Development of Mangrove Sediment Quality Index in Matang Mangrove Forest Reserve, Malaysia: A Synergetic Approach

: Sediment is an important part of heavy metal cycling in the coastal ecosystem, acting as a potential sink and source of inorganic and organic contaminants as environmental conditions change. The productivity of mangroves is utterly dependent on sediment enrichment. Moreover, mangrove sediment can trap pollutants discharged by households, industries, and agriculture activities. In this regard, it is essential to assess sediment quality in the presence–absence of heavy metals that are toxic to most living organisms. Thus, the question of how sediment quality is used as an index in the mangrove domain has arisen. Due to the many complex characteristics such as seasonal zones, tidal patterns, ﬂora and fauna, and water, no speciﬁc method is used in Malaysia for assessing and monitoring mangrove sediment quality. Thus, the current study intended to develop a mangrove sediment quality index (MSQi) in the Matang mangrove forest in Perak, Malaysia. An area was selected based on the distinct level of mangrove disturbances. At 1.5 m depth, sediments were sampled in ﬁve segments (0–15, 15–30, 30–50, 50–100, and 100–150 cm). All the sediment physicochemical properties were then analysed. Fourteen variables were chosen and included in MSQi. This index categorises mangrove sediment levels as I = Very Bad, II = Bad, III = Moderate, IV = Good, and V = Excellent. MSQi will be used as a guideline in monitoring mangrove sediment pollution. In conclusion, the data analysis showed that the Sepetang River (SR) was highly disturbed, followed by the Tinggi River (TR) (moderately disturbed), and the Tiram Laut River (TLR) (least disturbed).


Introduction
Mangroves are one of the most productive wetlands globally [1] and can be found in the intertidal zones along tropical and subtropical coastlines [2,3]. Mangroves are vital in providing breeding and nursery grounds [4] for commercially and recreationally important fish [5]. Mangroves also help to protect coastlines from erosion, storm damage, wave action [6], and tsunamis [7]. The mangrove ecosystem consists of several significant components, including forest, soil, and the marine ecosystem [8]. Mangrove sediments are complex and highly variable, composed of the river and marine alluvium, transported as sediment and deposited in rivers and seas [9].
Mangrove sediment is an abiotic matrix made up of residues, inorganic, and organic particles that is relatively heterogeneous in terms of physicochemical and biological characteristics [10]. Sediment is also vital in the heavy metals cycling in the coastal ecosystem [11,12]. It acts as a potential sink and source of inorganic and organic contaminants [3,13,14] during changes in environmental conditions [15][16][17][18]. In MMFR, working plans or management have been revised and implemented. The ten-year program provides detailed resources and schedules for harvesting, yield regulation, silvicultural operations, protection, and conservation. MMFR has been managed sustainably based on five work plans since Malaysia's Independence Day in 1957 [17]. However, the MMFR, with its large expanse of sheltered waters, is home to 7666 floating fish cages, and cockle culture covers an area of 4726 ha, both within and outside the estuaries [25]. Mangrove forest ecosystems provide productive and complex marine habitats for diversified marine life. There are 163 species of fish, 37 species of shrimps and prawns, and 45 species of crabs that have been identified and recorded in the Sixth Revision of the Working Plan. The following are the rivers' specific characteristics: TLR is located near the sea mouth at 4°52′30.30″ N and 100°38′8.04″ E ( Figure 2). The river's length is approximately 8.98 km. TLR is classified as least disturbed since most of this area was converted to open water, dryland forest, and waterways for fishing boats [17]. In MMFR, working plans or management have been revised and implemented. The ten-year program provides detailed resources and schedules for harvesting, yield regulation, silvicultural operations, protection, and conservation. MMFR has been managed sustainably based on five work plans since Malaysia's Independence Day in 1957 [17]. However, the MMFR, with its large expanse of sheltered waters, is home to 7666 floating fish cages, and cockle culture covers an area of 4726 ha, both within and outside the estuaries [25]. Mangrove forest ecosystems provide productive and complex marine habitats for diversified marine life. There are 163 species of fish, 37 species of shrimps and prawns, and 45 species of crabs that have been identified and recorded in the Sixth Revision of the Working Plan. The following are the rivers' specific characteristics: TLR is located near the sea mouth at 4 • 52 30.30" N and 100 • 38 8.04" E ( Figure 2). The river's length is approximately 8.98 km. TLR is classified as least disturbed since most of this area was converted to open water, dryland forest, and waterways for fishing boats [17]. TR is located near Kampung Pasir Hitam between 4°52′30.30″ N and 100°38′8.04″ E ( Figure 3). The river's length is ~8.1 km. Despite being closest to human development, this river is moderately disturbed due to minimal changes in mangrove land to water bodies, dryland forests, human development, agriculture, and aquaculture activities [17]. SR is near the Kuala Sepetang town, at latitude 4°52′30.30″ N and longitude 100°38′8.04″ E (Figure 4). The river's length is ~20.4 km. As observed during sampling activities, this river is highly disturbed due to its proximity to human settlements, agriculture, aquaculture, industrial operations, and a jetty. The land had been converted into oil palm plantations, horticulture, paddy fields, aquaculture, urban settlements, and dryland forests [17]. TR is located near Kampung Pasir Hitam between 4 • 52 30.30" N and 100 • 38 8.04" E ( Figure 3). The river's length is~8.1 km. Despite being closest to human development, this river is moderately disturbed due to minimal changes in mangrove land to water bodies, dryland forests, human development, agriculture, and aquaculture activities [17]. TR is located near Kampung Pasir Hitam between 4°52′30.30″ N and 100°38′8.04″ E ( Figure 3). The river's length is ~8.1 km. Despite being closest to human development, this river is moderately disturbed due to minimal changes in mangrove land to water bodies, dryland forests, human development, agriculture, and aquaculture activities [17]. SR is near the Kuala Sepetang town, at latitude 4°52′30.30″ N and longitude 100°38′8.04″ E (Figure 4). The river's length is ~20.4 km. As observed during sampling activities, this river is highly disturbed due to its proximity to human settlements, agriculture, aquaculture, industrial operations, and a jetty. The land had been converted into oil palm plantations, horticulture, paddy fields, aquaculture, urban settlements, and dryland forests [17]. SR is near the Kuala Sepetang town, at latitude 4 • 52 30.30" N and longitude 100 • 38 8.04" E ( Figure 4). The river's length is~20.4 km. As observed during sampling activities, this river is highly disturbed due to its proximity to human settlements, agriculture, aquaculture, industrial operations, and a jetty. The land had been converted into oil palm plantations, horticulture, paddy fields, aquaculture, urban settlements, and dryland forests [17].

