Longitudinal variations of phytoplankton compositions in lake-to-river systems

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Limnologica xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Limnologica journal homepage: www.elsevier.com/locate/limno

Longitudinal variations of phytoplankton compositions in lake-to-river systems Qian Yu a,b,c , Yongcan Chen c , Zhaowei Liu c,∗ , Dejun Zhu c , Haoran Wang c a

State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin, Beijing 100038, People’s Republic of China Research Center on Flood & Drought Disaster Reduction of the Ministry of Water Resources, China Institute of Water Resources and Hydropower Research, Beijing 100038, People’s Republic of China c State Key Laboratory of Hydroscience and Engineering, Tsinghua University, Beijing 100084, People’s Republic of China b

a r t i c l e

i n f o

Article history: Received 10 February 2015 Received in revised form 12 September 2015 Accepted 10 February 2016 Available online xxx Keywords: Eutrophication The Dianchi Lake The Tanglang River Functional groups Longitudinal variations Phytoplankton compositions

a b s t r a c t To test the hypothesis of longitudinal variations in phytoplankton compositions from a eutrophic lake to its river downstream and determine the length of the transition zone, we applied functional groups as well as taxonomical methods to this coupled aquatic system, which is composed of the Dianchi Lake upstream and the Tanglang River downstream, by sampling at 9 stations during Microcystis blooms in the Dianchi Lake in 2013. The longitudinal variations in phytoplankton compositions from lacustrine species to fluvial species were reflected by: (1) the shift from Microcystis to Chlorococcales green algae and centric diatoms; (2) the shift from the dominance of codon M to the coexistence of a variety of coda without one outstanding codon; and (3) except for codon M, the shift from lacustrine coda (H1, LO , T) towards coda that are adapted to both lacustrine and fluvial circumstances (MP, X1, X2). The prominent difference of phytoplankton compositions between the Dianchi Lake and the lower reaches of the Tanglang River revealed that there was a transition zone in between. The upper and middle reaches of the Tanglang River with a length of approximately 26.4 km were considered the transition zone because: (1) the dominant lentic codon M in the Dianchi Lake disappeared at the lower reaches of the river; (2) the amount of codon P that is sensitive to stratification rose at the beginning of the river; and (3) the codon T, which is well adapted to the persistently mixed layer or epilimnia of lakes, lost a large number of biomass at the upper and middle reaches of the Tanglang River. In this study, we found that the eutrophic lake had a significant influence on the river downstream. In addition, we found that functional groups were sensitive to the changes of external aquatic conditions and helpful in determining the length of the transition zone. © 2016 Elsevier GmbH. All rights reserved.

1. Introduction River eutrophication has attracted worldwide attention for decades. In the last thirty years, eutrophication has been a dominant pollution pressure on large rivers in Europe (Mischke et al., 2011). The Water Framework Directive (EC, 2000) of the European Parliament (WFD, 2000) recommends that researchers use phytoplankton data to assess river ecology. Generally, if the residence time is long enough, nutrient concentrations are high, water temperature and light intensity are satisfied, blooms may develop at the middle or lower reaches of large rivers rather than the upper reaches due to the small initial inocula at the sources of rivers (Hilton et al., 2006). The middle and lower reaches of rivers are usually dominated by centric diatoms, chlorococcalean

∗ Corresponding author. E-mail address: [email protected] (Z. Liu).

colonial green algae (Bahnwart et al., 1998), Chrysophyta and occasionally Cyanobacteria when flow is slow (Jeppesen et al., 2005; Phillips et al., 2008). Researchers consider the accumulations of phytoplankton biomass along rivers to be a significant source of phytoplankton biomass for lakes (Mihaljevic´ et al., 2010; Bridgeman et al., 2012; Soares et al., 2012; Kufel and Le´sniczuk, 2014) or estuaries (Li et al., 2007). However, few researchers have studied the impact of a eutrophic lake on the river downstream, which is also an important and serious issue. A large number of Cyanobacteria at the source of the river that is coming from a eutrophic lake upstream will stretch for a long distance in the river downstream (Chen et al., 2014). This eutrophic lake upstream and the outlet downstream comprise a coupled laketo-river system. Today, Cyanobacterial blooms are a major problem in lakes (Paerl and Huisman, 2008; Bridgeman et al., 2012). During high bloom periods, the Cyanobacteria will be transferred by the flow to the river downstream, leading to visible blue-green assemblages at the beginning of the river (Chen et al., 2014), which is

http://dx.doi.org/10.1016/j.limno.2016.02.007 0075-9511/© 2016 Elsevier GmbH. All rights reserved.

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

2

very different from other rivers without eutrophic lakes upstream as aforementioned. Because some Cyanobacteria can produce a toxin that is more harmful to aquatic organisms and humans than diatoms and other fluvial algae (Soares et al., 2012), polluted rivers would lead to a drinking water crisis for surrounding communities (Qin et al., 2010). However, draining water from eutrophic lakes to the rivers downstream is still a common method used in China to control lake eutrophication. Therefore, we should figure out the impacts of eutrophic lakes on the rivers downstream. Traditional researches on eutrophication are based on Chlorophyll a (Chla) (Mischke et al., 2011), but Chla cannot reveal the spatial variations in phytoplankton compositions and the responses of individual species to changes in the external environment. Reynolds et al. (2002) first put forward functional groups. The original idea is to classify species into ecologically similar groups on the basis of phytoplankton’s traits in lentic systems. The species belonging to the same subdivision share similar characteristics regarding the resources and energy, such as nutrients and light intensity (Reynolds et al., 2002). The functional groups have been successfully applied to many lakes and reservoirs (Becker et al., 2010; Xiao et al., 2011). Their application to lotic systems followed (Tavernini et al., 2011; Abonyi et al., 2012) after Devercelli (2006) had applied them to the Middle Paraná River and proved that they were also useful in rivers during extreme drought periods. On Reynolds’ basis, Borics et al. (2007) proposed several new groups that can be applied to rivers. Additionally, Padisák et al. (2009) expanded the initial 31 groups to 40 groups by reviewing related published papers. Despite these successful cases and improved methods, they are rarely used in lake-to-river coupled systems or river-to-lake coupled systems. In view of their sensitive responses to changes of the external environment, the compositions of functional groups should show differences in lakes and rivers. In other words, we can detect the changes of external aquatic environment by examining the compositions of coda, which will be tested in this study. In this research, we set nine sampling stations in the Dianchi Lake-to-Tanglang River system to study spatial variations in phytoplankton compositions. The objectives of this research were: (1) to judge if the phytoplankton assemblages (surface blooms) at the source of the Tanglang River came from the Dianchi Lake upstream and if the initial phytoplankton compositions would change along the Tanglang River; (2) to evaluate the influence of the eutrophic lake on the river downstream by evaluating the length of transition zone; and (3) to test the usefulness of functional groups in the eutrophic lake-to-river system. To achieve these goals, Reynolds’ functional groups as well as traditional taxonomical methods, were applied to the Dianchi Lake-to-Tanglang River system. 2. Materials and methods

