Aquatic Worm Assemblages along the Danube: A Homogenization Warning

: In this study, we analyzed the impacts of di ﬀ erent environmental conditions on aquatic worm communities along the Danube River, based on two longitudinal surveys, the Joint Danube Surveys 2 and 3 (JDS; 2007 and 2013). We identiﬁed the most important environmental factors (among analyzed groups) that shape worm communities: hydromorphlogical alterations, ﬂow velocity and substrate (HYMO group), dissolved oxygen, nitrates and nitrites (physico-chemical parameters), zinc and nickel (metals), monobutyltin cation, benzo(b) ﬂuoranthene and benzo(k)ﬂuoranthene, polychlorinated biphenyls PCB 77 and PCB 118 (selected chemical determinants—organotin compounds, Polycyclic aromatic hydrocarbons—PAHs and PCBs). A homogenization of species composition of Oligochaeta assemblages along the Danube was conﬁrmed. As one of main factors related to biotic homogenization, hydromorphological alterations represented by similar changes in ﬂow velocity and substrates along Danube’s course could be singled out. Our results indicate that Oligochaeta could be used for the identiﬁcation of the level of hydromorphological degradation in large rivers (homogenization), rather than for stressors classiﬁed as nutrient and organic pollutants. Our results provide additional evidence in risk assessment of the environment, contributing in water management and monitoring of the ecological status as proposed by the Water Framework Directive.


Introduction
The Danube is a large water resource for more than 80 million people. As one of the longest rivers in Europe, it represents an important connecting factor for biodiversity conservation [1][2][3][4]. It is therefore essential to collect reliable and comparable data on the functioning of this river and to adequately translate the results from research to the management level. To this end, the examination of the relation between environmental parameters and structures of communities and the subsequent identification of the typology based on selected natural characteristics of water types must be undertaken. These activities will provide the foundation for effective water management and monitoring of the ecological status, as proposed by the Water Framework Directive (WFD) [5]. Grouping of similar investigations of the Danube and other large lowland rivers in Europe, where the highest diversity was displayed by the tubificins [11][12][13][14][15][16].
Many hydraulic constructions have been built on the Danube River and its tributaries, and about 70 reservoirs have been constructed by 1980 [17]. The upper course of the Danube (Germany and Austria) has more weirs than the middle course-44 impoundments from 2600 to 2000 Rkm and three more dams from 1950 Rkm downstream [2]. Many of these dams were built to improve the navigation conditions and were unable to regulate water flow, however, they contributed to a certain decrease in suspended sediment runoff [17]. Poff et al. [18] showed that natural flow regimes have been significantly reduced in regulated rivers, and that postdam homogenization is statistically significant, with a broad range of hydrology alterations caused by dams. The consequence of homogenization on aquatic systems is the replacement of regional biotas with cosmopolitan species [19].
Aquatic worms, as one of the main components of communities in large rivers, are dominated by widely distributed taxa. Therefore, the information collected from analyses of the Danube could be used for other large rivers, not only in Europe, but in Palearctic and Nearctic areas. The basic faunistic features of the oligochaetes assemblages and the distribution of Oligochaeta (Annelida) in the Danube River were discussed in Atanacković et al. [16], where it was pointed out that the structure of the bottom sediment is an important factor which influences species composition. The value of oligochaetes as indicators in Europe is well known and these organisms have been included in different monitoring programs [20][21][22][23]. In contrast to some other organisms such as Diptera, which are used in monitoring, oligochaetes are more suitable because they filter the mud. The sediment is not only a passive reservoir of different pollutants, but oligochaetes themselves through burial, substrate plowing, nutrition, and respiration, represent significant accumulators in aquatic ecosystems, especially as they play an important ecological role in sediments when occurring at high densities, as we recorded in the Danube. Pollutants indirectly become toxic to animals at a higher trophic level.
Further, hydrophobic organic micropollutants (such as organotins, PAHs, and PCBs) in aquatic ecosystems have increased towards the end of the 20th century, and several of these organic pollutants belong to the Water Framework Directive priority substances (EQS Directive, 2013/39/EU), and have hydrophobic properties. PAHs, PCBs, and organotin compounds are therefore usually associated with suspended particulate matter that settles in the bottom sediment, and an extensive database on aquatic oligochaetes has demonstrated their sensitivity to a range of chemicals (metals, pesticides, PAHs, etc.) [23].
