Until the early 1980s the only surviving population of the greater one-horned rhinoceros Rhinoceros unicornis in Nepal was in Chitwan National Park. Between 1986 and 2003 87 rhinoceroses from Chitwan were translocated into Bardia National Park and Suklaphanta Wildlife Reserve in the western terai region to establish founder populations and reduce the threat of local extinction from natural catastrophic events, disease and/or poaching. The founder populations increased in number through births but a rise in poaching during the period of civil strife in Nepal during 1996–2006 resulted in a dramatic decline in the populations, including in Chitwan. In 2001 the Terai Arc Landscape programme was initiated to connect 11 protected areas in Nepal and north-west India and facilitate dispersal of megafauna and manage them as metapopulations. Corridors that were restored under the programme and that connect Bardia and Suklaphanta with protected areas in India are now used by the greater one-horned rhinoceros. The successes and failures of the last 2 decades indicate that new paradigms for protecting rhinoceroses within and outside protected areas are needed, especially with reference to managing this species at a landscape scale.

We assessed the abundance and distribution of the greater one-horned or Indian rhinoceros Rhinoceros unicornis in all its potential habitats in Nepal, using block counts. In April 2011 5,497 km were searched in 3,548 elephant-hours over 23 days. The validity of the block count was assessed by comparing it with counts obtained from long-term monitoring using photographic identification of individual rhinoceroses (ID-based), and estimates obtained by closed population sighting–mark–resighting in the 214 km2 of Chitwan National Park. A total of 534 rhinoceroses were found during the census, with 503 in Chitwan National Park (density 1 km−2), 24 in Bardia National Park (0.28 km−2) and seven in Suklaphanta Wildlife Reserve (0.1 km−2). In Chitwan 66% were adults, 12% subadults and 22% calves, with a female : male ratio of 1.24. The population estimate from sighting–mark–resighting was 72 (95% CI 71–78). The model with different detection probabilities for males and females had better support than the null model. In the Sauraha area of Chitwan estimates of the population obtained by block count (77) and ID-based monitoring (72) were within the 95% confidence interval of the estimate from sighting–mark–resighting. We recommend a country-wide block count for rhinoceroses every 3 years and annual ID-based monitoring in a sighting–mark–resighting framework within selected subpopulations. The sighting–mark–resighting technique provides the statistical rigour required for population estimates of the rhinoceros in Nepal and elsewhere.