Mapping landscape scale variations of forest structure, biomass, and productivity in Amazonia

Abstract

Abstract. Landscape and environmental variables such as topography, geomorphology, soil types, and climate are important factors affecting forest composition, structure, productivity, and biomass. Here, we combine a network of forest inventories with recently developed global data products from satellite observations in modeling the potential distributions of forest structure and productivity in Amazonia and examine how geomorphology, soil, and precipitation control these distributions. We use the RAINFOR network of forest plots distributed in lowland forests across Amazonia, and satellite observations of tree cover, leaf area index, phenology, moisture, and topographical variations. A maximum entropy estimation (Maxent) model is employed to predict the spatial distribution of several key forest structure parameters: basal area, fraction of large trees, fraction of palms, wood density, productivity, and above-ground biomass at 5 km spatial resolution. A series of statistical tests at selected thresholds as well as across all thresholds and jackknife analysis are used to examine the accuracy of distribution maps and the relative contributions of environmental variables. The final maps were interpreted using soil, precipitation, and geomorphological features of Amazonia and it was found that the length of dry season played a key role in impacting the distribution of all forest variables except the wood density. Soil type had a significant impact on the wood productivity. Most high productivity forests were distributed either on less infertile soils of western Amazonia and Andean foothills, on crystalline shields, and younger alluvial deposits. Areas of low elevation and high density of small rivers of Central Amazonia showed distinct features, hosting mainly forests with low productivity and smaller trees.

Quantifying Forest Biomass Carbon Stocks From Space

“…As mentioned before, the use of airborne data is only feasible at national level or below due to logistic and economic constrains. A range of non-parametric machine-learning algorithms are used at national level to extrapolate AGB measured from forest plots [11,21,62,63,99,[102][103][104]. This approach can combine different types of data from high to coarse resolution.…”

Influence of landscape heterogeneity on spatial patterns of wood productivity, wood specific density and above ground biomass in Amazonia

“…such as the RAINFOR network Peacock et al, 2007). Studies of these plots have generated significant insights into the spatial and temporal dynamics of the processes of forest carbon cycling, including biomass (Baker et al, 2004a;Chave et al, 2005;Malhi et al, 2006;Saatchi et al, 2009), wood productivity Baker et al, 2009;Mercado et al, 2009), wood specific density (Baker et al, 2004b), changes in growth Lewis et al, 2004), turnover (Phillips et al, 1994, and diversity (Phillips et al, 1994;ter Steege et al, 2000;Honorio et al, 2009).…”

“…Amazonia is an important part of the global biosphere, playing a major role in influencing climate systems (Andreae et al, 2004;Malhi et al, 2008), and providing vital ecosystem services such as carbon storage and cycling (Houghton et al, 2001;Malhi et al, 2004;Aragão et al, 2009;Baker et al, 2009;Chave et al, 2009;Saatchi et al, 2009) and water cycling (Betts et al, 2004;Marengo, 2005). It is also a repository for vast stores of biodiversity (Condit et al, 2002;ter Steege et al, 2006).…”

Forest response to increased disturbance in the central Amazon and comparison to western Amazonian forests

“…While wood density is typically found to be higher in the central Amazon and lower in the "west and south" (Baker et al, 2004a;Saatchi et al, 2009), high wood density is also found in northern Peru (Patiño et al, 2009;Saatchi et al, 2009). Next we compared the same disturbance scenario in CLM-CN 4.5 and found with regards to AGB response to disturbance, CLM performed in a very similar behavior to the gap model.…”