Foraging by consumers acts as a biotic filtering mechanism for biodiversity at the trophic level of resources. Variation in foraging behaviour have cascading effects on abundance, diversity, and functional trait composition of the community of resource species. Here we propose diversity at giving-up density (DivGUD), when foragers quit exploring a patch, as a novel concept and simple measure to quantify these effects at multiple spatial scales. In experimental landscapes, patch residency of wild rodents decreased local α-DivGUD (via elevated mortality of species with large seeds) and regional γ-DivGUD, while dissimilarity among patches in a landscape (ß-DivGUD) increased. Thus, DivGUD provides a framework linking theories of adaptive foraging behaviour with community ecology allowing to investigate cascading indirect predation effects across multiple trophic levels e.g. the ecology-of-fear framework; feedbacks between functional trait composition of resource species and consumer communities; and effects of inter-individual differences among foragers on the biodiversity of resource communities.

We created resource landscapes with nine experimental food patches, 10 m apart from each other in a 30 x 30 m grid (deployment). Each patch contained a 30 cm x 30 cm x 3 cm seed tray with 2 litres fine grained sand (0.1-0.5 mm grain size) and a mix of seeds from 8 different plant species, varying in functional traits of seeds (Table 2) such as size, mass, nutritional value, or presence of husks (initial density of 20 seed items of 8 seed species per patch). Trays were equipped with a camera trap to identify foraging animals and covered with Perspex corrugated plastic as rain protection. After exposure of food patches to foragers overnight, we sifted the remaining seeds from the sand and counted them by species to obtain seed-specific giving-up density (GUD, items per liter of sand). Data represents seed counts from food patches after exploitation by wild Norway rats that extracted seeds from the sand. PatchID reports replicate (R3-R9, chronologically, missing numbers indicate replicates without foraging events), grid ID (E4 or E5), and tray location (T1-T9); residency (minutes a rat spent in pach) were observed with camera traps (in two occasions (*) cameras were foggy during parts of night and residency was extrapolated from unspecific, cumulative GUD), followed by item counts of the 8 seed species.

Missing numbers in patch ID indicate replicates without foraging events, i.e. replicates where no seeds were taken and the initial density of 20 seed items of 8 seed species remained.