Models that assess the vulnerability of freshwater species to shifting environmental conditions do not always account for short-duration extremes, which are increasingly common. Life cycle models for Pacific salmon (Oncorhynchus spp.) generally focus on average conditions that fish experience during each life stage, yet many floods, low flows, and elevated water temperatures only last days to weeks. We developed a process-based life cycle model that links coho salmon (Oncorhynchus kisutch) abundance to daily streamflow and thermal regimes to assess: (1) “How does salmon abundance respond to short-duration floods, low flows, and high temperatures in glacier-, snow-, and rain-fed streams?” and (2) “How does the temporal resolution of flow and temperature data influence these responses?”. Our simulations indicate that short-duration extremes can reduce salmon abundance in some contexts. However, after daily flow and temperature data were aggregated into weekly and monthly averages, the impact of extreme events on populations declined. Our analysis demonstrates that novel modeling frameworks that capture daily variability in flow and temperature are needed to examine impacts of extreme events on Pacific salmon.

Dissolved inorganic carbon (DIC) represents an important but poorly constrained form of lateral carbon flux to the oceans. With high precipitation rates, large glaciers, and dense temperate rainforest, Southeast Alaska plays a critical role in the transport of carbon to the Gulf of Alaska (GOA). Previous estimates of DIC flux across the Southeast Alaska Drainage Basin (SEAKDB) are poorly constrained in space and time. Our goal was to incorporate recent measurements of DIC concentrations with previous measurements from the U.S. Geological Survey in order to model the spatial and temporal patterns of riverine DIC transport from SEAK to the GOA. We aggregated DIC concentration measurements from 1957 to 2020 and associated measurements of mean daily discharge. We then constructed load estimation models to generate concentration predictions across 24 watersheds. By spatially matching measurements of DIC with SEAKDB watersheds, we extrapolated concentration predictions across 2,455 watersheds encompassing approximately 190,000 km2. Models were aggregated according to two factors, the presence of karst and the discharge regime. Finally, monthly flux predictions were generated for each watershed using predicted concentrations and runoff estimates from the Distributed Climate Water Balance Model. Mean annual DIC flux from the SEAKDB was 2.36 Tg C with an average yield of 12.52 g C m−2. Both karst presence and flow regimes modified DIC flux and speciation across coastal marine areas. The high resolution of DIC flux estimates will provide useful inputs for describing seasonal C dynamics, and further refines our understanding of C budgets in the Pacific temperate rainforest and the surrounding marine environment.

1. Pulsed subsidy events create ephemeral fluxes of hyper-abundant resources that can shape annual patterns of consumption and growth for recipient consumers. However, environmental conditions strongly affect local resource availability for much of the year, and can heavily impact consumer foraging and growth patterns prior to pulsed subsidy events. Thus, a consumer's capacity to exploit pulse subsidy resources may be influenced by antecedent environmental conditions, but this has rarely been shown in nature and is unknown in aquatic ecosystems. 2. Here, we sought to understand the importance of hydrologic variation and a salmon pulse subsidy on the foraging and growth patterns of two stream salmonids in a coastal southeast Alaska drainage. 3. To do this, we sampled fish stomach contents at a high temporal frequency (daily–weekly measurements) and analyzed fish consumption rates in relation to streamflow and pulse subsidy resource availability. We then explored the influence of interannual hydrologic variation on access to pulse subsidy resources (i.e. whether fish exceeded an egg consumption gape limit) in a bioenergetic simulation. 4. Prior to Pink Salmon spawning, Dolly Varden and Coho Salmon displayed distinct and nonlinear flow-foraging relationships, where forage for both species consisted primarily of macroinvertebrates. During this time period, consumption maxima coincided with baseflow and the highest observed flow conditions, and consumption minima were observed at severe low-water and intermediate flow values. After salmon spawning began, forage was not significantly related to flow and consisted primarily of salmon eggs. Further, consumption rates increased overall, and foraging patterns did not appear to be affected by flow in either species. Bioenergetic simulations revealed that patterns of interannual hydrologic variation may shift Coho Salmon growth trajectories among years. 5. Together, our results suggest that access to marine pulse subsidy resources may depend on whether antecedent hydrologic conditions are suitable for juvenile salmonids to grow large enough to consume salmon eggs by the onset of spawning.

