Estimating population density and survival of ocelots in six study sites over multiple years in Belize, Central America

The elusive and nocturnal nature of the ocelot Leopardus pardalis poses difficulty in gaining basic information on demographic parameters needed to better inform conservation. My study used camera trapping data from long-term monitoring of ocelot populations on six different protected areas in Belize over a time span ranging from 1 to 12 years, with 1,700 ocelot detections in 65,157 total trap nights. I used classical and spatially explicit methods, including multi-session robust design, to estimate and compare ocelot density and survival across sites and time. Full likelihood single session models estimated densities ranging from 6.4 - 22.5 individuals/100km2 in the broadleaf forested sites. Robust design models estimated densities from 8.8 - 22.8 individuals/100 km2 and ocelots had high annual survival (71-79%) in 2 broadleaf sites. Contrary to predictions, robust design models had higher precision than full likelihood models less than half the time. Spatially explicit models estimated density ranging between 7.2 – 22.0 individuals/100 km2 in broadleaf sites, and much lower estimates at 0.9 individuals/100 km2 in the pine forest site. Accounting for sex in spatially explicit methods, which directly incorporate locations of captures into the model, increased precision in density estimates by reducing individual heterogeneity in capture probability. The spatial models also demonstrated that males moved larger distances than females and had slightly higher detection rates. Ocelot populations remained relatively stable over time at the long term sites. My study produced methodologically rigorous abundance/density estimates for ocelots in Belize and the first ever ocelot survival estimates.