>>Department of Ecological Modelling >> Personal homepage Thorsten Wiegand >> Research
Main research projects:
Gaucho and sheep in the Patagonean steppe between Governador Costa and Tecka
Declining biodiversity, land degradation, global climate change and in particular desertification are subjects of worldwide concern. In arid and semiarid plant communities, the mismatch between observation times (years) and time scales of vegetation change (centuries), event driven dynamic behaviour, unpredictable and low rainfall and complicated interactions between species make it especially difficult to fully understand their long-term dynamics. The effects of grazing or sparing management on such natural communities of long-lived plants generally take decades to become evident, and it is difficult to assess probabilities and time scales of vegetation change, and thus to develop sustainable management. A promising way to deal with these problems is to combine long-term monitoring, the indispensable basis of understanding and management, with other approaches such as experiments and modelling studies.
Ecological modeling has witnessed the rapid development of advanced new scientific tools, notably individual-based, spatially explicit population models, which offer scope for quantitatively exploring the long-term vegetation dynamics of arid and semiarid ecosystems. Although there is usually little long-term field data available on the full dynamics of semiarid plant communities, information on life-history attributes and interactions between individual species are relatively easy to observe on shorter time-scales. The basic idea of this bottom-up approach is to incorporate short-term knowledge in form of mechanistic rules into a model and to simulate the fate and the interactions of individual plants within the plant community. By using long-term climatic data and plausible management scenarios, the model extrapolates from the known short-term behaviour of individual plants to long-term vegetation dynamics. An important advantage of advantage of these type of simulation models is that they include the essential biological information in the form of structural realism (rules defining our current knowledge on the biology of the species) rather than mathematical equations, and thus avoid problems with deformation or mutilation of information due to mathematical constraints, or problems with abstract compound model parameters that have no direct biological meaning. Especially in more complex problems and applied case studies, this allows for the direct inclusion of the biological knowledge and real-world spatial information in the model. The combined approach of long-term field research and carefully constructed, spatially explicit simulation models has the potential to gain the best possible understanding about natural long-term dynamics and the possible consequences of anthropogenic impact.
Modeling spatial and temporal dynamics of plant communities
In arid regions, the effects of grazing or sparing management on natural communities of long-lived plants generally take decades to become evident. Event driven dynamic behavior, unpredictable and low rainfall and complicated interactions between species make it difficult to assess probabilities and time scales of vegetation change, and thus to develop sustainable management. To gain a better understanding of the main processes and mechanisms involved in vegetation change of the Karoo shrubland, we have developed a spatially explicit individual based model that simulates changes in plant communities over long time spans. (Wiegand et. al 1995, 1996, 1997, 1998, 1999, 2000)
The aim of this project is to obtain an understanding of the long-term dynamics of the semiarid grasslands in South Africa, and to quantify the relative roles of grazing and drought on long-term vegetation patterns. The model refines the techniques of individual-based and spatially-explicit modeling I developed for the Karoo shrubland model, especially through consequent application of the pattern-oriented modeling modeling strategy for adapting the model optimal to the data. We model the dynamics of four well studied species representing the basic functional groups in this grasslands which differ in life-history traits and their response to grazing. The four species include three perennial grasses Cymbopogon plurinodis, Themeda triandra, and Eragrostis curvula, and the annual Aristida congesta . The model combines information on life-history traits of the species, competitive interactions between species, and moisture availability with long-term rainfall data and management scenarios to simulate grassland dynamics over several decades. The model is a grid based model where grass tufts can cover several cells and competition is modeled with an expaned zone of influence approach. (Wiegand et al. 2004a)
The objective of this project is to gain an understanding of the processes of desertification and rehabilitation in the shrub-grass steppe of western Patagonia, and especially to investigate the long-term effect of grazing and climatic fluctuations on vegetation dynamics. The individual-based and spatially-explicit model is closely related to the grassland model and makes extensive use of the pattern-oriented modeling modeling strategy. A water balance model (Dinaqua) calculates from data on daily rainfall and plant cover (biomass) total transpiration of grasses and shrubs, and the water status of the upper and deeper layers (Paruelo and Sala, 1995). This information is used by the spatial explicit vegetation dynamics model that simulates the processes seed production, germination, establishment, competition and facilitation, growth, and mortality for the 3 dominant shrub species Mulinum spinosum, Adesmia campestris, and Senecio filaginoides, and for the three grass species Stipa speciosa, humilis, Poa ligularis, and Poa lanuginosa. A spatial pattern analysis of the shrub part of the community is presented in Wiegand et al. (2006). P. Ciprioti is working on integrating the small-scale model into a regional larger-scale model.
Reduced cover of the dominant native grass species Festuca pallescens is one of the major and widespread aspects of degradation in Patagonian Festuca grasslands. We study the grassland response to livestock grazing based on tussock level processes including phytomass and vegetation dynamics, as well as defoliation by grazing livestock. The aim of this project is to reach a quantitative understanding of the grazing impacts on vegetation dynamics at the landscape scale. Based on our understanding of tussock level vegetation dynamics and controls of the defoliation regime, we use a grid-based spatially explicit stochasitc simulation model for upscaling to landscape level of vegetation dynamics. Ultimately, we want to study the long-term effects of livestock grazing with respect to two groups of factors: factors open to direct management (utilization intensity, timing of defoliation), and non-manageable factors (grazing selectivity and heterogeneity). Next to understanding the links of livestock grazing and degrading changes in vegetation cover and structure, opportunities and timescales for recovery are of interest. See Paruelo et al. (2008).
The Patagonian steppes in Argentine are endangered by overgrazing. Landuse modifies the vegetation structure and promotes soil erosion leading to the desertification of these ecosystems. Although these changes have been noticed since the 1950s, processes, conditions, and time scales which lead to the degradation of vegetation in Patagonia are not well understood. The pathways of degradation and fragility of the main ecosystem types (meadows, grass steppes and shrub steppes) differ among them. Moreover, changes in one particular unit is supposed to be highly dependent on the structure of the landscape, i.e. type and proportion of the other ecosystem types. The aim of this project is to identify how climate variability, grazing management and landscape structure affect the degradation of single ecosystem types. For this purpose we develop a simulation model at the spatial scale of large paddocks (1000 to 8000 hectares) that integrates, in a spatially explicit way, the three small scaled simulation models on the grass-shrub steppe, the Festuca grassland, and meadow, which were developed independently for each ecosystem type.
|Modified: 05.10.2008||Resp.: Thorsten Wiegand||webmaster|