Doctoral Thesis
Refine
Year of publication
Document Type
- Doctoral Thesis (43) (remove)
Language
- English (43) (remove)
Has Fulltext
- yes (43)
Is part of the Bibliography
- no (43)
Keywords
- climate change (6)
- Jahresring (5)
- Moor (5)
- dendrochronology (5)
- Dendrochronologie (4)
- Klimawandel (4)
- dendroecology (4)
- forest ecology (4)
- Ökologie (4)
- Buche (3)
Institute
- Institut für Botanik und Landschaftsökologie & Botanischer Garten (43) (remove)
For many years, rangeland ecologists have debated about whether the state of semi-arid and arid rangelands is the expression of an ecological equilibrium or non-equilibrium dynamics reached in response to grazing livestock. Since the problem has been considered at different spatial scales, it is recognised that the competing concepts of equilibrium and non-equilibrium dynamics need to be integrated. Furthermore, the role of environmental variables as vegetation driving factors has long been ignored in the discussion on grazing effects on ecosystems. Present thesis, examines the dependence of plant communities on environmental in particular site-ecological conditions in three ecosystems of Western Mongolia established along a precipitation gradient to detect the vegetation-driving ecological factors involved. Furthermore, grazing impact is exemplary assessed in a desert steppe at additional spatial scales of plant communities and population. At the landscape level, a classification of plant communities in dependence on environmental conditions is carried out. Additionally, the investigations focused on the impact of grazing on soil and on the occurrence of grazing-mediated plant communities. Data were sampled along an altitudinal gradient between 1150 m to 3050 m a.s.l. from arid lowland with desert steppe via semi-arid mountain steppe to humid alpine belt. Within each altitudinal belt, data sampling was carried out along grazing gradients, established from grazing hot spots to areas distant from them. By means of an environmentally based vegetation classification, factors with highest explanation values for largest variation in vegetation were identified and considered as most responsible for vegetation patterns. To validate and affirm the classification, three different statistical methods are applied: environmentally adjusted table work of vegetation relevés supported by cluster analysis of species distribution, detrended correspondence analysis of vegetation data separately from environmental data, and the principle component analysis of only environmental data. Vegetation-driving factors change along the altitudinal gradient from abiotic forces in the desert steppe, as e.g. altitude and soil texture, to abiotic and biotic forces in the alpine belt represented by soil texture, soil nutrients and grazing. Vegetation and soil of all ecosystems respond to grazing but with different patterns and to a different extent. While desert steppe does not indicate grazing communities, mountain steppe demonstrates grazing communities at fertilised sites and alpine belt at nutrients depleted sites. Thus, the grazing sensitiveness of the ecosystems is assumed to be linked with plant productivity and the role of vegetation as site-determining factor (Chapter 2). To examine grazing impact at lower spatial scales on desert steppe as the ecosystem with lowest grazing sensitiveness at the landscape scale, at community scale the total number of species, the total vegetation cover, the percentage of annual species, the cover of annual species, and properties of soil nutrient along gradients of grazing intensity within three different communities were assessed. Vegetation parameters respond to grazing in different ways, and the responses of the same parameters vary between plant communities. Correlations with grazing intensity indicate only partly statistical significance. Significant correlations of grazing intensity with concentrations of soil nutrient point to eutrophication in two communities. A comparison of vegetation and soil properties refers to a greater indirect influence of grazing via increased soil nutrients than the direct effect on vegetation (Chapter 4). At the population level, data about stand density, aboveground biomass, individual plant weight, and the proportion of flowering plants of the dominant dwarf semi-shrub Artemisia xerophytica were collected along a grazing gradient. Soil data were used to distinguish between grazing and edaphic influences. All parameters of Artemisia xerophytica reflect the assumed gradient of grazing intensity up to 800 m distance from the grazing hot spot. As grazing pressure decreases, plant density and total biomass per plot increase. The average shrub weight, an indicator of plant vitality, is related to both: distance from the grazing hot spot and stand density, which may be explained by additional intraspecific competition at higher densities. At a longer distance, these effects are masked by variations in soil parameters determining water availability, leading to quite similar degradation forms. These results are in contrast to other studies carried out at the scale of plant communities which did not detect significant changes along a grazing gradient. One explanation is the different map scale: the study took place only within a single plant community comparing populations of one species (Chapter 3). The comparative study demonstrates that even arid desert steppes of western Mongolia display equilibrial and non-equilibrial properties, depending on the observational scale: while no grazing mediated plant communities could be identified at the landscape scale as predicted by the non-equlilibrium model, at the community level vegetation parameters imply an intermediate position between equilibrium and non-equilibrium system. At the population level, the results clearly reflect the grazing gradient as predicted by the equilibrium model (Chapter 4). As a consequence, the assessment of vegetation dynamics and grazing impact in rangelands requires a multiple-scale approach that duly considers different vegetation properties responding differently to grazing, climatic and edaphic variability at different spatial scales. It is further suggested, that future research should draw comparisons between landscapes that co-evolved with herbivory, and those that did without (Chapter 4).
