Lessons from A Degradation of Planted Kandelia obovata Mangrove Forest in the Pearl River Estuary, China

Abstract

Kandelia obovata (S., L.) Yong and Sonneratia caseolaris (L.) Engl. are two dominant mangrove species in the subtropical coastlines of the Pearl River Estuary, China. The main aim of this study was to investigate the specific causes of K. obovata mortality versus S. caseolaris vitality on the west coast of Bao’an, Shenzhen, China and to propose sustainable management strategies for mangrove protection and future ecological planting restoration. Results showed that although both mangroves possessed simple and unstable community structures, S. caseolaris had a more tenacious vitality than the native species K. obovata, indicating that the former possesses stronger adaptability under adversity conditions. Moreover, the salinity of the seawater collection point 5 from the K. obovata plot was found to be lower than that of seawater collection point 1–3 from the S. caseolaris sample plots, indicating that no hydrologic connectivity existed in the K. obovata plots. In addition, the location of the drain outlet (seawater collection point 8) might be another potential risk factor for the dead of near K. obovata forests, implying that they were badly affected by poor oxygen and serious inorganic pollution, such as ammonium nitrogen, total phosphorus, and other inorganic substances. Depending on local circumstances, we should consider strengthening infrastructure construction to activate hydrological connectivity, reinforcing the stability of man-made mangrove communities, and controlling the pollution sources for sustainable mangrove protection and management on the western coast of Bao’an, Shenzhen, China.

1. Introduction

Mangroves are special coastal ecosystems which mainly occur globally in the intertidal estuaries of tropical and subtropical regions and function as major biologically active areas in coastal ecosystems [1,2]. Mangroves are transitional wetland ecosystems between marine and terrestrial environments, occupying approximately 170,600 km² of tropical and subtropical seacoasts and covering 60–75% of coastlines worldwide [3,4]. Mangrove ecosystems have various characteristics, such as high productivity, a high restitution rate, a high decomposition rate, and a high stress tolerance [5]. Mangrove forests can provide large-scale ecosystem services, thereby sustainably maintaining the stability of intertidal and marine ecosystems—for instance, they are used as food and fuel by local communities; contribute to climate regulation via high sequestration and storage capacities; lessen the impacts of sea waves, tides, and storm events on coastlines, embankments and tidal-flat areas; and purify the ocean, thereby maintaining availably and sustainably the stability of intertidal and marine ecosystems [2,6,7].

Unfortunately, because of their location between the sea and the land, mangrove forests are subjected to natural and anthropogenic disturbances, such as climate changes (extreme temperature and precipitation and their related interferences, e.g., sea level rise and extreme climatic events), disease, insect pests, sediment property changes, seawall construction, aquaculture activity, pollution, and loss of tidal connectivity, resulting in serious degeneration of mangrove ecosystems in a global context [8,9,10,11,12]. Particularly, sewage pollution resulting from human activities seriously threatens the health of mangrove forests. For example, excess nutrients from polluted water, e.g., nitrogen (N) and phosphorus (P), might give rise to eutrophication and lead to a series of problems, such as changes in bacterial community structures, decreased carbon storage, and a lack of oxygen in sediments, eventually destroying the behavior and function of wetland ecosystems and causing mangrove death [13,14,15]. Indeed, previous studies reported that approximately 35% of the worlds’ mangrove forests vanished from 1980 to 2000 [11], although the annual rate of mangrove loss was on average 0.26–0.66% globally from 2000 to 2012 and slowed to 0.13% from 2000 to 2016 [10,12]. In China, according to the report of the National Investigation of Forest Resource, the mangrove area once covered 250,000 ha; however, it sharply decreased to 42,000 ha by 1956 [16]. Due to land reclamation from 1970 to 1980 in China, the mangrove forests reduced to 21,000 ha by the end of 1980s and further, to 15,000 ha, by the end of 1990s [16]. Fortunately, China started a 10-year restoration project for mangrove forests in the early 1990s, by which its mangrove reserves have raised to 22,000 ha [16]. Furthermore, over 80% of these mangroves are degraded secondary forests [16]. To sum up, through a series of natural and artificial threats, China has decreased around 50% reserves of mangrove forests during 1950 to 2001, which has been considered as one of the countries with the greatest losses of mangrove forests of the world [16,17].

