© EDP Sciences, 2019

1 Introduction

Intensive interspecific competition for limited resource can often result in the exclusion of inferior competitors, decrease of the species diversity and can alter the structure of the zooplankton community (Lynch, 1979; DeMott, 1989; Dumont, 1994; Bonsall and Hassell, 1997; Chase et al., 2002), but fluctuating environmental conditions such as variability of available resources can promote coexistence of zooplankton competitors (DeMott, 1989; Rothhaupt, 1990; McCauley et al., 1996; Kirk, 1997; Nisbet et al., 1997).

Competition among rotifer species is influenced by various factors such as body size, feeding habits, food type and nutritional quality, temperature, salinity, food level, inoculation density and diapause (Rothhaupt, 1988, 1990; DeMott, 1989; Boraas et al., 1990, Sarma et al., 1999, 2002; Fernández-Araiza et al., 2005; Montero-Pau and Serra, 2011; Divya et al., 2012; Li and Niu, 2015; Rebolledo et al., 2018). Body size is an important factor in determining competitive ability and controlling the outcome of competition in rotifers under different food levels (Sarma et al., 1996, 1999; Ooms-Wilms, 1998; Ciros-Pérez et al., 2001; Divya et al., 2012). Field studies have revealed that under food-limited conditions, smaller rotifer species are able to reproduce and maintain a population, and thus outcompete the larger species (Sarma et al., 1996, 1999), but larger rotifer species are unable to maintain a population because of differences in the available energy versus the maintenance costs (Downing and Rigler, 1984). On the other hand, at very high food levels, smaller species are unable to utilize food since their filtration systems get clogged, leading to poor population growth. However, at medium food levels, the outcome of competition in rotifers needs more investigation.

Food levels may modify the outcome of competition between two rotifer species which differ widely in their growth rates. When two species of rotifers compete for limited food, normally the one with higher population growth rate may be expected to outcompete the other with lower growth rate (Rothhaupt, 1990). However, this does not appear to hold true for many species (Sarma et al., 1996, 1999, 2007; Sarma and Nandini, 2002; Divya et al., 2012). When two planktonic rotifers, Brachionus angularis and B. calyciflorus, compete for limited food resource, the relationship between the outcome of competition and their population growth rates remains unknown.

Initial inoculation density is also an important factor affecting the population growth rates and controlling the outcome of competition in rotifers. High food levels usually lead to smaller rotifer species poor population growth, but their sufficiently initial density may reverse this condition (Sarma et al., 1996, 1997, 1999; Divya et al., 2012). Similarly, a low Chlorella level leads to larger rotifer species B. patulus poor population growth, but a higher ratio of B. patulus to Euchlanis dilatata at the onset of the experiments permits both rotifer species to coexist until the end of the experiments (Nandini and Sarma, 2002). At different food levels, what is the initial inoculation density required by B. angularis and B. calyciflorus to coexist?

B. angularis and B. calyciflorus are two common herbivorous rotifer species in freshwater lakes and ponds. Compared to B. calyciflorus, B. angularis is smaller and has a lower threshold food level but a higher population growth rate (Stemberger and Gilbert, 1987; Pan et al., 2017, 2018). The present study examined the combined effects of algal food level and initial inoculation density on the competitive interaction between the two herbivorous rotifer species, with the aim of testing the following two hypotheses: (i) at the low food level, B. angularis outcompeted B. calyciflorus and vice versa at the high food level, according to the available results (Sarma et al., 1996, 1999; Ooms-Wilms, 1998; Divya et al., 2012); and (ii) similar to the outcome of competition between the rotifers E. dilatata and B. patulus (Nandini and Sarma, 2002), a certain food level and inoculation density might permit the competition persistence between B. calyciflorus and B. angularis.

