The extent to which nutrients from Pacific salmon are transported to riparian areas may be influenced by differences in spawning behavior among species. Chum salmon Oncorhynchus keta, pink salmon O. gorbuscha, and sockeye salmon O. nerka typically spawn in dense aggregations, while species like steelhead O. mykiss and coho salmon O. kisutch spawn at lower densities. The contribution of nutrients to riparian vegetation was compared at two watersheds in western Washington, Griffin Creek (used by coho salmon) and Kennedy Creek (used by chum salmon). Salmonberry Rubus spectabilis foliage was collected at the channel edge above and below barriers to spawning salmon and at 20, 50, and 100 m upslope from the stream and analyzed for nitrogen stable isotope ratio (δ15N, an indicator of salmon‐derived nitrogen), total nitrogen (N), and phosphorus (P) content. Cover, plant density, and the species richness of shrub and understory vegetation were compared between sites with and without salmon. The δ15N values in salmonberry leaves were higher at sites with salmon than at corresponding distances from the channel at sites without salmon at Kennedy Creek but not Griffin Creek. Salmonberry foliage adjacent to salmon spawning reaches possessed significantly higher levels of total N and P in both watersheds. Nitrogen content was positively associated with δ15N values at the Kennedy Creek sites but not at the Griffin Creek sites. At Kennedy Creek, shrub species diversity and understory plant density and species diversity were higher at sites with salmon than at sites without salmon. These results suggest that areas bordering streams utilized by high‐density‐spawning species like chum salmon receive a substantial nutrient contribution from the fish and that this subsidy influences the vegetation. We did not see clear evidence for a similar nutrient contribution from coho salmon.

No abstract

Red alder (Alnus rubra Bong.) commonly dominates the overstory in managed riparian forests of the Pacific Northwest. Although it is assumed that competition with a dense shrub layer inhibits succession from hardwoods to shade-tolerant conifers within these forests, this assumption has not been explicitly tested. We hypothesized that seed availability rather than competitive interactions largely shape patterns of conifer regeneration in managed riparian forests. At nine locations in western Washington, we established paired transects in riparian sites adjacent to abundant versus few or no sources of seed to examine the importance for conifer regeneration of dispersal limitation, rooting substrate, and vegetation cover. Frequency and density of regeneration were significantly greater in transects within 100 m of remnant forest patches than in paired sites at greater distances (mean frequencies of 59 vs. 18% and densities of 0.23 vs. 0.16 trees/m2, respectively). Where seed sources were present, regeneration was positively associated with coarse woody debris and negatively associated with fine litter. Regeneration was most abundant in plots with <10% herb or shrub cover; however, for most species, recruitment occurred across the full range of understory plant cover. We detected no relationship between regeneration density and overstory conifer or hardwood cover. Our results suggest that, in managed forests, conifer regeneration is largely limited by seed availability and only secondarily by competitive interactions or substrate conditions. In managed landscapes, conventional strategies of vegetation control are not likely to be as effective in increasing conifer regeneration in riparian forests as managing for seed sources through green-tree retention.

No abstract