Variation in sediment characteristics between canopy gap and surrounding forest in Can Gio Mangrove Biosphere Reserve, Vietnam

This study presented the difference in the sediments under a large gap formed by the Durian typhoon in 2006 and its surrounding intact forest in Can Gio Mangrove Biosphere Reserve, Vietnam. Twelve plots previously settled in the original gap and its surrounding forest were used to collect sediments in the dry and wet season in 2019 to assess the difference in sediment properties and nutritional state caused by the self-recovery of the vegetation. Sediment temperature and pH were measured in field, while the contents of total organic carbon (Corg) and total nitrogen (Ntot) were subsequently quantified in the chemical laboratory. Sediment temperature, salinity, and pH between the gap and the surrounding forest were not statistically different. The variation of Corg concentration in the gap sediments seemingly increased with depth from 1–15 cm in the dry season. The Ntot in the gap sediments showed that it was continuously consumed in the underlying sediments during both seasons. In conclusion, the variation in vegetation composition between the gap and its surrounding forest was responsible for the differences in sediment C and N concentrations.


Introduction
Mangroves are plant communities presenting in intertidal zones along the coasts of tropical and subtropical regions [1].They adapt well to disturbances and are sustainable ecosystems that provide various ecosystem services for humans, such as foods, fuels, materials, medicines, honey, pollination, the nursery for valuable fishes, and coastal protection from natural disasters rising from oceans [2][3][4][5][6][7][8][9][10].Mangroves are a significant ecological component in the carbon cycle; hence, they are regarded as carbon storage with an average stock of 956 t C/ha [11].
Tropical storms can severely damage mangrove ecosystems, and several studies suggested that their intensity might increase due to climate change impacts caused by anthropogenic activities [24,25].The coastal exposure of mangrove forests and the low diversity in their structure make them particularly vulnerable to storm disturbance.Tree mortality in mangroves attacked by tropical cyclones varied between 25 and 100 % of the forest area and lasted for several months [26].Canopy gap formation may lead to differences in physical and biological processes compared with the surrounding forest, including increasing radiation intensity and promoting the growth rate of seedlings.Biodiversity and complexity of plant community structure also increase in gaps due to habitat partitioning [27][28][29][30].Differences in plant age and species composition will cause significant differences in yield, composition, and quality of plant litter between the gap and its surrounding forest, and these differences could be reflected in the sediment's nutritional state and other properties.The sediment in the mangroves is formed by the deposition of materials of different compositions and origins, both inorganic and organic, brought by flows into the area.The mangrove sediments are characterized by unconsolidation in the upper sediments, which can be several meters deep.Once the sediment's conditions change, it will lead to the response of mangrove vegetation which can be revealed in the plant growth rate, species composition, and productivity.Understanding the coupling between mangrove secondary succession after disturbance and its sediment floor characteristics is a requisite for mangrove management to maintain the ecosystem's health in the context of global climate change.
The difference in sediment physico-chemical properties between the canopy gap and its surrounding forest has been studied extensively.However, those studies focused on the early stages of gap formation when the difference was caused mainly by sudden changes in light and evaporation intensity in the gaps [29][30].This research is part of the long-term monitoring of mangrove ecosystem secondary succession in Can Gio Mangrove Biosphere Reserve after being disturbed by the Durian typhoon in December 2006.In this study, we examine the difference in the sediments under a large gap formed by the Durian typhoon and its surrounding intact forest after 13 years of vegetation selfrecovery, when the dissimilarity in plant diversity and structure between the gap and its surrounding forest was pronounced, to find out the response of the sediment floor to the vegetation succession.

