Transcript Network Perspective of Resilience in Social

Network Perspective of Resilience
in Social-Ecological Systems
Based on:
Janssen, M. A., Ö. Bodin, J. M. Anderies, T. Elmqvist, H.
Ernstson, R. R. J. McAllister, P. Olsson, and P. Ryan.
2006. Toward a network perspective of the study of
resilience in social-ecological systems. Ecology and
Society 11 (1): 15.
Resilience of Social-Ecological Systems
Challenge for Network Analysis
• How to do quantitative analysis when SES includes and mixes both
social
– Nodes: individuals, communities, organizations, farmers, etc.
– Links: trust, power, etc.
• and ecological entities?
– Nodes: Land properties, lakes, forests, species, etc.
– Links: water flow, cattle movement, seed dispersal, etc.
• Archetypical social-ecological networks
– ecosystem networks that are connected by people via information or
physicial flows,
– ecosystem networks that are disconnected and fragmented by
people,
– ecosystem networks that connect people.
• Two broad metrics: levels of
– Connnectivity
– Centrality
Network Metrics
Measures of Connectivity
• Density of Links
– Number of links / Max. possible number of links
• Reachability
– The extent to which all nodes are accessible to one
another
• Neither independent nor the same.
– High density implies high reachability (normally)
– High density with low reachability possible with high
degree of clustering.
Network Metrics
Measures of Connectivity
Min. number of
links connects
all nodes
All possible links are
included and each node
is a neighbor of every
other node
No node
can reach
all other
nodes
Different types of networks as a
function of reachability and density.
Resilience and Network Metrics
Measures of Connectivity
• Resilience of individuals to disease
– Low reachability  high resilience
• Resilience of network to removal of links
– High density (link redundancy)  high resilience.
• Network of managers
– R  balance between learning from others and
innovating on one’s own.
Sensitivity of System Performance to
Differences in Density
Density
High
Advantages
Good information exchange/learning
gives better management (e.g., Pretty
and Ward 2001)
Disdvantages
Potential for systems to become
superconnected and brittle
(Redman and Kinzig 2003)
Enhanced diffusion of innovations
(e.g.,Abrahamson and Rosenkopf 1997)
Low
Increased diversity in management
practices, low risk for lock-ins and global
coherence (Bodin and Norberg 2005)
Limited spread of information
Network Metrics
Quantifying Reachabilty
• Network Diameter
– Minimum path length connecting any pair of
nodes in the network
• Size of the Largest Component
– A set (or cluster?) of nodes in which any two
of the nodes are linked by a path.
Sensitivity of System Performance to
Differences in Reachability
Reachability
High
Advantages
Disdvantages
Access to distant information
(Granovetter 1973)
Spread of contaminants
over large distances
Increased ability to respond to changes
(see Aldrich 1999 and references therein)
Increased spread of
diseases such as HIV
(Friedman et al. 1997)
Union of different social actors, e.g.,
government agencies and local users, to
better match ecological and social
boundaries (Schneider et al. 2003)
Enhanced possibilities of long-range
interpatch dispersal (Urban and Keitt 2001)
Low
Potential for the formation of coherent
and efficient groups/clusters
Difficult recolonization (Keitt
1997, Nystrom and Folke 2001)
Implications of disturbances such as, e.g.,
extinction of single species, do not extend
beyond the local neighborhood in food webs
Inaccessibility of distant
information (Granovetter 1973)
Network Metrics
Measures of Centrality
• Includes
– Distribution of links among nodes in the
network
– Structural importance of links
• High level of centrality
– Well-connected (high ranking) nodes or Hubs
with higher than average number of links
and/or links that span larger than average
distances (inter-cluster links, for example)
Network Metrics
Quantifying Centrality
High Centrality
Low Centrality
• Degree Centrality
– The number of links possessed by a node in
the network
• Betweenness Centrality
– A node’s importance measured as it
contribution to decreasing network distance
(path length between nodes)
Resilience and Network Metrics
Measures of Centrality
• Centrality: impact on resilience
– If Network = Information exchange then
Centrality facilitates coordination and control
but reduces node diversity.
– Scale-free networks are vulnerable to
disappearance of hubs (social leaders,
keystone species)  disintegration into
unconnected sub-nets
Sensitivity of System Performance to
Differences in Centrality
Centrality
High
Advantages
Efficient coordination when solving
simple tasks (see Langan-Fox 2001
and references therein)
Potential to be more accountable, i.e.,
the central actors can to some extent
be held responsible for the group
Low
Possible perception as more fair and
open to group participation
Robustness to removal of nodes
(e.g., social leader or species)
High group efficiency when solving
complex tasks (see Langan-Fox 2001
and references therein)
Disdvantages
Reduced distribution of
information (e.g. Shaw 1981)
Possible perception as
undemocratic and unfair
Greater vulnerability to targeted
attacks (Albert et al. 2000)
Possible lack of control and
accountability
Inefficiency when solving
simple tasks