2 Materials and methods

2.1 Sample collection and culture

Individuals of B. calyciflorus Pallas and B. angularis Gosse, which were obtained by hatching the resting eggs in sediments from Lake Jinghu (31°36′11″ N, 118°38′23″ E) and identified morphologically under a microscope, were clonally cultured in rotifer culture medium (Gilbert, 1963). All clones were cultured in the laboratory for over 1 year, and one clone of each species was randomly selected for the experiments. The average lorica length (in µm, mean ± standard deviation, based on 100 individuals) of B. calyciflorus and B. angularis were 217 ± 4 and 99 ± 1, and the average lorica width were 167 ± 2 and 80 ± 3, respectively. Prior to the experiments, the two rotifer clones were mass-cultured. For clonal and mass cultures of the rotifers, an illumination incubator with a 16:8-h light:dark photoperiod at 130 lx at (23 ± 1)°C was used, and 1.5 × 10⁶ cells ml⁻¹ of Scenedesmus obliquus (Turp.) Kütz was supplied as food. The algal cells were semicontinuously cultured in HB-4 medium (Li et al., 1959), and those at the exponential phase of growth were harvested by centrifugation at 3,000 rpm for 5 min, resuspended in rotifer culture medium and stored at 4 °C. The density of algal cells was determined by counting using a haemocytometer.

2.2 Competitive experiments and parameter calculation

Competitive experiments between B. calyciflorus and B. angularis were conducted in 25 ml glass beakers each containing 20 ml rotifer culture medium with the chosen density of S. obliquus. Prior to the competitive experiments, the two rotifer clones were maintained at the designated food levels for more than 5 days to allow acclimation. For the competitive experiments, three S. obliquus densities (0.5 × 10⁶, 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹) and four initial inoculation densities (numerically, 100% B. calyciflorus; 75% B. calyciflorus and 25% B. angularis; 50% each of the two species; 25% B. calyciflorus and 75% B. angularis, and 100% B. angularis) were chosen. In all, the starting density of rotifers (alone or combined) was five individuals ml⁻¹. Four replicates were set up for each treatment.

Following inoculation, every day three aliquot samples of 0.5–1 ml were taken from each of the beakers, and the number of the rotifers was counted. After counting, the three aliquot samples were returned to the original beaker, and the culture medium was changed using fresh culture medium with appropriate algal density. The experiment was terminated after 13 days when all populations nearly completed one cycle.

The population growth rate was obtained using the following exponential equation:where N ₀ and N_t are the initial and final population densities; t is time in days. We used varying data points along the growth curve to calculate the mean per replicate. In general, we took four to six data points during the exponential phase of the population as documented in Dumont and Sarma (1995).

2.3 Statistical analyses

All statistical analyses were performed using SPSS 11.5. The Levene's test was performed to test the homogeneity of variances. One-way analysis of variance (ANOVA) was conducted to identify the significant effect of initial inoculation density and food level on each of the population growth variables of the rotifers cultured at each algal density and initial inoculation density, respectively. Two-way ANOVA was conducted to analyse the significant effects of initial inoculation density, food level and their interactions on each population growth variable. Multiple comparisons of the least significant difference were performed to determine which groups were significantly different among the four groups at each food level as well as the three groups at each initial inoculation density. Results with P values of less than 0.05 were considered statistically significant.

3 Results

The population growth curves of B. calyciflorus and B. angularis showed increased abundance in relation to food level. When both species were introduced together, and at 0.5 × 10⁶ cells ml⁻¹ of S. obliquus, B. angularis outcompeted B. calyciflorus regardless of initial inoculation density. This trend reversed as the food density offered increased to 2.0 × 10⁶ cells ml⁻¹. At 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, and when the initial inoculation density of B. angularis was one-third and the same as that of B. calyciflorus, B. angularis was displaced by B. calyciflorus; but when the initial inoculation density of B. angularis was three times that of B. calyciflorus, both species coexisted until the termination of the experiment (Fig. 1).

