QIN Hai-Ming ， YAN Jun-Wei NIE Xue WANG Wen-Juan and SHEN Xiao-Yuan

(1. Center for Watershed Ecology， Institute of Life Science， Nanchang University， Nanchang 330031， China; 2. Key Laboratory of Poyang Lake Environment and Resource Utilization， Ministry of Education， Nanchang University， Nanchang 330031， China;

3. National Ecosystem Research Station of Jiangxi Poyang Lake Wetland， Nanchang 330038， China)

Abstract: Our study examined spatial distribution differences of rotifer resting eggs (RRE) in lakes of the Poyang Lake Wetland Area during dry season， January 2015. Field sampling was conducted in both Nanshen Lake and Baisha Lake. We found that densities of RRE in different habitats were significantly different(P＜0.05)， and their distribution had an obvious gradient. The highest mean density was in the marshland vegetation area of Nanshen Lake， where it was (3.34±1.28) ind./cm3. The density of resting eggs in the Zizania latifolia area was up to 5.45 ind./cm3. The density of RRE was different spatially， with the greatest in the vegetation area， then in the muddy area， and then the lowest in the water area. ANOVAs results showed that water depth， vegetation， and sediment had significant effects on RRE (P=0.001， P=0.007， P＜0.001，respectively). The trend of resting egg densities in different aquatic habitats was: emergent aquatic plant ＞marshland vegetation ＞ floating-leaf plants ＞ submerged plants. The distribution of RRE in areas of different water depths was inconsistent. The density of RRE was higher in the hard than soft sediment.

Key words: Rotifers; Resting eggs; Spatial distribution; Shallow lakes; Poyang Lake Wetland

Rotifers form a major group of freshwater zooplankton and mostly feed on bacteria， organic debris and microalgae. They play an important role in the transformation process of primary productivity and have great implications for the stability of aquatic ecosystems. Some species are food for a variety of economic aquatic animals and have wide applications in aquaculture[1]. They also represent a class of indicator organisms used extensively in the monitoring of ecological environments and the study of ecotoxicology[2]. Resting eggs， a diapause structure formed by Monogononta rotifers in adverse ecological conditions， are the product of sexual reproduction. The RRE are characterized by resistance against harsh external environments， simple and rapid diffusion， and production of genetic diversity through recombination. These features are of great significance in enabling rotifers to withstand adverse environments， protect populations， reproduce offspring and spread[3，4].The formation of RRE is influenced by numerous factors， some of which are inherent， such as genetic characteristics and maternal age， as well as external factors such as ambient temperature， salinity change，pH change and food abundance. At present it is generally accepted that external factors have a greater impact on its formation. Resting eggs are formed in the life cycle of rotifers， except for few species (e.g.，some species in the order Bdelloidea). In the surface sediment (0—5 cm) of fishponds massive numbers of RRE may be found， from several tens of thousands to several millions per square meter[5]. Studies also have shown that complex habitat structure may provide more opportunities to rotifers for use of the resources and increase species diversity[6—8]. It is generally believed that the main environmental factors for stimulating and producing RRE are water sources drying up， sudden change in temperature， high population density and shortage of food resources[9，10]. These hatch and propagate in specific conditions and are the material basis for natural proliferation of rotifers in ponds and an important “provenance” for their concentrated cultivation. Resting eggs play a key role as a connecting link in the rotifer life cycle， and so studying them is particularly important. However， existing studies of RRE have mainly investigated their morphology and germination， and their spatial distribution has rarely been investigated.

Poyang Lake is the largest freshwater lake in China and has been recognized as one of the world’s important wetlands. The Poyang Lake wetland provides an excellent habitat for numerous migratory birds each winter and has an important ecological role[11]. Nanji wetland lies in the fan-shaped alluvial area where a branch of the Gan River flows into Poyang Lake. The wetland connects the river-lake ecosystems， and plays a major role in maintaining the ecological balance of the water and surrounding land areas[12]. Intensifying the ecological study of Nanji wetland has great implications for protecting the wetland ecosystem of Poyang Lake. In the present study we investigated the spatial distribution characteristics of RRE and their differences in the Nanji wetland of Poyang Lake. The reason for the distribution differences of dormant rotifer eggs was explored to provide data to support the ecological protection of Poyang Lake wetland.

