Introduction

Plum cultivation is a promising horticultural activity in the Brazilian southern states (Paraná, Santa Catarina e Rio Grande do Sul), due to the high and currently unmet demand in the national market, with 27,541 tons of imported fruit, valued at US$ 34 million in 2017 (Faostat 2020). Plum production in Brazil is based on Japanese cultivars (Prunus salicina Lindl, Rosaceae), mainly because of their lower chilling requirement compared with European cultivars (Prunus domestica L., Rosaceae) (Ojima et al. 1983).

Since 1970, some orchards in Brazil have been decimated by plum leaf scald (PLS). As a result, disease is the main factor limiting the expansion of the plum crop in the country (Eidam et al. 2012). The disease is caused by the bacterium Xylella fastidiosa (Wells et al. 1987) subsp. multiplex, which colonizes the foregut of xylem-sap feeding insects and the xylem of infected plants (Chatterjee et al. 2008). In general, disease symptoms (leaf scorch) occur only after a long incubation period when the pathogen titers achieve high density (Daugherty et al. 2011), weakening, and in more severe cases, causing premature plant death (Raju et al. 1982).

The spread of X. fastidiosa occurs through infected propagative material used for grafting (He et al. 2000) and by insects with piercing-sucking mouthparts that are specialized xylem feeders (Almeida and Nunney 2015), such as sharpshooter leafhoppers (Hemiptera: Cicadellidae: Cicadellinae) and spittlebugs (Hemiptera: Cercopoidea) (Redak et al. 2004). Sharpshooters are highly polyphagous, feeding on a wide range of herbaceous and woody plants (Hopkins and Purcell 2002; Mizell et al. 2008), showing greater preference for some plants over others (Tertuliano et al. 2012). Differences in host preference can be attributed to the nutritional status of the plant species, particularly regarding the composition of the xylem sap (Brodbeck et al. 1993). Longer feeding periods on preferred hosts may favor transmission, by increasing the chances for both acquisition and inoculation to occur (Daugherty et al. 2009; Sandanayaka et al. 2013; Markheiser et al. 2019).

Nowadays, PLS is still a limiting factor for crop expansion, mainly due to the lack of research information required to develop efficient management strategies. Control is mainly based on planting healthy nursery trees, roguing, avoiding establishing new orchards in proximity to infected areas, chemical and biological control of vectors and alternative hosts (ex: rye grass, macela, creeping signal grass, Brazilian pusley, among others) (Dalbó et al. 2010; Leite et al. 1997). The use of resistant cultivars would be an advantageous method to control PLS. While some cultivars (‘Carazinho’, ‘Sanguinea’, ‘Chatard’ and ‘Piamontesa’) have been reported as resistant, they usually produce low-quality fruit (Dalbó et al. 2018). In addition, the inheritance of resistance is polygenic and predominantly recessive (Dalbó et al. 2010).

Recently, the breeding program of the Agriculture Research and Rural Extension of Santa Catarina (EPAGRI) identified genotypes (‘SC7’ and ‘SC13’) that do not manifest PLS symptoms and have great fruit qualities, such as appearance, flavor and lower acidity at maturity (Dalbó et al. 2018). Detection trials of X. fastidiosa in these field-established plants was regularly performed by polymerase chain reaction (PCR) and the results were consistently negative for 10 years, indicating that these genotypes were not infected under field conditions, despite being close to naturally-infected cultivars, such as ‘Laetitia’ (widely cultivated in southern Brazil). The ‘SC13’ genotype was shown to be colonized by X. fastidiosa only when scions were grafted on infected plants (xylem vessels connected) (Dalbó et al. 2018). On the other hand, ‘SC7’ was considered immune, since symptoms were not observed after grafting this genotype on infected plants (Dalbó et al. 2016). Interestingly, monitoring observations with yellow sticky traps in experimental fields indicated a lower number of sharpshooters trapped on ‘SC7’ and ‘SC13’ plants compared with ‘Laetitia’(control host) (M. A. Dalbó, unpublished data), suggesting that the field resistance of the former two plum genotypes could be related to vector preference.

