Table 2.
Summary of baseline hedgehog occupancy models where detection rate was modelled as constant (did not vary between weeks within each season) versus survey specific (did vary between weeks within each season). Seasons are illustrated in Figure 2.
| Season | Model | QAIC | ΔQAIC | AIC Weight | Model Likelihood | K | Detection Rate | Naïve Ψ | True Ψ |
|---|---|---|---|---|---|---|---|---|---|
| Autumn | Ψ(.), p(survey-specific) | 270.59 | 0.00 | 1 | 1.0000 | 8 | 0.8234 | 0.5397 | 0.5403 |
| 0.8225 | |||||||||
| 0.8225 | |||||||||
| 0.6756 | |||||||||
| 0.6756 | |||||||||
| 0.2938 | |||||||||
| 0.2938 | |||||||||
| Ψ(.), p(.) | 294.26 | 23.67 | 0.0000 | 0.0000 | 2 | 0.6138 | 0.5397 | 0.5411 | |
| Winter | Ψ(.), p(.) | 213.63 | 0.00 | 0.9852 | 1.0000 | 2 | 0.2626 | 0.3016 | 0.3224 |
| Ψ(.), p(survey-specific) | 222.02 | 8.39 | 0.0148 | 0.0151 | 10 | 0.2481 | 0.3016 | 0.3198 | |
| 0.2481 | |||||||||
| 0.2481 | |||||||||
| 0.397 | |||||||||
| 0.1489 | |||||||||
| 0.1489 | |||||||||
| 0.3474 | |||||||||
| 0.1489 | |||||||||
| 0.4467 | |||||||||
| Spring | Ψ(.), p(.) | 64.13 | 0.00 | 0.8006 | 1.0000 | 2 | 0.6459 | 0.3810 | 0.3870 |
| Ψ(.), p(survey-specific) | 66.91 | 2.78 | 0.1994 | 0.2491 | 5 | 0.5377 | 0.3810 | 0.3838 | |
| 0.5377 | |||||||||
| 0.6204 | |||||||||
| 0.9100 |
Ψ = occupancy, p = detection probability, K = number of parameters. ΔQAIC is the change in quasi-likelihood adjusted Akaike’s information criterion. For each season, the variance inflation factor ĉ was adjusted based on goodness-of-fit tests of the most parameterised models (1.3226, 1.3385, and 3.5534 for autumn, winter, and spring, respectively). Naïve occupancy is the number of gardens where hedgehogs were detected and true occupancy is the number of gardens estimated to be occupied by hedgehogs after accounting for the false-absence error rate.