Introduction

Aquatic macroinvertebrates are one of the most important bioindicators of environmental river quality (Metcalfe 1989). Over time, the use of macroinvertebrates as bioindicators has become widely adopted across Europe due to their ease of collection, identification, and well-known indicator properties (Kolkwitz and Marsson 1909; Hellawell 1986; Rosenberg and Resh 1993). Initially, macroinvertebrate indices based on saprobic systems and indicator traits were used in biological assessment studies. Subsequently, multimetric indices using more than one metric were developed in the United States (Wilhm and Dorris 1968; Barbour et al. 1996, 1999) and in the European Union AQEM project (Hering et al. 2004). The utilization of multimetric indices took a different format with the adoption of "Water Framework Directive" (WFD) (Council of European Communities 2000/60/EC) in 2000. A typological framework was also defined for assessing the ecological quality of water bodies in the future. The scope of WFD was to use the type-specific multimetric index method, which has also been recommended for biological assessment. With this perspective, AQEM and STAR projects have been developed and implemented biological assessment methods in Europe (Furse et al. 2006, AQEM 2002). In this study, the previously developed Ceyhan Basin Multimetric Index (MMI-C) was carried out within the scope of this objective of the WFD.

In developing countries like Turkey, rivers are under intense domestic and industrial pollution pressure. Biological methods play an important role in the integrated management of water resources and have several advantages over physicochemical methods (Rosenberg and Resh 1993). Biological monitoring provides valuable information allowing estimates of deleterious influences on lotic habitats, at low cost and with minimal technical requirements. The use of macroinvertebrates in biological assessment studies of lotic systems, in combination with mathematical indices began at the end of the twentieth century in Turkey (Kazancı et al. 1997; Kazancı and Dügel 2000). Gradually, the number of studies started to increase using different biotic indices (Dügel and Kazancı 2004; Duran 2006; Kazancı et al. 2010, 2013; Kalyoncu and Zeybek 2011; Zeybek et al. 2014; Yorulmaz et al. 2015; Arslan et al. 2016; Zeybek 2017).

The method of developing multimetric indices for rivers was demonstrated by European studies (e.g., Hering et al. 2006) and in the United States of America (e.g., Hughes et al. 2009). In these countries, stream types were delineated and mapped with clear boundaries (Omernik 1995) (Verdonschot and Nijboer 2004). As a result of the utilization of stream types in the multimetric index development process, type-specific indices have been developed and used. Studies including the multimetric index concept have been initiated within the framework of harmonization laws between Turkey and the European Union (Dügel 2016). In these studies, country-specific typologies were determined (Digitizing Project 2022) but could not be used in the development of multimetric indices. The developed and used multimetric indices were mostly basin-specific indices or a single index covering all of Turkey instead of type-specific multimetric indices (Dügel 2016; Akay and Dalkıran 2020; Odabaşı et al. 2022; Koyuncuoğlu et al. 2023; Öztürk et al. 2023).

River basins provide an optimal setting for researching the assessment and monitoring of streams, as well as evaluating the environmental implications of both biotic and abiotic factors (Dawei and Jingsheng 2001). Water management issues, at the scale of whole river basins, have been of significant public concern in Turkey. Therefore, integrated basin management studies in Turkey have gained momentum since 2009. After that, conducted and continued river basin-related studies followed the rules of the WFD within the scope of the harmonization laws between the European Union and Turkey. In this study, we expect that evaluations performed with biological data in the Ceyhan River basin will contribute to holistic basin management. Additionally, the present study will provide data for multimetric index development studies and contribute to a robust biological assessment system for Turkey in the future. The multimetric index values of the sampling sites were calculated on a seasonal basis using four selected metrics in a previous study (Dügel 2016). The objectives of the current study were: (1) to make ecological evaluations of the streams in the Ceyhan River basin; (2) to increase multimetric index applications in Turkey with new data; (3) to provide data for type-specific multimetric index development studies; and (4) to contribute to the knowledge of macroinvertebrate fauna of the Ceyhan River Basin and Turkey.

Material and Methods

Study area

The sampling sites belong to the Ceyhan River Basin (Table 1, Fig. 1). The Ceyhan River, formerly the Pyramos, is one of the largest rivers in Anatolia. It is located between 36°55' and 38°72' north latitudes and 35°45' and 37°81' east longitudes in the Eastern Mediterranean Region of Turkey (Fig. 1). The basin covers 2.73% of Turkey's surface area, and the river flows through a mountainous catchment area in the eastern Taurides, characterized by Palaeozoic, Mesozoic, and tertiary karstic carbonate bedrock. The Ceyhan River has a tributary length of 510 km. It originates in the mountains around the Elbistan Plain (Kahramanmaraş) and joins large tributaries such as the Aksu and Göksun streams. The river flows in a southwestern direction, passing through the Çukurova floodplain and delta, before entering the Mediterranean Sea at the Bay of Iskenderun (CHKYP 2019; Akbulut et al. 2022).