The focus of this paper is on the relation of Oligochaeta assemblages and the available data on environmental variables, such as physico-chemical parameters and chemical determinants in water and sediment and the level of hydromorphological degradation. Since the information of concentrations of pollutants and the macroinvertebrate community have been published in the Joint Danube Survey Reports [7,8], the novelty of this paper is the detailed analysis of oligochaetes assemblages along the Danube River and the statistical analysis that was undertaken in order to find relationships between the chemical and hydromorphological environmental variables and biological data. The construction of dams and river regulation has led to the homogenization of habitats and thus to an increase in limno(rheo)philic taxa that prefer slow-flowing and lentic zones. As there is a lack of papers dealing with the homogenization (biotic and hydrological) of large European rivers, the present paper is a contribution to this field. We hypothesized that the Oligochaeta are suitable bioindicators of the deterioration of environmental conditions, and we recognized these organisms as typology descriptors (indicators for the homogenization of the river) for sectioning of the Danube.

Study Area and Selection of Sampling Points
Oligochaeta assemblages were collected in the frame of the Joint Danube Surveys JDS2 [7] and JDS3 [8], during 2007 and 2013, respectively. More than 2580 km of the Danube were investigated, including 96 sites during the JDS2, and 68 sites during the JDS3 (Figure 1).

Study Area and Selection of Sampling Points
Oligochaeta assemblages were collected in the frame of the Joint Danube Surveys JDS2 [7] and JDS3 [8], during 2007 and 2013, respectively. More than 2580 km of the Danube were investigated, including 96 sites during the JDS2, and 68 sites during the JDS3 (Figure 1). While keeping in mind the size of the watercourse and the significant number of analyzed parameters, an a priori classification of sectors was applied [24], which implies division of the river into 10, biologically meaningful section types, as follows: Sector 1-Upper Course of the Danube A discussion regarding the selection of sampling sites was presented in Liška et al. [7,8].
The overall descriptions of the hydromorphology during the surveys have remained comparable/essentially unchanged. The channel patterns have been significantly altered along the entire Danube by navigation and hydroelectric power plants, with only a very few sections retaining natural banks (in 6.45% of sites). The banks of the Danube have been entirely transformed in urban areas.
A significant stretch of the Danube (21.28% of the river course) is affected by hydromorphological changes (backwater/impoundment), especially in the upper reach in Germany and Austria [7,8]. About 77% of the river course belongs to a good hydromorphological class (free flowing). One percent of the investigated sites (Sector 1) exhibit a significantly reduced water flow. Studies during JDS3 [8] showed that 21% of the analyzed reach was slightly modified, 39% was While keeping in mind the size of the watercourse and the significant number of analyzed parameters, an a priori classification of sectors was applied [24], which implies division of the river into 10, biologically meaningful section types, as follows: Sector 1-Upper Course of the Danube A discussion regarding the selection of sampling sites was presented in Liška et al. [7,8].
The overall descriptions of the hydromorphology during the surveys have remained comparable/essentially unchanged. The channel patterns have been significantly altered along the entire Danube by navigation and hydroelectric power plants, with only a very few sections retaining natural banks (in 6.45% of sites). The banks of the Danube have been entirely transformed in urban areas.
A significant stretch of the Danube (21.28% of the river course) is affected by hydromorphological changes (backwater/impoundment), especially in the upper reach in Germany and Austria [7,8]. About 77% of the river course belongs to a good hydromorphological class (free flowing). One percent of the investigated sites (Sector 1) exhibit a significantly reduced water flow. Studies during JDS3 [8] Water 2020, 12, 2612 5 of 21 showed that 21% of the analyzed reach was slightly modified, 39% was moderately modified, and 40% fell into extensively and severely modified sites, while "near natural" sites could not be found at all.
The assessment results from the JDS3 have confirmed the main findings of the JDS2 in 2007. The bed material of the investigated sampling sites with strong impoundments is mainly composed of fine sediment. Fine, medium, and coarse gravel is present with a participation of 31.92%, while only 6% of the riverbed is covered with bedrock. These sites, which are mostly downstream of dams, are missing fine fractions. Some areas (9.57% of sites) have a significant amount of organic matter. In certain areas, the riverbed is covered by macrophyte vegetation (about 34% of the riverbed in the littoral zone). Significant changes in river processes (erosion/deposition) that are caused by various degrees of hydromorphological degradation along the Danube were observed, such as the decrease of suspended sediment concentration along the impounded sections, deposition areas upstream of the barrier, and the absence of sediments in the downstream direction, especially a deficit of fine sediments downstream of the Iron Gate [8].

