Data-Driven Analysis of Tree Structure Variation Across Forest Types in Tapajós National Forest, Brazil

Understanding how tree structural traits vary across forest conditions is essential for assessing ecological integrity, detecting anthropogenic disturbances, and informing conservation strategies. This study investigates the architectural characteristics of trees within primary (PF), secondary (SF), and selectively logged forests (SLF) of the Tapajós National Forest, a representative landscape in the Brazilian Amazon. Although forest inventory datasets are increasingly available, few studies explicitly analyze the relationship between phenotypic tree traits and forest condition, or employ geometric modeling to estimate structural volume. To address this gap, we analyzed biometric data from 30 permanent plots surveyed in 2010, using a combination of statistical and geospatial techniques implemented in Python. Correlation and scatterplot analyses revealed strong associations between total tree height and maximum crown width (r > 0.8), and moderate to strong correlations among diameter at breast height (DBH), height, and crown dimensions (0.40 < r < 0.80). Comparative analyses among forest types showed that PF plots consistently exhibited greater DBH, crown depth, and total height than SF and SLF plots, reflecting structural degradation linked to anthropogenic disturbance. Distributions of wood density and crown morphology further highlighted ecological differences among forest types. To enhance structural assessment, we incorporated crown shape coefficients and directional crown radii to estimate individual tree crown volumes using ellipsoidal geometry. These volume estimates followed patterns consistent with other structural metrics and provided a scalable proxy for canopy structure, enabling spatially explicit comparisons. Overall, this integrative approach offers a robust framework for quantifying tree architecture and supports improved forest monitoring, carbon modeling, and biodiversity evaluation in tropical forest ecosystems.