Estimating foliage surface area index in 8-year-old stands of Pinus elliottii var. elliottii x Pinus caribaea var. hondurensis of variable quality

Abstract

Comparisons were made between allometric and gap fraction estimates of foliage surface area index (foliage SAI) and spectral reflectance characteristics of 8-year-old stands of Pinus elliottii Engelm. var. elliottii x Pinus caribaea var. hondurensis Barr. et Golf of variable quality. Allometric foliage SAIs, estimated using an empirically derived relationship between basal (1.3 m) sapwood area and leaf surface area (R2 = 0.88. n = 11), ranged from 4.46 to 9.77. Gap fraction estimates (LI-COR LAI-2000 plant canopy analyser, PCA) of foliage SAI, when corrected for the nonrandom distribution of foliage elements, correlated well (R2 = 0.90) with direct estimates, being, on average, 11% lower. Relationships between foliage SAI (allometric) and spectral reflectance characteristics were tested. Strong, inverse nonlinear relationships between foliage SAI and thematic mapper (TM) bands 5 (1550-1750 nm, R2 = 0.9) and 3 (630-690 nm, R2 = 0.71), principal component (PC) 1 of bands 3, 4, and 5 (PC1, R2 = 0.86), and the composite of bands 3, 4, and 5 (R2 = 0.92) were observed. Positive linear correlations with the normalised difference vegetation index (NDVI. R2 = 0.81) and the simple ratio of TM band 4 to band 3 (SR, R2 = 0.78) were also observed. However, foliage SAI was poorly related (R2 = 0.02) to TM band 4 (760-900 nm). Strong relationships observed between various spectral reflectance characteristics and the more commonly measured forest structural parameters basal area and predominant height suggest that these remote sensing techniques are promising for use in early-aged inventory systems.