Dissolved organic matter (DOM) composition in small watersheds depends on complex antecedent conditions that ultimately influence DOM generation, processing, and stability downstream. Here, we used ultrahigh resolution Fourier-transform ion cyclotron resonance mass spectrometry and total dissolved nitrogen and dissolved organic carbon concentrations to investigate how DOM is produced in distinct sub-catchment types (poor fen, forested wetland, and upland forest) and transported through a watershed in the northeast Pacific coastal temperate rainforest (NPCTR). We traced a suite of previously identified source-specific marker formulae from vegetation and soil downstream and used them to test models of terrestrial DOM inputs. Marker formulae escaped microbial degradation and were exported from the watershed, demonstrating strong land-to-ocean connectivity through the transfer of unmodified tree DOM from specific tree species into the marine environment. Simple hydrologic and temperature variables were better able to predict inputs of soil-sourced DOM into the stream network than tree-sourced DOM, highlighting the role of antecedent conditions (e.g., plant growth stage) in DOM source availability and hydrologic flow connectivity, particularly for plant-derived material. Forested wetland pore waters featured thousands of nitrogen-containing molecular formulae that potentially provide a path of direct organic nitrogen uptake to organisms. The modified aromaticity index peaked in midsummer (up to 0.55 for fen headwaters) suggesting DOM inputs from freshly produced vegetation provide a strong summertime terrestrial signal. As the climate changes, new watershed-scale conditions may further complicate predictions of DOM source availability, flow connectivity, and downstream fate in NPCTR watersheds.

To investigate how source and processing control the composition of “terrestrial” dissolved organic matter (DOM), we combine soil and tree leachates, tree DOM, laboratory bioincubations, and ultrahigh resolution Fourier-transform ion cyclotron resonance mass spectrometry in three common landscape types (upland forest, forested wetland, and poor fen) of Southeast Alaska’s temperate rainforest. Tree (Tsuga heterophylla and Picea sitchensis) needles and bark and soil layers from each site were leached, and tree stemflow and throughfall collected to examine DOM sources. Dissolved organic carbon concentrations were as high as 167 mg CL−1 for tree DOM, suggesting tree DOM fluxes may be substantial given the hypermaritime climate of the region. Condensed aromatics contributed as much as 38% relative abundance of spruce and hemlock bark leachates suggesting coniferous trees are potential sources of condensed aromatics to surface waters. Soil leachates showed soil wetness dictates DOM composition and processing, with wetland soils producing more aromatic formulae and allowing the preservation of traditionally biolabile, aliphatic formulae. Biodegradation impacted soil and tree DOM differently, and though the majority of source-specific marker formulae were consumed for all sources, some marker formulae persisted. Tree DOM was highly biolabile (> 50%) and showed compositional convergence where processing homogenized DOM from different tree sources. In contrast, wetland and upland soil leachate DOM composition diverged and processing diversified DOM from different soil sources during bioincubations. Increasing precipitation intensity predicted with climate change in Southeast Alaska will increase tree leaching and soil DOM flushing, tightening linkages between terrestrial sources and DOM export to the coastal ocean.

Dissolved organic matter (DOM) transported from terrestrial to aquatic ecosystems is an important source of C, N and energy for the metabolism of aquatic heterotrophic bacteria. I examined the concentration and chemical quality of DOM exported from coastal temperate watersheds in southeast Alaska to determine if wetland soils are an important source of biodegradable dissolved organic carbon (BDOC) to aquatic ecosystems. I addressed this question through a combination of high resolution temporal and spatial field measurements in three watersheds near Juneau, Alaska by using a replicated experimental design that characterized DOM export from three different soil types (bog, forested wetland and upland forest) within each of the watersheds. PARAFAC modeling of fluorescence excitation-emission spectroscopy and BDOC incubations were used to evaluate the chemical quality and lability of DOM. Overall, my findings show that wetland soils contribute substantial biodegradable DOM to streams and the response in BDOC delivery to streams changes seasonally, with soil type, and during episodic events such as stormflows. In particular, the chemical quality of DOM in streamwater and soil solution was similar during the spring runoff and fall wet season, as demonstrated by the similar contribution of protein-like fluorescence in soil solution and in streams. These findings indicate a tight coupling between wetland DOM source pools and streams is responsible for the export of BDOC from terrestrial ecosystems. Thus, seasonal changes in soil-stream linkages can have a major influence on watershed biogeochemistry with important implications for stream metabolism and the delivery of labile DOM to coastal ecosystems. Soil DOM additions in small streams draining the three soil types showed that DOM leached from watershed soils is readily used as a substrate by stream heterotrophs and at the same time modified in composition by the selective degradation of the proteinaceous fraction of DOM. These findings indicate terrestrial DOM inputs to streams are an important source of C to support stream heterotrophic production. Thus, the production of protein-rich, labile DOM and subsequent loss in stream runoff has the potential to be an important loss of C and N from coastal temperate watersheds.