Ecology, Epidemiology, and Evolution of Parasitism in Daphnia

  1. Introduction to the Ecology, Epidemiology, and Evolution of Parasitism in Daphnia
  2. Introduction to Daphnia Biology
  3. Some Parasites of Daphnia
  4. Parasitism in Natural Populations
  5. The Effects of Daphnia Parasites on Host Fitness
  6. Host Adaptations against the Costs of Parasitism
  7. Host Range of Daphnia Parasites
  8. Epidemiology
  9. Population Dynamics and Community Ecology
  10. Experiments with Daphnia and Parasites

8 Epidemiology

Epidemiology of infectious diseases attempts to describe the patterns and processes by which diseases are distributed in the host population. Here I present what is known about the transmission of Daphnia parasites, about the factors that influence transmission, and how they work together in shaping parasite dynamics. I further discuss two general models of parasite epidemiology, one for Daphnia populations in fishless ponds, another for Daphnia populations in lakes with planktivorous fish.

  1. Transmission
    1. Modes of Transmission in Daphnia: Parasite Systems
      1. Horizontal Transmission from the Living Host
      2. Horizontal Transmission from the Dead Hosts and Sediments
      3. Horizontal Transmission with a Two-Host Life Cycle
      4. Vertical Transmission
    2. Survival of Transmission Stages Outside the Host
    3. Uptake of Transmission Stages from Pond Sediments
    4. Factors Influencing Parasite Transmission
      1. Parasite Transmission Is Density Dependent
        1. Density-dependent Transmission in Natural Populations
        2. Experimental Evidence for Density-dependent Transmission
        3. Conclusions on Density-dependent Transmission
      2. Parasite Transmission Can Be Limited by Low Temperatures
      3. Host Stress Might Facilitate Parasite Spread
      4. Resistance May Limit the Spread of Diseases
      5. Summary of Transmission Limiting Factors
  2. Epidemiology of Daphnia Microparasites
    1. The Fishless Pond Model
    2. Suggestion for a Lake Model
  3. Conclusions and Open Questions

8.1 Transmission

In a parasitological context, epidemiology is the study of infectious diseases and disease-causing agents at the population level. It seeks to characterize the patterns of distribution and prevalence of the disease and the factors responsible for these patterns. In a more applied context, it also strives to identify and test prevention and treatment measures. The key factor to understanding the epidemiology of diseases is to understand transmission, or the movement of parasites from one host to the next.

In the following, I focus on four aspects of transmission: the mode of transmission, the survival of transmission stages, the uptake of transmission stages from sediments, and the factors that may limit transmission in natural populations.

8.1.1 Modes of Transmission in Daphnia: Parasite Systems

An important component of epidemiology is the parasite's mode of transmission, or how it moves from one host to the next. Unfortunately, surprisingly few scientific reports include information on parasite transmission. To my knowledge, the first description of a plankton parasite life cycle that tested mode of transmission was the description by Chatton (1925) of the amoeba Pansporella perplexa in Daphnia pulex. This parasite is transmitted between hosts via waterborne infective stages, which are released from infected hosts and are ingested by the same or other host individuals during filter feeding.

The modes of transmission of Daphnia parasites can be grouped into four types; these do not, however, exclude each other, because some parasites can be transmitted by more than one method (Figure 8.1

Figure 8.1
). Horizontal Transmission from the Living Host

This form of transmission is the typical mode of transmission for many human and livestock infectious diseases. Infected hosts release infective particles, which then infect other hosts (Figure 8.1A

Figure 8.1
). Influenza and measles are typical examples. This mode of transmission is frequently found among Daphnia parasites, particularly gut parasites, but also epibionts. Daphnia parasites that use this mode of transmission are the amoeba Pansporella perplexa, the microsporidia Glugoides intestinalis and Ordospora colligata, and the protozoan Caullerya mesnili. With these gut parasites, infected hosts carry comparatively few transmission stages at any one time (compared with the parasites falling into the next group), although they may produce many transmission stages during the lifetime of an infection.

