CLUE Philippines 

A new modelling tool was developed called CLUE-S: The Conversion of Land Use Change and its Effects at Small regional extent. The CLUE-S model is applied to the island Sibuyan in the province of Romblon, Philippines. The island measures 28 km east to west at its widest point and 24 km north to south, with a land area of approximately 456 km² surrounded by deep water. In the centre of the island lies a large protected area (Mount Guiting-Guiting Natural Park). It is characterised by its steep mountain slopes, covered with forest canopy. The land surrounding the high mountain slopes gently to the sea and is used for natural and plantation forest and agricultural, mining and urban activities.

Sibuyan is an so-called "Ice-Age" Island, one of the few Philippine islands that had already separated by deep water from any other land mass some 10000 years ago. Because of this separation the plant and animal life on islands like Sibuyan evolved independently, producing many unique life forms.

The island is believed to be completely covered by forest until the 1940's. From then on the forest has been cleared from the footslopes. Highest on the footslopes are the grassland derived from deforestation, used for pastures. They are regularly burnt to stimulate new grass growth. Rice paddies are common at low-lying land. Most cleared areas are however used for coconut plantation. The island is surrounded by some mangrove forests, sandy beaches and coral reefs. See also figure 2A.

CLUE-S evaluates a base line scenario for Sibuyan for 15 years (1997-2012). Higher resolutions used here demand a different data representation and modelling approach. Where the low-resolution data made use of non-homogeneous grid cells, this model version is based on homogeneous grid cells. The scale is 250x250 m.

A statistical analysis based on logistic regression is used to find the most important determinants of the land use patterns. Five to ten driving factors are necessary to explain the four dynamic land use types, coconut, forest, grass and rice. The population density, which acts over large distances, is taken into account by using an upscaling of grid information by a focal function.

The demands that are used are the linear extensions of the development between 1940-1997. The assumption is that the island is first opened up from 1940 on and that a linear development has occurred until 1997. The land use changes during these years are visualised in figure 1.

The model predicts the further development along the footslopes of the mountains especially in the west and the north (figure 2B). Especially the coconut plantations expand towards the northern part of the island. The northwestern part consists of very steep slopes, too steep for coconut, but some patches of grassland are developing. Some of the rice paddies will move to the southwest, and also more new paddies will develop here. The hot-spots of change are indicated in figure 3 (black-spots).

Figure 3 Hot-spots of change at Sibuyan (differences between figure 2A and B indicated in black).

The black line in figure 2C indicates the NIPAS-area. This is the border of the Mount Guiting-Guiting National Park. Without taking the border into account, the park will on the long term slowly be invaded for agricultural purposes, especially on the east side. However, it also shows that new agricultural developments mainly occur at the west side of the island.

The model is also capable of excluding areas from land use change. This can be used in the case of a protected area, such as the Mount Guiting-Guiting National Park. 2 simulations are done using this tool. The first is using the current park as exclusion zone (figure 2C); the second is using the area of an imaginary park that might arise under a more strict policy of nature conservation (figure 2D).

These results are compared to the results of the default scenario. Like was expected, there are no differences between figure 2B and 2C. But even the differences between 2B and 2D are not large. Rice that would be grown at the west side inside the park borders will now be grown at the east side. There is a new side where grass will appear, near the border of the park in the south. The results are an indication that the areas were forest occurred in 1997 are unfavourable for other land use types, for instance due to steeps slopes and soil with potentially toxic concentrations of magnesium and nickel. The changes that other land use types will cross the borders of the present park in a serious way are not high.

Download full report (1.8 Mb)

back