"Development of time-lapse and drone imagery for population monitoring of North Atlantic seabirds"

我的研究将使用无人机和延时图像，由研究人员、公民科学家和机器学习相结合来分类，以研究古北区受到威胁的海鸟物种。我将使用现有的延时相机网络来聚焦于一种悬崖筑巢物种--黑腿猫尾鸟Rissa tridactyla，目的是了解猫尾鸟可能会如何受到环境变化的影响。我还将部署相机研究一种洞穴筑巢的物种--北极大西洋海雀，以尝试和改进测量洞穴筑巢海鸟生产力的方法。通过开发可重现的技术来从照片中提取数据，我希望提供一种在足够的空间尺度上持续长期监测的方法，以了解海鸟衰退的情况。1)比较公民科学、研究人员和机器学习在时间推移图像中识别猫尾鸟的准确性。2)了解非生物因素(例如纬度、群落大小、年份)和生物因素(例如天气条件、捕食、到渔业的距离、人类干扰)如何影响猫尾鸟的物候，3)开发利用延时图像测量海雀生产力的方法，并与根据“传统”野外观测计算的生产力进行比较。4)审查无人驾驶飞行器(UAV)用于海鸟普查的使用情况，并制定利用无人机调查海鸟的“最佳做法”指南。5)开发有效的方法来从无人机图像中计算地面和悬崖筑巢的海鸟群体，并利用这些数据测试与种群趋势有关的群体结构的空间影响。

My research will use UAV and time-lapse imagery, classified by a combination of researchers, citizen scientists and machine learning, to study threatened seabird species across the Palearctic. I will use an existing time-lapse camera network to focus on one cliff-nesting species, the Black-legged Kittiwake Rissa tridactyla, with the aim of understanding how kittiwakes may be impacted by a changing environment. I will also deploy cameras to study a burrow-nesting species, the Atlantic Puffin Fratercula arctica, to try and improve methods for measuring productivity of burrow-nesting seabirds. By developing reproducible techniques to extract data from photographs, I hope to provide a means of sustained long-term monitoring across sufficient spatial scales to inform understanding of seabird decline.OBJECTIVES1) To compare accuracy of citizen science, researcher and machine learning identification of kittiwake in time-lapse images.2) To understand how abiotic (e.g. latitude, colony size, year) and biotic (e.g. weather conditions, predation, distance to fisheries, human disturbance) factors influence kittiwake phenology, breeding success and breeding site attendance across a latitudinal gradient using time-lapse imagery.3) To develop methods to measure puffin productivity from time-lapse images and compare with productivity calculated from 'traditional' field observations.4) To review the use of Unmanned Aerial Vehicles (UAVs) for seabird censuses and develop a guide to 'best practice' for seabird surveys with UAVs.5) To develop efficient methods for counting ground- and cliff-nesting seabird colonies from UAV imagery, and to use these data to test for spatial effects of colony structure linked to population trends.