Methane Flux in Cropland and Adjacent Riparian Buff ers with Diff erent Vegetation Covers

While water quality functions of conservation buffers established adjacent to cropped fields have been widely documented, the relative contribution of these re-established perennial plant systems to greenhouse gases has not been completely documented. In the case of methane (CH(4)), these systems have the potential to serve as sinks of CH(4) or may provide favorable conditions for CH(4) production. This study quantifies CH(4) flux from soils of riparian buffer systems comprised of three vegetation types and compares these fluxes with those of adjacent crop fields. We measured soil properties and diel and seasonal variations of CH(4) flux in 7 to 17 yr-old re-established riparian forest buffers, warm-season and cool-season grass filters, and an adjacent crop field located in the Bear Creek watershed in central Iowa. Forest buffer and grass filter soils had significantly lower bulk density (P < 0.01); and higher pH (P < 0.01), total carbon (TC) (P < 0.01), and total nitrogen (TN) (P < 0.01) than crop field soils. There was no significant relationship between CH(4) flux and soil moisture or soil temperature among sites within the range of conditions observed. Cumulative CH(4) flux was -0.80 kg CH(4)-C ha(-1) yr(-1) in the cropped field, -0.46 kg CH(4)-C ha(-1) yr(-1) within the forest buffers, and 0.04 kg CH(4)-C ha(-1) yr(-1) within grass filters, but difference among vegetation covers was not significant. Results suggest that CH(4) flux was not changed after establishment of perennial vegetation on cropped soils, despite significant changes in soil properties.

While water quality functions of conservation buff ers established adjacent to cropped fi elds have been widely documented, the relative contribution of these re-established perennial plant systems to greenhouse gases has not been completely documented.In the case of methane (CH 4 ), these systems have the potential to serve as sinks of CH 4 or may provide favorable conditions for CH 4 production.Th is study quantifi es CH 4 fl ux from soils of riparian buff er systems comprised of three vegetation types and compares these fl uxes with those of adjacent crop fi elds.We measured soil properties and diel and seasonal variations of CH 4 fl ux in 7 to 17 yr-old re-established riparian forest buff ers, warm-season and cool-season grass fi lters, and an adjacent crop fi eld located in the Bear Creek watershed in central Iowa.Forest buff er and grass fi lter soils had signifi cantly lower bulk density (P < 0.01); and higher pH (P < 0.01), total carbon (TC) (P < 0.01), and total nitrogen (TN) (P < 0.01) than crop fi eld soils.Th ere was no signifi cant relationship between CH 4 fl ux and soil moisture or soil temperature among sites within the range of conditions observed.Cumulative CH 4 fl ux was -0.80 kg CH 4 -C ha -1 yr -1 in the cropped fi eld, -0.46 kg CH 4 -C ha -1 yr -1 within the forest buff ers, and 0.04 kg CH 4 -C ha -1 yr -1 within grass fi lters, but diff erence among vegetation covers was not signifi cant.Results suggest that CH 4 fl ux was not changed after establishment of perennial vegetation on cropped soils, despite signifi cant changes in soil properties.T he global atmospheric concentration of CH 4 has increased from a preindustrial value of about 715 to 1774 μL L -1 in 2005, likely a result of anthropogenic activities such as agricultural production and fossil fuel use (IPCC, 2007).Soils have been shown to both produce and consume CH 4 (Topp and Pattey, 1997;Le Mer and Roger, 2001).In a recent review, Dutaur and Verchot (2007) summarized net CH 4 fl ux as the result of the balance between the two off setting processes of methanogenesis (microbial production under anaerobic conditions) and methanotrophy (microbial consumption).Th ese authors identifi ed methanotrophy as the dominant process in upland soils, where oxidation generally exceeds production with a resulting net uptake of atmospheric CH 4 by soil.It is well known that forest soils are the most active sink of CH 4 , followed by grass lands and cultivated soils, and that the CH 4 uptake potential of many upland soils is reduced by cultivation and application of ammonium N fertilizer (Topp and Pattey, 1997;Le Mer and Roger, 2001;Dutaur and Verchot, 2007).It has been reported that land-use change can also infl uence CH 4 uptake rates.For instance, higher rates of CH 4 oxidation have been observed in soils aff orested from croplands or pastures (e.g., Ball et al., 2002;Merino et al., 2004;Tate et al., 2007).Observed increases in CH 4 uptake resulting from land-use change are attributed to changes in soil porosity, moisture content, and methanotroph population (Priemé et al., 1997).

