Data from: Managing seagrass resilience under cumulative dredging affecting light: predicting risk using dynamic Bayesian networks

  • Paul Pao Yen Wu (Creator)
  • Kathryn McMahon (Creator)
  • Michael A. Rasheed (Creator)
  • Gary Kendrick (Creator)
  • Paul H. York (Creator)
  • Kathryn Chartrand (Creator)
  • M. Julian Caley (Creator)
  • Kerrie Mengersen (Creator)



Coastal development is contributing to ongoing declines of ecosystems globally. Consequently, understanding the risks posed to these systems, and how they respond to successive disturbances, is paramount for their improved management. We study the cumulative impacts of maintenance dredging on seagrass ecosystems as a canonical example. Maintenance dredging causes disturbances lasting weeks to months, often repeated at yearly intervals. We present a risk-based modelling framework for time varying complex systems centred around a dynamic Bayesian network (DBN). Our approach estimates the impact of a hazard on a system's response in terms of resistance, recovery and persistence, commonly used to characterise the resilience of a system. We consider whole-of-system interactions including light reduction due to dredging (the hazard), the duration, frequency and start time of dredging, and ecosystem characteristics such as the life-history traits expressed by genera and local environmental conditions. The impact on resilience of dredging disturbances is evaluated using a validated seagrass ecosystem DBN for meadows of the genera Amphibolis (Jurien Bay, WA, Australia), Halophila (Hay Point, Qld, Australia) and Zostera (Gladstone, Qld, Australia). Although impacts varied by combinations of dredging parameters and the seagrass meadows being studied, in general, 3 months of duration or more, or repeat dredging every 3 or more years, were key thresholds beyond which resilience can be compromised. Additionally, managing light reduction to less than 50% can significantly decrease one or more of loss, recovery time and risk of local extinction, especially in the presence of cumulative stressors. Synthesis and applications. Our risk-based approach enables managers to develop thresholds by predicting the impact of different configurations of anthropogenic disturbances being managed. Many real-world maintenance dredging requirements fall within these parameters, and our results show that such dredging can be successfully managed to maintain healthy seagrass meadows in the absence of other disturbances. We evaluated opportunities for risk mitigation using time windows; periods during which the impact of dredging stress did not impair resilience.,Validation dataValidation data used to validate the DBN model for Amphibolis in Jurien Bay, Halophila at Hay Point, and Zostera at Gladstone (supporting information S4 through S6, respectively).data_supp.xlsxLateral Growth from Existing IndividualsDBN CPT (visualised in supporting information S7)Physiological Status of PlantsDBN CPT (visualised in supporting information S7)Rate of Recovery in Shoot DensityDBN CPT (visualised in supporting information S7)Recruitment Rate from SeedsDBN CPT (visualised in supporting information S7)Seed DensityDBN CPT (visualised in supporting information S7),
Date made available25 Sept 2018
Date of data production25 Sept 2018

Cite this