Eco-assessment of streams of Konya closed river basin (Turkey) using various ecoregional diatom indices

Ecological assessment of freshwater ecosystems based on diatom metrics is an important issue for attaining environmental sustainability. The present study aimed to evaluate differences in the diatom–stressor relationship in relatively least disturbed streams in the Konya closed river basin using multivariate analyses and to bio-assess streams by the application of different ecoregional diatom indices. Cocconeis euglypta, Cymbella excisa, Cocconeis placentula, and Achnanthidium minutissimum are the most contributing species to the dissimilarity of sampling stations between rainy (spring) and dry (summer and fall) seasons and also between altitude (A2 800- < 1600 m and A3 ≥ 1600 m) groups. The first two axes of canonical correspondence analysis revealed a significant (82.8%) relationship between diatom species and stressors. Diatom species displayed distinct responses to environmental variables (electrical conductivity, Ni, Cu, B, and altitude) playing important roles on the distribution of species. Diatom indices indicate different ecological statuses of stations, from bad to high. European diatom indices except Duero Diatom Index (DDI) and Trophic Diatom Index (TDI) showed good responses to the eco-assessment of streams and indicated high ecological status for the least disturbed sampling stations symbolized as S16, S20, S24, S25, S27-29, S37, and S39. These results were also supported by abiotic evaluation. Although TIT was more competitive in the bio-assessment of streams among diatom indices, it is necessary to increase its species list by determining their trophic weights in future studies. Therefore, the use of ecoregion-specific diatom indices is suggested along with increasing the number of used species to correctly interpret the water quality.


Introduction
Freshwater resources are mostly impacted by human activities and climate change, but there is not enough information about the multi-pressure environmental impacts on aquatic communities. Therefore, ecological monitoring and bio-assessment are needed to see the effects of the abovementioned factors on the status of surface waters (Charles et al. 2021;Çelekli et al. 2021a). For this aim, the ecological assessment should be integrated with the biological quality tools such as phytobenthos (particularly diatoms), benthic macroinvertebrates, fishes, macrophytes, and phytoplankton into routine freshwater biomonitoring with respect to the European Union Water Framework Directive (WDF) (Directive 2000).
Ecological assessments of freshwater ecosystems based on bioindicator metrics are one of the crucial issues to attain environmental sustainability worldwide. Among them, diatoms are robust bioindicators of spatiotemporal changes in Responsible Editor: Philippe Garrigues environmental conditions at any time of the year (e.g., Rott et al. 1999;Lobo et al. 2015;Delgado and Pardo 2015) and also, they are important primary producers in the energy and nutrient cycling of the biosphere (Smol and Stoermer 2010). Diatoms are valuable indicators of environmental conditions in lotic ecosystems because they quickly respond to spatial and temporal changes in environmental conditions in ecosystems, including nutrient concentrations, temperature, electrical conductivity, and pH (Charles et al. 2021;Çelekli et al. 2021b). Besides, ecology and taxonomy of diatom communities are reasonably well understood, and each diatom species provides important information about the environment where they live (Rott et al. 1999;Kelly et al. 2008;. Diatom species have strong relationships with nutrients especially total phosphate (TP), soluble reactive phosphorus, electrical conductivity, salinity, etc., which are quantifiable in the different trophic gradients from reference sites to highly disturbed areas to develop diatom indices (Charles et al. 2021;Çelekli et al. 2021a). Trophic Index-TI in Austria (Rott et al. 1999), Specific Pollution-Sensitive Index-IPS in France (Cemagref 1982), Trophic Diatom Index-TDI in the United Kingdom (Kelly et al. 2008), Eutrophication and/or Pollution Index Diatom-EPI-D in Italy (Dell'Uomo 2004), Duero Diatom Index-DDI in Spain (Álvarez-Blanco et al. 2013), and Trophic Index Turkey-TIT in Anatolia (Çelekli et al. 2019) in Europe; Trophic Water Quality Index-TWQI in Brazil (Lobo et al. 2015), Diatom Ecological Quality Index-DEQI in Mexico (Salinas-Camarillo et al. 2021) in South America; and Diatom Species Index Australian Rivers-DSIAR (Chessman et al. 2007) and Richmond River Diatom Index-RRDI (Oeding and Taffs 2017) in Australia are some of the indices developed in different ecoregions to evaluate the ecological status of water resources.
