Forest fuel load modeling in the Iguaçu National Park, PR - Brazil

Abstract

The objectives of this research were to measure and model the surface forest fuel in the Iguaçu National Park. A total of 133 plots (1.0mx1.0m) were located in the Park area in order to collect the information needed to develop the investigation. Live and dead woody surface fuels were collected, separated and classified, generating the following groups: miscellanea (MIS); dead fuel, diameter ≤ 0.7 cm (L1M); dead fuel, diameter > 0.7cm (L2M); total woody dead fuel (LTM = L1M + L2M); total dead fuel (MLTM = MIS + LTM); live fuel, diameter ≤ 0.7cm (L1V); live fuel, diameter > 0.7cm (L2V); total woody live fuel (LTV = L1V + L2V); and total fuel load (CTC = MLTM+LTV). The independent variables measured were: litter depth (EMO); average DBH (DAP); basal area (G); and trees average height (H). The total fuel load measured through the sample plots varied from 1.75 to 21.72 ton.ha^-1, with an average of 11.74 ton.ha^-1. Models generated through the forward stepwise methodology could be considered a first approach to estimate the fuel load based on some easily obtained independent variables in the Iguaçu National Park. The best model to estimate total dead fuel load was (R² = 0.57): MLTM = 951.639 + 20.179 (EMO³) – 800.441 (1/EMO) – 0,090 (G² *EMO) + 14.982(G)and the best model to estimate the total fuel load was (R² = 0.58): CTC = 1162.193 + 19.355 (EMO³) – 925.90 (1/EMO) – 0.086 (G² *EMO) + 13.980 (G). It was not possible to generate a model to estimate live fuel load due to the lower determination coefficients obtained.