A network analysis of global cephalopod trade

The global trade in cephalopods is a multi-billion dollar business involving the fishing and production of more than ten commercially valuable species. It also contributes, in whole or in part, to the subsistence and economic livelihoods of thousands of coastal communities around the world. The importance of cephalopods as a major cultural, social, economic, and ecological resource has been widely recognised, but research efforts to describe the extent and scope of the global cephalopod trade are limited. So far, there are no specific regulatory and monitoring systems in place to analyse the traceability of the global trade in cephalopods at the international level. To understand who are the main global players in cephalopod seafood markets, this paper provides, for the first time, a global overview of the legal trade in cephalopods. Twenty years of records compiled in the UN COMTRADE database were analysed. The database contained 115,108 records for squid and cuttlefish and 71,659 records for octopus, including commodity flows between traders (territories or countries) weighted by monetary value (USD) and volume (kg). A theoretical network analysis was used to identify the emergent properties of this large trade network by analysing centrality measures that revealed key insights into the role of traders. The results illustrate that three countries (China, Spain, and Japan) led the majority of global market movements between 2000 and 2019. Based on volume and value, as well as the number of transactions, 11 groups of traders were identified. The leading cluster consisted of only eight traders, who dominated the cephalopod market in Asia (China, India, South Korea, Thailand, and Vietnam), Europe (the Netherlands, and Spain), and the USA. This paper identifies the countries and territories that acted as major importers or exporters, the best-connected traders, the hubs or accumulators, the modulators, the main flow routes, and the weak points of the global cephalopod trade network over the last 20 years. This knowledge of the network is crucial to move towards an environmentally sustainable, transparent, and food-secure global cephalopod trade.


Supplementary
The Strength of a node i is defined as the sum of the weights of the in-coming links and the weights of the out-going links.
where: w = weighted matrix w wij = weight of the directed link from node i to node j wji = weight of the directed link from node j to node i Also named weighted degree. The node degree is the number of relations (edges) of the nodes. In weighted networks, node Strength is the sum of weights of links connected to the node.
Strength could indicate if a trader is involved in important (by weight) trades with other traders. Traders with high Strength can be acting as keystones since they are connected by imports and exports to many neighbouring traders.
The In-strength of a node i is defined as the sum of the weights of the in-coming links.
where: w = weighted matrix w wij = weight of the directed link from node i to node j In directed networks, the Instrength is the sum of inward link weights.
Traders with a high in-strength could act as important importers or hubs for the distribution of raw materials. High in-strength could also be targeting traders acting as major consumers of products. The Out-strength of a node i is defined as the sum of the weights of the out-going links.
where: w = weighted matrix w wji = weight of the directed link from node j to node i In directed networks, the Outstrength is the sum of outward link weights.
Traders with high out-strength may be acting as raw producers with high export flows. This may indicate the geographical origin of the commodities and essential habitats for the species. Outstrength can also indicate whether a trader is a major exporter of processed products.
Closeness Freeman, 1979 3 Closeness centrality indicates how close a node is to all other nodes in the network. It is calculated as the average of the shortest path length from the node i to every other node in the network.
where: Ci = Closeness centrality of node i dij = distance between nodes i and j N = number of activity nodes in the network Closeness centrality indicates how long it will take for information from a given node to reach other nodes in the network.
Traders with a higher closeness have a high probability of exporting to the nearest neighbouring traders. These traders could be important in trade at regional or continental geographic scales.

Betweenness Freeman, 1979 3
Betweenness centrality is calculated with the number of shortest paths (between any couple of nodes in the networks) that go through the target node i. The score is moderated through the total number of shortest paths between any pair of nodes in the network. The target node will have a high betweenness centrality if it appears in many shortest paths. Traders with high Betweenness centralities have been called "bottlenecks" or "bridges" and prevent network fragmentation. A trader that acts as a bridge between two well differentiated groups of traders usually has a high Betweenness.
Edge betweenness Girvan and Newman, 2002 4 Edge betweenness centrality is a measure of the centrality of an edge in a network based on the number of shortest paths through the given edge.
σ %--% EBe = edge betweenness centrality of edge e σij = number of shortest paths between nodes i and j σij (e) = number of shortest paths between i and j that go through the edge e Edge betweenness centrality identifies the edges of the network that are crucial for information flows.
An edge with a high edge betweenness centrality score represents a bridge-like connector between two countries or territories in the global market, and whose removal may affect the flow of goods between many pairs of partners through the shortest paths between them.

PageRank
Brin and Page 1998 5 A variant of Eigenvector Centrality, primarily used for directed networks. PageRank considers (1) the number of in-coming links (i.e., nodes that link to a target node), (2) the quality of the linkers (i.e., the PageRank of nodes that link to the target node), and (3) the link propensity of the linkers (i.e., the number of nodes the linkers link to).