The identification of plum genotypes that do not become infected under field conditions has opened up new prospects to control PLS in Brazil. However, it remains unknown whether vector behavior activities related to host plant selection (e.g. host searching, landing, probing, acceptance and suitability) play any role on the very marginal PLS prevalence observed on some plum genotypes under field conditions. Landing and settling behavior can be studied by choice and no-choice assays (Smith and Capinera 2005; Marucci et al. 2005). The electrical penetration graph (EPG) technique (Tjallingii 1978) has been widely used to evaluate the insect stylet penetration (probing) behavior and acceptance of plants for feeding (Sandanayaka et al. 2007; Sandanayaka and Backus 2008; Miranda et al. 2009; Sandanayaka et al. 2013; Cornara et al. 2018).

The goal of this study was to describe the settling preferences, probing behavior and feeding of sharpshooter vectors on plum genotypes that differ in field resistance to leaf scald caused by X. fastidiosa. The non-affected genotypes ‘SC7’ and ‘SC13’ were compared with the closely-related ‘Laetitia’, which shows natural infections (control host). Two sharpshooter species known as vectors of X. fastidiosa in citrus were used in the study, Bucephalogonia xanthophis (Berg) (Miranda et al. 2009) and Sibovia sagata (Signoret) (Esteves et al. 2019), because they are commonly found in plum orchards of Southern and Southeastern Brazil (Azevedo Filho et al. 2016) and are likely associated with PLS spread.

Materials and methods

Biological material: Sharpshooters and plants

Adults of B. xanthophis were collected from plants of Lagerstroemia indica (L.) (Verbenaceae) and Duranta repens L. (Verbenaceae) (approximately 50% from each host) located in the municipality of Piracicaba-SP. Adults of S. sagata were collected from plants of Peumus boldus Molina (Monimiaceae), located in the municipality of Tatuí, Brazil. The insects were reared in aphid-proof cages (55 × 55 × 32 cm) covered with nylon screen (mesh size 96 × 26 | 680 μm aperture) on plants of Vernonia condensata Baker (Asteraceae), as described by Marucci et al. (2003), in a greenhouse equipped with a pad-fan type cooling system and thermostat-activated heater for temperature control (25 ± 5 °C, natural light) in Piracicaba, SP, Brazil.

Cuttings of approximately 40 cm of V. condensata were planted in plastic pots (Nutriplan® 15 cm × 20 cm × 14.5 cm) containing a substrate mix composed of shredded pine bark, peat, and expanded vermiculite at 60% w/w moisture content (Tropstrato HT Vida Verde, Mogi Mirim, SP, Brazil). The plants were kept in a greenhouse protected with anti-aphid screen, without temperature control, until they were used for sharpshooter rearing.

Two Japanese plum genotypes that showed no symptoms of PLS in the field [‘SC7’: crossing with (‘Laetitia’ × ‘Piamontesa’); and ‘SC13’: crossing ‘SC7’ × Fortune] and the control cultivar ‘Laetitia’, which is widely cultivated in southern Brazil and naturally infected with X. fastidiosa, were selected for the study. Healthy nursery trees of these genotypes were grafted on ‘A9’ rootstock and planted in plastic pots (Nutriplan® 12.5 cm × 17 cm × 12 cm; 2 L) containing the aforementioned substrate mix and fertilized by weekly fertigation [magnesium sulfate (0.83 g/L), potassium nitrate (0.36 g/L), monoammonium phosphate (0.14 g/L), zinc (0.02 g/L), copper (0.03 g/L), iron (0.09 g/L), ammonium nitrate (0.08 g/L) and calcium nitrate (0.91 g/L)]. The plants were kept in a greenhouse as described above until they reached approximately 40 to 50 cm in height.

Settling preference assay

Settling preference assays were conducted for B. xanthophis and S. sagata on plum genotypes (‘SC7’, ‘SC13’ and ‘Laetitia’), by releasing 40 adults into each observation cage (base of 75 × 75 cm, height of 115 cm) with a translucent plastic cover (Bugdorm type, model BD2400F, MegaView Science Co, Taichung Taiwan). Inside the cage, six healthy nursery trees (two of each plum genotype) were arranged randomly in a circle, forming an experimental arena, according to the design described by Fereres et al. (1999). Before releasing the insects, they were starved for 20–30 min in 50-ml conic plastic tubes (Falcon ™) (17 mm O.D.; 120 mm length) closed with screw caps and covered with black tape, in groups of 10 individuals per tube (four tubes per replicate). The tubes were placed on a flight platform positioned in the top of the cage (20 to 30 cm above the top of the plants) and then were opened to allow the insects to exit. The total number of insects settled on each plant was assessed at 3, 6, 9, 24, 27, 30, 48, 51, and 54 h after the release, which was always started between 8:00 and 9:00 a.m. local time. Dead insects and those that were perched either on the walls or at the base of the cage over time were grouped on a new level of treatment, named ‘others’.

