Biodiversity in communities is changing globally, including the gain and 1 loss of host species in host-pathogen communities. The dilution effect argues for a mecha-2 nistic link between increased host diversity and decreased disease in a focal host. However, 3 we currently have a limited understanding of how the pathogen transmission mechanism 4 and between-host interactions influence whether increased host diversity leads to increased 5 (amplification) or decreased (dilution) infection prevalence. We use a two-host-one-pathogen 6 model to unify theory for pathogens with environmental transmission and density-dependent 7 and frequency-dependent direct transmission. We then identify general rules governing how 8 the pathogen transmission mechanism and characteristics of the introduced host (disease 9 competence and competitive ability) influence whether the introduction of a second host 10 species increases or decreases disease prevalence in a focal host. We discuss how our results 11 yield insight into how specific biological mechanisms shape host biodiversity-disease patterns.
13Biodiversity in communities is changing across the globe, with species introductions in some 14 and extirpation in others. Altered biodiversity can influence patterns of infectious diseases 15 because most pathogens can infect multiple host species and most communities are made 16 up of multiple host species (Cleaveland et al., 2001; Pedersen et al., 2005; Rigaud et al., 17 2010). These effects of biodiversity on disease levels can occur as a result of the ways each 18 host species interacts with the pathogen (e.g., within-species transmission) and interspecific 19 interactions between host species (e.g., resource competition and between-species transmis-20 sion).
21The dilution effect argues that increased host biodiversity decreases disease prevalence 22 (i.e., the proportion of infected hosts in a population) (Keesing et al., 2006). However, when 23 and whether increased host biodiversity reduces disease prevalence (dilution) or increases 24 disease prevalence (amplification) in a focal host population has been debated in the litera-25 ture (e.g., Lafferty and Wood 2013; Ostfeld and Keesing 2013; Wood and Lafferty 2013 and 26 reviewed in Rohr et al. 2019). Empirical evidence is mixed: a recent meta-analysis found 27 general empirical support for dilution (Civitello et al., 2015), but amplification also occurs 28 (Wood et al., 2014; Venesky et al., 2014; Searle et al., 2016). This implies that increased host 29 biodiversity likely has context-dependent effects (Salkeld et al., 2013), motivating calls for 30 theory that identifies specific biological mechanisms promoting amplification versus dilution 31 (Buhnerkempe et al., 2015; Halsey, 2019; Rohr et al., 2019).
32Current theory (Keesing et al., 2006(Keesing et al., , 2010 predicts that amplification versus dilution gen-33 erally depends on how host biodiversity affects host-pathogen encounter rates, susceptible 34 host densities, and transmission, mortality and...