Environmental Factors Affecting the Phytoplankton Composition in the Lake of Tibetan Plateau

Abstract

Due to the high altitude, unique geographical location, difficult accessibility and low temperature, the environmental factors influencing phytoplankton composition have rarely been investigated in the Selin Co Lake, which is the largest lake in the Tibetan Plateau. Phytoplankton composition can indicate aquatic ecosystem conditions, which may be sensitive to environmental factors in the Tibetan Plateau. In this study, we investigated the main environmental factors that influence phytoplankton species in the Selin Co Lake by analyzing the spatial distribution and applying statistical analyses. We also compared the influential environmental factors in this lake with other lakes around the world. The results suggest that the eleven environmental variables can explain about 46.78% of the phytoplankton’s composition. DO and fluoride were the most significant environmental variables, followed by arsenic and COD, and the other variables had comparatively smaller and more insignificant influences on phytoplankton composition. There were five dominant phytoplankton species in the Selin Co Lake, namely, Microcystis sp., Navicula spp., Chlorella vulgaris, Ankistrodesmus falcatus, and Westella sp. Some of these dominant species were also found in other tropical lakes, suggesting that the phytoplankton community could adapt to environmental changes. A clear understanding of the influential environmental variables affecting phytoplankton composition could help us to make proper water quality protection strategies in future climate change scenarios.

1. Introduction

Phytoplankton is one of the most important primary producers in lakes and reservoirs. It can influence energy transfer, the circulation of biological materials, and the absorption and release of carbon, nitrogen, and phosphorus in the ecosystem [1]. Phytoplankton plays an important role in maintaining the stability and integrity of aquatic ecosystems [2]. As they are at the base of aquatic food webs, phytoplankton composition, abundance, and biomass are widely used to identify water quality [3,4]. There has been a lot of research on the relationships between phytoplankton composition and the physico-chemical characteristics of inland lakes [5,6,7,8], while such research in the alpine region is rare. A former study in southeast Fujian in China found that algae communities responded relatively quickly to changing conditions, and were sensitive to a localized or recent disturbance [9]. The availability of light and nutrients influence phytoplankton standing biomass, community structure and stage of succession [10]. The abundance and types of algae in different lakes can represent their community structure, which in turn determines diversity indices [11]. The phytoplankton community is affected by many factors, such as climate conditions, nutrient loads, and seasonal variations, and the diversity indices change accordingly as the phytoplankton community changes.

Phytoplankton species composition and the factors influencing them in alpine mountainous lakes have rarely been investigated, due to harsh meteorological conditions, high altitudes, and the remoteness and inaccessibility of the lakes. The Selin Co Lake is the largest lake in the Tibetan Plateau, with an average elevation of higher than 4000 m. Some studies have investigated lake water quality in terms of water chemistry in the Tibetan Plateau [12,13,14,15]. The phytoplankton in the nearby lakes mainly included Bacillariophyta, Cyanophyta, and Chlorophyta [16]. The absence of human activities surrounding the Selin Co Lake [17] makes it a valuable natural laboratory for the study of relationships between phytoplankton communities and water quality.

In this study, we thoroughly investigated the influencing environmental factors affecting phytoplankton composition in the Selin Co Lake by analyzing the spatial distribution and applying redundancy analysis. We also compared the distributions of dominant phytoplankton species and their influencing factors in other lakes around the world. The results can help us to better understand the evolution of phytoplankton composition in the Selin Co Lake, and thus make strategies to prevent the possible occurrence of algal blooms under future climate warming scenarios.

