Summary 1. In large deep oligotrophic lakes, the shallow nearshore waters may provide the most important habitat for animals to feed and breed, and it is this area of the lake where humans are most likely to have initial impacts as the shoreline is developed. Nutrients in fertilizers, sediments and sewage effluents are likely to be rapidly intercepted by nearshore algae at the lake edge, having heterogeneous effects nearshore before offshore effects are noted. 2. Here we examined the spatially explicit effects of residential development on nearshore periphyton communities in three large deep oligotrophic lakes that have all experienced modest residential development in the Pacific Northwest of the United States. We demonstrate that substantial nearshore changes in the basal food web are detectable even with low levels of shoreline development. These changes can potentially affect whole‐lake food web dynamics. 3. For our primary study site (Lake Crescent, Washington, USA), we found that algal biomass and accumulation of detritus were higher at developed sites. In addition, both macroinvertebrate and periphyton communities exhibited a shift in composition with more detritivores and filamentous green algae at developed sites. These differences were more pronounced during the spring than at other times of year. 4. A complementary investigation of field patterns in Priest Lake and Lake Pend Oreille (Idaho, USA) suggested that, although spatial and temporal patterns were idiosyncratic, indicators of productivity and the presence of filamentous green algae were generally higher at developed sites across lakes. 5. Stable isotope signatures and water column nutrients were not useful in distinguishing developed and undeveloped sites, increasing the potential usefulness of periphyton monitoring during early stages of lake development. 6. A laboratory investigation suggested that common macroinvertebrate grazers assimilated a much greater proportion of diatoms than the filamentous green algae that are associated with fertilization at developed sites. 7. These findings have at least two clear implications: (i) periphyton may be used to detect human impacts before disturbance is evident in offshore monitoring programmes and (ii) nearshore impacts in response to modest residential development have the potential to disrupt lake food web dynamics.

In large deep oligotrophic lakes, multiple lines of evidence suggest that the shallow nearshore water provides disproportionately important feeding and breeding habitat for the whole-lake food web. We examined the trophic importance of the nearshore environment, human impacts nearshore, and several approaches to disturbance detection in a deep (190 m) oligotrophic lake with relatively modest residential development. In Lake Crescent, on the Olympic Peninsula of Washington (USA), stable isotope analysis demonstrated that apex salmonid predators derived more than 50% of their carbon from nearshore waters, even though this nearshore water accounted for only 2.5% of total lake volume. Unfortunately, it is this land–water interface that is initially degraded as shorelines are developed. We hypothesised that under these conditions of relatively modest disturbance, the effects of residential development would be strongly localised near to shore. Indeed, we found striking differences between developed and undeveloped sites in periphyton and associated organic matter, though there were no offshore signals of human impact in water nutrient analysis or paleolimnological investigations. Together, these results suggest that nearshore biological monitoring should be integrated in lake management plans to provide ‘early warning’ of potential food-web repercussions before pollution problems are evident in open water and comparatively intractable.