Despite their high species diversity and endemism, island ecosystems are extremely vulnerable to invasive plant species (IPS). Reunion Island, located in the Indian Ocean, represents the largest area of intact vegetation in the Mascarene archipelago. However, the flora of Reunion Island is mostly threatened by IPS. Ecological modelling is crucial for studying IPS dynamics and for guiding management decisions on islands. To improve the capacity of IPS management in Reunion Island, this study aimed: (1) to predict the future spatio-temporal dynamics of IPS at island scale, (2) to better understand IPS spread factors. This work was conducted through an active “research-action” partnership. We designed a hybrid model to simulate the spatial spread over time of IPS on Reunion Island. We selected three species to be modelled: Anthoxanthum odoratum, Hiptage benghalensis and Solanum mauritianum. The hybrid model predicted an increase of 30 % of the current invaded surface by A. odoratum in Key Biodiversity Areas (KBAs); an increase of 20 % for S. mauritianum; and a spread outside of KBAs for H. benghalensis. The framework of the model, based on generic life-cycle mechanisms and species distribution models, can be adapted to other IPS of different plant types (such as lianas, shrubs and trees). The model framework is also adaptable and can be used with different climate change scenarios. With the implementation of IPS control actions also planned, the model could be used as a support tool by managers for analysing IPS trends spreading into KBAs and for taking appropriate actions.

Reunion Island, a French overseas region in the Indian Ocean, has endorsed policies targeting food and electricity self-sufficiency. This objective implies balancing different land-uses (food, feed, bioelectricity, urbanisation, etc.) which we explore in a set of scenarios towards 2035. Through participatory structural analysis, we modelled drivers of change as processes using Ocelet, a spatially explicit and dynamic modelling platform. We built a detailed land-use map for our initial state and calibrated relevant processes through four scenarios ranging from "business-as-usual" to "implementation of ambitious territory planning policies". To improve local self-sufficiency, our results support the need for large-scale land planning policies, suggesting partial sugarcane conversion into food crops, urbanisation control, farmlands expansion onto fallows and photovoltaic increase. Our context-specific approach addresses food and electricity self-sufficiency as a whole and understands its inner dynamic and spatial processes from stakeholders' viewpoint. Moreover, our model recognizes small-scale spatial heterogeneity and contributes to mediate controversial issues related to territory foresight and land-use planning.

Increasing nutrient circularity and use efficiency is a leading topic in the search for more sustainable agri-food-waste systems (AFWS). This paper proposes a method to assess the role of crops and livestock in nutrient circularity and use efficiency in an AFWS. The method is based on the analysis of nutrient flows, a detailed typology of flows, and a set of 3 groups of indicators to characterize (i) circularity between sub-systems, (ii) the process efficiency of the sub-systems and (iii) the efficiency of the AFWS. The method is illustrated using the nitrogen metabolism of the AFWS of a tropical Island, French Reunion Island. The island's current nitrogen use efficiency is very low (0.7%). Crops and livestock are major sources of inefficiencies due to their processes, they account for respectively, 42% and 9% of total AFWS inefficiency. However, crops and livestock are involved in circularity, as they play, respectively, a recycling receiver and recycling supplier role. Among the internal recycling routes between all sub-systems, 41% go to crops and 31% come from livestock. The paper argues that circularity and process efficiency are not objectives per se but means to achieve AFWS efficiency, and that the distinction between these three elements enable a systemic multi-level understanding of the roles of crops and livestock.

The availability of crop type reference datasets for satellite image classification is very limited for complex agricultural systems as observed in developing and emerging countries. Indeed, agricultural land use is very dynamic, agricultural censuses are often poorly georeferenced and crop types are difficult to interpret directly from satellite imagery. In this paper, we present a database made of 24 datasets collected in a standardized manner over nine sites within the framework of the international JECAM (Joint Experiment for Crop Assessment and Monitoring) initiative; the sites were spread over seven countries of the tropical belt, and the number of data collection years depended on the site (from 1 to 7 years between 2013 and 2020). These quality-controlled datasets are distinguished by in situ data collected at the field scale by local experts, with precise geographic coordinates, and following a common protocol. Altogether, the datasets completed 27 074 polygons (20 257 crops and 6817 noncrops, ranging from 748 plots in 2013 (one site visited) to 5515 in 2015 (six sites visited)) documented by detailed keywords. These datasets can be used to produce and validate agricultural land use maps in the tropics. They can also be used to assess the performances and robustness of classification methods of cropland and crop types/practices in a large range of tropical farming systems. The dataset is available at https://doi.org/10.18167/DVN1/P7OLAP (Jolivot et al., 2021).