Study area
The study area is located in Can Gio Mangrove Biosphere Reserve (10°22′-10°40′ N, 106°46′-107°00′ E), in an estuary formed by Saigon and Dong Nai River.The mean annual temperature is 25.8 o C. The minimum and maximum temperatures are 18.8 o C and 35 o C, respectively.The mean annual precipitation varies between 1300-1400 mm•yr -1 .Most of the precipitation was reported in the rainy season (May -October).Can Gio is characterized by the semi-diurnal tidal regime.Tidal amplitude ranges from 2 m at neap tide to 4 m at spring tide.The reserve is affected by wastewater from Ho Chi Minh City through Long Tau and Soai Rap River [32].
Typhoon Durian in 12/2006 created several large gaps, with a total affected area of approximately 27 ha, inside the Rhizophora apiculata Blume plantations in the southern part of Can Gio.The study area locates in the 17 th compartment of the mangrove biosphere reserve.At the time of the disturbance, the tree's height was 18 -20 m.The initial gap in the study area was 60,000 m 2 .According to field observation, the vegetation coverage area almost completely recovered after 13 years (2006-2019).R. apiculata dominated almost entirely in the surrounding intact forest, while the diversity of plant species increased significantly in the gap, with the presence of woody plants of Avicennia, Xylocarpus, Sonneratia, Lumnitzera, and Ceriops.The dominance of the vegetation in the gap was to Avicennia (A. alba Blume & A. officinalis L.).Besides, there were non-woody plants in the gap, including Finlaysonia obovate Wall., Derris trifoliata Lour., and Nypa fruticans Wurmb.

Figure 1.
The study area in Can Gio Mangrove Biosphere Reserve [33].The yellow and red markers show the sampling sites in the gap and its surrounding forest, respectively.

Methods
A piston corer was used to collect one sediment core (50 cm in length) from each plot during the ebbs.The temperature and pH values of each five-centimeter depth were measured in situ by the immediate insertion of electrodes into the sediment core via inlets on the wall of the corer.The sediment pH was measured with a sulfide-resistant electrode SEA/SE® (Schott, Germany) containing a sensor for temperature measurement.
After the measurements, the sediment cores were sectioned into five-centimeter subsamples by a knife sterilized with absolute alcohol to avoid organic contamination between depths.A tight correlation between the sedimentary nutrient levels at 0-15 cm and 30-35 cm and leaf nutrient concentrations was previously reported in Can Gio Mangrove Biosphere Reserve [34].Consequently, the subsamples at these four depths were stored at 4 o C during transportation to the laboratory for further analyses.
A quantity of 5 g fresh sediment was suspended in 25 mL distilled water for 12 h at room temperature.The conductivity and temperature of the suspension were measured (TetraCon 96, WTW, Germany) afterward, and sediment salinity was calculated according to Ensminger (1996).Large amounts of fresh sediments (60-70 g) were used for humidity determination.The fresh sediments were homogenized in the container by a spatula prior to spreading in a sterile petri plate and dried at 60 o C to constant weight.Humidity was calculated by the difference between the weight of sediment before and after drying.The dry sediment was subsequently homogenized by grinding and passing through a sieve that was 250 μm in the mesh size for the chemical analyses.
To quantify total nitrogen (Ntot) and total organic carbon (Corg), 10 mg of dry and homogenized sediment was weighed into a tin cup and wrapped.Silver cups were used for Corg analysis because the sediment had to be acidified by 1 N HCl to remove carbonate.Standard Leco 1013 was utilized for a fifteen-point calibration and as a quality control after every five samples.Corg and Ntot were quantified with an elemental analyzer Fisons NA 2100 (Germany).
The data analysis was carried out with Statgraphic Centurion XVI.The normal distribution of the data was tested by the Shapiro -Wilks test.The differences between 2 seasons, as well as between the gap and its surrounding intact forest, were tested by one-way ANOVA analysis with α = 0.05.Correlations between the properties are tested by the Pearson product-moment correlation.The correlations are accepted when p-values are less than 0.05.

