Comparing stakeholders’ representations of social-ecological systems: mental models of the
functioning of contrasted French rural landscapes
Carole Vuillot1* and Raphaël Mathevet1
1CEFE CNRS MONTPELLIER, FRANCE
* Presenting author: [email protected]
1
AAG Annual Meeting
21st APRIL 2015
Chicago, Illinois
Evolution of European farmed landscapes
• Intensification of agricultural systems and practices
Landscape simplification
Biodiversity loss
Evolution of European farmed landscapes
• Intensification of agricultural systems and practices
Landscape simplification
Biodiversity loss
• EU Common Agricultural Policy
Land sparing
FamLand research projet
Cro
p co
mpo
sitio
n
Crop configuration
Armorique
Plaine-Val de Sèvre
East Anglia Goettingen
Camargue
Gascony Valleys and Hills Lleida
Ontario
https://farmland-biodiversity.org
?
FamLand research projet
Cro
p co
mpo
sitio
n
Crop configuration
https://farmland-biodiversity.org
?
?
How to deal with complex social-ecological systems ?
• Agricultural landscapes = social-ecological systems
• You can’t manage or care for interactions you don’t perceive
• Conservation policies in farmed landscape inefficient due to lack of relevance to on-ground context
“The soundness of a politic often lies upon our ability to understand the various types of knowledge and worldviews of the relevant stakeholders “ (Ostrom 2005)
C. Vuillot 2013
eliciting farmers’ world views
Three contrasted landscapes
Gascony Valleys and Hills Plaine et val de Sèvre Camargue
A gradient of territorial coordination and institutionalization of biodiversity conservation
Hypotheses
consensus on the system
Integrated vision of links between nature and agriculture
European N2000 area Regional Nature Park
Armorique
Plaine-Val de Sèvre
East Anglia Goettingen
Camargue
Gascony Valleys and Hills Lleida
Ontario
Three contrasted landscapes
Gascony Valleys and Hills Plaine et val de Sèvre Camargue
A gradient of territorial coordination and institutionalization of biodiversity conservation
EU N2000 site Regional Nature Park
Three contrasted landscapes
Gascony Valleys and Hills Plaine et val de Sèvre Camargue
A gradient of territorial coordination and institutionalization of biodiversity conservation
Hypotheses
consensus on the social-ecological system
Integrated vision of links between nature and agriculture
EU N2000 site Regional Nature Park
Theoretical background: a mental model approach
Adapted from Clayton and Myers (2009) , Cosquer (2012)
Biodiversity
Behavior/ practices
Mental model
Local social interactions
Natural and economical assets and disadvantages
• Internal representation of an external system • Individual cognitive structure
• socially elaborated • on which , reasoning, attitudes, and behavior are based
Carley and Palmquist (1992) Lynam and Brown (2011)
Hoffman (2015) Agricultural landscape
Affective components
attachment – fear - denial
Socio-economic background
Social norms values
Cognitive components Knowledge – beliefs
Theoretical background: a mental model approach
Adapted from Clayton and Myers (2009) , Cosquer (2012)
Biodiversity
Behavior/ practices
Mental model
Local social interactions
Natural and economical assets and disadvantages
• Internal representation of an external system • Individual cognitive structure
• socially elaborated • on which , reasoning, attitudes, and behavior are based
Carley and Palmquist (1992) Lynam and Brown (2011)
Hoffman (2015) Agricultural landscape
Affective components
attachment – fear - denial
Socio-economic background
Social norms values
Cognitive components Knowledge – beliefs
Network of linked concepts
Mental models elicitation technique
• Iterative elicitation of a conceptual model of the system functioning :
Özesmi et Özesmi (2004)
Mathevet et al. (2011)
Etienne et al. (2011) C. Vuillot. 2014
How would you describe the functioning of the agricultural landscape ?