Vegetation dynamics on abandoned terraces of Sicily: the course and driving factors of succession
(2007)
Secondary succession processes have been widely studied in Europe for some agroecosystems, but not for terraced ones. The first part of the present study focuses on a description of the plant communities involved in secondary succession processes on Sicily (Italy) a) from a floristic and structural point of view and b) from a species diversity point of view. In order to obtain these results, 129 vegetation relevés (sensu Braun-Blanquet) were made on abandoned terraces in five of the main terraced areas of Siciliy: 1) the Aeolian Islands, 2) Pantelleria Island, 3) Mt. Etna, 4) the Palermo Mts. and 5) the Hyblaean Plateau. Only abandoned vineyards or grain crop fields were selected as sample plots, always 50 m2-sized. The results of biodiversity evaluation by t-tests and ANOVA showed that vascular plant diversity is linked to disturbance regime and to abiotic factors (especially geological substrate). Especially grazing increases species richness. Moreover, it was found that on limestone species richness is higher than on volcanic substrates. Vegetation relevés were also analysed with DCA and TWINSPAN. The resulting 14 sample plot groups (= clusters) were then used to check the dynamic relations. From a floristic point of view, plant communities involved in secondary succession processes on Sicilian terraces are quite different between and within the five study areas. This is mainly due to different substrate and bioclimatic conditions. Moreover, vegetation is strongly influenced by abandonment age and disturbance status. If no disturbance biases succession, then plant communities evolve rather rapidly (30-50 years) to maquis communities. If frequent fires or intense grazing occur, secondary succession is blocked in a "steady state". The second part of the present study focuses on the colonization mechanisms of old fields by woody species. In a first section, the existence of 1) the neighbourhood effect and 2) the safe-site effect are checked by analyzing 51 transect relevés, made up of 357 subplot relevés (1x1m). The transects were made in target fields 1) with older neighbour (i.e. old succession stage characterized by maquis communities) and 2) with older neighbour absent within a 100 m-distance. All woody species individuals were counted, recording if they grew within the influence of a potential safe site (former crop plants of vine and the terrace wall base). Data evaluation by Kruskal-Wallis ANOVA and Mann-Whitney Rank Sum confirmed the existence of the two effects. Moreover, it was shown that animals as dispersal vectors strongly influence these effects. For the neighbourhood effect, seed dispersal distance is the crucial point, while for the safe site effect 1) passive facilitation (i.e. animals tend to create heterogeneous seed rain patterns because they frequent certain microhabitats more often than others) and 2) active facilitation (i.e. the positive influence of an existing woody or herbaceous plant individual on the establishment or the growth of another one) are crucial. The second section describes the performance of establishment of Quercus ilex L. in different microsites of terraced old fields. In November 2004, acorns were buried on a North-facing slope and on a South-facing slope in five different microsites: 1) under vine plants, 2) at wall bases, 3) under the canopies of isolated shrubs, 4) between small rock accumulations and 5) in open spaces (i.e. outside of any of the previously named microsites). In monthly checks, seedling emergence, survival, height and leaf number were recorded. Moreover, in April and July were measured air temperature and air humidity in the different microsites. Overall emergence rate was 52.4% (n = 1,020). More seedlings emerged on the South-facing slope (S; 59.8%) than on the North-facing slope (N; 45.0%). Emergence was higher when acorns were buried under vine plants and at the wall base than in other microsites of the old fields. At the end of the experiment (September 2006), 45.3% of all emerged seedlings were still alive (29.2% on N, 58.9% on S). Survival was higher in general on the South-facing slope, and higher under vine plants and at the wall base than in the open spaces of the old fields. From literature, it is known that seed vitality, seed germination and seedling survival of Quercus ilex are favoured by shady, wet and fresh conditions. The temperature and air humidity measurements showed that at the wall base, under vine plants and under isolated shrubs environmental conditions are milder than in open spaces. However, even if temperature and relative air humidity seem to play an important role for Quercus ilex seedling emergence and survival, they did not unambiguously explain the differences between the safe site types. A factor of major importance is probably soil moisture. As a last part, the present study discusses what does the obtained results mean for terrace landscape conservation and biodiversity management.
