The biochemical composition and calorific content of a rotifer and its algal food: comparison of a two stage chemostat and batch culture

It is often assumed that the use of a two-stage chemostat yields algal food with a well-defined nutritional composition that can maintain herbivores in a steady state of growth. In this study I investigated two bacteriafree culture techniques, continuous flow chemostats and batch cultures, to determine whether the biochemical composition of the rotifer Encentrum linnhei differed in the two cultures. Changes in the biochemical composition and calorific content of the algal food were also examined. In the rotifer reaction vessel only the lipid content of the algal food increased significantly with dilution rates, while significant decreases in protein and carbohydrates were detected at increasing algal densities. A different pattern was observed in the response of the unused algal cells to variables such as dilution, algal input and algal densities in the sump of the rotifer chemostat. In the chemostat the biochemical composition of the rotifers varied as expected with dilution rates, algal input and food availability but significant differences were found in the biochemical composition of the animals growing in the reaction vessel and those collected from the sump. In contrast, the biochemical content of batch-grown E. linnhei varied with time in a way that depended upon food availability and also on the biochemical state of the algal food. However, at the end of the exponential phase of growth, when maximum densities had been achieved, batch-grown rotifers were more biochemically nutritious than chemostat-grown animals in their steady-state phase.     

Effect of Food Quantity and Quality on Population Growth Rate and Digestive Activity in the Euryhaline Rotifer Brachionus plicatilis Müller

We investigated the nutritional effects of both food quantity and quality on Brachionus plicatilis. Decomposition of particulate and dissolved organic matter by rotifer digestive enzymes play a crucial role in rotifer nutrition. Among other enzymes, rotifers produce phosphatases, non‐specific enzymes that allow for the release of orthophosphate from a variety of organic phosphorus compounds. Phosphatase saturation was measured in B. plicatilis homogenates using the spectrofluorimetric method. We examined population growth rate, reproduction and phosphatase activity in the homogenate of rotifers (PA_RH) fed by nutrient‐replete algal food supplied at different quantities. Population growth rate, number of eggs per individual and PA_RH were affected by food quantity. Growth rate and number of eggs per individual significantly increased in rotifers fed by food supplied at the highest quantity. The highest population growth rate was reached by rotifers fed by nutrient‐replete food, while it did not significantly differ between rotifers fed on nitrogen (N)‐depleted and phosphorus (P)‐depleted food. The number of eggs per individual was more affected by N than P supply. PA_RH and rotifer RNA content were not influenced by different food quality. The results indicate that B. plicatilis is not able to regulate its digestive apparatus in terms of efficiently getting access to essential nutrients when scarce, but do this when nutrient‐replete food is available in different quantity. (© 2009 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Intensive rotifer cultures using chemostats

Continuous production of the rotifer Brachionus plicatilis rotundiformis (S-type) in an intensive chemostat culture system has been investigated. The production dynamics of rotifers in relation to different flow rates and feed regimes show that the growth rate and production depends on the type of algal feed and flow rate utilized in the culture system. It was possible to achieve a mean production of up to 318.84 × 10⁶ rotifers m^–3 d^–1 at a flow rate of 6 1 h^–1 in 100 1 chemostats and up to 261.21 × 10⁶ rotifers m^–3 d^–1 at a flow rate of 40 1 h –1 while using 1 m³ capacity rotifer chemostats as production units. The 3 fatty acid composition of rotifers while using Chlorella and Nannochloropsis in the culture system has been described. The results of this investigation show that the rotifer productivity in the continuous culture system is considerably higher than in any of the conventional culture systems described to date for aquacultural purposes.This research was financed by the Kuwait Foundation for the Advancement of Sciences (KFAS), Kuwait, under a contract research project code 86-04-02.     

Steady-state rotifer growth in a two-stage, computer-controlled turbidostat

Steady-state populations of rotifers (Brachionus calyciflorus)were maintained in twostage, continuous-flow turbidostatic cultureon the green alga Chlorella pyrenoidosa. In this system, themaximum specific growth rate,µ^max of the rotifers wasmaintained by using a computer to control the concentrationof algae, as rotifer food, in the rotifer culture. As rotifersconsumed algae, the turbidity decreased until a set-point wasreached. Then fresh algal suspension (supplied from a steady-statealgal chemostat) was metered into the rotifer culture, whichwas held in the dark. Rotifer and algal populations, as wellas rotifer µ^max entered steady states. These steady-stateresults were consistent with previous data from chemostat studies,but growth transients indicated that the of the µ^maxrotifersmay be subject to selection. The system is unique in providinga means to explore population dynamics of a metazoan maintainednear its µ^max.     

