The Ocean Environment-Bacterioplankton Nexus

Marine ecologists have sought to decode the complex and dynamic nexus that is the marine environment-microbiome “holobiont” given its importance for supporting ecosystem function and services. However, current approaches to marine ecohealth characterization misrepresent the environment as a disjointed assortment of stressors and fail to effectively incorporate the collective dynamics of inter-species bacterioplankton interactions.
Using marine microbiome data from the Great Barrier Reef, we introduce a robust methodology for marine ecohealth signalling which considers the independent and joint probability distribution functions of oceanic environmental and bacterioplankton interactions quantified as transfer entropy among each system’s components. Functional networks are constructed from these interactions, whose keystone nodes have the most interaction and connectivity magnitudes, and whose keystone links have the highest network-wide importance.
Suboptimal marine ecosystems present low exponential Zipf’s law scaling parameters among these interactions, corresponding to a more exponential distribution of environmental stressors and of species collective abundance. Keystone links consistently connects a highly central and effectively distant node, providing a more robust classification of node importance. For the bacteriome, keystone species are rare with relatively unstable abundances. A new Kleiber's law-like exponent is discovered between species pairwise interactions and abundance, where the scaling exponent is proportional to the centrality of the microbial species in isolation. In a collective sense the Kleiber's scaling exponent is the highest for the impacted suboptimal habitats such as the Northernmost marine reefs. Overall, Marine inshore reefs are suboptimal marine ecosystems likely resulting from temperature shocks, while the estuarine inshore reefs are most optimal despite impact from river runoff.
The method provides a metric which quantifies departure from optimal ecosystem functioning by emphasizing the importance of functional organization- in terms of the interaction network topology of microbial and environmental components- rather than purely taxonomic approaches or historically derived thresholds of environmental measurements. Thus, ranking of microbes and attribution of environmental factors to ecosystem changes must be done considering the collective distribution of interactions versus the consideration of macroecological indicators that do not consider ecosystem function.