Network analysis of ballast-mediated species transfer reveals important introduction and dispersal patterns in the Arctic

Mandana Saebi, Jian Xu, Salvatore R. Curasi, Erin K. Grey, Nitesh V. Chawla, David M. Lodge

Abstract

Rapid climate change has wide-ranging implications for the Arctic region, including sea ice loss, increased geopolitical attention, and expanding economic activity resulting in a dramatic increase in shipping activity. As a result, the risk of harmful non-native marine species being introduced into this critical region will increase unless policy and management steps are implemented in response. Using data about shipping, ecoregions, and environmental conditions, we leverage network analysis and data mining techniques to assess, visualize, and project ballast water-mediated species introductions into the Arctic and dispersal of non-native species within the Arctic. We first identify high-risk connections between the Arctic and non-Arctic ports that could be sources of non-native species over 15 years (1997-2012) and observe the emergence of shipping hubs in the Arctic where the cumulative risk of non-native species introduction is increasing. We then consider how environmental conditions can constrain this Arctic introduction network for species with different physiological limits, thus providing a tool that will allow decision-makers to evaluate the relative risk of different shipping routes. Next, we focus on within-Arctic ballast-mediated species dispersal where we use higher-order network analysis to identify critical shipping routes that may facilitate species dispersal within the Arctic. The risk assessment and projection framework we propose could inform risk-based assessment and management of ship-borne invasive species in the Arctic.

Introduction

As climate change accelerates the melting of Arctic sea ice, the region is experiencing a dramatic surge in global shipping activity. This shift represents more than just a geopolitical or economic transformation; it is a significant biological event. Ships utilize ballast water for stability, which inadvertently transports thousands of non-native marine species across the globe. If these species establish themselves in the fragile Arctic ecosystem, they can devastate local fisheries, disrupt infrastructure, and permanently alter biodiversity.

Traditional risk assessments often rely on simple "point A to point B" models, evaluating the risk of species moving directly between two ports. However, the global shipping network is a complex, interconnected web. A species might survive a journey through multiple intermediate ports before finding a suitable environment. This research addresses the need for a holistic, "higher-order" understanding of how species move into and within the Arctic.

Methodology

To capture the complexity of species movement, the study utilized advanced data mining and network science techniques. The researchers developed a Species-Flow Higher-Order Network (SF-HON). Unlike standard first-order networks that only consider direct connections, the SF-HON accounts for "memory"—recognizing that a ship’s next destination is often dependent on its previous locations. This allows the model to track indirect dispersal pathways where species are carried through intermediate hubs to a final destination.

This network was integrated with massive global datasets, including Lloyd’s List Intelligence (LLI) for ship movements and the National Ballast Water Information Clearinghouse (NBIC) for discharge volumes. Furthermore, the researchers applied Environmental Filtering by categorizing species into groups based on their tolerance for temperature and salinity. This allowed them to determine which pathways are biologically viable for specific organisms, such as freshwater versus marine species.

The Rise of Arctic Shipping Hubs

A key finding of the network analysis is that Arctic shipping is not just increasing; it is consolidating. While the number of voyages grew significantly between 1997 and 2012, the number of unique pathways only increased by 2%. The network is effectively "rewiring" itself around a few major hubs.

Global map showing introduction pathway risk and aggregated risk at ports. Major hubs like Murmansk and Narvik show high aggregated risk.

As shown in the spatial analysis, ports such as Murmansk (Russia) and Narvik (Norway) have emerged as primary gateways for species introduction. Interestingly, the model reveals that indirect risk is a major factor; for instance, Reykjavik (Iceland) ranks relatively low for direct introduction risk but is the 3rd highest for indirect dispersal, explaining the appearance of non-native species there despite low direct traffic from known source regions.

Statistical analysis of the shipping data from 1997 to 2012 confirms a steady upward trend in both activity and biological risk. The number of voyages into the Arctic increased by approximately 128 per year, while the total deadweight tonnage (DWT) grew by roughly 2.52 × 10⁶ per year.

Time series plots showing the increase in voyages, DWT, and average risk per pathway from 1997 to 2012.

Crucially, the average risk per individual pathway has increased significantly. This trend is driven by the use of larger ships and more frequent trips on established routes rather than the creation of new routes. The concentration of traffic into specific hubs means that the risk of invasion is becoming more localized and intense at these critical nodes.

Discussion and Limitations

The "memory" effect captured by the SF-HON model demonstrates that traditional models miss hidden risks. For example, ships coming from Tromsø through Murmansk are statistically much more likely to head to Bodø than to Nuuk. A first-order model would treat these destinations as equally likely, potentially leading to inefficient management strategies.

However, the study acknowledges several limitations. The primary shipping dataset ends in 2012, and the rapid pace of Arctic melting necessitates more recent data to capture the current reality. Additionally, the model focuses exclusively on ballast water, omitting the risk of biofouling (species clinging to ship hulls). Future work aims to integrate real-time environmental data and environmental DNA (eDNA) monitoring to validate these network predictions in the field and provide decision-makers with more precise tools for managing ship-borne invasive species.

Summary updated

Abstract

Rapid climate change has wide-ranging implications for the Arctic region, including sea ice loss, increased geopolitical attention, and expanding economic activity resulting in a dramatic increase in shipping activity. As a result, the risk of harmful non-native marine species being introduced into this critical region will increase unless policy and management steps are implemented in response. Using data about shipping, ecoregions, and environmental conditions, we leverage network analysis and data mining techniques to assess, visualize, and project ballast water-mediated species introductions into the Arctic and dispersal of non-native species within the Arctic. We first identify high-risk connections between the Arctic and non-Arctic ports that could be sources of non-native species over 15 years (1997-2012) and observe the emergence of shipping hubs in the Arctic where the cumulative risk of non-native species introduction is increasing. We then consider how environmental conditions can constrain this Arctic introduction network for species with different physiological limits, thus providing a tool that will allow decision-makers to evaluate the relative risk of different shipping routes. Next, we focus on within-Arctic ballast-mediated species dispersal where we use higher-order network analysis to identify critical shipping routes that may facilitate species dispersal within the Arctic. The risk assessment and projection framework we propose could inform risk-based assessment and management of ship-borne invasive species in the Arctic.
BibTeX
@inproceedings{saebi2020network,
  author = {Mandana Saebi and Jian Xu and Salvatore R. Curasi and Erin K. Grey and Nitesh V. Chawla and David M. Lodge},
  title = {{Network analysis of ballast-mediated species transfer reveals important introduction and dispersal patterns in the Arctic}},
  booktitle = {Scientific},
  year = {2020},
  doi = {10.1038/s41598-020-76602-4},
}
Figure