Experimental study on the impact of dinoflagellate Alexandrium species on populations of the rotifer Brachionus plicatilis

To investigate harmful effects of the dinoflagellate Alexandrium species on microzooplankton, the rotifer Brachionus plicatilis was chosen as an assay species, and tested with 10 strains of Alexandrium including one known non-PSP-producer (Alexandrium tamarense, AT-6). HPLC analysis confirmed the PSP-content of the various strains: Alexandrium lusitanicum, Alexandrium minutum and Alexandrium tamarense (ATHK, AT5-1, AT5-3, ATCI02, ATCI03) used in the experiment were PSP-producers. No PSP toxins were detected in the strains Alexandrium sp1, Alexandrium sp2.Exposing rotifer populations to the densities of 2000 cells ml⁻¹ of each of these 10 Alexandrium strains revealed that the (non-PSP) A. tamarense (AT-6) and two other PSP-producing algae: A. lusitanicum, A. minutum, did not appear to adversely impact rotifer populations. Rotifers exposed to these three strains were able to maintain their population numbers, and in some cases, increase them. Although some increases in rotifer population growth following exposures to these three algal species were noted, the rate was less than for the non-exposed control rotifer groups.In contrast, the remaining seven algal strains (A. tamarense ATHK, AT5-1, AT5-3, ATCI02, ATCI03; also Alexandrium sp1 and Alexandrium sp2) all have adverse effects on the rotifers. Dosing rotifers with respective algal cell densities of 2000 cells ml⁻¹ each, for Alexandrium sp1, Alexandrium sp2, and A. tamarense strains ATHK and ATCI03 showed mean lethal time (LT₅₀) on rotifer populations of 21, 28, 29, and 36h, respectively. The remaining three species (A. tamarense strains AT5-1, AT5-3, ATCI02) caused respective mean rotifer LT₅₀s of 56, 56, and 71 h, compared to 160 h for the unexposed “starved control” rotifers. Experiments to determine ingestion rates for the rotifers, based on changes in their Chlorophyll a content, showed that the rotifers could feed on A. lusitanicum, A. minutum and A. tamarense strain AT-6, but could graze to little or no extent upon algal cells of the other seven strains. The effects on rotifers exposed to different cell densities, fractions, and growth phases of A. tamarense algal culture were respectively compared. It was found that only the whole algal cells had lethal effects, with strongest impact being shown by the early exponential growth phase of A. tamarense. The results indicate that some toxic mechanism(s), other than PSP and present in whole algal cells, might be responsible for the adverse effects on the exposed rotifers.     

Experimental studies on the seasonal variation of the rotifer Keratella cochlearis (Gosse)

A culture containing the rotifer Keratella cochlearis was divided into two aliquots, one reared at + 20°C and the second at + 5°C, all other conditions being identical. The experimentl lasted 74 days. At 20°C the animals did not undergo any noticeable morphological changes, while at 5°C the posterior spines grew longer, a development which had nearly reached its maximum as early as 30 days after the experiment began, i.e. after very few generations. The experiment shows that the temperature (and its direct consequences, such as slower growth) plays an important role in causing seasonal morphological variation in this rotifer. The results are in accordance with field experience as well as with culture experiments on other brachionid rotifers.     

Characterization of the bacterial flora of mass cultivated Brachionus plicatilis

Bacterial density and composition in association of mass cultivated rotifers (Brachionus plicatilis, SINTEF-strain) was investigated, during experimental conditions identical to the procedures used for preparing rotifers as live food for marine cold water fish larvae. These procedures include cultivation, enrichment with squid meal and acclimation to low temperature by storage of the rotifer culture at 6 °C. Large variations were observed in the number of rotifer associated (1.8–7.6 · 10³ colony forming units per rotifer^–1) and free-living (0.6–25 10⁷ cells·ml^–1) bacteria. An increase of 50–150% in the bacterial number was normally observed after feeding the rotifer with squid meal, but after three days of acclimation at 6 °C, the bacterial numbers decreased to the initial level.After enrichment of the cultures with squid meal, the similarity in the composition of the bacterial flora between the rotifers and water was reduced. However, acclimation of the culture at 6 °C resulted in better agreement of the rotifer associated flora and that in water. Enrichment of the cultures induced a shift in the bacterial composition from Cytophaga/Flavobacterium dominance to Pseudomonas/Alcaligenes dominance. The bacterial flora of the rotifer cultures are dominated by presumably opportunistic species after enrichment, which may have detrimental effects when rotifers are fed as live food to marine fish larvae.     

