Vegetation cover, plant mortality and leaf nitrogen content
The vegetation cover before the start of the experiment was similar among plots (p >0.05 for all species, Table S2). After 6 years of starting the experiment, shrub cover did not change (Table 3), but it tended to be lower in the treatments with nitrogen addition compared to the control (Fig. 2A). The +N and +NW treatments increased the percentage of dead shrub (p =0.02 and p =0.007 with respect to the control, respectively), being higher in the year prior to the extreme rain event of 2017 compared to the following years (p <0.01) (Table 3, Fig. 2B).
The grass cover after 6 years under resources addition (water and nitrogen) was similar among treatments, however, nitrogen addition modified grass species composition (Table 3, Fig. 3). The +N and +NW treatments resulted in a decrease in P. ligularis grass cover (p <0.0001) and an increase in P. humilis grass cover (p <0.0001), with no change in P. speciosacover (p>0.05), although this species also tended to decrease in the +N and +NW plots (Fig. S4). P. ligularis tussock size in the treatments with nitrogen addition was also significantly reduced compared to the control (p =0.03) (Fig. S5, Table S3). The +W treatment had no effect on tussock size (p>0.05) or grass cover (p >0.05). The +NW treatment had no differential effect with respect to the effect of adding nitrogen alone (p >0.05).
All studied species (shrubs and grasses) significantly increased the leaf nitrogen content in the nitrogen addition treatments (Table S4).Adesmia volckmannii and P. ligularis were the species with the highest and the lowest leaf nitrogen content, respectively, in the control plots. The higher change in leaf nitrogen content was observed in P. ligularis (1.08% in C and 2.41% in +N) and the lower change was found in A. volckmannii (2.36% in C and 2.81% in +N).