Tracing sewage-derived organic matter in fish trophic pathways in the lower Doce River basin, Brazil

Patrícia Santos Fráguas¹ , Débora Reis de Carvalho^1,2, Frederico Fernandes Ferreira³, Carlos Frankl Sperber³ and Paulo Santos Pompeu¹

PDF: Download Here | Supplementary: Sup | Cite this article

Associate Editor: Francisco Araújo

Section Editor: Fernando Pelicice

Editor-in-chief: José Birindelli

Abstract​

---

Introduction​

---

The input of untreated sewage into watercourses is a growing threat to freshwater ecosystems worldwide (Dudgeon et al., 2006; Malmqvist, Rundle, 2002; Wear et al., 2021). In addition to posing serious risks to public health (Akpor, Muchie, 2011; Sojobi, Zayed, 2022), it also affects associated fauna (Malik et al., 2020; Rangel-Buitrago et al., 2024). The presence of these effluents can alter water quality, including shifts in salinity and increased concentrations of nitrates and phosphates (Malik et al., 2020). The increased nutrient availability, in turn, stimulates excessive primary production, which leads to reduced dissolved oxygen levels in the water (Malik et al., 2020; Rangel-Buitrago et al., 2024), negatively impacting aquatic fauna (Devlin, Brodie, 2023; Feng et al., 2023).

These disruptions in primary production and aquatic fauna reflect trophic web changes, as fish feeding habits shift in response to alterations in the availability and quality of food resources (Horka et al., 2023; Neves et al., 2024). In polluted environments, omnivorous species may gain a competitive advantage, as their opportunistic feeding strategies enable them to tolerate fluctuations in resource availability (de Carvalho et al., 2019a, 2020a; Neves et al., 2024). Notably, the omnivorous species Poecilia reticulata has been reported to feed directly from sewage in polluted sites (de Carvalho et al., 2019b). Detritivorous species may also be favored by sewage pollution (de Carvalho et al., 2020a), as the discharge of organic matter increases detrital food sources (Terra et al., 2013), and they are capable of adjusting their diets according to pollution (Prado et al., 2020). Seasonality also influences food availability. For instance, sewage may become more diluted during the wet season due to increased rainfall, reducing its influence. In contrast, during the dry season, reduced water flow can lead to higher sewage concentrations and greater ecological impacts (Yunus, Nakagoshi, 2004; Kütter et al., 2023).

In this context, carbon (δ¹³C) and nitrogen (δ¹⁵N) stable isotopes provide valuable insights into the feeding ecology of fish in polluted environments. Nitrogen isotopes reflect the organism’s trophic position, as δ¹⁵N increases at each trophic transfer (Fry, 2006; Anderson, Cabana, 2007). In addition, δ¹⁵N can serve as a tracer of anthropogenic nitrogen inputs, since sewage and other pollution sources often elevate nitrogen concentration in aquatic systems, resulting in distinct isotopic compositions (Wang et al., 2016; de Carvalho et al., 2021). Studies show that, depending on the type of sewage treatment, impacted systems may exhibit depleted δ¹⁵N compositions (Loomer et al., 2015; de Carvalho et al., 2019b, 2020b). Carbon isotopes, in turn, are more conservative across trophic levels and are used to identify the primary basal sources that support the food web (Fry, 2006; Anderson, Cabana, 2007). As sewage-derived carbon becomes an additional basal energy source for aquatic communities, it can be traced through the food web (Loomer et al., 2015; de Carvalho et al., 2019b, 2020b).

The Doce River basin has historically been affected by multiple anthropogenic pressures, including mining activities, deforestation, untreated sewage discharge, intensive land use, and industrial development (ANA, 2016; Hatje et al., 2017; Macêdo et al., 2024). These cumulative impacts were further intensified in 2015 by one of Brazil’s most severe environmental disasters: the collapse of the Fundão tailings dam in Mariana, Minas Gerais (Garcia et al., 2017; Vergilio et al., 2021; Yamamoto et al., 2023b). This event drew national and international attention to the basin, prompting numerous studies focused on assessing the impacts of the disaster on the river’s ichthyofauna (Bonecker et al., 2019; Andrades et al., 2020, 2021; Ferreira et al., 2020; Yamamoto et al., 2023a; de Carvalho et al., 2024; Paiva et al., 2024; da Silva et al., 2024; Fráguas et al., 2025). Human occupation continues to exert pressure on the basin. The lower portion of the Doce River basin lies in the state of Espírito Santo and includes two major cities: Colatina, with a population of 120,033, and Linhares, with 166,786 inhabitants (IBGE, 2022). Colatina, situated upstream in the lower basin, treats only 14.45% of its sewage before discharging it into the Doce River (SNIS, 2023). Although the presence of sewage pollution has been documented in the literature for the Doce River basin (Fraga et al., 2021; Kütter et al., 2023; Yamamoto et al., 2023b), few studies have investigated its impacts on the trophic ecology of fish assemblages.

This study aimed to assess the incorporation of urban sewage-derived carbon into the aquatic food web of the lower Doce River basin. Additionally, we aimed to identify which trophic guilds and species assimilate the highest proportion of sewage-derived carbon. We tested the following hypotheses: 1) during the dry season, sewage has a greater assimilation by the fish assemblage, 2) species that assimilate more sewage usually occupy lower trophic positions, 3) Detritivorous and omnivorous fish are those that most assimilate sewage-derived carbon.

Material and methods

---

Study area. This study was conducted in the lower Doce River basin, within the municipality of Colatina. Colatina has a population of 120,033 inhabitants (IBGE, 2022) and is located 132 km from Vitória, the capital of Espírito Santo State, Brazil. The main economic activities in the region include commerce, services, and a strong industrial and agricultural sector (INCAPER, 2023). The city lies within the Atlantic Forest biome, although most of its area has been altered to support agriculture and livestock activities (INCAPER, 2023). The region receives an average annual rainfall of approximately 1,100 mm, with the rainy season extending from October to April, and the dry season from May to September (INCAPER, 2023). In addition to land-use changes, another anthropogenic pressure is the discharge of untreated sewage into the river. According to data from 2022, only 14.45% of the sewage generated in the city is treated before being discharged into the Doce River (SNIS, 2023). Sampling of fish and potential food sources was conducted once during the dry (September) and wet (February) seasons of 2022.

