Abstract: With the support of ArcGIS and IDRISI platform, in combining with landscape ecology principles, and introducing landscape index and Kappa coefficient, this paper analyzed the landscape pattern and its dynamic degree of the Sanchuan basin in the southern mountainous region of Ji’nan, Shandong Province of East China in 1990-2010, and predicted the dynamic trend of the landscape pattern of Sanchuan basin with CA-Markov model. The landscape index analysis showed that the main landscape types in the basin were cropland, forestland, urban/built-up land, and grassland. Under the effects of human activities, the cropland and grassland in 1990-2010 decreased, while the others presented an increasing trend. Overall, the landscape fragmentation degree and patch complexity in the basin increased, and various landscape types tended to be distributed uniformly. The landscape pattern change was more obvious in the first ten years than in the later ten years, and human activities had more obvious effects on the local ecological environment. Kappa index suggested that the landscape change turned gradually from the variation of landscape quantitative proportion to the variation of the landscape spatial position under the conditions of the landscape quantity keeping relatively stable. Under the policy guidance since 2000, the overall landscape change gradually became more stable, and the effects of human activities on the local ecological environment decreased. The prediction results of CA-Markov model showed that in 2010-2020, the dominance degree and fragmentation degree of all landscape types would be declined, the balanced characteristics would be improved, and the amplitude of landscapes moving in or out would be not large, only occupying 3.08% of the total area. The transformation of landscape types would mainly occur at the patch borders. Overall, the landscape pattern change in the basin was mainly manifested in decrease of cropland and forestland and the increase of urban/builtup land, with the ecological environment becoming more fragile.

Key words: effective leaf area index, SPOT5, texture parameter, geostatistics, montane evergreen broad-leaved forest.