Soil Nitrogen Dynamics in Organic and Mineral Soil Calcareous Wetlands in Eastern New York

C of wetlands is often highly problematic due to the need to consider soil, plant, and Calcareous wetlands are of great interest in conservation biology. Previous research has suggested that there are significant differences hydrologic factors as fundamental components of these in soil microbial N cycle processes between calcareous wetlands develecosystems (Bridgham et al., 1996). The term wetland oped on organic vs. mineral soils. In the study presented here, we refers to ecosystems that are characterized by hydric measured potential net N mineralization and nitrification, denitrificasoils and plant and animal species adapted or partially tion enzyme activity and soil inorganic N levels at 25 calcareous adapted to life in saturated conditions (National Rewetland sites with variable substrate types. We also evaluated the search Council, 1995). However, there are often multiresponse of N cycling to livestock grazing by sampling at two sites ple names for specific wetland types, reflecting the diwith heavy grazing activity. All N cycle variables were significantly versity of approaches that have been taken to wetland higher in organic soils than mineral soils on a weight basis; however, evaluation and classification through decades and centuthere were very few differences when results were expressed on an areal (volume) basis because of the low bulk density of the organic ries (Mitsch and Gosselink, 1993). Classification probsoils. The areal results suggest that organic and mineral soil calcareous lems have taken on new importance in recent years due wetland sites have similar N water quality maintenance values, that to intense interest in functional evaluation and classifiis, the ability to absorb N from upland land areas. Heavily grazed cation of wetlands (i.e., their effects on water quality, sites had significantly decreased pH and increased NO2 3 levels relative atmospheric chemistry and biodiversity) (Ehrenfeld, to undisturbed sites, but the differences were small. The lack of strong 1986; Johnston, 1991; Brinson, 1993; Ainslie, 1994; Beddifferences in N cycle variables between mineral and organic soil sites ford, 1996). raises questions about the need to make a classification distinction Calcareous wetlands have proven to be particularly between calcareous peatlands (fens) and calcareous mineral soil wetchallenging to classify and evaluate in a functional conlands in nutrient cycling and water quality maintenance contexts. text. These wetlands have developed on limestone or other Ca-rich parent materials, are characterized and D. Van Hoewyk and P.M. Groffman, Institute of Ecosystem Studies, classified by the presence of calcicole vegetation (i.e., Box AB, Millbrook, NY 12545; E. Kiviat, G. Mihocko, and G. Stevens, Hudsonia, Ltd., Bard College Field Station, Annandale, NY 12504. plants that normally grow on soils high in Ca), and lack Received 15 Feb. 2000. *Corresponding author (groffmanp@ a closed tree canopy (Boyer and Wheeler, 1989; Johnson ecostudies.org). and Leopold, 1994; Motzkin, 1994). Many investigators make a distinction between calcareous peatlands (fens) Published in Soil Sci. Soc. Am. J. 64:2168–2173 (2000). VAN HOEWYK ET AL.: NITROGEN IN ORGANIC AND MINERAL SOIL CALCAREOUS WETLANDS 2169 30-yr mean annual temperature at Poughkeepsie, ≈10 to 50 and calcareous wetlands on mineral soil substrates. km from our study sites, is 9.58C (Thomas, 1985). Average However, there have been few evaluations of functional monthly temperatures in the Hudson Valley region of New differences between calcareous wetlands with mineral York State in June, July, and August 1997 were 19.5, 21.2, and vs. organic soils. Functional evaluation of calcareous 20.28C respectively, all within 0.58C of normal. Precipitation in wetlands is needed because in many areas, they support June through August 1997 was 26.3 cm, which was 3.3 cm a variety of rare plant and animal species and are of below the 30-yr average (Northeast Regional Climate Cengreat interest in conservation biology (Kiviat et al., 1993; ter, 1997). Johnson and Leopold, 1994; Almendinger and Leete, Potential calcareous wetland sites were first identified on 1998). Conservation of these ecosystems is problematic soil maps (Case, 1989, unpublished soil survey of Dutchess County, New York) as areas that were “somewhat poorly because they are often intensively used for grazing or drained” or wetter with a surface or