INTRODUCTION

The lake smelt (a dwarf form of the European smelt Osmerus eperlanus) and the Black Sea-Caspian sprat (also known as Black and Caspian Sea sprat) Clupeonella cultriventris are two small short-cycle invasive species of the Rybinsk Reservoir. The lake smelt dominated in the pelagial of the reservoir from the 1950s to the mid-1990s, while the sprat dominated from the early 2000s until now.

Before the regulation of the Volga River, in its currently flooded part, the lake smelt was periodically found (Arnold, 1925); however it did not naturalized at that time in riverine environmental conditions. After the Volga regulation, the first records about the occurrence of the smelt in the Rybinsk Reservoir were made by local fishermen in 1943, i.e. the species appeared already a year after the filling of the reservoir and after that was constantly noted in the control catches. In 1949, commercial catching of lake smelt was organized using small-mesh fishing gear (Vasil’ev, 1950). During the entire period of the species presence in the Rybinsk Reservoir, there were three periods when the smelt population sharply decreased. The reason for this was the abnormally hot summer months and low water levels. The temperature significantly influences the lake smelt as it is a cold-water species, which upper (lethal) level of water temperature is 26–27°С (Ivanova and Lapkin, 1982). In the anomalously hot summer of 1972, the water temperature did not fall below 25°С; the average monthly temperature in July exceeded the long-term norm (19.6°С) by almost 6°С (Butorin et al., 1982). Such high summer water temperatures for the Rybinsk Reservoir led to the death of most of the smelt population. Catches of this species have decreased by 35 times (Ivanova, 1982).

For the lake smelt populations (as it is a psammophilic species), very low water levels of the reservoir are also crucial: the sandy shallow waters, on which smelt spawns, remain unwatered, preventing the spawning. In 1952, during the spawning period of the lake smelt, the water level turned out to be lower than the long-term average by 1.58 m. Thus, more than 700 km2 of the coastal area were not flooded, which resulted in drainage of almost all areas suitable for the smelt spawning. As a result, the population of the species has sharply decreased, the catches have fallen 10 times, from 150 t in 1952 to 15 t in 1953 (Vasil’ev, 1955).

A similar situation was observed in 1996: water level in the Reservoir was even lower than in 1952. In May 1996, 854 km2 of shallow-water areas were not flooded; during the lake smelt spawning period, the water level was lower than the long-term average by 1.7 m. As a result, only 19 individuals aged 1+ and 2+ were detected in the control catches that autumn, with complete absence of underyearlings. Smelt catches decreased from 1045–2117 ind./10 min trawl in 1994–1995 to 37 ind./10 min trawl in 1996. In 1998, a further decrease in the abundance of the species was associated with the changes in its distribution: for the first time in several previous decades, the lake smelt was found only at 50% of control trawl stations, and in 2002 it completely disappeared from Rybinsk Reservoir.

The described events turned out to be crucial for the lake smelt population, which has not yet restored in the Rybinsk Reservoir (Ryby…, 2015).

The Black Sea sprat was first found in the Rybinsk Reservoir in 1993 in the catch of a pelagic trawl (three mature individuals aged 3+) at a control station near the western shore of the reservoir. That year, the lake smelt was caught at all trawl control stations, and its catch per 10 min trawl with a pelagic trawl ranged from 1000 to 2000 individuals (Ryby…, 2015).

Starting from its first findings in the Reservoir (and also against the background of the lake smelt population decrease after 1996) and until 1998, the sprat was found only sporadically in the catches at one or two control trawl stations along the western coast of the Rybinsk Reservoir. During this period juveniles of representatives of Leuciscidae (Cyprinidae) and Percidae dominated in the pelagial of the Rybinsk Reservoir: bream Abramis brama (13.8%), blue bream Ballerus ballerus (14.4%), common roach Rutilus rutilus (24.0%), pikeperch Sander lucioperca (19.0%), and common perch Perca fluviatilis (6.0%). Their share (abundance) in the catches of the pelagic trawl was ~80% (Gerasimov et al., 2018a).

A noticeable increase in the number of the Black-Sea sprat was noted in 1999 (up to 33 ind./10 min trawl). Moreover, it formed pelagic aggregations in the flooded channels of the Mologa and Sheksna rivers. Its occurrence increased up to 60%. In 2000, it was already encountered at all control trawl stations; its catches reached an average of 300 ind./10 min trawl. In 2002, its catches reached 1400 ind./10 min trawl (Ryby…, 2015).

