Influence of meteorological variables in litter production and decomposition at Ferreira Penna Scientific Station, Caxiuanã, PA.
Silva, Museu Paraense Emílio Goeldi, email@example.com
da Costa, Universidade Federal de Viçosa, firstname.lastname@example.org
Maria de Lourdes
Ruivo, Museu Paraense Emílio Goeldi, email@example.com
da Costa, Universidade Federal do Pará, firstname.lastname@example.org
Almeida, Museu Paraense Emílio Goeldi, email@example.com
The evaluation of influence of exclusion of rainwater on the variation in litter production was made at the forest reserve of Caxiuanã-Ferreira Penna Scientific Station (1o 42’ S; 51o 31’ W) during the period of March, 2001 to February, 2003, to identify the main meteorological and monthly water balance variables that affected the production and the decomposition of litter. This work is part of a sub-project drought experiment in the forest (ESECAFLOR) which aims to study the long term impact of drought on the mass and energy fluxes in the forest. The seasonality of litter production and its components (leaves, twigs and reproductive parts was well characterized, with the occurrence of greater production during the months of less wet season. The total monthly production of litter ranged during the experimental period from 297.78 kg ha-1 to 1,758.69 kg ha-1, with an average value of 777.70 kg ha-1 distributed in the fractions of leaves, twigs and reproductive parts in 61.40%, 18.45% and 20.14%, respectively. The results obtained in the plot under natural conditions were approximately 25% higher than the litter production values obtained in the plot under drought stress due to the rainwater excluded in the plot. The litter decomposition rates for the wide, average and thin mesh were described satisfactorily by the exponential model submitted by Olson (1963). The variables more strongly correlated with the litter production and or its components were the wind speed, the global radiation, the photosynthetic active radiation, the soil temperature (at 5 cm depth) and the precipitation. Among the water balance variables the better correlations were obtained with the soil moisture content, water surplus and water deficit. The regression analysis between total litter production and its components with the meteorological and or monthly water balance variables were not satisfactory for prediction purposes.