subsoil pH .7.0. Aerial are surrounded by heavily fertilized agricultural or resiphotos and air and roadside surveys were then used to elimidential areas (Koerselman et al., 1990). Nutrients from nate areas that clearly did not support fen plant communities. adjacent land uses can disturb the unique biogeochemisFinal sites were selected to have ,50% coverage of 1-m-tall, try of calcareous wetlands, resulting in changes in plant woody vegetation and the presence of fen indicator plants (at and animal populations (Verhoeven and Schmitz, 1991; least four fen species per m or 10% coverage by fen species). Boeye et al., 1995; Kooijman and Bakker, 1995). The sites had to be large enough to encompass three 5 by 5 m In a previous study (Groffman et al., 1996), we evalusampling plots separated by at least 2.5 m. ated microbial biomass and activity in calcareous wetDetailed characterization of plant communities at these lands and three other wetland types—red maple (Acer sites is presented elsewhere (Kiviat et al., unpublished data). The calcareous wetlands in this study had open, low vegetation rubrum L.) swamps, clay meadows, and woodland characterized by shrubby cinquefoil (Potentilla fruticosa auct. pools—in eastern New York. The objective of the previnon L.) and low, narrow-leaved sedges (Carex spp.). These ous study was to determine if different wetland types sites belong to the Palustrine Emergent Persistent and Palushave distinctive patterns of microbial activities related trine Scrub-Shrub Broad-leaved Deciduous categories of to nutrient cycling and water quality maintenance funcCowardin et al. (1979). Soils at the sites were mapped as Sun tions (e.g., the ability to absorb pollutants, especially (coarse-loamy, mixed, active, nonacid, mesic Aeric EpiaNO 3 , from upland areas). In that study, the calcareous quepts), Carlisle (euic, mesic Typic Haplosaprist), Limerick wetlands were the most problematic of the four wetland (coarse-silty, mixed, nonacid, mesic Typic Fluvaquents), Lintypes evaluated. Variation between calcareous wetland lithgo (fine-loamy over sandy or sandy-skeletal, mixed nonacid, sites was extreme, making it difficult to determine mesic Aeric fluvaquents), Alden (fine-loamy, mixed, active, nonacid, mesic Mollic Endoaquepts), or Wayland (fine-silty, whether there were characteristic levels of microbial mixed, active, nonacid, mesic Fluvaquentic Endoaquept) sebiomass and activity in this wetland type relative to the ries. It is interesting to note that none of these soil families has other wetlands. Two of our calcareous wetland sites had a designation as calcareous. However, we identified calcareous mineral soil substrates, while one had an organic soil wetlands in this study by the presence of distinctive vegetation substrate. Although vegetation and water table levels known to be adapted to calcareous conditions. Organic and were similar among the three sites, microbial biomass mineral soil sites were evenly distributed throughout the study and activity were much higher at the site with organic area. Two of the sites had evidence (hoof prints and manure) soil. of recent heavy grazing by horses (Equus caballus) and cattle In the study presented here, we tested the hypothesis (Bos taurus). that microbial N cycling in calcareous wetlands differs All sites were sampled in June and July 1997 (except for bulk density). At each site, the extent of calcareous wetland, significantly with substrate type (organic vs. mineral fen vegetation was delineated as described above and three soil). We tested this hypothesis by measuring potential 5 by 5 m sampling plots were located using a random number net N mineralization and nitrification, denitrification table. Sites ranged in size from 0.01 to 1.29 ha. A 5.0 cm wide enzyme activity, and soil inorganic N levels at 25 sites by 50 cm deep hole was augered in the center of each plot. with different substrate types. We also hypothesized The top 10 cm of soil from each hole was taken for soil analysis that N cycling differs significantly in calcareous wetlands and transported to the laboratory on ice. After water table that are heavily grazed by livestock. We tested this hylevels in the holes had stabilized (required several hours at pothesis by comparing N cycle variables at two sites some sites), the depth to water table was measured and a with heavy grazing activity with three ungrazed sites water sample was taken for analysis. Conductivity and pH of with very similar water table and soil organic matter the