A COMPARISON OF DIRECT AND INDIRECT METHODS FOR ESTIMATING FOREST CANOPY LEAF-AREA

Two indirect gap fraction methods for estimating leaf area index (LAI) are compared with estimates from litterfall collections in a mixed-age oak-hickory forest. One indirect method uses averaged, direct beam penetration data obtained with a moving tram. The second uses a portable light sensor system that measures diffuse light penetration for five sky sectors between zenith angles 0 and 75-degrees. Data were collected from September 1989 to January 1990. The Poisson model and the negative binomial model of gap frequency were applied to estimate LAI from observed transmittances. With the Poisson model, an assumption of a random leaf spatial distribution contributes to an underestimation of LAI by as much as 45%; this is because leaves at this site are actually clumped at both large and small scales. The negative binomial, which requires determination of a clumping parameter, produces estimates comparable with those of the litterfall method. Both indirect techniques accurately describe temporal changes in leaf area using either the Poisson or negative binomial model. The portable system also allows easy estimation of the spatial variation in leaf area within the site or between sites, and it can be used to obtain a vertical profile of leaf area.