polluted section of the Pudu River. One of the reasons is that the Tanglang River flows through densely populated farmlands while the downstream of the Pudu River flows through sparsely populated mountainous areas. Another more important reason is that the Tanglang River accepts water directly from the Dianchi Lake upstream. As the single greatest outflow river from the Dianchi Lake, it receives most of outflows with large amounts of Microcystis biomass. Despite the many studies on eutrophication in the Dianchi Lake, phytoplankton identification at the Tanglang River was performed only once, by Qian et al. (1985), in 1982, and even that was qualitative rather than quantitative.

2.2. Sampling and analysis Sampling was done in the daytime of June 19, 2013 at nine monitoring stations during Microcystis blooms in the Dianchi Lake (Fig. 1). There was no rain during the samplings. The nine stations included two stations at the mouth of the Dianchi Lake, at −2.04 km (D1) and 0 km (D2, reference station), and seven others along the Tanglang River: 1.47 km (T1), 4.83 km (T2), 5.46 km (T3), 6.46 km (T4), 11.38 km (T5), 27.84 km (T6) and 114.60 km (T7). Among them, T6 was located at the converging point of a small tributary and the main stream. In addition, there was a large number of emergent plants between stations T3 and T4. The samples of surface water (0.5 m) were taken from the shore of the lake and the middle of the river. They were immediately preserved in situ with Lugol’s iodine solution. The one-litre samples of each station were brought back to the laboratory in darkened bottles. After 24 h of standing in the laboratory, the supernate was siphoned and the remaining 30-mL residue was prepared for further phytoplankton identifications. All phytoplankton samples were identified and counted by using a Nikon E100 biological microscope (10 × 40) at species level according to Hu and Wei (2006). Water temperature was measured in situ at all stations using a YSI6600 multiparameter water quality sonde (Yellow Spring Instruments, USA). Flow velocity was detected in situ using an LS300 portable flow meter (Nanjing, China). Phytoplankton species were classified into corresponding functional groups according to Reynolds et al. (2002) and Padisák et al. (2009). The method is to put species that share similar features together. Reynolds et al. (2002) named the subdivisions in capital letters. Padisák et al. (2009) has summarized phytoplankton species that have corresponding coda. To avoid mistakes, we classified only those species summarized by Padisák et al. (2009) into the corresponding coda. Because of the difficulty in the identification of some phytoplankton at species level, such as centric diatoms and Oscillatoriales, we classified them into corresponding coda based on their morphologies.

2.1. Study area The study area, including the mouth of the Dianchi Lake and the entire Tanglang River (Fig. 1), lies in Yunnan Province, Southwest China (N24◦ 45 –N25◦ 22 and E102◦ 30 –E102◦ 37 ). The Tanglang River is a part of the Pudu River, and it flows 120 km north, covering approximately 3187 km2 . The surrounding communities of the river are seriously affected by industry and agriculture (Yu et al., 2013). The Dianchi Lake is one of the most polluted lakes in China. Toxic Microcystis blooms frequently occur in this lake. During bloom periods, the blue-green surface scum may produce abundant microcystin toxin. Additionally, the local warm and humid subtropical highland variety of the oceanic climate may lengthen the bloom period, leading to more lasting harms. According to previous studies and field experiments during 2010–2011 (Yu et al., 2013), the Tanglang River is the most seriously

3. Results 3.1. Environmental variables Flow velocities at three stations (T4, T5 and T7) were detected. The flow velocities at T4 and T5 were 0.19 m/s and 0.17 m/s, respectively. The flow velocity reached its highest value at T7 (1.90 m/s). The velocities at T1–T3 were not detected because the flow velocities were below the lower limit of the LS300 (0.10 m/s). The flow velocity at T6 was not detected in situ because the velocity was significantly disturbed by the tributary. The surface water temperature ranged between 25.5 ◦ C and 28.0 ◦ C at the nine stations in June 2013. The highest water temperature was at T1, while the lowest water temperature was at T5.

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

3

Fig. 1. Map of the Pudu River catchment and sampling stations at the mouth of the Dianchi Lake and along the Tanglang River in June 2013.

