PARENT SESSION
WA1 - Bioavailability
Chair: Griscom, Sara1, 1 48 Main Street, Southborough, MA
8:00 AM to 12:00 PM - Wednesday, 20 November 2002
Room Ballroom G-I
(492) Plants' Dynamic Interactions with Soil: A Feedback-Control Model for Metal Accumulation in Plants.
Saxe, Jennifer*,1, Impellitteri, Christopher2, Allen, Herbert3, 1 Gradient Corporation, Cambridge, MA, USA2 USEPA National Risk Management Research Laboratory, Cincinnati, OH, USA3 University of Delaware, Newark, DE, USA
ABSTRACT- Predictive models for the determination of plant-bioavailable fractions of potentially toxic metals in soil are useful for screening-level ecological risk assessment and in determining the feasibility of phytoremediation for a site. Much of the research toward this end has focused on the physico/chemical aspects of metal binding and mobility in soil, often as a function of pH and other soil properties. Such "supply-side" models mainly consider metal-soil interactions and can only indirectly consider plant-soil and plant-metal interactions, although they may be important. It is well understood that root exudates, excreted primarily in response to macronutrients, significantly alter soil chemistry in the vicinity of plant roots. Some soil properties that influence metal solubility in soil, and are often included in supply-side models, also influence plant root responses (e.g. pH). However, there are other soil factors that exclusively influence plants (e.g. N) and are typically not measured in supply-side models. We developed a dynamic feedback-control model of plant roots, considering soil chemistry, its influence on root exudations, and the resulting quantitative impact of those exudations on metal uptake. The model was developed using Simulink (The MathWorks) to simulate feedback-control mechanisms in root-soil interactions. The model uses principles of both soil chemistry and plant physiology, and was calibrated using data from several plant species grown in field-contaminated soils with and without nutrient amendments. A wide range of soils from three continents, some historically contaminated from anthropogenic sources, were used. This model allows estimation of the concentration of metals in plant tissues based on soil properties. Feedback-control parameters differ substantially between models for the three different plant species studied and for the two metals for which our data set is most robust, copper and zinc. This result explains why supply-side models are more successful predictors for certain species-metal combinations than for others.
Key words: plant bioavailability, copper, zinc, predictive model