Single-Camera Trap Survey Designs Miss Detections: Impacts on Estimates of Occupancy and Community Metrics

Abstract

The use of camera traps as a tool for studying wildlife populations is commonplace. However, few have considered how the number of detections of wildlife differ depending upon the number of camera traps placed at cameras-sites, and how this impacts estimates of occupancy and community composition. During December 2015–February 2016, we deployed four camera traps per camera-site, separated into treatment groups of one, two, and four camera traps, in southern Illinois to compare whether estimates of wildlife community metrics and occupancy probabilities differed among survey methods. The overall number of species detected per camera-site was greatest with the four-camera survey method (P<0.0184). The four-camera survey method detected 1.25 additional species per camera-site than the one-camera survey method, and was the only survey method to completely detect the ground-dwelling silvicolous community. The four-camera survey method recorded individual species at 3.57 additional camera-sites (P = 0.003) and nearly doubled the number of camera-sites where white-tailed deer (Odocoileus virginianus) were detected compared to one- and two-camera survey methods. We also compared occupancy rates estimated by survey methods; as the number of cameras deployed per camera-site increased, occupancy estimates were closer to naïve estimates, detection probabilities increased, and standard errors of detection probabilities decreased. Additionally, each survey method resulted in differing top-ranked, species-specific occupancy models when habitat covariates were included. Underestimates of occurrence and misrepresented community metrics can have significant impacts on species of conservation concern, particularly in areas where habitat manipulation is likely. Having multiple camera traps per site revealed significant shortcomings with the common one-camera trap survey method. While we realize survey design is often constrained logistically, we suggest increasing effort to at least two camera traps facing opposite directions per camera-site in habitat association studies, and to utilize camera-trap arrays when restricted by equipment availability.

Influence of Forest Structure and Composition on Summer Habitat Use of Wildlife in an Upland Hardwood Forest

“…Our camera trap survey results were comparable to other regional camera trap surveys, where 9 to 28 species were detected, suggesting sufficient survey effort [31,45,46]. Detectability probabilities in our study were also comparable to similar regional and international mammal occupancy studies.…”

A comparison of camera trap and permanent recording video camera efficiency in wildlife underpasses

“…Another way to improve the detection rate would be the simultaneous use of several camera traps (Pease et al., 2016). This can be easily achieved in the underpasses by placing one camera at each end of them, as we do in this study, rather than one on the middle.…”

“…In order to estimate the efficiency of camera traps, some studies compared this method to others (Diggins et al, 2016;Dillon & Kelly, 2008;Glen et al, 2013;Janečka et al, 2011;Li, McShea, Wang, Huang, & Shao, 2012;Lyra-Jorge, Ciocheti, Pivello, & Meirelles, 2008;Monterroso, Rich, Serronha, Ferreras, & Alves, 2013;Silveira, Jácomo, & Diniz-Filho, 2003;Villette, Krebs, Jung, & Boonstra, 2016). Other authors compared different models of camera traps (Hughson, Darby, & Dungan, 2010;Meek & Vernes, 2016;Rovero, Zimmermann, Berzi, & Meek, 2013;Swann, Hass, Dalton, & Wolf, 2004;Weingarth, Zimmermann, Knauer, & Heurich, 2013), their technical parameters (Kelly & Holub, 2008;Pease, Nielsen, & Holzmueller, 2016), and the different installation and placement methods (Foster & Harmsen, 2012;Guil et al, 2010;Smith & Coulson, 2012). As the efficiency of camera traps also depends on the targeted species and their characteristics (Ariefiandy, Purwandana, Seno, Ciofi, & Jessop, 2013;Lyra-Jorge et al, 2008;Rowcliffe, Carbone, Jansen, Kays, & Kranstauber, 2011;Tobler, Carrillo-Percastegui, Leite Pitman, Mares, & Powell, 2008;Welbourne, MacGregor, Paull, & Lindenmayer, 2015), the use of lures (Diete, Meek, Dickman, & Leung, 2016;MCCleery et al, 2014;Read, Bengsen, Meek, & Moseby, 2015) and the associated bias (Meek et al, 2014b;Newey et al, 2015;Rocha, Ramalho, & Magnusson, 2016) were also investigated.…”

Framing pictures: A conceptual framework to identify and correct for biases in detection probability of camera traps enabling multi‐species comparison

“…Meek, Dixon, Dickman and Leung, 2016;Satterfield et al, 2017; Suárez-Tangil and Rodríguez, Reis et al, 2017;Lesmeister et al, 2015;Pease, Nielsen and Holzmueller, 2016;Welbourne et al, 2016 Welbourne et al, ) al., 2017aHowe et al 2017; Rowcliffe et al.al., 2007;Wearn et al, 2017) Landscape features channeling animal movement (e.g.,…”