The least disturbed environmental conditions (i.e., minimal or no human impacts) of riverine ecosystems are necessary to determine the classification systems and the ecological quality class boundaries based on the bio-assessment approach of WFD (Directive 2000) to achieve a "good ecological status." Diatom metrics give strong reflectivity to different environments (physical, chemical, and hydromorphological) from the least disturbed to the worst conditions (Karr and Chu 1998;Feio et al. 2014).
Ecoregional factors such as climate, geology, land use, and anthropogenic activities have significant effects on the ecological preferences of diatom species (Lobo et al. 2015;Çelekli and Kapı 2019;Salinas-Camarillo et al. 2021). Different stream typologies (e.g., altitude, catchment area, geology, flow regime, precipitation, and hydro-morphological dynamics) can also affect the diatom composition and their abundance Charles et al. 2021;Salinas-Camarillo et al. 2021). In addition, each diatom species has a fundamental ecological niche with its optima and tolerance for stressors in complex environmental conditions (Hutchinson 1957). The trophic weight values and indicator scores of diatom species may show changes in the realized niches between different ecoregions (Rott et al. 1999;Çelekli and Kapı 2019;Pajunen et al. 2020). This highlights the importance of ecoregion-based indices.
The Konya closed river basin includes streams with different hydro-morphological properties under the pressures of natural and human activities. Until now, no study has been conducted to investigate the bio-assessment of stream conditions in the Konya closed river basin. Therefore, bio-assessments of streams in the Konya closed river basin by a few different ecoregional diatom metrics and by the application of multivariate analyses were done for the first time according to the implementations of the WFD. Accordingly, aims of the present study were to evaluate the differences in the diatom-environment relationship in relatively least disturbed streams using multivariate analyses and to dia-assess the sampling stations with different typologies using various ecoregional diatom indices.

Study area
The Konya closed river basin in central Anatolia covers a total area of 49,786 km 2 with plains and plateaus, surrounded by the Taurus, Geyik, Sultan, and Melendiz mountains (Fig. 1). It is mainly constituted by 9 provinces (Konya, Niğde, Aksaray, Karaman, Ankara, Isparta, Nevşehir, Mersin, and the north part of Antalya). This closed basin consists of various streams in the upper 1100 m asl. with different typologies (Table 1).
The Konya closed river basin is dominated by a central Anatolia continental climate as a semi-arid continental climate, which is characterized by seasonal differences (hot, dry summers and cold, snowy winters). Most of the basin has relatively low annual precipitation (340-380 mm). The Taurus Mountains (up to 3404 m asl.) are in the south of the basin, which have low temperatures, a lot of snow, and heavy rain. On the other hand, the plateau and plain areas are warmer with relatively low rainfall and more evaporation.

Sampling
In the study, the least disturbed sampling areas with minimal human impacts were selected. Hydro-geographical features of 39 sampling stations symbolized with S1-S39 from 38 streams ( Fig. 1) are given in Table 1. A geographical positioning system (GPS, Garmin Vista HCx model) was used to obtain information about the locations of sampling stations. The sampling stations were classified according to typological criteria (altitude with A2 800-< 1600 m and A3 ≥ 1600 m, geology with G1-high and G2-low mineralization, drainage area with D1-wet and D2-dry regions, flow regime with F1-the seasonal and F2-continuous, slope with S1 < 2% and S2 > 2%, and precipitation with P1 < 400 mm and P2 ≥ 400 mm) of the Ministry of Agriculture and Forestry, General Directorate of Water Management of Turkey.