1 Materials and Methods

1.1 Study sites

Jiangxi Poyang Lake Nanji Wetland National Nature Reserve is located in the southwestern part of the main lake area of Poyang Lake. The reserve is part of the delta area formed by alluviation when the northern， middle and southern branches of Gan River flow into the open water of Poyang Lake. It is within the geographical range of 28°52′21″—29°06′46″ N and 116°10′24″—116°23′50″ E. The reserve covers a total area of 333 km2and belongs to the subtropical humid monsoon climate zone. The climate is warm and humid with a hot and rainy summer， while winter has low temperatures and less rainfall. Average annual precipitation ranges from 1450 to 1550 mm，mainly concentrated in April to June[12]. As it is affected by seasonal hydrological changes in Poyang Lake， the water area of the reserve shows large intraannual and inter-annual changes. The rainy season lasts from April to September in the reserve. Except for Nanshan Island and Jishan Island， the sub-lakes and grass islands are submerged by floods in the reserve and connected in a typical lacustrine state. The dry season starts after October. As lake water retreats，river channels and plate-like depressions appear in the reserve. Rivers crisscross and sub-lakes are dotted about. Marshlands emerge sequentially and are gradually covered by vegetation. The reserve shows a natural wetland landscape with rivers， sub-lakes and marshlands spread about[12]. The sampling sites， Nanshen Lake and Baisha Lake， are two core sub-lakes in the reserve. They are typical shallow plate-like lakes，with flat beds. The depth changes greatly with water level fluctuation， and the maximum water level difference is more than 6 m[12]. Local fishermen conduct fishery activity by controlling water flow from the lakes in the dry autumn season. The water gate is opened to discharge the water in mid-October every year and the lake water is generally drained by January the following year. The lake surface drops drastically after drainage and small puddles are formed in lower-lying areas. Water depth declines to approximately 0.2—0.3 m.

1.2 Sampling methods

Sediment samples were taken from two sublakes of the Poyang Lake Wetland Area， Nanshen Lake and Baisha Lake， in the Jiangxi Poyang Lake Nanji Wetland National Nature Reserve (116°10′—116°23′ E， 28°52′—29°06′ N) on 17 and 18 January 2015. Sampling points were set in the vegetation area，muddy area and water area in the two sub-lakes (Fig.1). A total of 10 sampling points was set in Nanshen Lake. Of the 10 sampling points， two were set in vegetation area (one inCarexspp. habitat and another inZizania latifoliahabitat)， another two points were in muddy area， and six sampling points were set in the water area. About the 6 sampling points in the water area， two points were at 2 cm depth area (one inMyriophyllum spicatumarea and another in bare muddy area). One point was at 5 cm depth area without vegetation. Two points were set in 8 cm depth area (one inNymphoides peltatumhabitat and another in bare mudflat habitat). The last one sampling point was set in 10 cm depth area， where no aquatic plants growing. A total of 8 sampling points was set in Baisha Lake. Of the 8 sampling points， two were set in marshland vegetation (one inCarexspp.habitat and another inZizania latifoliahabitat). One sampling point was set on bare mudflat. And sampling points were set in the water area， of which one point was at 2 cm depth area (a bare area)， one point was at 5 cm depth area (Ceratophyllum demersumhabitat)， other two points were at 6 cm depth area(one atC.demersumhabitat and anther at bare area，respectively). The last sampling point was set at 15 cm depth area (a bare area). Sediment hardness was also classified at each sampling point. The soft habitat was recorded where the ankle was submerged after taking a step， and the hard habitat was recorded where the ankle was not submerged. One replication sample was set at 3 m intervals for each sampling point. A columnar sediment collector (inner diameter， 7 cm; height，10 cm) was used to take sediment samples. The collector comprised a columnar body and a cylindrical metal rod， with a mobile handle in the middle of the metal rod. Sediment samples were sealed into plastic bags and stored. Vegetation type， sediment hardness and water depth were recorded at each sampling point. Samples were taken back to the laboratory and stored in a refrigerator at 5℃.