Experimental Design
The soil sampling was conducted in three rivers using the normalised difference vegetation index (NDVI) at different levels of mangrove disturbance (green = least disturbed, yellow = moderately disturbed, and red = highly disturbed) ( Figure 1). A systematic sampling [26] was applied in this study, with three main plots of 450 m × 25 m established as the primary study plot from the landward, central, and seaward zones of each river. Each main plot contained five 5 m × 5 m subplots, with a distance of 100 m between each subplot. Five 1 m × 1 m mini subplots were established for sediment sampling ( Figure 5). GPSMAP ® 60CSx Garmin was used to record the sampling points.

Experimental Design
The soil sampling was conducted in three rivers using the normalised difference vegetation index (NDVI) at different levels of mangrove disturbance (green = least disturbed, yellow = moderately disturbed, and red = highly disturbed) ( Figure 1). A systematic sampling [26] was applied in this study, with three main plots of 450 m × 25 m established as the primary study plot from the landward, central, and seaward zones of each river. Each main plot contained five 5 m × 5 m subplots, with a distance of 100 m between each subplot. Five 1 m × 1 m mini subplots were established for sediment sampling ( Figure 5). GPSMAP ® 60CSx Garmin was used to record the sampling points.

Experimental Design
The soil sampling was conducted in three rivers using the normalised difference vegetation index (NDVI) at different levels of mangrove disturbance (green = least disturbed, yellow = moderately disturbed, and red = highly disturbed) ( Figure 1). A systematic sampling [26] was applied in this study, with three main plots of 450 m × 25 m established as the primary study plot from the landward, central, and seaward zones of each river. Each main plot contained five 5 m × 5 m subplots, with a distance of 100 m between each subplot. Five 1 m × 1 m mini subplots were established for sediment sampling ( Figure 5). GPSMAP ® 60CSx Garmin was used to record the sampling points.

Sediment Sampling and Laboratory Analysis
Seven hundred and fifty sediment samples were collected from five mini subplots along the same transect. The sediment samples were taken using a peat auger [27,28] in two seasons: November and December 2017 (wet season) and March and April 2018 (dry season). This study obtained a total of 2250 sediment samples at five depths, i.e., 0-15, 15-30, 30-50, 50-100, and 100-150 cm, because sediment depths can also influence pollution [2,28]. The sediment samples were placed into a labelled plastic bag before being transported to the soil laboratory for analysis.
Sediment samples were characterised for physical and chemical properties. Sediment texture was determined using the hydrometer method [29,30]. Sediment pH was measured in a 1:2.5 ratio (sediment: distilled water) using an electrode pH meter (Model MW 100, Milwaukee, Italy) [31,32]. Total Nitrogen was analysed using the Kjeldahl method [26,33]. Phosphorus was determined using the blue method and a double acid method [26,32,34]. Subsequently, samples were examined using an ultraviolet-visible (UV/Vis) spectrophotometer with a specific wavelength (Model Cary 50 Scan UV/V Spectrophotometer) [35]. Aqua regia method was used to extract and digest the sediments [32,36]. Finally, samples were analysed for heavy metals and base cations using an atomic absorption spectrophotometer (AAS, Model Shimadzu AA-6800) with specific flame and wavelength settings.

