Introduction

Studies on heavy metal pollution in aquatic ecosystems have been a major environmental focus especially in the last decade (Weis and Weis 1995; Ongeri et al. 2009a, b; Öztürk et al. 2009; Omwoma et al. 2010) particularly due to their potential toxic effects and ability to bioaccumulate in aquatic ecosystems (Censi et al. 2006; Batvari et al., 2007). Surface waters are most exposed to pollution due to their easy accessibility for disposal of wastewaters. Anthropogenic influences such as urban, industrial and agricultural activities, increasing exploitation of water resources as well as natural processes which include precipitation inputs, erosion and weathering of crustal materials degrade surface waters and damage their use for drinking as well as for industrial, agricultural and other purposes (Simeonov et al. 2003). Sediments serve as reservoirs for heavy metal residues and play a significant role in their remobilization and distribution in aquatic systems through interactions between water and sediment under favorable conditions (Klavinš et al. 2009). In aquatic ecosystems, water will have lower concentrations of heavy metals whose magnification increases in the recipient sediment and biota (through food chain transfer) (Öztürk et al. 2009). However, high concentrations of heavy metals in the aqueous phase indicate more recent input into the aquatic environment. Since heavy metals in the aquatic systems are usually predominantly associated with sediment, their concentrations in sediment are more sensitive and accurate indicators of contamination of the aquatic system when compared with their concentrations in the aqueous phase. Among the animal species, fish are the inhabitants that cannot escape from the detrimental effects of these pollutants (Olaifa et al. 2004) and are therefore widely used to evaluate the health of aquatic ecosystems as pollutants which build up in the aquatic food chain can accumulate and cause adverse effects including mortality (Farkas et al. 2002). Due to lack of adequate capacity, heavy metal contamination monitoring has been rare in Lake Naivasha in Kenya despite frequent calls by researchers for such monitoring. Cadmium (Cd), lead (Pb) and copper (Cu), selected in this study, are regarded as serious pollutants of aquatic ecosystems because of their environmental persistence, toxicity and ability to be incorporated into food chains (Forstner and Wittmann 1983) and hence the required obligatory monitoring of their concentrations by WHO and FAO (Staniškienė et al. 2009). In addition Pb and Cd are some of the heavy metals linked most with human poisoning. Cu is required in small amounts as an essential element, but is toxic in certain critical doses. Although some research has been done in Lake Naivasha, including speciation studies on some of the heavy metals in selected sites around the lake, including the Malewa river Mouth, in 2003 (Kamau et al. 2007) and analysis of As, Co, Cr, Cu, Ni, Pb and Zn to determine the background geochemical contributions (Tarras-Wahlberg et al. 2002), none has directly correlated heavy metal concentrations in sediments, water and fish. Further, no studies have addressed the potential impact of the recent commercial horticultural/floricultural activities along the lake with respect to these heavy metals. The objective of this study was to determine the concentrations of Cd, Cu and Pb in surface water, sediment and C. carpio muscle tissue samples from selected sites within Lake Naivasha, Kenya, and to assess the potential impact of the recent anthropogenic activities on their concentration levels.

