Content uploaded by Vardan Asatryan
Author content
All content in this area was uploaded by Vardan Asatryan on Nov 16, 2023
Content may be subject to copyright.
Citation: Asatryan, V.; Vardanyan, T.;
Barseghyan, N.; Dallakyan, M.;
Gabrielyan, B. Experimental
Validation of Suitability of a River for
Natural Reproduction of Trout of
Lake Sevan Using Egg Incubation.
Water 2023,15, 3993. https://
doi.org/10.3390/w15223993
Academic Editor: Antonia Granata
Received: 24 September 2023
Revised: 2 November 2023
Accepted: 7 November 2023
Published: 16 November 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
water
Article
Experimental Validation of Suitability of a River for Natural
Reproduction of Trout of Lake Sevan Using Egg Incubation
Vardan Asatryan , Tigran Vardanyan, Nelli Barseghyan, Marine Dallakyan * and Bardukh Gabrielyan
Scientific Center of Zoology and Hydroecology, National Academy of Sciences of the Republic of Armenia,
7 P.Sevak str., Yerevan 0014, Armenia; asatryanvardan@gmail.com (V.A.); vardtigran@mail.ru (T.V.);
nelli.barseghyan@yahoo.com (N.B.); gabrielb@sci.am (B.G.)
*Correspondence: dallakyan.marine@gmail.com; Tel.: +374-99999175
Abstract:
The endangered endemic species Sevan trout (Salmo ischchan Kessler, 1877) is under the
threat of extinction and its survival is dependent on restocking by smolts. Thus, there is an urgent
need to find an effective solution for restocking wild populations. After developing a theoretical
methodology to reveal the most suitable parts of spawning rivers for natural reproduction, we
consider the incubation of eggs in the wild as an appropriate method for the validation of the results
of the theoretical assessment. Thus, we estimate the eggs’ survival differences in the wild and
microcosm conditions which found significant differences in the sac-fry emergence rate in the zones
assessed as “partly suitable” and “unsuitable”. Also, the results show that the emergence of eggs
in the wild was less than in the microcosm conditions and the success rate of emergence of eyed
eggs in the wild was higher than that of the green eggs. The results of one-way ANOVA, correlation
and general linear mixed model analyses have in general proven the theoretical assessment but raise
some questions on the completeness of using a set of abiotic criteria as the range of success during
different experiments varies significantly.
Keywords: Sevan trout; the Masrik River; endemic species; restocking; incubation; egg survival
1. Introduction
Freshwater ecosystems play a major role in the maintenance of sustainability of the
planet through the provision of several ecosystem services [
1
]. Although freshwater habitats
comprise about 0.0091% of the hydrosphere, some 15,200 strictly freshwater species of
fish live there [
2
]. High levels of endemism are typical for freshwater fish due to the
fragmentation or complete isolation of their habitats. Many endemic species are heavily
threatened due to ongoing climate change [
3
] and human-made disturbances [
4
]. The
Caucasus region is one of the biodiversity hotspots on the planet [
5
]. Sevan trout (Salmo
ischchan Kessler, 1877) is an endemic species native for only Lake Sevan (Armenia) [
6
,
7
].
Considering its commercial value, Sevan trout were regularly introduced to other water
bodies throughout the territory of the former Union of Soviet Socialist Republics (USSR)
for commercial trout fishery. But only the stocking of Lake Issyk Kul (Kyrgyzstan) was
successful in the long term [8,9].
State of Sevan trout.
Since the morphology of Lake Sevan was established in the
Pleistocene and Holocene [
10
], the lake and its endemic species are relatively young. How-
ever, human intervention into natural processes of the lake during the 20th century led to
the degradation of its ecosystem and increased the vulnerability of native species [
11
,
12
].
According to the most recent estimation, Sevan trout are a critically endangered species
(corresponds to International Union for Conservation of Nature (IUCN) category CR
A2cd) [
13
]. Until the 1970s, it was represented by four subspecies (morphs) differing in
terms of breeding times and places, as well as growth rates [
14
,
15
]. Particularly, winter
bakhtak (
S. ischchan ischchan
Kessler, 1877) (growing up to 90 cm and reaching 15 kg) and
Water 2023,15, 3993. https://doi.org/10.3390/w15223993 https://www.mdpi.com/journal/water
Water 2023,15, 3993 2 of 12
bodjak
(S. ischchan danilewskii
Iakowlev, 1988) (dwarf, slowly growing lacustrine fish, not
exceeding 33 cm and 250 g) bred in the lake while summer bakhtak (S. ischchan aesti-
valis Fortunatov, 1926) bred both in the tributaries or near the river mouth; gegharkuni
(S. ischchan gegarkuni
Kessler, 1877) bred exclusively in the tributaries [
15
–
18
]. While in-
creased water diversion for irrigation and energy generation has led to extinction of winter
bakhtak and bojak [
19
] due to loss of spawning ground in the littoral zone, poaching in river
mouth parts has significantly reduced the populations of summer bakhtak and gegharkuni.
Thus, only these two subspecies have remained in Lake Sevan today [20].
Theoretical background of the experimental validation.
Several human-induced
changes in the spawning rivers like the establishment of migration barriers [
21
] and in-
creased level of pollution [
22
] make the possibilities for a natural reproduction of Sevan
trout very limited. Thus, all the relevant authorities and academia in Armenia have been
looking for effective solutions to protect Sevan trout since the 1980s. However, conservation
measures had small success due to socio-economic problems in the country after the col-
lapse of USSR. The main strategy of restocking wild population by farm-raised smolts also
has little success as the natural reproduction of stocked fish in the rivers hardly occurs [
23
].
