The Editor’s Choice for issue 55:4 is written by Senior Editor, Phil Stephens. The selected article is Grower and regulator conflict in management of the citrus disease Huanglongbing in Brazil: A modelling study by Craig et al.

Plant disease is already recognised as a major driver of crop yield losses. With a huge proportion of the growing human population’s food intake dependent on a relatively small number of crop species, plant disease is also a significant concern for food security. Policies to ensure that plant diseases are contained when detected, and regulations to ensure that those policies are followed effectively, are absolutely essential in reducing the risks plant diseases pose to economies and food security.

Citrus production (Fig. 1) is an important contributor to the economies of several regions in the Americas and elsewhere. Over the last decade, however, major impacts on citrus growing areas have resulted from the spread of ‘citrus greening’ disease (Fig. 2). Citrus greening is caused by the bacterial disease Huanglongbing, which is vectored by plant lice from the family Psyllidae. As its name suggests, citrus greening leads to trees producing green fruits; they are also bitter and misshapen. Huanglongbing has been described as the most damaging bacterial disease for trees and has a potentially devastating impact on citrus production. As one example, it led to a drop in citrus production in Florida by close to a third in just two years.

Huanglongbing can spread rapidly and, at present, there is no cure for the disease (which is thought to have arisen relatively recently, providing limited opportunity for plants to evolve resistance). As a result, containment of the disease is essential and is the only available management strategy. However, containment is made particularly problematic by the fact that infected trees may remain asymptomatic for several years, providing ample opportunity for the disease to spread before infected trees can be identified and removed.

In 2008, based on the best available evidence at the time, the Brazilian government issued a specific regulatory instruction to the effect that symptomatic trees must be removed at quarterly intervals. It also directed that any production unit (such as an orange grove) with at least 28% symptomatic trees must be removed completely. The regulation has caused conflict with growers because even symptomatic trees can produce some profitable yield and, hence, many growers would prefer not to remove them (at least, when symptoms are minor). Moreover, although the regulation concerns containment of the disease, it does not require vector control. The conflict between growers and regulators, as well as the potential for including vector control in management of the disease, are the focus of this Editor’s Choice article by Andrew Craig and colleagues. Specifically, the authors set out to determine whether there were better ways to manage the system in order to reduce infections (the aim of the regulators) and avoid major yield losses (the aim of the growers).

Craig and colleagues used Bayesian analyses of field data to determine plausible distributions for key parameters determining the spread of the disease, such as the rates of primary infection (from outside the grove) and secondary infection (from within the grove). These parameters were then used in a spatially-explicit stochastic model of the fates of thousands of individual trees moving between an uninfected status and various stages of infection (from exposed to asymptomatic and then symptomatic infection). The model accounted for imperfect detection of disease, and the relationship between infection and yield loss. Various management approaches were assessed, including the removal of symptomatic trees, the removal of larger production units, and spraying for vector control.

The authors found that, unmanaged, the spread of Huanglongbing is very rapid. By the time disease is detected in 28% of trees within a grove, typically all trees in the grove are diseased (the association that underlies the regulatory threshold of 28% infection for removal of the entire grove). However, disease detection is a very poor indicator of the prevalence of the disease, which can be 90% when disease is detected in as few as 4% of trees. For this reason, Craig and colleagues advocate against the use of a threshold for removing a whole production unit. Instead, they advocate a combination of area-wide spraying (to substantially reduce abundances of tree lice and, with that, rates of both primary and secondary infections) and rigorous removal of symptomatic trees. If spraying translates into expected reductions in infection, fewer symptomatic trees will have to be removed, which should be more acceptable to growers.

It is particularly good to see this combination of rigorous analysis of field data and management-relevant modelling used to inform both policy for, and reduce conflict surrounding, the containment of such an economically-damaging and worrying disease.

Journal of Applied Ecology offers authors the opportunity to replace the final ‘Synthesis and applications’ point of their abstract with a point summarising the ‘Policy implications’ of their work.  As we noted in a recent Editorial, it is important that those implications should be relevant to specific existing policy. The work by Craig and colleagues is a great example of applied ecological work that relates directly to specific regulation, with strong potential to inform its development.

The full Editor’s Choice article, Grower and regulator conflict in management of the citrus disease Huanglongbing in Brazil: A modelling study is free to read for a limited time in issue 55:4 of Journal of Applied Ecology.