Quantication of Litter Fall and Estimation of Nutrient Release Through in-Situ Decomposition of Leaf Litter From Some Important Mangrove Species of Indian Sundarbans

Looking into the importance of mangrove leaf litter in regulation of sediment carbon sequestration and nutrient ux in Sundarbans ecosystem, an experiment was conducted at Jharkhali island of Sundarbans. In this experiment, collection of leaf litter-shedding from nine dominant mangrove species during December 2012 to November 2013 was donemonthly using ‘litter traps’ (1 m 2 ) in Eco-garden on the bank of Herobhanga creek.Seasonal litter fall was highest in Geon (Excoecariaagallocha)(103 gm -2 ) followed by Keora (Sonneratiaapetala) (98.5 gm -2 ). Kal Bain (Avicennia alba) produced the highest amount (414.37 gm -2 ) of total annual leaf litter followed by Bruguieragymnorrhiza (410.43 gm -2 ). Kankra (Bruguieragymnorrhiza), Garjan (Rhizophoramucronata) and Geon (Excoecariaagallocha) dry leaf litters contained more than 50% carbon (oven dry basis). Litter from Avicennia group contained more nitrogen and carbon. Decomposition rates of various mangrove litters were estimated through twoshort-term (30 days and 52 days) in-situ experiments using mangrove leaf litter in nylon net bags (0.6 mm mesh) subjected to periodical diurnal submergence by tidalriver water at Jharkhali. During decomposition process,observation said thatmost susceptible and resistant litter with respect to mass loss were Geon (Excoecariaagallocha) (81±5.5%) and Taura (Aegialitisrotundifolia) (26±4%) respectivelyafter 30 days. The biomass retained after decomposition losses (average45±15.2%after 30 days and 56±20.2 % after 52 days) indicated the amount of carbon retained in mangrove soil and ultimately determines the carbon sequestered in soil through mangrove litter fall. The study gives important insight into contribution of different mangrove species in carbon sequestration and nutrient dynamics in mangrove ecosystem of Indian Sundarbans.


Introduction
Mangrove forests, a productive ecosystem, support wildlife of high abundance and diverse variety (Ong 1995). The contribution of mangrove in the Global carbon cycle is quite signi cant. Researchers calculated the total global mangrove biomass estimate as approximately 8.7 gigatons dry weight, which is equivalent to 4.0 gigatonnes of carbon (Twilley et al. 1992). Mangrove ecosystems are considered highly productive because of their e ciency in trapping suspended material from the water column. The detritus developed from mangrove serves as a useful food source for many macroinvertebrates,e.g., sesarmid crabs, which can consume mangrove litter (Fratiniet al. 2000;Cannicciet al. 2008). Degradation of litter by microbes involves both aerobic and anaerobic pathways. A signi cant portion of the organic matter present in detritus is generally "exported" to adjacent coastal water through regular tidal ooding of the forest oor. The undegraded fraction of litter gets permanently buried under the sediments of the same system or any adjacent ecosystem. Mangrove productivity indirectly measured by its annual litter fall; however, this, in reality, underestimates the net primary productivity (NPP) because it excludes the biomass produced from underground roots and wood. Nevertheless, nutrients, e.g., carbon, nitrogen, and phosphorus, are recycled back to the mangroves through the litter fall when it undergoes decomposition in the sediments where it gets deposited rst.
Besides, this also gives support tomarine food web by providing ahouse for an aquatic nursery (Srisunontet al. 2017). Different stages of leaf litter break down are (i) loss of weight due to physical fragmentation (abiotic), (ii) feeding by animals and (iii) microbial activity and leaching (Steward and Davies 1989). Three signi cant factors generally control the litter decomposition rates, viz., temperature, moisture, and litter quality, and sometimes an additional fourth factor of in uence of earthworm also got importance (Bohlen et al. 1997;Dechaineet al. 2005). Soil microbes, which do the heterotrophic decomposition of litters, are in uenced by both temperature and moisture; microbial activity has a positive correlation with temperature, and sometimes a doubling of activity was noted with a 10°C increase of temperature (Kirschbaum 1995).Soil moisture (and precipitation) also play an essential role in the maintenance of higher microbial activity at enhanced temperature (Peterjohnet al. 1994; Meentemeyer 1978).
Besides temperature and moisture, chemistry and some other physical characteristics of litter also decide how susceptible that is to decomposition. Three hypotheses are available in the literature to explain the in uence of initial litter quality on litter decomposition. According to the rst hypothesis, C:N ratio can better predict mass loss and release of nitrogen from litter. The second one, the decay lter hypothesis states that in the early stages, decomposition and release rate is guided by the initial litter quality, like the ratio of lignin and nitrogen or lignin and cellulose. Lastly, the third hypothesis talked about the negative correlation of litter decomposition rates with N-based estimates of initial litter quality (Karberget al. 2008).
Productivity and also to a good extent, functions of mangrove ecosystem may be better assessed through the estimation of litter production followed by litter decomposition study. Several pieces of literature are available describing research ndings on litter fall production from mangroves, its decomposition study and also the nutrient analysis in subtropical (Twilley et al. 1986;Tam et al. 1998;Sánchez-Andréset al. 2010;Kamruzzamanet al. 2012) and tropical (Robertson 1988;Wafaretet al.1997;Silva et al. 2007; Srisunontetet al.2017)ecosystems. However, similar study is lacking in world's largest mangrove ecosystems, the Sundarbans. So,the present experiment was planned in the middle part of Indian Sundarbans to get an estimate of leaf litter fall from the predominant mangrove plants of the area and nutrient ux through in-situ litter decomposition study.

