2.3. Laboratory transmission experiments

We used a series of targeted, laboratory experiments that manipulated different variables independently with the aim of testing the alternative transmission functions (Table 1) across a range of parameter space. While other studies have investigated transmission with respect to system specific mechanisms, such as age of hosts or feeding behaviour (D'Amico et al., 1996; Goulson et al., 1995), we sought to manipulate factors relevant across host–pathogen systems and modes of transmission. We also took care to avoid confounding variables such as densities and numbers of hosts, noting that some of the most common experimental designs—i.e. varying total parasite number while keeping host and total volume constant—often do not reveal any distinction between transmission functions (see also Antonovics & Alexander, 1992; Rachowicz & Briggs, 2007).

We used P. regilla tadpoles raised in the laboratory from eggs and R. ondatrae from naturally infected snails to conduct experiments that independently varied (1) parasite number (4, 13, 30, 63 and 144 parasites), (2) host density (0.48, 0.95 and 1.9 tadpoles/L), (3) duration of exposure (15, 30, 60, 120 and 240 min) and (4) parasite density (5, 15, 30, 45 and 60 parasites/L), with each treatment replicated 10 times. For all experiments, the baseline conditions were 30 parasites/L, 30 min of exposure time and one tadpole/L. Total parasite number was manipulated independently of density by increasing volume while parasite density was held constant (30 parasites/L, range of volumes 0.12–4.8 L). Finally, our experimental design minimized heterogeneity in hosts and parasites by using tadpoles that were size‐ and stage‐matched and cercariae collected within a narrow age range (Appendices S1 and S2), thereby limiting the mechanisms that could drive nonlinear transmission dynamics.