Table_4_Warm beach, warmer turtles: Using drone-mounted thermal infrared sensors to monitor sea turtle nesting activity.xlsx

For decades sea turtle projects around the world have monitored nesting females using labor-intensive human patrolling techniques. Here we describe the first empirical testing of a drone-mounted thermal infrared sensor for nocturnal sea turtle monitoring; on the Osa peninsula in Costa Rica. Preliminary flights verified that the drone could detect similar sea turtle activities as identified by on-the-ground human patrollers – such as turtles, nests and tracks. Drone observers could even differentiate tracks of different sea turtle species, detect sea turtle hatchlings, other wildlife, and potential poachers. We carried out pilot flights to determine optimal parameters for detection by testing different thermal visualization modes, drone heights, and gimbal angles. Then, over seven nights, we set up a trial to compare the thermal drone and operators’ detections with those observed by traditional patrollers. Our trials showed that thermal drones can record more information than traditional sea turtle monitoring methods. The drone and observer detected 20% more sea turtles or tracks than traditional ground-based patrolling (flights and patrols carried out across the same nights at the same time and beach). In addition, the drone operator detected 39 other animals/predators and three potential poachers that patrollers failed to detect. Although the technology holds great promise in being able to enhance detection rates of nesting turtles and other beach activity, and in helping to keep observers safer, we detail challenges and limiting factors; in drone imagery, current cost barriers, and technological advances that need to be assessed and developed before standardized methodologies can be adopted. We suggest potential ways to overcome these challenges and recommend how further studies can help to optimize thermal drones to enhance sea turtle monitoring efforts worldwide.

数十年来，全球各地的海龟保护项目一直采用劳动密集型的人工巡逻技术，对筑巢雌性海龟开展监测。本研究首次在哥斯达黎加奥萨半岛（Osa peninsula）开展了搭载热红外传感器（thermal infrared sensor）的无人机用于夜间海龟监测的实证测试。初步飞行测试验证，该无人机可识别出与地面巡逻人员所发现的同类海龟活动，包括海龟、巢穴及其爬行轨迹。无人机观察员甚至能够区分不同海龟物种的爬行轨迹，还可检测到海龟幼崽、其他野生动物以及潜在偷猎者。我们开展了试点飞行，通过测试不同的热成像模式、无人机飞行高度及云台角度（gimbal angles），以确定最优检测参数。随后在七个夜晚的试验中，我们设置了对照实验，将热成像无人机与操作员的检测结果，与传统巡逻人员的观测结果进行比对。试验结果表明，热成像无人机能够记录到比传统海龟监测方法更丰富的信息。在同一夜晚、同一海滩的相同时段开展飞行与巡逻的前提下，无人机及观察员发现的海龟或爬行轨迹数量较传统地面巡逻高出20%。此外，无人机操作员还检测到39只其他动物/捕食者以及3名潜在偷猎者，而这些均未被巡逻人员发现。尽管该技术在提升筑巢海龟及其他海滩活动的检测率、助力保障监测人员安全方面极具应用前景，但我们也详细说明了其面临的挑战与限制因素：包括无人机影像相关的现有成本壁垒，以及在落地标准化监测流程前，仍需评估与开发的技术提升方向。我们提出了克服这些挑战的可行方案，并建议后续研究应如何优化热成像无人机，以助力全球范围内的海龟监测工作。