Water and epilithic diatom samples were collected from sampling stations in three seasons (spring, summer, and fall) of 2017 where there are two hydrological periods (rainy (spring) and dry (summer and fall) seasons). Water samples were maintained under cooler conditions until transferred to the laboratory for chemical analyses. Epilithic diatom samples were gathered by brushing the upper surfaces of at least five stones taken from the riffle parts of streams in 100 mL of distilled water, and then, they were fixed with a Lugol-glycerol solution (EC 2014).

Physicochemical analysis
Environmental variables (e.g., water temperature (°C), electrical conductivity (EC μS/cm), pH, dissolved oxygen concentration (DO, mg/L), and salinity (ppt) were measured by an oxygen-temperature meter (YSI professional plus) in situ. Values of chemical variables like TN (total nitrogen), N-NO 3 (nitrate-nitrogen), TP (total phosphorus), and TOC (total organic carbon) and biochemical oxygen demand (BOD 5 ) of water samples were determined according to standard methods of APHA (2012). The amount of metal ions in water was measured by an inductively coupled plasma-optical emission spectrometer (Perkin Elmer, Optima 2100 DV).

Fig. 1
Location of sampling stations in the Konya closed river basin. Full names of sampling stations symbolized with S1-S39 are given in Table 1 Identification and preparing permanent slides of epilithic diatoms Hot acids (H 2 SO 4 : HNO 3 as 2: 1) were used to digest diatom samples and to clean the organic materials, and then, diatom samples were washed following the European EN 13,946 (EC 2014). The Naphrax, with a refractive index of at least 1.74, was used to mount diatoms on each slide under a light microscope (Olympus BX53 model with a DP73 attached) at 1000 × magnification and then at least 500 valves of diatom species were counted per slide. Species identification was done according to taxonomic keys provided in Krammer (2000Krammer ( , 2002, Lange-Bertalot (2001), Bey andEctor (2013), andLange-Bertalot et al. (2017). Besides, the nomenclature Table 1 Hydro-geographical features of streams A, altitude (A2 800-< 1600 m and A3 ≥ 1600 m); F, flow regime (F1: intermittent and F2: permanent); S, slope (S1 < 2% and S2 > 2%); G, geology (G1: high and G2: low mineralization); P, precipitation (P1 < 400 mm and P2 ≥ 400 mm), and D, drainage area (D1: wet and D2: dry regions) was checked using AlgaeBase (Guiry and Guiry 2021) and diatoms of North America (Spaulding et al. 2019). The threat status and ecological attributes of diatom species were assigned by following Hofmann et al (2018).

Statistical analyses
Differences in mean environmental data between/among sampling stations were revealed by one-way ANOVA and Duncan's multiple range test (SPSS version 15.0, USA). The Spearman correlation test (SPSS version 15.0, USA) was applied to elucidate diatom metrics-environment correlations. The percentile analysis (SPSS version 15.0, USA) was used to figure out the 25 th , 50 th , and 75 th percentiles of the data. An analysis of similarities (ANOSIM) test was used to evaluate whether there were significant differences between the sample groups from Konya closed river basin stations. An analysis of similarity percentage (SIMPER) test was applied to determine the similarity status of the sampling groups depending on the diatom composition with the contributions of diatom species affecting these dis/similarity situations when the results of ANOSIM were significant. Bray-Curtis (similarity coefficient) was used to test the null hypothesis (there is no difference in the diatom composition of sampling stations among sampling seasons). ANOSIM and SIMPER analyses were performed using Community Analysis Package version 4.1.3 software (Seaby and Henderson 2007). Detrended correspondence analysis (DCA) test was used to estimate gradient lengths. A gradient length greater than 3.0 was calculated for the first two axes (> 6.0) that indicates the suitability of a direct gradient analysis technique. Therefore, a canonical correspondence analysis (CCA) using CANOCO 4.5 software was applied to explain relationships between five transformed environmental variables (ln(x + 1) and 98 diatom species in the Konya closed river basin (ter Braak and Smilauer 2002). Monte Carlo permutation test (CANOCO 4.5 software) was used to reveal which environmental factors significantly affect the distribution of diatom species. Thus, only significant explanatory factors were exhibited in the ordination. The weighted average (WA) regression model of the CALIBRATE program was carried out to predict the optima of diatom species for significant environmental stressors (Juggins and ter Braak 1992). Diatom species (98 taxa) occurred three or more times were used in multivariate statistical analyses (Supplementary 1).

