Repellency of forty‐one aromatic plant species to the Asian citrus psyllid, vector of the bacterium associated with huanglongbing

Abstract Huanglongbing (HLB) is the most devastating citrus disease worldwide. The organism associated with the disease is spread by an insect vector, Diaphorina citri, commonly known as Asian citrus psyllid (ACP). Current management of HLB relies either on physical removal of the infected plants or on chemical control of ACP. Both methods are costly and not overly effective. In addition, public concerns regarding insecticide residues in fruit have greatly increased in recent years. It has been hypothesized that plant volatiles could act as repellents to ACP, thus reduce the incidence of HLB. To test this hypothesis, the repellency of fresh tissues of 41 aromatic plant species to ACP was investigated. The repellency of individual species was determined using a Y‐tube olfactometer. Our results showed that volatiles of five plant species were highly effective in repelling ACP with repellency as much as 76%. Among these, the tree species, Camptotheca acuminate, and the two shrubs, Lantana camara and Mimosa bimucronata, could potentially be planted as a landscape barrier. The two herbs, Capsicum annuum and Gynura bicolor, could potentially be used as interplantings in orchards. This is the first time that the repellency of fresh tissues from a diverse range of plant species to ACP has been determined. Although further field evaluation of various interplanting regimes and landscape barriers are needed to assess their effectiveness, our results showed that these aromatic species, being highly repellent to ACP, offer great potential as more cost‐effective and environmentally sustainable alternatives to the current methods of managing HLB.


| INTRODUC TI ON
or greening disease is the most devastating citrus disease worldwide and is considered a major threat to global citrus industry (Bove, 2006). The disease is prevalent in all major citrus growing regions, including Africa, Asia, and North and South America (Bove, 2006;Yan et al., 2015), and causes considerable economic losses. In China, the disease is prevalent in 11 of the 19 citrus growing provinces; in Guangdong alone, the disease causes annual economic loss of over 10 billion CNY (or 1.5 billion USD) (Qin, 2018).
In Florida, United States, where HLB was first recorded in 2003 and has since been found in 80% of commercial orchards; it has caused a 71% decline in fruit production ( (Gottwald, 2010). The Asian strain of the disease, prevalent in China and Southeast Asia, is caused by Ca. L. asiaticus and is spread by an insect vector, the Asian citrus psyllid (ACP), Diaphorina citri (Gottwald et al., 2014;Halbert & Manjunath, 2004;Yan et al., 2015). The disease affects fruit production by causing premature fruit drop and producing fruit with low economic value (Bassanezi et al., 2009;Plotto et al., 2010) or causing whole plant death in extreme cases (Zheng et al., 2018).
Currently, there is no adequate treatment for infected plants.
Management of the disease relies either on physical removal or pruning of the infected plants, or on chemical control of the insect vector (Bove, 2006;Wang, 2019). Pruning of infected parts of plants is costly and not overly effective given the long latent period of the disease (Bassanezi et al., 2013;Rouse et al., 2017) and is considered a non-viable approach for HLB management (Rouse et al., 2017;Vashisth & Livingston, 2019) Using insecticides to control ACP is presently the primary option for HLB management. However, increase in the resistance of ACP to various insecticides has reduced their effectiveness (Hall et al., 2013;Ichinose et al., 2010;Naeem et al., 2016). In addition, insecticides can have adverse effects on non-target species (Beloti et al., 2015) and increasingly becoming a health hazardous both to the citrus workers and to the general public (Beloti et al., 2015;Wang, 2019 (Diaz, 2016;Hill et al., 2007;Karunamoorthi et al., 2009;Moore et al., 2002) and stored-product pests (Tribolium castaneum and Liposcelis bostrychophila) (Yang et al., 2015). In horticulture, volatiles from a garlic-pepper combination were effective in controlling Guatemalan potato moth (Tecia solanivora) in potato crops (Jimenez & Poveda, 2009). Aromatic species, Saturela hortensis and Agerarum houstonianum, interplanted with pears (Pyrus pyrifolia) halved the abundance of scarab beetles (Serica orientalis) (Tang et al., 2013).
Similarly, volatiles have been used for ACP control. In laboratory and field studies, volatiles from guava (Psidium guajava) were effective in repelling ACP and reducing the incidence of HLB (Barman et al., 2016;Gottwald et al., 2014;Zaka et al., 2010). It is therefore hypothesized that using volatiles from aromatic plants in HLB management could be a cheaper, more effective, and more environmentally friendly alternative to current methods, since these volatiles could repel ACP.
To test the hypothesis that volatiles of aromatic plants can effectively repel citrus psyllids, the repellency of volatiles of 41 aromatic plant species to ACP in Ganzhou, Jiangxi Province, China were investigated. Most of the selected plant species occur locally in Ganzhou. Our objective was to identify plants with volatiles that are effective in repelling ACP in order to improve the control of HLB in citrus.
Ganzhou is one of the major citrus growing regions of China.
HLB is prevalent in the region and has caused significant economic loss. Finding a cost-effective, and environmentally friendly solution to HLB is imperative. This study formed part of the national program for scientific management of HLB in the Chinese citrus industry.

