Microclimate of Developed Peatland of the Mega Rice Project in Central Kalimantan

Microclimate of Developed Peatland of The Mega Rice Project in Central Kalimantan (A Jaya,T Inoue, SH Limin, U Darung and IS Banuwa) : In Indonesia peatland covers an area of 16 to 27 Mha and this ecosystem is vitally linked to environmental and conservation issues, as well as its economic value for human survival. These peatlands are, however, the subject of various land use pressures, including forestry, agriculture, energy and horticulture. A field study was carried out 6 years after the end of failed peatland development project shows that inappropriate and unsustainable forms of peatland management have resulted in degradation of the natural forest vegetation, draw-down of the peat water table, increase of peat surface and air temperatures and recurrent surface and ground fires. Implications of microclimate for possible restoration options.


INTRODUCTION
Peatlands are wetland ecosystems characterised by accumulation of organic matter that is produced and deposited at a faster rate than it can be decomposed (Gore 1983). Peat formation in the tropics commences under conditions of constant waterlogging or in wet coastal areas where organic matter is produced in abundance by an adapted vegetation of mangroves, grasses or swamp forest trees (Driessen 1977;Radjagukguk 2000).
Approximately 12 percent of the global peatland area occurs in humid tropical zones, mainly in mainland East Asia, Southeast Asia, the Caribbean and Central America, South and southern Africa (Rieley et al. 1996). Peatland in Indonesia covers about 16 to 27 Mha (Radjagukguk 1992;Rieley et al. 1996). These ecosystems are vitally linked to conservation issues such as carbon sequestration affecting global climate change, and provision of key habitat for a diverse range of the world's flora and fauna. They are also the source of a significant portion of the freshwater and many economic resources vital to human survival. In addition, peatland ecosystems are important for water resources conservation since their ability to store water is very high and can be up to 8 times of the peat volume (Widajaja-Adhi 1997). Tropical peatlands, however are also the subject to land use pressures including forestry development and agriculture on them as well as extraction for energy and horticulture (MacDicken 2002). In 1996, 1.5 Mha of Peatland in Central Kalimantan was developed for rice production. This project failed and was closed down in 1999 but left on legacy of habitat destruction and fire occurrence.
Conservation of vast peat land area for the carbon stored in them as well as for water protection seems to be one solution for the future. It seems, however that even of conservation show a great and significant success, the loss of forest still occurs at the alarming rate. The involvement of local communities therefore is important to the success of conservation of tropical peat land.
The purpose of this research is to study the environmental condition of the developed peat land area and various impacts after the commencement of development of the ex-Mega Rice Project (the MRP) in 1996 especially on microclimate aspect.

MATERIALS AND METHODS
Research was conducted in the area of ex-MRP Central Kalimantan, which opened in 1996 through a Presidential Decree 82/1995 to develop the area of more than one million hectares for food production, especially rice. The Mega Rice Project covers an area of 1,457,100 ha area with a boundary Sebangau River in the West, Java Sea in the South, Barito River in eastern and Primary Channel Master (SPI) in the northern part ( Figure 1). The area of the MRP is divided into 4 areas of work, Block A (227,100 ha), Block B (161,480 ha), Block C (568,635 ha) and Block D (162,278 ha) plus the Block E (337,607 ha) as a buffer zone adjacent to North of SPI.
Data of peat soil surface temperature was measured with automatic recording devices at intervals of every one-hour observation, and conducted in forested peat, burnt forest (re-growing forest) in the northern part of Block C of the ex-MRP and agricultural areas in the Village of Kalampangan. Rainfall was measured using a measuring type of tipping bucket rain gauge set in forest areas, and open areas ex-fire in Block C of the ex MRP and Kalampangan village farming areas. Thermometer to measure soil and air temperature set at the same location with the logger and set to record every onehour intervals. Soil moisture was only installed in agricultural areas of Kalampangan village and observation sensors for soil moisture were installed at a depth of 10, 20, 30, and 40 cm.

