Host density can increase infection rates and reduce host fitness as increasing population density enhances the risk of becoming infected either through increased encounter rate or because host condition may decline. Conceivably, potential hosts could take high host density as a cue to up-regulate their defence systems. However, as host density usually covaries with food availability, it is difficult to examine the importance of host density in isolation. Thus, we performed two full-factorial experiments that varied juvenile densities of Daphnia magna (a freshwater crustacean) and food availability independently. We also included a simulated high-density treatment, where juvenile experimental animals were kept in filtered media that previously maintained Daphnia at high-density. Upon reaching adulthood, we exposed the Daphnia to their sterilizing bacterial parasite, Pasteuria ramosa, and examined how the juvenile treatments influenced the likelihood and severity of infection (Experiment I) and host immune investment (Experiment II). Neither juvenile density nor food treatments affected the likelihood of infection; however, well-fed hosts that were well-fed as juveniles produced more offspring prior to sterilization than their less well-fed counterparts. By contrast, parasite growth was independent of host juvenile resources or host density. Parasite-exposed hosts had a greater number of circulating haemocytes than controls (i.e., there was a cellular immune response), but the magnitude of immune response was not mediated by food availability or host density. These results suggest that density dependent effects on disease arise primarily through correlated changes in food availability: low food could limit parasitism and potentially curtail epidemics by reducing both the host’s and parasite’s reproduction as both depend on the same food.
Citation: Schoebel CN, Auld SKJR, Spaak P, Little TJ (2014) Effects of Juvenile Host Density and Food Availability on Adult Immune Response, Parasite Resistance and Virulence in a Daphnia-Parasite System. PLoS ONE 9(4): e94569. https://doi.org/10.1371/journal.pone.0094569
Editor: Nicole M. Gerardo, Emory University, United States of America
Received: October 22, 2013; Accepted: March 18, 2014; Published: April 15, 2014
Copyright: © 2014 Schoebel et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: For the present study CNS and PS were supported by the Swiss National Science Foundation (grant 3100 AO-116470; http://www.snf.ch/E/Pages/default.aspx); and the Swiss ETH-Board (CESS-Gedihap; http://www.cces.ethz.ch). SKRJA was funded by NERC (http://www.nerc.ac.uk/funding/available/fellowships/); and TJL by the Wellcome Trust (http://www.wellcome.ac.uk). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Host fitness decline due to parasitism (often termed virulence) is commonly context dependent –. For example, recent theoretical studies suggest that the expression of virulence depends on host population density, such that infected hosts have a higher sensitivity to density, and hence reach their carrying capacity earlier than uninfected hosts , . Moreover, since increased host density is thought to enhance the potential of parasite transmission , , elevated juvenile host densities may predict increased likelihood of infection at the adult stage, and thus act as a cue for hosts to shift investment into immune defences. Yet, increased immune preparedness can potentially come with costs - either energetic cost of investment or immunopathological cost when responses are launched . Indeed, immune functions have been shown to trade off with other life history traits. For example, survival was reduced in bumblebees and beetles with challenged immune systems , , and in Indian meal moths and sticklebacks, an increase in resistance was correlated with longer development time , .
Changes in host density are likely to be accompanied by changes in food availability and hence host condition (i.e. fitness). This is termed negative density dependence, and is commonly observed in animal and plant populations , , . Moreover, the incidence and severity of parasitism is often highest when the host is under stressful conditions, for example very low food conditions. This was discussed for the Daphnia galeata – Caullerya mesnili host-parasite system in a Swiss lake, where infection levels peak in autumn, when host density is still high, but food level is low . An increase in parasite-induced effects under food stress has also been shown in other systems (e.g. parasite specific mortality rates under starvation in snail hosts infected with either of two different micorparasites ), and this may partly be explained by the difficulty of maintaining energetically expensive immune functions under low food conditions , . However, for butterfly larvae immune parameters were negatively affected by high-density, while starvation did not have any effect  and in females of a parthenogenetic freshwater snail, a reduction in reproduction and growth (key fitness traits) was detected under high-density and constant food conditions, compared to low-density .
