Abstract In order to clarify the regular pattern of community succession of planktonic diatoms and the relationship between the diatoms and the physicchemical parameters， the species， abundance， and biomass of planktonic diatoms， and environmental factors in Baiguishan Reservoir （BGR） were investigated in Baiguishan Reservoir from February to November in 2016. Trophic state index （TLI） was used to evaluate the trophic state， and factors affecting the dynamics of planktonic diatom community were analyzed by Redundancy analyses （RDA）. The results showed that Baiguishan Reservoir belonged to mesotrophic water body， and all observed phytoplankton belonged to 69 genera of 25 families. The sampling points were divided into the same 2 groups in the clustering analysis based on environmental factors and Nonmetric MultiDimensional Scaling （NMDS） analysis， indicating that the community structure of planktonic diatoms had the same change trends with the changes of environmental factors. RDA showed that the main factors affecting community structure were total phosphorus （TP）， water temperature （WT） and the ratio of ammonia nitrogen to total phosphorus （NH3N/TP）. Therefore， TP was the most important factor influencing the dynamics of planktonic diatom community in mesotrophic reservoir.
　　Key words Planktonic diatom； Community structure； Baiguishan Reservoir； Redundancy analyses
　　The rapid hydrological changes in freshwater systems make it difficult to evaluate[1]. As an important primary producer in the water body， diatoms are sensitive to the response of physicochemical factors and have a short life cycle， so they can quickly respond to changes in the water ecosystem. The species composition and species diversity of diatom community also changes with the water body[2]. Therefore， diatoms are often used to monitor various water systems. The Water Frame Work Directive of the European Union recommended diatoms as an effective biological indicator that could be used to evaluate water nutrient levels in 2000[3].
　　The reservoir ecosystem system is complex and has attracted much attention due to its high primary productivity， water supply， and purification of sewage and other functions[4]. In reservoirs with different trophic degrees， the community structures of planktonic diatoms are also different correspondingly： in areas with great climate changes， water temperature has a significant effect on diatoms， and nutrient salts and their proportions also have a great influence on planktonic diatom community[5]. In mesotrophic reservoirs， there are few studies on the changes of planktonic diatom community. Therefore， the purpose of this paper was to study the annual changes of planktonic diatom community structure in Baiguishan Reservoir in the following ways： ① evaluating the water quality of a mesotrophic reservoir， ② using Redundancy Analysis （RDA） to study the changes of planktonic diatom community structure with temporal and spatial variation as well as the affecting factors， ③ exploring the relationship between dominant planktonic diatom species and major nutrients， so as to provide some theoretical bases for the protection of planktonic diatom diversity protection and the biological monitoring and protection of water environment in Baiguishan Reservoir. 　　Materials and Methods
　　Research area
　　Baiguishan Reservoir is located in the southwestern suburb of Pingdingshan City， Henan Province. It is an inland freshwater reservoir in the main stream of Shahe River of the Shayinghe River system in Huaihe River Basin. It is a drinking water source for local residents and a water source for industrial and agricultural use. Therefore， it is very important to monitor and protect the planktonic diatoms in Baiguishan Reservoir. In this study， focusing on the planktonic diatoms in Baiguishan Reservoir， 5 sampling sites were set up in the whole survey area， namely， S1 （33°43′5.50″N， 113°13′22.28″E）， S2 （33°44′0.40″N， 113°06′28.81″E）， S3 （33°43′29.90″N， 113°10′11.63″E）， S4 （33°45′5.39″N， 113°10′4.14″E） and S5 （33°44′13.96″N， 113°09′5.97″E）， which were respectively located in the east， west， south， north and central parts of Baiguishan Reservoir （Fig. 1）.
　　Sample collection and determination of physical and chemical indicators of water
　　From February to November 2016， the samples were collected at the end of each month according to the "Water Quality Sampling Technical Guidelines" （HJ 4942009）. Qualitative samples were slowly collected to 1 L on water surface in the "∞" shape using plankton net （mesh diameter of 0.064 mm）， and after fixed in Lugols solution， the samples were brought back to the laboratory for microscopic examination. For the collection of quantitative samples， 1 L of water was taken respectively at the 0.5 m and 1.5 m of the reservoir using a 1 L plexiglass water sampler， after mixing， the collected water was fixed in Lugols solution and then placed still for 24-36 h to precipitate， and then concentrated to 30 ml. The planktonic diatoms were quantified using a 0.1 ml phytoplankton counting box under a Nikon eclipse 80i microscope. Each sample was counted twice and each counting was done at about 100 fields. If the error of the 2 counting was more than 15%， a third time was repeated. The identification of planktonic diatoms was mainly based on the corresponding methods of Hu et al.[6]， while the planktonic diatom abundances and biomass calculations referred to Jin et al.[7].
