COS 303-2 - Evaluating methods for measuring the leaf area index of encroaching shrubs in grasslands: from leaves to optical methods, 3-D scanning, and airborne observation

Abstract: Leaf area index (LAI) is a key variable describing ecosystem structure and regulating processes from carbon exchange, light extinction, and evapotranspiration to canopy interception and energy balance. LAI is commonly evaluated with indirect methods. However, the accuracy of indirect methods varies with canopy structure and is not always generalizable across ecosystems. While the accuracy of indirect methods is well characterized for woody plants in forest ecosystems, it is not well established for woody plants in open ecosystems, despite having varied canopy structure. In this study, we compared direct LAI measurements of dense grassland shrub canopies to three indirect methods: a ceptometer, a contemporary handheld 3D scanner, and airborne hyperspectral imaging. To our knowledge, this is the first study to assess the accuracy of leaf area data in woody plants using a handheld 3-D scanner and one of few studies assessing the accuracy of the National Ecological Observation Network’s (NEON) Airborne Observation Platform (AOP), which was our source of hyperspectral imaging. Data were collected in tallgrasss prairie undergoing woody encroachment and three treatments: no herbivore disturbance, bison present, and simulated browsing. We found that (1) the LAI of non-browsed C. drummondii canopies using direct methods averaged ~8.0 and the coefficient of light extinction varied with leaf area density. (2) The ceptometer produced accurate LAI estimates in control and grazed canopies but overestimated the LAI of browsed canopies by 38%. (3) The leaf area of individual ramets measured with a handheld 3D scanner had a strong relationship with direct measurements (r²=0.86), but underestimated ramet leaf area at greater values. (4) LAI estimated from airborne spectral data underestimated LAI by 55%. We conclude that a ceptometer was adequate for measuring LAI in dense clonal shrub canopies when browsing was not present, the handheld 3D scanner was adequate for measuring leaf area of individual C. drummondii ramets, and the NEON AOP remote sensing was not suitable for estimating LAI of dense shrub islands in the open grassland.