Methodology

Study area

Lake Naivasha and its catchment characteristics

Lake Naivasha (0.45°S, 36.26°E), located at an elevation of 1,890 m above sea level, is a shallow freshwater lake in Kenya’s Rift Valley, with an approximate area of 100 km2 and a depth of 3–6 m (Hickley et al. 2004). It is a Ramsar site threatened by recent extensive human development activities in its catchment. Some of the activities that threaten the lake’s ecosystem include the disposal of untreated effluents from industrial and human settlements, discharge of contaminated irrigation wastewater from small- and large-scale horticultural farms and inflow of rivers from its upper catchment where wheat, fodder, maize and cattle farming is practised (Kitaka et al. 2002; Becht et al. 2005; Mireri 2005; Owiti-Otianga and Osewe 2007). The effluents which get into the lake vary in composition and quantity and contain contaminants such as heavy metals. Some of the main sources of heavy metals into the lake include natural geochemical processes and heavy usage of agrochemicals such as fertilizers and pesticides in its catchment (Tyagi and Mehra 1990; Alloway and Aynes 1993). The lake is shallow and has no surface outlet and therefore there is potential accumulation of contaminants in sediment, water and biota. Fish, particularly the benthivorous C. carpio, which forms 93% by composition of all fishes in the lake, is likely to be impacted by these contaminants and therefore provides a good biological indicator (Hickley et al. 2004; Batvari et al. 2007). The lake gets its water from three rivers; Malewa (contributing about 90% of surface inflow), the Gilgil and Karati (Figs. 1, 2). Malewa is perennial, has tributaries originating from the slopes of Nyandarua (3,700 m above sea level), passes through agricultural areas and discharges directly into the lake. Some of its sediment-laden plumes reach up to about 500 m into the lake (Everaard et al. 2002). Gilgil, arising from 2,700 m above sea level, is also perennial and passes through Gilgil town carrying effluent from municipal wastes and pastoralist and agricultural activities. Karati enters the lake through various distributaries/channels which are not perennial. The river starts from Kinagop plateau (2,600 m above sea level) and passes through a catchment with various agricultural farming activities. The natural geochemical and erosion processes, wash off during rainfall as well as anthropogenic disturbances in the upper catchments of these rivers influence the heavy metal loads in the northern part of Lake Naivasha. Distribution within the lake water due to water currents is also evident and heavy metal-laden particulate matter can be transferred to other parts within the lake. The southern part of the lake (Fig. 1) is exposed to human activities including municipal and industrial waste disposal, abstraction of water for irrigation and intensive horticultural farming and use of fertilizers and pesticides which would influence heavy metal concentrations and distribution in the lake. The physical and biological changes as well as the various activities within the Lake Naivasha catchment which form a strong background of this study, have been described in detail in numerous earlier publications including those of Harper et al. (1995), Becht et al. (2005), Everaard et al. (2002), Everaard and Harper (2002), Kitaka et al. (2002), Tarras-Wahlberg et al. (2002) and Mireri (2005). In particular, some of the changing characteristics such as drop in water levels through imbalance in evaporation losses and depletion of the papyrus resulting in lack of water filtration can influence heavy concentrations as well as their distribution and bio-accumulation in various trophic levels in the food chain within the lake.

Fig. 1
figure 1

Socio-economic activities surrounding Lake Naivasha (WWF 2010)

Full size image
Fig. 2
figure 2

Lake Naivasha: the study area

Full size image

The catchment of Lake Naivasha is approximately 3,400 km2 in area (WWF 2010) and has a mean annual temperature, varying with altitude, ranging from 25°C on the lake shore (1,890 m above sea level) to 16°C in the Aberdare mountains (3,700 m above sea level). The catchment has an annual bimodal rainfall pattern, with two rainy seasons in April–May (long rains) and October–November (short rains) and the annual average rainfall ranges from 1,350 mm in the higher altitude upper catchment to 600 mm near the shores of the lake. The sampling period in this study (late March–early June 2010) captured the long rainy season when agricultural activities were at their peaks and there was surface runoff carrying contaminants into the lake

The sampling sites

The sampling points (Fig. 2) were purposively chosen based on the possible routes of heavy metal residues into the lake and included undisturbed areas which served as control. Areas with intensive agricultural activities and wetland discharge outlets around the lake represented disturbed areas due to the potent influx of contaminants into the aquatic ecosystem. Three chosen sites represented the discharge points for the floriculture farms which form a major agricultural activity around the lake and included Sher Karuturi, Kasarani and Oserian discharge outlets, respectively. The Mouths of Rivers Malewa and Karati served as principal sample sites together representing upstream natural processes, anthropogenic activities and possible contamination sources from their catchments as they channel used surface water into the lake. In their course downstream, many small-scale farmers divert water from portions of the two rivers into their farms for irrigation purposes and could potentially contaminate their waters with agrochemical residues which they eventually discharge into the lake. Hippo Point and Crescent Island characterized by minimal anthropogenic activities; mostly boating and fishing, served as undisturbed areas and hence control sites.

Sampling and sample preparation

Samples were randomly collected in triplicates from the seven sampling sites along Hippo Point, Kasarani, Mouth of Malewa, Mouth of Karati, Crescent Lake, Sher Karuturi and Oserian discharge outlet. In situ measurements were also undertaken for water quality parameters namely dissolved oxygen (DO), pH, water temperature (°C), electrical conductivity (μS/cm) and total dissolved solids (TDS) using the Hydro lab YSI meter model 85. A total of 84 C. carpio were assayed.