High density culture of the freshwater rotifer,Brachionus calyciflorus

The freshwater rotifer, Brachionus calyciflorus is one of the live food organisms used for the mass production of larval fish. In this study possibility of obtaining high density cultures of the freshwater rotifer B. calyciflorus were investigated. The two culture systems used differed in their air and dissolved oxygen supplies using three temperatures in each case: 24, 28 and 32 °C. Rotifers were batch-cultured using 5 l-vessels and fed with the freshwater Chlorella. The growth rate of rotifers significantly increased with an increase in temperature. The maximum density of the rotifers with air-supply at 24 °C, 6500 ind. ml^–1, was significantly lower than those cultured at 28 and 32 °C, i.e. 8600 and 8100 ind. ml^–1, respectively. Dissolved oxygen levels decreased with time and ranged from 0.8 to 1.4 mg l^–1 when the density of freshwater rotifer was the highest at each temperature. The highest density (19200 ind. ml^–1) of freshwater rotifer was obtained in cultures with a supply of oxygen at 28 °C. Densities of 13500 and 17200 ind. ml^–1 were found at 24 and 32 °C, respectively. Levels of NH₃-N increased with time and a dramatic increase of NH₃-N was observed at high temperatures. Levels of NH₃-N at 24, 28 and 32 °C were 13.2, 18.5 and 24.5 mg l^–1, respectively. These levels coincided with the highest rotifer density at each of the three temperatures. When rotifers were cultured with an oxygen-supply and pH was adjusted to 7, the maximum density of rotifer reached 33500 ind. ml^–1 at 32 °C . These results suggested that high density culture of freshwater rotifer, B. calyciflorus could be achieved under optimal conditions with DO value of exceeding 5 mg l^–1 and NH₃-N values of lower than 12.0 mg l^–1.     

Physiological adaptations in the survival of rotifers (Brachionus plicatilis,O. F. Müller) at low temperatures

The biology of the rotifer Brachionus plicatilis has been studied extensively in recent years, due to its importance in aquaculture. Today, the culture of several marine fish species relies completely on the daily production of live rotifers.In the present paper we explore the conditions that facilitate maintaining live rotifers for extensive periods at low temperatures. In addition to its possible contribution in providing reserve stocks for commercial application, these studies may be of ecological importance. They could explain some of the physiological adaptive mechanisms that are involved in the survival of rotifers under adverse environmental conditions.Experimental results showed relatively high survival rates (82–85%) in rotifers that were cultured at 25 °C and exposed later to -1 °C for 12–14 days. During this period, rotifers were kept without food and their media were not changed. The survival was found to depend on the rotifer culture conditions, prior to exposure to -1 °C. These included the type of food fed to rotifers (yeast or algae), the salinity in which they were cultured, and an essential acclimation period of 2–6 day at 4 °C or 10 °C. The acclimation period was associated with the synthesis of at least one specific protein and accumulation of lipids. Profiles of protein synthesis in rotifers incubated at 10 °C revealed a 94 kD protein, which did not appear in rotifers cultured at 25 or 37 °C. Immunoisolation, using a polyclonal antibody that was prepared against HSP60, revealed that this protein was synthesized in rotifers kept at 10, 25 or 37 °C. However, this antibody did not react with the 94 kD peptide.In addition, rotifers kept at 10 °C accumulated substantial amounts of lipids, including eicosapentaenoic acid (EPA), which is found in the algae fed to them. These results support the hypothesis of specific adaptations to survival at low temperatures during an acclimation period.     

Combined Effects of Food Concentration and Temperature on Competition Among Four Species of <Emphasis Type="Italic">Brachionus </Emphasis>(Rotifera)

