No abstract

Trophic magnification factors (TMFs) provide a method of assessing chemical biomagnification in food webs and are increasingly being used by policy makers to screen emerging chemicals. Recent reviews have encouraged the use of bioaccumulation models as screening tools for assessing TMFs for emerging chemicals of concern. The present study used a food web bioaccumulation model to estimate TMFs for polychlorinated biphenyls (PCBs) in a riverine system. The uncertainty associated with model predicted TMFs was evaluated against realistic ranges for model inputs (water and sediment PCB contamination) and variation in environmental, physiological, and ecological parameters included within the model. Finally, the model was used to explore interactions between spatial heterogeneity in water and sediment contaminant concentrations and theoretical movement profiles of different fish species included in the model. The model predictions of magnitude of TMFs conformed to empirical studies. There were differences in the relationship between the TMF and the octanol-water partitioning coefficient (KOW ) depending on the modeling approach used; a parabolic relationship was predicted under deterministic scenarios, whereas a linear TMF-KOW relationship was predicted when the model was run stochastically. Incorporating spatial movements by fish had a major influence on the magnitude and variation of TMFs. Under conditions where organisms are collected exclusively from clean locations in highly heterogeneous systems, the results showed bias toward higher TMF estimates, for example the TMF for PCB 153 increased from 2.7 to 5.6 when fish movement was included. Small underestimations of TMFs were found where organisms were exclusively sampled in contaminated regions, although the model was found to be more robust to this sampling condition than the former for this system.

Understanding variability of contaminant bioaccumulation within and among fish populations is critical for distinguishing between the chemical and biological mechanisms that contribute to food web biomagnification and quantifying contaminant exposure risks in aquatic ecosystems. The present study examined the relative contributions of chemical hydrophobicity (octanol-water partition coefficient [KOW ]) and habitat use as factors regulating variability in polychlorinated biphenyl (PCB) congener bioaccumulation in 3 lower trophic level cyprinid species across spatial and temporal scales. Bluntnose minnows (Pimephales notatus), spottail shiners (Notropis hudsonius), and emerald shiners (Notropis atherinoides) were sampled at 3 locations in the Detroit River, Ontario, Canada. Variability in PCB concentration was evaluated with respect to several factors, including chemical hydrophobicity, site, season, species, and weight using sum of squares and Levene's test of homogeneity of variance. Individual variability in bioaccumulated congener-specific residues depended on chemical hydrophobicity with mid- and high-range KOW congeners (log KOW >6.0), demonstrating the highest amount of variance compared with low KOW congeners. Different feeding strategies also contributed to the variance observed for mid-range KOW congeners among species. In the present study, benthic feeding specialists exhibited lower variance in PCB concentrations compared with the 2 generalist species. The results indicate that chemical hydrophobicity and feeding ecology not only contribute to differences in the biomagnification potentials of fish, but also regulate between-individual variation in PCB concentrations both across and within fish species.

The persistence of whole communities hinges on the presence of select interactions which act to stabilize communities making the identification of these keystone interactions critical. One potential candidate is omnivory, yet theoretical research on omnivory thus far has been dominated by a modular theory approach whereby an omnivore and consumer compete for a shared resource. Empirical research, however, has highlighted the presence of a broader suite of omnivory modules. Here, we integrate empirical data analysis and mathematical models to explore the influence of both omnivory module (including classic, multi‐resource, higher level, mutual predation and cannibalism) and omnivore–resource interaction type on food web stability. We use six classic empirical food webs to examine the prevalence of the different types of omnivory, a multi‐species consumer–resource model to determine the stability of these different kinds of omnivory within a module context, and finally extend these models to a 50 species, whole food web model to examine the influence of omnivory on whole food web persistence. Our results challenge the concept that omnivory is broadly stabilizing. In particular, we demonstrate that the impact of omnivory depends on the type of omnivory being examined with multi‐resource omnivory having the largest correlation with whole food web persistence. Moreover, our results highlight that we need to exercise caution when scaling modular theory to whole food web theory. Cannibalism, for example, was the most persistent and stable omnivory module in the modular theory analysis, but only demonstrated a weak correlation with whole food web persistence. Lastly, our results demonstrate that the frequency of omnivory modules are more important for whole food web persistence than the frequency of omnivore–resource interactions. Together, these results demonstrate that the role of omnivory often depends both on the type of omnivory being examined and the food web within which it is nested. In whole food web models, omnivory acts less as a keystone interaction, rather, specific types of omnivory, particularly multi‐resource omnivory, act as keystone modules. Future work integrating module and whole food web theory is critical for resolving the role of key interactions in food webs.

