Supporting information file.

Document containing 3 supplementary texts, 7 supplementary tables, 5 supplementary figures, and references. <b>S1 Text</b>. Detailed bait design and methods. <b>S2 Text.</b> Description of <i>Tapinoma cf. sessile</i>. <b>S3 Text</b>. Analysis of variation in bait heights. <b>S1 Table.</b> Nutritional content of tuna and jam baits. <b>S2 Table.</b> Complete list of taxonomic keys used in this study. <b>S3 Table.</b> Substitute species biomass values. List of species for whom we could not calculate average individual weights and the species’ values used as their substitute weights. <b>S4 Table.</b> Occurrence of ant species in arboreal and terrestrial baits across seven forest and seven urban habitats in central North Carolina, USA. <b>S5 Table.</b> Pairwise comparisons of changes in ant community composition. <b>S6 Table.</b> Contrasts of the three-way interaction between stratum, habitat, and nutrient from the model predicting ant biomass. <b>S7 Table</b>. Species specific biomass estimates and model results for four urban biomass-dominant species. <b>S1 Fig.</b> Bait height varied with habitat and tree species. <b>S2 Fig.</b> Ant diversity at baits did not significantly differ between habitats and strata. <b>S3 Fig.</b> Changes in ant communities between the ground and canopy were due to species turnover. <b>S4 Fig.</b> Changes in ant communities between the forest and urban habitats were due to species turnover. <b>S5 Fig.</b> Ant biomass was significantly greater on the ground than in the canopy. (DOCX)