Physical and chemical variables of sampling stations
Physical and chemical variables varied among sampling stations and their results are given in Table 2. Streams in the Konya closed river basin had alkali waters with a pH range of 7.5 at Asın Stream (S32) and 8.8 at Gök Stream (S1). The mean EC values ranged from 64 μS/cm at Baldıran Stream (S7) to 723 μS/cm at Anaçay Stream (S2). Sampling stations S1, S2, S10-12, S15, S22, and S23 had EC values of more than 500 μS/cm.
Relatively low variations were found in the nutrient gradients of sampling stations. The highest mean TP was found in Asın Stream (S32) with 142 μg/L, followed by Uluçay Stream (S34) with 116 μg/L. On the other hand, several streams (e.g., Avşar, Kurucaova, and Baldıran streams) had TP values smaller than 10 μg/L. NO 2 -N was not detected in most of the sampling stations (see Table 2). Mean TN in the Konya closed river basin varied between 1.98 mg/L in Seydihasan Stream (S21) and 0.28 mg/L in Çarşamba Stream (S25) Concerning heavy metal variations, the highest mean nickel (Ni) concentration was determined in Kovalık Stream (S9) with 13.10 μg/L, followed with 11.67 μg/L in Güvercinlik Stream (S4), whereas the lowest mean concentration of Ni (1.6 μg/L) was found in 17 sampling stations (Table 2). Some metals were measured in a few regions, such as relatively high iron (Fe) values were measured in S30-S34 (> 231.0 μg/L) at Erenler Mountain, S3 (632.1 μg/L) and S5 (501.7 μg/L) at Melendiz Mountains, and S14 (417.8 μg/L) at Taurus Mountains. The highest mean boron (B) value was found in Ilısu Stream (S3) with 601.7 μg/L, followed by Yağıl Stream (S6) with 309.5 μg/L and Okçu Stream (S5) with 208.9 μg/L. Ilısu Stream also had relatively high concentrations of metals like arsenic (As), boron (B), iron (Fe), Ti, and V (see Table 2).

Bio-assessment of sampling stations
Dia-assessments of sampling stations in the Konya closed river basin are given in Table 3. Ecological conditions ranging from high to bad were found in the Konya closed river basin according to responses of different ecoregional diatom indices. Percentages of the ecological quality status of diatom indices are provided in Fig. 4. The highest percentage of high ecological status (30.8%) was determined by the TIT, followed by EPI-D (23.1%), DEQI (20.1%), and IPS (10.8%) (Fig. 4). DSIAR and DDI could not separate the ecological status of sampling stations.
Diatom indices showed different ecological conditions for sampling stations (Table 3). For instance, S32 had five different environmental conditions, such as a bad status due to TIT and TWQI; a poor condition according to TI; a moderate status after being evaluated by EPI-D, IPS, and TDI; and a good status after being assessed by DEQI. Similarly, four different ecological conditions were found in S33. European indices displayed a good status for S9, whereas it was assigned a bad condition according to DEQI. European diatom indices except DDI mainly displayed similar behavior in eco-assessment of sampling stations in the Konya closed river basin (Table 3 and Fig. 4). European diatom indices indicated that the least disturbed areas are S16, S20, S24, S25, S27-29, S37, and 39 with high ecological status in the Konya closed river basin.
