You are here: Home Publications Abstract "Heft 38"
Document Actions

Abstract "Heft 38"

Freiburger Bodenkundliche Abhandlungen

Schriftenreihe des

Institut für Bodenkunde und Waldernährungslehre
der Albert-Ludwigs-Universität Freiburg i.Br.
Schriftleitung: P. Trüby

Heft 38

Martin Armbruster

Zeitliche Dynamik der Wasser- und Elementflüsse in Waldökosystemen
- Zeitreihenanalysen, Simulationsmöglichkeiten und Reaktion auf experimentell
veränderte Stoffeinträge in den ARINUS-Wassereinzugsgebieten

Freiburg im Breisgau 1998

ISSN 0344-2691


Analyses of time series, possibilities of Simulation and reaction to experimentally altered element inputs to the ARINUS-watersheds.

This thesis summarizes results of measurements of water and element fluxes at the ARINUS experimental sites Schluchsee and Villingen (Black Forest, SW Germany) for the period November 1987 through October 1996. Mid- to long-term element fluxes in untreated control watersheds are characterized as well as fluxes in watersheds experimentally treated with ammonium sulfate, sulfatic magnesium/potassium fertilizer and dolomitic limestone. Research plots were neighboring and comparable watersheds covered with Norway spruce. In each watershed apart from the measuring of open land precipitation plots for measuring canopy throughfall, soil see-page water and soil moisture existed. Also runoff was measured in each watershed. At Villingen permanent wells allowed the sampling of groundwater. Geological parent material of both sites -granite and Mesozoic quartz sandstone - represents two parent materials of the Black Forest. The 50 to 80 year-old spruce stand at Schluchsee grows on highly permeable loamy Sandy podzols. At Villingen, the 90 to 130 year-old spruce Stands Stocks on sandy-loamy brownearths with a tendency to periodical waterlogging - respectively the stagnogleys. One aim of this thesis was to analyze the temporal changes of element fluxes at the untreated control watersheds. A second aim was the investigation of the changes induced by experimental nitrogen and sulfur treatment, liming with a dosage used in forest practice and fertilizing with magnesium/potassium. Also the nutrient cycling model NuCM was tested for the suitability for Simulation of water and element fluxes at the catchment scale and the possibilities for simulating the experimental treatments.

Long-year water-balances are plausible and accurate for the calculation of element fluxes. Differ-encesbetween long-year averages of potential evaporation and the water-balance are especially at the site Schluchsee are a consequence of the underestimation of snow and fog precipitation. The study period was hydrological representative. As a tool to quantify element fluxes in soil seepage the model WHNSIM was used for the Simulation of soil water fluxes. Rates of watershed runoff were compared with simulated seepage rates to test if the soil moisture measurement plots repre-sent the watershed. All waterfluxes revealed no temporal tendency in the study period.

Temporal changes of element fluxes in the untreated control watersheds were investigated by application of a seasonal regression model. The ARINUS-sites have a low atmospheric deposition rate compared to other sites in Central Europe. During the study Ca2+ and SO42" inputs decreased 20 to  25 %.  Nitrogen  input with  canopy throughfall remained  almost  unchanged.  As  a consequence of decreasing SO42- input a decreasing of SO42- concentrations in soil seepage water and streamwater was observed. The lower decrease of runoff concentration compared to soil water concentration is attributed to the incomplete reversibility of stored S in the soil accumulated in former times during higher deposition rates. Sulphur seemed to be stored to a great extent in the deeper mineral soil and decomposed rock zone. Decreasing SO42-concentrations resulted in an increase of the acid neutralizing capacity of the deeper soil seepage water. Acid neutralizing capacity in runoff remained unchanged. Despite unchanged atmospheric nitrogen inputs no tendency of increasing NO3" Outputs were observed at both sites. This could also be observed at the site Schluchsee which is a “N  saturated" site according to the definition.

Apphcation of sulfatic fertilizer leading to an improvement of the Mg and K nutrition Status at both sites resulted in higher concentrations of applied elements (S, Mg and K) in seepage water. Also high concentrations of AI and a strong decrease of the acid neutralizing capacity especially in the year following the application were observed. The effects of fertilization to streamwater chemistry differed distinctively between both sites. At the site Schluchsee with a dominating vertical waterflux and marked buffering processes in the deeper seep zone (release of base cations, precipitation of Al-complexes), no sustained negative influences of the MgSO4 treatment on the streamwater chemistry were observed. Losses of Mg with streamwater add up to 20 % of applied amount 8 years after treatment and were relatively low despite high water fluxes. In contrast, the site Villingen, characterized by periodical waterlogging and dominating lateral water fluxes, showed stronger and longer lasting effects on streamwater chemistry. Besides the exchange capacity for Mg2+ (K+) and SO42- of the upper soil effects and risks of apphcation of sulfatic fertilizer are additionally depending on the direction of water flow and the buffering capacity of the deeper subsoil. Lower additional Outputs of SO42- seven years after fertilization of Villingen (22 % of applied amount) in contrast to Schluchsee (50 % of applied amount) are caused by lower water fluxes (average catchment Output: Viningen 465 mm; Schluchsee 1417 mm) and slower seepage velocity.

Application of (NH4)2SO4 three times revealed distinctive differences in N turnover between both sites. Applied NH/ was nitrified in considerable amounts at the site Schluchsee despite the low soil pH. After the first application of 150 kg ha-1 N N-outputs increased to six times the value of the untreated catchment. Whereas only the third application led to N-outputs in the range of the inputs at the N-limited site Villingen that showed nearly zero N-output in the untreated catchment. Effects of experimentally induced acidification in streamwater did not only depend on the nitrification potential. Higher proton loadings induced by nitrification at the Schluchsee site were buffered in considerable amounts in the deeper subsoil. Difference in additional SO42--outputs can also be explained with differences in water fluxes. However the site Villingen had a marked S-retention in the upper mineral soil whereas SO42--adsorption at the site Schluchsee is limited by organic compounds.

As expected application of 4 t ha-1 dolomitic limestone led to only small losses of the applied elements Ca and Mg (1 respectively 3 % of applied amount). NO3--concentrations in seepage water increased distinctively during the first three years after liming. However watershed Outputs (streamwater) showed no significant increase in NO3- -concentrations. This could be explained by reduction of NO3- and gaseous N-losses within regions with high groundwater level along the stream Mg2+-concentrations had increasing trends in seepage water even six years after Hming. Therefore only for Mg2+ remarkable increases in streamwater concentration could be observed. A drastic increase of buffer capacity and pH in streamwater did not appear after liming. Application of dolomitic limestone had some effects on high flood chemistry. After liming tendencies of increasing acid neutralization capacity and decreasing Al-concentrations compared to the control catchment were observed.

The nutrient cycling model NuCM was calibrated for the Schluchsee site to simulate element fluxes. This model describes sufficiently water fluxes and annual element Outputs. However, N-budget and the temporal variations of other elemental concentrations in seepage water ard streamwater were not simulated satisfactory. Agreement of annual element Outputs were insufficient to characterize the quality of a Simulation. The experimental MgSO4-treatment was simulated with the model after calibration and validation. This Simulation gave additional insights about the quality of model predictions that could be directly compared to the measurements. Especially during the Simulation of SO42- and Mg2+ Sorption and element retention of S and Mg in the biomass after treatment discrepancies occurred. Forecasts of long term element fluxes seemed to be not advisable generated with the NuCM model for the site Schluchsee. Models are a valuable tool to extend knowledge about forest ecosystem processes but they cannot replace intensive field observations.

Personal tools