Physico-chemical properties of the sediments
The range, average values and standard deviations of the physico-chemical properties from each depth in the gaps and its surrounding forest acquired in the dry and wet season are presented in table 1.The ranges of sediment temperature in the gap and its surrounding forest were identical in both seasons (table 1).The sediment temperature did not significantly varied with depths although they tended to increase slightly at 30-35 cm (table 1).The temperature values in the forest section were constantly lower compared to the gap (table 1) but the differences between two sections (gap and the surrounding forest) were not statistically significant (p > 0.01; table 2).The temperature values recorded in the wet season were significantly higher than in the dry season (table 1), except for those acquired from the surface (0-5 cm) sediment in the forest section (table 2).
The sediment pH values in the gap and its surrounding forest tended to constantly increase with depth in the dry season, but in the wet season, a reverse tendency was found in the forest section (table 1).However, the differences between the depths were not remarkable (p > 0.05; table 1).No considerable difference in pH between the gap and its surrounding forest was recorded (table 2).On the contrary, the differences between the two seasons were more evident in the gap compared to the forest section (table 2), although the pH values in the dry season were consistently higher than in the wet season (table 1).The sediment humidity significantly decreased against the depths in both sections in the dry season (table 1).This tendency was found again in the wet season, but the significant difference between the depths was only reported in the forest section (table 1).In the dry season, the lower sediment humidity in the gap was found at all surveyed depths, but in the wet season, similar findings were acquired at 5-15 cm (table 1), and all of the differences between the two sectors were not substantial (table 2).The humidity in the forest section did not vary significantly in the wet season, but in the dry season, the humidity values collected from the depth of 0-5 cm and 30-35 cm were substantially higher in the wet season (p < 0.05; table 1).
No clear trends of depth variations in salinity were found in both sections in the sampling year, except for a significant increase against the depths in the gap in the wet season (table 1).The salinity in the gap and its surrounding forest were identical in both seasons (table 1), but the salinity in both sections decreased remarkably in the wet season compared to the dry season (table 1 and 2).

Elemental compositions
The range and significance of depth variations of the elemental compositions in the gap in its surrounding forest are presented in table 3.In the dry season, the gap sediment Corg contents continuously increased against the depths while the downward increase was restricted to 0-15 cm in the forest sediment (figure 2a).Nevertheless, these differences were not statistically significant (table 3).No visible tendency was found in the depth variations in the rainy season (figure 2a), and a considerable difference between the depths was found only in the forest section (table 2).No remarkable difference was found between the gap and its surrounding forest (table 4), although the Corg contents in the forest sections were higher than the gap except for the values acquired at 30-35 cm in the wet season (figure 2a).The Corg contents in the forest section expressed no substantial changes between the dry and the wet season (table 4), but in the gap, the Corg contents increased drastically in the wet season relative to the dry season (figure 2a).However, the seasonal difference found at 10-15 cm in the gap was not significant (table 4).The Ntot contents did not vary notably in the study area in the sampling year (table 3).The gap and forest sediment Ntot contents constantly decrease downward in both seasons (figure 2b), but the significant difference between the depths was found only in the forest section in the dry season (table 3).The Ntot contents in the gap and its surrounding forest were not notably different from each other in the wet season (table 4), but in the dry season, the Ntot contents in the forest section were substantially higher compared to the gap with an exception at 30-35 cm (figure 2b; table 4).Though the Ntot contents in the wet seasons were higher relative to the dry season (figure 2b), no significant seasonal difference was found in the gap and its surrounding forest (table 4).The gap and forest sediment Ptot contents consistently decreased significantly against the depth in both seasons (figure 3).The Ptot contents in the gap were higher than in the forest section, but the significant difference was found only at 0-5 cm and 30-35 cm in the dry and wet season, respectively (figure 3; table 4).In the gap section, the Ptot contents were higher in the dry season compared to the wet season, and a similar vague trend was recorded in the forest section (figure 3), but all of these seasonal differences were not statistically significant (table 4).