Sampling bottles were pre-cleaned thoroughly with detergent, soaked overnight in 6% nitric acid and thereafter rinsed well with distilled water. These were then stored in a dust-free place to avoid contamination. In the lake, a sample of 500 mL of surface water was taken by immersing the bottles and lifting them up. Immediately after, 4 mL of nitric acid was then added to the water samples to adjust the pH and maintain the heavy metals in solution. In the laboratory, the water samples were filtered into 100 mL pre-cleaned plastic bottles through Whatman No. 42 to remove any suspended organic materials and further prevent microbial growth, flocculation and adsorption on container surfaces. The water samples were then stored at 4°C in the refrigerator awaiting analysis.

Bottom surface sediment samples were taken from the sites (at an average depth of 3 m) in triplicates using a stainless steel Ekman grab sampler, stored in PTFE polythene bags to avoid contamination for transportation to the laboratory. In the laboratory, samples were dried in the oven overnight at 110°C; before crushing in a mortar and pestle and sieving using a 2 mm mesh to remove plant parts and debris. One-gram (dry weight) portions were then weighed into a 50-mL Pyrex digestion tube with 10 mL mixture of concentrated nitric acid and concentrated hydrochloric acid (4:1, aqua regia). The mixture was digested for 3 h in a digestion block at 100°C, the contents filtered through a 0.45 μm Whatman No. 42 filter paper into a 50-mL volumetric flask and then made up to volume with double-distilled water after addition of 1.5 mg/mL of strontium chloride (analytical grade, SrCl2·6H2O). The extracts were analyzed for Cd, Cu and Pb by FAAS at wavelengths of 228.8, 324.7 and 283.3 nm, respectively, and calibration standards in the range of 0.1–0.6 mg/kg.

Freshly caught C. carpio were bought from local fishermen who were fishing at the different sampling sites and placed in a cool box before being taken to the laboratory for further processing and analysis. Research scientists from Kenya Marine and Fisheries Research Institute (KEMFRI) were consulted in the field for purposes of C. carpio identification. In the laboratory, C. carpio samples were washed with distilled water, scales removed and only muscle tissues without bone were taken for extraction and analysis. Two portions of 10 cm2 area of wet fish muscle were cut and homogenized in a pestle and mortar. The digestion mixture for fish sample analysis was prepared consisting of 0.42 g selenium powder (catalyst), 14 g lithium sulphate (to improve the FAAS efficiency), 350 mL hydrogen peroxide and 420 mL sulphuric acid in an ice bath (Okalebo and Gathua 1993). Then, 10 mL of the digestion mixture was added to the 2 g wet weight (ww) of fish muscle in the digestion tubes for digestion in a digestion block for 2 h at 360°C until clear solutions were obtained. Digests were then left to cool to ambient temperatures (25°C) and the resulting solutions filtered using 0.45 μm Whatman No. 42 filter paper and volume topped to 100 mL using distilled water for analysis as explained above.

Chemical and data analysis

A flame atomic absorption spectrophotometer (FAAS), a Thermo Jarrell Ash AA/AE spectrophotometer S11 model was used in the analysis. All sample preparations and analysis were carried out using standard methods of analysis. The concentrations of the metals were assayed in triplicates by use of FAAS with acetylene flame. The accuracy of the instrument was checked by triplicate analysis of same samples. In addition, a standard and a blank sample were run after every seven samples to check instrumental drift. A series of standards were prepared for instrument calibration by serial dilution of a working solution (100 mg/L) prepared from analytical grade stock solutions (1,000 mg/L) obtained from BDH Poole and Sigma Aldrich. The percentage recoveries were determined for all metals. Calibration curve method was used to quantify the heavy metal concentrations. The concentrations of the heavy metals in various matrices were presented as arithmetic mean with standard deviation (mean ± standard deviation). All statistical analyses were done at p = 0.05 significance using ANOVA SIGMA STAT.