Study Area
Study area was situated in the Sundarbans, the world's most substantial 'mangrove chunk'. The Sundarbans mangrove forest covers an area of about 10,000 km 2 (3,900 sq mi), and in West Bengal, they extend over 4,260 km 2 Vidyadhari rivers. Jharkhali was in the mesohaline zone, and sometimes in summer, it also shifted to polyhaline zone. The map of the study area is shown in Fig. 1.

Estimation of Litter fall
We had surveyed the experimental site and selected nine predominant mangrove species for our studies. Those are Avicennia alba (AA), Avicennia o cinalis (AO), Avicennia marina (AM), Bruguiera gymnorrhiza (BG), Excoecariaag allochha(EA), Ceriops decandra(CD), Rhizophora mucronate (RM), Sonneratia apetala (SA), and Aegialitisrotundifolia (AR).Then three representative plants from each species were selected with similar height and diameter at breast height (DBH). For the collection of litter fall, we got prepared litter traps indigenously. First, 1 m 2 frame was prepared with wood sticks available locally; the frames were reinforced with the help of nails and ropes. Then mosquito nets of < 1 mm mesh size was xed to the frame with the help of nails (Fig. 2). We then xed litter traps below canopies of the selected mangroves by hanging those from tree branches, maintaining heights above the reach of the highest tide to prevent inundation of traps (Fig. 3). Plants were selected in such a way that all 27 traps (nine mangrove species × three replications for each species) were distributed in about 2000 m 2 areaof the monitoring site.Litter fall in the traps was collected monthly from December 2012 to November 2013. As in this experiment, we were concerned only with leaf litters; therefore, the collected total litter fall from each trap was taken to the laboratory, and leaves were separated from the mixed litter (stipules, branches, owers, and ower buds, fruits, and other parts of plants). The leaves were rst air-dried and weighed on a digital balance to get replication-wise litter fall estimate.
Thereafter a portion of litter from each replication was dried at 65-70°C, cooled to room temperature in a desiccator and weighed on a digital balance. The dried litter was then ground to ne powder, and kept for nutrient analysis.
Litter decomposition study Senescent leaves were collected from different traps installed for litterfall study;those are freshly fallen leaves from mangrove plants. The leaves were collected for eight species except for Keora(S. apetala). These leaves were airdried in the laboratory for 48 hours. The dried leaves were mixed thoroughly for individual species, and then 50 g sample picked up and put in litterbags (Stewart and Davies, 1989). Litterbags (20 cm × 25 cm) were made of high strength nylon mosquito net of 2 mm mesh size (Fig. 4). For each mangrove species, three litter bags were taken as replications for each test species making the total number of litterbags to 24. The litterbags were then placed on the sediment in the Jharkhali Mangrove Eco-garden itself in August 2013, where those were exposed to periodical tidal inundation. Bags were little pushed in the soft soil and tied with the nearby mangrove trunks or roots to prevent their owing away with the tide. After 30 days,the litterbags were collected from the site and taken to the laboratory. In the laboratory, rst the litterbags were carefully washed with tap water to remove the soil particles sticking to the bags.
Then the inside samples were gently washed to remove additional soil particles before putting those in paper bags. Models: Kelplus® -KES 12L VA for digestion, Kelplus® -Classic DX for distillation, and Kelplus®-Kelvac VA for scrubbing acid fumes). The loss of mass from the leaf litter was determined by comparing it with its initial dry weight.Lignocellulose estimation in selected leaf litter was carried out following the detergent approach, developed by Van Soest and coworkers for the analysis of ber-rich forages and feed, which is currently most frequently used one (Van Soest et al. 1991). The fractions of the ber that are insoluble either in neutral detergents or in acid detergent are measured, and the residue after treatment of the ADF fraction with 12 mol/L sulfuric acid is considered to be acid detergent lignin (ADL). By difference, hemicellulose (NDF -ADF) and cellulose (ADF -ADL) were calculated.

