Trophic Network Stability in Coral Reef Systems

Authors

  • Lea Dubois Assistant Professor, Department of Machine Learning, European Institute of AI, Berlin, Germany Author
  • Hugo Rossi Research Scientist, Department of Machine Learning, European Institute of AI, Berlin, Germany Author
  • Helena Costa Associate Professor, Institute of Intelligent Systems, Central European Tech University, Vienna, Austria Author

Keywords:

coral reef food web, trophic network stability, connectance, keystone species, trophic cascade, herbivorous fish, network topology, ecosystem resilience

Abstract

Coral reef ecosystems support approximately 25% of all described marine species while covering less than 0.2% of ocean surface area, but their extraordinary biodiversity is underpinned by trophic networks of extraordinary complexity that are now under severe threat from bleaching, overfishing, and eutrophication. This study characterised the trophic network structure and stability of 24 coral reef systems spanning the Indo-Pacific, Caribbean, and Red Sea, using quantitative food web models built from 12,847 predator-prey interaction records and biomass estimates from underwater visual censuses, stable isotope analysis, and diet studies. Network topology metrics -- connectance, mean trophic level, omnivory index, and modularity -- were quantified for each system and correlated with reef health indices and bleaching history. Network stability was assessed via eigenvalue analysis and simulated species deletion experiments. Highly degraded reefs (live coral cover < 15%) showed significantly lower network connectance (mean C = 0.087 vs. 0.124 for healthy reefs; t = 8.47, p < 0.001), reduced mean trophic level (2.84 vs. 3.41; t = 12.84, p < 0.001), and greater network instability (dominant eigenvalue magnitude: 1.84 vs. 1.12; t = 9.24, p < 0.001). Apex predator removal simulations produced cascading secondary extinctions affecting 47.4% +- 8.2% of species in degraded networks compared with 18.7% +- 4.8% in healthy networks, demonstrating that degradation amplifies trophic cascade severity. Keystone species analysis identified herbivorous fish guilds (parrotfish and surgeonfish) as the critical functional group whose removal most destabilised network dynamics in all 24 reef systems regardless of health status. These findings provide mechanistic understanding of reef ecosystem collapse dynamics and prioritise herbivore protection as the single most important functional conservation intervention across reef systems.

Author Biographies

  • Lea Dubois, Assistant Professor, Department of Machine Learning, European Institute of AI, Berlin, Germany

    Assistant Professor, Department of Machine Learning, European Institute of AI, Berlin, Germany

  • Hugo Rossi, Research Scientist, Department of Machine Learning, European Institute of AI, Berlin, Germany

    Research Scientist, Department of Machine Learning, European Institute of AI, Berlin, Germany

  • Helena Costa, Associate Professor, Institute of Intelligent Systems, Central European Tech University, Vienna, Austria

    Associate Professor, Institute of Intelligent Systems, Central European Tech University, Vienna, Austria

Downloads

Published

2024-12-15

Versions

How to Cite

Trophic Network Stability in Coral Reef Systems. (2024). International Journal of Animal Biodiversity, Conservation and Systematics ( IJABC), 4(4), 36-43. https://stanfordgroup.org/index.php/IJABC/article/view/267 (Original work published 2026)

Similar Articles

11-20 of 88

You may also start an advanced similarity search for this article.