PARENT SESSION
Organized Oral Session 31: Structure and function of tropical rainforest canopies
Organizer(s): MG Ryan and DA Clark
Wednesday, August 10, 8:00 AM - 11:30 AM, Meeting Room 516 A, Level 5, Palais des congrès de Montréal
A comparison of three methods for assessing Leaf Area Index (LAI) across a tropical rain forest landscape.
Olivas, Paulo*,1, Oberbauer, Steven2, Clark, David1, 3, Ordonez, Harlyn1, 1 La Selva Biological Station, Puerto Viejo, Heredia, Costa Rica2 Florida International University, Miami, Florida, USA3 University of Missouri-St. Louis, St. Louis, Missouri, USA
ABSTRACT- Many structural and functional properties of a forest depend on its total leaf area. LAI (m2/m2) can only be measured directly by harvest, a daunting challenge in tall forests. As a result, leaf area is generally estimated by indirect methods involving measurements of light absorption. Such indirect methods have never been calibrated against direct measurements in tropical rain forest (TRF), either at point or landscape scales. Here we take advantage of a unique dataset of direct LAI measurements over a TRF old-growth upland landscape to examine the performance of two common methods for indirectly estimating LAI, digital fisheye photographs and the LI-COR LAI-2000 Plant Canopy Analyzer. LAI was measured directly using a portable aluminum tower erected at 45 points sited with a random design stratified by total soil phosphorus and slope. At each site all plant biomass within 4.5 m2 was harvested from the forest floor to the top of the canopy. Prior to tower installation, a fisheye photograph was taken within the tower footprint and subsequently analyzed for estimated LAI using the program Gap Light Analyzer. After the tower completion, five LAI 2000 measurements were taken in a direction perpendicular to the non-disturbed side of the tower to provide a local understory LAI estimate referenced using a LAI-2000 mounted above the canopy on the tower. For point estimation there was no significant relation between each tower's total LAI (N = 41) and the understory LAI estimates for either the LAI 2000 (r 2= 0.05, NS) or fisheye photographs (r 2 = 0.09, NS). At the landscape scale, total direct harvest LAI averaged 5.69 + 0.35 (1 SEM), while fisheye photographs averaged 3.46 + 0.18 and the LAI 2000 averaged 5.17 + 0.15. Based on these results, it appears that both indirect methods underestimate the magnitude and variance of actual LAI over this old-growth landscape, and in the case of fisheye photographs the underestimate is large (39%). However, the LAI estimate with the LAI 2000 was marginally low (9%). When used in a stratified random sampling design, the LAI 2000 may be an adequate tool for quickly characterizing the mean LAI of a sizeable landscape if an above-canopy or open-site reference is available.
Key words: Leaf area index, LAI, LAI 2000, Fisheye photographs