Environ Dev Sustain DOI 10.1007/s10668-016-9819-2

Investigating soils retention ratios and modelling geochemical factors affecting heavy metals retention in soils in a tropical urban watershed Ce´lestin Defo1,2,3 • Bernard Palmer Kfuban Yerima2 Nestor Bemmo4

•

Received: 3 February 2016 / Accepted: 1 June 2016  Springer Science+Business Media Dordrecht 2016

Abstract This study aimed at investigating the retention of Pb and Cd in soils and the geochemical factors influencing the adsorption of these pollutants. Soil samples were airdried and ground to pass through a 2-mm sieve, and different soil extracts were prepared for chemical analysis (organic matter, cation exchange capacity and pH). Total Pb and Cd were extracted with diacid using digestion method and determined by atomic adsorption spectrophotometer (AAS) after filtration. Results revealed that the heavy metals retention ratio (RR) of the Rhodic ferralsol, Xanthic ferralsol and Mollic gleysol (2) were very high for Cd ([80 %) and was relatively low (generally \ 60 %) for Pb. In contrast, RRs for the Plinthic gleysol and the Mollic gleysol (1) were relatively low (\60 %), regardless of the heavy metal concerned. Multiple regression equations indicated for Pb and Cd concentrations different linear relationships over simple linear regression, when pH, organic matter, clay percentage and cation exchange capacity (CEC) were used as independent variables. Results indicate that organic matter exerts major influences on the retention of Pb and Cd in soils, while CEC, clay content and pH have a minor influence in this process in the Ntem watershed. From these observations, the application of soil organic matter could be a solution in protecting shallow aquifers from heavy metal pollution and thus insuring that they are not a hazard to public health.

& Ce´lestin Defo [email protected] Bernard Palmer Kfuban Yerima [email protected] 1

Water Resources Management Laboratory, Department of Agricultural Engineering, University of Dschang, PO Box 222, Dschang, Cameroon

2

Faculty of Agronomy and Agricultural Sciences, University of Dschang, PO Box 222, Dschang, Cameroon

3

Laboratory of Water and Food Quality, Water Technology Centre, Indian Agricultural Research Institute, New Delhi 110012, India

4

National Advanced School of Engineering, PO Box 8390, Yaounde´, Cameroon

123

C. Defo et al.

Keywords Regression analysis  Heavy metals  Urban soils pollution  Principal component analysis  Cluster analysis

1 Introduction Heavy metal contamination in urban areas has become a growing concern in recent years. Urban soils receive deposits of metals from various sources such as vehicular emissions, industrial discharges, domestic and commercial wastes, urban agriculture and municipal wastes (Atiemo et al. 2011; Hariprasad and Dayananda 2013; Pius and Orish 2013; Cheng et al. 2014; Defo et al. 2015; Yerima et al. 2013; Gruszecka and Wdowin 2013; Beata and Mirosław 2014; Nwaichi et al. 2014; Purushotham et al. 2012; Tahra and Hameed 2013; Pingguo et al. 2014; Lu et al. 2014; Al-Jaboobi et al. 2014). In urban areas, the contamination of streets, soils along the roadsides and major lands by heavy metals from atmospheric deposition, runoff, or other sources has been reported (Elliot 2002; Biggan and Linsheng 2010; Pius and Orish 2013). However, long-term deposition of metals at the soil surface leads to accumulation, transport and bio-toxicity caused by the mobility of these pollutants in the environment (Adriano 2001; Hariprasad and Dayananda 2013). Surfacedeposited metals infiltrate into soil and are either transferred to the water table by percolation or adsorbed by the solid matrix of the soil system. As observed in different cities in developing countries (Zhang 2008; Sudha et al. 2013), the population of Yaounde´ has experienced an explosive growth (averagely 4 %/year; Nguendo 2010) for fifty years. Between 1957 and 2005, the population increased from 58,099 to 1,817,524 (Zogning Moffo et al. 2011). The spectacular growth of the population has had an impact on the spatial evolution of the city which itself has been multiplied by nine (09) times between 1956 and 2000 (from 1740 to 15,900 ha). Spatial dynamics of the city at the expense of existing forest has contributed to increase in the frequency of floods in the city and in some pericentral areas. In the city of Yaounde´ and particularly in the Ntem watershed, the consequences of this urbanization are visible through the increase in impermeable surfaces (buildings and tarred roads), extension of urban agriculture, increase in sewage and solid wastes and improper waste dumping sites. Furthermore, as reported in many cities of developing countries, Yaounde´ is characterized with a progressive increased rate of vehicular pollution due to the congested urban circulation (Matcheubou et al. 2009; Rana 2011). Some reports also revealed that drinking water provisioning in the city of Yaounde´ by the public water distribution network can satisfy less than 65 % of the city’s needs; with the rest of the population using groundwater from wells/springs. The interactions between soil properties and pollutants are essential for determining their fate in soils system (Chuan et al. 1996; USEPA 2008). The transfer of metals from soil surface to the saturated materials below the water table generally occurs during and after rainfall events (Atteia 2005; Yerima et al. 2013; Shivakumar et al. 2012; Guala et al. 2013). Soil geochemical properties play a significant role in the retention of metals. Silveira et al. (2003) and Yerima et al. (2013) showed that the capacity of sorption of heavy metals in soils is generally influenced by some soil properties such as pH, carbonates, phosphates, organic matter, amount and type of clay, CEC and ionic strength. For the efficient soil pollution management in developing countries, models are useful and necessary to simulate the relationship between soil properties and pollutants over an extended

123

Investigating soils retention ratios and modelling…

period of time and better understand the mobility of these pollutants in soils (Smith et al. 1985; Yerima et al. 1989, 2013). Geochemical factors influencing the mobility of metals in soils have been discussed in previous reports. Yerima et al. (2013) showed that extractable micronutrients and heavy metals generally present a decrease in concentration with depth associated with decreasing organic carbon contents and increasing pH and CaCO3 contents that exert a major influence on the availability of the micronutrients and heavy metals in vertisols and vertic inceptisols of Ethiopia. Mariana and Guimaraes (2015) found that some soil properties (OM, CEC, clay) are suitable for the retention of contaminants in soils in Brazil. Many other investigations have reported similar observations in different other places (Fifi 2010; Violate et al. 2010; Xu et al. 2015; Tomasz et al. 2014; Xiao et al. 2015; Rajmohan et al. 2014). Nonetheless, very limited efforts have been devoted on heavy metals retention capacities of tropical soils (ferralsols and gleysols) and on the investigation of geochemical factors influencing the retention of these pollutants in soils in tropical environment. The objective of this study is to investigate the variation of soil retention ratios and the soil geochemical factors influencing the retention of heavy metals such as Pb and Cd in soils of the Ntem watershed in Yaounde´, Cameroon. Pb and Cd are chosen in this paper for the preliminary study of heavy metals occurrence in urban soils of Yaounde´, while another metal species will be investigated in further studies. Besides this, different types of pollution sources like vehicular and industrial emissions, flaking paint, fossil fuel combustion, smelter emissions, waste incineration and use of fertilizers in urban agriculture supported the presumption of the occurrence of Pb and Cd in these soils. Beyond Yaounde´, the knowledge from this paper is of interest to many cities of the world built on tropical soils, without sewage system, which are experiencing rapid population growth, increasing flood frequency, proliferation of improper wastewater and solid waste dumpsites, progressive rate of motorization and industrialization and low coverage of basic services of community water supply and electricity.

