Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Salmon increase forest bird abundance and diversity

  • Marlene A. Wagner ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Earth to Ocean Research Group, Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia, Canada, Hakai Institute, Heriot Bay, British Columbia, Canada

  • John D. Reynolds

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Supervision, Writing – review & editing

    Affiliations Earth to Ocean Research Group, Department of Biological Sciences, Simon Fraser University, Burnaby, British Columbia, Canada, Hakai Institute, Heriot Bay, British Columbia, Canada

Salmon increase forest bird abundance and diversity

  • Marlene A. Wagner, 
  • John D. Reynolds


Resource subsidies across ecosystems can have strong and unforeseen ecological impacts. Marine-derived nutrients from Pacific salmon (Onchorhycus spp.) can be transferred to streams and riparian forests through diverse food web pathways, fertilizing forests and increasing invertebrate abundance, which may in turn affect breeding birds. We quantified the influence of salmon on the abundance and composition of songbird communities across a wide range of salmon-spawning biomass on 14 streams along a remote coastal region of British Columbia, Canada. Point-count data spanning two years were combined with salmon biomass and 13 environmental covariates in riparian forests to test for correlates with bird abundance, foraging guilds, individual species, and avian diversity. We show that bird abundance and diversity increase with salmon biomass and that watershed size and forest composition are less important predictors. This work provides new evidence for the importance of salmon to terrestrial ecosystems and information that can inform ecosystem-based management.


Resource availability and movement are major processes shaping ecosystem structure and function [1]. Resource subsidies are prevalent across landscapes [2,3], and can have profound direct and indirect impacts on recipient community structure, influencing primary productivity, trophic interactions, and predator-prey relationships [1,46]. Coastal streams provide bidirectional highways for nutrient transport [7], and support the anadromous and semelparous life-history of Pacific salmon (Oncorhynchus spp.). Salmon deliver an annual and predictable flux of nutrient subsidies from marine to terrestrial systems, creating an opportune natural experiment to examine effects of variation in resource subsidies across ecosystem boundaries.

Pacific salmon acquire 99% of their body mass after leaving freshwater streams to grow and mature at sea [8]. When they return to spawn in natal streams, they bring a seasonal influx of marine-derived nutrients that enhance both freshwater and terrestrial productivity by fertilizing otherwise nutrient-poor watersheds with nitrogen and phosphorous [912]. Salmon carcasses are transferred to adjacent terrestrial habitat by bears, wolves, and other primary consumers, as well as through flooding and hyporheic flow [1315].

After transfer, marine-derived nutrients move through several trophic pathways. First, they enhance primary production, favoring plant growth and structural complexity [5,16,17] and they influence the diversity of understory vegetation [12,18]. Birds respond to plant structure and composition, and discriminate habitat at fine scales associated with foliage density and geometry [1921]. Furthermore, herbivorous insects are attracted to foliage with elevated levels of nitrogen [2224], and both terrestrial and aquatic invertebrates can be more abundant on salmon streams [10,25,26] (but see [27,28]). As the combination of habitat (vegetation) and food (invertebrates) are essential resources for birds, salmon may play a role in shaping avian communities.

Songbird densities have been positively associated with artificial fertilization (N) of lakes [29]. Previous work also suggests that songbirds achieve higher densities in the presence of salmon across 15 small streams with and without salmon in Alaska [30], and this was supported in a comparison of bird densities above and below salmon barriers in two salmon-bearing and one stream without salmon at our study streams in British Columbia [31]. Additionally, both density and (Shannon’s) diversity of bird communities increased in estuaries in British Columbia along with salmon biomass [32,33]. However, no study has explored how salmon-derived nutrients affect birds in riparian forests across a range of salmon-spawning magnitudes. Our comparisons involved a wide span of salmon densities, as opposed to comparisons where salmon are either presence or absent, and we test for correlations with bird communities across as opposed to within watersheds [31]. We also standardize elevation and other factors that may influence bird communities inland by restricting site locations to forests near the mouths of streams.

Materials and methods

Study system

Our research was conducted under Simon Fraser University Animal Care Protocol #1044B-12, and approval from Heiltsuk Tribal Council. We conducted our study in Heiltsuk First Nation territory near Bella Bella, along the Central Coast of British Columbia, Canada (52.1619N, 128.1450W). This region is in the Western Hemlock Coastal Biogeoclimatic Zone, characterized by a cool, maritime climate, heavy rainfall (>3 m per year), and forests dominated by coniferous tree communities of western hemlock (Tsuga heterophylla), and Sitka spruce (Picea stichensis), western red cedar (Thuja plicata), and Amabilis fir (Abies amabilis). The only deciduous tree in the region is red alder (Alnus rubra). Understory vegetation consists of stink currant (Ribes bracteosum), blueberry and huckleberry (Vaccinium spp.), salmonberry (Rubus spectabilis), salal (Gaultheria shallon), false azalea (Menziesia ferruginea), red elderberry (Sambucus racemosa), and devil’s club (Oploplanax horridus). Further habitat descriptions are in Mathewson et al. 2003 and Hocking & Reynolds 2011.