– « Stakeholders»
– « Biophysical Components»
– « Drivers of change »
– « Interactions» : directed arrow + verb
Individual concept map (ICM)
TO
Crops Farmers …
Crops 0 0 …
Farmers 1 0 …
… … … …
FRO
M
ICM
Aggregated ICMs
Analysis of directed graphs and group conceptual maps
Transmitter variables (T) forcing function, givens
Receiver variables (R) utility variables, ends
Ordinary variables (O) means
Bougon et al., (1977) Vanwindekens et al., (2014)
R/T ratio
Crops Farmers …
Crops 0 1 …
Farmers 1 0 …
… … … …
Crops Farmers …
Crops 0 1 …
Farmers 1 0 …
… … … …
GCM = group concept map
Concept categorization
Individual adjacency matrices
Sampling
Gascony Valleys and Hills Plaine et val de Sèvre Camargue n = 29 n=30 n=30
Fert 2014 Education
Age
N= 89
None Middle School High School High School + 2 License degree Master degree
Comparing map structures at the individual level: R/T ratio
R>T Emphasis on consequences
R<T Emphasis on causes
Qualitative comparative analysis at the site level : Patterns in Group Concept Maps
• Qualitative Pattern detection B
S S
Qualitative comparative analysis at the site level : Patterns in Group Concept Maps
• Qualitative Pattern detection B
S S
Gascony Valleys and Hills
Ø
Qualitative comparative analysis at the site level : Patterns in Group Concept Maps
• Qualitative Pattern detection B
S S
Gascony Valleys and Hills Plaine et Val de Sèvre
Qualitative comparative analysis at the site level : Patterns in Group Concept Maps
• Qualitative Pattern detection B
S S
Camargue Gascony Valleys and Hills Plaine et Val de Sèvre
3 SITES; minimum weight = 50%;
farmers Annual crops
Content analysis: what consensus across and within sites ?
Consensus seem slightly stronger in the Camargue
• 200 different concept categories 1275 different directed links
Plaine et Val de Sèvre Camargue Gascony Valleys and Hills
Fisher’s Exact test : p <0,001
A1
A2
A
3
one ICM
Comparing contents of maps at the site level
• Selecting the most consensual links in Group Concept Maps by study site
Frequency threshold
• Can we find any significant difference in frequency of citation between study sites for some key links ?
Fisher’s Exact test for count data
Fisher exact test for count data ** p < 0.01 * p < 0.05 Gascony Valleys and Hills
minimum link frequency = 20%
Plaine et Val de Sèvre
*
Fisher exact test for count data ** p < 0.01 * p < 0.05
minimum link frequency = 20%
Qualitative comparison in a nutshell
Camargue
Gascony Valleys and Hills
Plaine et Val de Sèvre
• Great emphasis put on production means like agricultural machine and inputs feeling of path dependency
• Farmers exploit woodlands
• Emphasis put on hedges and trees in a plain where they almost disappeared
• Central and local authorities play a significant role
• Irrigated crops organize the landscape • More links between landscape components • Protected area management structures play
a significant role
What visions of links between nature and agriculture ?
Integrated vision
Fauna in the margins of the farmed areas
Gascony Valleys and Hills
Plaine et Val de Sèvre
What visions of links between nature and agriculture ?
Integrated vision
Fauna in the margins of the farmed areas
Absence of fauna; Ø
Camargue
Gascony Valleys and Hills
Plaine et Val de Sèvre
What visions of links between nature and agriculture ?
Integrated vision
Fauna in the margins of the farmed areas
Absence of fauna; Segregated from farmed landscape and associated with wetlands in farmers’ mind ?
Camargue
Gascony Valleys and Hills
Plaine et Val de Sèvre
Conclusion Gascony Valleys and Hills Plaine et val de Sèvre Camargue
Integrated vision of links between nature and agriculture
territorial coordination and institutionalization of biodiversity conservation
consensus on the system
Discussion on mental model methods
• PROS – Stay close to people’s own perception – Transparency kept all along data processing – Facilitate research on complex social-ecological
systems • CONS
– Time consuming – Critical categorization process – Great number of variables vs small samples, need
for further research on data analysis techniques
Take home messages
• Great diversity of mental models of the landscape between and within regions
• Conservation institutions and concertation processes do not necessary lead to social learning
• Conservation policies in French farmed landscape still fail to promote effective land sharing integrated approaches
• Perspectives = collectively build models of the socio-ecosystem functioning in participatory fashion