Hydrochemistry of shallow groundwater and surface water in the Ndop plain , North West Cameroon

1 Department of Chemistry, School of Science, Tokai University, Hiratsuka, 259-1211, Japan. 2 Institute of Mining and Geological Research, P.O. Box 4110, Yaounde, Cameroon. 3 Department of Environmental Science, University of Buea, Box 63, Buea, Cameroon. 4 G.B.H.S. Atiela, Nkwen, P.O. Box 5147, Bamenda, Cameroon. 5 Department of Earth and Planetary Sciences, Tokyo Institute of Technology, Tokyo, 152-8551, Japan.


INTRODUCTION
Groundwater and surface water geochemical studies can provide a better understanding of potential water quality variations due to geology and land use practices (Edmunds and Smedley, 1996;Deutsch, 1997;Appelo and Postma, 2005;Ramesh and Elango, 2012).In Africa, groundwater is a vital resource with nearly 80% of the continent's population relying on it as a main source of drinking water.However, in many parts of the continent, reaching basic health requirements is still a concern (Adelana and MacDonald, 2008;Adelana et al., 2011;Edet et al., 2011;Tay, 2012).
Cameroon is endowed with numerous water resources (Katte et al., 2003;Molua and Lambi, 2006), but an assessment of these resources, including their reliability, quality and sustainability has not been systematically carried out (Katte et al., 2003).Improved drinking water sanitation in semi-urban and rural communities is nearly the same from 37% in 1990 to 36% in 2010 (WHO/UNICEF, 2012).With increasing population, the high demand for water, especially in most poor rural communities, is met by shallow groundwater of generally unknown quality.This is usually the best option all year round as most surface water sources and springs are unreliable and run dry or reduce in size during the dry season, or become polluted by heavy floods in the rainy season.
Only few hydrochemical studies have been carried out to assess the geochemical characteristics of groundwater and its suitability for use in Cameroon (example, Njitchoua and Ngounou-Ngatcha, 1997;Katte et al., 2003;Fantong et al., 2009;Eneke et al., 2011;Ako et al., 2012).The Ndop plain is a semi-urban community where groundwater, as well as surface water, is a critical resource for human livelihood.The plain has an estimated population of over 200,000 people with ca.70% involved in subsistence farming.Many ethnic groups have been attracted to the area as a result of fertile soils (Ndzeidze, 2008;Fonge et al., 2012).This has been followed by high demand for water and other natural resources (Fonge et al., 2012).Over 70% of the population depends solely on shallow groundwater, as well as surface water of poor microbial quality for drinking and other domestic purposes (Wirmvem et al., 2013).A hydrochemical study of water quality based on major ion composition has not been carried out.Since the development of the Bamendjin dam in the area in the 1970s, rice cultivation is the main cash crop using the dam and river waters.Unfortunately, there is no information on the suitability of the water sources for irrigation despite the on-going programme by the Cameroon government to revamp rice cultivation in the area.The chemical composition of water and its mineralization process are imperative in classifying and assessing drinking water quality (Deutsch, 1997;WHO, 2004;Kozisek, 2005) while irrigation water quality criteria can be used as a guideline by farmers for selecting appropriate management practice to overcome potential salinity hazard (Gupta et al., 2009;Ramesh and Elango, 2012).
The present study, therefore, presents the first hydrochemical baseline dataset of groundwater and surface water in the Ndop plain.The objectives were to (1) determine the physico-chemical characteristics of the water, (2) define the factors that control major ion chemistry, and (3) evaluate their suitability for drinking and irrigation.The data will provide a hydrochemical baseline data for water quality assessment, management, and monitoring of changes in the hydrochemical system from the prevailing human activities in the area.The study will also contribute to the limited knowledge on natural water composition in North West Cameroon.With the on-going programme to revamp rice cultivation in the area, an assessment of the suitability of the water sources for irrigation is invaluable.