The assay was performed in a randomized complete block design. Three independent experiments were carried out for each vector species with four replicates in the first experiment and three replicates in the second and third experiments. Plant positions were switched during the experiments.

Statistical analysis of settling preference assay

The response variable in the analysis was the average number of insects settled per experimental unit (plant). Since plants were observed over time, we have a longitudinal study and, consequently, successive counts are not independent. Therefore, this correlation structure should be included in the model. Although the number of insects per cage was fixed, the number of insects observed on each cultivar (treatment) was random, resulting in a Poisson process.

Thus, a Poisson model with random effect of block (cage) and an additional parameter to account for overdispersion was used to analyze settling preference (Appendix 1). The statistical analyses were made in R statistical computing environment (R Core Team 2019).

Evaluation of sap ingestion (confinement assay)

The ingestion rate was indirectly measured by collecting the liquid excreted by insects per unit of time (Marucci et al. 2005). For the evaluation of sap ingestion rates by B. xanthophis and S. sagata on healthy plum genotypes (‘SC7’, ‘SC13’ and ‘Laetitia’), three adults of each sharpshooter species were confined on branch sections using a cage similar to that described by Andersen et al. (1992). The cage was composed of a 50 ml transparent conic plastic tube (Falcon ™) (30 mm O.D.; 115 mm length) with sponge stopper, coupled in a 15 mL scaled conic plastic tube (Falcon ™) (17 mm O.D.; 120 mm length). These cages were positioned on leaves at the middle third of the plant, attached at an angle to allow the collection tube to remain hung vertically, in order to favor the honeydew flow to the scaled recipient. Each plant contained a single cage. This assay was performed in a climatized room at 25 ± 2 °C under fluorescent light (150 W, photophase of 14 h).

The experimental design was completely randomized, with seven replicates (seven plants with three insects per cage) for S. sagata and eight replicates (eight plants with three insects per cage) for B. xanthophis per cultivar (‘SC7’, ‘SC13’ and ‘Laetitia’). The volume of excreted honeydew (mL) in each replicate was recorded visually 24, 48 and 72 h after the sharpshooters were confined on the branches.

Statistical analysis of sap ingestion assay

A mixed-effects linear regression model was fitted to measure the associations between honeydew volume and predictor variables. Hence, honeydew volume was assumed to follow a normal distribution conditional on the random effect of cage, and fixed effects of cultivar and time. A parametric bootstrap approach with 10.000 pseudo-samples was used to compute the 95% simultaneous confidence intervals for all parameters in the model. Furthermore, the intraclass correlation coefficient under mixed-effects model assumptions was used to access similarity between honeydew volumes measured on each cultivar at different time periods (Appendix 2). The statistical analyses were made in R statistical computing environment (R Core Team 2019).

Probing behavior assay with B. xanthophis

The Electrical Penetration Graph (EPG) technique (McLean and Kinsey 1964; Tjallingii 1978; Backus and Bennett 2009) was used to monitor sharpshooter stylet and probing activities of B. xanthophis on three plum genotypes (‘SC7’, ‘SC13’ and ‘Laetitia’). The sharpshooter B. xanthophis was selected for this experiment, because a previous study of EPG waveform correlations with stylet activities is available for this species (Miranda et al. 2009).

Adult females were anesthetized with CO2 for 5–8 s, and then immobilized under a dissecting microscope by using a vacuum device, similar to that described by Van Helden and Tjallingii (2000). The tip of a 37-μm gold wire, 3 cm long was attached to the sharpshooter pronotum with a droplet of conductive silver paint (Ted Pella, Inc., Colloidal Silver Liquid, Redding, CA, USA). The opposite end of the gold wire was attached with a droplet of silver paint to a thin copper wire (3 cm long × 1 mm diameter), which was connected to the EPG head stage amplifier. Another copper electrode (10 cm long, 2 mm in diameter) was inserted into the soil of the plant container. After a 1-h starvation period, the wired sharpshooters were placed individually on the stem of a plant, inside a Faraday cage (for electrical noise isolation) under laboratory conditions (25 ± 1 °C; fluorescent light (80 W)).

