Estimating leaf area index of a degraded mangrove forest using high spatial resolution satellite data

Abstract

Leaf Area Index (LAI) values from 124 mangrove plots were acquired within a degraded mangrove forest of the Agua Brava Lagoon System of Nayarit (Mexico) using a hand held LAI-2000 Plant Canopy Analyzer. For each plot, two values of LAI were calculated to represent approximate half radii (180°) ground coverage of 8 m and 15 m. The location of each plot was recorded at sub-meter accuracy using an Ashtech SCA-12 GPS. Using a geometrically corrected IKONOS satellite image, the mean values for both the Normalized Difference Vegetation Index (NDVI) and the simple ratio (SR) vegetation indices were also calculated for each plot. Regression analyses of the in situ LAI with both vegetation indices revealed significant positive relationships (LAI versus NDVI at 8 m ( R 2 = 0.71); LAI versus NDVI at 15 m ( R 2 = 0.70); LAI versus SR at 8 m ( R 2 = 0.73); LAI versus SR at 15 m ( R 2 = 0.72)) at the 8 m and 15 m plot sizes. Standard errors, derived from the testing of the regression models with a random sample, revealed little difference between the models. Moreover, F-tests of the residual variances also indicated no significant difference between the SR and NDVI models at both plot sizes. Consequently, the results indicate that each model could be used to successfully predict LAI. It is thus suggested that high spatial resolution IKONOS data can be employed as a valuable tool for monitoring LAI in less than ideal mangrove forests (i.e. disturbed stands).