Lateral eye width for 240 species of Andean cloud forest birds

We quantified eye size using two sources. For 80% of the species (N = 192) we passively captured birds using mist nets at sites located within cloud forest tracts and regenerating second growth. One observer (IJA) used calipers to measure lateral eye width to the nearest 1.1mm. We compared our field measurements to published measurements of traverse eye diameter from preserved specimens for a subset of species (N =78; Ritland 1982) and found the two metrics to be highly correlated (r = 0.92, P &lt; 0.001). Hence, our field measurements can be considered accurate measures of true eye size. For the remaining 48 species (20%) we downloaded six photos per species of individuals perched in the field. We scaled photos in ImageJ using measurements of mean bill height, the position of which is clearly identified by the nares in photos. We only selected photos of birds at eye level, with the beak and eye positioned on the same focal plane. We averaged eye size across the six photos for each species and compared photo measurements from 20 species for which we also had field data. The two methods were strongly correlated (beta = 0.97, P &lt; 0.001, R-squared= 0.96). Photo measurements were on average 3% smaller than field measurements, so we calibrated mean eye size for the 48 species by multiplying by 1.03. The attached data set follows the taxonomy of the 2019 Clements List and the global Bird Tree (Jetz et al. 2012) and included raw measurements and residuals of eye size on body mass. See associated publication for details on analysis of eye residuals. <br>

本研究通过两种途径量化鸟类的眼球尺寸。
针对80%的物种（样本量N=192），我们在云雾林斑块及恢复次生林的调查点位，采用雾网被动捕获鸟类。由观测者IJA使用游标卡尺测量鸟类侧眼宽度，测量值精确至最接近的1.1毫米。我们将野外实测数据与78个物种馆藏标本的公开眼横径测量结果（Ritland，1982）进行比对，发现两类测量指标具有极强相关性（r=0.92，P<0.001）。由此可见，本研究的野外实测数据可作为真实眼球尺寸的精准量化指标。
剩余20%的物种共计48个，我们下载了每个物种6张野外停歇个体的鸟类照片。在ImageJ软件中，我们以喙部平均高度为参照对照片进行比例尺校正——照片中鼻孔可清晰标识喙高位置。筛选标准限定为：鸟类眼部与拍摄视角平齐，喙部与眼部处于同一焦平面。我们对每个物种的6张照片的眼球尺寸取平均值，并针对同时拥有野外实测数据的20个物种，对比了照片测量结果与野外实测结果。两类方法的相关性极强（beta=0.97，P<0.001，决定系数R²=0.96）。照片测量的眼球尺寸平均较野外实测结果偏小3%，因此我们将48个物种的平均眼球尺寸乘以1.03完成校正。
本附数据集遵循2019年版克莱门茨鸟类名录（Clements List）与全球鸟类系统树（Jetz et al. 2012）的分类体系，包含原始测量数据及眼球尺寸相对于体重的残差值。有关眼球尺寸残差的分析细节，请参阅相关发表文献。