The EPG waveforms were recorded and observed in real time on a computer screen, using an 8-DC EPG device, model Giga-8d, adjusted to 50x gain, +2 V to −3 V. EPG data acquisition was conducted using Stylet+ for Windows software (EPG Systems, Wageningen, Netherlands). Each sharpshooter was monitored for 6 h and a total of 15 individuals (replicates) were analyzed on each plum genotype (‘Laetitia’, ‘SC7’ and ‘SC13’).

EPG waveform types previously described for B. xanthophis (Miranda et al. 2009) and Philaenus spumarius L. (Hemiptera: Aphrophoridae) (Cornara et al. 2018) were identified and associated with stylet activities, as follows: np (non-probing period: stylets out of the plant); C (pathway through epidermis and parenchyma, including both stylet penetration and withdrawal); R (resting during xylem activity); N (interruption within xylem phase); Xc (xylem contact/pre-ingestion); and Xi (active intake of xylem sap).

Statistical analysis of EPG assay

Data from the 6 h recording period were analyzed in this research. The EPG sequential and non-sequential variables were manually calculated with an Excel Workbook. The following variables (mean ± SE) were calculated and compared between treatments, as previously described by Backus et al. (2007): NWEI, number of waveform events per insect; WDI, waveform duration (min) per insect, and WDE, waveform duration (min) per event.

The EPG data consisted of 24 response variables, of which 10 were discrete variables (counts) and 12 were continuous ones. To analyze count data, the Poisson model was a reasonable first choice. However, the residual deviance and generalized Pearson statistics were approximately equal to the residual degrees of freedom, indicating that the model may not fit the data well. The poor fit of the Poisson model in this study was due to under-dispersion or over-dispersion. To account for over-dispersion, a quasi-Poisson model was used, through a simple extension in the variance function of the Poisson model (Ver Hoef and Boveng 2007). To account for under-dispersion, a COM-Poisson model was used as described by Sellers and Shmueli (2010). For continuous variables, a generalized linear model using Gamma or Gaussian families were used (McCullagh et al. 1989). To test for the significance of treatments, we used likelihood-ratio tests for nested models. Goodness-of-fit was assessed via half-normal plots with simulation envelopes (Moral et al. 2017) (Appendix 3). All analyses were carried out using software R (R Core Team 2019).

Results

Settling preference

For both sharpshooter species (B. xanthophis and S. sagata), significant interaction effects between three choice experiments and plant genotype was observed at the 5% level of significance (Appendix 1- Table S2). Moreover, significant interactions between cultivar and time were verified only for S. sagata (P value = 0.015) (Appendix 1- Table S2).

Both sharpshooter species settled on all three plum genotypes (‘SC7’, ‘SC13’ and ‘Laetitia’). ‘SC7’ was less preferred for settling by B. xanthophis within experiments 1 and 3, whereas ‘Laetitia’ and ‘SC7’ were less preferred for settling in experiment 2. On the other hand, ‘SC13’ was the most preferred genotype by B. xanthophis in all choice tests (Table 1). For S. sagata, ‘SC7’ was less preferred for settling in experiments 1 and 2, while in the experiment 3 no differences were observed among cultivars (Table 1). However, ‘Laetitia’ was the genotype most preferred for settling in experiment 2.

Table 1 Overall mean number (±SE) of insects perched on treatments (plum genotypes) and on the walls or at the base of the cage (Other) within experiments for Bucephalogonia xanthophis and Sibovia sagata settling preference
Full size table

The number of S. sagata individuals perched on ‘Laetitia’, ‘SC7’ and ‘SC13’ did not change over time (Table 2). ‘SC7’ was less preferred than ‘SC13’ until 24 h, and equal to ‘Laetitia’ from the 6 h onwards (Table 2). From the 27th h, there was no difference in settling preferences among the tested genotypes (‘Laetitia’, ‘SC7‘and ‘SC13’ treatments) (Table 2).

Table 2 Mean number (±SE) of Sibovia sagata perched on treatments (plum genotypes) at successive periods after release in settling preference assays
Full size table

Evaluation of sap ingestion (confinement assay)

For S. sagata, a significant interaction effect between cultivar and time was observed (Appendix 2 - Table S2), whereas for B. xanthophis only time was significant (P value <0.0001) (Appendix 2 - Table S2).