Varying the environmental quality experienced by juvenile hosts may shed light on investment in immune defences and patterns of parasitism in adult hosts. But to what extent is variation in juvenile host condition driven by host density itself (as opposed to correlated effects of food availability)? Progress in understanding the consequences of changing density for parasitism will be aided by experimental designs that simultaneously and independently study variation in host density and variation in host condition (as determined by food). This is a challenging task, because these two factors are normally not independent in natural and experimental systems. Therefore, a treatment of simulated high-density (SHD) could help to disentangle the effects of actual crowding and sensed crowding onto host condition and immunity. For example, for aquatic hosts, a SHD treatment could involve maintaining low densities of hosts in filtered media that previously contained large populations of hosts , . In aquatic systems, SHD treatments allow hosts to sense and release waterborne chemical cues without actual physical crowding . However, SHD treatments are not feasible for many model systems and, to our knowledge no study has achieved this goal. We use this experimental approach with a natural host-parasite system: Daphnia magna and its bacterial parasite Pasteuria ramosa, because previous studies have shown that Daphnia are able to sense and react to chemical cues dissolved in the surrounding water that indicate crowding . We present the results of two cross-factored experiment with four different food and three density treatments, including low, high and simulated high-density (a total of 12 treatments). In the first experiment, we record the proportion of hosts that suffer infection following parasite exposure during the adult stage, as well as measures of host fitness in healthy and parasite-infected hosts. In the second experiment, we record immune investment (haemocyte number) in control and parasite-exposed adult hosts.
Materials and Methods
We used the cyclically parthenogenetic freshwater crustacean D. magna and its parasite, the spore-forming bacterium P. ramosa, an obligate endoparasite that infects via horizontal transmission of mature spores from dead hosts by ingestion of spores . Successful infection can be easily seen by eye, since Pasteuria causes gigantism, red colouration and obvious bacterial growth in the haemolymph of the Daphnia (symptoms are visible 8–25 days post-exposure). Infection severely curtails host fitness, with hosts typically becoming completely sterile . We used a single genotype of D. magna (GG4) and a single parasite isolate (Sp1). Both the host genotype and parasite isolate were collected in 1997 from a pond near Gaarzefeld, Northern Germany and maintained in laboratory populations ever since (first studied by ). Sampling permission was not required, as neither D. magna, nor P. ramosa are protected or endangered in Germany.
We performed two separate experiments: Experiment I examined how the probability of infection and the virulence (defined here as the reduction in host fecundity due to infection) were affected by juvenile food and density conditions (following a fully-factorial design); Experiment II was identical in setup, but was dedicated to testing how juvenile food and density conditions affected the numbers of circulating haemocytes in parasite-exposed and non-exposed hosts , . Throughout the experiments, Daphnia were kept in artificial Daphnia media (ADaM ) in a 20°C incubator with a 12∶12 hour light: dark cycle. They were fed daily with 1.0 absorbance (abs) of chemostat-grown Chlorella vulgaris per Daphnia (one abs is the optical absorbance of 650 nm white light by the algae culture). Prior to the experiments, independent host replicates were maintained for three generations in order to minimize variation in maternal effects. Daphnia were kept under standard conditions in groups of 5 animals in 200 ml of media. Three times per week they were transferred to fresh media and any offspring were discarded. Second-clutch neonates from the third generation were used to set up the experimental units. The experimental treatments consisted of four pre-exposure food treatments - from excess food - 2 abs, down to low food conditions at 0.25 abs, with intermediate levels of 0.5 and 1.0 abs algae per day and Daphnia.