The study area
The study area (Figure 1) is an inter-mountain basin (ca.1,100 km 2 ) in the Western Highlands of Cameroon (WHC), along the Cameroon Volcanic Line (CVL).From the surrounding CVL rages, the relief drops steeply from an average altitude of ca.1,899 m to ca. 1,200 m at the margins of the plain, and then gently to ca. 1,180 m above sea level in the south central portion.The flanks of the mountain chain are discharge points for numerous rivers and streams flowing dendritically into the plain and recharging the Bamendjin dam (Figure 1).Details on the drainage are described by Ndzeidze (2008).According to Neba (1999), the area has an Equatorial monsoon climate type characterized by high rainfall and relative humidity.This is attributed to the effect of the mountainous relief that characterises much of the WHC (Neba, 1999;Molua and Lambi, 2006).Two distinct seasons are differentiated, a long rainy season (mid-March to mid-November) and a short dry season (mid-November to mid-March).Mean annual rainfall and temperature, in 2012, were 1,540 mm and 26°C, respectively (Wirmvem et al., unpublished data), but with cooler climate in adjacent high elevations (Ndzeidze, 2008;Ndenecho, 2009).The heaviest rainfall occurs from July to September and often results in floods due to the poor drainage (Ndzeidze, 2008).
Geologically, the area is a shallow Cenozoic 'sedimentary' basin, underlain mainly by a consistent Precambrian granitic basement (Figure 1).This basement forms part of the tectonically inactive African shield consisting mostly of Precambrian Basement Complex rocks (Neba, 1999;Marzoli et al., 1999).Chemical weathering of the basement and surrounding volcanic rocks has produced thick unconsolidated sediments, mainly of clay to sand sizes.The basement is largely covered with these sediments, but outcrops in certain areas.The surrounding volcanoes are represented by voluminous Q-trachytes and minor rhyolitic ignimbrites with slight to moderate alkaline basalts and minor basanites (Marzoli et al., 1999).The voluminous Q-trachytes are enriched in alkali feldspars (Na 2 O+K 2 O).In-situ weathered basement and alkali-rich fluvial sediments constitute the aquifer material in the plain.From the margins of the plain towards the south-central portion, there is a progressive change from sandy to clayey sediments.The unconsolidated sandysediments form the groundwater aquifer system.This explains the concentric model of human settlement around the dam water (Ndzeidze, 2008) tapping groundwater through wells from this aquifer.The poor microbial quality of water with consequent water-borne infections (Wirmvem et al., 2013) has prompted the chlorination of water in wells by some members of the community.

Water sampling and physico-chemical measurements
Water was collected from 72 water sources in January 2012 (dry season) within 13 communities (Figure 1).The sources comprised 32 open wells and 14 'boreholes' (hereafter termed groundwater); and 14 streams, 4 rivers, 5 dam waters, and 3 lakes (Lakes Bambili, Oku, and Ber) (hereafter called surface water).The sample location points (Figure 1) and altitudes were noted in the field by using a Garmin Vista CX GPS.Four sporadic rainfall events in the plain and its environs were also sampled.Field measurements and water sampling followed methods described by Karklins (1996).New 100 ml narrow mouth and low-density polyethylene (LDPE) bottles were properly rinsed with distilled water and dried in the Laboratory (in order to avoid contamination) and used for sampling the water.At each site, water was collected into a plastic bucket after thorough rinsing with water to be sampled.Groundwater from active wells was collected using a rope tied to a bucket and/or permanently installed hand-pumps (after several minutes of pumping).Surface water was sampled as far as possible from the edges of the water bodies and as deep as possible along the flow path.The water collected was then filtered directly into 2 LDPE bottles using a plastic syringe, fitted with a 0.20 μm filter after thorough rinsing with the same water that was sampled.Collected samples were properly capped and preserved in a cooler container containing ice blocks.Water for major ion analyses was put in one bottle and water for alkalinity determination in a separate bottle.The following physical parameters: activity of hydrogen ion (pH), electrical conductivity (EC), total dissolved solids (TDS), and water temperature were measured immediately from the remaining collected water using a multipurpose HI991300 hand held pH/EC/TDS/C Hanna Meter.The meter was calibrated before and during the field campaign using buffer solutions recommended by the manufacturer.The depths to standing water level and total depth of water column were determined with a water level indicator.Alkalinity/bicarbonate (HCO3 -) was determined within 8 hours of sampling by acid titration from a volume of 0.02 M HCl, added to the sample and stirred with an automatic stirrer until the end-point, marked by a pH of 4.5.

Chemical analyses
Chemical analyses were carried out at Tokai University, Japan within two weeks of sample collection.Major cations: Na + , K + , Mg 2+ and Ca 2+ were determined by a Flame Technique in a High Resolution Continuum Source AAS (ContrAA 700) as described by Welz et al. (2006).Samples with EC values >100 μS/cm were diluted 5 or 10 times to acquire an absorbance within the range of the used standards for Na + , K + and Ca 2+ .The anions: F -, Cl -, NO3 -, SO4 2-, PO4 3-, NO2 -and Br -were determined by using an Ion Chromatography (Dionex ICS-900) while dissolved Silica (Si(OH)4), hereafter considered as SiO2, was determined by using an ICP-AES.The reliability of the chemical analyses was verified by using an ionic balance error (IBE) equation (Appelo and Postma, 2005).The values were within ±5 %; hence, suitable for geochemical interpretations.

