Stöckl Anna Lisa, O'Carroll David Charles, Warrant Eric James
Department of Biology, University of Lund, Sölvegatan 35, 22362 Lund, Sweden.
Department of Biology, University of Lund, Sölvegatan 35, 22362 Lund, Sweden; Adelaide Centre for Neuroscience Research, The University of Adelaide, Adelaide, SA 5005, Australia.
Curr Biol. 2016 Mar 21;26(6):821-6. doi: 10.1016/j.cub.2016.01.030. Epub 2016 Mar 3.
Most of the world's animals are active in dim light and depend on good vision for the tasks of daily life. Many have evolved visual adaptations that permit a performance superior to that of manmade imaging devices [1]. In insects, a major model visual system, nocturnal species show impressive visual abilities ranging from flight control [2, 3], to color discrimination [4, 5], to navigation using visual landmarks [6-8] or dim celestial compass cues [9, 10]. In addition to optical adaptations that improve their sensitivity in dim light [11], neural summation of light in space and time-which enhances the coarser and slower features of the scene at the expense of noisier finer and faster features-has been suggested to improve sensitivity in theoretical [12-14], anatomical [15-17], and behavioral [18-20] studies. How these summation strategies function neurally is, however, presently unknown. Here, we quantified spatial and temporal summation in the motion vision pathway of a nocturnal hawkmoth. We show that spatial and temporal summation combine supralinearly to substantially increase contrast sensitivity and visual information rate over four decades of light intensity, enabling hawkmoths to see at light levels 100 times dimmer than without summation. Our results reveal how visual motion is calculated neurally in dim light and how spatial and temporal summation improve sensitivity while simultaneously maximizing spatial and temporal resolution, thus extending models of insect motion vision derived predominantly from diurnal flies. Moreover, the summation strategies we have revealed may benefit manmade vision systems optimized for variable light levels [21].
世界上大多数动物在昏暗光线下都很活跃，并且依靠良好的视觉来完成日常生活中的任务。许多动物已经进化出视觉适应性，使其表现优于人造成像设备[1]。在昆虫这个主要的视觉系统模型中，夜行性物种展现出了令人印象深刻的视觉能力，从飞行控制[2,3]到颜色辨别[4,5]，再到利用视觉地标[6 - 8]或昏暗的天体罗盘线索进行导航[9,10]。除了在昏暗光线下提高其敏感度的光学适应性[11]外，理论[12 - 14]、解剖学[15 - 17]和行为学[18 - 20]研究表明，光在空间和时间上的神经总和——以牺牲更嘈杂的精细和快速特征为代价，增强场景中更粗糙和缓慢的特征——有助于提高敏感度。然而，目前尚不清楚这些总和策略在神经层面是如何发挥作用的。在这里，我们对一种夜行性天蛾的运动视觉通路中的空间和时间总和进行了量化。我们表明，空间和时间总和超线性地结合，在四个数量级的光强度范围内大幅提高对比度敏感度和视觉信息率，使天蛾能够在比没有总和时暗100倍的光线下视物。我们的结果揭示了在昏暗光线下视觉运动是如何在神经层面进行计算的，以及空间和时间总和如何在提高敏感度的同时最大化空间和时间分辨率，从而扩展了主要源自昼行性果蝇的昆虫运动视觉模型。此外，我们所揭示的总和策略可能会使针对可变光水平进行优化的人造视觉系统受益[21]。