Fish sampling. Seines, gill nets, and mosquito nets were used for fish sampling during wet and dry seasons in the Doce River. Shallow and littoral zones were sampled using seines (5 m long and 1mm mesh). We used two sets of gillnets, with 10 nets in each set (10 m long, with mesh sizes ranging from 3 to 16 cm stretched mesh). Nets were exposed for 14 h overnight. Near-shore macrophyte banks were sampled using mosquito nets (80 cm in diameter and 1mm mesh) for a period of approximately 1 h. After capture, fish were measured, weighed, and identified. Samples for isotopic analysis were collected for up to five individuals per species. Larger fish had a portion of the muscle tissue removed for isotopic analysis, while smaller fish were analyzed whole, with only the digestive tract removed. Samples were frozen until subsequent laboratory processing. Specimens were deposited at the Ichthyological Collection of Universidade Federal de Lavras(CI-UFLA) (Tab. S1).

Resource sampling. We collected five samples per food resource in each season:aquatic macrophytes (MA), fine particulate organic matter (FPOM), coarse particulate organic matter (CPOM), periphyton (PE), filamentous algae (AL), terrestrial invertebrates (TI), aquatic invertebrates (AI), suspended matter (SM), C3 vegetation (VC3), C4 vegetation (VC4), sewage after treatment (SEWA), and sewage before treatment (SEWB). D-frame nets were used to sample AI, while FPOM was sampled by passing the same net through resuspended sediment. SM was collected by placing a phytoplankton net (45μm mesh) in the water for three minutes. Stones were scraped with a small brush in order to collect PE, and the material was stored with distilled water in plastic bottles. FPOM, MA, TI, AL, VC3 and VC4 were randomly collected in the sampling site. Samples of SEWA and SEWB were collected at the wastewater treatment plant in Colatina, Espírito Santo, with authorization from the Environmental Sanitation Company of Colatina (SANEAR). All samples were kept frozen until laboratory processing. 

In the laboratory, fish muscle and dry food resources were lyophilized for 24 hours. Liquid samples (PE, FPOM, SM, SEWA and SEWB) were filtered through calcined quartz fiber filters (Whatman® QMA), using a filtration apparatus attached to a vacuum pump. Subsequently, all samples were dried and ground to a homogeneous powder and stored in Eppendorf tubes for further analysis. Samples were sent for isotopic analysis at the Central de Análises e Prospecção Química (CAPQ) of the Universidade Federal de Lavras (UFLA).

Isotopic analysis. A total of 114 samples of fish and 115 samples of food resources were analysed isotopically. Isotopic ratios were determined using a Delta V Plus Isotope Ratio Mass Spectrometer (Thermo Fisher) coupled with a Flash IRMS elemental analyzer interfaced with ConFlo IV. Results were expressed as relative differences from the international reference standard, with delta notation (δ ‰), according to the following formula: δX = [(R sample/R standard) – 1] x 10³, where X is ¹³C or ¹⁵N, and R is the isotopic ratio ¹³C/¹²C or ¹⁵N/¹⁴N (Barrie, Prosser, 1996). We applied a mathematical correction to isotopic ratios, to account for lipid variation in fish tissues. The correction formula proposed by Post et al. (2007) was used: ∆δ¹³C = –3.32 + 0.99 × C:N.

Statistical analysis. The trophic structure was seasonally characterized using bi-plot graphs (x axis: δ¹³C and y axis: δ¹⁵N). The bi-plot displays mean values for all species alongside the means and standard deviations for the food resources. This plot allows the visualization of the most important food resources for the community.

Trophic position of each species was estimated for each season using the Bayesian package “tRophicPosition” (Quezada‐Romegialli et al., 2018). We employed the oneBaseline model, using periphyton as the baseline source. Isotopic fractionation factors were set at ∆¹⁵N: 2.9 ± 0.32‰ and ∆¹³C: 1.3 ± 0.3‰ based on values reported by McCutchan et al. (2003).

To evaluate the resource contribution for each species, we used Bayesian mixing models, with the simmr package (Govan et al., 2023). Isotope mixing models included concentration dependence of basal sources to account for differences in %C and C:N ratios. Models were run with 3 Markov chains of 100.000 iterations each, a burn-in of 50.000 iterations and a thinning interval of 50, resulting in 3.000 posterior samples per model. Convergence was assessed using the Gelman-Rubin diagnostic (R̂) (Gelman, Rubin, 1992). Trophic enrichment factors (TEF) were set at ∆¹⁵N: 2.9 ± 0.32‰ and ∆¹³C: 1.3 ± 0.3‰ (McCutchan et al., 2003). Two types of priors were used. For models with ≥4 individuals, we used the default uninformative Dirichlet prior implemented in simmr. For models with ≤3 individuals, informative priors were specified using the simmr_elicit function to improve parameter and identifiability under small sample sizes. Food resources included in the model were defined for each feeding guild. The food resources selected for detritivores were: PE, SM, SEWA, and SEWB. For herbivores, we selected MA, PE, VC4, and CPOM. For omnivores, we selected AI, SEWA, SEWB, and TI. For invertivores, we selected AI and TI. For the piscivore guild, we selected AI, TI, and Fish. The Fish used as a resource comprised small-sized species and species documented in the literature as potential prey of the piscivorous species analyzed in this study. The selected species were Astyanax lacustris, Astyanax sp., Geophagus brasiliensis, Oreochromis niloticus, Hoplias malabaricus, Oligosarcus argenteus, Knodus moenkhausii, Poecilia vivipara,and Deuterodon taeniatus. To estimate the sewage contribution through all compartments of the food web, we also applied the mixing models for FPOM, SM, AI and TI. For FPOM and SM, the resources used in the model were: CPOM, PE, and SEWB. The TEF for these two food resources was set at ∆¹⁵N: 0.5 ± 0.4‰ and ∆¹³C: 0.5 ± 0.4‰ to account for potential isotopic shifts associated with microbial processing. For aquatic invertebrates, the resources used in the model were FPOM, PE, SEWB, and SM. For terrestrial invertebrates, the selected resources were VC3 and VC4. TEF for aquatic and terrestrial invertebrates was the same as that used for fish.

