Occurrence and risk assessment of organochlorine pesticides in River Rwizi sediment in Western Uganda

Abstract

Fifteen organochlorine pesticides (OCPs) were determined in sediments from 14 sites in upper, middle and lower reaches of River Rwizi. Duplicate samples were collected at each site, during three sampling regimes covering the wet and dry seasons. Sediments were extracted using acetone: hexane (1:1 mixture) and clean-up was achieved using C18 cartridge. OCPs were determined using gas chromatography-mass spectrometry (GC–MS). Principal component analysis (PCA) was used to assess spatial relationships. To assess ecological health risk, the risk quotient (RQ) based on exposure and toxicity was calculated. Hazard index (HI), based on the sum of ratios of the OCP doses to permitted levels was calculated assess risk to human health. Recoveries of 54.4–93%, linearity (R²) > 0.9988, limits of detection (LoD) of 0.34–0.66 ng/kg and limits of quantification (LoQ) of 1.02–2.01 ng/kg were obtained during method validation. The total OCPs concentration was 84.34–159.5 µg/kg dry weight. Presence of OCPs in the sediment may be attributed to environmental persistence, however the presence of endosulfans implies continuing input since they have a relatively short half-life. Ecological risk quotient (RQ) values were 0.01–9780.37 while hazard index (HI) values were all less than 1. The HI for dermal exposure in children was 0.2695 and in adults it was 0.1497, exceeding the negligible risk threshold of H < 0.1. While the risk to human health was low, the high risk posed to ecology by heptachlor (RQ = 9780.37) and endosulfan II (RQ = 835.98) requires immediate mitigation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s44274-025-00465-7.

Introduction

Organochlorine pesticides (OCPs) are synthetic chlorinated organic compounds which are widely used for control of pests in agriculture [1, 2]. OCPs are generally chemically stable, lipophilic and persistent [2, 3].Their presence in the environment is concerning because of their potential harm on human health and their contribution to ecological instability [2–4]. On exposure, OCPs bio-accumulate in fat tissues of animals where they affect pulmonary, circulatory, neural, hormonal, endocrine functions and fertility [5] leading to undesirable effects on life.

Pesticides are ordinarily applied in agriculture to protect yield [6, 7]. In Uganda, OCPs and organophosphates (OPs) are used in agriculture, forestry, and horticulture to enhance crop production [8] and in vector control to protect human health against vector borne diseases like malaria and dengue [9]. Pesticides enter the environment through spray drifts, agricultural runoffs, accidental spills and leakages [10, 11]. Improper use, overuse and surface runoff of pesticides can also lead to dispersion to non-target environment [12, 13] resulting into negative impacts on environmental and human health [13, 14].

While the importation of OCPs into Uganda was restricted between 1993 and 1999 [15], DDT was re-introduced for indoor residual spraying against mosquitoes in 2006 [16]. OCPs and their metabolites have been found in various matrices in Uganda including honey [17, 18], serum and urine [19] soil [17, 20] and in Lake Victoria sediment [21]. This raises concern about their potential ecological and human health impact. Risk assessment and identification of OCP pollution hotspots can support development of specific control measures to protect human and ecological health.

While OCPs are banned in most of the developed nations in Europe [22] and North America [22, 23], they are still exported to low and middle income countries [24], where OCP use continues due to lack of suitable alternatives and weak legislation [25]. In these low to middle income countries they are used in the control of disease vectors in areas where malaria is endemic [26, 27] and have been detected in various environmental matrices throughout the world [14, 26, 28] including water, fish [29] and sediment in Uganda [21].

Sediments are environmental sinks for different pollutants including pesticides [30]. Retention of pollutants in sediment is a potential risk to aquatic life. Furthermore, pollutants can be re-suspended by natural processes or human actions resulting into sources of pollution [30, 31]. The presence of OCPs in sediment therefore is a potential ecological health and human health risk [14]. Hence, analysis of sediment is crucial for understanding fate, transport and assessment of potential environmental risk.

In western Uganda, River Rwizi is main source of water for Mbarara City and its surroundings, supporting livelihoods of about 2.5 million people in twelve districts including Rakai, Lyantonde, Isingiro, Lwengo, Kiruhura, Mbarara, Bushenyi, Buhweju, Rwampara, Sheema and Rubirizi [32–34]. It originates from the hills of Buhweju in Western Uganda and eventually pours its water into Lake Victoria through the smaller lakes of Mburo, Kachera, Nakivale, and Kijanebarola [33, 35, 36]. The river is currently threatened by pollution, climate change and land use [36–39]. Cropland in River Rwizi catchment increased from 23.0 to 51.6%, grassland and wetlands reduced from 63.3% to 37.8%, and 8.1% to 4.7% respectively for the period 2014–2020 [40] posing an existential threat to the river. This river plays a crucial role in the Lake Victoria basin as a habitat for macroinvertebrates [41] and other aquatic life, but also as an important source of water for various purposes in the region. Contamination of the sediment not only impacts the ecological health of the river and water quality but may have potential health risk of the surrounding communities.

