Understanding the ecology of boreal birds through automated soundscape analysis

点击翻译按钮获取中文摘要

Sound is a ubiquitous & important component of the biotic environment. Ecologists use sound to track the state of wildlife by counting vocalizations made by species. However, people cannot be everywhere all the time, so many questions about the timing & spacing of ecological processes, particularly at large spatial scales & over long time-frames are unanswerable using human observers. We will test several key ecological theories  like ideal free vs. despotic selection, abundant centre hypothesis, range-phenology shifts & landscapes of fear using data from the sounds made by boreal forest songbirds. Testing all of these theories requires accurate density, space use & timing information that is far more precise than human observers typically collect. Our technological advance is the use of autonomous recording units (sound recording devices) to do this. We will develop deep learning tools to improve measures of density, seasonality, return rates, occupancy dynamics & breeding state through sound. We are also changing how data is collected, collated & most importantly shared between researchers through innovative online tools. This allows us to provide the information needed by decision makers about pressing large-scale conservation issues. Habitat selection is one of the most studied processes in ecology and is central to applied conservation and theoretical biology. However, current approaches used by most researchers tend to emphasize how local scale patterns of birds vary based on human footprint & vegetation. Our work advances this paradigm by studying variation in habitat selection & the interactions between species that may alter patterns of habitat selection at large scales & over time that are only possible through ARUs. HQP are central to my research program. In the past 6 years I have provided opportunities to over 200 HQP. I foresee similar numbers of students being trained in the next cycle. The success of my program is dependent on developing a strongly-coordinated system of collaborative student projects that individually are unique but when integrated provide greater opportunities to reveal fundamental insights into behavioral, population, community & landscape ecology. As evidenced by my publication record and the employability of past students, this approach creates well-rounded individuals with competencies in both applied and theoretical ecology. The DG program provides critical funds that allow my students to expand applied work to test specific hypotheses about ecological theory and the processes that drive patterns of avian diversity at different spatial scales.