A hub and spoke network model to analyse the secondary dispersal of introduced marine species in Indonesia

ICES Journal of Marine Science ICES Journal of Marine Science (2015), 72(3), 1069– 1077. doi:10.1093/icesjms/fsu150 Contribution to the Themed Section: ‘Risk Assessment’ Original Article A hub and spoke network model to analyse the secondary dispersal of introduced marine species in Indonesia 1 National Centre for Marine Conservation and Resource Sustainability, University of Tasmania, Locked Bag 1370, Newnham, TAS 7250, Australia School of Science, Faculty of Science and Engineering, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand 2 *Corresponding author: tel: +64 27 456 3930; e-mail: Azmi, F., Hewitt, C. L., and Campbell, M. L. A hub and spoke network model to analyse the secondary dispersal of introduced marine species in Indonesia. – ICES Journal of Marine Science, 72: 1069 – 1077. Received 1 April 2014; revised 6 August 2014; accepted 10 August 2014; advance access publication 5 September 2014. Indonesia is a biodiversity hotspot threatened with new introductions of marine species. As with many countries, Indonesia has a stratified shipping network of international ports linked to a large suite of domestic ports. We developed a hub and spoke network model to examine the risk associated with the secondary transfer of introduced marine species from the port hub of Tanjung Priok in Jakarta Bay to the 33 Indonesian provinces (including other ports in the Jakarta province). An 11-year shipping dataset was used (vessel next port of call records for maritime vessels that originated in Jakarta Bay and that remained in domestic waters) to derive a province ranking of vulnerability. Fifteen provinces represented almost 94% of the traffic frequency, with East Java and Jakarta provinces dominating. All urban provinces featured within the top seven highest frequency traffic provinces. Traffic patterns reflect an intra-coastal reliance on shipping, with traffic frequency decreasing with distance from Jakarta Bay. Provinces were regionalized into three categories (Lampung to East Java, Makassar Straits, and Malacca Straits) each with different vulnerabilities based on their values. Keywords: dispersal, non-indigenous species, risk management, risk model, shipping, vector, vulnerability. Introduction Once an introduced speciesestablishes within a port, management shifts its focus from prevention (e.g. Hewitt et al., 2004a) to control, eradication, and management (e.g. Wotton and Hewitt, 2004; Campbell, 2008). Of significant concern is that once established, an introduced marine species can continue to disperse (dispersal via natural or secondary human-mediated transfer; Carlton and Hodder, 1995; Carlton, 2001; Floerl and Inglis, 2005; Wyatt et al., 2005; Minchin et al., 2006; Ruiz et al., 2011) to new regions within domestic borders, leading to the need for management to expand its focal range. This spread effect can be a drain on resources as managers attempt to make practical decisions about potential loss of values caused by the spread of the introduced species (Johnson et al., 2001; Ashton et al., 2006). To proactively manage the secondary dispersal of introduced species, biosecurity managers can implement marine vessel traffic analyses (pathway connection and vector strength) that examine vessel movements between ports within domestic borders. From these analyses, hub and spoke network models can be created to identify potential risky pathways that may require “future watch” activities (vigilant surveillance of a pathway for species introductions) or more engaged management. These models can be useful in regions where limited species data exist but where shipping patterns are well known. To test the utility of hub and spoke network models for introduced marine species, we examined Jakarta Bay (Port of Tanjung Priok), Indonesia. The Port of Tanjung Priok is the largest port in Indonesia and has strong vector connections with the major port hub cities of Singapore and Tanjung Pelepas (Malaysia; Azmi, 2010). Tanjung Priok is within the Coral Triangle Initiative region (http://www. cti-secretariat.net/) that focuses on conserving biodiversity, developing food security, and establishing a sustainable future for the region. Domestic shipping from other Indonesian provinces constitutes the greatest amount of ship traffic into this port (62% of port calls; Azmi, 2010). Shipping patterns in the Port of Tanjung Priok follow an intracoastal transport model (Lee et al., 2008); being connected to # International Council for the Exploration of the Sea 2014. All rights reserved. For Permissions, please email: Fauziah Azmi1, Chad L. Hewitt 1,2, and Marnie L. Campbell 1,2* 1070 applied to shipping, logistic delivery, and other transportation activities (e.g. Aykin, 1995; Bendall and Stent, 2001; Bryan and O’Kelly, 2005; Hsu and Hsieh, 2007; Imai et al., 2009), as well being applied to human health services to examine the epidemiology of disease spread and healthcare management (Richards et al., 1997; Sibthorpe et al., 2005). The epidemiological aspect of the models makes them ideal for risk assessment within a biosecurity context. The vectoring of introduced aquatic species has been conceptualized in a hub and spoke network context (e.g. Carlton, 1996; Johnson et al., 2005), yet few have created or used hub and spoke network models to examine realized vector connectivity (except see Lavoie et al., 1999; Muirhead and MacIsaac, 2004). To examine this in a marine context, we developed a hub and spoke network model to assess the strength of transport pressure (frequency of maritime vessel transfers) from the Port of Tanjung Priok (the hub, gateway, or feeder port) to other domestic Indonesian ports (spokes or outports). The model provides an assessment of possible secondary dispersal of introduced marine species within the Indonesian region, recognizing that this is limited to the Port of Tanjung Priok as the single point of entry into the Indonesian domestic system. Based on the model outcomes, vulnerable provinces or regions are identified. Although this paper is focused on Indonesia, the model can be applied to efficiently analyse the domestic transfer of introduced marine species in other countries, especially when species data may be limited, but shipping strength is known. For example, this model is currently being used to examine both international and domestic connections for ports in Australia and the Galapagos Islands (Campbell et al., 2013). Methods We used an 11-year (1999–2009) shipping dataset purchased from the Lloyd’s Maritime Intelligence Unit (Azmi, 2010). The records of next port of call (NPOC) for ships that departed the Port of Tanjung Priok provide the pathway connections and vector strength to determine the dispersal patterns. Due to the large number of Indonesian ports, the analysis was undertaken at the level of province (Figure 1), which is the most likely management level considering jurisdictional and political boundaries. Each NPOC w (...truncated)