To uncover the genetic structure of Populus euphratica forests along the Tarim River in Xinjiang, China, a PCR set of eight microsatellite markers was established. 18 primer pairs originally developed for P. tremuloides and P. trichocarpa were screened for amplification in P. euphratica. The eight most variable loci were selected for further genotyping experiments. Subsequently, two multiplex PCR assays, each containing four loci, were set up and optimized. Three populations containing altogether 436 trees were used to characterize the selected loci. The set was found to be moderately polymorphic (mean expected heterozygosity = 0.57). The resolution was sufficient to discriminate even siblings with high confidence (PID = 1.81x10-5). Cumulative exclusion probabilities were 0.89 (single parent), 0.98 (paternity), and 1.00 (parent pair) and proved the set’s suitability for parentage analysis. Practical and theoretical analysis of consequences of genotyping errors in this semi-clonal plant showed that the vast majority of errors (62.1%) lead to division of identical genotypes. Merging of different genotypes was found to be a very rare case (0.4%). This always leads to an overestimation of genotypes. A similarity threshold of one allele difference between two genotypes to be regarded as being identical lead to an underestimation of clonal richness and genotype number of one per cent compared to an overestimation of more than 20 per cent without such a threshold. Allowing a certain amount of variation is therefore expected to reflect the clonal structure better than an analysis that considers exact matches only. Using a combination of morphological and molecular analyses, a first study demonstrated that root suckers are clearly distinguished from seedlings in their root architecture. Root suckering starts when trees are 10–15 years old and bridges distances of up to 40 m at a time. Root suckers depend on their parent tree for at least five years and are expected to have a higher mortality than generatively grown trees. Molecular analysis of old growth stands revealed a highly variable proportion of clonal growth between different stands. In the study area, the proportion of clonality decreases with distance to the main river bed (R = 0.31 at the site closest to the main river, R = 0.97 at the site farthest away from the river). An analysis of the history of river movements at different sites indicates a dependency of clonal growth on the frequency of ground water replenishment by the yearly floods. Genetic differentiation among the stands in the study area is low (FST = 0.055), and isolation by distance was not detectable (P = 0.058). Also, the river does not function as a vector for directed gene flow in downstream direction (P > 0.11). The forests are therefore considered to be one large panmictic metapopulation with unrestricted gene flow. Clonal growth does not lead to higher final stand densities (P = 0.99) and is obviously not of crucial importance for stand survival. Furthermore, analysis of vitality measures and size differences indicate that root suckers are in disadvantage both in vitality and in survival rate compared to seedlings. In this light, a possible function of clonal growth as a luxury strategy to enhance a genetic individual’s reproduction success under good site conditions can be discussed. The genetic structure of the (meta)population bears direct implications for management and conversation of the Tugai forest in Xinjiang. Due to the low degree of differentiation and the unhindered gene flow even small, fragmented, or isolated populations have conservational value, thereby clearly answering the SLOSS question (a single large or several small protected areas) in the latter sense. More than that, non-clonal stands with the highest amount of genotypic diversity can be easily identified on satellite and aerial images. Selection of such stands for conservation is therefore possible without expensive and time-consuming molecular analyses.
