Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Australie → 2007 → Development of Processes and Tools to Support Adaptive Management in Complex Rangelands Systems

University of Queensland (2007)

Development of Processes and Tools to Support Adaptive Management in Complex Rangelands Systems

Bashari, Hossein

Titre : Development of Processes and Tools to Support Adaptive Management in Complex Rangelands Systems

Auteur : Bashari, Hossein

Grade : Doctor of P0hilosophy PhD (2007)

Université de soutenance : University of Queensland. School of Natural and Rural Systems Management

This thesis begins by identifying appropriate indicators (compositional and functional) of grazing impact for the subtropical grasslands in south-east Queensland that will help landholders to better evaluate rangeland condition. Species composition, Landscape Function Analysis (LFA), and visual condition assessments were used to assess grazing impact at 70 sites in the Ironbark-Spotted Gum Woodland of Gatton. Multivariate analyses, including Principal Component Analysis (PCA) and Multidimensional Scaling (MDS), were used to identify species’ responses on the grazing gradient. Regression techniques were used to explore the correlation between LFA functional indices (stability, infiltration and nutrient cycling) and grazing gradient. The frequencies of six tall tussock, short tussock and lawn species, and the LFA stability index were found to be effective indicators of grazing impact in the study area. The slow response of soil function indices to grazing pressure indicates that the Ironbark-Spotted Gum Woodland examined in this study is a fairly “robust” and stable system with a high resilience to grazing. Although the species composition of the grassland changes under the stress of selective or heavy grazing, trampling and nutrient enrichment, there is no appreciable loss of landscape functions such as nutrient cycling and infiltration. Hence, there was no evidence of major deterioration and degradation. Methods of modelling rangeland dynamics were reviewed to determine their strengths and weaknesses with respect to assessing the effects of management on vegetation change. State and Transition Models (STMs) were found to provide a versatile way of describing vegetation dynamics. An iterative model development process was used to construct an STM for the Ironbark-Spotted Gum Woodland of the study area. This process utilised multiple information sources to identify possible vegetation states and transitions, including experiential knowledge of scientists and land managers familiar with the study area, and results from multivariate analysis to characterise vegetation states in terms of species composition and soil condition. The proposed STM is a three dimensional model showing five vegetation states and seventeen transitions, mapped according to grazing intensity, grassland palatability and soil nutrient content. Despite the advantage of STMs, they are traditionally descriptive diagrams and are unable to be used for predictive modelling and scenario analysis. They also handle uncertainty associated with causes of vegetation change poorly. A Bayesian Belief Network (BBN) was used to assist in the development of a dynamic and predictive STM by providing a graphical modelling framework for building a probability-based cause and effect model. The results indicated that the BBN approach is a highly useful mechanism for adding value to descriptive STMs. First, it allowed the uncertainty in transitions to be expressed by using probabilistic relationships. Second, the approach provided a scenario and sensitivity analysis tool for both scientists and landholders to assess the probable vegetation outcomes of rangeland management decisions, and to identify those management options most likely to improve or degrade vegetation condition. Third, it is particularly complementary to the adaptive management process, because monitoring records can be used to update probability relationships within the BBN model over time. Therefore, the modelling approach supported the planning, monitoring and review steps of the adaptive management cycle. This is an advantage over current rangeland management simulation models that are good at supporting management planning through their predictive capabilities, but poor at supporting monitoring and evaluation steps. Sensitivity analysis using the BBN model constructed for the Ironbark -Spotted Gum Woodland STM highlighted that grazing pressure is the main factor driving almost all transitions. The stocking rate has a great influence on grazing pressure, but drought and the use of dry season supplements magnifies the influence of stocking rate on grazing pressure. Selective grazing is an important factor determining the transitions from or to unpalatable tall tussock grasses. The model also indicates that drought and the occurrence of good seasons, which were the two factors beyond the control of land managers, have a clear effect on the vegetation state via their direct impact on the recovery of tall tussock grasses and their indirect impact on the frequency of fire. Soil nutrient content is an important variable that influences transitions to and from lawns. The thesis concludes by applying the concepts and approaches used in this study to a rangeland area in Iran, in order to explore their applicability in developing a decision support tool (DST) for another area and under different conditions. Particular focus was on identifying potential stumbling blocks to the development and use of the rangeland management DST in another area. The results suggest that the method is transferable to other situations and ecosystems. In Iran, however, the lack of a long-term rangeland monitoring program, which can provide the data needed to update the Bayesian Network Model over time and support adaptive management, was identified as a key stumbling block towards successful implementation of the DST.


Version intégrale (pdf) ADT (program)

Page publiée le 29 février 2008, mise à jour le 30 mai 2017