2 Materials and methods 2.1 Study area The study area is the Ntem drainage basin (Fig. 1), situated in the Eastern of the city of Yaounde´, between latitudes N03550 0600 and N03520 3300 and longitudes E011310 2600 and E011330 0200 . Yaounde´, the capital city of the Republic of Cameroon is divided into six (06) communes such as Yaounde´ 1 (Y1), Yaounde´ 2 (Y2), Yaounde´ 3 (Y3), Yaounde´ 4 (Y4), Yaounde´ 5 (Y5), Yaounde´ 6 (Y6) and Yaounde´ 7 (Y7). Yaounde´ city is drained by a number of perennial rivers including the Ntem River. The Ntem drainage basin surface area is 554 ha, unequally distributed in the communes of Yaounde´ 1, 2, 4 and 5 as shown in Fig. 1. This basin has an average altitude of 744.14 m and a moderate hydrographic network (Kalla 2007). The climate is the Equatorial Guinean type with four seasons. The mean annual rainfall in the city of Yaounde´ is about 1600 mm, while the mean annual temperature is 24.23 C. The primary vegetation (formerly equatorial forest) has been transformed by urbanization into tertiary forests. The geology of Yaounde´ reveals the existence of metamorphic rocks, dominated by gneiss. In the Ntem drainage basin, the hydrogeology is characterized by a shallow aquifer overlying a deep aquifer resting on faulted gneiss which is vulnerable to urban pollution (Kalla 2007).

123

C. Defo et al.

LOCATION OF THE REGION OF CENTRE IN

CAMEROON

Legend Internaonal limits Limits of the region Region of Centre

Legend Main Road Secondary Road Rail way Stream Lake Ntem drainage basin

Fig. 1 Location of the Ntem drainage basin within the Yaounde´ city, Cameroon (Adapted from Mabou 2013)

123

Investigating soils retention ratios and modelling…

2.2 Soil prospection, cartography of soil units and soil sampling For soil, prospection studies were carried out on-site, in the watershed. A soil auger was used to sample the various types of soils during prospection studies, and soil color was determined using the Munsell Color Chart following the procedures described by Yerima and Van Ranst (2005a, b). An Etrex GARMIN Geographical Positioning System (GPS) was used to designate boundaries between the different soil types. The watershed was divided into an appropriate number of small representative homogenous soil units (15). Each soil unit was homogeneous in slope, color and texture. The distribution of soil units was carried out (Fig. 2), using the software infographic (Map Info 8.5). Based on the distribution of soil units, soil samples were collected at surface (0–40 cm) in each soil unit, around 5 representative profiles excavated within the watershed (up to the water table) at sites P1, P2, P3, P4 and P5. As such, each soil profile was representative of a

Legend Limits of the watershed Stream Gleysols Xanthic Ferralsols Rhodic Ferralsols Main road Secondary road Hospitable wastewater dumpsite Sampling sites Church Mosque Rail way

Fig. 2 Presentation of the Ntem watershed, the sampling points and the distribution of the soil units [according to the World Reference Base for Soil Classification System (IUSS/FAO/ISRIC 1998)]

123

C. Defo et al.

soil unit (Fig. 2): Sampling sites were selected from different topographic levels (P1, P2, P3 and P4 were located at slope \3 %; P5 was at semi-slope \20 %), their proximity to pollution sources: 1. 2. 3. 4.

Tarred roads and railway (P1, P2, P3, P4 and P5), Waste dumpsites and wastewater effluents (P1, P2, P3, P4 and P5) Urban agricultural soils and commercial activities (mechanics, repair shops and painting) (P1, P2, P3, P4 and P5) Upstream of Ntem River (P1 and P2), its downstream (P4) and Ntem River tributary (P2 and P3).

One more important criteria was the proximity of sampling sites to water points that are most solicited by the populations. Site P1 corresponds to a Xanthic ferralsol; P2 to a Plinthic Gleysol; P3 and P4 to Mollic gleysols, and P5 to a Rhodic ferralsol. Sites P1 and P2 are both located upstream close to the ‘‘General Hospital’’ effluent drainage way near a solid waste dumpsite and agricultural farms and tarred roads, respectively. Sites P2, P3 and P4 are located downstream, representing the converging area of all the effluents from hospitals, domestic wastes, runoff and liquid effluents from solid waste dumpsites. Each sampling site was geo-referenced with a portable Global Positioning System (GPS) to locate the site and to set up future field surveys. In this study, data were collected according to the type of analysis targeted: on one hand, soil sampling for estimating the degree of purge of soil (Mina et al. 2007) or the soil retention ratio (RR), and on the other hand, soil sampling for linear regression analysis, during 3 months (August–October 2012) (in rainy season). For estimating the soil retention ratio (degree of purge), data were collected during the big rainy season in Yaounde´ (from August to September). Before the profile realization (at P1, P2, P3, P4 and P5), a simple indigenous water conservation system was temporary prepared around each sampling site. It was a simple contour slope wall constructed with stones across the slopes, thereby intercepting the surface runoff and maximizing the accumulation of pollutants transported by the runoff at that point. Thereafter, soil profiles were dug (up to the water table) after rainy events and total water infiltration. Soil samples were collected in the surface horizons (0–20 cm) and at the bottom of the unsaturated zone of the soil profiles (with variable depth). In each profiles, 5 samples were collected at surface horizon and 5 samples at the bottom of the unsaturated zone, which yields 10 samples per profile and a total of 50 samples to make the dataset necessary for estimating soil retention ratios. However, for linear regression analysis, a total of 110 soil samples were collected from the soil surface (0–20 cm) in the watershed according to 5 soil units. This surface sampling was distributed as follows: 20 samples around the sampling site P1 in Xanthic ferralsol; 20 samples around P2 in Plinthic Gleysol; 40 samples around P3 and P4 in Mollic gleysols; 20 samples around P5 in Rhodic ferralsol; and 10 composite soil samples (hence 2 per soil unit). All the soil samples collected were stored in new polythene bags, labeled and transported to the laboratory for analysis. Afterward, all the data were mixed to perform the linear regression analysis between heavy metal (Pb and Cd) concentrations and geochemical properties (OM, Clay, CEC and DpH).

123

Investigating soils retention ratios and modelling…

2.3 Analysis of soil samples The total concentrations of Pb and Cd in the soil samples were determined in soil extracts from soil air-dried \2 mm samples and total digestion in acid media following NF Standard ISO 11464/X31-412 procedures (NF ISO 2006) and Houba et al. (1989). Concerning soil samples preparation, the procedure was the following: air-dry the soil samples in shade, discard the plant residues, gravels and other materials if any, crush the soil clods lightly and grind with the help of wooden pestle and mortar, pass the entire quantity through a 2-mm stainless sieve (as well as 0.2–0.5 mm for organic carbon analysis) and remix the entire quantity of sieved soil thoroughly before analysis. Afterward, 10 mg of dried and homogenized (properly mixed) soil samples were placed in acid-washed 50-mL Teflon vessels, and 15 mL of diacid [HNO3/HClO4 (perchloric acid) at 0.1 N; 9 V:4 V] was added into the Teflon (PTFE) bombs using a microwave oven (AntonPaar MW 3000) for digestion. After digestion the solution was brought up to the 50-mL mark with distilled deionized water, filtered through a Whatman No 42 filter paper and diluted to 50 mL (Houba et al. 1989). Whatman No 42 provides a maximum degree of fine particle filtration, retaining the fine precipitates encountered in chemical analysis (excellent clarifying filter for cloudy suspensions and for water and soil analysis). The reagent blanks were monitored throughout the analysis to check the analytical results for reliability. The heavy metal concentrations were obtained through atomic absorption spectrophotometer (AAS) [Perkin–Elmer AAnalyst 800 Atomic Absorption Spectrometer, detection limit: Pb: 1.6 (lg/ kg); Cd: 0.1 (lg/kg)] after calibration of the device with standard solutions and the results reported in mg/kg. After heavy metal determination, organic matter, particle-size distribution, CEC and pH H2O and pH KCl in the soil samples were analyzed: Organic carbon (OC) was determined according to the method of Walkey and Black as described by Pauwels et al. (1992). Organic matter (OM) was obtained by the formula: OM (%) = OC (%) 9 1724. Particle-size distribution was determined by the pipette method and cation exchange capacity (CEC) by the ammonium acetate method buffered at pH 7 according to standard procedures of the Soil Conservation Service (Soil Conservation Service 1972). Besides this, pH H2O and pH KCl in soil samples were analyzed for soil pH (H2O and KCl) using the ISO 11464 method, with a glass electrode in a suspension of 1:2.5 soil: solution ratio in distilled deionized water and 1 N KCl solutions, respectively, after equilibration.