Physico-chemical composition of groundwater and surface water
Table 1 presents results of the 72 water samples with a statistical summary shown in Table 2.A statistical summary of rain data is presented in Table 3. Mean depth to the water table (<27 m) and depth of the water column (<8.4 m) were low, suggesting a shallow aquifer system.The mean temperature of the water was closer to the mean atmospheric temperature of 26°C, suggesting present day climatic influence and recharge.Ground-water had a lower and narrow range in pH with a mean of 5.9, moderately acidic, while surface water had a relatively higher and wider range in pH with a mean of 6.8, slightly acidic.About 88% of the 72 samples had pH<7, suggesting an acidic aquifer system.The EC and TDS values of all water sources were low with mean values of 60 µS/cm and 39 mg/l, respectively suggesting lowmineralised and freshwater.Major ion concentrations, like TDS, were low (Table 2) and in the order (meq/l) Na 2+ > Ca 2+ >Mg 2+ >K + for cations and HCO 3 ->>NO 3 -≥Cl ->SO 4 2for anions.The small range in their concentrations and higher dispersions around their respective mean values as reflected by the small STD values, except for NO 3 -and Cl -(Table 2), suggests a unique source of most ions and a single aquifer system.Mean concentration of SiO 2 was higher than that of individual cations but less than the mean of HCO 3 -(Table 2).This suggests that SiO 2 was released more than the cations into solution.The high SiO 2 suggests the abundance of silicate minerals (Srinivasamoorthy et al., 2008) in agreement with the dominant granitic basement (Figure 1) and as observed in similar terrains by Njitchoua and Ngounou-Ngatcha (1997) and Fantong et al. (2009).
As expected, the mean TDS increased progressively from a minimum in rainwater (20 mg/l) to surface water (30 mg/l) and a maximum in groundwater (44 mg/l).Similarly, mean relative concentrations of dissolved ions showed a progressive increase from the lowest in rainwater, to surface water, and the highest in groundwater (Figure 2a).These results suggest chemical evolution from rain to surface water and groundwater.The relatively high TDS in groundwater indicated the influence of the unconsolidated sediments and granitic bedrock material in chemically enriching the water as it percolates through the under-saturated zone (or flows through the aquifer).The small difference in mean TDS of groundwater and rainwater by only 53 % is likely because these are early rains, usually enriched in dissolved ions (Takounjou et al., 2011).The low major ion concentrations in groundwater depict low water-rock interactions in the granitic basement, short residence time, the shallow nature of the aquifer and its acidic nature (Edmunds and Smedley, 1996).Similar low pH and major ions have been reported from granitic basement complex aquifers in Ghana (Adomako et al., 2011) and Nigeria (Edet et al., 2011).

Hydrochemical facies
The meq/l concentrations of major ions plotted on a pie chart (Figure 2b), showed that Na + , HCO 3 -and to some extent Ca 2+ were the major ionic species.Low K + may be Wirmvem et al. 521 due to its low geochemical mobility (Hem, 1989;Srinivasamoorthy et al., 2008) while low Mg 2+ is possibly due its to low content in rocks of the study area.On a Piper diagram (computed using GW_Chart Version 1.23.7.0), most samples plotted in the field of alkali metals (Na+K) (Figure 3).This indicates the dominance of alkali metals in the groundwater aquifer system.The water types and their proportions are shown (Figure 3).The observed spatial distribution of the water types (Figure 1) showed no discernible pattern from Ca-HCO 3 to Na-HCO 3 .Given the gentle relief in the plain, this suggests a mixed aquifer system as revealed by the mixed Na-Ca-HCO 3 water.Similarly, no clear pattern, in water type, was observed from surface water to groundwater (Figure 3).Nonetheless, there is chemical evolution from Ca+Mg+SO 4 +Cl (rain), to Ca-HCO 3 (largely groundwater and lakes), and to Na-HCO 3 (basically groundwater, streams, rivers, and dam waters) as indicated by the dotted arrow (Figure 3).It is probable that the dissolution of Na + into surface waters from alkali-rich trachytic rocks in the surrounding highlands tends to mask the chemical evolution.The presence of Ca-HCO 3 water in groundwater within the plain, which is sporadic, in surface waters (streams and rivers) discharging from the highlands suggests that the main recharge of groundwater may not necessarily occur at higher altitudes but rather through direct precipitation to the surface and subsequent percolation through the under-saturated zone into the aquifer.

