Environmental Risks of Pesticide Residues in the Lake Tana Sub-basin, Ethiopia: A Review

Table S2 and Fig. S1 show the various types of pesticides used across 13 districts (Fig. 2). Among these, Libokemkem, Fogera, and Gondar Zuria exhibit the highest number of active ingredients ( > 30), while Mecha and Dembia also show high levels (30). Alefa, Robit, North Achefer and South Achefer stand out with the lowest numbers (10-13). Overall, farmers across these districts applied 66 different active ingredients (Table S2), with high numbers particularly linked to the recent surge in small-scale irrigation projects in the area (Worqlul et al. 2015). The Fogera district was identified by Taye et al. (2021) as having the largest irrigated area and ranking second in the number of pesticides used, after Libokemeken (Fig. S1). Some of the active ingredients are classified as hazardous by the WHO (2020). For example, highly hazardous (Class Ib) active ingredients include methomyl, terbufos, 2,4-D, endosulfan, and dichlorodiphenyltrichloroethane (DDT). Moderately hazardous (Class II) active ingredients comprise chlorpyrifos, dimethoate, cypermethrin, deltamethrin, profenofos, propiconazole, mancozeb, bifenthrin, and imidacloprid. Additionally, glyphosate and malathion are categorised as slightly hazardous (Class III) (Table S2).

Based on the number of active ingredients, insecticides outnumbered those of herbicides, fungicides, and rodenticides (Fig. S1), indicating an extensive reliance on insecticides to manage common pest incidents. The predominant insects targeted in the area include aphids, African bollworm caterpillars, thrips, termites, spider mites, stock borers, fruit borers, and soil-borne insects affecting crops like cabbage, onion, tomato, pepper, maize, chickpea, grass pea, wheat, teff, fruits, and khat (Agmas and Adugna, 2020). Agmas and Adugna (2020) observed that farmers commonly apply multiple pesticide types, with insecticides being the most frequently used, followed by herbicides and fungicides. Similarly, the study by Abera et al. (2022) reported that most of the pesticides identified in the Lake Tana basin were insecticides. Additionally, Sishu et al. (2022) reported that all identified pesticides in their Lake Tana study were insecticides used specifically to control aphids. Herbicides are the second-most-used pesticides, particularly in the Fogera district, where they play a key role in rice production. Farmers in this region face significant weed challenges, prompting widespread herbicide use to protect their crops (Chauhan and Johnson 2010; Assefa et al. 2011; Jaipaul et al. 2014).

In this study, fungicides and rodenticides were found to have a lower utilization compared to insecticides and herbicides. This could be attributed to the inherent resistance of the predominant crops to fungal diseases, reducing the need for fungicides. Additionally, certain farming practices, such as crop rotation or intercropping, may naturally suppress fungal growth, thereby reducing the need for fungicide applications.

However, our results contrast with survey-based reports from the Central Rift Valley (CRV) of Ethiopia. In the CRV, fungicides are utilized much more intensively due to the region’s focus on horticultural crops like tomatoes and onions, which are highly susceptible to fungal blights (Mengistie et al. 2017). Furthermore, the high reliance on herbicides in the Lake Tana sub-basin, particularly for rice production in the Fogera district, differs from the CRV, where manual weeding by hired laborers is more prevalent in vegetable plots. This underscores how regional crop specializations and specific pest pressures (e.g., rice weeds versus vegetable fungi) dictate the pesticide profile of a watershed.

To synthesize pesticide use patterns with environmental occurrence and ecological relevance, a conceptual table was developed linking active ingredients reported in the Lake Tana sub-basin to their main crop uses, environmental detection locations, and reported ecological risk levels (Table S3). This synthesis integrates survey-based studies documenting pesticide use across different districts with monitoring studies reporting pesticide detections in water, sediment, and fish tissues. Pesticides reported as being used by local farmers but not detected in environmental matrices were not subjected to ecological risk characterization, as the absence of measured environmental concentrations precludes quantitative risk assessment. However, this synthesis also highlights a critical spatial bias in current research. While districts such as Libokemkem, Fogera, and Gondar Zuria in the eastern and northeastern plains have been relatively well-monitored due to their intensive rice and vegetable production, other highly irrigated areas remain significantly underrepresented. Notably, the southern and southwestern parts of the basin, which are home to the Koga irrigation scheme (6,508 ha in Mecha) (Taye et al. 2021) and the Tana Zegie and Tana Wenjeta pump irrigation scheme (Mequanent et al. 2019), lack systematic environmental monitoring. Despite the high agricultural intensity and use of active ingredients in these southern watersheds, the lack of site-specific monitoring data prevents an accurate assessment of the total contaminant load entering the lake from these tributaries. Consequently, there is an urgent need for more intensive, basin-wide monitoring that bridges these geographic gaps to ensure that conservation strategies for Lake Tana’s endemic biodiversity are based on a representative dataset covering all major irrigation entry points.

