**************************************************************************************************************************** Data used for to fit linear mixed-effect models (package lme4, R v. 3.2.3) to assess the effect of nitrogen addition, warmer temperatures and browsing intensity on radial growth of Betula glandulosa (Table 3, figure 3) **************************************************************************************************************************** *Protocol* We conducted factorial manipulations of growing season temperature, soil nitrogen and browsing intensity over 5 years (2009-2013) in a split-plot experimental design (Fig. 1). The design included two levels of nitrogen (with or without nitrogen addition) in primary plots (4 m x 24 m), and two levels of temperature (ambient and warmed) crossed with 3 levels of summer caribou browsing (control: 0%, moderate: 25% and heavy: 75% of available shoots stripped of their leaves) in 1 m x 1 m quadrats in the center of secondary plots (4 m x 4 m). Each combination of treatments (12) was replicated in five exclosures designed to exclude caribou (12 x 26 m with a 1.5 m fence; hereafter refers as blocks) for a total of 60 secondary plots. In August 2014, we cut three to six live birch primary stems (i.e. emerging from the ground) at their base in each secondary plot (60), for a total of 202 stems. Although the stems were sampled inside 1m2 plots, it was not possible to ascertain if they were or not from the same individuals. Samples were dried at room temperature for at least three months. Afterwards, the samples were boiled for at least 4 hours before being sliced with a sledge microtome (ca. 20 µm; WSL-Core-Microtome, Zürich, Switzerland), stained with safranin (1% solution, Safanin O; Fischer Science Education, Hanover Park, IL, USA), dried and permanently mounted with a 66% toluene solution (SHUR/mountTM liquid cover glass; Triangle biomedical sciences, Cincinnati, OH, USA). We then took digital photographs of each sample with a binocular-mounted camera (Olympus SZ61 with a SC100 camera, Richmond Hill, ON, Canada). Ring widths of each stem were measured with LIGNOVISION, a dendrochronological software (v. 1.36; Rinntech, Heidelberg, Germany). When possible (198/202 stems), ring widths were measured along two radii for each stem. The two radii were than cross-dated with Dendro 2015 (v. 2.5.4; Centre d’études nordiques, QC, Canada), a statistical cross-dating program. When growth rings were partially missing or too narrow to be measured, we assigned them a width of 1 µm, i.e. the lower limit of precision, to obtain accurate chronologies in relation to age (Stokes and Smiley 1968). _____________________________________________________________________________________________________ *Definitions of column headings* block fert temp brows plot stem id year gr A) block: Blocking factor. Each combination of treatments (12) was replicated in 5 exclosures designed to exclude caribou (12 x 26 m with a 1.5 m fence; hereafter refers as blocks) Value A, B, C, D, E B) fert: Fertilization treatment level Value: N (with nitrogen addition) or S(without nitrogen addition) C) temp: Temperature enhancement treatment level Value: H (warmed with open top chambers), L (ambient temperature) D) brows: Simulation of browsing by caribou treatment level Value:0 (control: 0%), 1 moderate: 25% of available shoots stripped of their leaves) or 3 (heavy: 75% of available shoots stripped of their leaves) E) plot: Exclusive ID of each secondary plot Value: Concatenation of block|Fertilization|Temp|Browsing F) stem: ID of individuals within a plot Value: 1 to 10 G) ID: Individual ID from each stem in the experiment Value: Concatenation of block|Fertilization|Temp|Browsing|stem H) year: Value: 1966 to 2013 I) gr: width of each annual growth rings in µm Value: 1 to 350 µm