This dissertation evaluates the effects of site conditions and livestock grazing on the vegetation of Azerbaijan’s winter pastures. We improved methods to estimate grazing intensity in vast rangelands and enhanced an approach to detect discontinuities in vegetation changes along environmental gradients. All analyses use field data from the semi-arid rangelands of Gobustan and Jeiranchel, at the foothills of the Greater Caucasus mountains. The data set comprises 313 vegetation relevés, each sized 100 m², based on a pre-stratification using topographical parameters. Additionally, we included data from farm transects and exclosure experiments. For each plot, selected site and soil variables were determined. VEGETATION AND SITE CONDITIONS: By means of cluster analysis, we derived 16 vegetation types with a total of 272 vascular plant species. Our vegetation classification, which is closely linked to site conditions, is an important groundwork for adapted rangeland management and monitoring. The study areas are dominated by semi-deserts with a high coverage of dwarf shrubs, and the mean number of vascular plant species was found to be about 28 per 100 m². According to ordination analysis (NMDS), species composition changes primarily along the altitudinal gradient, gradually proceeding from the Salsola nodulosa semi-deserts of the lowest parts (below 300 m a.s.l.) to the Salsola ericoides and Artemisia lerchiana semi-deserts of the upper regions (300–650 m a.s.l.). Soil salinity and carbonate concentration decrease as altitude increases. A second gradient reflects grazing intensity. One plant community that is typically found on intensively grazed sites in the vicinity of farmyards stands apart from the rest, which are subject to lower grazing and trampling pressures. A third factor that differentiates plant communities is the sand concentration of the soils. Additionally, communities that occur on steep slopes differ from communities that occur on level terrain. EXCLOSURE EXPERIMENTS: Exclosure experiments revealed that short-time abandonment of grazing leads to an increase in the number of annual species, in vegetation coverage, and in the heights of forbs and grasses. Clipping experiments indicated that the herbaceous species show hardly any compensatory growth in response to grazing. ESTIMATING GRAZING INTENSITY: A recurrent theoretical problem in rangeland research is the spatial modelling of grazing intensity around grazing hotspots like farms or watering places, the so called piospheres. In a widely used approach, grazing intensity is assumed to decrease in direct proportion to the distance from a hotspot. The resulting response patterns, which relate characteristics of the vegetation or site conditions to grazing intensity, are often nonlinear, and have been interpreted as indicating threshold changes or diff erent state-and-transitions along grazing gradients. However, we show that these ‘thresholds’ are usually geometrical artefacts. Taking into account the concentric structure of grazing hotspots, we suggest a new approach that approximates grazing intensity as the ratio of the total number of livestock kept at the farm to the distance between a given plot and the hotspot centre. Our approach is a simple yet significant improvement over current approaches because it enables us to merge or compare data from different sampling sites and because the approximation is in direct proportion to other grazing indicators like dung density or soil salinity. SPECIES TURNOVER PATTERNS: Combining our new grazing pressure model with species presence/absence data, we modelled vascular plant species responses, patterns of species richness and species turnover along grazing gradients on farm transects in Gobustan. The derived typical species response pattern along the finite grazing gradient is a sigmoid decrease. Species richness declines monotonically with increasing grazing intensity and thus conforms to generally acknowledged assumptions on the relationship between species richness and grazing pressure in semi-arid rangelands. Species turnover along the gradient was calculated using the slopes of species response curves. At first sight, the resulting pattern gives evidence for a discontinuous change. However, it ranges within the 95 % confidence interval of a null model based on assumptions of the individualistic continuum concept. Thus, species composition seems to change continuously along grazing gradients in Gobustan. This new null model approach can probably be adapted and applied to all ecological gradients and is useful for the validation of individualcontinuum or community concepts.
Peatlands cover only about 3% of the terrestrial surface but are significant players in the global carbon (C) cycle and the climate system, since they store roughly one quarter of the global soil carbon (C) and are among the largest natural sources of methane (CH4). Since the resulting feedbacks on the climate system are uncertain, research efforts aim at identifying key processes and quantifying the C exchange from ecosystem to regional and global scales. To identify peatland ecosystem dynamics requires analysis of yet different scales. The key scale for their C dynamics is the microform scale, which is the smallest entity of the system. To estimate ecosystem dynamics, up-scaling from the microform scale is needed. Up-scaling demands (1) a correct estimation of the spatial heterogeneity and (2) the correct aggregation. In this thesis, the traditional spatial weighting of microform fluxes by the microform distribution is evaluated by (1) analyzing the flux calculation procedure, (2) investigating the effect of the resolution of the landcover maps on the up-scaling and by (3) cross-evaluating the up-scaling result with the directly measured ecosystem flux. Eventually, it is evaluated how these dynamics are considered in a mechanistic ecosystem model (LPJ-WHyMe). CH4 fluxes were measured on the microform scale with the closed chamber technique and on the ecosystem scale with the eddy covariance (EC) technique. The quantification of microform fluxes relies on the correct flux calculation. Since only few gas samples are taken during the closure period, traditionally the linear regression is applied when calculating CH4 fluxes from chamber measurements. Still, the chamber itself affects the