Savola and Ebert BMC Ecol (2019) 19:14 https://doi.org/10.1186/s12898-019-0230-3 BMC Ecology RESEARCH ARTICLE Open Access Assessment of parasite virulence in a natural population of a planktonic crustacean Eevi Savola1,2 and Dieter Ebert1* Abstract Background: Understanding the impact of disease in natural populations requires an understanding of infection risk and the damage that parasites cause to their hosts ( virulence). However, because these disease traits are often studied and quantifed under controlled laboratory conditions= and with reference to healthy control hosts, we have little knowledge about how they play out in natural conditions. In the Daphnia–Pasteuria host–parasite system, feld assessments often show very low estimates of virulence, while controlled laboratory experiments indicate extremely high virulence. Results: To examine this discrepancy, we sampled Daphnia magna hosts from the feld during a parasite epidemic and recorded disease traits over a subsequent 3-week period in the laboratory. As predicted for chronic disease where infections in older (larger) hosts are also, on average, older, we found that larger D. magna females were infected more often, had fewer ofspring prior to the onset of castration and showed signs of infection sooner than smaller hosts. Also consistent with laboratory experiments, infected animals were found in both sexes and in all sizes of hosts. Infected females were castrated at capture or became castrated soon after. As most females in the feld carried no eggs in their brood pouch at the time of sampling, virulence estimates of infected females relative to uninfected females were low. However, with improved feeding conditions in the laboratory, only uninfected females resumed reproduction, resulting in very high relative virulence estimates. Conclusions: Overall, our study shows that the disease manifestation of P. ramosa, as expressed under natural condi- tions, is consistent with what we know from laboratory experiments. However, parasite induced fecundity reduction of infected, relative to uninfected hosts depended strongly on the environmental conditions. We argue that this efect is particularly strong for castrating parasites, because infected hosts have low fecundity under all conditions. Keywords: Host-parasite interactions, Virulence, Castration, Daphnia, Pasteuria ramosa Background population density depends in a non-linear way on how Along with other environmental factors, parasitic infec- it infuences individual host fecundity and survival: par- tions—and especially virulence, the ftness costs from asites that mainly reduce host fecundity suppress host such infections—play a key role in limiting and con- density to a greater degree than parasites that induce trolling host populations [1–5]. Te study of infectious host mortality [11, 12]. Terefore, understanding the disease virulence in humans [1, 6] and diverse animal expression of virulence in individual hosts is vital for our populations [7–10], has revealed a vast range of conse- understanding of the processes that govern host popu- quences for individual hosts and host populations. For lation and disease dynamics in the natural world [13]. example, the degree to which a parasite regulates host However, assessing virulence under controlled laboratory conditions may not accurately refect how the host sufers from infections under natural conditions, which are not *Correspondence: [email protected] controlled and are often highly variable. By better under- 1 Department of Environmental Sciences, Zoology, Basel University, standing how disease symptoms are expressed under Vesalgasse 1, 4051 Basel, Switzerland Full list of author information is available at the end of the article natural conditions, we can gain a better quantitative © The Author(s) 2019. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Savola and Ebert BMC Ecol (2019) 19:14 Page 2 of 11 understanding of the causes and consequences of wildlife Daphnia pulex, while even weaker and non-signifcant epidemics, which could, for example help design biologi- efects were reported for D. magna and D. longispina cal pest control programs [14, 15]. [27]. A study of a Belgian D. magna population, found For several reasons the efect of parasites on individ- a signifcant, but weak, fecundity reduction in asso- ual hosts is difcult to assess under natural conditions. ciation with P. ramosa infections [28]. In a D. magna First, parasite-induced host mortality and host fecundity population in northern England, Duncan and Little require prolonged observation of individual hosts, and [29] observed similar weak efects, but did not quan- infected and uninfected hosts need to be compared to tify them. Likewise, a study of D. longispina reported observe the efect of infection. Tis comparative or “rela- a weak and non-signifcant reduction of fecundity tive” aspect of defning virulence is important, although upon infection with P. ramosa [30]. How can this dis- frequently overlooked [16, 17]. Without it, we cannot crepancy—between fecundity reductions of over 80% distinguish parasitic efects on host wellbeing from other in laboratory experiments and small fecundity reduc- efects, such as environmental conditions, which may tions in the feld—be explained? Additionally, how do afect the wellbeing of infected and uninfected hosts other disease symptoms compare between the feld and alike. For example, resource limitation may increase mor- laboratory? Here we attempt to answer these questions. tality and decline fecundity in infected and uninfected Based on our observations of experimental infections hosts equally. Second, because infections may not be eas- and disease progression in controlled laboratory condi- ily detected (for example, in recently infected animals), tions, we suggest and test the following predictions: the infection status of individual hosts may not be accu- rately assessed. Tird, infections may increase mortality, • First, in advanced infection stages, females are cas- thus removing from the population those animals that trated by the parasite, i.e. they carry no sexual or show the strongest symptoms and consequently reduc- asexual eggs [19]. We expect this to hold true even ing estimates of virulence in the surviving population. under improved feeding conditions, where most Fourth, as infections by multiple parasites are common infected females usually do not reverse castration, under natural conditions, but not always easily detected, although some may produce a few small clutches the link between disease symptoms and a specifc para- after being initially castrated [26, 31]. site is not always clear [10, 18]. Finally, infected individu- • Second, in the laboratory, P. ramosa symptoms (col- als may not be “typical” for the study population. Hosts our change, castration, gigantism, loss of transpar- in poor shape, for example, may become infected more ency) appear about 10 to 20 days after infection [19]. easily and appear therefore to be more strongly afected Time to maturity takes about 6 to 12 days (longer by the disease than other hosts. Taken together, these under poor feeding conditions). Tus, we expect issues make working with natural populations challeng- that infected juveniles will not show disease symp- ing. Here, we undertake a study comparing estimates of toms on the day of their capture. In addition, since disease parameters caused by a bacterial parasite in a nat- their infections will be recent, we expect their disease ural zooplankton population with results from laboratory symptoms to show up towards the latter end of the experiments under controlled conditions. time spectrum (i.e. closer to 20 days than 10). Juve- Pasteuria ramosa is a common, obligate, bacterial niles caught in the feld and kept under good feeding endoparasite of the planktonic crustacean Daphnia conditions are, thus, more likely to produce ofspring (reviewed in [19]). In the laboratory, its waterborne before castration than older, larger animals, whose spores can infect animals of any age, class, and sex infections are probably older as well. that are exposed to them, although susceptibility does • Tird, in the laboratory P. ramosa causes chronic dis- vary [19, 20]. Infection changes the host’s body colour, ease. Tus, we expect infection prevalence in natu- and causes loss of transparency, castration of females, ral populations to increase with host age, which in enhanced body growth (gigantism) and increased mor- Daphnia can be recognized by their larger size, as tality [19]. Under controlled laboratory conditions, they grow indeterminately. Tis prediction assumes infected juveniles lose about 90 to 100% of their resid- that the parasite neither causes increased mortality ual life-time reproductive value (relative to uninfected in older animals nor that it infuences host body size. females), whereas adults lose about 80 to 90% [9, 21– Both these assumptions are violated by this system. 26]. Tese observations, however, contrast markedly