To the best of our knowledge, all of the parasites in this category enter their hosts with the food. Food uptake by Daphnia is through filter feeding, and the rate at which Daphnia filter their food therefore plays an important role in the spread of a disease (Fels 2005).

Brood parasites have been observed to occur in numerous Daphnia populations. These are typically transmitted from one living host to the next. The most devastating ones are certain fungi, which kill the entire brood while it is developing in the brood chamber. Brood parasitic copepods may also be listed here. In contrast to the other parasites in this group, they actively search for their host and enter the brood pouch from behind. Horizontal Transmission from the Dead Hosts and Sediments

Parasites that infect tissues other than the host gut or body surface may have more problems leaving their hosts. These parasites often produce many transmission stages that are only set free after the host's parasite-induced death (Figure 8.1B

Figure 8.1
). By the time of the host's death, these obligate killers (Ebert and Weisser 1997) may produce up to 100 million transmission stages, which are all released at once. Killing the host to achieve transmission is common among insect parasites (many viruses and bacteria) but seems uncommon among parasites of vertebrates (Alien, the deadly extraterrestrial from the movie with the same name, which killed the human crew of a spaceship, is the only exception known to me). Examples of Daphnia parasites with this mode of transmission include the blood parasitic bacteria Pasteuria ramosa and White Fat Cell bacterium, the yeast Metschnikowia bicuspidata, and the microsporidium Octosporea bayeri.

Once set free from the dead host, the spores of parasites that kill obligately must reach another host to achieve transmission. If the pile of transmission stages left by a decaying host is stirred up, spores may be suspended in the water and infect filter-feeding hosts. Some Daphnia species tend to browse over substrates and thus come into contact with very high local concentrations of transmission stages in the sediments, which may then be ingested.

Although we do not currently know how P. ramosa enters the host, the closely related parasite P. penetrans enters its nematode host through the cuticula (Preston et al. 2003), which may be the same route used by P. ramosa. In this case, it would be the only known parasite of Cladocera that does not enter the host with the food.

Despite the apparent advantage of killing the host early to achieve transmission, parasite virulence with transmission from dead hosts varies greatly and ranges from rapid killers (e.g., White Fat Cell Disease) to parasites that have only a modest impact on host survival (e.g., P. ramosa, O. bayeri). The reasons for this large variation may be found in the specific biology of the parasites (Ebert and Herre 1996; Ebert and Weisser 1997), but our knowledge about the evolution of virulence is still rather rudimentary. Horizontal Transmission with a Two-Host Life Cycle

A number of parasites cycle through two or more host species to complete their life cycle (Figure 8.1C

Figure 8.1
). Among the Cladocera, however, there are only a few known examples of parasites with multi-host life cycles. This is surprising, because life cycles with two hosts are well known among parasite systems where at least one host lives in freshwater, including a number of human parasites, such as the medina worm (Dracunculus) and Schistosoma. The only known examples of Daphnia parasites with a two (or more) host life cycles are the nematode Echinuria uncinata, the cestode Cysticercus mirabilis (Green 1974), and an undescribed trematode parasite of D. obtusa (Schwartz and Cameron 1993). It is possible, however, that some of the microsporidian parasites of Daphnia that appear untransmissable in the laboratory, such as Flabelliforma magnivora, have a second host (Mangin et al. 1995) (Mangin et al. called this species Tuzetia sp.).

Although the uptake of the parasites by the second host species is likely to happen via deliberate or accidental ingestion of infected Daphnia, the uptake of the parasite by Daphnia is currently unknown for all helminth parasites. It is possible that Daphnia pick up, with their food, transmission stages that are released from the second host. Vertical Transmission

Vertical transmission describes the movement of a parasite from the mother (seldom the father) to the offspring (Figure 8.1D

Figure 8.1
). This transmission may occur directly, i.e., while the mother and the offspring have a physical connection (e.g., transovarial or transuterine), or indirectly, i.e., when mother and offspring remain close to each other after the birth. Mechanistically, the latter is a form of horizontal transmission because other susceptible hosts close to the mother could become infected as well.