In addition to all the artificial barriers, there is a risk that stocked fish may not return to the
spawning rivers [
24
]. According to the information from the Foundation for restoration of
Sevan trout stocks and development for aquaculture (STF) (personal communication with
T. Vardanyan) about 3.818 million specimens of gegharkuni and 2.136 million specimens
of summer bakhtak were released in total into the drainage basin of Lake Sevan by the
Ministry of the Environment during 2005–2015 and by the STF during 2015–2020. Possible
restocking by the population from Lake Issyk Kul was also disputed as the recent findings
based on molecular and morphometric analyses are rather controversial. Particularly,
molecular analysis shows significant difference between the populations from various
farms in Armenia and from Lake Issyk Kul [
17
] while morphologically Kyrgyz and wild
populations from Lake Sevan are closer than the wild populations and trout released from
fish farms in Armenia [
25
]. Considering also the risk of domestication when restocking
with farm-grown fish, it was supposed that incubation of eggs directly in the river would
be an alternative and rather efficient solution for restocking Sevan trout [
24
] taking into
account the homing reflex of fish.
Thus, pilot projects have been launching since 2015 for the development of method-
ologies to assess the recent state of Lake Sevan tributaries as spawning grounds for Sevan
trout (hereafter theoretical assessment) [
22
,
26
]. Assessments rely on the analysis of core
hydro-physical, hydro-chemical and hydro-morphological parameters that support natural
reproduction of Sevan trout. In particular, Sevan trout prefer the gravel, cobble and pebble
to settle the nest where depth is about 15–50 cm; velocity is 0.15–0.55 m/s; dissolved oxygen
concentration is high (8–15 mg/L) and pH is between 6 and 9. However, gegharquni spawn
in cold periods when water temperature drops below +13
◦
C and summer bakhtak spawn
in a warmer period when temperature increases to more than +10 ◦C [15,26].
The recent experimental validation of the results of theoretical assessments were
conducted in the Masrik River system. According to the results of the theoretical assess-
ment [
26
], only two categories of river stretches could be distinguished in the Masrik
River (Figure 1): 1. partly suitable for natural reproduction and 2. unsuitable. “Suitable”
spawning grounds are missing in the system of the Masrik River because of two major
reasons: (1) artificial barriers for the spawning migrations; and (2) anthropogenic pressures
on aquatic ecosystems in terms of wastewater discharge, solid waste, etc.
Water 2023,15, 3993 3 of 12
Water 2023, 15, x FOR PEER REVIEW 3 of 13
Figure 1. The distribution of experiment sites.
Considering the lack of recent opportunities to observe the natural reproduction of
Sevan trout in the rivers, we launched an experiment with semi-natural reproduction via
nesting in a simple box-incubator in the wilderness and regularly maintained it until the
possible emergence of sac-fry. Such method is also common for increasing salmonid
densities in different countries [27]. Thus, the aim of the current study was to validate the
results of theoretical assessment of the suitability of rivers as spawning grounds for Sevan
trout through incubation of eggs in the wild. One more objective was to estimate the
difference of the egg survival until the sac-fry stage in the wild and microcosm conditions
created in the farm for restocking Sevan trout. Because there are only two categories of
stretches, the main hypothesis tested was that sac-fry should emerge from the eggs of
gegharkuni in the zones assessed as “partly suitable”, while there should be no emergence
in the “unsuitable” zones. At the same time the mortality rate of eggs in the wild should
be higher than in the microcosm conditions. Also, we have hypothesized that the success
rate of the experiments should be higher in case of planting eyed eggs (a stage when black
spot of the eyes of embryo are already visible) in the wild compared with planting green
eggs (just fertilized egg stage).
2. Materials and Methods
Experiment sites. Due to the conservation status of Sevan trout and a very limited
availability of their eggs, we have strictly limited many aspects of wild experiments. For
instance, they were conducted only in the system of the Masrik River which provides a
wide range of abiotic conditions and is recently beer studied than other spawning rivers
Figure 1. The distribution of experiment sites.
Considering the lack of recent opportunities to observe the natural reproduction of
Sevan trout in the rivers, we launched an experiment with semi-natural reproduction
via nesting in a simple box-incubator in the wilderness and regularly maintained it until
the possible emergence of sac-fry. Such method is also common for increasing salmonid
densities in different countries [
27
]. Thus, the aim of the current study was to validate
the results of theoretical assessment of the suitability of rivers as spawning grounds for
Sevan trout through incubation of eggs in the wild. One more objective was to estimate the
difference of the egg survival until the sac-fry stage in the wild and microcosm conditions
created in the farm for restocking Sevan trout. Because there are only two categories of
stretches, the main hypothesis tested was that sac-fry should emerge from the eggs of
gegharkuni in the zones assessed as “partly suitable”, while there should be no emergence
in the “unsuitable” zones. At the same time the mortality rate of eggs in the wild should be
higher than in the microcosm conditions. Also, we have hypothesized that the success rate
of the experiments should be higher in case of planting eyed eggs (a stage when black spot
of the eyes of embryo are already visible) in the wild compared with planting green eggs
(just fertilized egg stage).
2. Materials and Methods
Experiment sites.
Due to the conservation status of Sevan trout and a very limited
availability of their eggs, we have strictly limited many aspects of wild experiments. For
instance, they were conducted only in the system of the Masrik River which provides a wide
range of abiotic conditions and is recently better studied than other spawning rivers [
22
].
Two sites (Station N1—40
◦
11
0
15
00
N; 45
◦
42
0
30
00
E; Station N2—40
◦
13
0
4
00
N; 45
◦
39
0
26
00
E) were
chosen at so-called “partly suitable” zones with slightly different abiotic conditions [
28
] and
Water 2023,15, 3993 4 of 12
one at an “unsuitable” zone for comparison (Station N3—40
◦
8
0
3
00
N; 45
◦
49
0
3
00
E) (Figure 1).
Microcosm experiments were conducted in parallel in the Karchaghbyur fish farm located
nearby the Masrik River which specialized in hatching Lake Sevan trout.
Conditions of experiments.
For all experiments, eggs and milt were collected from the
same parents simultaneously to avoid any bias on a quality of eggs or success of fertilization
and all the experiments were conducted by the same order of steps. The fertilization
was strictly performed in frames of annual fertilization works in STF by their specialists.