The costs of degradation in mangrove ecosystems are tremendous, e.g., reduction of genetic diversity, species abundance and diversity, population density, and community structure, negative shifts in ecosystem functions, and loss of valuable ecosystem services [18,19]. More importantly, degradation poses a threat to the restorability of damaged mangrove ecosystems, making them more vulnerable to external environmental factors [20]. Thus, accurately recognizing the exact causes of degradation is one of the most vital steps in mangrove protection.

Overall, public awareness of natural resource conservation, sustainable development, and the necessity for proper management is of great significance to boost coastal environmental quality [21]. Specifically, people in Southeast Asian countries, such as India and Bangladesh, have started to realize and learn how to protect mangrove ecosystems efficiently, which are under severe threat due to natural and anthropogenic causes [21,22,23]. In addition, due to the ineffectiveness of tree planting reforestation in Kuala Gula, Malaysia, the Ecological Mangrove Rehabilitation Workshop was established to subsidize, organize, and monitor the mangrove rehabilitation programs for stabilizing coastal regions [24]. However, how can mangrove forests be protected? The basic principle of adaptation to local conditions for mangrove protection can be especially useful in addressing this issue. Making clear the history of local mangrove development is vitally important for determining the main deforestation drivers and for working out effective management plans [25]. The improvement of mangroves’ surroundings includes both protection and restoration. For example, if hydrological and sediment conditions are well conserved or rehabilitated, deforested mangroves are able to self-recover [26], although at a lower growth rate than man-made forests [27]. Additionally, removing the abandoned shrimp farms and salt ponds would obviously boost the water and sediment states and allow the generation of waterborne mangrove propagules [28]. However, the strategies of management, protection, and rehabilitation in Chinese mangroves are seldom elucidated and need to be pointed out clearly.

Kandelia obovata (S., L.) Yong and Sonneratia caseolaris (L.) Engl., two dominant mangrove species characterized by fast growth and proliferation, are widely distributed in similar tidal zones along the coastline in South China [29,30]. The difference between the two mangrove plants is that K. obovata is a common indigenous species in Shenzhen, Guangdong Province, China, but S. caseolaris is introduced and planted outside and north of its native habitat in Hainan Province, China [29,31]. On the west coast of Bao’an, Shenzhen, China, S. caseolaris and K. obovata were well cultivated on a considerable scale for mangrove restoration in 1999–2003. Interestingly, K. obovata did not thrive, while S. caseolaris showed healthy growth in recent years. In the present study, we aimed to investigate in detail the causes of degradation of the introduced mangrove species K. obovata and those that allowed S. caseolaris to grow well and to provide valuable suggestions for mangrove protection and management on the western coast of Bao’an, Shenzhen, China. This study, as a typical case, can provide theoretical references for specific mangrove protection and recovery in the future.

2. Materials and Methods

2.1. Research Area

This study was conducted on the Bao’an west coast of Shenzhen, Guangdong Province, China (22°42′ N, 113°45′ E, Figure 1), during the growing seasons of 2017 and 2018. Bao’an District of Shenzhen is located on the east coast of the Pearl River Delta and has a typical subtropical monsoonal climate. This site has a total area of 393.54 km², and a coastline 45.30 km in length, and its mangrove area is the second-largest coverage area in Shenzhen [32]. There are 10 mangrove species widely distributed in Bao’an District: K. obovata, S. caseolaris, Bruguiera gymnorrhiza, Avicennia marina, Sonneratia apetala, Aegiceras corniculatum, Acanthus ilicifolius, Heritiera littoralis, Acrostichum aureum, and Excoecaria agallocha [32]. On the west coast of Bao’an, the mangrove communities are mainly K. obovata, A. ilicifolius, A. Aureum, and S. caseolaris, the trees are 3–4 m high, the oldest tree at this site is more than 15 years old, and most trees are 7 to 9 years old [32]. Additionally, some herbaceous plants, such as Phragmites australis and other Gramineous sp., were also widely distributed in this coastal region. At present, to further expand the city scale, the Bao’an west coast has continued to reclaim land from the sea, resulting in the beach area in this district shrinking to 49 hm² [32].

