A meta-analysis of how parasites affect host consumption rates

Parasites are known to mediate trophic interactions and can, for example, modify how consumers acquire resources. These modifications of host feeding behaviour can be imposed through three interconnected mechanisms affecting: (1) host food acquisition, (2) host food digestion or (3) host energy budgets. As a result, infected hosts may consume more, less or the same amount of food compared to their uninfected conspecifics. It is commonly assumed that infected hosts have lower consumption rates than uninfected hosts, but a comprehensive quantitative synthesis investigating the effects of parasites on host consumption rate has been lacking thus far. To fill this knowledge gap, we systematically searched for experimental studies that evaluated changes in consumption rate of infected vs uninfected hosts. In total, we extracted 158 effect sizes from 68 studies. We then performed meta-analyses of mean differences in host consumption rates and their variation. The analyses were carried out for all taxonomic groups as well as separately for vertebrate and invertebrate hosts. The main-effects meta-analyses confirmed a generally negative effect of parasites on host consumption rates; infected hosts consumed on average 25% less food than their uninfected conspecifics. In addition, there was a significant increase in the variability in host consumption rate, on average by 25%, indicating that parasites can have variable effects on the foraging behaviour of their hosts. The meta-regression models revealed that several moderator variables related to host and parasite characteristics influence host consumption rate. Experimental infection had a stronger influence on variance effects than natural infection. Parasitic infections reduced consumption rate of vertebrate hosts by 28% and thus more strongly than those of invertebrates, which were reduced by 22%. We conclude with recommendations to facilitate future ecological research syntheses on host-parasite interactions and beyond.
Methods
We systematically searched for experimental studies that evaluated changes in consumption rate of infected versus uninfected hosts. The literature search was performed on the Web of Science Core Collection (WoS) following the guidelines of the PRISMA statement (Moher et al., (2010) Int J Surg 8, 336–341). The detailed description of the search strategy can be found in the associated publication.

Articles were included in the meta-analysis based on pre-defined eligibility criteria. The final dataset consists of 68 studies published between 1973 and 2019, from which we extracted 158 effect size estimates. From each study, we also recorded year of publication, parasite taxonomy (phylum and species), type of parasite (micro- or macroparasite), parasite life cycle (complex or direct), host taxonomy (group [vertebrate or invertebrate], class and species), details about the host (i.e., sex, age), infection type (natural or experimental), infected organ (body cavity, circulatory, digestive, muscular, sensory, whole body), prey mobility (immobile, mobile, unknown), study environment (freshwater, marine or terrestrial) and the geographic place where the experiment was performed (continent and country).

We performed meta-analyses of mean differences in host consumption rates and their variance. Two types of models were applied: (1) main-effects models to test for the general effect of parasites on host consumption rates and (2) multilevel meta-regression models to assess the contribution of the moderator variables to the magnitude and direction of the effect sizes.