Profondeur d'extraction racinaire et signature isotopique de l'eau prélevée par les racines des couverts végétaux

We seek to identify the depth to which water is extracted by the roots in the soil. Indeed, in an isotopic steady-state condition of leaf water, transpiration introduces into the atmosphere a vapour whose isotopic signature is identical to that of root water. In the isotopic models of atmospheric general circulation, it is classically allowed that the signature of transpiration belongs to the meteoric water line. This supposes that the water taken by the roots has escaped with the evaporation of the soil and comes thus from the deep layers of the soil. At the time of experimentation carried out on maize plants (Nemours, Seine-et-Marne, France), this extraction depth was inferred from the comparison between the signature of the water measured on the level of the first internode of the stems of the plants and the isotopic profile of water in the soil. When the flow of transpiration reaches a maximum value, the plant uptakes water resulting from precipitations and which preserves its non-evaporating character after having quickly infiltrated in the deep layers of the soil. This relates to only 55% of the flux transpired by the canopy, the remainder presenting an evaporating character more or less marked according to ambient conditions. This experiment invalidates the classical hypothesis used in isotopic models of general atmospheric circulation in temperate regions. In fact, only half the amount of water vapour transpired by the canopy during the day presents a signature similar to that of the rainwater sampled in deep soil layers. To cite this article: Z. Boujamlaoui et al., C. R. Geoscience 337 (2005).     

The response in floodplain respiration of an alpine river to experimental inundation under different temperature regimes

The respiratory potential [i.e. electron transport system activity (ETSA)] of soils and sediments from five floodplain habitats (channel, gravel, islands, riparian forest and grassland) of the Urbach River, Switzerland, and actual respiration rate (R) of the same samples exposed to experimental inundation were measured. Measurements were carried out at three incubation temperatures (4°C, 12°C and 20°C), and ETSA/R ratios (i.e. exploitation of the overall metabolic capacity) were investigated to better understand the effects of temperature and inundation on floodplain functional heterogeneity. Furthermore, ETSA/R ratios obtained during experimental inundation were compared with ETSA/R ratios from field measurements to investigate the exploitation in total metabolic potential at different conditions. Lowest ETSA and R were measured in samples from channel and gravel habitats, followed by those from islands. Substantially higher values were measured in soils from riparian forest and grassland. Both ETSA and R increased with increasing temperature in samples from all habitats, while the ETSA/R ratio decreased because of a rapid response in microbial community respiration to higher temperatures. The metabolic capacity exploitation (i.e. ETSA/R) during experimental inundation was lowest in predominantly terrestrial samples (riparian forest and grassland), indicating the weakest response to wetted conditions. Comparison of experimentally inundated and field conditions revealed that in rarely flooded soils, the metabolic capacity was less exploited during inundation than during non‐flooded conditions. The results suggest high sensitivity in floodplain respiration to changes in temperature and hydrological regime. ETSA/R ratios are considered good indicators of changes in metabolic activity of floodplain soils and sediments, and thus useful to estimate the impact of changes in hydrological regime or to evaluate success of floodplain restoration actions. Copyright © 2015 John Wiley & Sons, Ltd.     

Modeling Urban Land Use Changes Using Support Vector Machines

Support Vector Machines (SVM) is a machine learning (ML) algorithm commonly applied to the classification of remotely sensing data and more recently for modeling land use changes. However, in most geospatial applications the current literature does not elaborate on specifications of the SVM method with respect to data sampling, attribute selection and optimal parameters choices. Therefore the main objective of this study is to present and investigate the SVM technique for modeling urban land use change. The SVM model building procedure is presented together with the detailed evaluation of the output results with respect to the choice of datasets, attributes and the change of SVM parameters. Geospatial datasets containing nine land use classes and spatial attributes for the Municipality of Zemun, Republic of Serbia were used for years 2001, 2003, 2007 and 2011. The Correlation‐based Feature Subset method, kappa coefficient, Area Under Receiver Operating Characteristic Curve (AUC) and kappa simulation were used to perform the model evaluation and compare the model outputs with the real land use datasets. The obtained results indicate that the SVM‐based models perform better when implementing balanced data sampling, reduced data sets to informative subsets of attributes and properly identify the optimal learning parameters.