To visualize trophic interactions between food resources and feeding guilds, we constructed flow diagrams with the DiagrammeR package (Iannone, Roy, 2016), using the grViz() function. The flow diagram represents directed relationships, where arrows indicate the relative contribution of each food resource for the feeding guilds and the contribution of sewage to food resources. The width of the arrows is scaled proportionally to indicate the estimated contribution of each food resource summarized by feeding guilds. All analyses were performed using the R software (R Development Core Team, 2024).

Results​

---

We sampled a total of 114 individuals from 29 species (Tab. 1), 46 individuals from 15 species in the dry season and 68 individuals from 20 species in the wet season. Six species were sampled in both seasons, whereas nine species were exclusive to the dry season and 14 to the wet season (Fig. 1). Detritivore fish were only sampled during the wet season (Fig. 2A; Tab. S2), while herbivores were only sampled during the dry season (Fig. 2B; Tab. S3).

TABLE 1 | Mean and standard deviation of ẟ¹³C and δ¹⁵N in fish species and feeding resources in the dry and wet seasons.

Season	Dry					Wet
Species/Resource	N	ẟ13C		ẟ15N		N	ẟ13C		δ15N
		Mean	SD	Mean	SD		Mean	SD	Mean	SD
Detritivore
Hoplosternum littorale						5	-24.33	3.17	10.8	1.17
Hypoptopomatinae sp.						4	-24.64	3.08	8.33	2.34
Hypostomus affinis						1	-26.25	0	12.51	0
Prochilodus argenteus						1	-17.67	0	12.26	0
Prochilodus costatus						1	-16.14	0	12.58	0
Herbivore
Metynnis lippincottianus	2	-21.12	3.19	9.5	3.08
Invertivore
Pimelodus maculatus	1	-18.61	0	11.99	0	5	-23.13	2.83	13.7	1.34
Saxatilia lepidota						2	-20.39	1.46	11.73	1.29
Microphis lineatus	5	-17.26	1.07	5.73	1.12
Pachyurus adspersus	5	-20.19	3.56	12.91	0.53
Phalloceros sp.	1	-17.83	0	5.46	0
Omnivore
Astyanax lacustris						5	-21.81	1.90	11.05	1.83
Awaous tajasica						5	-18.05	1.90	11.98	0.96
Characidium sp.						1	-24.12	0	13.51	0
Clarias gariepinus						1	-25.54	0	13.44	0
Deuterodon cf. taeniatus						5	-21.40	2.03	12.69	1.06
Geophagus brasiliensis	5	-16.36	2.35	8.69	3.63	1	-20.56	0	12.48	0
Knodus moenkhausii	5	-16.77	1.38	10.19	0.73	5	-19.37	1.13	11.81	1.62
Oreochromis niloticus	5	-14.12	1.22	6.07	0.36	10	-22.80	1.62	7.92	2.7
Poecilia vivipara	2	-16.61	1.05	6.24	1.06	5	-22.88	3.25	10.44	2.12
Astyanax sp.	2	-16.36	0.17	4.75	0.75
Gymnotus cf. sylvius	2	-25.83	0.74	11.49	3.00
Piscivore
Centropomus sp.						1	-26.27	0	16.17	0
Cichla kelberi						1	-19.24	0	14.04	0
Oligosarcus argenteus						1	-20.96	0	11.13	0
Pygocentrus nattereri	6	-20.05	0.58	13.48	0.69	8	-20.21	0.39	13.62	0.99
Hoplias intermedius	3	-18.59	0.15	9.09	1.85
Hoplias malabaricus	1	-20.63	0	13.96	0
Salminus brasiliensis	1	-19.50	0	14.04	0
Feeding Resources
Aquatic Invertebrates	5	-22.31	1.15	9.49	2.04	5	-21.51	1.29	10.75	1.15
CPOM	5	-30.53	1.87	4.98	3.05	5	-30.47	0.32	2.62	1.04
FPOM	5	-23.05	0.52	5.18	0.98	5	-23.75	0.71	4.37	0.77
Macrophyte	5	-29.26	0.96	8.21	0.57	5	-26.54	0.64	9.07	0.99
Periphyton	5	-18.59	2.98	7.07	0.71	5	-23.23	0.52	7.23	1.69
Sewage – After treatment	5	-21.59	0.25	0.79	0.27	5	-23.47	0.28	1.46	0.78
Sewage – Before Treatment	5	-24.91	0.53	2.41	0.52	5	-23.86	0.24	3.58	0.23
Suspended Matter	5	-26.6	3.7	7.32	1.18	5	-25.6	2.07	5.86	0.5
Terrestrial Insect	5	-25.07	2.85	5.97	4.22	5	-21.5	6.08	6.99	1.1
Vegetation C3	8	-29.75	0.89	3.86	2.69	8	-29.5	1.59	6.83	1.34
Vegetation C4	2	-13.21	1.64	6.76	2.05	2	-13.83	0.18	3.6	0.91
Algae	5	-23.41	4.45	6.31	1.33

FIGURE 1 | Trophic structure of fish assemblage in Colatina in the dry (A) and wet (B) seasons. Graphics A, B show the mean values for each fish species and the mean and standard deviation for food resources. Food resources: Filamentous algae (AL), periphyton (PE), coarse particulate organic matter (CPOM), aquatic invertebrates (AI), terrestrial invertebrates (TI), suspended matter (SM), macrophyte (MA), fine organic particulate matter (FPOM), C3 vegetation (VC3), C4 vegetation (VC4), sewage before treatment (SEWB), sewage after treatment (SEWA).

FIGURE 2 | Percentage of food resources sustaining each feeding guild during wet and dry seasons in Colatina. A. Detritivore, B. Herbivore, C. Invertivore, D. Omnivore, E. Piscivore. Food resources: Aquatic invertebrates (AI); Coarse particulate organic matter (CPOM); Fish; Macrophytes (MA); Periphyton (PE); Sewage before treatment (SEWB); Sewage after treatment (SEWA); Suspended matter (SM); Terrestrial invertebrates (TI) and C4 Vegetation (VC4).

