• Rural Digital Europe
• Publications close
• Research data close
• Research software close
• French National Research Agency (AN... close
• HAL-Pasteur close
• HAL Descartes close

The limited availability of soil information has been recognized as a main limiting factor in digital soil mapping (DSM) studies. It is therefore important to optimize the joint use of the three sources of soil data that can be used as inputs of DSM models, namely spatial sets of measured sites, soil maps and soil sensing products.In this paper, we propose to combine these three inputs, through a cokriging with a categorical external drift (CKCED). This new interpolation technique was applied for mapping seven soil properties over a 24.6 km2 area located in the vineyard plain of Languedoc (Southern France), using an hyperspectral imagery product as example of a soil sensing data. Cross-validation results of CKCED were compared with those of five spatial and non-spatial techniques using one of these inputs or a combination of two of them.The results obtained in the La Peyne Catchment showed i) the utility of soil map and hyperspectral imagery products as auxiliary data for improving soil property predictions ii) the greater added-value of the latter against the former in most situations and iii) the feasibility and the interest of CKCED in a limited number of soil properties and data configurations. Testing CKCED in case study with soil maps of better quality and soil sensing techniques covering more area and depths should be necessary to better evaluate the benefits of this new technique. International audience

Efficient management of agricultural management should consider multiple services and stakeholders. Yet, it remains unclear how to guarantee ecosystem services for multiple stakeholders' demands, especially considering the observed biodiversity decline following reductions in semi-natural habitat (SNH), and global change. Here, we use an ecosystem service model of intensively-managed agricultural landscapes to derive the best landscape compositions for different stakeholders' demands, and how they vary with stochasticity and the degree of pollination dependence of crops. We analyse three groups of stakeholders assumed to value different ecosystem services most-individual farmers (crop yield per area), agricultural unions (landscape production) and con-servationists (biodiversity). Additionally, we consider a social average scenario that aims at maximizing mul-tifunctionality. Trade-offs among stakeholders' demands strongly depend on the degree of pollination dependence of crops, the strength of environmental and demographic stochasticity, and the relative amount of an ecosystem service demanded by each stakeholder. Intermediate amounts of SNH deliver relatively high levels of the three services (social average). Our analysis further suggests that the current levels of SNH protection lie below these intermediate amounts of SNH in intensively-managed agricultural landscapes. Given the worldwide trends in agriculture and global change, current policies should start to consider factors such as crop type and stochasticity, as they can strongly influence best landscape compositions for different stakeholders. Our results suggest ways of managing landscapes to reconcile several actors' demands and ensure for biodiversity conservation and food production. International audience

Highlights • We emphasize the end of the boom-and-bust development pattern in the Brazilian Amazon. • Deforestation and socioeconomic development follow an inverted U-shape relationship. • Household incomes are greater in stabilized areas than in areas undergoing deforestation. • Environmental governance efforts fostered the emergence of EKC. Abstract Socioeconomic development in the Brazilian Amazon is currently reaching national averages although deforestation activity has been declining for a decade. As a consequence, recent studies rejected the widely agreed boom-and-bust development hypothesis that deforestation first generates an economic boom, which is then followed by a collapse as forest resources are depleted. Here, we confirm these studies that there is no boom-bust cycle and suggest that a new pattern of relationship between deforestation and socioeconomic development has emerged following an environmental Kuznets curve (EKC). In this scenario, environmental degradation increases in the early stages of economic development and decreases in later stages as the economy develops and wellbeing increases. To validate this assumption, we conducted the first sub-municipal analysis of socioeconomic development and deforestation in the Brazilian Amazon for the 2000–2010 period. Our results confirm the emergence of an EKC relationship with a turning point beyond which socioeconomic growth does not appear anymore to be a driver of deforestation. We also emphasize that areas subjected to active deforestation in 2010 present lower socio-Author manuscript version economic indicators than stabilized areas, pointing to the precarious socioeconomic situation of areas still undergoing active deforestation. We put these results in perspective by considering Brazilian efforts to ensure a transition in environmental governance with the objective of promoting land use sustainability through control of deforestation at the same time as supporting socioeconomic development. International audience

Circum-Galactic Medium surrounding galaxies has been punctually detected, but its morphology remains largely unknown. The Multi-Unit Spectroscopic Explorer (MUSE) spectro-imager provides for the first time both spectral and spatial resolution to spatially map such features. The problem lies in the statistical detection of faint spatially-extended sources in massive hyperspectral images such as provided by MUSE, and has not been previously handled. This paper presents a statistical detection method based on hypothesis testing tackling this problem. The proposed strategy is step-by-step validated over alternative ways with simulations. Then, results on MUSE observations are presented. International audience

AbstractDouglas fir trees presumable stimulate nitrification in the soil. We studied in 21 French Douglas fir forests if and how nitrification is modulated by soil properties, past land use and current forest management. Soil (0–10 cm depth) was collected and initial concentrations of N-NH4+ and N-NO3−, potential net nitrogen mineralization (PNM) and net nitrification (PNN) rates and microbial biomass were measured. At 11 of the 21 sites, annual nitrate fluxes in the soil were measured using anion exchange resin bags. Soils contained between 2.3 to 29.4 mg N-NO3− kg soil−1. About 86% (±14%) of mineral N was nitrate. The proportion of nitrate increased to almost 100% during incubation. PNN varied from 0.10 mg N kg soil−1 day−1 to 1.05 mg N kg soil−1 day−1 (21 sites). Neither the initial nitrate concentration nor PNN was related to soil chemistry (pH, % C, %N, P, CEC), microbial biomass, texture, past land use or thinning. In situ net nitrate accumulation (NNA) estimated with resins beds varied from 4 to 100 kg N-NO3− ha−1 yr−1 (11 sites). It was positively correlated with base saturation, clay content, ELLENBERG N, temperature and negatively with soil organic N, C/N ratio and precipitation.

Despite many existing approaches, modeling karst water resources remains challenging as conventional approaches usually heavily rely on distinct system knowledge. Artificial neural networks (ANNs), however, require only little prior knowledge to automatically establish an input-output relationship. For ANN modeling in karst, the temporal and spatial data availability is often an important constraint, as usually no or few climate stations are located within or near karst spring catchments. Hence, spatial coverage is often not satisfactory and can result in substantial uncertainties about the true conditions in the catchment, leading to lower model performance. To overcome these problems, we apply convolutional neural networks (CNNs) to simulate karst spring discharge and to directly learn from spatially distributed climate input data (combined 2D-1D CNNs). We investigate three karst spring catchments in the Alpine and Mediterranean region with different meteorological-hydrological characteristics and hydrodynamic system properties. We compare the proposed approach both to existing modeling studies in these regions and to our own 1D CNN models that are conventionally trained with climate station input data. Our results show that all the models are excellently suited to modeling karst spring discharge (NSE: 0.73-0.87, KGE: 0.63-0.86) and can compete with the simulation results of existing approaches in the respective areas. The 2D models show a better fit than the 1D models in two of three cases and automatically learn to focus on the relevant areas of the input domain. By performing a spatial input sensitivity analysis, we can further show their usefulness in localizing the position of karst catchments. International audience