際際滷

際際滷Share a Scribd company logo

Dr Lael Parrott at the Landscape Science Cluster Seminar, May 2009

Download as PPT, PDF
P
pdalby

This document summarizes how concepts from complex systems studies can inform natural resource management. It discusses how ecosystems and landscapes are complex systems with emergent properties arising from local interactions. Agent-based models are useful for modeling ecological complexity across scales. Examples shown include a model of grassland resilience under disturbance and the relationship between grazing and spatial complexity. Understanding community assembly is explored through a spatial model linking local communities. The document concludes that embracing complexity requires new tools like multi-scale models and monitoring to manage social-ecological systems.

1 of 38
Download to read offline
Concepts & methods from complex systems studies that can help inform NRM Dr. Lael Parrott Director, Complex Systems Laboratory G辿ographie, Universit辿 de Montr辿al, Canada [email_address] http:// www.geog.umontreal.ca/syscomplex Presented to the Landscape Science Cluster, University of Adelaide, 28 May 2009.
How do we deal with complexity in the design, management & restoration of ecosystems and landscapes? Ecosystems and landscapes are complex systems
APPLICATIONS ecosystem & landscape design, management & restoration environmental indicators THEORY ecological complexity METHODS ecological informatics (models, analysis, data management & visualisation) Complex Systems Laboratory Research Program
What is a complex system? Complexity: Implications & challenges for natural resource management Modelling ecological complexity Examples : Modelling ecological resilience; Understanding community assembly Conclusion Outline
Complex Systems
What is a complex system? The whole is more than the sum of the parts. A system composed of multiple interacting elements having a comportment that is difficult to analyse or describe using only one scale or resolution. (Parrott, 2002. Trans. of the ASAE .)
What is a complex system?
What is a complex system? Locally interacting entities Emergent structures & processes Feedback What is a complex system?
Complex Systems Key characteristics Hierarchy and scaling Emergence Self-organisation Uncertainty Memory Adaptation
Accepting complexity means accepting that ecosystems and landscapes are Dynamic and subject to rapid change; Infused with memory of past events; Self-organised in space and time. (Parrott & Li, 2005, Ecological Society of America Evening Session on Ecological Complexity ) Complexity Implications for natural resource management
Integrating concepts from complex systems research into management paradigms (e.g., acceptance of multiple stable states; harnessing self-organisation; dealing with uncertainty) Determining how to treat multi-scale interactions Development of new methods to monitor and describe complex structures and dynamics Complexity Challenges for natural resource management
Modelling Ecological Complexity
Use a 束bottom-up損 approach: Model local interactions and let the higher level dynamics emerge Individual- and agent-based models Importance of space Modelling ecological complexity Considerations
Example: Modelling ecological resilience
Model: WIST (Weather-driven, Individual-based, Spatially-explicit, Terrestrial ecosystem model) Fully configurable, multi-species, multi-trophic individual-based model Includes environment (soil, atmospheric conditions), topography & weather Landscape is modelled as a grid MPI version available for parallel processing infrastructures (Parrott & Kok 2001, 2002 in Ecological Modelling ) Modelling ecological resilience
In WIST: Animal individuals: respire; search for food; eat; move about; reproduce Plant individuals: grow; photosynthesize; disperse seeds Modelling ecological resilience
WIST is a virtual ecosystem in which... Processes and interactions are deterministic and rule-based Complex, spatiotemporal dynamics emerge at the landscape level Modelling ecological resilience
Initial conditions: 500m x 500m surface 50 species of annual & perennial grasses and herbs 3 herbivore species Simulations for 100 years at 10 minute time steps Introduce 束clear-cuts損 randomly over the landscape (Parrott, 2004, Ecological Complexity) Case 1: Study of the resilience of a grassland ecosystem subject to frequent disturbance Modelling ecological resilience