Water quality parameters
Water samples were collected from adjacent Herobhanga creek during peak high tide and peak low tide.The water from this creek was actually responsible for periodical wetting of the litter bags through tidal cycles during the whole decomposition process. Surface water temperature, pH, transparency, salinity, dissolved oxygen, and total alkalinity were measured onboard, and nutrients (nitrate, phosphate, silicate, sulfate) were analyzed in the laboratory following recommendations of APHA (2012). Water temperature was measured by using a degree centigrade thermometer; pH with a digital pH meter (HANNA instruments), and transparency was measured by employing a Secchi disc (Strickland and Parsons 1972). The dissolved oxygen, salinity, and total alkalinity were determined by titrimetric methods (APHA 2012).

Statistical analysis
The litterfall data were subjected to one-way analysis of variance (ANOVA) and post hoc tests using SPSS v.21. The mean values of water quality parameters (overall annual and related to two litter decomposition periods)with standard deviations. These basic descriptive statistics (mean and standard deviation) and column diagrams were performed in MS Excel 2010.

Results And Discussion
Litterfall Estimation It was observed from analyzed data in Table 1 that,the highest total litterfall from all nine mangrove  Table 1, it could be noted that during the month ofhighest leaf litter production, the maximum contribution was made by Rhizophora mucronata (173.0 g) followed by Bruguiera gymnorrhiza (148.3 g), but post hoc test showed these two mean values were not signi cantly different at 5% level. Those were only signi cantly different from Avicennia marina (45.7 g) which contributed the least. The top litter producer amongst the nine species, Avicennia o cinalis, produced higher leaf litter during consecutive three months of December 2012 (85.9 g), January 2013 (113.3 g) and February 2013 (115.5 g), which were signi cantly different (5%) from the rest of the year (Table 1). Overall Avicennia alba produced highest amount of leaf litterfall during the collection period (414.37 gm − 2 or 4.14 Mg ha − 1 ) followed by Bruguiera gymnorrhiza (410.43 gm − 2 or 4.10 Mg ha − 1 ).  1986). Mangrove litter fall rate was also considered as a function of water turnover within the forest (Pool et al. 1975). In general, higher litterfall was observed during dry summer months as this time reduced canopy volume helps in decreasing the rate of transpiration; also, in the rainy season, when the plants receive more nutrient supplements (Wafar et al. 1997). In our experimental result, we, observed bimodal pattern of total litterfall, rst peakin March and another smaller peakin November also. However, contrary to the general observation, during raining months, litter fall did not increase but somewhat decreased. In this connection, we found mention in the literature about variations in litter fall quantity was also dependent on individual species besides ambient conditions (Kathiresan 2012).
Moreover, in our experiment, we had concentrated on leaf litters only, and so production was less than where all other plant components, e.g., twigs, barks, owers, fruits, were also considered. The contribution of leaves in the total litter production of mangroves was reported as 58.4% (Wafar et al.1997). Lee

Litter fall Decomposition
The mean decomposition percentages of various mangrove leaf litters (selected for the study) from two different periods were graphically presented in Fig. 2. The 30 days experiment was conducted during August-September and the 52days experiment was conducted during January-March. Interestingly, higher decomposition percentage was recorded in 30 days experiment as compared to 52 days experiment for all the species except Kalbain (Avicennia alba). The ambient temperature and humidity during period-1(70-90% RH, 32-37°C) was more than period-2(40-60% RH, 16-25°C) and that may justify the overall higher decomposition percentages during the earlier period. It was also observed by researchers earlier that,leaf litter decomposition of Rhizophora mucronata in Kenya was higher during the rainy season as compared to dry season (Woitchik et al. 1997). In both the experiments,maximum decomposition percentage was noticed for Geon (Excoecaria agallocha); mean decompositions were 81% and 69% for 30 days and 52 daysrespectively.In those experiments, Taura leaves (Aegialitis rotundifolia) was the most resistant to decomposition; only 26% decomposed during the 30 days period, whereas in 52 days experiment it was only 12%. A positive correlation between temperature and decomposition of plant material was reported by  Table 2 to have an idea about the tidal water in which decomposition experiments were conducted. The decomposition of leaves and nutrient release also depends on the type of carbon compounds initially present in the litter. In our experiment, we considered three representative mangrove leaves based on their decomposition rate -Excoecaria agalochha (highest decomposition), Avicennia o cinalis (intermediate decomposition) and Aegialitis rotundifolia (most resistant to decomposition) and analyzed the lignocellulose contents of those three plant leaves. The decomposition percentage of leaf litters (after 30 days) was dependent on its initial lignocellulose contents and they were in an inverse relationship as observed from result (Fig. 8  Con ict of Interest: The authors declare that they have no con ict of interest.
Declaration of Interests:The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this manuscript.
Availability of data and material: Data will be provided on request to the corresponding author.