Study sites in 14 watersheds that span approximately 60 km of the coast were chosen because they were adjacent to streams supporting a wide range of salmon (0 to 122,454 fish; Fig 1). Selective logging occurred along some streams in the mid-twentieth century, however no streams had been clear-cut and other modern-day anthropogenic disturbance is slight or nonexistent. All sites were only accessible by boat from the sea.

Fig 1. Location of 14 study streams along the central coast of British Columbia.

Breeding birds

We conducted standard 10-minute point-count surveys [34] over two years across the streams in 14 watersheds, where all birds seen or heard were recorded. Point-count stations were positioned in the riparian forest on alternate sides of the stream, beginning 50 m inland from the estuary. Each point count was located 50 m upland from the stream, for a total of 5 point-counts on each stream, located 100 m apart. Each station was surveyed 4 times between May 16 and July 23, 2012, and 3 times between May 16 and July 17, 2013. All surveys were completed between 6:30 and 10:20 am PST. Each stream was visited once in a rotation before continuing the process over again with starting points reversed on subsequent visits. Point-count surveys were not conducted in heavy rain or wind (>3; Beaufort Scale), or at stream noise levels rated more than moderate on a standard ordinal scale (0–5; [31]). If conditions prevented censusing, we returned during the same rotation under better conditions.

Birds detected as fly-overs, non-forest dwelling birds (e.g. seabirds, herons, gulls, etc.), and birds detected on fewer than two surveys (early spring migrants and other non-breeders) were excluded from analyses. We also excluded American Dipper (Cinclus mexicanus), an obligate riverine songbird that will include salmon eggs and fry in their diet [35], because we were interested in the influence that salmon may have on birds through indirect pathways. We truncated the bird data to include only detections within a 100 m radius due to the small overall plot size (~10 ha), distance between point-counts, and to ensure that resources within the riparian zone were available to our sample population. If birds were detected at more than one count, we removed subsequent detections from analyses. We calculated total bird density as the relative abundance at each point-count for all bird species combined [36]. We then separated the analyses into generalist, insectivore, and frugivore foraging guilds, and we examined the six most commonly detected species across all study sites for each year. Finally, because local bird species richness and available energy resources should be positively correlated, we calculated two diversity indices, the effective number of species [37], and species richness (the number of species observed at each point count station). The effective number of species represents the number of equally common species, is the exponential of Shannon’s Diversity, and has the advantage of scaling linearly with species richness [3840]. Subsequently, bird response variables were grouped into three general categories: 1) relative abundance of all birds and foraging guilds, 2) relative abundance of individual bird species, and 3) avian diversity measures.


Collaboration between the Department of Fisheries and Oceans, the Heiltsuk Integrated Resource Management Department, and our research group at Simon Fraser University has resulted in the creation of a large salmon dataset within the region, with salmon counts conducted annually each fall. On each stream, the number of pink (O. gorbuscha) and chum (O. keta) salmon, both live and dead, were counted by walking streams before, during, and after the peak spawning period. Other species of salmon were excluded from analyses because they comprise less than 5% of the salmon populations in these streams and tend to spawn farther upstream.

We used salmon count data from 2009 to 2011 to quantify a 3-year average salmon biomass metric (Table 1). On streams where 3 counts had been repeated, the area-under-the-curve (AUC) method was used to estimate the total number of spawning fish for the year [41]. Otherwise, the peak counts of live + dead were used as totals, which result in very similar population estimates to the AUC method [12]. Biomass was estimated using: where N = the number of adult salmon, i = the salmon species; pink or chum, and W = regional salmon mass estimates (1.2 and 3.5 kg, respectively; [12]). We also calculated two alternative metrics of salmon density: and where L = the spawning length of the stream and W = the mean bankfull width (the maximum width of a stream channel before flooding) of the stream, as these are common measures of salmon indices found in the literature. Preliminary analysis using univariate linear models with each salmon metric indicated that they were highly correlated and that our salmon biomass estimates (rather than density) explained the most variation in our bird data, so we used this measurement in all subsequent analyses (S1 Table).

Table 1. Stream-specific features.

Watershed, stream, and salmon metrics across 14 streams the central coast of British Columbia. Salmon biomass was calculated from 2009–11 mean counts of spawning adults.