The occurrence of the sprat the reservoir in the first half of the 1990s coincided with the warming process. The noticeable increase of the water temperature in the reservoirs of the upper Volga, including the Rybinsk Reservoir (Litvinov and Zakonnova, 2012), probably, is one of the reasons for a failure of the cold-water smelt population recovery after the unfavorable 1996. Since 2002, the sprat, which spread into the water body during the period of the smelt population depression, has remained the dominant species in the pelagial of the Rybinsk Reservoir, amounting in some years up to 99% of the total number of fish in pelagic trawl catches.

However, since 2000, there were two periods when smelt, against the background of a high abundance of the Black-Sea sprat, was found (although in small numbers) in almost all parts of the Rybinsk Reservoir. This opens a door for a possible restoration of the lake smelt population in the Rybinsk Reservoir under favorable climatic conditions. Significant average annual and seasonal changes in temperature and water level during the 23-year observation period have no visible negative effect on the dominating sprat in the pelagial of the Rybinsk Reservoir. It is assumed that the most likely scenario will be the cohabitation of the lake smelt and the Black-Sea sprat in the pelagial of the reservoir. To assess the possible level of competition between these species, it is necessary to study their feeding and distribution patterns in cohabitation conditions. For the first time, such an opportunity has arisen in 2019–2020, when the number and occurrence of the smelt reached their maximum values for the entire period of the sprat dominance.

The aim of the work is to study the distribution of lake smelt and Black-Sea sprat in the Rybinsk Reservoir during the period of their cohabitation since 2000, to study their diet in 2019 and 2020 (with largest number of smelt for the 2000s), and to estimate the probability of restoration of the smelt population in the Reservoir.

MATERIALS AND METHODS

Distribution density and size-species composition of the fish population of the Rybinsk Reservoir were determined by trawl-acoustic survey during the regular (in June, July, and October) ship expeditions on the research vessel of the Papanin Institute of Biology of Inland Waters of the Russian Academy of Sciences. The vessel was equipped with hydroacoustic equipment and mid-water trawls.

The trawl surveys were carried out at 20 standard trawl stations throughout the water area of the reservoir. Fish were caught in a pelagic zone with a pelagic trawl with a horizontal opening of 17 m, vertical opening of 1.8 m, and mesh size in the cod end of 4 mm. The length of the warps was 75 m, the duration of trawling was 10 min, and the speed of the ship was approx. 4.5 km/h.

Two trawls were carried out at each station: one in the surface layer and the second, at horizons from 2 to 6 m, depending on the vertical distribution of fish, which was estimated from the data of hydroacoustic surveys carried out during each trawl survey.

Hydroacoustic equipment in 1980 included the Simrad EY-M scientific echo sounder (Simrad, Norway) (operating frequency 70 kHz, beam pattern 10°) and in the 2000s, Simrad EY500 scientific echo sounder with ES120-7C antenna (operating frequency 120 kHz, split beam, omnidirectional pattern, beam angle 7°). The surveys were carried out according to modern methods and recommendations (Yudanov et al., 1984; Simmonds and MacLennan, 2005; Parker-Stetter et al., 2009).

The lake smelt and the Black-Sea sprat for the diet studies were taken during the 2019 and 2020 surveys only if the both species occurred in the trawl catches. The average size and weight of fish, as well as the species composition of the aggregations were determined from trawl catches in the research vessel’s laboratory according to the generally accepted method (Pravdin, 1966), except for individuals taken for feeding studies.

For the subsequent studies of the diet patterns, the collected specimens of the smelt and sprat were frozen. Their bioanalysis and study of the contents of their gastrointestinal tracts (GIT) were performed after defrosting in the laboratory at the end of the research ship expedition. In total, 122 specimens the lake smelt were caught in October 2019, and 60, 47, and 23, in June, July, and October 2020, respectively. All these individuals were taken to study their diet. The same amount of the Black-Sea sprat specimens was taken from the catches. During the expeditions of 2021 and 2022, the smelt was not found in the catches of the pelagic trawl surveys.