Thus far, transmission from mother to offspring has been observed only in two parasites of Daphnia, both microsporidians infecting D. magna (Flabelliforma magnivora and O. bayeri). O. bayeri is also horizontally transmitted after the death of the host (Vizoso and Ebert 2004). For both parasites, it seems likely that transmission is transovarial.

As mentioned above, it is important to note that, mechanistically, horizontally transmitted parasites may appear to be vertically transmitted. If horizontally transmitted parasites can infect host offspring in the brood chamber or shortly after birth, they are functionally vertically transmitted. It is not clear how commonly this form of transmission occurs in Daphnia. The vertical transmission of parasites that are horizontally transmitted mechanistically is, however, common in other host–parasite systems (Ebert and Herre 1996).

A vertically transmitted parasite that has attracted a lot of attention for its high prevalence across arthropod taxa, including several crustaceans, is the intracellular bacterium Wolbachia. Wolbachia is transovarially transmitted and may be the most common parasite of arthropods worldwide. S. West and D. Ebert (unpublished observations) tested three clones from D. magna and three clones of D. pulex (each from a different population in southern UK) for the presence of either Wolbachia clade A or B (methods as in West et al. 1998). Although positive and negative controls confirmed that the PCR protocols worked properly, none of the Daphnia samples tested positive. The absence of Wolbachia was also reported by Fitzsimmons and Innes (2005), who tested D. pulex from the Great Lakes region of North America. Although the absence of evidence should not be taken as evidence for absence, I consider it highly unlikely that further investigations would reveal Wolbachia in Daphnia. Given our current knowledge of the mechanisms Wolbachia uses to maintaine itself in host populations (male killing, feminization, induced parthenogenesis, and cytoplasmic incompatibility), it seems unlikely that populations of cyclic parthenogens such as Daphnia could support Wolbachia.

8.1.2 Survival of Transmission Stages Outside the Host

An important factor for parasites with waterborne transmission is the lifetime of transmission stages outside of the host. The longer they can survive outside the host, the higher their likelihood of transmission. The longest surviving Daphnia parasites known thus far are the heavily protected endospores of the bacterium P. ramosa. In sediment cores of shallow ponds, spores more than 20 years of age have been found to be infectious (Decaestecker et al. 2004). Resting stages of Daphnia epibionts were even found to be viable after more than 60 years in the sediments (Decaestecker et al. 2004). Bacteria and microsporidian parasites can also be stored in freezers (-20ºC) for several years without apparent loss of infectivity. Spores of the microsporidians G. intestinalis and O. bayeri survive for at least 6-12 months in dry conditions at room temperature (H.J. Carius, unpublished observations; D. Ebert, unpublished observations). O. bayeri survives summer droughts in rock-pool populations in southern Finland (S. Lass and D. Ebert, manuscript in preparation).

It seems plausible that parasites in aquatic systems face fewer problems surviving outside their hosts than their terrestrial counterparts, because the most common causes of transmission-stage mortality for air- and soilborne parasites do not exist for waterborne transmission stages. Desiccation, for example, is irrelevant in the aquatic environment. Furthermore, water not only provides protection from UV radiation to a large degree, but its high heat capacity also buffers the effects of rapid temperature changes and prevents overheating. Because it is costly to produce protective structures for transmission stages (e.g., thick spore wall), aquatic parasites (as opposed to terrestrial parasites) may be able to shift the trade-off between quantity and quality of spores toward the production of more transmission stages.

8.1.3 Uptake of Transmission Stages from Pond Sediments

Planktonic populations typically undergo tremendous fluctuations in density, often over several orders of magnitude. Some plankton organisms might even temporarily disappear from their habitat and survive in the form of resting stages. Because these bottlenecks in host density pose a problem for horizontally transmitted parasites, Green (1974) suggested that plankton parasites should have persistent transmission stages to endure phases of low host density. He suggested that pond sediments form spore banks for these infective stages, similar to the way they harbor resting stages of many plankton organisms.

