Fig. 2: Measurements of parameters used to calculate parasite density in the environment and to estimate the degree to which potential parasite exposures are elevated near water relative to matrix sites.

a–f Key transmission steps noted in the central figure. a Fecal egg counts measured in this study (green triangles) compared to those reported in studies across the African continent (mean of individual studies in blue; studies weighted by sample size in red). b Average dung density contributed by each major species in both experimental (OPC) and observational systems (Mpala). c Comparisons of estimated parasite eggs contributed by each species at permanent water sources (either dams or pans, shown in blue) and matrix sites (calculated from both experimental and observational data sets, shown in orange) show considerable consistency across species. Estimates from the experimental system compare filled pans only to matrix sites. Note that both graphs are visualized on the log₁₀ scale. Bars and centers represent means ± SE calculated from parasite density at the site level, averaged across all periods (observational system: n = 20 per species and treatment, experimental system: n = 5 per species and treatment). d Four theoretical scenarios of parasite mortality at water sources relative to matrix sites, ranging from an assumption of reduced mortality (low; in blue) due to potential increases in ground moisture, to an assumption of greatly increased mortality (very high; in red) due to decreases in vegetation cover. e Herbivore grazing activity at water sources (in blue) and matrix sites (in orange) for both experimental and observational systems. Note that both graphs are visualized on the log₁₀ scale. Bars and centers represent means ± SE calculated from average daily grazing activity at the site level, averaged across all periods (observational system: n = 12 per species and treatment, experimental system: n = 5 per species and treatment). f Relative number of potential parasite exposures at water relative to matrix sites for each species and parasite mortality scenario from d. Significant (p < 0.05) and marginally significant (p < 0.1) differences for two-sided t-tests with Holm adjustment for multiple comparisons are bordered by solid and dotted lines respectively. Source data are provided as a Source Data file. Centre panel and icon artwork by G. Titcomb.