Decomposition rate and litterfall dynamics of Tembawang agroforestry area , West Kalimantan , Indonesia

Understanding the carbon cycle could be the basis and best way to management practices in traditional agroforestry. In this study, we will estimate the litterfall dynamics and decomposition rate in one of the traditional agroforestry ‘Tembawang’ practiced by local people in Kalimantan. Litterfall dan decomposition monitoring was conducted every two weeks for three months with litterfall traps and litterbags. From the results, we found litterfall production of Tembawang Agroforestry was 91.23-95.03 gr m month or 10.95-11.40 tons ha year. The estimated litter decomposition rate in the first of 14 days was 3.92-4.43 gram day after that decomposition rate decreased (on 84 day) at 0.75-0.81 gram day. A positive Spearman correlation occurred with the rainfall and temperature variables on the dynamics of litterfall but the value was not significant. Tembawang agroforestry litter production is higher than another agroforestry and the value is approaching the primary tropical rainforest. Litter decomposed more than 50 percent in the initial two weeks and the value is correlated to humidity.


Introduction
The role of tropical forests in the carbon cycle in the atmosphere is determined by the amount of carbon contained in biomass and carbon sequestration rate per year. The carbon of tropical forest biomass is generally released through respiration into the atmosphere (Murdiyarso and Satjapradja, 1997). Tropical forest ecosystems play an important role in global carbon sequestration. Carbon intake in different vegetation occurs through photosynthesis and enhances by environmental factors and good forest management Original Article practices. Deforestation is an activity that can change forest land use. This activity will directly impact the heat transfer from and to the surface of the forest and thus have an impact on energy balances, water balances, and nutrient balances of the forest (Murdiyarso and Satjapradja, 1997). Litter decomposition is a very important process in nutrient dynamics in forest ecosystems (Regina and Tarazona, 2001). The litter decomposition process involves many factors including environmental factors such as pH (Van Breemen, 1995); climate (temperature, humidity) (Guo and Sims, 1999); chemical composition of litter (Aerts and Caluwe, 1997), and soil microorganisms (Saetre, 1998). In general, litter decomposition rate is generally slower under low pH soil than neutral pH (Murayama and Zahari, 1992). Moreover, Murayama and Zahari (1992) stated that a higher C/N ratio in the litter was more difficult to decompose than a lower C/N litter ratio. In general, litter from a plant that grows in nutrient-poor environments is more difficult to decompose and will cause a slower process of nutrient cycling in the environment than litter are from plants in nutrient-rich environments (Van Breemen, 1995;Aerts and Caluwe, 1997). Litter in highly microorganisms concentration areas tend to decompose more quickly than fewer microorganisms concentration areas (Saetre, 1998). The litter decomposition rate also takes place more rapidly under aerobic conditions than anaerobic conditions (Johnson and Damman, 1991). The Tembawang Agroforestry area in West Kalimantan is a tropical forest which is managed by Dayak's tribe traditional communities. Tembawang Agroforestry is a forest plantation which is far located from the settlement. Tembawang Agroforestry can be classified into a forest garden because they look like a forest and contains various types of fairly large diameter trees (Arifin et al., 2003). Rafdinal and Pitopang (2019) reported that Tembawang Agroforestry has 43 species of plants consisted of 32 genera and 13 families. This agroforestry area is a potential non-woody timber forest that was preserved and maintained by the local community. The results of the study of Rafdinal and Pitopang (2019) also showed that there was a correlation between the density and basal area of the stand with the condition of soil fertility. The productivity and sustainability in the nutrient cycle of Tembawang Agroforestry are interesting to study. A prospective analysis in this study is very important to determine the key factors in developing the optimal economic, social, and ecological functions of the Tembawang Agroforestry area in the future. The rehabilitation program of tembawang was carried out in several villages in several districts of West Kalimantan also has not been able yet to overcome the problem of adding forest value. The increasing land productivity through the study of the dynamics of decomposition and litterfall production related to carbon stocks and nutrient cycles is an appropriate alternative to increase Tembawang Agroforestry productivity, especially in assessing of global warming issue and optimizing the function of forests as environmental services, as well as increasing community income, especially in Sekadau Regency, West Kalimantan. This research aimed to estimate litterfall production, to analyze litter decomposition rate and to analyze litterfall dynamics of the Tembawang agroforestry area in Sekadau Regency, West Kalimantan, Indonesia.