Discussion
Can Gio was an estuarine-bay delta region formed 6000-5000 years ago, which is covered with marine clay and frequently influenced by the sea over the ages [36].Therefore, the sediments in the study area showed an increase in pH against the depths (table 1).The preliminary results of research taking place in the study area showed that the plant root density in the forest section was higher compared to the gap (Le Dinh Anh Vu pers.comm.).Consequently, the pH values in the forest were lower than the gap, especially in the dry season (table 1), probably due to the effects of root exudation.The predominance of fresh water from the hinterland resulted in a decrease in the pH values in the wet season (table 1 and 2).Due to the strong effects of the tidal water, propagules are continuously transported to the study area.However, in the first years of the succession in the gap, the growth of the propagules was inhibited by the thick cover of the trunk necromass [37].In 2019, despite the increase in plant diversity in the gap, the surficial sediment temperature in this section was higher than in the forest (table 1).Most of the gap area was covered by vegetation in 2019, but seedlings and saplings, whose density was low, were the major components in the vegetation structure with low density [37].On the contrary, the surrounding forest floor was covered by high adult trees with large canopies [37].The difference in vegetation structure resulted in the higher temperature in the gap sediment, but the influence of tidal water clouded the difference between the two sections.On the other hand, the anaerobic decomposition of organic matter in the forest section was probably more vigorous than the gap due to the larger quantity of litterfall, which might lead to a slight increase in temperature at 30-35 cm (table 1).This finding was supported by a significantly higher flux of CH4, a product of anaerobic decomposition, emitted from the forest section relative to the gap [38].Mangroves have been usually considered an effective carbon sink due to the predominance of high longevity and great biomass plants.The sediments with high water content suppress the CO2 emission from the aerobic decomposition of organic matter but release CH4, whose global warming potential is 27-30 times higher than CO2.The high CH4 flux in the forest section indicated that the reductive condition was predominated in the sediment, and hence, the concentrations of plant toxins could rise.The higher water contents at the upper 10 cm of the gap sediment (table 1) implied the influence of tidal water on the sediment humidity.The constantly lower salinity values acquired in the wet season (table 1 and 2) emphasized the salt dilution by precipitation, and the dilution effect was limited with the surface sediments (table 1).The dilution by tidal water might explicate the weak correlation (p < 0.01; r = -0.2547)(figure 4) between the sediment humidity and salinity in the study area.This study agreed with Amir et al. ( 2019) that there was no significant difference in pH and salinity between the gap and its surrounding forest.The weak correlation between humidity and salinity (figure 4) in the study area showed that evaporation was not a major driving force for physical and chemical processes in sediments.In the first years of the succession (2007 -2008), the Corg contents in both sections were higher than other mangroves in Vietnam, while the Ntot contents showed a severe deficiency [37].The high contents of tannins in the trunks and branches left in the gap [39] resulted in the high Corg and low Ntot contents in the gap, while the gap widening after the typhoon might explain the similarity in the abundance of Corg content between the two sections [37].In 2019, the vegetation coverage could explain the resemblance in Corg contents between the gap and its surrounding forest (table 4).However, the forest section is still a monoculture of old Rhizophora apiculata trees, while the diversity in the gap was higher with the younger trees [37].Substantial changes in species composition between the large gap and the surrounding forest were suggested by Ewel et al. [40] and demonstrated by Vogt et al. [29].The higher tree age caused an increase in the amount of plant litter and the proportion of woody materials in the forest [41], while its quality regulates the decomposition rate [41][42][43].An abundance of low-quality and degradation-resistant organic materials caused not only the increase of the Corg but also the decrease of Ntot contents in the surrounding forest, compared to the gap.The C:N ratio of Rhizophora and Avicennia leaf litter in Can Gio were 60 and 22, respectively [44].Therefore, the Corg contents in the gap were lower than the forest section (figure 2a).The increase of Corg and Ntot contents in the wet season (figure 2a and 2b) reinforced the influence of organic matter from the hinterland transported to the study area by tidal water.The differences between the gap and its surrounding forest in the wet season were lower than in the dry season (figure 2a and 2b), probably resulting from the spread of abundantorganic matter tidal water, which could be restrained in the gap due to its lower topography compared to the forest section [37].The high water contents suppressed the organic matter decomposition in the dry season and resulted in the accumulation of Corg and Ntot contents in the sediments.This finding can be seen via the correlation between the sediment humidity and Corg contents (p < 0.05; r = 0.3866), and Ntot contents (p < 0.05; r = 0.7228) (figure 5a  and 6a), but these correlations were erased by the organic matter supplement in the wet season (figure 5b and 6b).The decrease in Ntot contents against the depths in the dry season (figure 2b) showed nitrogen usage by the sediment microbial community, whose structure and function are tightly related to the abundance and quality of litterfall in the upper layers.The diversity of the microbial community, in turn, enriched the Ntot contents in the sediments via their excretion.According to Nguyen Thi Lan Thi et al. (2022), the mean values of the Corg and Ntot contents in the whole study area were 3,63 ± 0,21 % and 0,054 ± 0,003 %.Our results showed that after 13 years of the vegetation self-recovery, the Corg contents decreased to 3.03 ± 0.28 % while the Ntot contents drastically rose to 0.297 ± 0.052 %.The amelioration of the sediment's nutritional state in the study resulted from the changes in the organic matter source to the sediment.In the first years of the succession, the primary source of organic matter sediment was the trunks and branches abandoned in the gap, and the tree uprooted afterward in the surrounding forest.In 2019, leaf litter predominantly contributed to the litterfall of a more diverse plant community in the study area.The lower C:N ratio of the leaf litter compared to the trunks and branches [39] led to the higher turnover rate of the organic matter in the sediments.