Correlations between diatom indices and environmental variables are represented in Table 4. IPS, DSIAR, and DDI showed inverse correlations with environmental pollution, while others indicated direct correlations. A few diatom indices such as TIT (p < 0.01, r = 0.506), TI (p < 0.01, r = 0.386), EPI-D (p < 0.01, r = 0.367) had significant positive correlations with the TP gradient, whereas IPS (p < 0.01, r = − 0.477) and DDI (p < 0.05, r = − 0.304) displayed significantly negative correlations with TP (see more in Table 4). There was no significant correlation between the diatom indices and EC and BOD5, which are not shown in Table 4. TIT, TI, EPI-D, and TDI had positive correlations with B and V, whereas IPS and DDI were negatively correlated (Table 4).  Table 1 and Supplementary 1, respectively

Discussion
It is the first study to describe the diatom composition of river ecosystems in the Konya closed basin and to provide the eco-assessment of sampling stations as a benchmark for future limno-ecological studies. According to Turkish surface water quality regulation standards (TSWQR 2016), sampling stations except S32 and S34 have class I water quality based on TP, TN, and NO3-N variables. Similar nitrogen values were found in the Antalya River Basin (Çelekli et al. 2021b), which is a contiguous basin to the Konya closed river basin. In contrast, TP values are higher than those of the Antalya River system. The Antalya River basin has different ecological areas with different altitudes and higher annual precipitation compared to the Konya closed river basin. TP values in the European Mediterranean Rivers with the least disturbed conditions (Feio et al. 2014) were lower than those measured in the present study. The maximum threshold values of total phosphorus as a limiting factor (Reddy et al. 1999) are 70 μg/L for European Mediterranean Rivers (Feio et al. 2014), 30 μg/L for New York State (Smith and Tran 2010), and 80 μg/L for Turkey (TSWQR 2016). Increasing of phosphorus in streams is closely related to human activities as a result of runoff Fig. 4 The percentages of ecological quality status of diatom indices in the Konya closed river basin Table 4 Spearman's rank correlations between diatom indices and environmental factors The correlation between diatom indices and TP (a relevant chemical to quantify the trophic weight of species) is given in bold * and ** show significant levels at 0.05 and 0.01, respectively. Abbreviations of environmental variables and diatom indices are given in Tables 2 and 3, respectively   TIT  TI  EPI-D IPS  TDI  DSIAR  DEQI  TWQI  fertilizer from agricultural lands, wastewater discharge, etc. (Feio et al. 2014). Turkish regulation standard for surface water indicates that sampling stations except S1, S2, S10, S11, S12, S15, S19, S21-23, S32, S36, and S38 have highwater qualities according to EC gradient (EC < 400 μS/cm) (Das et al. 2006;. Relatively high values of some heavy metal ions were found in some stations: Ni in S4, S8-10 located at Melendiz and Taurus mountains, Cu in S33 and S32 located at Erenler Mountain and S14 and S15 at Taurus Mountains, B in S3, S5, and S6 at Melendiz Mountain, and Fe in S3-S5, S7, and S8 at Melendiz Mountain and S30-34 at Erenler Mountain. There are no industrial activities in these mountainous regions, but mineral mining near sampling stations on these mountains might be the cause of these heavy metals in the southeast of Anatolia (Çelekli et al. 2017). Heavy metal values in the present study were higher than those of rivers in the Antalya river system (Çelekli et al. 2021b). Turkish regulation standard for surface water represents that S6, S12, S18-S21, S27, S28, and S36-S39 have high-water qualities when the remaining sampling stations are of water quality II according to Fe ion gradient (TSWQR 2016).