The environmental risks of pesticide residues to fish species (Labeobarbus megastoma, Labeobarbus tsanensis, and Oreochromis niloticus) in the Gumara River (Fogera district), Ribb River (Fgera abd Libo Kemkem districts) on the eastern side of Lake Tana, and Lake Tana itself were assessed using risk quotients (RQs) derived from measured body burdens (Table S4; Fig. 3). In the Gumara River, oxamyl and flazasulfuron presented the highest risks, posing very high and high risks to L. megastoma, respectively (Fig. 3). Flazasulfuron also posed a high risk to L. tsanensis and O. niloticus. Other pesticides, including pyrimethanil and carbaryl, presented low or no risk (RQ < 1) to the species in this river. In the Ribb River, imidacloprid was the primary concern, posing a high risk to L. megastoma and O. niloticus and a very high risk to L. tsanensis (Fig. 3). Flazasulfuron posed a high risk to L. tsanensis and O. niloticus and a low risk to L. megastoma. The measured concentration of oxamyl posed a low risk to L. tsanensis, while pyrimethanil and carbaryl levels posed no risk (RQ < 1) to all species (Fig. 3). In Lake Tana, O. niloticus was at risk from measured organochlorines concentrations such as endosulfan, endrin, and lindane posed a very high risk to fish populations, while DDT and dieldrin also posed high risks (Fig. 3).

Bild vergrößern

These results are consistent with broader African studies that emphasize (i) legacy OCPs as ongoing risks for freshwater fish and (ii) currently used pesticides (carbamates, neonicotinoids, sulfonylureas) as emerging concerns. For example, in South African freshwater systems, Ansara-Ross et al. (2012) also reported that OCPs (endosulfan, DDT) remain a priority threat to fish despite bans, while Olisah et al. (2022) showed a strong spatial and seasonal variation of organophosphates in estuarine fish, stressing the role of timing of monitoring and the hydrology of the system under investigation. Similarly, Ogbeide et al. (2015) found that fish from Nigerian rivers accumulated multiple pesticide residues, with carbamates and OCPs frequently exceeding ecological risk thresholds. Together, these studies confirm that the high risks of OCPs to O. niloticus in Lake Tana are part of a continent-wide pattern of persistence and bioaccumulation of these legacy pesticides. Although RQs for organochlorine pesticides were explicitly found only for O. niloticus, the endemic Labeobarbus species of Lake Tana (L. tsanensis and L. megastoma) may be equally or more vulnerable to these compounds based on their trophic ecology. L. tsanensis is predominantly insectivorous while L. megastoma, as a largely piscivorous species (Zelalem et al. 2023), occupies a higher trophic position and is therefore expected to experience enhanced exposure through biomagnification. Studies have shown that organochlorine pesticide concentrations increase with trophic level, with insectivorous and piscivorous fish exhibiting higher body burdens than omnivorous species due to dietary transfer and lipid affinity of these compounds (Erdogrul 2005; Eqani et al. 2013; Zhou Shan Shan et al. 2018). Consequently, the high risks observed for O. niloticus likely represent a conservative estimate of ecosystem-level risk, and the endemic Labeobarbus species may face comparable or greater long-term threats from legacy OCP contamination in the Lake Tana food web.