However, considering some differences in temperature conditions between Station N3 and
the two other stations (Supplementary Table S1), we launched the experiment for Station
N3 on 23 October 2018 and for Stations N1 and N2 on 30 October 2018.
In the case of the experiment in Station N3, about one third and, in the case of the
experiment in Stations N1 and N2, about half of the green eggs were directly transported to
the experimental sites using special box-refrigerators. The remaining part was transported
to the farm for the incubation. The time between fertilization and planting in the experi-
mental sites fluctuated from 1 to 2 h. Because we set two stages for the experiments, when
the eggs emerge to the eyed stage in the farm, we transported almost half of the eggs to the
wild experiment sites using the same box-refrigerators. In all cases, semi-natural planting
of eggs was performed by putting the eggs into the incubator boxes and burying them into
the artificially established nests made by the cobble, gravel and pebble taken in situ.
In order to set the best possible conditions for the wild experiments, the spawning
habits of gegharkuni were followed as much as possible. In particular, the experiments
were launched when optimum temperatures for breeding were established in the cho-
sen stretches. Nests were established at optimal depths following the observations of
Dadikyan [
29
] and, in optimal water velocity, following the observations of Savvaitova
et al. [30].
Microcosm conditions. Planting was conducted in two special incubators of 42 cm
in width, 360 cm in length and 17 cm in depth. Sieves had 39
×
39 cm size. Water was
supplied from a nearby source that also fed the Karchaghbyur River. During the experiment
Dissolved oxygen (DO) fluctuated from 8.5 to 9 mg/L, water temperature 8.3–9.5
◦
C and
pH was 7 (Supplementary Table S1).
All experiments in the wild were conducted using Whitlock-Vibert Boxes (WVB’s
from the Fly Fishers International) as their effectiveness for incubation of gegharkuni eggs
in the tributaries of Lake Sevan was already proven [
24
]. During the experiments, some
basic abiotic parameters as well as survival rates were monitored on a weekly basis from
23 October
to 11 November at Station N3 and from 30 October to 15 February at Stations N1
and N2. In particular, temperature, pH and dissolved oxygen conditions were measured by
handheld Hanna HI9813-5N pH/EC/TDS and HI9147-10 DO meters twice per day in the
early morning and in late afternoon, while the fluctuations in water level near nests were
measured by a meter stick once per visit. All the measurements were strictly conducted
near the places where the WVB’s were installed. Then, the mean values were calculated for
each visit. Success of hatching in all sites was estimated through empirical observation of
discolored eggs’ proportion in each WVB, which shows the mortality rate.
Experiments with green eggs. The number of eggs planted in each WVB was derived
by dividing the total mass of planted eggs by the average mass of gegharkuni egg (0.07 g).
Thus, in total, six WVB’s (two per station) with 2657 green eggs were installed in three
chosen wild stretches and 2374 eggs remained in the farm.
Experiments with eyed eggs. A second set of experiments was launched on 6 December
2018 when the green eggs from the first set of experiments reached the eyed stage in the
farm. As the average weight of such eggs was 0.09 g, it was calculated that we have
installed two WVBs with a total number of 744 eyed eggs only in the “partly suitable”
zones (one per site). Considering the loss until the eyed stage, about 890 eggs remained in
the farm after the launch of the second stage of the experiments. Observations of abiotic
conditions in Station N3 (unsuitable zone) have shown no chance for egg survival as the
Water 2023,15, 3993 5 of 12
water was mostly frozen. Thus, it was preferable to interrupt the experiment there to avoid
loss of eyed eggs.
Statistical processing of data.
All statistical calculations were conducted using the
IBM SPSS 17 software. Series of one-way ANOVA tests were performed to compare
means for the results of different experiments to check the results of theoretical assessment
and to reveal the part of the Masrik River where further implementation of artificial
incubation of eggs could be more effective. In all tests, the survival rates were set as a
dependent parameter. Basically, the hypothesis of homogeneity of variance was checked
through Levene’s test and the equality of means was checked through the Welch and
Brown–Forsythe tests. In the first set of analyses, we separately check the hypotheses
of significance of differences for mean values of survival rates of (1) green and (2) eyed
eggs between two stations and also the fish farm. The experiment sites were set as an
independent factor here. As a post hoc test for the results of survival rates between the
stations, the Tukey test was used. Because in the first set of experiments the eggs were
planted using two WVB’s in each station, the mortality rates were derived by calculating the
average numbers of discolored eggs per station for each visit. In the second set of analyses,
the means for green and eyed egg survival rates were compared for the same stations in
the wild. Here, we assumed that if the variance between the groups is significantly higher
than the variance within the groups (F factor), then there are proven differences in the
effectiveness of hatching green and eyed eggs. The F value was derived through univariate
analysis of variance.
As abiotic parameters and their fluctuations were considered decisive for the spawning
success in theoretical assessment, we performed General Linear Model analysis to estimate
a significance of effects of measured basic parameters on hatching success. The main
assumption made was that if the mean difference of the parameter is significantly different
among two stations, then it could lead to differences in success rates of the experiments.
For that, stations were set as an independent factor and the measured abiotic parameters–
dependent. Then, nonparametric correlation analysis (Spearman) was conducted between
the values of abiotic parameters and the recorded mortality rates.
Ethics statement.
All eggs for the experiments were supplied by the STF in frames
of their restocking and breeding action plans; all the experiments were approved by the
STF and Scientific Center of Zoology and Hydroecology of National Academy of Sciences
(NAS) of Armenia in frames of the contract with the Ministry of Environment of RA on
annual assessment of the stocks of fish and crayfish in Lake Sevan basin.
3. Results
Hatching of green eggs
. The success rate of the experiments varies in a wide range
between the stations (Figure 2). Null success of hatching was recorded in Station N3.