Mixing model diagnostics indicated satisfactory performance. All parameters across all models showed R̂ <1.004, indicating adequate convergence and mixing of chains. Among models including three or more sources, most posterior correlations were moderate. However, five out of 35 cases showed strong negative correlations (r ≤ -0.85; Tab. S4). Two-source models were excluded from this assessment because perfect negative correlation is mathematically constrained in such models. Posterior median source contributions and their 95% credible intervals (2.5%–97% quantiles) for all species are reported in Tabs. S2, S3.

Among invertivorous fish, Pimelodus maculatus showed greater assimilation of aquatic invertebrates in the wet season (Fig. 2C). Sewage-derived assimilation through aquatic invertebrates was higher for Pachyurus adspersus in the dry season (Tabs. S2, S3). Among the omnivores, sewage (both before and after treatment) accounted for approximately half of the species assimilation (Fig. 2D). Oreochromis niloticus showed high sewage assimilation in both seasons, especially during the wet season, which accounted for up to 90% of its assimilated diet. For piscivores, the assimilation of food resources had similar ranges between seasons, and Pygocentrus nattereri had higher assimilation of aquatic invertebrates in both seasons (Fig. 2E).

Trophic positions ranged from 2.02 (95% CI: 2.00–2.82) to 5.07 (95% CI: 2.04–9.47) in the dry season, and from 2.21 (95% CI: 2.00–3.66) to 5.67 (95% CI: 2.14–9.55) in the wet season (Tab. S5). The highest trophic positions were observed in piscivorous species, except Hoplias intermedius and O. argenteus in the dry and wet seasons, respectively. Among the six species sampled in both seasons, Oreochromis niloticus consistently had the lowest values. Geophagus brasiliensis and Poecilia vivipara had lower values in the dry season compared to the wet season. The other three species (Knodus moenkhausii, Pimelodus maculatus,and Pygocentrus nattereri) presented similar values in both seasons (Fig. 3).

FIGURE 3 | Trophic position occupied by each species in the dry (A) and wet (B) seasons in Colatina, lower Doce River, Brazil.

Sewage-derived carbon contributed to multiple trophic levels in both seasons. Assimilation was similar between seasons within each guild: invertivores (~22%), omnivores (~60%), and piscivores (~30%). Among guilds, the omnivores had the highest assimilation, accounting for more than 60% in both seasons (Fig. 4). Regarding sewage contributions to food resources, FPOM showed the highest values in both seasons, ranging from 44% in the dry season and 60% in the wet season (Fig. 5).

FIGURE 4 | Food resources that sustain the fish community in Colatina, lower Doce River, Brazil. Food resources: Aquatic invertebrates (AI); Coarse particulate organic matter (CPOM); Fine particulate organic matter (FPOM); FS* (specific fish species that are common preys to the piscivores); Macrophytes (MA); Periphyton (PE); ewage before treatment (SEWB); Sewage after treatment (SEWA); Suspended matter (SM); Terrestrial invertebrates (TI); C3 Vegetation (VC3), and C4 Vegetation (VC4). The percentages shown inside the circles indicate the proportion of sewage-derived material contributing to each component of the trophic web, where 100% corresponds to untreated and treated sewage samples.

FIGURE 5 | Estimated sewage contribution for fine particulate organic matter (FPOM) (A), suspended matter (SM) (B), terrestrial invertebrates (TI) (C) and aquatic invertebrates (AI) (D). Food resources considered in the partitioning model: Coarse particulate organic matter (CPOM); Fine particulate organic matter (FPOM); Periphyton (PE); Sewage before treatment (SEWB); Suspended matter (SM); C3 Vegetation (VC3) and C4 Vegetation (VC4). Boxplots represent posterior distributions of source contributions from mixing models. Central line indicated the median, boxes represent the interquartile range (25–75%), and whiskers indicate the 95% credible interval (2.5–97.5%).

Discussion​

---

Our results indicate that sewage–derived resources were assimilated at similar levels across seasons, rejecting our first hypothesis. Species with higher assimilation of sewage-derived resources occupied the lowest trophic positions within the assemblage, supporting our second hypothesis. The third hypothesis was partially confirmed, as omnivorous species assimilated high proportions of sewage-derived resources, but detritivorous species were not sampled in the dry season.

Although sewage assimilation did not differ between seasons in our dataset, the influence of sewage inputs is often expected to be higher during the dry season, due to reduced rainfall, which makes sewage less diluted than in the wet season (Yunus, Nakagoshi, 2004; Kütter et al., 2023). The absence of detritivores in the dry season limits a full evaluation, as detritivores could increase sewage-derived carbon assimilation and potentially enhance its importance during the period.

In our study, untreated sewage samples showed depleted δ¹⁵N values, averaging 3.5‰ in the wet season and 2.4‰ in the dry season, which were reflected in the fish isotopic values. Sewage was also reflected in the resources, since we found that sewage-derived carbon accounts for 25% to 60% of key basal resources such as suspended matter and FPOM. These low δ¹⁵N values likely reflect strong sewage influence, given that over 80% of local sewage is discharged untreated (SNIS, 2023). Effluents receiving only primary treatment, or none, typically carry elevated ammonia concentrations, with depleted δ¹⁵N values (Gaston, Suthers, 2004; Hicks et al., 2017), whereas secondary treatment tends to enrich δ¹⁵N through nitrification and denitrification, which preferentially remove ¹⁴N (Heaton, 1986; Loomer et al., 2015; Alldred et al., 2023). The depleted δ¹⁵N values observed in Colatina are consistent with a neighboring basin reporting values around 1.5‰ (de Carvalho et al., 2019b, 2020a) and with Canadian studies documenting depleted δ¹⁵N in fish downstream of poor-quality (Loomer et al., 2015). Furthermore, the isotopic compositions of fish in our study closely mirror those recorded in the same area two years earlier where δ¹⁵N ranged from 5 to 15‰ (de Carvalho et al., 2024), indicating that sewage assimilation remains a persistent trend.