3.2. Phytoplankton compositions Most of the samples involved in this study had great total abundances with more than 108 cells/L (Fig. 2). A total of 90+ species were identified (+ means that more than one species were found but could not be detected), which belonged to seven phyla: Cyanobacteria, Chlorophyta (green algae), Bacillariophyta (diatoms), Xanthophyta, Cryptophyta, Euglenophyta and Dinophyta (dinoflagellates). Among them, Cyanobacteria contributed the most to the total abundances at almost all stations, followed by Chlorophyta and Bacillariophyta. These three phyla contributed over 90% of the total abundance from station D1 to T6 and approximately 86.7% at T7. They were referred to as the main phyla in this study. Additionally, Bacillariophyta contributed the most species (36), followed by Chlorophyta (24+) and Cyanobacteria (19+). Among all of the species, 80+ species were classified into seventeen coda (Table 1). The other 10+ species were not categorized because they were not covered by the work of Padisák et al. (2009). Each functional group included one or more species. The most species-rich functional groups were codon J (19 spp.), codon M (10 spp.) and codon MP (9+ spp.). The most frequent groups were the

Fig. 2. The compositions of phytoplankton at all stations. The other four phyla (Xanthophyta, Cryptophyta, Euglenophyta and Dinophyta) are presented in the figure as one group because their combined abundance contributed less than 3% to the total abundance from station D1 to T6.

coda LO , P, MP, J, and X1, and they were present at every station. Additionally, codon M contributed the most to the total abundances from station D1 to station T5 (Fig. 2). 3.3. Spatial distributions of phytoplankton Distinct spatial variations in phytoplankton abundances were found in these samples (Fig. 2). The phytoplankton compositions at riverine station T1 and lacustrine station D2 were similar at the genus level. From station D2 to T1, the total abundance declined by only 11.6%, accompanied by lacustrine dominant Cyanobacteria dropping by 12.8%. The dominant Cyanobacteria in the Dianchi Lake were transported downstream by the flow and contributed the Table 1 List of the species and corresponding functional groups in the Dianchi Lake-toTanglang River system. Cyanobacteria Microcystis spp. Anabaena spp. Chroococcus minutus Chroococcus minor Coelosphaerium kuetzingiaaum Merismopedia minima Oscillatoria spp. Aphanizomenon flos-aquae Phormidium sp. Bacillariophyta Aulacoseira granulata Aulacoseira distans Navicula spp. Stephanodiscus neoastraea Cyclotella meneghiniana Nitzschia acicularis Nitzschia spp. Eunotia sp. Synedra acus Fragilaria sp. Gomphonema parvulum Euglenophyta Euglena spp. Phacus spp.

M H1 LO LO LO LO MP H1 TC P C MP B C D D MP D TB TB W1 W1

Chlorophyta Pediastrum spp. Scenedesmus spp. Planctonema lauterbornii Chlorella vulgaris Staurastrum spp. Ankistrodesmus sp. Closterium aciculare Closterium acutum Closterium pronum Closterium spp. Crucigenia spp. Coelastrum spp. Chlamydomonas spp. Planktosphaeria gelatinosa Actinastrum sp. Cosmarium sp. Xanthophyta Tribonema spp. Cryptophyta Cryptomonas pyrenoldfera Chroomonas sp. Dinophyta Peridinium sp.

J J T X1 N X1 P P P P J J X2 F J N T X2 X2 LO

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

4

Fig. 3. The distributions of main coda along sampling stations in June 2013 (a) with the lacustrine dominant codon M, (b) without the codon M. The bottom five coda with low proportions (<1%) are not presented in these figures.

greatest abundance at the upper five stations along the Tanglang River, accounting for more than 80% of the total abundance at these stations. Among them, we found that the abundance of Cyanobacteria decreased significantly from station T3 to T4 because the emergent plants prevented light from penetrating into water. At T6, although it was not the phylum with the greatest abundance, Cyanobacteria still made up 41.6% of the total abundance, of which 85.2% was Microcystis. At T7, Cyanobacteria contributed only 3.8% to the total abundance, whereas the Microcystis disappeared. The existing species of Cyanobacteria at this station were Phormidium and Chroococcus. In contrast, there were no Phormidium and only small amounts of Chroococcus at D1 and D2. Hence, all of the Cyanobacteria species that were coming from the Dianchi Lake upstream disappeared at the lower reaches of the Tanglang River. The abundances of Chlorophyta and Bacillariophyta increased accompanied with the decreasing abundance of Cyanobacteria along the Tanglang River (Fig. 2). Chlorophyta reached its maximum abundance at T6, while Bacillariophyta peaked at T7 (Fig. 2). At all stations, Chlorococcales green algae, such as Chlorella, Pediastrum and Scenedesmus, contributed the most to Chlorophyta. Centric diatoms followed by pennate diatoms provided the greatest contributions to Bacillariophyta. The abundances of the other four phyla were not as large as the abundances of the three main phyla. They all had some sporadic disappearances at different stations. Additionally, Dinophyta (Peridinium sp.) was present only at T7. 3.4. Longitudinal distributions of functional groups Among the seventeen detected coda, coda M, MP, X1, H1, and J appeared as the most successful groups and dominated at different stations (Fig. 3). Codon M, consisting of all Microcystis spp. in this study, prevailed at the upstream stations (68.42–90.18%), while it lost the dominant place at stations T6 and T7. Cyanobacteria H1 had relatively high proportions at upstream stations, together with codon M. The percentage of codon H1 reached its peak at

T2. The percentage of Chlorophyta codon J remained relatively stable at stations D1 to T2, then decreased slightly at station T3 and finally peaked at T7 (16.64%). A unimodal pattern of codon MP was obvious, with the greatest percentage at T4. Codon X1 (41.7%) replaced codon M (35.6%) as the dominant codon at T6. The number of co-occurring prevailing coda (>1%) reached the maximum value at T7. A total of ten prevailing coda coexisted at this station. Additionally, in nearly half of these coda, each contributed more than 10 per cent of the total abundance. In contrast, the number of prevailing coda was smaller at the stations upstream. 4 and 5 prevailing coda co-occurred at D1 and D2, respectively. 7 prevailing coda coexisted at T2 and T4, while 5 coda co-occurred at T1 and T3. On the basis of descriptions of habitat templates (suitable environments for the species belonging to the codon) in Reynolds et al. (2002) and Padisák et al. (2009) for different coda, we divided these coda into several classes in two ways to analyze the influence of eutrophic lakes on the rivers downstream. (1) We separated the seventeen coda into three classes: the first class included coda fit for lakes only, coda M, H1, LO , P, C, F, W1, and B; the second class included coda without explicit descriptions for lakes or rivers, coda MP, D, J, X1, X2, TC , T, and N; and the third class included a codon adapted to rivers only, just codon TB . (2) We separated the seventeen coda into two groups: Borics’ fluvial group (2007), coda TB and TC ; and Reynolds’ lacustrine group, the other fifteen coda. According to Fig. 4(a), the percentage of functional groups that are adapted only to a lacustrine environment decreased continuously towards downstream with one exceptional peak at station T5. In contrast, the percentage of functional groups that are adapted to both lacustrine and fluvial circumstances rose. The functional groups that are adapted to rivers appeared only at station T6. However, the abundance was too small to be presented in the figure. According to Fig. 4(b), the coda suitable only for rivers were present at station T2, station T6 and station T7. However, the abundance of fluvial phytoplankton was too small in comparison to other lacustrine species. Nevertheless, there was a trend that phytoplankton that are well adapted to lakes and rivers gained proportions along the river and

