Fin-Search for potential vectors of ‘ Candidatus Liberibacter solanacearum’: population dynamics in host crops

‘ Candidatus Liberibacter solanacearum’ has recently been reported to be associated with vegetative disorders and economic losses in carrot and celery crops in Spain. The bacterium is a carrot seedborne pathogen and it is transmitted by psyllid vector species. From 2011 to 2014 seasonal and occasional surveys in carrot, celery and potato plots were performed. The sticky plant method was used to monitor the arthropods that visited the plants. The collected arthropods were classified into Aphididae and Cicadellidae, and the superfamily Psylloidea was identified to the species level. The superfamily Psylloidea represented 35.45% of the total arthropods captured on celery in Villena and 99.1% on carrot in Tenerife (Canary Islands). The maximum flight of psyllid species was in summer, both in mainland Spain and the Canary Islands, reaching a peak of 570 specimens in August in Villena and 6,063 in July in Tenerife. The main identified psyllid species were as follows: Bactericera trigonica Hodkinson, B. tremblayi Wagner and B. nigricornis Förster. B. trigonica represented more than 99% of the psyllids captured in the Canary Islands and 75% and 38% in 2011 and 2012 in Villena, respectively. In addition, Trioza urticae Linnaeus , Bactericera sp. , Ctenarytaina sp., Cacopsylla sp., Trioza sp. and Psylla sp. were captured. ‘ Ca. L. solanacearum’ targets were detected by squash real-time PCR in 19.5% of the psyllids belonging to the different Bactericera species. This paper reports at least three new psyllid species that carry the bacterium and can be considered as potential vectors.


Introduction
'Candidatus Liberibacter solanacearum' , which is also known as 'Ca. Liberibacter psyllaurous' (Hansen et al., 2008) (Bacteria: Proteobacteria: Alphaproteobacteria: Rhizobiales: Rhizobiaceae), is a Gram-negative bacterium restricted to plant phloem and insect hemolymph that currently cannot be cultured in vitro . Five 'Ca. L. solanacearum' haplotypes (designated A, B, C, D and E) have been described to affect several crops worldwide. Haplotype A has been found from Central to North America and New Zealand, haplotype B has been found in Mexico and the United States of America, haplotype C is present in Finland, Sweden and Norway, 2 detected in eggs, different nymph instars stages and adults of B. cockerelli (Hansen et al., 2008). The bacterium was also detected by conventional PCR in field-collected and laboratory-reared T. apicalis in southern Finland (Munyaneza et al., 2010b). Squash real-time PCR is a useful tool for the detection of nucleic acid targets in insect vectors and was successfully used to detect 'Ca. L. americanus' and 'Ca. L. asiaticus' in D. citri specimens . The squashing of individual psyllids on membranes is a direct method of sample preparation in which neither extract preparation nor nucleic acid purification is necessary. The main drawback of these systems based on target immobilisation is the small amount of sample that can be loaded onto the support. This limitation is overcome by coupling these preparation methods with highly sensitive techniques such as realtime PCR (De Boer & López, 2012). In addition, the immobilisation of targets on paper is simpler and much faster than DNA extractions and can be used with quarantine pathogens without risks. The presence of DNA targets in individual psyllids can be accessed from fresh and previously captured individuals stored in alcohol and/ or squashed on paper (Marroquín et al., 2004).
The knowledge of arthropod species that visit crops and the seasonal fluctuations of their populations is basic for the identification of putative vector species and the development of control strategies (Cambra et al., 2006). Thus, the main goal of this study was to evaluate the psyllid species that landed on carrot, celery and potato crops, with high prevalence of 'Ca. L. solanacearum', grown in different Spanish regions to determine the population dynamics. We also investigated whether the bacterium was associated to any of the different captured psyllid species.