Data Collection
The employed sampling methodology is explained in detail in Liška et al. [7,8]. The biological component was collected, identified, and analyzed by the authors. The following sampling methods during the surveys were used: air-lift samples were used in the study of the faunal composition of deep water habitats and a modified Multi-Habitat-Sampling (MHS) approach [25] was performed to highlight the importance of specific micro-habitats. A combination of techniques was used in both surveys: kick and sweep multihabitat sampling or K&S (FBA hand net, mesh size 500 µm) [26] and dredging.
The following environmental factors were analyzed during the surveys JDS2 and JDS3: hydrological alterations, flow velocity, substrate type, and bank modifications (HYMO group). The physical and chemical parameters were analyzed too. Data on physical and chemical parameters from both surveys are available in the ICPDR Water Quality Database (http://www.icpdr.org/wq-db/). We used information from the database for common parameters such as water temperature, pH, dissolved oxygen, conductivity, ammonia (NH 4 + ), nitrites (NO 2 − ), nitrates (NO 3 − ), orthophosphates (PO 4 − ) together with the concentrations of WFD priority substances: diethylhexyl phthalate (DEHP), PAHs, PCBs, organotin compounds, dioxins and dioxin-like compounds (polychlorinated dibenzo-p-dioxins-PCDDs, polychlorinated dibenzofurans-PCDFs, and dioxin-like polychlorinated biphenyls-PCB-DL), and metals (copper (Cu), arsenic (As), cadmium (Cd), chromium (Cr), nickel (Ni), and lead (Pb)) in the sediments. The measured concentrations of heavy metals were estimated in relation to target values evaluated for heavy metals [27]. It should be mentioned that there are currently no commonly accepted limit values for chemical pollutants in sediments in the European water legislation [1].

Data Analyses
The frequency of occurrence (F) for each species in oligochaetes assemblages was calculated using the formula: F = n/N, where n is the number of samples in which a taxon was found, and N is the total number of samples.
The percentage participation of species for each sector was calculated and is presented in Tables 1 and 2. Taxa that were not identified up to the species level are not shown in the tables.
In order to determine the species variability in different sectors along the Danube, the preliminary discriminant analysis (DA) [28,29] was applied. This analysis was used on species data to check whether specific sectors could be identified. An input matrix consists of 66 Oligochaeta taxa and 590 samples (collected using K&S, airlift and dredging, with 183 and 407 samples gathered during the JDS2 and JDS3, respectively). As the first step, standardization of the raw species data was performed (species percentages per sample/relative abundance) to avoid inconsistencies of different data formats obtained by different sampling techniques.   Canonical correspondence analysis (CCA) [30] served to determine the relation between species distribution and environmental variables (six groups in total: hydromorphological alterations, physico-chemical parameters in water samples, and metal concentrations as well as organotins, PAHs, and PCBs in sediments). After preliminary testing of the analyzed variables, ammonium and dioxins (with a variance close to zero) were excluded from further analyses. The significance of CCA, with all environmental variables included, was tested to check the applicability of forward selection (FS) [31] of the environmental variables. The obtained results for the R correlation test were considered significant at p < 0.05. Forward selection was used to select environmental variables that explained most of the variations observed in the primary taxa matrix. The variables were processed by an automatic FS with a Monte Carlo permutation test [32] by using "Flora" Software (Version 2011) [33]. The same software was used to perform DA and CCA.

Environmental Data
During both surveys the water temperatures varied from 16.9 to 26.6 • C. Values of pH showed a slight increase in the upper section of the Danube (i.e., a slightly more alkaline environment) but as Liška et al. [8] observed, the pH and dissolved oxygen content demonstrated an overall good balance between primary production and decomposition of organic matter. Concerning the dissolved oxygen content, the longitudinal profile accompanied the pH profile with an increase in the upper Danube. Conductivity was relatively constant along the main watercourse, exhibiting several distinctive profiles, such as a significant decrease in the measured values in the Upper Danube during 2013 (from 566 µS/cm at Böfinger Halde, to 320 µS/cm at Oberloiben).
Along the river course, a markedly declining profile for ammonium was obtained during both years of investigation, excluding a few specific peaks in the upper course. In the majority of sampling sites, the concentration of ammonium was below the limit of detection. The spatial variations in nitrite and nitrate concentrations displayed a significant decrease profile in the upper course of the Danube. This was also observed along the middle stretch, while downstream, from the Iron Gate dams, a nitrate increase was determined (similar to the ammonium, nitrite, and dissolved oxygen).
