Water Balance in Oil Palm Plantation with Ridge Terrace and Nephrolepis biserrata as Cover Crop

The existence of oil palm plantations as a possible cause of drought of the surrounding areas in Indonesia is a critical issue. Therefore, information related to the effects of oil palm plantations on the surrounding environment in terms of soil water content (SWC) availability is needed. Soil and water conservation techniques in the form of ridge terracing and cover crops, such as Nephrolepis biserrata, can be expected to potentially improve soil water reserves, especially in the dry-season, by accumulating water in the rainy season. This study aimed to study the effects of N. biserrata as cover crop, together with the potential effects of ridge terraces, on the water balance in mature oil palm plantations. The research was conducted in mature oil palm plantations, Afdeling III block 375 (planted in 1996) and block 415 (planted in 2005), Rejosari Unit, PT Perkebunan Nusantara (PTPN) VII in Natar District, South Lampung Regency, Indonesia, from August 2014 to January 2015. The research was based on of setting up 15 m x 20 m experimental plots with the following treatments: (i) without ridge terraces and without N. biserrata (G0T0); (ii) without ridge terraces but with N. biserrata (G0T1); (iii) with ridge terraces but without N. biserrata (G1T0); (iv) with ridge terraces and with N. biserrata (G1T1). Hydrology parameter data were collected for each treatment plot; water balance was calculated using a water balance equation. The results showed that the use of the cover crop N. biserrata in combination with ridge terraces helped improve SWC reserves by approximately 71% and 12%, respectively. The use of N. biserrata as a cover crop reduced the rate of water loss by percolation and run-off, by approximately 36% and 80%, respectively, in an area where the annual rainfall is above 2,400 mm per year. The presence of N. biserrata shortened the period of SWC defi cit by extending the period of a water surplus by 70 days when compared with ridge terracing alone (which reduced the period of SWC by 50 days).


Introduction
Oil palm (Elaeis guineensis Jacq.) is one of plantation crops that is important as a source of foreign exchange for Indonesia.The main product of oil palm is oil, for which the production value of production needs to be further developed in the interest of the national economy of Indonesia.Data from the Direktorat Jenderal Perkebunan (2014) showed that CPO production in 2014 was 27.7 million tons from an area of approximately 10.9 million ha.
The increase in the production of oil palm has been the result of a combination of strategies, particularly the adoption of technology improvements in the fi eld.Soil and water conservation techniques have been adopted in oil palm plantations to improve the carrying capacity of those two factors (soil and water), with resulting improvements in plant growth and development, that have, in turn, improved production.According to Murtilaksono (2007), soil and water conservation technique using siltpit could delay dryness 3.5 months rather than contour ridge which could delay only 2.5 months more than control.
Studies of the potential benefi ts of cover crops in oil palm plantations have received increased emphasis to help in soil and water conservation of oil palm cropping.Cover crops have several potential functions including: a reduction in soil density (Cock, 1985); a place for storage of carbon (Reicosky and Forcella, 1998); a reduction of soil erosion by water and wind, through improvements in soil hydrology (Battany and Grismen, 2000); and an increase in the rate of water infi ltration (Archer et al., 2002).
Nephrolepis biserrata (Sw.)Schott is a weed that commonly found in many oil palm plantations especially in mature oil palm plantations (Syahputra et al., 2011), and it is capable of serving as a cover crop (Ariyanti, 2016).N. biserrata is shade tolerant and, as such, can be planted under mature oil palm trees.N. biserrata is characterized as follows: it has fl aky soft petioles and is brown in color; it has a rough leaf surface with lush foliage and a tapering leaf shape.The spores are located evenly on the edge of the leaf.The stem is round, slim, elongated and brown.Roots are fi brous roots and black (Romaidi et al., 2012).In oil palm plantations, N. biserrata very useful in maintaining a humid environment.It is also believed to be a host plant for fi re caterpillar predators, although this has not yet been supported by scientifi c research and data.
The adoption of cover crop cultivation is acknowledged to be one vegetative method for soil and water conservation.Other techniques are mechanically based and include soil tillage, ridges, terracing, reservoirs, and drainage and irrigation improvement (Arsyad, 2010).Ridge terracing is a mechanical conservation method for retaining water, so that water can be absorbed by the soil.Cover crops serve to improve the capability of soils in retaining water through the root system and plant debris, which also improve soil organic matter content after decay.Ridge terraces are based on the construction of is a soil mound along the contour or across the slope, with channels being created in the upper slopes that follow the direction of the soil mounds.Murtilaksono et al. (2007) stated that the application of mounds and silt pilt completely with vertical mulching has positive effects on the number of midrib leaf, the number of oil palm bunches, the average weight of bunches, and the production of fresh fruit bunches (FFB) of oil palm.This conservation technique is useful in improving SWC reserves to meet water needs during the dry-season, so that palm oil production may be maintained.The state of hydrology in a landscape can be described through the water balance.Broadly speaking, the water balance refl ects the relationship between the water in-fl ow and outfl ow in an area.
The water balance can also be defi ned as the difference between the amount of water received by plants and water loss from the plant and soil through evapotranspiration process (Mayong, 2006).Rainfall is the input variable in the water balance, while runoff, interception, evapotranspiration and percolation are output variables, all of which play important roles in determining the soil water content reserves of oil palm plantations.