Only when S. obliquus density was 2.0 × 10⁶ cells ml⁻¹ was the time required to reach the maximum population density by B. calyciflorus significantly affected by inoculation density (One-way ANOVA, P < 0.05). The time required to reach the maximum population density by B. calyciflorus at the inoculation densities of 100, 75 and 25% were similar but were shorter than that at 50%. When the inoculation densities of B. calyciflorus were 100 and 50%, the time required to reach the maximum population density by B. calyciflorus was affected by food level (One-way ANOVA, P < 0.01). At both inoculation densities, the time required to reach the maximum population density at 0.5 × 10⁶ and 1.0 × 10⁶ cells ml⁻¹ of S. obliquus were similar but were shorter than that at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus (Fig. 1).

Only when S. obliquus density was 2.0 × 10⁶ cells ml⁻¹ was the time required to reach the maximum population density by B. angularis significantly affected by inoculation density (One-way ANOVA, P < 0.05). The time required to reach the maximum population density by B. angularis at the inoculation densities of 100, 75 and 50% were similar but were longer than that at 25%. Similarly, only when the inoculation density of B. angularis was 25% was the time required to reach the maximum population density by B. angularis affected by food level (One-way ANOVA, P < 0.05). The time required to reach the maximum population density at 0.5 × 10⁶ cells ml⁻¹ of S. obliquus was longer than that at 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, with the latter two being similar (Fig. 1).

Two-way ANOVA indicated that the time required to reach the maximum population density by both species was significantly affected only by food density (Tab. 1). In general, at 0.5 × 10⁶ and 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, B. calyciflorus reached the peak population abundance earlier than B. angularis regardless of initial inoculation density; but at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, this trend reversed. In the absence of competing species, B. calyciflorus reached the peak population abundance on days 3–4, 4–5 and 6–7 following inoculation at 0.5 × 10⁶, 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, respectively. The corresponding values for B. angularis were days 6–7, 5–7 and 6 (Fig. 1; Tab. 1).

At all the three food levels, the maximum population density achieved by B. calyciflorus was not significantly affected by inoculation density (one-way ANOVA, P > 0.05). However, at each inoculation density, the maximum population density achieved by B. calyciflorus was markedly influenced by food level (one-way ANOVA, P < 0.01). When the inoculation densities of B. calyciflorus were 100, 75 and 50%, the maximum population density achieved by B. calyciflorus increased with increasing food level. When the inoculation density of B. calyciflorus was 25%, the maximum population densities at 0.5 × 10⁶ and 1.0 × 10⁶ cells ml⁻¹ of S. obliquus were similar, but were lower than that at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus (Fig. 2).

At all the three food levels, the maximum population density achieved by B. angularis was significantly affected by inoculation density (one-way ANOVA, P < 0.01). At 0.5 × 10⁶ cells ml⁻¹ of S. obliquus, the maximum population density achieved by B. angularis was the highest at the inoculation density of 100% and the lowest at the inoculation densities of 50 and 25%. At 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, the maximum population density was the highest at the inoculation density of 100% and the lowest at the inoculation density of 25%. The maximum population density at the inoculation of 50% was similar to those at the inoculation densities of 75 and 25%. When the inoculation densities of B. angularis were 100 and 25%, the maximum population density achieved by B. angularis was markedly influenced by food level (one-way ANOVA, P < 0.01). The maximum population densities at 0.5 × 10⁶ and 1.0 × 10⁶ cells ml⁻¹ of S. obliquus were similar, but were lower than that at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus (Fig. 2).

Two-way ANOVA indicated that the maximum population densities achieved by B. calyciflorus and B. angularis were significantly affected by food level, inoculation density and their interaction (Tab. 1). In general, at any given food density, B. angularis was numerically more abundant than B. calyciflorus. When grown alone, B. angularis reached the peak abundance values of 265 ± 8 (mean ± standard error), 330 ± 30 and 802 ± 87 individuals ml⁻¹ at 0.5 × 10⁶, 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, respectively. The corresponding values for B. calyciflorus were 34 ± 4, 69 ± 5 and 101 ± 9 individuals ml⁻¹ (Fig. 2).

