Distance and size matters: A comparison of six wildlife camera traps and their usefulness for wild birds

Abstract

Camera traps are increasingly used in ecological research. However, tests of their performance are scarce. It is already known from previous work that camera traps frequently fail to capture visits by animals. This can lead to a misinterpretation of ecological results such as density estimates or predation events. While previous work is mainly based on mammals, for birds, no data about if and how camera traps can be successfully used to estimate species diversity or density are available. Hence, the goal of our study was an empirical validation of six different camera traps in the field. We observed a total number of N = 4567 events (independent visits of a bird) in 100 different sessions from March 2017 until January 2018 while camera traps were deployed. In addition, N = 641 events are based on a comparison of the two close‐up camera traps especially designed for birds. These events were all directly observed by the authors. Thus, the cameras can be compared against the human observer. To give an overall assessment and a more generalizable result, we combined the data from the six camera traps and showed that bird size category (effect size = 0.207) and distance (effect size = 0.132) are the most important predictors for a successful trigger. Also, temperature had a small effect, and flock size had an impact with larger flocks being captured more often. The approach of the bird, whether it approached the camera frontally or laterally had no influence. In Table 8, we give some recommendations, based on our results, at which distances camera traps should be placed to get a 25%, 50%, and 75% capture rate for a given bird size.

Component processes of detection probability in camera-trap studies: understanding the occurrence of false-negatives

“…Our trials with two frequently used models of camera-trap demonstrate important limitations in PIR sensitivity. Similar poor capture at close distance (1 m) has also been found in a study of birds (mean of 60% across six size classes of bird and six CT models), where CTs were programmed to capture still images and high sensitivity (Randler and Kalb 2018). We suggest that imperfect triggering at close distances for small to medium homoiotherms may be ubiquitous in CT technology and thus needs to be evaluated prior to distance sampling and other quantitative studies, with a CCTV control being a useful method.…”

“…A CT trigger is not sufficient alone to record an animalthe animal must also be visible on the CT image or video. Trigger latency or trigger speed is the interval of time between PIR trigger and initiation of the camera (Rovero et al 2013) which can vary widely between CT models (Randler and Kalb 2018). A slow trigger speed coupled with fast moving animals means that not all triggers lead to registration as the animal has passed through the field-of-view before the camera has been activated (Rovero et al 2013).…”

Baiting/Luring Improves Detection Probability and Species Identification—A Case Study of Mustelids with Camera Traps

“…Lures were set 3–4 m directly in front of cameras within their detection zone. At this distance, a mammal the length of a marten (>40 cm, about 1.6 kg) has a high probability of being detected by the camera traps (see, [ 10 ]).…”

“…The following camera traps were used: SecaCam Raptor, Dörr SnapShot Extra Black 5.0 black (model 204401), Bushnell Natureview (model 119740). For further details on all cameras, see Randler and Kalb [ 10 ].…”

“…The detection probability is affected by the study design, including the number and position of traps as well as the number of active camera days [ 6 , 7 , 8 ]. Moreover, the size of the target species and its distance to the camera can affect detection probability [ 9 , 10 ]. Another problem in studies surveying mammals with wildlife camera traps is the insufficient image quality for species or individual recognition, as detected animals are for example very small, too far away or move too quickly in front of the camera trap.…”