Introduction

Rivers and streams constitute integral components of terrestrial landscapes. Covering approximately 0.6% of the Earth’s land surface, equivalent to an area of 773,000 ± 79,000 km2 (Lehner 2024), they serve as efficient conduits for natural water circulation on Earth (Sánchez-Montoya et al. 2023). They also exhibit notable dynamism across various aspects, including hydrology, geomorphic, and chemical processes (Jørgensen 2016). Hydrologically, for example, these environments experience diverse conditions, ranging from inundation during floods to the loss of water flux connectivity during droughts (Omosulu et al. 2023). Despite their high dynamicity, rivers and streams have historically assisted humans in procuring water for agricultural, drinking, and navigational purposes, providing a significant benefit to inland civilizations (Wang and He 2022).

Rivers and streams depend on their individual and collective extents to function naturally and effectively (Ferreira et al. 2023). Categorized as vulnerable ecosystems, any disturbance upstream may impact every component of the downstream watershed (Büttner et al. 2020). Similar to numerous ecosystems, human activities impose increased stress on these systems, frequently exceeding their ecological carrying capacity. For example, 46% of rivers and streams in the USA are reported to be in poor biological condition and incapable of supporting healthy aquatic communities (Mihelcic and Rains 2020). Globally, many of these water bodies receive effluents from sewage systems, leading to elevated pollutant levels (Pascual-Benito et al. 2020; Singh et al. 2020). In numerous instances, the mere existence of humans is associated with an elevation in microplastic pollution within these ecosystems (Mora-Teddy and Matthaei 2020). Therefore, the heightened sensitivity and impact of human activities on these ecosystems emphasize the imperative requirement for proficient management to guarantee their optimal functionality.

The health and functioning of rivers and streams are influenced by a multitude of factors and indices, intricately interwoven within a complex network of relationships. A significant proportion of these attributes can be linked to hydrological conditions, such as flow patterns (Shanafield et al. 2021), geomorphic features, including channel morphology and sediment transport (Paixão et al. 2023), as well as nutrient levels and pollutant concentrations (Asgarian and Soffianian 2023). Moreover, the biological constituents of these ecosystems serve as a key indicator of their overall health. This component spans a broad spectrum of species ranging from microbial communities to aquatic flora and fauna (Burdon et al. 2020; Hilderbrand et al. 2020). It constitutes an essential aspect of river water quality assessment, involving the examination of the type, abundance, and diversity of various organisms inhabiting or relying on these aquatic environments (Sadeghi et al. 2021). In this field, for example, Akamagwuna et al. (2021) observed a taxonomy-based response pattern of macroinvertebrates to a pollution gradient in the Tsitsa River, South Africa. Their findings indicated that the abundance of sensitive taxa decreases with the increasing deterioration of river water quality. In a study conducted on the Hau River in Vietnam, Mutea et al. (2021) illustrated the impact of Pangasius fish farms and agricultural runoffs on the deterioration of water quality downstream. This deterioration was manifested by elevated levels of biological oxygen demand and total coliforms. Exploring subtropical river drainages, Durán-Rodríguez et al. (2022) demonstrated that a two-decade decline in dissolved oxygen (DO) and water pH resulted in a reduction in the overall abundance and structure of aquatic insects.

Rivers and streams in Iran are facing significant pollution issues, attributed to the country’s arid climate and the concentration of human activities along these water bodies (Vayghan et al. 2022). Despite the recognition of these challenges and recent government initiatives toward environmental monitoring and conservation (Asgarian et al. 2018), there remains a critical need to address specific gaps in understanding the ecological impacts of agricultural practices on river ecosystems. In this context, our research focuses on a complex interconnected stream network in South-west Iran, traversing extensive agricultural lands and urban centers before reaching the Maharlu Wetland. Unlike previous studies that have broadly assessed water quality, our research uniquely identifies 12 specific locations along the Maharlu stream network (MSN), covering diverse agricultural activities from upstream to downstream. Through a comprehensive analysis of chemical properties and biological indicators, our study aims to discern the primary factors influencing water quality and the associated ecological responses. The novelty of our approach lies in the detailed spatial and ecological characterization of agricultural impacts on river ecosystems in this region. By filling these specific gaps in knowledge, our findings will contribute significantly to the ongoing conservation efforts and underscore the urgency of implementing targeted measures to mitigate the adverse effects on rivers and streams in Iran.

