Long-term dynamics of photosynthetic pigments in plankton of a large plains reservoir

Long-term observations are necessary to analyze and predict changes occurring in biological communities under global and regional climatic changes. The data on seasonal and long-term dynamics of chlorophyll in plankton of the Rybinsk Reservoir (Upper Volga, Russia) obtained in May – October 2009–2019 at six standard stations using the fluorescence method are presented. In the years with contrasting hydroclimatic conditions and water regime, the chlorophyll content varied from 1–3 to >100 μg/L. The significant variability of mean seasonal values (7.9 ± 0.5 μg/L in 2009 to 27.6 ± 1.7 μg/L in 2013 with variation coefficients of 52–134%) indicates the low resistance of the community. The total chlorophyll content is associated with the development of three main phytoplankton divisions i.e., diatoms, cyanoprokaryots, and green algae. The trophic status of the reservoir was characterized as mesotrophic in 2009 and 2017, eutrophic in 2011–2014, and moderately eutrophic in other years. In the long-term seasonal cycle of phytoplankton, there are five periods with stable temperature conditions and transparency, but variable chlorophyll content. A moderate positive relationship was found between the seasonal dynamics of chlorophyll and water temperature, but a moderate negative relationship with transparency. The priority factors regulating the long-term dynamics of chlorophyll include the NAO indices, Wolf numbers, temperature, and underwater light conditions, as well as the inflow volume and water level. Water regime limits the development of phytoplankton.


Introduction
It is well known that global climatic changes that have a significant impact on the structure and dynamics of biological communities of aquatic ecosystems (Adrian et al., 2009;Bertani et al., 2016;Özkan et al., 2016) continue to take place the beginning of the XXI century (Vtoroy, 2014). An increase in temperature is considered as a eutrophication factor that changes the availability of nutrients, promotes increase in the internal phosphorus load, as well as a more abundant and prolonged vegetation of cyanoprokaryotes (blue-green algae) (Jeppesen et al., 2005;Winder & Hunter, 2008). To analyze and predict changes occurring in biological communities under these conditions, long-term observations which are carried out in many water bodies of the world (Ruggiu et al., 1998;Kangur et al., 2002;Chen et al., 2003;Babanazarova & Lyashenko, 2007;Canfield et al., 2018;Lamont et al., 2019;Gao et al., 2020) are required. Photosynthetic pigments, which are the universal ecological and physiological characteristics of the development and photosynthetic activity of algae, as well as the ecological state of water bodies, are widely used in the study of the autotrophic community in aquatic ecosystems. The main plant pigment, chlorophyll а (CHL), is a physiological marker that effectively indicates changes in the external environment.
The study of plant pigments in the water of the Rybinsk Reservoir has been carried out in IBIW RAS since the middle of the 20th century using the standard spectrophotometric method (SCOR-UNESCO, 1966). Longterm data have made it possible to study the seasonal and interannual dynamics of pigments, its relationship with regional and global environmental factors (Pautova & Rosenberg, 1999;Kopylov, 2001;Sygareva et al., 2016;Lazareva, 2018). In 2009, we began additionally use fluorescent diagnostics, which allowed us to determine the chlorophyll content directly in natural water, obtain a number of phytoplankton characteristics without damaging its integrity, and quickly analyze a large data set (Mineeva, 2016;Mineeva & Semadeni, 2020). Differentiated determination of chlorophyll in cyanoprokaryotes, diatoms and green algae made it possible to obtain new data on the seasonal and interannual dynamics of large taxonomic groups of phytoplankton, their contribution to the total chlorophyll content and their relationship with environmental factors in the years with different hydroclimatic conditions. The objective of this work was a comparative analysis of these data, which continue and supplement a long-term series of observations.

Materials and methods
The data were collected at six standard stations in the Rybinsk Reservoir ( Fig. 1) during the growing seasons of 2009-2019 with a frequency of 1-2 times a month.