In addition, hosts were kept under three different density pre-exposure treatments: high-density (HD), with 15 Daphnia in a 200 ml jar of fresh media, low-density (LD), with 5 Daphnia in 200 ml of fresh media and simulated high-density (SHD), with 5 Daphnia in 200 ml of filtered media that previously maintained 15 Daphnia. The SHD media was filtered through pore size 45 µm inert filters (Sartoban 300, Sartorius Stedim Biotech, Germany) to remove particles such as excess food or moulted Daphnia carapaces. In order to collect enough SHD media, we recycled and filtered water from all HD replicates (except the ones fed 2.0 abs of food) as well as additional SHD “water factories” consisting of age matched Daphnia kept in HD conditions outside the experiment. We measured host age at first reproduction and size of first clutch to assess the effects of food and density conditions during the juvenile stage.
Parasite exposure was carried out in an identical way for both experiments. Each jar had its own day of maturity. On the day Daphnia within a jar reached maturity (defined as the day that over half of the Daphnia in each jar had deposited eggs into their brood chamber), five Daphnia (all Daphnia from the LD and SHD jars and five randomly-chosen Daphnia from the HD jars) were exposed to the parasite treatment for five hours as follows. Five Daphnia of each replicate-treatment combination were placed in one well of a 24 well cell plate (Costar, Corning Inc., NY) containing 1 ml of media. Parasite-exposed replicates received 50 000 P. ramosa spores, a dose commonly used in D. magna experiments –, , . Non-exposed control replicates did not receive any spores, but were placed in a cell plate well for the same amount of time. This resulted in 12 (8 for Experiment II) replicates per food, per parasite, and per density treatment, leading to 288 (Experiment I) and 192 (Experiment II) experimental units (i.e. jars). Permits are not required in order to conduct laboratory experiments with D. magna and P. ramosa.
After parasite exposure, a single randomly-chosen Daphnia per replicate was kept in an individual jar each containing 60 ml of fresh media and fed 1 abs of C. vulgaris per day; the others were discarded. Throughout the experiment, jars were randomly distributed within trays of 24, and tray position within the incubator was randomized daily to reduce the impact of any positional effects. Jars were checked daily for offspring and mortality. If a female had a clutch, the offspring were counted, and the mother (experimental individual) was placed into fresh media. Media was changed every 3 days regardless of whether or not the female had a clutch. On day 25 post-exposure, the proportion of Daphnia that had become infected was recorded.
This experiment was terminated on day 35 post-exposure. Surviving hosts were frozen individually in a 1.5 ml Eppendorf tube for spore counting. Daphnia that died prior to termination of the experiment were frozen for spore counting purposes on the day of death. Parasite spore number was taken as a measure of parasite fitness, and was determined as follows: the host body was crushed in 500 µl de-ionised water using a plastic Pellet Pestle; this solution was then vortexed and spores were counted using a CASY Cell Counter (Model TT) according to the manufacturer’s protocol. Subsequently, the number of transmission spores per Daphnia was calculated from the respective dilutions. For a schematic drawing of the experimental protocol see Fig. 1.
Both experiments included three pre-exposure host densities: high-density (striped bar), low-density (white bar) and simulated high-density (grey bar), and D. magna were fed four different food levels 0.25, 0.5, 1.0 and 2.0 absorbances (abs; one abs is the optical absorbance of 650 nm white light by the algae culture). On the day that D. magna reached maturity (between day 8–12) they were exposed to the bacterial parasite P. ramosa for 5 hours. Each jar had its own day of maturity. On day 35 post exposure (p.e.) all infected Daphnia were sacrificed and transmission spores were counted in A, while day X indicates the day the last healthy Daphnia had died and life history measures were terminated.
Five hours after parasite exposure, all five Daphnia of each replicate were placed in a Petri dish and the Daphnia hearts were pierced with a 25-guage needle (BD Microlance, Drogheda, Ireland). From each individual 0.5 µl of haemolymph was taken up with a 10 µl TipOne Repel Polymer Technology pipette tip (StarLab, Ahrensburg, Germany), pooled per replicate and mixed with 4 µl of ice-cold anticoagulant buffer (98 mM NaOH, 186 mM NaCl, 17 mM EDTA and 41 mM citric acid, pH adjusted to 4.5, ). Of this suspension, 4 µl were placed in a fertility-counting chamber 0.001 mm2×0.100 mm, Hawksley, Lancing, Sussex, UK), and the number of haemocytes was counted. For a schematic drawing of the experimental protocol see Fig. 1B.
