Vulnerabilidade mecânica de árvores relacionada a tempestades convectivas na Amazônia Central, Manaus (AM)

Abstract

Storms producing high winds are frequent and major natural disturbance events in the Central and Western Amazon, affecting forest structure and diversity at the landscape level. The stability of Amazonian tree species against storm-induced wind-throw has never been evaluated using direct experimental approaches such as winching. This has typically been related to the distribution of wind speeds, butt little is known about difference in the susceptibility to wind damage cause by site-related difference in growth form. Previous work in the Central Amazon found that tree mortality caused by storm-induced blowdowns, primarily by snapping and uprooting, was highest on plateaus and slopes, while much lower in valley forests. In this study we investigated the critical turning moments (Mcrit) of 60 trees, ranging from 19.0 to 41.1 cm in diameter at breast height (DBH) and located in different topographic positions, using a cable-winch load-cell system. Our approach used torque as a measure of tree failure to a point of no return (snapping or uprooting). Size controlled variation in Mcrit was quantified for cardeiro (Scleronema mincranthum (Ducke) Ducke), mata-matá (Eschweilera spp.), and a random selection of trees across a number of species. Our analysis of Mcrit revealed that tree resistance to failure increased with size and differed among species, but no effects of topography or failure mode were found for the species groups either separately or pooled. For the random species, total explained variance increased from an R2 of 0.49 for DBH alone, to 0.68 when including both DBH and stem wood density (SWD) in a multiple regression model. For the random selection of trees those that have interlocked grain seems not to increase Mcrit whit increases in size ,when compared to similar sized trees with other grain orientation. The slenderness factor (total tree height/DBH, both in m) influence Mcrit for the different groups, overall values bellow 80 show higher resistance. Depending on the size attribute (DBH or above ground biomass (AGB)) for similar sized trees on group is just more resistant than the others. This mechanics approach allows the comparison of tree vulnerability induced by wind damage across ecosystems and facilitates the use of forest structural information in ecosystems models. Our results indicate that topographic differences in wind-throw vulnerability are likely due to elevational differences in wind velocities and crown exposure, rather than by differences in soil properties related to topography