Sensory strategies and behavioural algorithms in predator-prey-interactions

动物的感知和行动形成了一个紧密结合的循环。感觉系统为动物提供有关周围环境的信息，运动系统执行行为动作。行为算法是将感觉输入映射到自适应运动输出的规则。捕食者-猎物相互作用在自然界中无处不在，对捕食者和猎物都构成了强大的选择压力。感觉和行为策略是捕食者-猎物相互作用的关键部分，尤其是在视觉系统中有充分记录的，包括伪装和视觉引导的追捕和逃跑。追赶和逃跑是研究捕食者和猎物的感觉策略和行为算法的好模型，因为它们有一个明确可识别的目标，可以通过实验跟踪和数学描述。在这里，我提出了一种结合经验和理论的方法来研究捕食者和猎物在听觉信息指导下的感觉策略和行为算法，即回声定位蝙蝠和耳蛾。研究听觉引导的捕食者-猎物相互作用与更知名的视觉策略相比，对于获得行为算法的一般理解至关重要，包括共同原则以及感觉系统之间的基本差异。我将开发、校准和描述先进的技术，用于自由飞行蝙蝠的3d跟踪，以及基于虚拟现实和仿生机器人猎物的飞行蝙蝠的实时闭环实验。我将根据经验测量飞蛾提供的感觉声学线索，包括飞行声音和时变回声信息，并量化尺度对声学伪装的贡献。在野外和实验室对自由飞行的回声定位蝙蝠进行的实验中，我将对不同的反捕食者策略（声学伪装、停止拍翼和落地）的有效性进行实证测试。此外，我将分离猎物产生的声音线索，回声定位蝙蝠依靠它来探测和选择自由飞行的飞蛾，它们是如何追捕逃避的飞蛾的，我将建模它们攻击猎物的控制规则。在建模方法中，我将把负声母性（这是耳蛾逃避防御的第一道防线）划分为三个连续的行为算法。我将使用社区范围内的听力阈值、飞行运动学和蝙蝠呼叫参数数据来建模相互检测距离，并将蝙蝠搜索和飞蛾逃避飞行模型参数化。在不同的建模方法中，我将测试负声性的恒定缓冲假设，预测飞蛾成功逃脱所需的最低听力阈值，并开发一个大规模的感觉运动模型来描述飞蛾逃避飞行的有效性和成本。

Sensing and acting of animals form a tightly integrated loop. Sensory systems provide animals with information about their surroundings, and motor systems implement the behavioural actions. Behavioural algorithms are the rules that map the sensory input to adaptive motor output. Predator-prey-interaction occur everywhere in nature and pose a strong selection pressure on both predators and prey. Sensory and behavioural strategies are crucial part of predator-prey-interactions, particularly well documented in the visual system, including camouflage and visually-guided pursuit and escape. Pursuit and escape are great models to study the sensory strategies and behavioural algorithms of predators and prey, as they serve a clearly identifiable goal and can be experimentally tracked and mathematically described. Here, I propose a combined empirical and theoretical approach to investigate sensory strategies and behavioural algorithms in predators and prey that are guided by auditory information, namely echolocating bats and eared moths. Investigating auditory-guided predator-prey-interactions in comparison to the better-known visual strategies is crucial for obtaining a general understanding of behavioural algorithms including common principles as well as fundamental differences between sensory systems.I will develop, calibrate and describe advanced technologies for the 3D-tracking of free-flying bats and for real-time closed-loop experiments with flying bats based on virtual realities and biomimetic robotic prey. I will empirically measure sensory acoustic cues provided by moths, including flight sounds, and time-variant echo information, and quantify how much scales contribute to acoustic camouflage. In experiments with free-flying echolocating bats in the field and lab, I will empirically test the effectiveness of different antipredator strategies (acoustic camouflage, cessation of wingbeats, and dropping to the ground). Furthermore, I will isolate the prey-generated acoustic cues that echolocating bats rely on to detect and select free-flying moths, how they pursue evading moths, and I will model the control rules underlying their attacks on prey.In a modelling approach, I will divide negative phonotaxis, which is the first line of evasive defence of eared moths, into three consecutive behavioural algorithms. I will use community-wide data of hearing thresholds, flight kinematics and bat call parameters to model mutual detection distances, and to parametrize models of bat search and moth evasive flight. In different modelling approaches, I will test the constant-buffer-hypothesis of negative phonotaxis, predict the minimally required hearing thresholds of moths for successful escape, and develop a large-scale sensory-motor-model to delineate the effectiveness and costs of moth evasive flight.