Environmental stress may favour facilitative interactions among plants but whether these interactions are positive for the benefactor and how this depends on stress factors, remains to be determined. We studied the effect of beneficiary cover and biomass on reproduction of the benefactor cushion plant Laretia acaulis (Apiaceae) in the central Chilean Andes during three years. Study sites were situated along an elevational gradient at 2600, 2800, 3000 and 3150 m a.s.l. This range comprises a cold- and a drought-stress gradient, with moisture increasing and temperature decreasing with elevation. We studied the effect of natural gradients in beneficiary cover and of experimental cover removal on cushion flower and fruit production. Beneficiary cover had a negative effect on flower production but not on infructescence and fruit densities or fruit weights. A positive effect of beneficiaries on the fraction of flowers converted into fruits was detected for hermaphrodite cushions. The effect of beneficiary cover on flowering was independent of elevation or cushion gender, although these latter factors explained most of the variation. Removing the aboveground parts of the beneficiaries positively affected flowering at 2800 m a.s.l. but not at the other elevations. Our results suggest negative effects of facilitation on L. acaulis flowering, but these are neutralized in fruit production. Surprisingly, this conclusion holds along the entire elevational or stress gradient. This suggests that this system of facilitation is evolutionarily stable and not very sensitive to environmental change. It remains to be tested, however, whether facilitation affects fitness via growth and long-term survival in these slow-growing alpine cushions.

Data set of Laretia acaulis reproductive outputs in 2012, 2013, and 2014 in the high Andes of central ChileDetails of the spreadsheet: 1. Data set of Laretia acaulis reproductive output in 2012, 2013, and 2014 of intact cushions. See also Fig. 1 of the original article. 2. Dataset of Laretia acaulis reproductive output in 2012 as functions of A) beneficiary cover (n=80, but n=60 at 2800 m a.s.l.) or B) biomass (n=40, only 2600, 3000 and 3150 m a.s.l.). For a statistical summary see Tables 1 and 2, and Fig. 2 and 3 of the orginal article. 3. Data set of Laretia acaulis reproductive output as a function of experimental beneficiary removal. A) for three elevations (2600, 3000 and 3150 m), one or two years after removal, B) for 2800 m a.s.l., one, two or three years after removal. For a statistical summary see Table 3 andFig. 4 of the original article.15.06.09_Dataset.xlsx