| Adult psyllids (Diaphorina citri)
Live adult psyllids were collected from the orchards of Ganzhou Citrus Research Centre, Jiangxi Province, China. These psyllids were collected on the day of the experiment to ensure that they were active and vigorous during the experiment. Before the experiment, the psyllids were deprived of food for 2 hr to ensure that they were actively seeking food. We did not examine individual psyllids for bacterial infection. As these psyllids were collected in orchards of HLB prevalent areas, it is likely that some of

| Plant species selection
Forty-one species of plants were selected for evaluation (Table 1)

| Repellency assessment
The repellency of 41 plant species to psyllids was determined using a Y-tube olfactometer (Figure 1) in laboratory conditions. The design of TA B L E 1 List of plant species tested for repellency to Asian citrus psyllids (Diaphorina citri) the Y-tube olfactometer followed that of Li et al. (2008). The Y-tube olfactometer was made of an air pump, two air-purifiers (activated carbon), two humidifiers, test and control chambers, two gas flow meters, two Y-tube (clear glass), a light globe, and an insect-holding chamber. The light globe was positioned above the two Y-arms of the olfactometer to ensure that lighting to both arms was similar. The single arm of the Y-tube was wrapped with black cloth. During the experiment, air was supplied by the air pump, filtered and humidified before passing through the test or control chambers. Air flow through the chambers to the two arms of the Y-tube was regulated by an airflow meter, which was set at 400 ml/min for the duration of the experiment. The Y-tube olfactometer was sterilized at 100°C for 30 min before the experiment. During assessment, the Y-tube olfactometer was placed in horizontal position within a 1.0 × 0.6 × 0.6 m cardboard box. A fluorescent 1,600 lux daylight (5,000-6,500 K) globe, was mounted on the box to provide uniform lighting for the two Y-tube arms.
The olfactometer was set up similarly to that of Zaka et al. (2010), but only leaves were used as volatile source materials. Leaves of the test and control plants were cleaned with distilled water, blot-dried, and cut to small pieces of less than 2 mm; 5 g of leaves of test plants and 5 g of leaves of control plants were placed in the test chamber; 5 g of leaves of control plants only was placed in the control chamber.
The experiment was conducted in a laboratory at 25°C. At the start of each assessment, the freshly cut leaves were allowed to sit for 5 min to allow the plant volatiles to permeate throughout the Y-tubes, then 40 healthy and active psyllids were released in the insect-holding chamber. The number of psyllids entering the arms of the Y-tube, passing the two-third mark, and remaining for >30 s were recorded separately for both the test and control chamber arms.
Each assessment was conducted for 20 min, starting from the time the psyllids were released in the holding chamber. All psyllids were removed after each assessment. The Y-tube and the test and control chambers were cleaned, deodorized after each assessment by spraying with 75% alcohol then rinsing in distilled water three times before oven drying. For each test plant species, the assessment was repeated three times. Time laps between tests were approximately 15 min. Fresh psyllids were used in each assessment. We did not identify the sex of the collected psyllids. Each group of 40 individuals was randomly selected from a pool of psyllids collected a few hours before the experiment. We assumed that sex ratio between groups of psyllids (40) was similar and any sex-related phenomenal response difference was therefore minimum.
In all, 4,920 psyllids were used in the assessment of the 41 test plant species.

| Data analysis
An independent student's t test was used to determine the difference in number of psyllids entering the test and control arms of the Y-tube for each plant species tested. ANOVA was used to examine the difference between test plant species using plant species as an independent factor and repellence as a dependent variable. The repellence was calculated as Wang et al. (2005): where T and C are the number of psyllids in the test and control Y-tube arm, respectively.
The repellence ranged from 0 to 1. When T equaled C, that is, psyllid counts at the two Y-tube arms are equal, repellence was 0, indicating the test plant species was not repellant. When C equaled 0, that is, no psyllids present in the test Y-tube arm, repellence equaled 1, indicating the test plant species was strongly repellant. Student's t test and ANOVA were performed using SPSS 21.0 for Windows (SPSS Inc., Chicago, IL, USA). Cluster analysis was conducted using R version 3.6.3 (R-Team, 2020) based on Euclidean distance and average linkage clustering.