At each food level, the population growth rate of B. calyciflorus was not significantly affected by inoculation density (One-way ANOVA, P > 0.05). Similarly, at each inoculation density, the population growth rate of B. calyciflorus was not influenced by food level (one-way ANOVA, P > 0.05). However, at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, the population growth rate of B. angularis was significantly affected by inoculation density (P < 0.01). The population growth rate of B. angularis was the highest at the inoculation density of 25% and the lowest at the inoculation densities of 50 and 75%. At each inoculation density, the population growth rate of B. angularis was influenced by food level (One-way ANOVA, P < 0.01). When the inoculation density of B. angularis was 100 and 25%, the population growth rate increased with increasing food level. When the inoculation density of B. angularis was 50 and 75%, the population growth rates were similar at 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹ of S. obliquus and were higher than at 0.5 × 10⁶ cells ml⁻¹ of S. obliquus (Fig. 3).

Two-way ANOVA indicated that the population growth rate of B. calyciflorus was significantly affected only by food level but that of B. angularis was influenced by food level, inoculation density and their interaction (Tab. 1). In general, at 0.5 × 10⁶ cells ml⁻¹ of S. obliquus, B. calyciflorus had a higher population growth rate than B. angularis; but at 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, the reverse was also true. At 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, their population growth rates were similar. When grown alone, B. angularis had population growth rates of 0.623 ± 0.020, 0.770 ± 0.036 and 0.871 ± 0.013 d⁻¹ at 0.5 × 10⁶, 1.0 × 10⁶ and 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, respectively. The corresponding values for B. calyciflorus were 0.608 ± 0.032, 0.654 ± 0.033 and 0.518 ± 0.039 d⁻¹ (Fig. 3).

Fig. 1 Population growth curves of B. calyciflorus (filled circle) and B. angularis (unfilled circle) cultured at different Scenedesmus obliquus levels and inoculation densities. Series I: 100% B. calyciflorus or B. angularis; II: initial proportion of 75% B. calyciflorus + 25% B. angularis; III: 50% B. calyciflorus + 50% B. angularis; IV: 25% B. calyciflorus + 75% B. angularis. Shown are the mean + standard error values based on four replicate recordings.

Fig. 2 The maximum population density (mean + standard error, based on three replicates) achieved by B. calyciflorus and B. angularis cultured at different Scenedesmus obliquus levels and inoculation densities. The inoculation densities used for B. calyciflorus are as follows: I − 100% B. calyciflorus; II − 75% B. calyciflorus; III − 50% B. calyciflorus; IV − 25% B. calyciflorus. For B. angularis, the inoculation densities A, B, C and D represent 100, 75, 50 and 25% B. angularis, respectively. Shown are the values mean + standard error based on four replicates. Small letters indicate means that are similar (same letter) or different (different letters) among three food levels when inoculated at each of four densities (LSD multiple comparison), and capital letters indicate means that are similar (same letter) or different (different letters) among four inoculation densities when fed 0.5 × 106 (white bars), 1.0 × 106 (grey bars) and 2.0 × 106 (black bars) cells ml−1 of S. obliquus, respectively (LSD multiple comparison).

Fig. 3 The population growth rate (mean + standard error, based on three replicates) achieved by B. calyciflorus and B. angularis cultured at different Scenedesmus obliquus levels and inoculation densities. The inoculation densities used for B. calyciflorus are as follows: I − 100% B. calyciflorus; II − 75% B. calyciflorus; III − 50% B. calyciflorus; IV − 25% B. calyciflorus. For B. angularis, the inoculation densities A, B, C and D represent 100, 75, 50 and 25% B. angularis, respectively. Shown are the values mean + standard error based on four replicates. Small letters indicate means that are similar (same letter) or different (different letters) among three food levels when inoculated at each of four densities (LSD multiple comparison), and capital letters indicate means that are similar (same letter) or different (different letters) among four inoculation densities when fed 0.5 × 106 (white bars), 1.0 × 106 (grey bars) and 2.0 × 106 (black bars) cells ml−1 of S. obliquus, respectively (LSD multiple comparison).