Degradation of physical–chemical [36] and microbial parameters [38] of water obtained from River Rwizi as well as presence of toxic metals [37, 42] have been reported. Information on occurrence of other pollutants including pesticides is scarce, yet they are pollutants of emerging concern in the environment [43]. Over 80% of the Ugandan population relies on agriculture, mostly on subsistence farming [8] however, agricultural inputs available on the market for use in enhancing productivity are often poor quality, counterfeits or adulterated due to ineffective regulation, liberalization of market and limited farmer knowledge [20, 44, 45].When pesticides are used, environmental transport mechanisms can lead to their distribution in the environment eventually reaching sediment [4, 30] where they may influence the ecological health. There is a need to determine the presence of OCPs in River Rwizi sediment, the potential hot spots of pollution as well as the effect of their presence in the river ecosystem.

The impact of a pollutant on the environment can be determined through a risk assessment. Sediment quality guidelines (SQGs), threshold effect concentration (TEC), probable effect concentration (PEC) are some of the parameters used to evaluate sediment toxicity [46, 47]. Limited work has been done in Uganda on the potential risk posed by presence of organochlorine pesticides in sediment; for example aldrin and dieldrin posed a threat on lake Victoria ecosystem based on the threshold effect concentration (TEC) for fresh water ecosystems [21]. There is no information on the risk posed by OCPs in River Rwizi sediment, hence the need to assess it to protect the ecosystem health.

The risk quotient (RQ) value method which is based on the ratio of the average concentration of a single chemical in the sediment to the predicted no effect concentration (PNEC) remains one of the most popular methods for ecological risk assessment of pollutants such as organochlorine pesticides in sediment [47, 48]. The risk to human health is often assessed using the hazard quotient (HQ) based on the lifetime average daily dose (LADD) [31, 49]. Due to limited information on OCPs environmental occurrence and their potential risk in Uganda, this study determined the OCPs contamination profile of River Rwizi sediment and assessed the potential ecological and human health risk posed by their presence.

Materials and methods

Materials

A mixture of OCP standards at 200 μg/mL in hexane: toluene (1:1) (EPA 8081 pesticide standard mix) containing: Aldrin, α-BHC, δ-BHC, Lindane, β-BHC, α-Chlordane, γ-Chlordane, 4,4′-DDD, 4,4′-DDE, 4,4′-DDT, Decachlorobiphenyl, Dieldrin, α-Endosulfan, β-Endosulfan, Endosulfan sulfate, Endrin, Endrin aldehyde, Endrin ketone, Heptachlor, Heptachlor exo-epoxide, Methoxychlor and 2,4,5,6-Tetrachloro-m-xylene was obtained Sigma Aldrich.

Analytical grade acetone (99.5%, SPAN Chemie) and n- hexane (99%, Pallav) were used in preparation of extraction solvent. C18 cartridges (500 mg, 6 mL), for clean-up were obtained from Hypersep, Thermofisher, USA. Working standard solutions were prepared daily by diluting stock solution in n-hexane solvent in glass vials. All glassware that were used for the sample preparation were washed with soap, solvent rinsed and kept at 100 °C in convection laboratory oven until use.

Sampling and sample treatment

Sample collection

Study sites were purposively selected along the River Rwizi from the source Buhweju hills, downstream through the rural to Urban areas of Mbarara City; as shown in the site map in Fig. 1. The river which is about 55 km long [35, 37] was divided into three sections covering the upper, middle and lower reaches. Sediments were judiciously collected from 14 sites ensuring representation. The coordinates, description of site activities are listed in Table 1.

Fig. 1.

Map showing Uganda (A) and the sampling sites at a bigger scale (B and C) along River Rwizi in Western Uganda. KIJ, KAK, KAT represent wastewater treatment ponds located at Kijungu, Kakoba and Katete in Mbarara City respectively

Table 1.