2.2. Experimental Design

First, we selected the plant sample plots for ecological investigations mainly due to the target mangrove plants, namely, the dominant species in their respective areas (Ko Plots 1–5 and Sc Plots 6–8). In addition, we also chose some sample points for evaluating the qualities of seawater and sediments in/around plots 1–8 along a walking line (A-G for sediment collection and S1-S8 in red circles for seawater collection, respectively). Since the research area was in the intertidal zone, where it was difficult to set up sample plots with uniform size, we established 8 plots (size: 75–150 m²) for vegetation investigation. The distance between two adjacent sampling plots was 0.5–1 km. The field of the two forests were recorded and parameters of the investigation were as follows: plant species and health conditions, community form and composition, height, arborous layer coverage, and tidal conditions.

2.3. Calculations of Several Ecological indices

The importance value is a quantitative index used to investigate the status and function of a certain species in a community [33]. The importance value was calculated as follows:

$$\mathrm{Importance} \mathrm{value}=\frac{\mathrm{relative} \mathrm{density}+\mathrm{relative} \mathrm{frequency}+\mathrm{relative} \mathrm{dominance}}{3}$$

(1)

where relative density is the density of one species (the density of all species × 100%), relative frequency is the frequency of one species (the frequency of all species × 100%), and relative dominance is the coverage of one species (the dominance of all species × 100%).

We also assessed the Patrick richness index (R), Margalef index (E), Simpson index (D), Shannon-Wiener index (H’), and Pielou community evenness index (Jsw) to manifest the mangrove species diversity in this study. These indices were calculated as follows:

$$\mathrm{R}=\mathrm{S}$$

(2)

$$\mathrm{E}=\frac{\mathrm{S}-1}{\mathrm{lnN}}$$

(3)

$$\mathrm{D}=1-\sum {\left(\frac{\mathrm{Ni}}{\mathrm{N}}\right)}^2$$

(4)

$$\mathrm{H}\prime =-{\sum }^{ }\left(\frac{\mathrm{Ni}}{\mathrm{N}}\right)\ln \left(\frac{\mathrm{Ni}}{\mathrm{N}}\right)$$

(5)

$$\mathrm{Jsw}=\frac{\mathrm{H}\prime }{\mathrm{lnS}}=-{\sum }^{ }\frac{\mathrm{Ni}}{\mathrm{N}}\frac{\ln \frac{\mathrm{Ni}}{\mathrm{N}}}{\mathrm{lnS}}$$

(6)

where N is the total number of plant individuals in each plot, Ni is the number of individuals of a species in each plot, and S is the number of species in each community.

2.4. Sampling Collection and Processing

In the present study, we collected samples of roots, stems, and leaves of mangrove plants, sediments, and seawater. Specifically, samples from K. obovata and S. caseolaris were chosen as randomly as possible, and sediment and seawater samples were both collected from the top layer (0–10 cm). Thereafter, all the collected samples were transported to our laboratory for analysis, using at least three replicates from each plot.

The fundamental elements, such as carbon (C) and N, were measured via an automated elemental analyzer (Vario MACRO cube, Elementar, Germany). Heavy metals were determined using inductively coupled plasma-mass spectrometry (ICP-MS, Agilent 7700x, Mulgrave, Australia). The TN in seawater was tested according to the Kjeldahl method [34]. The total phosphorus (TP) was analyzed via the molybdenum blue method [35]. The concentrations of ammonium nitrogen/nitrate nitrogen (NH₃-N/NO_x-N) in seawater were determined by a flow injection analyzer (FIA, Lachat QuikChem 8500 Series II, USA). The six heavy metals were determined by inductively coupled plasma–mass spectrometry (ICP-MS, Agilent 7700x, America) according to one of our previous studies [36]. The biological and chemical oxygen demand (BOD and COD, respectively) analyses were performed according to the protocols described by the American Public Health Association [37]. The salt content (SC), pH, redox potential (RP), electric conductivity (EC), and dissolved oxygen (DO) in sediment were measured using a salimeter (AR8012, Jiangsu, China), a pH meter (PHS3E, Shanghai, China), an oxidation–reduction potential (ORP) meter (P330, Guangzhou, China), a conductivity meter (YQ-012, Shanghai, China), and a water analyzer meter (P330, Guangzhou, China), respectively.

