Interpreting Carnivore Scent-Station Surveys

Abstract

The scent-station survey method has been widely used to estimate trends in carnivore abundance. However, statistical properties of scent-station data are poorly understood, and the relation between scentstation indices and carnivore abundance has not been adequately evaluated. We assessed properties of scentstation indices by analyzing data collected in Minnesota during 1986-93. Visits to stations separated by <2 km were correlated for all species because individual carnivores sometimes visited several stations in succession. Thus, visits to stations had an intractable statistical distribution. Dichotomizing results for lines of 10 stations (0 or ?1 visits) produced binomially distributed data that were robust to multiple visits by individuals. We

A monitoring program for Patagonian foxes based on power analysis

“…Among these components, it was particularly relevant choosing a visitation index whose statistical properties included a relatively low variance. In this regard, index 6 (a single check after 48-h operation) and especially index 7 (one check after 72 h) showed the best behavior and performed much better than the other indices, including index 1 (one check after 24 h) which is widely used in scent station and bait station surveys of many carnivores (Sargeant et al 1998). The superior performance of index 7 seems to be density dependent.…”

“…We computed visitation indices for the first (index 1), second (index 2), and third (index 3) night separately. Index 1 was identical to the one used in most single-night scent station studies (Sargeant et al 1998;Travaini et al 2003a, b). As resident foxes could encounter bait stations on successive nights, index 2 and index 3 could be positively or negatively affected by previous experience with baits.…”

“…As recommended by Roughton and Sweeny (1982), scent stations and bait stations have been operated just for one night in most studies (Linscombe et al 1983;Smith et al 1994;Travaini et al 1996;Sargeant et al 1998). Singlenight operation may restrict the chances of individuals to encounter at least one station during its activity period that particular night, especially if their home ranges are large compared with station spacing.…”

“…The probability of contagion in visits has been emphasized for adjacent stations or even clustered lines (Sargeant et al 1998). Temporal contagion is also likely since individuals may visit a station with higher probability if they have previously obtained a bait as a reward.…”

“…Scent stations and bait stations are standard methods to survey carnivores at broad spatial scales (Harrison et al 2002;Sargeant et al 2003;Zielinski et al 2005). These methods can detect population trends when substantial changes in numbers occur across wide areas (Diefenbach et al 1994;Sargeant et al 1998). One of their drawbacks, however, is that a high proportion of visitation rates equal to zero produces concomitant high variability and reduces statistical power to detect trends (Gibbs 2000).…”