Moreover, the main experiment was interrupted after two weeks due to frozen water in
the station. The results of ANOVA test for hatching success claim the significance of mean
differences between the fish farm and both Stations N1 and N2 (Table 1). In particular,
there was no survival of green eggs in Station N2 while the success rate in Station N1 was
10%. Also, it should be considered that while there was a yield from Station N1, the success
rate of the hatching in the farm was 57.38% in the same time period. Thus, environmental
conditions in Station N1 significantly affected survival. This also proves the results of
theoretical assessment and the assumption that incubation in the wild will have a lower
success rate than in the farm.
Water 2023,15, 3993 6 of 12
Table 1. The results of one-way ANOVA test for the hatching success of green eggs.
Compared Pairs for Mean
Difference
Mean Difference of Survival
Rate ±SE
Significance of Mean
Difference (Tukey’s HSD)
Station N1-Station N2 16.2 ±10.663 0.292
Station N1-Fish farm −27.45 ±10.663 0.036
Station N2-Fish farm −43.65 ±10.663 0.001
Note: The mean difference is significant at the <0.05 level.
Water 2023, 15, x FOR PEER REVIEW 6 of 13
Table 1. The results of one-way ANOVA test for the hatching success of green eggs.
Compared Pairs for Mean
Difference
Mean Difference of
Survival Rate ± SE
Significance of Mean Difference
(Tukey’s HSD)
Station N1-Station N2 16.2 ± 10.663 0.292
Station N1-Fish farm −27.45 ± 10.663 0.036
Station N2-Fish farm −43.65 ± 10.663 0.001
Note: The mean difference is significant at the <0.05 level.
Figure 2. Cumulative mortality rates (%) of green eggs until emerging into eyed eggs on 6 December
2018 and later until emerging into sac-fry in different stations.
Hatching of eyed eggs. The experiments with the eyed eggs were launched after about
five weeks from fertilization. Considering the emergence of green eggs at the farm was
68.78%, only a 7.25% loss was recorded during the eyed stage (Figure 3). Unlike the green
eggs, the yield from the eyed eggs in Station N2 was 5%, while, in Station N1, it was 80%.
However, compared with the 92.75% yield in the farm, it is also quite low. The results of
ANOVA claim the significance of difference of eyed egg survival between Station N2 and
the remaining two sites of the experiments (Table 2). However, a null hypothesis was
rejected for the pair Station1-Farm. Considering the results of theoretical assessment, such
result raises some questions on its precision as both stations were previously concluded
“partly suitable”.
Figure 2.
Cumulative mortality rates (%) of green eggs until emerging into eyed eggs on 6 December
2018 and later until emerging into sac-fry in different stations.
Hatching of eyed eggs
. The experiments with the eyed eggs were launched after about
five weeks from fertilization. Considering the emergence of green eggs at the farm was
68.78%, only a 7.25% loss was recorded during the eyed stage (Figure 3). Unlike the green
eggs, the yield from the eyed eggs in Station N2 was 5%, while, in Station N1, it was 80%.
However, compared with the 92.75% yield in the farm, it is also quite low. The results
of ANOVA claim the significance of difference of eyed egg survival between Station N2
and the remaining two sites of the experiments (Table 2). However, a null hypothesis was
rejected for the pair Station1-Farm. Considering the results of theoretical assessment, such
result raises some questions on its precision as both stations were previously concluded
“partly suitable”.
Water 2023,15, 3993 7 of 12
Table 2. The results of one-way ANOVA test for the hatching success of eyed eggs.
Compared Pairs for Mean
Difference
Mean Difference of Survival
Rate ±SE
Significance of Mean
Difference (Tukey’s HSD)
Station N1-Station N2 37.2 ±9.093 0.001
Station N1-Fish farm −6.644 ±9.093 0.748
Station N2-Fish farm −43.844 ±9.093 0.000
Note: The mean difference is significant at the <0.05 level.
Water 2023, 15, x FOR PEER REVIEW 7 of 13
Table 2. The results of one-way ANOVA test for the hatching success of eyed eggs.
Compared Pairs for Mean
Difference
Mean Difference of
Survival Rate ± SE
Significance of Mean
Difference (Tukey’s HSD)
Station N1-Station N2 37.2 ± 9.093 0.001
Station N1-Fish farm −6.644 ± 9.093 0.748
Station N2-Fish farm −43.844 ± 9.093 0.000
Note: The mean difference is significant at the <0.05 level.
Figure 3. Cumulative mortality rates (%) of eyed eggs installed on 6 December 2018 in different
stations.
Considering the faster emergence in the farm, the observations continued on yolk
sacs after the 13th week. During the last two weeks of the experiments, some 4.15% of sac-
fry had also died in the farm. However, it was impossible to measure the semi-natural
mortality of yolk sac in the river to compare as they were aacked by the crustaceans of
the genus Gammarus (Fabricius, 1775). Thus, we conclude it inappropriate to involve such
comparisons in the further discussion.
The effectiveness of incubation in the wild. Comparison of green and eyed egg incubation
success. It was initially assumed that the survival rate of green and eyed eggs at the same
stations will be different because of the sensitivity criterion. The results of Univariate
Analysis of Variance partially prove the assumption (Table 3). In particular, differences
between the groups were significantly higher than within the groups in Station N1 but not
in Station N2. This means that the environmental conditions almost equally influence the
Figure 3.
Cumulative mortality rates (%) of eyed eggs installed on 6 December 2018 in different stations.
Considering the faster emergence in the farm, the observations continued on yolk
sacs after the 13th week. During the last two weeks of the experiments, some 4.15% of
sac-fry had also died in the farm. However, it was impossible to measure the semi-natural
mortality of yolk sac in the river to compare as they were attacked by the crustaceans of
the genus Gammarus (Fabricius, 1775). Thus, we conclude it inappropriate to involve such
comparisons in the further discussion.
The effectiveness of incubation in the wild.
Comparison of green and eyed egg incubation
success. It was initially assumed that the survival rate of green and eyed eggs at the same
stations will be different because of the sensitivity criterion. The results of Univariate
Analysis of Variance partially prove the assumption (Table 3). In particular, differences
between the groups were significantly higher than within the groups in Station N1 but
not in Station N2. This means that the environmental conditions almost equally influence
the hatching of eggs in both stages in Station N2. Considering the low success rates of
hatching there, it could be concluded that the area is not well suited for the proposed
method in general.