Soil Moisture
Data for hourly soil moisture recorded from 30 September 1999 until 19 February 2004 in the agriculture area of Kalampangan village adjacent to the northern area of Block C of ex-the MRP are presented in Figure 2.
Data analysis using paired t-tests (Tables 1 and 2) show that the thickness of the peat layer has highly significant effects on soil moisture. The volumetric soil moisture at 10 cm depth varies from 13.50 to 68.98%, with an average of 39.87%. The soil moisture in this layer has the highest fluctuation as shown by the highest standard deviation (13.76%). At 20 cm depth soil moisture ranges from 37.84 to 68.64% with an average of 58.56%, showing a higher difference of 18.69% of soil water content compared to 10 cm depth. The results of correlation analysis between volumetric soil moisture in several peat layers and water table and air temperature are shown in Table 3. The volumetric soil moisture at all peat layers measured has a significantly negative correlation with the depth of water table. The deeper the site of measurement the stronger the correlation is. In contrast, temperature only shows a correlation with soil moisture at peat depths of 10 and 20 cm.
The peat in the agriculture area, based on continuous measurement, has a wide range of soil moisture from 0.1350 m 3 m -3 at a depth of 10 cm to 0.7107 m 3 m -3 at 30 cm. In comparison, the measurement at a depth of 10 cm, Hatano et al. (2004) showed that soil moisture in the agriculture area ranged from 0.326 to 0.367 m 3 m -3 , which was higher than the natural forest (0.175 m 3 m -3 ) and regrowing forest (0.216 m 3 m -3 ). Kurnain et al. (2002) stated that land use clearly influences the gravimetric field water content of peat because, in pristine forest of Central Kalimantan, Indonesia, it was 574% and significantly higher compared to clear cut peat forest (203%), cultivated peatlands (438%) and burnt peat forest (305%). The water table level has a significant effect upon soil moisture, which is higher closer to the water table. In contrast, temperature only affects soil moisture to a depth of 20 cm below the surface.
The moisture condition of peat soil has an important role in soil management since most physical characteristics are related to the moisture conditions. Soil moisture in peat soil directly influences the degree of subsidence, pore geometry, buffer capacity and soil thermal characteristics (Bouman and Driessen 1985). Soil moisture has a significant role in carbon balance. The emission rate of CO 2, for example, depends on the moisture condition (Toyota and Okazaki 2004). For the tropical peatland of Central Kalimantan, Indonesia, it was found that the largest CO 2 emission occurs at a water content of 65% in both natural and re-growing forest while, in the burnt area, it takes place with a water content of 85%. In addition, based on work in the same region, Hatano et al. (2004) found that soil moisture positively affects methane flux, but has only a moderately significant correlation with NO 2 flux.

Temperature Air Temperature
Hourly measurements of air temperature were recorded from September 2000 to May 2005 within the forest, regrowing forest, destroyed forest and agriculture area. A continuous record of data is not available because the equipment malfunctioned at certain times. From the data available the variations in average daily temperatures under the four types of land cover are presented in Figure 3. Table 4 shows the average daily temperature and its fluctuation in the agriculture area and destroyed forest are higher than in the forest and regrowing forest. In the agriculture area, the average temperature ranged from 23.79 to 31.    Temperature (  The paired sample t-tests, presented in Table 5, show that there are significant differences in the means of daily temperature between each location, except between agriculture area and destroyed forest. Table 6 shows the data of average hourly air temperatures were obtained within regrowing forest, agriculture area, destroyed forest and forest area. The highest average hourly temperature and its fluctuation were in the regrowing forest (  The pattern of diurnal temperature fluctuation in all four land cover types was similar. Figure Figure 4. Frequency analysis of daily air temperature within several land-uses.