water samples were measured immediately using a portalevels. The work was associated with an effort to evaluble conductivity meter and pH probe, respectively. Four sites (two organic, two mineral) were sampled for bulk ate plant and animal biodiversity functions of calcareous density in fall 1998. Six intact soil cores (5-cm diam., 10-cm wetlands and to address fundamental questions about depth) were taken at each site, returned to the laboratory, functional differences between calcareous wetlands on dried, and weighed. These sites were chosen by convenience mineral and organic soils. (ease of access) and a subjective assessment of mineral vs. organic soil conditions. Soils at the four sites ranged from 12 MATERIALS AND METHODS to 43% organic matter. Rock fragments were not present in these soils. The 25 calcareous wetlands were located in Dutchess and For microbial analyses, soils were sieved (,4 mm), homogeColumbia Counties in eastern New York State and in Litchnized by hand mixing, and held at field moisture at 48C in field County, in northwestern Connecticut. Mean annual presealed plastic bags between sampling and analysis (,7 d). Soil cipitation averaged across five weather stations in Dutchess County is ≈104 cm, 6.6 to 10.2 cm mo2 (Thomas, 1985). The moisture content was determined by drying at 608C for 48 h 2170 SOIL SCI. SOC. AM. J., VOL. 64, NOVEMBER–DECEMBER 2000 Table 2. Soil (0–10 cm) variables on a volumetric basis in organic Table 1. Soil (0–10 cm) variables on a weight basis in organic and mineral soil calcareous wetland sites sampled in summer 1997. and mineral soil calcareous wetland sites sampled in summer 1997. Values are mean (standard error) of 12 organic and 13 Values are mean (standard error) of 12 organic and 13 mineral soil sites, except for bulk density values, which are mean (stanmineral soil sites. dard error) of two organic and two mineral soil sites. Variable Organic Mineral Variable Organic Mineral Soil organic matter, kg m22 16.4 (11.2)† 10.7 (6.8) NH4–N, kg ha21 5.9 (0.7)† 3.4 (0.5) Organic matter, % (w/w) 57 (6)† 18 (1) NO3–N, kg ha21 0.41 (0.15) 0.52 (0.14) Bulk density, g cm23 0.26 (0.05)† 0.65 (0.05) Denitrification enzyme activity, Water table depth, cm 8.2 (2.7)† 21.1 (4.8) mg N m22 h21 44 (10) 40 (6) Conductivity, mS 480 (30)† 580 (20) Potential net N mineralization, pH 7.0 (0.04)‡ 7.2 (0.06) kg N ha21 d21 23.2 (0.7)† 20.7 (0.6) Moisture content, % (w/w) 78 (2)† 54 (2) Potential net nitrification, kg N NO3, mg N kg21 1.3 (0.4) 0.8 (0.2) ha21 d21 0.7 (0.3) 0.6 (0.2) NH4, mg N kg21 20.4 (2.8)† 5.8 (0.9) Denitrification enzyme activity, † Indicates significant difference between organic and mineral sites in a mg N kg21 h21 1519 (376)‡ 711 (130) one-way analysis of variance at P , 0.01. Potential net N mineralization, mg N kg21 d21 211.2 (2.5)† 21.4 (1.0) Potential net nitrification, mg N Given large differences in bulk density, it is important kg21 d21 2.3 (1.0) 0.9 (0.4) to compare differences between mineral and organic † Indicates significant difference between organic and mineral sites in a soil sites on both a weight and area (volume) basis. On one-way analysis of variance at P , 0.01. a weight basis, most N cycle variables measured (KCl ‡ Indicates significant difference between organic and mineral sites in a one-way analysis of variance at P , 0.05. extractable NH 4 , denitrification enzyme activity, and potential net N mineralization) were higher (P , 0.01 or 0.05) at organic soil sites than at mineral soil sites (McInnes et al., 1994). Soil organic matter content was determined by loss on ignition at 4508C for 4 h (Nelson and Som(Table 1). However, on a volume basis, several of these mers, 1996). Mean bulk density values for mineral and organic differences were not present (Table 2), and only soil soils were used to convert data on organic matter, inorganic organic matter, NH 4 , and potential net N mineralization N, denitrification enzyme activity, and potential net N mineraldiffered among organic and mineral soil sites on a voluization and nitrification from a weight to a volume basis. metric basis. There were no correlations between N Denitrification enzyme activity was measured using the cycle variables and soil organic matter content on a short-term anaerobic assay described by Smith and Tiedje volumetric basis, but there were strong correlations (1979). Sieved soils were amended with NO 3 , dextrose, chloramong the N cycle variables themselves (Table 3). Poamphenicol, and acetylene, and were incubated under