2. Materials and Methods

2.1. Study Area

The Selin Co Lake is in the central Tibetan Plateau (88.5–89.4° E, 31.5–32.1° N) (Figure 1), and it covers a water area of 2389 km² (June 2017) and a drainage area of 45,530 km² [18]. The total lake volume of the Tibetan Plateau increased by 102.64 km³ during 1976–2013, with an average annual rate of 2.77 km³ year⁻¹ [19]. The lake area of the Selin Co Lake surpassed that of Nam Co in 2014 and it became the largest lake on the Tibetan Plateau [20]. The main factor driving the expansion of Selin Co Lake’s area is the increase in non-glacial runoff caused by precipitation, with the weakening of lake evaporation as a second factor. The land cover types of the Selin Co basin include forest, grassland, water, built-up land, and other non-utilized land. Grassland is the most important cover type, accounting for 83.82% of the total watershed area, while the water area is 9.44%. The lake area increased by 48.51 km² during 1990–2015, inundating grassland and saline–alkali land. Animal husbandry is the main form of land use in this area [18]. The annual average precipitation is approximately 315 mm, and this mainly occurs from May to September [21]. There are several rivers in the plateau that flow into the lake, providing water and sediment to the lake. The Zagen Zangbo River to the west is the largest endorheic river on the Tibetan Plateau, with a length of 355 km. The Zajia Zangbo to the north is the longest (409 km) endorheic river in the Tibetan Plateau. The Ali Zangbo and Boqu Zangbo are located in the southwest and the east, with total lengths of 245 km and 85 km, respectively. The climate is affected by the South Asia monsoon [22]. Besides river inflow, rainfall and melting glaciers or snow also recharge the lake water [23].

2.2. Field and Laboratory Analyses

We conducted fieldwork in July 2023 to collect water samples and to perform phytoplankton investigations in Selin Co Lake. There were 21 sampling points for collecting and analyzing the lake water quality and phytoplankton species. Water pH, dissolved oxygen (DO, in mg/L), water temperature (°C), electrical conductivity (EC, in µS/cm), total dissolved solids (TDS, in mg/L), and salinity (Salt, ppt) were measured in situ with a portable multi-parameter water quality analyzer (YSI ProQuatro, Columbus, OH, USA). Water transparency was estimated by Secchi depth (SD) using a 30 cm-diameter Secchi disc. Total nitrogen (TN) and total phosphorus (TP) were measured using unfiltered water samples. Samples for the measurement of chemical oxygen demand using potassium permanganate as the oxidizing agent (COD_Mn) were stored in polypropylene bottles after filtration and measured in the National Engineering Laboratory for Lake Pollution Control and Ecological Restoration, Chinese Research Academy of Environmental Sciences. According to discussions with a local environmental agency, levels of fluoride and arsenic in the lake water have increased in recent years, so we measured their concentrations in the lake water. The fluoride (F) was measured with ion chromatography (Integrion ICS-5000, Thermo, Waltham, MA, USA). Arsenic (As) speciation was performed using high-pressure liquid chromatography (HPLC: Thermo Scientific Dionex UltiMate 3000; Thermo, Waltham, MA, USA) with an anion exchange column. After chromatographic separation, column effluent was pumped into the inductively coupled plasma mass spectrometer (ICP-MS) for arsenic calculation [24]. The measured samples all have units in mg/L.

2.3. Phytoplankton Collection and Identification

We collected phytoplankton samples using a 1-L polymethyl methacrylate sampler at 0.5 m below the water surface [25,26]. The samples were stored in bottles and fixed with 1.5% Lugol’s iodine solution [8]. Phytoplankton species assessment and quantitative phytoplankton analysis were completed using microscopy [27]. The taxonomy and nomenclature information of the phytoplankton was updated based on the AlgaeBase (https://www.algaebase.org; accessed on 1 March 2024) and [28]. The biovolumes were estimated using a geometric shape and mathematical equation for each taxon unit [29]. The phytoplankton biomass (mg/L) was estimated through biovolumes [8].

2.4. Data Analyses

2.4.1. The Trophic Level Index

The trophic level index (TLI) was calculated using TP, TN, SD, and COD_Mn [30]. Chlorophyll-a was not measured in the Selin Co Lake. The formulae are as follows:

$$\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathrm{T}\mathrm{P}\right)=10\times [9.436+1.624\ln \left(\mathrm{T}\mathrm{P}\right)]$$

(1)

$$\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathrm{T}\mathrm{N}\right)=10\times [5.453+1.694\ln \left(\mathrm{T}\mathrm{N}\right)]$$

(2)

$$\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathrm{S}\mathrm{D}\right)=10\times [5.118-1.94\ln \left(\mathrm{S}\mathrm{D}\right)]$$

(3)

$$\mathrm{T}\mathrm{L}\mathrm{I}\left({\mathrm{C}\mathrm{O}\mathrm{D}}_{\mathrm{M}\mathrm{n}}\right)=10\times [0.109+2.661\ln \left({\mathrm{C}\mathrm{O}\mathrm{D}}_{\mathrm{M}\mathrm{n}}\right)]$$

(4)