RESULTS

Population Dynamics and Distribution

From 2000s to 2020s the abundance of the lake smelt in the water area of the Rybinsk Reservoir was very low (Fig. 1). Its average abundance in the catch per 10 min trawl with a pelagic trawl was 0.8 ± 0.5 ind. (Fig. 1a), in particular, from 2005 to 2022, it varied from 0 to 18 ind./min trawl. The proportion of smelt (by abundance) in the catches of the pelagic trawl averaged 0.18 ± 0.12% (Fig. 1b). After the first disappearance in 2002, the lake smelt reached its maximum abundance in 2018–2019. In 2002–2004, 2010–2011, and 2021–2022 lake smelt was not found in trawl catches.

Fig. 1.
figure 1

Abundance of the lake smelt Osmerus eperlanus and the Black-Sea sprat Clupeonella cultriventris in the catches of the pelagic trawl and water temperature in the Rybinsk Reservoir in 2000–2020: (a, c) average catches of smelt and sprat, respectively; (b, d) dynamics of the proportion of the smelt and sprat, respectively, in the total number of fish in the catch; (e) average monthly water temperature in July.

Full size image

All the cases of the smelt disappearance from the Reservoir occurred in years with a high average July water temperature (Fig. 1e), which is quite consistent with long-term data (from the 1950s to 1990s) on the dependence of the yield of the smelt population of the Rybinsk Reservoir on the July water temperature (Fig. 2). The greatest catches were observed in years with an average monthly water temperature in July of 18–19°С.

Fig. 2.
figure 2

Dependence catches of underyearlings of the smelt Osmerus eperlanus population in the Rybinsk Reservoir on the summer (July) water temperature (r = −0.43, p = 0.01) (according to: Ryby…, 2015): (—), dependence curve; (– –), confidence interval at a significance level of 0.01; (···), confidence interval at a significance level of 0.05.

Full size image

The abundance of the Black-Sea sprat, the species that dominated in the pelagial from the 2000s to 2020s, was several orders of magnitude higher than that of the smelt, and its share in the catches of the pelagic trawl averaged 69 ± 11% (Figs. 1c, 1d) . The average numbers of the sprat in the catch for 10 min of trawling reached 401 ± 208 ind. As for any short-cycle species, the abundance of the sprat varies significantly. Its maximum abundance was noted in 2015, with its average catch of 1873 ind./10 min trawl. The lowest level of this species abundance for the entire study period was in 2010, 48 ind./10 min trawl. That year, as in the second half of the 1990s, the pelagic zone of the reservoir was dominated by juveniles of cyprinids (33%) and perch (48%), but already in the following year the sprat regained its dominant position: its average catch during the 2011 survey amounted to 583 ind./10 min trawl.

However, after the complete disappearance of the smelt in the Rybinsk Reservoir (in 2002), there were two periods when it reappeared in the reservoir. Both times these periods were preceded by the appearance of the lake smelt in the upper reaches of the Sheksna Reach of the Reservoir, since the donor reservoir both in the 1950s (Vasil’ev, 1950; Poddubnyi, 1971) and in the 2000s is Beloye Lake, from which the smelt penetrates into the Rybinsk Reservoir along the Sheksna River. In subsequent years, smelt spread to the central part of the reservoir, occurring in trawl catches at 50% of standard trawl stations. However, the correlation analysis did not reveal a statistically significant relationship between the total number of smelt and sprat (r = −0.12, R2 = 9%) from 2000 to 2022.

After disappearing in the hot year of 2002 (Fig. 1b), the lake smelt reappeared in 2005 and was found in the catches at several trawl stations in the upper reaches of the Sheksna Reach and already in 2008–2009 it was found in catches at one to three stations in the central part of the reservoir (Fig. 3). A similar situation was repeated after the lake smelt disappearance in 2010, which was also characterized by high water temperatures in July, up to 29°С (Lazareva et al., 2012). The smelt reappeared in 2012; in 2013 it was encountered at two stations (in the upper reaches and in the middle part of the Sheksna Reach), but in 2014 the situation of 1996 was repeated: ~700 km2 of shallow waters were not flooded, during the spawning period of the species, the water level at these areas turned out to be lower than the long-term average by 1.4 m. In 2015, the situation with the filling level of the reservoir was repeated: ~900 km2 of spawning grounds of the species were not flooded, since during the spawning period of the smelt, the water level was lower than the long-term average by 1.9 m. The state of the smelt population in 2014 remained unknown, since the autumn level of the Reservoir turned out to be 2.2 m lower than the long-term average and prevent the research vessel of the Papanin Institute of Biology of Inland Waters of the Russian Academy of Sciences to set out on a voyage. In 2015, control trawling yielded in 30 individuals aged 3+ in 10 min in the near-dam part of the central reach of the reservoir; in the upper reaches of the Sheksna Reach, no smelt individuals were found in the catches.

