University of Natural Resources and Life Sciences, Vienna (BOKU) - Research portal
Gewählte Doctoral Thesis:
Grzegorz Gruszczynski
(2007):
ADAPTATION AND APPLICATION OF AGROECOSYSTEM MODELS TO DROUGHT RELATED PROBLEMS IN AGRICULTURAL GRASSLAND PRODUCTION.
Doctoral Thesis,
BOKU-Universität für Bodenkultur,
pp 253.
UB BOKU
obvsg
- Abstract:
- The content of this study was the adaptation and application of agroecosystem models to drought related problems in Austrian grassland production. Crop models, drought indices and information derived from meteorological observations can contribute to improved use of resources, assessment of agricultural production risk, assessment of potential climate change impacts, identification of sustainable use of water reserves in agriculture and more. Drought indices, crop models and backpropagation neural networks were tested, compared and partly adapted for grassland at the experimental stations in Gumpenstein, Piber and Kirchberg am Walde in Austria. The most suitable model, the MACROS model, was adapted and improved and calibrated and validated for the experimental sites. The adapted MACROS model was named GGLawn. Specific attention was paid to simulation of drought effects on grassland yield and to a user friendly software. The results show that grassland growth under optimal and water-limited conditions can be simulated reasonably well with GGLawn. GGLawn was successfully validated to estimate grassland dry matter at Gumpenstein (RSq= 0,82). However, when the calibrated program was applied to other locations the quality of results was reduced (Piber RSq= 0.48 and Kirchberg am Walde RSq=0.31). GGLawn seems to be a useful tool to investigate in an integrated manner the sensitivity of grassland to overall effects of climate change if calibrated for a location. The model is able to simulate ecosystem processes over the whole year, including wintertime, and can be a useful tool to investigate responses of grassland to climate change scenarios. However, there are limitations, such as the consideration of various grassland species and of multi-year effects, which should be reduced to improve validation results.