TLI was calculated as

$$\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)={\sum }_{\mathrm{j}=1}^{\mathrm{m}}{\mathrm{w}}_{\mathrm{j}}\times \mathrm{T}\mathrm{L}\mathrm{I}\left(\mathrm{j}\right)$$

(5)

$${\mathrm{w}}_{\mathrm{j}}={\mathrm{r}}_{\mathrm{i}\mathrm{j}}^2/\left({\sum }_{\mathrm{j}=1}^{\mathrm{m}}{\mathrm{r}}_{\mathrm{i}\mathrm{j}}^2\right)$$

(6)

where TLI(j) is the trophic level index of j, ${\mathrm{w}}_{\mathrm{j}}$ is the corresponding weight, ${\mathrm{r}}_{\mathrm{i}\mathrm{j}}$ is the specific coefficient for each parameter (TP, 0.84; TN, 0.82; SD, 0.83; COD_Mn, 0.83), and m is the number of indicators.

The water body trophic level is classified into five categories according to $\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ values, which range from 0 to 100. Trophic levels correspond to oligotrophic ($\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ < 30), mesotrophic (30 ≤ $\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ ≤ 50), lightly eutrophic (50 < $\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ ≤ 60), moderately eutrophic (60 < $\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ ≤ 70), and hypertrophic ($\mathrm{T}\mathrm{L}\mathrm{I}\left(\mathsf{\Sigma }\right)$ > 70) [31].

2.4.2. Phytoplankton Species Diversity

The species diversity of phytoplankton is estimated using the Shannon–Wiener diversity index (H), where $\mathrm{H}=-\sum \left({\mathrm{n}}_{\mathrm{i}}/\mathrm{N}\right)\times \ln \left({\mathrm{n}}_{\mathrm{i}}/\mathrm{N}\right)$. The Pielou’s evenness index (J) is calculated as $\mathrm{J}=\mathrm{H}/\ln \left(\mathrm{S}\right)$. In the formula, i equals $\mathrm{i}=1$ to S, with S being the number of groups, where ${\mathrm{n}}_{\mathrm{i}}$ is abundance of the i-th group, and N is the total abundance [32]. The Margalef richness index is calculated as $\mathrm{D}=(\mathrm{S}-1)/\ln \left(\mathrm{N}\right)$ [33].

The dominance value of each species was calculated using Equation (7),

$$\mathrm{Y}=\left({\displaystyle \frac{{\mathrm{n}}_{\mathrm{i}}}{\mathrm{N}}}\right)\times {\mathrm{f}}_{\mathrm{i}}$$

(7)

where ${\mathrm{n}}_{\mathrm{i}}$ is the number of individuals of species i at each sampling location, N is the total number of individuals of all species in the sampling time, ${\mathrm{n}}_{\mathrm{i}}/\mathrm{N}$ represents the relative proportion of each species, and ${\mathrm{f}}_{\mathrm{i}}$ is the occurrence frequency of species i among all sampling locations [34].

2.4.3. Statistical Analyses

The Pearson’s correlation coefficient (r) between the environmental variables was calculated. A significant correlation was defined as having a p-value of less than 0.05.

The detrend correspondence analysis (DCA) or redundancy analysis (RDA) was used to analyze the relationship between environmental factors and phytoplankton groups [35]. When the length of the sorting axis was less than three, the relationship between phytoplankton groups and environmental factors was investigated by RDA analysis [36]. The environmental variables were used as the predictor variables and were standardized before the RDA analysis. The abundances of phytoplankton groups were used as the response variables and were transformed using the Hellinger’s distance method [37]. Co-linearity between environmental variables was identified using variance inflation factors (VIF), and the redundant environmental variables with VIF > 10 were removed before the RDA. The ability of the environmental variables to explain the variance in species data in the RDA was tested using Monte Carlo simulation under 999 permutations. Variables were significant when the p-value was less than 0.05. All analyses were performed in the R software version 4.3.3 using the “corrplot”, “vegan” and “Hmisc” packages [38,39].

3. Results

3.1. Characteristics of Environmental Parameters

The statistics of the physico-chemical parameters of the Selin Co Lake are presented in

Table 1. Statistics of physico-chemical parameters of Selin Co Lake.