Fig. 3.
figure 3

Occurrence of the lake smelt Osmerus eperlanus at trawl stations in the Rybinsk Reservoir in 2005–2020 (total number of trawl stations, 18). Stations: (), in the central part of the Reservoir, (), in the upper reaches of the Sheksna Reach.

Full size image

In the same years, some troubles with spawning were also observed in phytophilous fish species, the basis of the ichthyofauna of the Rybinsk Reservoir, since coastal vegetation was not flooded at such a low water level. All this results in the absence of the juveniles in the pelagic zone in 2014–2015, and pelagic fish aggregations almost completely consisted of sprat, a species with pelagic spawning that does not depend on the water level (Fig. 1c). Thereby, in 2015, the maximum abundance of the Black-Sea sprat was recorded for the entire period of its occurrence in the Rybinsk Reservoir (Fig. 1d).

In 2016, after stabilization of the water level of the Reservoir, in the upper reaches of the Sheksna Reach in summer, control trawl catches of smelt yielded 60 ind./10 min trawl. In October of the same year, the smelt was already caught throughout the Sheksna Reach, the average catch was 24 ind./10 min trawl. In 2017, it continued to be found only in the Sheksna Reach, and since 2018, the species has been caught at five to six stations in the central reach of the Reservoir (Figs. 3, 4).

Fig. 4.
figure 4

Distribution of the lake smelt in the water area of the Rybinsk Reservoir (according to the catches of the fry trawl) in 2018–2020: (→), flow direction.

Full size image

Correlation analysis did not reveal a significant relationship between the abundance of lake smelt and Black-Sea sprat at the control trawl stations, where both species occurred simultaneously in the catches. In 2019, the correlation coefficient was −0.25 (R2 = 8%); in 2020, r = −0.20 (R2 = 4%).

The last disappearance of the smelt caused by a summer increase in water temperature (Fig. 1e) occurred in the abnormally warm 2021 (Figs. 1c, 3), when the average water temperature in June was 24°С, and in the last ten days of June and in the first days of July exceeded 26°C. The species was not recorded in the control catches of the pelagic trawl in 2022 either.

Vertical Distribution

In the 1980s when smelt dominated in the Rybinsk Reservoir, its largest catches were noted in the middle layer of the water column at a depth of 5–9 m. An echogram of 1982 (Fig. 5a) shows multispecies aggregations of juvenile fish dominated by smelt at a depth of 5–8 m. Big echo markers (bream, pike perch, and blue bream) were found throughout the thickness. In the surface horizons (up to a depth of 5 m), an adult smelt was not found during daylight hours. The water temperature of the layer with the smelt was did not exceed 19.8°C. At the surface, it reached 20.5°С, and below 9 m, it reached 18.5°С.

Fig. 5.
figure 5

Echograms of the vertical distribution of the lake smelt Osmerus eperlanus (before the introduction of the Black-Sea sprat Clupeonella cultriventris) in the 1980s (a) and of the Black-Sea sprat (in the absence of smelt) in the 2000s (b): (a) 1982, Simrad EY-M echo sounder; (b), 2016, Simrad EY500 echo sounder. (1) aggregations of smelt and juveniles of other species; (2) flocks of sprat; (3) echo marks from large fish.

Full size image

In the 2000s, when the Black-Sea sprat already dominated in the Rybinsk Reservoir, its accumulations were noted closer to the water surface. On the echogram of 2016 (Fig. 5b), large rounded echo marks from monospecies sprat flocks were noted at a depth of 3–6 m. More than that, the density of pelagic fish accumulations in the 1980s (with the smelt dominance), according to hydroacoustic data, was 8.4 ± 4.7 g/m2 and in the 2000s (with the sprat dominance) it was 0.8 ± 0.4 g/m2. The accumulations of smelt were sparse, while the sprat kept in dense flocks (Fig. 5).

