Spatially overlapping salmon species have varied population response to early life history mortality from increased peak flows

Abstract

In the Pacific Northwest, USA, climate change is expected to result in a shift in average hydrologic conditions and increase variability. The relative vulnerabilities to peak flow changes among salmonid species within the same basin have not been widely evaluated. We assessed the impacts of predicted increases in peak flows on four salmonid populations in the Chehalis River basin. Coupling observations of peak flows, emissions projections, and multi-stage Beverton–Holt matrix-type life cycle models, we ran 100-year simulations of spawner abundance under baseline, mid-century, and late-century climate change scenarios. Coho (Oncorhynchus kisutch) and spring Chinook salmon (Oncorhynchus tshawytscha) shared the highest projected increase in interannual variability (SD = ±15%). Spring Chinook salmon had the greatest reduction in median spawner abundance (–13% to –15%), followed by coho and fall Chinook salmon (–7% to –9%), then steelhead (Oncorhynchus mykiss) (–4%). Our results show that interspecies and life history variability within a single basin is important to consider. Species with diverse age structures are partially buffered from population variability, which may increase population resilience to climate change.

How does habitat restoration influence resilience of salmon populations to climate change?

“…Fall-run Chinook and steelhead were less responsive to modeled restoration actions. Projected climate-related increases in flood flow reduced modeled median spawner abundance by −4% for steelhead and −15% for spring-run Chinook by late century, with intermediate reductions for coho and fall-run Chinook (Table 1) (Nicol et al, 2022).…”

“…In previous HARP model analyses in the Chehalis River basin, we quantified habitat change from historical to current conditions (Beechie et al, 2021), modeled potential changes in equilibrium spawner abundance for seven individual restoration action types (Jorgensen et al, 2021), and modeled two individual effects of climate change (increasing flood flow and increasing stream temperature) (Fogel et al, 2022; Nicol et al, 2022). We estimated changes in habitat conditions for seven specific impacts to salmon habitat: migration barriers, increased fine sediment, reduced wood abundance, decreased stream shade, channel straightening and bank armor, beaver removal, and disconnection of floodplain habitats (Beechie et al, 2021).…”

“…Fall‐run Chinook and steelhead were less responsive to modeled restoration actions. Projected climate‐related increases in flood flow reduced modeled median spawner abundance by −4% for steelhead and −15% for spring‐run Chinook by late century, with intermediate reductions for coho and fall‐run Chinook (Table 1) (Nicol et al, 2022). Increasing temperature generally had larger effects on each species/run by late century, ranging from −12% for fall‐run Chinook salmon to −87% for spring‐run Chinook salmon (Table 1) (Fogel et al, 2022).…”

“…In this paper, we compare the potential effectiveness of these strategies for three species of Pacific salmonids with four different life histories in the Chehalis River basin of the Pacific Northwest, USA. We use the Habitat Assessment and Restoration Planning (HARP) model (Beechie et al, 2021; Jorgensen et al, 2021), which was designed to evaluate the effects of alternative habitat restoration actions and climate change on salmonid populations (Fogel et al, 2022; Jorgensen et al, 2021; Nicol et al, 2022). The HARP model quantifies degradation and potential improvement in habitat conditions, translates environmental or habitat conditions into life stage parameters for each species, and uses salmon life cycle models to evaluate change in population size as a function of modeled habitat changes (Beechie et al, 2021).…”

“…Steeper slopes indicate greater sensitivity per 1% increment of change, and greater total magnitude of change indicates greater restoration potential. Model uncertainties are briefly discussed in Appendix S2, and additional details on uncertainties are in prior HARP model publications(Beechie et al, 2021;Fogel et al, 2022;Jorgensen et al, 2021;Nicol et al, 2022).…”