Study sites on River Rwizi from where sediment was collected

Site	Coordinate	Activities in the vicinity
SITE 1, Karungu, Buhweju District	0° 20′ 58.344"S30° 28′ 8.0034"E	Farming, local distillation
SITE 2, Kamajumba, Buhweju, District	0° 21′ 6.732"S 30° 28′ 24.7434"E	Farming, school, human settlement
SITE 3, Kanyabukanja, Buhweju District	0° 23′ 20.652"S 30° 28′ 7.2474"E	Farming
SITE 4, Kyenyonyi, Sheema District	0° 25′ 8.904"S 30° 26′ 23.352"E	Local distillation, farming, high way, Sub–County Offices
SITE 5, Nyakambu, Sheema District	0° 26′ 4.74"S 30° 27′ 36.8274"E	Wetland, high way, farming
SITE 6, Koga, Sheema District	0° 34′ 54.0114"S 30° 27′ 3.6"E	Highway, farming, wetland
SITE 7,Nyeihanga, Rwampara District	0° 41′ 54.492"S 30° 22′ 45.192"E	Trading centre, farming, wetland
SITE 8, Nyamitanga, Mbarara City	0° 37′ 6.8514"S 30° 38′ 29.76"E	Highway, trading centre, farming
SITE 9, Kashanyarazi, Mbarara City	0° 37′ 12.756"S 30° 38′ 58.9554"E	Residential area, farming
SITE 10, Katete, Kitobero, Mbarara City	0° 37′ 13.8"S 30° 39′ 44.028"E	Farming, boat rides
SITE 11, Katete (after wastewater stabilization pond)	0° 37′ 13.8"S 30° 39′ 44.028"E	Next to WSP, farming
SITE 12, Kakoba near Bishop Stuart, Mbarara City	0° 36′ 29.6994"S 30° 41′ 48.984"E	Residences, farming
SITE 13,Kyahi forest reserve, Mbarara City	0° 36′ 42.732"S 30° 44′ 5.244"E	Forest, farming
SITE 14, Rwebicuncu, Mbarara City	0° 36′ 33.4794"S 30° 45′ 14.796"E	Farming

Duplicate samples were collected to make a total of 28 surface sediment samples from each site during each sampling of the three sampling regimes (October 2023, March 2024 and June 2024). These periods were selected to represent, the short wet season, dry season and long wet seasons in Uganda [50]. Surface sediments (0–10 cm) were collected using a hand held trowel, as described in the field sampling guidance document number 1215 [51]. The sediment was transferred into a plastic zip-lock bag and transferred on ice in a cool box to the laboratory for processing.

Sample extraction

Sediment samples were air-dried in the fume hood for 7 days and any debris was manually removed. Sediments samples were extracted following the United States Environmental Protection Agency (US EPA) method 3550 C [52]. The sample (100 g) was weighed into a beaker and a mixture of analytical grade1:1 acetone: n- hexane (100 mL) was added. The sample was mixed thoroughly and left overnight. The next day, the mixture was extracted using an orbital shaker (Grant OLS200, England) set at 125 rpm for 40 min in order to transfer non-volatile and semi volatile organic compounds from sediment to the liquid phase. The extract was decanted and the extraction was repeated twice more with additional 100 mL portions of clean solvent (1:1 acetone: hexane). All the extract fractions (3) were collected into a borosilicate reagent bottle. Extract cleaned up was achieved using solid phase extraction (SPE) [53]. A C18 cartridge (Hypersep, 500 mg, 6 mL, Thermofisher). The cartridge was conditioned with 10 mL of 1:1 acetone: hexane mixture followed by 10 mL of distilled water. The extract was loaded onto the SPE cartridge and the cartridge was dried under vacuum. The analyte was eluted using 5 mL acetone: hexane which was dried under vacuum. Samples were later reconstituted by dissolution in hexane before GC–MS analysis.

Method validation

Method validation was realised through assessment for linearity, recovery, limits of detection (LoD), and quantitation (LoQ). The linearity of the method was studied with using standard pesticide solutions of 0.1 µg/mL, 0.4 µg/mL, 1.0 µg/mL, 5.0 µg/mL and 10 µg/mL for each analyte. Recoveries were determined by spiking α-BHC, γ-BHC, heptachlor, heptachlor-oxide, α-chlordane, endrin, endrin ketone and endusulfan II at different concentrations into sediment collected from site 1. Spiked samples were left overnight to allow pesticide adsorption into the sample. They were then extracted according to the extraction procedures described above and recoveries assessed using equation 1.

$$Recovery=\frac{Amountdetected-Amountunspikedsite1}{\mathit{Amountspiked}}\times 100$$ 1

Limits of detection (LoD) and quantification (LoQ) were determined using regression. The standard deviation of the response (σ) and the slope of the calibration curve (S) were used to calculate LoD (3.3σ/S) and LoQ (10σ/S). Quality control was assured by running laboratory blanks and a standard after every six real samples using the same analytical procedures. All samples were analysed in triplicate.

Gas chromatographic analysis

Samples were analysed using method 8081 B [54]. OCPs residues were analysed using a gas chromatograph (Finigan Trace GC ultra) equipped with the autosampler (AS 3000) and mass spectrometer (DSQ II Thermo Fisher Scientific). Briefly 1 µl was injected in a split mode onto a Rtx-5 ms gas chromatograph (GC) capillary column (30 m, 0.25 mm ID, 0.25 µm, Restek). The injector temperature was constant at 225 °C, split flow (mL/min) was 10, split ratio was 10, split-less time was1.00 min. The GC temperature programme was: Initial C, 100 °C, hold 2.00 min, ramp 1 was: 15.0 °C /min to 160 °C, held for 1 min. Ramp 2, was 4.0 °C /min to 280, hold 2 min. The mass spectrometry (MS) transfer line temperature was 250 °C while ion source was maintained at 220 °C. Helium was used as the carrier gas at 1.0 mL/ min flow rate. Data acquisition and processing was achieved using Xcalibur Software (Thermo Fisher Scientific).