Forest habitat

Forest habitat variables were quantified from data collected on 50 m vegetation transects running perpendicular from the stream to point-count stations. Diameter at breast height (DBH), and percent cover by species and height class (1.3–15 m, 15–25 m, and >25 m) were recorded for all trees greater than 2.5 cm DBH in a 6 m belt along each transect. Shrub cover was recorded as the percentage of cover by species and height class (<0.5 m, 0.5–1.3, and >1.3 m; including saplings with DBH <2.5 cm) in five 1 m2 quadrats located at 5, 15, 25, 35, 45 m along the transects. Vegetation sampling methods were adopted from modified protocols described in Christie and Reimchen (2008) and Field and Reynolds (2011).

The DBH data were used to calculate stand basal area per plot for each dominant conifer tree species and subjected to a principal components analysis (PCA) to determine the major axes of change in conifer composition. Mean percent cover estimates from the six height classes for each transect were used to determine foliage height diversity (FHD; [42]). We also calculated the mean percent cover of red alder, a nitrogen-fixing species and the only deciduous tree in the region, and the mean percent cover of shrubs per transect (Table 2).

Table 2. Forest composition features.

Mean stand basal area for conifer species, mean percent cover for red alder and shrubs, and foliage height diversity (FHD) across 14 streams along the central coast of British Columbia. The FHD represents the distribution in amount of vertical canopy within plots.

Principal components of conifer composition PC1 and PC2 explained 23.4% and 21.2%, respectively, of the cumulative variation in forest species across 70 vegetation plots (S2 Table). PC1 for conifer composition indicates a shift from spruce-dominated to cedar-dominated riparian areas, with negative loadings for stand basal area of Sitka spruce (-0.65) and Amabilis fir (-0.38), and high stand basal area of western red cedar (0.64). PC2 represents a shift from low Amabilis fir (-0.45) to high Western hemlock (0.61) and snags (0.63).

Watershed size

Variation in geomorphology shapes riparian structure within a watershed [43]. Watershed size can influence cross-boundary nutrient transfer by mediating both predator access to salmon and terrestrial inputs [44,45]. Watershed catchment area (km2) was calculated in a Geographical Information System using iMAPBC ([46], Table 1). To account for stream size and length, we used field measurements of bankfull width and the mainstream salmon spawning length [12]. We combined these three variables to compute a PCA for overall watershed size for each of our study streams. The first principal component axis explained 82% of the overall variation in catchment area, bankfull width, and length of stream used by salmon, with all variables loading positively (S2 Table). This was therefore the only metric used in subsequent analyses to describe watershed characteristics.


To test hypotheses about the influence of salmon-spawning biomass on birds, we used hierarchical partitioning and Akaike’s Information Criterion adjusted for small sample sizes (AICc) with mixed-effects models. First, we modeled the individual contribution of each of our five forest habitat covariates (conifer composition PC1, conifer composition PC2, FHD, red alder cover, and shrub cover), on each of our bird response variables. The top-ranked forest habitat variable was then used to build candidate model sets. We used this method to limit insertion of forest habitat covariates in final candidate model sets [47], as our intent was to test overall influences of factors that may drive bird distribution and diversity in the watershed, and not quantify fine-scale habitat selection by various guilds or species.

We created a final candidate suite of seven models for each bird response variable, using the unique forest habitat covariate retained in our initial model selection process (S3 Table). We fit models to describe bird communities as a function of salmon biomass, forest habitat, and watershed size with site and point as random factors in all models to account for the repeated measures and spatial autocorrelation inherent in our study design [48]. We tested for multicollinearity among predictor variables. All variance inflation scores were less than two, and correlation coefficients were below 0.6. Yearly variation was included in all models as a two-level factor (2012 and 2013). We examined the variance structure of residuals to ensure assumptions of normality were met. Models were averaged to obtain weighted parameter estimates for highly competitive models (ΔAICc < 2) using the natural method [49]. We standardized individual coefficients to enable direct comparison of effect sizes across variables [50]. All analyses were completed in the R statistical program, version 3.2.2 [51], using packages AICcmodavg [52], Mumln [53], nlme [54], and vegan [55].


We detected 55 species of birds across our point-count surveys over both years, of which 35 were retained for final analyses (S4 Table). The six most commonly detected species were, in order of decreasing abundance; Pacific-slope Flycatcher (Empidonax difficilus), Pacific Wren (Troglodytes pacificus), Golden-crowned Kinglet (Regulas satrapa), Townsend’s Warbler (Dendroica townsendi), Varied Thrush (Ixoreus naevius), and Swainson’s Thrush (Catharus usulatus).

