Wetland characteristics linked to broad-scale patterns in Culiseta melanura abundance and eastern equine encephalitis virus infection

Table 3 Best-performing Generalized Additive Model (GAM) estimating EEEV presence/absence in Cs. melanura. All explanatory variables included in the model are listed in the left-most column, including the most important spatial scale or temporal lag for that variable. The proportion of the deviance explained by each explanatory variable is listed in the right-most column. The proportion of the total deviance explained by the full model is listed in the left-most column header. For the sixth variable listed, there were two time periods when hydrological conditions were important. The variable for the second important time-period and the output values that change between models are listed in brackets

Response: EEEV Pres/Abs (Deviance explained: 0.49 [0.46])	χ ²	P	Deviance explained
Emergent Wetland 2000 m	8.2 [4.6]	0.01 [0.06]*	0.01
Evergreen Forested Wetland 1000 m	0 [1.4]	0.3 [0.1]	0
Deciduous Forested Wetland 5000 m	6.1 [15.6]	0.005 [< 0.001]*	0.03 [0.01]
Scrub/Shrub Wetland 5000 m	0 [0]	0.8 [1]	0
Average Forested Wetland Size 1500 m	2.1 [0.01]	0.08 [0.2]	0.01 [0.02]
Relative semi-permanent wetland 500 m* PHDI (10 lag) OR [Relative semi-permanent wetland 500 m* PHDI (0 lag)]	43.5 [24.7]	< 0.001*	0.06
Impervious Surfaces 200 m	3.7 [3.9]	0.1 [0.1]	0.01
Month	50.1 [51.6]	< 0.001*	0.17
Mean Cs. melanura abundance	70.9 [53.8]	< 0.001*	0.15
% Gravid Traps (linear)	-4.3 [−4.7] (z)	< 0.001*	0.05

ISSN: 1756-3305