anaerotential net N mineralization was positively correlated bic conditions for 90 min. Gas samples were taken at 30 and 90 min, stored in evacuated glass tubes, and analyzed for N2O with water table depth, and negatively correlated with by electron capture gas chromatography. NH 4 . Nitrification was negatively correlated with pH. Amounts of inorganic N (NO 3 and NH 4 ) in soil were deterA comparison of the two disturbed (heavily grazed) mined by extraction with 2 M KCl followed by colorimetric sites with the three undisturbed sites with the most analysis with a Perstorp Flow Solution Analyzer. Potential closely matched organic matter contents and water table net N mineralization and nitrification were measured from levels found lower pH (P , 0.05) and higher NO 3 (P , the accumulation of NO 3 plus NH 4 and NO 3 alone during 0.10) in the disturbed relative to the undisturbed sites a 7-d aerobic incubation at room temperature. (Table 4). Differences between mineral and organic (.30% organic matter) and between disturbed (evidence of heavy grazing) and undisturbed site samples were analyzed by one-way analyDISCUSSION sis of variance of site means using the General Linear Models (for normally distributed data) or NPAR1WAY (for nonnorAre There Functional Differences between mally distributed data) routines of the Statistical Analysis Organic and Mineral Soil Calcareous Wetlands? System (SAS Institute, 1988). Relationships among variables Differences in inorganic N availability and denitrificawere analyzed by computing Pearson (linear) and Spearman tion between organic and mineral soil sites were much (nonparametric) correlation coefficients (SAS Institute, 1988). more marked on a weight than volumetric basis. The We report the higher of the two correlation coefficients because we were interested more in the strength of the relationweight basis results are consistent with our previous ships between variables than in their linearity. study (Groffman et al., 1996) that found very high levels of microbial biomass and activity in an organic soil calcareous wetland compared with two mineral soil sites RESULTS and with many other studies that have shown that soil biological activity increases with organic matter content Using a breakpoint of 30% organic matter, our sampling design produced 13 mineral and 12 organic soil (Paul and Clark, 1996). However, when considered on a volumetric basis, the results suggest that there are sites. Two of the mineral soil sites were disturbed; that is, they were heavily grazed, with obvious manure input. very few differences in inorganic N availability and denitrification between organic and mineral soil calcareous The organic soil sites had higher (P , 0.01) water tables and water content and lower bulk density (P , wetlands sites in our region. It is important to note that our analysis was restricted to the top 10 cm of the soil 0.01) and porewater pH (P , 0.05) than the mineral soil sites (Table 1). Conductivity did not differ between profile. However, biological activities are generally highest near the soil surface (Paul and Clark, 1996), sites with mineral or organic soils. VAN HOEWYK ET AL.: NITROGEN IN ORGANIC AND MINERAL SOIL CALCAREOUS WETLANDS 2171 Table 4. Soil (0–10 cm) variables (mean and standard error) on Table 3. Significant correlations between volumetric estimates of potential net N mineralization, net nitrification, denitrification a weight basis in two heavily grazed and three undisturbed mineral soil calcareous wetland sites sampled in summer 1997. enzyme activity and other variables in 25 organic and mineral soil (0–10 cm) calcareous wetlands sampled in summer 1997. The undisturbed sites were closely matched to the grazed sites by water table and soil organic matter data. Values are Pearson product moment or Spearman nonparametric correlation coefficients, whichever was higher. Variable Grazed Undisturbed Potential net N Potential net Denitrification Organic matter, % (w/w) 0.17 (0.02) 0.18 (0.02) mineralization nitrification enzyme activity Water table depth, cm 7.4 (3.0) 8.7 (2.0) Conductivity, mS 610 (10) 600 (20) NH4–N (20.75)*** DEA (0.48)** Nitrification pH 6.9 (0.02)† 7.2 (0.07) (0.48)** NO3, mg N kg21 1.4 (0.5)‡ 0.2 (0.1) Water table depth (0.55)** Mineralization (0.48)** Mineralization NH4, mg N kg21 7.5 (0.3) 5.4 (1.7) (0.37)* Denitrification enzyme activity, DEA† (0.37)* pH (20.43)* mg N kg21 h21 1020 (95) 679 (226) * Significant at the 0.05 probability level. Potential net N mineralization, ** Significant