Water 2023,15, 3993 8 of 12
Table 3.
The results of univariate analysis of variance between green and eyed eggs in the wild.
Survival rates are dependent parameter and green/eyed eggs are independent.
Station Egg Type Mean SE F Value Sig.
N1 Green 44.67 7.75 19.94 0.000
Eyed 88 2.02
N2 Green 28.47 9.87 2.21 0.151
Eyed 50.8 10.89
Note: F value is significant at the <0.05 level.
Fluctuations in abiotic parameters. In general, Station N1 has the most stable conditions
for all the measured parameters except oxygen saturation, while Station N3 has the most
unstable conditions (Supplementary Table S1). Moreover, all monitored parameters were at
the margins of optimum during the experiments besides temperature (Table 4).
Table 4. The ranges of chosen abiotic parameters in the wild and the farm.
Station Ranges of Values of Abiotic Parameters and Standard Deviations (SD)
pH T (◦C) DO (mg/L) DO (%) Depth (cm)
N1 (n = 30) 8–8.2 (SD = 0.05) 5.7–8.5 (SD = 0.96) 11.3–12.2 (SD = 0.26) 90–104 (SD = 3.7) 16–20 (SD = 1.45)
N2 (n = 30) 8.4–8.6 (SD = 0.8) 3.8–7.4 (SD = 1.35) 11.2–12.8 (SD = 0.55) 92–102 (SD = 3.29) 36–38 (SD = 0.62)
N3 (n = 6) 8.4–8.5 (SD = 0.46) 0.2–3.5 (SD = 1.22) 14–14.2 (SD = 0.1) 96–104 (SD = 2.49) 18–26 (SD = 3.6)
Farm (n = 135) 7(SD = 0.0) 8.3–9.5 (SD = 0.17) 8.5–9 (SD = 0.08) - 12 (SD = 0.0)
Optimum
conditions 6–9 5–12 8–15 80–120 15–50
Note: n—number of measurements during the experiment.
For instance, during 15 visits to the experimental sites, the optimum temperatures
were never exceeded in Station N1, were exceeded seven times in Station N2 and were
always exceeded in Station N3 (Supplementary Table S1). The results of the general linear
mixed model analysis show that the null hypothesis of equality of error variance of the
dependent variable across groups is rejected for pH, DO (mg/L) and depth parameters.
Thus, the results of Tukey’s HSD are not accurate for them. Therefore, Tukey’s test was
performed only for temperature and DO (%) parameters (Table 5), and revealed significant
differences only in terms of the temperature parameter.
Table 5.
The results of Tukey’s test for the significance of mean differences of abiotic parameters
between wild stations. The results of parameters that reject the null hypothesis of equal variances
checked by Levene’s test are shown.
Dependent Variable (I) Station (J) Station Mean Difference (I-J) SE Sig.
Temperature
Station N1 Station N2 1.26 *
0.432 0.017
Station N1 Station N3 5.80 *
0.518 0.000
Station N2 Station N3 4.54 *
0.518 0.000
DO%
Station N1 Station N2 0.20
1.217 0.985
Station N1 Station N3 −2.38
1.459 0.245
Station N2 Station N3 −2.58
1.459 0.194
Note: * The mean difference is significant at the <0.05 level.
However, considering time lag between visits, the expected negative impact on the
incubation should be rather small in Station N1 but high in Stations N2 and N3. Also, the
gap between optimal and registered average temperatures for Station N1 was only 0.5
◦
C
and from 0.2 to 1.2
◦
C (SD
±
0.38) in Station N2. Such a situation with the temperature
conditions also led us to state that the results of theoretical assessment for the remaining
upstream parts of the Masrik River system were also valid. When comparing the results
for the wild with the farm conditions, it was obvious that the biggest gap in all parameters
was in Station N3 and relatively closer conditions were in Station N1.
Water 2023,15, 3993 9 of 12
Correlations with hatching success. Because it has already been shown that only tem-
perature and DO (%) parameter effects on hatching could be considered confident, the
correlation analysis (Table 6) was also performed for these two parameters only.
Table 6. The results of nonparametric correlation.
Parameter
Station N1 Station N2
Mortality Rate (%)
of Green Eggs
Mortality Rate (%)
of Eyed Eggs
Mortality Rate (%)
of Green Eggs
Mortality Rate (%)
of Eyed Eggs
TCorrelation 0.930 ** 0.997 ** −0.414 0.888 **
N 15 10 15 10
DO (%) Correlation −0.070 −0.963 ** 0.677 ** −0.522
N 15 10 15 10
Note: ** Correlation is significant at the 0.01 level (2-tailed).
The results show the existence of strong links between tested parameters and hatching
success. Particularly, it could be derived that temperature changes correlate strongly with
the mortality of both green and eyed eggs in Station N1 but only with the mortality of eyed
eggs in Station N2. Meanwhile, oxygen saturation correlates well with the mortality of
green eggs in Station N1 and with the mortality rate of eyed eggs in Station N2. However,
because the DO parameter was always in the optimum range, only the temperature was
influential. Such results do not neglect the hypothesis that other environmental conditions
could have a strong effect on the natural reproduction success of trout of Lake Sevan in
the rivers.
4. Discussion
Environmental factors
. The results of abiotic factors prove that the temperature
conditions generally remain the main constraining factor for the successful growth of eggs
in the Masrik River as was also shown during the theoretical assessment [
26
]. In particular,
compared with Station 1 where temperatures remained in the optimum range throughout
all the experiment, there was less success in hatching in Station 2 for both green and eyed
eggs. For the failure of hatching in Station 3, the temperature factor was decisive. The
results of analyses also revealed that various environmental parameters or features should
be involved in the theoretical assessment. Particularly, solid waste and organic matter were
regularly covering nests in Stations N1 and N2. This issue is regularly reported [
22
] but no
action from the authorities to prevent solid waste dumping directly into the rivers. If not
maintained properly, this issue will lead to closure of pores and reduced aeration in the
nest or even their destruction.
