Enter the Mosaic: Aquatic-Terrestrial Reciprocal Fluxes and Food Webs Are Dynamically Interdependent Across Space and Through Time

Abstract

1. Decades-old research describes dynamic interdependence among aquatic and terrestrial food webs, leading to calls for integrating cross-ecosystem linkages with landscape ecology to evaluate dynamics of spatially-subsidised food webs. Though development of meta-community theory has suggested that such spatial dynamics may help sustain biodiversity, empirical data remain limited. In northern Yellowstone National Park, over a century of terrestrial wildlife dynamics, including the extirpation and subsequent reintroduction of wolves, have contributed to a habitat mosaic in which stream-riparian ecosystems are dominated by either woody or herbaceous vegetation. In the context of this habitat mosaic, we addressed the overarching questions: (1) Are habitat mosaics associated with spatial and temporal variation in reciprocal fluxes and linked food webs and (2) how do biodiversity, organism traits and species interactions influence, and are they influenced by, that spatial and temporal variation?
2. From 2019 to 2021, we intensively sampled eight headwater streams to characterise reciprocal fluxes of aquatic and terrestrial invertebrates and the patterns of potential responses by fish, birds, bats and spiders. We evaluated sites individually as well as how they contributed to a meta-community.
3. We found that local stream-riparian ecosystems contributed to a mosaic in which reciprocal fluxes of invertebrates among local patches were asynchronous and tracked by both aquatic and terrestrial consumers in ways mediated by organism traits. Within sites, aquatic and terrestrial invertebrate fluxes were seasonally asynchronous with each other, but these patterns varied from site to site. Across the mosaic, comparisons of daily aquatic insect emergence varied from 25% to 167% among streams and did so variably throughout the year, revealing asynchronous dynamics created at the meta-community scale. Daily inputs of terrestrial invertebrates were similarly asynchronous across the mosaic, varying from 14% to 170%. These asynchronies were positively correlated with invertebrate beta diversity and associated with varying riparian vegetation, stream temperature, and flow regimes. In turn, in situ consumers tracked the allochthonous invertebrate prey in ways that were mediated by site context (i.e., local habitat characteristics) and consumer traits (e.g., range, foraging strategy and breeding requirements).
4. Based on these observations as an example, we infer there is not one way for food webs to be reciprocally and spatially linked, but multiple ways that can vary both across a spatial mosaic and through time. Our findings provide empirical evidence suggesting potential relationships between habitat complexity and the maintenance of biodiversity via aquatic-terrestrial reciprocal fluxes and dynamic interdependence across mosaics.