Data from: Raccoon density estimation from camera traps for rabies virus management

During the springs of 2019 and 2020, we conducted camera trapping at 12 sites at the Savannah River Site in South Carolina. These sites were the same locations where Hill et al. (2023) conducted live trapping. The objective of our study was to determine if we could reliably estimate raccoon densities from images of unmarked animals. We conducted the camera trap study in the location where a live-trap mark-recapture study of raccoons was being conducted (Hill et al. 2023) to know what densities at each site should be.At each site we installed camera traps (Reconyx PC900 Hyperfire Professional IR cameras; Reconyx, Holmen, WI, USA) ~150 cm above the ground using stainless steel camera mounts and we angled cameras to ensure our bait was in the center of the frame. We baited each camera with a single plaster tab scented with fish oil, which we replaced each week. We deployed 34 cameras at each site. We used a 25-camera array (5×5 array with cameras at 100-m spacing) located in the same locations where box traps were deployed during live-trapping sessions. In addition, we added a 9-camera array (3×3 array with 200-m spacing) next to the 25-camera array to allow for a broader examination of effort reduction. We set cameras to take 3 pictures per motion-trigger event on a 1-second interval with no quiet period. We ran camera traps from 21-24 days for each site (exception for first round of 2019, which ran for 15 days). We recorded the date and time (DateTimeOriginal), if the photo used infrared (InfraredIlluminator - 1=Yes, 0=No), the species observed and the number of each (ID), the site name designation (Site), the camera name designation (camera), and an indicator if the image was the first in the motion capture burst (firstImage) for each photo. We also gave each image record a unique record ID (RecordID). The species and number of species in the record are denoted by a short species code designation (see SpeciesIDCode table for interpretation). The number of repeats of a species code indicates the number of individuals of that species seen on an image. Related study: Hill, J. E., J. L. Helton, D. A. Bernasconi, W. C. Dixon, M. T. Hamilton, R. B. Chipman, A. T. Gilbert, J. C. Beasley, G. Dharmarajan, and O. E. Rhodes Jr. 2023. Raccoon densities across four land cover types in the southeastern United States. Journal of Wildlife Management 87:e22480.

2019年与2020年春季，我们在南卡罗来纳州萨凡纳河站点（Savannah River Site）的12个监测点位开展了红外相机诱捕调查（camera trapping）。本次调查的点位与Hill等人（2023）开展活体诱捕调查的点位完全一致。本研究的核心目标为验证：能否通过未佩戴标识标记的浣熊影像，可靠估算其种群密度。我们选择Hill等人（2023）开展浣熊标记重捕活体诱捕研究的区域部署红外相机，以获取各点位的真实种群密度作为参照基准。
每个点位均采用不锈钢相机支架，将红外相机陷阱（camera trap，Reconyx PC900 Hyperfire Professional IR红外相机；Reconyx公司，美国威斯康星州霍尔门市）安装于距地面约150厘米处，并调整相机角度以确保诱饵处于画面中央。每台相机均投放一块浸有鱼油香味的石膏诱饵块，且每周更换一次诱饵。每个点位共部署34台相机。
相机布设采用两种阵列：其一为25台相机组成的5×5网格阵列，相机间距为100米，布局位置与活体诱捕阶段布设笼式诱捕器的区域完全一致；其二为在上述25台相机阵列旁额外增设的9台相机组成的3×3网格阵列（间距200米），以更全面地探究监测投入缩减的可行性。相机参数设置为：每次运动触发事件拍摄3张照片，拍摄间隔为1秒，无静默期。各点位的红外相机监测时长为21至24天不等，仅2019年首轮监测为例外，仅持续15天。
我们为每张照片记录以下元数据：拍摄日期与时间（DateTimeOriginal）、是否启用红外补光（InfraredIlluminator，1=是，0=否）、观测到的物种及其个体数量（ID）、点位名称标识（Site）、相机名称标识（camera），以及该照片是否为运动触发连拍的首张照片（firstImage）。我们还为每条影像记录分配了唯一的记录编号（RecordID）。记录中的物种及个体数量通过简短的物种代码标识（具体释义详见SpeciesIDCode表），物种代码的重复次数即代表该图像中观测到的对应物种个体数。
相关参考文献：Hill, J. E.、J. L. Helton、D. A. Bernasconi、W. C. Dixon、M. T. Hamilton、R. B. Chipman、A. T. Gilbert、J. C. Beasley、G. Dharmarajan 及 O. E. Rhodes Jr.，2023。美国东南部四种土地覆盖类型下的浣熊种群密度。《野生动物管理杂志》（Journal of Wildlife Management）87卷:e22480。