Beta Diversity
To deal with unequal sequence coverage and compositional variation
within the data, we used Cumulative Sum Scaling (CSS) normalization on
the filtered dataset using the R package metagenomeseq (Paulson
et al., 2013) following the methods in Maraci et al., (2021). We then
log (x + 0.0001) transformed the data and later corrected the
transformed values by subtracting the log of the pseudo count (Thorsen
et al., 2016). Dissimilarity matrices were computed based on
Bray–Curtis (Bray & Curtis, 1957), unweighted UniFrac (C. Lozupone &
Knight, 2005), and weighted UniFrac (Lozupone et al., 2007). To
visualize the dissimilarities between nestlings based on parasite
treatment and urbanization, we used a principal coordinate analysis
(PCoA) using the ordinate function in the phyloseq package
(McMurdie & Holmes, 2013). To determine the effect of parasitism and
urbanization on beta diversity metrics, we used PERMANOVAs with parasite
treatment, location, year (2018, 2019), and the interaction between
parasite treatment and location as fixed effects and beta diversity
metrics as response variables. Nest identification was included as a
random effect in all models. We used the adonis2 function with thevegan package for these analyses (Oksanen et al., 2022).