Several species assimilated a greater proportion of sewage and occupied the lowest trophic positions. Among them, the omnivorous O. niloticus deserves special attention, as it assimilated the highest proportion of sewage-derived carbon in both seasons. During the wet season, when sewage assimilation was more pronounced, this resource accounted for up to 90% of the species’ diet. Oreochromis niloticus is non-native to the Doce River basin (Bueno et al., 2021), and its success may be partly attributed to its trophic plasticity (Abelha et al., 2001), which allows it to shift its primary food items according to environmental availability. Similar dietary shifts have been observed in other non-native species, such as Poecilia reticulata,which has been reported feeding directly on sewage in the Rio das Velhas River basin (de Carvalho et al., 2019b). These patterns highlight the ability of invasive species to establish themselves under adverse environmental conditions and also increase pressure on native fauna. Most omnivorous species assimilated at least 50% of sewage-derived carbon, confirming that this feeding guild is particularly adaptable to environmental changes (de Carvalho et al., 2019a,b; Neves et al., 2024).

Detritivorous species were only sampled during the wet season. Two species, Hypoptomatinaesp. and Hoplosternum littorale,showed higher assimilation of sewage-derived carbon, with values of 35% and 28%, respectively. For the remaining species, the contribution of sewage-derived carbon was below 10%, with periphyton and suspended matter being the main food sources. In Colatina, the Doce River flows through a broad channel. Therefore, the influence of riparian vegetation on shading is reduced, allowing greater light penetration and potentially stimulating periphyton growth (McCall et al., 2017). However, during the wet season, increased water flow contributes to higher levels of suspended organic material, an important food resource for fish; however, the resulting turbidity can negatively affect periphyton growth (Ren et al., 2021). It is important to note that the consumption does not necessarily reflect resource availability, which was not quantified in this study. Although detritivores can adjust their diets to environmental conditions (Prado et al., 2020), they are often specialized due to morphological constraints (Bowen, 1983) and may continue to consume periphyton even when it is less abundant.

Beyond freshwater species, sewage inputs also affect estuarine species. In our sampling, three estuarine species were recorded: Awaous tajasica, Centropomus sp., and Microphis lineatus. These species move into the lower Doce River, where they use the area for feeding, shelter, and nursery functions (Sarmento-Soares, Martins-Pinheiro, 2022). Some, such as Centropomus sp., also have economic importance (Sarmento-Soares, Martins-Pinheiro, 2022). The invertivore Microphis lineatus assimilated sewage-derived carbon, representing 11% of its diet during the dry season. Considering that sewage also introduces additional pollutants (Strauch, Erzinger, 2021; Roveri et al., 2025), the assimilation of sewage-derived carbon by estuarine species may still lead to broader impacts on estuarine species and the marine environment.

Stable isotope analysis revealed that sewage inputs into Colatina have been incorporated into the local food web during both seasons. This is consistent with fish being effective indicators of nitrogen sources in freshwater (de Carvalho et al., 2020b) and estuarine systems (Schlacher et al., 2005), as their distribution across trophic levels enables the tracking of isotopic signals across the food web (Schlacher et al., 2005).

Domestic sewage also introduces multiple contaminants into the ecosystem, including pathogenic microorganisms, detergents, heavy metals, and pharmaceutical compounds (Strauch, Erzinger, 2021; Gomes et al., 2022; Roveri et al., 2025). Their presence alongside sewage raises concern about their movement through the food web and potential impacts on both ecosystem integrity and public health. In addition, industrial effluents may also contain high metal concentrations (Strauch, Erzinger, 2021), and, together with the 2015 Fundão dam collapse, have intensified metal pollution in the region (Queiroz et al., 2018), and its impacts have already been documented (Ferreira et al., 2020; Vergilio et al., 2021; Yamamoto et al., 2023a).

Although our findings are restricted to Colatina, the Doce River flows through other major urban centers, such as Ipatinga, Governador Valadares, and Linhares. As these cities are also influenced by sewage discharge (SINISA, 2025), future research should evaluate pollution gradients along the river and test whether the patterns observed here are consistent across the basin. At the local scale, public policies in Colatina should prioritize increasing the proportion of treated sewage and expanding treatment infrastructure capacity to reduce pollutant inputs and mitigate their ecological impacts.

Acknowledgments​

---

The authors thank Ingryd Faria and the Fish Ecology Laboratory at UFLA for their assistance with laboratory processing, and the Central of Analysis and Chemical Prospecting of UFLA (CAPQ – UFLA) for its support and collaboration in conducting stable isotope analyses.

References​

---

Abelha MCF, Agostinho AA, Goulart E. Plasticidade trófica em peixes de água doce. Acta Sci Biol Sci. 2001; 23(2):425–34.

Akpor OB, Muchie M. Environmental and public health implications of wastewater quality. African J Biotechnol. 2011; 10(13):2379–87.

Alldred FC, Gröcke DR, Leung CY, Wright LP, Banfield N. Diffuse and concentrated nitrogen sewage pollution in island environments with differing treatment systems. Sci Rep. 2023; 13(4838). https://doi.org/10.1038/s41598-023-32105-6

Agência Nacional das Águas e Saneamento Básico (ANA). Encarte especial sobre a Bacia do Rio Doce: rompimento da barragem de Mariana/MG. Conjuntura dos Direitos Hídricos no Brasil – Informe 2015. Brasília, DF: Superintendência de Planejamento de Recursos Hídricos – SPR; 2016.

Anderson C, Cabana G. Estimating the trophic position of aquatic consumers in river food webs using stable nitrogen isotopes. J North Am Benthol Soc. 2007; 26(2):273–85. https://doi.org/10.1899/0887-3593(2007)26[273:ETTPOA]2.0.CO;2

Andrades R, Guabiroba HC, Hora MSC, Martins RF, Rodrigues VLA, Vilar CC et al. Early evidences of niche shifts in estuarine fishes following one of the world’s largest mining dam disasters. Mar Pollut Bull. 2020; 154:111073. https://doi.org/10.1016/j.marpolbul.2020.111073

Andrades R, Martins RF, Guabiroba HC, Rodrigues VLA, Szablak FT, Bastos KV et al. Effects of seasonal contaminant remobilization on the community trophic dynamics in a Brazilian tropical estuary. Sci Total Environ. 2021; 801:149670. https://doi.org/10.1016/j.scitotenv.2021.149670

Barrie A, Prosser S. Automated analysis of light-element stable isotopes by isotope ratio mass spectrometry. New York: Marcel Dekker; 1996.

