Search

For more than three decades, major efforts in sampling and analyzing tree diversity in South America have focused almost exclusively on trees with stems of at least 10 and 2.5 cm diameter, showing highest species diversity in the wetter western and northern Amazon forests. By contrast, little attention has been paid to patterns and drivers of diversity in the largest canopy and emergent trees, which is surprising given these have dominant ecological functions. Here, we use a machine learning approach to quantify the importance of environmental factors and apply it to generate spatial predictions of the species diversity of all trees (dbh ≥ 10 cm) and for very large trees (dbh ≥ 70 cm) using data from 243 forest plots (108,450 trees and 2832 species) distributed across different forest types and biogeographic regions of the Brazilian Amazon. The diversity of large trees and of all trees was significantly associated with three environmental factors, but in contrasting ways across regions and forest types. Environmental variables associated with disturbances, for example, the lightning flash rate and wind speed, as well as the fraction of photosynthetically active radiation, tend to govern the diversity of large trees. Upland rainforests in the Guiana Shield and Roraima regions had a high diversity of large trees. By contrast, variables associated with resources tend to govern tree diversity in general. Places such as the province of Imeri and the northern portion of the province of Madeira stand out for their high diversity of species in general. Climatic and topographic stability and functional adaptation mechanisms promote ideal conditions for species diversity. Finally, we mapped general patterns of tree species diversity in the Brazilian Amazon, which differ substantially depending on size class., (© 2023 John Wiley & Sons Ltd.)

Due to changes in the Amazon forest dynamics after the first cutting cycle, non exploited species become dominant in the forest. The lack of technological knowledge makes it hard to commercialize these woods, making the understanding of their physical-mechanical properties a fundamental step to properly define their applications. This study aimed to characterize physically and mechanically the wood of Pseudopiptadenia psilostachya and Eschweilera ovata from the second cutting cycle of the Tapajós National Forest, intending to commercially promote and to identify usages for them, as well as to evaluate the viability of replacement of highly commercialized species. The tests were performed accordingly to the Brazilian standard NBR 7190. P. psilostachya, presented bulk density of 0.683 g.cm-3, medium levels of shrinkage and anisotropy, as well as medium hardness (7366 N) and high strength on compression parallel to grain (71.63 MPa) and on static bending (103.9 MPa). It was generally superior to Euxylophora paraensis, but inferior when compared to Bagasssa guianensis and Apuleia leiocarpa, with possibilities to replace these three species. E. ovata, on the other hand, presented bulk density of 0.798 g.cm-3, high shrinkage and anisotropy values, but high values for hardness (12089 N) and strength on compression parallel to grain (68.67 MPa) and on static bending (127.1 MPa). This species exhibited, in general, similarities with Mezilaurus itauba, although it is a little inferior in relation to Hymenaea sp. and Astronium lecointei. Both species, P. psilostachya and E. ovata fit in the highest strength class described on the Brazilian Standard NBR7190, C60. By means of the results found, it was concluded that although the species studied were unknown, they presented timber with enough quality to replace some Amazon timber species widely commercialized on both national and international market.