From Branch to Branch: Unravelling Primate Vision in Arboreal Leaping and Navigation

灵长类动物，包括人类和他们的非人类亲属，拥有高度衍生的视觉任务，如物体识别，空间视觉，检测和处理社会线索。在冠群灵长类动物的起源阶段，早期视觉皮层经历了一次重大的转变，在初级视觉皮层（V1）中出现了一种近乎不变的视觉回路形式。这种结构设计在灵长类动物的大脑进化过程中一直被保留下来，并伴随着专门用于视觉的新皮层的扩张。在鼠狐猴（Microcebus murinus）中，V1覆盖了20%的新皮层。然而，这种结构所提供的视觉能力和行为与人类、猿类和旧大陆猴有很大不同。在这里，我们提出了一个研究计划，以调查早期灵长类动物视觉的关键特征。使用EthoLoop平台，我们将在一个受控的自然主义实验室环境中研究自由放养的鼠狐猴的行为，认知和感知，类似于它们的树栖栖息地。该平台可以对复杂任务中的视觉引导导航、觅食行为和神经元记录进行定量研究。通过调整环境条件，我们可以探索视觉参数并推断计算原理。该项目结合了量化生态行为，理论神经科学和计算机视觉以及自由放养鼠狐猴神经记录的专业知识。本研究旨在了解在树栖导航和觅食过程中的视觉处理和认知。目标1：了解在树上跳跃的主动视觉的策略：为了表征视觉引导跳跃中涉及的不同模式和策略，我们将设计一种新颖的，可参数化的行为跳跃任务，受马达加斯加自然环境中实际观察到的物理特征的约束。目标二：在不同的视觉条件下表征大规模导航策略：为了更好地了解视觉条件如何影响导航策略，我们将在野外观察自由移动的鼠狐猴，模拟它们的行为，并在实验室环境中的受控条件下测试我们的预测。目标3：揭示小鼠狐猴大脑中视觉，凝视，姿势和空间的神经元代码：为了表征视觉和非视觉特征（包括运动和姿势参数）的编码，我们将使用电生理学和便携式微型显微镜成像记录自由移动条件下单个神经元的活动。综上所述，我们的研究将为灵长类动物的视觉活动开辟一个新的视角，它定义了人类大脑中仍然使用的视觉回路的大部分形式，并解决了几个关于灵长类动物大脑和密切相关的啮齿动物大脑中视觉信息处理之间潜在差异的主要开放问题。

Primates, including humans and their non-human relatives, possess a highly derived sense of vision for tasks such as object recognition, spatial vision, and detecting and processing of social clues. The early visual cortex underwent a significant transformation at the origin of crown group primates, with the emergence of a near invariant form of vision circuitry in the primary visual cortex (V1). This architectural design has been conserved throughout primate brain evolution and was accompanied by an expansion of neocortex dedicated to vision. In the mouse lemur (Microcebus murinus), a basal primate resembling the primate last common ancestor, V1 covers 20% of the neocortex. However, the visual capabilities and behaviors served by this architecture differ greatly from those in humans, apes, and Old World monkeys. Here we propose a research program to investigate key features of early primate vision. Using the EthoLoop platform, we will study behavior, cognition, and perception in free-ranging mouse lemurs in a controlled naturalistic laboratory setting resembling their arboreal habitat. This platform enables quantitative studies of visually guided navigation, foraging behaviors, and neuronal recording during complex tasks. By adjusting environmental conditions, we can explore visual parameters and infer computational principles. The project combines expertise in quantifying ecological behaviors, theoretical neuroscience and computer vision, and neural recording in free-ranging mouse lemurs. The research aims to understand visual processing and cognition during arboreal navigation and foraging. Aim 1: Understanding the strategies of active vision during arboreal leaping: To characterize the different modalities and strategies involved in visually guided leaping, we will design a novel, parameterizable behavioral leaping task, constrained by physical features actually observed in natural settings in Madagascar. Aim 2: Characterizing large scale navigation strategies under varying visual conditions: To gain a better understanding of how the visual conditions affect navigational strategies we will observe freely moving mouse lemurs in the wild, model their behavior and test our predictions under controlled conditions in the laboratory settings. Aim 3: Revealing the neuronal codes for vision, gaze, posture and space in the mouse lemur brain: To characterize the coding of visual and non-visual features (including motor and posture parameters) we will record the activity of individual neurons under freely moving conditions using electrophysiologyy and portable miniature microscope imaging. Taken together, our studies will open a new perspective on primate vision in action that defined much of the form of vision circuitry still used in the human brain and address several major open questions about potential differences between visual information processing in the primate brain and in that of closely related rodents.