Table 1 Codes and geographical information of sampling sites in the present study
Full size table
Fig. 1
figure 1

Location of sampled sites in the Ceyhan River Basin. Full names and codes of sites are available in Table 1

Full size image

Sampling

Before macroinvertebrate sampling, the coordinates of each site were noted using a global positioning system (GPS), and details of the sampling sites are provided in Table 1. A YSI Professional Plus multi-probe was used to measure dissolved oxygen concentration (DO, mg L−1), water temperature (Tw, °C), electrical conductivity (EC, μS cm−1), pH, salinity (ppt), total dissolved solids (TDS, mg L−1) and oxidation and reduction potential (ORP, mV). For chemical analyses and estimation of biological oxygen demand (BOD5, mg L−1), 500 mL water samples were collected from each site in polyethylene bottles and preserved in a container maintained at a temperature of + 4 °C during the field survey.

Macroinvertebrate sampling was carried out using the multihabitat sampling method with a hand-net featuring a 500 μm mesh size (STAR 2003). Each site was sampled three times: in April (for the spring) and October (for the autumn) of 2021 and August 2022 (for the summer). Collected specimens were placed on a large tray, and sensitive specimens were manually collected and stored in 50 ml falcon tubes containing ethanol (70°). The remaining substrate was placed in 500 ml plastic containers and fixed with 95° ethanol for transportation to Limnology laboratory, Bolu Abant İzzet Baysal University.

Laboratory analyses

Sulfate (SO42−, mg L−1), total nitrogen (TN, mg L−1), total phosphorus (TP, μg L−1), nitrate (NO3−, mg L−1), nitrite (NO2−, mg L−1), fluorine (F, mg L−1) and chloride (CI−, mg L−1), ammonium (NH4+, mg L−1), iron (Fe, mg L−1), calcium (Ca2+, mg L−1), magnesium (Mg2+, mg L−1), potassium (K+, mg L−1) and sodium (Na+, mg L−1) and total organic carbon (TOC, mg L−1) of water samples were measured using the Hach LT 200 Thermoreactor, Hach Cuvette tests and Hach Lange DR 5000 spectrophotometer. For BOD5 (mg L−1), water samples in 420 mL opaque dark bottles were incubated at 20 °C for 5 days, and then BOD5 values were recorded using a Hach BOD Trak 2 device (Hach 2015). The standard procedure provided in APHA (2012) was used to record the total hardness (mmol CaCO3 L−1) values of each water sample.

Samples were gently rinsed to separate macroinvertebrate specimens from sediments, both with the naked eye and under a stereo microscope (Olympus ACH 1X) when necessary. Specimens were sorted into groups (e.g., Ephemeroptera, Odonata, Gastropoda), and each group was then placed into 50 ml falcon tubes and fixed with 70° ethanol. Some specimens were dissected in glycerol solution to prepare permanent slides for identification. Macroinvertebrate specimens were identified using both stereo (Olympus ACH 1X) and light (Olympus BX-51) microscopes to the lowest possible systematic category, following suitable taxonomic keys (Hliley 1976; Franke 1979; Carchini 1983; Brinkhurst and Wetzel 1984; Vepsäläinen and Krajewski 1986; Zimmermann 1987; Sivec et al. 1988; Nilsson 1989; Müller 1990; Savage 1990; Zwick 1991, 2004; Schmedtje and Kohmann 1992; Bauernfeind 1994; Engblom 1996; Jansson 1996; Solem and Gullefors 1996; Norling and Sahlén 1997; Jensen 1997; Butler 1998; Bauernfeind and Humpesch 2001; Eggers and Martens 2001; Malicky 2004; Neu and Tobias 2004; Eiseler 2005, 2015; Vallenduuk and Moller Pillot 2007; Moller Pillot 2009, 2013; Tachet et al. 2010; Waringer and Graf 2011; Bauernfeind and Soldan 2012; Andersen et al. 2013; Cranston and Epler 2013; Dobson 2013; Van Haaren and Soors 2013). Specimens were stored in the Limnology Laboratory of Bolu Abant İzzet Baysal University, Turkey, and are available upon request.

