No abstract

In recognition of the influence of flow on riverine habitats and organisms, stream ecologists have devoted considerable effort to the development of quantitative predictive relationships describing ecological responses to flow variability, i.e. flow‐ecology relationships. Methods used to generate flow‐ecology relationships can be thought of as a continuum bookended by pure states approaches on one end and by rates approaches on the other. In pure states approaches, the ecological response is a snapshot of a condition or property (i.e. a state) derived from a single measurement in time. In contrast, ecological responses in rates approaches reflect temporal change (i.e. a rate) and are thus reliant on repeated measurements made over time. Here, we elaborate on the fundamental characteristics of different approaches (pure states, rates and an intermediate approach we call repeated states) for generating flow‐ecology relationships, examine how commonly the different approaches are used in the flow‐ecology literature, conduct an independent analysis to illustrate the different outcomes achieved by applying repeated‐states and rates approaches using a dataset for stream fish diversity in relation to flow magnitude, and identify some of the different ways ecologists are applying rates approaches in flow ecology. Our literature review revealed that repeated‐states approaches (53% of reviewed studies) were used far more commonly than either pure states (19%) or rates (12%) approaches to generate flow‐ecology relationships. The remaining hybrid studies (17%) used both state and rate responses, and thus also relied on repeated measurements over time. Despite frequent collection of data suitable for rates approaches, flow‐ecology relationships have generally been developed using states approaches that relate changes in ecological states to different long‐term average flow conditions, rather than to specific flow sequences over much shorter time intervals. Such flow‐ecology relationships cannot generate temporally specific predictions of ecological responses to changing flow conditions (i.e. the expected change in state following a specific flow sequence), nor can they describe demographic processes underlying observed changes. While there are different scenarios in which a pure or repeated‐states approach would be useful, more frequent use of rates approaches would increase our ability to test flow‐ecology hypotheses and our mechanistic understanding of flow‐ecology relationships.

Summary Consumer excretion can be an important component of nutrient cycling in aquatic ecosystems. Uncertainty concerning the functional role of many freshwater organisms remains, including those with migratory life‐history strategies that may introduce nutrients to recipient systems. We quantified the magnitude, variability and importance of excretion by migratory fish in the context of stream nutrient cycling. In 2011–12, we collected data from tributaries of a central Utah reservoir used by two potamodromous fishes (Bonneville cut‐throat trout – BCT, Oncorhynchus clarkii utah; kokanee salmon – KOK, Oncorhynchus nerka) with temporally separated spawning migrations. To quantify the contribution of the two migratory freshwater fishes to nutrient cycling, we extrapolated measurements of per capita nitrogen (N) and phosphorus (P) excretion rates to the population level within the local environmental context of two tributaries. We observed differences in excretion subsidies between species and tributaries. BCT excretion rates and ratios were significantly greater than those for KOK. Estimates of the ratio of population‐level migrant excretion to tributary nutrient export were highly variable through time and between tributaries. These estimates were influenced by spatiotemporal hydrologic variation and positively related to ratios of migrant biomass to discharge. During migrations, daily migrant excretion loading comprised a maximum of 6–859% and 1–388% of tributary NH4‐N and soluble reactive phosphorus (SRP) export, respectively. Measurements of nutrient uptake suggested that migrant excretion could meet a substantial portion of ecosystem nutrient demand. Migrant excretion fluxes comprised 46–188% of ecosystem NH4‐N demand and varied between streams and species. In contrast, the proportion of SRP demand supplied by migrant excretion (34–37%) was relatively invariant. These results demonstrate an important role for potamodromous fishes as nutrient sources in recipient ecosystems. Furthermore, our data support predictions about when and where effects of fish‐derived nutrients will be strongest, thereby advancing the understanding of context‐dependent migratory fish effects in riverine ecosystems. Although widespread and common, potamodromous fishes are overlooked but important organisms capable of substantially affecting stream nutrient cycling.

The ability for migratory fishes to move commonly limiting resources such as nitrogen (N) and phosphorus (P) between discrete environments can have pronounced effects on recipient ecosystems. To further understand the geographic and taxonomic scope of migratory fish resource subsidies, we quantified N and P subsidies delivered by adfluvial suckers (Smallmouth Buffalo, Ictiobus bubalus) via excretion, eggs and carcasses to a small oligotrophic stream during their annual spawning migration. We also compared nutrient inputs from migrant buffalo with watershed nutrient export to assess the likelihood that delivered nutrients were ecologically important. We estimated that approximately 67,000 buffalo delivered 730 kg of N and 80 kg of P to Citico Creek as a result of excretion and egg subsidies across three migration waves. We estimated that carcasses delivered negligible amounts of N and P due to extremely low retention. The ratio of migrant inputs (M w ) to system export (E w ; M w /E w ) varied amongst three migration waves and compounds (i.e. dissolved inorganic nitrogen, ammonium and soluble reactive phosphorus), with values for M w /E w ranging from 0.25 to 5.10, reflecting the potential of nutrient subsidies to exceed nutrients exported from the system under certain conditions. Our findings suggest that suckers have the potential to deliver large resource subsidies to their spawning habitats and that these subsidies may be ecologically important, thus warranting additional consideration of the functional relevance of nongame fishes and their migrations.

Migratory fishes can affect tributary ecosystem properties given their potential to introduce nutrients (fertilize) and physically modify habitat (engineer) during spawning. Nonetheless, migrant effects are frequently context-dependent, and it is useful to understand their strength relative to other potential ecosystem drivers. We examined whether tributary ecosystem properties varied in response to migrations of two adfluvial salmonids, taking advantage of differences in migration timing and reproductive strategy between species, as well as hydrogeomorphic differences between a pair of tributaries. For analyses, we used a model comparison approach to evaluate migrant effects relative to other possible drivers. We observed that Bonneville cutthroat trout (Oncorhynchus clarkii utah) engineered benthic chlorophyll a in redds, with reduction (51% ± 16% decrease) generally occurring during migrations. Contrary to expectations, migrant fertilization effects were not pronounced even in the more retentive tributary during migration by species (kokanee, Oncorhynchus nerka) that exhibited high postspawning mortality. Based on multimodel comparisons, isolated migrant effects were not the primary influence on measured ecosystem properties. Our findings underscore the need to consider different biotic and abiotic conditions that can mediate migratory fish effects.