Results and discussion

Water quality

Table 1 shows the results of the water quality parameters that had been determined in situ at the study sites. Data show an average of four measurements taken at each of the seven sampling sites. The recorded limnological parameters were within the required WHO drinking water desirable limits in terms of pH range, electrical conductivity, total dissolved solids and dissolved oxygen concentration, respectively. The main observations that can be mentioned include lower dissolved oxygen concentration at Kasarani and lower transparency of the water at Mouth of Malewa which could be due to presence of algae, as this part of the lake is least disturbed. The limnological data are also compared with those of other freshwater systems in Kenya. These parameters indicate that Lake Naivasha is suitable for aquatic life and for human use. The lake is slightly alkaline but the pH level is comparable with that recorded earlier by Ochieng et al. (2007). The electrical conductivity represents the quantity of dissolved salt in the freshwater and has been used by researchers as a general indication of their potential productivity. The conductivities in Kenya are known to be influenced by the carbonate salts which account for up to 80% of the conductivity in most cases, the remaining electrolytes including phosphates, and nitrates existing in relatively lower concentrations (Ochieng et al. 2006). It is expected that the higher the electrical conductivity of the water, the higher the water pH (Ochieng et al. 2006) and this is also evident in the other closed Rift Valley lakes including Lakes Nakuru, Bogoria and Elementaita. These limnological parameters can also vary from one location to another on the same water body and therefore the slight differences seen in the values of these Lake Naivasha parameters are expected. The massive fish kills, observed in February 2010, caused by rapid changes in water quality including a decrease in oxygen levels to zero leading to hypoxic conditions as well as an increase in TDS from 242 and 453 mg/L, respectively, indicate the important need to avoid localized degradation of water quality (Gichuki et al. 2010). These rapid physico-chemical changes were reportedly caused by a drop in water level to 1.5 m and an inflow of floodwaters containing decaying vegetation, animal dung and other wastes during El Nino rains (Gichuki et al. 2010).

Table 1 Limnological parameters obtained during sampling in Lake Naivasha compared with those of other lakes in Kenya
Full size table

Heavy metal concentrations in water

Overall, high % recoveries of 88, 89, 87 (sediment), 89, 92, 90 (water) and 91, 93, 96 (fish muscle) were obtained. The results in Table 2 indicate that lead is the most predominantly detected metal in water samples but only in samples from the Mouths of Rivers Malewa and Karati and from Sher Karuturi and Oserian discharge outlets. The concentrations were, however, low and in the ppb range. Cu and Cd concentrations were only detected at the Sher Karuturi discharge canal indicating possible significant inflows from the flower farms. Most of the metal concentrations were below quantification limits (μg/L) of 100 (Pb), 30 (Cu) and 10 (Cd) as indicated in Table 2. The amounts of Pb, Cu and Cd at Hippo Point, Kasarani and Crescent Island were, as expected, below quantification limits. At Sher Karuturi, all the three metals, Pb, Cd and Cu were detected, whereas at Oserian outlet only Pb was detected, with Cd and Cu being below quantification limits. Elevated levels of Pb concentrations in water samples from Sher Karuturi and Oserian outlet are attributable to inflow from the flower farms. The high concentrations of Pb in water at these sites are of concern because Pb is among heavy metals, such as Hg, As, and Cd, detected frequently in fertilizers and linked most to human poisoning. These results also show the impact of anthropogenic activities upstream in the catchment of the lake. Although Cd and Cu were below quantification limits at the Mouths of Rivers Malewa and Karati, high Pb levels were detected at these sampling points which could indicate Pb coming in through wash off from upstream; most likely from the roads because of the use of leaded gasoline. The use of leaded gasoline was banned in 2006 in Kenya but Pb may still be persisting in the environment even though no new residues are likely to be added into the environment from this source now. Pb residues could also be coming from upstream bound to sediment and other particulates from natural sources and from anthropogenic activities including farming. High levels (μg/L) of Cd (range 3–40), Cu (range 5–100) and Pb (range 106–563), attributable to anthropogenic sources, have also been reported in L. Nakuru (Ochieng et al. 2007). The water at Kasarani, Hippo Point and Crescent was, as expected, found to be relatively clean with respect to the three heavy metals.

Table 2 Heavy metal concentrations in water and sediment
Full size table

The concentrations of the three metals in water were also compared with those of other lakes in Kenya as well as with Kenyan and WHO acceptable standards which indicated that Lake Naivasha water, especially at the Mouths of the Rivers Malewa and Karati and at Sher Karaturi and Oserian outlets, may be polluted with Pb since the concentrations in water samples from these sampling points are higher than the KEBS drinking water limits as well as the WHO highest desirable level in drinking water, the WHO maximum permissible level in drinking water and WHO threshold concentration for aquatic life tolerance, respectively (Table 2). The water concentrations of Pb at these contaminated sites were only comparable with Lake Nakuru but were above all the levels detected in most other Kenyan lakes (Table 2). Cu and Cd were, however, within acceptable limits because their concentrations were lower than the Kenyan and WHO endpoints and comparable with those of other Kenyan rift valley lakes.