Origin of solutes and hydrochemical controls
The major ion chemistry of groundwater and compositional relations among ionic species can reveal the origin of solutes and processes that generated an observed water composition (Hounslow, 1995;Demlie et al., 2007;Cendon et al., 2011;Tay, 2012).Table 4 shows correlation coefficients between the ionic species in water.The good positive correlation of NO 3 -with Cl -suggests their anthropogenic origin (Demlie et al., 2007;Fantong et al., 2009).Main potential sources of NO 3 -are numerous shallow pit toilets and oxidation of organic matter given the proximity (>15 m) to shallow groundwater and agricultural dominance, respectively.Fonge et al. (2012) identified relatively high concentrations of NO 3 and PO 4 3-in soil water (<5 cm deep) from rice farms in the study area.The low content of NO 3 -and the absence of PO 4 3-in groundwater suggest their sorption on organic colloids (in the under-saturated source of zone) or the location of rice fields rather in low-lying swampy areas.

The likely Cl
-is atmospheric inputs given the positive correlation with Na + and disinfection of wells through chlorination as noted in the field.Sulfur occurs in a reduced form in igneous rocks and is oxidized to SO 4 2when weathered in contact with aerated water (Hem, 1989).The significant positive correlations of SO 4 2-with K + , Ca 2+ and HCO 3 -possibly suggest its igneous origin in  agreement with the geology.Fluoride showed a distinct positive correlation with Na + relative to Ca 2+ (Table 4).
The likely source of F -in the granitic basement is fluorite, CaF 2 , the most common F --bearing mineral (Edmunds  and Smedley, 1996).However, the released Ca 2+ is possibly exchanged for Na + on clay surfaces, resulting in a relative increase in F -and a decrease in Ca 2+ , hence, the weak correlation between Ca 2+ and F -.
High F -concentrations in groundwater have been identified in North Cameroon, along the CVL, in a graniticalkaline environment from fluorite, amphiboles and micas (Fantong et al., 2009).The low concentrations of F -in groundwater of the study area may be due to its acidic nature which renders F -immobile (Hem, 1989;Edmunds and Smedley, 1996).For example, the highest value of F - (0.39 mg/l) was identified in BH29 ('borehole') with a pH of 6.4.The strong positive correlations of HCO 3 -with Ca 2+ , Mg 2+ , K + and SO 4 2-further indicate similar origin of ions.Given the Na-HCO 3 dominance (Figure 3), the weak positive correlation of Na + with HCO 3 -suggests that varied processes accounts for its concentration in solution.
Table 5 shows a summary of computed ratios between some ionic species.The observed Na + /Cl -molar ratios were high, relative to that of sea water, 0.86 (Millero et al., 2008).A Na + /Cl -molar ratio approximately equal to 1 is usually attributed to halite dissolution, whereas >1 is typically interpreted as reflecting Na + released from silicate weathering reactions (Fisher and Mullican, 1997;Cendon et al., 2010).Therefore, the excess Na + as further shown (Figure 4) is likely from silicate weathering, possibly feldspar mineral dissolution.On a plot of total cations, Na + +K + +Ca 2+ +Mg 2+ (TC) versus alkalinity, most of the 72 samples (94 %) plotted on or near the 1:1 line (Figure 5a).The high mean HCO 3 -/∑anions ratio in the water (Table 5) and the 1:1 relationship (Figure 5a) suggest silicate or carbonate weathering (Hounslow, 1995;Fisher and Mullican, 1997;Kim, 2002), but acid waters (as prevailed in the study area) usually results from dissolution of non-carbonate rocks such as granites (Edmunds and Smedley, 1996), the dominant rock in the  area.
Based on the geology and observed correlations between ionic species, the main mineral assemblage in the granitic basement under dissolution is possibly the plagioclase mineral series, with two end members, Albite (NaAlSi 3 O 8 ) and Anorthite (CaAl 2 Si 2 O 8 ).In the dissolution process, which involves the release of SiO 2 and HCO 3 into solution and the formation of clay minerals (Equation 1), Ca 2+ is preferentially released into solution more than Na + (Clayton, 1988;Deutsch, 1997).As expected, Ca 2+ would be greater in groundwater, but as earlier observed, Na-HCO 3 dominated Ca-HCO 3 waters.Calculated Weathering Index (R E ) using the formula of Tardy (1971) are shown in Table 5.About 88% of samples had R E values <2, while 12% had values <2.Accordingly, the weathering process in the study area resulted mainly in montmorillonite clay (R E >2) and minor kaolinite (R E >2) (Tardy, 1971).Montmorillonite has a high cation-exchange capacity (Appelo and Postma, 2005) and will probably result in ion exchange of Ca 2+ in solution for Na + in rocks given the freshwater aquifer system (Hem, 1989) as des-cribed in this equation (Njitchoua and Ngounou-Ngatcha, 1997): In order to determine the significance of base-exchange in enriching the water chemistry, a plot of Ca 2+ +Mg 2+ -HCO 3 --SO 4 2-versus Na + -Cl -was examined.If cation exchange is a significant groundwater composition-controlling process, the relation between these two parameters will be linear with a slope of -1.0 (Fisher and Mullican, 1997;Adomako et al., 2011;Tay, 2012).About 89% of the 72 samples plotted along the line with a slope of -0.99 (Figure 6a).The 9 samples that diverted most from the line had relatively anomalous concentrations of cations or anions.A plot excluding these samples gave atrolling process, the relation between these two parameters will be linear with a slope of -1.0 (Fisher and Mullican, 1997;Adomako et al., 2011;Tay, 2012).About 89% of the 72 samples plotted along the line with a slope of -0.99 (Figure 6a).The 9 samples that diverted most from the line had relatively anomalous concentrations of cations or anions.A plot excluding these samples gave a slope of -1.03 with an improved linearity (Figure 6b), indicating that ion exchange reactions occurred in a greater portion of the aquifer.The nature of ion exchange was determined by calculating Chloro-Alkaline Indices (CAI) (Schoeller, 1965).All samples gave negative values from CAI 1 and CAI 2 (  and Ca 2+ in groundwater were being exchanged with Na + and K + in rocks; hence, favouring cation-anion exchange reactions (chloro-alkaline disequilibrium) (Gupta et al., 2009).The Na + enrichment (addition) by ion exchange probably explains its observed weak positive correlation with other cations and SO 4 2-.Thus, besides the chemical dissolution of plagioclase feldspar minerals, cation-exchange reactions between water and montmorillonite in the aquifer also play a role in modifying the water composition.
The mean P CO2 in the groundwater (Table 5) was higher than the mean atmospheric value of 0.0003 atm (Appelo and Postma, 2005), suggesting that the major source of CO 2 governing the above reactions was probably from the oxidation of organic matter in the undersaturated zone (Karanth, 1987;Appelo and Postma, 2005).The high P CO2 in groundwater are consistent with findings along the CVL (Njitchoua and Ngounou-Ngatcha, 1997;Fantong et al., 2009;Ako et al., 2012), which has been attributed to the same process.Unlike in these related studies, groundwater in the study is comparatively acidic.The acidic nature probably results from reduction in acid buffering capacity of infiltrating rainwater and leaching of organic constituents from soils into the shallow groundwater.