Hosts and monitoring sites
Seasonal surveys were carried out in carrot and celery crops at Villena (Alicante) and Tenerife (Canary Islands) during different growing seasons between 2011 and 2012. Occasionally, carrot crops were also surveyed in the La Rioja region (Santo Domingo de la Calzada), and these surveys were extended to potato in Tenerife and Valencia from 2012 to 2014. The cultivars and monitored crops were as follows: Loretta and Golden var. dulce (Mill.) Pers. in celery plots grown in Villena; cv. Maestro in Villena and cv. Bangor in La Rioja in carrot plots; cv. Vivaldi in Valencia in a potato plot; cv. Bangor in carrot plots and cv. Slaney in Tenerife in a potato plot (Table 1).
land (Munyaneza et al., 2010a) and Spain (Alfaro-Fernández et al., 2012a) to be associated with vegetative disorders in carrot causing leaf curling, yellow and purple discoloration of leaves, stunted growth of shoots and roots and the proliferation of secondary roots. More recently, the bacterium was also associated in celery with symptoms such as an abnormal amount of shoots, curling of stems and yellowing (Teresani et al., 2014), producing relevant yield reductions and important economic losses to the carrot and celery industries.
The bacterium is primary transmitted by carrot seeds (Bertolini et al., 2015) and is afterwards transmitted by different psyllid species in a persistent way. Bactericera cockerelli Sulc is the vector of haplotypes A and B in solanaceous crops (Nelson et al., 2011). Trioza apicalis Förster was first described in carrot in Finland (Munyaneza et al., 2010b) transmitting the haplotype C (Nelson et al., 2011) and B. trigonica Hodkinson is associated with the transmission of 'Ca. L solanacearum' haplotype D in carrot and likely with haplotype E in carrot and celery in Spain (Alfaro-Fernández et al., 2012b;Nelson et al., 2012;Teresani et al., 2014). In another 'Ca. Liberibacter' associated disease, Diaphorina citri Kumayama, Trioza erytreae Del Guercio and Cacopsylla citrisuga Yang & Li have also been described as 'Ca. Liberibacter' spp. vectors of huanglongbing (HLB), an important citrus disease (McLean & Oberholzer, 1965;Capoor et al., 1967;Bové, 2006;Cen et al., 2012). In addition, there are reports of 'Ca. L. solanacearum' detection in nonidentified species of psyllids in the genera Acizzia and Trioza collected in New Zealand (Munyaneza, 2012).
Several trapping methods have been used in surveys to determine the arthropod species that are present on crops or visit crops to establish the population dynamics. These methods include the observation of established colonies, suction, water and sticky fishing-line traps as well as the sticky plant or shoot method (Cambra et al., 2006). The sticky shoot method, which uses glue-covered bait leaves, shoots or the entire plant, has been extensively used to determine the arthropod species that visit crops and monitor aphid species in adult trees (Cambra et al., 2000;Marroquín et al., 2004;Vidal et al., 2012). This method is the most efficient for estimating the numbers of insects landing on the plants (Hermoso de Mendoza et al., 1998). In Spain, arthropods occurring in carrot and celery crops have been previously studied using Moericke´s yellow traps (Villaescusa et al., 2011), but the captured arthropods were not identified at the species level. No data of psyllid species that visit potato crops have been available until now.
Conventional and real-time polymerase chain reaction (PCR) protocols have been described for 'Ca. L. solanacearum' detection in plant material and insect vectors. Using conventional PCR 'Ca. L. solanacearum' was 3 Potential 'Ca. L. solanacearum' vectors veys. Monitoring was conducted sporadically during the occasional surveys.
For the seasonal surveys, the same 20 celery or 50 carrot plants were randomly selected in the plot and nondestructively sampled at weekly intervals. The whole plant was initially sprayed with glue (Souverode aerosol, Scotts, France) however, as the plants grew larger (4 weeks) only 1-2 fully developed leaves were sprayed. The sprayed leaves were detached after a week, and the new sticky leaves were prepared. The removed sticky leaves with arthropods stuck on the surface were placed in turpentine to dissolve the glue, and collected specimens were washed in soapy water to remove the solvent (Marroquín et al., 2004). The collected arthropods were kept in 70% alcohol for later counting and identification.
For the occasional surveys, a sampling site was randomly selected within the field, and 10 consecutive plants were monitored. One to two fully developed leaves from each plant were sprayed every 10 days. The attached arthropods were treated as previously described. The carrot surveys in La Rioja (2012 and 2013) and Villena (2012) were performed in autumn (September to November), whereas the potato survey in Valencia was performed in spring (March to May) and in Tenerife in spring-summer (May to July).