In 2019 and 2020 when smelt and sprat both inhabited the Reservoir, the species were present in trawl catches in the layer from the surface to a depth of 7 m. At the same time, 10% of the total catch of smelt was caught in the layer from 0 to 4 m, although in this layer it was found only in 12% of trawl catches. In the layer from 4 to 7 m (with maximum catches of smelt) its occurrence reached 42%. The vertical distribution of the sprat in the layer from 0 to 7 m was more uniform. Up to 75% of its total catch was caught in the surface layer (0–4 m), while its occurrence practically did not depend on the depth of trawling. In the surface layer, the occurrence of the sprat was 84%, and in the 4–7 m layer, 79% of trawl catches.

Feeding Habits (Figs. 6, 7)

Fig. 6.
figure 6

Feeding spectrum of the lake smelt Osmerus eperlanus (a–d) and the Black-Sea sprat Clupeonella cultriventris (e–h) in the Rybinsk Reservoir in the summer of 2020. Size groups of fish, mm: a, e, 60–69; b, f, 70–79; c, g, 80–89; d, h, >90.

Full size image
Fig. 7.
figure 7

Feeding spectrum of the lake smelt Osmerus eperlanus (a–c) and the Black-Sea sprat Clupeonella cultriventris (d–f) in the Rybinsk Reservoir in autumn 2019–2020. Size groups of fish, mm: a, 70–79; b, e, 80–89; c, f, >90; d, 50–59.

Full size image

The maximum similarity between the lake smelt and the Black-Sea sprat according to the prevalent prey species dominating in their food in summer was observed only in individuals less than 80 mm in size (age 0+…1+) (Figs. 6a, 6e). Such sized individuals of both species were characterized by the dominance of Bosmina sp. in the contents of the gastrointestinal tract. However the second most abundant group in the smelt consisted of various representatives of Daphniidae, while in the sprat, of Cyclopoida (Fig. 6). In both species, these groups of food objects accounted for >90% of the GIT content. The main difference is that the GIT content of smelt individuals with the size less than 70 mm included undetectable fragments of larvae of late spawning (late May–early June) pelagophilic fish (Fig. 6a). Under the conditions of the Rybinsk Reservoir, these larvae may be attributed to sabrefish Pelecus cultratus or to the sprat, whose eggs and juveniles develop in the sub-surface water layers.

In larger sprats, the food spectrum remained practically unchanged: Bosmina sp. was a dominating food object. In larger smelt individuals, the share of Bosmina sp. decreased and larger food objects, Bythotrephes longimanus, Leptodora kindtii, and large Daphniidae, appeared. In the specimens over 100 mm in size, perch and Volga zander larvae sized 19–20 and 13–14 mm, respectively, were noted (Fig. 6d). The qualitative similarity of food objects (Jaccard coefficient) of the sprat and smelt in summer was 40–54%.

In autumn (Fig. 7), the species composition of the GIT content of both species decreased. In smelt specimens of all captured size groups, the main food objects were B. longimanus with a 15–25% admixture of Bosmina sp. and representatives of Daphniidae. Fish objects were not found in the smelt GIT content in autumn.

The main food object of the sprat of all size groups was Bosmina sp.; in small individuals with a 25% admixture of representatives of Cyclopoida, and in large individuals, with B. longimanus and L. kindtii. The species similarity of diet (Jaccard coefficient) of the sprat and smelt in autumn, as well as in summer, was highest in smaller individuals (~80%). In large fish, this figure did not exceed 25%.

DISCUSSION

Distribution

All of the abovementioned indicates that nowadays the main limiting factor negatively affecting the spread of smelt in the water area of the Rybinsk Reservoir is occasional anomalous heating of the water column and a decrease in the water level during the spawning period.

The 23-year period of the Black-Sea sprat dominance in the pelagial of the Rybinsk Reservoir was characterized by significant average annual and seasonal changes in temperature and water level; however, no negative impact of these factors on the sprat population was noted. The lowest level of its abundance for the entire study period was noted in 2010, 48 ind./10 min trawl (Figs. 1c, 1d). The main difference of this year was the high average monthly water temperature in July, which reached 29°C, furthermore, that period was also characterized by an intensive development of cyanobacteria (Lazareva et al., 2012). Since Black-Sea sprat is a southern invader, summer temperatures (even anomalous for the Rybinsk Reservoir) could not affect it crucially (r = −0.03, p = 0.86). Most likely, the reason for such a significant decrease in the size of the sprat population in 2010 was the consequences of intense water bloom and subsequent decay of algae, which caused oxygen deficiency in the autumn-winter period, resulted in elimination of part of the sprat population (Karabanov, 2013).