For heavy metal testing, the ICP-MS parameters were set according to Hao et al. (2021) and Wang et al. (2021) [38,39]. To be specific, the calibration standards of quality assurance (QA) and quality control (QC) with concentration levels ranged from 0.001 to 1 mg/L, which were prepared from a standard solution with multi-elements (ICP-MS-CAL2-1, AccuStandard, Germany), and 0.5% quality fraction of nitric acid (HNO₃) were used to bracket the abundance of elements in the digests. An internal standard, e.g., 50 μg/L of Rh and In, was used to correct the instrumental drift and matrix effects, and it was applied to blanks, calibration standards, and samples. The recovery rate and the detection limit were operated at 90–110% and 0.002–0.054 mg·kg⁻¹, respectively. If the recovery rate outstripped the tolerance range, the samples would be measured again.

2.5. Description of Statistical Analysis

All the acquired data were analyzed by Excel 2010. In addition, the data from Figure 2, Figure 3 and Figure 4 were subjected to one-way analysis of variance (ANOVA) with Student’s t test for statistical analysis and indicated as “±SD” by SPSS 20.0. When p < 0.05, differences between the two averages were considered statistically significant unless otherwise stated. Principal component analysis (PCA) was applied on various parameters of seawater or three mangrove trees (K. obovata, S. caseolaris, and B. gymnorrhiza), and visualized as a biplot of principal component 1 and 2 using the R package. All the figures in this study were drawn using OriginPro 2018 or R package.

3. Result

3.1. Vegetation Characteristics, Ecological Indices, and Element Contents of Planted Mangrove Communities on the Bao’an West Coast of Shenzhen

To assess the growth conditions of planted mangroves on the Bao’an west coast of Shenzhen, herein, we first investigated the vegetation characteristics of K. obovata and S. caseolaris communities. Specifically, the accompanying species of the two types of mangrove communities were similar in that they both mainly consisted of A. ilicifolius and A. aureum (

Table 1. The states of mangrove community characteristics and hydrological connectivity along the western coast of Bao’an, Shenzhen.

Sample Number	Acreage (m2)	Community Composition	Height (m)	Coverage (%)	Community Morpha	Growth Conditions	Tidal Conditions
Ko 1	120	K. obovata A. ilicifolius	4.0–5.0	10	Small Arbor	Arbor Death	No Tides
Ko 2	110	K. obovata A. ilicifolius A. aureum	3.5–4.5	10	Arbor + Shrub	Arbor Death	No Tides
Ko 3	100	K. obovata A. ilicifolius A. aureum	3.5–4.5	15	Arbor + Shrub	Arbor Death	No Tides
Ko 4	75	K. obovata A. aureum	3.5–4.0	8	Small Arbor	Arbor Death	No Tides
Ko 5	100	K. obovata A. ilicifolius A. aureum	3.0–4.0	20	Arbor + Shrub	Arbor Death	No Tides
Sc 6	150	S. caseolaris A. aureum	5.5–6.5	98	Small Arbor	Normal Growth	Normal Tides
Sc 7	100	S. caseolaris A. ilicifolius	5.0–6.0	95	Small Arbor	Normal Growth	Normal Tides
Sc 8	110	S. caseolaris	6.0–7.0	95	Arbor	Normal Growth	Normal Tides

Note: “Ko 1”: means K. obovata 1 and “Sc 1” means S. caseolaris 1, and so on.

). In addition, the morpha of the K. obovata and S. caseolaris communities both belonged to small arbors or mixtures of arbors and shrubs (Table 1). However, the height and coverage in the K. obovata sample plots were 3.0–5.0 m and 8–15%, respectively, which were notably less than those in the S. caseolaris sample plots (5.0–7.0 m and 95–98%, respectively). Interestingly, the individuals of K. obovata communities in plots 1–5 were all dead in the arbor layer, while S. caseolaris in these plots were still alive (Table 1). Moreover, tidal conditions hardly existed in the K. obovata communities, but those in the S. caseolaris communities were normal (Table 1).

The importance value, which serves to measure the status and functions of plant species in communities, is the most widely applied metric to assess species diversity [34]. As depicted in

Table 2. Importance values of the arbor layer in the K. obovata and S. caseolaris communities on the west coast of Bao’an, Shenzhen.