Suitability for drinking and irrigation
The drinking water guidelines of the WHO (2004) were used to evaluate the quality of water in the study area for human consumption.Most water sources had a cream colour, likely due to suspended organic colloids.Regardless of the cream colour, the range of all major ions and TDS were below the maximum values of the guidelines for drinking water.However, the F -content was low (0 to 0.39 mg/l) with a mean of 0.08 mg/l.At low concentrations (<0.2 mg/l) in drinking water, dental caries may result (Edmunds and Smedley, 1996).The water was soft, that is, low total hardness (Table 5) given the low Ca 2+ and Mg 2+ .Their concentrations were mostly below the minimum required: 10 mg/l (Mg 2+ ), 20 mg/l (Ca 2+ ) and 60 mg/l total hardness (Kozisek, 2005).Evidence is now available to confirm the existence of health risk from drinking water deficient in Ca 2+ and Mg 2+ (Rylander et al., 1991;Dissanayake and Chandrajith, 1999;Kozisek, 2005).Thus, the general water softness could be a health problem since soft water has been linked to cardiovascular ailments (Kozisek, 2005).About 69% of the water sources had pH values below 6.5.For effective disinfection of water with chlorine, the pH should preferably be less than 8, but lower pH water (<6.5) is likely to be corrosive (WHO, 2004).High trace-metal contents, a health problem, are also likely from such low pH waters (Edmunds and Smedley, 1996), hence, a need to assess potentially harmful trace element.The low TDS in groundwater are consistent with the observation that most exploited groundwater resources in the weathered Precambrian basement rocks of Africa area low in dissolved ions and generally suitable for consumption (Adelana et al., 2011).However, the microbial quality of groundwater in the tropi-cal Africa is a serious health issue (Roche, 1993) as reported in the study area (Wirmvem et al., 2013) and elsewhere (Katte et al., 2003).