Statistical Analyses, Experiment I
We first examined if juvenile density or food availability, or both affected the number of offspring in the first clutch and the age at first reproduction. As exposure to the parasite only occurred once the first clutch was laid in the brood chamber, these first clutch traits cannot have been influenced by the parasite. This analysis thus allowed us to detect if our pre-exposure food and density treatments were generally effective and impacted Daphnia life history. Size of the first clutch was analysed with a univariate general linear model (GLM), and age at first reproduction, being a time to event variable, was analysed with Cox regression (proportional hazards). Finally we studied the proportion of infected hosts and how infection was influenced by density and food with a generalized linear model (GLM) with binomial error distribution.
To determine how our treatments (juvenile density and food) affected the performance of infected hosts relative to uninfected hosts, we added infection status (two levels, healthy or infected) as an explanatory variable to the model that also included density and food level. Thus, it was the interactions between infection status and food and density levels that were the main explanatory variables of interest. We analysed the following response variables: number of offspring, host survival and parasite transmission spore production. The number of offspring and parasite transmission spores production were ln-transformed and studied with a univariate GLM. Time to host death and time to castration (measured as the day that offspring production ceased due to parasite infection) were analysed using a Cox-regression (proportional hazards). For the survival data we censored our data with 1 being individuals that were dead and 0 individuals still alive when last seen. The number of offspring per Daphnia did not include the first clutch, as it was unlikely to have been affected by the parasite treatment because of the timing of parasite exposure. All analyses were performed in SPSS 19.
Daphnia life history was significantly affected by both juvenile food and density treatments. There was also a food by density interaction for both these response variables (size of first clutch: F6, 384 = 4.82, p<0.001; age at first reproduction: N = 408, Wald-Chi2 = 20.59, p = 0.002) and higher food levels led to earlier age at first reproduction (N = 408, Wald-Chi2 = 9.78, p = 0.020, Fig. 2) and larger first clutches (F3, 384 = 55.47, p<0.001, Fig. 3); juvenile density only affected the age at first reproduction (N = 408, Wald-Chi2 = 13.42, p = 0.001, Fig. 2).
Note that this trait is not yet influenced by parasite infection. Black symbolizes the high-density treatment, light grey the low-density treatment, and dark grey the simulated high-density treatment.
Note that this trait is not yet influenced by parasite infection. Black symbolizes the high-density treatment, light grey the low-density treatment, and dark grey the simulated high-density treatment.
The mean prevalence of infection was 85% with prevalences ranging from 94.3% (0.5 abs food) to 76.1% for (1.0 abs). However, this was not significantly affected by the pre-exposure food (F3, 197 = 2.393, p = 0.070) or density treatments (F2, 197 = 1.003, p = 0.369). Time to castration took 10.9 days ±0.25 (SE) and was independent of both food treatment (Wald-Chi2 = 2.19, p = 0.534) and host juvenile density (Wald-Chi2 = 0.492, p = 0.782). Specifically, 50% of hosts were castrated by day 10, and only 2.4% of infected hosts had more than 3 clutches before castration. Uninfected hosts had 6.9±0.06 (SE) clutches.
Unexposed hosts had 65.07±11.2 (SE) offspring, whereas infected individuals produced only 21.13±6.0 (SE). However, we were most interested in how pre-exposure food or density modified consequences of infection, which would be evident as an infection status by treatment interaction with one of the response variables. Such interactions were not apparent for parasite transmission spore production or host mortality, but an interaction between infection status and juvenile food level was observed for total host fecundity (Table 1); infected hosts had fewer offspring in each clutch they had (Fig. 4), but this was modified by food such that infected hosts receiving the higher pre-exposure food levels had clutch sizes that approached those of uninfected hosts.
