1

Figure S1. Harvested aboveground biomass (g m-2) of grasses, forbs, sedges, and legumes displayed separately for the five N levels 4, 9, 14, 29, and 54 kg N ha-1 yr-1 and the seven experimental years 2004-2010 (4 kg N h-1 yr-1 = ambient deposition). For significance of the treatment effects see Tab. 1.

Figure S2. Relative contribution to the total harvested aboveground biomass of the three functional groups grasses, forbs, sedges, displayed for the five N levels (4, 9, 14, 29, and 54 kg N ha-1 yr-1) and the seven experimental years 2004-2010. Displayed are fitted lines based on the lme model (see Tab. 1).


Figure S3. Relative abundance of the 17 most dominant plant species and a pool of the remaining other 89 species as measured by the point quadrat method in the 36 monoliths of the a) N4 treatment in 2004, b) N4 treatment in 2010, c) N54 treatment in 2010, d) O3 control treatment 2010, and e) O3++ treatment in 2010 in grassland monoliths at Alp Flix.

Table S1. Effects of block, ozone (O3), nitrogen (N), time and selected interactions among them on the harvested aboveground biomass [g m-2] of grasses, forbs and sedges and on the proportion of grasses [%], forbs [%], and sedges [%] in the monoliths of the N4 and N9 treatment. Data were analysed with linear mixed regression, and inference on the variables was done using likelihood ratio tests ( statistic). Biomass and proportion of legumes was not analysed, since data were not normally distributed even after transformation. Initial soil pH (2003) is a continuous covariable. * P < 0.05, ** P < 0.01, *** P < 0.001

Variable / Grasses biomass / Forbsbiomass / Sedgesbiomass / Grassesproportion / Forbsproportion / Sedgesproportion
Df / / P / Df / / P / Df / / P / Df / / P / Df / / P / Df / / P
Block / 4 / 8.0 / * / 4 / 3.6 / 4 / 0.5 / 4 / 3.3 / 4 / 0.6 / 4 / 1.8
O3 / 4 / 2.5 / 4 / 0.8 / 4 / 1.1 / 4 / 1.5 / 4 / 1.4 / 4 / 1.1
N / 59 / 0.1 / 59 / 0.2 / 59 / 11.3 / ** / 59 / 3.6 / 59 / 0.0 / 59 / 8.2 / **
Time / 425 / 63.8 / *** / 424 / 117.3 / *** / 424 / 163.5 / *** / 425 / 51.9 / *** / 424 / 25.8 / *** / 424 / 47.9 / ***
pH / 59 / 11.5 / ** / 59 / 1.3 / 59 / 2.3 / 59 / 19.5 / *** / 59 / 6.3 / * / 59 / 4.3 / *
N x N
O3 x N / 59 / 0.3 / 59 / 4.9 / 59 / 1.2 / 59 / 1.6 / 59 / 5.4 / 59 / 1.2
N x Time / 425 / 5.2 / * / 424 / 1.9 / 424 / 18.2 / *** / 425 / 0.4 / 424 / 0.6 / 424 / 5.0 / *
O3 x Time / 425 / 4.9 / 424 / 3.7 / 424 / 1.4 / 425 / 14.4 / ** / 424 / 9.5 / ** / 424 / 1.1
Time x Time / 425 / 3.7 / 424 / 18.7 / *** / 424 / 69.8 / *** / 425 / 25.4 / *** / 424 / 7.3 / ** / 424 / 45.2 / ***
N x Time x Time / 425 / 1.4 / 424 / 3.3 / 424 / 1.3 / 425 / 0.1 / 424 / 3.9 / * / 424 / 4.2 / *

Table S2. Effects of block, ozone (O3), nitrogen (N), time and selected interactions among them on the number of point quadrat hits for the six most abundant grass species in monoliths of a free-air fumigation experiment at Alp Flix in 2004-2010. Data were analysed with linear mixed regression, and inference on the variables was done using likelihood ratio tests ( statistic). Initial soil pH (2003) is a continuous covariable. * P < 0.05, ** P < 0.01, *** P < 0.001

