Microcystis aeruginosa strengthens the advantage of Daphnia similoides in competition with Moina micrura

Microcystis blooms are generally associated with zooplankton shifts by disturbing interspecific relationships. The influence of Microcystis on competitive dominance by different sized zooplanktons showed species-specific dependence. We evaluated the competitive responses of small Moina micrura and large Daphnia similoides to the presence of Microcystis using mixed diets comprising 0%, 20%, and 35% of toxic M. aeruginosa, and the rest of green alga Chlorella pyrenoidosa. No competitive exclusion occurred for the two species under the tested diet combinations. In the absence of M. aeruginosa, the biomasses of the two cladocerans were decreased by the competition between them. However, the Daphnia was less inhibited with the higher biomass, suggesting the competitive dominance of Daphnia. M. aeruginosa treatment suppressed the population growths of the two cladocerans, with the reduced carrying capacities. Nonetheless, the population inhibition of Daphnia by competition was alleviated by the increased Microcystis proportion in diet. As a result, the competitive advantage of Daphnia became more pronounced, as indicated by the higher Daphnia: Moina biomass ratio with increased Microcystis proportions. These results suggested that M. aeruginosa strengthens the advantage of D. similoides in competition with M. micrura, which contributes to the diversified zooplankton shifts observed in fields during cyanobacteria blooms.

Biomass inhibition of species and biomass ratio in competition. The biomass inhibition of the two species changed with time depending on the Microcystis proportion ( Fig. 3a-c). Without Microcystis addition, the Daphnia biomass in competition was sharply inhibited by ~57% during the initial 3 days, whereas that of Moina was promoted by the presence of Daphnia, as indicated by the negative values of biomass inhibition rate. Nonetheless, the biomass inhibition of Moina dramatically increased to ~61%, and was higher than that of Daphnia after 7 days (Fig. 3a). With 20% Microcystis in food, although the biomass inhibition of Daphnia increased along with time, it reached a maximum value of ~59% at the end of experiment, which was rapidly achieved on day 3 in groups without Microcystis. The biomass inhibition rate of Moina varied around that of Daphnia (Fig. 3b). With 35% Microcystis, although negative values were observed for the biomass inhibition of Moina at the initial 7 days, it sharply increased to ~49% from day 9. Nonetheless, the biomass inhibition rate of Daphnia was negative and approached zero during the experiment period (Fig. 3c).
The Daphnia: Moina biomass ratio provided an intuitive understanding for the differences in population dynamics between the two cladocerans (Fig. 3d). In cocultures, the biomass of Daphnia was always higher than that of Moina, regardless of the Microcystis treatment, as indicated by the ratio values >1. There were remarkable increases in the biomass ratio with increased Microcystis proportions in food. Nonetheless, the peak value of the ratio at 35% Microcystis appeared later than that at 20% Microcystis did. Moina carrying capacity ratio in cocultures was always higher than that in monocultures, although decreasing trends were observed with increased Microcystis proportions in food (Fig. 4b).
The population growth rates (r) of the two cladocerans were significantly affected by the Microcystis treatment, but not the competition (Table 1). When fed 100% Chlorella, the r of Moina was higher by ~42% than that of Daphnia. Microcystis treatment increased the r of Daphnia, but decreased the r of Moina. The r of Daphnia was remarkable higher by 50.7% at 20% Microcystis and by 77.6% at 35% Microcystis than those of Moina in cocultures, leading to the increased Daphnia: Moina growth rate ratio with increased Microcystis proportions in food (Fig. 5).

Discussion
The present study revealed that the large-sized D. similoides was superior to the small-sized M. micrura in competition under favorable food condition. When fed 100% Chlorella, D. similoides had higher biomass than the M. micrura did in cocultures, although the biomasses of the two cladocerans were suppressed by the competition (Figs 1 and 2). This is in accordance with the previous conclusion that abundant edible food favours the larger species to be superior competitor [25][26][27] . On condition that the carbon levels satisfy the food requirements of the animals, large species gathers food more efficiently, thereby decreasing the food availability of small species 28 . The large animals also generally have the stronger ability to ingest the food. Other mechanisms, such as age at first reproduction and embryonic developmental time, also contribute to the competitive outcomes among zooplankton species 29 .
