ID do Resumo: 633

Towards quantifying the effect of light limitation on tree growth in Amazonia using small footprint Lidar

Light strongly limits the productivity and function of tropical forests. Quantifying the magnitude of the effect of light limitation on productivity is difficult; however, we can begin by understanding the relationship between light and individual tree growth rates. We can combine measurements of light availability, tree leaf area, as well as other factors (e.g., leaf physiology and vascular network volume) to predict observed growth rates. Major problems with this approach, though, are that (i) light availability and (ii) individual tree leaf area are difficult to measure, particularly for large individuals (> 10cm dbh). Here we develop and begin to test an approach based on small footprint lidar. Established techniques can estimate three-dimensional maps of vegetation density in the canopy from small footprint. We develop a model of light penetration based on three-dimensional vegetation maps and apply it to lidar data collected from the Tapajós National Forest (FLONA), PA, and the Ducke Reserve, AM, Brazil in June and July 2008. Preliminary light measurements suggest that this technique can predict up to 50% of the variability in light environments at the forest floor. We also assess the ability of lidar to predict the leaf area of individuals trees (>10 cm dbh) by comparing a preliminary dataset of crown percent cover estimates collected in FLONA with lidar pulse transmission rates through theses crowns (we expect cover and lidar pulse transmission rate to be similar; crowns are spatially delineated from the ground). We conclude that small footprint lidar based light transmission modeling and leaf area estimation hold exceptional potential for the investigation of light limitation in tropical tree communities from local to regional scale.

Sessão:  Carbono - Integrando fluxos de carbono para a região a partir de medidas em pontos de coleta, torres e aeronaves.

Tipo de Apresentação:  Oral