Zooplankton Communities of the Middle River Part of the Cheboksary Reservoir and Factors Influencing Their Species Structure

Modern methodological approaches were applied to analyze the zooplankton community spa- tial distribution (with the example of the middle river part of the Cheboksary Reservoir). Zooplankton communities were sampled in the middle river part of the Cheboksary Reservoir (from the city of Nizhny Novgorod to Vasilsursk town) in the summer low-water period in 2018. The boundaries between the communities in the Cheboksary Reservoir were gradually changing during the history of the Cheboksary Reservoir from the time of its construction to present. In the middle river part of the Cheboksary Reservoir there are two distinct spatially stable zooplankton commu- nities associated to the Oka and Volga streams. The distinction between these two zooplankton communities was demonstrated by hierarchical cluster analysis. Redundancy analysis has shown that chlorophyll- a and pH were the main factors influencing the specific zooplankton structure. Chlorophyll- a concentration reflects the meso-scale heterogeneity of the horizontal phytoplankton distribution and hence the distribution of the zooplankton’s food sources. The relation of zooplankton to pH level reflects the high sensitivity of the species of the genus Brachionus Pallas, 1766 to high acidity. The influence of pH as an environmental factor was less evident. However, this variable is well known as one of the leading factors determining the structure of zooplankton communities. Its role in zooplankton community assembly of lowland reservoirs deserves further investigation.


INTRODUCTION
There is a great interest in the problem of identification of spatial distribution of zooplankton communities and assembly of their species structure under the impact of various natural and anthropogenic factors (Leitao et al., 2006;Sokolova, 2012;Bolotov et al., 2013;Lazareva, Sokolova, 2015;Presnova, Khulapova, 2015;Joniak, Kuczynska-Kippen, 2016;Shurganova et al., 2018). The creation of reservoirs in the river beds transforms the original river ecosystem into a lentic ecosystem (Simões et al., 2015) and modifies the functioning of hydrobiocenoses. The dynamics of spatial distribution of zooplankton communities during long-term succession and changes of influence of various environmental factors is of particular interest in hydroecology (Shurganova, 2007). Species structure and abundance of zooplankton in reservoirs depends on a lot of environmental factors (Kiselev, 1969;Hayrapetyan et al., 2016) including food resources which are regarded by some authors as a leading ecological factors in water communities (Leibold et al., 1997;Dai et al., 2014).
Cheboksary Reservoir is the fifth stage in the cascade of reservoirs on river Volga. It was filled in 1981. Its area is 1080 km 2 , maximal depth is 21 m (mean depth 5.8 m), water exchange coefficient is 19.8 per year (Shurganova, 2007), trophic status is stable eutrophic (Korneva, 2015). Cheboksary Reservoir is formed by two streams coming from the Gorky Reservoir and Oka River. These water masses differ from each other considerably in physical and chemical properties (Shurganova, 2007;Shurganova et al., 2018). Studies of the zooplankton of the Cheboksary Reservoir have been conducted since its construction (Kuznecova et al., 1991;Shurganova et al., 2003Shurganova et al., , 2014Shurganova et al., , 2018Shurganova, 2007;Shurganova, Cherepennikov, 2010). However, the current spatial distribution and species structure of zooplankton communities in the middle river part of the Cheboksary Reservoir requires additional investigation because zooplankton communities change their structure and boundaries during succession (Shurganova, 2007). That's why the study of the dynamics of the spatial structure of hydrobiocenoses and evaluation of the impact of environmental factors on their assembly and functioning at the modern successional stage is an actual task.
The goal of the present work is to characterize the modern spatial distribution of zooplankton communities in the middle river part of Cheboksary Reservoir and to study the dependence of zooplankton communities on environmental factors.

