Effects of sardines as an attractant on carnivore detection and temporal activity patterns at remote camera traps

Context. Adding an attractant to remote camera traps has become a popular method to increase detections of cryptic carnivores. However, there is ongoing debate about whether this practice can bias the behaviour and activity patterns of these species.Aims. Our aim was to determine how using a popular attractant at camera traps could affect carnivore detection probabilities and temporal activity.Methods. We used an experimental design in which we randomly set half of our camera traps on arrays at two sites in western North Carolina with an attractant: canned sardines in oil. Halfway through the survey season, we switched which camera traps had sardines and which did not. We estimated detection probability for each carnivore species observed using occupancy models, and we used kernel density estimations to evaluate changes in diel activity patterns between camera traps with and without an attractant.Key results. We found that when sardines were used at camera traps, detectability of bobcats (Lynx rufus), coyotes (Canis latrans), northern raccoons (Procyon lotor), Virginia opossums (Didelphis virginiana), and eastern spotted skunks (Spilogale putorius) more than doubled, but there was little or no effect on striped skunk (Mephitis mephitis) detectability. Of the species we most frequently detected (coyotes, raccoons, and opossums), activity patterns between camera traps with and without a sardine attractant overlapped moderately to highly, and a significant effect on diel activity patterns was observed only for raccoons.Conclusions. Use of attractants can greatly increase the probability of detecting nearly all carnivores at camera traps. The effects of attractants on diel activity patterns are species-specific, with two of our three most-detected species unaffected by their use, suggesting that attractants can be used to effectively study these behaviours in some carnivore species.Implications. A sardine attractant can increase the detection of many carnivore species, using camera traps, without causing a significant deviation of diel activity patterns, thus allowing for unbiased investigations into most species' spatio-temporal behaviour in the Appalachian Mountains - and likely other systems.