Data from: Influence of visual perception on movement decisions by an ungulate prey species

Visual perception is dynamic and depends on physiological properties of a species’ visual system and physical characteristics of the environment. White-tailed deer (Odocoileus virginianus) are most sensitive to short- and mid-wavelength light (e.g. blue and green). Wavelength enrichment varies spatially and temporally across the landscape. We assessed how the visual perception of deer influences their movement decisions. From August to September 2019, we recorded 10-min locations from 15 GPS-collared adult male deer in Central Florida. We used Hidden-Markov models to identify periods of movement by deer and subset these data into three time periods based on temporal changes in light environments. We modeled resource selection during movement using path-selection functions and simulated 10 available paths for every path used. We developed five a priori models and used 10-fold cross validation to assess our top model's performance for each time period. During the day, deer selected to move through woodland shade, avoided forest shade, and neither selected nor avoided small gaps. At twilight, deer avoided wetlands as cloud cover increased but neither selected nor avoided other cover types. Visual cues and signals are likely more conspicuous to deer in short-wavelength-enriched woodland shade during the day, while at twilight in long-wavelength-enriched wetlands during cloud cover, visual cues are likely less conspicuous. The nocturnal light environment did not influence resource selection and likely has little effect on deer movements because it's relatively homogenous. Our findings suggest visual perception relative to light environments is likely an underappreciated driver of behaviors and decision-making by an ungulate prey species.

视觉感知具有动态性，既依赖于某一物种视觉系统的生理特性，也受所处环境的物理特征影响。白尾鹿（*Odocoileus virginianus*）对短波长与中波长光线（如蓝光、绿光）最为敏感。景观内的波长富集状况存在空间与时间上的异质性。本研究探究了鹿的视觉感知如何影响其移动决策。2019年8月至9月，我们对佛罗里达州中部15只佩戴GPS项圈的成年雄性白尾鹿进行了定位记录，每10分钟获取一次位置数据。本研究采用隐马尔可夫模型（Hidden-Markov Models）识别鹿的移动时段，并依据光环境的时间变化将数据集划分为三个时间段。本研究采用路径选择函数构建移动过程中的资源选择模型，并为每条已记录的移动路径模拟出10条可供选择的替代路径。本研究构建了5个先验模型，并采用10折交叉验证（10-fold cross validation）法评估各时间段最优模型的拟合性能。日间时段，白尾鹿偏好穿过林地遮蔽区域，规避森林阴影区域，而对小型林隙既无明显偏好也无规避行为。暮光时段，随着云量增加，白尾鹿会规避湿地生境，但对其他遮蔽类型无明显选择倾向。日间处于短波长富集的林地遮蔽区域时，视觉线索与信号对白尾鹿而言通常更为醒目；而暮光时段在云量较多的长波长富集湿地生境中，视觉线索则相对难以被察觉。夜间光环境相对均匀，因此不会对白尾鹿的资源选择产生影响，且大概率对其移动行为的作用有限。本研究结果表明，相较于光环境的视觉感知，可能是这类有蹄类猎物物种行为与决策中未被充分重视的驱动因素。