LB 10-116 - Sculpted by the winds: late-successional species grow shorter in wind-exposed sites in a tropical forest

Background/Question/Methods:

Chronic wind-exposure is a vital factor in shaping tree architecture and crown structure, changing carbon allocation away from height towards mechanical stability. The shape of a tree, in turn, can have implications for its vulnerability to severe disturbances like cyclones. Although some studies have examined the relationship between wind and tree architecture, only a few have focused on tropical forests. Here we use high-resolution LiDAR data in the cyclone-prone, 16-ha Luquillo Forest Dynamics Plot (LFDP) in Luquillo Experimental Forest, Puerto Rico, to ask: does chronic wind-exposure reduce individual tree height and crown volume across species? We hypothesize that longer-lived species will be shorter and have smaller crowns in wind-exposed areas, whereas shorter-lived species will be unaffected.



We used a 30m DEM with 20 years of wind data from the roof of a nearby station to map exposure in the LFDP using EXPOS. We then randomly sampled two common tree species (n=30) stratified by wind-exposure, and extracted individual tree height (m) and crown volume (m3) using data from NASA’s G-LiHT 2017 flyover. For each species, we ran linear regressions of structure as a function of wind-exposure, tree size, microtopography, and neighborhood crowding (NC), using AIC to select the best models.

Results/Conclusions:

Results for two species - Buchenavia tetraphylla (late-successional) and Cecropia schreberiana (early-successional) - demonstrate that wind alters tree structure differently across succession. The best height model for B. tetraphylla included all predictors (adjusted R2 =0.51), and indicated that DBH and wind-exposure were the only significant predictors. Height increased by 1m for every 10cm increase in DBH, and decreased by 4m on average between wind-protected and wind-exposed sites. The best model for crown volume only included DBH as a predictor (R2=0.3).

The best height model for C. schreberiana included DBH, topography and NC (adjusted R2 =0.52) showing DBH as the sole significant predictor, with 1m increase in height for every 3cm increase in DBH. The best crown volume model (R2=0.36) included all predictors and showed DBH as the sole significant predictor.

These results indicate that late-successional species, which are longer-lived and experience regular wind disturbance, have shorter canopies in wind-exposed areas, potentially increasing their resistance to repeated cyclones. Conversely, short-lived species, which are shorter in stature and potentially less exposed in the canopy, do not exhibit any structural differences based on wind-exposure. This highlights a potential biomechanical difference in life-history strategy in this cyclone-prone tropical forest ecosystem.