Community	Main Species	Quantity	Relative Density, %	Relative Frequency, %	Relative Dominance, %	Importance Value, %
K. obovata	K. obovata	3396	99.24	36.71	99.98	78.64
	B. gymnorrhiza	16	0.47	28.57	0.01	9.68
	S. apetala	7	0.20	21.43	0.01	7.21
S. caseolaris	S. caseolaris	78	78.08	50.15	76.86	68.36
	S. apetala	19	19.01	33.33	21.66	31.85
	B. gymnorrhiza	3	3.31	16.67	1.49	7.16

, in addition to the dominant species, K. obovata and S. caseolaris, the accompanying woody species in the two target communities were also mangroves, such as S. apetala and B. gymnorrhiza (Table 2). In plots Ko 1–5, the average quantity of K. obovata was 3396, which was dominant among all the mangrove plants (Table 2). However, S. apetala and B. gymnorrhiza were distributed sporadically in the K. obovata and S. caseolaris communities, respectively (Table 2). Therefore, it was found that the importance values of K. obovata and S. caseolaris were the highest in their own communities (Table 2). Namely, the importance value of K. obovata was 8.12 times higher than that of B. gymnorrhiza and 10.91 times higher than that of S. apetala in its community, and the importance value of S. caseolaris was 2.15–9.55 times higher than those of S. apetala and B. gymnorrhiza in sample plots Sc 6–8 (Table 2).

We also analyzed the importance values of plants from the shrub-grass layer in the two target mangrove communities. There were no great differences between the two mangrove communities in terms of the composition of the shrub layer; the shrub layers of both communities were composed of A. ilicifolius and A. corniculatum (

Table 3. Importance values of the shrub-grass layer in the K. obovata and S. caseolaris communities on the west coast of Bao’an, Shenzhen.

Community	Main Species	Quantity	Relative Density, %	Relative Frequency, %	Relative Dominance, %	Importance Value, %
K. obovata	A. ilicifolius	30	28.30	23.53	19.32	23.72
	A. corniculatum	6	5.67	11.76	12.97	10.13
	A. aureum	28	26.42	29.41	58.22	38.02
	P. australis	8	7.54	5.88	5.52	6.31
	other Poaceae sp.	34	32.08	29.41	39.74	33.74
S. caseolaris	A. ilicifolius	24	31.58	20.32	17.44	23.11
	A. corniculatum	14	18.42	20.03	34.16	24.20
	A. aureum	19	25.14	31.21	44.60	33.65
	P. australis	2	2.63	11.57	1.56	5.25
	other Poaceae sp.	17	22.37	18.78	2.24	14.46

). However, the importance value of the shrub layer from the K. obovata community was 33.76% lower than that from the S. caseolaris community (Table 3). In contrast, the importance value of plants from the herbaceous layer in the K. obovata community, including A. aureum and P. australis, which have the largest ecological niches in herbal plants from the mangrove community, was 1.46 times higher than that in the S. caseolaris community (Table 3).

As shown in Table 3, the degrees of E, D, H’, and Jsw in the S. caseolaris community were all far greater than those in the K. obovata community, being 1.58 times greater in E, 11.06 times in D, 7.99 times in H’, and 7.99 times in Jsw (

Table 4. Biodiversity indices of K. obovata and S. caseolaris communities on the west coast of Bao’an, Shenzhen.

Community	R	E	D	H’	Jsw
K. obovata	8	0.8573	0.0674	0.2078	0.0999
S. caseolaris	8	1.3538	0.7456	1.6604	0.7985

Note: “R”: means Patrick richness index, “E” means Margalef index, “D”: means Simpson index, “H’”: means Shannon-Wiener index, and “Jsw”: means Pielou community evenness index.

). However, the Patrick richness indices (R) in the S. caseolaris and K. obovata communities were identical, meaning that the variation in the species diversity level was dependent on species richness (Table 4).

