6533b7cffe1ef96bd1258c0e

RESEARCH PRODUCT

Modélisation des effets des systèmes de culture sur la levée des adventices à partir de relations fonctionnelles utilisant les traits des espèces

subject

description

The aim of the Ph. D. Was to adapt the monospecific model ALOMYSYS to a multi-specific flora, in order to predict the effects of cropping systems on the timing and quantities of emerging weed seedlings. Determining the parameters for each additional species is unfeasible. A second objective was thus to estimate the parameters from easily measurable species traits. After having performed a large literature analysis, 18 weeds were studied for the processes involved into the seedbank dynamics and emergence. The seed mortality in the soil could not be related with species traits and can be parameterised with literature data. A generic and simplified formalism represens the seasonal variations of seed dormancy and can be parameterised with literature data and expert knowledge. Seeds then germinate as a function of the soil hydrothermal conditions and of the species moisture and temperature requirements, which were correlated to expert knowledge on weed emergence periods in the field. The seed germination rate was predicted from the nature of their reserves and from the species temperature requirements. A correlation was found between the seed mass and its maximal emergence depth. The seedling mortality caused to soil compaction decreases with increasing shoot diameter. The formalisms and relationships were introduced into the code of the FLORSYS model. A partial evaluation of the model showed than it can be used to predict the emergence of new species, when their parameters are estimated from their traits. Finally, a few simulations show the interest of the model for designing and evaluating innovative cropping systems for weed management in a context of a changing agriculture.