SIZE AND SHAPE OF SAMPLE UNIT FOR LITTER QUANTIFICATION OF A SUBTROPICAL SEASONAL FOREST FRAGMENT

To study the nutrient cycle in a forest, a precise estimation of the amount of litter produced and its fractions is necessary. Among the recognized ways to minimize variability between samples is the use of optimal sampling size units to litter collection. However, there are a few recommendations found in another studies for this parameter, which generates variation in the size of the sampling unit used in similar experiments. The aim of this work is to determine the optimal size and shape of the sampling unit for quantification of litter in a Subtropical Seasonal Forest fragment. For this purpose, five plots of 3.0 x 2.0 meters in each season were randomly allocated in the area, subdivided into 96 basic units of 0.25 x 0.25 meters. In each basic unit all organic matter above ground was collected and separated in the fractions leaves, twigs and waste. The estimative of the basic units optimal size was obtained by two different methods. It was observed that the coefficient of variation decreases as the size of the sample unit is increased; there are differences in the production variability of litter between the plots in the same collection station; rectangular sampling units are more homogeneous than the square ones; the use of sample units 0,40m(2) (1.5m x 0.25m) is recommend for litter quantification produced in Subtropical Seasonal Forest fragment.