Methane Flux in Cropland and Adjacent Riparian Buff ers with Diff erent Vegetation Covers
Nonpoint source (NPS) pollutants such as sediment, N, P, and pesticides are major causes of water quality problems around the world (Duda, 1993;Tonderski, 1996;Sabater et al., 2003).Riparian buff ers have been recommended as one of the most eff ective tools for mitigating NPS pollution (Hubbard et al., 2004;Mayer et al., 2007).Some of the important functions of riparian buff ers related to NPS pollution control are fi ltering and retaining sediment and immobilizing, storing, and transforming chemical inputs from uplands (Schultz et al., 2000).Generally, riparian buff ers re-established on cultivated crop fi elds consist of combinations of grasses, forbs, shrubs, and trees (Schultz et al., 2004).While these systems have been well documented for their water quality functions, little is known about other ecosystem processes such as their relative greenhouse gas fl ux.If these systems perform similar to perennial plant systems in upland positions, it would suggest that riparian buff ers re-established on cropped soils may produce less and consume more CH 4 than crop fi elds.However, riparian buff ers are often fl ooded and also sustain relatively high soil moisture conditions caused by high water tables, long residence time and slow discharge (Schultz et al., 2000).Th ese conditions may be favorable for CH 4 production.For example, Ambus and Christensen (1995) reported that CH 4 was produced in temporarily fl ooded riparian areas at rates of 78.8 kg CH 4 -C ha -1 yr -1 .Methane was also produced from riparian areas of ponded depressions in northern Germany at rates of 0.33 to 330.3 kg CH 4 -C ha -1 yr -1 (Merbach et al., 1996).In contrast, rates of CH 4 consumption in temperate regions have been estimated as 1.29 ± 0.16 kg CH 4 ha -1 yr -1 in crop fi elds (n = 48), 5.75 ± 0.59 kg CH 4 ha -1 yr -1 in grasslands (n = 24), and 2.40 ± 0.40 kg CH 4 ha -1 yr -1 in forests (n = 91) (data extracted from Dutaur and Verchot (2007)).Th ese results suggest that reestablished riparian buff ers may produce more CH 4 than crop fi elds and natural lands, at least when they are fl ooded, and the benefi ts of reduced nonpoint-source pollution from riparian buff ers may be off set by increased greenhouse gas emissions.
Numerous studies have emphasized the role of vegetation in soil biogeochemical processes within natural or re-established riparian buff ers, with many studies demonstrating an improvement in soil quality indicators (e.g., Tufekcioglu et al., 1999;Bharati et al., 2002;Marquez et al., 2004).However, most studies have often found confl icting results regarding the eff ect of vegetation type on biogeochemical process regulation.For example, there are uncertainties about the eff ect of vegetation type on groundwater NO 3 -removal or denitrifi cation in riparian buff ers (e.g., Groffman et al., 1991;Schnabel et al., 1996;Hefting et al., 2003).With respect to CH 4 fl ux, several studies have compared rates among vegetation types, (Topp and Pattey, 1997;Le Mer and Roger, 2001;Chan andParkin, 2001a, 2001b;Dutaur and Verchot, 2007).However, few studies of CH 4 fl ux have focused on riparian soils, particularly those re-established to perennial vegetation, or on the relationship between observed changes in soil quality on conditions regulating methane fl ux.Specifi c objectives of this study were to compare CH 4 fl ux from riparian buff er systems comprised of forest, warm-season grasses, and cool-season grasses and an adjacent crop fi eld, and to relate these fl uxes to changes in soil properties after re-establishment of perennial plants.

Study Site
Th e study area consisted of three forest buff ers, three warmseason grass fi lters, one cool-season grass fi lter, and one adjacent crop fi eld located in the Bear Creek watershed, Story County and Hamilton County, Iowa (42° 11´ N, 93° 30´W).Th e Bear Creek watershed (6810 ha) is a predominantly agricultural watershed typical of north central Iowa, with a mean annual air temperature of 8.7°C and mean annual precipitation of 810 mm over the period of record (United States Department of Commerce-National Oceanic and Atmospheric Administration, 2009).Most of the area was originally covered with prairie and wetland vegetation except for riparian forests along higher order streams.Most of the area is now rain-fed agriculture cultivated with soybean [Glycine max (L.) Merr.] and corn (Zea mays L.), which are usually grown in rotation.Re-established forest buff ers, and warm-season and cool-season grass fi lters were previously under row-crop cultivation or livestock grazing.Th e forest buff ers and grass fi lters ranged in age from 7 to 17 yr since re-establishment.

