This piece was written together with Julia Köninger from University of Vigo, Spain, who co-led the corresponding article - we navigated this adventure together!
All of us know that pesticides can affect above-ground organisms beyond their intended targets (e.g., pollinators such as social bees) but their effects on soil organisms have received very little attention. Yet, soils host approximately 60% of the world's total biodiversity and these organisms are not just passive inhabitants. They underpin essential ecosystem functions, including food production, carbon storage, erosion control, and water regulation.
One of the challenges in assessing the broad-spectrum effects of pesticides on soil biodiversity is the lack of empirical data collection in the field. This gap motivated a collaboration between ten research institutions from across Europe, allowing us to explore pesticide impacts on soil biodiversity at an unprecedented continental scale.
Our continent-wide study was made possible by a unique survey, namely LUCAS Soil, of which we considered 63 common pesticide residue measurements combined with soil biodiversity data from 373 sites in croplands, grasslands and woodlands (i.e., forests) across 26 European countries (Figure 1). DNA-based methods allowed to assess the biodiversity of major organism groups, including archaea, bacteria, fungi, protists, nematodes, and arthropods. In addition, we examined functional groups, like pathogens and beneficial symbionts, as well as genes involved in key ecosystem processes such as nitrogen and phosphorous cycles.
We especially accounted for a number of key environmental drivers, such as soil properties, climate, and ecosystem type, allowing us to evaluate the relative importance of pesticide residues compared to other abiotic factors in shaping soil biodiversity. This is the first study of this extent, and the first to include not only croplands but also grasslands and woodlands.
Although we expected to detect pesticide residues at many sites, we had hoped that this concern would be disproven by the data. Instead, 70% of the investigated sites tested positive for at least one residue, including some grasslands and woodlands. This led us to first analyse the data across all ecosystem types together before repeating the analyses exclusively in croplands. As expected, the patterns became even more pronounced when considering croplands alone. We anticipated clear negative effects on soil organisms in croplands and weaker effects when also including grasslands and woodlands. While this expectation was partly met, several results surprised us.
Our results demonstrate that in croplands and beyond, pesticides were, after soil properties, the second major driver of soil biodiversity patterns. Besides, pesticide effects were not uniformly negative. In particular, some bacterial groups benefited from higher pesticide concentrations, probably because they can use pesticides as a food source or benefit from altered microbial interactions in the soil. Therefore, the responses of soil biodiversity were variable depending on ecosystem type, organism group, functional role, and pesticide type, indicating that inferences for one ecosystem or soil type cannot be extrapolated to other regions.
Particularly concerning was the suppression of beneficial non-target taxa, including arbuscular mycorrhizal fungi and bacterivore nematodes, which are essential for plant growth, nutrient cycling and pest control. While fungicides accounted for more than half of all detected residues, increasing fungicide concentrations were negatively associated with a wide range of non-target groups from bacteria, nematodes and arthropods.
Furthermore, our functional gene analyses revealed that certain pesticides affected particular functional capacities of soil biota. This is because some pesticides are difficult to break down and they remain in the soil for years after application. If the well-functioning of soil is reduced, then other costly interventions (e.g., additional fertilization) are necessary to maintain yields in the long-term. Therefore, to improve pesticide risk evaluation, future frameworks should set acceptable exposure limits based on effects on biological communities and ecosystem functioning, e.g. through declines in biodiversity or the loss of important ecological functions.
Current risk assessment frameworks indeed primarily evaluate single substances using a limited number of test species and functions, often under simplified laboratory conditions, which may incompletely capture pesticide-driven risks to key ecosystem functions. However, our results indicate that more realistic approaches including more (complex) soil communities, evaluating their responses under field conditions, and accounting for the presence of pesticide residues from past practices or mixtures, will help to better capture the complexity of pesticide effects across different environments.
More transparent pesticide use data are urgently needed – also because our study lacked detailed information on pesticide application rates, timing, and mixtures at each site, preventing us from fully disentangling pesticide effects from those of intensive land management more broadly. While patterns found in our study represent early-warning signals, only through such integrative approaches we can better understand how pesticide use shapes soil biodiversity and the ecosystem functions upon which we depend.
References
Köninger, J., Labouyrie, M., Ballabio, C. et al. Pesticide residues alter taxonomic and functional biodiversity in soils. Nature (2026). https://doi.org/10.1038/s41586-025-09991-z
Poster image from © AdobeStock ref. 188819540