Introduction

This dataset was used for analyses and figures in Ankori-Karlinsky et al., 2025 in New Phytologist.

The main question of this paper is whether exposure to chronic winds reduces individual trees' vulnerability of stem-break within species.

In order to evaluate this question, we use airborne LiDAR-derived crown metrics from 124 geo-located trees in the 16ha Luquillo Forest Dynamics Plots (LFDP) in the Luquillo Experimental Forest (LEF) in El Yunque National Forest, Puerto Rico.

The LiDAR data come from geo-located 35cm resolution G-LiHT (https://gliht.gsfc.nasa.gov/) LiDAR point-clouds taken in 2017 (before Hurricane Maria) over the LEF at 12 pulses/m2 density.

We focus on 4 abundant (~40% of forest basal area) species that differ in life-span and wind-resistant life-history traits:

• Buchenavia tetraphylla (~40-50 year lifespan, intermediate hurricane resistance)
• Cecropia schreberiana (~20-40 year lifespan, low hurricane resistance)
• Dacryodes excelsa (~400 year lifespan, high hurricane resistance)
• Manilkara bidentata (~400 year lifespan, high hurricane resistance)

To understand how chronic wind-induced architectural imprints may impact a tree’s vulnerability to storms, we calculated mechanical wind vulnerability for each individual tree i of species s ($rs_{is}$) at storm speeds (e.g., 25 m/s) based on the GALES wind vul model from Gardiner et al., (2000) and based on Jackson et al., (2021):

$$vul_{is} \sim \frac{H_{is}}{dbh_{is}^{3}} ⋅\frac{A_{is}}{\rho_{s}} $$

where H is the vertical height of the tree in m, dbh is the diameter of the stem in cm at 1.3 m Height, A is the surface area of the crown in m$^{{2}$, and $\rho$ is species-level wood weight to volume in $\frac{kg}{cm}{3}}$ from Swenson et al., (2012)

We therefore estimate how chronic wind-exposure (exposure to NE-SE trade winds) may reduce wind vulnerability by leading to either shorter trees, wider stems, or reducing crown area.

We also explores how a number of other covariates may influence tree architecture and therefore, vulnerability:

• Topography - wetness index, elevation, slope, curvature within a radius of 20 m$^{2}$
• Neighborhood Crowding using the crowding index from Uriarte et al., (2004) within a radius of 20 m$^{2}$
• Previous Neighborhood Hurricane Damage within a radius of 20 m$^{2}$

References

Gardiner, B., Peltola, H., & Kellomäki, S. (2000). Comparison of two models for predicting the critical wind speeds required to damage coniferous trees. Ecological Modelling, 129(1), 1–23. https://doi.org/10.1016/S0304-3800(00)00220-9

Jackson, T., Shenkin, A. F., Majalap, N., Jami, J. B., Sailim, A. B., Reynolds, G., Coomes, D. A., Chandler, C. J., Boyd, D. S., Burt, A., Wilkes, P., Disney, M., & Malhi, Y. (2021). The mechanical stability of the world’s tallest broadleaf trees. Biotropica, 53(1), 110–120. https://doi.org/10.1111/btp.12850

Swenson, N. G., Stegen, J. C., Davies, S. J., Erickson, D. L., Forero-Montaña, J., Hurlbert, A. H., Kress, W. J., Thompson, J., Uriarte, M., Wright, S. J., & Zimmerman, J. K. (2012). Temporal turnover in the composition of tropical tree communities: Functional determinism and phylogenetic stochasticity. Ecology, 93(3), 490–499. https://doi.org/10.1890/11-1180.1

Uriarte, M., Canham, C. D., Thompson, J., & Zimmerman, J. K. (2004). A Neighborhood Analysis of Tree Growth and Survival in a Hurricane-Driven Tropical Forest. Ecological Monographs, 74(4), 591–614. https://doi.org/10.1890/03-4031

Metadata

Each row is one individual tree stem.
Below is information on each column:
UniqueID: Tag number of tree in the Luquillo Forest Dynamics Plot (LFDP)
Species: Species 6-letter abbreviation
Quadrat: 20m² quadrat number
Q5: 5m² quadrat number
X: Relative x coordinate of tree stem
Y: Relative y coordinate of tree stem
DBH: Diameter at 1.3m height in cm
Elev_20m: Average elevation in m in 20m radius around tree stem
Slope_20m: Average topographic slope in degrees in 20m radius around tree stem
Curve_20m: Average topographic curvature in 20m radius around tree stem
TWI_20m: Average topographic wetness index in 20m radius around tree stem
NCI_2000_20m: Average crowding index in 20m radius around tree stem from 2000 census
Wind_Binary: Chronic wind exposure (protected or exposed)
Wood_Density: Mean species-level wood specific gravity according to Swenson et al. 2012 (kg/m³)
SLA: Mean species-level specific leaf area according to Swenson et al. 2012
Hugo_damage: Proportion severely damaged trees from Hurricane Hugo in 1989 in 20m radius around tree stem
Height: Maximum height of tree crown from ALS data in m
Crown_Volume: Volume of tree crown from ALS data in m³
Crown_area: Area of tree crown from ALS data in m²
Mean_aspect: Average direction of tree crown from ALS data in degrees
AGR_all: Absolute growth rate of tree from 2016 census to earliest census observed
Slenderness: Height/DBH
DBH3: DBH to the third power to model cantilever beam
Wind_Vul: Mechanical vulnerability of wind-induced stem-break calculated according to Gardiner et al. 2000 GALES model
Comment: Comment in Geo7x handheld Trimble device indicating direction of measurement relative to the location of the tree stem
Corr_Type: Correction type (if any) used by Pathfinder software from Trimble
Rcvr_Type: Recovery type of information
GPS_Date: Date measurement was taken
GPS_Time: Time of day measurement was taken
Unfilt_Pos: Number of measurements (1 per second) taken before processing
Filt_Pos: Number of measurements (1 per second) taken after processing
GNSS_Heigh: Elevation above sea level measured in m
Vert_Prec: Vertical precision accuracy in m
Horz_Prec: Horizontal precision accuracy in m
Std_Dev: Standard deviation of precision accuracy
Northing: Northing coordinate (in UTM -20N Puerto Rico projection)
Easting: Easting coordinate (in UTM -20N Puerto Rico projection)

Sharing/access Information

Please cite Ankori-Karlinsky et al., 2025 in New Phytologist if you use these data.