We evaluated the effect of algal food density (1.5 × 10⁶, 3.0 × 10⁶ and 4.5 × 10⁶ cells ml⁻¹ of Chlorella) and temperature (22° and 28 °C) on competition among the rotifers Brachionus calyciflorus, Brachionus havanaensis, Brachionus patulus and Brachionus rubens, based on population growth experiments for 24 days. The growth experiments were conducted seperately for each individual rotifer species (i.e., controls), and in mixtures of all four species in equal initial proportions (i.e., under competition). The population growth of B. calyciflorus, B. havanaensis, B. patulus and B. rubens grown separately at two temperatures and at three algal food densities showed typical patterns of lag, exponential and retardation phases in the controls. This pattern differed considerably under competition. In general, we observed that in all of the test species, the highest growth rates were observed at higher food levels and in the absence of congenerics. At 22 °C, under the lowest food level, the differences in the population abundances of B. havanaensis, B. patulus and B. rubens grown alone and in the presence of competition were large. However, these differences reduced as food density was increased from 0.5 × 10⁶ to 4.5 × 10⁶ cells ml⁻¹. At 28 °C and at the lowest food level, all of the other rotifer species eliminated B. havanaensis in mixed cultures. Each brachionid species had a higher rate when grown alone than when cultured with other species. The highest r (mean ± standard error: 0.54 ± 0.01 day⁻¹) was recorded for B. havanaensis at 28 °C under 4.5 × 10⁶ cells ml⁻¹ of algal food density. At 28 °C at low algal food density, the presence of competitors resulted in negative population growth rates for three of the four rotifer species tested.     

Combined influences of particulate and dissolved factors in the toxicity of Microcystis aeruginosa (NRC-SS-17) to the rotifer Brachionus calyciflorus

The rotifer Brachionus calyciflorus is subject to the combined influence of particulate and dissolved factors derived from the toxic cyanobacterium Microcystis aeruginosa. Toxicity manifests itself through reduction of cohort survivorship and probably though reduction of cohort reproduction. Toxic effects are seen only when there is a particulate food material available in suspension, either M. aeruginosa itself or another food type which may act as a carrier for a dissolved toxic factor. Regardless of exact mechanism, the presence of a toxic strain of this blue-green may lead to diminution of population growth of co-occurring rotifer populations.     

Interactions between red tide microalgae and herbivorous zooplankton: effects of two bloom-forming species on the rotifer <Emphasis Type="Italic">Brachionus plicatilis</Emphasis> (O.F. Muller)

Experiments were carried out to investigate interspecific interactions between the rotifer Brachionus plicatilis and two harmful algal bloom (HAB) species using single and mixed culture methods. B. plicatilis populations and the growth of two algae were compared at different algal cell densities. The results demonstrate that B. plicatilis obtained sufficient nutrition from Alexandrium tamarense to support net population increase. When exposed to a density of 8 × 10⁴ cells ml⁻¹ A. tamarense, the number of B. plicatilis increased faster than it did when exposed to other four algal densities (16 × 10⁴, 24 × 10⁴, 32 × 10⁴, and 40 × 10⁴ cells ml⁻¹). Cell densities of A. tamarense decreased due to the grazing of B. plicatilis. In contrast, Heterosigma akashiwo had an adverse effect on the B. plicatilis population and its growth was largely unaffected by rotifer grazing. In this case, the B. plicatilis population decreased and H. akashiwo grew at a rate similar to that of a control without addition of rotifers. Mixed culture experiments showed that A. tamarense could partly counteract the effect of H. akashiwo in limiting the rate of population increase of rotifer. In addition, the effect of different initial cell densities on interspecific competition between A. tamarense and H. akashiwo in mixed culture(s) was also investigated. The results show that A. tamarense competed very successfully when the inoculation proportions of A. tamarense and H. akashiwo were 40:5 and 40:30. Handling editor: D. Hamilton     

Nutrient Utilization Strategies of Algae and Bacteria after the Termination of Nutrient Amendment with Different Phosphorus Dosage: A Mesocosm Case

The impacts of nutrient amendment termination on the growth strategies of algae and bacteria were conducted in experimentally designed mesocosm in which two different phosphorus (P) dosages were treated. The algal community composition did not change greatly in Group A (low phosphorus) and Group B (high phosphorus). In Group A, the secretion of bacterial alkaline phosphatase (AP) after nutrient termination stimulated bacterial phosphorus acquisition, which caused the decrease in algal phosphorus levels, in terms of the increase of bacterial abundance and bacterial production, as well as the decrease in chlorophyll a and particulate organic carbon. The algal collapse resulted in dissolved organic carbon secretion, further fuelling bacterial growth. In Group B, excess phosphorus input urged algae to store phosphorus as poly-phosphate. When phosphorus input ceased, in order to maintain their used high phosphorus demand, algae strengthened to gain phosphorus through the hydrolysis of dissolved organic phosphorus in water column and ploy-phosphate inside the cells by AP, evidenced by high algal alkaline phosphatase activity, algal growth continuation, and bacterial growth decline. These facts indicated that phosphorus content should reduce to a lower level than expected, so that algal bloom can be effectively controlled in eutrophic water bodies.     