Orthophosphate exhibited high spatial variability along the river. Decreasing concentration was noted in the upper course. In the middle course, we obtained a slightly ascending profile, especially in the backwaters of the Iron Gates reservoir, whereas in the lower course, an increase was observed.
According to Liška et al. [8], the total nitrogen and phosphorous contents were comparable in the two JDSs.
Significant variations in cadmium contents were observed downstream of the two major tributaries, Tisa and Sava. At 75.54% of sites, the concentration value was greater than 0.8 mg/kg (the standard value for the concentrations in sediment [27]). In the case of lead, the longitudinal profile indicated low concentrations along the Upper Danube, and high variation downstream from the confluence of the Tisa and Sava rivers. The sediment standard was exceeded (85 mg/kg) in 21.08% of the samples. During the JDS2 the concentration of nickel in bottom sediment at most sampling sites (86.41%) exceeded the 50 mg/kg (standard value for the concentration in sediment) and significantly increased downstream from the confluence of the Tisa and Sava rivers. Most of the results for arsenic were between 50 and 100 mg/kg, however, a very high concentration of 432.27 mg/kg was measured at sampling site JDS 60 (Iron Gate reservoir). The overall distributions of concentrations of copper, chromium, and zinc were similar to that of nickel.
In general, the concentrations of metals in the bottom sediments determined during the JDS3 were similar to those observed during the JDS2 [7]. The longitudinal concentration profiles for most metals in sediments of JDS3 declined compared to the JDS2 (especially for Cu, Ni, and Zn).
DEHP was present in most sediment samples, with a maximum concentration of 26 mg/kg at sampling site JDS 9 (Klosterneuburg). The sediment downstream from the confluence of the Velika Morava was contaminated with 16.7 mg/kg DEHP. In JDS3, the concentrations of DEHP were higher in the middle part of the Danube compared to the observations during the JDS2. The presence of PAHs in the sediment was observed in more than 50% of sites, the maximum values ranging between 57 and 489 µg/kg.
Monobutyltin, dibutyltin, and triphenyltin were the most abundant organotin compounds in the sediment samples detected in both longitudinal surveys with similar concentrations. The different dioxins and dioxin-like compounds are converted into toxic equivalents (according to WHO 2005 Toxic Equivalence Factors-TEF) and summed up. Most PCDDs and PCDFs were quantified in all samples. These compounds, analyzed during the JDS3 survey, can be considered as emerging substances in sediment samples [8]. Comparison of the concentrations in the sediment revealed results similar to those obtained during the JDS2.
Due to the large amount of data collected during the surveys, the concentrations for most important environmental parameters are presented in Figure 2.

Fauna
The distribution of Oligochaeta taxa along the Danube during JDS2 and JDS3 is presented in Tables 1 and 2, respectively. Oligochaeta with 66 recorded taxa, represent a significant component of the macroinvertebrate community of the Danube. The species richness was similar during both the JDS2 and JDS3.

Fauna
The distribution of Oligochaeta taxa along the Danube during JDS2 and JDS3 is presented in Tables 1 and 2, respectively. Oligochaeta with 66 recorded taxa, represent a significant component of the macroinvertebrate community of the Danube. The species richness was similar during both the JDS2 and JDS3.
The most diverse family was Naididae, with 26 species of subfamily Tubificinae and 17 species of subfamily Naidinae, followed by families Lumbriculidae (five taxa) and Enchytraeidae (two taxa). The families Propappidae, Haplotaxidae, and Criodrilidae were represented by one species each, which is in accordance with their distribution in Europe, while the family Lumbricidae was not identified to the species level, except Eiseniella tetraedra.
During the JDS2, the highest number of aquatic worm taxa was recorded in the middle stretch of the Danube, the Hungarian Danube Bend, and the Western Pontic Danube (sectors 5 and 8, respectively). During the JDS3, the diversity of oligochaetes assemblages was uniform along the entire stretch. Lower numbers of species were in the upper course of the Danube (JDS2) and the Danube Delta (both JDSs).

Multivariate Analysis
During the first analysis stage, estimation of the DA for the community data (relative abundance) was performed in order to define the main community patterns along the longitudinal profile of the Danube. Based on the performed analysis, (DA diagram, Figure 3), the majority of the defined sectors [24] could not be separated, but a distinction between the samples from the lower Danube (sectors 7-10) and samples from all other sectors (upper and middle Danube) was observed.
