The Concentration and Spatial Distribution of Mercury, Lead, and Cadmium in Surface Sediments of Mangrove Forests Using Geostatistics in GIS Environment

Document Type : Research Paper


1 MSc. Environment, Department of Environment, Faculty of Natural Resources, Tarbiat Modares University Tehran, Iran

2 Associate Professor, Department of Environment, Faculty of Natural Resources, Tarbiat Modares University, Tehran, Iran

3 Assistant Professor, Department of GIS & RS, Faculty of Environment and Energy, Islamic Azad University of Tehran, Iran

4 Assistant Professor, Department of Environment, Faculty of Natural Resources, Tarbiat Modares University, Tehran, Iran


Environmental pollutants are considered among the factors disturbing natural ecosystems. Among them, heavy metals due to their toxic effects and having a high bio-accumulation are known as one of the most dangerous pollutants. This leads to the concentration of these metals in the food chain, in the top of the pyramid. Many of these metals are natural components of aquatic ecosystems and some of them play a critical role in the survival of living organisms. However, if the concentration of heavy metals exceeds a certain limits, aquatic life will be threatened and ecosystem degradation will be occurred. Heavy metals have not the ability to be refined in aquatic ecosystems, so with gradual blurring of these ecosystems, they can easily accumulate in sediments. In fact, it can be said that the marine sediments are considered often as a final repository for the accumulation of metals. So, we can say that the sediments are considered as an important indicator of pollution and it is used to estimate the amount of pollution in the environment, especially in aquatic ecosystems. The spatial distribution of toxic metals in marine sediment in the explanation on the contamination history of aquatic ecosystems and location of pollution sources is very important and effective. Spatial and visual assessment of the pollutants is important for a better understanding of the threats of pollution sources. This can be achieved by Geographical Information System (GIS) techniques. A GIS-based and geostatistical approach provides the possibility of the spatial datasets processing. Using of the geostatistical principles, in addition to describing the spatial pattern of the observed data, provides the possibility of creating contamination maps with minimum variance. It is believed that the spatial assessment and visualization of the pollutants is essential to better understanding of threats of the pollution sources. For this reason it is recommended to use GIS techniques in the studies about the distribution of pollutants in the environment.
Persian Gulf supply major part of the global oil and gasenergy and is known as a potentially oil contaminated ecosystem. Since there is only a narrow exchangeable path between the Persian Gulf and Oman Sea, it takes a very long time to transfer water in the entire Persian Gulf into open seas. One of the most important and sensitive ecosystems in the Persian Gulf is mangrove forests. Mangrove ecosystems have many environmental and social-economical functions. Therefore, metal contaminants entering to Persian Gulf through the exploration of oil and tanker shipping, due to the high toxicity and bioavailability, are one of the main concerns for these ecosystems. The purpose of this study is an investigation on metals of lead, mercury and cadmium concentrations in surface sediments of mangrove forests. This is in order for getting a correct spatial distribution pattern of these pollutants in this very sensitive ecosystem. The spatial distribution modeling involves using GIS and principles of geostatistics.
Materials and methods
Study area of this research is the mangrove forests in the Hormozgan province (in 55˚33'42" to 55˚47'23" E and 26˚46'21" to 26˚58'49" N). Sampling was conducted from 42 stations in three parts of Qeshm Island, Khamir Port, and middle part from the surface sediments 0-5c min, March 2010. Geographic locations and characteristics of each station were recorded and finally samples were transported to the laboratory in boxes of ice-containing. To determine the concentrations of lead and cadmium after preparation and digestion of the samples, they were analyzed by graphite furnace atomic absorption spectrophotometer Model AA-67OG. To determine the concentrations of mercury, the samples were also placed into the freeze dryer at -63°C for 48 hours. They were crushed to be prepared for analysis. In order to measure the total mercury, the 0.03 to 0.05 g of each sample was placed directly in the Advanced Mercury Analyzer Model AMA254. This is designed specifically for determination of mercury concentration in liquid and solid samples. To determine the level of contamination in surface sediments of mangrove forests, the mean values of the calculated concentrations for metals were compared with the NOAA and SQGs standards. In fact, these standards represent concentrations of pollutants that in the lower values of them, biological effects are rarely observed. On the other hand, if the concentration of pollutants is greater than this amounts, many incompatible biological effects will be occurred.
