Plant-light ecological relations have been studied since the early 20th century. Light within forests has been studied since then. At community level, light environment analysis has been carried out by indirect cover measurement, but without any photogrammetric methods. This work aims to apply the canopy photogrammetric method in light environment studies of understory. Correlations among edaphic, cover and species variations were investigated to understand the understory environment. These studies were carried out in a sampled area for understory phytosociological analysis. The black and white canopy photographs were taken for cover estimation. Photographs were taken in dry and rainy seasons at 100 sample quadrats. Soil samples were taken at these quadrats. The cover average for sampled species was calculated and "t" student test was applied for testing the significance of difference between specific and populational means. Correspondence Canonical Analysis was applied to determine correlations among species, edaphic factors and cover. Cover was not significantly different in dry and rainy seasons. Only three species Heisteria silviani, Calathea brasiliensis and Psychotria conjugens, had mean cover greater than the populational mean. Other three species, Olyra micrantha, Lacistema pubescens and Pteris denticulata, had mean cover smaller than the populational mean. The sample quadrats frequency distribution and the gap size frequency distribution were similar to those found in other tropical forests. Thus, the canopy photogrammetric method was appropriate for the indirect evaluation of light environmemt of tropical seasonal and rain forests. The Correspondence Canonical Analysis showed that calcium, magnesium and potassium grades were positively correlated among themselves and negatively correlated to aluminium grades. Cover was negatively correlated to aluminium grades. Five species groups were detected by preferences to calcium, magnesium, potassium and aluminium grades and cover percentage. The aluminium grades are the main variable relatated to the herb-shrub vegetation changes. The edaphic factor variations were more important than the cover percentage variations in determining structural alterations of forest herb-shrub layer.
canonical correspondence analisys; canopy gap; light; seasonal semidecidual forest; Atlantic Forest; phytosociology
