Peak shift and epidemiology in a seasonal host-nematode system

Abstract

Insight into the dynamics of parasite-host relationships of higher vertebrates requires an understanding of two important features: the nature of transmission and the development of acquired immunity in the host. A dominant hypothesis proposes that acquired immunity develops with the cumulative exposure to infection, and consequently predicts a negative relationship between peak intensity of infection and host age at this peak. Although previous studies have found evidence to support this hypothesis through between-population comparisons, these results are confounded by spatial effects. In this study, we examined the dynamics of infection of the nematode Trichostrongylus retortaeformis within a natural population of rabbits sampled monthly for 26 years. The rabbit age structure was reconstructed using body mass as a proxy for age, and the host age-parasite intensity relationship was examined for each rabbit cohort born from February to August. The age-intensity curves exhibited a typical concave shape, and a significant negative relationship was found between peak intensity of infection and host age at this peak. Adult females showed a distinct periparturient rise in T. retortaeformis infection, with higher intensities in breeding adult females than adult males and non-breeding females. These findings are consistent with the hypothesis of an acquired immune response of the host to a parasite infection, supporting the principle that acquired immunity can be modelled using the cumulative exposure to infection. These findings also show that seasonality can be an important driver of host-parasite interactions.

Figure 1

Host age–parasite intensity profiles for each cohort of individuals born from February (M 2) to August (M 8). The smooth profiles represent a cubic spline curve fitted to the relationship between geometric mean of parasite intensity and host cohorts, averaged over 26 years of data (symbols).

Figure 2

Host age–parasite intensity profiles by sex for rabbits sampled every month. Each profile represents a cubic spline curve fitted to the relationship between geometric mean of parasite intensity and host age classes, averaged over 26 years of data (symbols). The standard error bars of the mean intensity are reported. For simplicity, snap shots for February, May, June and August are presented.

Figure 3

(a) Parasite density profiles by host cohort in each month of birth, from February (M 2) to August (M 8), estimated using the Weibull probability density distribution. We have rescaled the curves, multiplying each one by the parasite intensity in each age class of every month. (b) Relationship between cumulative (±s.e.) parasite intensity and estimated mean host age at peak from February (M 2) to August (M 8) months of birth.