Mapping the ghost: Estimating probabilistic snow leopard distribution across Mongolia

Aim Snow leopards are distributed across the mountains of 12 countries spread 1.8 million km2 in Central and South Asia. Previous efforts to map snow leopard distributions have relied on expert opinions modelling presence-only data. Expert opinion is subjective its reliability difficult assess, while analyses data tended ignore imperfect detectability this elusive species. The study was conducted prepare first ever probabilistic distribution Mongolia addressing challenge detection. Location We sign-based occupancy surveys 1,017 grid-cells covering 406,800 Mongolia's potential range. Methods Using a candidate model set 31 ecologically meaningful models that used six site seven sampling covariates, we estimate probability sites being by entire country. Results Occupancy increased with greater terrain ruggedness, lower values vegetation indices, less forest cover, were highest at intermediate altitudes. Detection higher for segments walked foot, those more rugged terrain. Our results showed broad agreement maps developed using but also highlighted important differences, example northern areas deemed largely unfavourable previous analyses. Main conclusions This reports national-level survey highlights methodological opportunities can be taken scale support conservation planning. assessments indicated <12% land mass, 42% habitat has high (>0.5) leopards. emphasize utility modelling, which jointly detection use, achieving these goals.