Statistical analyses

The species diversity of each site during different seasons (spring, summer, and autumn) was estimated by calculating Shannon–Wiener diversity (H) (Shannon and Weaver 1949) and Shannon Evenness (E) density indices with the aid of the Species Diversity and Richness Package (SDR) (Seaby and Henderson 2006). Analysis of Similarity (ANOSIM) and Similarity Percentages (SIMPER) analyses (Clarke 1993) were applied using the Community Analysis Package (CAP 4.0) (Henderson and Seaby 2007) to understand the level of possible differences between the seasonal occurrence of species and to evaluate the taxa that contribute to these differences, respectively. Dissociation between seasons was visualized by non-metric multidimensional scaling (NMDS) analysis (CAP 4.0). The relationships between environmental variables and macroinvertebrate species that occurred two or more times were explored by Canonical Correspondence Analysis (CCA) using CANOCO 4.5, and statistical significance between axes of ordination was tested with the Monte Carlo Permutation test (999) (ter Braak 1988).

The water quality of the sites was evaluated by a multimetric biotic index based on macroinvertebrate metrics (Multimetric Index-Ceyhan, MMI-C). To determine the ecological water quality, seasonal ecological quality ratios were calculated over the abundance values of the determined macroinvertebrate groups. Four metrics were used; i) Biological Monitoring Working Party Score System-Spanish version (BMWP-Sp, sensitivity/tolerance metric), ii) Shannon Wiener Diversity index (richness/diversity metric), iii) [%] epirhithral (functional metric) and iv) EPT (Ephemeroptera, Plecoptera and Trichoptera) Taxa (%, composition/abundance metric). To calculate the multimetric index, the 75th percentile of the metric values obtained was used for standardization processes, then the final index values of the sites were determined. Class boundaries between ecological quality ratios were used for the MMI-C index (Dügel 2016). All community metrics were calculated using the ASTERICS 4.0.4 software (AQEM/STAR Ecological River Classification System) (AQEM 2002).

Results and discussion

A total of 166 taxa and 14,839 macroinvertebrate individuals were collected from the 44 sampling sites across the three seasons. Among the reported 18 groups, the highest number of taxa was observed in the Chironomidae and Ephemeroptera. The highest number of individuals was found among amphipods, while the lowest count was recorded for isopods (Table 2). The distribution of macroinvertebrate taxa among the sampled sites is presented in Table 3.

Table 2 Number of taxa and individuals of macroinvertebrate groups encountered in the Ceyhan River Basin streams in three seasons
Full size table
Table 3 Names and codes of macroinvertebrate taxa collected in the sampling sites in different seasons
Full size table

Analysis of macroinvertebrate communities by ANOSIM shows significant differences between species compositions in spring–autumn and spring–summer pairwise comparisons (R = 0.17, p < 0.05). The ordination chart shows the sampling sites in the spring period are positioned differently from the other sampling sites. (Fig. 2).

Fig. 2
figure 2

Graphic representation of ANOSIM results using Non-metric Multidimensional Scaling of the similarity (Bray–Curtis) for the macroinvertebrate composition among the spring, summer, and autumn seasons

Full size image

Results of the SIMPER analysis to assess the degree of dissimilarity within groups (seasonal sampling sites) are shown in Table 4. According to the SIMPER results, within-group similarity rates for each of the spring, summer, and autumn seasons showed approximately 15% similarity. Spring sampling sites differed significantly from autumn sampling sites (ANOSIM: R = 0.23, p < 0.05, average dissimilarity of 88.61%), with Baetis rhodani as the major contributor (20.42%) to those differences. Similarly, spring sampling sites differed significantly from summer sampling sites (ANOSIM: R = 0.24, p < 0.05, average dissimilarity of 88.21%), with B. rhodani as the major contributor (22.42%) to those differences. Although the dissimilarity between autumn and summer was not statistically significant, two species, B. rhodani (19.47%) and G. balcanicus (21.10%), contributed the most to the 84.60% dissimilarity between both seasons. The contribution of species other than B. rhodani and G. balcanicus to the discrepancies between seasons was less than 6% (Table 4).

Table 4 SIMPER and ANOSIM results for macroinvertebrate community composition between autumn (A), spring (Sp) and summer (Su) sampling sites
Full size table

Baetis rhodani and Gammarus balcanicus are taxa that characterize the autumn and summer seasons, but G. balcanicus was completely absent from spring sampling sites. This indicates that this species colonizes in early summer and is not abundant in the cold spring.