Data analysis

Quantitative data was presented as means and standard deviation using Microsoft Excel. OCPs data was subjected to principal component analysis (PCA) to understand spatial variation and potential hotspots of OCP pollution. PCA is a multivariate statistical technique used when variables are highly correlated. It reduces the number of observed variables to a smaller number of principal components which account for most of the variance of the observed variables [31, 55]. PCA was performed using IBM SSPS statistics 26 software with varimax rotation to minimize the number of variables with high loading on each component to facilitate easier interpretation [31].

Risk assessment

The deterministic approach of ecological risk assessment was used to compare toxicity and environmental exposure. A risk quotient (RQ) was calculated based on exposure and toxicity. Exposure was obtained from the measured average environmental concentration (MEC) and toxicity values were based on effect level or endpoint obtained from eco-toxicity testing. The toxicity endpoint for assessment of acute toxicity is the lowest tested EC₅₀ or LC₅₀ while that for chronic toxicity is the lowest no observed adverse effect concentration (NOAEC) [56]. Since the concentrations of OCPs in sediment were low, equation (2) was used.

$$RQ=\frac{\text{MEC in Sedment}}{\text{PNEC sediment}}$$ 2

where MEC is the mean environmental concentration of a pesticide in sediment

PNEC is the predicted no-effect concentrations (PNEC) of the corresponding pesticide in sediment.

The lowest PNEC values for the organochlorine pesticides were obtained from NORMAN Ecotoxicology Database. The calculation used for determination of PNEC in sediments (equation 3) [57] is:

$$Lowest{\mathit{PNEC}}_{\mathit{fw}}\ast 2.6\ast (0.615+0.019\ast K_{\mathit{oc}})$$ 3

Human risk associated with inhalation, ingestion and dermal exposure was assessed using the hazard quotient (HQ) was calculated using equations (4–7) [25, 48, 58]

$$HQ=\frac{\mathit{CDI}}{\mathit{RfD}}$$ 4

$${\mathit{CDI}}_{\mathit{ingestion}}=\frac{C\times IR\times EF\times ED}{BW\times AT}$$ 5

$${\mathit{CDI}}_{\mathit{Dermal}}=\frac{C\times SA\times CF\times EF\times ED\times ABS\times AF}{BW\times AT}$$ 6

$${\mathit{CDI}}_{\mathit{inhalation}}=\frac{C\times \frac{1}{\mathit{PET}}\times IAR\times EF\times ED}{BW\times AT}$$ 7

where:

CDI is the chronic daily intake (mg/kg/day); RfD is the reference dose (μg/ kg), C is the concentration of pesticide in sediment (μg/ kg), IR is the Ingestion rate (200 mg/day for a child and 100 mg/day for and adult), CF is the conversion factor (1× 10^-6 kg/mg), EF is the exposure frequency (350 days/year), ED is the exposure duration (6years for a child, 24 years for an adult), BW is the body weight (15 kg for a child, 65 kg for an adult), AT is the average life span (non-carcinogenic ED × 365 days which is 2190 children and 8760 for adults; carcinogenic 70 × 365 days for both adults and children), PEF is the particle emission factor (1.32×10⁹ m³ /kg), IAR is the inhalation rate (for 7.5 child; 15 for adult, SA is the area of dermal contact with sediment (1800 for a child; 5000 for an adult), ABS is the dermal absorption factor (0.1), AF is the dermal surface factor (0.2 child, 0.04 adult) [58].

Hazard index (HI) which is used to assess health risks from combined exposures [17, 59, 60] was calculated from equation 8

$$HI=\sum HQ$$ 8

Results and discussion

Analytical method validation

Eleven OCPs were used in determination of recovery which was found to be 54.4–93% as shown in Table 2. Generally, this recovery was in the acceptable ranges [53], hence, sediments were extracted using the method described. The EPA OCP mix was used in instrument calibration and the chromatogram was as shown in Fig. 1S (Supplementary data). There was no significant difference (P > 0.05) between the data obtained from the same sites during the three sampling regimes except for sites 6, 7 and 13 which had a P < 0.05 as shown in Table 2S.

Table 2.

