Jolliet O. 1, Charles R. 2 and Margni M. 1
1 Swiss Federal Institute for Technology,
2 Swiss
research Station for Plant Production,
The evaluation of pesticide residue
in agricultural commodities is usually performed by analytical processes.
Alternative and complementary approaches by modelling the fate of pesticide in
agricultural plants are needed
-
to calculate pesticide transfer factor to food and exposure
efficiency to human
-
for a better reliability and admittance of Life Cycle
Assessment (LCA) methodology for chemicals in agriculture
-
for evaluating concentration levels under the limit of
detection
-
for wide-reaching conditions beyond local and specific
circumstances
To this purpose, the following
challenges need to be addressed. How can the direct application of pesticides
be modelled ? What is the importance of pesticide
residue deposited on cuticle compared to residues in soil and in air ? What is the dynamic behaviour of pesticide in plants ? How does this affect the final residues in plants
depending on the time interval between application and harvest
? These points must be addressed by identifying principal processes
responsible for the transfer of substances applied directly on plants and in
the near environment and for the dynamic behaviour of substances in plant
system.
Developments
were brought in the identification of critical factors for the fate of
pesticides and pollutants in agricultural plants and for the residue evolution
till harvest. Environmental models are usually running in steady state. The
development of a time dynamic plant model opens prospects in the evaluation and
quantification of xenobiotics in food. Methodological
implements were brought according to agricultural
processes determining the fate of pesticides: treatment products sprayed
directly on the plant, time dynamic processes according to varying treatment
period. The functioning of the system phytosanitary
measures – plant – environment was studied in details. Exchanges between
environmental and plant compartments are described in form of transfer rates. A
dynamic model calculates the concentrations evolution during crop growth.
The
model was tested on wheat crop. Experimental data of pesticides residues were
used to assess the initial concentration of substance in plant just after
spraying and the residue evolution till harvest. A second approach consisted in
studying the functioning of the model according to different types of plant
treatment strategies.
Results
show that main processes determining the fate of pesticides in crops concern
initial concentration, time from spraying to harvest and degradation rate.
Pesticides sprayed on the plant are generally rapidly absorbed, so that initial
concentration in the plant may reach high levels. The absorption efficiency
from the air is inferior, basically due to the high dilution volume. Transfers
of substances from the soil last longer according to the higher persistence of
substances in this compartment. Accumulation processes are counterbalanced by
the substance dissipation in the system, due to degradation and plant growth.
Due to the narrow diversity of approved pesticides, physical properties
(partition coefficients) of substances play a minor role on the variation
between pesticides fate, except for some atypical substances and for
applications shortly before harvest.
Concentrations
of pesticide at harvest varies by a factor
100, 1000, up to 105 between substances with identical action.
Identification of alternative substances to substitute problematic ones can be
more objectively supported.
Further
developments concern simplification of the model to the main processes and
extension to different agricultural conditions. Integration of main identified
plant processes in environmental multimedia models are
also ongoing, so that the significance of pesticides by contrast with
environmental pollutants can be compared according to human exposure and
toxicity.
The
plant model is incorporated into the multimedia fate&exposure
model "Impact 2002" to calculate the pesticide intake fractions: the
fraction of the applied pesticide, which is taken in by a population (Bennett
et al., 2002). Impacts on human health are then calculated using an effect
factor. This factor is based on the risk assessment concept of a benchmark dose
and the concept of Disability Adjusted Life Years (DALYs)
and is detailed in
Crettaz et al. 2002. Dose-response slopes based on
benchmark dose ED10s (Effect Dose 10%) are given for more than 600 carcinogenic
compounds, and average DALY's determined on the basis
of 17 types of cancer. For non-carcinogenic impacts, Pennington et al. (2002)
presented slope factors from bioassay data for 12 chemicals and, using
best-estimate extrapolation factors from NOAEL and LOAEL data, facilitated the
initial calculation of slope factors for an additional 403 compounds. This
approach is finally illustrated through results on the human health impacts of
several key pesticides.