Material and methods

Study stations

This study was conducted on the MSN system, situated in the south-west region of Iran (Fig. 1). The MSN functions as the principal inflow source to the Maharlu Wetland (257 km2), receiving water from the thawing of snow and seasonal rains originating in the elevated terrains of the Central Zagros. Prior to reaching the wetland, the MSN traverses an extensive plain terrain extensively employed for large-scale irrigated agricultural practices. The cumulative agricultural area within the study network is estimated at approximately 580 km2, relying on the stream network to meet its water requirements. Relying on the stream network to fulfill its water requisites. Potable water for the populace residing along this network is predominantly supplied through groundwater wells, encompassing Shiraz, the fourth most populous Iranian city with a population of 1.7 million (Shahsavani et al. 2023). The climate in this region is characterized as a moderate semi-arid climate, with annual average rainfall, humidity, and temperature recorded at 335 mm, 64%, and 18 °C, respectively (Abbasi and Turner 2021). To encompass a range of environmental conditions, sampling stations were strategically positioned at the initial segments of the stream network, characterized by minimal human influence, at the terminal sections leading to the wetland, and additional stations were distributed along the intermediate sections.

Fig. 1
figure 1

Distribution of sampling stations along the north-west stream network reaching the Maharlu Wetland in the south-west of Iran

Full size image

Measurement of water quality parameters

During sample collection, factors of DO, pH, EC (electrical conductivity), TDS (total dissolved solids), and turbidity (in Nephelometric Turbidity Units-NTU) were measured using the aquaread multi-meter device (Italy). In the laboratory, pH was determined using the HACH device. Biological oxygen demand (BOD) was measured using a BOD meter (HANA, USA). Chemical oxygen demand (COD) was measured using the reflux distillation method, phosphate (\({\text{PO}}_{4}^{3-}\)) under acidic conditions was assessed through a reaction with ammonium heptamolybdate, and nitrate (\({\text{NO}}_{3}^{-}\)) was determined through reduction with cadmium followed by a reaction with sulfanilic acid (Eaton et al. 2005). Total coliform count was conducted using the most probable number (MPN) method with EC medium in the general microbiology laboratory. Ammonium (\({\text{NH}}_{4}^{+}\)) was measured using spectrophotometry and the UV 2150 spectrophotometer (Unico, USA).

Species identification and diversity analysis

Statistical analysis

Principal component analysis (PCA) serves as a widely utilized statistical tool in the assessment of water quality, revealing intricate interconnections among a diverse array of water parameters (Teixeira de Souza et al. 2021). This analytical approach transforms water parameters into a new set of uncorrelated variables, effectively reducing data dimensionality while preserving crucial information. In this study, our dataset comprising m water parameters and n observations underwent standardization by subtracting each variable’s mean and dividing it by the standard deviation. Subsequently, a correlation matrix was computed to gauge the strength and direction of linear relationships (Güner et al. 2021). The eigenvalues, representing the variance captured by each principal component (PC), were calculated, and results were arranged in descending order, selecting k components with eigenvalues exceeding 1. Additionally, the proportion of total variance explained by each chosen component was employed to assess the information retained in the reduced-dimensional space (Anastasova et al. 2020). Biplots were employed to depict the relationships between variables and observations, providing a visual representation of orientations.

Canonical correlation analysis (CCA) was also employed to investigate and quantify relationships between the primary PCs derived from the PCA analysis and the abundance of species within each taxon. CCA focuses on identifying canonical variables, including correlation coefficients and loadings, which represent linear combinations of the original variables in each set. Finally, it performs a canonical variate analysis to scrutinize the patterns of relationships between the two sets of variables (Jung et al. 2020). Consistent with PCA, a biplot was utilized to showcase the orientation of both sets of variables in the canonical space.