2.4 Estimation of the heavy metals retention (R) and the retention ratio of soils The retention (R) of a heavy metal in a soil is the concentration of the pollutant retained by the soil during the infiltration calculated by the formula: R ¼ Cs  Cw where Cs is the concentration of a given heavy metal at the subsurface of the soil and Cw is the concentration of the given heavy metal reaching the groundwater table (at the bottom of the unsaturated zone).

123

C. Defo et al.

The retention ratio (RR) in the unsaturated zone of a soil of a given heavy metal was calculated according to the following formula (Mina et al. 2007): RRð%Þ ¼

Cs  Cw  100 Cs

When RR = 100 %, the metal is totally retained in the unsaturated zone, but when this value is\100 %, the unretained portion of the heavy metal is released into the water table.

2.5 Statistical analysis The relationships among physical and chemical properties were investigated using correlation and regression analysis. The regression equations were computed using the ordinary least square method and the assessment of normality for the data set using the Shapiro– Wilk test, while multicollinearity test was computed between OM, CEC, clay percentage and DpH using tolerance and VIP (Variance Inflation Indicator) as indicator. The dependent variables selected included: Pb and Cd, while the independent variables were: OM, CEC, clay percentage and DpH where DpH = (pH H2O - pH KCl). When the probability of the statistics of Fischer (risk to be mistaken on the total significance) is lower than 10 %, the regression tested is significant for the given number of independent variables selected. Additionally, the principal component analysis (PCA) was implemented to further identify metals and geochemical parameters having similar affinity patterns, while hierarchical cluster analysis (CA) was used in this study to identify the relatively homogeneous groups of heavy metals and soil characteristics. Statistical analyses were performed using SPSS Software and Microsoft Excel.

3 Results and discussions 3.1 Morphological properties The main soil groups in the Ntem drainage basin are reported to be ferralsols and gleysols. Within the ferralsols (Station P5), Munsell colors showed a B horizon with a hue redder than 5 YR, a moist color value\3.5 and a dry color value[1, qualifying the soil samples at P5 (Fig. 2) as a Rhodic ferralsol. The other group of ferralsols have ferralitic horizons with yellow to pale yellow colors (Station P1) qualifying them as Xanthic ferralsols. At stations P2, P3 and P4, soils generally had gleyic properties within 100 cm from the soil surface qualifying them as Gleysols. The soil structure of surface horizons is dominantly granular. The subsurface horizons generally have prismatic and angular to subangular blocky structures parting to granular. The representativity of soil units in the surface area of the watershed showed the following distribution: Rhodhic ferralsol (85.2 %), Xanthic ferralsol (7.9 %) and gleysol (6.9 %). These observations are consistent with those of Yerima and Van Ranst (2005b) for tropical ferralsols and gleysols. Soil geochemical properties and distribution of heavy metals concentration were presented in this section according to the soil types of the watershed.

123

Investigating soils retention ratios and modelling…

3.2 Soil geochemical characteristics and heavy metals concentrations in soil layers in the Ntem drainage basin Table 1 presents the range of values of geochemical factors recorded a different sampling point P1, P2, P3, P4 and P5 distributed along between the soil surface and the bottom of the unsaturated zone (no replication was done for these parameters in soil profiles). Table 1 shows some chemical properties of different layers of soils investigated which are associated with the heavy metals retention. According to this table, the quantities of clays and organic matters are high irrespective of the soils investigated. In soils, heavy metals are divided up between the solid phase and the liquid phase which make up these soils. The quantity existing in the solution of the groundwater represents only a small amount of the totality of the pollutants (Perrono 1999). Thus, metals concentrate in the solid fraction of the earth, where they are divided up in the various organic and mineral fractions. An important fraction of heavy metals of the soil is found in the clayey fraction: they are fixed in silicate networks under a little available shape or are adsorbed on the clay surface (Yerima and Van Rants 2005a, b). Organic matter participates effectively in the retention of heavy metals which can be retained under exchangeable form or in form of complexes in which they are more energetically fixed (Yerima and Van Rants 2005a, b; Fifi 2010). For all the soils investigated, the CEC values indicate that there are net negative charges on the soil exchange complex, which attract and fix positively charged elements like heavy metals. The values of CEC (10.7–14.7 meq/100 g) recorded for these soils were in the same range like those obtained (12.71 meq/100 g) by Ngo-Mbogba et al. (2015) for the ferralsols of the southern Cameroon. In contrast, these CEC values were lower than the values (72.9–82.7 meq/100 g) obtained by Shabtai et al. (2014) in vertisols in the North of Ethiopia, and similarly to the observations of Temgoua et al. (2014) who obtained 30.1–32.2 meq/100 g for Andisols and Andic Ferralsols from Mount Bambouto in Cameroon. According to Yerima and Van Rants (2005a, b), the CEC is due to negative colloidal substance such as the clay minerals, the organic matter and colloidal silica. The concentrations of Pb recorded were higher than the permissible levels allowed for soils, at the stations P3 and P4 at the surface horizon (Table 2), while for Cd, all the mean concentrations were found below the norms at all the stations investigated. The highest concentrations of pollutants (especially Pb) were recorded in the gleysols (P3 and P4), located at the lowest elevation within the watershed, where the greatest quantities of pollutants are deposited by runoff and riparians. However, the norms used in this case are just indicative, since the values are for agricultural lands, and there are no local norms for soil pollution in Cameroon. Table 2 shows that the distribution of heavy metal varies from one site to another within the watershed. Additionally, the values of heavy metal concentrations recorded at the soil surface are higher than those of the bottom of the unsaturated zone, regardless of the sampling site. Soil layers play an important role on the immobilization of pollutants in the environment (Violate et al. 2010; Yerima et al. 2013). As previously discussed by Defo et al. (2015), several authors have shown the metal occurrence in different places in the world. But, these studies results fluctuate from one place to another according to local conditions. Table 1 indicates soils retention ratios have not been sufficiently used as a parameter to assess the heavy metal retention capacities of soils. Defo et al. (2015) indicated that several studies have shown the pollution and contamination of soils by heavy metals in many places in the world. The range of values of concentrations obtained are as follows: in the present study (Pb: 33.2–162 mg/kg; Cd:

123

123

b

a

24.50

51.00

5.44

5.36

3.20

9.30

22.68

41.00

4.01

3.76

CEC (meq/100 g) clay

Clay (%)

pH (H2O)

pH (KCl)

Organic matter (%)

CEC (meq/100 g) soil

CEC (meq/100 g) clay

Clay (%)

pH (H2O)

pH (KCl)

3.57

3.85

50.00

20.6

10.30

7.40

3.75

4.11

51.00

26.27

13.40

9.80

Xanthic ferralsol at (P1)