The very high risk of oxamyl (a carbamate) to L. megastoma as assessed in the Gumara River is similar to a Ghanaian risk assessment, where Onwona-Kwakye et al. (2020) predicted carbamate insecticides as key contributors to aquatic risk using regulatory ecological risk assessment scenarios. Similarly, in their systematic review for sub-Saharan Africa, Fuhrimann et al. (2021) emphasized carbamates among the most frequently applied and risk-relevant pesticide groups. For imidacloprid, the high risks observed for both L. tsanensis and O. niloticus in the Lake Tana sub-basin are similar to risks assessed using Ethiopia-wide surface water modeling (Teklu et al. 2015; Asefa et al. 2024), which identified neonicotinoids as high-risk drivers for aquatic invertebrates and fish. Risks from flazasulfuron body residues varied by species and river. While high risks from this herbicide were calculated for several fish-sampling site combinations, the background values on which its low PNEC (0.043 µg/kg ww) is based are unknown, as it is likely based on a confidential report supplied by industry to one of the EU regulators (https://www.norman-network.com/nds/ecotox/). Similar results were reported by Tyohemba et al. (2021), who showed that herbicides in South Africa bioaccumulate in fish and potentially disrupt food-web energy flow. Although pyrimethanil body residues posed low or no risk, fungicides have been recognized as sub-lethal stressors in other African ERA studies. For example, Asefa et al. (2024) noted that fungicides pose lower direct risks compared to insecticides but can still impair community dynamics when combined with other pesticide stressors. The low RQs for carbaryl (carbamate) in this study might mask concerns for chronic and mixture effects as observed in Nigerian rivers (Ogbeide et al. 2015). The very high risks to O. niloticus posed by endosulfan, lindane, endrin, DDT, and dieldrin body residues are part of a broader African pattern observed with legacy pesticides. Across the continent, OCP residues remain a major ecological threat in South African estuaries, where endosulfan continues to pose risks (Olisah et al. 2022; Ansara-Ross et al. 2012). Their residues also pose risks to fish in Nigerian rivers, where DDT and endosulfan persist at hazardous levels decades after bans (Ogbeide et al. 2015). Legacy OCPs remain ubiquitous drivers of risk in other African aquatic systems as well (Fuhrimann et al. 2021; Asefa et al. 2024), as they bioaccumulate in fish, herewith posing health risks to human consumers, a concern also raised by local farming communities (Orou-Seko et al. 2024).

Unlike most monitoring-based studies performed in Ghana, Nigeria, and South Africa, and the modeling or meta-analyses performed in Ethiopia and across Africa, this assessment was based on measured fish body burdens used to calculate RQs. By capturing organ-specific body residues in muscle and liver tissues, reflecting both bioaccumulation and metabolic processing of pesticide residues. Together, these findings stress the need for African ERA frameworks to integrate both exposure modeling and monitoring and biota-based body-residue approaches to fully characterize risks.

THQ values for all detected pesticides are presented in Table S5. At the Gumara and Ribb sites, THQs for current-use pesticides ranged from 0.000001 (carbaryl) to 0.00064 (oxamyl). In Lake Tana, THQs for current-use pesticides ranged from 0.0000033 (imidacloprid) to 0.0002791 (pyrimethanil) (Fig. 4). For legacy organochlorine pesticides in Lake Tana, THQs ranged from 0.000516 (DDT) to 0.1269 (dieldrin). Particularly, dieldrin showed the highest THQ among all pesticides, highlighting its dominant contribution to non-carcinogenic risk (Fig. 4). The cumulative non-carcinogenic risk (Hazard Index, HI) for legacy pesticides was 0.171, while the cumulative HI for current-use pesticides was < 0.001, indicating that overall non-carcinogenic risk remains below the safety threshold (HI < 1).

Bild vergrößern

The Incremental Lifetime Cancer Risk (ILCR) was calculated for DDT, lindane, and dieldrin from dietary exposure to O. niloticus from Lake Tana. The cumulative ILCR for these three carcinogenic pesticides was 0.000109, slightly above the USEPA action level of 0.0001, indicating potentially unacceptable lifetime cancer risk. Dieldrin is the primary driver of carcinogenic risk, far exceeding the ILCR of other detected pesticides (Fig. 5, Table S5). This study provides a dual-perspective health risk assessment, distinguishing between risks posed by current-use pesticides and those from persistent legacy organochlorines in Lake Tana’s fish. The non-carcinogenic risk assessment showed that all individual THQs were substantially below 1, and the cumulative Hazard Index (HI = 0.171 for legacy pesticides; <0.001 for current-use) confirms that chronic consumption of fish from Lake Tana, Gumara, and Ribb is unlikely to cause systemic health effects. These results are consistent with studies from other Ethiopian water bodies, such as Lake Ziway and northern Rift Valley lakes, where non-carcinogenic risk indices for both organochlorines and current-use pesticides were generally below the threshold of concern, indicating limited immediate systemic risk (Yohannes et al. 2014; Tarekegn et al. 2025). risks, While DDT and lindane posed negligible or tolerable risks, dieldrin contributed the largest carcinogenic risk, slightly exceeding the USEPA ILCR benchmark of 0.0001. The cumulative ILCR (0.000109) indicates that lifetime exposure to carcinogenic residues in O. niloticus may be unacceptable. This pattern is similar to findings from Lake Ziway, where legacy organochlorines such as lindane and DDT and certain OCPs were the primary drivers of cumulative cancer risk in fish (Yohannes et al. 2014; Tarekegn et al. 2025). In contrast, lindane and DDT concentrations reported in the Volta Basin of Ghana were within U.S. EPA regulatory limits, indicating lower associated health risks compared to those observed in the present study (Magna et al. 2022).

