Aquatic food webs use C derived from 2 sources: autochthonous (derived from inside the water body) and allochthonous (derived from outside the water body). Various autochthonous sources are available to consumers (e.g., algae, mosses, macrophytes), and stable isotope analysis can be used to differentiate among C sources and to determine which sources sustain aquatic communities. The goal of our study was to determine which C sources contribute to the autochthonous pathways in 2 oligotrophic rivers, the Upper Salmon and Big Salmon, in New Brunswick (Canada). Samples were taken from headwaters to river mouth (above head of tide) at 17 study sites distributed across subbasins and stream orders. Vegetation samples included macrophytes, filamentous algae, biofilm, and aquatic bryophytes. Macroinvertebrate taxa in the scraper (grazer) functional feeding group were collected to represent primary consumers that feed primarily on autochthonous food sources. Macrophyte and cyanobacteria d 13 C values were correlated with scraper d 13 C values (r = 0.55 and 0.47, respectively), but d 13 C of most scrapers was more depleted than d 13 C of these food sources (.40% and .90%, respectively). Trophic fractionation is primarily an enrichment process, so these results indicated that macrophytes and cyanobacteria were not important food sources for primary consumers. Filamentous algae and biofilm d 13 C values were poorly correlated with scraper d 13 C values (r = 0.36 and 0.40, respectively), and .60% of the values were too depleted for these sources to be important food sources for scrapers. Ninety-eight percent of scraper d 13 C values were enriched relative to bryophyte d 13 C values, and bryophyte d 13 C values were correlated with scraper d 13 C values (r = 0.53), particularly when samples from slow-flowing habitats were removed from the analysis (r = 0.76). Spatial distributions of 2 important bryophyte taxa differed. Fontinalis sp. was abundant in headwater streams, and Drepanocladus sp. was abundant in low-order streams. Scrapers in low-order streams seemed to depend more on Fontinalis sp. than scrapers in highorder streams depended on Drepanocladus sp. Our results suggest that in low-productivity, nutrient-limited rivers, reduced availability of more preferred food sources (e.g., epilithic algae) might cause primary consumers to switch to marginal food sources, such as bryophytes. Further research is required to confirm this hypothesis and to eliminate confounding factors.