Research Method
This research was conducted in Afdelling III block 375 and block 415 Rejosari Unit, PT Perkebunan Nusantara (PTPN) VII, District of Natar, South Lampung Regency.Soil analysis was conducted in Soil Laboratory, Bogor Agricultural University.The research was conducted from August 2014 to April 2015.
The research was conducted on 9 and 18-yearold mature oil palm with N. biserrata as cover crop.Scoring of soil water content was conducted using measuring device DFRobot SEN0114 soil moisture sensors/probes and a custom-built multimeter to read the sensors (Yogaswara, 2015).
The research was implemented by designing of experimental plots measuring 15 m x 20 m with treatments combination: without ridge terrace, without N. biserrata (G 0 T 0 ), without ridge terrace with N. biserrata (G 0 T 1 ), with ridge terrace without N. biserrata (G 1 T 0 ), with ridge terrace and N. biserrata (G 1 T 1 ).The experimental plots with cover crops treatment were fully covered by N. biserrata (100%).

PERK = percolation
Daily change of SWC (∆SWC) was calculated by the average value of SWC differences between SWC at measurement with the previous day.Oil palm interception was calculated by equation : Int(KS) = (0.1513 x rainfall) + 0.8885 (Purba, 2007).Interception of N. biserrata was calculated based on the equation : if 0 ≤ Lai ≤3 then Int (Nb) = (1.27/3)x Lai (Nb), if Lai > 3 then Int (Nb) = 1.27 mm (Zinke, 1967).N. biserrata evapotranspiration is calculated based on changes in daily SWC average at soil depth of 30 cm in the dry months (no rain at all).SWC was measured using a sensor that connected with multimeter (Yogaswara, 2015).Percolation and runoff occur when SWC value higher than total soil porosity.All hydrological variables were based on ten days data and calculated at 30 cm of soil depth.The observed physical property was bulk density required for calculation of SWC with the equation: SWC= bulk density x SWC read on multimeter