Site description
This research was conducted at the Tembawang Agroforestry area, Nanga Pemubuh Village, Sekadau Hulu District, Sekadau Regency, West Kalimantan. The Tembawang Agroforestry area has two types, namely fruit Tembawang and rubber Tembawang. Sekadau Hulu Subdistrict is one of seven subdistricts in Sekadau Regency which has ±869.7 km 2 with generally highland and fewer lowland areas. Geographically, Sekadau Hulu Subdistrict is located at 0 o 8'25'-0 o 21'30" S and 110 o 8'45"-111 o 15'20" E. The average rainfall in the Sekadau Hulu was between 1,684 to 12,000 mm. The topography of Sekadau District was plain and hill forms with altitude 0-1,000 m above sea level (masl) (Fig 1).

Litterfall dynamics and nutrient cycle
There are four blocks, each block consists of four sampling 50 m x 50 m plots in relatively flatted Tembawang Agroforestry area. Each block was selected as a sample plot for research of litter and nutrient content. Litterfall traps were placed randomly in four samplings 50 m x 50 m plots (each plot has 10 traps). The 1m x 1m traps were made from a plastic net (0.1 cm hole) with a frame placed 1 m above the ground. The collected litter in each litter trap was put into a wheat bag and air-dried for at least 2 days. After drying, the litters of each bag were sorted into 4 parts of plants, following Proctor et al. (1983), namely (1) leaf (2) branch/twig (diameter <2 cm), (3) flower + fruit, and (4) mixed litters. The sorted litters were then weighted and combined into a litter combination for each part per month from each plot. The data analyzed was average litter gram m -2 month -1 . Then, each part was oven-dried at 105°C and re-weighed to get ovendried weight. Chemical analysis was carried out for all parts (leaf, branch, flower + fruit, and mixed litters). Before being analyzed, the samples were crushed using an electric grinding machine. At least, 0.2 grams of oven-dried weight of each sample was crushed and dissolved into 4.4 ml of a solution made from 350 ml of sulfuric acid, 420 ml of hydrogen peroxide, 0.42 grams of selenium as a catalyst and 14 gram of lithium sulfate (Allen, 1989).

The rate of nutrient decomposition and release in the Tembawang agroforestry
The rate of nutrient decomposition and release of the Tembawang agroforestry area was measured by putting 100 g fresh litter into a net bag (with a size of 25 x 22 cm) placed on the forest floor, this intended to make the process of decomposition as natural as possible (Moore et al., 1984). The monitoring of litterbags was carried out at certain intervals to estimate the rate of litter mineralization (Haraguchi et al., 2002). 30 litterbags were placed at each research location where the litters were taken. Five replicate samples were taken every two weeks. After weighing, the dried leaf litters were oven-dried for 48 hours at 70 o C to determine the ratio between air-dry weight and oven-dried weight for each sub-type of Tembawang Agroforestry, namely fruit Tembawang and rubber Tembawang. The nutrient release was calculated based on the formulation of Guo and Sims (1999); Guo and Sims (2001

Data Analysis
A comparison of litter production and decomposition rate for each pick-up period was analyzed using oneway ANOVA. The differences were evaluated at P <0.05. The relationship between the dynamics of litterfall and climate was analyzed using Spearman's regression and correlation. All statistical analyzes used the SPSS Graduate Pack TM 14.0 program for Windows. A difference was evaluated at P <0.05.

Results and Discussion
The litterfall dynamics There were several tree species i.e. Durio zibethinus Arthocarpus elasticus (Terap/Kepuak) at rubber-Tembawang. These plants contributed to the production of forest litter and natural soil nutrients. Durian tree (Durio zibethinus) was the dominant species found at the fruit Tembawang Agroforestry site and had the largest contribution to the production of litters at the location compared to other species. The mean production of litterfall at the research location is presented in Table 1. Estimation results show the total litter production at the Tembawang fruit agroforestry location was 3.04±1.07 gram m -2 day -1 , and at the rubber-Tembawang location was 3.17±3.01 gram m -2 day -1 . If estimated in a year, the total litter production at study sites was estimated at 10.95-11.40 tons ha -1 year -1 . Analysis of litter production based on parts of plants was divided into vegetative parts (leaves and branches), generative parts (flowers and fruits), and others (Fig 2). The leaf portion was the highest component in both locations (60-80 percent of all litter components). This was similar to Alhamd and Hagihara (2004), Salim and Budiadi (2014) where leaves were a dominant component of the whole litterfall components (more than 50 percent). This is related to the leaf is a dominant plant organ and is easily released organ due to direct contact with climatic factors such as rain and wind.