Bild vergrößern

The detected pesticides in the water of the Lake Tana sub-basin included the legacy pesticides DDTs, p,p’-DDE, lindane, endosulfan, endrin, dieldrin and O,P’-DDT, as well as cypermethrin and alachlor (Table S6). Bifenthrin and chlorothalonil were reported at levels below the limit of detection. Alachlor was detected only at very low concentrations, resulting in negligible PAF and msPAF values, and was therefore excluded from the risk figures to improve graphical clarity. In contrast, O,P′-DDT could not be included in probabilistic risk calculations due to the absence of an SSD, which prevented PAF estimation.

The probabilistic risk assessment approach (PAF and msPAF) indicated that endosulfan concentrations posed high ecological risks, particularly in the Ribb River (in Fogera and Libokemkem district) (PAF acute: 0.60, PAF chronic: 0.76), Megech River (in the Dembia district) (PAF acute: 0.59, PAF chronic: 0.74), and Gumara River (Fogera) (PAF acute: 0.56, PAF chronic: 0.72). Within Lake Tana, the highest acute PAF values for endosulfan were observed at Bahir Dar (PAF acute: 0.57, PAF chronic: 0.72) (Fig. 6). Cypermethrin, detected at Zegie (PAF acute: 0.26, PAF chronic: 0.31) and St. George church (LT3) (PAF acute: 0.45, PAF chronic: 0.44), exhibited substantial risks. The highest risks of measured dieldrin (PAF acute: 0.17, PAF chronic: 0.51), endrin (PAF acute: 0.42, PAF chronic: 0.79) and lindane (PAF acute: 0.03, PAF chronic: 0.10) concentrations were all observed for Ribb River, followed by Megech River, Gumara River, Gorgora, Sekelet, Bahir Dar and Deke. p,p’-DDE was frequently detected in both lake and river water bodies, but at lower risk levels (PAF acute: 0.01–0.02, PAF chronic: 0.08–0.10), with the highest concentrations in the northern and southern parts of Lake Tana and at river mouths (Table S6). Wetlands also exhibited traces of p,p’-DDE, indicating pesticide runoff into adjacent aquatic habitats. p,p’-DDE is a breakdown product of DDT and is known for its persistence and bioaccumulation in aquatic ecosystems. Its presence, even at low concentrations, can have long-term effects on aquatic organisms (Giroux et al. 2024).

Bild vergrößern

When multiple substances were combined (msPAF), the ecological risk increased significantly. In the Ribb River, the combined presence of DDTs, dieldrin, endosulfan, endrin, and lindane resulted in high risk levels (msPAF acute: 0.86, msPAF chronic: 0.99), with similar values also calculated for the other sites (Fig. 7). Only for one site in the lake region (Zegie), where only cypermethrin and p,p’-DDE were detected, the combined risk was lower (msPAF acute: 0.27, msPAF chronic: 0.37) (Table S6). The high ecological risks indicated by the PAF and msPAF analyses in the Lake Tana sub-basin are consistent with findings from other African freshwater ecosystems impacted by agricultural pesticide use. For example, in Lake Ziway, Ethiopia, Merga et al. (2021) reported moderate to high ecological risks from current-use pesticides using a probabilistic risk assessment approach, although legacy organochlorine pesticides were not included in that study. The substantially higher msPAF values observed in the present study highlight the additional contribution of co-occurring legacy OCPs to mixture toxicity in the Lake Tana system. Importantly, there is very limited research in Ethiopia and across African freshwater systems that applies mixture-based probabilistic risk assessment approaches such as PAF and msPAF, which limits regional comparisons. These findings also align with the growing consensus that mixture toxicity is a dominant driver of pesticide impacts in aquatic ecosystems. Dietrich et al. (2022) demonstrated that regulatory risk assessments based solely on single-substance evaluations frequently underestimate ecological risks and that pragmatic mixture-based approaches, grounded in concentration addition and Species Sensitivity Distributions (SSDs), provide a more realistic assessment of impacts on aquatic communities.

