Um inventário foi conduzido em Três Barras, Santa Catarina, para quantificar e modelar material combustível superficial vivo e morto em plantações de Pinus taeda. De forma sistemática foram estabelecidas 20 parcelas para cada povoamento de 3, 5, 7, 9, 11, 13, 15, e 17 anos de idade, nas quais foram medidos a carga de combustível, o DAP, a altura e o diâmetro dominantes, a área basal e a espessura da liteira. Nove modelos foram ajustados através dos métodos Stepwise e “todas as regressões possíveis”. O melhor modelo para estimar a carga de acículas foi Wa = 2,356794 * S (R2 = 0,9563), ajustado com base na espessura (S) da camada de acículas. Para a carga de combustível morto de até 2,5 cm de espessura o melhor modelo foi W10 = 1,889811 + 1,569117 * S + 0,145218 * S 2 (R2 = 0,8577). O modelo considerado mais adequado para determinação da carga total de combustível foi WT = 3,35892 * S − 1,87453 (R2 = 0,7034). Três gráficos para estimativa indireta do combustível e um calibrador para estimar a carga de acículas são também apresentados.
A fuel inventory was conducted in Três Barras county, state of Santa Catarina, Brazil (26o15’ S latitude and 50o48’W longitude), in order to develop predicting models for fuel loading in loblolly pine (Pinus taeda) plantations. Sampling was done in 03, 05, 07, 09, 11, 13, 15, and 17-year-old stands, covering the whole rotation of the plantations. Twenty (20) plots of 1.0m2 (1.0x1.0m) were randomly located in each stand, totalizing 160 plots. The independent variables measured in the stands were age (in years), mean DBH (cm), dominant DBH (cm), dominant height of the trees (m), basal area (m2), and fuel bed depth (cm). The dependent variables were live surface fuel, dead foliage (needles), dead woody fuel (separated by size classes), and total fuel load, all measured in ton.ha-1 (oven dry weight). Results showed good correlation between fuel bed depth and age and most of the dependent (fuel related) variables. Live surface fuel only presented significant correlation with mean DBH and basal area. However, live surface fuel was only significant in the 3-year-old plantation, and practically disappeared when the pines canopy intercepted most of the sun light, usually after the 5th year. Models were developed to estimate the following variables: dead foliage (Wa) dead fuel up to 10-hour (∅ ≤ 2.5cm) timelag (W10), and total fuel load (WT). Independent variables were chosen not only by the correlation coefficients, but also for the measuring facility, and based on these principles, fuel bed depth (S), age (I), dominant height (hdom), and dominant DBH (Ddom) were selected. The models were built through the Stepwise method, using the Statistic 5.0 software. Fuel bed depth was the variable that provided best fits for all predicting models, and the inclusion of other independent variables did not improve the models precision. The best model to estimate the needles load was Wa = 2,356794 * S (R2 = 0,9563); to estimate dead fine fuel load was W10 = 1,889811 + 1,569117 * S + 0,145218 * S 2 (R2 = 0,8577); and to estimate the total fuel load was WT = 3,35892 * S − 1,87453 (R2 = 0,7034). Total fuel load in the loblolly pine plantations in the studied site ranged from approximately 6.0 to 21.0 ton.ha-1.