During the period from 2000 to 2022, the lake smelt distribution in the water area of the Rybinsk Reservoir was successful only under favorable temperature and level regimes, which persisted for several years. Furthermore, its distribution is not influenced by the abundance of the Black-Sea sprat. For example, in 2018–2020 the smelt was found throughout the central part of the Rybinsk Reservoir and did not occur only at stations in the upper reaches of the Volga and Molozhsky reaches, where it simply did not manage to settle due to the counteraction of the runoff current and insufficient abundance.

As in the 1980s in 2019–2020, 90% of the total catch of smelt was caught in the horizon below 5 m, while only 1/4 of the total catch of sprat accrued to the same horizon. Moreover, the water temperature in the habitat layer of the smelt was lower than near the surface for ~1°С.

From 2018 to 2020 (the period with the largest yields of the lake smelt in the 2000s in the water area of the Rybinsk Reservoir) there was no statistically significant relationship between the total number of smelt and sprat at the control trawl stations, where both species were found in catches simultaneously. Even despite the fact that sprat is more competitive than the smelt. For example, the average catch per effort of a pelagic trawl in the 1980s (smelt domination) reached 7851 ± 5834 ind./10 min trawl, and in the 2000s (sprat domination), 2008 ± 1017 ind./10 min trawl. At the same time, the proportion of juveniles of other fish species in these catches in the 1980s was 33%, while in the 2000s, only 4%. The higher competitive ability of the Black-Sea sprat is also proved by the fact that at a relative abundance of sprat in the catch of the pelagic trawl (300–700 ind./10 min trawl), the admixture of juveniles of other fish species was rare (6 ± 3 ind. per 100 ind. of sprat ). However, during the period of smelt dominance, such a low density of juveniles of other fish species was observed only when the abundance of the dominant in the catch was >900 ind./10 min trawl (Gerasimov, 2018b).

The absence of dependence of the number of smelt on the number of sprat from 2018 to 2020, during the smelt spread over the water area of the Reservoir, was obviously associated with the observed spatial vertical segregation due to different thermal preferences of these species.

Diet

The study of the smelt and sprat diets during periods of their cohabitation in the Rybinsk Reservoir with a significantly lower abundance of smelt showed that they have a certain similarity in the composition of food planktonic organisms found in their gastrointestinal tracts, especially in specimens <80 mm in size (age 0+…1+). With increasing size, the differences increase: the food object spectrum of sprat changes slightly, while in the smelt these changes are more significant: large planktonic organisms (B. longimanus, L. kindtii, and large Daphniidae) begin to predominate among the food objects of the individuals >80 mm in size and fish objects appears in food of individuals from the age of 1+. As a result, the diet similarity (Jaccard coefficient) of sprat and smelt in 2020 in individuals less than 80 mm in size in summer was 40–50%, and increased to 80% in autumn. In larger individuals, the diet similarity in summer was 45–54%, and in autumn, 23%.

A similar age-related variability in the diet of the smelt was also observed in the 1950s–1990s, before the Black-Sea sprat introduction to the Reservoir, when small food organisms were replaced by larger ones as the fish grew (Ivanova, 1982). The smelt fattened most intensively at the end of May, in June, and July. The basis of its food bolus during this period consisted of Bosmina sp., L. kindtii, Daphnia sp., and other cladocerans. In individuals aged 2+, such large objects as L. kindtii and B. longimanus accounted for up to 70% of the GIT contents (20 and 48%, respectively), and 20% were fish objects. In individuals aged 3+, the proportion of invertebrates decreased (20 and 8%), and juvenile fish accounted for up to 70% of the GIT content (Ivanova, 1982).

The studies of the sprat diet after its introduction into the Rybinsk Reservoir (Kiyashko et al., 2012) showed that the proportion of large crustaceans (Heterocope, L. kindtii, and B. longimanus) increases in adults, although their proportion is lower than that of large smelt. Small representatives of plankton (Bosmina, Chydorus, and Cyclopoida) are qualitatively and quantitatively more abundant in the sprat diet. The age-related feeding habits of sprat from the Rybinsk Reservoir remain unchanged in years with different hydrological conditions (Kiyashko et al., 2012).

