Table 3.
Model (year) | AICc | R2 | ANOVA |
---|---|---|---|
OLS† | GWR† | OLS | GWR | F-value | |
2001 | 59.87 | 70.311 | 0.18 | 0.47 | 3.72* |
2002 | 66.77 | 71.613 | 0.37 | 0.53 | 3.89* |
2003 | 60.47 | 62.913 | 0.28 | 0.51 | 4.28* |
2004 | 51.017 | 56.7913 | 0.33 | 0.47 | 2.94 |
2005 | 54.67 | 64.011 | 0.16 | 0.49 | 3.85* |
2006 | 43.97 | 43.513 | 0.57 | 0.74 | 5.70* |
2007 | 46.97 | 72.58 | 0.41 | 0.86 | 9.29** |
Analysis of variance (ANOVA) tests the null hypothesis that the GWR model represents no improvement over the global OLS model. A Monte Carlo test compares the difference in the residual sums of squares of the OLS model with the residual sum of squares of the GWR model. R2 represents the adjusted coefficient of determination, and AICc is the corrected Akaike information criteria. Model variables were population density, terrain elevation, terrain slope, number of aquatic habitats, distance to the nearest breeding site, and distance to the main road.
P < 0.05; ** P < 0.01.
DF of OLS and GWR residuals, respectively.