Fig. 4. (a) The longitudinal variations of the lacustrine class, lacustrine and fluvial class, and fluvial class in functional groups along the sampling stations and (b) the spatial variations in functional groups belonging to Reynolds’ class and Borics’ class, respectively.

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

finally replaced the algae that are adapted only to lakes according to these two classifications. 4. Discussion 4.1. Special phytoplankton structures in eutrophic lake-to-river systems The Dianchi Lake was the main phytoplankton source for the algae at the upper reaches of the Tanglang River. Lentic species, e.g., Microcystis, rather than benthic or lotic species (Istvánovics et al., 2010), were the main algae at the upper reaches of the Tanglang River. Additionally, the planktonic coda (M, H1) dominated at station T1, which were similar to the codon compositions at station D2. In general, the residence time of the upper reaches may determine the survival or the mortality of lacustrine algae from the lakes upstream (Hilton et al., 2006; Devercelli and O’Farrell, 2013). If the residence time is long, phytoplankton blooms occur in the rivers downstream due to the high initial inocula from the eutrophic lakes (Hilton et al., 2006). Otherwise, the mortality of phytoplankton entering the river downstream can be high (Wetzel, 2001), which means that most of the lacustrine algae from the lake would be lost rather than be transported downstream. Considering that low water velocities are associated with high residence time (Devercelli and O’Farrell, 2013), the slow flow velocities at the upper reaches of the Tanglang River supported that the lacustrine algae was from the Dianchi Lake upstream. The small loss of the total abundance and the Microcystis from station D2 to T1 elucidated that the mortality was not high (Wetzel, 2001). The majority of the dominant phytoplankton from the Dianchi Lake entered the Tanglang River and was transported downstream the river. Moreover, the distance between stations T1 and D2 was short (1.47 km). Therefore, we could conclude that the great mass of phytoplankton at T1 came from the Dianchi Lake upstream. The phenomenon observed in the Tanglang River (Fig. 2) was very different from that observed in other rivers without eutrophic lakes upstream. Generally, there is a positive relationship between the lengths of rivers and their phytoplankton abundances (Salmaso and Zignin, 2010; Tavernini et al., 2011), which means that the total abundance of phytoplankton would increase with the reach length. However, the greatest total abundance appeared at the beginning of the Tanglang River, and the total abundance decreased downstream the river. In addition, the algae abundances in many large low-gradient rivers are usually inversely correlated with discharges because of the dilution effect (Sullivan et al., 2001; Wetzel, 2001). However, the Tanglang River revealed the exact opposite phenomenon. Here, most of the phytoplankton at the upper and middle reaches came from the Dianchi Lake upstream rather than growing in the river. Consequently, in contrast to the phytoplankton concentration dilution in other large rivers caused by inflows from tributaries or enhanced discharges (Salmaso and Braioni, 2008; Devercelli and O’Farrell, 2013), the water from the Dianchi Lake enhanced the input of algae biomass instead of dilution. Such an enriching biomass effect has been described in other articles (Hilton et al., 2006), but few instances of it have been studied. 4.2. Spatial change in phytoplankton compositions in eutrophic lake-to-river systems The original abundance of Cyanobacteria coming from the Dianchi Lake accounted for the spatial variations in phytoplankton compositions in the Tanglang River downstream (Bahnwart et al., 1998). Microcystis and other Cyanobacteria were dominant phytoplanktons in the Dianchi Lake. As aforementioned, the Dianchi Lake was the main contributor of phytoplanktons to the Tanglang