3.,60

4.11

43.00

19.76

8.50

7.40

3.58

4.28

44.00

24.31

10.70

6.90

Plenthic gleysol at (P2)

3.69

4.56

46.00

24.13

11.10

6.20

4.58

5.96

45.00

32.66

14.70

8.70

Mollic gleysol (1)a at (P3)

3.83

4.27

46.00

15.65

7.20

7.50

3.69

5.35

49.00

23.46

11.50

7.70

Mollic gleysol (2)a at (P4)

Unsaturated zone

Soil sampling preformed in the same soil type (Mollic gleysol) at two different stations (sites P3 and P4 in Fig. 2) and specified by (1) and (2) respectively

Bottom UZb

9.00

12.50

Organic matter (%)

Soil surface

Rhodic ferralsol at (P5)

Types of soils

CEC (meq/100 g) soil

Soil parameters

Soil layers

Table 1 Geochemical properties in different layers of soils

C. Defo et al.

28.32 ± 9.43

udl

Cd

udl

26.59 ± 3.8

164

0.14 ± 0.02

56.53 ± 4.09

0–20

Xanthic ferralsols at P1

0.14 ± 0.05

55.29 ± 10.30

50

0.19 ± 0.02

57.85 ± 10.01

0–20

Plinthic gleysol at P2

0.09 ± 0.03

83.51 ± 12.02

40

0.24 ± 0.08

104.70 ± 12.02

0–20

Mollic gleysol (1)# at P3

udl

65.16 ± 12.4

75

0.29 ± 0.05

162.8 ± 12.67

0–20

Mollic gleysol (2)# at P4

#

1.00*

100.00*

NA

1.00*

100.00*

NA

Norms applicable to soils (mg/kg)

Soil sampling preformed in the same soil type (Mollic gleysol) at two different stations (sites P3 and P4 in Fig. 2) and specified by (1) and (2), respectively

* NSW EPA (1997); Mean ± SD corresponds to the mean and standard deviation of 5 values. NA = Not Applicable, udl = under detected limit

430

Depth (cm)

0.25 ± 0.02

Cd

Pb

33.42 ± 11.28

Bottom UZ

0–20

Depth (cm)

Pb

Soil surface

Rhodic ferralsols at P5

Heavy metals (mg/kg)

Soil layer

Table 2 Heavy metals concentrations (mg/kg) in soil surface, bottom of the unsaturated zone of soil profiles

Investigating soils retention ratios and modelling…

123

C. Defo et al.

0.00–0.29 mg/kg) which is similar to those found by Rajmohan et al. (2014) in Indian agricultural soils in the Ganges Basin: Pb (17.79 mg/kg) and Cd (0.08 mg/kg) in soils. These observations are in the same range with those of Pingguo et al. (2014) found in agricultural soils of China (0.15 mg/kg for Cd). But the values of Pb (18.8 mg/kg) were lower than those found in tropical soils in Yaounde´ (Defo et al. 2015). In opposition, Gruszecka and Wdowin (2013) found high values of Pb (922 mg/kg) and 10 mg/kg of Cd in soils in vicinity of post-flottaison waste site in Poland. In Morocco, Al-Jaboobi et al. (2014) have found high values of Pb (107 mg/kg); Cd (1.17 mg/kg) in wastewater irrigation soils. In many cases, the soil pollution by diverse contaminants and its implication in water resources pollution has been associated to the consequences of land use due to human activities (Elhag et al. 2013). In the present study (in tropical soils), attention have been paid to show the variation of heavy metals concentrations between the soil surfaces and the bottom of unsaturated zones with respect to the types of soil, classification of soils (soil taxonomy) and soil retention ratios variation according to different soils units and geochemical factors affecting heavy metal retention in urban soils.

3.3 Retention ratios of soils in the Ntem drainage basin Figure 3 presents the variation of the different retention capacities (RR) (%) of the different soils. The graph presented in Fig. 3 has been performed using the data presented in Table 1. Figure 3 indicates that the retention capacities of these soils vary from one heavy metal to another. The retention capacities of the Rhodic ferralsol, Xanthic ferralsol and Mollic gleysol (2) were very high regardless of the heavy metal examined (RR [ 80 %), except Pb, for which (RR) was relatively low (generally \ 60 %; about 10 % for Rhodic ferralsols; 40 % for Xanthic ferralsols and 60 % for Mollic gleysols). It was 39.8 % for Pb and 100 % for Cd. In contrast, the retention ratio values of the Plinthic gleysol and the Mollic gleysol (1) were relatively low \60 %, regardless of the heavy metal concerned. This could be due to the small thickness of the unsaturated zone observed in these stations (\0.5 m). The high values of retention ratios observed for the Mollic gleysol (2) and the Xanthic and Rhodic ferralsols (from 75 to 100 %, for Cd) could be associated with the very thick unsaturated zone (from 1.65 to [5 m) and the much higher organic matter contents (Table 1) consistent with observations by Yerima et al. (2013).

120

Soil Retention Ratio (%)

Fig. 3 Distribution of retention ratios (RR) of the different soils within the Ntem drainage basin (Mean ± SD corresponds to the mean and standard deviation of 5 values)

100 80 Pb 60

Cd

40 20 0 Rodhic Ferralsol

Xanthic Ferralsol

Plenthic gleysol

Mollic Mollic gleysol(1) gleysol(2)

Types of soils

123

Investigating soils retention ratios and modelling…

These observations indicate that soils constitute a barrier which protects the groundwater resources from surface pollution (Mohamed and Elsayed 2007; Alloway 1995). Thus, the retention ratios of heavy metals vary as a function of soil type.

3.4 Linear regressions explaining relationships between heavy metal concentrations and geochemical soil properties 3.4.1 Range of data recorded at the soil surface and used in the linear regression analysis Regardless of the types of soil, all the values recorded (for 110 samples) were mixed for metals (Pb and Cd) on one hand and for geochemical parameters (OM, Clay, CEC and DpH) on the other hand for performing regression analysis. It can be observed that the values of all the parameters used above to calculating soil retention ratios (Table 2) are higher than those presented in Table 3 (for regression analysis). As stated above, 110 samples were collected in soil surface horizons (0–40 cm) and used for regression analysis, while 50 soils samples used for estimating the soil retention ratio (RR) were collected on sites where the stone walls were used for intercepting runoff and maximizing infiltration at the soil surface. This fact affected the results observed and showed a difference between the two cases. This difference is due to the fact that the sites of soil profiles were arranged to store the maximum load of pollutants during raining period, while no particular consideration and no replication at a particular point were taken for all the samples collected at the soils surface in the watershed for linear regression development. Instead of presenting the raw data for 110 samples in this section, Table 3 shows the range (minimum and maximum values) of values of each parameter used in regression analysis. The model investigated linear relationships between heavy metals (Pb and Cd) concentrations considered as dependent variables (Y) and soil characteristics considered as independent variables, Xi (OM, CEC, Clay and DpH).

3.4.2 Assessment of normality and multicollinearity of data The assessment of the normality of the variables of the dataset has been performed using Shapiro–Wilk test. Regardless of the variable tested (Pb, Cd, OM, CEC, Clay and DpH), the Shapiro–Wilk normality test showed for different variables that generally, W & 0.957, p value & 0.06642 [ 0.05, indicating that all datasets follow the normal distribution. Multicollinearity analysis were performed between the variables OM, CEC, Clay percentage and DpH in order to test whether one variable could be predicted from the others with a fair degree of accuracy. Each R-squared value explains how well that X variable is predicted from the other X variables (Y not considered). VIF [Variance inflation factor

Table 3 Range of values of the soil properties and heavy metals concentrations used for developing linear regression analysis Range

Clay (%)

OM

CEC (meq/ 100 g soil)

pH H2O

pH KCl

Pb (mg/kg)

Cd (mg/kg)

Minimum

11

1.47

1.81

4.01

3.19

7.5

0.027

Maximum

67

5.69

4.47

5.96

5.36

25

0.17

123

C. Defo et al.

(VIF = 1/(1 - R2)] is calculated from R–squared value. Results from indicate that all R2 values were \0.60, with the VIF \ 5, indicating that the X variables are independent.