　　Another water sample was taken to determine the physical and chemical indicators. Ammonium molybdate spectrophotometric method was used to measure total phosphorus （TP）， alkaline potassium persulfate digestion uV spectrophotometric method to measure total nitrogen （TN）， Nesslers reagent colorimetric method to measure ammonia nitrogen （NH3N）， Iodometric method to measure dissolved oxygen （DO）， acid process to measure permanganate （CODMn）， UV spectrophotometric method to measure chlorophyll a （Chl.a）； and water temperature （WT）， pH and transparency （SD） were measured on the spot. 　　Data statistics and analysis
　　First， samples with severe planktonic diatom damage in the water due to the large amount of sediment were removed， so the sampling sited with SD<70 were kicked off， thus， the samples collected in AprilSeptember at sampling site S2 were excluded.
　　BergerParker species dominance index （D）[8]， ShannonWiener diversity index （H′）[9]， Pielou species evenness index （J）[10]and trophic state index （TLI）[11]were used respectively to calculate the dominance degree of species， evaluate the planktonic diatom biodiversity， determine the water pollution status， and assess the degree of eutrophication according to the recorded methods. A species was determined as the dominant species if it had D>0.1 in at least 2 samples. The standards were as follows： H′>3， clean； 1-3， medium pollution； 0-1， heavy pollution； J> 0.5， clean； 0.3-0.5， medium pollution； 0-0.3， heavy pollution； TLI<30， oligotrophic； 30-50， mesotrophic； over 50， eutrophic.
　　In the SPSS 19.0， Pearson correlation analysis was used to analyze the correlation between physical and chemical factors and biological indicators.
　　Among the physical and chemical factors， those with values of more than 10 but less than 100 were square root converted， those more than 100 were logarithmically converted. The abundance of planktonic diatoms were all logarithmically converted for the following analysis because after converted， the data could provide more weights for the species with low abundance to show the changes in diatom community[12].
　　In software Primer 5， clustering analysis based on BrayCurtis similarity coefficient was used for the physicochemical factors TP， TN， NH3N， Chl.a， CODMn， and DO using the method of averaging between groups[13].
　　Redundancy Analyses （RDA） based on linear regression were used to study the effect of environmental factors on diatom composition changes[14]. In order to reduce the influence of lowabundance species on the ranking， the common species with dominance D>0.1 appearing in at least 1 sampling site were selected for numerical analysis. A total of 12 major diatom species were used as response variables for analysis. The physicochemical factors of the reservoir were used as the explanatory variables. There may be highly autocorrelated variables in the explanatory variables， so it is necessary to eliminate variables with an expansion coefficient greater than 20， and therefore， a total of 10 variables were selected for the analysis， namely， WT， pH， SD， TP， TN， NH3N， NH3N/TP， Chl.a， DO and CODMn）. Forward selection and Monte Carlo permutetinontest （P<0.05， n=999） were used to further select the significant variables explaining the change of diatom community. The above statistical analysis was done in the program CANOCO 4.5 version. 　　Results and Analysis
　　Community structure of planktonic diatoms in Baiguishan Reservoir
　　There were abundant planktonic diatoms in Baiguishan Reservoir. A total of 69 species of diatoms of 25 genera （including variants and mutant） were identified. Synedra acus took a certain advantage during the investigated time， which was the dominant species through the year except for AprilJune； Cyclotella meneghiniana was the dominant species from February to June； some of the other species also showed some dominance in some months （Table 1）.
　　In terms of abundance and biomass， in March， the increase in the abundance of planktonic diatoms was responsible for the massive growth of C. meneghiniana and S. acus， and the biomass was also slightly increased. The growth of C. meneghiniana and C. stelligera was the reason for the increase in the abundance in May， but due to the low wet weight of individuals of Cyclotella， the increase in biomass was not significant. In July-November 2016， S. acus and S. ulna had explosive growth， which was responsible for the dramatic rise in water abundance and biomass.