Figure-2. Percentage of litter production at fruit Tembawang agroforestry (left) and Rubber Tembawang agroforestry (right).
Litterfall production at both locations was related to climatic factors, especially rainfall, but it was not significant (Fig 5 and Table 3). In dry months such as June and August, litterfall production at both locations increases. In line with research from Tongkaemkaew et al. (2018) wherein the dry season litter production increases and vice versa. Litter production of Tembawang agroforestry was from 10.95 to 11.40 tons ha -1 year -1 . This litter production was smaller than litter production in the tropical rainforest ecosystem reported by Songwe et al. (1988) and Triadiati et al. (2011). However, litter production at the study site was higher than the other ecosystems ( Table 2). The litter production was mainly influenced by the fertility of the ecosystem where the higher the fertility of an ecosystem, the greater the production of litter (Triadiati et al., 2011). Ecosystem fertility is related to the high intensity of the generative phase of the types of ecosystems. When entering the generative phase (flowering), plants tend to fall more of their leaves due to the process of concentration of energy and growth components to support the formation of flowers.

Litter decomposition of Tembawang agroforestry
The estimated results showed that process of litter decomposition of fruit Tembawang agroforestry was significant decomposition (more than 50 percent) on the first observation (fourteenth day) with an average decomposition rate was 4.13 grams day -1 (Table 3). In the next observation, the decomposition rate tended to constantly decrease so that at the last observation (eighty-fourth day) the decomposed litter was approximately 68 percent. The average litter decomposition rates of fruit Tembawang agroforestry are presented in Table 3. The average litter decomposition rates of rubber Tembawang Agroforestry location are presented in Table 4. Based on the estimated results, the highest decomposition rates occurred in the first two weeks, with average decomposition litter was 3.92 grams day -1 . The decomposition rate tended to decrease constantly every week so that at the end of the observation (eighty-fourth day), the dry litter weight was 36.76 grams or an average decomposition percentage was 63.24 percent of the initial weight. The rate of litter decomposition is related to climatic factors at the location such as rainfall. In conditions of increased rainfall, the rate of decomposition increases and vice versa, but is not significant (Fig 5 and Table  3). In line with the research of Singhal et al. (2019) where the rate of litter decomposition increases during the rainy season. This is related to the physical properties and microbial activity in humid conditions, so the microbial activity in decomposing will increase. The decomposition process that occurred in both locations tended to have a similar trend where the significant decomposition process occurs in the first two weeks with the percentage of litter decomposition was more than 50 percent. Furthermore, the rate of decomposition at both locations was constantly declined and at the end of the observation, more than 60 percent of the litter had been decomposed. These results were similar to the decomposition process in natural forests and cocoa agroforestry in Central Sulawesi (Triadiati et al., 2011). At the last observation, decomposed litter was about 60 percent. In line with the research of Triadiati et al. (2011) and Songwe et al. (1988), the decomposition process was very quickly at the beginning of the observation and was significantly slowdown in subsequent observations. This was due to the initial content of litterfall still contained a lot of water or dissolved compounds in water that were easily decomposed.

Figure-3. Climate factor (Rainfall, Humidity and Temperature) at Sekadau Regency in 2016-2017
The results of Spearman correlation analysis between litter production and decomposition with climate variables were presented in Table 5. All test variables have no significant correlation (p<0.05) except for the correlation between humidity factor and litter decomposition was significant. The rainfall and temperature factors have a positive correlation to litter decomposition of fruit Tembawang Agroforestry with a correlation value of more than 0.7 but not significant.

Conclusion
Litter production of Tembawang Agroforestry in the Sekadau Regency was sufficient to contribute to the nutrient cycle. Although litter production was smaller than litter production in tropical rainforest ecosystems, the contribution of litter production can be an alternative to increase the natural fertility of the forest floor. Ecosystem fertility is related to the high intensity of the generative phase of types of ecosystems. When entering the generative phase (flowering), plants tend to fall more leaves due to the process of concentration of energy and growth components to support their flower formation. The decomposition process of two types of Tembawang Agroforestry tended to have a similar trend where a significant decomposition process occurred in the first two weeks with the percentage of litter decomposed more than 50 percent. Furthermore, the rate of decomposition at both sites was constantly declined and, at the end of the observation, more than 60 percent of litters were decomposed.