3.4.3 Simple linear regression equations between heavy metal concentrations and soil properties a.

Simple linear regression equations between Pb concentrations and soil properties

From Table 4, it was observed that the coefficient of determination value (R2) is 0.750, indicating that there is a significant (p value \ 00.1) strong positive relationship between Pb concentrations in soils and amount of organic matter. For the clay percentage and CEC, the coefficients of determination (R2) values are 0.533 and 0.410. This indicates that there is a significant (p value \ 0.01) positive (low) relationship between Pb concentrations and the clay content of soils on one hand and with Pb concentration and CEC on other hand. In contrast for delta DpH, the coefficient of determination is very low, R2 = 0.011. From these observations, there is an extremely low positive relationship and not significant (p value[ 0.01) between Pb concentrations and pH. This is consistent with the behavior of ferralsols that have pseudo-sand and pseudo-silt textures and composed of sesquioxidic minerals with low surface areas and hence low retention capacities when organic matter content is low (Yerima and Van Ranst 2005a). From these analyses, the retention of Pb in these soils is little influenced by DpH. In this study, DpH represents the difference between pH H2O and pH KCl. A DpH [ 0 indicates that the soil exchange complex is dominated by layer silicate clays with a net negative charge with the potential to attract and retain positively charged cations (Pb and Cd) (Yerima and Van Ranst 2005a). This confirms that these soils have very low negative charges on their surfaces and thus cannot retain many cations. From the linear regressions developed to measure the relationship between the concentrations of lead (Pb) retained in soils and geochemical properties such as OM, CEC, Clay percentage and DpH, the following observations can be made: the major geochemical factors influencing the retention of Pb in the soils of the watershed is OM (major influence), followed by Clay and CEC which exert a minor influence. Besides, the latter factor (pH) has no influence on the retention of lead in the soils of the watershed. b.

Simple linear regression between Cd concentrations and soil properties

Table 5 presents a coefficient of determination of the linear equations between Cd and OM with R2 = 0.634 and p value \ 0.01, indicated that there’s a significant positive relationship between Cd concentrations in soil and OM. The coefficient of determination of the relationship between Cd and CEC, R2 = 0.327, Cd and Clay (R2 = 0.385) indicates that there is a low significant (p value\ 0.01) positive

Table 4 Simple linear regression equations between Pb concentrations and OM, Clay, CEC and DpH Dependent variable

Regressor

Equation

R2

p value

Significance

Pb

OM

Pb = 13.01*OM

0.750

6.367 9 10-13

Yes*

Clay

Pb = 2.911*Clay - 5.705

0.533

7.404 9 10-09

Yes*

CEC

Pb = 22.61*CEC ? 30.230

0.410

5.569 9 10-07

Yes*

DpH

Pb = 10.16*DpH ? 77.190

0.011

0.460

No

* p value \ 0.01

123

Investigating soils retention ratios and modelling…

relationship between Cd concentrations in soil, amount of clay and CEC confirming the minor influence on clay and CEC on the retention of Cd in the soils of the watershed. The low R2 value indicates that the clay fraction of these soils is dominated by low activity clays. The retention of Cd in soil is very low or not influenced by DpH. The coefficient of determination, R2 = 0.001, indicates a very low positive relationship that is not significant (p value [ 0.01) between the two variables confirming that these soils have a very low negative charge. The simple linear regression models developed between Cd concentrations and some geochemical parameters indicate that the major factor which influence the retention of Cd in the soils of the Ntem watershed are OM and CEC; amount of clay is a minor factor, while DpH plays very little or no role.

3.4.4 Multiple regression analysis between heavy metal concentrations and soil properties Multiple linear regression equations developed between heavy metals concentrations and soil properties (clay content, organic matter, cation exchange capacity and DpH) indicate that the regression coefficients obtained between Pb, Cd concentrations and different regressors were 0.82 and 0.72, respectively (Table 6). These results indicate that there is generally a significant relationship between Pb and Cd concentrations and geochemical factors such as OM, CEC, amount of clay and pH. Results obtained in this study fit the postulated hypothesis as all the equations developed (Yi = Xi Bi ? e) are acceptable as all the regression coefficients (Bi) were very significant (p values \ 0.0001). a.

Multiple linear regression equations between Pb concentrations and soil properties

The overall equation that explains the relationship between Pb concentrations and geochemical factors influencing its retention is as follows: Pb ¼ 36:814 þ 125:094 Cd þ 6:436 OM þ 5:034 CEC þ 1:074 Clay In this equation, significant codes are: 0.0001 ‘***’ 0.001 ‘**’ 0.05. Each p value compares the full model with a simpler model omitting one variable. It tests the effect of one variable, after accounting for the effects of the others. This model was consistent with the conclusions obtained for the simple linear regression equations. From the above equation, the regression coefficients of Pb (dependent variable) with regressors (Cd, OM, CEC and Clay) were positive and were all significant (p value\ 0.05). Hence, contributing to the Pb retention in soils except for DpH was not significant (p value[ 0.1) and was excluded from

Table 5 Simple linear regression equations between Cd concentrations and OM, Clay, CEC and DpH Dependent variable

Regressor

Equation

R2

p value

Cd

OM

Cd = 0.040*OM

0.634

3.271 9 10-12 Yesa

Clay

Cd = 0.008*Clay

0.385

2.358 9 10-07 Yesa

CEC

Cd = 0.050*CEC ? 0.148

0.327

1.420 9 10-05 Yesa

DpH

Cd = 0.004*DpH ? 0.265

0.001

0.898

a

Significance

No

p value \ 0.01

123

C. Defo et al.

the model. These confirmed the observations of Violate et al. (2010) and Silveira et al.(2003) who reported that sorption of heavy metals in soils is strongly influenced by soil geochemical factors such as the nature of the sorbents, redox reactions, and presence of organic and inorganic ligands. Besides this, these observations showed that the retention of a given metal is influenced by effects of competitive sorption (due to the presence of other metals in the system), and then explaining similar observations by Silveira et al. (2003), Yerima et al. (2013) and Fifi et al. (2013) who approved on one hand that organic matter is the most important soil constituent retaining heavy metals, though pH, CEC, clay and on the other hand that the presence of competing ions also affect heavy metal adsorption and speciation in soils. In addition, Dong et al. (2008) attested similar observations while assessing the adsorption of Pb, Cd and other metals in sediments, using the column experiments. Results showed that metals became more mobile in multimetal than in monometal conditions, as a consequence of the competitive adsorption among metals increases the mobility of these metals. b.