at the 0.01 probability level. mg N kg21 d21 22.0 (1.0) 1.4 (2.6) ** Significant at the 0.001 probability level. Potential net nitrification, mg N † DEA is denitrification enzyme activity. kg21 d21 1.6 (0.8) 0.9 (0.4) † Indicates significant difference between grazed and undisturbed sites in a one-way analysis of variance at P , 0.05. so our sampling was probably representative of a high ‡ Indicates significant difference between grazed and undisturbed sites in percentage of the N cycling activity of the soil profile. a one-way analysis of variance at P , 0.10. Moreover, differences between organic and mineral soil sites should be most marked at the top of the soil profile community composition between our mineral and oras soil organic matter levels generally decline sharply ganic soil sites (Kiviat et al., 1998, unpublished data). with depth in most soil profiles (Paul and Clark, 1996). Investigators in other regions have also noted a lack The equal denitrification enzyme activity and the fact of marked difference in plant community composition that both wetland types exhibited net immobilization between organic and mineral soil calcareous wetland (i.e., negative net mineralization) suggest that organic sites (Tyler, 1984; Motzkin, 1994; Nekola, 1994). Previand mineral soil wetland sites have similar water quality ous studies in eastern New York have found no differmaintenance values, at least as far as N is concerned. ences in bog turtle (Clemmys muhlenbergii) habitat Denitrification is considered to be an important sink value of mineral and organic soil sites (Kiviat, 1978). for NO 3 , a drinking water pollutant and agent of marine Even though they do not have the organic soil that many eutrophication (Howarth et al., 1996). Nitrate is the would consider to be a requirement for classification as most commonly detected groundwater pollutant in the a fen (Bridgham et al., 1996), the fact that our mineral USA and often moves from upland land uses into wetsoil sites have similar denitrification enzyme activity, lands (USEPA, 1990; Johnston, 1991). Net mineralizanet N mineralization, plant community composition, tion is the product of microbial production and conand the ability to support bog turtles as organic soil sumption of inorganic N. The fact that all of our sites sites suggests that they should be in the same functional had negative net mineralization suggests that there is class as the organic soil sites in nutrient cycling, water potential for microbial absorption of exogenous N in quality maintenance, and rare plant and wildlife habitat these wetlands. contexts in our region. Some caveats to this conclusion It is interesting to note that the organic soil sites had include the fact that our work did not consider all possihigher levels of soil NH 4 , but lower rates of potential ble wetland functions and that we did not consider soil net N mineralization, than the mineral soil sites on both depth and its effect on hydrologic flow paths and rates. a weight and a volumetric basis. These results suggest These topics will be addressed in future research. that there may be subtle differences in N cycling between organic and mineral soil sites. The relatively high Regulation of Nitrogen Cycling NH 4 levels in the organic soil sites suggest that plant in Calcareous Wetlands available N may be higher in organic soil than mineral soil wetland sites. The high rates of potential net immoThe strong correlations among the N cycle variables bilization in the organic soil sites suggest that these (mineralization, nitrification, and mineralization) and sites may be less susceptible to nutrient loss following the lack of correlation between these variables and soil disturbance. Our potential net mineralization–immobiorganic matter or spatial variation in water table depth lization assay was done with sieved (i.e., disturbed) samsuggest, in a preliminary way, that the inherent ability ples. This disturbance probably increased C availability of the soil to supply inorganic N to plants and microbes and stimulated immobilization more in the organic soil is a stronger controller of N cycling among sites than is sites than the mineral soil sites. hydrology or wetland soil type (organic vs. mineral). The lack of strong functional difference with regard Mineralization was positively correlated with water tato N dynamics between mineral and organic soil sites ble depth, which is a logical result as the aerobic condiraises questions about the need to make a