Identification of arthropods
Aphids, leafhoppers and psyllids were kept in alcohol, and the other arthropods were discarded. The se-A total of 16 commercial plots located in different regions where 'Ca. L. solanacearum' prevalence was high were selected since 2011 to 2014. For the seasonal monitoring of arthropods that visit celery plants in 2011, three plots of approximately 1 ha were selected in Villena, representing each celery cycle of cultivation. In 2012, another six celery fields of 1 ha were also seasonally monitored in Villena, one during the early cycle, two in the meddle cycle and three in the late cycle in an attempt to cover the possible differences between plots over time. Two fields (1.3 and 0.4 ha, respectively) were also se-  Table 1).

Monitoring of arthropods
Arthropods monitoring was focused on Hemiptera belonging to Cicadellidae and Aphididae and the superfamily Psylloidea. The sampling for the seasonal surveys was performed weekly since the emergence of the plants until harvest during all the cycles of cultivation. It was done sporadically during occasional sur- positive when an exponential amplification curve occurred and the Ct value was below 45.

Arthropod monitoring
A total of 18,751 arthropods were captured during the seasonal surveys. From this total, 2,695 arthropods were caught on celery in 2011 in Villena. The superfamily Psylloidea (1,373 individuals) and the family Aphididae (762) were the most frequently found, followed by Cicadellidae (560) ( Table 2). In 2011, the higher numbers of captured arthropods were observed in the middle cycle of cultivation (June 3 rd to September 12 th ). A total of 1,533 specimens were caught in 2012 on celery. Cicadellidae was the most frequently found family in the sticky plants (924 specimens), followed by Aphididae (483) and Psylloidea (126) ( Table 2). In 2012, a low number of psyllid species were caught in comparison with 2011, and no remarkable peaks of the populations of any the species were observed.
A total of 14,523 arthropods were caught on carrot crops in Tenerife. The superfamily Psylloidea was the predominant (14,401 specimens) followed by Aphididae (62) and Cicadellidae (60) ( Table 2).

Psyllid species composition
The overall numbers of captured psyllid species (15,900 individuals) in the seasonal surveys are shown lected families were counted by date of capture and then identified. Only the superfamily Psylloidea was identified to the species level due to the important role they may play in the transmission of 'Ca. L. solanacearum'.

Detection of 'Ca. L. solanacearum' targets in psyllid species
The identified individual psyllid specimens were squashed on Whatman 3MM membranes (GE Healthcare Europe) using the round bottom of an Eppendorf tube until the complete disruption of the insect (Olmos et al., 1996;Bertolini et al., 2014). The membrane containing squashed psyllids was carefully cut around the sample (~ 0.5 cm 2 ) and inserted into Eppendorf tubes. DNA was released from the piece of the membrane by adding 100 μL of distilled sterile water and vortexed. Three μl were analysed by specific 'Ca. L. solanacearum' real-time PCR according to Teresani et al. (2014) using a CaLsol/100 kit (Plant Print Diagnòstics, Valencia, Spain). Positive and negative controls (5 μL of crude extract of infected and healthy plant material spotted on a piece of membrane, respectively) and PCR reagents were used. Psyllids were considered A total of 811 psyllids were caught in the occasional surveys on carrot and potato from 2012 to 2014 (Table 4) Table 3. The population dynamics of psyllid species captured in celery plots in Villena (2011) are presented in Figure 1A. A total of 1,499 psyllids were caught on celery in Villena, 1,373 in 2011 and 126 in 2012. In 2011, the predominant species were as follows: B. trigonica (1,085 specimens) followed by B. tremblayi Wagner (225) and B. nigricornis Förster (2). Sixty-one specimens from other psyllid species (Bactericera sp., Trioza sp. and Psylla sp.) were also caught in 2011. Two maximum population peaks of B. trigonica were observed during the summer of 2011. The most important peak occurred from August 10 th to 17 th , with 570 specimens captured. The second peak occurred from August 24 th to 30 th , with 293 specimens captured (Figure 1A). The population dynamics of psyllid species captured on celery plots in Villena (2012) are presented in Figure 1B. The two species found in the second year of seasonal surveys were B. trigonica (48) and B. tremblayi (46). Thirty-two specimens from other non-identified psyllid species (most likely Bactericera sp.) were also captured. B. nigricornis was not found on celery plants in the Villena area in 2012.