Comparison of our data with the results of earlier studies shows that the feeding spectrum of the Black-Sea sprat practically does not change. The same is true also for the lake smelt, whose feeding spectrum in the 1980s was similar to that obtained in 2019–2020. In larger individuals of the smelt, as in the 1980s, L. kindtii and B. longimanus dominate and fish food appears. The basic food objects for large smelt were abundant Daphnia and a few large crustaceans Heterocope, B. longimanus, and L. kindtii, as well as large Cyclopoida with body length >1 mm (Kiyashko et al., 2007).

Significant qualitative similarity of the smelt and sprat diets has been revealed. Certain differences are observed in the proportions of consumed species and in the range of food spectrum. Obviously, the high density of sprat leads to intensive consumption of plankton, especially of such large and less abundant species as L. kindtii and B. longimanus. The selectivity indices of these species for sprat were positive and reached 0.8–0.9 (Kiyashko et al., 2007). In addition, being predators, these species may lose in competition with sprat for smaller species of planktonic crustaceans (Herzig, 1995; Barbiero et al., 2004). In the water horizon inhabited by smelt (less populated by fish), the density of these two species of large crustaceans and their food items is higher (Kiyashko et al., 2007, 2012). Obviously, the difference in the abundance of L. kindtii and B. longimanus in the water layers preferred by smelt and sprat determines the differences in their proportions in the GIT of these fish species. This also explains the wider diet range of the Black-Sea sprat compared to the lake smelt due to smaller representatives of planktonic crustaceans.

Therefore, it can be assumed that if the number of smelt increases significantly, these differences will most likely disappear and the similarity of the qualitative and quantitative indicators of the diet of sprat and smelt of older ages will increase, which will lead to increased competition between them. Especially significantly it will be pronounced in years with moderate water temperatures, when the difference in temperature preferences will not contribute to the formation of pronounced vertical segregation.

Predatory Feeding Patterns of the Lake Smelt

In the 1980s, cannibalism was common for the smelt in relation to its own early juveniles of the first year, especially in the autumn-winter period. Consequently, the smelt can potentially act as a predator, actively consuming not only its own, but also early juvenile sprats. In the summer of 2020, predator smelt individuals accounted for 3% of the total number of examined individuals. Fish food was found both in smelts of approximately 70 mm long (age 1+) and in specimens >100 mm in size (age 3+).

In the 1950s–1990s fish objects were constantly found in the food of the smelt of the Rybinsk Reservoir (Ivanova, 1982, Gerasimov et al., 2018a). In spring and summer (until September), the proportion of individuals feeding on fish varied from 6 to 15%. The first fish food objects are usually found in smelt of the age of 1+ and averaged ~5%, in individuals aged 2+, its proportion is 20%, and at age 3+, up to 70%. In May–July, the main consumed fish objects are the larvae of the ruffe Gymnocephalus cernua, perch, zander, and its own juveniles. In the second half of summer, fast-growing perch juveniles become out of the influence of smelt predators, and smelt underyearlings become the main food for adult smelt.

The mass consumption of its own juveniles, apparently, is explained by the high availability of the latter, which is determined by size characteristics and similar habitats. At the same time, even among their own juveniles, smelt choose underyearlings that are lagging behind in growth (Ivanova, 1982). Therefore, on the one hand, the high availability of early sprat juveniles for smelt will be associated with its size and similar habitats, since the sprat is a pelagophile and its eggs and early juveniles develop in the pelagic zone, and on the other hand, juvenile sprat will get out of the smelt pressure faster than its own juveniles due to their high body size. As a result, as in the 1950s–1990s, in the second half of the summer and in the autumn–winter period, it will be forced to switch to feeding on its own juveniles. This is also indicated by the lack of fish food objects in the large smelt individuals in the fall of 2020, since the grown juveniles of sprat and other fish species got out of its predatory pressure by autumn, and their own juveniles were not abundant.

Thus, it can be assumed that if the smelt population reaches a comparable abundance levels to that of the sprat, its predation and cannibalism can reduce the tension of competitive relations due to the active consumption of its own juveniles as well as sprat juveniles at early stages of development.