Fig. 1 Sketch maps of Jiangxi Poyang Lake Nanji Wetland National Nature Reserve and sampling points for rotifer resting eggs

1.3 Isolation, identification and enumeration of RRE

Identification followed the method of Li，et al.[13]for morphological and structural identification and description of RRE. Most were oval or spherical and covered by a thick， dark， hard shell， with ratchet on the surface[14]. They were isolated using the hyperosmotic suspension method[15]. Each sediment sample was weighed (0.01 g) and three 50 g sub-samples were selected. The selected samples were placed into 100 mL Erlenmeyer flasks， followed by the addition of a hyperosmotic solution (saturated saline solution +20% sucrose solution) to 4/5 of the flask volume. The mixture was stirred using a glass rod to make a homogeneous slurry suspension. The suspension was allowed to stand for 20min. When the sediment settled，the suspension was stirred again and the hyperosmotic solution was added to about 1 cm below the flask mouth. Subsequently， a dropper was used to slowly add the hyperosmotic solution， until the liquid surface was slightly above the flask mouth without overflowing. The suspension was allowed to stand for another 20min. A glass slide was gently placed on the flask mouth in contact with the liquid surface， three times， to allow RRE on the liquid surface to adhere.The RRE were counted under a microscope (Olympus CX23， Korea) after staining with iodine. The supernatant (1/5 of the flask volume) was then carefully decanted. The remaining sediment sample was stirred and the hyperosmotic solution was added for suspension and counting. A glass slide was used to adhere RRE twice. Since not all RRE could be extracted from the sediment sample by the five counts in the two suspension processes， we estimated the number of RREin each selected sample using the method proposed by Zippin[16]， as follows:

whereCtwas the number at thettime of count， andtwas the number of consecutive counts. Zippin provided the schematic method to solveaccording to theRvalue. When the data of RRE in a certain sampling could not meet the condition of Zip-pin’s formula， their density at this sampling point was estimated using the actual count number (N) multiplied by the mean ofat other sampling points in the same habitat.

1.4 Data analysis

Multi-factor analysis of variance (factorial ANOVA) was used to test the interaction effects of vegetation cover， water and sediment hardness on RRE.One-way ANOVA was used to test the effect of different habitat types on the density of RRE， and the difference in their density across different habitat types， water depth areas and vegetation types. Statistical analysis was conducted using Statistica7.0 (Stat-Soft Inc， Tulsa OK， USA)， withP＜0.05 considered statistically significant.

2 Results

2.1 Differences in densities of RRE in vegetation,mudflat and water areas

The density of RRE showed a gradually decreasing trend in the marshland vegetation， mudflat and water areas. Density in the marshland vegetation was markedly higher than in the two other areas (Fig. 2).In particular， it was significantly higher in the marshland vegetation compared with the other two habitats of Nanshen Lake (P＜0.01). The highest density of RRE (3.34±1.28) ind./cm3was found in the marshland vegetation of Nanshen Lake. In particular， their density reached 5.45 ind./cm3in theZ. latifoliahabitat. The lowest density (0.22 ind./cm3) appeared in the water area of Baisha Lake. Comparison of the two sub-lakes revealed that the density of RRE in all three types of habitat was higher in Nanshen Lake than in Baisha Lake.

The effect of habitats on RRE was tested by oneway ANOVA. Results showed that presence or absence of water， aquatic vegetation and soft and hard sediment had significant effects on the density of RRE (Tab. 1). The results of factorial ANOVA showed that only presence or absence of water and sediment had highly significant effects on RRE. Presence or absence of vegetation had a certain effect on RRE， but did not reach the level of significance. Interaction of the three habitat factors had marginally significant effects on RRE.

