Detecting mechanisms that structure ecological communities often requires investigations at appropriate spatial and temporal scales. A 15—mo, whole—lake experiment revealed direct and indirect mechanisms by which a fish predator structure its prey population, primarily by affecting recruitment of young—of—the —year (YOY) fish. Piscivorous Eurasian perch (Perch fluviatilis) were added to two of four sections of a divided lake in eastern Finland previously stocked with crucian carp (Carassius carassius). Although the stocked crucians quickly grew beyond vulnerable sizes, recruitment of their offspring through their 1st yr of life was reduced by 90% in sections with perch. Surviving YOY in predator sections were completely confined to vegetated inshore areas. This restriction to refuges lowered their growth rates in comparison with YOY in predator—free sections, which can further reduce recruitment by prolonging susceptibility to size—limited predation and decreasing energy reserves needed for anaerobic overwintering. In the presence of piscivores, surviving prey subsequently benefit from reduced intraspecific competition, increasing growth rates to achieve large, invulnerable body sizes rapidly. Resulting populations of crucian carp, consisting of relatively few but relatively large individuals, contrast with populations living in the absence of piscivores, which are characterized by high densities of small, stunted individuals. Such divergent population patterns are displayed by a number of prey species in Scandinavia and North America, suggesting that a combination of direct and indirect effects of predation on younger age classes, such as documented for crucian carp, may be common in fish populations of small northern lakes. Detecting these patterns and identifying their underlying mechanisms may often require a combination of extensive and intensive studies.

1. Community concordance measures the degree to which patterns in community structure in a set of sites are similar between two different taxonomic groups. Although seldom incorporated into studies of lake ecosystems, aquatic birds can be influenced by the same environmental features of lakes which affect fish and invertebrates, and can interact with these organisms directly as predators, competitors or prey. We surveyed lakes in north‐central Alberta, Canada, to determine if co‐occurring fish and aquatic bird assemblages displayed concordance, and assessed the relative importance of environmental and biotic factors in contributing to observed concordance. 2. In 41 lakes (3–305 ha), we encountered seven species of fish and thirty‐one avian taxa which subsequently were used in multivariate analyses. Fish assemblages dominated by large piscivores were in large deep lakes, whereas fishless lakes and lakes with only small‐bodied fish were small and shallow, and thus, prone to winter hypoxia. Bird assemblages displayed three general patterns: (a) small shallow lakes supported a ‘core’ of widespread species (between three and eight species per lake); (b) large, deep lakes supported more species (between 11 and 16), including large, aerially foraging piscivores; and (c) large, shallow lakes supported the most species (between 15 and 23), including many ducks. 3. Randomization tests of matrix concordance and Mantel tests both showed that fish and bird assemblages were significantly concordant. Concordance reflected the fact that both groups were strongly affected by the same key environmental factors, principally lake size and maximum depth, and to a lesser extent, productivity and geographic isolation. Direct interactions between birds and fish, such as predation and competition, appeared to play much smaller roles in shaping the two assemblages.

Stable isotope analysis has emerged as an important tool in aquatic ecology. For fish, dorsal muscle from sacrificed individuals has traditionally been used in stable isotope studies; however, there are many instances when lethal sampling is undesirable. We evaluated the feasibility of using adipose and caudal fin clips as alternatives to muscle in stable isotope studies for five species of salmonids. Because fish size and water temperature can affect stable isotope ratios, we also determined whether fish length and sampling date affected the difference in isotope signatures between fins and muscle. Biopsied muscle plugs and fin clips were collected from rainbow trout Oncorhynchus mykiss, brook trout Salvelinus fontinalis, and lake trout S. namaycush as well as lake whitefish Coregonus clupeaformis and pygmy whitefish Prosopium coulterii and analyzed for stable isotopes of carbon and nitrogen. The isotope signatures of both adipose and caudal fins were significantly correlated (0.33 < R2 < 0.97) with those of dorsal muscle from the five salmonid species studied, and in some cases the signatures of fins did not differ from those of muscle. Fish length and sampling date occasionally had a small effect (0.042 < R2 < 0.49) on the relationship between the isotope signatures of fin and muscle. Although muscle biopsy provides a viable, nonlethal method of collecting muscle tissue from suitably sized fish, the strong relationships between the isotope signatures of fin and muscle demonstrate that fin clips should be considered good surrogates for muscle in stable isotope studies of salmonids.

We compared population densities and size structures of central mudminnows (Umbra limi) occurring with and without yellow perch (Perca flavescens) in six small lakes in northern Wisconsin over a 6-yr period. Our observations suggested that a size-dependent predator–prey interaction was a major mechanism organizing these assemblages. Densities of mudminnows cooccurring with perch averaged 12% of populations in mudminnow-only lakes. Large mudminnows (≥76 mm total length) predominated in populations with perch whereas smaller fish predominated in single-species assemblages. Following severe winterkills of adult perch in two lakes, mudminnow densities increased significantly due to high survival and recruitment of yearlings. Increases in small mudminnows lasted only a year: densities and size distributions returned to prewinterkill patterns when perch that had survived winterkill attained lengths >13 cm. To examine the role of size-limited predation, we performed an experiment in 2000-L tanks, varying the initial size distributions of mudminnows and perch and measuring size-specific survival among mudminnows. Predation in the experimental populations was greatest when large perch (121–180 mm) were with populations of predominantly small mudminnows (<76 mm); smaller perch (114–132 mm) were significantly less successful as predators whereas large mudminnows were significantly less vulnerable to predation by even large perch. Despite this interaction, central mudminnows coexist at low densities with yellow perch; possible factors permitting coexistence are fish sizes and spatial and structural refuges.

No abstract