No abstract

Resource managers have traditionally had to rely on simple hydrological and habitat‐association methods to predict how changes in river flow regimes will affect the viability of instream populations and communities. Yet these systems are characterized by dynamic feedbacks among system components, a high degree of spatial and temporal variability, and connectivity between habitats, none of which can be adequately captured in the commonly employed management methods. We argue that process‐oriented ecological models, which consider dynamics across scales and levels of biological organization, are better suited to guide flow regime management. We review how ecological dynamics in streams and rivers are shaped by a combination of the flow regime and internal feedbacks, and proceed to describe ecological modeling tools that have the potential to characterize such dynamics. We conclude with a suggested research agenda to facilitate the inclusion of ecological dynamics into instream flow needs assessments.

We surveyed 53 stream reaches from the eastern slopes of the Canadian Rocky Mountains and examined the distribution of native and nonnative salmonids as related to habitat variables measured at the reach scale (100 m). The most common fishes encountered in these surveys were cutthroat trout Oncorhynchus clarki, bull trout Salvelinus confluentus, brook trout S. fontinalis, and rainbow trout O. mykiss. Of these salmonids, only cutthroat and bull trout are native to Kananaskis Country; however, cutthroat trout have also been extensively stocked throughout the region. Reach elevation, which strongly influenced mean summer stream temperatures, was the only habitat variable that was significantly related to the presence of all four salmonids. Both cutthroat and bull trout were more likely to occur in the higher elevations, whereas brook and rainbow trout were more likely to occur in the lower elevations. Because the distribution of stocked fishes is not independent of their original stocking locations, we tested the hypothesis that their distribution was simply an artifact of past stocking. Based on the historical stocking record for the surveyed region, brook and rainbow trout would be more prevalent in higher elevations if stocking location only dictated their presence. This expectation directly contradicts our observed results, suggesting there has been a preferential downstream movement of brook and rainbow trout to colonize streams at lower elevations. In contrast, the distribution of cutthroat trout predicted from the stocking record and observed from the stream surveys did not differ, suggesting their current distribution may reflect past stocking.

Regulations designed to protect recreational fisheries from overexploitation can fail. Regulations such as size and bag limits restrict harvest by individual anglers but not angler effort and therefore not total harvest. Even when individual harvest limits are set to zero (i.e., catch and release), a combination of hooking mortality and noncompliance may lead to fishing mortality rates that are not sustainable if angling effort is sufficiently high. These assertions were tested and quantified by using simulation experiments on a size‐ and age‐structured model developed for a fishery on an adfluvial bull trout population. The functions and rates describing the biology and fishery were derived from a variety of sources, including published and unpublished information on bull trout and, where such sources were unavailable, from other salmonid species. The model predicts that a 40‐cm minimum size limit for harvest would maintain viable populations at an annual effort up to 4 angler‐hours · ha−1 · year−1, a 65‐cm minimum size limit up to 10 angler‐hours · ha−1 · year−1, and a catch‐and‐release fishery up to 18 angler‐hours · ha−1 · year−1. The quality of the fisheries that developed under these three alternative regulations varied substantially with the amount of angler effort imposed. Uncertainty in the minimum population size necessary to ensure sustainability, recruits per unit stock, catchability, hooking mortality rate, and noncompliance rate modifies quantitative predictions, but the qualitative patterns are general. If anglers respond dynamically to variation in the quality of fishing, then the ability of size limit regulations to sustain fisheries is further compromised. The combination of life history and fishery traits such as slow growth, late age at maturity, low fecundity, longevity, and high catchability render adfluvial bull trout particularly susceptible to overfishing, even within relatively narrow bounds of angler effort.

Trout stocking in the mid-1960s eliminated the calanoid copepod Hesperodiaptomus arcticus and other largebodied crustaceans such as Gammarus lacustris, Daphnia middendorffiana, and Daphnia pulex from many alpine lakes in the Rocky Mountain Parks of Canada. H. arcticus frequently dominates the plankton communities of fishless lakes, preying on rotifers and nauplius larvae. Following the extirpation of H. arcticus, rotifers and small-bodied cyclopoid copepods dominate the zooplankton assemblages of alpine lakes.We studied the zooplankton community of Snowflake Lake, Banff National Park, from 1966 to 1995. H. arcticus was eliminated following stocking of the lake with trout in the 1960s. It failed to become reestablished after the disappearance of the fish population in the mid-1980s. Several species of rotifers and small-bodied crustaceans, species originally rare or absent from the plankton, became abundant following fish stocking and remained so after the fish population declined.In 1992, we reintroduced H. arcticus to Snowflake Lake. The H. arcticus population grew exponentially for 4 yr, but had not reached stable densities typical of unmanipulated alpine lakes by 1995. By 1994, however, even the small population of Hesperodiaptomus was beginning to suppress populations of rotifers, copepod nauplii, and large diatoms. Because H. arcticus is omnivorous, a simple model of cascading trophic interactions did not predict the outcome of trophic manipulations in this alpine lake.

An exploited bull trout, Salvelinus confluentus, population experienced a 28-fold increase in adult density during a 10-year period from a minimum of 60 individuals. This demonstrates the extent to which this population was overharvested. Its ability to respond in fewer than two generations to the implementation of zero-harvest regulations suggests this population was growth-overfished not recruitment-overfished. Examination of stock–recruitment relationships of various life stages indicates that recovery of this population was regulated by the density-dependent survival of juveniles in the rearing creek. This compensatory response occurred between egg deposition and age-1 and regulated the number of fish recruiting into the adult population. A second population bottleneck became apparent later in the recovery process when density-dependent survival of the adult population resulted in its approach to an asymptote, highlighting the necessity of long-term data sets for examining these compensatory responses. Results from this study demonstrate the importance of understanding the influence of individual life stages on the ability of overexploited populations such as threatened bull trout to recover and for their future management.

No abstract

Paleolimnology, bioenergetics modelling, and mesocosm experiments were used to quantify changes in phytoplankton following introduction of trout into fishless alpine lakes in the Canadian Rocky Mountains. During the 1960s, Snowflake and Pipit lakes were stocked with brook (Salvelinus fontinalis), cutthroat (Oncorhynchus clarkii) and rainbow trout (O. mykiss) either singly or in combination. Stocked trout eliminated large invertebrates (Daphnia spp., Hesperodiaptomus arcticus, Gamrnarus lacustris), but the fish died within 15 yr. High performance liquid chromatographic analysis of carotenoids and chlorophylls in sediments inferred that algal abundance increased 4- to 10-fold shortly after fish stocking. In contrast, phytoplankton composition and biomass were constant in nearby, unstocked Harrison Lake, as inferred from fossils. Pigment analysis of mesocosms showed that phytoplankton were sensitive to moderate fertilization: 11 μg P∙L−1 resulted in four- to six-fold increases in algal biomass. Bioenergetics modelling was used to estimate phosphorus (P) excretion from trout. The flux of excreted P was highly correlated (r2 = 0.76, p < 0.0001, N = 12) to changes in algal biomass, as estimated from fossil pheophytin b. Consequently, we infer that nutrient recycling by stocked trout was one of several mechanisms that contributed to increased algal biomass.