Variables	Minimum	Maximum	Mean ± Std
Water temp. (°C)	12.80	14.87	13.68 ± 0.52
Water depth (m)	9.60	50.19	27.84 ± 12.09
DO (mg/L)	2.11	5.37	3.82 ± 1.17
TDS (mg/L)	6984.65	8011.21	7750.21 ± 300.66
pH	9.35	9.70	9.54 ± 0.09
Water transparency (m)	1.50	4.30	3.10 ± 0.87
Salinity (ppt)	6.10	7.10	6.85 ± 0.29
EC (µS/cm)	8339.30	9812.71	9344.76 ± 377.96
TP (mg/L)	0.01	0.03	0.02 ± 0.01
TN (mg/L)	0.13	0.61	0.34 ± 0.13
COD (mg/L)	3.30	7.89	6.09 ± 1.19
TLI	32.17	43.11	36.93 ± 3.15
Fluoride (mg/L)	0.58	0.93	0.76 ± 0.10
Arsenic (µg/L)	55.72	104.99	83.90 ± 13.30

Note: Std represents standard deviation.

. The spatial distribution of environmental parameters in the lake is presented in Figure 2. The water depth was deepest in the mid-south area of the lake (S19 was 50.19 m), and it was shallowest in the northern part of the lake (S5 was 9.60 m). The water depth decreased from the mid-south and mid-west to the north and east of the lake. The average depth of lake water was 27.84 m. Water transparency was highest in the south of the lake (S20 and S21 were 4.3 m), and it was lowest in the north of the lake (S5 was 1.5 m). The average transparency of the lake water was 3.1 m. The north of the lake had the least transparency, while the east and south of the lake had the highest water transparency. Dissolved oxygen was highest in the west (S2 was 5.37 mg/L), and decreased to the lowest in the east and mid-south regions of the lake (S19 was 2.11 mg/L). The average DO value was 3.82 mg/L. Water temperature was highest in the southwest part of the lake (S3 was 14.87 °C), and it decreased to its lowest in the eastern part of the lake (S16 was 12.8 °C). The average temperature of the lake water was 13.68 °C. The TN concentration of lake water was highest in the west of the lake (S1 was 0.61 mg/L), and it decreased to lowest in the northeast of the lake (S8 was 0.13 mg/L). The average TN concentration in the lake was 0.34 mg/L. The TP concentrations were highest in the mid-west and mid-south parts of the lake (S2, S4, and S19 were 0.03 mg/L), and were lowest in the west, middle, and east of the lake (S1, S13, and S17 were 0.01 mg/L). The average TP concentration in the lake was 0.02 mg/L. The TP and TN concentrations of lake water were within the Grade III level according to the China Surface Water Environmental Quality Standard (0.05 mg/L and 1.0 mg/L, respectively). The COD_Mn was highest in the southwestern part of the lake (S3 was 7.89 mg/L), and decreased to the lowest in the mid-northern part of the lake (S9 was 3.30 mg/L). The average COD_Mn was 6.09 mg/L. Trophic level indices were high in the west and mid-north of the lake (S6 was 43.11), and they decreased to their lowest in the east of the lake (S17 was 32.17). The average TLI of lake water was 36.93. TDS was high in the southeast of the lake (S21 was 8011.16 mg/L), and it decreased to its lowest in the north of the lake (S5 was 6984.65 mg/L). The average TDS of lake water was 7750.21 mg/L. The salinity was highest in the south and southeast of the lake (S21 was 7.1 ppt), and decreased to the lowest in the north of the lake (S5 was 6.1 ppt). The average salinity of lake water was 6.85 ppt. The arsenic content was highest in the northeast of the lake (S12 was 104.99 µg/L), and was lowest in the west and east of the lake (S17 was 55.72 µg/L). The average arsenic content was 83.90 µg/L. Fluoride content was highest in the east of the lake (S15 was 0.93 mg/L), and was lowest in the northeast, west, and south of the lake (S8 was 0.58 mg/L). The average fluoride content was 0.76 mg/L.

The correlation heatmap of the environmental variables in the Selin Co Lake is presented in Figure 3. TN was significantly positively correlated with COD_Mn, DO, and TLI. COD_Mn was significantly positively correlated with fluoride, EC, and TLI. Fluoride and arsenic showed weakly negative correlations with DO. Water temperature was significantly positively correlated with DO, and was weakly correlated with EC and TLI. DO showed weakly negative correlations with water transparency and water depth. TDS was significantly positively correlated with water transparency, water depth, salinity, and EC. Water transparency was significantly positively correlated with water depth, salinity and EC, and was weakly negatively correlated with TLI. Water depth was significantly positively correlated with salinity.

