Is Pesticide Risk in France Declining? It Depends on Which Risk You Look At

Authors: Thomas Perrot, Colin Fontaine and Niklas Möhring

Is France making progress in reducing pesticide risks? The answer depends heavily on how risk is measured. In a new study, we apply the Pesticide Load Index to pesticide sales in France between 2012 and 2022. Overall, pesticide risk fell by 29%. Behind this headline number, however, the picture is mixed: risks to biodiversity decreased markedly, risks to human health show no clear trend, and the potential for long-term environmental contamination increased.

Different indicators, different stories

Reducing pesticide risks has become a central policy objective. The UN Convention on Biological Diversity has adopted a goal of halving risks from pesticides and hazardous chemicals by 2030 (CBD, 2022), and France’s Ecophyto 2030 strategy aims to halve pesticide use. France matters a lot in this context: it is the EU’s largest agricultural producer and one of its largest pesticide users.

Whether such targets are met depends on how progress is measured, and the choice of indicator is far from trivial (Möhring et al., 2023). In France, the official indicators tell contrasting stories. The NODU, which measures use intensity in standardised doses, shows a stable trend between 2008 and 2022. The EU’s Harmonised Risk Indicator 1 (HRI-1), in contrast, fell by 36% between 2011 and 2022 (Garnault et al., 2024). Neither indicator uses substance-specific toxicity data: the NODU weights substances by their authorised dose, HRI-1 by four broad hazard categories. Yet toxicity and environmental behaviour vary widely between active substances, and quantity-based indicators can miss the most extreme risks (Möhring et al., 2019).

Measuring three dimensions of risk

In our recent study (Perrot et al., 2026), we use the Pesticide Load Index. It was developed in Denmark, where it has served for more than a decade as an official indicator and as the basis for a risk-based pesticide tax (Kudsk et al., 2018). The index combines three sub-indicators for each commercial product:

• Ecotoxicity Load: acute and chronic toxicity to birds, mammals, fish, daphnia, algae, aquatic plants, earthworms and bees
• Human Health Load: based on the hazard statements and formulation of the product
• Fate Load: persistence in soil, bioaccumulation and risk of leaching to groundwater

We combined sales data at department level from the French national pesticide sales database with product composition data from the E-Phy database of the French agency ANSES. We computed the index with the PesticideLoadIndicator R package (Möhring et al., 2021). The analysis covers 2,977 commercial products containing 293 active substances, which account for 99% of the product mass sold. It starts in 2012 because reporting of seed-treatment sales only became mandatory that year. To assess changes over time, we compare average values for 2012–2014 and 2020–2022 and test for monotonic trends.

Biodiversity risk down, environmental fate up

The Pesticide Load Index decreased by 29% between 2012–2014 and 2020–2022. This is not explained by changes in farmland extent, as total crop area declined by only 1.9% over the period. The three dimensions of risk, however, developed very differently (Figure 1). The Ecotoxicity Load fell by 37%. The Human Health Load was 34% lower in 2020–2022 than in 2012–2014, but it fluctuated strongly between years; over the longer period 2008–2022 it was essentially unchanged (−4%). The Fate Load moved in the opposite direction: it was 13% higher in 2020–2022 than in 2012–2014, and 18% higher over 2008–2022. Given the large year-to-year variation, only the decline in the Ecotoxicity Load is statistically significant over 2012–2022.

Figure 1. Temporal trends of (A) the Pesticide Load Index, (B) commercial product mass sold, and the three sub-indicators (C) Ecotoxicity Load, (D) Human Health Load and (E) Fate Load in France, 2012–2022, by pesticide class. Source: Perrot et al. (2026), CC BY 4.0.

Small quantities, large risks

Insecticides account for only 11% of the product mass sold, but for almost half (48%) of the Pesticide Load Index and more than 80% of the Ecotoxicity Load. They also explain most of the overall decline: without insecticides, the index would have fallen by only 15% instead of 29%. This decline is mainly driven by lower use of two organophosphates, chlorpyrifos-ethyl and chlorpyrifos-methyl, and the neonicotinoid imidacloprid. All three were banned during the study period: imidacloprid in France in 2018, chlorpyrifos in the EU in 2020.

Herbicides show the reverse pattern. They make up half of the product mass sold and more than two-thirds of the Fate Load, and their contribution to the index grew over time: without herbicides, the index would have declined by 39%. The rising Fate Load is mainly driven by increasing use of pendimethalin, which is highly persistent and bioaccumulative, and prosulfocarb, which is highly volatile and frequently detected in non-target compartments such as surface waters. For the Human Health Load, herbicides and fungicides contribute about equally (46% and 43%), with chlorothalonil and metaldehyde as the largest single contributors.

These results also show why indicators diverge. Glyphosate, the most sold herbicide in France, is the main driver of HRI-1 (Garnault et al., 2024), but contributes only 1.7% to the Pesticide Load Index.

