Abstract:
This thesis was conducted within the scope of a graduation fellowship from the state of Baden-Württemberg, Germany (Grant No.: VI 4.2-7631.2/Baumann) in cooperation with the Depart-ment of Ecology, Peking University, Beijing. Scientists specialised in both ecology and soil science investigated the same sites, thus allowing an interdisciplinary approach to evaluate soil properties, C and N cycles as well as geomorphological processes in close connection to ecosys-tem interrelations on the Tibetan Plateau. The research sites are located along a 1,200 km long north-south transect at altitudes between 2,925 and 5,105 m ASL. Two thirds of the Tibetan Plateau is influenced by permafrost. Due to the high sensitivity to global climate warming and land use changes, permafrost degradation processes are widespread, increasing the heterogeneity of soil formation, soil hydrology, and related soil chemical processes (i.e. C and N cycling). In order to account for the resulting extremely diverse ecosystem, investigations at different spatial scales related to large-scale climate patterns were performed. The scales comprise the total main transect, the split transect into an eastern and western section, diverse catenas along distinct geomorphological relief units, and finally the single site soil profiles.
The first part of this work examines C and N contents as well as portions of plant available min-eralised nitrogen in relation to their main influencing parameters. For investigations on land-scape scale, soil moisture was found to have the strongest effect on C and N cycling, followed by CaCO3-content and soil texture. Altogether, the general linear model explains 64% and 60% of the variation of soil organic carbon (SOC) and total nitrogen (NT) contents, respectively. Thereby, two aspects are important: (1) temperature variables have no significant influence and (2) indicators for soil development (i.e. CaCO3 and soil texture) are included besides commonly con-sidered ecological (i.e. moisture, temperature and biomass) parameters. It could be shown that in the highly diverse permafrost-affected ecosystem of the Tibetan Plateau, other factors than precipitation mainly control soil moisture contents and distribution, with permafrost and relief position being the most dominant parameters. Since pedogenic parameters turned out to be important predictors, the degree of soil development can be regarded as an additional control quantity, indicating higher C and N contents of topsoils with longer duration of undisturbed and stable soil development. Mineralised plant available N can be almost exclusively found as am-monium-N, which is closely related to higher soil moisture contents and frigid climate condi-tions, showing by far the highest contents in the permafrost main soil group. As nitrification is strongly temperature dependent, nitrate-N contents are correspondingly very low. The results provide clear evidence that limitation in plant available nutrients as a negative feedback to lower soil moisture is crucial for plant growth in nutrient-limited alpine grassland ecosystems, even though higher temperatures occur with respect to climate warming. Importantly, these strong feedback mechanisms between altering permafrost conditions (degradation and higher active layer thickness) and hence reverse influence of rising temperatures (further decay of permafrost and related dryer conditions) could be only detected by conducting this study on landscape scale. These dependencies are based on the overall limitation of moisture, because evaporation exceeds precipitation by far at all investigated sites. Degraded permafrost profiles show low C and N contents combined with distinct depth patterns, mainly caused by higher mineralisation rates and deposition of proximal airborne sediments. This is can be exemplarily shown for the Shule River basin located at the very north-eastern margin of the Tibetan Plateau, where soils under desert-type vegetation have their highest SOC density in soil depths between 20 and 40 cm, but not in the top 20 cm as evident for all other vegetation types. The main reason for these patterns are most likely such syngenetic soil forming processes. Results of soil respiration measurements basically confirm the findings observed for C and N contents. Belowground biomass and soil moisture explain 82% of the variation, whereas no direct effects of temperature could be described. Respiration values of alpine meadows were 2.5 times higher than of alpine steppes, which is a consequence of higher biomass and productivity in alpine meadows.
Besides the relations to control variables, SOC was further analysed with regard to its stocks and composition. The comparison of two main investigation sites for discontinuous and continuous permafrost, respectively, clearly shows higher SOC stocks for discontinuous (10.4 kg m-2) than for continuous permafrost (3.4 kg m-2). Highest values occur at water-saturated profiles (19.3 kg m-2), causing positive feedbacks to even higher SOC accumulation, if in turn denser vegetation isolates the soil. At the same time, these soils contain substantial higher portions of easily de-composable particulate organic matter fractions, which are especially vulnerable to climate change owing to shorter turnover rates. The colder and dryer climate in continuous permafrost areas leads to a lower productivity and an allocation of belowground biomass mainly in the top 10 cm. This can be approved by studies conducted in the Shule River basin, also characterised by low mean annual temperature and precipitation, showing average SOC stocks of 7.7 kg m-2. Moreover, different vegetation types can be distinguished very clearly, ranging from 4.4 kg m-2 under desert vegetation to 19.8 kg m-2 under partly water-saturated alpine swamp meadow (cf. above-mentioned corresponding respiration rates related to vegetation type patterns).
Moreover, it could be shown that soil inorganic carbon (SIC) and SOC are influenced by different parameter sets. Whereas soil physical and chemical properties are most appropriate to describe SIC, biotic and climatic factors are more important for SOC. Soil pH was found to predict 42% of SIC variation, leading to lower contents with decreasing pH. However, the overall effect of the released carbon under scenarios of potential soil acidification is assumed to be compensated, since SOC reacts vice versa to increased soil acidity.
Since pedological processes proved to have significant influence on C and N contents, it is im-portant to specifically qualify related weathering and sedimentation processes depending on the state of permafrost as well as land surface stability. To address this issue, weathering indices and pedogenic Fe-oxides were applied to particular sampling groups, distinctly influenced by the Indian and Asian Monsoon systems. The chemical index of alteration (CIA) represents the most useful weathering index, best describing large scale climate trends, varieties of substrates, and specific permafrost patterns. For pedogenic Fe-oxides, (Fed-Feo)/Fet ratio best illustrates small-scale shifts of pedogenesis. This can be confirmed by the differentiability of the main soil groups, which cannot be obtained by CIA. Essentially, groundwater and permafrost influenced soils can be clearly distinguished by distinct parameter sets best explaining each soil group: climate pa-rameters for the permafrost soil group (climate-zonal soil formation), and site-specific variables for groundwater-influenced soils (azonal soil formation). Moreover, the two soil groups can be significantly differentiated by Fep, even though both show high soil moisture and SOC contents. Therefore it can be assumed, that particular redoxi-morphic and soil formation processes with corresponding soil organic matter structures evolve under the influence of permafrost. Alto-gether, the application at various spatial scales give strong evidence that weathering indices and pedogenic Fe-oxides are useful tools to depict states of permafrost distribution and its degrada-tion features.
Summarising, the described geochemical patchwork (manuscripts 1-5) can be disentangled by applying weathering indices and pedogenic oxides ratios, depending on the scale and process. Together with the evaluation of the prevailing main influencing parameters, they proved to be crucial for assessing C and N cycles and ecosystem functioning on the Tibetan Plateau.