Multiple linear regression equations between Cd concentrations and soil properties

The overall equation that explains the relationship between Pb concentrations and geochemical factors influencing its retention is as follows: Cd ¼ 0:334 OM þ 0:01 Pb In this equation, significant codes are: 0.001 ‘***’ 0.05 ‘**’. The principle of this analysis is that each p value compares the full model with a simpler model omitting one variable. It tests the effect of one variable, after accounting for the effects of the others. For equation, it was observed for this model that p value is 0.01(\0.05), indicating that the model is globally significant. The regression coefficients of Cd with CEC, Clay and DpH were not significant and were excluded from the model. Only OM and Pb were significant, p value = 0.0001 \ 0.001 (for Pb) and p value = 0.01 \ 0.05 (for OM). Therefore, OM was, therefore, the major geochemical factor influencing the retention of Cd in the soils of the Ntem watershed, coupled to the presence of the other metals like Pb (which showed the significant contribution of the competitive adsorption) in the soil exchange complex. Additionally, other parameters like CEC, clay and DpH have minor or negligible contribution. These observations are in agreement with those of Silveira et al. (2003) who

Table 6 Multiple regression parameters of Pb, Cd and top soil geochemical characteristics Variables

Lead

Cadmium

Coef

SD

Coef

SD

Clay (%)

1.3693***

(0.3275)

0.0028120**

(0.0009869)

OM (%)

9.9425***

(1.5656)

0.0273032***

(0.0047180)

CEC (meq/100 g)

6.51290*

(2.8695)

0.0101919

(0.0086475)

DpH

7.0110

(6.2838)

-0.0020637

(0.0189367)

Intercept

-43.7575***

(10.1898)

-0.0208601

(0.0307080)

Observations

150

150

R2

0.8083

0.72

Signif. codes: 0.0001 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

123

Investigating soils retention ratios and modelling…

reported that OM is one of the major factors that influence Cd retention in the soils, while clay content and pH play negligible roles. Equally, these observations are in agreement with the simple regression models described in the previous sections. In agreement with Silveira et al. (2003), Yerima et al. (2013) and Fifi et al. (2013), organic matter is the most important soil constituent retaining heavy metals, though pH, CEC and the presence of competing ions also affect heavy metal adsorption and speciation in soils. The equations found in this work can be used to predict mobility and only estimate concentration of heavy metal in soil, but not to predict concentration (owing the fact that the predicted concentration depends on many more factors than the interaction with soil components or the selected parameters, such as periodical metal concentration, climate, soil moisture, etc.) If no resolutions are taken to mitigate the occurrence of these metals in the watershed, a long-term deposition of these pollutants in soils can lead to accumulation, transport and bio-toxicity caused by mobility of significant fractions of the metals, enter the food chain and consequently become harmful for living organisms and public health. This study can be also applicable to design containment technology using soils to restrict the movement of pollutants (heavy metals) in a polluted area.

3.5 Detection of metals and geochemical parameter having similar distribution (affinity) patterns and the relative homogeneous groups The principal component analysis was used to identify heavy metals and geochemical factors with similar affinity. Principal components analysis with varimax rotation was conducted to assess how five ‘‘achievement’’ variables clustered. These variables were Pb, Cd, CEC, OM, and clay and DeltapH (DpH). (The assumption of independent sampling was met. The assumptions of normality, linear relationships between pairs of variables and the variables being correlated at a moderate level were checked, and DeltapH (DpH) test did not meet the assumptions, and it was correlated at a low level with each of the other variables.) Two components were rotated, based on the eigenvalues over 1 criterion and the scree plot. After rotation, the first component accounted for 61.37 % of the variance, and the second component accounted for 17.02 % of the variance. Table 7 displays the items and component loadings for the rotated components, with loadings less than 0.30 omitted to improve clarity. Results suggest, in keeping with zero-order correlations, that DeltapH (DpH) test scores are not substantially related to the other measures and should not be aggregated with them but that the other measures form a coherent component. Figure 4 presents the component

Table 7 Component loadings for the rotated components

Item

Component loading 1

2

Communality

Pb

0.945

0.567

Cd

0.913

0.653

OM

0.878

0.765

Clay

0.790

0.902

CEC

0.749

DeltapH (DpH)

0.834 0.991

0.982

123

C. Defo et al.

Fig. 4 Component plot in rotating space showing metals and geochemical parameter with similar affinity

plot in rotating space obtained from the analysis. The component plot in rotated space gives a visual representation of the loadings plotted in a 2-dimensional space (Fig. 4). This plot of the component loadings shows that Pb, Cd, CEC, OM, and clay load highly and positively on the first component. DeltapH (DpH) has a loading closed to zero on the first component, but loads highly on the second. This plot confirmed all the information discussed in the previous sections using multiple linear models to describe relationships between Pb, Cd, OM, clay, CEC and DpH. The component 1 describes better these interactions showing affinity between metals and geochemical parameters studied except DpH which was excluded in the models. Figure 5 shows the Dendogram using average linkage between groups. The branching-type nature of the Dendrogram gives an idea of how great the distance was between groups that are clustered in a particular step, using a 0–25 scale along the top of the chart. Figure 5 shows that Pb and Cd were significantly correlated with each other and formed a cluster. Strong association was also observed between Pb, Cd and OM, which were associated with clay and CEC at the later stages. As expected, DeltapH (DpH) was isolated from the other geochemical parameters, which is indicative of lack of association with the others in the soils. From these observations, Pb concentrations influence the retention of Cadmium and vice versa, as well as geochemical factors like OM, clay content and CEC. OM appears to be the major factor followed by clay and CEC, while DpH plays minor or no role. These observations are in agreement with those of Hu et al. (2013) who showed strong correlation between metals concentrations in soils of Guangdong Province of China (Cu, Zn, Pb and Cd) forming clusters. Equally, Fifi (2010) indicated that the

123

Investigating soils retention ratios and modelling…

Fig. 5 Dendogram identifying metals and geochemical parameters relative homogeneous groups

presence of one metal influences the co-precipitation of other heavy metals in the soil matrix, and the adsorption process is affected by soils geochemical properties. Numerous studies have been conducted to determine geochemical parameters influencing heavy metal mobility in soils in different places in the world. The following section compares the findings of the present study to the findings of some previous authors in different conditions in the world. As it is observed, many studies investigated geochemical factors influencing heavy metal retention in soils worldwide. But very little attention had been paid to assess relationships between soil geochemical parameters and heavy metal retention in ferralsols and gleysols (Xiao et al. 2015; Chandrasekaran and Ravisankar 2015; Xu et al. 2015; Gruszecka and Wdowin 2013; Yerima et al. 2013). Xiao et al. (2015) used multiple linear regression between heavy metals (Cu, Pb, Cd and Zn) and OC, soil texture and CEC of loess, sandy and loamy soils in surface horizon (0–2 m) in Hannover, Germany. From the relationships, results did not provide any conclusion concerning major/minor factor influencing heavy metals retention in soil. The authors suggested that further investigations are necessary to provide final conclusion concerning these factors. Equally in Tamil Nadu, India, Chandrasekaran and Ravisankar (2015) showed multiple regression equations between Al, Co, Fe, Mn, Cu, Zn, Cr, Cd and Pb and pH, EC, bulk density, porosity texture and color without indicating major or minor factor influencing the retention of heavy metals in soils. Nevertheless, in Poland, Gruszecka and Wdowin (2013) did a similar work between Pb, Cd, Zn, As, Tl and pHH2O; pHKCl; OM; Eh; mineral composition of Podzo soils and concluded that OM exerts major influence and mineral composition no influence.

123

C. Defo et al.

In Ethiopia, Yerima et al. (2013) stated that parameters such as pH, EC and organic carbon and the mutual interrelationships among these metals are important for development of predictive models for available indices of micronutrients and some heavy metals as demonstrated in this study. They added that as the parameters like pH, EC, organic carbon and CaCO3 equivalent are generally available from most soil survey data or can be determined at minimal cost, they can be used for prediction of the available micronutrients Fe, Mn, Zn and Co with a fair degree of accuracy at little cost.

