Table 4. Mixed effect models conducted on theg1 estimates. Square root transformation was applied to meet ANOVA assumptions. Campaign representing day of measurement was used as random variable. Experimental factors are represented as follow: sites (Cloquet Forestry Center – in Cloquet, MN and Hubachek Wilderness Research Center – Ely, MN), warming (ambient T and +3.3°C), reduced rainfall (ambient and ~40% of summer rainfall removed), species (see Table 3). The environmental treatment (i.e., open and closed canopy) are analyzed separately.