Fig. 4.1 Increase in the number of published records (citation counts) from 5 bibliographic databases addressing host-parasite ecology or evolution. Databases include BIOSIS Previews (covering a wide range of life sciences journals; OVID Technologies, Inc.), PubMed (focusing on the biomedical literature; National Center for Biotechnology Information), Zoological Record Plus (ZooRecord, focusing on publications in animal biology; CSA Illumina) and PrimateLit (covering references in primatology; Wisconsin Primate Research Center and Washington National Primate Research Center). Search terms used in each case were as follows: (infectious disease or parasite or pathogen) and (dynamics or ecology or evolution), together with the years of publication as shown on the X-axis. For PrimateLit and ZooRecord, only journal publications were extracted (i.e. not websites, conferences, books, or other sources).
and from increased interest in the joint evolution of hosts and pathogens (Hamilton and Zuk 1982; Lively 1992; Clayton and Moore 1997; Lively 1999). For example, long-term studies on the dynamics of cecal nematodes (Trichostrongylus tenuis) in red grouse (Lagopus lagopus scoticus) showed that parasites can drive host population cycles, in part due to their sublethal impacts on host fecundity and the persistence of parasite infectious stages during periods of low host abundance (Hudson et al. 1985, 1992, 1998a; Dobson and Hudson 1992). Similarly, long-term studies of genetic interactions between trematode parasites (Microphallus sp.) and freshwater snails (Potamopyrgus sp.) demonstrated that parasites can adapt to infect common host genotypes, and that frequency-dependent selection can generate evolutionary cycles in host and parasite allelic frequencies (Dybdahl and Lively 1998; Lively 1999).
Surprisingly, virtually no detailed epidemiological studies have focused on parasite dynamics in wild primate hosts. Even for primate infections of great concern to human health, such as lentiviruses, spumaviruses, and Schistosoma parasites, a great number of studies report data from "wild" primate hosts sampled long after they were captured in the wild and far away from the capture sites, thus calling into question whether patterns reflect those found in the wild and reducing the usefulness of prevalence data for epidemiological analysis. The shortage of studies on parasite dynamics in wild primates, including those that join modeling approaches with field data, is perhaps the greatest challenge for developing a better understanding of infectious diseases in these species.
4.1.2 Basic terminology and measures of infection
Several basic epidemiological parameters are important for describing parasitism in natural populations. As mentioned briefly in Chapter 1, prevalence is a primary measure of pathogen occurrence that refers to the proportion of hosts infected with a micro- or macroparasite. Prevalence can be estimated using any methods that discriminate infected from non-infected hosts, such as (1) outward signs of disease linked to changes in physical appearance or behavior of the host, (2) direct evidence of parasites in the blood, feces, or other host tissues, including use of PCR-based methods for detecting parasite DNA or RNA, and (3) serological methods that use antigen-antibody reactions to infer the past history of exposure to a particular agent (i.e. seroprevalence). These different approaches for sampling individual animals capture different phases of infection in the host (Table 4.1).
Some warnings are in order when interpreting prevalence estimated from field sampling protocols. First, inferences of disease status will depend on the method used and the hosts' stage of infection, since hosts could be sampled before developing disease or antibodies, and outward signs of infection might persist after hosts are no longer infectious (Table 4.1). Thus, multiple methods for assessing host infection status are often needed, especially when modeling approaches require information
Table 4.1 Stages of infection or disease status assignments based on different methods for examining hosts for signs of infection, including actual presence of the pathogen (using microscopy, culture, or PCR-based methods), physical or outward signs of disease, and the presence of host antibodies using serological techniques
Stage of infection Test method
Presence of Outward signs Host antibodies pathogen of disease to infection
Exposed and infectious Yes
Diseased and infectious Yes
Diseased, infectious, Yes and host immune response
Recovering but still No diseased
Recovered and immune No
Asymptomatic carrier Yes state
Unrelated cause of No disease
Different combinations of presence/absence information can be used to infer the infectiousness of the host and potential impacts of disease at the individual level. Note that each test method alone could correspond to multiple phases of infection in the absence of other information.
on the infection status of different individuals (see below). A second potential caveat for interpreting prevalence data is that accurate estimates of prevalence might require correcting for uneven sampling of healthy and diseased animals by researchers (Jennelle et al., provisionally accepted). This issue arises because behavioral or physical changes in diseased animals could make them more or less apparent to observers, leading to systematic biases in estimated prevalence (Faustino et al. 2004). Third, further complications occur when the infection status of animals is quantified with uncertainty, as many methods for detecting parasites are accompanied by imperfect specificity and sensitivity. In this case, specificity refers to the ability of a test to discriminate between true versus false positives, and sensitivity refers to the power of a test to detect true versus false negatives (Burr and Snodgrass 2004). Detecting infection status at different stages following host exposure can influence the sensitivity and specificity of the test.
For microparasites, ecologists frequently assume that it is sufficient to know whether or not a host is harboring a given parasite, rather than counting the actual number of viral or bacterial particles per host. In some cases, however, high levels of pathogens in the blood or other host tissues (e.g. viremia, bacteremia, or parasitemia) can indicate particularly severe infections, with impacts on pathogen transmission and the likelihood of host survival. A second useful measure of infection status commonly employed for many macroparasites is the intensity of infection, or numbers of parasites per infected host (see Chapter 1). Not surprisingly, intensity is more likely to be used for parasites that can be readily counted, such as ticks per animal or numbers of worms inhabiting a section of the gastrointestinal tract. Estimates of intensity can be measured by quantifying parasite life stages in feces, blood smears, muscles or other organs (Bush et al. 1997). Another measure related to intensity is parasite abundance or the average number of parasites across all hosts. Because this measure also includes non-infected hosts in the calculation, it is a composite measure of both intensity and prevalence (i.e. abundance = intensity * prevalence).
Measures of intensity and abundance are closely tied to the fundamental ecological question: are individual organisms clumped, randomly dispersed, or evenly dispersed within host populations? Parasite aggregation is critically important to understanding the population-level impacts of parasites on their hosts (Anderson and May 1978; May and Anderson 1978). A large number of empirical studies have shown that macroparasites are almost always aggregated or clumped, with most parasites in a population found in a small number of hosts, and most hosts harboring light infections (Fig. 4.2). A key measure of parasite dispersion (or aggregation) is the ratio of the variance to the mean in parasite numbers (Shaw and Dobson 1995; Shaw et al. 1998). This ratio should be close to one when parasites are randomly distributed among hosts and much greater than one when the majority of parasites are clumped in just a few hosts. Another measure of parasite aggregation uses the negative binominal distribution (see Fig. 4.2).
Some studies of wild primate populations have provided statistics on parasite aggregation (Fig. 4.2(b)). Several processes could generate aggregated populations of parasites, including parasite recruitment into already-infected hosts through continual re-exposure. Differences in exposure could also result from behavioral variation in food or habitat preferences (Müller-Graf et al. 1996, 1997), or from patterns of dispersion linked to variation in primate mating and social systems (Stuart and Strier 1995). Heterogeneity in parasite burdens might further reflect differential susceptibility arising from genetic variation in host resistance (Patterson et al. 1998; Coltman et al. 2001), or from variation in host social status, nutrition, or stress. Traits of the parasites themselves, such as parasite mobility (Shaw and Dobson 1995), could also affect levels of aggregation in populations.
Was this article helpful?