Corresponding to the increased Microcystis proportion, different population responses to Microcystis were observed: faster to reach the carrying capacity with the subsequent increased population growth rate in D. similoides, but the opposite case in M. micrura. This result enriches the species-specific responses in zooplankton population growth to cyanobacteria 30 . Exposure to Microcystis would promote the large zooplankton (e.g., D. magna) to reach its maturity earlier, with the shortened reproduction age 23 . At a population level, these changes in life history traits facilitate the large species reaching its carrying capacity faster. Nonetheless, the biomasses and carrying capacities of the two cladocerans were finally decreased by Microcystis treatment (Figs 1, 2 and 4). This is highly related to the nutritional deficiencies and the toxicity of microcystins of the cyanobacteria for the zooplankton 7, 31, 32 . Given the biomass inhibition by Micorcystis treatment, the biomass difference between the two cladocerans in competition was enlarged with increased Microcystis proportion, as indicated by the increasing  Daphnia: Moina biomass ratio (Fig. 3). It is concluded that the competitive advantage of D. similoides over M. micrura was strengthened by Microcystis treatment.
Because of the relatively larger gape size with higher filtration on the filamentous cyanobacteria, large Daphnia is generally assumed to be more vulnerable to Microcystis than small species 33,34 . Nonetheless, the present strain of M. aeruginosa grows as unicell in laboratory, and the D. similoides and M. micrura are supposed to graze both the toxic and non-toxic cells equally due to their non-selective filtration. M. micrura assimilated little Microcystis when fed only Microcystis or even a mixture of Microcystis and Chlorella 35 . Given the present low proportions (<35%) of Microcystis in diet, large Daphnia can minimise the negative influence from cyanobacteria via microcystins detoxification 36 . Zhang et al. 37 studied that the large D. similoides assimilates low abundance of Microcystis with improved reproduction. Using a combined stable-isotope and fatty-acid approach, de Kluijver et al. 38 found that D. similis consumes live Microcystis cells. This supplies additional material and energy for the D. similoides growth in comparison with M. micrura. Repeated toxic cyanobacteria-exposure can also increase the tolerance of large Daphnia population to toxic Microcystis via improving antioxidant systems 39,40 . In addition, the low nutritional value of cyanobacteria for Daphnia promotes the offspring tolerance to toxic Microcystis 41 . Gustafsson et al. 13 studied that previous-exposure to toxic Microcystis increased the offspring fitness in Microcystis environment. This adaptation may result from improved survival, enhanced reproduction or faster development of offspring 42,43 . Although some small cladocerans are also studied to develop adaptation 34 , the severer biomass inhibition by Microcystis in M. micrura indicated its weaker adaptation compared with that in D. similoides in the current study. A comparative study on the phenotypic adaptations between the two species based on individual performances will be performed in the next work. In the presence of M. micrura, we surprisingly observed that the biomass of D. similoides at 35% Microcystis was slightly higher than that at 20% Microcystis. As the Moina was severer inhibited by competition with increased Microcystis, it is presumed that Daphnia in the 35% Microcystis-treated cocultures consumed more good food, thereby leading to the relatively higher biomass.
There is a great variation in the influence of cyanobacteria on zooplankton competition. The present observation is not consistent with the general recognition that Microcystis promotes the dominance of small-sized cladocerans, but instead supports several investigations demonstrating that large-sized cladocerans are superior competitors in cyanobacteria environment 22,24,44 . Besides the concentration tested in present study, many bloom-related variables, such as morphology and toxic property, affect the zooplankton shifts 7, 45-47 . In natural systems, the zooplankton composition during bloom is also regulated by other factors such as temperature and planktivorous fish 48,49 . Under the background of global warming, increasing temperature would enhance the effects of cyanobacteria on zooplankton with expansive blooms [50][51][52] . For example, the cladocenran offspring tolerance to toxic Microcystis can be promoted by maternal warming 53 . Fish predation also drives the zooplankton fluctuation 54 . The planktivorous fishes affect zooplankton via directly predation or by reducing edible phytoplankton abundance to zooplankton 55 . Recent study showed that planktivorous fishes associated with cyanobacteria promote the zooplankton community shift towards species with good escape ability and r-strategy in survival 56,57 . Therefore, besides the exploration on bloom-relevant factors influencing zooplanktons alone or in combinations, the trophic interactions between planktivorous fish, zooplankton and cyanobacteria require deeper studies to assess the zooplankton community structures during the expansive blooms under warming climate.