     In order to model the spatial variations of toxic metals of mercury, lead and cadmium in surface sediments of mangrove forests, seven of different geostatistical methods were used. These methods were: Radial Basis Functions, Local Polynomial Interpolation, Global Polynomial Interpolation, Inverse Distance Weighting, Simple Kriging, Universal Kriging, and Ordinary Kriging. The cross-validation method was used to compare the methods used in this study and selection of the most appropriate geostatistical method. For performance evaluation of geostatistical methods, the statistical parameters such as Mean Bias Error (MBE), Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) were examined.
Results and discussion
The results showed that the highest concentrations for the metals lead and cadmium were in Qeshm Island and the highest concentration of mercury was found in the Kamir port. However, Duncan's test results at 1% significance level were indicative of the lack of significant differences between the estimated concentrations for each metal in the three investigated parts. The mean concentrations of the metals lead and cadmium in the mangrove forests were obtained 1.86 and 0.21 µg/g, respectively, and that of mercury obtained 8/04 ng/g. In order to understand and evaluate the contamination of surface sediments of mangrove forests in these metals, specified concentrations for them were compared with NOAA and SQGs guidelines and the results showed a significant difference for all metals in the study area (P<0.05) and the concentrations were lower than these standards. In order to make a modeling the spatial variations in the studied variables, before any calculations, test of the normality was conducted with Kolmogorov-Smirnov test on data sets. The results of the tests showed that mercury and lead follows a normal distribution trend. Cadmium values were transformed to their logarithms to obey normal distribution.
     Finally, after the implementation of geostatistical methods on a dataset, the simple kriging because of having the lowest values for RMSE and MAE and nearing the Mean Bias Error (MBE) parameter to zero was selected as the best approach for modeling the spatial variations of the variables. Superiority of kriging methods can be due to the nature of these methods in minimizing variance of the estimation and use of the variogram techniques in the modeling the spatial distribution of the pollutants. The best experimental variogram was plotted for each of the metals after fitting and checking of various models. The spatial structures of these metals were studied and the results showed that the spatial distribution of mercury, lead and cadmium are following the gaussian, exponential and circular models, respectively.
Although concentration of the metals in mangrove forests in terms of toxicity for organisms is not threatening, but this is one of the most serious dangers that is threatening the mangrove ecosystems in Persian Gulf. The sediments of these forests are low-oxygen and water-saturated that takes a time to clean up oil pollution in these areas. This makes the oil floats on water that may cause death of the tree roots in mangrove forests. Finally, the structures and relationships in the mangrove forests may be broken down.
     But as laboratory analysis of sediment samples in this study showed, fortunately a number of natural factors reduce the accumulation of toxic pollutants in mangrove forests. First, the flow is counterclockwise in the Persian Gulf. This can convey clean waters into Persian Gulf from northern parts of the Qeshm Island. After crossing the northwest and western parts of the Persian Gulf, where there are the highest concentration of oil wells and oil exploration projects, it gets out from southern parts of the island. Therefore, it is expected that the sediments of northern parts of Qeshm have less pollution in comparison with those of southern parts. This is also confirmed by the results of previous studies. It is also noteworthy that the issue of the mangrove forest floor is full of microorganisms that can degrade petroleum compounds into simpler substances. This can be revealed by the toxic metals found in crude oil before accumulation in surface sediments, by marine counterclockwise currents driven toward the inner parts of the Persian Gulf,i.e., where they deposited. On the other, Qeshm Island is located in such a way that serves as a protective barrier to protect physically a large portion of the mangrove forests against contaminants.
In this study, due to the lack of extensive sampling of the total mangrove forest areas, we used of geostatistical methods to model the distribution pattern of toxic metals. In general, we can say that employing human development in data processing techniques and using remote sensing and GIS, mapping of pollutions in environment (water, soil and air) is easier and faster and that geostatistical methods are very useful due to spatial extent and the problems associated with sampling.


Main Subjects