A total of 14,401 psyllids were caught on carrot fields in Tenerife, with B. trigonica as the dominant psyllid species, followed by Bactericera sp. and Trioza urticae Linnaeus ( Figure 1C). In the middle cycle of cultivation, 14,255 B. trigonica and seven specimens from other psyllid species (Bactericera sp.) were caught. In the late cycle, a lower number of psyllids than in the previous cycle was found: 119 B. trigonica and 20 specimens of T. urticae, Ctenarytaina sp. and Cacopsylla sp. (other species in Table 3) were also identified in Tenerife. Table 3. Different psyllid species captured in seasonal surveys on sticky celery and carrot plants in 2011 and 2012. Estimation of the number of specimens carrying the bacterium was determined by real-time polymerase chain reaction (PCR).

C)
Tenerife were analysed (  -Fernández et al., 2012b). In 19.5% of the psyllid species tested, 'Ca. L. solanacearum' targets were amplified, suggesting the high prevalence of the bacterium in the celery plants grown in the monitored areas. This fact could justify the higher prevalence of symptoms and the important crop losses in the middle and late cycles of cultivation. In 2012, the most frequent visitors were included in the families Cicadellidae, followed by Aphididae and the superfamily Psylloidea. In this year, although there were two celery plots sampled in the middle cycle and three in the late cycle of cultivation, the prevalence of arthropods was very low compared with the previous year. Consequently, a lower prevalence of symptoms in celery was observed, whereas the presence of symptomatic carrots was similar to the previous years. This was likely, due to the primary infection caused by carrot seed transmission that was recently demonstrated (Bertolini et al., 2015). In both of the surveyed years, the arthropods collected on sticky celery plants showed the same psyllid species structure: B. trigonica was always the prevalent species and B. nigricornis was the less frequently observed species, with only two caught specimens in 2011. The same psyllid species are visiting early potato crops in Valencia, representing a threat for the crop if non-solanaceaous haplotypes are able to colonise potato plants. In addition, non-identified Trioza sp., Psylla sp. and Bactericera sp. (but not B. cockerelli and T. apicalis), were found on carrot, celery and potato in mainland Spain. specimens out of seven Bactericera sp., positive amplification was observed. In the late cycle of carrot cultivation, 38 out of 119 B. trigonica were positive against the bacterium. None of the 20 specimens from the other psyllid species tested positive (Table 3) (Table 4).

Discussion
The knowledge of the seasonal dynamics and abundance of arthropods in crops is key to determine the species responsible for the natural spread of 'Ca. L. solanacearum'. In addition, due to little or no available information on the arthropod species that lands on economically important crops, is necessary to identify the species that visit the potential hosts of the bacterium in different ecological areas in Spain. This is a basic foundation that is necessary to design strategies to mitigate the natural spread of the bacterium.
