Table 2. Model parameter estimates (with constant abundance) for the detection probability (p) of sloth bear, obtained from the N-mixture model in Sanjay Tiger Reserve, Madhya Pradesh, India, during 2016–2017.
| Model description | β estimates | SE (Standard error) | P value | AIC | p | SE (Standard error) |
|---|---|---|---|---|---|---|
| Poisson | ||||||
| [p(fruit_density)λ(.)] | 0.27 | 0.11 | 0.01 (<0.05) | 1,020.22 | 0.08 | 0.01 |
| [p(shrub_density)λ(.)] | 0.16 | 0.14 | 0.25 (>0.05) | 1,025.15 | ||
| [p(termitemounds)λ(.)] | 0.09 | 0.11 | 0.42 (>0.05) | 1,025.85 | ||
| [p(trapdays_total)λ(.)] | 0.03 | 0.14 | 0.80 (>0.05) | 1,026.42 | ||
| NB | ||||||
| [p(fruit_density)λ(.)] | 0.28 | 0.14 | 0.04 (<0.05) | 991.86 | 0.01 | 0.004 |
| [p(shrub_density)λ(.)] | 0.09 | 0.17 | 0.57 (>0.05) | 995.44 | ||
| [p(termitemounds)λ(.)] | 0.08 | 0.15 | 0.59 (>0.05) | 995.47 | ||
| [p(trapdays_total)λ(.)] | 0.07 | 0.17 | 0.68 (>0.05) | 995.59 | ||
| ZIP | ||||||
| [p(fruit_density)λ(.)] | 0.29 | 0.11 | 0.007 (<0.05) | 1,017.04 | 0.06 | 0.02 |
| [p(shrub_density)λ(.)] | 0.16 | 0.13 | 0.22 (>0.05) | 1,022.56 | ||
| [p(termitemounds)λ(.)] | 0.10 | 0.11 | 0.377 (>0.05) | 1,023.30 | ||
| [p(trapdays_total)λ(.)] | −0.04 | 0.16 | 0.81 (>0.05) | 1,024.04 |
Note:
Three distributions [Poisson, negative binomial (NB) and zero-inflated Poisson (ZIP)] were used to model the detection probability of sloth bear; Covariates considered: fruit_density–density of fruiting trees per 10 m-radius plot around each camera trap location, shrub_density–density of shrubs per 5 m-radius plot around each camera trap location, termitemounds- number of termite mounds per 10 m-radius plot around each camera trap location, trapdays_total–total number of days or occasions for which camera trap (s) was active; AIC, Akaike Information Criterion; Detection probability estimates (p) were provided for the best-selected models of each distribution.