Conclusion
A hydrochemical study, in the Ndop plain, to determine the physico-chemical characteristics, controls on water chemistry and suitability for drinking and irrigation has been carried out by analysing 72 water samples.The first physico-chemical baseline data of groundwater and surface water for water quality assessment, management and monitoring of changes in the hydrochemical system has been made available in the area.About 69% of the water sources had pH values below 6.5, classifying the water as barely acidic.Therefore, careful attention is necessary at all stages of pipe borne water treatment and supply in the area as this could lead to corrosion and scale formation in pipes.The low mean EC (60 µS/cm) and TDS (39 mg/l) classify the water as low-mineralised and freshwater.Major ion concentrations were low, and below maximum values of the WHO standards for drinking water, but the low Ca 2+ and Mg 2+ classify the water as soft (<<57 mg/l CaCO 3 ).This could be a health problem given the relationship between soft water and cardiovascular ailments.The low F -content in the water sources (<0.39 mg/l) could lead to dental caries.In spite of these, the water sources are considered suitable for drinking as per WHO standards.
The relative abundance of major cations and anions in water (meq//l) is Na , respectively.The main water types and proportions are Na-HCO 3 (53%), Ca-HCO 3 (35%) and mixed Na-Ca-HCO 3 (11%).Alkali metals (Na+K) exceed alkali earth metals (Ca+Mg) indicating their dominance in the aquifer system.The major controls on water chemistry are incongruent dissolution of silicate minerals (mainly plagioclase feldspars) and cation-exchange of Na + in rocks for Ca 2+ in solution.The dissolution is governed by biogenic CO 2 and results to montmorillonite clay, which Wirmvem et al. 529 enhances the ion exchange.The low NO 3 -and general absence PO 4 3-in groundwater suggest their sorption by organic matter in the under-saturated zone.The low EC (TDS) and major ions in groundwater depict low solubility of the silicates, short residence time of groundwater, shallow nature of the aquifer and its acidic nature.The water sources are suitable for irrigation.Potentially harmful trace elements should be investigated given the acidic nature of the water.Hydrochemical investigations in the rainy season should also be carried out, to determine the seasonal controls on water chemistry.

Figure 1 .
Figure 1.The geology and drainage map of Ndop Plain (modified after Ndzeidze, 2008) showing sample points/spatial distribution of water types.

Figure 2 .
Figure 2. A, Relative mean concentrations of silica and cations (mg/l) in various water sources of the study area including rain.Si was not measured in rain; Si>Na>K>Ca>Mg.B, Pie chart of mean concentrations of ions (meq/l), showing high Na+ and dominant HCO3-in water sources of the study area.

Figure 3 .
Figure 3. Piper's diagram showing the water types and proportions in the Ndop Plain.Alkali metals (Na+K) exceed alkali earth metals (Ca+Mg).Weak acids (CO3+HCO3) greatly exceed strong acids (SO4+Cl).The dash line in the diamond section indicates the path of chemical evolution from Ca-Mg-SO4-Cl (rain), to Ca-HCO3 (groundwater, lakes and dam water), and to Na-HCO3 (groundwater, streams and rivers).Ca-Na-HCO3 represents a transition (mixing zone) from Ca-HCO3 to Na-HCO3 water.

Figure 4 .
Figure 4. Plot of Na+ as a function of Cl-showing an excess of Na+ over Cl-in water sources.

Figure 5 .
Figure 5. A, Plot of total cations as a function of alkalinity showing a 1:1 relationship in water sources of the study area.B, Plot showing deficiency in Ca 2+ +Mg 2+ in relation to SO4 2-+HCO3 -in water sources.

Table 2 .
Statistical summary of physico-chemical data of groundwater and surface water in the study area.

Table 3 .
Physico-chemical data of January 2012 sporadic rainfall events in the Ndop Plain and its Environs.
EC = uS/cm, TDS and all ions = mg/l, Na/Ca

Table 4 .
Pearson correlation coefficients of all analysed water sources (n=72).

Table 5 .
A statistical summary of ratios between some ionic species in the analysed water.

Table 5 )
. This indicated that Mg