5

River. Therefore, Cyanobacteria dominated at the upper and middle reaches of the Tanglang River. However, these lacustrine phytoplanktons are suitable to live in stagnant water or slow flow (Mitrovic et al., 2011) rather than turbulent flow (Hilton et al., 2006). Thus, Cyanobacteria decreased gradually, while Chlorophyta and Bacillariophyta increased in abundance and dominated at the lower reaches of the river. As the flow was turbulent before the tributary merging into the Tanglang River in the field observation, the main algae in this section of the tributary were probably green algae and diatoms as in other similar rivers (Phillips et al., 2008). Hence, the significant increase of the green algae and diatoms abundances at T6 was likely due to the inflow of the tributary. Further studies are necessary to study whether this deduction is true or not. The phytoplankton structure at station T7 was similar to that of other rivers where dominant species are centric diatoms and Chlorococcales (Bahnwart et al., 1998; Friedrich and Pohlmann, 2009; Tavernini et al., 2011). Another sampling was taken in September 2013 for studying the influencing environmental variables explaining the spatial variations (Yu et al., 2015a). The phenomenon observed in September was similar to that in June that the lacustrine species were gradually replaced by fluvial species. Microcystis is a typical lacustrine algae, while Chlorococcales green algae and centric diatoms are typical fluvial algae (Istvánovics et al., 2010; Devercelli and O’Farrell, 2013). These phytoplankton species that are adapted to similar habitats have various common characteristics. The majority of typical lacustrine algae are suitable to live in slow flows. Stable lakes create a good environment for Microcystis and other buoyant algae to regulate their buoyancy. The buoyancy helps the Microcystis migrate to the surface of water for full photosynthesis (Huisman et al., 1999; Yu et al., 2015b). In contrast, fluvial phytoplankton’s growth rate is usually faster than lacustrine algae’s (Wetzel, 2001) such that the doubling time of the species has to be shorter than the residence time to replicate themselves. In general, the residence time in rivers is much shorter than that in lakes/reservoirs (Devercelli and O’Farrell, 2013). Additionally, the time to absorb nutrients is limited for fluvial phytoplankton, as these nutrients are transported downstream by the flow. Consequently, the fluvial phytoplanktons have to acquire and absorb nutrients faster and more efficiently. In this case, the Microcystis and other Cyanobacteria with slower growth rates were suppressed by the fast flow (Wetzel, 2001) at the lower reaches. Consequently, light intensity in the water would be abundant for fluvial algae due to lack of the Microcystis on the surface of water. Centric diatoms and chlorococcalean green algae can have enough light for full photosynthesis and increase in abundance. Phytoplankton presented a clear functional response to the changes of the complex external aquatic environment from the lake to the river downstream. The compositions of functional groups were very different from those in rivers without eutrophic lakes upstream. The slow flow at the upper reaches of the Tanglang River allowed Cyanobacteria (M, H1, LO ) and Euglenophyta (W1) flowing in from the Dianchi Lake upstream to thrive. The highly lotic codon TB showed up only at station T6. Bacillariophyta functional groups (C, D) had higher abundances at the middle and lower reaches of the Tanglang River, as in many other enriched rivers (Soares et al., 2007; Devercelli and O’Farrell, 2013). A large number of Microcystis assemblages on the surface of water would keep light from penetrating into the water (Paerl and Huisman, 2008), further preventing their competing algae from acquiring light (Huisman et al., 2004). Under these conditions, those coda that are sensitive to low light could hardly coexist with codon M. Hence, the number of coexisting coda was small when codon M dominated. There were only 3 and 4 other coda that contributed more than 1% to the total abundance coexisting with codon M at stations D1 and D2, respectively. However, the number soared to 10 when codon M disappeared at station T7. The environment at D1 and D2 was analogous to

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

6

the habitats that Reynolds et al. (2002) described for codon H1 as “eutrophic, both stratified and shallow lakes,” but the environment at the lower river reaches was very different where the flow was turbulent, which was similar to the optimum environment of codon MP. In addition, the basic growth rate of Oscillatoria, representative of codon MP, is larger than that of Anabaena, the representative of codon H1 (Reynolds, 1989). Therefore, the faster-growing Oscillatoria showed up at fluvial station T5 while not appearing at two lacustrine stations. On the contrary, the slower-growing Anabaena appeared at D1 and D2 in lakes but disappeared at station T5 in the lower reaches. All of these observations revealed that the functional groups were sensitive to the changes in aquatic environment and that the compositions of functional groups would change with external aquatic environment. Levels of nutrients, water temperature and grazing were not the main reasons for phytoplankton variations in the space. The lowest concentrations of total phosphorus and total nitrogen detected in the Tanglang River during 2010–2011 were 71 ␮g/L and 863 ␮g/L, respectively. The values were much higher than those limiting phytoplankton growth, 3–6 ␮g/L and 100 ␮g/L (Reynolds, 2006), as in many other rivers (Salmaso and Zignin, 2010; Devercelli and O’Farrell, 2013). In other words, nutrients were abundant and would not limit the growth of phytoplankton in the Tanglang River. The surface water temperature was higher than 25.5 ◦ C at all nine stations, which were suitable for the growth of Microcystis and other species (Paerl and Huisman, 2008). Finally, the grazing effect in rivers is much lower than that in lakes (Chen et al., 2014) because only rotifers and other crustaceans that have fast growth rates but low filter rates can live in rivers (Allan and Castillo, 1995; Twiss et al., 2010). In that case, the increasing velocity might be the main reason. The flow velocity was much faster at T7 than at other upper stations. Higher velocities refer to more turbulent flow (Devercelli and O’Farrell, 2013), and thus the environment at station T7 was not suitable for lentic algae. Although we did not measure the light intensity underwater in situ, we assumed that the light intensity was inversely linked with the total suspended solids (Huisman et al., 2002). As aforementioned, large amounts of Microcystis on the surface of upper reaches increased the concentrations of the suspended solids and further blocked the light penetration. However, the abundances of buoyant Microcystis were low at stations T6 and T7. Accordingly, the penetrated light was enough and free to other competing algae, resulting in increased abundances of Chlorococcales green algae and centric diatoms. Therefore, intensifying the flow turbulence at the upper reaches of the Tanglang River could be useful to control the lentic and toxic Microcystis from the Dianchi Lake. The environmental variables (i.e., velocity, turbidity, total phosphorus, total nitrogen, etc.) were detected in September 2013. The detailed analysis of influencing factors for spatial variations in phytoplankton compositions was presented in Yu et al. (2015a), and we refer interested readers to this paper for more information. 4.3. Transition zone In a eutrophic lake-to-river system, there is a transition zone where both lacustrine and fluvial phytoplanktons coexist under the influence of the eutrophic lake upstream. Although Chla was often regarded as a proxy of the phytoplankton biomass, it is hard to detect the spatial variation of the algae between the Dianchi Lake upstream and the Tanglang River downstream because Chla is included in every kind of phytoplankton. Therefore, we could only get the variation in the total phytoplankton in the longitudinal rather than the spatial variation in the phytoplankton compositions. Consequently, it is hard to evaluate the transition zone between the Dianchi Lake and the Tanglang River downstream. On the contrary, the method of functional groups is to put species that share similar features together. Hence, if the nutrients are abundant

