Table 1.

Results of the full and reduced GLMs used to assess how landscape variables impacted adelgid numbers

Variables	Estimate	Standard error	z-value	p
Full model
(Intercept)	18.284	5.489	3.330	8.67 × 10−4
Northeast	1.152	0.514	2.24	0.025
East	−0.66	0.565	−1.169	0.242
Southeast	0.395	0.538	0.735	0.462
South	0.045	0.584	0.077	0.938
Southwest	0.597	0.492	1.213	0.225
West	0.169	0.478	0.354	0.723
Northwest	0.672	0.466	1.442	0.149
Slope	0.006	0.046	0.134	0.893
Elevation	−0.077	0.029	−2.597	0.009
EucDist	−0.004	8.21 × 10−4	−5.263	1.42 × 10−7
(AIC = 832.4)
Reduced model
(Intercept)	17.295	4.974	3.477	5.07 × 10−4
Elevation	−0.069	0.027	−5.139	2.76 × 10−7
EucDist	−0.004	8.04 × 10−4	−2.624	0.009
(AIC = 830.31)

The full model used trap elevation, slope, aspect, and Euclidean distance (EucDist) to the nearest HWA-infested hemlock tree as explanatory variables for adelgid number caught by a motorized trap. The reduced model used only trap elevation and EucDist as explanatory variables for the number of adelgids caught. For the aspect variable, North was considered the reference variable in the GLM.