4 Discussion

Classical population growth curves of rotifers in batch cultures with living algal cells as food show a lag phase, an exponential growth phase, a post-exponential growth phase and a final declining phase. The stationary phase, if it appears, is very short (Yuferal and Navarro, 1995). With the rotifer species B. plicatilis as test animals, Yoshinaga et al. (2001) observed a stationary phase instead of the decline phase. The present results showed that the population growth curves of both B. calyciflorus and B. angularis differed with food level and initial inoculation density.

It is well known that the increase in food availability results in increased population abundance and higher growth rates of rotifers (Sarma et al., 1996, 1999; Ooms-Wilms, 1998; Divya et al., 2012). The present study showed that in general the population abundance of both B. calyciflorus and B. angularis increased with increasing food availability. The population growth rate of B. angularis also increased with increasing food level but that of B. calyciflorus was the highest at 1.0 × 10⁶ cells ml⁻¹ of S. obliquus. The diverse responses in population growth rate to increasing food level might be attributed to different rotifer species.

When two species of rotifers compete for limited food resource, normally the one with higher population growth rate may be expected to outcompete the other with lower growth rate (Rothhaupt, 1990). However, this does not appear to hold true for many species (Sarma et al., 1996, 1999, 2007; Sarma and Nandini, 2002; Divya et al., 2012). The present study showed that at 0.5 × 10⁶ cells ml⁻¹ of S. obliquus, B. calyciflorus had a higher population growth rate than B. angularis, but the former was displaced by the latter. At 2.0 × 10⁶ cells ml⁻¹ of S. obliquus, B. calyciflorus had a lower population growth rate than B. angularis, but the former outcompeted the latter. At 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, their population growth rates were similar but the outcome of competition between them differed with their initial inoculation densities.

Stemberger and Gilbert (1985a) defined the food threshold for rotifers as the concentration of food required to maintain a zero population growth rate. When competitors were continuously exposed to food depleting situations, species with an ability to maintain at least a zero population growth rate should persist over the others that could not do so. It is known that smaller rotifer species have lower threshold food levels than larger rotifer species (Stemberger and Gilbert 1985b), which implies that under food-limited conditions, smaller rotifer species should be able to reproduce and maintain a population and thus outcompete the larger species. Under food-limited conditions, both A. fissa and B. patulus with relatively smaller lorica sizes outcompeted the larger species B. calyciflorus (Sarma et al., 1996, 1999), and B. rotundiformis with a smaller lorica length formed superior competitor than B. plicatilis with a larger lorica length (Divya et al., 2012). Identical to these results, the present study showed that at 0.5 × 10⁶ cells ml⁻¹ of Scenedesmus, smaller species B. angularis outcompeted larger species B. calyciflorus. At 2.0 × 10⁶ cells ml⁻¹ of Scenedesmus, B. angularis initially increased its population but thereafter declined due to increased food limitation, not only from its own population but also from continuously growing B. calyciflorus, which was identical to the conclusion of DeMott (1989) that the higher growth rates of some zooplanktonic species would allow them to reach densities at which inter- and intraspecific competition would become important. The outcome of competition between B. angularis and B. calyciflorus at the low and high food levels supported the hypothesis that at the low food level, B. angularis outcompeted B. calyciflorus and vice versa at the high food level.

The inoculation density of two competing species is generally thought to play a decisive role in the competitive outcome, but the accumulating evidences showed that its impact was also related to food level and growth characteristics of the competing species. At higher food levels, higher inoculation densities of inferior competitors helped themselves coexist with superior ones (Sarma et al., 1996; Nandini and Sarma, 2002; Divya et al., 2012). Similarly, the present study showed that at 1.0 × 10⁶ cells ml⁻¹ of S. obliquus, and when the initial inoculation density of B. angularis was three times that of B. calyciflorus, both species coexisted until the termination of the experiment, which supported the hypothesis that a certain food level and inoculation density might permit the competition persistence between B. calyciflorus and B. angularis.