For the particular case of the Masrik River, it seems meaningful to also use a criterion
of a suspension level, because a fast stream current is eroding away significant amount of
clay from the floodplain [
31
] which blocks the pores of WVB and we assume will do the
same with the pores of the nests.
One more feature to consider is the wind chill effect on water temperatures [
32
].
Because most of the tributaries of Lake Sevan flow through gorges and V-shape valleys in
their upper and middle course parts [
33
], obviously wind chill significantly contributes to
early drop of temperatures, which makes such areas unsuitable for gegharkuni’s spawning.
Benthic macroinvertebrates were considered in the theoretical assessment, but only
from a perspective of food base for the fry and the indicators of ecological status. However,
as our observations show, gammarids are also predating eggs, which should be considered
in the theoretical assessment. Moreover, some zooplankton species are also known as
predators of fish eggs [
34
]. Thus, the next important step towards the restoration of
spawning rivers in Lake Sevan basin should reflect such concerns.
Hatching success.
Natural and Semi-Natural mortality. One of the most complicated
issues is the precise estimation of eggs’ hatching success in the wild. In general, it is
stated that the emergence of smolt of some salmonids (involving Rainbow and Brown
Water 2023,15, 3993 10 of 12
trouts) varies between 85 and 95% in hatcheries and between 1 and 5% in the wild [
35
].
According to Hoitsy et al. [
36
], the fertility rate of eggs of Brown trout—a close relative
of Sevan trout—is 95–100% in the farms while the hatching rate from green eggs is 90–
100% if incubated at 10
◦
C. Such results correspond to the experimental results from the
Karchaghbyur farm. However, natural mortality of Sevan trout eggs in the wild has not yet
been studied sufficiently. Some hyperlocal studies in this issue could be found in the works
dated back to the first half of 20th century. Particularly, Vladimirov has found a 7–65% yield
from green egg [
37
], while Fortunatov has found a 0–50% [
38
] yield in the Gavaraget River.
Vladimirov has also found a 64–93% yield of green eggs in the Karchaghbyur River [
37
].
On average, he estimates the success rate of the natural reproduction of gegharkuni in
tributaries of Lake Sevan at a level of 10%. However, considering the changes in the basin
of Lake Sevan during 20th century [
39
], these data are less relevant today. Moreover, the
Masrik River cannot be mechanically compared to the Karchaghbyur or Gavaraget rivers.
Thus, there are insufficient data to compare with our results. Moreover, our results cannot
be considered as natural mortality because we used some artificial incubation in man-made
nests thus we prefer to call the results semi-natural. In general, hatching rate in Station N1
corresponds to the estimations of Vladimirov. The success rate in Station N2 claims that
some natural reproduction there is still possible, but with a lower yield compared with
Station N1 and average survival rate given by the Vladimirov. At the same time, in the
current climatic conditions, the natural reproduction of Sevan trout is impossible in the
middle or the upper course parts of the Masrik River since mid-November given the low
temperatures that lead to freezing water.
Impact of different conditions. Considering the artificial conditions established for the
experiments, there should be some differences from natural hatching success rates in the
wild. More probably, natural reproduction would be more successful which means that it
still can happen in Station N2 but with a significantly lower success rate than in Station
N1. However, taking into account the issue with the high level of suspension, the overall
success rate of natural reproduction of Sevan trout in the Masrik River is expected to be
lower than that measured for the Gavaraget and Karchaghbyur rivers back in the 1940s [
37
].
Suspension in the Masrik River is able to block the pores significantly and lessen the
aeration and thus increase mortality [
40
]. Considering the results of analyses, we conclude
it meaningful to revise the “partly suitable” class of a river for natural reproduction of
Sevan trout to address the range of impacts of environmental conditions more precisely.
The effectiveness of proposed method for restocking Sevan trout
. The results of hatch-
ing rates for the experiments have just proven the general pattern of sensitivity of green
eggs towards environmental conditions is higher than in case of eyed eggs, e.g., [
27
,
41
,
42
].
However, the results of experiments for the different stretches of the Masrik River and
the results of almost similar experiments conducted in the Lichq River [
24
] have shown
obvious differences. In particular, the emergence of green eggs to the eyed eggs in the
Lichq River was about 34%, while it was 40% in Station N1 and 20% in Station N2 (please,
see the reading for the 12 June 2018 on Figure 2). In the eyed stage, the emergence at the
end of the experiment in Station N1 was 2% less and in Station N2 77% less than in the
Lichq River. Moreover, the eggs were treated in a different way in the farm and in the wild,
which further contributes to the drop of the success level in the wild experiments. Thus, if
treated more properly, the success rate can increase in the wild.
It can be concluded that the method of artificial incubation of eyed egg can be quite
effective in restocking Sevan trout population and a significantly cheaper solution compared
with restocking by smolt or reintroduction from Lake Issyk Kul. However, the effectiveness
of this method is low in the Masrik River given the environmental conditions and the
current state of water resource management in the area. Thus, further action on river
restoration is still necessary to conduct prior to introducing this method in a wider scale in
Lake Sevan basin. Considering the importance of conservation of an endemic and native
fish species in the world, this work also comes to show that egg incubation in the wild is
Water 2023,15, 3993 11 of 12
still an effective method for restocking salmonids in the rivers with limited human-induced
modifications in river regime and an ecological status.
Supplementary Materials:
The following supporting information can be downloaded at https:
//www.mdpi.com/article/10.3390/w15223993/s1. Supplementary Table S1: The measurements
spreadsheet with raw data.
Author Contributions:
All authors contributed to the study conception and design. Material prepa-
ration, data collection and analysis were performed by V.A., T.V. and M.D. The first draft of the
manuscript was written by V.A., T.V., N.B., M.D. and B.G., and all authors commented on pre-
vious versions of the manuscript. All authors have read and agreed to the published version of
the manuscript.