Soil Sampling and Analysis
Six intact soil cores (5.3-cm diam.) were collected to a depth of 15 cm around each of three gas sampling points in a forest buff er, a warm-season grass fi lter, a cool-season grass fi lter, and an adjacent crop fi eld in October 2006 and September 2007.A plastic sleeve liner was placed inside the metal core tube and the liner and intact soil core pulled from the tube and capped for transport to the laboratory.Soil samples were transported back to lab in a cooler and stored at 4°C until analysis.Soil pH was determined by using 1:1 diluted soil solution.Gravimetric moisture content was determined by oven drying a subsample at 105°C for 24 h.Bulk density was estimated using the core method (Grossman and Reinsch, 2002).For C and N analysis, soils were air dried at room temperature, and sieved (2 mm).Total C and TN were measured using a Flash EA 2000 (Th ermoFinnigan, Italy) elemental analyzer.Soil inorganic N was extracted with 2 mol L -1 potassium chloride (KCl) within 4 h of sampling and stored at 4°C until fi ltration (Van Miegroet, 1995).Filtrates were frozen and stored until analysis.Nitrate (NO 3 -) and ammonium (NH 4 + ) contents were analyzed by colorimetric method (Mulvaney, 1996) with an auto analyzer (Quikchem 8000 FIA+, Lachat Instruments, Milwaukee, WI).

Field Gas Sampling, Methane Gas Analysis, and Flux Calculation
Soil CH 4 fl ux from riparian forest buff ers, warm-and coolseason grass fi lters and one crop fi eld was measured from January through December 2007.To assess the temporal variation of fl ux, fi ve locations were randomly selected in each of three forest buff ers, three warm-season grass fi lters, one cool-season grass fi lter, and one crop fi eld with the distance between gas sampling points ranging from 5 to 10 m.A polyvinyl chloride (PVC) ring (30 cm diam.by 15 cm height) served as base for gas chambers and was installed to a depth of approximately 10 cm.In the crop fi eld, rings were placed either between plants within the row or between rows.Th ese rings were left in place between sampling periods, but were removed for fertilization, planting, and tillage events in the crop fi eld.Vegetation inside the rings was cut before gas sampling in the forest buff ers and grass fi lters during growing seasons.Gas samples were collected within static vented chambers (PVC, 30 cm diam.by 15 cm height with a vent) weekly or biweekly during the mid-morning.Chambers were equipped with a thermometer to measure air temperature within the chambers at the time of sampling.Ten milliliters of air was sampled from the chamber with a polypropylene syringe at 15min intervals for 45 min (four samples 0, 15, 30, and 45 min) and the gas stored in pre-evacuated glass vials (6 mL, fi tted with butyl rubber stoppers) until analysis.Glass vials were prepared by alternately evacuating the vial headspace and fl ushing with helium to remove air.In addition to the regular measurements, diel variation of CH 4 fl ux was measured on 16-17 July 2007.For this assessment, three gas sampling points were randomly selected in a forest buff er, a warm-season grass fi lter, a cool-season grass fi lter, and an adjacent crop fi eld and gas samples were collected every 3 h for 24 h.Methane concentrations were determined with a gas chromatograph (GC) (Model GC17A; Shimadzu, Kyoto, Japan) equipped with a fl ame ionization detector (FID) and a stainless steel column (0.3175 cm diam.× 74.54 cm long) with Porapak Q (80-100 mesh).Samples were introduced into the gas chromatograph using an autosampler described by Arnold et al. (2001).Th ree diff erent CH 4 standards (0, 2.0, and 10 μL L -1 ) were used to perform calibration curves and fi eld ambient samples and CH 4 standards were analyzed every 20 gas samples to verify accuracy in GC results.Methane fl uxes were obtained by applying linear regression to the CH 4 concentration vs. time data (Holland et al., 1999).Linearity with R 2 > 0.8 was accepted as a valid fl ux rate, which resulted in the inclusion of 90% of the fl ux rates in this study.Where removing a sample corrected a poor linearity (R 2 < 0.8) to R 2 > 0.9, the sample was eliminated from the calculation of fl ux rate (Altor and Mitsch, 2006).Th erefore, fl ux rate was determined using a minimum of three gas samples.
Th e minimum detectable CH 4 fl ux was calculated using an average of standard deviations of CH 4 concentrations of lab ambient air samples and CH 4 standards (n = 500) analyzed with collected gas samples, chamber volume, chamber footprint, and chamber enclosure time in the fi elds as following: Th e calculated minimum detectable CH 4 fl ux (μL CH 4 -C m -2 h -1 ) converted to mass unit (μg CH 4 -C m -2 h -1 ) through application of the universal gas law (Holland et al., 1999).Our estimated minimum detectable fl ux was 33.2 μg CH 4 -C m -2 h -1 .Some of the fl uxes measured from the individual chambers were smaller than our detection limit.In these situations, we followed the recommendation of Gilbert (1987) and Chan and Parkin (2001a) and included the measured values of these "nondetects" in computing mean fl uxes.Cumulative CH 4 fl uxes from each site over the 1-d study period (16-17 July 2007) and the 1-yr study period (January-December 2007) were calculated by linear interpolation and numerical integration between sampling times.Soil temperature (ST) and soil water content (SWC) were measured simultaneously with CH 4 gas collection around the chamber at a 5 cm depth using a digital thermocouple ( Th ermoWorks, Orem, UT) and a digital soil moisture meter (HydroSense, Campbell Scientifc, Inc., Logan, UT).Air temperature was measured simultaneously with CH 4 gas collection inside and outside the gas chamber.A soil temperature and soil moisture data logger (HOBO Micro station data logger with sensors, Oneset Computer Corporation, Bourne, MA) was installed at 5 cm soil depths around a chamber at each site to measure hourly ST and SWC at each site.Daily rainfall and snow data were provided by a nearby meteorology station (Colo, IA, 42° 1´ N, 93° 19´W).