Results

The analysis of water quality parameters (Table 1) provided a comprehensive insight into the diverse composition of the sampled river water. TH (mg CaCO3/l) ranged from 337.00 to 399.33, with a mean of 363.36 ± 18.51. NTU exhibited significantly high variability, with values spanning from 2.00 to 1568.00 (mean of 487.75 ± 582.94). Considerable variability was also observed in MPN, fluctuating between 21.33 and 6511.00, with a mean of 2335.97 ± 2842.50 per 100 mL water samples. \({\text{NH}}_{4}^{+}\) and \({\text{NO}}_{3}^{-}\) concentrations (mg/L) were notably present at the stations, averaging 8.31 ± 10.78 and 67.77 ± 79.72, respectively. Results also showed variable chemical and biological oxygen demands, indicative of a complex mixture of organic and inorganic substances and ecological stressors in specific sampling locations. Both low and high DO were observed in the samples, ranging from 0.20 to 9.26 mg/L. On average, the water samples showed to be slightly acidic, related to both acidic (pH = 5.21) and alkaline (pH = 8.15) waters. The results also suggest a diverse and variable composition of dissolved solids and ions in the sampled river water, with mean EC and TDS values of 2302.00 ± 1946.91 (µS/cm) and 927.61 ± 889.29 (mg/L), respectively.

Table 1 Range and mean ± standard deviation of the water parameters measured at 12 river stations reaching the Maharlu Wetland
Full size table

The results of the PCA analysis yielded three principal components (PC) with eigenvalues greater than 1 (Fig. 2). PC 1 had a much larger eigenvalue (8.10) compared to PC 2 (1.97) and PC 3 (1.48), suggesting that it captures a substantial amount of the overall variance in the dataset (Fig. 2). PC 1 explained the majority of the variance (57.90%), followed by PC 2 and PC 3 (collectively equaling 24.71%). PC 1 appeared to capture variations related to water quality indicators associated with organic pollution (BOD, COD) and \({\text{NO}}_{3}^{-}\) on one side, and DO, TMP, and pH on the other side. It suggests that high values of PC 1 indicate cleaner water due to being associated with lower concentrations of TH, MPN, \({\text{NO}}_{3}^{-}\), COD, BOD5, TDS, EC, and pH, while high DO levels. PC 2, however, represented a contrast between measures of TDS, EC, and TH. High values in PC 2 may indicate a high mineral content in the water. PC 3 was primarily influenced by TH, TMP, and pH, which might represent the overall water chemistry and environmental conditions of the river section.

Fig. 2
figure 2

Eigenvalues, percentage of variance, and loadings of water quality variables on the first three principal components in the PCA test

Full size image

The Chironomidae family, belonging to the order Diptera and class Insecta, emerged as the most prevalent in the collected samples, exhibiting a frequency of 266 species (Fig. 3). This constituted nearly 47% of the total abundance observed in the samples. Following closely were the Baetidae and Simuliidae families, both originating from the order Ephemeroptera and Diptera, respectively, within the class Insecta. These families ranked as the second and third most abundant, with abundances of 266 and 150 individuals, respectively. Subsequent to these, Dytiscidae (Order: Coleoptera), Empididae (Order: Diptera), and Tipulidae (Order: Diptera) represented the fourth, fifth, and sixth most abundant species in the water samples. Cumulatively, these taxa accounted for over 98% of the total species abundance identified in the samples.

Fig. 3
figure 3

Abundance and cumulative percentage of different species families belonging to the five classes Bivalvia, Crustacea, Gastropoda, Hirudinae, Insecta, and Oligochaeta

Full size image

The preeminent loadings in CC1 among the taxonomic families were ascribed to Odonata, Trichoptera, and Ephemeroptera. These families concurrently exhibited negative values in CC2, exclusively situating themselves in the upper-left quadrant of the ordination diagram (Fig. 4). Variables employed in constructing PC1 similarly manifested analogous loadings, with positive values on CC1 and negative values on CC2. This pattern suggests a plausible association between the Odonata, Trichoptera, and Ephemeroptera families and key water quality parameters, including DO, TMP, and pH. These associations were notably linked to upstream sampling stations S1 and S2, extending to a lesser extent to S3 to S5. Conversely, species within the taxonomic families Lamellibranchiata, Arhynchobdellida, Pulmonata, and Diptera were positioned in the lower-right quadrant of the diagram. This spatial arrangement signified negative loadings in CC1 and positive loadings in CC2, corresponding to PC2 and sampling stations S9, S10, and S12. Additionally, discernible trends were observed in Hemiptera and Coleoptera, in conjunction with the variables utilized in constructing PC3 and stations S8 and S6 (Fig. 4).

