Introduction

Mangroves play a major socio-economic role to human communities in developing countries. They not only provide protection from tidal erosion, storm surges and trap sediment for land accretion2 but also play an important role in biogeochemical transformations in coastal ecosystems3. These transformations are mainly microbially-mediated which catalyze various steps of the oxidative and reductive phases of elemental cycles. Reducing conditions in mangrove sediments are known to favour alternate respiratory pathways like denitrification, sulfate reduction, etc. Recently, it has been shown that denitrification and anammox operate in tandem resulting in N loss in tropical mangrove sediments1. However, denitrification is a more important process which effectively reduces N load from the system. Though, denitrification is a major mechanism for NO3− removal in coastal sediments4, it is also possible that its removal could proceed through other pathways. Nutrient regeneration could be important in N limited ecosystems like mangroves5 wherein the microbial community could be competing with the vegetation for inorganic N requirements. Internal regeneration could act as an efficient mechanism to meet the N demand from both the microbial and plant communities. As a sequel to earlier findings1, we examined the down-core variation (at every 2 cm interval within 0–10 cm depth range) in nitrate reducing activity (NRA), dissimilatory nitrate reduction to ammonium (DNRA) and net nitrous oxide (N2O) production in two tropical mangrove systems of Goa, India. The Divar mangrove ecosystem which is influenced by NH4NO3 input from ferromanganese mines located upstream6 was compared to Tuvem which is relatively pristine7. We hypothesize that mangroves are closed systems which efficiently conserve N through pathways like DNRA. Our observations reveal that ammonium is re-circulated within the mangrove systems through DNRA. This mechanism helps to overcome N limitation that could possibly arise due to demand from the biotic components.

Results

The Tuvem and Divar mangrove ecosystems are characterized by measurable pore water NO3- concentrations. Down-core variation in the concentraton of the nutrient showed a sub-surface maxima at Tuvem (36.62 (±2.91) μmol NO3--N L−1 at 2–4 cm) (Table 1). At Divar, the concentration of the nutrient was found to decrease with depth with a maximum of 19.90 (±1.66) μmol NO3--N L−1 at the surface. Examination of NRA at both the locations revealed that the activity at Divar was comparatively higher within 0–4 cm occurring at a rate of 3.52(±0.38) μmol g−1 h−1 (≈1.07 μmol cm−3 h−1) which is almost twice the rate recorded at Tuvem (Figure 1).

Table 1 Percentage of ammonium retention and N loss in mangrove sediments
Full size table
Figure 1
figure 1

Down-core variation in nitrate reducing activity (NRA) at the relatively pristine site Tuvem and anthropogenically-inflenced Divar.

Error bars represent SDs.

Full size image

Labelling with15N to measure DNRA showed a steady increase in 15NH4+ over time at all depths investigated at both the locations. Down-core observations showed NO3- removal through DNRA occured ~2x faster at the relatively pristine Tuvem than at the anthropogenically-influenced Divar ecosystem (Figure 2). Maximum NH4+-N retention in the Tuvem sediments was recorded at 2–4 cm occurring at a rate of 1.19 μmol g−1 h−1 i.e. ~13.44 mmol m−2 h−1. Co-occurring processes like anammox and denitrification were also measured1 in conjunction with NRA and DNRA measurements. However, these results have been published separately (Table 1).

Figure 2
figure 2

Down-core variation in rate of dissimilatory reduction of nitrate to ammonium (DNRA) at Tuvem and Divar.

Error bars represent SDs.

Full size image

In the mangrove sediments of Goa, production of the greenhouse gas N2O has been attributed to the denitrification pathway8. Apart from measuring N2 production arising from anammox (Anx) and denitrification activity (DNT)1, we measured net N2O production to account for N loss in these systems. Nitrous oxide production at both the locations was found to vary with depth. At Tuvem, maximum production of N2O occurred between 2–6 cm (Figure 3). At Divar, a steady decrease in N2O production with depth was observed. Here, N2O production occurred at a rate of 2.71 nmol g−1 h−1 which was almost ~2x lower than the rate recorded at Tuvem.

Figure 3
figure 3

Down-core profile of net nitrous oxide production at Tuvem and Divar.

Error bars represent SDs.

Full size image

For a more holistic view of the N cycle processes at Tuvem and Divar, the range of activities measured have been illustrated in Figure 4. If the maximum rate of occurrence is considered, it can be observed that NH4+-N retention through DNRA is almost 15x higher than combined N2O and N2 loss through DNT in the Tuvem sediments (Figure 4). However, at Divar, NH4+-N retention through DNRA proceeds only 3x higher than denitrification.