Result and Discussion
The water balance in conditions ridge terraces and N. biserrata as a cover crop in oil palm plantation PTPN VII, South Lampung, is presented in Figure 1 (October to December 2014) and Figure 2 (January to April 2015).The hydrological variables of plots without ridge terraces (G 0 ) and with ridge terraces (G 1 ), and planting with N. biserrata (T 1 ), includes precipitation, N. biserrata evapotranspiration, N. biserrata and palm oil interception, daily average of ∆ SWC and percolation.The hydrological variable of plots without ridge terraces (G 0 ) and with ridge terrace (G 1 ), but not planted with N. biserrata (T 0 ), include precipitation, oil palm interception, daily average of ∆SWC and percolation.The water balance was calculated based on the data of hydrological variables over 10 day periods.
During the period October to December 2014, the average monthly rainfall was 183.5 mm per month, while the SWC in treatments G 0 T 0 , G 0 T 1 , G 1 T 0 , and G 1 T 1 were -48.21 mm, -13.99 mm, -42.42 mm and -57.11 mm.respectively (Figure 1).This indicates that planting of N. biserrata as a cover crop was able to increase the water holding capacity, so that water soil content reserves increased by an average of 71%.This was largely achieved by the root systems of N. biserrata in creating soil pore space that can be fi lled with water, so that the water reserves increased.Ridge terraces also improved the SWC reserves but only by an average of 12.0%The combination of treatments (ridge terrace with N. biserrata) decreased runoff by 91.5% in the period when the rainfall was less than 200 mm/month.In the dry months, N. biserrata planting increased the daily average of SWC in ridge terrace plots at a depth of 30 cm, 60 cm, 90 cm by 47.9%, 27% and 38.9%, respectively (Ariyanti, 2016).
During the period of highest rainfall, January to April 2015 (225.62 mm per month), the SWC of treatments G 0 T 0 , G 0 T 1 , G 1 T 0 , and G 1 T 1 were -14.14 mm, 1.08 mm, 10.08 mm, and -17.63 mm, respectively (Figure 2).Ridge terrace was able to increase the SWC reserves higher than N. biserrata planting alone.
In this period, the planting of N. biserrata was able to shorten the period of SWC defi cit, with a longer period of water surplus period (70 days) compared to ridge terrace alone (50 days).In addition, N. biserrata reduced the percolation loss to 36.15% and run-off by 80.4%.Land with ridge terraces and planted with N. biserrata, land without ridge terraces but planted with N. biserrata, and land ridge terraces but without N. biserrata, reduced surface water run-off by 95.7%, 80.0% and 63.4%, respectively (Ariyanti, 2016).
During the period October 2014 to April 2015 there was one dry month, this being October 2014 (47.6 mm), while the wettest month was December 2014 (278.9 mm).The water balance in the dry month for treatments G 0 T 0 , G 0 T 1 , G 1 T 0 , G 1 T 1 was -9.08 mm, 103.3 mm, 17.9 mm and 31.1 mm, respectively, for SWC, and 0 mm, 0 mm, 41.52 mm and 34.98 mm, respectively, for percolation (Figure 1).This shows that in the dry months, N. biserrrata was able to maintain the SWC.N. biserrata affected the water balance in mature oil palm by reducing the SWC defi cit during the dry months, with an average of SWC defi cit amounting to 36.71% (Ariyanti et al., 2015a).There was an increase in SWC on average in plots planted with N. biserrata compared to without N. biserrata (Ariyanti et al., 2015b).SWC is affected by land cover, in this case by N. biserrata.In order to maintain SWC during dry-season, it is suggested to improve the condition of vegetation growing above the soil, so that shading reach above 80% and 100% litter coverage of the soil (Suhardi et al., 2012).Land cover with Asystasia gangetica is able to increase SWC in the range 33% -66% (Junaedi, 2014).
Water balance in the wet month December 2014 for treatments G 0 T 0 , G 0 T 1 , G 1 T 0 , G 1

Conclusion
Planting cover crop N. biserrata and ridge terraces were able to improve SWC reserves by an average of 70.98% and 12.01%, respectively.N. biserrata reduced the rate of percolation and run-off by 36.15% and 80.42%, respectively, during periods when the level of precipitation above 2400 mm/year.N. biserrata shortened the period of SWC defi cit and gave longer water surplus (70 days) compared to the use of ridge terraces alone (50 days).However, the optimum treatment for sustainable production is the combination of the use of ridge terraces and N. biserrata as a cover crop.

