Abstract
Parts of south-west China are heavily exposed to acid deposition. This is the cause for the IMPACTS (Integrated Monitoring Project on Acidification of Chinese Terrestrial Systems) project, to which this thesis is related. Five study sites have been established in south-west China by the IMPACTS project. The investigations that constitute the basis for this thesis were conducted at one of these sites, the Liu Chong Guan (LCG) catchment outside Guiyang in the Guizhou province. The catchment is characterised by a subtropical climate, highly weathered soils and heavy acid deposition.
The main objective of this study was to gain a preliminary understanding of the way environmental factors determine the distribution of the ground vegetation in this type of ecosystem. Special emphasis was given to the effects of different soil characteristics, especially relative to the changes in these soil characteristics with respect to acidification processes. Major differences compared to boreal systems were also addressed. Investigations were made in 50 mesoplots distributed randomly within ten macroplots that were subjectively placed in order to represent the variation along presumably important ecological gradients. Recordings of species abundances and several environmental factors were performed by others in the field. Soil samples were collected close to each mesoplot and were analysed for the variables dry matter, loss on ignition, soil pH, exchangeable cations, total N, total C and adsorbed sulphate.
Prior to analysis of the data from LCG, two soil analytical issues were addressed: Intercalibration of soil parameters and comparison of CEC determination methods. An intercalibration was performed between the laboratory at the Chinese Research Academy of Environmental Science (CRAES) in Beijing and Norwegian laboratories. CRAES is analysing all the soil samples collected in connection with the IMPACTS project. The objective of the intercalibration was to discover possible discrepancies between the results and to sort out possible new artifacts when analysing the types of soil found in south-west China. The conclusion was that the determinations of the variables already mentioned was satisfactory, while for the variables particle size distribution, Al- and Fe-pools and adsorbed phosphate, the analytical procedures used by CRAES must be revised. Cation exchange capacity (CEC) is one of the key parameters in studying soil quality in regards to acidification. An experiment was set up to investigate in what way various determination methods give different results when analysing soils from south-west China. As expected, the major difference was found between buffered and unbuffered extracting agents, with the buffered agents giving a much higher total acidity and slightly lower calcium saturation. The effect was most pronounced for samples with a high organic content, meaning that for such samples the difference in clay types between the soils samples seemed to be of minor importance. A similar experiment including only samples with a low organic content should be set up to further investigate the importance of variable charge in clays. There were only minor differences between the unbuffered methods. However, there were indications that the Ba2+ ion is a slightly superior exchanger due to its higher ionic potential, and that the extraction procedure applied for extraction with NH4NO3 could be insufficient, because the soil suspension is not shaken. In the study of the extraction procedures it was confirmed that a single extraction is sufficient and that pH measurement of the extract is an appropriate measure to determine exchangeable H+ in acid soils. The determination of exchangeable cations with BaCl2 selected in the IMPACTS project can therefore generally be recommended for acids soils.
The vegetation and environmental data from LCG were analysed by methods commonly applied in vegetation ecology. Two main gradients in species composition were found, of which the first seemed to be governed by a complex-gradient in nutrient conditions and possibly also a complex-gradient in soil moisture, and the second by a complex-gradient related to litter depth. The saturation of manganese on the exchanger in all horizons and the magnesium-, potassium- and base saturation in the B1 horizon only were the nutrient variables that most strongly determined the distribution of the vegetation along the nutrient complex-gradient. The span in these variables could be related to different stages or degree of soil acidification. Manganese seemed to play a key role in the distribution of the vegetation, which could be explained by its role as an indicator of a gradient in different pH buffering systems. In some parts of the catchment soil water buffering appears to be managed by the dissolution and reduction of Mn-oxides, while in other parts the buffering is dominated by the dissolution of Al- and Fe-sesquioxides. The role of manganese should be further investigated be means of Mn-determinations in water samples and a survey of the Mn mineralogy and Mn content in plants in the catchment. Further acidification could result in a change in the species composition towards that typical for vegetation on nutrientpoor sites. Increased mobilization of Mn could possibly produce Mn-deficiency in parts of the catchment and/or toxic levels in other parts.
The quality of the deeper horizons seemed to be more important to the ground vegetation species abundance and composition than what is found in boreal systems. This can be explained by the deeper penetration of the roots and the thin upper horizons encountered several places in LCG. Calcium was less important in LCG, likely due to the high deposition of alkaline dust. Soil pH was not empirically related with the nutrient complex-gradient. This can be assigned to the only minor importance of buffering by adsorption of H+ in the catchment. There was a gradient in soil moisture in the catchment, but it seemed to run parallel with the nutrient complex-gradient.
The distribution of ground vegetation along the litter complex-gradient can mainly be related to the difficulty of establishment for many species, especially bryophytes, when the litter layer is thick. pH was strongly connected with the complex-gradient, but it was probably mainly passively following the gradient in litter depth, due to the production of humic acids in the decomposition of litter. There were, however, indications that this production of humic acids could affect the nutrient conditions along this gradient as well.
This investigation of the relationships between vegetation and environment in LCG must be regarded as preliminary. More information should be extracted by applying additional data treatment techniques. Furthermore, the compilation with results from the other IMPACTS catchments and re-analyses and continued monitoring will give a broader understanding of these ecosystems and the possible effects of acid deposition.