Streams with higher salmon biomass had a greater relative abundance of all birds detected, and this was true within each of our foraging guilds of generalists, insectivores, and frugivores (Figs 2 and 3). We also observed a positive relationship between salmon biomass and effective number of species and species richness (Fig 3). Each individual species, except for Pacific-slope Flycatcher and Varied Thrush, showed strong evidence of higher relative abundances as salmon biomass increased (Fig 4).

Fig 2.

Relationships between relative abundance of all birds and A) salmon biomass, B) conifer composition, and C) watershed size. Salmon biomass was log-transformed and 2012 (blue triangles) and 2013 (green circles) data points were jittered to prevent over-plotting. Conifer composition and watershed size metrics are based on PCA (see methods; S2 Table).

Fig 3. Standardized coefficients with 95% confidence intervals from the averaged mixed effects model (ΔAICc < 2) for all birds, generalists, insectivores, frugivores and diversity metrics of effective number of species and richness as a function of salmon, forest habitat, and watershed size.

Fig 4. Standardized coefficients with 95% confidence intervals from the averaged mixed effects model (ΔAICc < 2) for the 6 most common species as a function of salmon, forest habitat, and watershed size.

Most top-ranked models in our final analysis retained a forest habitat covariate (Table 3). Abundances for all birds combined, Pacific Wren, Townsend’s Warbler, Golden-crowned Kinglet, and both diversity measures increased with salmon biomass and decreased conifer PC1, indicating a preference for stands that are fir- and spruce-dominated (Figs 24). Generalist and frugivore abundances also increased with salmon biomass and with shrub cover (Fig 3). Insectivore and Swainson’s Thrush abundances increased with salmon biomass but decreased with red alder (Figs 3 and 4; S5 Table).

Table 3. Model selection results (ΔAICc < 2) depicting avian response to stream, salmon, and habitat features on 14 streams along the central coast of British Columbia.

Year was included as a covariate in all models but is not included in the table for clarity.

Overall watershed size characteristics that included width and length of the streams, and catchment size, explained little of the variation in our bird metrics. However, overall watershed size was included in two highest-ranking models (insectivores and Pacific-slope Flycatcher; Figs 3 and 4; Table 3), and occurred across most averaged models (except Pacific Wren, Townsend’s Warbler). However, confidence intervals for effect sizes overlapped zero, leading to ambiguity in results.


Our study confirms that salmon may provide an important indirect resource subsidy for songbirds. We controlled for both watershed size and forest habitat while examining the influence of salmon on avian communities across 14 streams that supported a wide range of variation in biomass of spawning salmon. Many of the species in this study are migratory, and therefore they do not have direct exposure to salmon in the fall (S4 Table), but are responding to their influence in the spring. This provides evidence for a seasonal legacy effect of salmon. Every bird metric correlated positively with salmon biomass, with the exception of the relative abundance of Pacific-slope Flycatcher (which was not explained well by any model), and the relationship with salmon was greater than either forest habitat or watershed size. Our study corroborates with previous work on two of our study streams that found increased abundance of Pacific Wren, Swainson’s Thrush, and Golden-crowned Kinglet on salmon-bearing reaches compared to non-salmon bearing reaches above waterfalls [31]. Additionally, our results also provide strong evidence that inputs of salmon even at low levels on streams appear to elevate bird populations above those without salmon.

There are likely several pathways that provide birds with the benefits from salmon biomass [56]. Salmon enhance primary productivity in aquatic ecosystems [57], thus increasing the density of common invertebrate taxa, at least in the spring [10,26,58,59]. In the spring, emergent aquatic insects can comprise 50–90% of resident bird diet, at a time when other invertebrate populations are low [60,61]. Concentrations of both migrant and resident birds correlate strongly with the timing of emergence [62,63] and emergent population abundance [64]. Many terrestrial invertebrates oviposit directly on salmon carcasses [65]. For example, Calliphoridae exhibit high productivity when breeding on chum carcasses and contribute to a significant increase in overall invertebrate numbers in the fall [30,45]. Furthermore, herbivorous insects select vegetation with high levels of nitrogen and may attain higher densities on vegetation subsidized by salmon [2224,66]. We also recorded higher numbers of terrestrial invertebrates at nests in forests along streams with more fish in the breeding season during field collections (Wagner, unpublished data).

Forest habitat influenced many of our avian response variables. Overall, conifer composition was the strongest forest habitat influence on abundance for all birds and for Pacific Wren, Townsend’s Warbler, and Golden-crowned Kinglet. These results indicate a preference for spruce- and fir-dominated forests and an avoidance of cedar. Tree species composition is important to insectivores [67,68], and members of the cedar family have secondary chemicals that reduce invertebrate colonization of their bark and branches. Cedars host fewer beetles compared with both Western hemlock and Amabilis fir, and this in addition to repellant phenolics may reduce the prevalence of bark-gleaning or other foraging birds [69,70]. Surprisingly red alder, which fixes atmospheric nitrogen and is the only deciduous tree present at study sites, was a negative predictor for both insectivores and Swainson’s Thrush, which was contrary to our expectations, though 95% confidence intervals overlapped zero in both analyses. Shrub cover was a positive predictor for both generalists and frugivores, and this may reflect an attraction to berry-producing shrubs such as salmonberry, which are more dense along streams with more salmon [12].