First clutch was removed from analysis since it was deposited before parasite exposure and thus does not influence the cost of infection. Black depicts P. ramosa infected animals, and grey uninfected D. magna. See Table 1 for statistical details.
Parasite-exposed Daphnia had more circulating haemocytes than controls (N = 238, F1, 214 = 10.44, p = 0.001). The pre-exposure density treatments affected haemocyte counts, with the number of haemocytes being highest in Daphnia that experienced high juvenile host density (HD) previous to parasite exposure (F2, 214 = 3.88, p = 0.022, Fig. 5). However, variation in pre-exposure food treatment did not affect haemocyte numbers (F3, 214 = 2.03, p = 0.111).
Environmental heterogeneity, such as variation in population density and food levels, may affect the expression of infection-related traits and thus alter host and parasite fitness , . The present study analysed the effects of varying juvenile host densities and food availability on D. magna fitness under infection with the sterilizing parasite P. ramosa. This study is among the first to experimentally test for the host density dependence of virulence, which has been proposed by recent theoretical models , , , and to disentangle pre-exposure host density from pre-exposure food availability.
Food treatment had a large impact on host fitness: juvenile Daphnia from the lowest food treatment produced 24% fewer offspring than those from the highest food treatment. However, neither the food nor the density treatments affected the probability of becoming infected. As expected, the infected hosts were eventually sterilised by the parasite. Even prior to complete sterilisation, infected hosts showed reduced fecundity. This reduction in fecundity was dependent on juvenile food treatment: the parasite-induced reduction in fecundity was large in the low food environment, but when food was abundant, infected and healthy Daphnia differed only slightly in their early reproduction (Fig. 4). It is important to note that sterilisation has the biggest effect on host fitness. However, small changes in numbers of offspring in early clutches could have large effects on population size in the future, given that early population growth is exponential. A decline in food availability could thus reduce the supply of susceptible hosts, lower the incidence of infection and potentially terminate epidemics. This is in line with results by Vale et al.  who studied the environmentally mediated tolerance to infection in the same host-parasite system. Vale et al.  observed a much more benign parasitic interaction under high food conditions, whereas under food-limited conditions, the parasite severely damaged its host. Whilst well-fed mothers will suffer less virulence, the offspring may in fact suffer more. Specifically, well-fed D. magna tend to produce smaller, lower-quality offspring that are also relatively susceptible to parasites, compared to offspring of mothers in poor condition –. And yet these offspring of well-fed mothers may reproduce more in the absence of parasites . The population dynamic and epidemiological consequences of these interactions between maternal food, current food, reproduction and susceptibility (or virulence) remain to be determined.
We estimated parasite fitness by measuring the production of transmission spores within infected hosts. Unlike host fecundity, parasite transmission spore production was not affected by the food treatment experienced by juvenile hosts. These results suggest that past low food conditions increase virulence without affecting parasite transmission potential (Fig. 4). These results were surprising, as low food for hosts should reduce host reproduction as well as parasite reproduction since both are using the same food source. Based on past studies , , we expected parasite fitness to be linked to host quality, and in particular that there would be many more spores produced under high food conditions. Both Ebert et al.  as well as Vale et al.  continued with food treatments after parasite exposure and collected parasite transmission spores on the day the hosts died. Hence, they collected spores from Daphnia that died “naturally”, i.e. once host death was induced by the parasite. In contrast, we stopped our food and density regimes upon parasite exposure (as we had to sacrifice all individuals from Experiment II at this point) and collected parasite spores at a fixed point in time, prior to host death. Past studies have shown that the number of spores produced will increase with time, but show similar patterns with respect to treatment effects . This is why we presumed that it was not strictly necessary for hosts to die before counting transmission spores. However, it remains conceivable that methodological differences between studies account for these discrepancies.