Percentage recoveries based on spiked sediment R²values, limits of detection and quantification of OCPs in sediment

OCP	Retention time (minutes)	LoD(3.3 s/S)	LoQ (10 s/S)	R2	% recovery
α-BHC	10.80	0.44	1.32	0.9995	69.87
γ-BHC	12.80	0.35	1.07	0.9997	57.02
Heptachlor	16.60	0.42	1.28	0.9995	57.85
Heptachlor-Epoxide	19.61	0.66	2.01	0.9988	93.28
α- Chlordane	20.62	0.66	2.01	0.9988	70.87
p,p’-DDE	22.32	0.34	1.02	0.9997	54.44
Endrin	23.25	0.53	1.60	0.9992	73.40
Endosulfan II	23.68	0.66	2.01	0.9988	73.40
Endosulfan sulfate	25.57	0.46	1.40	0.9994	57.90
DDT	25.80	0.46	1.40	0.9994	57.90
Endrin Ketone	28.40	0.66	2.01	0.9988	83.68

Occurrence of organochlorine pesticides in sediment of River Rwizi

Fifteen OCPs were detected in sediments collected from fourteen sites along the River Rwizi in the concentration range not detected (ND) to 16.71 µg/kg dry weight (dw) as shown in Table 1S. The total of OCPs concentration ranged between 84.34 and 159.5 µg/kg dw. This concentration was slightly higher than that reported in Lake Victoria sediment [21]. The amount detected in this study was was comparable to reports about sediments from Nigeria [61], Lakes Tana and Hayqe in Ethiopia [48] Slovenia, Switzerland and Denmark [14]. Higher concentrations of OCPs were reported in Spain and Czech republic [14, 30] South Africa [62], Kenya [63] and Gulf of Mexico [64] as shown in Table 3.

Table 3.

Comparison of OCP concentrations determined in different Countries

Comparison of OCP concentration levels in sediment
Location	Number of OCPs determined	Concentration range (µg/kg dw)	References
River Rwizi, Uganda	15	ND–6.71	This Study
Lake Victoria, Uganda	16	ND–15.96	[21]
Lake Tana and Hayqe, Ethiopia	18	10.11–125.65 and 10.96–45.1	[48]
Ebro River Delta, Spain	12 (out of 69 pesticides)	0.01–89.7	[30]
Lake Fuxian, China	20	0.35–3.83	[65]
Msunduzi River, South Africa	13	464.65–3773.66	[62]
Gulf of Mexico	16	86.13–232.93	[64]
Reservoirs and lakes in Slovenia, Switzerland, Croatia, and Denmark	10 (out of 99 pesticides)	80–120	[14]
Asa River, Nigeria	07	0.0045–0.947	[61]
Nairobi River, Kenya	17	0.01–41.9	[63]

Table 4 Hexachlorocyclohexane also known as benzene hexachloride (BHC) was detected at all sites (100% frequency) with mean concentrations lower than the calculated PNEC of 0–128.523 (Fig. 2, Table 5) but higher than the drinking water recommendation of 2 µg/L [66]. α-BHC, γ-BHC and δ-BHC were in the concentration range of 1.66–9.90 µg/kg dw, which was higher than that previously detected in lake Victoria sediment [21]. These concentrations were also higher than the lowest effect level (0.003–0.006 pg/g) as recommended by the guidelines for the protection and management of aquatic sediment quality in Ontario [67]. γ-BHC commonly referred to as lindane was historically used as a broad spectrum insecticide to control the cotton ball worm although it is reported to have still been in use around 1998 [68]. Lindane continues to be detected in different environmental compartments and in food products in Uganda [18, 20, 21, 69] demonstrating its environmental persistence and accumulation in the food chain. α and δ- BHC are isomers of γ-BHC which can be found in its technical grade form, so they are likely to have been generated from the γ-BHC. Heptachlor, endrine and p,p’- DDD also had a 100% detection frequency. The least detected OCP was Aldrin at 14% as shown in Fig. 2.

Table 4.

Pearson correlation of OCPs

	Alpha-BHC	Gamma-BHC	Delta-BHC	Heptachlor	Aldrin	Heptachlor-Epoxide	Alpha-Chlordane	Endosulfan	PP-DDE	Endrine	Endosulfan II	PP-DDD	Endosulfansulfate	DDT	Endrine Ketone
Alpha-BHC	1
Gamma-BHC	0.729876	1
Delta-BHC	0.007926	0.23801808	1
Heptachlor	0.507702	0.76295543	0.742369	1
Aldrin	− 0.27106	0.02174762	0.805715	0.590572	1
Heptachlor-Epoxide	0.502677	0.53128282	− 0.12925	0.114345	− 0.42728	1
Alpha-Chlordane	0.225183	0.69509613	0.376553	0.517886	0.096351	0.55812	1
Endosulfan	− 0.23606	0.12360607	0.853732	0.658507	0.97194	− 0.39503	0.180772	1
PP-DDE	0.207221	0.2973135	0.243095	0.380422	0.054771	− 0.05131	0.096209	0.142755	1
Endrine	0.459608	0.48708836	0.364768	0.463758	0.06647	0.537803	0.282516	0.162532	0.093132	1
Endosulfan II	− 0.41593	0.07706967	0.14824	− 0.00326	0.084602	0.197562	0.555056	0.171255	− 0.15897	− 0.05607	1
PP-DDD	− 0.07225	0.3193183	0.839955	0.768295	0.902388	− 0.27507	0.281078	0.966232	0.231155	0.280773	0.193526	1
Endosulfansulfate	0.491206	0.42102978	0.223125	0.352687	0.057543	0.728304	0.202182	0.06753	− 0.04098	0.733699	− 0.10979	0.14655	1
DDT	0.561077	0.50786154	0.221042	0.506464	0.184562	0.269966	0.10514	0.202129	0.25147	0.437977	− 0.18965	0.384812	0.496486	1
Endrine Ketone	0.30299	0.45106121	0.481947	0.5054	0.292654	0.241132	0.376428	0.403989	0.418804	0.467895	− 0.06942	0.474777	0.441746	0.326563	1

