报告时间：2015年5月13日（周三）3:30 p.m.
报告地点：五号楼附3—202报告厅
报 告 人：Nicolas Parisey, Ph.D.,
法国农业研究所（French National Institute for
agricultural Research）高级研究员.
|
 |
报告摘要：
More and more agricultural practitioners and decision makers are interested in alternative crop protection strategies, such as biological pest control to replace pesticide. However, how to use biological pest control for crop protection, and at the same time, preserve yield production, is quite an important problematic. For pest management, developing these alternative crop protection strategies requires an extensive knowledge on pest colonisation process and their reproduction strategies. Current research in mathematical biology and remote sensing offers new leads for pest management. In the former one, we can apply state-space spatio-temporal models to identify pest invasive mechanism by using information like geo-referenced time-series of pest characteristics (count data, relative occurrence rate, etc.) and associated photo-interpreted argonomical maps. In the latter one, recent publications show a great interest in mixing optical and radar data to directly infer the pest abundance or/and richness [3]. Therefore, similar to the way how data assimilation methods are used in meteorology, we intend to construct quantitative ecology models for the direct inference and selection of most likely pest invasive mechanism at landscape scale based on remote sensors and sparse abundance and/or genetic data of pest.