To determine the specific reason for the death of K. obovata, we first tested the contents of C and N from mangrove plant samples in S. caseolaris and K. obovata communities to distinguish the differences between the dead K. obovata and other live mangrove plants. In Figure 2, the C content (approximately 48% in both mangrove forests) in the two mangrove forests was much higher than N (approximately 1.5% in S. caseolaris forests and 0.8% in K. obovata forests, the variance analysis results were in Table S1, the same below). In addition, we also found that the trends of C and N contents between forests of S. caseolaris and K. obovata were totally different. The content of C in S. caseolaris forests showed little difference (p = 0.82) from that in K. obovata forests, while the N content in S. caseolaris forests was obviously higher (p = 0.03) than that in K. obovata forests (Figure 2). In the S. caseolaris community, the content of C was more highly distributed in stems than in leaves; in contrast, N was mainly found in leaves instead of stems (Figure 2). Additionally, in K. obovata forests, the leaves from mangrove trees that were still alive, such as K. obovata and B. gymnorrhiza, were noticeably richer in N compared to the dead samplings (Figure 2). Interestingly, there were almost no differences between dead or live K. obovata plants from the K. obovata community in terms of the contents of C and N (Figure 2).

Additionally, we determined the concentrations of six heavy metals, Fe, Al, Cu, Zn, Cr, and Ni, from mangrove plant samples in S. caseolaris and K. obovata communities. Specifically, the Fe concentrations in the two mangrove forests were far higher (p = 1.09 × 10⁻¹¹) than those of the other five heavy metals (Figure 3). Moreover, except for Fe, the concentrations of Al, Cu, Zn, Cr, and Ni in most cases in S. caseolaris forests were higher than those in K. obovata forests (Figure 3). In the K. obovata community, the concentrations of Al, Cu, Zn, Cr, and Ni from B. gymnorrhiza plants were apparently higher than those in the dominant species, K. obovata (Figure 3). Similar to the contents of essential elements (C and N), the concentrations of six heavy metals between dead and alive K. obovata were not remarkably different in the K. obovata community (Figure 3).

From principal component 1 (PC1) of the PCA analysis, which explained 41.0% of the variance in the concentrations of six heavy metal data separated out stems and leaves of K. obovata, S. caseolaris, and B. gymnorrhiza (Figure S1). Heavy metal samples from S. caseolaris leaves and dead B. gymnorrhiza stems clustered far from others at PC1. Principal component 2 (PC2) further divided stems and leaves samples with 31.6% of the variance in the data (Figure S1).

3.2. Physicochemical Properties of Seawater and Sediments in or near the Planted Mangrove Communities on the West Coast of Bao’an, Shenzhen

In terms of seawater samples near the K. obovata and S. caseolaris communities, we investigated some vital physicochemical properties, such as salinity, pH, RP, EC, DO, TN, TP, NH₃-N, NO_x-N, BOD, and COD, to further elucidate the reason for K. obovata plant mortality. In

Table 5. Physicochemical properties of seawater near K. obovata and S. caseolaris communities on the west coast of Bao’an, Shenzhen. Mean ± SD of three replicates were shown. Different small letters indicate significant differences among various physicochemical parameters from the 8 seawater collection points at p < 0.05.

Seawater Collection Points	Salinity (‰)	pH	Redox Potential (mV)	Electrical Conductivity (ms/cm)	Dissolved Oxygen (mg/L)
S1	12.47 ± 1.01 b	7.43 ± 0.05 a	435.33 ± 8.50 a	21.53 ± 0.23 bc	5.76 ± 0.37 a
S2	13.33 ± 0.42 ab	7.38 ± 0.04 a	426.60 ± 0.52 a	22.23 ± 0.72 abc	5.75 ± 0.98 a
S3	11.80 ± 0.44 b	7.95 ± 0.02 a	403.73 ± 1.46 a	19.92 ± 0.36 c	7.37 ± 0.33 a
S4	15.23 ± 0.76 a	7.53 ± 0.04 a	441.87 ± 1.07 a	24.80 ± 1.28 ab	6.52 ± 0.08 a
S5	6.64 ± 1.52 c	7.87 ± 0.79 a	448.70 ± 34.78 a	11.85 ± 2.50 d	8.64 ± 3.13 a
S6	14.00 ± 0.10 ab	7.48 ± 0.14 a	442.30 ± 2.95 a	23.17 ± 0.31 abc	4.83 ± 1.03 a
S7	15.63 ± 0.06 a	7.55 ± 0.06 a	449.53 ± 5.85 a	25.47 ± 0.23 a	7.00 ± 0.11 a
S8	5.08 ± 0.22 c	7.51 ± 0.03 a	292.50 ± 16.16 b	9.27 ± 0.22 d	0.07 ± 0.02 b

Note: “S1” means seawater collection point 1, and so on.

, we found that there were few differences among the physicochemical properties of seawater samples near the K. obovata and S. caseolaris forests except the drain outlet (S8, Table 5); namely, the salinity, RP, EC, and DO in S8 were far lower than those in other seawater collection points (S1–S7, Table 5). In addition, the salinity of S5 also showed a significant difference from other seawater collection points (Table 5). Regarding seawater pollution, we discovered that the TN, TP, NH₃-N, and BOD at S8 was highest among all the seawater collection points (Figure 4). Moreover, the concentration of NH₃-N in S6 also had the highest value in S1–S8 (Figure 4). In contrast, S5, which was collected from the K. obovata community, had the minimum values of all the tested seawater pollution indices (Figure 4).

As for various physicochemical parameters in seawater collection points, the PCA plot showed that S8 separated well along PC1, which explained 57.7% of the variation in data of physicochemical parameters from the seawater (Figure S2). S5 clustered away from other seawater collection points (except S8) along PC2, which explained 22.5% of the variance in the data of physicochemical parameters.

The physicochemical properties of sediments in the two mangrove communities, such as salinity, pH, MC, and OMC, were also tested in the present study. The results showed that the salinity in sediments from plot Ko 3 was significantly higher than that in other mangrove sample plots (Figure S3A). However, there were no obvious differences among the sediment points of A-H in terms of pH and MC (Figure S3B–C). The A and D sediment points had the highest OMC content of all the sediment points (Figure S3D).

Unlike the heavy metals in mangrove plants, the concentration of Al in the sediments was the highest in all of the tested heavy metals (Figure S4). In addition, the orders of magnitude from Fe and Al concentrations were far greater than those from the concentrations of Cu, Zn, Cr, and Ni (Figure S4). Compared to S. caseolaris forests, the heavy metals in sediments of K. obovata forests were not different except for the Al and Cu concentrations from sediment collection point H.

4. Discussion

It is generally accepted that mangroves are an important part of coastal shelterbelt systems in tropical and subtropical regions and play an important role in coastal protection and maintenance of coastal ecosystem function [1,2]. Unfortunately, mangrove wetlands have declined by 30–50% worldwide in the last century due to human disturbances such as urban development and aquaculture, with particularly severe degradation in Asia, including China [40,41]. In this study, we tried to discover the specific causes of mangrove forest degradation and raise some sustainable management strategies via a case study on dead planted K. obovata forests using live S. caseolaris forests nearby as a control on the western coast of Bao’an, Shenzhen, China. This research may allow us to better understand how to determine the precise causes of mangrove degeneration and adequately protect mangrove forests. The environment, which provides all material resources for living plants, plays a key role in vegetation development [42]. In this research, we first found that two nearby mangrove forests on the west coast of Bao’an, Shenzhen had totally different growth conditions, namely, dead K. obovata and live S. caseolaris, indicating that the climate might not be the main environmental factor in this case (Figure 1). Indeed, our previous study on the relationships between three different kinds of medicinal plants and their environments found that other environmental factors, such as aspects and soil types, rather than the climate, mainly determined plants’ growth conditions [43]. Another finding of this research on the plant community structures was that planted mangrove forests had a low degree of diversity, for instance, the low values of E, D, H’, and Jsw of K. obovata communities (Table 4), which suggested that these man-made mangrove communities might lack ecological stability. Compared to a natural forest, a planted forest with lower biodiversity is more vulnerable to suffering dramatic disasters and is less able to adapt when faced with external challenges [44]. Therefore, we hold the view that lack of biodiversity may be another reason for the death of K. obovata forest. Regardless, compared with the K. obovata forest in the same area, why did the S. caseolaris forest nearby still survive regardless of the external disturbances? We suggest that the ecological characteristics of S. caseolaris could account for this. For instance, as an introduced mangrove species, S. caseolaris has faster growth percentages, higher survival rates, stronger stress resistance, and a better capacity to establish populations in a variety of habitats than native mangrove species, such as K. obovata, which is sensitive to adverse situations [45]. Indeed, we also discovered that the N in leaves from S. caseolaris was markedly higher than that in leaves from K. obovata, indicating that the development of S. caseolaris might be better than that of K. obovata (Figure 2).