For S. sagata, no significant difference was observed between the plum genotypes in the first evaluation period (24 h), but in further evaluations, the highest volume of excreted honeydew was recorded when the insects fed on ‘Laetitia’ (5.73 mL), differing statistically from ‘SC13’ (0.73 mL) (P < 0.0001) (Fig. 1 and Appendix 2 - Table S4). On the other hand, no statistical difference was observed for B. xanthophis in the evaluation periods (24, 28 and 72 h) between the three genotypes (Fig. 2).

Fig. 1
figure 1

Cumulative honeydew excretion (mL) for Sibovia sagata in three plum genotypes (‘Laetitia’, ‘SC7’, ‘SC13’) after 24, 48 and 72 h. The bars represent the standard error of the mean. The dots were shifted to the best view of the error bars

Full size image
Fig. 2
figure 2

Cumulative honeydew excretion (mL) for Bucephalogonia xanthophis in three plum genotypes (‘Laetitia’, ‘SC7’, ‘SC13’) after 24, 48 and 72 h. The bars represent the standard error of the mean. The dots were shifted to the best view of the error bars

Full size image

Probing behavior of B. xanthophis

The analysis of the 6-h EPG recordings of B. xanthophis on the three plum genotypes showed significant differences for the number (NWEI) of Xi (xylem ingestion), the total waveform duration per insect (WDI) of C (pathway phase) and the waveform duration per event (WDE) of np (non-probing), C, R (resting during xylem activity), N (interruption of sustained ingestion during xylem phase) and Xi.

B. xanthophis performed more Xi when held on the ‘SC13’ genotype (11.87 ± 2.17) than on ‘Laetitia’ (6.20 ± 1.53) or ‘SC7’ genotypes (6.60 ± 0.92) (Table 3). Besides that, the WDI of C was longer for insect held on ‘SC13’ (483.38 ± 149.54) than on ‘SC7’ (204.88 ± 34.06). However, the analysis did not show differences between ‘Laetitia’ (346.91 ± 63.27), ‘SC13’, and ‘SC7’ genotypes (Table 3). The WDE of np was shorter for insects held on the cultivar ‘Laetitia’ (5.12 ± 1.57), differing statistically from the ‘SC13’ genotype (9.58 ± 1.99). The WDE of C was shorter for insects held on the ‘SC7’ genotype (12.44 ± 1.55), distinguishing their response from insects held on the other genotypes. The WDE of R was longer for insects held on the ‘SC7’ than on the ‘SC13’ genotype, but did not differ from the WDE of R for insects held on the cultivar ‘Laetitia’ (14.83 ± 1.55). The WDE of waveform N was significantly shorter for insects held on ‘Laetitia’ (7.01 ± 0.69) than on ‘SC7’ (9.54 ± 0.86). In addition, the WDE of Xi for insects held on ‘SC13’ (6.11 ± 0.89) was approximately two times shorter than the WDE of Xi for insects held on ‘SC7’ (10.89 ± 1.88), and three times shorter than the WDE of Xi for insects held on ‘Laetitia’ (16.98 ± 2.98) (Table 3). No significant differences were found among the three plum genotypes for the other non-sequential and sequential variables analyzed (Table 3 and Table 4).

Table 3 Mean non-sequential EPG (± SE) variable values for the probing behavior of Bucephalogonia xanthophis on three plum genotypes (treatment) during a 6-h recording. Time is expressed in minutes, except N expressed in seconds
Full size table
Table 4 Mean sequential EPG variables (± SE) values for the probing behavior of Bucephalogonia xanthophis on three plum genotypes (treatments) during a 6-h recording
Full size table