Understanding the response of organisms to multiple stressors is critical for predicting if populations can adapt to rapid environmental change. Natural and anthropogenic stressors often interact, complicating general predictions. In this study, we examined the interactive and cumulative effects of two common environmental stressors, lowered calcium concentration, an anthropogenic stressor, and predator presence, a natural stressor, on the water flea Daphnia pulex. We analyzed expression changes of five genes involved in calcium homeostasis - cuticle proteins (Cutie, Icp2), calbindin (Calb), and calcium pump and channel (Serca and Ip3R) - using real-time quantitative PCR (RT-qPCR) in a full factorial experiment. We observed strong synergistic interactions between low calcium concentration and predator presence. While the Ip3R gene was not affected by the stressors, the other four genes were affected in their transcriptional levels by the combination of the stressors. Transcriptional patterns of genes that code for cuticle proteins (Cutie and Icp2) and a sarcoplasmic calcium pump (Serca) only responded to the combination of stressors, changing their relative expression levels in a synergistic response, while a calcium-binding protein (Calb) responded to low calcium stress and the combination of both stressors. The expression pattern of these genes (Cutie, Icp2, and Serca) were nonlinear, yet they were dose dependent across the calcium gradient. Multiple stressors can have complex, often unexpected effects on ecosystems. This study demonstrates that the dominant interaction for the set of tested genes appears to be synergism. We argue that gene expression patterns can be used to understand and predict the type of interaction expected when organisms are exposed simultaneously to natural and anthropogenic stressors.

Accurate predictions on the bioaccumulation of persistent organic pollutants (POPs) are critical for hazard and ecosystem health assessments. Aquatic systems are influenced by multiple stressors including climate change and species invasions and it is important to be able to predict variability in POP concentrations in changing environments. Current steady state bioaccumulation models simplify POP bioaccumulation dynamics, assuming that pollutant uptake and elimination processes become balanced over an organism's lifespan. These models do not consider the complexity of dynamic variables such as temperature and growth rates which are known to have the potential to regulate bioaccumulation in aquatic organisms. We contrast a steady state (SS) bioaccumulation model with a dynamic nonsteady state (NSS) model and a no elimination (NE) model. We demonstrate that both the NSS and the NE models are superior at predicting both average concentrations as well as variation in POPs among individuals. This comparison demonstrates that temporal drivers, such as environmental fluctuations in temperature, growth dynamics, and modified food-web structure strongly determine contaminant concentrations and variability in a changing environment. These results support the recommendation of the future development of more dynamic, nonsteady state bioaccumulation models to predict hazard and risk assessments in the Anthropocene.

Understanding factors influencing mercury (Hg) bioaccumulation in fish is important for examining both ecosystem and human health. However, little is known about how differing ecosystem and biological characteristics can drive Hg bioaccumulation in top predators. The present study compared and contrasted Hg bioaccumulation in multiple age classes of lake trout (Salvelinus namaycush) collected from each of Lake Huron's Georgian Bay, North Channel, and Main Basin regions. Mercury concentrations exhibited a basin specific pattern with Main Basin fish having the highest average concentration (0.19 ± 0.01 mg/kg), followed by Georgian Bay (0.15 ± 0.02 mg/kg), and North Channel (0.07 ± <0.01 mg/kg) fish. Age-related increases in Hg concentrations were observed across the 3 basins with North Channel fish exhibiting the slowest rate of Hg bioaccumulation. No significant difference was determined between the relationships describing Hg concentration and age between Main Basin and Georgian Bay fish (p < 0.05). Mercury biomagnification factors (BMF) determined between lake trout and rainbow smelt, lake trout's primary prey, were significantly correlated with fish age and differed across the 3 basins (p < 0.05). Specifically, Georgian Bay fish exhibited the greatest age related increase in Hg BMF followed by Main Basin and North Channel fish, and these differences could not be attributed to trophic level (δ(15)N) effects or lake trout growth rates. A highly significant negative relationship was determined between Hg BMFs and basin specific prey fish densities indicating that ecological factors associated with food acquisition and foraging efficiencies play an important role in Hg bioaccumulation in feral fish communities.

Local dynamics are influenced by regional processes. Meta‐ecology, or the study of spatial flows of energy, materials, and species between local systems, is becoming increasingly concerned with accurate depictions of species movements and the impacts of this movement on landscape‐level ecosystem function. Indeed, incorporating diverse types of movement is a major frontier in metacommunity theory. Here, we synthesize literature to demonstrate that the movement of organisms between patches is governed by the interplay between both a species’ ability to move and the combined effects of landscape structure and physical flows (termed abiotic controls), which together we refer to as abiotic‐dependent species connectivity. For example, two lakes that share geographic proximity may be inaccessible for mobile fish species because they lack a river connecting them (landscape structure), but wind currents may disperse insects between them (physical flows). Empirical evidence suggests that abiotic controls, such as ocean currents, lead to abiotic‐dependent species connectivity and that, in nature, this type of connectivity is the rule rather than the exception. Based on this empirical evidence, we introduce a novel mathematical framework to demonstrate how species movement capabilities and abiotic conditions, can interact to influence metacommunity stability. We apply this framework to predict how incorporating abiotic‐dependent species connectivity applies to classic empirical examples of aquatic, aquatic‐terrestrial, and terrestrial experimental metacommunities. We demonstrate that incorporating abiotic‐dependent species connectivity into metacommunity models can lead to a much broader range of dynamics than models previously predicted, including a wider range of metacommunity stability. Our framework fills critical gaps in our basic understanding of organismal movement across landscapes and provides testable predictions for how such common natural phenomena impact landscape‐level ecosystem function. Finally, we present future perspectives for further development of meta‐ecological theory from questions about fragmentation to ecosystems. Anthropogenic change is not only leading to habitat loss from the damming of rivers to denuding the landscape, but altering the physical flows that have historically connected communities. Thus, recognizing the importance of these processes in tandem with species’ movement abilities is critical for predicting and preserving the structure and function of ecological communities.