Variable / Agrostis capillaris / Anthoxanthum alpinum / Briza media / Festuca spp. / Helictotrichon versicolor / Nardus stricta
Df / / P / Df / / P / Df / / P / Df / / P / Df / / P / Df / / P
Block / 4 / 0.2 / 4 / 1.7 / 4 / 0.9 / 4 / 2.4 / 4 / 3.3 / 4 / 0.5
O3 / 4 / 0.8 / 4 / 7.7 / * / 4 / 4.5 / 4 / 1.3 / 4 / 4.3 / 4 / 7.2 / *
N / 31 / 5.9 / * / 31 / 0.7 / 31 / 5.9 / * / 31 / 0.1 / 31 / 0.2 / 31 / 0.4
Time / 264 / 3.2 / 264 / 0.0 / 262 / 66.3 / *** / 264 / 8.5 / ** / 264 / 4.0 / * / 264 / 0.8
pH / 31 / 1.1 / 31 / 1.9 / 31 / 0.5 / 31 / 1.3 / 31 / 1.5 / 31 / 7.1 / **
N x N / 31 / 0.2 / 31 / 0.2 / 31 / 0.2 / 31 / 0.4 / 31 / 0.0 / 31 / 0.6
O3 x N / 31 / 0.3 / 31 / 2.8 / 31 / 1.7 / 31 / 3.7 / 31 / 0.4 / 31 / 0.2
N x Time / 264 / 0.0 / 264 / 0.4 / 262 / 21.8 / *** / 264 / 8.3 / ** / 264 / 0.0 / 264 / 4.0 / *
O3 x Time / 264 / 0.6 / 264 / 0.6 / 262 / 18.7 / *** / 264 / 1.2 / 264 / 0. 6 / 264 / 9.2 / *
Time x Time / 264 / 15.4 / *** / 264 / 0.8 / 262 / 18.5 / *** / 264 / 2.8 / 264 / 1.1 / 264 / 4.4 / *
N x Time x Time / 264 / 2.9 / 264 / 1.1 / 262 / 0.1 / 264 / 0.1 / 264 / 0.1 / 264 / 0.1

Table S3. Effects of block, ozone (O3), nitrogen (N), time and selected interactions among them on the number of point quadrat hits for the eight most abundant forb species in monoliths of a free-air fumigation experiment at Alp Flix in 2004-2010. Data were analysed with linear mixed regression, and inference on the variables was done using likelihood ratio tests ( statistic). Initial soil pH (2003) is a continuous covariable. * P < 0.05, ** P < 0.01, *** P < 0.001

