Models that quantify the effects of cropping systems on weed dynamics are useful tools for testing innovative cropping systems. In these models, seed mortality in the soil is a key parameter to account for the cumulated effect of cropping systems over time via the soil seed-bank. Since seed mortality is difficult to measure, our objective was to develop a method to estimate it from easily accessible information. Seeds of 13 weed species were buried 30cm deep in fields and were recovered regularly for 2 years to measure their viability. Seed mass, dimensions, shape, and protein and lipid contents as well as coat thickness were measured. To estimate seed mortality of species not included in the study, we searched for relationships between mortality rates and seed traits. Seed viability mainly decreased during the second year of burial, with mortality rates ranging from 0.01 to 0.63 seeds.seeds(-1).year 21, depending on the species. Seed mortality decreased with increasing seed coat thickness. No correlation was found with other measured traits or with seed persistence data in the literature. These results were confirmed when the effects of phylogenetic relatedness with phylogenetically independent contrasts were included. The thickness of the seed coat, which varied between 17 and 231 mm over the range of species studied, can protect the seed from external attacks in the soil and slow down seed decay. This trait can be easily measured via X-ray images and could be used to estimate the seed mortality rate for a wider range of species

To the best of our knowledge, this is the first soft X-ray radiographic study of the development of an entire young root system of western red cedar (Thuja plicata D. Don). The experimental device was adapted to radiographic constraints and allowed the plant to grow for several months. Successive radiographs revealed primary growth and differentiation, secondary growth, and primordium development. The younger (8-day-old) part of the root system appeared fully opaque. Afterwards, the stele and the cortex could be distinguished from each other progressively in a second zone where most of the primary differentiation had already occurred. Measuring early cambial growth was made possible by spacing the radiographs at 10-day intervals. Time-course developments of primordia and their histolysis cavities were traced for 8-15 days, depending on the sample. The acropetal sequences of primordia along the same adventitious root were shown. Primordia had different growth curves, most of which were nearly logistic. These meristems were produced by unequal primordia, which generated axes of unequal diameters. Our techniques provide new opportunities for understanding the differentiation of axes and defining potential root growth during primordium development.Key words: soft X-ray radiography, root development, primordia, root differentiation, radial growth, morphogenesis.