Statistical Analysis

All experiments were performed in technical duplicates with two experimental trials. Log reductions were calculated as -log10FinalPFUInitialPFU, and data were analyzed in R Studio using a generalized linear model (GLM) with Quasipoisson error distributions due to heteroscedasticity, non-normality, and overdispersion (R Studio Team, 2020). More specifically, the Q–Q plot indicated non-normality of the data based on deviation of the tails from the reference line. Regarding heteroscedasticity, the ratio between the largest and smallest fitted residual is 2748.83 indicating significant deviation from the threshold (1.50) for homoscedasticity. To quantify overdispersion, the residual deviance was divided by the residual degrees of freedom to yield a value of 53.26. Since this value is much greater than one, a Quasipoisson distribution was applied in place of a Poisson distribution. The treatment means and their associated standard errors were calculated using estimated marginal means. Statistical differences between treatments were determined using multiple pairwise comparisons and visualized using compact letter display. The data were analyzed in R (R Core Team, 2021) using the base, base, ggplot2 (Wickham, 2016), emmeans (Length et al., 2021), tidyverse (Wickham et al., 2019), ggpubr (Kassambara, 2020), gdata (Warnes et al., 2022), rstatix (Kassambara, 2021), lme4 (Bates et al., 2015), lmertest (Kuznetsova et al., 2017), multcomp (Hothorn et al., 2008), and multcompView (Graves et al., 2019) packages.