An investigation of non-steady-state algal growth. II. Mathematical modelling of co-nutrient-limited algal growth

The ability of mathematical models to simulate competition for nutrients between three algal species, the diatom Thalassiosira pseudonana, a marine raphidophyte Heterosigma carterae and the dinoflagellate Alexandrium minutum, was investigated. Transient growth models were parameterized and tested using a number of closely controlled laboratory data sets including batch monocultures, batch competition experiments and semi-continuous culture competition experiments. The cell quota model of algal growth was found to be adequate to simulate growth of both the raphidophyte and the dinoflagellate. Batch monoculture data for diatom growth obtained under either nitrogen (N) or silicon (Si) limitation could also be simulated with a quota-style model, which in this case included feedback inhibition of nutrient uptake. However, to simulate both batch and semi-continuous culture experiments (and competition between the species), it was necessary to consider diatom Si-N metabolism. A model was derived which contains a representation of both intracellular N ad Si, and of the interaction of these nutrients within the cell. The model used a co-nutrient limitation based on the perceived functional and structural role of N and Si, respectively, within the cell. Simulations indicated that models capable of adequately representing monoculture growth in batch culture may produce erroneous results when incorporated into models of competition. The co-nutrient model is a first step to producing tractable algal growth models which will represent multiple nutrient stress in transient growth conditions.      

Dependence of temperature on loss rates of rotifers,lipids, and ω3 fatty acids in starved Brachionus plicatilis cultures

Rotifer cultures of Brachionus plicatilis (SINTEF-strain, length 250 m) rich in 3 fatty acids were starved for > 5 days at variable temperature (0–18 °C). The net specific loss rate of rotifer numbers were 0.04 day^–1 (range 0–0.08 day^–1) at 5–18 °C, but reached values up to 0.25 day^–1 at 0–3 °C. The loss rate was independent on culture density (range 40–1000 ind ml^–1), but was to some extent dependent on the initial physiological state of the rotifers (i.e., egg ratio).The loss rate of lipids was 0.02–0.05 day^–1 below 10 °C, where the potential growth rate of the rotifer is low (0–0.09 day^–1). The loss rate of lipids increased rapidly for higher temperatures where the rotifer can maintain positive growth, and reached 0.19 day^–1 at 18 °C. The Q₁₀ for the lipid loss rate versus temperature was higher than the Q₁₀ for respiration found in other strains. This may suggest that other processes than respiration were involved in lipid catabolism. The content of 3 fatty acids became reduced somewhat faster than the lipids (i.e. in particular 22:6 3), but the fatty acid per cent distribution remained remarkably unaffected by the temperature during starvation.The results showed that rotifer cultures could be starved for up to 4 days at 5–8 °C without essential quantitative losses of lipids, 3 fatty acids, and rotifers. The rotifers exhausted their endogenous lipids through reproduction (anabolism) and respiration (including enhanced locomotion) at higher temperatures. At lower temperatures, the mortality rate became very high.     

Algae and rotifer turbidostats: studies on stability of live feed cultures

A two stage turbidostat was developed according toBoraas & Bennet (1988), but with highly improvedturbidity sensors. The first stage was an algalturbidostat where algal density was regulated byturbidity measurements. Algal density was also heldconstant in the second stage (rotifer production)according to turbidity measurements. Additionally,the growth rates were monitored. The regulation system allowed an effective on-line process control.Initially, the production of rotifers in long-timestudies was variable. However, after furtherimprovements of the turbidity measurement, fluctuations in the rotifer turbidostat decreasedsignificantly.     

A novel growth-promoting protein in the conditioned media from the rotifer <Emphasis Type="Italic">Brachionus plicatilis</Emphasis> at an early exponential growth phase

We confirmed the existence of growth-promoting substances in the conditioned media (CM) from the rotifer Brachionus plicatilis at an early exponential growth phase and isolated a novel protein with a growth-promoting activity from the crude extract (CE) of rotifer cells. CM was prepared from the culture media where rotifers had been cultured at an early exponential growth phase and filtered through a 0.22-μm filter membrane. The growth-promoting activity was determined using rotifers in CM for 5 days. As a result, the increase of rotifers added with CM was significantly higher than that of the control in artificial seawater (P < 0.001). Moreover, the growth-promoting activity of CM was dose-dependent and inactivated by heat treatment at 80°C for 60 min. Meanwhile, CM filtered through a <10 kDa ultrafiltration membrane showed a low activity, whereas proteinase K treatment resulted in a complete inactivation. These results suggest that the rotifer secrets growth-promoting proteins into CM. CE also contained a protein with the activity and properties similar to those found in CM. Then, CE was subjected to purification of a growth-promoting protein for convenience using various types of chromatography after fractionation with 30–80% saturated ammonium sulfate. Subsequently, a protein with an approximate molecular weight of 25000 was isolated, and its N-terminal amino acid sequence was determined to be PAVVDFTAVWFGPLQMIKP. An orthologue was found in the EST database of B. plicatilis, the full sequence of which showed about 50% identity to the corresponding regions of thioredoxins from other organisms.     

