Huanglongbing (HLB) is a disease caused by the phloem-inhabiting bacterium Candidatus Liberibacter asiaticus (CLas) or Ca. L. americanus (CLam) (Bové, 2006). It is the most devastating citrus disease and a worldwide threat to the citrus industry (Bové, 2006; da Graça et al., 2016) because of the lack of a cure or/and an effective control method (Chin et al., 2014). The most important and effective measure to control the spread of HLB in citrus is the management of the key vector, the Asian citrus psyllid, Diaphorina citri Kuwayama (Hemiptera: Liviidae) (Stansly et al., 2014; Qureshi et al., 2014; Monzo and Stansly, 2015). Diaphorina citri nymphs and adults can acquire CLas by feeding on infected plants (Inoue et al., 2009; Pelz-Stelinski et al., 2010). If the pathogen is acquired at the nymphal stage, adults are able to transmit it immediately after emergence (Inoue et al., 2009). Previous studies have demonstrated that feeding on infected plants could increase the fecundity and produce more offspring of D. citri (Pelz-Stelinski et al., 2010; Pelz-Stelinski and Killiny, 2016). Diaphorina citri has highly efficient CLas transmission, and even an individual adult can successfully transmit the pathogen to healthy plants within a short time period (Pelz-Stelinski et al., 2010). Currently, D. citri and HLB have spread to almost all citrus-growing regions worldwide (Ghanim et al., 2017).

Diaphorina citri is an oligophagous insect with a host range restricted to plants in the Rutaceae family (Li, 2011) that includes more than 50 species (Halbert and Manjunath, 2004). However, D. citri also tends to feed on weeds of non-Rutaceae surrounding citrus orchards, such as Bidens alba L. (Asteraceae), Eupatorium capillifolium (Lam.) Small ex Porter & Britton (Asteraceae), Ludwigia octovalvis (Jacq.) P.H. Raven (Onagraceae) (Johnston et al., 2019; George et al., 2020), Solanum nigrum L. (Solanaceae), Ageratum conyzoides L. (Asteraceae), and Praxelis clematidea R.M. King & Robinson (Asteraceae) (Lu et al., 2021). These weeds have been suggested as alternative or secondary host plants for D. citri and allow the insect to survive in the absence of suitable hosts, although they are unable to reproduce and survive for an extended period on these hosts (Tiwari et al., 2010; Johnston et al., 2018; Johnston et al., 2019). Thus, these alternative hosts could play an important role in area-wide management of D. citri and affect control measures such as insecticide applications. Diaphorina citri can utilize hosts in the Rutaceae, Asteraceae, Onagraceae, Solanaceae, Moraceae (Thomas and de León, 2011; Felisberto et al., 2019), and Boraginaceae (Arshad et al., 2019), suggesting that their olfactory system can sense chemical signals emitted by these hosts. It also suggests that these signals might not be specific but common to several plant species such as green leaf volatiles (GLVs) (Riddiford, 1967).

GLVs are a series of six-carbon aldehydes and alcohols, their esters emitted by plants and produced from polyunsaturated fatty acids through the lipoxygenase (LOX) pathway (Holopainen and Gershenzon, 2010), such as (Z)-3-hexenol, 1-hexanol, (Z)-3-hexenyl acetate, and (±)-2-hexanol (Frati et al., 2008; Zheng et al., 2014). They usually serve as signal molecules to herbivorous insects in host finding. Although D. citri has a relatively simple olfactory system (Onagbola et al., 2008; Zheng et al., 2020), there are rhinarial plates known as the principal odorant sensors (Park and Hardie, 2002; Couthinho-Abreu et al., 2014) found on their antennae (Onagbola et al., 2008; Zheng et al., 2020), containing plant volatile-sensing olfactory neurons (Chapman, 1982; Kristoffersen et al., 2008). Couthinho-Abreu et al., 2014 demonstrated that the antennal neurons in D. citri strongly respond to odorants found in citrus. Several studies with D. citri revealed that stimuli emitted by flushing shoots of the host are implicated in host location (Patt and Sétamou, 2010; Sétamou et al., 2012; Sule et al., 2012). Therefore, these studies point to the feasibility of developing an odorant-based approach for improving D. citri surveillance and control.

Nicotiana tabacum is an important commercial crop in China that serves as a natural source of active compounds with important applications in medicine and agriculture for centuries (Peng et al., 2022). For instance, nicotine (a natural compound from N. tabacum) and its synthetic derivatives were used as insecticides against sucking insect pests (Feyereisen, 2018). Fu et al. (2008) found that (E)-2-hexenal and (Z)-3-hexenyl acetate (GLVs from N. tabacum) significantly attracted Helicoverpa assulta Guenée and H. armigera Hübner to N. tabacum plants. Several studies have demonstrated that D. citri adults may move to surrounding non-host plants (at least temporarily) when host conditions are unfavorable and insecticide sprays induce their dispersal (Lewis-Rosenblum et al., 2015; Johnston et al., 2018; Johnston et al., 2019; Lu et al., 2021). In South China, it is a common practice to plant N. tabacum around citrus orchards, and our field survey found that D. citri adults on N. tabacum leaves that moved from nearby citrus orchards were all dead ( Supplementary Figure 1 ). While the applied significance of this finding for D. citri and HLB management can be significant, the mechanistic basis of this process remains unclear. We hypothesized that the GLVs from N. tabacum plants are involved in D. citri host location, especially when adults move to N. tabacum plants due to unfavorable host conditions. Following, adults on N. tabacum are killed by the plant’s secondary metabolites such as nicotine. Therefore, the main objective of this study was to determine the preference and performance of D. citri on N. tabacum that could offer some practical methods to control the spread of D. citri and HLB.