Introduction
Soil erosion on agricultural land results in the loss of soil organic matter (Chen et al., 2011).Erosion caused reduction in fertile surface soil layer which are rich in organic matter and nutrients (Blanco and Lal, 2008), and reduce land and plant productivity.According to Abdurachman et al. (2003) loss of 10 cm top soil on oil palm plantation can decrease production by more than 50% despite complete fertilization was applied, because top soil as the source of nutrients will be eroded.Erosion does not only affect soil organic matter content but also the soil major nutrients N, P, and K.The amount of nutrients lost to erosion was usually greater than the predicted values, because fi ne and fertile soil particles will be leached, resulting in an accelerated decrease in soil fertility (Arsyad, 2010).
Oil palm plantations in PTPN VII Rejosari, Regency of Natar, South Lampung is generally dominated by S-2 (moderately suitable) and S-3 (marginally suitable) Ultisol, which indicates that the production potential of oil palm is relatively low.One of the constraining factors in oil palm production is the sloping contour of the land (3%-8%), shallow solum (± 1 m), and limited rainfall throughout the year in the area of oil palm production.(Bunch, 2012).
Recently mature oil palm plantations no longer use legumes as cover crops due to their intolerance to shade, and because planted legumes were naturally overgrown by different types of shade-tolerant weeds, including A. gangetica.
A. gangetica is known as weeds to be controlled in oil palm plantations, because it produce seeds in large quantities (Adetula, 2004).However, A. gangetica, mainly subspecies gangetica, can be used as cover crop as it has no tendrils or spines (Ismail and Shukor, 1998), easy to grow and grow quickly (Yenni et al., 2015a), adaptable to different environmental conditions (Sandoval and Rodriguez, 2012), shadetolerant (Yenni et al., 2015a), they can even grow well under 90% shaded (Adetula, 2004).A. gangetica was reported to increase water availability in ultisol (Junedi, 2014), and increase the availability of N, P, K through the creating nutrient balance (Yenni et al., 2015b).
A. gangetica is able to grow well in full of light and low soil fertility (Samedani et al., 2013;Kiew and Vollesen, 1997), to restrain or reduce erosion due to raindrops and surface run off (Adetula, 2004), has high nutritional value (antioxidants) for animal feed and drugs (Adetula, 2004;Gopal et al., 2013;Mugabo and Raji, 2013), and can serve as bio-monitor for the presence of heavy metals such as mercury (Hg) (Chew et al., 2012).In addition A. gangetica contributed N, P, and K to the soil (Yenni et al., 2015b), rapidly decomposed (Yenni et al., 2014), and can serve as soil carbon stock (Yenni et al., 2015b).
Results of Fuady and Satriawan (2011) showed that planting cover crops such as corn and peanuts as well as ridge terracing was able to control run off and erosion to 63.5% and 90.3%, respectively, compared to the absence of cover crops and ridge terrace.Planting cover crops on palm oil plantations were effectively reduced run off and soil erosion, and prevent loss of nutrients (Fuady et al., 2014;Satriawan et al., 2011;Satriawan et al., 2012).
This study aims at examining the roles of A. gangetica as cover crop to minimize soil erosion and loss of nutrients in mature oil palm plantation in South Lampung, Indonesia.

Materials and Methods
This research was conducted in the fi eld using split block design in a randomized block design with two factors and six replications.The main plots were: ridge terrace consists of with and without ridge terrace.
The subplots were cover crop, consists of with and without cover crops A. gangetica.
Before the erosion plots were constructed ridge terraces were arranged in the same directions to contour on each vertical interval 80 cm.Height, width and depth of mounds channel was 30 cm (Figure 1).Erosion plots were made on each block experiment with an area ± 300 m 2 using ebonite tarpaulin material.Erosion materials from the erosion plots were collected using Tub A measuring 5 m x 1 m x 1 m, and on outwards facing side created 7 sinkholes, 6 cm in diameter.The center hole was connected by pipes Ø 6 cm into Tub B (Figure 1).Gauze was placed above the Tub A which serves to accumulate soil erosion by run off.A. gangetica for cover crop was planted after erosion plot has been constructed, with a spacing of 10 cm x 10 cm.
Erosion measurements were performed following every rainfall event during the study.The erosion measurement includes soil fi ltered on gauze and sediments that were dissolved in the tub A and B. Soil particles that were collected in gauze was weighed by draining in oven at 105 ºC until reaching a constant weight.The weight of soil sediment samples were weighed by fi ltering the water using fi lter paper.
Sediments that were left on the fi lter paper were dried in the oven at 105ºC until reaching a constant weight.The amount of soil and sediment (E) was calculated using the following equation : Where E : soil erosion (t.ha -1 ) C apA and C apB : concentration of sediment load in Tub A and B (kg.m -3) V A and V B : run off volume (m 3) A : area (ha) 10 -3 : conversion unit from kg to ton.
Soils and sediments analysis were performed to measure the concentration of organic C using Walkley & Black Method, Total N using Kjeldhal Method, P 2 O 5 available using Bray Method with spectro-photometer, and K 2 O available using Bray Method with fl ame photometer.The analysis results of organic C, Total N, P 2 O 5 and available K 2 O through erosion (soil and sediment) were calculated by the following equation: Where: X = Amount of organic C, N, P and K lost through erosion (kg.ha -1 ) Y = Concentration of Organic C, total N, P and K available in sediment E = Amount of total soil erosion (t.ha -1 ).
Data obtained from the amount of erosion and nutrients loss through erosion were analyzed using ANOVA; further testing used Least Signifi cant Difference (LSD) at 5% signifi cant level.Data were analysed using the Statistical Analysis System (SAS) Software 9.1.(SAS, 2004).

