Results of the generalised linear model deviance analysis on the number N of triatomines (R. ecuadoriensis) infected by T. cruzi.

The model was: N∼Date+habitat+host+nest height+house spraying+date×habitat+date×host+habitat×host+nest height×habitat. For each effect variable, Δ AIC corresponds to the difference between the AIC of the initial model and that of reduced model. Likelihood-ratio test (LRT) and associated P-value test the hypothesis that the suppression of the ‘effect’ term provides no better fit than the initial model.AIC = Akaike's information criterion of the initial model after the removal of the ‘effect’ term.