and other environment conditions are suitable, the species adapted to calm aquatic environment (such as lakes) would be gathered into one group and the species adapted to turbulent fluvial environment would be clustered into another group. Because different functional groups adapt to different aquatic circumstances, the changes of aquatic environment should be reflected in the compositions of functional groups if they are sensitive to the changing environment. Codon M was entirely composed of Microcystis, which was the bloom species in the Dianchi Lake. All of codon M came from the lake upstream because the Microcystis does not adapt to fast flow and rarely shows up in turbulent rivers (Mitrovic et al., 2011). In contrast, some Cyanobacteria adapt to fast flow. Hence, in comparison to Cyanobacteria, codon M was a more suitable symbol to evaluate the impact of the Dianchi Lake in this study. Codon M dominated the algae from station D1 to station T5. As aforementioned, the flow velocities at these stations were slow, far slower than that at station T7. The dominance of codon M demonstrated that the Dianchi Lake had a significant impact on the upper reaches of the Tanglang River. Additionally, although codon X1 dominated at station T6, codon M still made up 35.6% of the total abundance. It was not until station T7 that codon M disappeared because it is sensitive to flushing (Reynolds et al., 2002) in view of the highest velocity at station T7. Therefore, the end of the transition zone was located between stations T6 and T7. Additionally, codon P is sensitive to weak mixing. It contributed a small percentage to the total abundance at D1 and D2, while it had higher proportions since station T1 and achieved the highest percentage at station T7. Considering that stations D1 and D2 were located at the mouth of the Dianchi Lake, where the water was relatively stable, the increased abundances of the species that belonged to codon P at station T1 indicated that the stable status of the water was disturbed. In addition, codon T is suitable for “persistently mixed layers” (Padisák et al., 2009). The surface waters at T6 and T7 were persistently mixed. Therefore, the environments at these stations were suitable for species belonging to codon T. Correspondingly, functional group T made up a certain proportion at T6 and further reached its highest percentage at T7, but it made up lower proportions at stations at the upper reaches of the Tanglang River. The mixing in lakes is usually caused by wind, while in rivers it is usually caused by shear stresses on the riverbeds. Rivers with a high length-to-width ratio are much less susceptible to winddriven mixing than lakes (Kufel and Le´sniczuk, 2014). Moreover, the velocities at the upper reaches of the Tanglang River were so small that the corresponding small shear stresses would not arouse persistent mixing. Hence, the mixing was not complete at these upper stations. However, with a velocity of 1.90 m/s, the water at the lower reaches of the river (station T7) was persistently mixed, and the aquatic conditions were fit for codon T. The aforementioned phenomenon suggested that station T1 was the start of the transition zone. Moreover, the total abundance decreased by 62% within 12 km (T5) and by 88% within 115 km (T7) of the Tanglang River downstream from the Dianchi Lake. The abundance of the planktonic Microcystis also decreased by 63% within 12 km and declined by 100% within 115 km. In addition, the abundance of Microcystis decreased by 80% at station T6 compared to that at D2. Hence, the bloom species Microcystis from the Dianchi Lake lost the most abundance at station T6. In conclusion, we roughly considered the reaches between T1 and T6 as the transition zone, with a length of approximate 26.4 km. The analysis demonstrated that the toxic Microcystis would contaminate the river water over a long distance. In addition, colonial Microcystis would reduce the efficiency of water treatment plants. Hence, it is better to take water from beyond the transition zone for drinking and industry. However, there are three drawbacks to this sampling. First, we did not know if the spatial variations in phytoplankton abundances and compositions in June 2013 also occurred in other time. Second,

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

the inflow of a tributary at station T6 and the long distance between stations T6 and T7 might make the result lose its accuracy. Third, we did not detect environmental variables to analyze their influences. All of these problems have been settled in the second sampling in September 2013. We set more sites between stations T5 and T7 during the second round of sampling in September 2013. Additionally, we put station T6 downstream (0.81 km) from the intersection of the tributary and the mainstream to avoid the influence of the tributary. Besides, we detected the environmental variables. The phenomenon observed in the second sampling also proved that the Dianchi Lake had a big impact on the Tanglang River downstream and the lacustrine algae was gradually replaced by fluvial algae. The hydrodynamic condition was the main influencing factor, which was described in detail in Yu et al. (2015a). However, we still achieved the goals by analyzing the data of this sampling alone, despite of the three drawbacks mentioned above. 5. Conclusions With this study, we concluded that the Dianchi Lake was the main phytoplankton source for the Tanglang River downstream and had a significant impact on this river. The bloom species Microcystis was the main algae inflow to the river, but it dominated only a 26.4 km transition zone and disappeared at T7. Both the taxonomical method and functional groups revealed that there were longitudinal variations in phytoplankton compositions in this eutrophic lake-to-river system: the shift from lacustrine phytoplankton at the upper reaches to fluvial algae at the lower reaches and the shift from the planktonic coda (M, H1, LO ) to the lotic codon TB . Another significant difference that distinguished the Tanglang River from other rivers without eutrophic lakes upstream was that the total abundance decreased downstream. Our results demonstrated that functional groups were sensitive to the changes of aquatic environment, and thus they were useful to identify the spatial variations of phytoplankton in lake-to-river systems. They were also helpful in distinguishing the water conditions (stratified or mixed) via compositions of functional groups. Overall, the field observations demonstrated the existence of spatial variations in phytoplankton structures in the Tanglang River downstream from the eutrophic Dianchi Lake. The reason for this spatial variation was believed to be the change in hydraulic condition. But how flow affects the algae variation is not clear yet, and whether such variation occurs at other times or in other aquatic environments is still an open question. Further research is required to focus on these problems. Acknowledgements We acknowledge the financial support of the National Natural Science Foundation of China (No. 51039002 and No. 51279078) and State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University (No. 2012490811). References Abonyi, A., Leitão, M., Lanc¸on, A.M., Padisák, J., 2012. Phytoplankton functional groups as indicators of human impacts along the River Loire (France). Hydrobiologia 698 (1), 233–249, http://dx.doi.org/10.1007/s10750-012-11300. Allan, J.D., Castillo, M.M., 1995. Stream Ecology. Chapman and Hall, London. Bahnwart, M., Hübener, T., Schubert, H., 1998. Downstream changes in phytoplankton composition and biomass in a lowland river–lake system (Warnow River, Germany). Hydrobiologia 391 (1–3), 99–111, http://dx.doi.org/ 10.1023/A:1003558209411. ˜ Becker, V., Caputo, L., Ordónez, J., Marcé, R., Armengol, J., Crossetti, L.O., Huszar, V.L., 2010. Driving factors of the phytoplankton functional groups in a deep Mediterranean reservoir. Water Res. 44 (11), 3345–3354, http://dx.doi.org/10. 1016/j.watres.2010.03.018.