Food availability is a major driver of songbird productivity [71]. Limitation of food resources can mediate spatial aggregation, nest success, and alter life history strategies. A passerine that does feed on salmon and eggs when available, the American Dipper, is more likely to disperse away from breeding territories in the winter in habitats without salmon, and enjoy higher breeding success with salmon [72]. While they consume salmon fry and eggs, much of their diet consists of aquatic invertebrates [73,74], suggesting indirect salmon subsidies also influence habitat use by this species. There is also evidence of salmon-derived nutrients in the diet of Pacific Wrens in two of the streams that we studied [75]. Further research is needed to see what additional mechanisms lead to the increased abundance of the birds that do not consume salmon or their eggs. For example, they could have higher nesting success on streams with more salmon, or they may preferentially aggregate on such streams [76].


Our study shows that salmon biomass has a stronger relationship with bird density and diversity across watersheds than forest composition or watershed size. As salmon impact terrestrial taxa even in severely degraded habitats [77], and recently restored habitats [74], our results emphasize the strength and importance of cross-boundary subsidies. The current fishery management paradigm of maintaining stock levels for next generation recruitment does not consider essential landscape-scale processes that support ecosystem function. Our results also reflect the importance of considering cross-boundary interactions during the current trend towards ecosystem-based management [56,7880]. As salmon runs persist well below historic levels in many parts of their North American range [81], an ecosystem-based approach to managing salmon and river ecosystems is necessary to maintain holistic ecosystem function.

Supporting information

S1 Table. Preliminary univariate models comparing three salmon density metrics.


S2 Table. Principal component factor loadings for conifer composition and watershed size.


S3 Table. Model selection results of all models depicting avian response to salmon and habitat features on 14 streams along the Central Coast of British Columbia.


S4 Table. Species, their associated foraging guilds, and migratory status of birds detected on point-count surveys in 2012 and 2013 along the Central Coast of British Columbia.


S5 Table. Fixed-effect estimates (standardized regression coefficients) and standard errors for averaged candidate models (ΔAIC<2) describing relative bird abundance as a function of salmon biomass, forest habitat, and watershed size across 14 streams along the Central Coast of British Columbia.



We thank the Heiltsuk First Nation for allowing us to work in their traditional territory. We also thank Mathias Herriges, Howard Humchitt, Allen Larocque, Jane Pendray, Mandy Proudman, Ben Rabinovich, Taylor Watson, and Ben Zyla for assistance in the field. We thank Mike Reid, Sandra Vishloff, the Heiltsuk Integrated Resource Management Department, Pacific Wild, and Larry Jorgenson at Qqs Project Society for logistical help. We appreciate statistical support from Morgan Hocking, Carl Schwarz, Nicola Smith, and the Earth to Ocean’s Stats-beerz Club. We also thank Thomas E. Reimchen and an anonymous reviewer for their helpful reviews of the manuscript.