　　Assessment of water quality， biodiversity and their correlation
　　The average values of TLI， H′， and J were 45.07±2.00， 1.71±0.31， and 0.36±0.07， respectively. TLI did not change much yearround （Table 2）， indicating that the trophic level of the reservoir was relatively stable of mesotrophic level. Both the diversity index and evenness index indicated that Baiguishan Reservoir was in a light pollution state， and species were richer and more diverse from February to June， while the diversity was poorer from July to November.
　　Correlation analysis was performed to planktonic diatom biomass， dominant species abundance， index and physicochemical factors （WT， TP， NH3N， Chl.a， NH3N/TP， and TN/TP）. As shown in Table 3， TLI showed significant positive correlations with TP， TN and NH3N （respectively， r=0.362， P<0.05； r=0.388， P<0.01； r=0.387， P<0.01）， and significant negative correlation with TN/TP （r=-0.561， P<0.01）. The Shannon-Wiener diversity index was positively correlated with TP （r=0.381， P<0.05）. The biomass presented significant positive correlations with NH3N/TP and Chl.a （r=0.392， P<0.01； r=0.815， P<0.01）， and a significant negative correlation with TP （r=-0.659， P<0.01）. There was a significant positive correlation between C. meneghiniana and TP （r=0.306， P<0.05）， but it was negatively correlated with NH3N/TP （r=-0.350， P<0.05）. From July to November， the large growth of S. acus and S. ulna was found to have significant negative correlations with TP （r=-0.572， P<0.01 and r=-0.751， P<0.01）， and TN （r=-0.356， P<0.05 and r=-0.369， P<0.05）， but positively correlated with TN/TP （r=0.715， P<0.01； r=0.871， P<0.01， respectively）. 　　Cluster analysis and Nonmetric Multidimensional Scaling Analysis
　　Primer 5.0 was used to analyze the environmental variables of the 44 samples in Baiguishan Reservoir area， and the cluster analysis was performed using the method of averaging between groups based on the BrayCurtis similarity coefficient. The results show that the 44 sampling site in the cluster analysis could be divided into 2 groups. Group A included samples 1-22， falling in February-June 2016； the group B included samples 23-44， falling in July-November 2016 （Fig. 2）.
　　Redundancy analysis
　　The RDA results showed that the first coordinate axis and the second coordinate axis together accounted for 59% of the total variables （Table 4）， and the speciesenvironment correlation coefficient was high， indicating that planktonic diatoms had a strong relationship with environmental variables. As shown in Fig. 3a， the vectors of TP， Chl.a， NH3N/TP and TN/TP were relatively longer， and closer to the first coordinate axis； the vector of WT was also long and close to the second coordinate axis. As shown Fig. 3b， the samples were clearly segmented in a way that samples 1-22 fell in the second and third quadrants， while samples 23-44 in the first and fourth quadrants， which was consistent with the clustering results. In addition， the concentration of the nutrient salts showed great effects on the community structure of planktonic diatoms from February to June， when the concentration of TP and TN was the main factor affecting the community， while from July to November， the proportions of nutrient salts were the dominant factors.
　　Conclusion and Discussion
　　Water quality assessment and relationship between biological indicators in Baiguishan Reservoir
　　TLI is a commonly used method for evaluating water quality[11]， which evaluates water quality using 5 factors， including Chl.a， TP， TN， SD， and CODMn weighting. The evaluation results show that Baiguishan Reservoir was in the mesotrophic level through the monitoring period. However， from the perspective of the changes of nutrient salts， the level of nutrient concentration in the reservoir fluctuated greatly. From February to June， the concentration of TP averaged （0.22±0.08） ml， while from July to November， the average concentration was （0.02±0.01） ml. Therefore， even though TLI evaluation results indicate that the trophic level of the reservoir changes little， the community structure of planktonic diatoms varies greatly， which could inevitably affect the chlorophyll a content and planktonic diatom biodiversity and biomass. The Shannon-Wiener diversity index and evenness index indicate that the reservoir is in a mediumpollution state and needs further treatment， and it is also related to a small area of the reservoir surrounded by much agricultural land. 　　The correlation analysis shows that TP has significant correlations with biomass， chlorophyll a， Shannon-Wiener diversity index. Similarly， in RDA analysis， the TP vector is long， indicating that in the reservoir of this study， TP is one of the factors that significantly affects the change of planktonic diatom community. The change of TP concentration results in the change of dominant species and their abundance in the community， which in turn affects biomass， chlorophyll a， and ShannonWiener diversity index.
