Infection burdens and virulence under heat stress: ecological and evolutionary considerations

Abstract

As a result of global change, hosts and parasites (including pathogens) are experiencing shifts in their thermal environment. Despite the importance of heat stress tolerance for host population persistence, infection by parasites can impair a host's ability to cope with heat. Host-parasite eco-evolutionary dynamics will be affected if infection reduces host performance during heating. Theory predicts that within-host parasite burden (replication rate or number of infecting parasites per host), a key component of parasite fitness, should correlate positively with virulence-the harm caused to hosts during infection. Surprisingly, however, the relationship between within-host parasite burden and virulence during heating is often weak. Here, we describe the current evidence for the link between within-host parasite burden and host heat stress tolerance. We consider the biology of host-parasite systems that may explain the weak or absent link between these two important host and parasite traits during hot conditions. The processes that mediate the relationship between parasite burden and host fitness will be fundamental in ecological and evolutionary responses of host and parasites in a warming world. This article is part of the theme issue 'Infectious disease ecology and evolution in a changing world'.

Keywords: disease burden; pathogen evolution; thermal tolerance; trade-off; transmission; virulence.

Figure 1.

Virulence–transmission trade-off under ambient and stressful conditions. (a) Theoretical predictions from the trade-off hypothesis for the relationships between virulence and within-host parasite burden, host lifespan and parasite fitness—R₀ (box 1). (b) Under extreme heat, the relationship between virulence and within-host parasite burden disappears, which has the potential to shift the relationship between virulence and parasite transmission. Without a positive relationship between parasite burden/replication and virulence, one consequence could be for lower virulence to become optimal (particularly when demographic change is considered [54]). However, if virulence is not associated with within-host parasite burden (or at least fitness), parasite evolution could instead be constrained or dampened due to the random removal of genetic variation within the population. Note that these hypothetical predictions do not account for important complexities of host–parasite systems, such as transmission mode, which mediate the relationships between host and parasite fitness traits ([40,42,44,45]; see main text and box 1 for a discussion of the generality of the trade-off hypothesis).

Figure 2.

The impact of within-host Pasteuria ramosa spore burdens on the critical thermal limit (CT_max) of Daphnia magna from Hector et al. [19]. CT_max is the temperature causing mortality during a 0.06°C/min heating ramp from ambient temperature and was measured on two host genotypes (A or B) infected with one of two parasite genotypes (C1 or C20). The strength of the relationship depended on the specific host–parasite genotype combination—a clear negative relationship is only apparent for one genotype pair.

Figure 3.

(a) The mean number of parasite larvae (i.e. nauplii and cyprids) released by brood from Eurypanopeus depressus infected by Loxothylacus panopaei with either one (solid line, estimate +/– s.e. of minimum temperature = 6.01 + /– 2.3, and maximum temperature = 38.03 + /– 2.34) or two externa (dashed line, estimate + /–s.e. of minimum temperature = 5.14 + /– 1.23 and maximum = 27.47 + /– 0.07) within a two-week period, after having been acclimated to temperature treatments over 11 days and held at experimental temperature for a week (mean + /– s.e.). Two-week period for comparison was selected because at 20°C, L. panopaei will release approximately 1 brood a week. (b) The expected lifespan in weeks of E. depressus infected by L. panopaei with either one (green, estimate + /– s.e. of minimum temperature 4.99, and maximum temperature 32.10 [27]) or two externa (dashed line, estimate + /–s.e. of minimum temperature = 4.99 + /– 0.02 and maximum temperature = 35.05 + /– 2.95). Expected lifespan was calculated from a fit survival object, methods available in [27]. Due to logistical constraints, replication in the two externa groups was not equal across temperatures (replication at 5°C = 3, 10°C = 5, 15°C = 4, 20°C = 3, 25°C = 3, 30°C = 4, 35°C = 3). Additional information about experimental design and methods in [27]. Animals with co-infections were kept in the same conditions as those with single infections. (c) An E. depressus with two attached L. panopaei externa (arrows indicate the two different externa). (Online version in colour.)

Figure 4.

Ecological relationship between parasite transmission strategy and host–parasite population densities under extreme heating. (a) Direct transmission between living hosts or via vectors. (b) Transmission depends on host death or via environmental stages. Blue and red lines represent the dynamics of a hypothetical host and parasite population, respectively. Across time, we may see fluctuating host–parasite population dynamics depending on the system [153]—although our general point does not depend on the exact nature of these dynamics. After an extreme heat event (sun and arrow) population dynamics are interrupted, with subsequent dynamics depending on the host–parasite system: (a) excessive host mortality causes both host and parasite populations to shrink, with parasite reestablishment lagging behind any host population recovery; (b) excessive host mortality results in an overabundance of environmental parasite transmission stages, which could suppress host population growth or cause local extinction.