Conclusion
The thermo-humidity regime of sediments in the study area was strongly influenced by tides.Therefore, the influence of solar radiation and evapotranspiration was vague, leading to the inability to see significant differences in the physico-chemical characteristics of the background between the survey areas.The self-recovery of the vegetation led to the amelioration of the sediment's nutritional state by the supply of a higher quality source of organic matter to the sediments.However, according to studies conducted at the same time in the area on plant community structure [37], the community in the gap area has not reached the climax.Therefore, it is necessary to continue monitoring the coupling between the sedimentary nutritional state and the changes and litter production yield according to the succession process.Furthermore, it is necessary to conduct a research program to assess greenhouse gas emissions in the region to clarify the role of this mangrove forest in carbon accumulation, in order to have a basis for effective solutions to both protect the health of the mangrove ecosystem and contribute to mitigating the negative manifestations of climate change.

Figure 2 .
Figure 2. Variation of Corg (a) and Ntot contents (b) in the gap and forest sediments in the dry and wet season.

Figure 3 .
Figure 3. Variation of Ptot contents in the gap and forest sediments in the dry and wet season.

Figure 4 .
Figure 4.The weak correlation between sediment humidity and salinity in the gap and its surrounding forest in the dry and wet season.

Figure 5 .
Figure 5.The correlation between the sediment humidity and Corg contents in the dry season (a) and the wet season (b).

Figure 6 .
Figure 6.The correlation between the sediment humidity and Ntot contents in the dry season (a) and the wet season (b).

Table 1 .
The average values of sediment properties in the gap and surrounding forest in the dry and wet season.Data was presented as mean ± standard deviation."Dif."refers to the difference between depths with ** p < 0.01, * p < 0.05, ns: non-significant difference.

Table 2 .
Influence of gap/surrounding forest conditions and seasons on the physico-chemical properties of the sediments.

Table 3 .
Range and significance of depth variation of the elemental compositions in the gap and its surrounding forest's sediments."Dif."Refers to the difference between depths with ** p < 0.01, * p < 0.05, ns: non-significant difference

Table 4 :
Influence of sections and seasons on the elemental compositions and nutrient concentrations in the gap and forest sediments.