During the JDS2, the highest number of aquatic worm taxa was recorded in the middle stretch of the Danube, the Hungarian Danube Bend, and the Western Pontic Danube (sectors 5 and 8, respectively). During the JDS3, the diversity of oligochaetes assemblages was uniform along the entire stretch. Lower numbers of species were in the upper course of the Danube (JDS2) and the Danube Delta (both JDSs).

Multivariate Analysis
During the first analysis stage, estimation of the DA for the community data (relative abundance) was performed in order to define the main community patterns along the longitudinal profile of the Danube. Based on the performed analysis, (DA diagram, Figure 3), the majority of the defined sectors [24] could not be separated, but a distinction between the samples from the lower Danube (sectors 7-10) and samples from all other sectors (upper and middle Danube) was observed.  Especially, high heterogeneity between samples collected from the Hungarian and Serbian (sectors 5 and 6) and German stretch (sector 2) of the Danube was observed. Prominent species from the Upper Danube were S. heringianus, E. tetraedra, N. bretscheri, Lumbriculidae, P. moldaviensis, P. barbatus, and for the lower Danube: I. michaelseni, N. christinae, O. serpentina, B. sowerbyi.
CCA was used to analyze complex multivariate interactions between oligochaetes and the selected environmental variables. To assess whether the taxa matrix could be elucidated by the environmental matrix, a preliminary CCA (correlation coefficient (R) at p < 0.05) was performed for each group of environmental variables. Forward selection showed that the most important environmental variables that determine the taxa matrix/species distribution (from each environmental group) were as follows: hydrological alterations, flow velocity and substrate (HYMO group), dissolved oxygen, nitrates and nitrites (physico-chemical parameters), zinc and nickel (metals), monobutyltin cation (organotins group), benzo(b)fluoranthene and benzo(k)fluoranthene (PAHs), and PCB 77 and 118 (PCBs).
After CCA analysis of the selected environmental variables and taxa data, we obtained insight into the major environmental gradients. The results regarding the relations of the HYMO group and the Oligochaeta assemblages are presented in Figure 4a. The negative correlation between hydrological alterations on the one hand, and flow velocity and substrate on the other (more prominent alterations, lower flow and finer substrate, and vice versa) should be noted. Hydrological alterations were identified as the most important environmental variable overall, particularly along the first CCA axis; it was the main explainable variation in the taxa matrix/distribution. The vast majority of taxa seem to be unresponsive to these factors (the large taxa group situated near the coordinate center of the CCA diagram). Species that tolerate faster flow and a larger substrate (with more detritus/organic components) are on the right side of the diagram: N. simplex, N. communis, T. ignotus, and P. vejdovskyi. On the other hand, R. limosela, Limnodrilus sp., L. claparedeanus, Stylodrilus sp., and B. sowerbyi are taxa associated with more prominent hydrological alterations. Analysis of the relation of oligochaetes assemblages and these factors revealed that species that exhibit significant differences were the most numerous, as expected.
The impact of the main physico-chemical characteristic of the water on oligochaetes assemblages along the Danube is presented in Figure 4b. Dissolved oxygen was found to be the most important factor, and it is positively correlated with nitrates, whereas a negative correlation was noted with nitrites. Oligochaeta that favor more oxygenated water and tolerate higher concentrations of nitrates were L. profundicola, L. udekemianus, and L. hoffmeisteri. On the other side of the diagram, Limnodrilus sp., T. tubifex, and B. sowerbyi were associated with increased concentrations of nitrites and were more tolerant to lower concentrations of dissolved oxygen.
Analysis of the relations of metals in sediments and Oligochaeta revealed that zinc and nickel were the most important environmental variables. Oligochaeta assemblages in the multivariate space defined by selected factors are shown in Figure 4c. High concentrations of these metals were recorded during the JDS2 in the majority of sample sites belonging to sectors 6-10 (values over 400 and 100 mg/kg, respectively). The highest values for zinc and nickel were obtained at site 51, the confluence of the Sava River (745 and 210 mg/kg, respectively). Zinc and nickel are negatively correlated with the first CCA axis (69% of explained variance). The distinctive sample on the diagram was collected at Upstream Cernavoda (sector 9). Oligochaeta taxa tolerant to increased concentrations of these metals are as follows: A. japonicus, P. volki, A. limnobius, N. simplex, N. communis, P. albicola, and P. isochaetus. Among them A. japonicus, A. limnobius, and P. volki tolerated nickel (CCA axis 2), while N. simplex, N. communis, and P. albicola tolerated zinc.
