Éditeur(s) : HAL CCSD Royal Meteorological Society
Résumé : Assessing the impact/adaptation of human activities on/to climate change is a key issue, especially in the tropics that concentrate major anthropogenic dynamics such as deforestation and nearly two-thirds of the planetary rainfall. However, this task is often made tough because human activities such as agricultural dynamics are usually analysed at local or regional scale whereas climate related studies are led at large to global scales due to a lack of reliable data, especially in the tropics. In this article we argue that the increased spatial resolution of remote sensing-based rainfall estimates enables assessing the spatiotemporal variability of rainfall regimes at regional and local scales, thus allowing fine analysis of the interactions with human activities. We processed Tropical Rainfall Measuring Mission (TRMM) 3B42 daily rainfall estimates over the state of Mato Grosso (southern Brazilian Amazon) for the 1998-2012 study period in order to compute rainfall metrics such as annual rainfall and duration, onset and end dates of the rainy season based on the Anomalous Accumulation methodology (at a 0.25◦ spatial resolution). We then crossed these metrics with agricultural maps (produced at a 250m spatial resolution) and proved that the adoption of intensive agricultural practices such as double cropping systems is partly the result of a strategy to adapt practices to local climatic conditions. Finally, we discuss how such results raise important issues regarding the sustainability of the agricultural development model in the Southern Amazon.
International Journal of Climatology

halshs-00910825
https://halshs.archives-ouvertes.fr/halshs-00910825
DOI : 10.1002/joc.3863

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
Herbicides are generally the most extensively used of the pesticides applied to agricultural crops. However, the literature contains little evidence useful in assessing the potential sources of the general population's exposure to herbicides, including by residential proximity to crops. The objective of this study was to take advantage of data from the PELAGIE mother-child cohort to identify the main determinants of the body burden of exposure to the chloroacetanilide and triazine herbicides commonly used on corn crops in Brittany, France, before 2006. Urine samples from a randomly selected subcohort of women in the first trimester of pregnancy (n=579) were assayed for herbicide metabolites. The residential exposure resulting from proximity to corn crops was assessed with satellite-image-based scores combined with meteorological data. Data on diet, drinking tap water (from the public water supply), occupations, and household herbicide use were collected by questionnaires. Herbicides were quantified in 5.3% to 39.7% of urine samples. Alachlor and acetochlor were found most frequently in the urine of women living in rural areas. The presence of dealkylated triazine metabolites in urine samples was positively associated with residential proximity to corn crops (OR=1.38, 95% CI: 1.05-1.80). Urinary metabolites of both atrazine and dealkylated triazine were correlated with tap water consumption (OR=2.94, 1.09-7.90, and OR=1.82, 1.10-3.03, respectively); hydroxylated triazine metabolites were correlated with fish intake (OR=1.48, 1.09-1.99). This study reinforces previous results that suggest that environmental contamination resulting from agricultural activities may contribute to the general population's exposure to herbicides.
Environment International

hal-00966842
https://hal.archives-ouvertes.fr/hal-00966842
https://hal.archives-ouvertes.fr/hal-00966842/document
DOI : 10.1016/j.envint.2013.10.010
PUBMED : 24246238

Éditeur(s) : HAL CCSD
Résumé : International audience
GEOBIA 2012
Rio de Janeiro, France

hal-01107866
https://hal.archives-ouvertes.fr/hal-01107866