483 Environmental Toxicology and Chemistry, Vol. 22, No. 3, pp. 483–493, 2003  2003 SETAC Printed in the USA 0730-7268/03 $12.00 + .00 GENERAL FUGACITY-BASED MODEL TO PREDICT THE ENVIRONMENTAL FATE OF MULTIPLE CHEMICAL SPECIES THOMAS M. CAHILL,IAN COUSINS, and DONALD MACKAY* Canadian Environmental Modelling Centre, Trent University, 1600 West Bank Drive, Peterborough, Ontario K9J 7B8, Canada ( Received 1 April 2002; Accepted 26 August 2002) Abstract—A general multimedia environmental fate model is presented that is capable of simulating the fate of up to four interconverting chemical species. It is an extension of the existing equilibrium criterion (EQC) fugacity model, which is limited to single-species assessments. It is suggested that multispecies chemical assessments are warranted when a degradation product of a released chemical is either more toxic or more persistent than the parent chemical or where there is cycling between species, as occurs with association, disassociation, or ionization. The model is illustratively applied to three chemicals, namely chlorpyrifos, pentachlorophenol, and perfluorooctane sulfonate, for which multispecies assessments are advisable. The model results compare favorably with field data for chlorpyrifos and pentachlorophenol, while the perfluorooctane sulfonate simulation is more speculative due to uncertainty in input parameters and the paucity of field data to validate the predictions. The model thus provides a tool for assessing the environmental fate and behavior of a group of chemicals that hitherto have not been addressed by evaluative models such as EQC. Keywords—Multispecies model Environmental fate Chlorpyrifos Pentachlorophenol Perfluorooctane sulfonate INTRODUCTION Several multimedia environmental models have been de- veloped to predict chemical concentrations, distribution, and persistence in both real and evaluative environments using physical–chemical partitioning properties and reaction-rate es- timates [1–3]. Although these models calculate the rate of loss of the parent chemical, they do not consider the secondary products that are formed. Modeling the parent and secondary species separately may be feasible, but it is often unsatisfactory because the rate and location of the formation of the secondary species may not be obvious and the conversion may be re- versible. In many instances, multispecies environmental fate models are needed to establish a more comprehensive under- standing of chemical fate by including an assessment of the breakdown products as well as the parent chemical. There is a view that the assessment of a substance’s degradation in the environment should include all degradation products until complete mineralization occurs. Pesticides are often assessed on this basis. Previous multispecies models [4–6] have focused on esti- mating the overall persistence of a parent chemical and a series of degradation products. The models developed by Quartier and Mu ¨ller-Herold [4] and Gonza ´lez et al. [6] treat only a single compartment and are thus not multimedia in nature. The model by Fenner et al. [5] has three environmentally relevant compartments, but it has been applied only to a parent chemical and a single transformation product with the aim of relating the overall persistence of the parent and product pair to the persistence properties of the individual species. No environ- mental concentrations were predicted, although such a model can be applied to predict environmental concentrations. In ad- dition, the model assumes that the product yield of all reactions * To whom correspondence may be addressed (dmackay@trentu.ca). is unity (i.e., all the reactant transforms into a given product), which may limit the description of chemicals with more com- plex reaction pathways. None of these models allow for chem- ical species to cycle between forms, as is the case for ionizing chemicals. Diamond et al. [7] provided a conceptual frame- work for a multispecies model that can assess cycling chem- icals using forward and backward conversion rates. It was concluded that a simpler approach is to use the observed steady-state ratio of the chemical species and then write a pseudo–single-species equation for the sum of the chemical species. This approach has been successfully applied to mer- cury in a Canadian shield lake [8] and Lahontan Reservoir (NV, USA) [9], but it cannot be applied to dynamic situations where interconversion rates are relatively slow compared with other environmental processes or situations where one or more of the reaction pathways are unidirectional. The Exposure Analysis Modeling System [10], developed at the U.S. En- vironmental Protection Agency, is a multispecies aquatic fate model that can simultaneously simulate the fate of the parent compound and its degradation products, and it allows chem- icals to cycle between different forms (e.g., ionization). The Exposure Analysis Modeling System, however, was designed to simulate aquatic fate and exposure, so it treats only water and sediment compartments. Models linking mass transport and species transformation have also been widely used in the radionuclide literature for years, but radioactive decay is a unidirectional process, so these models consider only the for- ward reactions and do not consider the cycling of chemical species. There are at least five situations in which an investigation of multiple chemical species is warranted. The first is when one (or more) of the degradation products is more toxic than the parent chemical. Organophosphorous pesticides, such as malathion, chlorpyrifos, and parathion, are classic examples where a degradation product, namely the corresponding oxon