Fig. 1. Location of observation stations in the Rybinsk Reservoir
The samples were taken with a 1 m Elgmork bathometer from the entire water column (0 m -bottom). During the study period, more than 500 observations were made at the reservoir stations and more than 940 sam-ples were taken for the analysis of pigments. Chlorophyll а was determined by the fluorescence method, which makes it possible to estimate the total amount of pigment (ƩCHL) by its content in the main representatives of freshwater phytoplankton -cyanoprokaryotes, diatoms, and green algae (CHL Cyan , CHL Bac , CHL Chl , respectively) (Gold et al., 1986). Chlorophyll fluorescence was measured on a stationary PFL-3004 fluorimeter made at the Krasnoyarsk State University. The analysis procedure was described by us earlier (Mineeva, 2016). To assess the state of the community, the plasticity index was used (Shashulovsky & Mosiyash, 2010), which is numerically equal to the average value of the modules of ΣCHL correlation coefficients with the environmental parameters. Water temperature, transparency, and colour were measured at each station. Reservoir level and inflow data in Table 1 were taken from the website www.rushydro.ru/hydrology/informer/?date; the data on air temperature and rainfall -from the weather archive https://rp5.ru; NAO index and Wolf numbers -from www.cpc.ncep.noaa.gov/data/teledoc/nao.shtml and www.sws.bom.gov.au/Solar/1/6. The average long-term climatic, hydrophysical, and hydrochemical characteristics of the Rybinsk Reservoir are given based on the materials of monographs (Kopylov, 2001;Lazareva, 2018).
Rybinsk Reservoir (the third stage of the Volga River cascade), located in the southern taiga subzone (58°00'-59°05' N, 37°28'-39°00' E), is a large relatively shallow water body of slow water exchange (with surface area of 4500 km 2 , average depth of 5.6 m, and average coefficient of water exchange equaling 1.9 year -1 ). The area of the reservoir is divided into four heterogeneous areas (reaches) occupied by water masses with specific hydrophysical and hydrochemical characteristics. Three reaches are located along the flooded channels of the main tributaries, the Volga, Mologa and Sheksna rivers. River waters are gradually transformed into the water mass of the reservoir proper, which occupies a vast lake-like central part, the Main reach (Kuzin, 1972;Lazareva, 2018).
Standard software packages for a personal computer Statistic10 (StstSoft Inc., USA) were used for statistical data analysis. The data are given as mean value with standard error (x ± SE). To determine the relationships between chlorophyll content and environmental parameters we used Spearman rank coefficient (r S ).

Results
The development of phytoplankton largely depends on hydrometeorological conditions. Although the observation years in the long-term series were characterized as warm, they differed significantly in weather patterns. The average annual air temperature exceeded the climatic norm of 3.7 °C. With the average temperature for May -October 2000-2019 equaling 13.6 °C, the growing season was the warmest in 2010 and the coldest in 2017 (Table 1). The heating of the water column is related to the air temperature with coefficient of determination R 2 = 0.74. The water temperature in May -  Table 2). The annual amount of precipitation generally corresponded to the climatic norm (436-887 mm, average 655 mm). The average amount of precipitations for May -October 2000-2019 was 395 mm, 72% and 80% of this amount fell in 2013 and 2014, 140% in 2012. According to the water content, 10 out of 11 years of observations referred to the highwater period with the inflow volume significantly exceeding the norm in 2017. Only in the extremely low-water year of 2014, did the inflow sharply decrease. In accordance with the inflow volume, the reservoir level was minimal in 2014 and exceeded the normal level of 101 m BS in 2009-2012, 2016, 2017. Judging by the dynamics of Wolf numbers, the observations covered an 11-year cycle of solar activity ( Table 1). The transparency and colour of the water were characterized by values typical for the reservoir (Table 2). Over the 11-year observation period, there occurred a moderate linear increase in the mean annual air temperature (R 2 = 0.21), winter and annual NAO indices (R 2 = 0.31 and 0.25), but a decrease in the average seasonal and especially summer water temperature (R 2 = 0.27 and 0.56).