Bild vergrößern

The calculated risk quotients showed ecological risks associated with sediment-bound pesticides across all sampling sites in the Lake Tana sub-basin (Fig. 8). Among the assessed pesticides, lindane consistently exhibited the highest RQ values, exceeding 9000 at all sites, with maximum values observed at Ribb and Gumara (Fig. 8, Table S7). Similarly, endrin and dieldrin showed very high RQ values, ranging from 1000 to 3300. In contrast, DDT and endosulfan displayed moderate to high risks, with RQ values generally exceeding unity across sites. Although the Fogera site showed comparatively lower RQ values, DDT still exceeded risk thresholds.

Bild vergrößern

The high RQ values for organochlorine pesticides (OCPs) in sediments reflect the persistent, hydrophobic, and bioaccumulative nature of these pesticides, which promotes their long-term retention in sediments. Similar patterns have been reported in other Ethiopian freshwater systems, including the Awash River Basin (Dirbaba et al. 2018) and the Tekeze Dam (Teklit, 2016), Lake Ziway (Merga et al. 2021), Akaki River (Kassegne et al. 2020), and Lake Awassa (Yohannes et al. 2013; Wm-Bekele et al. 2024), which point out the widespread legacy of historical pesticides used. Notably, the RQ values of pesticides are markedly higher than those reported for other Ethiopian water bodies. For example, sediment concentrations in the Awash River were reported at 8.33 µg kg⁻¹ for lindane, 13.98 µg kg⁻¹ for DDT, 0.81 µg kg⁻¹ for endosulfan, and 1.0 µg kg⁻¹ for endrin (Dirbaba et al. 2018). Similarly, in Lake Ziway, where endosulfan levels reached 2.69 µg kg⁻¹ (Merga et al. 2021), the calculated risks remain substantially lower than those observed in the Lake Tana sub-basin. The high RQs for lindane, endrin, and dieldrin found in this study indicate a significant potential for adverse effects on benthic invertebrates, such as oligochaetes, chironomids, and amphipods. These organisms play critical roles in sediment food webs and nutrient cycling; consequently, their impairment could trigger cascading effects in the ecosystem (Zhao et al. 2009; Jayaraj et al. 2016).

Beyond Ethiopia, studies across Africa show comparable risks from sediment-associated pesticides in freshwater ecosystems. For instance, Darko et al. (2008) reported RQ values indicating high risk for lindane, moderate risk for DDT, and low risk for dieldrin and endosulfan in Ghanaian lake sediments. Similarly, the application of threshold-based sediment quality benchmarks identified dieldrin as a major ecological threat in the Ugandan portion of Lake Victoria (Wasswa et al. 2011). Furthermore, Buah-Kwofie and Humphries (2017) reported that OCP concentrations in sediments from the iSimangaliso Wetland Park, South Africa, were among the highest recorded both nationally and globally in recent decades. The authors emphasized that sediments within lakes and wetlands act as substantial reservoirs for contaminants, posing significant ecotoxicological threats to internationally recognized biodiversity hotspots. They underscored the urgent need for detailed bioaccumulation and toxicological assessments in such regions. These external findings are highly consistent with the present study, which identifies lindane, endrin, and dieldrin as the primary compounds of concern in the Lake Tana sub-basin.

The Lake Tana sub-basin is highly productive, supporting staple and cash crops such as teff, millet, maize, and vegetables (Merkuz 2017; Agmas and Adugna 2020; Taye et al. 2021). Intensive pesticide use, with sometimes up to twenty-six applications per season, is common, particularly on irrigated cash crops (Merkuz 2017; Agmas and Adugna, 2020; Abaineh et al. 2024). Despite national Integrated Pest Management (IPM) guidelines and awareness campaigns (FDRE MOA & RDCPD 2006), adoption remains limited.

Farmers in the sub-basin often lack formal education and training on pesticide safety, leading to misuse and risky application practices (Agmas and Adugna 2020; Asmare et al. 2022). Economic constraints further drive reliance on older, cheaper, and more toxic pesticides (Hu 2020). Improper application contributes significantly to environmental contamination in the region (Merkuz 2017; Agmas and Adugna 2020; Abaineh et al. 2024). Climate change is also increasing pest pressure, further incentivizing pesticide use (IPCC 2001; Palikhe 2007).

To reduce risks from current-use pesticides, strategies should focus on education, monitoring, and promoting safer practices:

For legacy pesticides, the following strategies could be applied: