Distribution and speciation of Fe, Mn, Zn, Cu, Pb and P in surface sediments of Lake Mariut, Egypt

Lake Mariut is a shallow brackish water body with no direct connection to the sea. The lake is located at the longitude of 29º56’E and 29º51’E and the latitude of 31º10’30”N and 31º3’30”N (Fig. 1). The feeding waters of the lake are agricultural waters from El-Umoum Drain, sewage and industrial wastewater from El-Qalaa Drain and fresh water from the Nubaria Canal. The lake is heavily polluted by industrial wastes, in addition to raw and treated sewage from point and nonpoint sources such as agricultural runoff containing fertilizers and pesticides through these drains (Donia 2016). The lake is divided into four basins (Fig. 1): the Main Basin (14.77 km²), the Aquaculture Basin (9.44 km²), the West Basin (11.59 km²) and the South Basin (33.77 km²). The four basins are interconnected with each other by several breaches in the dikes of El-Umoum Drain and the Nubaria Canal. The central portions of the Main Basin receive most of their water from El-Qalaa Drain, in addition to water from El-Umoum Drain via several fractures in its embankment. Lake Mariut surface is maintained at 2.8 m below mean sea level by pumping water from the lake (West Basin) to the sea through El-Max Pumping Station (Khalil et al. 2013). Properties of the water in Lake Mariut are presented in Table 1.

Figure 1

Table 1

Physical and chemical characteristics

Cited from El-Shabrawy and Goher (2011)

Ten sampling sites in Lake Mariut were selected for sediment collection from the top 20 cm layer using an Eckman Dredge device during winter 2013 (Fig. 1). The sites were distributed as follows: 2 sites in the Aquaculture Basin (AQ), 3 sites in the Main Basin (MB), 2 sites in the West Basin (WB) and 3 sites in the South Basin (SB). Samples were air-dried and passed through a 2-mm sieve to remove plant fragments and stones. The sieved samples were powdered and analyzed for P, Fe, Mn, Zn, Cu and Pb fractions, in addition to the total metal concentration. The summary of sieve test results as well as the content of organic matter (OM) and carbonates for Lake Mariut sediment samples are presented in Table 2.

Table 2

Results of grain size

Personal communication with Dr. Hassan Farahat (NIOF)

Extraction of 1 gm of each sediment sample was made through sequential steps (Tessier et al. 1979; Abdel-Satar & Goher 2015) as shown in Table 3. For the determination of Fe, Mn, Zn, Cu and Pb in the total content of metals and fractions, an atomic absorption spectrometer GBC (model SavantAA AAS) with a graphite furnace GF 5000 was used.

Table 3

Sequential extraction scheme for heavy metals

Sediment P fractions were characterized using a modification of the sequential fractionation procedure (Paludan & Jensen 1995). Phosphorus fractions were identified by this procedure based on the extractant. To estimate loosely sorbed P (Lsor-P), the sediment sample was shaken in 1 M NH₄Cl. Bicarbonate dithionite (BD) was used to extract P adsorbed to oxidized Fe and Mn (Fe-P) in the remaining soil pellets. To estimate P associated with clay and Al oxides (Al-P), remaining soil pellets were shaken with 0.1 M NaOH. The extracts were acidified to pH 1 and filtered to isolate P associated with humic acids (HA-P). To extract Ca-bound P (Ca-P), remaining soil pellets were shaken with 1 M HCl. In each step, soil residues were washed twice by the extractant and then by deionized water, where P was analyzed in the combined three extracts. Refractory organic P (Res-P) was estimated as P remaining in each soil pellet following the HCl-P extraction. Pellets were dried and subsamples were combusted at 550^oC in a muffle furnace, and Res-P was extracted with 1 M HCl in a block digester. The inorganic P (Inorg-P) content of each extract was colorimetrically determined as orthophosphate by the procedure of Murphy and Riley (1962). Lsor-P, Fe-P, and Al-P extracts were subsequently digested, and organic P (Org-P) content in these extracts was calculated as a difference between Inorg-P concentrations of digested and non-digested extracts. For Res-P and HA-P extracts, all P was considered to be Org-P, but for Ca-P extracts, all P was considered to be Inorg-P. Total P was calculated by summing the content of Inorg-P and Org-P.

The pollution phosphorus index was calculated as follows (Wang et al. 2013):

Si=CiCs $${S_i} = {{{C_i}} \over {{C_s}}}$$

where S_i is the pollution index, C_i is the phosphorus concentration of sample i (μg g^-1), and C_s is the phosphorus standard value (μg g^-1). Wang et al. (2013) considered 600 μg g^-1 as the phosphorus standard value which causes the minimum ecological risk. When S_i is more than 1, the phosphorus concentration exceeds the standard.

