Modeling the spatio-temporal dynamics of semiarid plant communities

CIPRIOTTI, P.A.; WIEGAND, T.

Simposio; Simposio Internacional EcoDesert sobre Geoecología y Desertificación - Joan Puigdefabregas; 2019

Arid and semiarid plant communities are under increasing pressure due to human land use activities and climatic variability in combination with complex ecosystem dynamics. According to the United Nations, approximately 70% of all dry lands is endangered by desertification. This raises concerns about their future ability to provide critical ecosystem services. Designing appropriate management requires a general understanding of the underlying mechanisms that drive ecosystem dynamics as well as detailed case-specific data-driven analyses. Modeling is crucial to this effort because it can enhance our understanding of the processes governing community dynamics and may allow us to predict how semiarid plant communities respond to land use (e.g. grazing) and climate change. However, understanding patterns and processes of vegetation dynamics in semiarid rangelands is an inherently difficult task: they are subject to internal and external stochasticity, community dynamics may be complex and event-driven, data are usually sparse, and the scales of data collection are often the management relevant scales. Additionally, capturing the particularities of the natural history of the system can be challenging. As a consequence, modeling is often case-driven with limited powers of generalization and no general theories akin to those in physics exist. These problems are not unique to modeling semiarid plant communities, but reflect general issues in modeling. There are ongoing discussion regarding (i) the appropriate level of detail required for modeling and understanding ecological systems, (ii) how to parameterize such models, and (iii) how to upscale the models from tractable, small-scale observations and experiments to prediction of large-scale patterns. I will present in my keynote some solutions to each of these points.