Variable / Arnica montana / Gentiana acaulis / Homogyne alpina / Leontodon helveticus / Ligusticum mutellina / Plantago alpina / Potentilla aurea / Ranunculus villarsii
Df / / P / Df / / P / Df / / P / Df / / P / Df / / P / Df / / P / Df / / P / Df / / P
Block / 4 / 1.4 / 4 / 1.4 / 4 / 0.7 / 4 / 2.5 / 4 / 0.9 / 4 / 1.1 / 4 / 0.8 / 4 / 4.1
O3 / 4 / 2.5 / 4 / 2.0 / 4 / 0.4 / 4 / 1.1 / 4 / 1.8 / 4 / 0.6 / 4 / 0.0 / 4 / 4.6
N / 31 / 2.7 / 31 / 0.7 / 31 / 0.7 / 31 / 2.4 / 31 / 0.0 / 31 / 4.2 / * / 31 / 5.7 / * / 31 / 2.1
Time / 264 / 26.0 / *** / 264 / 0.4 / 264 / 87.5 / *** / 264 / 10.6 / ** / 264 / 7.9 / * / 264 / 43.1 / ** / 264 / 12.2 / ** / 264 / 4.7 / *
pH / 31 / 10.7 / ** / 31 / 7.2 / ** / 31 / 1.0 / 31 / 11.9 / ** / 31 / 6.1 / * / 31 / 13.7 / *** / 31 / 10.1 / ** / 31 / 0.1
N x N / 31 / 0.6 / 31 / 0.4 / 31 / 0.2 / 31 / 0.5 / 31 / 0.2 / 31 / 0.1 / 31 / 0.1 / 31 / 2.8
O3 x N / 31 / 2.4 / 31 / 1.7 / 31 / 3.4 / 31 / 2.5 / 31 / 5.2 / 31 / 2.7 / 31 / 5.4 / 31 / 1.1
N x Time / 264 / 1.5 / 264 / 0.6 / 264 / 0.0 / 264 / 1.1 / 264 / 0.1 / 264 / 2.4 / 264 / 0.5 / 264 / 0.2
O3 x Time / 264 / 1.0 / 264 / 1.2 / 264 / 0.4 / 264 / 9.1 / * / 264 / 7.1 / 264 / 8.8 / * / 264 / 9.4 / ** / 264 / 4.4
Time x Time / 264 / 2.6 / 264 / 22.1 / *** / 264 / 19.9 / *** / 264 / 20.1 / *** / 264 / 2.1 / 264 / 7.1 / ** / 264 / 21.3 / *** / 264 / 1.7
N x Time x Time / 264 / 2.3 / 264 / 3.0 / 264 / 0.0 / 264 / 1.0 / 264 / 4.8 / * / 264 / 0.0 / 264 / 0.0 / 264 / 0.0

Table S4. Effects of block, ozone (O3), nitrogen (N), time and selected interactions among them on the number of point quadrat hits for the two most abundant sedge species and one legume species in monoliths of a free-air fumigation experiment at Alp Flix in 2004-2010. Data were analysed with linear mixed regression, and inference on the variables was done using likelihood ratio tests ( statistic). Initial soil pH (2003) is a continuous covariable. * P < 0.05, ** P < 0.01, *** P < 0.001

Variable / Carex sempervirens / Carex ornithopoda / Trifolium alpinum
Df / / P / Df / / P / Df / / P
Block / 4 / 0.0 / 4 / 1.6 / 4 / 1.0
O3 / 4 / 0.8 / 4 / 2.4 / 4 / 0.0
N / 31 / 20.8 / *** / 31 / 10.1 / ** / 31 / 6.7 / *
Time / 264 / 20.9 / *** / 264 / 0.0 / 264 / 10.4 / **
pH / 31 / 0.9 / 31 / 2.1 / 31 / 0.2
N x N / 31 / 0.5 / 31 / 0.8 / 31 / 0.4
O3 x N / 31 / 0.8 / 31 / 8.3 / * / 31 / 1.2
N x Time / 264 / 35.5 / *** / 264 / 2.9 / 264 / 0.2
O3 x Time / 264 / 2.6 / 264 / 1.2 / * / 264 / 1.4
Time x Time / 264 / 22.2 / *** / 264 / 23.5 / *** / 264 / 0.1
N x Time x Time / 264 / 3.3 / 264 / 9.5 / ** / 264 / 0.0

Methodological aspects

The high number of species present made the analysis of vegetation data difficult. Analyzing species abundance univariately by means of linear mixed effect models was not feasible for 75% of the species due to their uneven distribution among monoliths. Alternatively, multivariate techniques, namely Principle Response Curves (Van den Brink & Ter Braak, 1999) were performed tentatively on the complete species data set, but revealed only a coarse picture of what was found by means of univariate models. Finally, univariate linear mixed effect models were used with the aim to detect species-specific differences in the treatment response patterns.

VanDen Brink PJ, Ter Braak CJF. 1999. Principal response curves: analysis of time-dependentmultivariate responses of biologicalcommunity to stress. Environ ToxicolChem 18:138–48.