Table 2.

Goodness of fit statistics and improvements in AIC for 90% quantile regression models demonstrate a higher predictive value of deposition with climate compared to climate alone. Smaller AICs indicate better fit (ΔAIC must be >25). The larger the ΔAIC, the more improvement in the model compared to other models for that metric. R1 measures absolute model fit whereas AIC measures the relative fit of nested models (i.e., models in the same row). Model predictor variables: Dep = deposition only, Clim = all climate variables only, Dep + Clim = deposition and all climate variables.

	Dep 1		Clim		Dep + Clim 2		(Dep + Clim) − Clim
Lichen Metric	R1	AIC	R1	AIC	R1	AIC	Δ AIC2
Nitrogen Models
Species richness	0.11	33265	0.09	33425	0.16	32807	619
Oligotroph	0.19	28672	0.11	29520	0.26	27887	1633
Cyanolichens	0.19	29465	0.21	29202	0.28	28419	783
Forage lichens	0.26	31982	0.14	33291	0.29	31626	1665
Sulfur Models
Species richness	0.07	40123	0.07	39997	0.13	39308	689
Sensitive	0.23	32588	0.02	35055	0.25	32340	2715
Cyanolichens	0.08	37265	0.16	36297	0.22	35560	737
Forage lichens	0.19	39312	0.11	39562	0.23	38026	1536

¹Models used to assess risk.

²Models used to explore the behavior of the metric under variable (instead of optimized) climate.