Consumption of Smelt by Predatory Fish

Predation by mass species of predatory fish, primarily from the Percidae family (pikeperch and perch) also limits the number of smelt in the Reservoir. In addition, the smelt will be a more preferable prey than sprat for immature individuals of perch and zander because of the more slender body (the ratio of body length to its height in sprat is 3.3 versus 6.4 in smelt) (Gerasimov et al., 2013). It is known that the ratio of the lengths of zander underyearlings and their food objects is closely related to the body shape of the latter. Thus, in zander 25 mm long, the length of prey, perch, roach, and smelt, was 50, 73, and 96% of the predator’s length, respectively (Romanova, 1958). This indicates that the fish with the greatest relative length were slender fish, i.e. with the lowest body height. Consequently, the smelt remained available to the majority of zander juveniles throughout the summer (Gerasimov et al., 2013). Until the mid-1990s in its diet perch juveniles accounted for up to 43% of the number of fish eaten, smelt, up to 23%, ruffe, 14%, and roach, 13% (Gerasimov et al., 2013). Another part of the yearlings of zander for a very long time (up to October) fed exclusively on zooplankton. The ratio of planktophages and predators among zander underyearlings largely depended on climatic conditions. In some years, the proportion of underyearling predators increased to 60% by autumn (Romanova, 1958). However, in years with hot summers, the proportion of individuals that switched to carnivorous feeding decreased. The reason for it is a rapid growth of juvenile food fish species with an increase in water temperature. As a result, the prey quickly became inaccessible to juvenile zander (Poddubnyi, 1971; Ryby…, 2015).

The replacement of smelt with sprat enhance the problem of the transition of early juvenile pike perch to fish food in the Rybinsk Reservoir, since the Black-Sea sprat is much higher bodied than smelt. As a result, during the period of its dominance, the proportion of predators among pike-perch underyearlings in autumn does not exceed 10%. In abnormally hot years (for example, in 2010), there were no zander underyearlings in the catches that switched to predatory feeding. This affected the growth rate of immature zander. During the period of smelt dominance, zander individuals aged 1+ in the control trawl catches had an average length of 200 mm and a weight of 100 g, and during the dominance of sprat, they were 150 mm and 70 g, respectively.

Thus, the mass development of smelt will lead to its active consume not only by sprat but also by mass species of predatory fish of all size groups. Also it should be considered that in the 2000s there was a significant decrease in the number of these predators in the Rybinsk Reservoir. For example, zander (the main consumer of small pelagic fish) fishery catches in the 1980s averaged of 257 tons per year, while currently the catches have decreased to 53 tons per year (Gerasimov et al., 2018b). The fishing load on the perch population, on the contrary, has increased four times due to a decrease in the stocks of commercially more valuable fish species (bream, pike perch, and pike Esox lucius) (Gerasimov et al., 2018b). This also leads to a decrease in its abundance. All this should lead to a potential reduction in the pressure of mass predators on the smelt population compared to the 1980s.

Thus, the number of smelt will be limited due to its consumption by mass predators, but their press will be weaker compared to the 1980s. In addition, the Black-Sea sprat is also actively used by predators as a food object (Stepanov and Kiyashko, 2008; Gerasimov et al., 2018a). Active consumption of smelt and sprat by mass predators will obviously become the main factor contributing to the reduction of competition between these two species.

CONCLUSIONS

It can be assumed that if there will be a long period with the absence of years with an abnormally high summer water temperature, the abundance of the lake smelt in the Rybinsk Reservoir can reach values comparable to the abundance of the Black-Sea sprat.

However, it is unlikely that the smelt will be able to regain dominance. The co-dominance with sprat is most realistic. More than that, it can be facilitated by the formation of spatial vertical segregation of these species, due to different thermal preferences, and the regulation of their numbers by smelt predators and other mass species of predatory fish.

For example, the lack of dependence between the numbers of smelt and sprat from 2018 to 2020 during the spread of smelt over the water area of the reservoir was obviously associated with the observed spatial vertical segregation between these species.

On the other hand, a significant increase in the smelt number will increase similarity of qualitative and quantitative diet indicators of sprat and smelt of older ages, which in turn will lead to increased competition between them. This will be especially evident in years with moderate water temperatures, when the difference in thermal preferences will not contribute to the formation of vertical segregation.

Obviously, one of the factors that can reduce competition between sprat and smelt will be the active consumption by smelt of its own juveniles as well as sprat juveniles at early stages of development, but the main role in reducing competition between these species will be played by mass predators (zander and perch), actively consuming sprat, because its share in their diet is now higher than that of smelt in the 1980s.

The restoration of the lake smelt population will bring certain benefits to the populations of mass species of predatory fish in the Rybinsk Reservoir, since it will improve the nutritional conditions of the immature part of these populations, especially at the earliest stages of development, facilitating their transition from planktonic to predatory food objects.