South Africa is the leading pesticide user in Sub-Saharan Africa. However, little is known about the occurrence of pesticide mixtures in surface water and potential environmental risks in Africa. This study investigated the occurrence of pesticides mixtures in three watersheds during a drought year in South Africa. The study was conducted in the Krom River, Berg River and Hex River watersheds within larger agriculture systems in the Western Cape. Pesticide spray records were collected from 38 farms. A total of 21 passive water samplers (styrenedivinylbenzene disks (SDB)) were deployed, each for two weeks per month, over seven sampling rounds during the main pesticide application period between July 2017 and January 2018. Samples were analyzed for 248 pesticide compounds using LC-HR-MS/MS. Pesticide occurrence was analyzed for temporal agreement with pesticide spraying events (Cohen's κ) and correlation with rainfall patterns and river discharge (Pearson correlation (r p )). Pesticide time-weighted average concentrations were estimated and compared to environmental quality standards (EQS). According to the farm spray records, 96 different pesticides were sprayed during the sampling period and differed considerably between the three study areas, seasons and crops grown. In total, 53 compounds were detected in river water. We detected 39% of compounds from the spraying records and demonstrated close temporal correlations of seasonal patterns for 11 pesticide compounds between reported on spraying records and observations in the streams (κ = 0.90). However, 23 detected pesticides were not found on spray records, many of them being herbicides. Most of the estimated two-week average pesticide concentrations were below 40 ng/L. The insecticides imidacloprid, thiacloprid, chlorpyrifos and acetamiprid and the herbicide terbuthylazine exceeded at least once their EQS 58-fold (EQS 13 ng/L), 12-fold (EQS 10 ng/L), 9-fold (EQS 0.46 ng/L), 5-fold (EQS 24 ng/L) and 3-fold (EQS 220 ng/L), respectively. Our study substantially widens the view on pesticide pollution in surface water compared to previous studies in Sub-Saharan Africa by targeting more than 200 pesticides using passive sampling systems. This broad assessment revealed the presence of 53 compounds, some of them in high concentrations, indicating possible adverse effects on biota and the quality of the ecosystem. Whether the observed concentration levels in the year 2017 were exceptional due to the lowest ever recorded rainfall and river discharge needs to be tested with additional data to better understand how pesticide pollution levels manifest under average rainfall and river discharge conditions.

In December 2020, high soil concentrations of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) were discovered across large parts of Lausanne, Switzerland. Concentrations reached up to 640 ng TEQ_WHO-2005/kg dry weight. The most likely source was a former municipal waste incinerator. A three-step, multidisciplinary approach to human health risk assessment was conducted to determine the potential population exposure to PCDD/Fs and identify appropriate preventive measures. First, exposure scenarios were developed based on contaminated land uses. Second, the toxicological risks of different scenarios were evaluated using a toxicokinetic model estimating increases in blood serum PCDD/F concentrations over background concentrations from the general population's food consumption. Third, a detailed geostatistical mapping of PCDD/F soil contamination was performed. Stochastic simulations with an external drift and an anisotropic model of the variogram were generated to incorporate the effects of distance from emission source, topography, and main wind directions on the spatial distribution of PCDD/Fs in topsoil. Three main scenarios were assessed: i) direct ingestion of soil by children in playgrounds; ii) consumption of vegetables from private gardens by children and adults; and iii) consumption of food from livestock and poultry raised on contaminated soil. The worst exposure scenario involved the consumption of eggs from private hen houses, resulting in PCDD/F concentrations in serum an order of magnitude higher than might normally be expected. No relevant increases in serum concentrations were calculated for direct soil ingestion and vegetable consumption, except for cucurbitaceous vegetables. Combining mapping and exposure scenario assessment resulted in targeted protective measures for land users, especially concerning food consumption. The results also raised concerns about the potential unsafe consumption of products derived from animals raised on land with PCDD/F concentrations only moderately over environmental background levels.

Pesticides are intensely used in the agricultural sector worldwide including smallholder farming. Poor pesticide use practices in this agronomic setting are well documented and may impair the quality of water resources. However, empirical data on pesticide occurrence in water bodies of tropical smallholder agriculture is scarce. Many available data are focusing on apolar organochlorine compounds which are globally banned. We address this gap by studying the occurrence of a broad range of more modern pesticides in an agricultural watershed in Uganda. During 2.5 months of the rainy season in 2017, three passive sampler systems were deployed at five locations in River Mayanja to collect 14 days of composite samples. Grab samples were taken from drinking water resources. In these samples, 27 compounds out of 265 organic pesticides including 60 transformation products were detected. In the drinking water resources, we detected eight pesticides and two insecticide transformation products in low concentrations between 1 and 50 ng/L. Also, in the small streams and open fetch ponds, detected concentrations were generally low with a few exceptions for the herbicide 2,4-D and the fungicide carbendazim exceeding 1 ug/L. The widespread occurrence of chlorpyrifos posed the largest risk for macroinvertebrates. The extensive detection of this compound and its transformation product 3,4,5-trichloro-2-pyridinol was unexpected and called for a better understanding of the use and fate of this pesticide. Graphical abstract

Few epidemiological studies use exposure determinants specifically tailored to assess pesticide or plant protection product (PPP) exposures when assessing presumed association between occupational exposure and health outcomes among agricultural workers. This lack of exposure specificity could lead to results that fail to detect an association. It could be related to the lack of consensus on exposure assessment methods and the choice of exposure determinants. We conducted a meta-analysis following the PRISMA checklist to identify PPP exposure determinants used in occupational studies and identified exposure determinants that best characterized agricultural exposures to PPPs. Out of 1436 studies identified, 71 were included. The exposure determinants identified were active ingredients, chemical classes, types of PPP, crops, tasks, frequencies, duration, lifetime exposure days, and intensity-weighted exposure days. Only six over 17 associations between exposure determinants and health outcomes were found with moderate quality of evidence. Overall, epidemiological studies had difficulty defining relevant determinants to characterize PPP exposures for agricultural workers. We recommend that a standardized list of determinants for PPP exposures in occupational exposure studies should include information on formulations, intensity, duration, and frequency of PPP exposure. Harmonized data collection on exposure and health outcomes are required as well as standard units for each exposure determinant.