ETH Zurich, Institute of Environmental Engineering, Schafmattstrasse 6, Zurich, Switzerland.
Chemosphere. 2012 Oct;89(7):850-5. doi: 10.1016/j.chemosphere.2012.05.007. Epub 2012 Jun 4.
We evaluated the exposure to pesticides from the consumption of passion fruits and subsequent human health risks by combining several methods: (i) experimental field studies including the determination of pesticide residues in/on passion fruits, (ii) dynamic plant uptake modelling, and (iii) human health risk assessment concepts. Eight commonly used pesticides were applied onto passion fruits cultivated in Colombia. Pesticide concentrations were measured periodically (between application and harvest) in whole fruits and fruit pulp. Measured concentrations were compared with predicted residues calculated with a dynamic and crop-specific pesticide uptake model, namely dynamiCROP. The model accounts for the time between pesticide application and harvest, the time between harvest and consumption, the amount of spray deposition on plant surfaces, uptake processes, dilution due to crop growth, degradation in plant components, and reduction due to food processing (peeling). Measured and modelled residues correspond well (r(2)=0.88-0.99), with all predictions falling within the 90% confidence interval of the measured values. A mean error of 43% over all studied pesticides was observed between model estimates and measurements. The fraction of pesticide applied during cultivation that is eventually ingested by humans is on average 10(-4)-10(-6), depending on the time period between application and ingestion and the processing step considered. Model calculations and intake fractions via fruit consumption based on experimental data corresponded well for all pesticides with a deviation of less than a factor of 2. Pesticide residues in fruits measured at recommended harvest dates were all below European Maximum Residue Limits (MRLs) and therefore do not indicate any violation of international regulatory thresholds.
（i）包括测定百香果中/上农药残留量的田间实验研究，（ii）动态植物吸收建模，和（iii）人体健康风险评估概念。八种常用的农药被应用于在哥伦比亚种植的百香果上。在整个果实和果肉中定期（在施药和收获之间）测量农药浓度。测量的浓度与用动态和作物特异性的农药吸收模型（即 dynamiCROP）计算的预测残留量进行比较。该模型考虑了施药和收获之间的时间、收获和食用之间的时间、施药到植物表面的喷雾沉积量、吸收过程、作物生长导致的稀释、植物成分降解以及由于食品加工（去皮）造成的减少。测量和建模的残留量之间的相关性非常好（r(2)=0.88-0.99），所有预测值都落在实测值的 90%置信区间内。所有研究的农药的平均误差为 43%，在模型估计值和测量值之间。在平均水平上，人类最终摄入的在种植过程中应用的农药的比例为 10(-4)-10(-6)，这取决于施药和食用之间的时间间隔以及考虑的加工步骤。基于实验数据的模型计算和通过水果摄入的摄入量分数与所有农药都非常吻合，偏差小于 2 倍。在推荐的收获日期测量的水果中的农药残留均低于欧洲最大残留限量（MRLs），因此没有违反国际监管阈值。
