The value of by-catch data: how species-specific surveys can serve non-target species

Abstract

Camera trapping has a wide range of research application, but, while research designs are often focused on the study of a single focal species, cameras can also record other non-target species. Occupancy modeling using by-catch data can be a valuable resource to gain information on these species maximizing the scientific effort and efficiency of wildlife surveys. In this study, we used by-catch data from a European wildcat (Felis silvestris silvestris) survey in Southern Italy to assess the habitat covariates determinant for the occupancy of the crested porcupine (Hystrix cristata). We recorded 33 detections at 17 out of 51 cameras (naive occupancy = 0.33). The best models fitted the data well, and porcupine occupancy estimate was 0.58 (SE +/- 0.09) with a detection probability of 0.11 (SE +/- 0.03). Average model showed that woodlands and number of shrub patches increased porcupine occupancy, while the reverse was true for altitude. Our results have improved the insights on the habitat use and ecological needs of this understudied species, and it is the first study that develops occupancy models for the porcupine using the presence/absence data obtained from a camera trap survey. Our study is an example of how camera trap surveys are often an under-exploited source of valuable information on a wider spectrum of sympatric species beyond the focal species for which camera traps were deployed. Minimum requirements for a camera trap survey to provide robust occupancy estimates for non-target species are discussed.