During the seasonal arthropod surveys performed in celery crops in Villena in 2011 and 2012, 35.4% of the catches belonged to the superfamily Psylloidea although important differences were observed between both years in the number of specimens and the preva- a 'Candidatus Liberibacter solanacearum' positive psyllid specimens/total of analysed psyllid specimens. b Other indicates non-identified specimens, most likely B. trigonica, that had key morphological characteristics damage that was caused during the process of recovery from the sticky plant. 8 nomically important crops: D. citri, T. erytreae and C. citrisuga in citrus (Bové, 2006;Cen et al., 2012), B. cockerelli in potato and tomato (Hansen et al., 2008) and T. apicalis in carrot (Nissinen et al., 2014). Moreover, recent studies suggest that Liberibacter species may be more widespread than previously thought, and vector species play an important role in bacterial spread. In addition, psyllids such as B. cockerelli and T. apicalis are able to transmit the same bacterium in distant and different geographical areas, highlighting the importance of the local identification of psyllid species and putative vectors. For this reason, we focused our interest in psyllid species. Bertolini et al. (2015) reported that the presence of B. trigonica and other species was correlated with an increase in the prevalence of the bacterium from 2% to approximately 100% after six months of carrot cultivation. Other experiences with emerging diseases in the citrus and potato industries suggest that psyllid species, feeding briefly outside their normal plant host range, could introduce a pathogen to another crop (Nelson et al., 2013). This might present a serious threat for other economically important crops, such as potato, tomato and aubergine which could naturally be infected by 'Ca. L. solanacearum' if psyllid species carrying the bacterium have the opportunity to reach the phloem of a potential host species. Villaescusa et al. (2011) reported, using Moericke´s yellow traps, the arthropods occurring in the ambience of carrot and celery plots in Spain; psyllids represented 92.4% of the total number of captured arthropods. In our case, using sticky plants, psyllids landing on the plants represented 50.9% in 2011 and 8.2% in 2012 of the total number of the arthropods. Our data are essentially in agreement with the structure of the species found in the previous study, where Bactericera spp. (85% of the total psyllid caught), Cacopsylla sp. and Trioza sp. were caught. Althought the authors did not identify the Bactericera spp. found, it is likely that B. trigonica, B. tremblayi and B. nigricornis were already present in the Villena area at that time.
To make decisions regarding integrated disease management strategies, it is essential to identify the psyllid species that land on a particular crop and estimate the abundance of the different species and the percentage of specimens carrying the bacterium. In this context, the use of appropriate methodology is crucial. The use of the sticky host plants make it possible to more accurately determine the species that actually land on the plants. In addition, the squash protocol described by Bertolini et al. (2014) combined with real-time PCR described by Teresani et al. (2014) has demonstrated their potential for the detection of 'Ca. L. solanacearum' targets in psyllids. The squash procedure In total, 99.1% of the 14,401 captured insects on carrot during seasonal surveys performed in Tenerife in 2012 were psyllids. This high number is typically found in the middle cycle of cultivation, with a maximum peak of 6,063 specimens caught in summer. Almost all the psyllid species captured were B. trigonica, ranging from 99.9% in the middle cycle to 85% in the late cycle of carrot cultivation. T. urticae, Bactericera sp., Cacopsylla sp. and Ctenarytaina sp. were the other psyllid species captured in the monitored plots, showing a high population number and diversity of species in the Canary Islands. The population dynamics of psyllids in the potato plots in Tenerife were in agreement with the dynamics observed in carrots. The comparison of the number of psyllid species caught on celery in mainland Spain and on carrot in the Canary Islands suggests that carrot is the preferential host for the species found in the monitored areas. In fact, in Villena, where carrot and celery are grown in the vicinity, a higher prevalence of psyllid species was found in carrot than in celery (data not shown).
The Spanish mainland climate varies across the peninsula among the three main climatic zones, which can be distinguished according to the geographical location and orographic conditions. The typical Mediterranean climate is represented by Valencia, the continental Mediterranean climate by Villena and the last zone, with some oceanic characteristics is represented by the climate of La Rioja. The subtropical climate is the predominant climate in the surveyed carrot and potato areas in the Canary Islands. Dixon (1985) reported that the population dynamics of aphid species can vary depending on the year and other factors, such as whether conditions, natural enemies, abundance in previous years or human actions. Some of these factors could explain the variation in the number of arthropod species caught among the different years and cycles of cultivation in different areas. In fact, the rainfall in 2012 (352.4 mm) was approximately twice than that in 2011 (181.7 mm) in Villena. In Tenerife, the late cycle of cultivation occurs in autumn when the temperature is lower than in the previous cycles and the rainfall is more constant (data not shown). The mentioned factors could also justify why the population structure changed in the carrot crops in La Rioja between 2012 (B. trigonica was not captured and B. nigricornis was present) and 2013 (B. trigonica was the prevalent species and B. nigricornis was not captured). However, these factors did not affect B. tremblayi which maintained similar populations in both years.