Statistical Analysis
The data were analysed using a statistical analysis system (SAS) software version 9.4 for descriptive analysis. The statistical package for the social sciences (SPSS) version 25 was used for principal component analysis (PCA) to reduce the covariance and correlation matrix [37] and identify important MSQi parameters. Microsoft Excel 2020 was used to create the mangrove sediment degree of pollution table (MSDPT) and MSQi formulation model.
PCA was performed on all measured sediment variables to determine variables with the highest score, more significant than 0.75. PCA, when combined with a coefficient of linear correlation, provides a multi-dimensional statistical test of the studied variables [38]. PCA is widely used in various sediment fields, including sediment assessment [2,21,39]. The most significant variables were determined using PCA and are characterised by the highest score component of each principal component (PC). Each PC is derived from a linear combination of the p metrics. The first component has been extracted and accounts for the second-largest amount of variance that remains after associated with the first extracted component. The second extracted accounts for the second-largest amount of variance.

Development of MSQi
PCA was conducted on all measured sediment parameters to develop the index using the steps outlined below: STEP 1: Fourteen of the nineteen parameters were chosen for PCA analysis because they were grouped in one unit (mg/kg) to reduce bias in PCA analysis. The fourteen parameters are Nitrogen (N), Phosphorus (P), Potassium (K), Cal-cium (Ca), Magnesium (Mg), Sodium (Na), Manganese (Mn), Iron (Fe), Lead (Pb), Zinc (Zn), Copper (Cu), Cadmium (Cd), Chromium (Cr), and Nickel (Ni). STEP 2: Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy with KMO value >0.600 was tested to ensure that the relationship between the data in the observation is adequate. If the KMO value < 0.600, the data are insufficient to run the PCA [40]. STEP 3: Essential parameters in the component were identified, where the highest score of the PC grouped these parameters. STEP 4: In PCA, two results were obtained: component matrix and rotated component matrix. The proportion of the variability explained by each important component was identified. For example, four crucial parameter factors (PFs) were selected: P(Xi), P(Xii), P(Xiii), and P(Xiv).
Forests 2021, 12, 1279 7 of 21 STEP 5: Each suggested critical PF concentration in the sediment was rated as 0 (low), 1 (medium), and 2 (high), based on the permissible limit range for soil and plant [41], as presented in Table A1 (Appendix A). STEP 6: The PFs ratings were referred to the MSDPT (see Figure A1, Appendix A). STEP 7: These MSDPT was developed by summation of these parameters. Σ MSDPT = P(X i ) + P(X ii ) + P(X iii ) + P(X iv ). STEP 8: MSQi was developed by classifying MSDPT as I (very bad), II (bad), III (moderate), IV (good), and V (excellent) (see Table A2, Appendix A). STEP 9: A simplified format has been developed to facilitate MSQi modelling (see Table A3, Appendix A).

Development of MSQi
Since PCA was used to analyse fourteen parameters, only the highest four were extracted to strengthen MSQi development. The KMO test was performed; the results that were significant at <0.001 with adequacy of 0.747 (Table 1) were used in the PCA analysis [42]. MSQi development involved PCA to interpret sediment chemical composition and calculate the pollution score. PCA's primary function is to reduce the complexity of the loading factors [37]. PCA also works as an indicator of anthropogenic sediment pollution. Component loading greater than 0.750 indicates "strong", values between 0.500 to 0.750 indicate "moderate", and values between 0.500 to 0.000 indicate "weak". Only the "strong" values were taken in MSQi modelling. In this study, at least 200 data points were selected to run the PCA [37]. Other missing chemical parameters should be included to improve the PCA loading value in the MSQi formulation. Tables 2 and 3 presents the PCA results.     Figure 6 shows the component plot in rotated space. The red circle represents a strong correlation between all parameters.   [38]. Permissible limit PF values were divided into three categories: low, medium, and high concentrations using ratings 0, 1, and 2.   Table 4 shows the PF. The World Health Organisation (WHO) guidelines for soil and plants were used to calculate the PF values [38]. Permissible limit PF values were divided into three categories: low, medium, and high concentrations using ratings 0, 1, and 2.  Table 5 shows an example simulation using a random MSDPT table ( Figure A1) (Appendix A). The obtained code of MSDPT was 1 2 0 2. This score value was used in Table 5 to calculate the MSQi class. When the MSDPT was obtained, the value was then referred to the rating of MSQi Index at Table 6. Our result showed that the MSDPT score is 5 (Table 6) and falls into Class II ( Table 6), indicating that the description of the areas is highly polluted.