In D. magna, the cellular response is thought to occur as Pasteuria transmission spores pass from the gut to the haemocoel; it is a consequence of infection rather than a cause of resistance . We found juvenile Daphnia kept at high-density had the highest baseline (pre-exposure) haemocyte counts (Fig. 5). However, these high haemocyte counts were not associated with increased prevalence of infection in parasite-exposed hosts. The lack of statistical interaction between the juvenile density treatments and parasite exposure on haemocyte number shows that the juvenile density treatments neither strengthen nor weaken the host’s cellular response to parasite exposure. Still, we need to keep in mind that haemocytes are also involved in key physiological processes other than immunity , : higher baseline haemocyte numbers may thus reflect other physiological stresses associated with crowding. Testing very low densities (<5 Daphnia) while keeping food constant could help to disentangle the relationship between density and immunity for D. magna. A follow-up study could test haemocyte expression upon exposure to a parasite other than Pasteuria and to a non-pathogenic immune stimulant. Such an experiment could detect if there is a general immune system mechanism for which costs are expected to be high in a parasite free environment.
Simulated high-density (SHD) was included to study differences between actual, physical crowding and perceived crowding (which is likely mediated through chemical cues of conspecific individuals dissolved in the water). While the host’s cellular immune response significantly differed between SHD and HD, this was not in the direction we expected. Overall, the interpretation of the SHD results is not straightforward; for example, time to first reproduction under SHD (Fig. 2) was, compared to the other two juvenile host density treatments, shorter in low food conditions but longer in high food conditions. For the size of the first clutch (Fig. 3) we found the opposite: clutches under SHD were larger in low food conditions but smaller than those of the other density treatments when food level was high. Therefore, the reasons for the outcome of our experiment might be more complex than we expected, and SHD might not solely cue for crowded conditions. It could also signal anoxia and/or contain cues from degenerating algae, or there could be unknown ecological or physiological interactions between food level and crowding, including processes that we did not control for and of which we do not know the exact effects on Daphnia fitness.
Overall, the effect of pre-exposure host density was rather weak compared to other studies investigating host density without parasite exposure , . We did not detect a significant effect of juvenile host density (before parasite exposure) on host fecundity. However, unlike previous studies, we were able to disentangle juvenile host density from food availability. As we only found direct effects of food, our findings suggest that density-dependent effects act mainly through correlated effects on food availability. For the snail P. antipodarum, Neiman et al.  showed that negative-density dependence is mainly caused by food limitation, while Burns  detected depressed growth and lower reproduction in small-bodied Daphnia under crowded conditions. As in the present work, both studies observed complex effects between food level, host density, and host fitness. While the high and low host densities (15 and 5 individuals per 200 ml) we used lay within the range of a natural population , even higher host densities might be required to cause sufficient stress lasting over a long enough period in order to detect its responses in Daphnia life-history traits. Generally, there remains a lack of empirical studies combining (juvenile) food availability, (juvenile) host density and parasitism, and our results indicate how such multi-factorial interactions are much more complex than generally expected. To disentangle the effects of food availability, host density and parasite exposure we intentionally kept the genetic component constant for this study (one host clone and one parasite strain), but future experiments might profit from a higher number of host clones and parasite strains.
We thank Alena Gsell, Christian Rellstab, Nicole Gerardo and two anonymous reviewers for helpful comments on an earlier version of this manuscript, as well as Jen Scholefield and Phil Wilson for their assistance in the lab.
Conceived and designed the experiments: CNS TJL. Performed the experiments: CNS SKJRA. Analyzed the data: CNS TJL. Contributed reagents/materials/analysis tools: TJL PS. Wrote the paper: CNS SKJRA TJL PS.
- 1. Brown MJF, Loosli R, Schmid-Hempel P (2000) Condition-dependent expression of virulence in a trypanosome infecting bumblebees. Oikos 91: 421–427.
- 2. Jokela J, Taskinen J, Mutikainen P, Kopp K (2005) Virulence of parasites in hosts under environmental stress: experiments with anoxia and starvation. Oikos 108: 156–164.
