Use of camera traps to evaluate traffic on poultry premises.
By Erin Cortus, Marie Culhane, Eva Cornwell There are many pathways for poultry diseases like Salmonella and Highly Pathogenic Avian Influenza (HPAI) to transfer from one host to another, from one farm to another, or from wildlife to domestic flocks. In the environment around poultry farms, indirect transfers of disease depend on many factors including the frequency of visits and possible interactions between birds (wild and domestic), animals (wild and domestic), humans, vehicles, and equipment. Disease risk reduction usually involves farm biosecurity measures such as rodent control, restricting human and equipment traffic between poultry houses and the environment, and maintaining logs of activities. Visual observations of wild birds and other wildlife on the premises by workers and managers are limited to their time spent on farm. One of the goals of this work is to help poultry farmers refine biosecurity measures to reduce the risk of pathogen introduction from interactions with wild species or increased traffic on the farm. Camera traps, or trail cameras, are commonplace in hunting and trapping. More recently, the research community has made use of cameras in various countries and settings, often to document visits of wildlife to a potential disease transmission zone. Using camera traps in both wildlife and agriculture studies reduces the disturbance and time requirements of in-person observations. Modern trail cameras have variable settings that can enable image collection even in poor light or night conditions, motion detection, and still or video file formats. Although camera image review is time-consuming and generally a fixed position view, the data captured can be used to identify the frequency of wildlife, humans, and possible vectors of disease in and around a poultry farm. The cameras represent one tool to add to the biosecurity toolkit that includes farm records, restricted/controlled access and entry points, and training. It is important to recognize that neither farm records (i.e. logs) nor camera images denote the presence of disease, but they do provide evidence of the frequency and type of visitors that make up certain pathways.
A Minnesota study There were camera traps at six poultry farms in Minnesota during three migration seasons (Fall 2018, Spring 2019, Fall 2019). The cameras were Bushnell B-12 12 MP Trail Cameras, set in motion sensor mode. The camera position and view angle were fixed within a season for each barn but did change between seasons on some farms. The number of images was limited by the view angle for each farm and was not wholly indicative of all movement on the farms. A team of reviewers sorted the images based on traffic type (Mammals, Birds, Humans, Vehicles and No Traffic) and documented the number, type (as best possible), and location (i.e. on ground, perched on barn ridge, flying overhead, etc.) of traffic in each image. Birds included wild and domestic (i.e. escapees). The farm veterinarians provided visitor logs and rodent trap records for the camera-monitored periods for Farms 1, 2 and 3, for comparison to the images. Across the six barns and three seasons, the cameras captured over 9,000 images. Following review, approximately 5,500 images contained at least one form of “Traffic”. Snow, blowing vegetation, or other non-traffic movement triggered photos that were ultimately removed from the dataset, but still consumed time and energy during the review process. The number of images differed between sites and seasons. There were 73 total images of mammals on all farms in the view angles, and 79% of the instances occurred outside of typical working hours. On Farm 2, the mammals tended to be housecats, raccoons, coyotes, or other small animals during Fall 2018 and Spring 2019; there were frequent instances of white-tailed deer in Fall 2019. The greatest frequencies of birds were in the Spring 2019 season on all but Farm 4. Most bird images occurred during working hours.
The number of images with vehicles is similar in pattern, but generally greater in frequency than images with humans. The vehicles were generally farm trucks. Farms 3 and 5 view angles included a service road, resulting in higher frequencies of human and vehicle traffic images. The human and vehicle traffic pictures generally exceeded those of birds or mammals at most farms. With camera traps, one must consider maintaining the security on the farm, but also the right to privacy. This needs to be a discussion point for future camera use on farms. While visitor and rodent/pest logs are recommended biosecurity measures for any farm, their usefulness is limited by the completeness, accuracy, and consistency of the log entries. All rodent/pest logs reviewed for this case study suggested regular maintenance/inspection of rodent traps and/or bird nests, and visits by crews for vaccinations or bird movements. Periodically, there were notations of the number of mice trapped or bird nests. There was no noticeable agreement between the images and the logs for many reasons, particularly view angles. This suggests the two record types may serve more in complementary roles rather than as validation measures. Summary • Camera traps were easy to install and use. • The time for image review was considerable. Each image took approximately 60 s to review and code in this project. Recording image types is not necessary but recommended for pattern detection and comparison to management records. • View angle is critical. More cameras can capture more angles but adds to review time. Camera trap technology is evolving, and video may prove more useful for some situations than still images. • The goal of camera trap use should ultimately guide camera placement. For example, camera settings could limit images to off-work hours to capture wildlife or other unexpected human traffic but could miss bird movement. • Cameras are complementary to other biosecurity measures. Regardless of camera angle and placement, if the images are to be used to determine pathways of pathogen introduction into a turkey barn, regular and timely review of the images is highly recommended, not only to decrease the backlog of images to review but also provide more immediate feedback to the farm managers who can then make corrections to biosecurity protocols as needed.
Erin Cortus, PE, PhD, is Associate Professor and Extension Engineer, Bioproducts and Biosystems Engineering, at the University of Minnesota. Marie R Culhane, DVM, PhD, is Professor, Department of Veterinary Population Medicine, at the University of Minnesota. Eva Cornwell, BA, is a Research Assistant and Veterinary Student at the University of Minnesota.