Soil Incubation with Control and 10 Pa Acetylene (C 2 H 2 )
Aerobic CH 4 production and net CH 4 fl ux were estimated using the intact soil cores (0-15 cm depth) collected in September 2007.Soil samples were transported in a cooler and stored at 4°C until experiments, and incubation experiments with the intact soil cores were conducted within 6 h of sampling.All aboveground vegetation in the soil cores was cut off before the experiments.Six intact soil cores (5.3 cm diam.by 8 cm long) collected at each site were placed into 350-mL glass jars with gas-tight lids containing a gas-sampling port and all jars were sealed.Gravimetric moisture content of each soil was determined by oven drying a subsample at 105°C for 24 h.Th ree soil cores from each site were treated with 10 Pa C 2 H 2 and three were retained as controls (no C 2 H 2 ).Soil cores were incubated at 22°C, the on-site soil temperature.Ten milliliters of air was sampled from the jars with a polypropylene syringe at 3, 9, and 16 h, and stored until analysis.Storage, gas analysis, and fl ux calculations were as described above.Aerobic CH 4 production was estimated from soil incubations in which CH 4 oxidation was inhibited by 10 Pa C 2 H 2 (Chan andParkin, 2000, 2001b)

Statistical Analyses
Th e Shapiro-Wilk normality test was used to assess normality of data.A two-sample t test was used to evaluate diff erences in soil C measured in 1998-1999 and 2006-2007 in the same sites.One-way ANOVA was used to evaluate the diff erences in soil properties, and diel and seasonal CH 4 fl ux by site.When the standard assumptions of normality were violated, nonparametric Kruskal-Wallis one-way ANOVA on ranks was used.Diff erences were considered signifi cant at the P < 0.05 level.To determine the relationship between soil properties and CH 4 fl ux, correlation analysis using the GLM procedure was applied.Statistical analyses were conducted using SAS ver 8.1 (SAS Institute, 1999).

Soil Properties
Th e texture of all treatment site soils (Coland) was loam (Marquez et al., 2004) (Table 1).Soils within all riparian buff er vegetation types had signifi cantly lower bulk density (one-way ANOVA P < 0.01); higher pH (P < 0.01), and NH 4 + (P < 0.01) than crop fi eld soils.TC (P < 0.01) and TN (P < 0.01) within the pooled riparian buff er vegetation soils were signifi cantly higher that crop fi eld soils.However, this diff erence is apparently driven by the forest buff er soils, as indicated in pairwise comparisons.Soil NO 3 -was not signifi cantly diff erent among sites (Table 1).Within the same sites, soil carbon content (0-15 cm soil depth) was 30.4 ± 1.6 g C kg soil -1 (n = 6) in the forest buffer, 24.4 ± 1.0 g kg -1 (n = 6) in the warm-season grass fi lter, and 31.0 ± 1.8 g kg -1 (n = 6) in the cool-season grass fi lter in 1998 and 1999 (J. Raich, unpublished data, 1999).Comparing these data with those of this study (Table 1), soil C in the forest buff er (42.9 ± 3.2 g kg -1 , n = 6) in this study was signifi cantly higher than those in 1998 and 1999 (two sample t test P = 0.006, 95% CI for diff erence of means: 4.5 -20.5 g kg -1 ).