Water 2020, 12, x FOR PEER REVIEW 13 of 21 CCA was used to analyze complex multivariate interactions between oligochaetes and the selected environmental variables. To assess whether the taxa matrix could be elucidated by the environmental matrix, a preliminary CCA (correlation coefficient (R) at p < 0.05) was performed for each group of environmental variables. Forward selection showed that the most important environmental variables that determine the taxa matrix/species distribution (from each environmental group) were as follows: hydrological alterations, flow velocity and substrate (HYMO group), dissolved oxygen, nitrates and nitrites (physico-chemical parameters), zinc and nickel (metals), monobutyltin cation (organotins group), benzo(b)fluoranthene and benzo(k)fluoranthene (PAHs), and PCB 77 and 118 (PCBs).
After CCA analysis of the selected environmental variables and taxa data, we obtained insight into the major environmental gradients. The results regarding the relations of the HYMO group and the Oligochaeta assemblages are presented in Figure 4a. The negative correlation between hydrological alterations on the one hand, and flow velocity and substrate on the other (more prominent alterations, lower flow and finer substrate, and vice versa) should be noted. Hydrological alterations were identified as the most important environmental variable overall, particularly along the first CCA axis; it was the main explainable variation in the taxa matrix/distribution. The vast majority of taxa seem to be unresponsive to these factors (the large taxa group situated near the coordinate center of the CCA diagram). Species that tolerate faster flow and a larger substrate (with more detritus/organic components) are on the right side of the diagram: N. simplex, N. communis, T. ignotus, and P. vejdovskyi. On the other hand, R. limosela, Limnodrilus sp., L. claparedeanus, Stylodrilus sp., and B. sowerbyi are taxa associated with more prominent hydrological alterations. Analysis of the relation of oligochaetes assemblages and these factors revealed that species that exhibit significant differences were the most numerous, as expected.
(a)    Tables 1 and 2. Organotins, PAHs, and PCBs were tested to check whether their variability allowed for multivariate analyses and those with zero variance were excluded from further analyses. FS revealed that only one factor (monobutyltin cation) was important (p level 0.05) among organotins, thus in order to analyze these parameters in a multivariate space, CCA with all seven factors was used (Figure 5a). P. frici, P. blanci, U. uncinata, R. limosella, Enchytraeidae, and Lumbricidae were taxa that tolerated increased concentrations of monobutyltin cation. The highest concentrations of this pollutant were recorded during the JDS3 at the sampling sites from the Geisling power plant to Bratislava (sectors 3 and 4).
space defined by selected factors are shown in Figure 4c. High concentrations of these metals were recorded during the JDS2 in the majority of sample sites belonging to sectors 6-10 (values over 400 and 100 mg/kg, respectively). The highest values for zinc and nickel were obtained at site 51, the confluence of the Sava River (745 and 210 mg/kg, respectively). Zinc and nickel are negatively correlated with the first CCA axis (69% of explained variance). The distinctive sample on the diagram was collected at Upstream Cernavoda (sector 9). Oligochaeta taxa tolerant to increased concentrations of these metals are as follows: A. japonicus, P. volki, A. limnobius, N. simplex, N. communis, P. albicola, and P. isochaetus. Among them A. japonicus, A. limnobius, and P. volki tolerated nickel (CCA axis 2), while N. simplex, N. communis, and P. albicola tolerated zinc.
Organotins, PAHs, and PCBs were tested to check whether their variability allowed for multivariate analyses and those with zero variance were excluded from further analyses. FS revealed that only one factor (monobutyltin cation) was important (p level 0.05) among organotins, thus in order to analyze these parameters in a multivariate space, CCA with all seven factors was used (Figure 5a). P. frici, P. blanci, U. uncinata, R. limosella, Enchytraeidae, and Lumbricidae were taxa that tolerated increased concentrations of monobutyltin cation. The highest concentrations of this pollutant were recorded during the JDS3 at the sampling sites from the Geisling power plant to Bratislava (sectors 3 and 4).   Tables 1 and 2. Among PAHs, FS distinguished benzo(b)fluoranthene and benzo(k)fluoranthene as the most important environmental variables. The relations in the multivariate space among oligochaetes assemblages and selected variables are presented in Figure 5b. As it can be observed, these variables are negatively correlated with the CCA axes, particularly with the second axis. Taxa tolerant to increased concentrations of selected PAHs were as follows: S. ferox, A. pluriseta, N. alpina, and H. ventriculosa. Samples with higher concentrations of these PAHs are separated on the CCA diagram and mainly belong to the Danube sectors 2 and 3.
