This study aimed to demonstrate the practical advantages of the UBCSAND constitutive model in simulating the mechanical behavior of tailings sands through numerical analysis. For this purpose, a mixed approach of non-experimental design, applied type, and comparative scope was chosen. The population consisted of 20 monotonic triaxial tests for tailings sands and the purposive sample consisted of 3 tests performed with confining stresses of 5, 10, and 15 kg/cm2 under undrained conditions. Simulations were performed using the TNO DIANA finite element software, which incorporates the UBCSAND model. The validation of the model's implementation was achieved by comparing its predictions with both the experimental results and simulations conducted using the PLAXIS program. The results indicated that the numerical responses of stress-strain, volumetric change-strain, and stress path, showed a strong compatibility with the experimental data at a confining stress of 5 kg/m2, while for higher values (10 and 15 kg/m2) there was still a good agreement with some moderate observations. Consequently, the study proposes that the UBCSAND constitutive model is an efficient alternative approach to predict with high accuracy the mechanical behavior of tailings sands.

Tropical Amazonian forests serve as critical carbon reservoirs, although their storage capacity varies significantly depending on the topographic, hydrologic, and successional heterogeneity of the landscape. This study quantified aboveground biomass (AGB) and carbon stocks across four vegetation units in the Tacshitea sector, Sierra del Divisor Na-tional Park (Peru): semi-dense flooded forest (Z1), hill forest (Z2), hill forest associated with secondary growth (Z3), and Mauritia flexuosa palm forest (Z4). We established 100 plots of 20 × 20 m (1 ha per zone), assessing all individuals with diameter at breast height (DBH) ≥ 10 cm using allometric equations validated for Amazonian ecosystems. A total of 2,363 individuals belonging to 158 species from 41 families were recorded. AGB varied sig-nificantly among vegetation units (Kruskal-Wallis, p < 0.001), with the highest values in Z1 (414.09 Mg ha⁻¹; 194.62 Mg C ha⁻¹) and Z2 (332.01 Mg ha⁻¹; 156.05 Mg C ha⁻¹), whereas Z3 and Z4 recorded substantially lower values (145.27–153.18 Mg ha⁻¹; 68.28–71.99 Mg C ha⁻¹). The Fabaceae family dominated carbon storage at the landscape scale (33.82%). Floristic diversity showed positive but weak associations with carbon storage (R² = 0.045–0.056), and this relationship lost significance when controlling for spatial structure using linear mixed-effects models (p = 0.409), indicating that structural attributes—particularly basal area and wood density—are more robust predictors of storage capacity than taxo-nomic richness per se. Beta diversity analysis revealed that floristic differentiation among zones was dominated by species turnover, explaining 26.5% of the variation by vege-tation unit (PERMANOVA, F = 11.55; p = 0.001). These results support the relevance of Sierra del Divisor National Park as a priority carbon sink in the western Peruvian Amazon and underscore the necessity of conservation strategies that integrate the structural heterogeneity of the forest mosaic.

Amazonian forests store large amounts of carbon, but storage capacity varies widely with topography, hydrology and successional stage. We quantified aboveground biomass (AGB) and carbon stocks across four vegetation units in the Tacshitea sector, Sierra del Divisor National Park (Peru): semi-dense flooded forest (Z1), hill forest (Z2), hill forest associated with secondary growth (Z3), and Mauritia flexuosa palm forest (Z4). In each unit we established a contiguous 5 × 5 grid of 20 × 20 m plots (25 plots, 1 ha sampled per unit) and assessed all individuals with diameter at breast height (DBH) ≥ 10 cm, using allometric equations validated for Amazonian ecosystems. A total of 2363 individuals belonging to 156 species from 41 families were recorded. Aboveground biomass differed markedly among vegetation units, with the highest values in Z1 (401.15 Mg ha⁻¹; 188.54 Mg C ha⁻¹) and Z2 (328.92 Mg ha⁻¹; 154.59 Mg C ha⁻¹), whereas Z3 and Z4 recorded substantially lower values (142.25–144.53 Mg ha⁻¹; 66.86–67.93 Mg C ha⁻¹). The Fabaceae family dominated carbon storage at the landscape scale (34.7%). Comparing predictor blocks directly at the plot level, structural attributes explained 92.4% of the variance in carbon stock against 5.6% for taxonomic diversity (ΔAIC = 248), and variance partitioning attributed 87.3% exclusively to structure and 0.5% exclusively to diversity, indicating that basal area, wood density and stem size are far more robust predictors of storage capacity than taxonomic richness per se. Beta diversity analysis revealed that floristic differentiation among zones was dominated by species turnover, explaining 26.5% of the variation by vegetation unit (PERMANOVA, F = 11.54; p = 0.001), although within-unit dispersion was heterogeneous, so part of that signal reflects differences in internal variability rather than centroid position alone. Because each unit was sampled as a single contiguous block, unit effects are confounded with location, and the comparisons describe the four stands evaluated at Tacshitea rather than the vegetation types they represent. These results support the relevance of Sierra del Divisor National Park as a carbon reservoir in the western Peruvian Amazon and underscore the necessity of conservation strategies that integrate the structural heterogeneity of the forest mosaic.