Roguing with Replacement in Perennial Crops: Conditions for Successful Disease Management

Abstract

Replacement of diseased plants with healthy plants is commonly used to manage spread of plant pathogens in perennial cropping systems. This strategy has two potential benefits. First, removing infected plants may slow pathogen spread by eliminating inoculum sources. Second, replacing infected plants with uninfected plants may offset yield losses due to disease. The extent to which these benefits are realized depends on multiple factors. In this study, sensitivity analyses of two spatially explicit simulation models were used to evaluate how assumptions concerning implementation of a plant replacement program and pathogen spread interact to affect disease suppression. In conjunction, effects of assumptions concerning yield loss associated with disease and rates of plant maturity on yields were simultaneously evaluated. The first model was used to evaluate effects of plant replacement on pathogen spread and yield on a single farm, consisting of a perennial crop monoculture. The second model evaluated effects of plant replacement on pathogen spread and yield in a 100 farm crop growing region, with all farms maintaining a monoculture of the same perennial crop. Results indicated that efficient replacement of infected plants combined with a high degree of compliance among farms effectively slowed pathogen spread, resulting in replacement of few plants and high yields. In contrast, inefficient replacement of infected plants or limited compliance among farms failed to slow pathogen spread, resulting in replacement of large numbers of plants (on farms practicing replacement) with little yield benefit. Replacement of infected plants always increased yields relative to simulations without plant replacement provided that infected plants produced no useable yield. However, if infected plants produced useable yields, inefficient removal of infected plants resulted in lower yields relative to simulations without plant replacement for perennial crops with long maturation periods in some cases.

Conditional Vector Preference Aids the Spread of Plant Pathogens: Results From a Model

“…In agricultural settings, if such preferences were present, a manager would need to increase ␥, the rate at which the pathogen is removed from the system, or reduce the amount of vectors moving among plants p v . In practice, one may increase ␥ by roguing (removing diseased plants) (Jeger et al 2004, Sisterson andStenger 2013). Alternatively, insecticides or biological control can increase the death rate of the vectors .…”

Cost-Effective Control of Plant Disease When Epidemiological Knowledge Is Incomplete: Modelling Bahia Bark Scaling of Citrus

“…Ultimatately this derives again from the limited disperal ability of the pathogen that causes BBSC. For pathogens capable of faster and/or long-distance dispersal, synchronisation in control is acknowledged to be extremely important, since otherwise the pathogen is able to persist, bulk-up and repeatedly cause devastating reinvasion from uncontrolled areas that act as refugia [10]. Following common practice in the Brazilian citrus industry, removed plants were not replaced in our model, which again facilitated control.…”

“…Mathematical models of plant disease can be used to screen and assess control strategies [1]–[10]. Although work on plants is not subject to the ethical concerns that hamper experimentation targeting pathogens of animal or human hosts, mathematical modelling nevertheless becomes particularly compelling for plant diseases when logistic constraints mean that experimentation would be costly or difficult.…”

“…As little is known of the putative BBSC pathogen, and even less about any potential vector, it is difficult to reliably estimate the efficacy of any chemical [23] or biological [24] control. We therefore concentrate on cultural strategies [25], and focus on the effectiveness of reducing the density of planting [26] and of roguing [10] (i.e. searching for and removing infected plants).…”

“…The only constraint is that pathogens must cause symptoms that can be detected, either by visual inspection or by diagnostic testing. Roguing has been included in non-spatial mathematical models for a number of years [1], [2], [54], [55], and more recent work has embedded control by roguing in spatial models of pathogen spread [10], [56]–[58], although realistic parameterisation of pathogen dispersal is less common [6]–[8]. Typically these later models have also considered culling, in which all hosts within a particular distance of a symptomatic focal host are removed at the time of control.…”