Background
Six malaria parasite species (
Plasmodium spp.) have been shown to infect wild chimpanzees (
Pan troglodytes):
Plasmodium reichenowi,
Plasmodium gaboni and
Plasmodium billcollinsi (subgenus
Laverania), and the less common
Plasmodium vivax-like
, Plasmodium ovale-like and
Plasmodium malariae-like parasites [
1]. Another
Laverania species of chimpanzee malaria parasites,
Plasmodium billbrayi, is recognized by some, but not all, authors [
1,
2]. The distribution, diversity and phylogenetic relationships of
Plasmodium spp. have been the focus of several extensive studies conducted in the sub-Saharan African range of chimpanzees [
1‐
5]. Although these studies did not aim at producing directly comparable prevalence estimates, the resulting picture was one of extreme geographic variations, with values ranging from 0% in
Pan troglodytes schweinfurthii (Issa Valley, Tanzania) [
5] up to 48% in
Pan troglodytes troglodytes across multiple sites [
6]. This spatial heterogeneity may be explained by a complex combination of ecological factors influencing malaria parasite transmission, including the availability and abundance of vectors and/or the demographic, social, and behavioural characteristics of chimpanzee communities [
5,
7].
However, before the effect of such factors can be explored, clarifying temporal variation in
Plasmodium spp. prevalence in chimpanzees is necessary. For humans, as well as birds (the key wildlife model in malaria research), longitudinal studies have shown that malaria parasite prevalence often varies significantly between seasons and years [
8,
9]. Such unaccounted-for temporal variation may influence the apparent geographic distribution of chimpanzee malaria parasites (mainly deduced from cross-sectional sampling), complicating the identification of local drivers of transmission [
5‐
7,
10]. This study aimed at filling a fundamental gap in the understanding of the basic epidemiology of malaria parasites in chimpanzees by determining whether
Plasmodium spp. detection rate varied in a wild human-habituated community across four non-consecutive sampling periods.
Discussion
Marked, apparently stochastic variations in detection rates were observed at different temporal scales: monthly detection rates varied from 0 to 43%, whereas detection rates per sampling period varied from 4 to 27%. Strikingly, focusing on results of 2015 (N = 131) would have led to the erroneous conclusion that this chimpanzee community is malaria-free.
While shown for the first time in chimpanzees, such variations are not without precedent. Longitudinal studies of other host populations (e.g., humans, birds and lizards) showed both high seasonal and inter-annual variation, sometimes providing evidence for cyclical patterns with periods of 1–13 years [
8,
9,
24‐
30]. In particular, large complex geographical variation in short and long-term patterns of
Plasmodium falciparum has been observed worldwide [
8]. For example, rates of malaria cases in Venezuela exhibited multi-annual cycles of 2–6 years, with higher rates directly following an El Niño event and increased rainfall [
31]. Similarly, in Western Kenya, multi-annual malaria outbreak cycles of 2–4 years are associated with rainfall [
24]. Such climatic factors are probably the strongest drivers behind prevalence patterns, having direct effects on vector abundance and parasite development and exhibiting high temporal fluidity [
8,
9,
24,
28]. However, biotic factors most certainly also play a role, modifying the biology and behaviour of host, vector and/or parasite, and possibly operate at different temporal scales [
7,
8,
24]. For example, changing host demographics with the influx of naïve and highly susceptible individuals can alter the pool of infective individuals, driving variability through changing processes of transmission and immunity within a population [
24,
26,
31,
32]. The interplay between climatic and biotic factors therefore creates a spatiotemporally dynamic host-parasite system [
8,
24].
Within a year, detection rates followed seasonal patterns: peaking 3 months after the start of the wet seasons and reaching their lowest levels at the end of dry seasons or beginning of wet seasons. Such seasonal increases in prevalence have been observed for most malaria parasites, including the closest relative of the chimpanzee-adapted Laveranian parasites,
P. falciparum (the dominant cause of malaria-induced mortality in humans) [
8]. In Bangladesh, both
P. falciparum and
P. vivax exhibited seasonal patterns associated with temperature and monthly rainfall [
28]. This may be explained by higher abundance of vectors and/or higher prevalence of the parasites in vectors during rainy periods [
8,
9,
24,
25]. The combination of stochastic and seasonal variation in detection rate, and presumably prevalence [
8,
9,
27], has immediate practical consequences. These results caution against purely cross-sectional comparisons of detection rates across sites/habitat, which may over- or underestimate the abundance of parasites and oversimplify the dynamics of this host-parasite system.
Temporal variation also highlights the complexity of malaria parasite epidemiology in wild chimpanzee communities. This is partly due to the inherently complicated parasite life cycle and further complicated by fluctuating chimpanzee demography, community structure and behaviour [
7]. For example, since younger individuals, and in particular infants and juveniles (under 4 and 10 years old, respectively), are more likely to be infected (this study and [
5,
12]), changing group size and composition may alter infection patterns. As processes ultimately leading to the observed prevalence patterns are heterogeneous, future studies aimed at disentangling which of these factors determine the spatiotemporal distribution of malaria parasites in chimpanzees will require longitudinal sampling of a wide range of data on hosts, habitats, vectors, and parasite lineages [
7‐
9,
27].
Authors’ contributions
DFW participated in study design, laboratory work and data analysis, and drafted the manuscript. TL collected samples and participated in laboratory work. AS participated in laboratory work. RM participated in data analysis and writing of the manuscript. RMW participated in study design. SC-S participated in study design and writing of the manuscript. TD participated in study design. FHL designed and supervised the study and participated in writing of the manuscript. All authors read and approved the final manuscript.