Ecological Archives A025-021-A2
Hillary S. Young, Douglas J. McCauley, Rodolfo Dirzo, Jacob R. Goheen, Bernard Agwanda, Cara Brook, Erik Otárola-Castillo, Adam W, Ferguson, Stephen N. Kinyua, Molly M. McDonough, Todd M. Palmer, Robert M. Pringle, Truman P. Young, and Kristofer M. Helgen. 2015. Context-dependent effects of large-wildlife declines on small mammal communities in central Kenya. Ecological Applications 25:348–360. http://dx.doi.org/10.1890/14-0995.1
Appendix B. Camera trapping validation of dung surveys.
In addition to dung surveys detailed in the main text, we also surveyed relative activity levels of wildlife across sites using camera traps. These surveys are much more likely to detect small predators than dung surveys and were intended to provide a metric of predator activity across land-use types. We also used estimated activity levels of all wildlife garnered through camera traps, and compared this to the estimates of wildlife activity made via dung surveys as a way to confirm the usefulness of dung surveys in quantifying relative activity of wildlife.
For our camera trapping, we placed two cameras (Scout Guard Model SG750; HCO, Norcross, GA or Reconyx RM45) at each of our sites for approximately 21 days after each trapping session. Both cameras were located within the site, or within 100 m of the site, and were separated by at least 50 m from each other. Cameras were set opportunistically near animal trails and signs so as to maximize animal detection. Cameras took pictures 24 hours per day, with a 30 second delay between images. For analysis, only mammals larger than ~2 kg were counted. Photographs taken of the same species within a 30 minute window were considered to be the same animal (modified from Rovero and Marshall 2009). Camera trapping photographic rate was used as an index of relative activity of each focal group (Rovero and Marshall 2009).
For analysis of predator activity we considered only the following subset of animals likely to be predators of small mammals: domestic cat (Felis catus), African wild cat (Felis lybica), serval (Leptailurus serval), common genet (Genetta genetta), white-tailed mongoose (Ichneumia albicauda), dwarf mongoose (Helogale parvula), zorilla (Ictonyx striatus), and black-backed jackal (Canis mesomelas). We had between 300 and 1200 camera days for each land-use form. There were no significant differences in mammalian predator activity across habitat types. These results are consistent with a much more intensive camera trapping effort across the same area (Kinnaird and O'Brien2012). That study compared exclusive conservancies, to high and low intensity pastoral sites and fenced ranches (the last were not studied here), and found no effect of management category on most small carnivore activity but did document an increased abundance of one rodent predator (C. mesomelas) in pastoral sites. While this camera trapping only surveyed mammalian predators, given that site pairs were in such close proximity, we assume that avian predators of small mammals did not change in density within site pairs (although risk from avian predators may change based on changes to vegetation structure).
Results of both wildlife and domestic stock activity estimates from camera trap data were consistent with those estimated from dung surveys. There was a strong correlation between livestock (R² = 0.50, P < 0.0001) and wildlife (R² = 0.30, P < 0.0001) relative activity estimates obtained in these two methods. Only relative abundance of dung was used in all analyses as a proxy of wildlife activity.
Literature cited
Rovero, F., and A. R. Marshall. 2009. Camera trapping photographic rate as an index of density in forest ungulates. Journal of Applied Ecology 46:1011–1017.
Kinnaird, M. F., and T. G. O'Brien. 2012. The role of private lands, livestock management and human tolerance on diversity, distribution and abundance of large African mammals. Conservation Biology 6:1026–1039.