The Impact of canal blocking to reduce fire risks and carbon emissions on tropical peatland, Siak District, Riau Province

Tropical peatlands are wetland ecosystems formed from the accumulation of organic matter over thousand of years period. Indonesia has an area of about 13.5 million ha and play important roles for society and the environment. The development of drainage canals has caused peatlands to become dry and degraded, rendering them highly susceptible to fires. Peatland restoration through rewetting activities with canal blockings can restore the hydrological function of peatlands. In addition, groundwater level (GWL) also affects carbon emissions from peatlands. This study aims to determine the distance of canal blocking effect on groundwater level so that it can be known which areas have a lower risk of fire and carbon emissions in peatlands. This study compared areas affected by canal blocking with those without canal blocking. The results of this study show the significant effect of canal blocking in increasing the groundwater level in areas with <100 m distance from the canals and with different types of land use. The average GWL of peat during one year of monitoring period was around - 26.67 ± 2.4 cm at the location of the monitoring well close to the canal with canal blocking, equivalent to carbon emissions of 26.5 tCO2eq ha−1 year−1. This is lower than the average GWL of those areas without canal blocking that was - 58.67 ± 3.1 cm, which is equivalent to carbon emissions of 57.8 tCO2eq ha−1 year−1.


Introduction
Peatland is a type of soil formed from the accumulation of organic material deposits formed naturally from the weathering of vegetation that grows above the soil surface over a long period of time.Peatlands have multifunctional functions, namely hydrological, production and ecological functions that are vital for human survival [1].The decomposition of organic matter takes a long time due to the anaerobic and wet conditions of peatlands.Peat has unique physical properties, namely being able to absorb very large amounts of water.It is estimated that pristine peatlands consist of almost 90% water and the remaining 10% is the remains of decomposed plant material or vegetation [2].
In addition to playing an important role in regulating soil hydrology, especially in maintaining and regulating the water the soil, peatlands also play an important role in storing carbon reserves.The condition of peatlands, which asam dengan lowyang oxygen levels sehingga slowed down the 1315 (2024) 012058 IOP Publishing doi:10.1088/1755-1315/1315/1/012058 2 decomposition yang terjadi this has caused peatlands to become one of the largest carbon stocks in the world.Peatlands can store ten times more carbon than mineral soils depending on the thickness of the peat layer [3].Indonesia has 13.5 million ha of peatlands, about half of the tropical peatlands, which represents 3.5% of the 400 million hectares of peatlands in the world [4].The largest peatland area in Indonesia is in Sumatra, with an area of about 6.4 million ha, followed by the islands of Kalimantan (4.7 million ha) and Papua (3.7 million ha) [5][6].This potential peatland area makes Indonesia one of the countries as a high storage of terrestrial carbon reserves with around 57 Giga Tons [4].Riau is the province on the island of Sumatra that has the largest peatland, which is 3.89 million hectares out of 6.49 million hectares of total peatland area on the island of Sumatra.However, around 2.31 million hectares have been degraded .The total area of 934,130 ha of degraded peatland that has not been utilized, about 585,217 ha, has the potential to be developed as plantation, food and horticultural land.Deforestation and land degradation have threatened peatland ecosystems over the past few decades [7][8], indirectly affecting the ecosystem structure and reducing the productivity of the land [9].Previous analysis shows that the rate of forest loss in Indonesia is around 1,000 km 2 per year to 5,000 km 2 per year [10].This change in land cover has degraded the forest into various types including degraded forest, shrubs and agricultural and plantation cover types on peatlands.
The conversion and degradation that occurs on peatlands that do not apply a sustainable management system, with the construction of drainage canals, causes the condition of peatlands to dry easily due to the increase in the rate of water flow and reduce the water storage capacity.The dry condition of the peatland will make it vulnerable to fire.In addition, the degraded peatlands are one of the triggers for the peatland subsidence and the lowering of the water table.This has caused Indonesia's peatlands to shift from carbon sinks to carbon emitters from forest and peatland fires [11][12][13][14][15][16][17][18][19][20][21][22].Peatland fires occur almost every year in Indonesia, causing haze disasters that affect the economy and public health.Fires also cause a decrease in peat thickness to 10-15 cm and result in the death of flora and fauna [12].
The Groundwater level (GWL) in peatlands is an important parameter in determining the hydrological condition of peatlands.The water table in drained peatlands is known to significantly affect the level of greenhouse gas (GHG) emissions from peat decomposition [13][14].In addition, GWL is also a key parameter in determining or assessing fire risk in peatlands [8][9].Carbon loss models have been developed as a function of the water table from plantation areas on peatlands [11][12][13][14].In addition, few studies have investigated how land use and topography can affect the groundwater table (GWL) in the context of peatland domes [15][16][17].Presidential Decree No. 1 of 2016 regulates peatland management by restoring degraded peatlands through rewetting [18].Rewetting activities need to be carried out to restore the hydrological conditions of degraded peatlands.Canal blocking is one way to restore the hydrological conditions of the peatland.
The rewetting of peatlands with the construction of canal blocking has become a national agenda in restoring degraded peatlands with the main objective of overcoming land fires due to damage to the peat ecosystem.This study aims to examine the impact or influence of canal blocking on the dynamics of groundwater level (GWL) and carbon emissions in peatlands.This research was conducted by comparing the groundwater level on peatlands where canal blocking has been built with those where canal blocking has not been built.In addition, this research also aims to estimate the impact of land use change on peat groundwater level.

