3.2.1 Site discription
Measurements were carried out in agriculturally used peatland sites at two research areas in Lower Saxony in North Germany.
One research area is located in the “Nordhümmlinger Moore”, in the northwest part of Lower Saxony (see ch.4). Measurements were carried out at two locations, south of the nature reserve “Leegmoor”, on the southern edge of the peatland complex. This area (“Surwold”) is agriculturally used and well drained. Underneath the shallow bog peat resides fine sand with silt (Eggelsmann & Blankenburg 1990). The site S1 is a bog covered with a layer of sand (Sanddeckkultur) after it was deep ploughed in the end of the 50th. Crop rotation with winter wheat and maize as well as oilseed rape and mustard as catch crops takes place. The cropland was harrowed and ploughed and supplied with organic and mineral fertilizer. Weeds consist of Chenopodium album and Galinsoga. The site S2 is mainly used to grow maize, occasionally potatoes are cultivated. After WWII, peat was extracted at this site and then left lying fallow. In the 1970s sand was applied on top and the land has been used as cropland since then. Harrowing, ploughing and grubbing as well as fertilizing with organic and mineral fertilizer take place. Beside Zea mays and Solanum tuberosum, the vegetation consists of Atriplex, Chenopodium album, Galinsoga, Poa annua and Echinochloa crus-galli.
The second research area is located in the Dümmer peatland (see ch.2). The two examination sites were classified as histic gleysols (histosols with only small peat layers = Anmoorgley (AG Boden 2005)), covered with a layer of sand and are located in the “Ochsenmoor”, south of Lake Dümmer, on the southern edge of the fen area. One site (cropland; O1) is used to
79 grow maize and winter wheat. Yearly harrowing, ploughing and grubbing as well as fertilising with organic and mineral fertilizer take place. Vegetation of the cropland consists of Chenopodium album, Echinochloa crus-galli and Persicaria lapathifolia, beside Zea mays and Titricum aestivium. The other site (grassland; O2) is located beside and is extensively managed grassland with typical grassland-vegetation (Anthoxanthum odoratum, Bromus hordeaceus, Alopecurus pratensis and Poa trivialis). Grass is cut two to three times per year.
3.2.2 Measurements of site factors
Soil parameters: The methods for soil identification as well as determination of true density (s), pore volume (PV), gravimetric water content and water filled pore space (wfps) are described in chapter two.
The dry bulk density (ρt) was calculated using the formula in chapter two. At each site ten soil samples at the depth of 0 to 10 and 10 to 20 cm, respectively were taken with sampling rings (250 ml) in March 2011 and June 2010. The soil samples were heated to 105 °C in the drying oven to determine the dry mass (VDLUFA 1991). At O2 the dry bulk density determined in June 2010 was taken for the time period from April until September, and the value determined in March 2011 was taken for the time period from October until March. At S1, S2 and O1 a different procedure was chosen: From the date when soil cultivation took place until December, we took the value of the dry bulk density determined in June, and from January until the date when soil cultivation took place, we used the value determined in March.
With each CH4 und N2O flux measurement, ten soil samples were taken with a boring rod for mineralised nitrogen (Nmin-Bohrstock) in 0-20 cm depth and subsequently mixed. Analysis of nitrate and ammonium content was carried out in the laboratory of “Landwirtschaftliches Labor Dr. Janssen” with the Continuous-Flow-analyser. The compounds were extracted with a calcium chloride CaCl2 solution (VDLUFA 1991).
Meteorological parameters: Meteorological parameters such as temperatures (air temp., soil temp. at 2, 5 and 10 cm depth), photosynthetic active radiation (PAR), air pressure and precipitation were measured and saved half hourly at meteorological stations. The meteorological station for Ochsenmoor is located near the grassland site of chapter two. The station for Surwold is about 20 km northeast of Surwold (see ch.4).
80 Half hourly meteorological parameters at the sites were achieved by using the data from the meteorological station. The soil temperatures at each individual site were separately measured and saved half hourly with a datalogger (DN Messtechnik, Norderstedt).