7

Borics, G., Várbíró, G., Grigorszky, I., Krasznai, E., Szabó, S., Tihamer Kiss, K., 2007. A new evaluation technique of potamo-plankton for the assessment of the ecological status of rivers. Arch. Hydrobiol. 17 (Suppl. (3–4)), 465–486, http:// dx.doi.org/10.1127/Ir/17/2007/466. Bridgeman, T.B., Chaffin, J.D., Kane, D.D., Conroy, J.D., Panek, S.E., Armenio, P.M., 2012. From river to lake: phosphorus partitioning and algal community compositional changes in Western Lake Erie. J. Great Lakes Res. 38 (1), 90–97, http://dx.doi.org/10.1016/j.jglr.2011.09.010. Chen, Y.C., Yu, Q., Zhu, D.J., Liu, Z.W., 2014. Possible influencing factors on phytoplankton growth and decay in rivers: review and perspective. J. Hydroelect. Eng. 33 (4), 186–195 (in Chinese). Devercelli, M., 2006. Phytoplankton of the Middle Paraná River during an anomalous hydrological period: a morphological and functional approach. Hydrobiologia 563 (1), 465–478, http://dx.doi.org/10.1007/s10750-006-00360. Devercelli, M., O’Farrell, I., 2013. Factors affecting the structure and maintenance of phytoplankton functional groups in a nutrient rich lowland river. Limnologica 43 (2), 67–78, http://dx.doi.org/10.1016/j.limno.2012.05.001. Directive, W.F., 2000. Directive 2000/60. EC of the European Parliament and of the Council of, 23(1). Friedrich, G., Pohlmann, M., 2009. Long-term plankton studies at the lower Rhine/Germany. Limnologica 39 (1), 14–39, http://dx.doi.org/10.1016/j.limno. 2008.03.006. Hilton, J., O’Hare, M., Bowes, M.J., Jones, J.I., 2006. How green is my river? A new paradigm of eutrophication in rivers. Sci. Total Environ. 365 (1), 66–83, http:// dx.doi.org/10.1016/j.scitotenv.2006.02.055. Hu, H.J., Wei, Y.X., 2006. The Freshwater Algae of China: Systematic, Taxonomy and Ecology. Science Press, Beijing (in Chinese). Huisman, J., van Oostveen, P., Weissing, F.J., 1999. Species dynamics in phytoplankton blooms: incomplete mixing and competition for light. Am. Nat. 154 (1), 46–68, http://dx.doi.org/10.1086/303220. Huisman, J., Arrayás, M., Ebert, U., Sommeijer, B., 2002. How do sinking phytoplankton species manage to persist? Am. Nat. 159 (3), 245–254, http:// dx.doi.org/10.1086/338511. Huisman, J., Sharples, J., Stroom, J.M., Visser, P.M., Kardinaal, W.E.A., Verspagen, J.M., Sommeijer, B., 2004. Changes in turbulent mixing shift competition for light between phytoplankton species. Ecology 85 (11), 2960–2970, http://dx. doi.org/10.1890/03-0763. Istvánovics, V., Honti, M., Vörös, L., Kozma, Z., 2010. Phytoplankton dynamics in relation to connectivity, flow dynamics and resource availability—the case of a large, lowland river, the Hungarian Tisza. Hydrobiologia 637 (1), 121–141, http://dx.doi.org/10.1007/s10750-009-9991-6. Jeppesen, E., Søndergaard, M., Jensen, J.P., Havens, K.E., Anneville, O., Carvalho, L., Winder, M., 2005. Lake responses to reduced nutrient loading – an analysis of contemporary long-term data from 35 case studies. Freshw. Biol. 50 (10), 1747–1771, http://dx.doi.org/10.1111/j.1365-2427.2005.01415.x. Kufel, L., Le´sniczuk, S., 2014. Hydrological connectivity as most probable key driver of chlorophyll and nutrients in oxbow lakes of the Bug River (Poland). Limnologica 46, 94–98, http://dx.doi.org/10.1016/j.limno.2013. 10.008. Li, M., Xu, K., Watanabe, M., Chen, Z., 2007. Long-term variations in dissolved silicate, nitrogen, and phosphorus flux from the Yangtze River into the East China Sea and impacts on estuarine ecosystem. Estuar. Coast. Shelf S. 71 (1), 3–12, http://dx.doi.org/10.1016/j.ecss.2006.08.013. ˇ ´ M., Spoljari ´ D., Stevic, ´ F., Cvijanovic, ´ V., Hackenberger Kutuzovic, ´ B., Mihaljevic, c, 2010. The influence of extreme floods from the River Danube in 2006 on phytoplankton communities in a floodplain lake: Shift to a clear state. Limnologica 40 (3), 260–268, http://dx.doi.org/10.1016/j.limno.2009.09.001. Mischke, U., Venohr, M., Behrendt, H., 2011. Using phytoplankton to assess the trophic status of German rivers. Int. Rev. Hydrobiol. 96 (5), 578–598, http://dx. doi.org/10.1002/iroh.201111304. Mitrovic, S.M., Hardwick, L., Dorani, F., 2011. Use of flow management to mitigate cyanobacterial blooms in the Lower Darling River. Aust. J. Plankton Res. 33, 229–241, http://dx.doi.org/10.1093/plankt/fbq094. Padisák, J., Crossetti, L.O., Naselli-Flores, L., 2009. Use and misuse in the application of the phytoplankton functional classification: a critical review with updates. Hydrobiologia 621 (1), 1–19, http://dx.doi.org/10.1007/s10750-008-9645-0. Paerl, H.W., Huisman, J., 2008. Blooms like it hot. Science 320 (5872), 57–58, http:// dx.doi.org/10.1126/science.1155398. Phillips, G., Pietiläinen, O.P., Carvalho, L., Solimini, A., Solheim, A.L., Cardoso, A.C., 2008. Chlorophyll–nutrient relationships of different lake types using a large European dataset. Aqua. Ecol. 42 (2), 213–226, http://dx.doi.org/10.1007/ s10452-008-9180-0. Qian, C.Y., Deng, X.Y., Xu, J.H., Wang, R.N., Zhao, J.C., Zhu, Y.X., 1985. The investigation of the algae in the Tanglang River. J. Yunnan U. (Sci. Tech.). 7, 123–137 (in Chinese). Qin, B., Zhu, G., Gao, G., Zhang, Y., Li, W., Paerl, H.W., Carmichael, W.W., 2010. A drinking water crisis in Lake Taihu, China: linkage to climatic variability and lake management. Environ. Manage. 45 (1), 105–112, http://dx.doi.org/10. 1007/s00267-009-9393-6. Reynolds, C.S., 1989. Physical determinants of phytoplankton succession. In: Plankton ecology. Springer, Berlin, Heidelberg, pp. 9–56, http://dx.doi.org/10. 1007/978-3-642-74890-5 2. Reynolds, C.S., Huszar, V., Kruk, C., Naselli-Flores, L., Melo, S., 2002. Towards a functional classification of the freshwater phytoplankton. J. Plankton. Res. 24 (5), 417–428, http://dx.doi.org/10.1093/plankt/24.5.417.

