Linking host distributions, parasite specialization, and parasite distribution
We used generalized linear mixed effects models (GLMMs) to test how parasite specialization in experiments was related to 1) host community variation (crop area harvested, Q1), 2) parasite occurrence across hosts (from recorded natural history collections, Q2), and 3) predicted specialization from abiotic environment (ENM contrasts, Q3). For each of the three predictors of host specialization, we modeled relative emergence of an S. hermonthica population i , on a focal host as follows:
Emergenceij = β0 + β1Xi + S0j + 𝜖ij ,
where β0 and β1 are coefficients, Xi is the value for one of the predictors of specialization (i.e. crop area harvested, observed parasite occurrence, or ENM contrast), S0j is the random effect of host genotype j , and 𝜖ijis the irreducible error. Models were built using the ‘lmer’ function from the lme4 package (Bates et al. 2015) and corresponding model summaries and statistical parameters for the fixed effects were calculated with the lmerTest package using Satterthwaite’s method (Kuznetsova et al. 2017).