Back

Ecological coherence in abundance dynamics across terrestrial and marine assemblages

Fuster-Calvo, A.; Hebert, K.; Dolezal, A. J.; Parent, C.; Massol, F.; Waters, M. T.; Gravel, D.

2025-11-05 ecology
10.1101/2025.11.05.686795 bioRxiv
Show abstract

Understanding how communities respond to environmental change requires assessing not just overall variability but also the structure of co-variation among taxa. We frame this idea under the Ecological Coherence (EC) framework, which generalizes previous notions such as community synchrony or coherence. EC captures the structure of co-responses among taxa within assemblages and can be expressed through two complementary objects: (1) the co-response matrix C, which contains all pairwise correlations between taxa and can be used to identify clusters of taxa with coherent responses as well as the contributions of individual species to community-wide coherence; and (2) the EC distribution, which summarizes the overall profile of co-responses by capturing their shape, spread, and central tendency across the community. By combining these two views, EC moves beyond single summary metrics and provides a richer picture of how coherence is organized within communities. Analyzing the EC distribution across 341 terrestrial and 105 marine assemblages worldwide, we found a general prevalence of weak correlations and a few strong, directional correlations. We also found that it varies with community composition, sampling effort, and biogeographic region. Moreover, the C matrix consistently identified a small subset of taxa with strong correlations to many others, suggesting a promising path to detecting those that may play central roles in amplifying or buffering community responses to environmental change. Our findings on EC pave the way for deeper investigations into what drives the diversity of ecological responses to environmental changes and how it shapes community dynamics, while also underscoring the need for strategically distributing ecological monitoring across trophic guilds and functional roles.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.