Magnitude and decline of pesticide co‐formulant residues in vegetables and fruits: results from field trials compared to estimated values

Abstract BACKGROUND The application of plant protection products (PPPs) leads to the formation of residues in treated crops. Even though PPPs contain considerable amounts of co‐formulants, regulation and monitoring of residues normally focus on the active substances (a.s.) only. For our study we selected four commonly used co‐formulants (three anionic surfactants and one organic solvent) and investigated the formation and decline of residues in vegetables and apples under field conditions. The aims were to characterize the behavior of co‐formulant residues on crops and to provide a basis for future investigations on consumer exposure. RESULTS The development of robust and sensitive analytical methods allowed the quantification of residues in the low μg/kg‐level. After treatment with PPPs, co‐formulants were detected up to approximately 10 mg kg–1 in vegetables. In general, these residues declined fast with half‐lives of a few days. Wash‐off and volatilization were identified as important removal processes for anionic surfactants and the organic solvent, respectively. However, in specific crops (parsley and celery), organic solvent residues were still considerable (≈2 mg kg–1) 2 weeks after treatment. We further demonstrate that it is feasible to estimate co‐formulant residues using publicly available data on pesticide a.s. CONCLUSION To date no information on co‐formulant residues in food is available. The findings from our field trials, as well as the presented approach for the prediction of residues, provide key elements for future consideration of consumer exposure to PPP co‐formulants. © 2020 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.


Supporting information 1: Details on field trials, test sites, and selected substances
plot size: 1.5 m x 7 m. Seedlings were purchased from Jud Bio-Jungpflanzen AG, Tägerwilen. The application of all plant protection products (PPPs) was conducted with a self-constructed plot sprayer, equipped with IDK 120 04 nozzles (distance between nozzles 25 cm) at a velocity of 3.6 km/h and a pressure of 1.7 bar.       (middle), or SDS and docusate (right), respectively. The broad peak for DBNS likely is due to the presence of several, partially separated isomers. Likewise, spiroxamine consists of two diastereomers which are only partially resolved under the chosen chromatographic conditions. The small peak at 6.2 min. in the mass trace for trifloxystrobin is a matrix component which did not interfere with our measurements. Note: positive and negative ion transitions were acquired in separate chromatographic runs. Sigma-Aldrich, St. Louis, MO a Technical product Dispergator B Gran, which consists ≈40-50% of a mixture of various (butyl) naphthalene sulfonates (the reaction product of naphthalene and butanol, sulfonated and neutralised). The content of dibutyl naphthalene sulfonate in the dispersing agent was estimated from HPLC-UV analysis assuming identical response of all individual components. Dispergator B Gran was kindly provided by Syngenta.  The stability was determined by re-analysis of the samples from day 0, with the exception of some values that were determined from day 1 (#) or day 2 (##) samples. The indicated storage stability values represent the mean as calculated from two separately processed subsamples (value in stored sample expressed in % of the value before storage). n.a.: not applicable, as residues were below or close to limit of quantification (SDS in rondini) or stability experiments were not conducted (DMDA and spiroxamine in apple). Supporting Information 3: Further information on results and discussion

Notes to table:
For fitting of the decline curves of co-formulants and the two monitored a.s. in vegetables and apples the program CAKE (Version 3.4) was used (available at https://www.tessella.com/showcase/computer-assisted-kinetic-evaluation).
The kinetic model has two fitting parameters (initial residue and degradation rate=ln2/DT50); initial residues were not fixed. Residue values <LOQ were set to 0.5 x LOQ for the first time point with a concentration level <LOQ, consecutive time points were not considered then. SFO fits were run for all available co-formulant/crop combinations, provided there where at least five data points available. Examples see Figure SI 3.2.
n.a.: not available (no fits could be run because there were not enough values >LOQ available).
 2 (Chi-square) and r 2 are measures for the deviation between observed and predicted values, as described in the respective EU Guidance 4 . Further, the quality of the fits was checked and qualified (good, acceptable, or poor) by visual examination. 6.17 0.9670 good a g is the fraction of residues at time 0 which is subject to fast degradation

Notes to table:
The decline curves of docusate, SDS, and trifloxystrobin in apples were well described by bi-exponential kinetics (DFOP: double first order in parallel, Figure SI 3.2), assuming a substance is applied to two compartments with fast and slow decline, respectively. Further information and reference, see Notes to table SI 3.3a.
Decline curves of co-formulants and actives in vegetables did not exhibit a clear bi-phasic behaviour. Kinetics other than SFO were, therefore, not applied.