Dissertation Research: Disentangling drivers of community structure and composition of a tropical forest

该项目旨在解开和描述世界上最大的再生热带森林之一的动植物群落结构和组成的驱动因素。确定为什么物种能够在特定的时间和地点存在（特别是偶然与不可避免的角色）是生态学中最核心的任务之一，也是在快速变化的世界中保护物种的一个非常紧迫的问题。考虑到(1)热带森林令人难以置信的生物多样性，(2)被分类为再生的剩余森林的比例很高并且不断上升，(3)这些知识是将生态恢复作为保护受威胁物种和生态系统的有效工具进行部署的先决条件，了解热带森林再生的这些力量尤为及时。项目团队将评估群落组成和结构的三个最终驱动因素：土壤质量、景观组成和初始植被条件。利用植被样地网络，将评估陆生哺乳动物、蝙蝠、鸟类和鸣禽-昆虫组合的物种组成。这项任务的基础是开发一套工具，以便对这些动物群体进行快速、廉价和非侵入性的评估。为此，将部署超声波和音频记录器，研究团队正在开发技术，允许使用机器学习对录音进行半自动分析。为了进一步评估植被结构和物种相互作用作为组成的近似驱动因素，将使用同步记录仪阵列来确定关键的森林结构特征和某些物种的分布如何影响其他物种的存在。

This project seeks to disentangle and characterize drivers of plant and animal community structure and composition in one of the world's largest regenerating tropical forests. Determining why species are able to exist in a particular time and place (particularly the role of chance vs. inevitability) is simultaneously one of the most central tasks in ecology and a deeply pressing question for conservation in a rapidly changing world. Understanding these forces in regenerating tropical forest is a particularly timely given (1) the incredible biodiversity of tropical forests, (2) that the proportion of remaining forest classified as regenerating is high and ever-rising, and (3) such knowledge is a prerequisite for deploying ecological restoration as an effective tool for preserving threatened species and ecosystems.The project team will assess three ultimate drivers of community composition and structure: soil quality, landscape composition, and initial vegetative conditions. Using a network of vegetative plots, species composition of terrestrial mammal, bat, bird, and singing-insect assemblages will be assessed. Fundamental to this task is developing a set of tools that will allow for the rapid, cheap, and non-invasive assessment of these animal groups. To this end, ultrasonic and audible-frequency recorders will be deployed, and the research team is developing techniques that will allow for the semi-automated analysis of recordings using machine learning. To further assess vegetative structure and species interactions as proximate drivers of composition, arrays of synced recorders will be used to determine how key forest structural features and the distribution of some species affects the presence of others.