Change of Soil Properties after Re-establishment of Riparian Buff ers
Soils within forest buff ers and grass fi lters had signifi cantly lower bulk density, higher pH, TC, TN, and NH 4 + than those in adjacent crop fi elds.Th is suggests that the re-establishment of the perennially vegetated buff ers changed these properties in soils that were previously under row-crop cultivation.Th is conclusion is corroborated when comparing data collected from the same sites in 1998 and 1999 and indicate a 29% increase in soil C in the forest buff er over the last 9 yr.Decomposition of above and belowground litter, root exudates, and microbial C accumulation may contribute to the observed C increase.Increased soil C resulting from conservation practices such as conversion from crop lands to grasslands or forest has been reported in other studies (Gebhart et al., 1994;Knops and Tilman, 2000;Uri, 2000;Post and Kwon, 2000;Guo and Giff ord, 2002;McLauchlan et al., 2006).Johnson et al. (2005) reported that conversion of previous cropland to grass increased soil organic C by 4.2 ± 4.5 Mg C ha -1 yr -1 after 6-8 yr since establishment in the central United States.
We observed signifi cantly higher soil moisture and lower soil temperature in the soils of riparian buff ers compared to those of the crop fi eld.Th is may be the result of the vegetation within riparian buff ers providing more shade to prevent high temperatures in the summer and the lower soil bulk density and high organic matter of riparian buff ers holding more soil moisture.In contrast, soils in conventionally cultivated crop fi elds are more exposed to direct sunlight, have higher bulk density and lower soil organic matter and tend to hold less soil moisture compared with riparian buff ers soils.