The FS of PCBs singled out PCB 77 and 118 as the most important PCBs. The connection between oligochaetes and these variables is shown in Figure 5c. P. volki is the only species showing variations of selected variables. Samples from Downstream Arges, Oltenita, and Upstream Olt (during JDS2) had the highest concentrations of these pollutants. In a sample from Oltenita, P. volki was the dominant member of the Oligochaeta community with the highest abundance. To better evaluate the correlation between other species and selected environmental variables, after excluding the species P. volki and the Oltenita sample, CCA with FS was performed again (Figure 5d). PCB 167 was added as the third most important PCB and it positively correlated with PCB 77 and 118, while all variables negatively correlated with the CCA axes. Several taxa from samples from different sectors (sites Downstream Pancevo, Stara Palanka-Ram, Sulina, Jochenstein, Klosterneuburg, Bratislava, and Braila) were found to be associated with increased concentrations of PCBs; these were P. albicola, Stylodrilus sp., Limnodrilus sp., A. limnobius, E. tetraedra, P. moldaviensis.

Discussion
The investigated sectors of the Danube exhibited high faunistic similarities due to the dominant influence of ubiquitous tubificins with wide ecological valences, such as Limnodrilus, Potamothrix, and Psamoryctides species. In regard to the higher level of sectioning (the large river stretches-Upper, Lower, and Middle Danube), separation of samples from the lower stretches of the Danube could be observed, which has been explained in detail by Atanacković et al. [16]. However, performed DA based on the Oligochaeta community showed that sectioning of the Danube according to Robert et al. [24] could not be observed. This occurrence indicates the high homogenization of the habitat and community as well. It is obvious that in a multistressed environment as it is in large rivers, with short environmental gradients and homogenous water quality along the entire river course, a few species with wide ecological valences could dominate along the entire stretch of the river, thus diminishing the potential of the entire community to react properly [34]; it was confirmed for the Chironomidae community in the Danube [34], and could be observed from the presented results for the Oligochaeta community. Tubificins, particularly L. hoffmeisteri, predominated by frequency and abundance in almost all investigated sectors. One of the main causes of biotic homogenization (related to Oligochaeta communities) of the entire river were the HYMO parameters. Actually, regarding the main HYMO factors, CCA showed high habitat variability within each individual sector, i.e., there are similar microhabitats in each sector as a result of hydromorphological alterations, changes of the water regime, and substrate structure, thus the different sectors show habitat similarities. In each sector, there are sites influenced by dams, with a backwater effect (a slowing down of the current and increased sedimentation), sites with bank modifications and with natural banks, sites with a high nutrient/organic load, so that each individual sector was characterized by the presence of all substrate types. These conditions led to the "potamalization" of upper and middle river stretches, followed by changes in faunal structure [35], transforming these sections of the river flow into reservoirs and causing a change from rhithral to potamal type of communities. The existence of similar types of microhabitats within each sector causes homogenization of the oligochaetes community of the entire Danube. This means that there is a similar oligochaetes community with dominance of tubificins in each sector along the entire Danube. The slowing down of the current which has a backwater effect, increased sedimentation, rising water temperature and accumulation of organic matter caused by dams in upper and middle river stretches have led to an increase in the number of limnorheophilic species (characteristic for the lower part of the River) instead of reophilic species (characteristic for upper and middle parts of the River). Dumnicka [11] pointed out that damming had a marked influence on oligochaetes fauna. In addition, the predominance of potammal species in the Danube, observed by Moog [35], is in accordance with our results which confirmed that the high participation of tubificins that was recorded in each sector reveals the existence of well-developed muddy zones.
Many authors have pointed out that the bottom structure, which is closely related to the river current, is the most important factor for Oligochaeta fauna in the Danube [12,36]. This is confirmed by our analyses, and to which we can add one more factor-hydrological alterations. Most of the recorded species prefer fine, muddy, and sandy substrates, with slow currents, while only three species, N. simplex, N. communis, and T. ignotus, can be distinguished by their preference for large substrates and faster currents. The importance of substrate type, urban pollution, and hydromorphological alterations (changes in current velocity, riverbed structure, etc.) in the Slovakian and Hungarian Danube stretch was confirmed [12,37,38].