To test this hypothesis, mud samples were collected from different ponds that harbored parasitized populations of D. magna. Subsamples of these sediments were placed in beakers, and uninfected D. magna were added. When the hosts were later dissected, infections with different microparasites were found: among others, the bacterium P. ramosa, the yeast Metschnikowia bicuspidata , and the microsporidia G. intestinalis and O. bayeri (Ebert 1995; Decaestecker et al. 2002) (D. Ebert, unpublished observations). The results clearly confirm Green's (1974) hypothesis that pond sediments can serve as "parasite spore banks" and that parasites can survive periods of low host density in a "sit-and-wait" stage.

The uptake of spores from sediment is related in part to poor feeding conditions for the hosts and in part to their phototactic behavior. When feeding conditions deteriorate, some Cladocerans switch from filter feeding in the free water to browsing on bottom sediments. This behavior stirs up particles from the sediments, which are then ingested by filter feeding (Horton et al. 1979; Freyer 1991). What is important here is that spore uptake from pond sediments is primarily a density-independent form of transmission; it may only be linked to density indirectly, because high density may induce a switch in Daphnia's feeding behavior.

There is also evidence that the phototactic behavior of Daphnia clones also affects their likelihood of catching sediment-borne diseases (Decaestecker et al. 2002). D. magna genotypes with negative phototactic behavior are much more likely to come in contact with pond sediments and thus catch a disease than clones with a positive phototactic behavior (Figure 8.2

Figure 8.2
). Decaestecker et al. (2002) speculated that a trade-off between predator and parasite avoidance may be important in the evolution of habitatselection behavior. Negatively phototactic clones suffer less from visually hunting predators by residing in deeper and darker portions of the water column during the day, whereas positively phototactic clones, which are at a higher risk of predation, are less exposed to parasite spores in the sediment and consequently suffer less from parasitic infection. It was shown that increased infection rates near the sediments can be triggered by changing the daphniids' phototactic behavior, exposing them to chemical cues from fish (kairomone) and thus inducing a general behavioral shift toward lower positions in the water. This trade-off highlights a cost of predator-induced changes in the D. magna's habitat selection behavior and may help to explain genetic polymorphism for habitat selection behavior and disease resistance in natural Daphnia populations (Decaestecker et al. 2002).

8.1.4 Factors Influencing Parasite Transmission

After a parasite appears in a host population, it can only survive if each infection causes on average at least one secondary infection, that is, the basic reproductive rate of the parasite, R0, must be larger than 1 (Anderson and May 1986). There has been much discussion about what factors influence a parasite's transmission in a plankton population; I will summarize these below. Parasite Transmission Is Density Dependent

Density-dependent transmission, which is a central assumption of much epidemiological theory for horizontally transmitted parasites, has often been discussed with regard to plankton parasites (Canter and Lund 1951, 1953; Miracle 1977; Brambilla 1983; Ebert 1995; Bittner et al. 2002). Convincing data for density-dependent transmission and host population regulation under natural conditions were presented by Canter and Lund (1953), who observed strong fluctuations of the diatom Fragilariacrotonensis in an English lake. Whenever the density of these planktonic algae reached more than about 100 cells/ml, a fungal parasite (Rhizophidium fragilariae) spread rapidly, and host density dropped by two orders of magnitude. Density-dependent Transmission in Natural Populations

For Daphnia, no such example exists, although published data do not contradict density dependence. Brambilla (1983) observed that a microsporidian was generally present whenever the D. pulex density rose above 10 animals/liter, although the parasite suddenly disappeared one year in mid-summer despite high host densities. Vidtmann (1993) observed that the microsporidium Larssonia daphniae was present only when Daphnia density was high and yet was often absent during periods of high host density. Similar results were reported by Yan and Larsson (1988). Ruttner-Kolisko (1977) described a significant relationship between the density of a rotifer and prevalence, and even attributed a strong population decline in Conochilus unicornis to a microsporidian epidemic: "... Plistophora finally terminates its host species". Stirnadel (1994) was not able to detect density-dependent interactions between any of three Daphnia species and their numerous microparasites. The same was observed by Decaestecker (2002) in a very similar study on D. magna. Despite this paucity of published evidence to prove that density dependence plays a critical role for Daphnia epidemiology, many studies note that there is a minimum host density for parasite persistence, although the behavior at high densities has yet to be determined. For the time being, experimental approaches are more helpful than observations for investigating the role of density-dependent transmission. Experimental Evidence for Density-dependent Transmission

