Functional Ecology Appendices Appendix S1. Soil water content along the topographical gradient, from upper (1) to lower slope (40). Soil water content was measured twice (May 2013 and May 2014 ) on each quadrat with a ThetaProbe ML2 sensor and a Theta HH1 Meter (Delta OHM, Caselle di Selvazzano, Italy), and the two values for each plot averaged. To show the relationship between soil water content (expressed here as a percentage: 100 m3 m-3), we fitted a linear regression of soil water content on position on the gradient, which revealed an important increase of soil water content along the studied gradient. More for less: sampling strategies of plant functional traits across local environmental gradients. Carmona, CP., Rota, C., Azcárate, F.M. & Peco, B. Functional Ecology Appendix S2. Community weighted mean (CWM) and functional diversity (FD) along the topographical gradient, from upper (1) to lower slope (40). Black dots and lines indicate the values of these parameters obtained considering 10 individuals per species per quadrat (labour-intensive strategy) and the model fitted to these data, respectively. For each of the considered strategies, we present the envelopes that contain the 95% of the predicted values (considering 1,000 simulations) for three different sampling strategies. Note that for the GLO and MAX strategies, greater sampling intensities yielded more precise (i.e. narrower envelopes), but not more accurate (i.e. closer to the regressions estimated using the labor intensive strategy) predictions. More for less: sampling strategies of plant functional traits across local environmental gradients. Carmona, CP., Rota, C., Azcárate, F.M. & Peco, B. Functional Ecology Appendix S3. Boxplots representing the error of the different intensities for the LOC strategy. For each trait and indicator of functional structure (CWM and FD), we calculated the ratio between the CWM and FD values obtained following the different sampling strategies and the CWMI and FDI of each quadrat. Error is the absolute difference between 1 and this ratio. Values close to 0 indicate a low error (i.e. simulated values are very close to the ones attained using the labour-intensive strategy). The dashed line represents a 10% level of error. Boxplots are ordered from lower (top) to higher (bottom) sampling intensities. Numbers indicate the number of individuals and species per sampling unit (in parentheses, the total number of traits measured in the whole gradient for each sampling intensity). Note that the greatest increases in accuracy between consecutives sampling intensities happened at low levels of intensity, and successive increases in intensity resulted in smaller improvements. a) CWM Height Leaf Area SLA 1 (257) (257) (257) 2 (514) (514) (514) 3 (771) (771) (771) 4 (1028) (1028) (1028) 5 (1285) (1285) (1285) 6 (1542) (1542) (1542) 7 (1799) (1799) (1799) 8 (2056) (2056) (2056) 9 (2313) (2313) (2313) 0 5 10 15 0 1 2 3 4 5 6 0 2 4 6 8 10 0 1 2 3 4 5 b) FD 1 (257) (257) (257) 2 (514) (514) (514) 3 (771) (771) (771) 4 (1028) (1028) (1028) 5 (1285) (1285) (1285) 6 (1542) (1542) (1542) 7 (1799) (1799) (1799) 8 (2056) (2056) (2056) 9 (2313) (2313) (2313) 0 2 4 6 8 0 1 2 3 4 More for less: sampling strategies of plant functional traits across local environmental gradients. Carmona, CP., Rota, C., Azcárate, F.M. & Peco, B. Functional Ecology Appendix S4. Boxplots representing the reductions in error attained using the example of EFF strategy proposed in the main text (2 individuals per species and quadrat if the cover of the species in the quadrat is <25% and 4 individuals otherwise). This strategy involves the selection of 2.26 individuals per species and sampling unit. We also represent two levels of the LOC strategy (2 and 4 individuals per species and sampling unit). Boxplots are placed in the axis of ordinates according to decreasing (bottom to top) sampling intensity. Numbers in parentheses indicate the total number of traits measured in the whole gradient. Dashed grey lines indicate the expected values of error that would be attained for any given number of individuals measured (vertical axis), assuming a constant rate of improvement between the two selected intensities of the LOC strategy. Note that, with the exception of FD for SLA, the EFF strategy yields more accurate values than expected for its intensity level. a) CWM Height Leaf Area SLA LOC 2 (514) (514) (514) EFF 2.26 (580) (580) (580) LOC 3 (771) (771) (771) 0 5 10 15 20 0 1 2 3 4 5 6 7 0 2 4 6 8 10 12 b) FD LOC 2 (514) (514) (514) EFF 2.26 (580) (580) (580) LOC 3 (771) (771) (771) 0 2 4 6 8 0 1 2 3 4 0 1 2 3 4 5 More for less: sampling strategies of plant functional traits across local environmental gradients. Carmona, CP., Rota, C., Azcárate, F.M. & Peco, B.