Bonecker ACT, Castro MS, Costa PG, Bianchini A, Bonecker SLC. Larval fish assemblages of the coastal area affected by the tailings of the collapsed dam in southeast Brazil. Reg Stud Mar Sci. 2019; 32:100848. https://doi.org/10.1016/j.rsma.2019.100848

Bowen SH. Detritivory in neotropical fish communities. Environ Biol Fishes. 1983; 9(2):137–44. https://doi.org/10.1007/BF00690858

Bueno ML, Magalhães ALB, Andrade Neto FR, Alves CBM, Rosa D M, Junqueira NT et al. Alien fish fauna of southeastern Brazil: species status, introduction pathways, distribution and impacts. Biol Invasions. 2021; 23(10):3021–34. https://doi.org/10.1007/s10530-021-02564-x

de Carvalho DR, Alves CBM, Moreira MZ, Pompeu PS. Trophic diversity and carbon sources supporting fish communities along a pollution gradient in a tropical river. Sci Total Environ. 2020a; 738:139878. https://doi.org/10.1016/j.scitotenv.2020.139878

de Carvalho DR, Alves CBM, Flecker AS, Sparks JP, Moreira MZ, Pompeu PS. Using δ15N of periphyton and fish to evaluate spatial and seasonal variation of anthropogenic nitrogen inputs in a polluted Brazilian river basin. Ecol Indic. 2020b; 115:106372. https://doi.org/10.1016/j.ecolind.2020.106372

de Carvalho DR, Castro DMP, Callisto M, Chaves AJM, Moreira MZ, Pompeu PS. Stable isotopes and stomach content analyses indicate omnivorous habits and opportunistic feeding behavior of an invasive fish. Aquat Ecol. 2019a; 53(3):365–81. https://doi.org/10.1007/s10452-019-09695-3

de Carvalho DR, Ferreira FF, Dergam JA, Moreira MZ, Pompeu PS. Food web structure of fish communities of Doce River, 5 years after the Fundão dam failure. Environ Monit Assess. 2024; 196(3):300. https://doi.org/10.1007/s10661-024-12395-7

de Carvalho DR, Flecker AS, Alves CBM, Sparks JP, Pompeu PS. Trophic responses to aquatic pollution of native and exotic livebearer fishes. Sci Total Environ. 2019b; 681:503–15. https://doi.org/10.1016/j.scitotenv.2019.05.092

de Carvalho DR, Sparks JP, Flecker AS, Alves CBM, Moreira MZ, Pompeu PS. Nitrogen pollution promotes changes in the niche space of fish communities. Oecologia. 2021; 197(2):485–500. https://doi.org/10.1007/s00442-021-05029-z

Devlin M, Brodie J. Nutrients and Eutrophication. Mar. Pollution-monitoring, Manag. Mitig. 2023. p.75–100. https://doi.org/10.1007/978-3-031-10127-4_4

Dudgeon D, Arthington AH, Gessner MO, Kawabata Z, Knowler DJ, Lévêque C et al.Freshwater biodiversity: importance, threats, status and conservation challenges. Biol Rev. 2006; 81(2):163–82. https://doi.org/10.1017/S1464793105006950

Feng K, Deng W, Zhang Y, Tao K, Yuan J, Liu J et al. Eutrophication induces functional homogenization and traits filtering in Chinese lacustrine fish communities. Sci Total Environ. 2023; 857:159651. https://doi.org/10.1016/j.scitotenv.2022.159651

Ferreira FF, de Freitas MBD, Szinwelski N, Vicente N, Medeiros LCC, Schaefer CEGR et al. Impacts of the Samarco tailing dam collapse on metals and arsenic concentration in freshwater fish muscle from Doce River, Southeastern Brazil. Integr Environ Assess Manag. 2020; 16(5):622–30. https://doi.org/10.1002/ieam.4289

Fraga M S, Silva DD, Reis GB, Guedes HAS, Elesbon AAA. Temporal and spatial trend analysis of surface water quality in the Doce River basin, Minas Gerais, Brazil. Environ Dev Sustain. 2021; 23(8):12124–50. https://doi.org/10.1007/s10668-020-01160-8

Fráguas PS, de Carvalho DR, Ferreira FF, Dergam JA, Sperber CF, Pompeu PS. Assessing temporal shifts in trophic diversity in fish assemblages after the Fundão dam collapse. Environ Monit Assess. 2025; 197(4):390. https://doi.org/10.1007/s10661-025-13830-z

Fry B. Stable isotope ecology. Springer; 2006.

Garcia LC, Ribeiro DB, Roque FO, Ochoa-Quintero JM, Laurance WF. Brazil’s worst mining disaster: corporations must be compelled to pay the actual environmental costs: Corporations. Ecol Appl. 2017; 27(1):5–09. https://doi.org/10.1002/eap.1461

Gaston TF, Suthers IM. Spatial variation in δ13C and δ15N of liver, muscle and bone in a rocky reef planktivorous fish: The relative contribution of sewage. J Exp Mar Bio Ecol. 2004; 304(1):17–33. https://doi.org/10.1016/j.jembe.2003.11.022

Gelman A, Rubin DB. Inference from iterative simulation using multiple sequences. Stat Sci. 1992; 7(4):457–72. https://doi.org/10.1214/ss/1177011136

Gomes MP, Brito JCM, Vieira F, Kitamura RSA, Juneau P. Emerging contaminants in streams of Doce River watershed, Minas Gerais, Brazil. Front Environ Sci. 2022; 9:801599. https://doi.org/10.3389/fenvs.2021.801599

Govan E, Jackson AL, Inger R, Bearhop S, Parnell AC. simmr: a package for fitting stable isotope mixing models in R. 2023.

Hatje V, Pedreira RMAA, Rezende CE, Schettini CAF, Souza GC, Marin DC et al.The environmental impacts of one of the largest tailing dam failures worldwide. Sci Rep. 2017; 7(1):10706. https://doi.org/10.1038/s41598-017-11143-x

Heaton THE. Isotopic studies of nitrogen pollution in the hydrosphere and atmosphere: A review. Chem Geol Isot Geosci Sect. 1986; 59(C):87–102. https://doi.org/10.1016/0168-9622(86)90059-X

Hicks KA, Loomer HA, Fuzzen MLM, Kleywegt S, Tetreault GR, McMaster ME et al.δ15N tracks changes in the assimilation of sewage-derived nutrients into a riverine food web before and after major process alterations at two municipal wastewater treatment plants. Ecol Indic. 2017; 72:747–58. https://doi.org/10.1016/j.ecolind.2016.09.011

Horka P, Musilova Z, Holubova K, Jandova K, Kukla J, Rutkayova J et al. Anthropogenic nutrient loading affects both individual species and the trophic structure of river fish communities. Front Ecol Evol. 2023; 10: 1076451. https://doi.org/10.3389/fevo.2022.1076451

Iannone R, Roy O. Package “DiagrammerR”. 2016.