MSQi of TLR during Dry and Wet Seasons
The MSQi of TLR during the dry season is illustrated in Table A4 (Appendix A). The result showed that the MSQi of the landward zone at a sediment depth of 0-15 cm was in Class III, followed by a moderate rating, indicating that the sediment was moderately polluted. However, MSQi of sediment depths of 15-30, 30-50, 50-100, and 100-150 cm was obtained under Class IV, indicating a good rating. In central and seaward zones of the TLR, MSQi in all sediment depths was obtained under MSQi Class IV with a good rating, indicating that the sediment in this area was less polluted. During the wet season, MSQi at landward, central, and seaward zones changed to Class IV: less polluted (Table A5 Appendix A) due to heavy metal (HMs) content dilution [43] with flooded river water [39,44]. In TLR, all the heavy metal content (Pb, Zn, Cr, and Ni) during the wet season at all mangrove zones and sediment depths were below the permissible limit of WHO guidelines for soil and plants [41]. Therefore, mangrove sediment pollution in this river was at lower risk than in the surrounding area.

MSQi of TR during Dry and Wet Seasons
The MSQi of TR during the dry season is illustrated in Table A6 (Appendix A). The results showed that MSQi in landward and central zones in all sediment depths were obtained with Class III and a moderate rating. However, MSQi at seaward in sediment depths of 30-50 and 50-100 cm was obtained under MSQi Class IV with less pollution. Meanwhile, Table A7 (Appendix A) depicts MSQi during the wet season. The results revealed that MSQi in a landward zone at all sediment depths was obtained with a III score, indicating that the area was moderately polluted. In contrast, sediment depths of 30-50 cm were discovered in Class IV with less polluted conditions. The same result was obtained in the central and seaward of TR, where sediment in-depths of 0-15 and 15-30 cm were obtained under MSQi Class III, and the rating was moderate. However, under MSQi Class IV, sediment depths of 30-50, 50-100, and 100-150 cm were obtained, and the rating was good.

MSQi of SR during Dry and Wet Seasons
The MSQi of SR during the dry season is illustrated in

Discussion
TLR has been classified as least disturbed because most of this area has been converted to open waterand dryland forest [17]. This river also serves as a waterway for fishing boats from Kuala Sepetang Jetty, Kuala Trong Jetty, and Kg. Pasir Hitam Jetty. As this area was categorised as low polluted, the source of heavy metals from landward and seaward during the dry and wet seasons was runoff [39]. The top layer of sediment contained high metal concentrations due to pollutant migration from the landward zone [45]. This pollution migration is influenced by many factors, such as rainfall, tidal, sediment type, the porosity of sediment, type of vegetation cover, and others [43]. Moreover, sediment contaminations with Pb, Zn, Cr, and Ni are common in many environments. For example, lead came from the historical use of leaded fuels, zinc from galvanised steel, and increasing copper content from the passive leaching of antifouling paints [15,35].
From this study, the findings showed that pollution indeed had an impact on TR. Due to land changes into the water body, dryland forest, human settlement, agriculture, and aquaculture practices, this river is classified as moderately disturbed [14]. Furthermore, as a waterway, this river is opened to fishermen's boats. Overall, the Pb, Cu, Zn, and Ni content of TR during the dry season in all mangrove zones and sediment depths were below the WHO guidelines for soil and plants [41]. Therefore, mangrove sediment pollution in TR is at low risk than in the surrounding communities.
Meanwhile, SR is highly disturbed due to its proximity to human settlements, agriculture, aquaculture, industrial operations, and a jetty. SR had been converted to oil palm, horticulture, paddy field, aquaculture, urban settlement, and dryland [17]. Hence, the tidal process increased most of these metals content during dry seasons. During high tide, river water flows from seaward to landward and vice versa during low tide [43]. Under these circumstances, pollution remains suspended in sediment in both directions (seaward and landward).

Conclusions
During the dry season, TLR was moderately polluted (Class III). However, the MSQi became less polluted during the wet season (Class IV). On the other hand, TR was classified as Class III (moderately polluted) during the dry season, except at seaward at sediment depths of 30-50 and 50-100 cm (Class IV: less polluted). During the wet season, the TR was moderately polluted to less polluted. Lastly, during the dry season, the SR was classified as Class III with moderate pollution, while less polluted during the wet season. Therefore, based on the findings of this study, it can be concluded that sediment depths have an impact on pollution. The MSQi development will serve as an essential benchmark and guideline for assessing sediment pollution in Malaysia's mangrove ecosystem. The application of MSQi will reduce time and cost in monitoring the mangrove sediment quality compared with current practices.

Conflicts of Interest:
The authors declare no conflict of interest.