Search

The large surface area of Lake Victoria (about 68,800 km2) makes it vulnerable to high atmospheric deposition of chemical pollutants. We present measurements of polychlorinated biphenyls (PCBs) from the lake's atmospheric environment. High volume air (24 h) samples were collected within the northern Lake Victoria watershed in Uganda over two periods; 1999–2004 [at Kakira (KAK) and Entebbe (EBB)] and 2008–2010 (at EBB only). Precipitation samples were also collected monthly during the 2008–2010 period at EBB. Analysis for PCBs was done using GC-μECD in a dual column approach. The ranges of ΣPCB concentrations in the KAK air samples were 154–462 pg m−3 (KAK 1999–2000), 26.7–226 pg m−3 (KAK 2003–2004), 27.0–186 pg m−3 (EBB 2003), 46.8–174 pg m−3 (EBB 2004), 19.2–128 pg m−3 (EBB 2008), 45.8–237 pg m−3 (EBB 2009) and 65.6–244 pg m−3 (EBB 2010). The di-, tri-, tetra- and penta-PCBs were predominant in air sample sets while the tetra- and penta-PCBs were predominant in precipitation samples. The mean flux of ΣPCBs in the precipitation samples was 26.9 ng m−2 (range of 14.8–41.5 and median of 27.5). Concentrations at EBB were lower than those reported elsewhere for urban sites in the East and Central African region. Multivariate analysis and analysis of air mass movements suggested influence of combustion sources on the PCB profiles from the region, especially, from the major East African urbanized regions.

Two commercially important fish species, Nile perch (Lates niloticus) and Nile tilapia (Oreochromis niloticus) belonging to different trophic levels were collected from the Napoleon Gulf and Thurston Bay in Lake Victoria. Polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and dioxin-like polychlorinated biphenyls (dl-PCBs) were extracted from the fish muscles and livers using the (13)C isotope dilution method, followed by multiple column chromatography clean-up. Analysis was achieved by a high resolution gas chromatography coupled with a high resolution mass spectrometer. The concentrations of analytes ranged from 0.07 to 0.59pgg(-1) fresh weight (fw) and 0.3-19.0pgg(-1) in L. niloticus and 0.06-0.18 and 0.2-15.7pgg(-1) in O. niloticus, for ∑PCDD/Fs and ∑dl-PCBs, respectively. Differences in congener concentrations were observed between the two fish species and study sites, and this was attributed to differences in feeding habits and trophic levels. World Health Organization-toxic equivalents (WHO-TEQs) were in the range 0.01-0.16pgTEQg(-1) for the PCDD/Fs and 0.001-0.74pgTEQg(-1) for the dl-PCBs. The TEQ values in the present study were lower compared to those of most fish samples reported in literature and were within permissible levels recommended by the European Union, implying that the fish was fit for human consumption.

Fresh and pasteurized milk samples from Kampala markets were analyzed for organochlorine pesticides using a gas chromatograph equipped with an electron capture detector. Five organochlorine pesticides, namely; aldrin, dieldrin, endosulfan, lindane, DDT and its metabolites were detected in the milk samples and confirmed with a gas chromatograph equipped with a mass spectrometer [GC-MS]. The mean values are expressed in mgkg(-1) milk fat (mf) basis. The mean concentration in the fresh milk (n=54) were: 0.026 ± 0.003 mg kg(-1) mf; 0.002 ± 0.0003 mg kg(-1), below the detection limit; 0.007 ± 0.003 mg kg(-1), 0.009 ± 0.002 mg kg(-1) milk fat for lindane, endosulfan dieldrin and aldrin, respectively. The mean concentrations of p,p'-DDE; p,p'-DDT and o,p'-DDT were 0.009 ± 0.002 mg kg(-1); 0.033 ± 0.007 mg kg(-1) and 0.008 ± 0.001 mg kg(-1) mf, respectively in the fresh milk samples. In the pasteurized milk samples (n=47), the mean concentrations recorded were: 0.008 ± 0.003 mg kg(-1), 0.025 ± 0.004 mg kg(-1), and 0.007 ± 0.001 mg kg(-1), respectively for p,p'-DDE; p,p'-DDT and o,p'-DDT. Alpha and beta-endosulfan recorded the concentration below the detection limit and the mean of 0.022 ± 0.001 mg kg(-1) mf, 0.005 ± 0.002 mg kg(-1) mf, and 0.006 ± 0.0002 mg kg(-1) mf, respectively for lindane, dieldrin and aldrin. Although, most of the residues detected were above the residue limits set by the FAO/WHO (2008), bioaccumulation of these residues is likely to pose health risks to the consumers of milk in Uganda.

