- Humans have drastically altered the physical habitat and food web structure of stream ecosystems. Two major impacts humans have had on Pacific Northwest streams are modification of streamside forests (as a result of agriculture, land development, and timber harvest), and declines in the return of wild anadromous salmon to headwater ecosystems (due to a range of habitat degradation, dams, harvest, and hatcheries). Riparian forest impacts have altered stream light dynamics, while the loss of salmon has led to declines in the delivery of nutrients from the ocean to streams. While the initial impacts of the modifications took place decades or even centuries ago, they can have lasting effects on stream ecosystems and food webs. This dissertation evaluates 1) influences of long-term recovery from historic riparian harvesting on stream light, habitat, and food webs, and 2) how reduced salmon subsidies to streams may be impacting stream productivity and food webs.
Today most streams in the Pacific Northwest, and indeed across much of North America, have buffers of riparian forests that are regenerating from earlier land clearing. As stands recover, the trajectories of stand development will affect forest structure, which in turn affects stream light regimes. In the first half of my dissertation, I explore how stand age and structure relates to stream light availability and then how spatial differences and temporal changes in stream light influence stream food webs and higher trophic level biomass in headwater streams. In Chapter 2, I explore how stream light availability differs with the age and stage of riparian forests. I found that stream light flux was generally lower and less variable when bordered by second-growth forests compared to old-growth forests within a stream network and more broadly across forests west of the Cascade Mountains. Numerous studies have evaluated how large differences in light availability (e.g. fully forested compared to complete removal of riparian forests) influence stream food webs, but smaller differences in light availability, such as those found in Chapter 2, have received less consideration. In Chapter 3, I conducted surveys across 18 stream reaches and evaluated how variables associated with stream habitat, light, primary production, and macroinvertebrate biomass account for variability in the biomass of cutthroat trout and total vertebrates (fish and salamanders). Habitat metrics were not well correlated with higher trophic level biomass. In contrast, factors associated with resource availability — as regulated through bottom-up, autotrophic pathways — were closely related to the biomass of fish and other consumers. In Chapter 4, I quantified long-term responses of stream biota to the regeneration of riparian forests following clear-cut harvest. I resampled five stream reach pairs that were originally sampled in 1976 shortly after canopy removal. This initial survey showed that periphyton chlorophyll a, predatory invertebrate biomass, and cutthroat trout (Oncorhynchus clarkii clarkii) biomass were elevated in harvested reaches relative to reference reaches. After four decades of riparian regeneration, mean canopy openness, chlorophyll a, predatory invertebrate biomass, and cutthroat trout biomass declined in harvested reaches relative to paired old-growth reference reaches. Changes in canopy cover were consistent with biotic responses and suggest that changes in light availability as stands regenerated exerted control on biota through bottom-up pathways in these streams.
While spatial and temporal light dynamics appear as important regulators of stream food webs in small forested streams of western Oregon, other factors may emerge as important constraints on food web productivity across stream networks in other regions. In the second half of my dissertation, I explore bottom-up drivers of fish production in a river network in eastern Oregon where canopies are more open than small western Oregon streams. I focus on nutrient and carbon subsides in this study as the loss of returning anadromous fish has been hypothesized as a key factor contributing to poor recovery of ESA-listed salmonids. In chapter 5, I evaluate network-scale spatial patterns of primary production, potential drivers of primary production, and juvenile salmonid abundance throughout two NE Oregon sub-basins. Primary production rates increased with watershed area and we were able to explain 72% of the variation in primary production across these basins using a combination of fixed-effects (e.g. light, nutrients, and temperature) and spatial autocorrelation. In contrast to other studies, juvenile salmonid abundance was greatest in cool headwaters where nutrient concentrations and rates of primary production were very low. To test the hypothesis that growth of juvenile salmonids and other biota in these low-productivity stream sections may be inhibited by the reduction of returning adult salmon and the associated loss of nutrient subsidies, I conducted a carcasses addition experiment in three locations of the Upper Grand Ronde River. In chapter 6, I focused on the responses of juvenile Chinook (Oncorhynchus tshawytsca) and steelhead (O. mykiss). Chinook and steelhead consumed an abundance of eggs and carcass tissue which resulted in greater growth rates and body condition of fish in treatment reaches relative to controls. To contextualize potential effects of increased growth on Chinook survival, I used an 18 year tagging and detection dataset to evaluate Chinook length-survival relationships. The positive association between length and survival suggests that actions resulting in larger Chinook lead to increased survival rates. In chapter 7, I evaluate carcass addition effects on the broader food web. Periphyton, aquatic invertebrates, and non-salmonid fish assimilated carcass nitrogen, but enrichment was far less than observed in juvenile salmonids. In contrast to salmonids, diet analysis and stable isotope patterns indicated that non-salmonids were not consuming eggs and carcass material, suggesting carcass nitrogen assimilation occurred through bottom-up pathways. These results suggest that salmon subsidies have the potential to broadly impact stream food webs in this region, but that species able to directly consume eggs and carcass material (i.e. juvenile salmonids) clearly benefit more from these subsidies.