4 Conclusions This study showed that heavy metal retention in soil is actually an ongoing process, and the retention ratio (RR) varies from one soil to another. The RRs ranged from 10 to 100 % for all soils investigated regardless of the type of heavy metal, indicating that soils constitute a barrier which protects the groundwater resources from surface pollution. Organic matter exerts major influences on the retention of Pb and Cd contents in soils of the watershed, while CEC, clay content and DpH play only a minor role. The high amounts of low activity clays with pseudo-sand and pseudo-silt textures coupled with the sesquioxidic mineralogy with low surface areas are largely responsible for the low retention of Pb and Cd observed when amount of clay and pH are used as regressors. Moreover, soils enriched in organic matter can be used in containment technology to restrict the movement of Pb and Cd in controlled waste dumpsites. In view of the fact that environmental pollution is still on the increase in tropical cities, knowledge about the geochemical factors that affect the mobility of heavy metals in tropical soils could be useful in designing containment technologies that would enable the formation of barriers between the contaminated media and the soil surface, thereby shielding humans and the environment from the harmful effects of its contents and thus limiting the migration of the contaminants. Acknowledgments The first author received a research grant from the University of Dschang. The authors would like to thank IRAD-Yaounde´ and the Laboratory of Soil and Environment of the Faculty of Agronomy and Agricultural Sciences (FASA) of the University of Dschang for the soil analysis.

References Adriano, D. C. (2001). Trace elements in terrestrial environments: biochemistry, bioavailability and risks of metals. New York: Springer. Al-Jaboobi, M., Adelmajid, Z., Mohamed, T., Abdellah, E., Zakaria, M., Hasna, Y., & Mohammed, B. (2014). Evaluation of heavy metals pollution in groundwater, soil and some vegetables irrigated with wastewater in the Skhirat region ‘‘Morocco’’. Journal of Materials and Environmental Science, 5(3), 961–966. Alloway, B. J. (1995). Soil processes and the behavior of heavy metals. In B. J. Alloway (Ed.), Heavy metals in soils (pp. 122–127). New York: Blackie Academic and Professional Publication. Atiemo, M. S., Ofosu, G. F., Kuranchie-Mensah, H., Tutu, A. O., & Blankson, S. A. (2011). Contamination, assessment of heavy metals in road dust from selected roads in Accra, Ghana. Research Journal of Environmental and Earth Sciences, 3(5), 473–480. Atteia, O. (2005). Chimie et pollutions des eaux souterraines. Paris: Lavoisier Tec & Doc. Beata, M., & Mirosław, W. (2014). Trace metals content in soils along the state road 51 (northeastern Poland). Environmental Monitoring Assessment. doi:10.1007/s10661-013-3562-z. Biggan, W., & Linsheng, Y. (2010). A review of heavy metal contamination in urban soils, urban roads, urban dusts and agricultural soils from China. Microchemichal Journal, 94, 99–107.

123

Investigating soils retention ratios and modelling… Chandrasekaran, A., & Ravisankar, R. (2015). Spatial distribution of Physico-chemical properties and function of heavy metals in soils of Yelagiri hills, Tamilnadu by Energy dispersive X-ray florescence spectroscopy (EDXRF) with statistical approach. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 150, 586–601. doi:10.1016/j.saa.2015.05.083. Cheng, H., Li, M., Zhao, C., Li, K., Peng, M., Qin, A., & Cheng, X. (2014). Overview of trace metals in the urban soil of 31 metropolises in China. Journal of Geochemistry Exploration. doi:10.1016/j.gexplo. 2013.08.012. Chuan, M. C., Shu, G. Y., & Liu, J. C. (1996). Solubility of heavy metals in contaminated soil: Effects of redox and pH. Water, Air, and Soil pollution, 90, 543–556. Defo, C., Yerima, B. P. K., Kengne, N. I. M., & Bemmo, N. (2015). Assessment of heavy metals in soils and groundwater in an urban watershed in Yaounde (Cameroon-West Africa). Environmental Monitoring and Assessment, 187(1), 1–17. Dong, C. S., Yu, K., & DeLaune, R. D. (2008). Comparison of monometal and multimetal adsorption in Mississippi River alluvial wetland sediment: Batch and column experiments. Chemosphere, 73, 1757–1764. Elhag, M., Silovikos, A. P., & Sakellariou-Makrantonaki, M. (2013). Land use changes and its impacts on water resources in Nile Delta region using remote sensing techniques. Environment, Development and Sustainability, 15, 1189–1204. Elliot, E. (2002). Land application of sewage sludge and biosolids. Manchester: ACRC Press Company, Lewis Publisher. Fifi, U. (2010). Impacts des eaux pluviales urbaines sur les eaux souterraines dans les pays en de´veloppement. Me´canismes de tranfert des me´taux lourds a` travers un sol mode`le au Port-au-Prince, Haı¨ti. The`se de Doctorat/PhD en Cotutelle de l’INSA, France L’Universite´ Quisqueya (Haı¨ti). Fifi, U., Winiarski, T., & Evens, E. (2013). Assessing the mobility of lead, copper and cadmium in a calcareous soil of Port-au-Prince, Haiti. International Journal of Environmental Resources and Public Health, 10(11), 5830–5843. doi:10.3390/ijerph10115830. Gruszecka, A. M., & Wdowin, M. (2013). Characteristics and distribution of analyzed metals in soil profiles in the vicinity of a postflotation waste site in the Bukowno region. Poland: Environmental Monitoring and Assessment. doi:10.1007/s10661-013-3164-9. Guala, S., Vega Flora, A., & Covelo Emma, F. (2013). Modeling the plant–soil interaction in presence of heavymetal pollution and acidity variations. Environmental Monitoring and Assessment, 185(1), 73–80. Hariprasad, N. V., & Dayananda, H. S. (2013). Environmental impact due to agricultural runoff containing heavy metals—A review. International Journal of Scientific and Research Publications, 3, 2250–3153. Houba, V. J. G., Van Der Lee J. J., Novozamski, I., & Walinga, I. (1989). Soil and plant analysis. A series of Syllabi. Part 5: Soil analysis procedures, 5th edition. Agricultural University of Wagningen, Department of Soil Science and Plant Nutrition. Hu, Y., Liu, X., Bai, J., Shih, K., Zeng, E. Y., & Cheng, H. (2013). Assessing heavy metal pollution in the surface soils of a region that had undergone three decades of intense industrialization and urbanization. Environmental Science and Pollution Research, 20, 6150–6159. doi:10.1007/s11356-013-1668-z. IUSS/FAO/ISRIC. (1998). World reference base for soil resources. World soil Resources Report 84, FAO, Rome. Kalla, M. F. (2007). Contribution a` l’e´tude hydrodynamique des aquife`res du bassin du NTEM a` Yaounde´. Faculte´ des Sciences. Universite´ de Yaounde´ 1: Me´moire de DEA en Ge´ologie de l’inge´nieur. Lu, F., Yang, J., & Li, H. (2014). Risk evaluation of heavy metals in soil in the sewage irrigation area: A case study of Shijiazhuang. Journal of Chemical and Pharmaceutical Research, 6, 2353–2357. Mabou, P. B. (2013). Pouvoir public et ame´nagement urbain a` Yaounde´. Universite´ de Yaounde´ 1: The`se de Doctorat/Ph.D en Ge´ographie. Mariana, C. K., & Guimaraes, V. S. R. (2015). Geological and geotechnical characterization of the basal soil portion at a former mining waste disposal. EJGE, 20, 369–381. Matcheubou, A., Yamba, J., & Tamo Tatietse, T. (2009). Impact du parc automobile sur la congestion du trafic et la pollution de l’air dans la ville de Yaounde´. Colloque international Environnement et transports dans des contextes diffe´rents, Ghardaı¨a, Alge´rie, 16–18 fe´v. 2009. Actes, ENP ed., Alger, pp. 245–252. Mina, A., Saadia, A., Abdes, S. B., & Ben, Y. H. (2007). Cartographie de la vulne´rabilite´ a` la pollution des eaux souterraines: Application a` la plaine du Gharb (Maroc). Revue des Sciences de l’eau/Journal of Water Science, 20, 185–199. Mohamed, R., & Elsayed, A. S. (2007). Dispersion and deposition of heavy metals around two municipal solid waste (MSW) Dumpsites. Alexandria, Egypt. American-Eurasian Journal of Agricultural and Environmental Science, 2, 204–212.