Ecosystem biomass: spatiotemporal dynamics
Case 2: Relationship between grazing intensity and spatiotemporal complexity in a model ecosystem Initial conditions: 30 species of annual & perennial grasses and herbs Introduce herbivores (rabbits) Vary grazing intensity (# of herbivores and duration) 50-year simulations (Parrott, 2004, ESA Annual Meeting) Modelling ecological resilience
Modelling ecological resilience Measuring spatiotemporal complexity Dataset: biomass recorded monthly for 10mx10m grid cells
Modelling ecological resilience (Parrott, 2005, Ecological Complexity; Parrott et al., 2008, Ecological Informatics . ) Measuring spatiotemporal complexity
Results over many simulations: Grazing intensity (# of herbivores) Grazing duration (years) Spatiotemporal complexity of vegetation biomass (Parrott, 2004, ESA Annual Meeting) Modelling ecological resilience
Example: Understanding community assembly through modelling
Understanding community assembly Species Interaction
Understanding community assembly Spatial version of the individual-based 束Tangled Nature損 model of multi-species community dynamics Doctoral project of lise Filotas; Collaborators: Martin Grant, Physics, McGill University & Per Rikvold, Physics, Florida State University. Separate communities linked via dispersal (Filotas et al., Ecological Complexity , 2008)
Understanding community assembly dispersal Regional species pool (2 20 species; random interaction web) random species introductions Landscape of locally connected communities (typical grid size: 128 x 128 cells) Species in potentia Local community ( n interacting species)
Understanding community assembly Spatial patterns of community similarity (Filotas et al., in revision )
Understanding community assembly low dispersal regional species pool high dispersal Dispersal rate Dispersal rate Fraction of interacting pairs Structure of the interaction webs (Filotas et al., in revision )
Understanding community assembly Dispersal rate In this model... community and landscape level properties emerge from local level inter-species interactions. We demonstrate... the important role of biotic interactions in structuring communities and the emergence of mutualist-dominated interaction webs in low-dispersal landscapes; the role of facilitation as a building block for more biologically diverse communities.
Understanding community assembly Using modelling to find optimal community assembly sequences (C担t辿 & Parrott, 2005; C担t辿, Parrott & Sabourin, 2007)
Understanding community assembly
Understanding community assembly Using multi-criteria optimisation methods & genetic algorithms, we can find assembly sequences that maximize diversity & productivity while respecting constraints on food web properties such as connectance. (C担t辿 & Parrott, 2005; C担t辿, Parrott & Sabourin, 2007) Important implications for ecosystem design & restoration projects.
Conclusions
Embracing complexity requires new tools for natural resource management; models, monitoring programmes New modelling methods suited for studying complex ecological interactions across multiple scales are being developed; can simulate emergent properties in space and time Managing future landscapes and ecosystems as complex, coupled social-ecological systems will require multidisciplinary teams, combining approaches from the natural, applied and human sciences. Conclusions
Modelling of human and ecological components of landscapes for decision support in the Boreal Forest & the Saint Lawrence River estuary. - Come to the Place & Purpose symposium to learn more! Future Landscapes: Managing within complexity by L. Parrott and W. Meyer, manuscript in preparation. Work in progress...
Acknowledgements Natural Sciences and Engineering Research Council of Canada Les Fonds Qu辿b辿cois sur la Recherche en Nature et Technologies Canadian Foundation for Innovation Qu辿bec Supercomputing Network ALL of the students in the Complex Systems Laboratory, past & present.
Concepts & methods from complex systems studies that can help inform NRM Dr. Lael Parrott Director, Complex Systems Laboratory G辿ographie, Universit辿 de Montr辿al, Canada [email_address] http:// www.geog.umontreal.ca/syscomplex Presented to the Landscape Science Cluster, University of Adelaide, 28 May 2009.