Figure 4
figure 4

Nitrogen cycling in mangrove sediments where microbially-mediated activity has been expressed as nmol g−1 h−1.

Tuvem = (T); Divar = (D); nd = not detected; *Fernandes et al.1 **Krishnan & Loka Bharathi,12. At Tuvem, rates of N retention through DNRA are 15x higher than N loss through denitrification while at Divar they were only 3x higher. This observation indicates that NO3- reduction to NH4+ results in N conservation especially in mangrove ecosystems that are not prone to anthropogenically derived N inputs.

Full size image

To understand N retention versus loss at Tuvem and Divar, the percentage of NH4+-N retention through DNRA and loss as N2O/N2 through DNT and Anx was calculated based on the percentage of NO3- reduced. Our observations reveal up to 99% N retention in the sub-surface layers at both the locations (Table 1). Of the total NO3- reduced, the percentage of N loss was maximum at the deeper layer in Divar owing to elevated N2 production through Anx.

Discussion

Sediment characterization carried out in present investigation revealed that though Tuvem is relatively free from extraneous nutrient input, it is characterized by NO3- accumulation at depth ≥2 cm which could be attributed to it's intrinsic production. As a result, elevated NRA was observed at ≥2 cm at this location. At Divar, high rates of NRA were recorded within the first 4 cm. The top layers of the sediment at Divar are constantly replenished with NO3- either originating from internal generation through nitrification or from external anthropogenic sources viz.; sewage outfall, mining rejects, land-runoff during monsoon, etc. Elevated NRA at ≤4 cm indicates efficient transformation of NO3- through the reductive phase of the N cycle which helps to maintain low levels of the nutrient in this system. Values recorded for NRA at Tuvem and Divar are also in close range to those reported in coastal sediments9 where rates varying from 0.662–2.4 μmol cm−3 h−1 have been reported.

Dissimilatory nitrate reduction to ammonium is an important mechanism that supplies available N in the system10. Ammonium is known to be adsorbed easily onto clay particles11. In organically rich mangrove sediments12, NH4+ released through degradation of organic compounds could easily get bound to clay particles making it unavailable for biological uptake. In some areas, DNRA can remove more NO3- than denitrification13. Our observations reveal that DNRA values recorded in the present investigation were 2–3 orders higher as compared to rates reported in other marine sediment14. In the mangroves ecosystems of Goa, the process accounts for up to 99% of the NO3- reduced. In marsh sediments, DNRA has been estimated to account for up to 23% of the NO3- reduced15 which is far lower than observed in the current study. The DNRA process is probably responsible not only for NO3- removal, but also for a non-neglectable part of NH4+ production.

The partitioning of NO3- between denitrification and DNRA is affected by it's concentration and the quantity of carbon16. In carbon rich and N limited systems like Tuvem, the larger contribution of DNRA indicates that this ecosystem efficiently re-circulates available N and conserves it to overcome limitation. A similar scenario could be expected at Divar. However, this ecosystem receives extraneously-derived nutrients. As a result, the contribution of DNRA is relatively less as compared to ecosystems that need to conserve N. In anoxic estuarine sediments, degradation of organic matter results in sulfide enrichment17,18. Chemolithoautotrophic DNRA is also known to couple the reduction of NO3- to the oxidation of H2S/S2- for the generation of NH4+ which is a more readily utilisable form than NO3- and is less toxic than H2S. As DNRA provides an electron donor19, the process could be linked to lowering levels of reduced sulfur forms in the aquatic system.

Nitrous oxide production was also seen to occur in the Tuvem and Divar sediments. As NO3- removal is mainly routed through the DNRA pathway, it highlights the capacity of mangroves to buffer the climate against the production of N2O through incomplete denitrification and it's consequent flux to the atmosphere.

Until now, mangroves have been known to function as efficient buffer zones mitigating large amounts of intrinsically produced nutrients as well as extraneously derived anthropogenic inputs20 through denitrification. Our study shows that mangroves have the potential to buffer the climate by modulating the production of N2O resulting from incomplete denitrification. This is achieved in exchange for NH4+ that gets retained in the system perhaps within biologically acceptable limits through the DNRA pathway. The DNRA process is a major mechanism for NO3- removal rather than denitrification especially in N limited mangrove systems than those receiving nutrients through extraneous sources. Thus, these mangrove ecosystems have the potential to make a significant contribution to the N pool in coastal waters by accumulating and exporting inorganic N.