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007

G Model LIMNO-25501; No. of Pages 8 8

ARTICLE IN PRESS Q. Yu et al. / Limnologica xxx (2016) xxx–xxx

Reynolds, C.S., 2006. Ecology of Phytoplankton. Cambridge University Press, Cambridge. Salmaso, N., Braioni, M.G., 2008. Factors controlling the seasonal development and distribution of the phytoplankton community in the lowland course of a large river in Northern Italy (River Adige). Aquat. Ecol. 42 (4), 533–545, http://dx.doi. org/10.1007/s10452-007-9135-x. Salmaso, N., Zignin, A., 2010. At the extreme of physical gradients: phytoplankton in highly flushed, large rivers. Hydrobiologia 639 (1), 21–36, http://dx.doi.org/ 10.1007/s10750-009-0018-0. Soares, M.C.S., Huszar, V.L., Roland, F., 2007. Phytoplankton dynamics in two tropical rivers with different degrees of human impact (Southeast Brazil). River Res. Appl. 23 (7), 698–714, http://dx.doi.org/10.1002/rra.987. Soares, M.C.S., Marinho, M.M., Azevedo, S.M., Branco, C.W., Huszar, V.L., 2012. Eutrophication and retention time affecting spatial heterogeneity in a tropical reservoir. Limnologica 42 (3), 197–203, http://dx.doi.org/10.1016/j.limno. 2011.11.002. Sullivan, B.E., Prahl, F.G., Small, L.F., Covert, P.A., 2001. Seasonality of phytoplankton production in the Columbia River: a natural or anthropogenic pattern? Geochim. Cosmochim. Acta 65 (7), 1125–1139, http://dx.doi.org/10. 1016/S0016-7037(00)00565-2. Tavernini, S., Pierobon, E., Viaroli, P., 2011. Physical factors and dissolved reactive silica affect phytoplankton community structure and dynamics in a lowland

eutrophic river (Po river, Italy). Hydrobiologia 669 (1), 213–225, http://dx.doi. org/10.1007/s10750-011-0688-2. Twiss, M.R., Ulrich, C., Kring, S.A., Harold, J., Williams, M.R., 2010. Plankton dynamics along a 180 km reach of the Saint Lawrence River from its headwaters in Lake Ontario. Hydrobiologia 647 (1), 7–20, http://dx.doi.org/10. 1007/s10750-010-0115-0. Wetzel, R.G., 2001. Limnology: Lake and River Ecosystems. Academic, San Diego, CA. Xiao, L.J., Wang, T., Hu, R., Han, B.P., Wang, S., Qian, X., Padisák, J., 2011. Succession of phytoplankton functional groups regulated by monsoonal hydrology in a large canyon-shaped reservoir. Water Res. 45 (16), 5099–5109, http://dx.doi. org/10.1016/j.watres.2011.07.012. Yu, Q., Chen, Y.C., Zhu, D.J., Liu, Z.W., 2013. Pollutant source apportionment of the middle and lower reaches of the Pudu River in Southwest China. In: 2013 IAHR Congress, Chengdu. Yu, Q., Chen, Y.C., Liu, Z.W., van de Giesen, N., Zhu, D.J., 2015a. The influence of a eutrophic lake to the river downstream: spatiotemporal algal composition changes and the driving factors. Water 7 (5), 2184–2201, http://dx.doi.org/10. 3390/w7052184. Yu, Q., Liu, Z.W., Chen, Y.C., Zhu, D.J., 2015b. Modelling daily variation in the vertical distribution of Microcystis. China Environ. Sci. 35 (6), 1840–1846 (in Chinese).

Please cite this article in press as: Yu, Q., et al., Longitudinal variations of phytoplankton compositions in lake-to-river systems. Limnologica (2016), http://dx.doi.org/10.1016/j.limno.2016.02.007