The ALBOMAURICE software runs a mosquito population dynamics model to predict the temporal and spatial abundance of Aedes albopictus, the dengue disease vector in Mauritius. For each vector surveillance zone, it solves a system of ordinary differential equations describing different stages of the mosquito life cycle. ALBOMAURICE uses daily rainfall and temperature input data to produce abundance maps used operationally by health services for targeting areas where to apply vector surveillance and control measures. Model simulations were validated against entomological data acquired weekly during a year at nine locations. Different control options can also be simulated and their effects compared.

Nowadays, modern Earth Observation systems continuously generate huge amounts of data. A notable example is the Sentinel-2 Earth Observation mission, developed by the European Space Agency as part of the Copernicus Programme, which supplies images from the whole planet at high spatial resolution (up to 10 m) with unprecedented revisit time (every 5 days at the equator). In this data-rich scenario, the remote sensing community is showing a growing interest toward modern supervised machine learning techniques (e.g., deep learning) to perform information extraction, often underestimating the need for reference data that this framework implies. Conversely, few attention is being devoted to the use of network analysis techniques, which can provide a set of powerful tools for unsupervised information discovery, subject to the definition of a suitable strategy to build a network-like representation of image data. The aim of this work is to provide clues on how Satellite Image Time Series can be profitably represented using complex network models, by proposing a methodology to build a multilayer network from such data. This is the first work to explore the possibility to exploit this model in the remote sensing domain. An example of community detection over the provided network in a real-case scenario for the mapping of complex land use systems is also presented, to assess the potential of this approach.

Remote sensing data, crop modelling, and statistical methods are combined in an original method to overcome current limitations of crop yield estimation. It is then tested for timely estimation of maize grain yields and their year-on-year variability in Burkina Faso. Outputs from the SARRA-O crop model were used as a proxy for observed data for calibration. The final remote sensing-based yield model was constructed on the interaction between aboveground biomass at flowering (AGB-F) and crop water stress (Cstr) over the reproductive and maturation phases. Various vegetation and drought-related indices were derived from different spectral domains and tested. Model performance was evaluated by cross-validation against (a) simulated yields and (b) independent yields from ground surveys aggregated at village level. The results showed that the RF (Random Forest) model outperformed the MLR (Multiple Linear Regression) model for year-on-year yield estimation at the end of the season when compared to simulated yields. Surface soil moisture (SSM) information, as a proxy for soil water available for plant growth, together with information on the temperature of the canopy cover, helped to improve the RF maize yield model, impacting more particularly the estimation of crop water stress. Lastly, two months before harvest the RF model predicted 46% of the observed end-of-season maize grain yield variability. The combined remote sensing, crop model and machine learning method is thus an effective approach for estimating and forecasting inter-annual maize crop yields in environments where field data are scarce, such as in most parts of the African continent. However, more research is needed to better retrieve the spatial variability of yields in order to strengthen current agricultural monitoring systems, and to address societal challenges, such as declining food security.

Cropping systems’ maps at fine scale over large areas provide key information for further agricultural production and environmental impact assessments, and thus represent a valuable tool for effective land-use planning. There is, therefore, a growing interest in mapping cropping systems in an operational manner over large areas, and remote sensing approaches based on vegetation index time series analysis have proven to be an efficient tool. However, supervised pixel-based approaches are commonly adopted, requiring resource consuming field campaigns to gather training data. In this paper, we present a new object-based unsupervised classification approach tested on an annual MODIS 16-day composite Normalized Difference Vegetation Index time series and a Landsat 8 mosaic of the State of Tocantins, Brazil, for the 2014–2015 growing season. Two variants of the approach are compared: an hyperclustering approach, and a landscape-clustering approach involving a previous stratification of the study area into landscape units on which the clustering is then performed. The main cropping systems of Tocantins, characterized by the crop types and cropping patterns, were efficiently mapped with the landscape-clustering approach. Results show that stratification prior to clustering significantly improves the classification accuracies for underrepresented and sparsely distributed cropping systems. This study illustrates the potential of unsupervised classification for large area cropping systems’ mapping and contributes to the development of generic tools for supporting large-scale agricultural monitoring across regions.