Diatom species showed distinct responses to environmental variables such as EC, Ni, Cu, B, and altitude, driving the distribution of diatom species in the Konya closed river system. Relatively high EC values were found in Zorlak, Doğuca, Asın, and Seydihasan streams, which are characterized by some diatom species (e.g.,N. dealpina,G. schoenfeldii,C. cymbiformis,G. augur,G. angustius,C. silicula,D. oblongella,C. pseudolineata,N. dissipata,and N. cari (Fig. 2)) in agreement with WA's results (Fig. 3a). Navicula dealpina is a species in the Red List (RL) threat-category 2 in serious danger (Hofmann et al. 2018) and no information is available about its ecological status and biological condition gradient-BCG attribute (Hausmann et al. 2016). In addition, used diatom indices have no trophic weight for this species while it had a 93.3 μg/L TP optima in the Konya closed river basin. A medium EC optimum level of N. dealpina in the present study supported the finding of Cantonati et al (2016). Also, they reported that N. dealpina was assigned as a characteristic species of oligotrophic and calcium-bicarbonate-rich environments. Geissleria schoenfeldii is in the RL threat-category X and in serious danger lists (Hofmann et al. 2018), and its ecological status is eutrophic. The present study resulted in a 12.8 μg/L TP optimum for this species. This species (synonym Navicula schoenfeldii) is not found in the used diatom indices database and so more studies are needed to determine the ecological preference of this species in the riverine system. Cymbella cymbiformis is related to oligotrophic, predominantly slightly acidic, more or less humin-rich waters and species is in the RL threat-category 3 (endangered) (Hofmann et al. 2018). The present study revealed that its TP optimum is 25.1 μg/L. TIT indicated that this species prefers moderately deteriorated ecosystems, whereas TI, IPS, and EPI-D showed that it has a low trophic weight in the least disturbed waters. Gomphonema augur prefers eutrophic conditions (Hofmann et al. 2018). In the Konya closed river basin, this species occurred in deteriorated ecosystems with its high TP optimum of 63.1 μg/L, which is confirmed by TIT, TI, and EPI-D. Gomphonema angustius was shown to be a characteristic species of oligotrophic-calcareous waters (Reichardt 2009). This species had a 12.1 μg/L TP optimum in the Konya closed river basin. Among the used diatom indices, IPS only includes this species with low trophic weight (Cemagref 1982). Caloneis silicula preferred moderate environmental conditions based on TIT, TI, and DSIAR, whereas a good environment according to EPI-D and IPS. Navicula cari (36.8 μg/L TP optimum in the present study) is considered a characteristic species of the eutrophic environment (Hofmann et al. 2018) and has BCG attribute 4 as an intermediate tolerant species (Hausmann et al. 2016). Cantonati et al. (2016) reported that N. cari is related to eutrophic waters with higher electrolyte content. Besides, this species has a higher trophic weight in TIT (Çelekli et al. 2019) than those of TI (Rott et al. 1999), EPI-D (Dell'Uomo 2004), and IPS (Cemagref 1982). Cocconeis pseudolineata had a 15.8 μg/L TP optimum in the present study. It has a medium trophic weight in TDI (Kelly et al. 2008), whereas IPS indicates that it is a sensitive species to pollution (Cemagref 1982). S1, S8, S9, S14, and S15 were characterized by the association of N. veneta, N. palea, G. angustatum, S. ovalis, and A. ovalis with Ni ions (Fig. 2), which is also confirmed by the WA findings (Fig. 3b). These species and N. parvula had Cu optima higher than 75% in the Konya river basin (Fig. 3c). Navicula veneta prefers eutrophic environments (Hofmann et al. 2018) and is grouped under the BCG attribute 4 (Hausmann et al. 2016). Nitzschia palea is a characteristic species of the eutrophic environment (Hofmann et al. 2018), which is described in the BCG attribute 5 as a tolerant taxon (Hausmann et al. 