Methane Flux in Riparian Buff ers
Methane fl ux observed within the forest buff ers and grass fi lter soils (-0.5 to 0.9 mg CH 4 -C m -2 d -1 , n = 45-50) is similar to results of studies conducted in other riparian systems with infrequent saturation.McLain and Martens (2006) found the CH 4 sink averaged 26.1 ± 6.3 μg CH 4 m -2 h -1 in the semiarid riparian soils of southeastern Arizona.In a riparian alder stand in southern Estonia, Teiter and Mander (2005) observed an average CH 4 fl ux of 0.1 to 265 μg CH 4 -C m -2 h -1 .However, the CH 4 fl ux in the forest buff ers and grass fi lters soils was lower than those reported in other studies conducted in temporarily submerged areas such as rice (Oryza sativa L.) fi elds, wetlands, or riparian areas with frequent saturation.For example, Ambus and Christensen (1995) found CH 4 produced at rates of 7877 mg CH 4 -C m -2 yr -1 (78.8 kg CH 4 -C ha -1 yr -1 ) in a temporarily fl ooded riparian area in Denmark.Methane was produced from riparian areas in northern Germany at rates of 33 to 33,030 mg CH 4 -C m -2 yr -1 (0.33-330.3 kg CH 4 -C ha -1 yr -1 ) (Merbach et al., 1996).Altor and Mitsch (2006) reported that annual CH 4 fl ux from intermittently fl ooded zones was 13 g CH 4 -C m -2 yr -1 (130 kg CH 4 -C ha -1 yr -1 ) in the midwestern United States.Le Mer and Roger's (2001) review of the literature found that the median of CH 4 emissions were 0.72 kg CH 4 ha -1 d -1 (3 mg CH 4 m -2 h -1 ) in swamps, 0.43 kg CH 4 ha -1 d -1 (1.8 mg CH 4 m -2 h -1 ) in peat lands and 1.0 kg CH 4 ha -1 d -1 (4.2 mg CH 4 m -2 h -1 ) in rice fi elds.Th ese results suggest that riparian zones soils under certain conditions are not major sources of CH 4 compared to wetlands, rice fi elds, or riparian zones with more frequent saturation.In the case of riparian zones in many areas of the midwestern United States, changes in landscape hydrology resulting from the conversion to agriculture have resulted in incised stream channels and lowered riparian water tables, likely altering conditions favorable to CH 4 production.At our sites, some riparian buff ers were easily aff ected by fl ooding caused by snow melting (14 March) and heavy rainfall (26 April) and we conducted gas sampling when we were able to access the sites after fl ooding subsided.However, it is likely that because these conditions were so ephemeral, that observed CH 4 mass fl ux refl ect the hydrologic characteristics of riparian buff ers within this landform region.Interannual variation of greenhouse gas fl ux within this region can be signifi cant, as demonstrated by Chan and Parkin (2001a).Because this study was conducted for only 1 yr (January 2007-December 2007), the interannual variation of CH 4 fl ux cannot be assessed for these sites.However, a main objective of this study was to compare CH 4 fl ux among crop fi elds and adjacent riparian buff ers riparian buff ers re-established for water quality.Since all sites were in close proximity and experienced similar conditions, it can be assumed that annual (or interannual) climate variability did not aff ect study conclusions.Since climatic conditions in 2007 (mean air temperature 9.4°C; annual precipitation 1097 mm) were within the standard deviation of conditions over the last 37 yr of record (mean air temperature 8.7 ± 0.8°C; mean annual precipitation 914 ± 210 mm) results from this study could be considered representative of fl ux rates measured over multiple years.
Results of soil incubation experiments, and diel and seasonal CH 4 fl ux measurements indicate that CH 4 fl ux in the crop fi eld, forest buff ers, and grass fi lters were not signifi cantly diff erent from one another.In contrast, in the same region (central Iowa), Chan and Parkin (2001a) found that forest and prairie soils were net CH 4 consumers, with cumulative CH 4 fl uxes ranging from -0.27 to -0.07 g CH 4 m -2 (-2.7 to -0.7 kg CH 4 ha -1 ) over the 258-d sampling season, while agricultural sites were net CH 4 producers, with cumulative CH 4 fl uxes ranging from -0.02 to 3.19 g CH 4 m -2 (-0.2 to 31.9 kg CH 4 ha -1 ) over the same season.Th e prairie and forest soils were found to have the greatest potential to oxidize atmospheric concentrations of CH 4 (Chan and Parkin, 2001b).Within temperate regions globally, reported CH 4 consumption rates include 1.29 ± 0.16 kg CH 4 ha -1 yr -1 in crop fi elds (n = 48), 5.75 ± 0.59 kg CH 4 ha -1 yr -1 within grasslands (n = 24), and 2.40 ± 0.40 kg CH 4 ha -1 yr -1 within forests (n = 91) [data extracted from Dutaur and Verchot (2007)].Th ese reports indicate that CH 4 consumption within re-established riparian forest buff er and grass fi lter soils examined in this study were much lower than other reported CH 4 consumption rates within grasslands and forests in Iowa and the temperate regions.Such a contrast suggests that CH 4 soil oxidation capacity has not been improved during the 7 to 17 yr following re-establishment of perennial vegetation (forest buff ers and grass fi lters) on conventional crop fi elds, even when soil properties such as soil bulk density pH, TC, and soil moisture have changed signifi cantly.It is well known that CH 4 oxidation potential of upland soils is reduced by cultivation and ammonium N-fertilizer application (e.g., Topp and Pattey, 1997;Le Mer and Roger, 2001;Dutaur and Verchot, 2007).Le Mer and Roger (2001) summarized the eff ects of cultural practices on CH 4 oxidation as following: (i) an increase in NH 4 + content of soil by fertilizer application inhibits CH 4 oxidation because NH 4 + produces competition at the level of methane-mono-oxygenase, a transfer of the CH 4 oxidizing activity toward nitrifi cation (Castro et al., 1994;Nesbit and Breitenbeck,1992), and (ii) cultural practices that destroy micro-aerophilic niches suitable for CH 4 oxidizers reduce CH 4 oxidation (Hütsch et al., 1994;Sitaula et al., 2000).Slow recovery of CH 4 oxidation after land use change has been reported.In a range of successional sites on former arable land in Denmark and Scotland, CH 4 oxidation rates took more than 100 yr to reach precultivation levels (Priemé et al., 1997).Similarly, Suwanwaree and Robertson (2005) observed that rates of CH 4 oxidation in soils of 40 to 60 yr-old successional fi elds were be-tween those of the no-till and deciduous forest sites in southwest United States.Singh et al. (2007) reported that aff orestation and reforestation of pastures (30-50 yr later) resulted in changes in methane oxidation by altering the community structure of methanotrophic bacteria in these soils.In the case of the re-established riparian buff ers investigated in this study, it appears that, while soil properties have been altered, additional time is needed for changes in CH 4 fl ux to be manifested.