The precision of phosphorus and heavy metal analysis was controlled by using triplicate samples for each analysis. The obtained data were recorded as the average with relative standard deviations less than 5%, which was considered as a satisfactory precision. The correlation coefficients between the studied variables in the sediment samples were calculated. Also all sediment sampling data for P and metals were tested for significant differences between different sites by means of one-way ANOVA.

The distribution trends of the total heavy metals are shown in Table 4. Iron has the highest mean concentration at all sampling sites and was followed by Mn. The total metal (TM) concentrations at the investigated sites varied in a narrow range of values (Table 4).

Table 4

Distribution of heavy metals in Lake Mariut sediment

Although the lake is exposed to different pollution sources (agricultural, sewage and industrial effluents), the total Zn, Cu and Pb concentrations in Mariut sediments were mostly below the guidelines cited by the Canadian Council of Ministers of the Environment (CCME 2001) for the protection of aquatic life (Tables 4, 5). The prevalence of sand fractions in Lake Mariut sediment may be responsible for the decrease in the total concentration of metals (Table 2), where the greater the size of the sediment fraction, the smaller the amount of metals bound to this fraction due to the reduction in specific surface area of the larger fractions (Ong et al. 2013).

Table 5

Canadian Sediment Quality Guidelines (μg g^-1) for the protection of aquatic life (CCME, 2001)

Significant positive correlations at P < 0.05 between each of the total Zn, Cu and Pb concentrations and OM content (r = 0.68, 0.67 and 0.61, respectively, n = 10) were often related to the complexation and adsorption of metals by OM, indicating that the interaction between metals and OM is an important process for the fixation and removal of metals in Lake Mariut sediment (Kljaković-Gašpić et al. 2009). With respect to the role of carbonate content in the dispersion of metals, no trend was observed between the carbonate content and the concentrations of all the studied metals in the lake sediment.

The result of the speciation analysis of the metals is presented in Table 4. The summation of the five fractions is consistent with the total metal concentrations. High percentage recoveries were recorded for metals (84.9-104.9%), which proved that the sample preparation method and the analytical procedure described in this study were satisfactory (Table 4). The recovery of the sequential extraction procedure was computed as follows (Abdel-Satar & Goher 2015):

Recovery=F1+F2+F3+F4+F5TM×100 $$Recovery = {{F1 + F2 + F3 + F4 + F5} \over {TM}} \times 100$$

Cluster analysis can be used as an important tool for analyzing heavy metal data to understand the relationship between sampling sites. The cluster diagram shows two clusters (Fig. 2). The first cluster A is formed by WBN and MBE sites with the similarity index of 96.7%, while the cluster B has two subclusters containing the remaining sites with similarity indices ranging from 97.9 to 99.5%. These two clusters are accounted for by the prevalence of different environmental conditions (Rita et al. 2012).

Figure 2

The correlation coefficient matrix summarizes the strength of linear relationships between each pair of variables. Cu-Zn, Cu-Pb and Pb-Zn pairs are significantly positively correlated at P < 0.05 with each other (r = 0.75, 0.85 and 0.73, respectively; n = 10), which may suggest common pollution sources or similar mechanisms of transport and accumulation within the sediments (Ong et al. 2013). Furthermore, no significant differences (P > 0.05) in the concentrations of metal fractions were observed at the ten sites in Lake Mariut, despite the differences in sediment environment.

The potential environmental risk of trace elements in sediments is associated with both their total content and their speciation. The amount of metals coming from the anthropogenic input, the prevailing chemistry of the sediment-water interface and the species of metals represent serious factors to be considered when accounting for the levels of metals in a fraction (Ogunfowokan et al. 2013). Fe, Cu and Pb in Lake Mariut sediment were present mainly in the residual phase, which implies that these elements were strongly linked to the inert fraction of the sediments. The order of abundance of the Fe fractions was as follows: residual (45%) > Fe-Mn oxides (31%) > organic fraction (22%) > carbonate (1.7%) > exchangeable (0.8%). Residual phases of metals are generally much less toxic to organisms in the aquatic environment (Ogunfowokan et al. 2013; Moore et al. 2015; Abdel-Satar & Goher 2015). The residual fraction was the dominant Cu host in all analyzed samples and this phase carried approximately 63% of the total Cu concentration.