  1. 1. Polis GA, Anderson WB, Holt RD, Anderson B, Polis A. TOWARD AN INTEGRATION OF of ECOLOGY: The Dynamics Food Webs Subsidized Spatially. Annu Rev Ecol Syst. 1997;28: 289–316.
  2. 2. Loreau M, Holt RD. Spatial flows and the regulation of ecosystems. Am Nat. 2004;163: 606–615. pmid:15122506
  3. 3. Leroux SJ, Loreau M. Subsidy hypothesis and strength of trophic cascades across ecosystems. Ecol Lett. 2008;11: 1147–56. pmid:18713270
  4. 4. Lindeman RL. The Trophic-Dynamic Aspect of Ecology. Ecology. 1942;23: 399–417.
  5. 5. Helfield JM, Naiman RJ. Effects of Salmon-Derived Nitrogen on Riparian Forest Growth and Implications for Stream Productivity. Ecology. 2001;82: 2403–2409.
  6. 6. Knight TM, McCoy MW, Chase JM, McCoy KA, Holt RD. Trophic cascades across ecosystems. Nature. 2005;437: 880–883. pmid:16208370
  7. 7. Payne LX, Moore JW. Mobile scavengers create hotspots of freshwater productivity. Oikos. 2006;115: 69–80.
  8. 8. Quinn TP. The Behavior and Ecology of Pacific Salmon and Trout [Internet]. UBC Press; 2005. Available:
  9. 9. Willson S.M. Gende MF, Marston BH. Fishes and the forest: Expanding perspectives on fish-wildlife interactions. Bioscience. 1998;48: 455–462.
  10. 10. Wipfli MS, Hudson JP, Chaloner DT, Caouette JP. Influence of salmon spawner densities on stream productivity in Southeast Alaska. Can J Fish Aquat Sci. 1999;56: 1600–1611.
  11. 11. Hicks BJ, Wipfli MS, Lang DW, Lang ME. Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta, southcentral Alaska. Oecologia. 2005;144: 558–69. pmid:15891853
  12. 12. Hocking MD, Reynolds JD. Impacts of salmon on riparian plant diversity. Science. 2011;331: 1609–1612. pmid:21442794
  13. 13. Hilderbrand G V, Hanley T a., Robbins CT, Schwartz CC. Role of brown bears (Ursus arctos) in the flow of marine nitrogen into a terrestrial ecosystem. Oecologia. 1999;121: 546–550. pmid:28308364
  14. 14. Gende SM, Edwards RT, Willson MF, Wipfli MS. Pacific salmon in aquatic and terrestrial ecosystems. Bioscience. 2002;52: 917–928.
  15. 15. Buxton TH, Buffington JM, Tonina D, Fremier AK, Yager EM. Modeling the Influence of Salmon Spawning on Hyporheic Exchange of Marine-Derived Nutrients in Gravel Streambeds. Can J Fish Aquat Sci. 2015;72: 1146–1158.
  16. 16. Mathewson DD, Hocking MD, Reimchen TE. Nitrogen uptake in riparian plant communities across a sharp ecological boundary of salmon density. BMC Ecol. 2003;3: 4. pmid:12729462
  17. 17. Reimchen TE, Fox CH. Fine-scale spatiotemporal influences of salmon on growth and nitrogen signatures of Sitka spruce tree rings. BMC Ecol. BMC Ecology; 2013;13: 38. pmid:24093666
  18. 18. Hurteau LA, Mooers A, Reynolds JD, Hocking MD. Salmon nutrients are associated with the phylogenetic dispersion of riparian flowering-plant assemblages. Ecology. 2016;97: 450–460. pmid:27145619
  19. 19. Sharpe F. The biologically significant attributes of forest canopies to small birds. Northwest Sci. 1996;70: 86–93.
  20. 20. Johnson MD. Measuring Habitat Quality: a Review. 2007;109: 489–504.
  21. 21. Nur N, Ballard G, G r. Regional analysis of riparian bird species response to vegetation and local habitat features. Wilson J Ornithol. The Wilson Ornithological Society; 2008;120: 840–855.
  22. 22. Mattson WJ. Herbivory in Relation to Plant Nitrogen Content. Annu Rev Ecol Syst. Annual Reviews 4139 El Camino Way, P.O. Box 10139, Palo Alto, CA 94303–0139, USA; 1980;11: 119–161.
  23. 23. Price PW, Denno RF, Eubanks MD, Finke DL, Kaplan I. Insect Ecology: Behavior, Populations and Communities [Internet]. Cambridge University Press; 2011. Available:
  24. 24. Jackrel SL, Wootton JT. Cascading effects of induced terrestrial plant defences on aquatic and terrestrial ecosystem function. Proc R Soc B Biol Sci. 2015;282: 20142522–20142522. pmid:25788602
  25. 25. Reimchen TE, Mathewson DD, Hocking MD, Moran J, Harris D. Isotopic evidence for enrichment of salmon-derived nutrients in vegetation, soil, and insects in riparian zones in coastal British Columbia. Am Fish Soc Symp. 2002;34: 59–70. Available:
  26. 26. Verspoor JJ, Braun DC, Stubbs MM, Reynolds JD. Persistent ecological effects of a salmon-derived nutrient pulse on stream invertebrate communities. Ecosphere. 2011;2: 1–17.