diffusion gradient between peat and chamber atmosphere resulting in a theoretically non-linear concentration increase in the chamber. Using data with six data points per measurement from different microform types it is tested whether the linear or exponential regression fits the data better. In the majority of cases, the linear regression fits best. However, the exponential concentration change might still not be detectable resulting in an underestimation of the ’real‘ flux and the test of different techniqes to estimate the slope of a non-linear function with small sample amounts is recommended. To define the spatial heterogeneity of the peatland surface, the application of remote sensing techniques offer the advantage of supplying area-wide information with less uncertainty when compared to vegetation mapping along transects. However, the required resolution to resolve the microform distribution is <1m which in this study was derived from near-aerial photography. Besides for up-scaling, the resulting high-resolution landcover map was used in combination with a footprint model to analyze (1) the effect of landcover on the directly measured ecosystem flux and (2) its spatial representativeness. It was shown that fluctuations of the measured ecosystem flux over periods of several days could be explained by changes of the landcover composition in the source area of the EC measurements. The estimated budget was slightly biased towards the higher emissions from lawns which could be corrected. Still, the seasonal ecosystem CH4 budget was higher than the estimate derived from the up-scaling of microform fluxes. This is most likely due to an underestimation of microform fluxes by the chamber technique. Generally, the budget estimate derived from EC measurements was more accurate, i.e., characterized by less uncertainty than the up-scaled estimate. The developed approach depends on (1) identification and accurate measurements of all relevant microform types and (2) on spatial information which should be smaller than the footprint size of the EC measurements and available on the scale relevant for the studied process, i.e., the microform scale. The demonstrated effect of microform dynamics on the ecosystem flux highlights the importance of dealing with spatial heterogeneity of ecosystems in mechanistic modelling. For example, in LPJ-WHyMe, the ecosystem flux is simulated with mean input variables as water table level. To investigate its model performance, flux data from the rather homogeneous peatland margin and the more heterogeneous peatland centre were compared with the model output. At the homogeneous peatland margin, the ecosystem flux was clearly dominated (with a contribution of 91%) by one microform flux. In this case, one water table level as input variable could be used to estimate the ecosystem flux. However, for a heterogeneous site such as the peatland centre in this study, only one mean water table would simulate a mean microform flux but not the ecosystem flux. Consequently, it is recommended to incorporate at least one high-emitting and one low-emitting microform type in the model to increase the model performance.
Carbon dioxide (CO2) is one of the most important factors of the Earth’s carbon cycle. Peatlands are well-known to be a long term sink for atmospheric carbon dioxide. Under changing environmental conditions, the carbon balance and hence the CO2 fluxes can be significantly changed, and peatlands may even become a significant atmospheric carbon source. To be able to predict the changes in climatic conditions and their effects on ecosystems, it is important to understand the contemporary CO2 exchange of the ecosystems. Many studies on peatland CO2 fluxes have been conducted in the boreal zone of North America and Scandinavia. Still little scientific evidence is available from peatland ecosystems of boreal Russia. This dissertation presents the detailed investigation of CO2 dynamics and the relevant processes and environmental factors from the boreal peatland site Ust-Pojeg (61°56'N, 50°13'E) in Komi Republic, northwest Russia. On the small spatial scale (microform), the investigated peatland was characterised by high variability in vegetation composition and coverage as well as in water table level which resulted in large variability in CO2 fluxes not only between the microform types but also within one microform type. The cumulative flux over the investigation period for the different microforms ranged from strong CO2 sources to CO2 sinks. An area-weighted estimate for the entire peatland showed that it was a CO2 source for the investigation period, which was characterised by average conditions in terms of precipitation and temperature. The CO2 fluxes were measured at different scales: by the closed chamber method at the microform scale and by the eddy covariance technique at the ecosystem scale. Three different upscaling methods were used to compare the fluxes. Irrespective of the upscaling methods, the discrepancies between the estimates based on the upscaled chamber measurements and estimates based on measurements by the eddy covariance technique were high. The high spatial heterogeneity of the vegetation and the water table level and thus of the CO2 fluxes were recognised as reasons for high potential errors when upscaling CO2 fluxes from the microform to the ecosystem level. Large discrepancies were also observed in comparison between measured CO2 fluxes and CO2 estimates based on the mechanistic ecosystem model LPJ-GUESS. Insufficient model forcing may have led to errors in the timing of the onset and the end of the growing season, and the modelled vegetation did not always reproduce the observed vegetation. These two factors may have led to the discrepancies in the model-measurement comparison. Although the closed chamber technique is widely used for measurements of CO2 fluxes between ecosystems and the atmosphere, the errors which might occur during the measurement itself or which are associated with the used measurement devices as well as the flux calculation from chamber-based CO2 concentration data are still under discussion. The study showed that the CO2 fluxes measured by the closed chamber method can be overestimated during low-turbulence nighttime conditions and can be seriously biased by inappropriate application of linear regression for the flux calculation. The methodological studies were conducted at the boreal peatland Salmisuo in eastern Finland (62°46'N, 30°58'E). The methods developed in this dissertation could contribute significantly to improved CO2 flux estimates. VI
