The integration of an individual-based model into toxicokinetics to enhance ecological realism in evaluating population-level impacts of exposure to PCB Maryam Karim Pour* Great Lakes Institute for Environmental Research (GLIER), University of Windsor, Windsor, Canada Karimpo@uwindsor.ca Sourodeep Bhattacharjee School of Computer science, University of Windsor, Windsor, Canada bhattac1@uwindsor.ca Robin Gras Associate Professor in School of Computer science, Cross-appointed by the Biological Sciences Department and Great Lakes Institute for Environmental Research (GLIER), University of Windsor, Windsor, Canada rgras@uwindsor.ca Ken Drouillard Professor in Great Lakes Institute for Environmental Research (GLIER) and Biological Sciences Department, Adjunct Professor in UNU-INWEH University of Windsor, Windsor, Canada kgd@uwindsor.ca Abstract: Polychlorinated biphenyls (PCBs) are classified as one the most extremely regulated anthropogenic contaminants and they have been deeply probed in aquatic ecosystems. However, there is very limited understanding of the population level effects of exposure to PCBs on terrestrial animal species and this has been unanimously indicated as a critical gap in ecological risk assessment. To bridge this information gap, we integrated an individual-based model (IBM) framework into toxicokinetics resulting in a deeper ecological insight to simulate the accumulation of a hypothetical PCB in a terrestrial three-level food chain at the population level. We then validated our simulated system utilizing the observed field bioaccumulation factors in a well-studied terrestrial prey- predator, caribou-wolf. Key findings of the present study indicate that in a PCB- contaminated environment, where all food sources contain some amount of contaminants, producing more offspring results in lower toxic concentration in herbivores (prey) and higher concentration in carnivores (predator). Our novel contribution in this work is that we have achieved a validated system that enables us to investigate toxicokinetics in any animal species involved in a prey-predation interaction by providing lipid, non-lipid, and water fractions in their bodies. Additionally, we demonstrated how using IBM modelling approach could facilitate ecological risk assessment by offering detailed information of generations spanning as many years as required. Keywords: Individual-based model, Toxicokinetics, Ecological risk, Population-level assessment, and Polychlorinated biphenyls. I. Introduction Polychlorinated biphenyls (PCBs) are regarded as one of the most persistent environmental contaminants. As a result of stability and lipophilic properties, PCBs are chemically and biologically persistent. Consequently, they bioaccumulate in the lipid contents of organisms and biomagnifies throughout food chains. Although PCBs are no longer produced, owing to the relatively slow rate of degradation, they are still adversely affecting living organisms, their populations, their natural habitats, and the functioning of ecosystems [1]. The potential risks to ecological systems associated with exposure to these chemical stressors are mainly evaluated by means of ecological risk assessment. Individual’s exposure to PCB-contaminated environments and contaminated food resources leads to corresponding negative impacts at the population level. Thus, the primary concern in ecological risk assessment is predicting hidden threats of these persistent chemicals at the population level, which is difficult to determine through empirical studies regarding time and financial constraints. Fortunately, Individual-based modeling approaches (IBMs), where population level characteristics emerge from heterogeneous individuals interacting with their biotic and abiotic environments, are designed to facilitate this process. Therefore, there has recently been considerable growth in leveraging IBMs approaches across ecological risk assessment [2][3][4][5]. This increase stems from the improved predictive power of such modeling approaches in recent years and also recently issued appeals for more realistic evaluation of potential risks of contaminants [3][6]. By including a detailed consideration of spatial variation, demographic stochasticity, and population dynamics emerging from unique individuals interacting with each other and with their environment, IBMs pave the way for ecological realism and thereby improving their corresponding predictabilities [7][8][9]. Ultimately, the incorporation of detailed sub-models into IBM approaches would lead to more reliable predictions. This is because sub-models address and quantify the behavior of contaminants in abiotic media and consequently transfer this into living organisms, thereby accumulating throughout the food chains. These aspects, coupled with the invaluable ecological realism provided by IBMs, leads to far more reliable predictions [10]. Taking this into consideration, we applied a complex IBM framework called EcoSim [11][12] as a virtual laboratory 978-1-4673-9669-1/15/$31.00 ©2015 IEEE * Corresponding author