distinction tions associated with low water tables are known to between calcareous peatlands (fens) and calcareous foster mineralization (Williams, 1974; Humphrey and mineral soil wetlands in several specific contexts. It is Pluth, 1997). However, none of the other N cycle variinteresting to note that there were no differences in ables were correlated with water table depth, suggesting that spatial variation in hydrology was not a strong conporewater conductivity and few differences in plant 2172SOIL SCI. SOC. AM. J., VOL. 64, NOVEMBER–DECEMBER 2000 troller of N cycling among these sites. It is important other than N input from manure. Grazing has beenfound to have multiple effects on N dynamics and plantto note that we only measured water table depths atone date; therefore, these results must be evaluated with community composition (Holland and Detling, 1990;Shariff et al., 1994; Pastor et al., 1993; Frank and Groff-caution. However, while water tables are highly dynamicwithin and between sites and seasons, these data suggest man, 1998). Grazing may be critical to the maintenanceof the herbaceous vegetation of calcareous wetlands inthat variables other than water table should be consid-ered as strong controllers of N dynamics in calcareous humid regions like Eastern New York (Kiviat et al.,unpublished data). Denitrification response to grazingwetlands in our region.While calcareous wetlands are characterized by an may be important to controlling N availability and plantcompetition in these wetlands, but a full evaluation ofabundance of bases, N and P status can vary widely (vanWirdum, 1993; Motzkin, 1994). Total and available N its importance will require studies where grazing and Ninputs are manipulated in a controlled way.status has been shown to be influenced by external in-puts, grazing (see below) and harvesting (Koerselmanet al., 1990, 1993; Verhoeven and Schmitz, 1991; BoeyeCONCLUSIONSand Verheyen, 1994). It is important to note that vegeta-Although KCl extractable NH4 , denitrification en-tion in these wetlands is more likely to be limited by Pzyme activity, and potential net N mineralization wereand porewater conductivity rather than N (Kooijmanhigher at organic soil sites than at mineral soil sites onand Bakker, 1995; Bootsma and Wassen, 1996; Boeyea weight basis, the lower bulk density of organic soilset al., 1995, 1997; Verhoeven et al., 1996). Therefore,minimized most of these differences on a volumetricregulation of N cycling may not be under tight biologicalbasis.control in these sites (Vitousek et al., 1982).The lack of difference in observed values for N cyclevariables between organic and mineral soil sites on aResponse to Heavy Grazingvolumetric basis suggests that their N water qualityA critical issue in wetland ecology is how wetlands maintenance function (i.e., their ability to absorb exo-respond to nutrient inputs from the surrounding landgeneous N) on an areal basis may be similar. This simi-scape. This issue is of particular interest in calcareous larity should be considered in functional classificationwetlands that have unique vegetation with distinct nutriand comparison of these wetlands in water qualityent requirements and that have been shown to be sensicontexts.tive to nutrient inputs from surrounding areas (KoerselHeavy grazing did not induce marked differences inman et al., 1990; Bridgham and Richardson, 1993; inorganic N availability, mineralization, nitrification, orJohnson and Leopold, 1994; Boeye et al., 1995, 1997; denitrification in these wetlands.Verhoeven et al., 1996). Our comparison of heavilygrazed and undisturbed mineral soil sites suggests theseACKNOWLEDGMENTSwetlands have some capacity for a resilient responseThe authors thank Alan Lorefice and Sibylle Otto for helpto N inputs; specific processes increase to process andwith laboratory analyses. This research was funded by grantsremove the inputs. The decreased pH and increased from the U.S. Environmental Protection Agency (GrantNO3 levels in the grazed sites suggests that manure X992522-01-0), the National Science Foundation (Researchinputs from grazing have been nitrified (Schlesinger, Experiences for Undergraduates), and the Andrew W. Mellon1991). However, the lack of difference in extractable Foundation. Although the research described in this articleNH4 and the low levels of NO3 suggest that an increase has been funded in part by the U.S. Environmental ProtectionAgency, it has not been subjected to the Agency’s requiredin denitrification in the grazed relative to the undis-peer and policy review and therefore does not necessarilyturbed sites may have acted as a negative feedback re-reflect the views of the Agency, and no official endorsementsponse to N inputs, reducing the amount of extra Nshould be inferred.available to the plant community (Table 4). An increasein denitrification in response to N inputs has been ob-REFERENCESserved in several wetland studies (Bowden, 1987; Brod-erick et al., 1988; Warwick and Hill, 1988; Schipper et Ainslie, W.G. 1994. Rapid wetland functional assessment: Its role and utility in the regulatory arena. Water Air Soil Pollut. 