Table 1.
List of the study species with their abbreviation.
Table 2.
List of functional traits measured for 19 tropical dry forest species and their functional role as reported elsewhere.
Fig 1.
Principal component analysis of 12 traits in 19 tree species from the tropical dry forest, evaluated in seedlings.
Species and trait abbreviations are the same as in Tables 1 and 2, respectively. Traits: MPS: Minimum photosynthetic unit size, LDMC: Leaf dry matter content, Ψmin: Minimum leaf water potential, LRT: Leaf retention time, SLA: Specific leaf area, WD: Wood density, SWC: Stem water content, BWC: Bark water content, SRL: Specific root length, RWC: Root water content, VRER: Vertical root elongation rate, RB/LB: Total root biomass/total leaf biomass, FRL/LA: Fine root length/ total leaf area. Species: Acfa: Acacia farnesiana.; Amad: Amphipterygium adstringens; Appa: Apoplanesia paniculatta; Caer: Caesalpinia eriostachys; Capl: Caesalpinia platyloba; Ceae: Ceiba aesculifolia; Coal: Cordia alliodora; Coel: Cordia elaeagnoides; Cral: Crescentia alata; Ency: Enterolobium cyclocarpum; Glse: Gliricidia sepium; Guul: Guazuma ulmifolia; Ipwo: Ipomoea wolcottiana; Miar: Mimosa arenosa; Pico: Piptadenia constricta; Pidu:Pithecellobium dulce; Rupa: Ruprechtia pallida; Sppu: Spondias purpurea; Swhu: Swietennia hummillis.
Fig 2.
Physical variables during the dry-wet pulse experiments done in the common garden and the greenhouse.
I. simulation of dry-wet pulse scenarios in a common garden. Temporal course of soil water potential (mean and se) at different depths for: a) no drought (ND), b) short and frequent drought pulses (SFD; 15 days drought, 10 days wet), c) long and infrequent drought pulses (LID; 25 days drought, 10 days wet) and d) prolonged drought (PD; no wet period). Atmospheric conditions (mean and se) above the plants during the dry and wet periods imposed by the four treatments: (ND, SFD, LID, PD); e) air temperature, f) relative humidity, h) vapour pressure deficit. II. simulation of a dry-wet pulse scenario (SFD) in the greenhouse. i) no drought (ND), j) short frequent drought (SFD; 20 days drought, 10 days wet). Mean and se values at 20 cm depth. Atmospheric conditions shown in j), k), l). All variables measured at the end of the dry and the wet period, see details in methods.
Fig 3.
Performance of seedlings of 19 tropical dry forest tree species in response to different dry-wet pulse scenarios in field conditions: a) survival, and b) mean relative growth rate in each dry-wet pulse scenario. Points indicate individual plants, while boxes indicate 25% and 75% percentiles. ND: no-drought; SFD: short and frequent drought pulses, LID long and infrequent drought pulses; and PD: prolonged drought. Different letters indicate significant differences in the mean between dry-wet pulse scenarios at P < 0.05, according to a mixed model including species as a random factor.
Fig 4.
Seedling´s survival and RGR in relation to species’ functional strategies across dry-wet pulse scenarios simulated in field conditions, according to mixed models shown in Tables 2, 3.
PC1 scores represent a drought tolerance-drought avoidance continuum, and PC2 scores represent a belowground resource acquisition continuum. Colored shading indicates 95% prediction confidence intervals. * Indicates significant regressions. For a), b), different letters indicate significant differences of slopes between drought pulse scenarios. For c) and d), different letters indicate significant differences between the trends evaluated at three different values of the regressor. For PC1: low (-5), intermediate (0) and high (3.5); for PC2: low (-1.5), intermediate (0.5) and high (2.0).
Table 3.
Effects of species’ functional strategy (PC1 or PC2 scores) and dry-wet pulse scenario on survival, while controlling by plant height, in seedlings of 19 TDF species grown in a field common garden experiment.
(A) generalized linear mixed model for survival against PC1 (R2m = 0.27; R2c = 0.36) and B) generalized linear mixed model for survival against PC2 (R2m = 0.19; R2c = 0.39). X2 values correspond to Wald Type III test statistics. Each model considered 3463 individuals.
Table 4.
Effects of species functional strategy (PC1 or PC2 scores) and dry-wet pulse scenario on RGR, while controlling by plant height, in seedlings of 19 TDF species growing in a field common garden experiment.
A) quadratic mixed model for RGR against PC1 (R2m = 0.34; R2c = 0.60). B) quadratic mixed model for RGR against PC2 (R2m = 0.35; R2c = 0.61). X2 values correspond to Wald Type III test statistics. Each model considered 2862 individuals.
Fig 5.
Seedling´s survival in relation to plant functional strategies (PC1, PC2 species scores) under greenhouse conditions.
Panels a), b) plants subject to a simulated dry-wet pulse and non-drought scenarios, respectively. Trend lines and confidence intervals from mixed models are shown (red no-drought, blue dry-wet pulse). a) survival vs. PC1 (X2 < 0.01, p > 0.99 for dry-wet pulse or no-drought scenarios), b) survival vs. PC2 (X2 < 0.01, p > 0.99 for dry-wet pulse or no-drought scenarios). Panels c), d) survival after drought-provoked dieback. Trend lines and confidence intervals from mixed models are shown. c) trends against PC1 species scores describing a drought avoidance vs. tolerance continuum (X2 = 4.61 p < 0.05), d) trends against PC2 species scores describing an acquisitive resource use continuum (X2 = 0.20 p = 0.64).