Description of Experiment and Data Collection This experiment was conducted at the Danum Valley Field Centre (N04.96° E117.80°; 180 MASL), Sabah, Malaysia. The field centre is part of the Danum Valley Conservation Area (DVCA), a 43800-ha area of primary lowland forest dominated by the Family Dipterocarpaceae (dipterocarps). The experiment was planted in November 2004 throughout the primary forest south and west from the field centre. Twenty plots were placed across a large size range (12 – 400 m2) of natural canopy openings (gaps) and paired with identical plots under closed canopy (understorey) Understorey sites were placed 30 meters from each tree-fall gap at a random compass bearing. The combined gap and understorey plots represent a block. Blocks were placed across the range of topography found in Danum Valley whereby 6 blocks were in lowland riparian areas, 7 blocks were in lowland areas away from rivers and 7 blocks were on upslope sites (Fig. S1 in the Supplementary Information). The mean elevation across all plots was 249 m (range = 203 – 319 m) and the mean slope was 20 degrees (range = 3 – 56 degrees). Within each gap and understorey plot, two identical subplots were planted with seedlings of 25 tree species at 0.75 × 0.75 m spacing for a total of four subplots per gap-understorey pair and a total of 80 subplots (20 pairs × 2 gap/understorey plots × 2 subplots = 80 subplots). Due to the infrequent seed production of tree species in lowland Borneo forests, species were selected mainly based on availability in local nurseries (see Table S3 in Supplementary Information for species information). Twenty-four species were dipterocarps spanning five genera (Dipterocarpus, Dryobalanops, Hopea, Parashorea and Shorea) and one species of Bombaceae (Durian graveolens). Within each subplot, species were randomly assigned a planting point, therefore, all 80 subplots had a different planting pattern. Six months after planting—to allow for mortality due to planting stress (which was replanted after one and three months—seedlings were measured for height and diameter at base (10 cm from soil surface). From 2004 to 2007 seedlings were measured every 6 months after which they were measured approximately yearly from 2007 to 2017. One census was not included in this analysis because as it was measured only 3 months after the previous census, which led to a total of 16 censuses. These measurements of growth were used to calculate relative growth rate as the difference in log-diameter between two measurements divided by the number of days between measurements standardized to year by dividing days by 365.25. Description of each column in DataGrowthSurvivalClean.txt Block - Gap/understorey pair (Categorical variable from 1 to 20) pid - Plot ID (Categorical variable of 40 plots) Survey.date - Date of measurement (DD/MM/YYYY) Day - Julian day Year - Year of census Census - Census number Canopy - Gap or understorey designation (Categorical variable U/G) Plot - Plot number (Categorical variable from 1 to 40) Subplot - Subplot within plot (Categorical variable A/B) Subplot.no - Subplot number (Categorical variable from 1 to 80) Genus - Genus Species - Specific epithet Sp - Unique ID of first 3 letters of genus and first 3 letters of specific epithet Plant.id - ID for each individual (Categorical variable from 1 to 2000) Diam1 - Diameter at base measurement Diam2 - Diameter at base measurement DBH.1 - Diameter at breast height of taller than 1.3 m DBH.2 - Diameter at breast height of taller than 1.3 m Height.apex - Height to apical bud diam - Average of two diameter measurements dbh - Average of two DBH measurements light - Percent light in the plot measured at the start of the experiment deficit - Minimum 30-day average of rainfall deficit since last census zero - Number of days with no rain since last census dry - Number of days of the longest dry period since last census excess - Maximum 30-day average of rainfall excess since last census sun - Average sun hour per day since last census rain - Designation of rainfall conditions since last census (Wet > -50 mm deficit, Dry -100 mm deficit, Drought <-100 mm deficit) treat - Designation if plastic sheeting was applied to plot for the period (Everwet no plastic, Drought plastic rainfall exclusion) surv - Survival at census (1 = alive, 0 = dead) rgr - Relative growth rate for diameter at base elev - Plot elevation slp - Slope of plot topo - Designation of plot location (Categorical variable of Lowland, Riparian and Ridge) Description of RCode GrowthSurvivalGlobalFinal.R - Code for analysis of global growth and survival Figures 2 and 3 PlotsSpeciesResponses.R - Code to generate Figure 4 and 5 from Rstanarm output FitnessDifferences.R - Code to generate Figure 6 from Rstanarm output

Comunidad de Madrid Atracción de Talento Investigador2018-T1/AMB-11095 Agencia Estatal de Investigación Ramon y CajalRYC2021-032049-I

Peer reviewed