3.2. Phytoplankton Species Composition and Diversity

A total of 39 phytoplankton species were identified, belonging to six groups (Cyanophyta, Cryptophyta, Bacillariophyta, Euglenophyta, Chlorophyta, and Pyrrophyta). The phytoplankton abundance at each sampling point ranged from 5.86 × 10⁵ cells/L to 2.71 × 10⁶ cells/L. The average abundances of Euglenophyta, Pyrrophyta, Crytophyta, Bacillariophyta, Chlorophyta, and Cyanophyta were 4.49 × 10³ cells/L, 1.10 × 10⁴ cells/L, 2.92 × 10⁴ cells/L, 1.26 × 10⁵ cells/L, 2.71 × 10⁵ cells/L, and 1.30 × 10⁶ cells/L, respectively. The phytoplankton biomass at each sampling point ranged from 0.54 mg/L to 1.92 mg/L, with an average value of 1.01 mg/L. The average biomasses of Euglenophyta, Crytophyta, Cyanophyta, Chlorophyta, Pyrrophyta, and Bacillariophyta at the sampling points were 0.021 mg/L, 0.033 mg/L, 0.058 mg/L, 0.115 mg/L, 0.165 mg/L, and 0.578 mg/L, respectively. Figure 4 presents the relative abundances and biomasses of various phytoplankton groups at each sampling point. The phytoplankton abundance was dominated by Cyanophyta at most of the sampling points, while at S11, it was dominated by Chlorophyta (relative abundance of 50%). The phytoplankton biomass was dominated by Bacillariophyta at most of the sampling points, while it was dominated by Pyrrophyta at sampling points S5, S11, S12, and S17. These four points were in the north, northeast, and east of the lake.

There were five dominant species in Selin Co Lake, including Microcystis sp., Navicula spp., Chlorella vulgaris, Ankistrodesmus falcatus, and Westella sp. The degrees of dominance of Microcystis sp., Navicula spp., Chlorella vulgaris, Ankistrodesmus falcatus, and Westella sp. were 0.727, 0.022, 0.041, 0.023, and 0.021, respectively. The five species belonged to Cyanophyta, Bacillariophyta, and Chlorophyta. The Shannon–Wiener’s H index of the phytoplankton community varied from 0.68 to 2.32, with an average value of 1.17, suggesting that the water is in a light eutrophic state. The standard refers to Table S1. The Pielou’s evenness index (J) provides another metric to assess the water eutrophication state. The Pielou’s J index of the phytoplankton community in this lake varied from 0.26 to 0.80, with an average value of 0.44. The D index ranged from 0.70 to 1.35, with an average value of 0.94, suggesting that the lake water was in the eutrophic state.

3.3. Relationships Between Environmental Factors and Phytoplankton Composition

The axis length of DCA1 was less than 3, meaning that RDA analysis is more appropriate than CCA analysis. A redundancy analysis (RDA) was applied to analyze the main factors contributing to phytoplankton composition in Selin Co Lake. EC and salt content were highly correlated with TDS. TLI was highly correlated with TN and COD_Mn. Thus, the three variables EC, salt, and TLI were removed from the following analysis. The remaining eleven environmental variables had VIF values below 10, and were applied as the predictor variables in the RDA analysis. The abundances of six phytoplankton groups, including Cyanophyta, Crytophyta, Bacillariophyta, Euglenophyta, Chlorophyta, and Pyrrophyta, were used as the response variables. S6 had no water depth data, while S6 and S13 had no water transparency data. As such, there were 19 points applied in the RDA analysis. The RDA ordination results for phytoplankton groups and environmental variables on axes 1 and 2 are shown in Figure 5. The eleven environmental variables can explain 46.78% of the phytoplankton composition. Among the explained environmental variables, the first and second RDA axes represent 59.34% and 18.3% of the variance explained by the RDA model, respectively. The results show that DO (r² = 0.3654, p < 0.05) and fluoride (r² = 0.3212, p < 0.05) had a significant effect on phytoplankton composition. The other environmental variables were not significant environmental variables.