Considering that the entire course of the Danube has been altered under significant HYMO pressure, and that only 6.45% of the investigated sites had natural banks, it can be concluded that a homogenous environment of the entire river stretch has been created (the upper river stretches no longer differ from the middle and lower river parts), enabling the spread of similar ecological groups of oligochaetes species in each individual sector, thus making the oligochaetes community of the Danube uniform along the entire watercourse. Species with wider ecological valences, such as tubificins, predominate, so no obvious distinction between defined sectors regarding species composition of oligochaetes assemblages could be observed.
As the Danube flows through a relatively densely populated area of Europe with several large cities and extensive/intensive agriculture, a high input of organic load from different sources (communal waters, agriculture etc.) is evident. Tubificid species are expected to be recorded with higher abundance and frequencies at sites with higher nutrient values. Indeed, such sampling sites were the most numerous in our investigation, and are characterized by muddy substrates, slow currents, and increased sedimentation. Our findings are in accordance with these observations, with a positive correlation of tubificids (Limnodrilus genera) with nitrates. Moreover, turbificid species T. tubifex and B. sowerbyi are positively correlated with nitrites.
The results of the influence of metals on the distribution of Oligochaeta revealed that zinc and nickel were the most influential factors. Species that displayed a positive correlation with metal concentrations in the sediment, as expected belong to the Tubificinae (A. japonicus, A. limnobius, P. isochaetus, and P. albicola), with Naidinae (N. simplex and N. communis), and Propapidae (P. volki), exhibiting greater tolerance. No oligochaetes species were found to be sensitive to increased metal concentrations. Ecotoxicological significance of any heavy metal does not solely depend on its concentrations, but also on the form that it is present in the environment [21] and it is not surprising that no oligochaetes species were found to be sensitive to increased metal concentrations.
Several taxa, such as P. volki, P. albicola, Stylodrilus sp., correlated with elevated concentrations of organic pollutants such as PAHs and PCBs. In the case of PCBs, this could be due to the higher tolerance and more effective system of excretion of the taxa; on the other hand, it could be because of the lower degree of bioaccumulation due to the specificity of the habitats (organic matter, depth, etc.).
Aquatic oligochaetes have a long history of use in aquatic pollution assessments [22], though they rarely were used directly in ecological risk assessment due to complicated species identification and traditional linking to organic pollution only. Moreover, the whole group of Oligocaheta has been considered 'pollution tolerant' where this perception of 'tolerance' has been incorrectly projected to chemical pollution [22].
Water quality assessment as a tool in water management should deal with at least the major environmental conditions and certainly not be restricted to the saprobic part [39]. In this respect, we confirmed the significance of oligochaetes as indicators of hydromorphological changes, not just as pollution tolerant species, as was thought in the past.
According to our results, physico-chemical and chemical determinants in water and sediment (organic pollution and nutrient load) are not the main factors that shape oligochaetes communities and influence their distribution. The complexity of this issue is pronounced considering the structure of microhabitats caused by hydromorphological changes. Our results indicate that Oligochaeta could be used for the identification of the level of hydromorphological degradation in large rivers (homogenization), rather than for stressors classified as nutrient, heavy metal, and organic pollutants. The presence of metals in tissues can serve as an indicator of contamination of an organism [23] and these data provide additional evidence in risk assessment of the environment, but there is no standard response of specimens within a taxon to increasing concentrations of metals. Furthermore, it would be preferable to assess the effects of different pollutants found in aquatic environments to the level of DNA damage of Oligochaeta species like it was done for L. udekemianus [40] indicating that this species is suitable model organism in ecogenotoxicology.
The traditional distinction between Danube sectors is not applicable if we consider Oligochaeta communities. The main reasons causing biotic homogenization are hydromorphological alterations that are represented by similar changes in flow velocity and substrates along its course. As Oligochaeta are fully aquatic organisms with a low mobility, they are a good indicator of changes in aquatic habitats, which has been confirmed in our study. Moreover, if we consider Oligochaeta communities, this large river could be considered as "one habitat-one community river". Smutz and Moog [41] found that hydromorphological alterations (particularly damming) could have negative effects on aquatic communities of the hyporheic river zone resulting in impoverished fauna, dominated by few species and reducing the ecological status of this lacustrine sections, in reservoirs downstream dams. To get a more complete picture regarding aquatic biota and its relations to environmental pressures along the Danube, incorporating additional biotic components is needed.