In many ecosystems, a possibility of N limitation has been suggested to occur in the near future21. This is mainly attributed to elevated atmospheric CO2 concentrations which can reduce N mineralization22, consequently limiting the nutrient supply to plants. Besides, mangrove ecosystems are generally known to be rich in carbon but limited in N. Hence, N retention through DNRA could be an important strategy to overcome this constraint for ecosystem productivity. Most recent estimates using Global Land Survey (GLS) data and Landsat imagery have shown that the worldwide distribution of mangrove forests amounts to approximately 15x106 ha23. Thus, on a global scale, the prevalence of DNRA in mangrove and other carbon rich systems has critical implications for sustaining ecosystem productivity. We strongly recommend considering DNRA as a relevant process in future N cycling studies in mangrove ecosystems.

Methods

Sampling was carried out at two mangrove forests located along the Mandovi and Chapora rivers in Goa, west coast of India8. Sediment cores (inner diameter 7.5 cm, 20 cm length) for activity measurements were collected at low tide during May, 2008 from anthropogenically-inflenced site Divar (15°30′35′′ N and 73°52′63′′ E) which lies along the river Mandovi and the relatively pristine Tuvem (15°39′94′′ N and 73°47'65′′ E) along the river Chapora. The cores were maintained at 4°C until analyses. The cores were sectioned aseptically at 2 cm intervals to obtain representative samples at 0–2, 2–4, 4–6, 6–8 and 8–10 cm. For each sampling site, sediment corresponding to the same depth were pooled and homogenized. Each homogenized sample was further sub-divided as follows:

  1. i

    Duplicates (10 mL) for immediate analysis of NO3--N in pore water.

  2. ii

    For NRA and net N2O production, triplicate measurements were done at every time interval (0, 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 h; n = 21 for each section of the core).

  3. iii

    For DNRA measurement, duplicate samples were maintained at every time interval (0, 2, 4, 6, 8 and 10 h; n = 12 for each section of the core).

To measure nitrate reducing activity (NRA), approximately 1 g wet weight sediment obtained from each representative section was transferred to 60 mL serum bottles. Ambient seawater was collected from site for media preparation. This seawater contained approximately 4.5 µmol NO3--N L−1. The seawater was amended with allythiourea (ATU) at a pre-standardized concentration of 125 µmol L−1 to inhibit nitrification24. The sediment slurry was briefly vortexed and the bottles were then filled with filter sterilized seawater up to the brim to create micro-aerophilic conditions. The bottles were capped with butyl stoppers and the slurry was gently mixed and incubated in triplicates under static conditions for up to 3 h as the nitrification inhibitor used became ineffective beyond this period. At the end of the sampling period, the bottles were gently swirled. The contents were transferred to 50 mL centrifuge tubes and centrifuged (REMI Compufuge CPR-24) at 5000 rpm and 4°C for 10 min. Nitrate in the supernatant was measured spectrophotometrically25. The NRA was determined from the fall in NO3- level over time and has been expressed on a dry weight basis as µmol NO3--N g−1 h−1. Net N2O production was measured as described elsewhere8.

DNRA measurements were carried out in conjunction with pore water nutrient analysis and measurement of N2-fixation, Anx and DNT by mass spectrometry1. Four mL of homogenised sediment from each section was transferred into 22 mL headspace vials. Four mL of filter sterilized seawater containing NO3--N at a final concentration of 10 µmol L−1 was added. The vials were sealed with butyl stoppers, purged with He and pre-incubated for about an hour before addition of stock solution of 15NO3- (97.4 atom%, Isotech Mathesson, USA)26to obtain a final concentration of 80 µmol L−1. DNRA was measured by monitoring the progressive isotopic enrichment of 15NH4+ for up to 10 h in the dark. Two or three vials were sacrificed by adding HgCl2 (final concentration of 10 mmol L−1) at each point of the time series (0, 2, 4, 6, 8 and 10 h). NH4+-N in pore water and sediment was extracted by microdiffusion27 and the N was analyzed by mass spectrometry. Unlabelled ammonium (1 μmol L−1) was added to the filters after microdiffusion and this quantity was taken into account when calculating the DNRA activity. The samples were treated with a mild alkali (MgO) to convert NH4+ to NH3, which was trapped on acidified (50 μL, 0.5 N H2SO4) pre-combusted Whatman GF/C filters. To calculate the rate of flux from dissolved NO3- to dissolved NH4+, equations derived by analogy with that of Dugdale & Goering28 were used. DNRA was calculated using previously described equations27 and the rate has been expressed as µmol NH4+-N g−1 h−1.