More Related Content

Dr Lael Parrott at the Landscape Science Cluster Seminar, May 2009

  • 1. Concepts & methods from complex systems studies that can help inform NRM Dr. Lael Parrott Director, Complex Systems Laboratory G辿ographie, Universit辿 de Montr辿al, Canada [email_address] http:// www.geog.umontreal.ca/syscomplex Presented to the Landscape Science Cluster, University of Adelaide, 28 May 2009.
  • 2. How do we deal with complexity in the design, management & restoration of ecosystems and landscapes? Ecosystems and landscapes are complex systems
  • 3. APPLICATIONS ecosystem & landscape design, management & restoration environmental indicators THEORY ecological complexity METHODS ecological informatics (models, analysis, data management & visualisation) Complex Systems Laboratory Research Program
  • 4. What is a complex system? Complexity: Implications & challenges for natural resource management Modelling ecological complexity Examples : Modelling ecological resilience; Understanding community assembly Conclusion Outline
  • 6. What is a complex system? The whole is more than the sum of the parts. A system composed of multiple interacting elements having a comportment that is difficult to analyse or describe using only one scale or resolution. (Parrott, 2002. Trans. of the ASAE .)
  • 7. What is a complex system?
  • 8. What is a complex system? Locally interacting entities Emergent structures & processes Feedback What is a complex system?
  • 9. Complex Systems Key characteristics Hierarchy and scaling Emergence Self-organisation Uncertainty Memory Adaptation
  • 10. Accepting complexity means accepting that ecosystems and landscapes are Dynamic and subject to rapid change; Infused with memory of past events; Self-organised in space and time. (Parrott & Li, 2005, Ecological Society of America Evening Session on Ecological Complexity ) Complexity Implications for natural resource management
  • 11. Integrating concepts from complex systems research into management paradigms (e.g., acceptance of multiple stable states; harnessing self-organisation; dealing with uncertainty) Determining how to treat multi-scale interactions Development of new methods to monitor and describe complex structures and dynamics Complexity Challenges for natural resource management
  • 12. Modelling Ecological Complexity
  • 13. Use a 束bottom-up損 approach: Model local interactions and let the higher level dynamics emerge Individual- and agent-based models Importance of space Modelling ecological complexity Considerations
  • 14. Example: Modelling ecological resilience
  • 15. Model: WIST (Weather-driven, Individual-based, Spatially-explicit, Terrestrial ecosystem model) Fully configurable, multi-species, multi-trophic individual-based model Includes environment (soil, atmospheric conditions), topography & weather Landscape is modelled as a grid MPI version available for parallel processing infrastructures (Parrott & Kok 2001, 2002 in Ecological Modelling ) Modelling ecological resilience
  • 16. In WIST: Animal individuals: respire; search for food; eat; move about; reproduce Plant individuals: grow; photosynthesize; disperse seeds Modelling ecological resilience
  • 17. WIST is a virtual ecosystem in which... Processes and interactions are deterministic and rule-based Complex, spatiotemporal dynamics emerge at the landscape level Modelling ecological resilience
  • 18. Initial conditions: 500m x 500m surface 50 species of annual & perennial grasses and herbs 3 herbivore species Simulations for 100 years at 10 minute time steps Introduce 束clear-cuts損 randomly over the landscape (Parrott, 2004, Ecological Complexity) Case 1: Study of the resilience of a grassland ecosystem subject to frequent disturbance Modelling ecological resilience
  • 19. Ecosystem biomass: spatiotemporal dynamics
  • 20. Case 2: Relationship between grazing intensity and spatiotemporal complexity in a model ecosystem Initial conditions: 30 species of annual & perennial grasses and herbs Introduce herbivores (rabbits) Vary grazing intensity (# of herbivores and duration) 50-year simulations (Parrott, 2004, ESA Annual Meeting) Modelling ecological resilience
  • 21. Modelling ecological resilience Measuring spatiotemporal complexity Dataset: biomass recorded monthly for 10mx10m grid cells
  • 22. Modelling ecological resilience (Parrott, 2005, Ecological Complexity; Parrott et al., 2008, Ecological Informatics . ) Measuring spatiotemporal complexity
  • 23. Results over many simulations: Grazing intensity (# of herbivores) Grazing duration (years) Spatiotemporal complexity of vegetation biomass (Parrott, 2004, ESA Annual Meeting) Modelling ecological resilience
  • 24. Example: Understanding community assembly through modelling
  • 25. Understanding community assembly Species Interaction
  • 26. Understanding community assembly Spatial version of the individual-based 束Tangled Nature損 model of multi-species community dynamics Doctoral project of lise Filotas; Collaborators: Martin Grant, Physics, McGill University & Per Rikvold, Physics, Florida State University. Separate communities linked via dispersal (Filotas et al., Ecological Complexity , 2008)
  • 27. Understanding community assembly dispersal Regional species pool (2 20 species; random interaction web) random species introductions Landscape of locally connected communities (typical grid size: 128 x 128 cells) Species in potentia Local community ( n interacting species)
  • 28. Understanding community assembly Spatial patterns of community similarity (Filotas et al., in revision )
  • 29. Understanding community assembly low dispersal regional species pool high dispersal Dispersal rate Dispersal rate Fraction of interacting pairs Structure of the interaction webs (Filotas et al., in revision )
  • 30. Understanding community assembly Dispersal rate In this model... community and landscape level properties emerge from local level inter-species interactions. We demonstrate... the important role of biotic interactions in structuring communities and the emergence of mutualist-dominated interaction webs in low-dispersal landscapes; the role of facilitation as a building block for more biologically diverse communities.
  • 31. Understanding community assembly Using modelling to find optimal community assembly sequences (C担t辿 & Parrott, 2005; C担t辿, Parrott & Sabourin, 2007)
  • 33. Understanding community assembly Using multi-criteria optimisation methods & genetic algorithms, we can find assembly sequences that maximize diversity & productivity while respecting constraints on food web properties such as connectance. (C担t辿 & Parrott, 2005; C担t辿, Parrott & Sabourin, 2007) Important implications for ecosystem design & restoration projects.
  • 35. Embracing complexity requires new tools for natural resource management; models, monitoring programmes New modelling methods suited for studying complex ecological interactions across multiple scales are being developed; can simulate emergent properties in space and time Managing future landscapes and ecosystems as complex, coupled social-ecological systems will require multidisciplinary teams, combining approaches from the natural, applied and human sciences. Conclusions
  • 36. Modelling of human and ecological components of landscapes for decision support in the Boreal Forest & the Saint Lawrence River estuary. - Come to the Place & Purpose symposium to learn more! Future Landscapes: Managing within complexity by L. Parrott and W. Meyer, manuscript in preparation. Work in progress...
  • 37. Acknowledgements Natural Sciences and Engineering Research Council of Canada Les Fonds Qu辿b辿cois sur la Recherche en Nature et Technologies Canadian Foundation for Innovation Qu辿bec Supercomputing Network ALL of the students in the Complex Systems Laboratory, past & present.
  • 38. Concepts & methods from complex systems studies that can help inform NRM Dr. Lael Parrott Director, Complex Systems Laboratory G辿ographie, Universit辿 de Montr辿al, Canada [email_address] http:// www.geog.umontreal.ca/syscomplex Presented to the Landscape Science Cluster, University of Adelaide, 28 May 2009.