Materials and Methods
Cladocerans and algal food. Both the Daphnia similoides and Moina micrura were collected from Taihu Lake in China. The animals were then cultivated in laboratory by feeding 100% Chlorella pyrenoidosa at 25 °C for about three years. C. pyrenoidosa was pre-cultured in liquid BG-11 medium at 25 °C and illuminated at 45 μmol m −2 s −1 provided by fluorescent lamps in a light-dark period of 14:10 h. Log-phase C. pyrenoidosa were harvested by centrifugation at 6300 × g for 10 min and used as food. M. aeruginosa PCC7806 was obtained from the Freshwater Algae Culture Collection of the Institute of Hydrobiology (Wuhan, China). The cyanobacteria produce at least two types of microcystion (MC-LR and MC-RR) with a total content of 3.6 pg per cell via the high-performance liquid chromatography detection 58 . The cyanobacteria were axenically cultured under the same above conditions. Experimental protocol. Three food compositions were tested as 100% C. pyrenoidosa + 0% M. aeruginosa, 80% C. pyrenoidosa + 20% M. aeruginosa, and 65% C. pyrenoidosa + 35% M. aeruginosa, with the total carbon content of 1 mg L −1 . The two cladocerans all died in several days when fed on 50% Microcystis in diet based on our pre-experiment. The present-used proportions of M. aeruginosa were not highly toxic to cause elimination of either M. micrura or D. similoides, and thus suitable for competition experiments. Within each food composition, three cladoceran-treatments were set up: (1) 5 D. similoides (Ds) cultivated alone; (2) 5 M. micrura (Mm) cultivated alone; (3) 5 Ds and 5 Mm cultivated together. The initial sizes for D. similoides and M. micrura were averaged 0.65 mm and 1.2 mm in length on account of a Nis-elements image analyzer coupled with a Nikon light microscope. The experiment was carried out in 1-L beakers containing 500 mL culture media with one specific food composition. The experiment was performed in triplicate, resulting in 3 (food composition) × 3 (cultivation pattern) × 3 (replicate) = 27 beakers. All beakers were maintained at 25 °C in a temperature-controlled chamber and illuminated by 45 μmol m −2 s −1 fluorescent light with a light-dark period of 14:10 h. To maintain constant food concentrations, we replaced 50% of the medium in each beaker daily with fresh medium with appropriate food abundance. The dry biomass of the animals was estimated via measuring the body length on the basis of the regression curves by Culver et al. 59 . The experiment was not terminated until no remarkable increase was detected in the population abundance. When the experiment was finished, the cladoceran populations were composed of individuals in different ages with a length range of 0.7-1.6 mm for D. similoides and 0.4-0.9 mm for M. micrura.
The cladoceran biomass versus time was fitted by using the Logistic model: where B 0 and B t represent the cladoceran biomasses at initial time (t 0 ) and time t, r represents the population growth rate, and K represents the environmental carrying capacity. The biomass inhibition of species in competition was calculated as: biomass inhibition rate (%) = [(biomass in monocultures) − (biomass in cocultures)]/ biomass in monocultures × 100%. The biomass ratio of the two cladocerans in cocultures was defined as the biomass of Daphnia relative to that of Moina (Daphnia: Moina).
Statistical analysis. All data are presented as mean ± 1 SE. Two-way ANOVA was used to compare differences between groups in terms of population growth rate, maximum biomass, time to maximum biomass, and the carrying capacity with food combination and absence/presence of competitor as the fixed factors. Significant analyses were followed by Tukey's post-hoc tests to locate meaningful differences. Statistical analysis was performed using Sigmaplot 11.0 software.