Northern peatlands are ecosystems with unique hydrological properties, storing about 400-500 Gt of carbon. As the production rate of organic material is higher than its decomposition, which is slowed down in the wet and cold environment, peatlands store a great amount of carbon. Carbon assimilated from the atmosphere during photosynthesis by plants is partly lost due to autotrophic and heterotrophic respiration as carbon dioxide (CO2), as methane (CH4) or/and as dissolved organic carbon. The proportion of each carbon component is strongly controlled by environmental conditions as temperature, radiation, precipitation and subsequent water table changes and active role of vegetation. With predicted changes in the global climate, changes in the influence of environmental parameters on peatland ecology are expected. Thus thorough research is essential for a better understanding of mechanisms which influence carbon cycling in peatlands. In this thesis, various components of the carbon cycle were studied at two boreal peatland sites (Ust Pojeg in Komi Republic in Russian Federation and Salmisuo in Eastern Finland) using the micrometeorological eddy covariance method. The focus was placed on the temporal changes of the controlling parameters, ranging from a few days during short snow thawing through the rest of the year. At the Salmisuo site, two measurement seasons allowed to address possible inter-annual variation. We observed that diurnal variations in methane emissions which are typically controlled by vegetation during the growing season, might appear during snow melt as a result of the influence of physical factors rather than biological factors. The diurnal pattern in methane emissions was caused by the interaction of the freeze-thaw cycle and near urface turbulence. During the night time, when surface temperatures fell below zero and caused formation of the ice layer, methane emissions were only around 0.8 mg m-2 h-1, however after the increase in temperature and melting of the ice layer they reached peak values of around 3 mg m-2 h-1. The near surface turbulence had a significant influence on methane emissions, however only after the thawing of the ice layer. The effect of changing environmental parameters over the year was further elaborated on a carbon dioxide time series from the Ust Pojeg site. The generally accepted effects of temperature on ecosystem respiration during the night are not stable throughout the year and can change rapidly during the growing season. Using moving window regression analysis I could show that the strength of the exponential relationship between ecosystem respiration and temperature is changing during the year. This was in correspondence with recent publications elaborating on sub-seasonal changes of the controlling parameters. In general, measurements from the Ust Pojeg site represent estimates of annual CO2 and CH4 fluxes with an annual carbon balance of -94.5 g C m-2 and a new contribution to the quantification of trace gases emissions from a Russian boreal peatland. The inter-annual comparison of net ecosystem exchange (NEE) measurements with previously published data on CH4 and DOC flux from the Salmisuo site showed that the NEE of CO2 is the most important component of the carbon balance at this site. However, primary production was not responsible for the inter-annual changes in NEE. Rather, the effects of water table position during the year had a strong influence on ecosystem respiration, which was probably due to the influence on soil respiration, and higher NEE was observed during the year with smaller primary production, but higher water table levels. The effects of higher precipitation and higher water table during the wet year were shown to increase CH4 flux and the export of DOC, but their effects could not compensate for changes in ecosystem respiration. In the presented thesis intra- and inter- annual changes in carbon flux components and their controls, in our case attributed mostly to hydrological conditions in combination with other environmental parameters as temperature and the role of peatland vegetation, are discussed.
Recent climate change has affected the forest system comprehensively. Northern hemisphere elevational treelines are considered as a key environment for monitoring the effects of current anthropogenic climate change. Moreover, trees from these areas are also widely employed in paleo-climate reconstructions. The stability of the tree growth climate relationship under current scenario is crucial for all tree ring based climate researches. It is important to investigate how trees respond to this rapid environmental change at altitudinal treelines. Tree cores from 21 treeline sites of three species (Pinus tabulaeformis, Picea crassifolia, and Sabina przewalskii) from Northeastern Tibetan have been conducted in this thesis. The instable correlations between tree growth and climate are the general response pattern of trees from all study sites in NE Tibetan Plateau. Picea crassifolia shows the most instable response to climate factors (mean monthly temperature and total monthly precipitation). Pinus tabulaeformis and Sabina przewalskii just showed instable and divergent responses to their main limiting climate factors but no clear trend was found which is limited by the few sample sites. Corresponding to divergent responses of Picea crassifolia to mean monthly temperature, most radial growth of Picea crassifolia were inhibited by this climate change type drought, only few trees within same sites grew faster due to temperature increasing during recent decades. The divergence response mainly started in last 30 years in six of eleven sample sites over the Northeastern Tibetan Plateau. North-westerly drier sites showed a large percentage of trees per site with a negative correlation to temperature and mostly southerly moister sites showed more mixed responses with both negatively and positively responding trees within site. Concurrent with the regional pattern, low elevation sites show mostly negative correlations with temperature and high elevation sites