Direct and indirect impact of two common rotifer species (Keratella spp.) on two abundant ciliate species (Urotricha furcata, Balanion planctonicum)

1. We investigated the potential competition and feeding impact of the common rotifer species, Keratella cochlearis and K. quadrata , on the abundant prostomatid ciliates, Urotricha furcata and Balanion planctonicum, in laboratory batch culture experiments. All four species have similar feeding preferences, co-occur in many freshwater environments, and are thus potential competitors for the same algal food.
2. Two small Cryptomonas species served as food for the ciliates and the rotifers in the experiments. Growth rates of each ciliate species were measured when they grew alone and when they were paired with one of the rotifer species.
3. Both rotifer species reduced the growth rate of U. furcata , probably primarily by direct feeding on the ciliates. Growth rate of B. planctonicum was unaffected by K. cochlearis , but was drastically reduced by grazing and/or mechanical interference of K. quadrata .
4. These results suggest niche partitioning of the sympatric ciliates with respect to their rotifer competitors/predators.     

Amphora margalefii Tomàs var. lacustris P. Sánchez var. nova,a new brackish water diatom

The rotifer Brachionus plicatilis can grow in a wide range of salinities and temperatures, but rapid shifts in both salinity and temperature may result in immobilized, non-swimming rotifers. The goal of this study was to examine the effect of perturbations in temperature and salinity on the swimming pattern of the rotifer.Only slight changes in mobility were observed when rotifers were exposed to changes in temperature (from 20 °C to 8–30 °C) and to an increase in salinity (from 20% to 30%). When the salinity was reduced to 15% and 5%, the proportion of mobile rotifers was reduced to 50% and 5%, respectively. The rotifers were throughout more resistant to perturbations in temperature than to those of salinity.Combined temperature and salinity perturbations compared to perturbations in each factor separately suggested a synergetic effect of temperature and salinity on the rotifers locomotion. Transfer from cultivation conditions to low salinity (5%) and high temperature (28 °C) resulted in very low percent of mobile rotifers (0–10%). However, if the temperature was reduced to 8 °C concomitant with the changes in salinity, the percent of mobile rotifers was 85%.Rotifers use a high share of their metabolic energy for locomotion, and it is therefore not surprising that perturbations in salinity and temperature may result in partial or complete immobilization.     

Linking herbivore-induced defences to population dynamics

1. Theoretical studies have shown that inducible defences have the potential to affect population stability and persistence in bi‐ and tritrophic food chains. Experimental studies on such effects of prey defence strategies on the dynamics of predator–prey systems are still rare. We performed replicated population dynamics experiments using the herbivorous rotifer Brachionus calyciflorus and four strains of closely related algae that show different defence responses to this herbivore. 2. We observed herbivore populations to fluctuate at a higher frequency when feeding on small undefended algae. During these fluctuations minimum rotifer densities remained sufficiently high to ensure population persistence in all the replicates. The initial growth of rotifer populations in this treatment coincided with a sharp drop in algal density. Such a suppression of algae by herbivores was not observed in the other treatments, where algae were larger due to induced or permanent defences. In these treatments we observed rotifer population densities to first rise and then decline. The herbivore went extinct in all replicates with large permanently defended algae. The frequency of herbivore extinctions was intermediate when algae had inducible defences. 3. A variety of alternative mechanisms could explain differential herbivore persistence in the different defence treatments. Our analysis showed the density and fraction of highly edible algal particles to better explain herbivore persistence and extinctions than total algal density, the fraction of highly inedible food particles or the accumulation of herbivore waste products or autotoxins. 4. We argue that the rotifers require a minimum fraction and density of edible food particles for maintenance and reproduction. We conjecture that induced defences in algae may thus favour larger zooplankton species such as Daphnia spp. that are less sensitive to shifts in their food size spectrum, relative to smaller zooplankton species, such as rotifers and in this way contributes to the structuring of planktonic communities.