Abstract :
[en] In the context of climate change, population growth, and degradation of natural
resources, it is urgent to rethink our production systems to make them more
sustainable. In this regard, a better understanding of the environmental fate of
pesticides is essential. Most current studies rely on generic mobility parameters, which
can result in inaccurate assessments of contamination risks. Furthermore, soil column
leaching experiments are commonly conducted using a wide variety of
methodologies, which may significantly influence the outcomes. Additionally, the
long-term impact of agricultural practices and climatic conditions on soil
hydrodynamic properties is rarely assessed directly in the field.
This thesis aims to address these gaps. First, it evaluates the impact of several
commonly used leaching methodologies on experimental results. Second, it
investigates the fate of eight herbicides of concern for groundwater contamination in
Wallonia using undisturbed soil column experiments under various cropping systems
and soil depths. Site-specific mobility parameters were derived through dual-porosity
inverse modelling and compared to generic values used in Belgium and from
regulatory databases. Third, this work assesses the long-term effects of production
systems that incorporate sustainable agricultural practices under contrasting climatic
conditions on the temporal evolution of soil water retention, using in situ monitoring.
Finally, a comprehensive database was developed to document the influence of these
systems on pesticide, metabolite, and nitrate leaching, as well as on soil structure and
hydrodynamic properties, through high-frequency hydrological monitoring. The
evolution of soil water retention curves over a three-year period was analysed and
compared to laboratory-derived curves (ku-pF method) and pedotransfer function
predictions from the EU-HYDI database.
Results demonstrate that soil structure, column dimensions, and sampling
techniques greatly affected solute transport dynamics and water infiltration. Disturbed
columns tend to underestimate rapid contaminant transport and overestimate
retention, leading to biased contamination risk assessments. Shorter columns
overestimated leaching potential while underestimating degradation processes.
Furthermore, columns sampled with mechanical corers showed artificial preferential
flow caused by vibration, compromising the representativeness of water and solute
transport. These findings highlight the critical role of column design and sampling
methods in leaching experiments, emphasizing the need standardised experimental
protocols to improve the reliability of transport estimates.
Moreover, different soil properties and structure between soil depths had a greater
impact on pesticide leaching behaviour than the cropping systems. Significant
variations in pesticide transport and retention were observed between soil horizons,
illustrating the inadequacy of relying solely on surface parameters for the entire soil profile, which may lead to underestimating groundwater contamination risks. Root
architecture and surface tillage were found to affect pesticide leaching dynamics,
suggesting that cropping systems could serve as strategic levers to mitigate
groundwater contamination. Experimental transport parameters showed discrepancies
with established databases, which often overestimate retention and underestimate the
production of metabolites. These study underline the need of adjusting transport
parameters to site-specific conditions and systematically accounting for metabolite
behaviour. Future research should focus on long-term monitoring of the effects of
sustainable agricultural practices on pesticide behaviour over several seasons and for
a range of soil types.
The thesis also reveals that agricultural practices and crop types have a stronger
influence on soil water retention dynamics than seasonal wetting-drying cycles, plant
development or interannual climatic variability. Practices such as crop differenciation,
weed control, crop residue management, compaction during harvest, and the
introduction of temporary grasslands induced significant changes in soil water
retention capacity, persisting for more than two years in some cases. Comparisons of
SWRCs showed that theoretical curves derived from PTFs poorly represent actual
field conditions, especially under alternative agricultural systems. The laboratory
curves are closer with similar trends but are not optimal. These findings, which are
rarely documented in situ, demonstrate that soil water behaviour is far from static and
sensitive to management practices. They further highlight the potential of agricultural
systems as effective levers to enhance both water retention and food system resilience
under future climate conditions. Therefore, to assess the relevance of future
production systems, studies should focus on the impact of multi-cropping systems on
water retention dynamics, continuously and directly in the field.
Finally, the comprehensive database established provides a valuable resource for
supporting sustainable soil and water management decisions, support environmental
protection, and improve the predictive performance of hydrological and
agroecosystem models, addressing challenges such as climate resilience and food
security.