Discussion

In this work, we demonstrated that two sharpshooter vectors commonly found in plum orchards, B. xanthophis and S. sagata (Azevedo Filho et al. 2016), show variations in settling and feeding preferences for three plum genotypes (‘SC7’, ‘SC13’ and ‘Laetitia’) with different susceptibility levels to PLS. Both vector species showed low preference for settling on the resistant ‘SC7’ plum genotype in at least two out of three choice tests, compared to ‘Laetitia’, and especially to ‘SC13’ over the first 24 h. Honeydew excretion assays showed reduced feeding (lower xylem sap ingestion rates) of S. sagata on the resistant ‘SC13’ compared to ‘Laetitia’, although no statistical differences were found among the plum genotypes for B. xanthophis. EPG parameters indicated that ‘SC13’ also affected stylet activities of B. xanthophis in the xylem, since the sharpshooter performed a large number of waveform Xi (associated with xylem sap ingestion) but the mean duration of each Xi event was shorter than on the other genotypes. In some cases, Xc (stylet contact with xylem tissue) or Xi waveforms were performed just for a short period on plants of ‘SC13’, possibly because of xylem sap or vessel unsuitability, which affects the sustained sap ingestion by the vector (Miranda et al. 2009). In these situations, the sharpshooter starts a new pathway phase or withdraws its stylets from the plant (Miranda et al. 2009; Sandanayaka et al. 2017).

Parameters of plant acceptability should consider the duration of the stylet penetration phase (pathway phase) and duration of sustained ingestion periods (Tjallingii 1993). On suitable hosts, xylem sap-feeding insects usually spend less time in non-probing and stylet pathway phases and longer periods ingesting xylem sap (Sandanayaka et al. 2017). According to our EPG data, the duration of the stylet pathway phase (represented by waveform C) of B. xanthophis was longer on plants of the plum genotype ‘SC13’ than on plants of ‘SC7’, which indicates lower acceptability of the former genotype. In addition, the mean duration of non-probing (np) events was longer on ‘SC13’ than on ‘Laetitia’, reinforcing the lower acceptability of ‘SC13’. Although not statistically different, the mean time spent by the insects to start the first xylem sap ingestion event after the beginning of EPG was more than twice as long on ‘SC13’ (178 min) than on ‘Laetitia’ (80 min).

Piercing-sucking insects use a variety of chemical and visual cues during the host plant selection process (Bullas-Appleton et al. 2004; Fereres and Moreno 2009). Upon landing on a plant guided by visual and olfactory cues, this group of insects insert their stylets into the plant tissue and ingest cellular contents or sap, which are analyzed by gustatory sensilla located in the foregut (Backus and McLean 1985). If the plant is suitable, the insect usually remains on it for feeding and reproduction. Vector preferences for settling and feeding influence the population density of individuals and their residence time on vegetal species. Therefore, high settling and feeding preferences of vectors for hosts of X. fastidiosa should increase the probability of infection (Daugherty et al. 2009; Sandanayaka et al. 2013; Markheiser et al. 2019).

Based on the results obtained here and on the field observations made by Dalbó et al. (2018), it can be inferred that an antixenosis resistance mechanism against the sharpshooter vectors might be involved in the field resistance of ‘SC7’ and ‘SC13’ to PLS. For ‘SC7’, the antixenosis effect could be linked to visual and olfactory (volatile compounds) cues that mediate host searching and recognition by the insects (Lara 1991), making this genotype less preferred for sharpshooter landing and settling. On the other hand, the EPG and honeydew excretion results related to xylem feeding suggest that the resistance of the ‘SC13’ genotype occurs at the level of xylem chemistry or physiology, by reducing the duration of xylem sap ingestion (Xi) events and the volume of excreted honeydew (indirect measure of ingested sap). Considering that longer feeding periods increase the probability of transmission of X. fastidiosa (Almeida and Purcell 2003; Cornara et al. 2016), planting new orchards with resistant plum genotypes that affect vector settling and/or xylem sap feeding should have implications on leaf scald epidemiology by reducing pathogen spread.

Collectively, the results of the present study support the hypothesis that the field resistance of ‘SC7’ and ‘SC13’ genotypes against PLS may be associated with interference of sharpshooter vector settling and feeding behavior. These resistant genotypes can be recommended in PLS management program, because they produce very attractive fruits (Dalbó et al. 2018). Research is currently underway to identify volatile compounds that may be mediating the antixenosis effects of ‘SC7’ on the vectors (H. T. Kleina in review). Because X. fastidiosa is a xylem-limited bacterium (Purcell and Finlay 1979), the morphology of the leaves, the xylem structure and sap composition in resistant and susceptible plants also may play an important role in sharpshooter settling and feeding behavior and need to be addressed in future investigation. Finally, transmission experiments using potential vectors and artificial inoculation approaches (e.g. needle inoculation) should be carried out with susceptible and resistant cultivars to check if there is any level of resistance against bacterial infection and multiplication in plum genotypes that are unaffected by PLS in the field.