Fish ecosystems perform ecological functions that are critically important for the sustainability of marine ecosystems, such as global food security and carbon stock among others. Without lack of generality, by studying a coastal fish community in Japan, Dr. Convertino (Associate Professor at the Institute of Environment and Ecology) and his former Ph.D. student, Jie Li (now at the University of Amsterdam), investigated the causal effects of fluctuations of sea temperature in fish species interactions and community biodiversity (considering seasonal and long-term trends) to assess when, how, and which species are affected by climate change in the ocean (Graphical abstract).
Graphical abstract: time series of fish abundance and inferred species interaction networks over time. The variability of networks and their stability was assessed as a function of sea-surface temperature.
Observations show that a local 20% increase in temperature from 2002 to 2014 underpinned a 25% long-term reduction in fish diversity, driven by some native and invasive species (e.g. Chinese wrasse) becoming exceedingly abundant. The associated changes in the species interaction network caused a large decay in commercially valuable species that are also threatened by intense fishing (e.g. Japanese anchovy, also in certain areas in China; Fig. 1) despite an increase in total fish biomass. A bit of good news was reported: despite a negative collective response to climate alterations, certain species such as the near-threatened Trachurus japonicus (Horse mackerel) increased over time.
Fig. 1: Engraulis japonicus (Japanese anchovy) Native Range reviewed map (AquaMaps). Native species with cooperative behavior characterized by a decreasing abundance for growing temperature. Fishing peaks (in winter) are asynchronized with temperature fluctuations but synchronized with abundance; thus, temperature is a strong second-order environmental determinant of abundance coupled to fish community changes manifested by species interactions which are often neglected.
A previously developed model by the group (Optimal Information Flow model) was used to infer species interaction networks that are topologically different for distinct seasons. Networks for low temperatures are more organized and energetically efficient (‘’scale-free’’) compared to networks for intermediate (15-20◦C) temperatures in which the fish ecosystem experiences a sudden transition in interactions from locally stable to globally unstable. In the long-term, the dynamic dominant eigenvalue— which is an indicator of species interactions—shows increasing instability, especially in summer, which leads to enhanced community variability and ‘’critical slowing down’’ that is species disorganization till a potential ecological collapse.
More climate-fitted species (with higher eco-climatic memory and synchronization, such as the Japanese anchovy) follow temperature increase and cause larger species competition (Fig. 2). Decreasing dominant eigenvalues and higher disorganization in species interactions for warmer oceans indicate fishery more attracted toward persistent oscillatory states — yet unpredictable — with lower cooperation, diversity and fish stock. It is emphasized how changes in species interaction networks, primarily affected by temperature fluctuations, are the backbone of biodiversity and yet of functional diversity in contrast to simple taxonomic counts. Then species interactions are sensitive indicators of environmental chaoticity and sudden ecosystem shifts.
Fig.2: Abundance-interaction atemporal pattern. ε is the parameter governing the organization of fish abundance (where lower ε implies higher scale-free organization and energy efficiency) and OTE is the sum of all species interaction of each species. Cooperative species are more stable and less affected by temperature directly, which underlines the weak-interaction induced stability of ecosystems. Increasing temperature leads to a predominance of "small-world" high interactions of competitive species (whose abundance is synchronized with temperature); this causes the loss (or a shift) of Pareto salient links and the higher fitness of invasive species that tend to enter the ecosystem.
The work provides data-driven tools for analyzing and monitoring fish ecosystems under the pressure of global warming or other stressors. Abundance and interaction patterns derived by network-based analyses are useful to assess ecosystem susceptibility and effective change, and formulate predictive dynamic information for science-based fishery policy aimed to maintain marine ecosystems stable and sustainable. Dr. Convertino, PI of the bluEco lab at Tsinghua SIGS and a top-expert in ecosystem connectomics and systemic risk, strongly emphasizes the need to consider network interactions and organization when assessing ecological responses to environmental change versus single species or environmental factor approaches.
Link to Full Article: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0246222
Text and pictures: Matteo Convertino
Editing: A.S.，Yuan Yang