The ƩCHL content varied over a wide range. The minimum values >1-3 μg/L were shown by ~16% of the total samples, the ranges of 3-10 μg/L and 10-30 μg/L were 36% and 37%, the higher concentrations were ~10% and only 1.5% of them exceeded 60 μg/L (Fig. 2). The seasonal dynamics of ƩCHL, that is dynamics of phytoplankton biomass, were characterized by two or three peaks with differences in the timing of the onset, the duration, and the ratio of values in different years (Fig. 3).
The spring peak with average ƩCHL concentrations of 13-32 μg/L was recorded at the water temperatures of 6-9 °С to 11-15 °С in May   left axis -CHL Cyan (1) and CHL Bac (2), right axis -CHL Chl (3), mean values The analysis of 11-year data revealed five periods in the seasonal cycle of ƩCHL. Each period is characterized by uniform temperature conditions and transparency. The average values of temperature and transparency for each period are equal to their medians. The ƩCHL variability is minimal in early summer with the seasonal change of communities but is significantly higher during the spring and summer phytoplankton maxima; the arithmetic averages in all cases exceed the medians by 25-30%. The development of phytoplankton in spring and early summer is accompanied by a decrease in the content of mineral nitrogen, while changes in mineral phosphorus and total forms of biogenic elements during the entire growing season are small (Table 3). The fluorescence method makes it possible to estimate chlorophyll content of the three main divisions of algae (Table 4). In most cases concentration of ƩCHL was determined by the amount of CHL Cyan ; in 2010, 2016, 2017, 2019 the dominant influence belonged to CHL Bac , in 2017-2019 the role of CHL Chl was noticeable as well, and in 2011 ƩCHL equally depended on all three components (Table 5).

Table 4
Chlorophyll total amount and chlorophyll content in the main divisions of algae in plankton of the Rybinsk Reservoir in the years of study (x ± SE; in brackets -coefficient of variation (%) / median) The amount of CHL Cyan and CHL Bac did not correlate. With a wide range of values, the minimum chlorophyll content of each algal division did not exceed 1 μg/L. The predominant values of CHL Bac and CHL Cyan are limited to 10 µg/L (Fig. 2). In different years, the highest concentration of CHL Bac reached 18.4-92 μg/L, CHL Cyan 25.1-130 μg/L, CHL Chl 1-4 μg/L and 16 μg/L in 2011. A noticeable amount of CHL Cyan was present in the reservoir throughout the growing season, with the highest occurring during the summer phytoplankton maximum. The seasonal dynamics of CHL Bac was characterized by a spring peak in all the years, and by the second autumn peak in 2014. In 2010In , 2011In , 2013In , 2015In , 2017, high concentrations of CHL Bac were also noted in summer. At low CHL Chl concentrations, its seasonal dynamics are not pronounced (Fig. 3). CHL Bac and CHL Cyan made the main contribution to ƩCHL, their ratio varied in different years. The values averaged over the growing season show that CHL Cyan more often prevailed in the ƩCHL pool, particularly in 2013 and 2018. The difference between the shares of CHL Cyan and CHL Bac was not large in 2010 and 2015-2017, the percentage of CHL Chl increased in 2017-2019 (Table 4).   (Table 4). For all of the above indicators, the long-term dynamics of their mean values for the growing season is not linear and is approximated by a polynomial relationship (Table 6). The pigment characteristics during the growing season of each year varied greatly. The coefficients of variation were <100% in six of 11 cases for ƩCHL, in five cases for CHL Cyan , in three cases for CHL Bac and CHL Chl . Only in 2012 and 2015,was moderate variability of ƩCHL noted, as well as lower variability of CHL Cyan and CHL Bac against the general background. For all data, the average values exceeded their medians: 1.2-1.9 times for ƩCHL and CHL Cyan , 1.3 to over 2.0 times for CHL Bac and CHL Chl . Only in 2012 and 2015 are both characteristics equal for ƩCHL (Table 4).