Soil Erosion
Effect of ridge terrace and cover crops A. gangetica on erosion in oil palm plantation PTPN VII Rejosari, South Lampung is presented in Figure 2. Erosion is the loss of soil surface top layer along with run off caused by rain (Arsyad, 2010).Run off as the cause of soil erosion occurred due to heavy rainfalls, demonstrated in Figure 2. Erosion did not occur in August and September 2014 where there was no rain whereas erosion occurred in October 2014 when rainfall was 21.8-251.3mm.In December 2014 with rainfall of around 220.9 mm, erosion in plots with ridge terrace with A. gangetica as cover crop (G 1 T 1 ) was smaller than without ridge terrace and cover crop (G 0 T 0 ), i.e. 0.03 t.ha -1 and 3.3 t.ha -1 , respectively.This is because in the G 0 T 0 treatment rain droplets directly falling onto the unprotected soil surface, accelerating run off and caused soil erosion.Sinukaban (1989) stated that erosion will increase drastically with increased rainfall when the soil surface is not covered by vegetation, or contoured with ridge terrace due to limited opportunity for water infi ltration.
Treatment with ridge terrace and planting cover crop A. gangetica was able to reduce soil erosion despite the high rainfall; the plant canopy of the cover crops protected the soil surface from the kinetic energy of rain droplets.In addition, more rain water was intercepted by plants, and the ridge terrace improved water infi ltration to the soil through trenches and holes in the ridge terrace.Other studies show that oil palm planting + upland rice followed with soybean + Mucuna bracteata strips were able to minimize erosion in the fi ve to seven-year-old oil palm plantation (Fuady et al., 2014).Similarly, ridge terrace and cover crops signifi cantly supress erosion in coffee plantation compared to coffee without cover crops (Dariah et al, 2004).
Table 1 shows that interaction between ridge terrace and cover crops had signifi cant effects on erosion.Ridge terrace with cover crop A. gangetica (G 1 T 1 ) had the lowest soil erosion of 3.3 t.ha -1 per year, whereas those without ridge terrace and cover crops (G 0 T 0 ) had the highest soil erosion of 56.4 t.ha -1 per year.
Growing A. gangetica in the oil palm plantation improved the effectiveness of ridge terrace on reducing erosion from 47.1 % (G 1 T 0 ) to 94.1 % (G 1 T 1 ) (Table 1).Idjudin (2011) reported that the effectiveness of ridge terrace in reducing erosion will increase if this practice is combined with planting cover crops.Satriawan et al. (2015) also showed that combination of ridge terrace and cover crop reduced erosion 1.8 times more effective than without ridge terrace and cover crops, whereas Nursa'ban (2009) reported 100% soil protection from erosion by ridge terrace and cover crop.