Study site
Siak District is one of the districts in Riau Province with a peatland area of 56% of the total area of Siak District, which is 855,609 ha.In general, the types of land cover and land use of each peat area in Siak District have different types of land cover and land use.The land cover and land use conditions of peatlands in Siak District include oil palm plantations, secondary swamp forests, open land, industrial plantations, conservation areas/primary swamp forests [19].The climate of Siak District is classified as an isothermal climate with tropical rain according to the Koppen system.The average air temperature is between 25 °C and 32 °C and the average annual rainfall is 1099 mm (BPS Siak District 2015).
This research was conducted in Bunsur village, Sungai Apit sub-district, which covers an area of 12,386 hectares with almost 20% peatland, and Bunsur village itself is almost 80% peatland with varying depths.This research was conducted in an area of peatland that has been degraded by the construction of drainage canals and changes in the land cover of the peatland in Bunsur village.In this study, a groundwater level observation (GWL) was made by comparing canals that had been blocked with canals without blocked canals.

Groundwater table and carbon estimation
This research was conducted by measuring the characteristics of peatlands in various types of land cover namely: oil palm plantations, rubber plantations, sago plantations and shrubs.Observations of the peat water table were made by constructing monitoring wells at a distance of 5 m, 100 m, 400 m and 900 m from the canal or canal block at each observation transect.The monitoring well material (PVC pipe) was cut into 2 m lengths with an electric disk, and sharpened at one end.Each pipe had to be drilled to make holes along the pipe, spaced 25 cm apart on four lines.Aluminum mesh is used to cover the pipes, from the bottom to 20 cm before the top.The bottom of the pipe is covered with a lid that has been perforated in the center.The aluminum mesh and cover will limit the peat material from entering the pipe once it has been installed in the ground.At the measurement site, the peat should be drilled with an eijkelkamp drill to a depth of 2 m and the organic material removed.Then the PVC pipe is inserted into the hole, leaving only 0.2m of the pipe above ground level (figure 4).Another cap, without holes, was placed on top of the pipe to prevent rainwater from entering the monitoring well.Observations of the peat soil water table (GWL) were made manually by using a measuring stick [20].First, measure the height of the water table from the end of the pipe (B) (Figure 1).Then the height of the pipe from the ground must also be measured (A).GWT is the difference between B and A. Measurements are taken periodically, twice a month.

4
Observations of land elevation were also made to see the influence of land elevation on the dynamics of the water table (GWL) on peatlands.Observations of land elevation were made by direct levelling by measuring the difference in height between each point in a transect.Using a simple tool made from plastic pipes filled with water.The difference in height between each point or location becomes the difference in topography at that point (Figure 2).In addition to the measurement of groundwater level dynamics, the CO2 emission rate was also estimated using the carbon loss model of GWL adapted from Carlson et al [17] for palm oil (Elaeis guineensis) plantation: CL=5.4+0.21xGWLx-100*3.67 and Hooijer et al. [15] for acacia (Acacia mangium) plantation and deforested peatland: CL = 21-69 GWL and CL=9-84× GWL Where CL means total carbon loss (t CO2eq ha-1 year-1) and GWL means peat water table (GWL in m).The model was used because there are evidences that peat GWL significantly influences CO2 emissions in plantations on drained tropical peatlands [11,14].
This study examined differences in peat water table levels based on the impact of deforestation and canal blocking.The characteristics of the deforestation areas are plantations, shrubs and dryland agriculture.The characteristics of the deforested areas are plantation, shrubs and dry land farming, while the canal blocking is used to compare canals that are blocked and canals that are not blocked.