Fig. 2 Differences in densities of rotifer resting eggs in different habitats a and b indicate that there are significant differences in RRE densities in different habitats; the same applies below

2.2 Differences in densities of RRE in areas of different aquatic vegetation

Fig. 3 Differences in densities of rotifer resting eggs in areas of different vegetation

The density of RRE decreased successively in the aquatic vegetation ofZ. latifolia，Carexspp.，N.peltatum，C. demersumandM. spicatum. Density in theZ. latifoliaarea was markedly higher than with the other four aquatic plants (Fig. 3). In Nanshen Lake，the density of RRE was significantly higher in theZ.latifoliaarea compared with the other three aquatic plants (P＜0.01). The highest density of RRE (5.45 ind./cm3) occurred in theZ. latifoliaarea of Nanshen Lake. The lowest density of RRE (0.19 ind./cm3) was found in theM. spicatumarea of Baisha Lake. Comparison of the two sub-lakes showed that the density of RRE in all three habitats was higher in Nanshen Lake than in Baisha Lake. In the same vegetation areas (Z. latifoliaandCarexspp.) of the two sublakes， the density of RRE was also higher in Nanshen Lake than in Baisha Lake.

2.3 Differences in densities of RRE in areas of different water depths

The distribution of RRE in areas of different water depths showed opposite trends in the two sublakes. There was a positive correlation in Nanshen Lake， but a negative correlation in Baisha Lake (Fig. 4).In Nanshen Lake the density of RRE reached its highest level in the 10 cm water depth area (0.60 ind./cm3)， and was at its lowest in the 5 cm water area(0.10 ind./cm3). In Baisha Lake， the density of RRE reached its highest level (0.31 ind./cm3) in the 5 cm water depth area， and was at its lowest in the 15 cm area (0.06 ind./cm3). Comparison of the two sub-lakes revealed that the density of RRE in the 2 cm and 5 cm water depth areas was higher for Baisha Lake compared with Nanshen Lake， while the density in the 6 cm to 15 cm water depth areas was higher for Nanshen Lake compared with Baisha Lake.

2.4 Differences in densities of RRE in areas of different sediments

The distribution of RRE in sediments showed a clear pattern， being significantly higher in hard sediment than in soft sediment (P＜0.05， Fig. 5). In Nanshen Lake， the density of RRE in hard sediment (1.70 ind./cm3) was significantly higher than in soft sediment(0.20 ind./cm3). In Baisha Lake， the density of RRE in hard sediment (0.38 ind./cm3) was also significantly higher than in soft sediment (0.18 ind./cm3). Regardless of whether the sediment was hard or soft， the density of RRE was higher in Nanshen Lake than in Baisha Lake.

3 Discussion

3.1 Formation of RRE and effects of environmental differences on their distribution

Fig. 4 Differences in densities of rotifer resting eggs in areas of different water depth

The major environmental factors to stimulate and produce RRE are generally thought to be abrupt temperature change， excessive population density and food shortages[9，17]. The present study was conducted at low temperatures in the dry season and we found a relatively high density of RRE in sediments. This is mainly because the decrease in ambient temperature stimulated rotifers to form RRE for overwintering.Another factor was the lowering of the water level in Poyang Lake. This led to exposure of the majority of the lake basin， and loss of water environment was another important reason why numerous rotifers formed RRE. The density of RRE in water can directly affect their numbers in sediments[17]. The density of RRE in the sediment surface is positively correlated with the density of rotifers in the previous warm season. One researcher considered that the number of RRE is closely related to the maximum density of the rotifers in the previous year[10]. In the present study， we found that the density of RRE in three different habitat types ranked in the order: vegetation area ＞ bare area ＞ water area. There may be two main reasons for this: (1) During the wet season， the density of rotifers in the vegetation area was higher than in the water area with no distribution of vegetation (unpublished data). This result supports the view that the higher the density of rotifers in water， the higher the density of RRE in sediment. (2) Poyang Lake connects to rivers.With the arrival of the dry season， the water level gradually declines. In the vegetation area and shallow lakeside water， where more aquatic plants were distributed， the vegetation area with water recession was closer to the lake margin and was first exposed to the air after water level decline， forming an arid environment. Subsequently， water level declined in the bare area with no distribution of aquatic plants， leading to the formation of mudflats. This process of gradual change in water level resulted in a temporal gradient of environmental change， so the density of RRE was the highest in the first exposed vegetation area.

