Complex Adaptive Systems (CAS) An innovative method to work integrated towards sustainable systems Bas Kemp, Arnold Bregt and Tim Verwaart Theme coordinators CAS Wageningen UR (University & Research centre)

Welcome Introduction Why do we want CAS? What are the principles of CAS? Some examples of what you can do with CAS Udder Health Tail biting pigs Conclusion

Why do we want a CAS systems approach? We want to develop sustainable production systems Livestock Farming with Care (Scholten et al. NJAS, 2013)

Attitudes However, systems development is complex Heterogeneous stakeholder groups Within groups heterogeneous ‘agents’ with different underlying drivers ‘Interaction & adaptation’ between stakeholders. This leads to ‘no, partial or total’ adaptation for new systems developments. ‘Agent based modelling’ as model tool in CAS Heterogeneous stakeholders is modelled as an ‘agent’ with distinct drivers (‘identity’) In time ‘agents’ influence each other so they are adaptive In time this leads to new developments (emergence) Attitudes dependent on subjective beliefs about: measures effect, associated economic implications and penalties reinforced by communication campaign

So, systems development is complex Heterogeneous stakeholder groups Within groups heterogeneous ‘agents’ with different underlying drivers ‘Interaction & adaptation’ between stakeholders. This leads to ‘no, partial or total’ adaptation for new systems developments.

‘Agent based modelling’ as model tool in CAS Heterogeneous stakeholders is modelled as an ‘agent’ with distinct drivers (‘identity’) In time ‘agents’ influence each other so they are adaptive In time this leads to new developments (emergence) farmer

An example

Udder health management improvement Processors and governments are interested in motivating farmers to improve udder health Tools: penalty system based on BMSCC, communication campaign Implementation of various measures on the farm E.g. using gloves, milking mastitis cows last Valeeva and Verwaart., ICUHC, 2011

Objective To explore the extent of adoption of desirable measures over time and the influence of drivers of farmers’ decision making and external tools on this process

Farmers are modelled to be diverse ABM as integration of available knowledge on adoption of udder health measures - Modelling multiple heterogeneous dairy farmers (agents) with simple behavioural rules of action and interaction, so as to elicit emerging collective adoption behaviour

Conclusions To get adoption of measures to decrease mastitis you need: Economic incentives Active networks of farms for low motivated farmers At low economic incentive levels intense communication campaigns

An example

Modelling tail biting behaviour in pigs Fattening pigs & tail biting behaviour Complex interplay housing-pig-farmer No existing models for analysis of pig behaviour and related welfare & production issues Boumans et al. ISAE Benelux, 2013

Agents: Variable nr of fattening pigs Environment: Barren/enriched pen Space variation Feeder(s)

Results: effect of density on tail biting behaviour Tale biting outbreaks in pigs: Can be explained by increased by underlaying drivers that lead to frustation Farmers can learn what factors are involved in frustration(e.g. stocking rates) Develop a new model based on important behavioural motivations related to tail biting behaviour of pigs: Based on a conceptual model of stress as a state of motivational systems: describing internal & external stimuli causing motivation and feedback pathways (Jensen and Toates, 1997) Including feeding motivation, exploration motivation, coping with stress, biting motivation Use model to: Gain insight in mechanisms behind pig behaviour Test hypothesis of 3 types and mechanisms of tail biting behaviour (Taylor et al., 2011) Each variation simulated in 10 runs, every run 2000 time steps

Tail biting outbreaks in pigs: Can be explained by underlying drivers that lead to frustration Farmers can learn what factors are involved in frustration(e.g. stocking rates) External factors Internal factors Motivation Develop a new model based on important behavioural motivations related to tail biting behaviour of pigs: Based on a conceptual model of stress as a state of motivational systems: describing internal & external stimuli causing motivation and feedback pathways (Jensen and Toates, 1997) Including feeding motivation, exploration motivation, coping with stress, biting motivation Use model to: Gain insight in mechanisms behind pig behaviour Test hypothesis of 3 types and mechanisms of tail biting behaviour (Taylor et al., 2011) Behaviour

So systems are complex and many agents with different drivers influencing the development of sustainable systems With CAS (using ABM) we: Learn how heterogeneity in agents lead-to or hamper emergence of sustainability Learn which underlying drivers support or hamper emergence (virtual lab) Can connect the different aggregation levels (instrumentation of integration social and technical domain) Can be more successful in introducing sustainable system innovations So systems are complex and many agents with different drivers influencing the development of sustainable systems With CAS (using ABM) we learn how heterogeneity in agents lead-to or hamper emergence of sustainability We learn which underlying drivers support or hamper emergence (virtual lab) We can connect the different aggregation levels (instrumentation of integration social and technical domain) Use of CAS can lead to more successful introduction of sustainable system innovations

Thanks for your attention Complex Adaptive Systems An innovative method to work integrated towards sustainable systems

Wageningen UR Systems Approach