Like Fe and Cu, the average percentage of Pb in Lake Mariut sediments was bound mainly to the residual fraction (52.1%) and the fraction of carbonates (18.0%), and in a smaller amount – to hydrated oxides of iron and manganese (10.5%), organic matter (10.3%) and the exchangeable form (9.0%). Metals bound to different fractions will behave differently in the sedimentary environment, and thus will have different potentials for remobilization and uptake by biota (Fytianos & Lourantou 2004).

The high percentage of Zn in the non-residual fractions (≈65%) may be indicative of high cation exchange capacity of the sediment samples (Ogunfowokan et al. 2013). Zn is mainly associated with the reducible phase (F3 = 32.7%) due to the strong binding of Zn with hydrous Fe-Mn oxides forming stable complexes (Moore et al. 2015). The low content of Zn in the residual fraction was likely due to the low association or retention ability of the mineral’s crystal structure, such as with detrital silicates and resistant sulfides (Ogunfowokan et al. 2013).

Like Zn, sediment Mn was found abundantly in the first three fractions (F1+F2+F3 > 64%), which means that Lake Mariut has a Mn potential risk. The increases in exchangeable Mn species mostly resulted from anthropogenic activities (Akcay et al. 2003). Moreover, Mn was rather poorly absorbed from organic matter (F4 = 13.4%), so the bioavailability of Mn in organic matter of sediments was low (Ogunfowokan et al. 2013). The predominance of Mn in non-residual fractions of the samples from Lake Mariut was similar to that obtained from the East China Sea (≈85.5%) (Yuan et al. 2004) and Daya Bay (≈69.7%) (Gao et al. 2010).

A relatively low percentage of metals was bound to organic matter (average = 21.9%, 13.4%, 7.6%, 9.9% and 10.3% for Fe, Mn, Zn, Cu and Pb, respectively) in the sediments of Lake Mariut reflecting either a low percentage of organic matter itself or low retention capacity of organic matter (Abdel-Satar & Goher 2015). On the whole, organic matter and sulfides are important factors that control the bioavailability of heavy metals, but the low percentage of heavy metals in the oxidizable fraction suggested that organic matter is not the main factor affecting the behavior of metals in the sediment of Lake Mariut (Moore et al. 2015).

Phosphorus released from the sediments is the most important factor determining the P content in the overlying water (Abdel-Satar & Goher 2009). Such release may affect the water quality and may result in continuing eutrophication of the aquatic system (Aydin et al. 2009). The current results show that the West Basin in Lake Mariut, which collects water from different basins and discharges it into the sea, and the Main Basin that receives effluents from El-Umoum and El-Qalaa Drains (924 μg g^-1) recorded high average total P concentrations (1074 μg g^-1), while the South Basin showed the lowest values (693 μg g^-1). This suggests that different sediments had different P content due to their different drainage basins and pollution sources, where phosphorus gets into the marine environment from diffuse sources (particularly agriculture) and point sources (treated and untreated sewage effluents) showing significant differences (Aydin et al. 2009).

The sediment of Lake Mariut showed the decrease in Inorg-P percentage (48.6%) (Fig. 5) compared to other Coastal Delta lakes such as Manzalah that contains more than 55% of phosphorus originating from the inorganic form (Abdel-Satar & Goher 2009). This is consistent with the richness of phytoplankton in Lake Manzalah compared to Lake Mariut (Khairy et al. 2015), where dissolved inorganic phosphate is the major form absorbed by aquatic primary producers (aquatic macrophytes and phytoplankton) (Abdel-Satar & Sayed 2010).

Figure 5

NH₄Cl-P, BD-P, NaOH-P, HCl-P and Res-P content in the surface sediments of Lake Mariut are presented in Fig. 6. Phosphorus fractions showed different distribution patterns, where HCl-P contributed the highest fraction to the total sedimentary P. The HCl-P fraction was deemed as a relatively stable fraction of Inorg-P in the sediments and consists mainly of apatite P (natural and detritus), including P bound to carbonates (Abdel-Satar & Sayed 2010). The relative abundance of phosphorus forms followed the order:

HCL-P (31.6%) > NaOH-P (28.6%) > BD-P (21.1%) > NH4CI-P (12.7%) > Res-P (6.5%) $${\rm{HCL - P (31}}{\rm{.6\% ) > NaOH - P (28}}{\rm{.6\% ) > BD - P (21}}{\rm{.1\% ) > N}}{{\rm{H}}_4}{\rm{CI - P }}(12.7\% ){\rm{ > Res - P (6}}{\rm{.5\% )}}$$

