Mercury (Hg) is a potent toxin that biomagnifies in aquatic food webs. Large fish generally have higher Hg concentrations than small fish of the same species. However, models predict that fish that grow large faster will have lower Hg concentrations than small, slow‐growing fish owing to somatic growth dilution (SGD). We examined the relationship between Hg concentration and growth rate in fish by means of a large‐scale field experiment. Atlantic salmon Salmo salar fry hatched under uniform initial conditions were released at 18 sites in natural streams, collected after one growing season, and measured with respect to Hg concentration and growth. As expected for Hg accumulation from food, the Hg concentrations in the fish tracked those in their prey. Nonetheless, large, fast‐growing fish had lower Hg concentrations than small, slow‐growing ones, consistent with SGD. While the Hg concentration of the prey accounted for 59% of the explained variation in the Hg concentration in the Atlantic salmon across sites, the salmon growth rate accounted for 38%. A mass‐balance Hg accumulation model shows that such SGD occurs when fast growth is associated with high growth efficiency. Fish growth is tremendously variable and sensitive to anthropogenic impacts, so the SGD of Hg has important implications for fisheries management.

1 Empirical studies show that average growth of stream‐dwelling salmon and trout often declines with increasing density in a characteristic concave relationship. However, the mechanisms that generate negative density–growth relationships in populations in natural streams are not certain. 2 In a recent study, Imre, Grant & Cunjak (2005; Journal of Animal Ecology, 74, 508–516) argue that density‐dependent growth due to exploitative competition for prey causes the negative density–growth relationships for stream salmonids. They argue that the concave shape of empirical density–growth relationships is consistent with a simple model of exploitative competition and not consistent with interference competition for space. 3 We use a simple model to show that competition for space can yield concave density–growth relationships consistent with the empirical pattern when individuals compete for foraging sites that vary spatially in quality and lower‐quality sites predominate. Thus, the predictions of the exploitative competition and spatial competition models overlap. 4 The shape of the density–growth relationship does not differentiate between candidate mechanisms underlying density‐dependent growth for stream salmonids. Our results highlight the general problem with determining the mechanism driving an ecological process from patterns in observational data within the context of linking population demographics to habitat structure and animal behaviour.

Individual growth rates of fish often depend on both population density and resource availability, both of which are subject to anthropogenic impacts and can be manipulated in the interest of restoring or enhancing fish populations. However, direct tests of the relative contribution of these factors to growth variation are lacking for most populations and are critically needed to guide management. We used a large‐scale field experiment to evaluate the relative effects of increased population density and suppressed prey biomass on the growth of juvenile Atlantic salmon Salmo salar. We directly manipulated salmon population density by releasing newly hatched salmon fry from uniform initial conditions at three density treatments in natural streams (18 sites total, repeated over 2 years). We arrayed the density treatments across sites that encompassed a more than 10‐fold range in the biomass of benthic invertebrate prey for salmon, a low prey biomass being associated with heavy shading from riparian forest canopy. This controlled approach clearly demonstrated the enormous growth plasticity of juvenile Atlantic salmon; their mean body mass ranging from 1.2 to 14 g across sites after one growing season. Variation in prey biomass had the strongest effects on fish growth. The fish grew faster at sites with high prey biomass regardless of population density, and prey biomass alone accounted for 71% of the explained variation in their growth. Only after taking the variation explained by prey biomass into account was it apparent that the study fish grew faster at sites with low stocking density and at sites where high losses reduced Atlantic salmon population density. To maintain salmon growth and size‐dependent life histories, restoration and management efforts clearly need to consider anthropogenic factors that suppress prey abundance.