Research Article: The effects of network topology, climate variability and shocks on the evolution and resilience of a food trade network

Date Published: March 26, 2019

Publisher: Public Library of Science

Author(s): Alexander G. Dolfing, Jasper R. F. W. Leuven, Brian J. Dermody, Samir Suweis.


Future climate change will impose increased variability on food production and food trading networks. However, the effect of climate variability and sudden shocks on resource availability through trade and its subsequent effect on population growth is largely unknown. Here we study the effect of resource variability and network topology on access to resources and population growth, using a model of population growth limited by resource availability in a trading network. Resources are redistributed in the network based on supply and the distance between nodes (i.e. cities or countries). Resources at nodes vary over time with wave parameters that mimic changes in biomass production arising from known climate variability. Random perturbations to resources are applied to study resilience of individual nodes and the system as a whole. The model demonstrates that redistribution of resources increases the maximum population that can be supported (carrying capacity) by the network. Fluctuations in carrying capacity depend on the amplitude and frequency of resource variability: fluctuations become larger for increasing amplitude and decreasing frequency. Our study shows that topology is the key factor determining the carrying capacity of a node. In larger networks the carrying capacity increases and the distribution of resources in the network becomes more equal. The most central nodes achieve a higher carrying capacity than nodes with a lower centrality. Moreover, central nodes are less susceptible to long-term resource variability and shocks. These insights can be used to understand how worldwide equitable access to resources can be maintained under increasing climate variability.

Partial Text

Climate impacts all life on Earth, because variability in temperature and water availability affect primary productivity and thus food availability at all trophic levels. It is known that organisms compete for food resources in variable environments [1]. However, little is known about the fundamental dynamics underlying how humans access food resources in variable environments. Humans, like other organisms, are impacted by changes in food availability arising from climate variability [2]. Since the beginning of civilization, humans have developed methods to provide buffers against climate variability such as the use of irrigation, food stocks and resource redistribution via trade [3; 4]. Like other social organisms, such as ants, these strategies are associated with a fundamental reorganization of our societal structure [5].

To summarize our methodology; we generated a food distribution system of 10 and 100 nodes with a small-world network configuration. Population growth was simulated in these networks as a function of local resources at the node and competition for resources in the rest of the network. Resources varied in the network with varying amplitude and periodicity to reflect multiannual to decadal oscillation in the climate system where frequency occurs in phase or out of phase. Fig 1 and S1 Fig illustrate the experimental setup.

We studied the effect of climate variability, network topology and shocks on population growth in a model that simulates a resource redistribution in a small-world network. The results show that resource redistribution serves to increase carrying capacities and population size of the network when resource variability is out of phase between two parts of a network. This effect becomes stronger with increasing network size. The centrality of nodes (i.e. cities or countries) in the network correlates with population size, which is attributed to increased accessibility to resources through trade. These central nodes are also less vulnerable to resource variability and shocks, because they can access more resources as the paths to all the other nodes are relatively shorter. This means that centrality of a node improves the position of that node in the trade network, and therefore reduces their susceptibility to changes in the rest of the network. Nodes that depend on import from only a few other nodes, usually have a lower centrality. This enhances their risk of being affected by disturbances in the food supply of the global food trade network, especially when climate variability will increase in the near future.