A study on gammarids shows that their density increased in early summer (Pöckl et al. 2002). In another study in Yeşilırmak (Turkey) shows that gammarids are more common after the spring months (Duran 2007). Baetis rhodani, one of the common Ephemeroptera, was the species with the highest contribution percentages in spring (79.81%) to these within-group similarities, while its lowest contribution was observed in autumn with a rate of 46.96% (Table 4). The life cycle of B. rhodani changes between different environments (Bauernfeind and Humpesch 2001). In some studies conducted in karst areas, this species had a polyvoltine life cycle (Buffagni et al. 2003, Erba et al. 2003), although some researchers stated that this species has a univoltine life cycle (Bottová and Derka 2013). These results shows that B. rhodani has a flexible life cycle. Moreover, since this species has a wide ecological tolerance, it can survive in unpolluted and moderate organically polluted waters (Hellawell 1986).

Taxa are ranked according to their average contribution to similarity/dissimilarity values within (top of the table) or between (down of the table). Average abundances, ratio (similarity or dissimilarity) and percentage of cumulative similarity are also included. Only the three taxa that contributed the most to the distribution are shown.

In spring, the highest Shannon diversities (H) were calculated at sites 27 (2.46) and 17 (2.13), while the lowest were at sites 21 (0.23) and 33 (0.42). A similar situation was observed for sites 27, 17, 21 and 33 for evenness (E). In the summer season, the sites 39 (2.24) and 31 (2.20) had the highest diversities (H') while the lowest diversity values were found at sites 8 (0.10) and 16 (0.15), and evenness values showed similar patterns for these sites (Table 5).

Table 5 The number of individuals (N), taxa (S) and Shannon–Wiener diversity (H') and Evenness (E) values for macroinvertebrates collected in 44 sites according to season
Full size table

In autumn, a slightly higher diversity (H') value than in summer, but lower than in spring. This indicates the elimination of some species from the systems due to both the drying up of the waters and changing water parameters due to increasing air temperatures. The highest diversities (H') in autumn were calculated at sites 17 (2.28) and 13 (2.11). While the most homogenous distributions in autumn were observed at sites 17 (E = 0.50) and 13 (E = 0.46), non-homogenous distributions were observed at sites 8 (0.04) and 16 (0.11) (Table 5).

The difference in diversity among sites with the same taxon numbers depends on the different distribution patterns of individuals of the taxa. Although the same number of taxa were found (S = 16) in four sites (17, 18, 27, and 40) during the summer period, the species diversity started to decrease when the number of individuals increased. The "evenness" (E) values of the same sites decreased as the number of individuals increased. "Evenness" describes the variability in species abundances (Magurran 2004). A community in which all species have approximately equal numbers of individuals (or similar biomasses) would be rated as extremely even. Conversely, a large disparity in the relative abundances of species would result in the descriptor "uneven” (Magurran 2004). Although the lowest number of taxa were found in autumn, the high (E) value indicates a more homogeneous distribution of the species compared to summer. The Shannon diversity index calculates diversity values independently from the number of individuals, while the evenness index gives important results by testing whether the species are homogeneously distributed. As expected, the representation of species in a population with a relatively equal number of individuals is an indicator of ecosystem health. Considering taxon richness, it is seen that the most taxa were found in spring (114), and the fewest were found in autumn (93). However, when the diversity index values calculated according to all sites were examined, the highest value was observed in spring (3.01), and the lowest one in summer (2.45) (Table 5).

Considering the three-season sampling sites, in two sites in spring, ten sites in summer and six sites in autumn no macroinvertebrate specimens were found, which could be related to drought. Similar findings also show that diversity values in summer were lower than in the other seasons. When examining the total number of individuals, the highest number (6685) was found in summer and the lowest one (3340) was found in spring. During summer, the ratio of the number of individuals of gammarids to other taxa was 35%. Similarly, the ratio in autumn was 32%.

The total number of individuals in spring was lower than in the other periods as gammarids were not present in the spring sampling period.

The ecological quality ratio (EQR) of the sites was calculated using the Ceyhan Basin Multimetric Index (MMI-C) (Table 6). The highest ecological quality ratios were found at sites 17 (1.00) and 19 (0.99) in spring when the lowest EQR values were calculated for sites 1 (0.19) and 44 (0.26). In summer, the highest EQR ratios were calculated for sites 17 (0.93) and 30 (0.93), while the smallest values were calculated for sites 8 (0.04) and 16 (0.12) sites. Sites 17 (1.00) and 30 (0.98) had the highest EQS values in autumn, sites 8 (0.11) and 16 (0.19) displayed the lowest values. Considering the average of the three-period EQR values, the highest values were calculated for sites 17 (0.98) and 18 (0.94) and the lowest values for sites 8 (0.14 and 16 (0.19). As a result, site 17 had the best ecological status and site 8 the worst ecological status (Table 6).