Fig. 2.

Detection frequency of OCPs

Table 5.

Risk assessment

OCP	Mean Conc ug/kg	PNEC µg/kg dw	RfD µg/kg	RQ	HQ ingestion/ child	HQ ingestion/ adult	HQ inhalation/ child	HQ inhalation/ adult	HQ dermal/ child	HQ I dermal/ adult
Alpha-BHC	7.941	0.000	2.000	–	5.294 × 10–8	1.222 × 10–8	3.846 × 10–13	8.874 × 10–14	1.200 × 10–3	6.670 × 10–4
Gamma-BHC	6.880	17.613	2.000	0.390	4.580 × 10–8	1.058 × 10–8	3.332 × 10–13	7.689 × 10–14	1.040 × 10–3	5.779 × 10–4
Delta-BHC	2.583	128.523	2.000	0.020	1.722 × 10–8	3.974 × 10–9	1.251 × 10–13	2.886 × 10–14	3.906 × 10–4	2.169 × 10–4
Heptachlor	8.609	0.001	0.030	9780.370	3.826 × 10–6	8.829 × 10–7	4.169 × 10–13	9.621 × 10–14	8.678 × 10–2	4.821 × 10–2
Aldrin	5.645	0.000	0.030	–	2.509 × 10–6	5.789 × 10–7	2.734 × 10–13	6.308 × 10–14	5.690 × 10–2	3.161 × 10–2
Heptachlor-Epoxide	9.237	61.029	0.030	0.150	4.105 × 10–6	9.473 × 10–7	4.473 × 10–13	1.032 × 10–14	9.311 × 10–2	5.173 × 10 −2
Alpha-Chlordane	9.928	196.262	0.200	0.050	6.618 × 10–7	1.527 × 10–7	4.808 × 10–13	1.109 × 10–14	1.501 × 10–2	8.339 × 10–3
Endosulfan I	3.570	4.205	20.000	0.850	2.380 × 10–9	5.492 × 10–11	1.729 × 10–13	3.989 × 10–14	5.397 × 10–5	2.999 × 10–5
p,p’-DDE	6.421	4.744	1.000	1.350	8.561 × 10–8	1.976 × 10–8	3.109 × 10–13	7.176 × 10–14	1.941 × 10–3	1.078 × 10–3
Endrin	9.062	0.000	0.600	–	2.014 × 10–7	4.647 × 10–8	4.389 × 10–13	1.013 × 10–13	4.567 × 10–3	2.537 × 10–3
Endosulfan II	12.457	0.015	20.000	835.980	8.305 × 10–9	1.916 × 10–9	6.033 × 10–13	1.392 × 10–13	1.884 × 10–4	1.046 × 10–4
p,p’-DDD	1.804	17.906	1.000	0.100	2.405 × 10–8	5.550 × 10–9	8.737 × 10–13	2.016 × 10–14	5.455 × 10–4	3.031 × 10–4
Endosulfan sulfate	7.216	1.615	20.000	4.470	4.810 × 10–9	1.110 × 10–9	3.495 × 10–13	8.065 × 10–14	1.091 × 10–4	6.061 × 10 −5
DDT	7.094	895.311	1.000	0.010	9.458 × 10–8	2.183 × 10–8	3.436 × 10–13	7.928 × 10–14	2.145 × 10–3	1.192 × 10–3
Endrine Ketone	10.938	138.364	0.600	0.080	2.430 × 10–7	5.609 × 10–8	5.297 × 10–13	1.222 × 10–14	5.513 × 10–3	3.063 × 10–3
Hazard Index (HI)					1.188 × 10–5	2.742 × 10–6	5.297 × 10–12	1.222 × 10–12	0.270	0.150

PNEC values calculated using equation ii (from https://www.norman-network.com/nds/ecotox/lowestPnecsIndex.php) RfD from WHO drinking water guidelines [66]