In natural water bodies, zooplankton organisms are exposed to large-scale changes in their physical, chemical and biotic environments that may impact the demographic structure of populations. In consequence, many zooplankton species are often found during a restricted season (King and Serra, 1998). In Lake Jinghu, a subtropical shallow lake, B. angularis appeared during the period from the end of July to the end of October, 2008, and reached the peak population density of 0.075 individuals ml⁻¹ in early August (Wen et al., 2011), much lower than those obtained in the present study. Similarly, B. calyciflorus appeared during the period from the end of December, 2008, to the beginning of April, 2009, and reached the peak density of 0.035 individuals ml⁻¹ in mid-February (Wen et al., 2016), also much lower than those obtained in the present study. Because of seasonal distribution, interspecific competition between these two species might not occur.

5 Conclusion

The present study demonstrates that the outcome of competition between two differently sized herbivorous rotifer species is not only dependent on food density but also on relative initial population densities, and on the interaction of these two factors. In nature, it is likely that small B. angularis colonize oligotrophic water bodies more successfully than larger B. calyciflorus.

Acknowledgements

This work was funded by the Natural Science Foundation of China (31470015, 31170395) and the Foundation of Provincial Key Laboratory of Biotic Environment and Ecological Safety in Anhui Province.

References

All Tables

All Figures

Fig. 1 Population growth curves of B. calyciflorus (filled circle) and B. angularis (unfilled circle) cultured at different Scenedesmus obliquus levels and inoculation densities. Series I: 100% B. calyciflorus or B. angularis; II: initial proportion of 75% B. calyciflorus + 25% B. angularis; III: 50% B. calyciflorus + 50% B. angularis; IV: 25% B. calyciflorus + 75% B. angularis. Shown are the mean + standard error values based on four replicate recordings.

In the text

Fig. 2 The maximum population density (mean + standard error, based on three replicates) achieved by B. calyciflorus and B. angularis cultured at different Scenedesmus obliquus levels and inoculation densities. The inoculation densities used for B. calyciflorus are as follows: I − 100% B. calyciflorus; II − 75% B. calyciflorus; III − 50% B. calyciflorus; IV − 25% B. calyciflorus. For B. angularis, the inoculation densities A, B, C and D represent 100, 75, 50 and 25% B. angularis, respectively. Shown are the values mean + standard error based on four replicates. Small letters indicate means that are similar (same letter) or different (different letters) among three food levels when inoculated at each of four densities (LSD multiple comparison), and capital letters indicate means that are similar (same letter) or different (different letters) among four inoculation densities when fed 0.5 × 106 (white bars), 1.0 × 106 (grey bars) and 2.0 × 106 (black bars) cells ml−1 of S. obliquus, respectively (LSD multiple comparison).

In the text

Fig. 3 The population growth rate (mean + standard error, based on three replicates) achieved by B. calyciflorus and B. angularis cultured at different Scenedesmus obliquus levels and inoculation densities. The inoculation densities used for B. calyciflorus are as follows: I − 100% B. calyciflorus; II − 75% B. calyciflorus; III − 50% B. calyciflorus; IV − 25% B. calyciflorus. For B. angularis, the inoculation densities A, B, C and D represent 100, 75, 50 and 25% B. angularis, respectively. Shown are the values mean + standard error based on four replicates. Small letters indicate means that are similar (same letter) or different (different letters) among three food levels when inoculated at each of four densities (LSD multiple comparison), and capital letters indicate means that are similar (same letter) or different (different letters) among four inoculation densities when fed 0.5 × 106 (white bars), 1.0 × 106 (grey bars) and 2.0 × 106 (black bars) cells ml−1 of S. obliquus, respectively (LSD multiple comparison).

In the text