Back

Old growth attributes by chain saw: how between-patch heterogeneity changes the metacommunities of beetles in temperate forests

Mitesser, O.; Cadotte, M.; Mori, A. S.; van der Plas, F.; Chao, A.; Rothacher, J.; Baessler, C.; Bevanda, M.; Biedermann, P. H. W.; Bradler, P. M.; Gomez, A. C.; Decker, O.; Delory, B. M.; Dittrich, S.; Fichtner, A.; Kreis, A.; Koestler-Albert, L.; von Oheimb, G.; Pierick, K.; Thorn, S.; Vogelfaenger, L.; Weisser, W.; Wegmann, M.; Wild, C. M.; Mueller, J.

2025-09-04 ecology
10.1101/2025.08.30.673209 bioRxiv
Show abstract

Metacommunity theory has expanded our understanding of how spatial dynamics and local interactions influence species communities. Different assembly archetypes, reflecting different roles of species differences, habitat differences, and dispersal have been described, but we lack empirical studies specifically in terrestrial habitats testing which archetype is most important. In a replicated design we experimentally enhanced structural between-patch heterogeneity in homogeneous production forests and developed a statistical framework controlling for sample incompleteness to detect different metacommunity processes. Meta-analyses on >100K individuals of >1.3K beetle species showed an increase of [~]60 species in heterogenized forests at {gamma}-level promoted by increasing -diversity consistent with the mass-effect and an increase of {beta}-diversity by [~]10% supporting species-sorting. Additionally, we tested {beta}-deviations from random assembly as a proxy of neutral processes. Findings indicate that enhancing structural heterogeneity can shift forests from patch-dynamics dominance towards mass-effect and species-sorting, offering a promising pathway to restore biodiversity in managed landscapes.

Matching journals

The top 7 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.