Fig. 4
figure 4

Ordination diagram of canonical correspondence analysis depicting the first three principal components in the PCA test of water quality parameters and the abundance of species categorized by their families

Full size image

The diversity indices provided valuable insights into the ecological characteristics of the sampled stations (Fig. 5). Examining the Simpson index across the sampling sites revealed a range from 0.00 to 0.75. Remarkably, S5 stands out with a Simpson index of 0.00, indicating the dominance of a single species in this particular station. Conversely, S3 exhibited a high Simpson index of 0.75 which signifies a more even distribution of species within the community. The Shannon indexed variability from 0.00 to 1.69. Stations S3 and S6 showed the highest values (1.6 <), reflecting greater species diversity and a more balanced distribution of individuals among taxa. On the other hand, S5, with a Shannon index of 0.00, suggests a less diverse community with a dominant species. Evenness values ranged from 0.14 to 1.00, representing a perfectly even distribution in S5 and a highly skewed distribution with certain species dominating the community in S4. The highest Margalef index values (falling between 1.00 and 2.34) were seen in S4 and S6, suggesting greater species richness in proportion to the number of individuals present at the study stations.

Fig. 5
figure 5

Values of the Simpson, Shannon, Evenness, and Margalef indices measured for the species identified and counted at the sampling stations

Full size image

Discussion

Drivers of water contamination

The study stations were distributed across varied landscapes to investigate how human activities influence both the water quality and the composition of aquatic fauna in the MST. Consequently, varying levels of water characteristics such as NTU and MPN were acquired. Downstream observations revealed a significant variability and heightened concentration of \({\text{NO}}_{3}^{-}\) (67.77 ± 79.72 mg/L), indicating severe nutrient pollution. This concentration exceeds both the maximum contaminant level for nitrate in drinking water set by the U.S environmental protection agency (EPA) at 10 mg/L (USEPA 1976) and the World Health Organization (WHO) limit of 50 mg/l (Isiuku and Enyoh 2020). Several factors may contribute to this nutrient pollution, including agricultural runoff, limited sewage discharge, and potential atmospheric deposition. The primary contributor to this nutrient pollution appears to be agricultural runoff, evidenced by nitrate levels ranging from 5 mg/L in pristine upstream highlands to over 196 mg/L at stations within downstream extensive agricultural lands. This finding corroborates previous studies that have identified agricultural activities as a major source of nutrient pollution in diverse regions globally. For instance, Del Rossi et al. (2023) demonstrated that nutrient pollution from agricultural sources, primarily arising from the fertilization of row crops and livestock manure, significantly impacts ecological health in the USA through water pollution. Similarly, Strokal et al. (2016) reached similar conclusions, implicating China’s agricultural industry transition as a leading cause of river water pollution in their research.

Furthermore, organic pollution of the MSN, as evidenced by elevated mean COD and BOD concentrations, can be attributed to agricultural runoff carrying substantial amounts of fertilizers, pesticides, and livestock waste. Microbial contamination of the MSN is further exemplified by high MPN concentrations at specific stations, exceeding 6,000 MPN/100 mL. These observations are consistent with findings from other regions heavily influenced by agriculture, such as India (Nhantumbo et al. 2023), Africa (Gemmell and Schmidt 2013), and North America (Arce‐Navarro et al., 2021), indicating that agricultural activities significantly contribute to river contamination. This highlights the pressing need for conservation measures to mitigate the detrimental effects of intensive agriculture on the MSN’s species and integrity.

Interplay between water quality and aquatic fauna

Our study revealed a rich and abundant aquatic insect community within the MSN network, with the midges and non-biting mosquitoes (Chironomidae family) as the dominant taxon. This discovery aligns with the observations made by Gao et al. (2023), who reported the prevalence of Chironomidae in diverse river types with varying pollution levels. Additionally, Wang (2003) documented the ability of this taxon to thrive in heavily polluted water bodies, resembling the elevated nutrient (\({\text{NO}}_{3}^{-}\)), COD, BOD, and MPN concentrations observed at downstream stations in our study. The increased agricultural activity downstream likely induces substantial changes in water parameters, potentially disrupting the ecological integrity of the network. This shift appears to favor more tolerant species, leading to their dominance and the exclusion of sensitive taxa. Sueyoshi et al. (2017) referred to this phenomenon as environmental homogenization in stream ecosystems, where the resistance of taxon richness decreases with increasing agricultural development, and the abundance of all insects decreases synergistically with increasing agricultural development. Our findings corroborate this trend, with upstream stations (S3, S4) exhibiting higher diversity, as measured by Simpson and Margalef indices, respectively. Downstream stations (S7, S10), on the other hand, display reduced diversity, suggesting a transition toward the dominance of a few tolerant species.