The microspordian gut parasiteG. intestinalis in D. magna has proved to be an ideal system to test for the density dependence of transmission. The life cycle of G. intestinalis is direct, and transmission to new hosts occurs only 3 days after infection (Ebert 1994a, 1995). The waterborne spores of this parasite are transmitted with the feces. Laboratory experiments showed that the transmission of G. intestinalis is strongly density dependent and that the infection intensity (parasite load per host) increased more rapidly when hosts were more crowded (Figure 8.3

Figure 8.3
). Very similar experiments were conducted with the protozoan parasite C. mesnili, which infects D. galeata (Bittner et al. 2002). The higher the density, the more likely it was that C. mesnili was transmitted (Figure 8.3). These experiments were carried out by placing one infected and one uninfected host together in vials containing different volumes of medium. In smaller volumes, the likelihood of transmission was higher. Interestingly, however, the decline in transmission rate with increasing volume was much smaller than expected, assuming a dilution effect. A possible explanation for this result is that two Daphnia within a vial do not distribute themselves randomly and independently from each other but rather cluster in certain parts of the vial, e.g., the bottom or places with more or less light. Therefore, on average, they are closer to each other than volume alone would suggest. Whether clustering plays a role in the transmission dynamics of natural populations is not known, but nonrandom distributions have frequently been observed in natural Daphnia populations (Green 1955; Weider 1984; Watt and Young 1992). Therefore, it appears likely to me that local clusters of Daphnia may play an important role in parasite dynamics in natural populations.

For parasites that are transmitted after the death of their host, density dependence has to be tested in a different way. Here it is the density of free transmission stages in the water that is important (Anderson and May 1986), and density-dependent transmission is indicated by infection–dose response curves. This has been shown for the yeast M. bicuspidata, the parasite P. ramosa (Ebert et al. 2000b; Regoes et al. 2003), and the microsporidium O. bayeri (Vizoso et al. 2005). In a very rigorous and detailed analysis, Regoes et al. (2003) showed that the likelihood of P. ramosa infecting D. magna largely followed the mass action assumption of classic epidemiology, which states that the likelihood of transmission is linearly related to the product of susceptible hosts and transmission stages (Figure 8.4

Figure 8.4
). Conclusions on Density-dependent Transmission

From these experiments, one can conclude that density dependence is indeed a real phenomenon in the spread of horizontally transmitted parasitic infections in Daphnia populations. However, merely confirming that density-dependent transmission exists does not reveal its significance for epidemiology in natural populations. To date, little support has been found to verify that density dependence is an important factor in Daphnia parasite epidemics. Other factors that seem to play an important role in transmission may cloud the significance of density dependence. Among these factors may be the temperature dependence of transmission (Ebert 1995), host stress, the role of a spore bank in the sediments (Ebert et al. 1997), and the genetic structure of the host population with respect to susceptibility (Little and Ebert 2000; Carius et al. 2001). Parasite Transmission Can Be Limited by Low Temperatures

Plankton epidemics are predominantly found during the warm summer months (Green 1974; Brambilla 1983; Yan and Larsson 1988; Vidtmann 1993). Ruttner-Kolisko (1977), working with a microsporidian parasite in a rotifer population, proposed that transmission is impaired at low temperatures. I tested this hypothesis with G. intestinalis in D.magna and found that transmission was indeed impaired below 12°C (Ebert 1995). This is consistent with the observation that G. intestinalis decreased in late autumn in D. magna populations in southern England (Stirnadel 1994). Poor transmissability at temperatures below 25°C was reported for P. ramosa, which parasitizes the Cladoceran Moina rectirostris (Sayre et al. 1979). (Note: It is questionable whether this Moina parasite was indeed P. ramosa.) In contrast, P. ramosa in D. magna can be transmitted between 10 and 25°C in the laboratory (Ebert et al. 1996; Mitchell et al. 2005). Thus, temperature criterion appears to be species and strain dependent.