　　It has found that Chl.a represents the total biomass of phytoplankton to a certain extent[15]， and planktonic diatom abundance and biomass affect changes in Chl.a content. In the correlation analysis， there is a significant positive correlation between planktonic diatom biomass and chlorophyll a. In the results of RDA， the vector of Chl.a is also long， indicating that Chl.a has higher interpretation. The results suggest that planktonic diatoms are in the dominant position in the water body， which is also consistent with the predominance of planktonic diatoms in the mesotrophic lakes.
　　Community structure of planktonic diatoms and its spatial and temporal changes in mesotrophic reservoir
　　Environmental factors can directly affect the growth and reproduction of algae， and changes in water temperature and pH can stimulate the splitting speed of algae[16]. Nutrients play an important role in the growth and proliferation of algae[17]. In the study of Baiguishan Reservoir， the BrayCurtis similarity coefficient cluster analysis results of environmental variables divides the variables into Group A （February-June） and Group B （July-November）. In group A， the concentration of TP is relatively higher， and the number of planktonic diatom species was relatively abundant. Therefore， planktonic diatoms were predominant by some TPtolerant species， such as C. meneghiniana and Melosira granulate （Table 1）. Meanwhile， due to the low TN/TP [average of （6.76±0.26） mg/L]， the planktonic diatom abundance was limited during the period from February to June， resulting in low biomass and chlorophyll a content. In group B， the concentration of TP is lower than 0.1 g/ml， and the mean value of TN/TP is （64.2±0.15） mg/L. With appropriate water temperature， the community structure of planktonic diatoms tends to be stable， which is mainly composed of the species with higher individual biomass of S. acus and S. ulna. Moreover， since the 2 species also have high abundance， the planktonic diatoms also have high biomass， and the content of chlorophyll a increases. 　　The results of RDA （Fig. 3） show that TP， TN， WT， TN/TP and NH3N/TP are limiting factors in water. Samples 1-14 （February-April） are in the third quadrant， and they are in significant correlations with TP but negative correlations with WT， indicating that TP and WT are the main limiting factors in February-April. Samples 15-22 （May-June） are in the third quadrant， which is closely related to TN. From July to November， the main limiting factors are TN/TP and NH3N/TP. The dominant species in the water body， namely C. meneghiniana， Asterionella formosa， Cymbella tumida and C. stelligera， have high correlations with TP， TN and NH3N. A. formosa and M. granulate show significant positive correlations with TP， suggesting that they are more suitable for eutrophic water bodies， and therefore， mesotrophic water bodies， their dominance is not very high， and they are the dominant species only in a few months （Table 1）. C. meneghiniana and C. stelligera are mesotrophictolerant species， and they are also common in some eutrophic shallow water lakes in Europe and are used as the main indicator species for eutrophic water bodies[18].
　　Relationship between dominant planktonic diatoms and essential nutrients
　　The planktonic diatom species can be divided into nutrientsensitive and tolerant types. According to different levels of human interference in water bodies， Wu[19]selected sensitive and tolerant species by clustering， and then proposed the generic index of diatom （GI）. Similarly， Dong et al.[20]analyzed the eutrophic lakes in the Yangtze River basin and found that the TP tolerance of C. meneghiniana was from 0.1 mg/L to 0.3 mg/L， while the tolerance of S. acus and S. ulna was less than 0.1 mg/L. Cairns et al.[21]found that C. meneghiniana grew fastest at 5-25 ℃， and therefore， at appropriate TP concentration and water temperature， C. meneghiniana had a higher abundance level in February-June. S. acus and S. ulna have strong correlations with TN/TP， and are suitable for TP at a low concentration level， so that they grow in large quantities from July to October. For other nutrient salts， different species have different tolerance values. However， due to the small changes of other nutrients in Baiguishan Reservoir， no good correlation has been found.