Hotspots in the north and west

Pesticide risk is strongly concentrated in space (Figure 2). The highest values are found in northern and western France, with Aube, Pas-de-Calais and Marne at the top. All three sub-indicators show very similar spatial patterns. High values are associated with large areas of sugar beet, vineyards, vegetables and cereals; for the Human Health and Fate Loads, oilseed rape also plays a role. Per hectare, the index in French départements ranges from 3.5 to 5.6, compared with 1.9 to 2.4 in Denmark (Nielsen et al., 2023).

Figure 2. Spatial distribution of (A) the Pesticide Load Index, (B) Ecotoxicity Load, (C) Human Health Load and (D) Fate Load across French départements, averaged over 2012–2022. Source: Perrot et al. (2026), CC BY 4.0.

Bans moved the indicator, but not the system

Our results suggest that the largest changes in pesticide risk in France are linked to regulatory bans of a few highly toxic substances. They do not point to a broader transformation of crop protection practices or farming systems, despite substantial public investment in the Ecophyto programs over the past decade. How farmers responded to these bans, for example by substituting other products or changing crops, and what this meant for their incomes and for consumer prices, requires further causal analysis.

For policy, two lessons stand out. First, insecticides, herbicides and fungicides drive different dimensions of risk, so reducing one class does not address the others: the Fate Load rose while the Ecotoxicity Load fell. Second, a lasting reduction of risks likely requires more than bans alone, namely a combination of regulation, economic incentives, information on targeted applications, new technologies and the redesign of farming systems (Möhring et al., 2025). The spatial hotspots and high-risk crops we identify, such as sugar beet and vineyards, indicate where such measures could be targeted first. Monitoring should also be multidimensional: a single aggregate number can hide opposing trends in different types of risk.

Our study has limitations. The Pesticide Load Index measures potential risk based on substance properties from regulatory laboratory studies. It does not account for metabolites, mixture effects or real-world exposure conditions, and sales data may deviate from actual use, for example due to stockpiling. A decline in the index therefore does not automatically imply a decline in actual impacts on biodiversity and human health. Linking the indicator to field measurements of these impacts is an important next step. The data and R scripts of our study are openly available.

Authors, article and references

Thomas Perrot is at the Fondation pour la Recherche sur la Biodiversité (FRB-Cesab) and the Centre d’Etudes Biologiques de Chizé, France. Colin Fontaine is at the Centre d’Ecologie et des Sciences de la Conservation (CESCO; Muséum national d’Histoire naturelle, CNRS, Sorbonne Université), Paris. Niklas Möhring is at the University of Bonn. Contact: thomas.perrot@fondationbiodiversite.fr and mohring@uni-bonn.de

Article (open access): Perrot, T., Fontaine, C., & Möhring, N. (2026). France’s Pesticide Load Index: Decreasing Ecotoxicity, Persistent Human Toxicity, Rising Concern For Environmental Fate. Peer Community Journal, 6, e109. https://doi.org/10.24072/pcjournal.817

References

CBD (2022). Kunming-Montreal Global Biodiversity Framework (CBD/COP/15/L.25).

Garnault, M., Barbu, C., Aulagnier, A., … & Aubertot, J.-N. (2024). Evaluation of two indicators according to the objectives of the sustainable use of pesticides directive (SUD). A French case study. agriRxiv, 20240181241.

Kudsk, P., Jørgensen, L. N., & Ørum, J. E. (2018). Pesticide Load—A new Danish pesticide risk indicator with multiple applications. Land Use Policy, 70, 384–393.

Möhring, N., Gaba, S., & Finger, R. (2019). Quantity based indicators fail to identify extreme pesticide risks. Science of the Total Environment, 646, 503–523.

Möhring, N., Kudsk, P., Jørgensen, L. N., Ørum, J. E., & Finger, R. (2021). An R package to calculate potential environmental and human health risks from pesticide applications using the ‘Pesticide Load’ indicator applied in Denmark. Computers and Electronics in Agriculture, 191, 106498.

Möhring, N., Kanter, D., Aziz, T., … & Leadley, P. (2023). Successful implementation of global targets to reduce nutrient and pesticide pollution requires suitable indicators. Nature Ecology & Evolution, 7, 1556–1559.

Möhring, N., Ba, M. N., Braga, A. R. C., … & Finger, R. (2025). Expected effects of a global transformation of agricultural pest management. Nature Communications, 16, 10901.

Nielsen, H. Ø., Konrad, M. T. H., Pedersen, A. B., & Gyldenkærne, S. (2023). Ex-post evaluation of the Danish pesticide tax: A novel and effective tax design. Land Use Policy, 126, 106549.

Acknowledgements: This blog article was generated with Claude Opus 5.5 and manually edited by the authors afterwards.