Loss of Organic C, N, P, and K
Table 2 shows that soil erosion caused the loss of organic C, N, P and K in the soil, and that ridge terrace and cover crops has signifi cantly reduced the losses of organic C and total N, P 2 O 5 and K 2 O. Loss of organic C, total-N, P 2 O 5 and K 2 O by erosion were lower on ridge terrace (G 1 ) treatment compared to without ridge terrace (G 0 ).This is because ridge terrace will delay run off and provide channels for the run off, thereby increasing the rate of water infi ltration into the soil, reduce loss of organic C and soil nutrients through erosion.
Loss of organic C, total-N, P 2 O 5 and K 2 O through erosion in the plots with cover crops A. gangetica (T 1 ) were also less than without the cover crop (T 0 ), likely because the presence of canopy and root system from the cover crop were able to improve soil carrying capacity and facilitated water infi ltration into soil, in turn reduce loss of organic C and soil nutrients.
Combination of ridge terraces and cover crops also signifi cantly reduced the loss of organic C, total-N, P 2 O 5 and K 2 O through erosion.Cultivation of A. LSD.G 0 = without ridge terrace, G 1 = ridge terrace, T 0 = without cover crop; T 1 = with cover crop A. gangetica.The effectiveness to minimize erosion is calculated by comparing the erosion of G 0 T 0 treatment (control) with other treatment.Note : G0 = without ridge terrace, G1 = ridge terrace, T0 = without cover crop A. gangetica; T1 = with cover crop A. gangetica Values in the column and row followed by the same letter are not signifi cantly different at 5% LSD 1) The mean values in the same column and row followed by different letters show signifi cant differences at 5% LSD.
Loss of soil organic C also means losing of soil organic matter, because Organic C is the main constituent of soil organic matter.The loss of Organic C by erosion is a serious problem as it will accelerate soil degradation and declining soil fertility.Content of soil organic matter is one indicator of land resources sustainability (Wolf and Snyder, 2003).Organic materials serve to recycle nutrients back into the soil, and improve water holding capacity.The organic matter content of agricultural top soil is usually in the range of 1 to 6%.A study at Michigan demonstrated potential crop yield increases of about 12% for every 1% organic matter.In a Maryland experiment, researchers saw an increase of approximately 2 tons of maize per acre when organic matter increased from 0.8% to 2% (Magdoff, 2012).Soil organic matter affects soil biological, chemical and physical properties, and makes it of critical importance to healthy soils (Magdoff, 2012;Bunch 2012).
The loss of total N through erosion was higher than the loss of P and K.This might be because one of N sources is soil organic matter (Hardjowigeno, 2010), so the increased loss of organic C through erosion resulted in the higher leach of N. Loss of K by erosion is usually higher than P because by K is more susceptible to leaching compared to P (Havlin et al., 2005).
Information from this study shows that oil palm plantations may have experienced accelerated land degradation due to erosion, which resulted in decreased soil organic matter content and soil nutrients (Arsyad, 2010).However, with ridge terrace and cover crops A. gangetica erosion and loss of organic matter and nutrients can be controlled, as the loss of nutrients was directly related to the amount of erosion, and it is a function of organic C and nutrients concentration in the sediment (Sinukaban, 2007;Arsyad, 2010).
Similar results were reported in teak (Didjajani, 2012) and coffee plantation (Dariah et al., 2004) that higher soil erosion resulted in a higher loss of organic C, N, P, and K. Similarly, Henny et al. (2011) show that planting potato on ridges across the slope land decreased loss of organic C, N, P, and K due to reduced erosion.Other studies show that ridge terrace and intercropped areca nut with maize, and ridge terrace and intercropped cocoa with peanut reduced the loss of organic C, N, P, and K due to lower incidence of erosion (Satriawan, 2015).

Conclusion
A. gangetica as cover crops in mature oil palm plantations can effectively minimize erosion and reduce the loss of organic C, N, P, and K by 95.7%, 93.4%, 96.0% and 90.0%, respectively.Combination of ridge terrace with cover crop A. gangetica in mature oil palm plantations was more effective to reduce erosion and loss of organic C, N, P, and K, i.e. by 94.1%, 99.1%, 99.2%, 90.0% and 98.5%, respectively.
Figure 1.Daily cumulative water balance in plots (G1) and without ridge terrace (G0), with ridge terrace (T1) and without cover crop (T0) from October to December 2014 in Oil Palm Plantation PTPN VII Rejosari, South Lampung.

Figure 1 .Figure 2 .
Figure 1.The ridge terrace and sediment collector system at the experimental plots.

Anderson and Ridge Terrace in Reducing Soil Erosion and Nutrient Losses in Oil Palm Plantation in South Lampung, Indonesia
The Roles of Asystasia gangetica (L.) T. Anderson and Ridge Terrace in ..........The Roles of Asystasia gangetica (L.) T. AbstractAsystasia gangetica (L.) T. Anderson is a weed commonly found on oil palm plantations and can be used as cover crop for mature oil palm plantations due to its tolerance to shading.The use of cover crop is a soil conservation technique to support sustainable availability of soil nutrients by reducing erosion and nutrients loss, particularly during the rainy seasons.

Table 1 .
Effectiveness of ridge terrace and cover crop A. gangetica in reducing erosion in palm oil plantation PTPN VII Rejosari, South Lampung, from August 2014 to April 2015

Table 2 .
Effect of cover crops and ridge terraces to the loss of total organic C, total-N, P 2 O 5 and K 2 O throug erosion in oil palm plantation PTPN VII Rejosari, South Lampung from August 2014 to April 2015.