Analysis data
In this study, the analysis of differences in mean groundwater levels and land cover was tested by analysis of variance (ANOVA).The Kruskal Wallis test was also used for data that were not normally distributed.All statistical tests were performed using IBM SPSS statistics.

The tip of the water pipe is leaves open (1)
(2)

Ground water level and emission carbon
This research was conducted on degraded peatlands with various types of land cover .Based on the land cover map from the Ministry of Environment and Forestry 2020, the research plot was dominated by plantations, mixed land agriculture and swampy shrubs or bushes ( Figure 3).The type of land cover in the form of plantations consists of oil palm plantation with a productive age ranging from 3-10 years.Besides oil palm, other plantation commodities are sago (Metroxylon sago) and rubber (Hevea brasiliensis ) which are also categorized as productive.Meanwhile, the type of mixed dryland agricultural cover consists of commodities or pineapple plants (Ananas comosus ) and also mixed agriculture from paludiculture plants.This location is also an intervention area for Winrock International in sustainable peatland management activities by establishing paludiculture-based demonstration plots on peatlands.

Figure 3. Map of land use types in the research location
Siak district consists of 50% peatland, consisting of deep peatland and shallow peatland.In this study, most of the research plots were on deep peat with depths above 3 meters.Based on field measurements of peat depth at the research site, 90% of the plots were on deep peat with depths above 3 m, and only two plots were on shallow peat with depths ranging from 0-2 m (Figure 4).
The results of the analysis of the average peat water table at the research location showed significant differences with a p-value <0.05, this means that there is a significant difference between the peat water table height at one type of land use (Figure 5).The average water table height in the plantation type of land use was 44.5 ± 3.1 cm for palm oil, 24.4 ± 4.2 cm for sago and 19.6 ± 2.4 cm for rubber.As for the type of shrub land cover, the average groundwater level reached -25.7 ± 2.6 cm.For dry agricultural land cover types, namely in the form of mixed paludiculture agriculture, the groundwater level reaches -24.3 ± 2.2 cm.
The level of the peat water table also reflects the CO2 emissions from decomposition that occurs on peatlands [14][15].The significant difference between peat water table and land use type also reflects the difference in CO2 emission rate.Carbon estimation from the water table in different land covers can be seen in Table 1.From the Table 1, it can be seen the difference in carbon emissions generated from each type of land use type.The highest carbon emissions can be seen in the palm oil land cover type with an average groundwater level of -44.5 ± 3.1 cm with estimated carbon emissions of 43.1 t CO2eq ha -1 year -1 .While the lowest emissions are seen with the rubber land cover type with a groundwater level of -19.6 ± 2.4 cm and estimated carbon emissions of 19.2 t CO2eq ha -1 year -1 .In addition, paludiculture or dry mixed plantations and sago also have lower carbon estimates than oil palm and shrubs.

Impact of canal blocking on groundwater level dynamics
Groundwater level observations were conducted for one year starting from April 2022 to March 2023.Observations were made by comparing locations that have been built canal blocking (PT) compared to canals without canal blocking (PC) in one canal flow.The canals at this location are secondary canals with canal widths ranging from 2 -3 meters while canal depths range from 1-2 meters.The results of statistical analysis k on eight observation transects in general show that significant differences with a p value <0.05 are found in monitoring wells that have a close distance to the canal, namely 5 m.The average groundwater level in monitoring wells located 5 meters away from the canal, namely the plot with canal blocking (PT) -26.67 ± 2.4 cm or equivalent to carbon emissions of 26.5 t CO2eq ha -1 year - 1 and is more shallow than the location without canal blocking (PC) which is -58.67 ± 3.1 cm or equivalent to carbon emissions of 57.8 t CO2eq ha -1 year -1 .Meanwhile, monitoring wells that are far from the canal did not show a significant difference in water level between observation plots with and without canal blocking (p > 0.05).The average water level in monitoring wells with canal blocking (PT) was -31.7 ± 3.1 cm and the average in monitoring wells without canal blocking (PC) was -31.06 ± 1.9 cm [Figure 6].canal blocks in the drainage canals.Meanwhile, monitoring wells that are far from the canal block (100 m, 400 m and 900 m) are not significantly influenced by the canal block on the drainage canals.The results of elevation measurements taken at each observation transect show a difference in elevation at each monitoring well distance.The average elevation difference of all monitoring points is 45 cm, with the lowest elevation difference is 3 cm and the highest is 262 cm in monitoring dip wells that have a distance away from the canal (Figure 7).So it can be assumed that the influence of the canal block does not have much impact on the location of monitoring wells at a distance from the canal block, and there are other environmental characteristics that affect it.