show more mixed responses. As the hydrothermal conditions of the investigation area changed to a drier and warmer combination, drought stress on tree growth have been intensifying over time and expanding spatially from the middle to most of our study area during the last half century. The Picea crassifolia tree growth climate relationship conducted on an elevational gradient with four different levels from upper treeline to lower treeline at the NE Tibetan Plateau. Results show that upper treeline trees show divergent growth trends and divergent responses in recent decades. Trees from lower treeline show a strengthening drought stress signal over time and no divergent growth trends within sites. This potential ecological reaction of tree populations to changing environmental conditions shows an implications for using trees to reconstruct climate, since the indiscriminate use of tree ring data from sites showing opposite responses to increasing warming could cause mis-calibration of tree ring based climate reconstructions, and over- or underestimation of carbon sequestration potential in biogeochemical models. The physiological response of Sabina przewalskii tree growth to major limiting climate factors based on the Vaganov-Shashkin (VS) model indicated that precipitation during the early growing season, especially in May and June, has significant effect on tree growth, while temperature mainly affects tree growth by warming-induced drought and by extending the growing season in the NE Tibetan Plateau. Under current and projected climate scenarios, modeling results predict an increase in radial growth of Sabina przewalskii around the Qaidam Basin, with the potential outcome that regional forests will increase their capacity to sequester carbon. However, most Picea crassifolia trees growing at lower elevations than Sabina przewalskii might be continue stressed by the warming induced drought and might decrease radial growth in future.
The pollen record is a powerful proxy to reconstruct past terrestrial vegetation, but quantifying plant abundances is strongly limited because plants produce pollen in different amounts and pollen is dispersed differently. Further complications arise from the use of percentage data. Finally, a pollen grain deposited at a site may have arrived from proximate or distant sources, which implies that a single pollen sample may reflect very different vegetation scenarios. Present thesis suggests improving quantitative reconstructions of past vegetation by refined calibration of the pollen-vegetation relationship (paper I) and application of the downscaling approach (papers II-IV). Paper I primarily addresses the questions of pollen production and dispersal by calibrating the pollen-vegetation relationship. Data analysis employs the common extended R-value (ERV) approach and a new data-model comparison method, which appears more suitable than the ERV approach. For the first time PPEs have been calculated using three contrasting pollen dispersal options, including a Lagrangian stochastic (LS) model. The study proves that the underlying pollen dispersal model is a crucial parameter in PPE calculations and that the calculations with the LS model produce more reliable and realistic PPEs. Papers II to IV address quantitative reconstructions of past vegetation. Using the newly developed downscaling approach, the three studies explore fine scaled vegetation patterns in NE Germany during the Late Glacial and early Holocene. The main assumption of the downscaling approach is that the present day pattern of abiotic site conditions (e.g. the pattern of soil substrates) existed, at least to a large extend, also during the study periods. The basic principle of the approach is to test, whether pollen deposition in sites across a landscape is correlated to that site pattern. The first application of the approach (paper II) has shown a close correlation between PINUS pollen percentages and the distance weighted abundance of sandy soils and between BETULA pollen percentages and the distance weighted abundance of morainic till during the Allerød period, indicating that pine and birch formed rather separate stands on either substrate type. The cooling of the Younger Dryas induced significant changes in the vegetation of NE Germany. By combining pollen percentage and pollen accumulation rate data paper III identified a sharp vegetation boundary between the Mecklenburg and Brandenburg area at about 53 °N. The downscaling approach, here used with pollen accumulation rate data, suggests that in the North small tree stands could only exist in sheltered positions. The sharp vegetation boundary is possibly related to a climatic gradient and the southern permafrost limit, which itself may result from the formation of sea ice on the North Atlantic north of 53°N during winter. The warming of the Holocene again allowed the expansion of forests in the study area. Paper IV uses high resolution pollen (accumulation rate) data to study the successive forest formation, including the immigration of hazel, and explores vegetation patterns and composition during these successive stages using the extended downscaling approach. This approach addresses the problems related to differential pollen production, dispersal and the use of percentage data by applying simulations. It reveals that initially pine and birch established, as during the Allerød period, in largely separate stands with pine dominating on sandy soils and birch dominating on fine grained soils. Also open rich vegetation persisted, possibly due to seasonal drought, mainly on fine grained soils. Hazel later mainly spread on sites that received additional wetness from ground or surface water; it did not enter pine dominated forests on well drained sandy soils. Overall, the early Holocene vegetation of the study area was sharply differentiated by soil humidity and fertility. To conclude, present thesis has revealed vegetation patterns and species site preferences in NE Germany during three periods of the Lateglacial and early Holocene. The results improve our understanding of vegetation history in northern Central Europe, specifically for periods of rapid climate change. The approaches applied are flexible with respect to the type and quality of pollen data used and may be implemented using standard software packages.