In a context of fast land changes due to human activity, the old "Malthus vs. Boserup" debate about human pressure on natural resources is more than ever a prevalent subject. We illustrate this debate with an example based on the Tuy Province in West Burkina Faso, which has known an important development these last fifteen years, as mainly observable in the major regression of its natural vegetation. The objective of this article is to question the possible future scenarios of this region, by using a spatial model to understand past land change mechanisms and prospect plausible future ways of development. In particular, the spatial model describes the vegetation clearance processes identified during field campaigns, which we used to draw prospective scenarios and assess their possible effect on natural vegetation evolution. The processing of remote sensing images helped us reveal that one quarter of the total study area was cleared during the last fourteen years. Surveys carried out in the field enabled the identification of the three main processes responsible for these changes: farm size expansion, creation of new farms due to family nuclearization and migrant settlement. The model was then built to reproduce these three processes, and was validated by comparison with the land use classification of remote sensing images. Our model was also used to explore past clearance mechanisms: we found that 90% of the clearance was shared equally between farm size expansion and nuclearization processes, the settlement of migrants being responsible for less than 10% of the clearance. Model outputs also showed a shift in the clearance schemes compared to ancient practices: land characteristics are no longer considered when land is cleared and now clearance also occurs in neighbouring villages with available lands. These results suggest that the region has finally evolved similarly to a Malthusian rationale, even if the past dynamics resulted from a complex combination of factors. Finally, we analysed several prospective scenarios to assess the impacts of i) different population evolutions (normal demographic growth, emigrations and demographic regulation), ii) the implementation of protected areas in each village, iii) an intensification of fanning systems. Two possible solutions for reducing natural vegetation clearance in the region are discussed: Emigration to other regions or a demographic regulation accompanied with an intensification of agricultural systems. However, the question remains whether such changes can be accomplished rapidly enough to abate the pressing natural vegetation decrease threat and to maintain an acceptable livelihood in the region.

In the Sahel, crop growth and yield are strongly linked to climate fluctuations. The low and erratic rainfall the Sahel region has experienced for several years led to poor harvests, associated with dramatic food crises and famines. Consequently, numerous studies were conducted to develop innovative techniques to estimate crop yield based on satellite measurements. Unlike most approaches which use rainfall, temperature or vegetation indices to derive crop yield estimates, the present study investigates the potential of satellite-derived soil moisture products. This study focuses on millet, a major food crop in Africa. A first step was devoted to analyzing the relation between soil moisture and millet yield at the local scale using ground-based soil moisture and millet yield measurements obtained at ten site locations in Niger. Then, the statistical relationship obtained at the local scale was assessed at the regional scale (Niger, Mali, Senegal and Burkina Faso) using satellite-based soil moisture mapping (based on a simple land-surface model and a satellite precipitation product) and compared to millet yield estimates from the Food and Agriculture Organization (FAO) database. It was shown that millet yield variations are closely linked to soil moisture variations during two key periods of the plant growth: the "grain filling" and the "reproductive" periods. Soil moisture variations during these two periods led to explain 81% (R-2 = 0.81) of the FAO millet yield variations from 1998 to 2014 in the Sahel.

A wide range of environmental and societal issues such as food security policy implementation requires accurate information on biomass productivity and its underlying drivers at both regional and local scales. While many studies in West Africa are conducted with coarse resolution earth observation data, few have tried to relate vegetation trends to explanatory factors, as is generally done in land use and land cover change (LULCC) studies at finer scales. In this study we proposed to make a bridge between vegetation trend analysis and LULCC studies to improve the understanding of the various factors that influence the biomass production changes observed in satellite time series (using integrated Normalized Difference Vegetation Index [NDVI] as a proxy). The study was conducted in two steps. In the first step we analyzed MODIS NDVI linear trends together with TRMM growing season rainfall over the Sahel region from 2000 to 2015. A classification scheme was proposed that enables better specification of the relative role of the main drivers of biomass production dynamics. We found that 16% of the Sahel is re-greening—but found strong evidence that rainfall is not the only important driver of biomass increase. Moreover, a decrease found in 5% of the Sahel can be chiefly attributed to factors other than rainfall (88%). In the second step, we focused on the “Degré Carré de Niamey” site in Niger. Here, the observed biomass trends were analyzed in relation to land cover changes and a set of potential drivers of LULCC using the Random Forest algorithm. We observed negative trends (29% of the Niger site area) mainly in tiger bush areas located on lateritic plateaus, which are particularly prone to pressures from overgrazing and overlogging. The significant role of accessibility factors in biomass production trends was also highlighted. Our methodological framework may be used to highlight changing areas and their major drivers to identify target areas for more detailed studies. Finer-scale assessments of the long-term vulnerability of populations can then be made to substantiate food security management policies.