Water level: All sites were equipped with tubes perforated in the peat body, close to the collars. Water level (wl) at the grassland site was measured during each gas measurement campaign with an electric contact gauge during the entire measurement period and additionally continuously recorded every half hour using a Schlumberger MiniDiver from October 2010 until December 2011. At O1, measurements with the electric contact gauge have been conducted from March 2010 until June 2010 and measurements using a Schlumberger MiniDiver from October 2010 until December 2011. In Surwold the wl was measured during each gas measurement campaign with the electric contact gauge from October 2009 until August 2010 and continuously recorded using a Schlumberger MiniDiver from June 2010 until December 2011.
In addition to the continuous records with the MiniDiver, occasional measurements with the electric contact gauge have been performed for validation purposes.
In intervals of every three months we took samples from the ground water with a bailer and analysed for pH and electrical conductivity (Lf) with pH-electrode SenTix 950 (WTW) and standard conductivity measuring cell TetraCon 925 (WTW), respectively.
Biomass: For a description of examination of biomass refer to chapter two.
Carbon import and export: In case of organic fertilizer application at S1, S2 and O1, the collars and boardwalks were removed to make sure that the fertilizer was distributed evenly through the fertilizer spreader. The amount of fertilizer applied was estimated by the farmer.
We assumed a variation coefficient in spreading accuracy of less than 25 % (Frick 1999, Pöllinger 2006).
Carbon content of the dry biomass was assumed to be 45 % (KTBL 2005). C export through harvest was calculated accordingly. C/N ratio and nitrogen content of slurry and manure is 8 and 13.5 as well as 4 and 6 kg N t-1, respectively (KTBL 2005). Thus, carbon content of slurry and manure amounts to 32 kg C t-1 (= 32 kg C m-3) and 81 kg C t-1 (= 81 kg C m-3), respectively. The carbon content of fermentation residue was assumed to be similar to slurry.
81 3.2.3 Measurements and modelling of carbon dioxide exchange
A description of the determination of the CO2 exchange is performed in chapter two.
Measurement campaigns were held in intervals every four weeks, beginning in September 2009 and ending in December 2011. Additional measurements were conducted in case of management events (e.g. harvesting, tilling). If necessary, extensions for the chambers were applied (max. 230 cm).
3.2.4 Measurements of nitrous oxide and methane exchange
A description of the determination of the CH4 and N2O exchange is performed in chapter two.
The samples were analyzed in the gas chromatograph “Perkin Elmer Auto System”. A FID-Detector identified CH4, while an ECD-Detector was used to detect N2O.
Measurement campaigns were held in intervals every two weeks, beginning in September 2009 and end in December 2011.
3.2.5 Net ecosystem carbon balance and global warming potential
To obtain complete carbon balances of the examination sites, the net ecosystem carbon balance (NECB) was calculated (Chapin et al. 2006; see ch.2). DOC was estimated to 26 kg C ha-1a-1 according to Moore (1987). Values of DIC, CO and VOC are assumed to be negligible and not considered.
The global warming potential (GWP) was calculated according to IPCC (2007) (see ch.2). In general, the global warming potential over a time span of 100 years is taken (Drösler 2005).
Positive values represent efflux of CO2-equivalents into the atmosphere.
3.2.6 Statistical analyses
Unless otherwise stated, Microsoft® Excel was used.
Average values are arithmetic means +/- standard error.
Error analysis of CO2 gas fluxes was conducted by calculating the standard error for each calibrated regression model. Analogous to the interpolation of the half-hourly gas fluxes, standard errors were interpolated. The monthly and annual standard errors were calculated using appropriate error propagation equation. The standard errors of the means of the exported carbon through harvest were included.
82 For CH4 and N2O the standard error of the replicate chamber measurements of each measurement campaign were calculated and interpolated between the measurement campaigns analogous to the interpolation of the fluxes. The annual standard errors were calculated using appropriate error propagation equations.
Significant linearity of slope of the changes in gas concentration was tested following Huber (1984). To test if slopes are significantly different from 0, a t-test was performed (Neter et al.
1996). The variability of the slopes was calculated as the standard deviation of the residuals (syx). For the variability in PAR the coefficient of variability (cv %) was calculated.
Correlation and regression analysis was conducted providing the coefficient of determination (quadrate of Pearson Correlation Coefficient = R²) and tested for significance using a t-test.
Significant (p < 0.05) differences between the annual gas exchange balances were tested with the Permutation test “diffmean” (1000 permutations) using R script 0.97.237 (version 2.15.2) (simba package).