The integrated PCA–CCA analysis demonstrated downstream shifts in aquatic insect community composition and identified key environmental parameters influencing these changes at each station. The ecological interpretation of PCs aligned well with the characteristics of the associated species. For instance, PC1 appeared to reflect a balance between physical and nutrient parameters, with higher values indicating improved water quality, higher DO levels, lower organic pollution, and lower nutrient concentrations. Significantly, upstream stations S1–S4 exhibited higher PC1 values and greater species diversity, including Ephemeroptera, Trichoptera, and Odonata, which typically thrive in well-oxygenated environments. The decline in Ephemeroptera abundance observed in this study aligns with the findings of Akamagwuna et al. (2021) who identified these species as sensitive to changes in DO levels. Durán-Rodríguez et al. (2022) corroborate this association by demonstrating that decreasing DO concentrations over time significantly impacted the community structure of aquatic insects, particularly Ephemeroptera.

PC2 exhibited the highest loadings for EC and TDS, indicating the presence of dissolved salts in the water. These salts likely originated from soil mineral leaching and, potentially, agricultural runoff, as stations with stronger associations with PC2 (e.g., S9–S12) are located in agricultural regions and streams directly connected to the Maharlu Wetland. There is also substantial corroborating evidence that runoff from agricultural areas can elevate the levels of EC and TDS in river systems (Razali et al. 2020). The abundance of Lamellibranchiata, Arhynchobdellida, and Pulmonata species associated with PC2 in downstream stations aligns with their shared ecological roles as detritivores and scavengers. It is proposed that high TDS and EC conditions aid in the dissolution and dispersion of OM, making it more accessible to these taxa (Aduojo et al. 2020). Additionally, these species exhibit a higher tolerance for low oxygen levels, which is a characteristic frequently observed in streams with elevated TDS and EC concentrations. In addition to OM, these environmental conditions prove favorable for certain species, such as Arhynchobdellida. Specifically, they play a crucial role in maintaining the osmotic balance essential for the survival of these organisms. Different associations found among PCs, species, and stations highlight the role of human activities in changing the balance of the MSN. It seems that the disparity in the geographic locations of stations with differing levels of agricultural activities had a notable impact on the dominance and abundance of species, leading to the emergence of two comparable aquatic conditions primarily distinguished by agricultural practices in the vicinity of the Maharlu Wetland.

Limitations and future research directions

The current study focused on a limited number of sampling stations, potentially overlooking subtle regional variations in water quality and environmental conditions. Expanding the sampling network to encompass a broader range of locations within the MSN would provide a more comprehensive understanding of the basin’s heterogeneity and enhance the generalizability of the findings. Additionally, the separation of agricultural operations and municipal centers would allow for a more detailed exploration of the distinct impacts of each sector on the MSN’s water quality. Seasonal analyses of water quality parameters would provide insights into the dynamic nature of the MSN’s water resources. By tracking changes in nutrient levels, salinity, and other parameters throughout the year, it could be possible to identify seasonal patterns and their potential links to agricultural activities and climate variability. Furthermore, investigating the mechanisms governing sediment–water interactions within the MSN would provide a deeper understanding of how pollutants and nutrients are transferred between these two compartments.

Furthermore, it should be noted that the effect of urban runoff was not evaluated as part of this study, which represents an important area for future research in understanding the broader impacts on water quality within the MSN.

Conclusion

Our study highlights the significant impact of expanding agricultural activities on the water quality and aquatic fauna of the MSN network. We observed escalating nutrient and organic pollution downstream, attributed primarily to agricultural runoff carrying fertilizers, pesticides, and livestock waste. The dominance of the Chironomidae family in the aquatic insect community at downstream stations underscores the ecological shifts induced by these environmental changes. This transition toward tolerant species dominance observed through reduced diversity downstream reflects the phenomenon of environmental homogenization in response to increasing agricultural development. Our integrated analysis using PCA–CCA elucidated key environmental parameters influencing community composition, emphasizing the intricate interplay between water quality and species distribution. The associations identified between specific species and environmental factors underscore the ecological implications of human-induced alterations in the MSN. Conservation measures are urgently needed to mitigate the adverse effects of intensive agriculture on aquatic ecosystems and preserve the biodiversity and ecological integrity of the MSN.