Heavy metal concentrations in sediments

The heavy metal contents in sediments (mg/kg dry weight) are also presented in Table 2 with quantification limits of 0.1 (Pb), 0.01 (Cd) and 0.03 (Cu), respectively. The data indicate that Pb is the most concentrated followed by Cu and Cd. The highest Pb and Cu contents were found in sediment samples from the Mouth of Malewa River which could be due to the inflow of contaminated sediments from the catchment as explained in the previous section. The data also indicate that Cd and Cu concentrations in sediments are low which can be explained by assuming that most of the Cd and Cu inflows into the lake are still in dissolved form, still suspended in water and not settled at the bottom due to water turbulence and/or that most of the Cd and Cu is taken up by aquatic plants. The Crescent Lake and Hippo Point, as expected, had the lowest levels of these metals in the sediments, most likely due to lack of discharge canals and other inflows. However, the impact of anthropogenic activities is shown by the high concentrations of Pb. The mean heavy metal concentrations in all sediments showed wide variations (p < 0.05) among sites, with ranges (mg/kg) from 17.00–37.03 (Pb), 1.18–5.58 (Cd) and 2.99–8.578 (Cu).

The sediments from the Mouths of Rivers Malewa and Karati, Sher Karuturi and Oserian outlet indicated Pb concentrations above the TEC whilst Hippo Point, Crescent Island and Kasarani had concentrations below the LEL (Table 2). The TEC level marks the significance of this study as it is the concentration at which heavy metals are beyond permissible lowest effect level and can have significant effects on ecological components. The Mouths of Malewa and Karati are key sources of sediment deposition into the lake as they originate from the upper catchment areas, as described earlier, and are the major inlets into the lake. In addition, a major highway above the Malewa Bridge could also be a contributing factor with respect to motor vehicle Pb emissions and deposition on roadside soil. The Sher Karuturi and Oserian flower farm discharge outlets showed significant Cd and Cu metal loads as well, which could be attributed to the agrochemicals use, although Cu concentrations were below the LEL (Table 2). The Mouth of River Malewa still had the highest concentration of these metals in sediments showing that they emanate from natural processes through erosion and anthropogenic activities upstream.

Heavy metal concentrations in Cyprinus carpio muscle tissues

The results of the C. carpio assay presented in Table 3 indicate that Pb is the most predominant metal in the fish followed by Cd and Cu, respectively. The highest Pb content was found in C. carpio caught at the Mouth of Malewa river with the lowest concentration in C. carpio caught at Crescent Island. The ranges of heavy metals in muscle (in mg/kg wet weight) were 1.16–2.30 (Cu), 1.06–1.73 (Cd) and 5.12–58.16 (Pb).

Table 3 Heavy metal concentrations (mg/kg wet weight) in C. carpio muscle compared with other freshwater fishes from other lakes and with FAO acceptable standards
Full size table

The results also show that fish from Lake Naivasha are more exposed to these heavy metals than fish from Lake Victoria. The difference in contamination levels can be explained by the differences in the two species, L. niloticus and C. carpio, and differences in their feeding habits. However, considering water, sediment and fish concentration data, together, the results indicate that Lake Naivasha is more polluted in terms of Pb and Cd. The concentration levels of Pb and Cd in edible muscle of C. carpio are much higher than those in edible muscle of L niloticus from Winam Gulf of Lake Victoria and are also above the maximum allowable FAO limits for edible portions of fish.

Conclusions and recommendation

The study shows that Lake Naivasha, in some parts, is polluted with Pb, Cu and Cd which can be originating mostly from the lake’s upper catchment activities and the surrounding horticultural farming activities. The important potential pollution sources into the lake include the flower farms and Mouths of River’s Malewa and Karati. Based on the chemical composition of samples collected from Sher and Oserian discharge outlets, the horticultural farming sector is a significant source of Cu, Pb and Cd. This is thought to be due to the intensive use of agrochemicals containing these heavy metals. There is need, therefore, to carry out more investigations within the farms and in the canals in order to confirm this conclusion and initiate mitigation measures, if necessary. However, samples collected from Crescent Island, Hippo Point and Kasarani areas, which are not impacted by human activities, showed relatively low concentrations of Cu, Pb, and Cd.