Instituto Brasileiro de Geografia e Estatística (IBGE). Censo demográfico 2022: panorama dos municípios brasileiros 2022. Available from: https://censo2022.ibge.gov.br/panorama

Instituto Capixaba de Pesquisa, Assistência Técnica e Extensão Rural (INCAPER). Programa de assistência técnica e extensão rural. 2023:1–52.

Kütter VT, Martins GS, Brandini N, Cordeiro RC, Almeida JPA, Marques ED. Impacts of a tailings dam failure on water quality in the Doce River: the largest environmental disaster in Brazil. J Trace Elem Miner. 2023; 5:100084. https://doi.org/10.1016/j.jtemin.2023.100084

Loomer HA, Oakes KD, Schiff SL, Taylor WD, Servos MR. Use of stable isotopes to trace municipal wastewater effluents into food webs within a highly developed river system. River Res Appl. 2015; 31(9):1093–100. https://doi.org/10.1002/rra.2826

Macêdo AKS, Oliveira TCM, Brighenti LS, Santos HB, Thomé RG. Socio-environmental impacts on the Doce River basin, Brazil: a review from historic pollution to large disaster events. Int J Environ Sci Technol. 2024; 21(2):2339–56. https://doi.org/10.1007/s13762-023-05350-5

Malik DS, Sharma AK, Sharma AK, Thakur R, Sharma M. A review on impact of water pollution on freshwater fish species and their aquatic environment. Adv Environ Pollut Manag Wastewater Impacts Treat Technol. 2020:10–28. https://doi.org/10.26832/aesa-2020-aepm-02

Malmqvist B, Rundle S. Threats to the running water ecosystems of the world. Environ Conserv. 2002; 29(2):134–53. https://doi.org/10.1017/S0376892902000097

McCall SJ, Hale MS, Smith JT, Read DS, Bowes MJ. Impacts of phosphorus concentration and light intensity on river periphyton biomass and community structure. Hydrobiologia. 2017; 792(1):315–30. https://doi.org/10.1007/s10750-016-3067-1

McCutchan JH, Lewis WM, Kendall C, McGrath CC. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos. 2003; 102(2):378–90. https://doi.org/10.1034/j.1600-0706.2003.12098.x

Neves MP, Delariva RL, Perkins DM, Fialho CB, Kratina P. Trophic plasticity of omnivorous fishes in natural and human-dominated landscapes. Limnol Oceanogr. 2024; 69(1):189–202. https://doi.org/10.1002/lno.12467

Paiva LC, de Carvalho DR, Ferreira FF, Dergam JA, Moreira MZ, Sperber CF et al. Trophic ecology of a small characid reflects the degradation of a basin after the rupture of an ore tailings dam. Aquat Ecol. 2024; 59(1):349–62. https://doi.org/10.1007/s10452-024-10167-6

Prado MR, Carvalho DR, Alves CBM, Moreira MZ, Pompeu PS. Convergent responses of fish belonging to different feeding guilds to sewage pollution. Neotrop Ichthyol. 2020; 18(1):190045. https://doi.org/10.1590/1982-0224-2019-0045

Post DM, Layman CA, Arrington DA, Takimoto G, Quattrochi J, Montaña CG. Getting to the fat of the matter: Models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia. 2007; 152:179–89. https://doi.org/10.1007/s00442-006-0630-x

Queiroz HM, Nóbrega GN, Ferreira TO, Almeida LS, Romero TB, Santaella ST et al. The Samarco mine tailing disaster: a possible time-bomb for heavy metals contamination? Sci Total Environ. 2018; 637–638:498–506. https://doi.org/10.1016/j.scitotenv.2018.04.370

Quezada‐Romegialli C, Jackson AL, Hayden B, Kahilainen KK, Lopes C, Harrod C. tRophicPosition, an r package for the Bayesian estimation of trophic position from consumer stable isotope ratios. Methods Ecol Evol. 2018; 9(6):1592–99. https://doi.org/10.1111/2041-210X.13009

R Development Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2024. Available from: https://www.r-project.org/

Rangel-Buitrago N, Galgani F, Neal WJ. Addressing the global challenge of coastal sewage pollution. Mar Pollut Bull. 2024; 201:116232. https://doi.org/10.1016/j.marpolbul.2024.116232

Ren W, Yao Y, Zhang Z, Cao Y, Yuan C, Wang H et al. Changes of periphyton abundance and biomass driven by factors specific to flooding inflow in a river inlet area in Erhai Lake, China. Front Environ Sci. 2021; 9:680718. https://doi.org/10.3389/fenvs.2021.680718

Roveri V, Souza UP, Costa JHA, Toma W, Caivano SA, Correia AT et al. Occurrence of pharmaceuticals and cocaine in liver and muscle of different fish species from a Brazilian coastal zone. Reg Stud Mar Sci. 2025; 88:104266. https://doi.org/10.1016/j.rsma.2025.104266

Sarmento-Soares LM, Martins-Pinheiro RF. Endemicity analysis of the ichtyofauna of the rio Doce basin, Southeastern Brazil. An Acad Bras Ciênc. 2022; 94(3):1–18. https://doi.org/10.1590/0001-3765202220210646

Schlacher TA, Liddell B, Gaston TF, Schlacher-Hoenlinger M. Fish track wastewater pollution to estuaries. Oecologia. 2005; 144(4):570–84. https://doi.org/10.1007/s00442-005-0041-4

da Silva DAR, Carvalho DR, Ferreira FF, Dergam JA, Moreira MZ, Pompeu PS. Non-native fishes occupy broader isotopic niche than native fishes in an impaired river system. Hydrobiologia. 2024; 852:2149–69. https://doi.org/10.1007/s10750-024-05766-1