For heavy metal determination in plants and sediments, we found that there were no significant differences between the two kinds of sample plots, suggesting that heavy metals were not the main factors causing mangrove mortality (Figure 3 and Figure S4). However, in terms of seawater, the salinity of S5 from the K. obovata plot was much lower than that of S1–S4 (S1–S3 from S. caseolaris sample plots and S4 from the Pearl River Estuary); this result was mainly because no tides existed in the K. obovata sample plots (Table 1 and Table 5). Indeed, the original hydrological connectivity with the Pearl River was changed by local human activities, such as fish pond farming and construction, which made the mangroves flooded for a long time. Hydrological connectivity functions as a mover of matter, energy, or organisms within or between wetland elements or their hydrologic fluxes [46]. Jimenez et al. (1985) reported that hindering the development of wetland connectivity is one of the major causes of wetland degradation globally [47]. Combined with macroclimatic alterations, a lack of hydrologic connectivity can pose a highly detrimental threat to these wetland ecosystems [48]. In addition, it is worth mentioning that S. caseolaris possesses strong fingerlike respiratory roots, which can benefit a long vitality under anoxic flooding stress than K. obovata without those [49].

Except these, the location of the drain outlet (S8) might be another potential risk factor for the near K. obovata forests because of serious inorganic pollution (Figure 1 and Figure 4). The discharge of high loads of nutrients (including N) might give rise to eutrophication and lead to a series of problems, such as changes in bacterial community structures, decreased carbon storage, and a lack of oxygen in sediments, eventually destroying the behavior and function of wetland ecosystems and causing mangrove death [13,14,15]. Furthermore, excess N enrichment could also enhance the sensitivity to salinity and drought since N-elicited promotion of allocation to the canopy rather than roots could indirectly boost mortality rates due to enhanced susceptibility to water deficits [20].

Based on the above discussion, we propose some suggestions for sustainable management of mangrove forests on the west coast of Bao’an, Shenzhen: (1) Strengthening infrastructure construction to allow hydrological connectivity. We must respect the natural laws of the tides and understand that facilitating hydrological connectivity is key to survival for mangrove forests. The implementation of mangrove wetland ecological restoration projects should fully consider the construction of water conservancy facilities related to hydrological connectivity and learn from the existing technology of mangrove wetland restoration demonstration areas, such as fish pond restoration, river reconstruction technology, construction of water circulation ecological channels, and side weir gates regulating the tidal range at the estuary. (2) Reinforcing the ecological stability of man-made mangrove communities. The main reason why such continuous patches of death occurred in this project area, resulting in loss of wetland area and function, as well as adverse social impact, is that the number of tree species in the mangrove community is too low. Low diversity in community structure reduces the stability of wetland ecosystems and increases vulnerability to disasters [47]. Therefore, in subsequent construction of mangrove wetland protection areas, experimental areas, purification areas and ecological parks, we should pay attention to increasing plant species, rationally handling the relationship between exotic species and native species, and scientifically designing community structures. (3) Overall control of pollution sources is necessary, because unpolluted seawater is important for wetlands to thrive. Scientific sewage interception, rainwater diversion technology, and efficient comprehensive sewage treatment facilities are the infrastructures needed for regional ecological development. At the same time, mangrove plants (i.e., B. gymnorrhiza and A. ilicifolius) with high pollution resistance and decontamination capacity in coastal wetlands can also play a key role in the bioremediation of pollution [50,51].

5. Conclusions

Implementing concrete strategies of ecological restoration in mangrove wetlands should combine theory and practice. In this research, our findings indicated that there were three main risk factors for K. obovata mortality on the west coast of Bao’an, Shenzhen, China, namely, lack of hydrologic connectivity (lack of oxygen), the simple plant community structures (low biodiversity and ecological stability), and seawater pollution (excess N, P, and BOD in seawater). Therefore, to ensure sustainable management of this area or some other area with a similar problem, we recommend implementing several measures. These include improving infrastructure development to facilitate hydrological connectivity, enhancing the stability of the human-made mangrove communities, and managing pollution sources.