September 23, 2014

CTFS-ForestGEO grant project report from KC Cushman at Barro Colorado Island, Panama

CTFS-ForestGEO Grants Program recipient, KC Cushman, has been researching her grant proposal “Improving estimates of biomass change in buttressed trees using site-specific tree taper models" in Panama, Thailand, Singapore and Colombia. She recently completed her project and the following is her report:
KC with a buttressed tree in Khao Chong, Thailand
"The amount of biomass stored in any tropical forest can change over time as trees grow, produce new seedlings, and die. Measuring how tropical forest biomass changes over time is important for understanding the global carbon cycle; if tropical forests increase in biomass over time they act as a carbon sink (if trees are growing more than they are dying, on average), but if tropical forests decrease in biomass over time they act as a carbon source (if trees are dying more than they are growing, on average).

CTFS plot in Amacayacu, Colombia.
The diameter measurement height of
each tree is marked in yellow paint.
One tool for studying biomass change is measuring tree diameters in forest plots, such as those in the Center for Tropical Forest Science network. Diameter measurements taken at 1.3 m height are converted to an estimate of total biomass using equations developed in previous research. All trees in a plot are remeasured regularly to determine changes in biomass.
Picture of a tree, a 3-D model
of the same tree constructed
from 43 pictures using
Agisoft PhotoScan, and
trunk outlines extracted from
the 3-D model point using the
program Cloud Compare

However, the prevalence of buttressed trees in tropical forests presents some challenges for this approach. Trees with buttresses are not cylindrical at 1.3 m height, so their diameters are measured higher on the trunk. This practice creates a bias because tree trunks tend to decrease in diameter with height, or taper. Therefore, diameter measurements taken above 1.3 m height will be smaller and yield a lower value for biomass. The magnitude of biomass underestimation in a forest can change over time because measurement heights are often moved up a tree as it grows. This is problematic because biomass appears to change, but the change is not caused by real forest processes. In a previous study on Barro Colorado Island, Panama, we found that biomass trends change significantly after correcting for changing measurement heights in diameter data. Biomass decreased over time in uncorrected data but increased over time in corrected data (Cushman et al. 2014).

Walking to the plot in
Huai Kha Khaeng, Thailand
This year, I had the opportunity to follow up on our preliminary study on BCI through research supported by the CTFS-ForestGEO Research Grants Program. I visited four other CTFS plots to compare tree taper in different forests: Amacayacu (Colombia), Bukit Timah (Singapore), Huai Kha Khaeng (Thailand), and Khao Chong (Thailand). In this study, I am also using a novel tool to measure trees by creating 3-D models from digital pictures using Agisoft PhotoScan. I took 30-60 photos of each of approximately 100 trees per plot to create a model of each tree.  I am using these models to measure characteristics of trunk shape at each site, such as how diameter and trunk circularity change with height. Results from trunk shape measurement can then be used to improve estimates of biomass change when diameter measurement heights change over time.
Field Station at
Huai Kha Khaeng, Thailand

I am thankful to have had the opportunity to visit these plots and meet other researchers through the CTFS-ForestGEO network. This work would not have been possible without the support and guidance of collaborators Helene Muller-Landau, Stuart Davies, Sarayudh
Bunyavejchewin, Alvaro Duque, Somboon Kiratiprayoon, and Shawn Lum, and the assistance of Pablo Ramos, Paulino Villareal, Emily Francis, Juan Sebastian Barreto Silva, Gabriel, Pitoon Kongnoo, Mohamah Fairoz, Jonathan, and Reuben in the field."

To learn more about CTFS-ForestGEO, click here