Table 5. Comparison of the top fitting and null models in occupancy modeling under each survey method for species detected at southern Illinois, Dec 2015 –Feb 2016.
| Species | Method | Modela | Kb | AICc | ΔAICd | ωe |
|---|---|---|---|---|---|---|
| Bobcat | One Camera | Ψ (EDGE + ELEVATION) p (FULL + PRECIP) | 6 | 17.63 | 0.00 | 0.618 |
| NULLf | 2 | 30.95 | 13.32 | 0.001 | ||
| Two Cameras | Ψ (EDGE + ELEVATION) p (FULL X PRECIP) | 5 | 24.28 | 0.00 | 0.869 | |
| NULL | 2 | 36.14 | 11.86 | 0.002 | ||
| Four Cameras | Ψ (.) p (FULL X PRECIP) | 3 | 48.92 | 0.00 | 0.147 | |
| NULL | 2 | 53.00 | 4.08 | 0.019 | ||
| Coyote | One Camera | Ψ (UNDERSTORY + ELEVATION) p (TEMP X PRECIP) | 5 | 30.05 | 0.00 | 0.195 |
| NULL | 2 | 36.14 | 6.09 | 0.009 | ||
| Two Cameras | Ψ (UNDERSTORY + ELEVATION) p (TEMP X PRECIP) | 5 | 30.05 | 0.00 | 0.195 | |
| NULL | 2 | 36.14 | 6.09 | 0.009 | ||
| Four Cameras | Ψ (BA + ELEVATION) p (.) | 4 | 48.51 | 0.00 | 0.281 | |
| NULL | 2 | 53.00 | 4.49 | 0.029 | ||
| Virginia Opossum | One Camera | Ψ (EDGE + SLOPE) p (PRECIP + TEMP) | 6 | 26.92 | 0.00 | 0.760 |
| NULL | 2 | 46.40 | 19.18 | 0.000 | ||
| Two Cameras | Ψ (ELEVATION) p (PRECIP + TEMP) | 5 | 45.15 | 0.00 | 0.126 | |
| NULL | 2 | 53.00 | 7.85 | 0.002 | ||
| Four Cameras | Ψ (EDGE) p (PRECIP + TEMP) | 5 | 54.74 | 0.00 | 0.135 | |
| NULL | 2 | 68.98 | 14.24 | 0.000 | ||
| White-tailed Deer | One Camera | Ψ (BA + EDGE) p (.) | 4 | 58.25 | 0.00 | 0.552 |
| NULL | 2 | 67.63 | 9.38 | 0.005 | ||
| Two Cameras | Ψ (EDGE + SLOPE) p (.) | 4 | 82.99 | 0.00 | 0.960 | |
| NULL | 2 | 98.82 | 15.83 | 0.000 | ||
| Four Cameras | Ψ (SLOPE + ELEVATION) p (FULL X PRECIP) | 5 | 110.53 | 0.00 | 0.326 | |
| NULL | 2 | 115.98 | 5.45 | 0.021 |
Models were built in the statistical software R package ‘unmarked’[38]. Each survey method was offered identical detection and occupancy covariates. The top-ranked model for each species-specific survey method is given followed by the null model.
a BA, amount of basal area per camera-site; EDGE, distance to TTSF boundary; ELEVATION, meters above sea level at camera-site; FULL, unique detection probability per survey event; PRECIP, sum of precipitation recorded during survey period; SLOPE, the landscape grade of the camera-site; TEMP, average temperature recorded during survey period; UNDERSTORY, number of woody stems taller than 1.37 m and less than 5 cm dbh; (.), fixed, constant parameter among camera-sites or surveys.
bNumber of model parameters
cAkaike's Information Criterion
dChange in AIC value from the top-ranked model.
eModel weight; the probability of a model being the best approximating model among those evaluated.
fΨ(.)p(.)