2016). Nitzschia palea is also considered a pollution-tolerant species, with nutrient gradients ranging from low in the least disturbed areas Çelekli et al. 2021b) to high in the wastewater (Çelekli and Şahin 2021). Previously, close relationships of N. veneta and N. palea with Ni, Cu, and Cd ions were reported in Junction Creek, Ontario, Canada (Lavoie et al. 2018) and in Frood Branch and Nolin creeks of Greater Sudbury, Canada (Lavoie et al. 2018). The amounts of these metal ions are lower than those of the present study. Nitzschia palea is also known as a heavy metal tolerant species (Chen et al. 2014). Besides, N. veneta and N. palea are put in the species group with high tolerance to higher nutrient levels (Lavoie et al. 2018). S3 was characterized by N. flexa, N. gregaria, H. amphioxys, and R. abbreviata and showed a close relationship with B in the CCA ordination. This association was also supported by findings from WA (Fig. 3d). Navicula gregaria is related to the BCG attribute 4 as an intermediate tolerant and ubiquitous taxon (Hausmann et al. 2016) and has a relatively high trophic weight (Rott et al. 1999;Dell'Uomo 2004). Hantzschia amphioxys prefers aerophilic habitats (Hofmann et al. 2018) and is shown as a pollution tolerant organism (Rott et al. 1999;Dell'Uomo 2004). Rhoicosphenia abbreviata is commonly found in eutrophic environments (Hofmann et al. 2018), and it is indicated in the BCG attribute 3 as an intermediate sensitive species (Hausmann et al. 2016). These findings were also supported by the findings of TIT (Çelekli et al. 2019) and TI (Rott et al. 1999).
The eco-assessment of sampling stations in the Konya closed river basin indicated that diatom indices have various scores resulting in different ecological statuses ranging from a bad to a high environmental condition. TIT had the highest percentage of high ecological status, which was followed by EPI-D, DEQI, and IPS (Fig. 3).
Diatom indices had different behaviors in the bio-assessment of some sampling stations, such as S32, S33, and S9 in the Konya closed river basin. Similar results were also found in the study comparing European and American diatom indices (Lavoie et al. 2009), southeast Anatolia (Çelekli and Bilgi 2019) and in the Antalya river basin (Çelekli et al. 2021b). Five different ecological conditions were reported in S32: a bad status with TIT and TWQI; a poor condition with TI; a moderate status with EPI-D, IPS, and TDI; and a good status with DEQI. S32 had the highest values of TP, BOD, and TOC with medium EC, which suggests class II water quality according to Turkish regulation standards (TSWQR 2016). Urbanization, agricultural activity, and land use are present around S32. S33 had four different ecological conditions based on used diatom indices (Table 3). S33 was of a high ecological status based on TIT, whereas IPS, TI, EPI-D, and TDI indicated a poor environmental condition. Another index. DDI depicted a good ecological status for S33. Differences in the bio-assessment may be explained by the variation in the trophic weights of diatom species in diatom indices. For instance, M. circulare has a low trophic weight of TIT and EPI-D compared to TI, IPS, and DSIAR. Streambed (S33) consists of rocks, stones, and sand, which is surrounded by a small amount of agricultural land, including mostly poplar trees. European indices displayed that S9 had a good status whereas it was assigned as a station with bad condition based on the DEQI result. Mean values of EC, BOD, TN, and TP indicated that S9 has class I water quality according to Turkish regulation standards (TSWQR 2016). In terms of metal content, this stream had relatively high Fe and Ni ions, which could affect diatom composition and their abundances. Similar to the findings of the eastern Canadian diatom index-IDEC in Frood Branch and Nolin creeks, Canada (Lavoie et al. 2018), increasing in diatom scores could have been caused by metal ion contamination in the present study.
