Does Nest Density Matter? Using Novel Technology to Collect Whole-colony Data on Adelie Penguins

将开发部署协调无人驾驶飞行器的新方法，目的是获得可用来描述南极罗斯岛克罗齐尔角阿德利企鹅筑巢栖息地的全群落图像。这些信息将被用来检验关于地形特征、筑巢密度和繁殖成功之间关系的假设。这种种群可能是世界上最大的种群，处于该物种范围的南部极限，在过去20年中，其规模翻了一番，而该地区的大多数其他殖民地保持稳定或下降。从该项目获得的新信息将有助于理解气候驱动的陆地筑巢栖息地变化对未来影响阿德利企鹅的潜力。该项目将制作和记录开源软件工具，以帮助实现阿德利企鹅自动计数的图像处理自动化。该项目将培训研究生和本科生，并为基于相关企鹅科学项目的正在进行的教育推广项目贡献材料。从该项目获得的信息将有助于建立强有力、成本效益高的监测阿德利企鹅种群的议定书，这是罗斯海海洋保护区研究和监测计划草案中确定的一项关键生态系统指标。阿德利企鹅是南大洋生态系统功能和变化的重要指标。除了面临海冰的快速变化和远洋环境中的其他因素外，它们的陆地筑巢栖息地也在发生变化。了解物种对这种变化的反应对于评估其适应气候变化的能力至关重要。这个项目的目的是检验几个假设，关于细微规模的筑巢栖息地，巢密度，以及在罗斯海地区繁殖成功的阿德利企鹅的影响。为了实现这一目标，该项目将开发算法，以提高无人驾驶航空系统测量的效率和安全性，并开发和传播自动图像处理工作流程。在几次无人机调查中收集的图像将被用来估计产生的筑巢成虫和雏鸟的数量，以及估计南极洲罗斯岛上两个殖民地不同部分的筑巢密度，这两个殖民地的大小相差两个数量级。图像将被用于生成高分辨率的数字表面/高程模型，该模型将允许推导出洪水风险和地形复杂性等地形变量。将表面模型与两个群体的巢和雏鸟数量相结合，将提供栖息地协变量、巢密度和繁殖成功之间的关系。所制定的办法将使阿德利企鹅种群数量以及罗斯海海洋保护区研究和监测计划中的其他几项指标得以确定和评估。开发的飞行控制算法有可能用于许多类型的调查，特别是当需要在短时间内覆盖具有极端天气潜力和困难着陆选择的大片区域时。航空图像和视频将被用来制作有用的材料，包括在外联和教育计划中。自动图像处理工作流程和分类模型将作为开源软件开发，并将免费提供给其他解决类似野生动物监测挑战的人。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

New methodologies for the deployment of coordinated unmanned aerial vehicles will be developed with the aim of attaining whole-colony imagery that can be used to characterize nesting habitats of Adelie penguins at Cape Crozier, on Ross Island, Antarctica. This information will be used to test hypotheses regarding relationships between terrain characteristics, nesting density, and breeding success. This population, potentially the largest in the world and at the southern limit of the species' range, has doubled in size over the past 20 years while most other colonies in the region have remained stable or declined. New information gained from this project will be useful in understanding the potential ofclimate-driven changes in terrestrial nesting habitats for impacting Adelie penguins in the future. The project will produce, and document, open-source software tools to help automate image processing for automated counting of Adelie penguins. The project will train graduate and undergraduate students and contribute materials to ongoing educational outreach programs based on related penguin science projects. Information gained from this project will contribute towards building robust, cost-effective protocols for monitoring Adelie penguin populations, a key ecosystem indicator identified in the draft Ross Sea Marine Protected Area Research and Monitoring Plan. Adelie penguins are important indicators of ecosystem function and change in the Southern Ocean. In addition to facing rapid changes in sea ice and other factors in their pelagic environment, their terrestrial nesting habitat is also changing. Understanding the species' response to such changes is critical for assessing its ability to adapt to the changing climate. The objective of this project is to test several hypotheses about the influence of fine-scale nesting habitat, nest density, and breeding success of Adelie penguins in the Ross Sea region. To accomplish this, the project will develop algorithms to improve efficiency and safety of surveys by unmanned aerial systems and develop and disseminate an automated image processing workflow. Images collected during several UAV surveys will be used to estimate the number of nesting adults and chicks produced, as well as estimate nesting density in different parts of two colonies on Ross Island, Antarctica, that differ in size by two orders of magnitude. Imagery will be used to generate high resolution digital surface/elevation models that will allow terrain variables like flood risk and terrain complexity to be derived. Combining the surface model with the nest and chick counts at the two colonies will provide relationships between habitat covariates, nest density, and breeding success. The approaches developed will enable Adelie penguin population sizes and potentially several other indicators in the Ross Sea Marine Protected Area Research and Monitoring Plan to be determined and evaluated. The flight control algorithms developed have the potential to be used for many types of surveys, especially when large areas need to be covered in a short period with extreme weather potential and difficult landing options. Aerial images and video will be used to create useable materials to be included in outreach and educational programs. The automated image processing workflow and classification models will be developed as open source software and will be made freely available for others addressing similar wildlife monitoring challenges.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.