The surveys were focused on arthropod families in which several species are described as vectors of plant pathogens. Psyllid species are cited as efficient vectors of the fastidious bacterium 'Ca. Liberibacter' in eco-9 Potential 'Ca. L. solanacearum' vectors country side (Villena) and the Canary Islands (Tenerife), representing a broad spectrum of climatic conditions.
Currently, there are no effective control strategies for plant protection against natural 'Ca. Liberibacter' infection, except the potential use of cultivation under insect-proof facilities. The "three-pronged system" (TPS) (Belasque et al., 2010) is used for HLB management in perennial plants could be adjusted for horticultural crops. This system involves the removal of inoculum sources, the replacement of infected trees with healthy trees and insecticide treatments aimed to reduce psyllid vector populations to mitigate the spread of the disease. The use of 'Ca. L. solanacearum'-free carrot seed lots could be complemented with an accurate and timely detection of visitor psyllid species that may serve as a vector of the pathogen and with subsequent treatments to prevent transmission of the bacterium. The reduction of the psyllid population is critical independent of the efficiency of the transmission of the different vectors involved. Any vector species could play a role in the bacterium spread by compensating with their abundance poor transmission efficiencies.
This paper provides information about the psyllid species population that lands on celery, carrot and potato plants in Spain and reports, for the first time, B. tremblayi, B. nigricornis and Bactericera sp. as 'Ca. L. solanacearum' carriers and potential vectors of the bacterium. However, experimental transmission assays are necessary to assess the vector ability of the psyllid species that have not been previously described as vectors of 'Ca. L. solanacearum'. and subsequent detection by PCR-based methods yielded similar results using fresh or those preserved in alcohol for the detection of the viral targets (Marroquín et al., 2004). We assumed the non-effect of the treatment to remove the psyllids stuck on the plant and the preservation in alcohol. In fact, we were able to detect amplifiable 'Ca. L. solanacearum' targets in 95.2% of the captured psyllids in La Rioja in 2012 using this methodology. The use of this technique allowed the estimation of the percentage of psyllids carrying the bacterium that could transmit it if given the opportunity to feed on the phloem of a host species. Although we have not yet performed transmission trials, the detection of targets is a strong indication that the squashed arthropods are harboring 'Ca. L. solanacearum' acquired from infected plants; it is likely that it has multiplied in the insect to become detectable by real-time PCR. These facts represent the risk of bacterium transmission and disease spread.
B. tremblayi, B. nigricornis and B. trigonica are morphologically close psyllid species that belongs to the 'Bactericera nigricornis Förster group' (Hodkinson, 1981). They have polyphagous habits and show overlapping areas of distribution. These species were formally reported in Bosnia-Herzegovina, France, Greece, Iran, Italy, Serbia, Switzerland and Turkey (Ouvrard & Burckhardt, 2012). Here, we report the presence of B. tremblayi and B. nigricornis in Spain in addition to B. trigonica, which was already reported by Alfaro-Fernández et al. (2012b) and is widely distributed in the Mediterranean region (Haapalainen, 2014). 'B. nigricornis group' is composed of multivoltine species (Hodkinson, 2009), that feed on a variety of herbaceous plants, including beet, cabbage, carrot, onion, parsley or potato (Burckhardt & Lauterer, 1997), which are known hosts or potential hosts of 'Ca. L. solanacearum'. Adults have also been recorded to overwinter on conifers (Reuter, 1908). This level of polyphagy is exceptional in Psylloidea, which are usually host specific (Hodkinson, 1974).
B. trigonica, B. tremblayi and B. nigricornis were found to be current visitors of the surveyed crops in continental Spain. In the Canary Islands, B. trigonica is the predominant species, which is in agreement with previous reports (Font et al., 2010;Alfaro-Fernández et al., 2012a). All these visitors, which carry 'Ca. L. solanacearum', are potential vectors of the bacterium in different ecological areas. Only B. tremblayi is associated with Mediterranean climates; however, B. trigonica and B. nigricornis, which are also found in these regions, are associated with most temperate climates. This fact could suggest the climate adaptation of these species, which can be found in the north (La Rioja), Mediterranean coast (Valencia), continental