- 3. Ebert D, Carius HJ, Little T, Decaestecker E (2004) The evolution of virulence when parasites cause host castration and gigantism. American Naturalist 164: S19–S32.
- 4. Anderson RM, May RM (1981) The population dynamics of microparasites and their invertebrate hosts. Philosophical Transactions of the Royal Society of London Series B Biological Sciences 291: 451–524.
- 5. Lively CM (2009) The maintenance of sex: host–parasite coevolution with density-dependent virulence. Journal of Evolutionary Biology 22: 2086–2093.
- 6. Lively CM (2006) The ecology of virulence. Ecology Letters 9: 1089–1095.
- 7. Lafferty KD, Holt RD (2003) How should environmental stress affect the population dynamics of disease? Ecology Letters 6: 654–664.
- 8. Sheldon BC, Verhulst S (1996) Ecological immunology: Costly parasite defences and trade-offs in evolutionary ecology. Trends in Ecology & Evolution 11: 317–321.
- 9. Moret Y, Schmid-Hempel P (2000) Survival for immunity: The price of immune system activation for bumblebee workers. Science 290: 1166–1168.
- 10. Sadd BM, Siva-Jothy MT (2006) Self-harm caused by an insect’s innate immunity. Proceedings of the Royal Society B: Biological Sciences 273: 2571.
- 11. Barber I, Arnott SA, Braithwaite VA, Andrew J, Huntingford FA (2001) Indirect fitness consequences of mate choice in sticklebacks: offspring of brighter males grow slowly but resist parasitic infections. Proceedings of the Royal Society B-Biological Sciences 268: 71–76.
- 12. Boots M, Begon M (1993) Trade-offs with resistance to a granulosis virus in the Indian meal moth, examined by a laboratory evolution experiment. Functional Ecology 7: 528–534.
- 13. Turchin P, Taylor A (1992) Complex dynamics in ecological time-series. Ecology 73: 289–305.
- 14. Brook B, Bradshaw C (2006) Strength of evidence for density dependence in abundance time series of 1,198 species. Ecology 87: 1445–1451.
- 15. Bonenfant C, Gaillard JM, Coulson T, Festa-Bianchet M, Loison A, et al. (2009) Empirical evidence of density-dependence in populations of large herbivores. In: Advances in Ecological Research Caswell H, editor. 41: 313–357.
- 16. Schoebel CN, Wolinska J, Spaak P (2010) Higher parasite resistance in Daphnia populations with recent epidemics. Journal of Evolutionary Biology 23: 2370–2376.
- 17. Jokela J, Lively CM, Taskinen J, Peters AD (1999) Effect of starvation on parasite-induced mortality in a freshwater snail (Potamopyrgus antipodarum). Oecologia 119: 320–325.
- 18. Seppälä O, Jokela J (2010) Maintenance of genetic variation in immune defense of a freshwater snail: role of environmental heterogeneity. Evolution 64: 2397–2407.
- 19. Siva-Jothy MT, Thompson JJW (2002) Short-term nutrient deprivation affects immune function. Physiological Entomology 27: 206–212.
- 20. Pisek M, Karl I, Franke K, Fischer K (2013) High larval density does not induce a prophylactic immune response in a butterfly. Ecological Entomology DOI:10.1111/een.12024.
- 21. Neiman M (2006) Embryo production in a parthenogenetic snail (Potamopyrgus antipodarum) is negatively affected by the presence of other parthenogenetic females. Invertebrate Biology 125: 45–50.
- 22. Seiz A (1984) Are there allelopathic interactions in zooplankton? Laboratory experiments with Daphnia.. Oecologia 62: 94–96.
- 23. Cope NJ, Winterbourn MJ (2004) Competitive interactions between two successful molluscan invaders of freshwaters: an experimental study. Aquatic Ecology 38: 83–91.