Reports of natural Daphnia populations further indicate that certain parasites can be found under winter conditions (Stirnadel 1994; Bittner 2001). In Lake Constance, Daphnia parasites often occur predominately in fall and winter conditions (Bittner 2001), suggesting that temperature is certainly not universal in limiting parasite spread. The absence of parasites during summer in large lakes has been suggested to be related to intense predation during summer months (Duffy et al. 2005) and is unlikely to be a consequence of temperature effects on transmission. Host Stress Might Facilitate Parasite Spread

It has been claimed that stressed host populations are more susceptible to parasites and thus facilitate epidemics. This theory has been used to explain disease outbreaks in Cladocerans kept under poor laboratory conditions (Seymour et al. 1984; Stazi et al. 1994). Likewise, France and Graham (1985) observed higher rates of microsporidiosis among stressed crayfish in acidified lakes. For Daphnia, there is no support for the stress hypothesis but rather the opposite. Experimental transmission of G. intestinalis to individual D. magna appeared to be largely independent of the host's feeding conditions (and did not differ among age groups or sex) (Ebert 1995). Similar results were obtained for C. mesnili in D. galeata (Bittner et al. 2002). A direct test of the stress hypothesis was carried out in experimental populations of D. magna infected with G. intestinalis. When half of the experimental populations were stressed (reduced food level), parasite populations suffered more than the host populations (Pulkkinen and Ebert 2004) because mortality was disproportionately higher among the most heavily infected hosts (those that carried the most parasites). This result counters conventional wisdom about vertebrate populations, in which stress is thought to go hand-in-hand with disease outbreak. Experiments that tested the relationship between transmission stage production and host nutritional status further support the observation that Daphnia parasites do not fare well when their hosts are stressed. As in other invertebrate systems, parasites in poorly fed hosts produce fewer transmission stages than parasites in well-fed hosts (Ebert et al. 1998). Thus, although some observations have been interpreted to suggest that stress may lead to disease outbreaks, experimental results show clearly that this is not always the case, and this aspect of epidemiology needs further study. Resistance May Limit the Spread of Diseases

It has been long known that host genotypes differ in their susceptibility to parasites, as has been shown for several combinations of Daphnia populations and parasite species (Ebert et al. 1998; Little and Ebert 1999; Little and Ebert 2000; Carius et al. 2001; Decaestecker et al. 2003). Furthermore, there is good evidence for strong host–clone x parasite (isolate and species) interactions, both within and across populations (Ebert 1994b; Ebert et al. 1998; Carius et al. 2001; Decaestecker et al. 2003) (Figures 5.2

Figure 5.2
and 8.5). These studies also reported local parasite adaptation, noting that local parasites were more aggressive (more infective, more virulent, higher growth rate) than novel, introduced parasites (Ebert 1994b; Ebert et al. 1998).

The strongest evidence that infections within a population depend on host genotype was found by Little and Ebert (2000), who showed that in 3 of 4 tested populations, female D. magna infected with P. ramosa under natural conditions were genetically more susceptible to this parasite. To test this observation, they took field samples to the laboratory, divided them into infected and uninfected females, cured them with an antibiotic, and then cloned and reinfected the hosts with P. ramosa from the same population. The clonal offspring of the formerly infected females needed lower spore doses to become reinfected than the offspring of the formerly uninfected females (Figure 8.6

Figure 8.6
), thus indicating that genetic factors are clearly of crucial importance for the spread of diseases in natural Daphnia populations. Summary of Transmission Limiting Factors

The four factors discussed above may represent only a few of the many that influence the spread of diseases in Daphnia populations; however, I believe that they represent the most important ones. Other factors may be specific to certain diseases or may play minor roles. Although none of the factors discussed is likely to play a key role throughout the growing season, one or a few of them may become more influential at certain phases in epidemics. Furthermore, factors may interact to counterbalance or re-enforce each other. Genetic variation for resistance may, for example, be deflated by host stress. Thus, to understand the factors that influence the spread and dynamics of diseases in natural populations, it is necessary to conduct experiments that disentangle the complex interactions of host–parasite interactions. Experimental epidemiology is a particularly promising approach for addressing these questions (see Chapter 7 on Experimental Epidemiology and Evolution of Daphnia Parasites).