Heptachlor (7.04–11.56 µg/kg dw), heptachlor-epoxide (7.86–11.86 µg/kg dw), aldrin (01–11.19 µg/kg dw), alpha-chlordane (8.38–12.82 µg/kg dw), endosulfan I (0.03–16.7 µg/kg dw), endosulfan II (10.60–15.32 µg/kg dw), endosulfan sulfate (6.33–8.48 µg/kg dw), DDT (6.08–8.34 µg/kg dw), p,p’-DDD (1.14 -5.17 µg/kg dw), p,p’-DDE (5.54–8.01 µg/kg dw), endrin (7.57–11.66 µg/kg dw), and endrine ketone (9.12–14.81 µg/kg dw) were also detected in sediments from River Rwizi as shown in Table 1S. All of these concentrations were higher than sediment guidelines for no effect concentration [67]. These pesticides were banned worldwide and in Uganda by 2012, however they are persistent.

Endosulfan II, endrin ketone, chlordane, heptachlor epoxide, endrin, and heptachlor were present in high concentration as shown in Fig. 3. Endosulfan was imported and used for control of coffee borers and cotton boll worm however by 2012 all these pesticides were banned for importation and use in Uganda [70], their current detection may be attributed to their persistence in the environment, illegal use and/or atmospheric deposition [20].

Fig. 3.

Spatial distribution of OCPs

The concentration of endosulfan II detected in the River Rwizi sediment was 10.66–15.32 µg/kg dw, was slightly higher than that detected in the Lake Victoria sediment [21] implying potential environmental input. Furthermore, the half-life of total endosulfan and α and β-endosulfan is 1336 and 27.5 and 157 days respectively [71], therefore its detection in sediment shows potentially continued use in the River Rwizi basin Uganda. Endosulfan I and II concentrations were higher than the calculated PNEC of 4.205 and 0.015 respectively (Table 5) signifying potential threat to aquatic life, but were lower than the recommended drinking water guideline of 20 µg/L [66]. They were previously reported in honey at mean concentrations of 10.62 and 9.38 µg/kg respectively [17], in soils in Kihihi (western Uganda) at a concentration of 4–20 µg/kg dw [20] and in sediment from lake Victoria in the range 0.03–9.67 µg/kg dw.

Heptachlor was detected at all the sites (100%) at a concentration of 7.04–11.56 µg/kg dw, higher than the calculated PNEC of 0.001 µg/kg dw and the recommended value in drinking water of 0.03 µg/L [66]. In the environment, heptachlor is oxidised to heptachlor-epoxide. The high concentration of heptachlor and heptachlor epoxide may pose a risk to freshwater ecology. Heptachlor and its epoxide metabolite have also been detected in different matrices including serum [19] honey [18], fish [72] and the environment [15, 20, 21] showing its wide spread. p,p’-DDE a metabolite of DDT was detected at 5.54–8.01 µg/kg dw higher than the calculated PNEC of 4.744 µg/kg dw as shown in Table 5. It was higher than 1 µg/L the recommendation from drinking water guidelines [66]. The presence of p,p’-DDE in sediment is a potential threat to aquatic and human health.

Spatial distribution of OCPs

Site 6 (Koga) had the highest concentration of aldrin (Fig. 3) and was clearly distinct from other sites as shown by principal component analysis (PCA) in the scatter plot in Fig. 4 (A1 and B1). Sites 6, 5 and 4 were most impacted by pesticide presence as shown in the hierarchical cluster analysis result in Fig. 5. Other sites with a high presence of OCPs were 1, 3, 11, 13 and 14 (Fig. 3) and in all these sites there was evidence of farming activities in the river catchment as shown in Table 1, implying that the source pesticides could be agricultural activity in the areas, although persistence of the OCPs is a strong factor. The is a need for adequate farmer education at community level. This can be achieved through engaging community development officers and agricultural extension workers. Partnerships with research institutions may lead to development of green and safer agricultural inputs.

Fig. 4.

A PCA plot of distribution of organochlorine pesticides (OCPs) (A1), Scree plot of the OCPs (A2), spatial variation of sites (B1) and the Scree plot of sites (B2). Raw data was obtained from sediments collected along River Rwizi, Mbarara, Uganda

Fig. 5.

Hierarchical cluster showing sites most impacted by pesticides in sediment

Four principle components explained up to 84% of the variation in pesticides and three components explained a similar percentage variation of sites as shown in the scree plots in Fig. 4 (A2 and B2) as shown in the component matrices in Table 7S. Heptachlor. p,p’-DDD, δ-BHC, γ-BHC, endrin ketone and endrin contributed strongly to principle component (PC) 1 in the PCA of OCPs while site 13, site 12, site 3, site 8, site 9, site 14, site 1, site 10, site 7, site 2 were the major contributors to PC1 in evaluation of sites.