MATERIAL AND METHODS
Zooplankton communities were sampled in the summer low-water period (July) in 2018 in the middle river part of the Cheboksary Reservoir (from the city of Nizhny Novgorod to Vasilsursk) (Fig. 1). Samples were taken with plankton net (70 μm nylon sieve) and fixed with 4% formalin solution. Sample processing was conducted with standard protocol (Metodicheskie rekomendacii…, 1982). Zooplankton species were identified with keys and handbooks (Kutikova, 1970;Korovchinsky, 2004Korovchinsky, , 2018Key to Zoo-plankton…, 2010).
Water conductivity was measured at each sampling site with a YSI Pro30 (YSI Incorporated, USA) conductometer. pH value was measured with a YSI Pro10 pH meter (YSI Incorporated, USA). Water samples were taken at each site for the laboratory in-vestigation. HCO 3 concentration was determined by titration with a solution of hydrochloric acid in the presence of indicators (as the difference between total and carbonate alkalinity) in accordance with GOST 31957-2012GOST 31957- (2014 concentration of SO 4 and sodium was determined by ion chromatography. The photosynthetic pigments content (chlorophyll-a and -b) was determined according to GOST 17. 1.4.02-90 (1991) by spectrophotometry of the acetone extract from the precipitate obtained after filtering the sample through a membrane filter. Water turbidity was determined by an optical method with a HACH 2100 turbidimeter (Belozerova, Chalov, 2013) and was measured in nephelometric turbidity units (Hach Company, USA) (Grayson et al., 1996).
Zooplankton species were considered as dominant according to Kownacki index (Bakanov, 1987). Multivariate vector analysis was applied to classification if zooplankton samples (Shurganova et al., 2003;Shurganova, 2007). It is a version of hierarchical cluster analysis based on angle (cosine) between samples in multivariate abundance space as a measure of their dissimilarity. Clustering was performed with the average linkage algorithm. Silhouette width analysis and Mantel correlation coefficient between distance matrix and binary matrix representing partitions were applied to determine the optimal number of clusters (Yakimov et al., 2016). Redundancy analysis (RDA) was applied to consider the association of zooplankton community structure and environmental factors (Shitikov, Rozenberg, 2013;Legendre P., Legendre L., 2012) and to display it with ordination diagram. All calculations were performed in R (R Core Team, 2015).
Redundancy analysis was performed to consider the relation of the structure of zooplankton communities to the following environmental factors: water conductivity, pH, turbidity,  (Table 3). At the first step, separate models for each factor were considered. Statistical analysis has shown that all factors have a significant influence to zooplankton community structure (p < 0.05; Table 4). Next, full RDA model was constructed which included all factors as predictors. This model is characterized by exceedingly high correlation between predictors (variance inflation factors VIF > 20 for 5 out of 8 predictors). Therefore, our final analysis is based on parsimonious model which was constructed with a stepwise procedure of forward selection of explanatory variables. Parsimonious model included chlorophyll-a and pH as predictors. It significantly explained 35.84% of a total variation of zooplankton community structure (p < 0.001).  Fig. 4. Two groups of sites may be distinguished along the first axis which explain 26.39 % (р < 0.001) of the total variance. These groups correspond to zooplankton communities delineated on the basis of cluster analysis (Fig. 3, Oka and Volga streams).
Analysis of ordination results shows that rheophilic rotifera of genus Brachionus Pallas, 1766 are associated with waters of Oka stream which have high pH and high concentration of chlorophyll-a. On the other hand, limnophilous species such as D. galeata and P. truncatum as well as naupliar and copepodit stages of Cyclopoida are associated with Volga stream (Fig. 3). Sites of the Oka stream close to confluence point are characterized by the highest values of the first ordination axis. The more the distance from the confluence (and the more the number of the site, Fig. 1) the less the value of the first ordination axis. Thus, zooplankton composition of the Oka stream continually approaches zooplankton composition of the Volga stream. This tendency can be explained by the gradual alignment of environmental variables.
The two streams differed mostly in chloro- phyll-a concentration. But the difference gradually decreased downstream the middle river part of the Cheboksary Reservoir (Fig. 4). This suggests the homogenization of water masses and, as a result, the disappearance of the difference between the species structure of the zooplankton communities of the Oka and Volga streams downstream the Kremenki village (site number 11-12). Heterogeneity of other environmental variables (Table 3) also play a role in the formation of two separate zooplankton communities in the middle river part of the Cheboksary Reservoir (Oka and Volga streams). These communities differ in species composition until the point of water mass homogenization (close to Kremenki village).