- 24. Boersma M, Spaak P, De Meester L (1998) Predator-mediated plasticity in morphology, life history, and behavior of Daphnia: The uncoupling of responses. American Naturalist 152: 237–248.
- 25. Ebert D, Rainey P, Embley TM, Scholz D (1996) Development, life cycle, ultrastructure and phylogenetic position of Pasteuria ramosa Metchnikoff 1888: rediscovery of an obligate endoparasite of Daphnia magna Straus. Philosophical Transactions of the Royal Society of London Series B Biological Sciences 351: 1689–1701.
- 26. Carius HJ, Little TJ, Ebert D (2001) Genetic variation in a host-parasite association: Potential for coevolution and frequency-dependent selection. Evolution 55: 1136–1145.
- 27. Auld SKJR, Scholefield JA, Little TJ (2010) Genetic variation in the cellular response of Daphnia magna (Crustacea: Cladocera) to its bacterial parasite. Proceedings of the Royal Society B-Biological Sciences 277: 3291–3297.
- 28. Auld SKJR, Edel KH, Little TJ (2012) The cellular immune response of Daphnia magna under host-parasite genetic variation and variation in initial dose. Evolution 66: 3287–3293.
- 29. Klüttgen B, Dulmer U, Engels M, Ratte HT (1994) Adam, an artificial freshwater for the culture of zooplankton. Water Research 28: 743–746.
- 30. Vale PF, Wilson AJ, Best A, Boots M, Little TJ (2011) Epidemiological, evolutionary, and coevolutionary implications of context-dependent parasitism. American Naturalist 177: 510–521.
- 31. Ebert D, Zschokke Rohringer CD, Carius HJ (2000) Dose effects and density-dependent regulation of two microparasites of Daphnia magna. Oecologia 122: 200–209.
- 32. Lavine MD, Strand MR (2002) Insect hemocytes and their role in immunity. Insect Biochemistry and Molecular Biology 32: 1295–1309.
- 33. Lazzaro BP, Little TJ (2009) Immunity in a variable world. Philosophical Transactions of the Royal Society B: Biological Sciences 364: 15–26.
- 34. Wolinska J, King KC (2009) Environment can alter selection in host-parasite interactions. Trends in Parasitology 25: 236–244.
- 35. Lively CM (2010) Parasite viruleeecne, host life history, and the costs and enefits of sex. Ecology 91: 3–6.
- 36. Stjernman M, Little TJ (2011) Genetic variation for maternal effects on parasite susceptibility. Journal of Evolutionary Biology 24: 2357–2363.
- 37. Guinnee MA, West SA, Little TJ (2004) Testing small clutch size models with Daphnia. American Naturalist 163: 880–887.
- 38. Mitchell SE, Read AF (2005) Poor maternal environment enhances offspring disease resistance in an invertebrate. Proceedings of the Royal Society B-Biological Sciences 272: 2601–2607.
- 39. Guinnee MA, Gardner A, Howard AE, West SA, Little TJ (2007) The causes and consequences of variation in offspring size: a case study using Daphnia. Journal of Evolutionary Biology 20: 577–587.
- 40. Little TJ, Chadwick W, Watt K (2007) Parasite variation and the evolution of virulence in a Daphnia-microparasite system. Parasitology 135: 303–308.
- 41. Rowley AF, Ratclife NA (1981) Invertebrate blood cells. London: Academic Press.
- 42. Strand MR (2008) The insect cellular immune response. Insect Science 15: 1–14.
- 43. Neiman M, Warren D, Rasmussen B, Zhang S (2013) Complex consequences of increased density for reproductive output in an invasive freshwater snail. Evolutionary Ecology Doi:10.1007/s10682-013-9632-4.
- 44. Burns CW (1995) Effects of crowding and different food levels on growth and reproductive investment of Daphnia. Oecologia 101: 234–244.
- 45. Duncan AB, Little TJ (2007) Parasite-driven genetic change in a natural population of Daphnia. Evolution 61: 796–803.