class: title, smokescreen, shelf, no-footer background-image: url(web.jpg) # Networks models and socioecosystems --- class: fit-h1 # The resilience question   --- class: fit-h1 # The resilience question : a multidisciplinary concept   --- class: # Resilience, complex system and networks Resilience, robustness and vulnerability refers to structural characteristics --  -- **Diversity and connectivity are key structural descriptors** -- #### Scientific challenges : - What are diversity-stability relationships? - How connectivity affects resilience? -- .absolute.w-50pct.h-3.pa-1.l-1.b-4.ba.bc-orange.bw-2.br-2[] --- class: # Connectivity and resilience Connectedness impacts on resilience : both directions are possible! > more connections dilute impacts of strong changes but also propagate disturbances - depends on network properties (e.g. centrality) - depends on the characteristics of the connnected components - depends on the number of connections - depends on the strength of connections <!-------------------------------- --> --- class: title, smokescreen, no-footer background-image: url(nerve-cell.jpg) # Networks : from mathematical objects to socioecosystems --- class: # Networks and graph theory > graph = network of points connected by lines points = **nodes** (or vertices) and links/lines = **edges** --   -- Matrix formalism  --- class: compact, h3-cl # The expansion of graph theory ###### Koeninsberg Bridges ~1740 (Euler) -> preuve 1 siècle plus tard -> 1 siècle plus tard, graphs aléatoires   .absolute.w-50pct.h-3.pa-1.r-1.t-5["graph theory" (WoS)] .absolute.w-50pct.h-3.pa-1.r-4.b-4[2019----------------------------------------------------------------1995] --- class: col-2 # Interest of network approaches  1. depict systems with numerous interactions leading to complex and non-linear dynamics 2. an interdisciplinary object / bridges between disciplines .absolute.r-1.t-30pct[WoS Research Categories] --- class: compact # Networks for ecological systems ###Nodes = species -- #### Trophic networks Edges = biomass transfert <!-- Reconstruction from distributions and metaweb, traits -->  -- #### Bipartite/tripartite networks Edges = pollinisation, parasitism, herbivory  -- #### Trees (specific type of networks) Edges = phylogenetic/functional distances  --- class: compact # Networks for ecological systems #### Spatial ecological networks Nodes = habitat patches ; Edges = distance  .absolute.r-3.t-2[Gonzales et al. 2017] -- #### Metaecosystems (multi-layers networks) Nodes = species / ecosystems  .absolute.l-7.b-3[Gounand et al. 2018] --- class: # Networks for socio-economic systems ####Economic networks and concepts - stocks and flux - metabolic networks (territorial ecology)  -- ####Spatial networks Human mobility  .absolute.r-1.b-3[Barbosa 2018, Physics Reports] --- class: # Networks for socio-economic systems  ####Diversity of type of links between humans - goods - communication / information / knowledge transfert - spatial (co-presence) - physical vs non physical - emotional --- class: # Networks for socioecosystems ### Multi-layers --  .absolute.r-1.b-2[Quintescence Consorsium 2016] -- .absolute.l-6.b-4.bg-white.ba[**nodes=species; links=interactions**] -- .absolute.l-6.b-5.bg-white.ba[**nodes=individuals; links=common perceptions**] -- .absolute.l-6.t-5.bg-white.ba[**nodes=group of individuals; links=transactions**] -- .absolute.l-2.b-5.bg-white[What relationship<br>between layers?] --- class: # Networks for socioecosystems ### Pattern analysis Access to ressources 2-layers network  .absolute.r-2.b-2[Janssen et al. 2006, Bodin et al. 2009 (Ecology and Society) <br> Bodin and Tengo 2012 (Glob. Env. Change)] --- class: # Networks for socioecosystems ### Integrating multiple networks Objective oriented (resilience of an ES) Multiple type of links Interdependence   .absolute.r-2.b-2[Dee et al. 2017, TREE] --- class: roomy # Challenges for SES networks 1. How to include **space and time** to depict transitions/trajectoires ? -- 2. How to deal with the diversity of **types of links** ? -- 3. **Interdisciplinarity** : which perspective ? -- 4. How to integrate **scales** ? -- 5. How to represent **systems' environments** in a context of strongly connected systems ? How to choose optimal borders ? <!-------------------------------- --> --- class: title, smokescreen, no-footer background-image: url(Ecrins.png) # A theoretical exemple --- class: # Nodes : the human as another animal  <!--  --> **Habitat = Ecosystem/patch**: refers to a relatively homogeneous spatial unit and all its components (incl. species populations spatially restricted to the spatial unit) **User**: refers to individuals or group of individuals that are not spatially restricted to one spatial unit --- class: img-left, fit-h1 # Links : trophic web and metacommunities   -- User-user (direct) links: indirect beneficiary / higher trophic levels -- Habitat-Habitat links :<br> spatial connectivity  --- class: img-left # Network of networks   The network is built as **multi-layers**, including various **functions** that differ in their user-habitat primary link (herbivory, timber harvesting, fishing, vegetation structure, ...) --- class: img-left # Structural analysis   Explore interdependences and indirect interactions accross sectors analysing a metaweb --- class: img-left # Space and time : an ecological perspective   Using spatio-temporal model of habitats (e.g. landscape vegetation model) Rules for different uses Simulations and prospective --- class: img-right # Additional examples - Petri-nets : a powerful tool > A Petri net is a directed bipartite graph that has two types of elements, places and transitions, depicted as white circles and rectangles, respectively. A place can contain any number of tokens, depicted as black circles. A transition is enabled if all places connected to it as inputs contain at least one token  --- class: title, smokescreen, no-footer background-image: url(nerve-cell.jpg) # Discussion --- class: no-footer background-image: url(vallonpierre.png) # Thanks <br><br><br><br><br><br><br><br><br> <p align=right> Slides can be found at http://iboulangeat.github.io/Slides/ <br> Contact: isabelle.boulangeat@inrae.fr </p>