Our study revealed that N. tabacum and synthetic GLVs (Z)-3-hexenol and (E)-2-hexenol are attractive to adult D. citri, supporting the notion that this insect relies on olfaction to locate its host plants. GLVs, especially leaf alcohols, play an important role in host location for other phytophagous insects, for example (Z)-3-hexen-1-ol and 1-hexanol for Lygus rugulipennis Poppius (Frati et al., 2008) and (±)-2-hexanol for Aleurodicus dispersus Russell (Zheng et al., 2014). (E)-2-hexenal and (Z)-3-hexenyl acetate, the synthetic GLVs from N. tabacum, were also attractive to male D. citri adults ( Table 3 ), and (Z)-3-hexenyl acetate was reported to synergize with the sex pheromone of H. armigera (Kvedaras et al., 2007). Xia et al. (2017) found that D. citri adults prefer N. tabacum plants rather than Murcott tangor (host plant). In addition, previous studies have reported that D. citri adults may be able to use some non-host plants opportunistically as secondary host plants for acquiring water or alternate food when optimal hosts are scarce or absent (George et al., 2020; Lu et al., 2021). Thus, D. citri adults are likely to move from citrus orchards to N. tabacum if they grow next to each other.

Our EPG results showed that D. citri adults only secreted saliva and ingested sap from phloem in N. tabacum leaves and spent the longest duration in phloem sap ingestion (E2) ( Figure 2 ), which is evidence of a phloem-feeding habit in this psyllid species (Bonani et al., 2010). Diaphorina citri adults performed more phloem feeding activities (D, E1, E2) than other stylet activities (NP, C) on N. tabacum leaves, which agree with observations that adults more readily feed on phloem from citrus to obtain nutrition (Bonani et al., 2010; Zhu et al., 2010). Moreover, CLas can occur in N. tabacum (Francischini et al., 2007), and Xia et al. (2017) reported that D. citri adults can also transfer the bacteria to N. tabacum after feeding. The survival time of D. citri on N. tabacum was much shorter than that under starvation conditions, suggesting that the death of adults on N. tabacum was not caused by starvation but could be related to the toxic components of N. tabacum. Nicotine is the predominant alkaloid accumulating in the leaves of most N. tabacum varieties and represents 90–95% of the total alkaloid content (Wang et al., 2000; Shoji et al., 2010). Several studies have demonstrated that neonicotinoid insecticides derived from nicotine, such as thiamethoxam and imidacloprid, are highly toxic to D. citri (Boina and Bloomquist, 2015; Miranda et al., 2016).

Previous studies have reported that D. citri may disperse from citrus orchards and utilize secondary host plants as a reservoir, facilitating their survival when primary host conditions are unfavorable or when these hosts are sprayed with insecticides (Lewis-Rosenblum et al., 2015; Johnston et al., 2018; Johnston et al., 2019; Lu et al., 2021). However, our study showed that D. citri adults that move to N. tabacum from citrus orchards would die after feeding on these secondary plants. Thus, N. tabacum could be used as a dead-end trap for adult D. citri.

For N. tabacum to kill adult D. citri, two conditions must be met. (i) Nicotiana tabacum should be grown nearby citrus crops. Diaphorina citri is critical for the spread of CLas due to their strong capacity to move between infected and healthy trees within an orchard (Wu et al., 2018). Insecticide applications to control D. citri are the primary management strategy currently recommended for CLas (Stansly et al., 2012) and could promote this movement. (ii) Diaphorina citri adults must be pushed away from citrus trees by insecticide applications or as a result of poor host conditions. Basically, adult D. citri dispersal requires citrus trees to either lack citrus flushes or to be sprayed with pesticides. However, while moving to N. tabacum, D. citri are unable to survive ( Figure 1 , Xie et al., 2015; Xia et al., 2017). In this context, N. tabacum could be used as a dead-end trap plant, which has considerable potential in integrated pest management (IPM) for D. citri. Several kinds of dead-end trap plants have been suggested and used in IPM, such as N. tabacum and Barbarea vulgaris R. Br. for Plutella xylostella (L.) larvae (Gupta and Thorsteinson, 1960; Shelton and Nault, 2004; Møldrup et al., 2012) and Vetiveria zizanioides (L.) Nash for Chilo suppressalis Walker larvae (Lu et al., 2019).

Diaphorina citri eggs were only observed on M. exotica plants ( Figure 1B ). The pre-oviposition period of D. citri females ranged from 7 to 20 days (Li et al., 2019); thus, it is likely that females on N. tabacum died before laying eggs. Furthermore, adults on N. tabacum died within 30 h of feeding, whereas they initiated oviposition 5 days after feeding on M. exotica. In other words, N. tabacum is very effective as a dead-end trap because D. citri females were attracted to the plants, fed on them, and died shortly prior to laying any eggs. Our results suggest that it is worthwhile to grow N. tabacum in and around citrus orchards. To optimize area-wide management of D. citri, field studies are needed to evaluate the use of N. tabacum surrounding citrus orchards as dead-end traps.

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding authors.

LZ, WW, and WC designed the study. LZ performed the majority of experiments with the help of QX, GG, and YL. GG and WW analyzed the data, and LZ wrote the first version of the manuscript. MY, SS, and WC revised the manuscript and approved the submitted version. All authors contributed to the article and approved the submitted version.