　　References
　　[1]POTAPOVA M， CHARLES DF. Distribution of benthic diatoms in US rivers in relation to conductivity and ionic composition[J]. Freshwater Biology， 2003， 48（8）： 1311-1328.
　　[2]ROTT E. Methodological aspects and perspectives in the use of periphyton for monitoring and protecting rivers[J]. Use of algae for monitoring rivers， 1991： 9-16. 　　[3]WILLIAMS DF. Continuity response to orbital forcing from biogenic silica records in Lake Baikal[J]. Nature， 1995， 378： 769-771.
　　[4]LI QH， HE WT， CHEN C. Characteristics of the phytoplankton community in Wetlands of Macao[J]. Chinese Journal of Plant Ecology， 2009， 33（4）： 689-697.
　　[5]JAANUS A， TOMING K， HALLFORS S， et al. Potential phytoplankton indicator species for monitoring Baltic coastal waters in the summer period[J]. Hydrobiologia， 2009， 629（1）： 157-168
　　[6]HU HJ， WEI YX. Freshwater Algal in China. Beijing： Science Press， 2006.
　　[7]JIN XC， TU QY. The Standard Methods in Lake Eutrophication Investigation[M]. Beijing： China Environmental Science Press， 1990， 286-302.
　　[8]BERGER WH， PARKER FL. Diversity of planktonic foraminifera in deepsea sediments[J]. Science， 1970， 168（3937）： 1345-1347.
　　[9]SPELLERBERG IF， FEDOR PJ. A tribute to Claude Shannon （1916–2001） and a plea for more rigorous use of species richness， species diversity and the ‘Shannon-Wiener’ Index[J]. Global ecology and biogeography， 2003， 12（3）： 177-179.
　　[10]PIELOU EC. An Introduction to Mathematical Ecology[M]. New York & London， WileyInterscience， 1969， 1-286.
　　[11]WANG MC， LIU XQ， ZHANG JH. Evaluate method and classification Standard on lake eutrophication[J]. Environmental monitoring in China， 2002， 18（5）： 47-49.
　　[12]TISON J， PARK YS， COSTE M， et al. Typology of diatom communities and the influence of hydroecoregions： a study on the French hydrosystem scale[J]. Water Research， 2005， 39（14）： 3177- 3188.
　　[13]CLARKE KR， WARWICK RM. PRIMER v5： user manual/tutorial[M]. PRIMERE Limited， 2001.
　　[14]BORCARD D， LEGENDRE P， DRAPEAU P. Partialling out the spatial component of ecological variation[J]. Ecology， 1992， 73：1045–1055.
　　[15]SONDERGAARD M， LARSEN SE， JORGENSEN TB， et al. Using chlorophyll a and cyanobacteria in the ecological classification of lakes[J]. Ecological indicators， 2011， 11（5）： 1403-1412.
　　[16]LEHMAN JT， BOTKIN DB， LIKENS GE. The assumptions and rationales of a computer model of phytoplankton population dynamics[J]. Limnology and Oceanography， 1975， 20（3）： 343-364.
　　[17]WATSON SB， MACAULEY E， DOWNING JA. Patterns in phytoplankton taxonomic composition across temperate lakes of differing nutrient status[J]. Limnology and Oceanography， 1997， 42（3）： 487-495.
　　[18]BENNION H， APPLEBY PG， PHILLIPS GL. Reconstructing nutrient histories in the Norfolk Broads， UK： implications for the role of diatomtotal phosphorus transfer functions in shallow lake management[J]. Journal of Paleolimnology， 2001， 26（ 2）： 181-204.
　　[19]WU JT. A generic index of diatom assemblages as bioindicator of pollution in the Keelung River of Taiwan[J]. Hydrobiologia， 1999， 397： 79-87.
　　[20]DONG XH， YANG XD， WANG R. Diatom demonstrating species of eutrophication of the lakes in the middle and lower reach regions of Yangtze River[J]. China Environmental Science， 2006， 26（5）： 570-574.
　　[21]CAIMS J， BUIKEMA AL， HEATH AG， et al. Effects of temperature on aquatic plant sensitivity to selected chemicals[M]. Blacksburg， Virginia 24060： Virginia Water Resources Research Center， Virginia Polytechnic Institute and State University， 1978.
　　Editor： Na LI Proofreader： Xinxiu ZHU