Organochlorine (OC) residues were analysed in 117 sediment samples collected from four bays of the Uganda side of Lake Victoria. The sediments were collected with a corer at a depth of 0–20 cm, and extracted for OC residues using a solid dispersion method. The extracts were cleaned using gel permeation chromatography and analysed for pesticide residues using a gas chromatograph (GC) equipped with an electron capture detector. The results were confirmed using a GC equipped with a mass spectrometer (MS). A total of 16 OC residues, most of them persistent organic pollutants (POPs) were identified and quantified. The OC residue levels were expressed on an oven dry weight (d.w.) basis. Endosulphan sulphate, in the range of 0.82–5.62 μg kg−1 d.w., was the most frequently detected residue. Aldrin and dieldrin were in the ranges of 0.22–15.96 and 0.94–7.18 μg kg−1 d.w., respectively. DDT and its metabolites lay between 0.11–3.59 for p,p′-DDE, 0.38–4.02 for p,p′-DDD, 0.04–1.46 for p,p′-DDT, 0.07–2.72 for o,p′-DDE and 0.01–1.63 μg kg−1 d.w. for o,p′-DDT. The levels of γ-HCH varied from 0.05 to 5.48 μg kg−1 d.w. Heptachlor was detected only once at a level of 0.81 μg kg−1 d.w., while its photo-oxidation product, heptachlor epoxide, ranged between non-detectable (ND) to 3.19 μg kg−1 d.w. Chlordane ranged from ND to 0.76 μg kg−1 d.w. Based on the threshold effect concentration (TEC) for fresh water ecosystems, aldrin and dieldrin were the only OCs that seemed to be a threat to the lake environment.

Information on historical deposition of persistent organic pollutants (POPs) for African lakes is very limited. We investigated historical deposition trends and sources of POPs in sediment cores from Lakes Victoria (SC1), Bujuku (Buju2) and Mahoma (Maho2). The latter two lakes are situated in the Rwenzori mountain range in western Uganda. SC1 was taken from a central depositional area within the Ugandan part of the lake. Profiles in Buju2 and Maho2 were used as a reference for historical atmospheric deposition. For the post-1940 sediment deposits in SC1, average focusing factor-adjusted fluxes (FFFs) of ΣDDTs, polychlorinated biphenyls (ΣPCBs), hexachlorocyclohexanes (ΣHCHs) and chlordanes (ΣCHLs) were 390, 230, 210 and 120 ng m(-2) yr(-1). Higher fluxes of ΣDDTs, ΣPCBs, and ΣCHLs were observed in Buju2 and Maho2. The average FFF of HCB in Buju2 was the highest while the values for Maho2 and SC1 were similar. The endosulfan FFFs in SC1 were lower than in the alpine lake cores. Generally, Buju2 was a better reference for historical atmospheric deposition of POPs than Maho2 probably due to distortion of the latter's profile by Lake Mahoma's forested catchment. Profiles of p,p'-DDE, ΣCHLs and HCB in SC1 were consistent with atmospheric deposition while profiles of PCBs and HCHs were indicative of particle-bound loadings from additional sources. Profiles of endosulfans, DDTs, and chlordanes were consistent with influence of other factors such as anoxia, and dilution. Further studies of spatial resolution of historical deposition, especially in near-shore deposition areas of the lake are recommended.

Organochlorine (OC) residues were analysed in soil samples collected from Kihiihi sub-county, Kanungu District, which is located in south-western Uganda. Mabira Central Forest Reserve which is located in central Uganda was used as a reference site in this study. The samples were collected at 15-20 cm depths below the soil surface and extracted with organic solvents. The extracts were cleaned using florisil, and were analysed using a gas chromatograph (GC) equipped with an electron capture detector (ECD). The results were confirmed using a GC equipped with a mass spectrometer (MS). The levels of the OC residues in soil from Kihiihi sub-county varied from non-detectable (ND) to 59 microg kg(-1) dry weight. The frequencies of detection for p,p'-DDT, p,p'-DDE and p,p'-DDD in the total soil samples were 47%, 24% and 11%, respectively. The o,p'compounds were detected in the following frequencies: o,p'-DDT (23%), o,p'-DDE (19%) and o,p'-DDD (8%). The low p,p'-DDE/p,p'-DDT residue ratios (0.2-0.9) in Kihiihi samples suggest recent inputs of DDT in Kihiihi sub-county. The detection frequencies for dieldrin, endosulfan-alpha, -beta and -sulphate in the soil samples were 21%, 26%, 31% and 19%, respectively. Mabira Forest Reserve soils showed detection frequencies of p,p'-DDT (5%) and p,p'-DDE (9% of the soil samples), varying from ND to 9 microg kg(-1). Although the use of OCPs has been banned, our results show that they can still be detected in the environment. Their presence may be attributed to adulteration of pesticides which are not banned and also atmospheric deposition.

This paper presents results based on determination of residue levels of 1,1,1-trichloro-2,2-bis[p-chlorophenyl]ethane (DDT) and its metabolites in five fish species from Rwenshama landing site on Lake Edward, Uganda. The residue levels were analysed by using a GC-ECD and confirmed by GC-MS. The DDT residues detected in fish samples from Lake Edward were p,p'-DDT, o,p'-DDT, p,p'-DDE, o,p'-DDE, p,p'-DDD and o,p'-DDD. All the analysed samples of fish presented mean DDT residues ranging from non-detectable levels to 68 microg kg(-1) fresh weight. High levels of p,p'-DDT were detected in comparison to the metabolites (p,p'-DDE and p,p'-DDD). The detection of higher levels of p,p'-DDT than p,p'-DDE and p,p'-DDD, in most fish samples, suggests recent exposure of fish to DDT. Generally, most of the fish samples had residue levels below the maximum residue limits (MRL) recommended by FAO/WHO Codex Alimentarius Commission.