A correlation table (Table 3S) showed a strong positive correlation between in sites 1, 3, 7, 11, 13 and 14. Sites in lower reaches (7–14) also had a strong correlation while site 6 was unique. Table 4S shows that α-BHC and γ-BHC were strongly positively correlated. γ-BHC, heptachlor and α-chlordane were also positively correlated. Strong positive correlations have been previously reported between total organic matter and total pesticide residues [31], thus the relationship between organic matter content from sites along River Rwizi [73] were compared with the OCP data in this study. Table 5S shows the organic matter data used in the comparison. There was a strong positive correlation between percentage organic matter an δ-BHC, endosulfan I, p,p’-DDE and p,p’-DDD as shown in Table 6S. This strong positive correlation can be attributed sequestration of OCPs by organic matter [31, 74].

Risk assessment of organochlorine pesticides in River Rwizi sediment

Results from risk assessment are shown in Table 5. Hazard quotient and index values less than one are considered acceptable [75]. Generally, all HQ and HI values were less than 1 as shown in Table 5 thus, the detected concentrations of OCPs were considered not risky. The highest HI value was based on dermal exposure. HI for dermal exposure in children (0.2695) and in adults (0.1497) exceeded the risk negligible threshold of H < 0.1 [76]. Swimmers and children who play at the river banks may be at risk of dermal exposure. Generally, when the chronic daily intake (CDI) is higher than the reference dose value (RfD), the contaminant is unlikely to cause adverse human health effects [58]. In this study, the CDI was higher than RfD resulting in low hazard quotient (HQ) and hazard index values, therefore the concentrations measured in sediment in this study did not pose a significant health risk to humans. The HQ reported in Nigeria was higher [61] perhaps due to the presence of DDT and its metabolites which pose a high risk.

Some pesticides however posed a risk to fresh water ecosystem. When the RQ is less than 0.1, the contaminant is considered no risk; an RQ of 0.1–1 is low risk, RQ of 1 – 10 is medium risk while an RQ greater than 10 is for pollutants or contaminants of a high risk [47]. The calculated RQ values showed that heptachlor (9780.37) and endosulfan II (835.98) posed a very high risk to fresh water ecology while endosulfan sulfate (4.47) and p,p’-DDE (1.35) posed medium risk to the ecosystem as shown in Table 5. Acute and chronic to heptachlor has been associated with histopathological changes in the gill, liver, kidney, and muscles of tilapia [77] and in larval and adult epidermis of Rana kl. esculenta [78]even at low concentrations. Therefore, care should be taken to limit entry of heptachlor in aquatic ecosystems.

Endosulfan I, endosulfan II and endrin showed high risk (RQ > 1) for fish in Lake Tana, Ethiopia but p,p’-DDD, p,p’-DDE, and lindane were of medium risk (0.1 < RQ < 1). The medium risk posed by p,p’-DDE was comparable to our study. Moderate to high ecological risk due to presence of OCPs in sediment was also reported in China [65] and Spain [30].

The deterministic approach of risk assessment is limited by its inability to consider or quantify the full range of possible outcomes. Risk is assessed assuming exposure from a specific source which may lead to under or over estimation [59, 79]. Further-more it does not account for risk evaluation in case of synergistic behaviour of the OCPs [59]. Despite these limitations, the HI and RQ values obtained show a need for mitigation. Community education to avoid potential environmental input of these persistent pollutants through illicit use or use of counterfeit/adulterated pesticides may reduce risk to aquatic life.

Conclusions and recommendation

The findings demonstrate continued presence of organochlorine pesticides in River Rwizi sediment in Western Uganda. While the concentrations detected are comparable to worldwide amounts in similar studies, the high risk quotients as seen for heptachlor, endusulfan II, endosulfate and p,p’-DDE are concerning for freshwater ecological health. Generally, mean concentrations were higher than the recommended drinking water standards and sediment guidelines for no effect concentration. Therefore, sediment may be act as a source of OCPs to water in case of perturbation. There is a risk of potential build-up of the OCPs in the food chain too which concerns public health. Therefore, continued environmental monitoring and public education to are paramount. Capacity building and enforcement of policy regulations by the ministries of agriculture and environment should be strengthened. In terms of limitation of the study, the deterministic approach of using HQ to evaluate risk is a simplistic model, it is best used as a preliminary tool for risk assessment. In future actual dose–response data can be used to assess true risk for example using mouse or cell-line models.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

G. B: Conceptualization, sample collection, analysis and manuscript writing; D. B, sample collection, analysis and manuscript writing. S.M and J.B Laboratory analysis, quality control and manuscript writing.

Funding

Research reported in this publication was supported by the Fogarty International Center of the National Institutes of Health under Award Number K43TW012594.

Data availability

All data is included in this manuscript and supplementary files uploaded.

Declarations

Ethics and consent to participate

Not applicable.

Consent to publication

Not applicable.

Competing interests

The authors declare no competing interests.