Invertebrate feeding and emergence timing vary among streams along a gradient of riparian forest composition

1. The quality of allochthonous organic matter influences the transfer of energy and nutrients through recipient food webs. We investigated the effects of variation in the composition of riparian forests (deciduous, mixed, coniferous) on the elemental imbalance between basal resources and consumers in streams, on consumer feeding and on potential feedbacks to riparian systems via emergent aquatic insects. 2. We tested for differences in elemental stoichiometry (carbon/nitrogen/phosphorus; C/N/P) and stable isotopes (?¹³C and ?¹⁵N) between deciduous (red alder, Alnus rubra) and coniferous litter (western hemlock, Tsuga heterophylla) and among abundant stream invertebrates from streams draining different riparian forests (deciduous, mixed, coniferous). We then assessed shredder feeding preferences (of the trichopteran, Lepidostoma unicolor) for litter incubated in streams with these different forest types and quantified differences in emergence of aquatic and semiaquatic insects among streams. 3. Both initial (non‐incubated) and stream‐incubated A. rubra litter had lower C/N and C/P and were more depleted in ?¹³C and more enriched in ?¹⁵N, than T. heterophylla litter. The stoichiometry of invertebrate tissue did not vary significantly among taxa or with riparian forest composition. A predator (the plecopteran Chloroperlidae) and a collector‐gatherer (the ephemeropteran Paraleptophlebia gregalis) from mixed and coniferous forest streams were more enriched in ?¹³C and ?¹⁵N isotopes than those from deciduous streams, suggesting that low availability of palatable, N‐rich A. rubra litter may constrain energy flow and nutrient fluxes up through the food web in systems with little or no A. rubra. 4. Consumption of A. rubra litter by L. unicolor was most rapid when the litter had been incubated in streams draining deciduous forests, whereas consumption of T. heterophylla litter was not influenced by the composition of the riparian forest. 5. Peak insect emergence from coniferous forest streams occurred 1 month earlier and at 2–3× higher density than from mixed and deciduous‐forest streams, but total biomass of emerging insects throughout the study period was not different between forest types. Assemblages of emerging insects were different between deciduous and coniferous forest streams, and taxon richness and diversity were nearly 2× greater from deciduous than from coniferous forest streams. 6. Forest composition influences stream invertebrate feeding and could have reciprocal feedbacks onto riparian systems via altered insect emergence.