Ecology, 90(1), 2009, pp. 132–144Ó 2009 by the Ecological Society of AmericaLong-term disease dynamics in lakes: causes and consequencesof chytrid infections in Daphnia populationsPIETER T. J. JOHNSON,1,2,5ANTHONY R. IVES,3RICHARD C. LATHROP,4AND STEPHEN R. CARPENTER21Ecology and Evolutionary Biology, Ramaley N122, Campus Box 334, University of Colorado, Boulder, Colorado 80309-0334 USA2Center for Limnology, University of Wisconsin, 680 North Park Street, Madison, Wisconsin 53706-1492 USA3Department of Zoology, University of Wisconsin, Madison, Wisconsin 53706 USA4Wisconsin Department of Natural Resources and Center for Limnology, University of Wisconsin, 680 North Park Street,Madison, Wisconsin 53706-1492 USAAbstract. Understanding the drive rs and consequences of disease epidemics is animportant frontier in ecology. However, long-term data on hosts, their parasites, and thecorresponding environmental conditions necessary to explore these interactions are oftenunavailable. We examined the dynamics of Daphnia pulicaria, a keystone zooplankter in lakeecosystems, to explore the long-term causes and con sequences of infection by achytridiomycete parasitoid (Polycaryum laeve). After quantifying host–pathogen dynamicsfrom vouchered samples collected over 15 years, we used autoregressive models to evaluate (1)hypothesized drivers of infection, including host density, water temperature, dissolved oxygen,host-food availability, and lake mixing; and (2) the effects of epidemics on host populations.Infection was present in most years but varied widely in prevalence, from ,1% to 34%, withseasonal peaks in early spring and late fall. Within years, lake stratification strongly inhibitedP. laeve transmission, such that epidemics occurred primarily during periods of water mixing.Development of the thermocline likely redu ced transmission by spatially separatingsusceptible hosts from infectious zoospores. Among years, ice duration and cumulativesnowfall correlated negatively with infection prevalence, likely because of reductions in springphytoplankton and D. pulicaria density in years with extended winters. Epidemics alsoinfluenced dynamics of the host population. Infected D. pulicaria rarely (,1%) contained eggs,and P. laeve prevalence was positively correlated with sexual reproduction in D. pulicaria.Analyses of D. pulicaria density-dependent population dynamics predicted that, in the absenceof P. laeve infection, host abundance would be 11–50% higher than what was observed. Byunderscoring the importance of complex physical processes in controlling host–parasiteinteractions and of epidemic disease in influencing host populations, our results highlight thevalue of long-term data for understanding wildlife disease dynamics.Key words: chytridiomycete; climate change; Daphnia pulicaria; epizootic; limnology; parasitoid;pathogen; Polycaryum laeve; time series; zooplankton.INTRODUCTIONParasites and pathogens are ubiquitous members ofall ecosystems. Because of their small size and crypticnature, however, pathogens have historically beensidelined in ecological studies relative to more obviousinteractions such as predation and competition. Short-term studies have established that parasites can havesignificant impacts on host behavior, fecundity, popula-tion dynamics, and even community composition.Parasites can also play a powerful, albeit cryptic, rolein ecological food webs (Lafferty et al. 2006, Wood et al.2007). Nevertheless, the significance of these effects overlonger time scales, and identification of the environmen-tal factors that may initiate epidemics in wildliferepresent important frontiers in ecology (NationalResearch Council 2001). Unfortunately, the data neces-sary to explore these questions are often lacking;pathogens are difficult to observe and quantify whilethe corresponding environmental information needed toidentify the drivers of infection is frequently unavailable.Interactions between Daphnia and their parasites mayoffer valuable insights into the long-term causes andconsequences of wildlife disease epidemics (see Plate 1).Daphnia, which are ubiquitous in pond and lakeecosystems, serve as hosts for numerous endo- andectoparasites, many of which produce conspicuousinfections easily recognized through the transparentcarapace of their hosts (e.g., Green 1974, Ebert 2005).Although zooplankton population dynamics are tradi-tionally thought to be controlled by food availabilityand predation (e.g., Lampert et al. 1986, Carpenter et al.1987), increasing evidence suggests that parasite infec-tion can have important effects on Daphnia, includingdecreased reproduction, increased mortality, enhancedManuscript received 18 December 2007; revised 4 April 2008;accepted 24 April 2008; final version received 20 May 2008.Corresponding Editor: K. D. Lafferty.5E-mail: [email protected] to predators, and altered migratory andgrazing behavior (Brambilla 1983, Yan and Larsson1988, Bittner et al. 2002, Decaestecker et al. 2005, Duffyet al. 2005, Ebert 2005). Because Daphnia are alsoconsidered a keystone component of lake food webs,epidemic diseases of Daphnia have the potential to affectcommunity- and ecosystem-level properties of lakes(Duffy 2007).Despite continued advances in the understanding ofDaphnia–parasite interactions in the laboratory, thefactors governing disease epidemics in natural hostpopulations and their long-term consequences remainpoorly understood. Most field studies have been short-term (1–3 years), focused on simplified systems withoutvertebrate predators (e.g., ponds and rock pools), andaccompanied by limited amounts of environmental data(e.g., Green 1974, Brambilla 1983, Bengtsson and Ebert1998, Ebert et al. 2000, 2001, Decaestecker et al. 2005).Few studies have examined the significance of Daphniapathogens in lake ecosystems with more complex foodwebs and physical processes (e.g., Duffy et al. 2005,Ca´ceres et al. 2006). What information is availableindicates that infection levels in Daphnia populations areoften highly variable with pronounced seasonality (e.g.,Lass and Ebert 2006). Hypotheses advanced to explainthese patterns include fluctuations in host density,temperature limitation, food availability, selective preda-tion of infected animals, size-specific susceptibility toinfection, and host genetic composition (Ruttner-Kolisko1977, Yan and Larsson 1988, Little and Ebert 2000, Duffyet al. 2005). Evidence for these