Funding:
This work was made possible by a research grant from the Armenian National Science
and Education Fund (ANSEF) based in New York, USA. This work was supported by the Science
Committee of RA (Research project No. 21T-1F208).
Data Availability Statement: All primary data are available from Supplementary Table S1.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Millennium Ecosystem Assessment. Ecosystems and Human Well-Being: Biodiversity Synthesis; World Resources Institute, Island
Press: Washington, DC, USA, 2005.
2.
Manel, S.; Guerin, P.-E.; Mouillot, D.; Blanchet, S.; Velez, L.; Albouy, C.; Pellissier, L. Global determinants of freshwater and
marine fish genetic diversity. Nat. Commun. 2020,11, 692. [CrossRef] [PubMed]
3.
Panlasiguia, S.; Davisa, A.J.S.; Mangiantea, M.J.; Darling, J.A. Assessing threats of non-native species to native freshwater
biodiversity: Conservation priorities for the United States. Biol. Conserv. 2018,224, 199–208. [CrossRef]
4.
Dudgeon, D.; Arthington, A.H.; Gessner, M.O.; Kawabata, Z.; Knowler, D.J.; Lévêque, C.; Naiman, R.J.; Prieur-Richard, A.H.;
Soto, D.; Stiassny, M.L.J.; et al. Freshwater Biodiversity: Importance, Threats, Status and Conservation Challenges. Biol. Rev.
Camb. Philos. Soc. 2006,81, 163–182. [CrossRef] [PubMed]
5.
Myers, N.; Mittermeier, R.A.; Mittermeier, C.G.; da Fonseca, G.A.B.; Kent, J. Biodiversity hotspots for conservation priorities.
Nature 2000,403, 853–858. [CrossRef]
6. Gabrielyan, B.K. An annotated checklist of freshwater fishes of Armenia. Naga ICLARM Q. 2001,24, 23–29.
7.
Pipoyan, S.K.; Malkhasyan, A.H. Ichthyofauna of Armenian state reserves and national parks. Biol. J. Armen.
2014
,66, 18–25.
(In Armenian)
8. Luzhin, B.P. The Issyk Kul Trout Gegarkuni; Izdatelstvo Akademii Nauk Kirgizskoy SSR: Frunse, Russia, 1956; 136p. (In Russian)
9.
Alpiev, M.; Sarieva, M.; Siriwardena, S.N.; Valbo-Jørgensen, J.; Woynárovich, A. Fish Species Introductions in the Kyrgyz Republic;
FAO Fisheries and Aquaculture Technical Paper No. 584; FAO: Rome, Italy, 2013; 108p.
10.
Wilkinson, I.P. Lake Sevan: Evolution, Biotic Variability and Ecological Degradation. In Large Asian Lakes in the Changing World;
Mischke, S., Ed.; Springer Water; Springer: Cham, Switzerland, 2020; pp. 35–63. [CrossRef]
11.
Levin, B.; Simonov, E.; Gabrielyan, B.; Mayden, R.; Rastorguev, S.; Roubenyan, H.; Sharko, F.; Nedoluzhko, A. Caucasian treasure:
Genomics sheds light on the evolution of half-extinct Sevan trout, Salmo ischchan, species flock. Mol. Phylogenet. Evol.
2021
,
167, 107346. [CrossRef]
12.
Khosrovyan, A.; Avalyan, R.; Atoyants, A.; Aghajanyan, E.; Hambaryan, L.; Aroutiounian, R.; Gabrielyan, B. Tradescantia-Based
Test Systems Can Be Used for the Evaluation of the Toxic Potential of Harmful Algal Blooms. Water 2023,15, 2500. [CrossRef]
13. Aghasyan, A.; Kalashyan, M. (Eds.) The Red Book of Animals of the Republic of Armenia; Zangak: Yerevan, Armenia, 2010; 368p.
14.
Gerasimov, Y.V.; Gabrielyan, B.K.; Malin, M.I.; Rubenyan, A.R. Long-term dynamics of fish reserves in Lake Sevan and their
current state. In Ecology of Lake Sevan during the Period of Water Level Rise. The Results of Joint Russian-Armenian Biological Expedition
on Hydroecological Observation of Lake Sevan (Armenia); Krylov, A., Ed.; Nauka DNC: Makhachkala, Russia, 2010; pp. 249–278.
(In Russian)
15. Gabrielyan, B.K. Fish of Lake Sevan; Gitutyun: Yerevan, Armenia, 2010; 252p. (In Russian)
16. Kalashian, M.; Aghababyan, K.; Zarikian, N.; Gabrielyan, B.; Arakelyan, M.; Ghazaryan, A. Fauna of Armenia. In Biodiversity of
Armenia; Fayvush, G., Ed.; Springer: Cham, Switzerland, 2023; pp. 165–282. [CrossRef]
17.
Bogdanowicz, W.; Rutkowski, R.; Gabrielyan, B.K.; Ryspaev, A.; Asatryan, A.N.; Mkrtchyan, J.; Bujalska, B.M. Fish introductions
in the former Soviet Union: The Sevan trout (Salmo ischchan)—80 years later. PLoS ONE 2017,12, e0180605. [CrossRef]
18. Hovhanissian, R.O. Lake Sevan Yesterday, Today, Tomorrow; Gitutyn: Yerevan, Armenia, 1994. (In Russian)
19.
Savvaitova, K.; Petr, T. Fish and Fisheries in Lake Issyk-Kul (Tien Shan), River Chuand Pamir Lakes; Petr, T., Ed.; FAO Fisheries
Technical Paper No. 385; Food and Agriculture Organization of the United Nations: Rome, Italy, 1999; pp. 168–186.
Water 2023,15, 3993 12 of 12
20.
Kuljanishvili, T.; Epitashvili, G.; Freyhof, J.; Japoshvili, B.; Kalous, L.; Levin, B.; Mustafayev, N.; Ibrahimov, S.; Pipoyan, S.;
Mumladze, L. Checklist of the freshwater fishes of Armenia, Azerbaijan and Georgia. J. Appl. Ichthyol.