Figure 6

where the South Basin recorded the highest NH₄Cl-P percentage (17.7%) and the lowest Res-P percentage (4.0%). No significant differences (P > 0.05) in the content of the sediment P fraction were observed between the studied sites. The distribution of HCl-P, NaOH-P, BD-P, NH₄Cl-P in the studied lake was similar to the rank observed in lakes Volvi and Koronia, Greece (Fytianos & Kotzakioti 2005), and totally different from those observed in Egypt for lakes Manzalah (Abdel-Satar & Goher 2009), Qarun and Wadi El-Rayan (Abdel-Satar & Sayed 2010).

NaOH-P is exchangeable, including P bound to metal oxides (mainly Fe and Al), and is an indicator of algal-available P (Abdel-Satar & Goher 2009). This fraction recorded a relatively high percentage of total sedimentary P (28.6%). It can be released for the growth of phytoplankton when anoxic conditions prevail at the water-sediment interface (Abdel-Satar & Sayed 2010). It is likely that the depletion of DO occurred in the water of the studied lake (Table 1), which resulted in the high contribution of NaOH-P to the overlying water. Compared to the other Coastal Delta lakes such as Lake Manzalah, NaOH-P dominated (48.4%) in the total sedimentary P. The significant positive correlation (r = 0.96, n = 10) between the total amounts of P released from Mariut sediments and those in the NaOH fraction indicated that NaOH-P was the main fraction that can release P easily.

The contribution of redox-sensitive P (BD-P) that includes P bound to Fe/Mn hydroxides to the total P was relatively low in comparison to HCl-P and NaOH-P in the studied lake. This result may be associated with the increase in organic matter, which inhibits the formation of BD-P by competitive bonding with Fe (Zhang et al. 2013). Also the organic matter mineralization resulted in pH changes, thus further affecting the P release (Wang et al. 2008). An increase in the pollution of lakes results in a significant decrease in DO and an increase in pH of the overlying water, which leads to a significant increase in the release of NaOH-P and/or BD-P from the sediments (Zhang et al. 2013). The Fe:P ratio is an indicator of P release from the sediments in shallow lakes, and it may be possible to control the internal P loading by keeping the surface sediment oxidized if the weight-based Fe:P ratio exceeds 15 (Zhang et al. 2013). The Fe:P ratio of Lake Mariut sediment was much lower than 15 (0.48-1.15), indicating that Fe cannot regulate the P release and the contribution of NaOH-P and BD-P to the P content in the overlying-water was high.

NH₄Cl-P content in the sediments was the lowest among the phosphorus fractions, accounting for approximately 13% of the total P. It might be released from calcium carbonate associated P or from decaying cells of bacterial biomass in deposited phytodetrital aggregates (Wang et al. 2013).

The Org-P fraction dominates in sedimentary P (average = 51.4%; Fig. 5). It is one of the phosphorus fractions, which is generally retained in sediment during mineralization (Topcu & Pulatsü 2008). The surface sediment from Lake Mariut contained 17.3% of Org-P as exchangeable P, in addition to 73.92 μg g^-1 of humic substances-bound P (average 16.6% of total Org-P). There is a significant positive correlation (r = 0.75, n = 10) between P in the humic fraction with both total P and Org-P.

Ca-P substantially contributed to the supply of Inorg-P in sediments of Lake Mariut (68.8%; Fig. 7). The average of Ca-P and Al-P accounted for 31.6% and 10.3% of the total P, respectively. Available-P and Fe-P were the lowest in the phosphorus fractions, which varied between 10.49 to 56.29 μg g^-1, and less than 4% in the total P content. Generally, available-P, Fe-P, Al-P and part of Org-P are easily desorbed from the lake sediments and released into the overlying water. These are referred to as the bioavailable fraction of P in the lake sediment (Wang et al. 2013).

Figure 7

The correlation between various phosphorus fractions is presented in Table 8. Highly positive significant correlation between total P content and Inorg-P, Org-P and Al-P content (P < 0.01) was determined. As one of the inorganic phosphorus fractions, Al-P and Ca-P content was also closely related to Inorg-P content (P < 0.01).

Table 8

Correlations between different phosphorus fractions in sediments

Pollution P indices at Lake Mariut sampling sites were higher than 1, except site SBS (0.88), while the maximum value was recorded at site WBN (1.90), indicating that Lake Mariut sediments were at a potential phosphorus risk, especially in the West Basin.