European diatom indices, TI (Rott et al. 1999), TIT (Çelekli et al. 2019), EPI-D (Dell'Uomo 2004), and IPS (Cemagref 1982) except DDI, mainly displayed similar behavior in eco-assessment of sampling stations in the Konya closed river basin (Table 3 and Fig. 4). Some indices (e.g., DDI, DSIAR, and TWQI) could not separate the ecological status of sampling stations in the Konya closed river basin, in agreement with results from the Antalya river basin (Çelekli et al. 2021b). This may be due to the insensitiveness of theses indices to the environmental variations between sampling sites. The high and good ecological quality class boundaries of the DDI are between 10.00 and 8.50-9.99, respectively (Álvarez-Blanco et al. 2013), which strongly affect the bio-assessment results as good ecological status for all stations. Similar results were obtained from the DDI evaluation in the Antalya river basin (Çelekli et al. 2021b). Thus, determination of class boundaries of diatom indices can accurately facilitate the bio-assessment of freshwater ecosystems. DSIAR and TWQI were developed from different ecoregions, which closely affect the trophic weights of diatom species in the database of diatom indices. Another important factor is how many diatom species found in the Konya river basin were used in the diatom indices. For instance, DEQI only included N. veneta (trophic weight equal to 5), which is used in the bio-evaluation of S9 as a bad condition. Besides, diatom species found in the station S9 are not used in the TWQI index database and so, the TWQI index could not be used to assess the ecological status of S9 (Table 3).
European diatom indices indicated that the least disturbed areas are S20, S24, S25, S27-29, S37, and 39 with high ecological status in the Konya closed river basin. Diatom species diversity and trophic weights (distinct tolerant/sensitivity values) are different in the tested diatom database, which closely affects results of the bio-assessment of streams in the Konya river basin. This is because different ecoregions (e.g., geology, human activities, climate, vegetation, wildlife, and hydromorphology) strongly affect the trophic weight values of species (Lobo et al. 2015;Salinas-Camarillo et al. 2021). Therefore, diatom indices developed in different ecoregions can give distinct ecological status for the same sampling site.
TI (Rott et al. 1999), TIT ), IPS (Cemagref 1982, and EPI-D (Dell'Uomo 2004) significantly correlated with environmental variables, especially TP, which is used as a relevant chemical to quantify the trophic weight of species. Thus, TI, IPS, and EPI-D are commonly used to assess the water quality of European water bodies incorporating nutrients and hydromorphology. Among them, TIT was shown to be a more competitive index for the bioassessment of rivers in the Antalya river basin (Çelekli et al. 2021b), the West Mediterranean basin of Turkey (Çelekli and Lekesiz 2020), eight creeks in the Lake Sapanca basin (Sevindik et al. 2021), the west of the Gaziantep catchment (Çelekli and Kapı 2019) and the North Aegean catchment (Çelekli et al. 2018) than those of the Konya closed river basin. The diatom list of the performed diatom indices does not include all the diatom species found in sampling stations in the Konya closed river basin, which is a limiting factor in assessing the water quality. Although TIT had the lowest species number (38%) among European diatom indices (e.g., IPS (71%), TI (62%), and EPI-D (60%)), it was found to be a competitive index. In light of this information, ecoregionally specific diatom indices like TIT need to be revised in the future by increasing the number of diatom species taking into account the ecological preferences of diatom taxa to make bioassessment more accurate.

Conclusion
The present study underlines that diatom species have distinct responses to environmental variables in the Konya closed river basin, significantly affected by EC, Ni, Cu, B, and altitude. Cocconeis euglypta, C. excisa, C. placentula, and A. minutissimum played significant roles in the dissimilarity of sampling stations not only between rainy and dry seasons but also between altitude typological criteria A2 and A3 in the Konya closed river basin. Results of the bio-assessment revealed that diatom indices have various scores, resulting in different ecological statuses from bad to high ecological conditions of sampling stations in the Konya closed river basin. European diatom indices except DDI and TDI indicated that the least disturbed areas are S16, S20, S24, S25, S27-29, S37, and S39 with high ecological status, which was also confirmed by physico-chemical and hydromorphological evaluations of streams. TIT, TI, EPI-D, and IPS could be suitable diatom metrics for assessing the ecological status of sampling stations in this region. Results also indicated that revision of ecoregional specific diatom metrics like TIT with the enhancement of diatom species number is needed along with the ecological preferences of diatom taxa to effectively interpret the water quality.