Fig. 5 Differences in densities of rotifer resting eggs in areas of different sediments

3.2 Differences in resting egg distribution in areas of different aquatic vegetation

The characteristics of the microhabitat are determinants for the basic structure of the zooplankton community in lentic freshwater ecosystems[18]. The heterogeneity of the microhabitat is influenced by biological factors: the composition and abundance of aquatic plants[19，20]. In a variety of microhabitat structures， aquatic plants play an important role in the construction of freshwater lake ecosystems. The area where the aquatic plant community is distributed provides a highly favorable habitat for zooplankton communities[6]. The various habitat patches created by aquatic plants provide different microhabitats for zooplankton[20，21]. A markedly higher density of zooplankton has been found in areas where submerged plants are distributed[7，8]. The complex habitat structure may thus offer more opportunities to zooplankton (including rotifers) for use of the resources， and so increase species diversity. Different microhabitats are formed by different aquatic plants， and these directly affect the population distribution of rotifers，further influencing the distribution of RRE in the dry season. In the present study we found great differences in the distribution of RRE across different areas of aquatic plants. The density distribution increased sequentially in the habitats ofZ. latifolia，Carexspp.，N.peltatum，C. demersumandM. spicatum. This may be directly attributed to differences in the types of these aquatic plants (Z. latifolia: emergent aquatic plant;Carexspp.: marshland plant;N. peltatum: floatingleaf plant; andC. demersumandM. spicatum: submerged plants) and in the microhabitats that these plants form.

3.3 Differences in resting egg distribution in areas of different water depths

The distribution of RRE in areas of various water depths showed completely opposite trends in the two sub-lakes， but identifying the exact reason is challenging. It is very difficult to achieve uniformity in variables in the water area of the two sub-lakes， except for water depth. For example， submerged plants grow at some sampling points in the water area. From the perspective of microhabitat difference， this had a great effect on the distribution of RRE. Additionally，there are large differences in the sediment type (soft and hard) in areas of different water depths， which may also affect the distribution of RRE to some extent. Moreover， the difference in zoobenthos and its activities in areas of different water depths may also influence the distribution of RRE. Finally， the activities of wintering waterfowl are an unknown factor affecting the distribution of RRE. Therefore， we could not obtain an exact correlation between water depth and density of RRE using the data obtained in the present study.

3.4 Differences in resting egg distribution in different sediment types

In the present study a higher density of RRE was found in hard sediment compared with soft sediment.This result shows a high consistency in the two sublakes. However， we could not identify the reason for this difference because of the simple design of the experiment (a simple division between hard and soft sediments based on the depth that the ankle was submerged in the sediment at sampling). The difference in sediment may cause differences in a variety of other factors， such as the distribution difference of aquatic plants and the diversity difference of benthos. Further research is required in depth to find the exact reason for this difference.

3.5 Effects of other factors on experiment results

Different isolation and enumeration methods may be important reasons for the differences in the results[22]. In the present study， we used the adjusted hyperosmotic suspension method， which is simple and convenient. The vast majority of RRE can be isolated after suspension is carried out twice. However， we cannot completely rule out the possibility that a small number of RRE are retained in the sediment and not suspended. Nipkow[23]used the direct counting method.During the counting， RRE may also be covered by sediment impurities and not included in the data. May[10]used the germination method for enumeration of rotifers hatched from RRE. However， this method only counted the RRE capable of germination， while the eggs that lost the ability to germinate or those in the resting stage were not taken into account. Furthermore， we did not measure physical and chemical factors at the sampling points. As a result we could not analyze the distribution differences in RRE from the perspective of physical and chemical factors.

3.6 Evaluation of the study of RRE

The emergence of various types of rotifers in water has a certain seasonality. Multiple sampling must take place in all four seasons to fully determine the species composition of rotifers in the water of an area. A survey of the distribution and diversity of RRE in the dry winter season can supplement the species numbers in the rotifer diversity survey and to some extent reflects the species composition of rotifers in the surveyed water[24，25]. The species composition of rotifers obtained by May[26]through germination experiments of RRE at different temperatures was very similar to the species composition structure of rotifers obtained by the survey of actual samples.Additionally， RRE can be preserved for a longer time without losing vitality than can live rotifers[17]， and this feature has been very convenient in carrying out research. Collecting sediment samples for germination of RRE is a feasible method to use when conducting a classification survey of rotifer fauna.

Acknowledgements:

Yushun Chen provided assistance in improving the early draft of this manuscript.