Conclusion
This study shows that land cover change on peatlands affects peatland degradation.The monitoring results showed a significant difference in the water table in each type of land use.The deepest average water table was seen in the oil palm cover type with an average of -44.5 ± 3.1 cm, and an estimated carbon emission estimate of 43.1 t CO2eq ha -1 year -1 .While the most shallow is found in the rubber and sago land cover types, with an average groundwater level of -19.6 ± 2.4 cm and -24.4 ± 4.2 cm or an estimated carbon emission estimate of 19.2 t CO2eq ha -1 year -1 and 23.9 t CO2eq ha -1 year -1 .The construction of canal blocks in drainage canals aims to rewet degraded peatlands by raising and maintaining the surrounding peat water table.The results showed a significant effect of canal blocking on the water table, especially in monitoring wells that are located close to the canals.The average groundwater level in locations with canal blocking (PT) is -26.67 ± 2.4 cm or equivalent to carbon emissions of 26.5 t CO2eq ha -1 year , -1 more shallow than the groundwater level in locations without canal blocking (PC) which is -58.67 ± 3.1 cm or equivalent to carbon emissions of 57.8 t CO2eq ha -1 year -1 .Meanwhile, monitoring wells that have a long distance from the canal do not show significant differences ( p>0.05).This is due to differences in elevation and land cover at the research location.The measurement results show that the location far from the canal has a higher elevation than the monitoring well location close to the canal block.The average elevation difference of all monitoring points is 45 cm, with the lowest elevation difference is 3 cm and the highest is 262 cm in monitoring dip wells that have a distance away from the canal .So it can be assumed that the influence of the canal block has little impact on the location of monitoring wells at a distance from the canal block, and there are other environmental characteristics that affect it.

Figure 4 .
Figure 4. Map of peat depth distribution in the research location

Figure 5 .
Figure 5. Average groundwater level at different land use types

6 .
Plot with canal blocking **PC = Plot without canal blocking Figure Average peat water table height at the observation transect (a).distance 5 m from the canal, (b).distance 100 m from the canal, (c).400 m from the canal, (d).Distance 900 m from the canal Differences in the dynamics of the water table at the study site are not only influenced by the presence of canal blocking but also by land cover and topography of the peatland.The average groundwater level in monitoring wells located close to the canal block (5 m) is significantly influenced by the presence of Apr Jun Aug Oct Dec Feb 0

Figure 7 .
Figure 7. Digital elevation model map based on field measurements of observation points of monitoring wells4.ConclusionThis study shows that land cover change on peatlands affects peatland degradation.The monitoring results showed a significant difference in the water table in each type of land use.The deepest average water table was seen in the oil palm cover type with an average of -44.5 ± 3.1 cm, and an estimated carbon emission estimate of 43.1 t CO2eq ha -1 year -1 .While the most shallow is found in the rubber and sago land cover types, with an average groundwater level of -19.6 ± 2.4 cm and -24.4 ± 4.2 cm or an estimated carbon emission estimate of 19.2 t CO2eq ha -1 year -1 and 23.9 t CO2eq ha -1 year -1 .The construction of canal blocks in drainage canals aims to rewet degraded peatlands by raising and maintaining the surrounding peat water table.The results showed a significant effect of canal blocking on the water table, especially in monitoring wells that are located close to the canals.The average groundwater level in locations with canal blocking (PT) is -26.67 ± 2.4 cm or equivalent to carbon emissions of 26.5 t CO2eq ha -1 year , -1 more shallow than the groundwater level in locations without canal blocking (PC) which is -58.67 ± 3.1 cm or equivalent to carbon emissions of 57.8 t CO2eq ha -1 year -1 .Meanwhile, monitoring wells that have a long distance from the canal do not show significant differences ( p>0.05).This is due to differences in elevation and land cover at the research location.The measurement results show that the location far from the canal has a higher elevation than the monitoring well location close to the canal block.The average elevation difference of all monitoring points is 45 cm, with the lowest elevation difference is 3 cm and the highest is 262 cm in monitoring dip wells that have a distance away from the canal .So it can be assumed that the influence of the canal block has little impact on the location of monitoring wells at a distance from the canal block, and there are other environmental characteristics that affect it.

Table 1 .
Estimated carbon emissions in different land use types