Leaf Area Assessments of Overstory and Understory Vegetation in Pine Plantations Located in South Georgia and North Florida, US

Abstract

Leaf Area Index (LAI) was estimated in summer 2005 and winter 2006 for overstory and understory in loblolly pine and slash pine plantations at ages 7 and 10 year-old and on poorly, somewhat poorly and moderately-well drained soils located in the flatwoods region. Additionally, stand and site factors such as basal area, pine dominant height, understory height and understory coverage were estimated for each of the 40 plots established, and leaf area index and vegetation indices (SR, NDVI, VI and EVI) were calculated using remote sensing imagery. The objectives of this study were to determine the understory (competing vegetation) and overstory (crop-trees) leaf area index, to relate the variation in understory and overstory LAI to stand and site factors and to examine the relationships among understory and overstory leaf area index and spectral reflectance data captured by satellite imagery. Leaf area index values observed for the overstory were low in most of the plots (around 2 m2m-2 in slash pine and around 3 m2m-2 in loblolly pine), while the understory LAI was very high (around 2 m2m-2), which can be attributed to the lack of canopy closure observed in all plots. A negative relationship was observed between the overstory and the understory, where the higher the understory LAI the lower the overstory LAI. No significant differences were found in the understory LAI values across soil drainage classes. Low heights and short crown lengths were generally observed and could be explained by nutrient deficiency in most of the sites; which could be attributable to the belowground competition for water and nutrients. LAI and basal area were not correlated. Total LAI (overstory LAI plus understory LAI) estimated values on the ground were high and weakly correlated with the Landsat-derived vegetation indices, and the LAI values estimated with a LAI model were typically half of the values estimated on the ground. These results could be influenced by the contribution of different backgrounds, such as soil moisture and understory vegetation, plus the saturated response of the vegetation indices at high LAI values. Significant correlations were observed between the vegetation indices (SR and NDVI) and stand and site factors, suggesting that the satellite derived indices were more related to the stand biophysical parameters than in situ LAI estimates.