8.2 Epidemiology of Daphnia Microparasites

The results discussed thus far indicate that the invasion, spread, and persistence of parasites in Daphnia populations cannot be attributed to a single factor. Rather, the relevant factors may vary over time and act together or against each other. This interplay shapes parasite dynamics. Although we do not currently have conclusive explanations for the seasonal dynamics of Daphnia parasites, what we do know can serve as a starting point for a better understanding of plankton epidemics.

8.2.1 The Fishless Pond Model

Most of what we know about Daphnia parasites comes from small, predominantly fishless water bodies. The epidemiology of most microparasites of pond-dwelling Daphnia in the temperate zone follows a similar pattern (Green 1974; Brambilla 1983; Vidtmann 1993; Decaestecker 2002). Prevalence is usually low in winter and early spring. After host densities peak in spring, parasite prevalence increases; it fluctuates throughout the summer and decreases in autumn, with parasites often disappearing completely in winter. Green (1974) suggested that some microparasite epidemics (e.g., the bacterium Spirobacillus cienkowskii) start when a benthic feeding host acquires a parasite from the mud. Once the cycle starts, other Cladocerans that are partially benthic and partially free-water foragers become infected and transmit the parasite to those Cladocerans that live in the free water. The parasites disappear from the pond when the hosts go into diapause at the end of the season.

Earlier I proposed a single species version of this model (in 1995; Ebert et al. 1997). Following diapause, Daphnia hatch from their ephippia and recolonize a pond. Under good feeding conditions, the population increases rapidly during spring until food shortages lead to a switch from filter feeding in the free water to browsing on the bottom sediments. Browsing supplements the food because it stirs up food particles (Horton et al. 1979; Freyer 1991), which are then ingested by filter feeding. However, browsing also stirs up parasite transmission stages, which may infect the daphniid. Once the first hosts are infected, the disease may spread further. The epidemic ends either when environmental conditions deteriorate (e.g., low temperature) or when the host population becomes sparse or disappears altogether.

A key feature of this model is the uptake of spores from the pond sediments, which has very important consequences for the epidemiology of the system, as was shown in a mathematical version of this model (Ebert et al. 1997). First, uptake of spores from the sediments is independent from host density. The basic reproductive rate R0 becomes redundant as a means of predicting parasite persistence when there is a large, nondepleting spore bank in the sediment. Instead, the feeding behavior of Daphnia and the properties of the resource determine parasite invasions. This may explain why longitudinal studies of Daphnia pond populations have failed to find a relationship between parasitism and host density. Second, the spore banks allow the parasites to survive long periods of low host density.

Although this epidemiological model was developed for pond dwelling zooplankton, its findings about density-independent infection could also be relevant to a number of soil-borne diseases. Fleming and colleagues (1986) investigated the density-dependent transmission of a virus in different populations of the soil-dwelling pasture pest Wiscana sp. (Lepidoptera: Hepialidae). Evidence for density-dependent transmission was found only in young pastures but not in old pastures, perhaps because in older pastures transmission occurred mainly from a spore pool that had accumulated over several generations. In laboratory populations of a virus–insect system, Sait and colleagues (Sait et al. 1994) failed to detect density dependence and attributed this result to the rapid accumulation and long persistence of virus transmission stages within the cages. Contamination of the soil has been repeatedly cited as the source of various infections (Kellen and Hoffmann 1987; Young 1990; Woods et al. 1991; Dai et al. 1996). Thus, it appears that durable transmission stages and their accumulation in pond sediments or soil might be a widespread phenomenon in natural host–parasite systems and may obscure any pattern of density-dependent host-to-host transmission.

The Daphnia–parasite model for fishless ponds offers only the most basic pattern of parasite dynamics, leaving many details unexplained. It cannot, for example, explain the dynamics of prevalence in lakes where there are likely to be no spore banks and may also fail to predict epidemics in ponds with permanent (without diapause) Daphnia populations. It is further unable to explain why certain parasite species show short-lasting epidemics of a few weeks. Clearly, our understanding of parasite dynamics in natural Daphnia populations is still very limited.

