Variation in aboveground tree live biomass in a central Amazonian Forest: Effects of soil and topography

Abstract

We used forest inventories, conducted in seventy-two 1-ha permanent plots to assess the variation in tree and palm aboveground live biomass (AGLB), and its relation with soil gradients (texture and nutrients) and topography (altitude and slope). Our plots, located at Reserva Florestal Adolpho Ducke, a 10,000 ha reserve in central Amazonia, near the city of Manaus, were systematically spread over 64 km2. The plots were long (250 m) and narrow (up to 40 m), following elevational contours. Chemical and physical soil analyses were undertaken using topsoil samples, collected in each plot. The plots covered a soil textural gradient ranging from 8% to 98% percent of sand, slopes varying from 0.5° to 27°, and the maximum difference of altitude among plots was 70 m, ranging from 39 to 109 m a.s.l. The mean total AGLB (palm + trees) for stems over 1 cm diameter breast height (dbh = 1.3 m), was 327.8 Mg/ha (ranging from 210.9 to 426.3) and was similar to other studies in the region. We found a two-fold variation in total AGLB estimates among plots, with soil or topography explaining around 20% of this variation. AGLB was positively related to either a soil textural gradient (represented by the percentage of clay) or altitude, but was insensitive to slope. Although slope had no detectable effect on the mean variation of AGLB per plot, it explained 14% of the AGLB in understory and emergent trees. On slopes, AGLB is concentrated mostly in tree classes of small size (1 ≤ dbh < 10 cm), whereas in flat areas, most of the biomass is concentrated in a few big trees. When considering palms and trees separately, we found that tree biomass tended to increase in clay-rich soils (mostly located in high areas), whereas palm biomass was higher in sandy soils (low slopes and bottomlands). This study shows that forests in central Amazonia grow in a heterogeneous environment in relation to soil and topography, and this heterogeneity seems to be in part responsible for differences in structure and AGLB accumulation. In the forest around Manaus, topography can be used to predict aboveground live biomass as an alternative to soil variables, with similar predictive power. Topographic variables can be easily assessed using satellite imagery and could improve current estimates of carbon stocks over large areas in the Amazon. © 2006 Elsevier B.V. All rights reserved.