Peat Surface Temperature
The data for peat surface temperature (0-20 cm) are only available for the agriculture area, forest and regrowing forest from 1 st March to 15 th July 2002 owing to malfunctioning of the thermometer. The peat and air temperatures at the three locations are depicted in Figure 8 and a summary of descriptive statistics is presented in Table 7. Peat surface temperature is higher than air temperature in the agriculture area and regrowing forest while, in the forest, peat surface temperature is lower than air temperature. The highest average surface peat temperature of 30.22 o C occurs in the agriculture area, following by regrowing forest (26.71 o C) and forest (22.88 o C).
The analysis of means by using paired samples test between air and peat temperature in agriculture area, forest and re-growing forest and also peat surface temperature between the three locations is presented in Table 8. This shows a highly significant different between the average air and peat surface temperatures within each location and between all locations.
The pattern of diurnal peat surface temperature fluctuations in the three land cover types differs. The  Vegetation cover on tropical peatland has a significant influence on the average daily air and peat surface temperature at the three study sites. Open peatland areas, such the agriculture area with an average of 27.62  1.38 o C and destroyed forest with an average of 27.20  1.12 o C have much higher temperatures and variation than those with a vegetation cover (Rieley et al. 1996;Takahashi and Yonetani 1997) such as regrowing forest (26.04 ± 0.81 o C) and forest site (25.72 ± 0.72 o C). Similar ranges of temperature were also found by Hirano et al. (2003) who obtained an air temperature of 26.5 o C in a forested area at a height of 41.7 metres above the peat surface at the northern area of Block C of ex-the MRP.
Within the forested area, the average peat surface temperature was 2.96 o C lower than the average air temperature while, for the developed peatland area (e.g. the agriculture and re-growing forest areas), the average peat surface temperature was between 1.09-2.97 o C higher than the average air temperature in the same locations. From these data, it is clear that the loss or changes of vegetation cover in peatlands is strongly influencing the peat surface temperature. The existence of plants that grow on the peat swamp alters the heat and water balance of the soil on which they grow. It has been found, for example, that peat swamp forest reduces the ground albedo significantly contributing to the frequency of rainfall that decreases when forest is cleared. With reference to radiation exchanges, there are marked differences between forested and deforested areas with respect to their solar radiation exchange (Silvius and Giesen 1992).  Variations also occurred in both air and peat surface temperature under different land conditions. The highest variation was in the agriculture area, followed by destroyed forest, re-growing forest and the forested area. This again supports the importance of vegetation cover in controlling temperature fluctuation. Compared to air temperature, however, peat surface temperature is higher than air temperature but shows a smaller variation for all sites.
The higher temperatures will have an effect to the increase of oxidation process and, furthermore, will lead to higher gas efflux rates, especially CO 2 and methane. In sub arctic/boreal fen, Moore et al. (1990) reported that soil temperature has a positive correlation with methane flux, while Otter and Scholes (2000) found a similar relationship between methane flux and temperature for the wetlands of South Africa. In tropical peatland of Central Kalimantan, Indonesia, Hatano et al. (2004) reported that soil temperature has a moderate significant correlation with CO 2 flux and N 2 O as well as a positive correlation with methane flux.

CONCLUSIONS
The results of this study showed changes in land use as a result of the MRP activities cause changes in microclimate conditions in the peatland area. Fluctuation of air temperature and peat surface temperature significantly affected by the condition of vegetation covered on it. The loss or reduction in vegetation cover is significantly affect the peat surface temperature and this will also related to the fluctuations on peat surface moisture. Peat surface moisture Table 8. Paired samples test between air and peat temperature within each land use and peat temperature between three land uses.
Note: AgrA = Air temperature at agriculture area, AgrP = Peat surface temperature at agriculture area, ForA = Air temperature at forest area, ForP = Peat surface temperature at forest area, RegA = Air temperature at regrowing forest area, RegP = Peat surface temperature at regrowing forest area.
fluctuations are correlated with fluctuations in air temperature and peat surface temperature.