Conclusions
Soil properties such as soil bulk density, pH, TC, and soil moisture in riparian forest buff er and grass fi lter soils were signifi cantly diff erent from those in adjacent crop fi elds, suggesting that soil properties have changed since re-establishment of perennial vegetation on previously cultivated crop fi eld soils.Soil incubation experiments provide some indication that CH 4 consumption was higher than CH 4 production in forest buff ers and grass fi lters soils, while crop fi eld soils showed the opposite response.However, none of the CH 4 fl uxes from incubation experiments were signifi cantly diff erent, nor were diel and seasonal variation of CH 4 fl uxes in forest buff ers, grass fi lters, and adjacent crop fi elds.Th e cumulative CH 4 fl ux -0.80 kg CH 4 -C ha -1 yr -1 in the crop fi eld, -0.46 kg CH 4 -C ha -1 yr -1 in forest buff ers, and 0.04 kg CH 4 -C ha -1 yr -1 in grass were also not signifi cantly diff erent.Th e CH 4 fl ux in forest buff ers and grass fi lter soils was less than that reported for wetlands, rice fi elds, or riparian areas with more frequent saturation, which are known to be sources of CH 4 .Th e CH 4 fl ux rates reported here are also greater than those reported for forests and grasslands, which are known to be sinks of CH 4 .Th ese results suggest that these re-established riparian forest buff ers and grass fi lters, possibly due to altered hydrology, cannot be considered as major sources of CH 4 as has been found in other riparian areas or systems with more frequent saturation.However, any potential benefi t as increased sinks of CH 4 , as has been found for other perennial plant systems within the region, has not yet been achieved after 7 to 17 yr since re-establishment.Th ese results have important management implications given the signifi cant eff ort to promote such systems for water quality improvement and other ecosystem services.
Dong-Gill Kim and Thomas M. Isenhart* Iowa State University Timothy B. Parkin USDA-ARS Richard C. Schultz and Thomas E. Loynachan Iowa State University detectable CH flux( L CH -C m h ) 2 average of standard deviation( L L ) chamber volume (L) = chamber footprint (m ) chamber enclosure time (h) . Net CH 4 fl ux was determined from CH 4 fl ux in soil incubations without C 2 H 2 .Aerobic CH 4 production = CH 4 fl ux under 10 Pa C 2 H 2 [2] Net CH 4 fl ux = CH 4 fl ux under no C 2 H 2 [3]

Fig. 1 .
Fig. 1.Aerobic CH 4 production and net CH 4 fl ux from incubated soil core under controlled laboratory conditions.Soil cores were obtained from crop fi eld, forest buff er, warm-season grass fi lter and coolseason grass fi lter soils.Each mean represents three observations (one observation for CH 4 production in warm-season grass fi lter soil) and bars are the standard error of the mean.

Fig. 2 .
Fig. 2. (A) Diel variation of CH 4 fl ux in crop fi eld, forest buff er, warmseason grass fi lter and cool-season grass fi lter soils on 16-17 July 2007 and (B) cumulative diel CH 4 fl ux.Each mean represents three observations and bars are the standard error of the mean.

Fig. 3 .
Fig. 3. (A,B) Methane fl ux, (C) daily precipitation, (D) daily soil moisture, and (E) soil temperature in crop fi elds (n = 1), forest buff ers (n = 3), and grass fi lters (n = 4) in 2007.Each mean represents observations and bars are the standard error of the mean.Gaps in the soil moisture data (D) were caused by either data loggers malfunction or removing a data logger during planting and harvesting periods in the crop fi eld.

Fig. 4 .
Fig. 4. CH 4 fl ux in crop fi elds (CF), forest buff ers (FB), warm-season grass fi lters (WGF), and cool-season grass fi lter (CGF) soils in 2007 (n = 40-49).I, II, and III indicate replicates.The boundary of the box closest to zero indicates the 25th percentile, a line within the box marks the median, and the boundary of the box farthest from zero indicates the 75th percentile.Whiskers (error bars) above and below the box indicate the 90th and 10th percentiles.Solid circles indicate outliers.