4. Discussion

4.1. Spatial Distribution of Environmental Factors

The lake water was deepest in the mid-south and mid-west of the lake. Water transparency was also high in the middle to mid-south of the lake. Pearson’s correlation analysis showed that water depth was significantly positively correlated with water transparency in this lake. Variations in lake water transparency are related to changes in water optical components, including suspended matter, chlorophyll-a concentration, and fluorescent dissolved organic matter [17]. DO and water temperature decreased from the west to the east of the lake. DO and water temperature were also significantly positively correlated with each other. The water temperature was relatively low in this region, meaning that increased water temperature could exacerbate the dissolution of oxygen in this lake. TLI, TN and TP showed similar spatial patterns, as they generally decreased from the west to the east of the lake. COD_Mn decreased from the west and mid-east to the north and south of the lake. TLI had similar spatial patterns to TN and TP, which were high in the west and low in the east. TDS showed similar spatial patterns to salinity, with both decreasing from the southeast to the west and north of the lake. The high levels of arsenic and fluoride in the northeast and east of the lake may be due to surface erosion into the lake.

4.2. Phytoplankton Species Distribution in the Tibetan Plateau

The combined assessment of the water eutrophic state by TLI and diversity indices suggests that the lake water was in a mesotrophic to light eutrophic state. The middle of the lake had greater depth than the inshore areas of the lake. The abundances of phytoplankton were dominated by Cyanophyta at these sampling points, suggesting that water depth had no significant influence on phytoplankton species. The abundance of phytoplankton species was dominated by Cyanophyta at most of the stations, except for S11, which was dominated by Chlorophyta. The spatial distribution maps show that S11 had high water transparency, TDS, and salinity, and low DO, which may be the reason for the change in phytoplankton composition. The phytoplankton biomass was dominated by Bacillariophyta at most of the sampling points, except for four points in the north (S5), northeast (S11, S12), and east (S17) of the lake, where the phytoplankton biomass was dominated by Pyrrophyta. These four points had low DO, but no similar pattern of other environmental factors. Thus, DO may be one of the significant environmental factors that influenced phytoplankton species composition in this lake.

Former studies have shown that in lakes of the Selin Co basin, phytoplankton diversity was related to salinity and nutrient concentrations [18]. The salt tolerance of diatoms (which belong to Bacillariophyta) and their adaptation to low temperature facilitate their wide distribution in the Tibetan Plateau’s lakes [27]. Ref. [40] investigated the phytoplankton community in August 2015 and 2016 in a potential construction area of the Selin Co-Puruogangri Glacier National Park of Tibet, and found that the phytoplankton community’s composition mainly comprised of Bacillariophyta, Chlorophyta, and Cyanophyta. The overall Shannon–Wiener’s H index was 3.16, and Pielou’s J index was 0.84. The relatively high diversity indices suggest that the phytoplankton species were diverse and the phytoplankton community structure was stable. The construction area is mainly dominated by animal grazing, so the ecosystem is barely affected by human activities. Selin Co Lake is located in the southwestern part of the construction area, and so the harsh environment may make the ecosystem vulnerable. Ref. [16] investigated the phytoplankton compositions of 12 salt lakes in Naqu region, Tibet, during April to May 2009, and found that the species varied in the 12 lakes. In general, they belonged to Bacillariophyta, Cyanophyta, and Chlorophyta. The Shannon–Wiener’s H index varied from 0.44 to 2.50, while the Pielou’s J index varied from 0.25 to 1.0. Former studies have found that the phytoplankton species were characterized by Chlorophyta and Bacillariophyta in lakes on the Tibetan Plateau, while we found that Cyanophyta was also the main group. These phytoplankton groups belong to cold-water species, which can adapt to the low temperature, high salinity and high alkalinity.

4.3. The Main Influencing Environmental Factors of Phytoplankton Composition

The results of the RDA analysis show that DO and fluoride were the most significant environmental factors that influenced phytoplankton composition, followed by arsenic and COD_Mn. The other environmental factors had comparatively smaller influences on phytoplankton composition. We just considered abiotic factors in this study, while other biotic factors such as macrophytes also play an important role in influencing the phytoplankton community [41]. On the RDA plot, S2 and S9 are the closest to DO, and these two points showed higher DO values than the other sampling points. This may indicate that the phytoplankton composition is influenced by the most significant environmental factor. The closeness of the arrows suggests that the abundance of Bacillariophyta was closely related to the presence of fluoride. The phytoplankton groups were influenced by different factors due to their physiological characteristics. Lake Xingkai and Lake Xiaoxingkai are located in northeast China, where the climate is characterized by a temperate continental monsoon with below-freezing temperatures from October to April. They are in cold regions, and the phytoplankton species may be influenced by similar factors as those in Selin Co Lake [42]. The results show that TN, TP, and pH were the dominant factors influencing phytoplankton species composition in the Xingkai Lake basin.