123

C. Defo et al. NF ISO. (2006). Qualite´ des sols, pre´traitement pour analyses physico-chimiques. NF X31-412/NF ISO 11464 de´cembre 2006. Ed (01/12/2006). Ngo-Mbogba, M., Yemefack, M., & Nyeck, B. (2015). Assessing soil quality under different land cover types within shifting agriculture in South Cameroon. Soil & Tillage Research, 150(2015), 124–131. Nguendo, Y. B. (2010). J Health Popul Nutr, 28(5), 424–435. NSW EPA. (1997). Environmental management guidelines for the use and disposal of biosolids products. New South Wales Environmental Protection Authority. Nwaichi, E. O., Wegwu, M. O., & Nwosu, U. L. (2014). Distribution of selected carcinogenic hydrocarbon and heavy metals in an oil-polluted agriculture zone. Environmental Monitoring and Assessment. doi:10.1007/s10661-014-4037-6. Pauwels, J. M., Van Ranst, E., Verloo, M., & Mvondo Ze, A. (1992). Manuel de laboratoire de pe´dologie. AGC D-INADER: Publications agricoles - 28 Bruxelles - Dschang. Perrono, P. (1999). Les micropolluants me´talliques des boues de stations d’e´puration urbaine et l’e´pandage agricole. Me´moire de D.U.E.S.S., D.E.P: Universite´ de Picardie, Amiens. Pingguo, Y., Miao, Y., Renzhao, M., & Hongbo, S. (2014). Multivariate-Statistical Assessment of Heavy Metals for Agricultural Soils in Northern China. Hindawi Publishing Corporation The Scientific World Journal. doi:10.1155/2014/517020. Pius, S. F., & Orish, E. O. (2013). Heavy metal pollution in Sub-Saharan Africa and possible implications in cancer epidemiology. Asian Pacific Journal of Cancer Prevention, 14, 3393–3402. Purushotham, D., Mahjoor, A. L., Mehnaz Rashid, A., Narsing, R., & Shakeel, A. (2012). Deciphering heavy metal contamination zones in soils of a granitic terrain of southern India using factor analysis and GIS. Journal of Earth System Science, 121(4), 1059–1070. Rajmohan, N., Prathapar, S. A., Jayaprakash, M., & Nagarajan, R. (2014). Vertical distribution of heavy metals in soil profile in a seasonally water logging agriculture field in Eastern Ganges Basin. Environmental Monitoring and Assessment, 186(9), 5411–5427. Rana, M. M. (2011). Urbanization and sustainability: Challenges and strategies for sustainable urban development in Bangladesh. Environment, Development and Sustainability, 13, 237–256. Shabtai, I. A., Shenker, M., Edeto, W. L., Warburg, A., & Ben-Hur, M. (2014). Effects of land use on structure and hydraulic properties of Vertisols containing a sodic horizon in northern Ethiopia. Soil & Tillage Research, 136, 19–27. Shivakumar, D., Srikantaswamy, S., Sreenivasa, S., & Kiran, B. M. (2012). Speciation and geochemical behaviour of heavy metals in industrial area soil of Mysore City, India. Journal of Environmental Protection, 3, 1384–1392. Silveira, M. L. A., Alleoni, L. R. F., & Guilherme, L. R. G. (2003). Biosolids and heavy metals in soils. Science and Agriculture, Piracicaba Brazil, 60, 793–806. Smith, C. W., Hadas, A., & Koyumdjiisky, H. (1985). Shrinkage and Atterberg limits in relation to other properties of principal soil types in Isreal. Geoderma, 35, 47–65. Soil Conservation Service. (1972). Soil survey laboratory methods and procedures for collecting soil samples. Soil Survey Invest. Rep. no. 1. U.S. Government Printing Office, Washington, DC. Sudha, M. C., Sakthivadivel, R., & Sakthivadivel, R. (2013). Water Bodies Protection Index for assessing the sustainability status of lakes under the influence of urbanization: A case study of south Chennai, India. Environment, Development and Sustainability, 15, 1157–1171. Tahra, T. A., & Hameed, S. (2013). Bioconcentration of heavy metals in Alfalfa (Medicago sativa) from farm soils around Sohar industrial area in Oman. APCBEE Procedia, 5, 271–278. Temgoua, E., Djoukouo, J., Likeng, J. D. H., & Tematio, P. (2014). Structural stability of Andisols and Andic Ferralsols from Mount Bambouto, West Cameroon highlands. Cameroon Journal of Experimental Biology, 10(01), 26–34. Tomasz, C., Grzegorz, K., Poluszyn´ska, J., & Katarzyna, K. (2014). Pollution of flooded arable soils with heavy metals and polycyclic aromatic hydrocarbons (PAHs). Water Air Soil Pollution. doi:10.1007/ s11270-014-2145-0. U.S. EPA. (2008). Report on the environment. Washington, DC: Environmental Protection Agency. Violate, A., Cozzolino, V., Perelomov, L., Caporale, A. G., & Pigna, M. (2010). Mobility and bioavailability of heavy metals and metalloids in soil environments. Journal of Soil Science and Plant Nutrition, 10, 268–292. Xiao, H., Bottcher, J., & Uterman, J. (2015). Evaluation of field scale variability of heavy metal sorption in soils by scale factors-scaling approach and statistical analysis. Geoderma, 11, 241–252. doi:10.1016/ jgeoderma.201411012. Xu, L., Tieyu, W., Wei, L., Ni, K., Shijie, L., Lin, W., et al. (2015). Factors influencing the contents of metals and arsenic in soils around the watershed of Guanting Reservoir, China. Journal of Environmental Sciences, 25(3), 561–568. doi:10.1016/S1001-0742(12)60095-3.

123

Investigating soils retention ratios and modelling… Yerima, B. P. K., & Van Ranst, E. (2005a). Introduction to soil science: Soils of the tropics. Canada: Trafford Publishing. Yerima, B. P. K., & Van Ranst, E. (2005b). Major soil classification systems used in the tropics: Soils of Cameroon. Canada: Trafford publishing. Yerima, B. P. K., Van Ranst, E., Sertsu, S., & Verdoodt, A. (2013). Pedogenic impacts on the distribution of total and available Fe, Mn, Cu, Zn, Cd, Pb and Co contents of vertisols and vertic inceptisols of the Bale Mountain area of Ethiopia. African Journal of Agricultural Research, 8(44), 5429–5439. Yerima, B. P. K., Wilding, L. P., Hallmark, C. T., & Calhoun, F. G. (1989). Statistical relationships among selected properties of Northern Cameroon Vertisols and Associated Alfisols. Soil Science Society of America Journal, 53, 1763–1771. Zhang, W. (2008). A forecast analysis on world population and urbanization process. Environment, Development and Sustainability, 10, 717–730. Zogning Moffo, M. O., Tsalefac, M., & Corneliu, I. (2011). Floods risks in the Mfoundi upstream drainage basin in Yaounde´: A response to climatic modifications or to human impacts? Present Environment and Sustainable Development, 5(2), 33–44.

123