In accordance with the average ƩCHL content, the trophic status of the reservoir is assessed as mesotrophic in 2009 and 2017, eutrophic in 2011-2014, and moderately eutrophic in other years. CHL Cyan and CHL Bac change in parallel with ƩCHL (R 2 = 0.92 and 0.31), while such a relationship was not revealed for CHL Chl . When ranking all parameters by ƩCHL, their regular increase is observed with an increase in trophic level. The relative content of CHL Cyan increases with decrease in the proportion of CHL Bac and CHL Chl . The coefficients of variation show that the variability of ƩCHL within the trophic categories ranges from insignificant in mesotrophic and eutrophic waters to moderate in oligotrophic waters; CHL Cyan variability is moderate, CHL Bac variability is higher, and CHL Chl variability is maximal (Table 7).   (Fig. 4).
In the long-term aspect, the relationship of chlorophyll with environmental factors is more pronounced. For the mean concentrations of ƩCHL, a positive dependence on temperature, as well as on Wolf numbers, the index of solar activity, was revealed, whereas the water content (rainfall, inflow, water level) are the limiting factor. The ƩCHL content is negatively related to the transparency and colour of the water. A moderate positive correlation was obtained between ƩCHL, CHL Cyan and the annual NAO index (Table 8). The use of stepwise multiple regression made it possible to identify priority factors of long-term development of phytoplankton. Of all the parameters shown in Table 8, these are transparency, inflow volume, air and water temperature, reservoir level, winter and annual NAO indices (R 2 = 0.99), from climatic parameters, these are the Wolf numbers, the mean seasonal air temperature and the annual NAO index (R 2 = 0.84); from the regional ones, these are the transparency and the volume of inflow (R 2 = 0.83).

Discussion
The development of phytoplankton in the Rybinsk Reservoir is characterized by significant seasonal and interannual variations, and the ƩCHL content varies within a wide range. The main part of ƩCHL is accounted for by cyanoprokaryotes and diatoms, which corresponds to the taxonomic composition of algocenoses (Korneva, 2015). A significant amount of CHL Cyan is present during most of the growing season due to an increase in the period of development of cyanoprokaryotes under climate warming (Jeppesen et al., 2005). The low content of CHL Chl is consistent with the low amount of CHL b determined by spectrophotometric (Minee-va, 2004) and chromatographic (Breton et al., 2000) methods. The increase in the contribution of CHL Chl to ƩCHL in recent years is a sign of an increase in the abundance of green algae, which contain CHL b and periodically enter into the dominant complex of phytoplankton (Korneva, 2015). , TP, BOD, COD, Соrg, pH, 1-9, respectively; hydrochemical parameters are given in Bikbulatov et al. (2011) Seasonal dynamics, occurence rate of ƩCHL, its maximum and average values, the contribution of the main divisions of algae to the ƩCHL varied over the11-year period depending on temperature conditions and water content (Mineeva, 2016;Mineeva & Semadeni, 2020). The variability of ƩCHL increased from moderate in 2012 and 2015 to strong and very strong in other years. The coefficients of variation of CHL Cyan , CHL Bac , and CHL Chl are higher than ƩCHL, which corresponds to the idea of a lesser variability of the total as compared to the variability of its constituent parts (Alimov, 1999). The variability of parameters increased under the influence of uncontrolled or unaccounted for factors. The latter was relevant for reservoirs where the development of biological communities was affected not only by local weather conditions and climate, but also by the operating mode of hydraulic engineering structures. Fluctuations of the parameters relative to their average reflect the stability of the ecosystem, which decreases with an increase in the range of fluctuations. The ecosystem of a large artificial moderately eutrophic or eutrophic water body should have low stability, but significant endurance, since it is adapted to seasonal and interannual fluctuations in external conditions (Alimov, 2001). The plasticity index, which is proposed to be used as a measure of stability (Shashulovsky & Mosiyash, 2010) varied within small limits and was low over the 11-year period. Its minimum value was obtained in 2017 with an extreme inflow volume, a high level and increased colour, while the maximum in 2015 with a small seasonal variability of ƩCHL.