  27. 27. Moore JW, Schindler DE, Scheuerell MD. Disturbance of freshwater habitats by anadromous salmon in Alaska. Oecologia. 2004;139: 298–308. pmid:14997375
  28. 28. Harding JN, Reynolds JD, Bay H. Opposing forces: Evaluating multiple ecological roles of Pacific salmon in coastal stream ecosystems. Ecosphere. 2014;5(12): 157.
  29. 29. Folkard NFG, Smith JNM. Evidence for bottom-up effects in the boreal forest: Do passerine birds respond to large-scale experimental fertilization? Can J Zool. 1995;73: 2231–2237.
  30. 30. Gende SM, Willson MF. Passerine Densities in Riparian Forests of Southeast Alaska: Potential Effects of Anadromous Spawning Salmon. Condor. 2001;103: 624.
  31. 31. Christie KS, Reimchen TE. Presence of Salmon Increases Passerine Density on Pacific Northwest Streams. Auk. 2008;125: 51–59.
  32. 32. Field RD, Reynolds JD. Sea to sky: impacts of residual salmon-derived nutrients on estuarine breeding bird communities. Proc Biol Sci. 2011;278: 3081–8. pmid:21325324
  33. 33. Field RD, Reynolds JD. Ecological links between salmon, large carnivore predation, and scavenging birds. J Avian Biol. 2013;44: 9–16.
  34. 34. Ralph CJ, Sauer JR, Droege S. Monitoring Bird Populations by Point Counts. USDA For Serv Gen Tech Rep. 1995; 1–181.
  35. 35. Willson MF, Kingery H. American dipper (Cinclus mexicanus). Birds of North America Online;
  36. 36. Nur N, Jones SL, Geupel GR. Statistical Guide to Data Analysis of Avian Monitoring Programs. 1999.
  37. 37. Jost L. Entropy and diversity. Oikos. 2006;113: 363–375.
  38. 38. MacArthur RH. Patterns of species diversity [Internet]. Biological Reviews. 1965. pp. 510–533.
  39. 39. Pielou ECC. Species-diversity and pattern-diversity in the study of ecological succession. J Theor Biol. Academic Press; 1966;10: 370–83. pmid:5964400
  40. 40. Ellison AM. Partitioning diversity 1. Ecology. Ecological Society of America; 2010;91: 1962–1963. pmid:20715615
  41. 41. English KK, Bocking RC, Irvine JR. A Robust Procedure for Estimating Salmon Escapement based on the Area-Under-the-Curve Method. Can J Fish Aquat Sci. NRC Research Press Ottawa, Canada; 1992;49: 1982–1989.
  42. 42. Macarthur RH, Macarthur JW. On Bird Species Diversity Published by: Ecological Society of America Linked references are available on JSTOR for this article: Ecology. 1961;42: 594–598.
  43. 43. Vannote RL, Minshall GW, Cummins KW, Sedell JR, Cushing CE. The River Continuum Concept. Can J Fish Aquat Sci. NRC Research Press Ottawa, Canada; 1980;37: 130–137.
  44. 44. Gende SM, Quinn TP, Willson MF. Consumption choice by bears feeding on salmon. Oecologia. 2001;127: 372–382. pmid:28547108
  45. 45. Hocking MD, Ring RA, Reimchen TE. The ecology of terrestrial invertebrates on Pacific salmon carcasses. Ecol Res. 2009;24: 1091–1100.
  46. 46. Government of British Columbia [Internet]. 2006. Available:
  47. 47. Anderson DR. Model Based Inference in the Life Sciences: A Primer on Evidence [Internet]. Springer Science & Business Media; 2007. Available:
  48. 48. Zuur AF, Ieno EN, Walker N, Saveliev AA, Smith GM. Mixed effects models and extensions in ecology with R [Internet]. New York, NY: Springer New York; 2009.
  49. 49. Burnham KP, Anderson DR. Model Selection and Multimodel Inference: a Practical Information-theoretic Approach, 2nd edn. Springer-Verlag, New York. New York Springer. 2002.
  50. 50. Grueber CE, Nakagawa S, Laws RJ, Jamieson IG. Multimodel inference in ecology and evolution: challenges and solutions. J Evol Biol. 2011;24: 699–711. pmid:21272107
  51. 51. R Development Core Team R. R: A Language and Environment for Statistical Computing [Internet]. Team RDC, editor. R Foundation for Statistical Computing. R Foundation for Statistical Computing; 2011. p. 409.
  52. 52. Mazerolle MJ. AICcmodavg: model selection and multimodel inference based on (Q) AIC (c). R Packag version. 2011;1: 1–15.
  53. 53. Barton K. MuMIn: multi-model inference, R package version 1.9.13 2015.
  54. 54. Pinheiro J, Bates D, DebRoy S, Sarkar D, R Development Core Team R. nlme: linear and nonlinear mixed effects models. R Packag version 31–122. 2015;R package: 1–3.
  55. 55. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, et al. vegan: Community Ecology Package [Internet]. R package version 2.3–1. 2015. p. 264.