| RE SULTS
The 41 plant species tested in this study were from 21 families ( A dendrogram grouping the 41 test plant species based on cluster analysis of their repellency to ACP is given in Figure 2. Three groups with >24% dissimilarity were clearly identified. The first group consisted of the five species that had the highest repellency (59%-76%) and lowest mean number of psyllids entering the testing arm of treatment plants (

| D ISCUSS I ON
There is no known treatment for Ca. L. asiaticus infected plants.
Unlike other diseases, in HLB-endemic regions, disease prevention is more important than treatment (Wang, 2019). Efficient control of the vector, ACP, is key to successful HLB prevention. In Brazil, controlling ACP has been effective in reducing HLB incidence to a level that is economically sustainable (Bassanezi et al., 2013). In citrusproducing areas of China, orchards remain productive and profitable with an infection incidence below 30% (Wang, 2019). However, the actual Ca. L. asiaticus incidence may be much higher due to the latent period (up to 5 years) between Ca. L. asiaticus infection and symptom expression. Thus, in these regions, the aim is to keep the number of Ca. L. asiaticus infected trees below the 30% threshold (symptom based).
Using aromatic plants to reduce the abundance of ACP so that Ca.
L. asiaticus incidence falls below this threshold is a promising management option. The practice of using plant volatiles to repel insects has been used for pest management in horticulture production such as potato (Berberich, 1988), pears (Tang et al., 2013) as well as in citrus (Fancelli et al., 2018;Yan et al., 2015). Incidentally, the repelling effects of volatiles of guava on D. citri were discovered by accident rather than by planned research. In 2004 in tropical Vietnam, citrus growers used guava as an intercropping plant to increase short-term cash flow as it fruits within 1 year compared to 3 or more years for citrus trees (Ichinose et al., 2012). They reported a lower incidence of HLB in these interplanted orchards. The efficacy of guava interplanting in citrus HLB was limited to 1 year. Further experiments demonstrated that interplanting of guava with citrus trees can effectively reduce the density of psyllids, leading to the lower level of HLB infected plants (Ichinose et al., 2012). These findings have been confirmed by several other studies that show volatiles of guava leaves are an effective repellent for psyllids and contribute to the reduced HLB incidence (Barman et al., 2016;Gottwald et al., 2014;Onagbola et al., 2011;Silva et al., 2016;Zaka et al., 2010).
In our study, the five plants in the group with the highest repellency of ACP ( Figure 2) could be useful for interplanting or as a landscape barrier to deter ACP from citrus orchards. Under laboratory conditions, volatiles of these species were able to reduce ACP numbers by 75%-87% (Table 2). Among these species, Camptotheca acuminate is a deciduous tree that can grow to 20 m. Both L. camara and M. bimucronata are shrubs that can grow to 2 and 6 m, respectively. These three species could be planted around the perimeter of citrus orchards as landscape barriers to ACP. Landscape barriers, consisting of trees or shrubs, can serve as windbreaks and have been reported to reduce the abundance of D. citri in orchards and protect them from HLB (Martini et al., 2015). Gynura bicolor and L. camara are both known for their use as medicinal plants. Gynura bicolor is a herb and has been used for treating diseases such as dysmenorrhea, dysentery, and ulcers. Lantana camara has been used for treating diseases such fever, mumps, and tuberculosis. The fifth species, C. annuum commonly grown for its fruit. These species all have higher repellency rate than Psidium guajava (73%-59% vs. 49%) ( Table 2), TA B L E 2 Psyllid counts (mean ± SE) and repellency (%) of 41 plant species sorted in order of decreasing repellence the species that has previously been reported to be repellent to ACP and capable of reducing the level of HLB (Ichinose et al., 2012;Silva et al., 2016), suggesting these species offer great potential for man-

| CON CLUS IONS
In conclusion, our results supported the hypothesis that plant volatiles can effectively repel ACP, the insect vector of the bacterium associated with HLB. Volatiles of five plant species were most effective in repelling ACP, with a repellency as high as 76%. Among these, the tree species, C. acuminate, and the two shrub species, L. camara and M. bimucronata could be planted around the perimeter of citrus orchards and used as a landscape barrier to the vector. The two herb species, C. annuum and G. bicolor, could be intercropped with citrus trees to repel ACP. Although further field evaluation of various interplanting regimes and landscape barriers are needed to assess their effectiveness, our results showed that these aromatic species, being highly repellent to ACP, offer great potential as more cost-effective and environmentally sustainable alternatives to the current methods of managing HLB. Zhengyun helped with plant nomenclature.

CO N FLI C T O F I NTE R E S T
There is no conflict of interest/competing interest. Jie Luo: Formal analysis (equal); methodology (equal).