摘要

Forest landscape restoration (FLR) is a process that aims to regain ecological integrity and enhance human well-being in deforested or degraded forest landscapes. To ensure that restoration efforts are successful, the first step is to understand the dynamics of the forest landscape and the dominant forces responsible for its change. Taking Yong'an city, Fujian province in China as a case for study, this paper constructed a Markov model to predict the dynamics of the forest landscape based on sample-plot data of Continuous Forest Inventory at a county level. The study area was divided into eight landscape element types based on FLR, including approximated primary forest, secondary broad-leaved forest, secondary forest of Pinus massoniana, natural bamboo forest, planted forest, non-timber product forest, degraded forest land and non-forestry land. The analysis showed the following: (1) the extent of reforestation of planted forest, non-timber product forest and secondary forest of Pinus massoniana would be greater than that of deforestation of approximated primary forest, broad-leaved secondary forest and natural bamboo forest. Therefore, the total area covered by forest would increase steadily. (2) Conversely, conversion among different landscape element types would occur frequently and have high transition proportions. (3) Remarkable decrease of the extent of approximated primary forest, together with the conversion from degraded forest land to secondary forest, would probably result in the decline of forest volume. (4) Forest productivity in the meantime will not be maintained or enhanced because of the conversion from secondary forest to planted forest. These results suggest that the direct and underlying driving force of landscape dynamics should be understood and addressed in the upcoming studies for remnant approximated primary forest protection, secondary forest management and degraded forest land rehabilitation. The conclusion is that the Markov model can be used to analyze the forest landscape dynamics for FLR based on sample-plot survey data of Continuous Forest Inventory at a county level.