The trophic status of the reservoir varied from mesotrophic to moderately eutrophic and eutrophic. Interannual fluctuations in trophic state depending on hydroclimatic factors have been identified for water bodies in different regions (Ruggiu et al., 1998;Kangur et al., 2002;Babanazarova & Lyashenko, 2007). With the growth of trophic state, the role of the main divisions of algae changed: the absolute concentrations of CHL Bac , CHL Chl , and CHL Cyan increased, and the relative amount of CHL Bac and CHL Chl decreased against the background of the growing proportion of CHL Cyan , which is typical for eutrophic conditions. Seasonal development of plankton is an annually repeating process influenced by external factors and internal interactions (Reynolds, 2006). The seasonal variation of chlorophyll in the Rybinsk Reservoir corresponds to the classical model (Sommer et al., 2012) and is characterized by two or three peaks. The timing of their onset, duration, and ratio of values vary in different years. A short spring maximum of ƩCHL with a predominance of CHL Bac was recorded in early May -early June in 2011, 2016. Apparently, in other years, it ended before the start of observations due to the earlier opening of the reservoir (Lazareva, 2018), which was a trigger of vegetation in spring forms (Reynolds, 2006). The summer peak, formed by cyanoprokaryotes and diatoms (Korneva, 2015), was insignificant in 2009 and 2017, when the state of the ecosystem was characterized as mesotrophic. The first year, which was not distinguished against the general background by either the temperature regime or the water content conditions, ended the cycle of the decline in the phytoplankton productivity in the reservoir (Lazareva, 2018). Chlorophyll content in the Rybinsk Reservoir, as well as in other reservoirs of the Volga River (Kopylov et al., 2012), was not too large in 2010. However, the conditions of the abnormally hot summer in 2010 were the driving force for the subsequent growth of trophic state and the formation of a prolonged summer phytoplankton maximum with high concentrations of ƩCHL and CHL Cyan in 2011-2016. Later, in 2017, lower temperatures combined with extreme inflow volumes, limited the mass vegetation of cyanoprokaryotes and the development of the summer maximum.
The autumn peak occurred during a favourable combination of light conditions and supply of the cells with mineral nutrients (Sommer et al., 2012). During the study period, autumn peak was observed in 2014 only, and earlier it was observed in the years with predominance of sunny weather, which ensures the penetration of a sufficient amount of light energy into the water column (Mineeva, 2004). In spring and autumn, with mixing of the water column, which contributes to the maintenance of cells in the suspended state and provides the uptake of nutrients (Yang et al., 2016), diatom peaks are observed in many water bodies of the temperate zone (Reynolds, 2006;Sommer et al., 2012).
Despite the interannual peculiarities of seasonal dynamics and a wide range of values, five stages were distinguished in the seasonal cycle of ƩCHL in the reservoir. Each stage was characterized by homogeneous temperature and transparency, indicating the adaptability of communities to the specific conditions of a certain season. The ƩCHL content within each stage was variable, the average concentrations were higher than their medians, which confirms the complex unaccounted for or uncontrolled external impact on the biota. In large, shallow water bodies that are a dynamic environment, the course of the seasonal succession of phytoplankton is subject to frequent disturbing external effects (Honti et al., 2007, Sommer et al., 2012Yang et al., 2016). Nonlinear dynamics of average chlorophyll concentrations demonstrate the community's response to changing external conditions and confirms the conclusion about the cyclical nature of these changes. The maximum development of algae is observed in dry years with calm weather, increased insolation and water temperature, and the minimum is observed under opposite conditions (Pyrina, 2000;Mineeva 2004;Pyrina et al., 2006). Depending on the characteristics of the water body, the most important factors in the development of phytoplankton can be physical, chemical, or biotic (Chen et al., 2003;Reynolds, 2006;Yang et al., 2016). During the years of the study, we observed a negative correlation of the mean seasonal concentrations of ƩCHL, CHL Cyan , and CHL Bac with the water content parameters (i.e., inflow volume, reservoir level, precipitation amount) limiting the development of phytoplankton. Similar relationships were found for the longterm dynamics of phytoplankton biomass (Korneva, 2015). The average values of ƩCHL and CHL Cyan moderately depend on the water temperature, while this effect was not significant for CHL Bac .
