Characterizing height-diameter relationships for Caribbean trees using mixed-effects random forest algorithm

Accurately characterizing the relationships between tree height and diameter at breast (DBH) is vital to quantify productivity predict growth for forest management. This study makes use of data from inventories in Caribbean islands Puerto Rico, Trinidad U.S. Virgin Islands compare efficacy parametric mixed models, random mixed-effects algorithms (MERF) modeling species-specific H-D relationships. MERF as a semi-parametric approach advantageous handling high dimensional complex datasets using algorithm, well provide predictions with effects model framework. A total 37,124 observations used analyses were collected 645 permanent sample plots across three islands, including 309 species 121 34 Trinidad. Results showed that can be reliably diverse species, which produced more accurate than other two models. In addition DBH, competition index was found most important variable among all stand environmental variables islands. Adding climate-related improve prediction accuracy especially when size small. Precipitation selected often, confirming overall importance precipitation height. The results this work will not only relationship models trees, but also advance knowledge broad application mixed-species forests.