Among the five dominant phytoplankton species, Microcystis sp. belongs to Cyanophyta. Microcystis growth remains unaffected in water with a salinity of up to 10 g/L, and their high salt tolerance may profit from rising salinities in freshwater ecosystems by gaining a competitive advantage over other freshwater phytoplankton [43]. Microcystis species (Cyanophyta) are commonly found in Lake Chaohu, which is located in the lower reaches of the Yangtze River system, indicating its wide range of water temperature tolerance [44]. Navicula spp. grows in an environment with appropriate nitrogen and phosphorus concentrations, while its growth and community structure may be negatively influenced when nitrogen and phosphorus contents are too high [45]. Chlorella vulgaris, Ankistrodesmus falcatus, and Westella sp. all belong to Chlorophyta. Chlorella vulgaris was found to be abundant in the rainy season in a tropical Brazilian reservoir, and it belongs to the eutrophic assignation that is related to eutrophic and shallow systems [46,47]. Chlorella vulgaris and Ankistrodesmus falcatus were the dominant species in all seasons in Changhu Lake, which is an important ecological zone in the middle and lower reaches of the Yangtze River [48] that was heavily eutrophicated during the investigated time period. It has been confirmed that a warm temperature is beneficial for the growth and reproduction of Bacillariophyta, Chlorophyta and Cyanophyta [49,50]. Westella sp. is not common in other lakes. A high TDS, as well as low temperature and high light intensity, resulted in poor phytoplankton assemblage. The variants of dominant species in Selin Co Lake, such as the variants of Microcystis sp., Navicula spp., Chlorella vulgaris, and Ankistrodesmus falcatus, were also found in tropical areas, suggesting that these phytoplankton species could adapt to both warm and cold environments. The abundance of phytoplankton increases with water temperature, such that warm lakes have more phytoplankton species and a greater species density, with an appropriate range. Some of the dominant phytoplankton species in the Selin Co Lake were also found in freshwater systems. Thus, the phytoplankton species in the Selin Co Lake may have originated in freshwater ecosystems initially, and after the evolution of environmental conditions, they underwent natural selection and survived. These species can endure high salinity and alkalinity [16].

5. Conclusions

In this study, we investigated the composition of phytoplankton species and the environmental factors influencing them in Selin Co Lake, which is the largest lake in the Tibetan Plateau. We have identified six phytoplankton groups and 39 phytoplankton species. The data show that phytoplankton abundance was dominated by Cyanophyta at most of the sampling points, while the phytoplankton biomass was dominated by Bacillariophyta at most of the sampling points.

The results of the RDA analysis suggest that the chosen eleven environmental variables can explain 46.78% of the phytoplankton species composition. The first and second RDA axes represent 59.34% and 18.3% of the variance explained by the predictor variables, respectively. The phytoplankton composition was significantly influenced by DO and fluoride. The other environmental factors had comparatively smaller influences on phytoplankton composition. There were five dominant species that were discovered, including Microcystis sp., Navicula spp., Chlorella vulgaris, Ankistrodesmus falcatus, and Westella sp. These dominant species were also found in other inland lakes, suggesting that the phytoplankton species could adapt to environmental changes. Our study could expand our knowledge of the spectrum of phytoplankton species and their preferred environments, from tropical to alpine lakes.

The results suggest that phytoplankton composition is affected by the close environments, thus manifesting spatial aggregations. This study performed one field campaign due to limited resources. Future research could incorporate technology such as remote sensing and unmanned aerial vehicles to obtain water quality parameters and aquatic biology data over a longer time scale, in order to investigate seasonal variations in phytoplankton composition and the environmental factors influencing it. These findings contribute to a wider understanding of phytoplankton evolution with environmental changes, especially in high-altitude lakes. Finally, we can formulate strategies for water quality protection in the context of a changing climate in the future.