  56. 56. Reimchen TE. Diverse Ecological Pathways of Salmon Nutrients Through an Intact Marine-terrestrial Interface. 2017; 350–368.
  57. 57. Richey JE, Perkins MA, Goldman CR. Effects of fokanee salmon (oncorhyncus nerka) decomposition on the ecology of a subalpine stream. J Fish Res Board Canada. 1975;32: 817–820.
  58. 58. Wipfli MS, Hudson J, Caouette J. Influence of salmon carcasses on stream productivity: response of biofilm and benthic macroinvertebrates in southeastern Alaska, U.S.A. Can J Fish Aquat Sci. 1998;55: 1503–1511.
  59. 59. Rinella DJ, Wipfli MS, Walker CM, Stricker C a, Heintz R a. Seasonal persistence of marine-derived nutrients in south-central Alaskan salmon streams. Ecosphere. 2013;4: 122.
  60. 60. Nakano S, Murakami M. Reciprocal subsidies: Dynamic interdependence. Proc Natl Acad Sci. 2001;98: 166–170. pmid:11136253
  61. 61. Iwata T, Nakano S, Murakami M. Stream meanders increase insectivorous bird abundance in riparian deciduous forests. Ecography (Cop). 2003;26: 325–337.
  62. 62. Sweeney BW, Vannote RL. Population Synchrony in Mayflies: A Predator Satiation Hypothesis. Evolution (N Y). 1982;36: 810–821.
  63. 63. Baxter C V, Fausch KD, Saunders WC. Tangled webs: Reciprocal flows of invertebrate prey link streams and riparian zones. Freshw Biol. 2005;50: 201–220.
  64. 64. Murakami M, Nakano S, Letters E. Indirect effect of aquatic insect emergence on a terrestrial insect population through by birds predation. Ecol Lett. 2002;5: 333–337.
  65. 65. Hocking MD, Reimchen TE. Consumption and distribution of salmon (Oncorhynchus spp.) nutrients and energy by terrestrial flies. Can J Fish. 2006;63: 2076–2086.
  66. 66. Feeny P, Jul N. Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology. 1970;51: 565–581.
  67. 67. Airola DA, Barrett RH. Foraging and habitat relationships of insect-gleaning birds in a Sierra Nevada mixed-conifer forest. Condor. 1985;87: 205–216.
  68. 68. Carey AB. Biocomplexity and restoration of biodiversity in temperate coniferous forest: inducing spatial heterogeneity with variable-density thinning. Forestry. 2003;76: 127–136.
  69. 69. Berenbaum M. Coumarins and caterpillars: A case for coevolution. Evolution (N Y). 1983;37: 163–179.
  70. 70. Panella N a, Dolan MC, Karchesy JJ, Xiong Y, Peralta-Cruz J, Khasawneh M, et al. Use of novel compounds for pest control: insecticidal and acaricidal activity of essential oil components from heartwood of Alaska yellow cedar. J Med Entomol. 2005;42: 352–358. pmid:15962787
  71. 71. Martin TE. Food as a Limit on Breeding Birds: A Life-History Perspective. Annu Rev Ecol Syst. 1987;18: 453–487.
  72. 72. Tonra CM, Sager-Fradkin K, Marra PP. Barriers to salmon migration impact body condition, offspring size, and life history variation in an avian consumer. Ecography (Cop). 2016;39: 1056–1065.
  73. 73. Obermeyer KE, White KS, Willson MF. Influence of salmon on the nesting ecology of American dippers in Southeastern Alaska. Northwest Sci. 2006;80: 26–33.
  74. 74. Tonra CM, Sager-Fradkin K, Morley SA, Duda JJ, Marra PP. The rapid return of marine-derived nutrients to a freshwater food web following dam removal. Biol Conserv. 2015;192: 130–134.
  75. 75. Christie K, Hocking M, Reimchen T. Tracing salmon nutrients in riparian food webs: isotopic evidence in a ground-foraging passerine. Can J Zool. 2008;86: 1317–1323.
  76. 76. Schindler DE, Smits AP. Subsidies of Aquatic Resources in Terrestrial Ecosystems. Ecosystems. Springer US; 2016; 1–16.
  77. 77. Merz JE, Moyle PB. Salmon, wildlife, and wine: marine-derived nutrients in human-dominated ecosystems. Ecol Appl. 2006;16: 999–1009. pmid:16826998
  78. 78. Fisheries and Oceans Canada. Canada’s Policy for Conservation of Wild Pacific Salmon. Fisheries (Bethesda). 2005.
  79. 79. Price K, Roburn A, MacKinnon A. Ecosystem-based management in the Great Bear Rainforest. For Ecol Manage. Elsevier B.V.; 2009;258: 495–503.
  80. 80. Xiang H, Zhang Y, Richardson JS. Importance of Riparian Zone: Effects of Resource Availability at Land-water Interface. Riparian Ecol Conserv. 2017;3: 1–17.
  81. 81. Gresh T, Lichatowich J, Schoonmaker P. An Estimation of Historic and Current Levels of Salmon Production in the Northeast Pacific Ecosystem: Evidence of a Nutrient Deficit in the Freshwater Systems of the Pacific Northwest. Fisheries. 2000;25: 15–21.