Table 6 MMI-C (Ceyhan Basin Multimetric index) – EQR (Ecological Quality Ratio) and ES (Ecological Status) values of sampling sites in the present study. EQR values are between 1 and 0
Full size table

While the number of sites with “good” and “high” ecological status (ES) values was 14 in spring, 18 sites were classified as "good" in summer. There were two sites with “high” and 14 sites with “good” ES in autumn. In total, there were 19 “good”, 16 “moderate”, 3 “poor” and 4 “bad” sites. Thus, it seems that a significant part of the sampled sites were far from achieving good ecological status. The altitude (a.s.l) of almost all sites with “good” ecological quality status is approximately over 1000 m, indicating that the sites located at high-altitude areas in the Ceyhan Basin are in better condition. Similarly, sites with "poor" or "moderate" ecological status appear to be located downstream of rivers and in urbanized areas (Tables 1 and 6). Ceyhan Multimetric Index was used for the first time in this study and provided important data for subsequent studies. Therefore, these results were not compared with previous findings.

According to the results of Canonical Correspondence Analysis (CCA), the eigenvalues of the first two axes were as follows: 0.45 and 0.20, respectively. The first two axes of CCA elucidated a moderately low relationship (51.4%) between species distribution and environmental variables. In the CANOCO 4.5 program, the importance of the variables was calculated according to the best variance explanation percentages by using the forward selection method (Leps and Smilauer 2003). The three most important variables were altitude, temperature and pH. With the "collinearity test" application in the ECOM program, the variables that were correlated with each other were eliminated, and distribution analysis was performed using eight variables suitable for these characteristics. According to the CCA diagram, dissolved oxygen is positioned separately from the other variables on the opposite side of temperature. This also indicates the negative relationships between dissolved oxygen and temperature (Fig. 3a). Coenagrion pulchellum (Odonata), Tanytarsus gergarius (Chironomidae), and Polypedilum convictum (Chironomidae) showed close associations with BOD, which is inversely correlated with altitude. It is known that these species can live in low-oxygen environments (Hynes 1970; Hellawell 1986). The species Epeorus caucasicus (Ephemeroptera), Prosimulium tomosvaryi (Simuliidae), Blepharicera fasciata (Diptera), and Rhithrogena semicolorata (Ephemeroptera) were located in the same direction as dissolved oxygen but in the opposite direction of temperature and other nutrient salts (total nitrogen and total phosphorus). These results support the indicator properties of these species in well-oxygenated, clean waters (Hynes 1970) (Fig. 3a).

Fig. 3
figure 3

Canonical Correspondence Analysis diagrams for species (a) and for sampling sites (b) showing relationships between macroinvertebrates, sites, and physicochemical variables. For species codes see Table 3. ●: Species; TN: total nitrogen; TP: total phosphorus EC: electrical conductivity; DO: dissolved oxygen; Temp: water temperature. BOD: biological oxygen demand

Full size image

The CCA diagram shows that sites with low ecological quality ratios (e.g., S01, S44, S43, S35, S39) are directly related to BOD values. Similarly, sites with high ecological quality ratios (e.g., S17, S18, S31, S30) appear to be directly related to their dissolved oxygen values (Fig. 3b). It is known that sensitive macroinvertebrates disappear at low dissolved oxygen values. In this case, indicator taxa for good ecological status disappeared, and EQR values decreased (Hellawell 1986).

Conclusion

Results from this study carried out in the Ceyhan River Basin reveal that sites with bad ecological status are located near cities. As the altitude increases, the number of sites in good status starts to increase. Altitude is a variable that affects the distribution of species in this area. Generally, MMI-C values were high in areas with high diversity (H') values. The biotic and abiotic data obtained in this study will provide essential information for type-specific multimetric index development studies inside and outside of Turkey. Additionally, the macroinvertebrate fauna of a basin was postulated for the first time in Turkey. It is necessary to carry out studies in developing countries such as Turkey to detect aquatic macroinvertebrates throughout the country and reveal the bioindicator characteristics of these organisms for use in ecological assessment studies. With such studies, it will be possible to conduct ecological assessment studies much faster, more effectively, and at a lower cost than by using abiotic data. Therefore, the number of studies similar to the current study should be increased in the future.