Modelling the key drivers of an aerial Phytophthora foliar disease epidemic, from the needles to the whole plant.

Abstract

Understanding the epidemiology of infectious diseases in a host population is a major challenge in forestry. Radiata pine plantations in New Zealand are impacted by a foliar disease, red needle cast (RNC), caused by Phytophthora pluvialis. This pathogen is dispersed by water splash with polycyclic infection affecting the lower part of the tree canopy. In this study, we extended an SI (Susceptible-Infectious) model presented for RNC to analyse the key epidemiological drivers. We conducted two experiments to empirically fit the extended model: a detached-needle assay and an in vivo inoculation. We used the detached-needle assay data to compare resistant and susceptible genotypes, and the in vivo inoculation data was used to inform sustained infection of the whole plant. We also compared isolations and real-time quantitative PCR (qPCR) to assess P. pluvialis infection. The primary infection rate and the incubation time were similar for susceptible and resistant genotypes. The pathogen death rate was 2.5 times higher for resistant than susceptible genotypes. Further, external proliferation of mycelium and sporangia were only observed on 28% of the resistant ramets compared to 90% of the susceptible ones. Detection methods were the single most important factor influencing parameter estimates of the model, giving qualitatively different epidemic outputs. In the early stages of infection, qPCR proved to be more efficient than isolations but the reverse was true at later points in time. Isolations were not influenced by the presence of lesions in the needles, while 19% of lesioned needle maximized qPCR detection. A primary infection peak identified via qPCR occurred at 4 days after inoculation (dai) with a secondary peak observed 22 dai. Our results have important implications to the management of RNC, by highlighting the main differences in the response of susceptible and resistant genotypes, and comparing the most common assessment methods to detect RNC epidemics.