DISCUSSION
Chlorophyll-a and pH are the main environmental variables driving variability of zooplankton communities species structure in the middle river part of the Cheboksary Reservoir. Chlorophyll-a concentration characterizes production processes in water bodies. Maximal zooplankton abundance was found in the sites with maximal chlorophyll-a concentration (Oka stream, Tables 1 and 3). Zooplankton species structure of the Oka stream is dominated by rotifers of genus Brachionus Pallas, 1766. These rotifers are bacteriophages and herbivores and they cannot feed on particles more than 18 μm in size (Monakov, 1998). It is known that phytoplankton of the mouth part of the Oka River is represented mostly by small-cell cyanobacteria and diatoms (Pautova et al., 2013) which are the optimal food source for rotifers generally and for Brachionus Pallas, 1766 species specifically.
Minimal chlorophyll-a values were found in the sites of Volga stream where zooplankton abundance was also minimal (tables 1 and 3). Species structure is dominated by large cladoceran species D. galeata here which feeds on diatoms and cyanobacteria as well as on organic debris with particle size less than 60 μm. D. galeata abundance is negatively related to chlorophyll-a (Fig. 4). It can be explained by the fact that most chlorophyll-a is generated by cyanobacteria  which can harm the filtering apparatus of D. galeata and cause their death. However, we shall note that cladoceran D. orghidani have a remote position on ordination diagram (Fig. 4) relatively to other species. This species is characterized by high resistance to clogging the filtering apparatus by large colonies of cyanobacteria which provides it with competitive advantage in the case of increasing intensity of water «blooming» (Lazareva, Bolotov, 2013). Relation of chlorophyll-a and zooplankton abundance is well-known in hydrobiology  (Orsi, Mecum, 1986). The difference in chlorophyll-a between Oka and Volga streams reflects mesoscale heterogeneity of the distribution of phytoplankton in the reservoir (Mineeva, Abramova, 2009) which is the main food source and one of the basic factors determining the species structure of zooplankton communities in the middle river part of the Cheboksary Reservoir. The second significant environmental variable related to zooplankton species structure is pH. Its change is accompanied usually by a change in the structure and functioning of zooplankton communities. The sudden changes in pH can lead to changes in quantitative and qualitative development of species vulnerable to high acidity. These species can fall out of community. Cladoceran species are especially vulnerable to such effects (Sandoy, Nilssen, 1986;Korhola, 1992;Frolova et al., 2013). In our study species of genus Brachionus Pallas, 1766 react positively to increase in pH. On the other hand, some cladoceran species (particularly D. galeata and D. orghidani) are negatively related to pH. Similar effects were found in other studies (Kudrin, 2016;Fetter, Yermolaeva, 2018). We hypothesize that such a reaction of Brachionus Pallas, 1766 and some cladocerans to pH is a general regularity deserving further investigation.

CONCLUSION
We applied modern multivariate techniques and showed that in the middle river part of the Cheboksary Reservoir there are two distinct spatially stable zooplankton communities associated to Oka and Volga streams. Zooplankton species structure was influenced by chlorophyll-a and pH mostly. Oka stream community was dominated by euplanctonic filter-feeding species B. calyciflorus and B. angularis, whereas Volga stream community was dominated by naupliar stages of Cyclopoida and by large euplanctonic primary filtrator D. galeata.
Boundaries between communities of Oka and Volga streams has been changing during the history of Cheboksary Reservoir from the moment of its construction to present. These changes are studied thoroughly by Shurganova and colleagues (Shurganova, 2007). In the studies performed in the first decade of the XXI century it was shown that Oka stream community occupied the right-bank side of the middle river part of Cheboksary Reservoir reaching the sampling site downstream the Kstovo town ( Fig. 1, site 6). At present (2018, this study) boundary between Oka and Volga stream communities moved downstream. Oka stream community reaches the site near Kremenki village (Fig. 1, site 12). Such changes in the boundaries between zooplankton communities may reflect changes in the hydrological regime of the reservoir.
Redundancy analysis of the influence of environmental variables on zooplankton community structure has shown that all variables were significant in single-variable analysis. However, only chlorophyll-a and pH were included in the parsimonious model due to high correlation among environmental variables. This model explained 35.84 % (p < 0.001) of variation in zooplankton species structure. The difference in chlorophyll-a concentration between Oka and Volga streams reflect mesoscale heterogeneity of phytoplankton distribution and, consequently, the basic difference in food sources for zooplankton communities associated to the streams. The influence of pH as environmental factor was less evident. However, this variable is well known as one of the leading fac-tors determining the structure of zooplankton communities (Sandoy, Nilssen, 1986;Nilssen, Sandoy, 1990;Korhola, 1992;Frolova et al., 2013;Fetter, Yermolaeva, 2018). Its role in zooplankton community assembly of lowland reservoirs deserves further investigation.