Sistema Nacional de Informações em Saneamento Básico (SINISA). Resultados SINISA 2025. Available from: https://www.gov.br/cidades/pt-br/acesso-a-informacao/acoes-e-programas/saneamento/sinisa/resultados-sinisa/resultados-sinisa-2025

Sistema Nacional de Informações sobre Saneamento SNIS). Diagnósticos SNIS 2023. Available from: https://www.gov.br/cidades/pt-br/acesso-a-informacao/acoes-e-programas/saneamento/snis/produtos-do-snis/diagnosticos-snis

Sojobi AO, Zayed T. Impact of sewer overflow on public health: a comprehensive scientometric analysis and systematic review. Environ Res. 2022; 203:111609. https://doi.org/10.1016/j.envres.2021.111609

Strauch SM, Erzinger GS. Anthropogenic pollution of coastal ecosystems in Brazil. Anthropog Pollut Aquat Ecosyst. Cham: Springer International Publishing; 2021. p.181–203. https://doi.org/10.1007/978-3-030-75602-4_9

Terra BF, Hughes RM, Francelino MR, Araújo FG. Assessment of biotic condition of Atlantic Rain Forest streams: a fish-based multimetric approach. Ecol Indic. 2013; 34:136–48. https://doi.org/ 10.1016/j.ecolind.2013.05.001

Vergilio C S, Lacerda D, Souza TS, Oliveira BCV, Fioresi VS, Souza VV et al. Immediate and long-term impacts of one of the worst mining tailing dam failure worldwide (Bento Rodrigues, Minas Gerais, Brazil). Sci Total Environ. 2021; 756:143697. https://doi.org/10.1016/j.scitotenv.2020.143697

Wang Y, Liu D, Richard P, Di B. Selection of effective macroalgal species and tracing nitrogen sources on the different part of Yantai coast, China indicated by macroalgal δ15N values. Sci Total Environ. 2016; 542:306–14. https://doi.org/10.1016/j.scitotenv.2015.10.059

Wear SL, Acuña V, McDonald R, Font C. Sewage pollution, declining ecosystem health, and cross-sector collaboration. Biol Conserv. 2021; 255. https://doi.org/10.1016/j.biocon.2021.109010

Yamamoto FY, Onishi K, Ralha TR, Silva LFO, Deda B, Pessali TYC et al. Earlier biomarkers in fish evidencing stress responses to metal and organic pollution along the Doce River basin. Environ Pollut. 2023a; 329:121720. https://doi.org/10.1016/j.envpol.2023.121720

Yamamoto FY, Pauly GFE, Nascimento LS, Fernandes GM, Santos MP, Figueira RCL et al.Explaining the persistence of hazardous chemicals in the Doce River (Brazil) by multiple sources of contamination and a major environmental disaster. J Hazard Mater Adv. 2023b; 9:100250. https://doi.org/10.1016/j.hazadv.2023.100250

Yunus AJM, Nakagoshi N. Effects of seasonality on streamflow and water quality of the Pinang River in Penang Island, Malaysia. Chinese Geogr Sci. 2004; 14(2):153–61. https://doi.org/10.1007/s11769-004-0025-z

Authors

---

Patrícia Santos Fráguas¹ , Débora Reis de Carvalho^1,2, Frederico Fernandes Ferreira³, Carlos Frankl Sperber³ and Paulo Santos Pompeu¹

^[1]    Laboratório de Ecologia de Peixes, Departamento de Ecologia e Conservação, Instituto de Ciências Naturais, Universidade Federal de Lavras, 37203-202, Lavras, MG, Brazil. (PSF) psfraguas@yahoo.com.br (corresponding author), (PSP) pompeu@ufla.br.

^[2]    Lancaster Environment Centre, Lancaster University, LA1 4YN Lancaster, United Kingdom. (DRC) deboracarvalhobio@gmail.com.

^[3]    Departamento de Biologia, Universidade Federal de Viçosa, 36570-000, Viçosa, MG, Brazil. (FFF) frederico.fernandes@ufv.br, (CFS) sperber@ufv.br.

Authors’ Contribution

---

Patrícia Santos Fráguas: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing-original draft, Writing-review and editing.

Débora Reis de Carvalho: Conceptualization, Formal analysis, Investigation, Methodology, Writing-review and editing.

Frederico Fernandes Ferreira: Investigation, Writing-review and editing.

Carlos Frankl Sperber: Funding acquisition, Project administration, Resources, Writing-review and editing.

Paulo Santos Pompeu: Conceptualization, Formal analysis, Methodology, Project administration, Resources, Supervision, Writing-review and editing.

Ethical Statement​

---

Sampling procedures were authorized by SISBIO – Biodiversity Authorization and Information System, Nº 10327–3, and the Animal Ethics Committee of UFLA (CEUA – UFLA), Nº 034/21.

Competing Interests

---

The author declares no competing interests.

Data availability statement

---

The data supporting the findings of this study are available from the corresponding author, Patrícia Fráguas, upon reasonable request.

AI statement

---

ChatGPT (OpenAI, free version) was used solely for English language editing to improve clarity and grammar. No scientific content, analyses, interpretations, or conclusions were generated by AI.

Funding

---

This study was supported by a contract between FAPEMIG and the RENOVA Foundation under Grant Agreement No 4800028061 and was financed in part by CAPES (Finance code: 32004010017P3). PSP was granted a research fellowship (302328/2022–0) by Conselho Nacional de Desenvolvimento Cientifico e Tecnológico (CNPq).

Supplementary Material

---

Supplementary material SUP

Peer Review

---

Peer Review File

How to cite this article

---

Fráguas PS, Carvalho DR, Ferreira FF, Sperber CF, Pompeu PS. Tracing sewage-derived organic matter in fish trophic pathways in the lower Doce River basin, Brazil. Neotrop Ichthyol. 2026; 24(3):e250144. https://doi.org/10.1590/1982-0224-2025-0144

Copyright​

---

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Distributed under

Creative Commons CC-BY 4.0

© 2025 The Authors.

Diversity and Distributions Published by SBI

Accepted April 27, 2026

Submitted August 9, 2025

Epub September 7, 2026