RESUMO

While around 20% of the Amazonian forest has been cleared for pastures and agriculture, one fourth of the remaining forest is dedicated to wood production. Most of these production forests have been or will be selectively harvested for commercial timber, but recent studies show that even soon after logging, harvested stands retain much of their tree-biomass carbon and biodiversity. Comparing species richness of various animal taxa among logged and unlogged forests across the tropics, Burivalova et al. found that despite some variability among taxa, biodiversity loss was generally explained by logging intensity (the number of trees extracted). Here, we use a network of 79 permanent sample plots (376 ha total) located at 10 sites across the Amazon Basin to assess the main drivers of time-to-recovery of post-logging tree carbon (Table S1). Recovery time is of direct relevance to policies governing management practices (i.e., allowable volumes cut and cutting cycle lengths), and indirectly to forest-based climate change mitigation interventions.

Assuntos

RESUMO

Utilizando um modelo estocástico, foi projetada a distribuição diamétrica futura de uma floresta submetida à exploração seletiva de madeira na Amazônia Ocidental. Foram utilizados dados de cinco parcelas permanentes localizadas no PC Pedro Peixoto, no Acre. A primeira medição das parcelas ocorreu em 1996, a exploração florestal em 1997 e as re-medições em 1999 e 2001. A principal variável utilizada foi o diâmetro à altura do peito (DAP). A matriz de transição probabilística (Cadeia de Markov) foi utilizada para fazer a projeção da distribuição diamétrica do número de árvores sobreviventes nas classes diamétricas. O modelo foi primeiramente testado para fazer a projeção para 2001, tendo como base as observações de 1999 e seu passado imediato (1997). Quando comparadas às projeções feitas para 2001 e as medições de campo (2001), o teste Qui-quadrado mostrou que não houve diferença significativa entre freqüências esperadas e observadas na distribuição diamétrica (p=0,05). A projeção para 2005 indica que a taxa de mortalidade será próxima a de 2001, e se repetida a taxa de recrutamento em 2005 o total de árvores será maior que o observado em 2001. Esse comportamento da floresta indica que não existe um padrão definido para a dinâmica nas classes diamétricas em termos de mortalidade ou crescimento, apresentando um comportamento aleatório ou probabilístico, justificando a eficiência da Cadeia de Markov para projetar a dinâmica da floresta estudada, podendo auxiliar na determinação do ciclo de corte ou mostrando as tendências que a floresta de hoje apresentará em um futuro próximo.

The diameter distribution of an experimental forest stand in the Western Amazon was projected using a stochastic model after selective logging. The study was developed using data from five permanent plots located in the colonization project Pedro Peixoto, in the state of Acre. Initial measurements of diameter at breast height (DBH) were taken in 1996. The forest was selectively logged in 1997 and DBHs were re-measured in two different occasions, 1999 and 2001. A probabilistic transition matrix (Markov Chain) was used to project the diameter distribution of the number of surviving trees in each diameter class. The model was first tested to project the diameter distribution in 2001, based on DBH measurements from 1997 and 1999. When the projected diameter distribution for 2001 was compared with the field data from the same year, a Chi-squared test (á = 0.05) showed that there was not significant difference between the expected and observed diameter distribution. After that, a projection for 2005 (four years in the future) was run using DBH measurements from 1997 to 2001, indicating that mortality rate was similar to 2001. If repeated the rate of recruitment of 2005, the total number of trees will be higher than observed in 2001. The dynamics of the studied forest suggests that there is not a definitive pattern to changes in diameter distribution and mortality, which indicates a stochastic or probabilistic pattern. This pattern is better modeled by the Markov Chain to project the forest dynamics of studied area, and can help on determination of timber harvesting or the tendencies of forest dynamics in a near future.

Assuntos