Modeling population-environment dynamics in the Ecuadorian Amazon

描述（由申请人提供）：该项目的主要目标是研究人口与环境动态之间的一些复杂的多维度关系，在热带雨林的背景下，由于农业前沿的推进，正在发生重大变化。更好地了解这些关系对于森林保护和可持续发展至关重要，特别是在北方厄瓜多尔亚马逊（NEA），这是一个具有非凡生物多样性的地区，其森林砍伐速度在世界上排名第九。虽然造成毁林的直接原因各不相同，但已确定农村移徙农民是主要的直接动因。我们建议在NEA进行研究的基础上，在1990年和1999年收集的面板数据从地块，构建一个层次框架的地块和社区，使用多级统计模型来检验假设。本文提出了三个相互关联的目标来研究人口、社会经济、制度和生态因素：（i）确定与家庭住户空间分布相关的因素;（ii）确定与家庭生育率变化相关的因素; ㈢将土地使用和土地覆盖随时间的变化分解为与人口增长、地点/与社区/市场的接近程度有关的因素，农业建立、自然资源和财富的年份。空间明确的模型将开发基于人口-环境理论的综合方法，重点是变化的近因和根本原因，结合人口统计学，Chayanovian农业家庭模型，Von Thunen经济土地利用模型以及利用制度和背景变量的最新方法。数据来自于1990年至2000年收集的地理编码纵向家庭和社区数据，并通过1972年至今的遥感图像时间序列加以补充。沿着来自家庭和社区的纵向地理编码数据的可用性，以及最近多层次模型估计的改进，为评估许多因果因素的相关性提供了一个独特的机会。该项目将有助于增进对人口与环境关系的了解，推动我们完善热带地区变化的理论，并为有兴趣实现森林保护和可持续发展的决策者提供重要信息。

DESCRIPTION (provided by applicant): The primary objective for this project is to examine some of the complex multidimensional relationships between population and environment dynamics in a tropical rainforest context that is undergoing substantial change due to the advance of the agricultural frontier. A better understanding of these relationships is vital for forest conservation and sustainable development, particularly in the Northern Ecuadorian Amazon (NEA), an area of extraordinary biodiversity that has the 9th fastest deforestation rate in the world. While the proximate causes of deforestation vary, rural-rural migrant farmers have been identified as the primary direct agents. We propose to undertake research in the NEA based upon panel data collected in 1990 and 1999 from land plots, constructing a hierarchical framework of land plots and communities to test hypotheses using multilevel statistical models. Three interrelated Aims are proposed to examine demographic, socioeconomic, institutional, and ecological factors: (i) Determine factors associated with the spatial distribution of family households overtime; (ii) Determine factors associated with changing household fertility overtime; (iii) Decompose changes in land use and land cover over time into factors related to population growth, location / proximity to communities / markets, year of farm establishment, natural resources, and wealth. Spatially-explicit models will be developed based upon an integrated approach to population- environment theory that focuses on proximate and underlying causes of change, combining theories from demography, Chayanovian agricultural household models, Von Thunen economic land use models, and recent approaches utilizing institutional and contextual variables. Data are derived from geo-coded longitudinal household and community data collected between 1990 and 2000, augmented by a time series of remotely sensed images from 1972 to the present. The availability of longitudinal geo-coded data from households and communities along with recent improvements in multilevel model estimation provide a unique opportunity to assess the relevance of many causal factors. This project will help advance knowledge of population-environment relationships, moving us towards refining theories of change in the tropics, and proffering vital information to policy-makers interested in achieving forest conservation and sustainable development.