77:433–444.al., 1991; Ambus and Christensen, 1993; Hanson et al.,Almedinger, J.E., and J.H. Leete. 1998. Regional and local hydrogeol-1994; Lowrance et al., 1995). However, the ultimateogy of calcareous fens inthe Minnesota River basin. Wetlandsimportance of this denitrification response to preserving18:184–302.the unique vegetation and wildlife of calcareous wetAmbus, P., and S. Christensen. 1993. Denitrification variability and control in a riparian fen irrigated with agricultural drainage water.lands is unclear. As stated above, fen vegetation maySoil Biol. Biochem. 25:915–923.be more limited by P and porewater conductivity thanBedford, B.L. 1996. The need to define hydrologic equivalence at theN, and N inputs may therefore not induce major changeslandscape scale for freshwater wetland mitigation. Ecol. Appl.in the plant community. Manure P inputs may be partic-6:57–68. Boeye, D., D. van Straaten, and R.F. Verheyen. 1995. A recent trans-ularly important. Still, our data suggest that grazing mayformation from poor to rich fen caused by artificial groundwaternot cause significant changes in N cycle processes inrecharge. J. Hydrol. 169:111–129.these wetlands.Boeye, D., and R.F. Verheyen. 1994. The relation between vegetationOur comparison of disturbed and undisturbed calcar-and soil chemistry gradients in a ground water discharge fen. J. Veg. Sci. 5:553–560.eous wetlands is also complicated by effects of grazing VAN HOEWYK ET AL.: NITROGEN IN ORGANIC AND MINERAL SOIL CALCAREOUS WETLANDS2173 Boeye, D., B. Verhagen, V. van Haesebroeck, and R.F. Verheyen. Kooijman, A.M., and C. Bakker. 1995. Species replacement in the bryophyte layer in mires: the role of water type, nutrient supply1997. Nutrient limitation in species-rich lowland fens. J. Veg. Sci. 8:415–424.and interspecific interactions. J. Ecol. 83:1–8. Lowrance, R., G. Vellidis, and R.K. Hubbard. 1995. DenitrificationBootsma, W.C., and M.J. Wassen. 1996. Environmental conditions and fen vegetation in three lowland mires. Vegetatio 127:173–189.in a restored riparian forest wetland. J. Environ. Qual. 24:808–815. McInnes, K.J., R.W. Weaver, and M.J. Savage. 1994. Soil water poten-Bowden, W.B. 1987. The biogeochemistry of nitrogen in freshwater wetlands. Biogeochem. 4:313–348.tial. p. 53–58. In R.W. Weaver (ed.) Methods of soil analysis. Part 2. SSSA Book Ser. 5. SSSA, Madison, WI.Boyer, M.L.H., and B.D. Wheeler. 1989. Vegetation patterns in springfed calcareous fens: Calcite precipitation and constraints on fertilMitsch, W.J., and J.G. Gosselink. 1993. Wetlands. Van Nostrand Reinhold, New York.ity. J. Ecol. 77:597–609. Bridgham, S.D., and C.J. Richardson. 1993. Hydrology and nutrient Motzkin, G. 1994. Calcareous fens of western New England and adjacent New York state. Rhodora 96:44–68.gradients in North Carolina peatlands. Wetlands 13:207–218. Bridgham, S.D., J. Pastor, J.A. Janssens, C. Chapin, and T.J. Malterer. National Research Council. 1995. Wetlands: Characteristics and boundaries. National Academy Press, Washington, DC.1996. Multiple limiting gradients in peatlands: A call for a new paradigm. Wetlands 16:45–65.Nekola, J.C. 1994. The environment and vascular flora of northeastern Iowa fen communities. Rhodora 96:121–169.Brinson, M.M. 1993. Changes in the functioning of wetlands along environmental gradients. Wetlands 13:65–74.Nelson, D.W., and L.E. Sommers. 1996. Total carbon, organic carbon, and organic matter. p. 961–1010. In D.L. Sparks (ed.) Methods ofBroderick, S.J., P. Cullen, and W. Maher. 1988. Denitrification in a natural wetland receiving secondary treated effluent. Wat. Res.soil analysis. Part 3. SSSA Book Ser. 5. SSSA, Madison, WI. Northeast Regional Climate Center. 1997. New York climate. Vol.4:431–439. Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classifi-97, no. 6–8. Cornell Univ., Ithaca, NY. Pastor, J., B. Dewey, R.J. Naiman, P.F. MacInnes, and Y. Cohen.cation of wetland and deepwater habitats of the United States.United States Fish and Wildlife Service FWS/OBS-79/31, Washing-1993. Moose browsing and soil fertility in the boreal forests of IsleRoyale National Park. Ecology 74:467–480.ton, DC.Ehrenfeld, J.G. 1986. Wetlands of the New Jersey pine barrens: The Paul, E.A., and F.C. Clark. 1996. Soil microbiology and biochemistry.2nd ed. Academic Press, New York.role of species composition in community function. Am. Midl.Naturalist 115:301–313.SAS Institute. 1988. SAS/STAT user’s guide. Release 6.03. SAS Inst.,Cary, NC.Frank, D.A., and P.M. Groffman. 1998. Ungulate versus landscapecontrol of soil C and N processes in grasslands of Yellowstone Schipper, L.A., A.B. Cooper, and W.J. Dyck. 1991. Mitigating non-point-source nitrate pollution by riparian-zone denitrification. p.National Park. Ecology 79:2229–2241.Groffman, P.M., G.C. Hanson, E. Kiviat, and G. Stevens. 1996. Varia-401–413. In I. Bogárdi et al. (ed.) Nitrate contamination: Exposure,consequences, and control. Springer-Verlag, Berlin.tion in microbial parameters in four different wetland types. SoilSci. Soc. Am. J. 60:622–629.Schlesinger, W.H. 1991. Biogeochemistry: An analysis of globalchange. Academic Press, San Diego, CA.Hanson, G.C., P.M. Groffman, and A.J. Gold. 1994. Denitrificationin riparian wetlands receiving high and low groundwater nitrate Shariff, A.R., M.E. Biondini, and C.E. Grygiel. 1994. Grazing intensityeffects on litter decomposition and soil nitrogen mineralization. J.inputs. J. Environ. Qual. 23:917–922.Holland, E.A., and J.K. Detling. 1990. Plant response to herbivoryRange Manage. 47:444–449.Smith, M.S., and J.M. Tiedje. 1979. Phases of denitrification followingand belowground nitrogen cycling. Ecology 71:1040–1049.Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski,oxygen depletion in soil. Soil Biol. Biochem. 11:262–267.Thomas, H. (ed.) 1985. Dutchess County, New York natural resources.K. Lajtha, J.A. Downing, R. Elmgren, N, Caraco, T. Jordan, F.Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and Z.L. Zhu.Dutchess County Department of Planning, Poughkeepsie, NY.Tyler, C. 1984. Calcareous fens in south Sweden. Previous use, effects1996. Regional nitrogen budgets and riverine N & P fluxes forthe drainages to the North Atlantic Ocean: Natural and humanof management and management recommendations. Biol. Con-serv. 30:69–89.influences. Biogeochemistry 35:75–139.Humphrey, W.D., and D.J. Pluth. 1996. Net nitrogen mineralization USEPA. 1990. National pesticide survey: Nitrate. Office of Water,Office of Pesticides and Toxic Substances, Washington, DC.in natural and drained fen peatlands in Alberta, Canada. Soil Sci.Soc. Am. J. 60:932–940.Van Wirdum, G. 1993. An ecosystems approach to base-rich freshwa-ter wetlands, with special reference to fenlands. HydrobiologyJohnson, A.M., and D.J. Leopold. 1994. Vascular plant species rich-ness and rarity across a minerotrophic gradient in wetlands of St.254:129–153.Verhoeven, J.T.A., W. Koerselman, and A.F.M. Meuleman. 1996.Lawrence County, New York, USA. Biodiversity Conserv. 3:606–627.Nitrogenor phosphorus-limited growth in herbaceous, wet vegetation: relations with atmospheric inputs and management regimes.Johnston, C.A. 1991. Sediment and nutrient retention by freshwater wetlands: Effects on surface water quality. Crit. Rev. Environ.TREE 11:494–497. Verhoeven, J.T.A., and M.B. Schmitz. 1991. Control of plant growthQual. 21:491–565. Kiviat, E. 1978. Bog turtle habitat ecology. Bull. Chicago Herpetologi-by nitrogen and phosphorus in mesotrophic fens. Biogeochemistry 12:135–148.cal Soc. 13:29–42. Kiviat, E., G. Stevens, and S. Barbour. 1993. Blossoms and clay: Vitousek, P.M., J.R. Gosz, C.C. Grier, J.M. Melillo, W.A. Reiners, and R.L. Todd. 1982. A comparative analysis of potential nitrifica-Landfill siting, wetlands, and biodiversity. News Hudsonia 9:1–7. Koerselman, W., S.A. Bakker, and M. Blom. 1990. Nitrogen, phospho-tion and nitrate mobility in forest ecosystems. Ecol. Monogr. 52: 155–177.rus and potassium budgets for two small fens surrounded by heavily fertilized pastures. J. Ecol. 78:428–442.Warwick, J., and A.R. Hill. 1988. Nitrate depletion in the riparian zone of a small woodland stream. Hydrobiology 157:231–240.Koerselman, W., M.B. Van Kerkhoven, and J.T.A. Verhoeven. 1993. Release of inorganic N, P and K in peat soils: Effect of temperature Williams, B.L. 1974. The effect of water table level on nitrogen mineralization in peat. Forestry 47:195–202.water chemistry and water level. Biogeochem. 20:63–81.