2020
,36, 501–514.
[CrossRef]
21.
Pavlov, D.S.; Savvaitova, K.S.; Markaryan, V.G. Ecological monitoring and the potential for saving trout Salmo ischchan (Kessler,
1877) in Lake Sevan. Vestn. Mosk. Univ. Ser. Biol. 1989,1, 73–78. (In Russian)
22.
Asatryan, V.L.; Dallakyan, M.R. Assessment of seasonal differences of ecological state of lotic ecosystems and applicability of
some biotic indices in the Basin of Lake Sevan (Armenia): Case study of Masrik River. Water Sci. Technol. Water Supply
2018
,19,
1238–1245. [CrossRef]
23.
Vardanyan, T.V.; Barseghyan, N.E. Reasons for the changes in morphoecological parameters of Sevan trout. Biol. J. Armen.
2015
,
67, 80–84. (In Russian)
24.
Barseghyan, N.; Vardanyan, T.; Asatryan, V.; Dallakyan, M.; Arakelyan, A.; Mamikonyan, V. Application of new method of
artificial incubation of Sevan trout (Salmo ischchan, Kessler) eggs in natural condition of tributaries of lake Sevan. Vestn. Astrakhan
State Tech. Univ. Ser. Fish. Ind. 2020,1, 103–110. [CrossRef]
25.
Vardanyan, T.V.; Barseghyan, N.E.; Aghasyan, A.L. Comparative analysis of some morph-metric and plastic parameters of Sevan
trout from the lakes Issyk-Kul and Sevan. Biol. J. Armen. 2019,71, 37–42. (In Russian)
26. Asatryan, V.; Dallakyan, M.; Vardanyan, T.; Barseghyan, N.; Gabrielyan, B. An integrative evaluation of suitability of a river for
natural reproduction of trout of Lake Sevan (Armenia). Environ. Sci. Pollut. Res. 2020,27, 6352–6361. [CrossRef] [PubMed]
27.
Johnson, J.H. Comparative survival and growth of Atlantic salmon from egg stocking and fry releases. N. Am. J. Fish. Manag.
2004,24, 1409–1412. [CrossRef]
28.
Zinchenko, T.D.; Golovatyuk, L.V.; Harutyunova, L.D.; Kachvoryan, E.A. Composition, structure of bottom-dwelling communities
and water quality assessment of the small rivers of lake Sevan (the Masrik and Argichi rivers). Izv. Samara Sci. Cent. Russ. Acad.
Sci. 2013,15, 203–212.
29. Dadikyan, M.G. Fishes of Armenia; AS of ArmSSR: Yerevan, Armenia, 1986; 245p. (In Russian)
30. Savvaitova, K.; Dorofeeva, E.; Markaryan, V.; Smoley, A. Trout of Lake Sevan. Ann. Zool. Inst. 1989,204, 180. (In Russian)
31.
Heblinski, J.; Schmieder, K.; Heege, T.; Kwaku Agyemang, T.; Sayadyan, H.; Vardanyan, L. Mapping of water constituents in high
mountainous Lake Sevan (Armenia). SIL Proc. 1922–2010 2010,30, 1453–1455. [CrossRef]
32.
Schlatter, T.W. Wind Chill. In Encyclopedia of World Climatology; Oliver, J.E., Ed.; Encyclopedia of Earth Sciences Series; Springer:
Dordrecht, The Netherlands, 2005. [CrossRef]
33. Mnatsakanyan, B.P. Lake Sevan Basin (Nature, Climate and Water); Asoghik: Yerevan, Armenia, 2007; 190p. (In Armenian)
34.
Woynarovich, E.; Horváth, L. The artificial propagation of warm-water finfishes—A manual for extension. FAO Fish. Tech. Pap.
1984,201, 183. Available online: http://www.fao.org/3/ac742e/AC742E05.htm (accessed on 10 August 2023).
35.
Reisenbichler, R.R.; Rubin, S.P.; Wetzel, S.; Phelps, S. Natural selection after release from a hatchery leads to domestication in steel
head, Oncorhynchus mykiss. In Stock Enhancement and Sea Ranching: Developments, Pitfalls and Opportunities; Leber, K.M., Kitada,
S., Blankenship, H.L., Svåsand, T., Eds.; Blackwell Science: Oxford, UK, 2004; pp. 371–383.
36. Hoitsy, G.; Woynarovich, A.; Moth-Poulsen, T. Guide to the Small Scale Artificial Propagation of Trout; The FAO Regional Office for
Europe and Central Asia: Budapest, Hungary, 2012; Available online: http://www.fao.org/3/ap341e/ap341e.pdf (accessed on
16 July 2023).
37.
Vladimirov, V.I. The study of biology of fingerlinks and breeding of gegharkuni trout (Salmo ischchan gegarkuni, Kessl. Biotypus
α
fort.). Ann. Sevan Hydrobiol. Stn. 1940,8, 87–118. (In Russian)
38. Fortunatov, M.A. Fishery in Lake Sevan. Ann. Sevan Stn. 1927,1, 131. (In Russian)
39.
Gabrielyan, B.; Khosrovyan, A.; Schultze, M. A review of anthropogenic stressors on Lake Sevan, Armenia. J. Limnol.
2022
,
81, 2061. [CrossRef]
40.
Soulsby, C.; Youngson, A.; Moir, H.; Malcolm, I. Fine sediment influence on salmonid spawning habitat in a lowland agricultural
stream: A preliminary assessment. Sci. Total Environ. 2001,265, 295–307. [CrossRef]
41.
Kelly-Quinn, M.; Tierney, D.; Bracken, J.J. Survival of salmon, Salmo salar, eggs planted in upland streams. Aquac. Fish. Manag.
1993,23, 791–796. [CrossRef]
42.
Wagner, E.J.; Arndt, R.E.; Roubidoux, R. The effect of temperature changes and transport on cutthroat trout eggs soon after
fertilization. N. Am. J. Aquac. 2006,68, 235–239. [CrossRef]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.