites. Similar to other studies [63,64], we found that sites characterized by an abundance of both available nitrate and labile carbon are likely to exhibit high N 2 O emissions.

Conclusions
Gross rates of N transformations, such as mineralization and nitrification, determined in the laboratory varied among sites and were not always similar to net rates determined in situ. Nitrate reduction by nitrifier denitrification produced gaseous N as both N 2 and N 2 O at rates not previously measured in these well-aerated soils, suggesting that this process may be relevant in the environmental conditions of the southern Appalachians. Our data suggest that DNRA plays an important role in N retention in the mixed-oak site, which exhibited low NO 3 and NH 4 availability and a relatively high ratio of DNRA to nitrification. While the role of N 2 O emissions under future deposition and climate scenarios remains unclear, we suggest that the mixed-oak and northern hardwood forest sites appear to have the greatest potential for increased emissions from the well-aerated soil of the southern Appalachian Mountains. Future regional studies are critically needed to confirm how widespread these novel findings are across this region and its implications for biogeochemical and climate change models.