In terms of climate change and climate change mitigation, the quantitative knowledge of global carbon pools is important information. On the one hand, knowledge on the amount of carbon cycling among – and stored in – global pools (i.e. Atmosphere, Biosphere, Cryosphere, Hydrosphere, and Lithosphere) may improve the reliability of models predicting atmospheric CO2 concentrations in terms of fossil fuel combustion. On the other hand, the carbon sequestration potential of specific ecosystems allows for estimating their feasibility regarding carbon trade mechanisms such as the Clean Development Mechanism or the Reducing Emissions from Deforestation and Degradation Program (REDD+). However, up to date, the majority of terrestrial carbon assessments have focused on forests and peatlands, leaving a data gap open regarding the remaining ecosystems. This data gap is likely to be explained by the relatively high carbon densities and/or productivities of forests and peatlands. Nevertheless, to get a precise as possible global picture, information on carbon pools and sequestration of other ecosystems is needed. Although desert ecosystems generally express low carbon densities, they may absolutely store a remarkable amount of carbon due to their large areal extent. In this context, Central Asian Deserts (in particular within the Turanian Deserts, i.e. Karakum, Kysylkum, Muyunkum) likely inhibit comparably high carbon pools as they express a sparse vegetation cover due to an exceptionally high annual precipitation if compared to the World’s deserts. In this dissertation, three important woody plant species – Populus euphratica and Haloxylon aphyllum and Haloxylon persicum – of Central Asian Deserts were investigated for their carbon pools and carbon sequestration potential. These species were chosen as they I) locally express high carbon densities, II) are dominant species, III) have a rather large spatial distribution, and IV) have experienced a strong degradation throughout the 20th century. Thus, they likely show a remarkable potential for carbon re-sequestration through restoration and thus for an application of carbon trade mechanisms (CHAPTER I). P. euphratica was investigated in the nature reserve Kabakly at the Amu Darya, Turkmenistan and in Iminqak at the Tarim He, Xinjiang, China. The assessment of Haloxylon species was restricted to the Turanian deserts west of the Tain Shan. To achieve a first scientific basis for large scale estimates, different methodologies, ranging from allometric formulas, over dendrochronology to remote sensing were combined (CHAPTERS II-V). In CHAPTER II allometric formulas were successfully developed for Haloxylon aphyllum and Haloxylon persicum and applied to six study sites distributed over the Turanian Deserts to represent the allometric variability of Haloxylon species in Central Asia. CHAPTER III derives another allometric formula (only based on canopy area) for H. aphyllum and combines it with a remote sensing analysis from the nature reserve Repetek. Thereby, a first large scale estimate covering the Northeastern Karakum Desert of carbon pools related to mono specific H. aphyllum stands is achieved. CHAPTER IV describes the wood structure of Populus euphratica forests in the nature reserve Kabakly (Turkmenistan) and in Iminqak (Xinjiang, China). In CHAPTER V a dendrochronological approach derives models for predicting the Net Primary Productivity (NPP) and the age of P. euphratica in the nature reserve Kabakly. Thereby, a first feasibility assessment regarding remote sensing analyses and the upscaling of the obtained NPP results is carried out. First estimates based on these local studies (CHAPTER VI), reveal carbon densities ranging from 0.1 – 26.3 t C ha 1 for the three investigated species. Highest maximum and median carbon densities were found for P. euphratica, but Haloxylon aphyllum expressed remarkable maximum carbon densities (13.1 t C ha-1), too. The total carbon pools were estimated at 6480 kt C for P. euphratica, 520 kt C for H. aphyllum stands and 6900 kt C for Haloxylon persicum shrubland. Accounting for the extent of degraded areas, the total re-sequestration potentials of the respective species were estimated at 4320 kt C, 1620 kt C and 21900 kt C, this highlighting the remarkable absolute re-sequestration potential of H. persicum shrubland despite its low average carbon densities. In the end, the main results were put into a broader context (CHAPTER VI), discussing the general feasibility of reforestations both in ecological terms as well as in terms of carbon trade mechanisms. A short example highlights the strong connection between the feasibility of reforestations and the global carbon market. Finally, open research questions are brought forth revealing the yet large research potential of Central Asian Desert ecosystems in general and in terms of carbon sequestration.