8.2.2 Suggestion for a Lake Model

As discussed above, lakes with fish predation seem to have lower rates of parasitism than fishless ponds (see Chapter 4 on Daphnia Microparasites in Natural Populations). The following model may be a starting point for understanding zooplankton epidemics in lakes with fish. My ideas are partially based on the work of Kerstin Bittner at Lake Constance (Bittner et al. 1998, 2002; Bittner 2001).

Fish predation can be a severe mortality factor for Daphnia and will certainly influence the abundance of parasites. If fish predation is high, parasites may not be able to spread in Daphnia populations, because the average life expectancy of a Daphnia (and thus of an infection) is too short (see Chapter 4, Are There Fewer Parasites in Lakes with Fish?). K. Pulkkinen and D. Ebert (manuscript in preparation) have shown high parasite extinction rates in artificially predated, experimental D. galeata populations. Thus, during periods of high predation, parasites are expected to be absent or found in low prevalence. Because predation pressure often varies over time, parasites may spread during periods when adult host mortality is relatively low. This theory coincides with findings that the prevalence of Daphnia parasites in lake populations is high in fall when fish predation is low, whereas parasites are absent or only found in low prevalence during summer time, when predation is high (Bittner et al. 2002; Duffy et al. 2005).

In fishless ponds, parasites survive the absence of their hosts in the sediments. Because lakes with fish are less likely to have ecologically important spore banks in the sediments (Daphnia are much less likely to come into contact with the sediment in lakes), a different hypothesis is needed to explain how these parasites can survive unfavorable conditions. A possible explanation might be the large size of plankton populations, which may enable parasites to survive long periods of negative population growth (R0 < 1). With a huge host population size, for example, a parasite population might decline considerably for several generations, reaching very low prevalence. But low prevalence in large lakes is hardly an indication of extinction. For example, in a lake the size of Lake Constance (volume, 50 x 109 m3), if the host density falls to 0.1 Daphnia per m3 and 1 in 100,000 hosts is infected, there would be still about 50,000 infected hosts, certainly enough to maintain the parasite population, although at levels far too low to be detected with conventional sampling methods. This argument needs careful evaluation, taking absolute host and parasite population sizes into account as well as year-round growth conditions.

An alternative hypothesis is that parasites go extinct locally but occasionally recolonize the lake. However, if only one or a few immigrant parasites are introduced into a large host population, their spread to detectable levels takes considerable time unless R0 is high (>> 1). Nevertheless, this mechanism may still explain some of the observed cases of parasite disappearance and reappearance.

As mentioned above, parasites in large lakes with fish predation may evolve certain strategies to reduce their mortality. The most obvious of these are fast development (even if it has costs in terms of high virulence) and low visibility to visually hunting fish. A comparative study between lakes with and without fish predation would allow these two predictions to be tested.

In summary, parasites may be able to survive in large lakes with fish predation by exploiting hosts at times of low predation pressure and outlasting unfavorable times in a state of extended negative population growth.

8.3 Conclusions and Open Questions

At present, we have no satisfactory model for the epidemiology of Daphnia parasites, nor of any other zooplankton parasite. The two models presented above are general frameworks that treat all parasite species of a community alike and thus lack many important features. A more profitable approach may be to focus on certain parasite species and attempt to understand their epidemiology. Research has shown unambiguously that although certain mechanisms work under controlled conditions, e.g., density-dependent transmission, they may not necessarily explain the relevant dynamics in the field. In my judgment, a combined laboratory and field research approach is needed to elucidate the epidemiology of parasites. It is not clear whether general principles will explain the dynamics of certain host–parasite interactions or whether biological details of the specific interaction are required to understand the most of the observed variance. Some milestones on the way may be the answers to these open questions:

  1. Which factors limit the spread of microparasites in natural populations?
  2. What role do spore banks in sediments play in natural systems?
  3. Does density-dependent transmission explain parasite dynamics in natural populations?