In the presence of weather anomalies, close attention is paid to the temperature factor in the study of aquatic biota (Jeppesen et al., 2005). Temperature does not limit the growth of diatoms, but affects the development of cyanoprokaryotes (Chen et al., 2003), for which the lower temperature limit is 12-16 °C, and the optimal one is 20 °C and higher (Tryfon & Moustaka-Gouni, 1997;Tan, 2011). The broad temperature range (5-20 °C), favourable for the diatoms of the Rybinsk Reservoir, corresponds to the vegetation of thermophilic forms in summer, coldloving ones in spring and autumn; for cyanoprokaryotes it is limited to 20-25 °C (Mineeva, 2016). In the summer of 2010, at the water temperatures >25 °С, which is not typical of temperate latitudes, CHL Bac but not CHL Cyan , formed the ƩCHL pool, although in southern water bodies the abundant development of cyanoprokaryotes was observed at 30 °C and above (Chu et al., 2007).
The content of ƩCHL, CHL Cyan , and CHL Bac is negatively related to the water transparency and colour, the characteristics of hydrooptical conditions, which are fundamentally important for the development and photosynthetic activity of phytoplankton (Chen et al., 2003;Reynolds, 2006). At the same time, chlorophyll itself belongs to the optically active components of the aquatic environment and affects the formation of underwater light field, deteriorating its characteristics with excessive development of algae (Mineeva, 2004).
With the complex impact of environmental conditions on the aquatic ecosystem and the difficulties in identifying the main factors, the publications of recent years analyze the relationship of ecosystem indicators with markers of global processes, the NAO index and Wolf numbers (Ottersen et al., 2001;George et al., 2004;Pyrina et al., 2006;Maksimov et al., 2009;Lazareva et al., 2013;Mineeva, 2019). Their influence on the productivity of phytoplankton is considered either as indirect or as integral (cumulative). Correlation analysis is considered the main method for studying this impact on ecological processes (Ottersen et al., 2001).
Comparison of averaged parameters for the growing season revealed positive dependence of ƩCHL, CHL Cyan , and CHL Bac on Wolf numbers, which was previously noted for the long-term dynamics of phytoplankton productivity in different water bodies (Pyrina, 2000;Yevstafiev & Bondarenko, 2002;Pyrina et al., 2006;Trifonova et al., 2008). A moderate positive correlation, as well as for the pigment characteristics in the Ivankovo and Uglich reservoirs (Mineeva, 2019), was obtained between ƩCHL, CHL Cyan and the annual NAO. The NAO index, Wolf numbers, temperature, and underwater light conditions, as well as the volume of inflow and water level are among the priority factors regulating the long-term dynamics of phytoplankton in the Rybinsk Reservoir.

Conclusion
In the years with contrasting hydroclimatic conditions and water content (2009-2019), the chlorophyll concentrations in the plankton of the Rybinsk Reservoir varied within a broad range with significant variability of the mean seasonal values, which indicates the low stability of the community. In accordance with the average ƩCHL values, the trophic status of the reservoir was assessed as mesotrophic in 2009 and 2017, eutrophic in 2011-2014, and moderately eutrophic in other years. With an increase in trophy, the concentrations of CHL Cyan , CHL Bac and the relative content of CHL Cyan increased, while the proportion of CHL Bac and CHL Chl in the ƩCHL fund decreased. In the seasonal cycle of phytoplankton, there were five periods with stable temperature conditions and transparency, but significant variability of ƩCHL. The effect of environmental factors on the development of phytoplankton varied in different years. A moderate posi-tive relationship with water temperature and a moderate negative relationship with transparency were seen for the seasonal dynamics of ƩCHL. The dependence of ƩCHL on the influence of external factors may be studied in more detail in the long-term aspect. The NAO indices, Wolf numbers, temperature and underwater light conditions, as well as the inflow volume and water level are among the priority factors regulating the long-term dynamics of phytoplankton in the reservoir.