In response to the need for generic remote sensing tools to support large-scale agricultural monitoring, we present a new approach for regional-scale mapping of agricultural land-use systems (ALUS) based on object-based Normalized Difference Vegetation Index (NDVI) time series analysis. The approach consists of two main steps. First, to obtain relatively homogeneous land units in terms of phenological patterns, a principal component analysis (PCA) is applied to an annual MODIS NDVI time series, and an automatic segmentation is performed on the resulting high-order principal component images. Second, the resulting land units are classified into the crop agriculture domain or the livestock domain based on their land-cover characteristics. The crop agriculture domain land units are further classified into different cropping systems based on the correspondence of their NDVI temporal profiles with the phenological patterns associated with the cropping systems of the study area. A map of the main ALUS of the Brazilian state of Tocantins was produced for the 2013–2014 growing season with the new approach, and a significant coherence was observed between the spatial distribution of the cropping systems in the final ALUS map and in a reference map extracted from the official agricultural statistics of the Brazilian Institute of Geography and Statistics (IBGE). This study shows the potential of remote sensing techniques to provide valuable baseline spatial information for supporting agricultural monitoring and for large-scale land-use systems analysis.

Un dispositif partenarial mobilisant un outil de simulation paysagère, Ocelet, a été conçu comme support d’un exercice de prospective territoriale sur une intercommunalité de l’île de La Réunion où le mitage agricole représente un enjeu important. La démarche consiste à coconstruire, au sein d’une arène d’une vingtaine d’institutions, des scénarios d’évolution de l’étalement urbain. Pour favoriser ces interactions, des modèles contextualisés sont développés. Dans la démarche, le modèle remplit différentes fonctions au sein de l’arène, de l’initiation à la mise en débat de controverses. Les acteurs du projet s’opposent sur les facteurs concourant à l’étalement urbain et sur les moyens de le réguler. Si ces positions évoluent peu durant le projet, en revanche certaines visions ou certains partis pris ont pu être révisés. La démarche collaborative et le travail de prospective territoriale sont jugés de manière très positive par une large majorité de participants, pour leur rôle de moteurs d’apprentissage collectif.

In response to the need of large scale spatial information for supporting agricultural monitoring, we present a new remote sensing object-based approach for objective and repeatable agricultural systems mapping at regional level. This approach is in two steps: 1. A segmentation of land units, based only on a remote sensing time series; These land units are assumed to be representative of the human activity and environmental conditions and thus of the in situ agro-ecosystems; 2. A semi-automatic land use classification, performed in each land unit with high-resolution images, to label the land units in terms of agricultural systems. To produce the land units, a principal component transformation was first applied to an annual dataset of MODIS (MODerate Imaging Spectroradiometer) normalized difference vegetation index (NDVI) images. A series of segmentations were then performed on the principal component images that contain the essential information on the physiognomy and phenology of the cover. An unsupervised evaluation method was used for identifying the optimum segmentation which successfully delineates homogeneous units in terms of agricultural activity, discriminating the different cropland and grassland areas at regional level. Then, for each land unit, cropping systems maps were produced at a field level through object-based analysis of a Landsat8 30 m resolution mosaic image, and spectral variables derived from the MODIS NDVI time series, and were validated with in situ data. Finally, a bottom-up spatial analysis of the extracted land use information at field level allowed definitive classification and characterization of the homogeneous regional level land units in terms of agricultural systems. A map of the main agricultural systems of the Brazilian state of Tocantins, an agricultural expansion region, has been successfully produced for the year 2015 following this approach. This study shows the potential of multi-resolution satellite images to provide valuable baseline spatial information for supporting agricultural monitoring. (Resume d'auteur)