Modeling virus inactivation on salad crops using microbial count data

Microbial counts of the persistent Bacteroides fragilis bacteriophage B40-8 from a virus decay experiment conducted under glasshouse conditions were used to model the decay of viruses on wastewater-irrigated lettuce and carrot crops. The modeling approach applied gave specific consideration to the discrete nature of microbial count data. The experimental counts were best fit by a negative binomial distribution indicating highly dispersed distribution of viruses on lettuce and carrot crops following irrigation with wastewater. In addition, there was evidence for biphasic inactivation of viruses, signifying the presence of a persistent subpopulation of viruses that decayed slowly, resulting in virus accumulation on the crop surface over subsequent irrigations. Maximum likelihood estimates of initial and persistent subpopulation inactivation rates were 2.48 day(-1) and 0.51 day(-1) for lettuces and 0.84 day(-1) and 0.046 day(-1) for carrots. Maximum likelihood estimates of the persistent virus subpopulation size were 0.12% and 2% for lettuce and carrots, respectively.