X. fastidiosa is a strictly aerobic, non-motile bacterium that can survive and proliferate within the xylematic tissue of 560 plant species of 80 plant families ^[1][5]. It is displaced between the host plants through the xylem-sap feeding activity of some polyphagous insect vectors belonging to “Hemiptera”, suborder “Homoptera”; “Aphrophoridae”, “Cicadellidae”, and “Cercopidae” families ^[2][3][6,7]. Even though it is considered a benign commensal in most of its hosts, in some circumstances, this bacterium colonizes crop and ornamental plant species, causing relevant economic losses in North, Central, and South America as well as in southern Europe ^[4][8]. Generally, its main characteristic is the incapability to freely live outside the plant or the insect.

However, there is putative evidence that X. fastidiosa inhabited unexpected ecological niches of the Earth many millions of years before the appearance of plants and insects. With the aim of reconstructing the evolution of Prokaryotes, through a genomic timescale sequencing of different prokaryotic taxa strains, Battistuzzi et al. ^[5][9] estimated that X. fastidiosa originated about 600 million years ago, about 500 million years before the most closely related taxon (i.e., Xanthomonas) ^[5][9]. Concerning plant evolution, the transition from water to land and the origin of plant vascular tissue date back to around 470 million years ago ^[6][10] when the early terrestrial flora starts to diversify by colonizing different ecosystems ^[7][11], facilitated by symbiotic fungi ^[8][12]. Similarly, the early insects evolved about 479 million years ago, the flying insects about 407 million years ago ^[9][13], and the Hemiptera, to which the currently known X. fastidiosa vectors belong, about 160–100 million years ago ^[10][14]. If the genomic timescale reconstruction of Prokaryotes is correct, this microorganism apparently survived for more than 100 million years in a completely different environment than the current one and in the absence of land plants and insects. Moreover, its currently known insect vectors appeared after about 440–500 million years from its origin.

The most primitive plants were the green algae (i.e., photosynthetic eukaryiotic organisms) that originated about 1,500 million years ago ^[11][12][15,16]. For about 1000 million years these unicellular or pluricellular algae lived in marine environments as the sole representatives of the Plant Kingdom. It is known that green algae and bacteria live together in complex microbial communities by performing physiological activities that would not be possible in the absence of such partnerships ^[13][17]. Whether X. fastidiosa was able to establish relationships with such primitive plants in the absence of the vascular plants or insect vectors is not known. However, it should be said that this bacterium, through horizontal gene transfer, has obtained a relevant portion of its genome (i.e., about 20%) from distantly related bacteria ^[14][18]. Generally, this evolutionary “old age” could explain the recognized versatility of this microorganism in establishing relationships with so many plant species, together with its ability to rapidly colonize the xylem of the host plants with few cells without blocking the flow of the xylem sap ^[15][19] and to move against the xylem flow ^[16][20].

There is evidence that ornamental coffee (Coffea arabica L. and C. canephora Pierre ex Frohen) plants that originated from Central America and were shipped to Europe hosted cells of X. fastidiosa. After 2013, the subspecies fastidiosa was intercepted in France from C. canephora imported from Mexico, whereas the subspecies pauca was found in C. arabica plants from Ecuador ^[17][21]. Similarly, X. f. subsp. fastidiosa strains were intercepted both in the Netherlands ^[18][22] and in three different localities of northern Italy ^[19][23] from the same lot of C. arabica plants introduced from Costa Rica. Indirect but striking evidence for an introduction in Salento (Apulia, Italy) of X. f. subsp. pauca from Central America through ornamental coffee plants has been obtained through genomic sequence comparison. The molecular analyses and comparison of multiple strains of X. f. subsp. pauca isolated in Central and South America from different host plants with the strain isolated in Italy from an olive tree showing symptoms of OQDS revealed a very high similarity between this strain and strains isolated from coffee in Costa Rica, indicating a strict genetic relationship between them ^[20][21][24,25]. Sequence type (ST) 53 was always found to be associated with the olive and oleander decline, allowing for the further characterization of the pathogen ^[22][26].

These findings were also supported by additional analyses that found C. arabica plants imported from Costa Rica and Honduras as a reservoir for X. f. subsp. pauca strains ^[23][27]. The study also ascertained that the sequence type found in these plants very often belongs to the same ST associated with OQDS in Salento, namely ST53 ^[23][27]. The same ST was also found in other plant species such as Prunus dulcis, Prunus avium, Westringia fruticosa, Polygala myrtifolia, and Catharanthus roseus, which have grown close to the infected olive groves ^[22][26]. The genome sequencing of three ST53 strains isolated from infected olive trees revealed a very high similarity among these strains and with the related ones in Central America, suggesting that a single introduction of infected plant material took place and pointing to a “founder effect” for the X. f. subsp. pauca population ^[24][28].

Moreover, through a population genomic approach, based on the sequencing and comparison of multiple X. f. subsp. pauca strains isolated in Central America and in Salento from olive trees showing symptoms of OQDS, it has been assumed that the introduction of the bacterium, through infected coffee plants from Central America, most likely occurred in 2008 ^[25][29]. It should be added that X. f. subsp. pauca, including ST53, has been reported on coffee in Costa Rica during recent years ^[26][27][30,31] and that the export of plants from this country to different continents, including Europe, in recent decades is quite relevant ^[28][32], with about 43 million exported ornamental plants in 2012 alone (https://www.focus.it/ambiente/natura/xylella-il-batterio-killer-degli-olivi, accessed on 3 September 2022). Moreover, since 2000, a relevant import from Central America to Europe of potential ornamental host plants for the bacterium has been observed ^[29][33]. These data imply that X. f. subsp. pauca ST53 was most likely introduced from Central America, through already infected but not necessarily symptomatic ornamental coffee plants, into some nursery(ies) of Salento, where it subsequently infected the nearby olive trees.

Owing to the relevant import of coffee plants from Central America that occurred during the first decade of the new millennium in Italy and the frequent interception in Europe of X. f. subsp. pauca ST53 in coffee plants originating from that area ^[23][27], one corollary facet of this plausible reconstruction on the pathogen introduction in Salento concerns the absence of OQDS in other areas of Apulia, in nearby Mediterranean regions such as Basilicata, Calabria, and Sicily, and in other regions of Central Italy where olive is also cultivated. Considering that insect vectors are present in these areas ^[30][31][32][34,35,36], this situation could be due to the absence of the pathogen in the imported coffee plants or to intrinsic factors that characterize the olive groves of Salento, such as cultivar susceptibility and/or other predisposing factors ^[33][37], which will be taken into account in an epidemiological context.

The date of 2008, however, does not establish whether the pathogen was introduced in Salento in that year, whether it became adapted to the new host plant in that year, or whether the date represent the initial outbreak of OQDS ^[25][29]. Similarly, other studies, based on the records of monitoring data of the phytosanitary service of the Apulia region, together with logistic functions coupled with the best fitting models, have estimated that the initial OQDS spread occurred in 2008 with a margin of error of plus or minus five years ^[34][38]. Another analysis, based on the measurement of the land surface temperature index revealed that this index started to increase from 2006 to 2010, revealing a relevant increase in olive grove mortality during those years ^[35][39]. Within this scenario, there is other evidence, based on farmer observations, of symptoms of olive decline during the years 2004–2006 ^[36][40]. To these reconstructions, it should also be added that the first symptoms of OQDS (i.e., twig and branch dieback) can be visibly scored after 1–2 years from the initial inoculation of the pathogen within the leaf by the vector ^[37][41]. Altogether, these data could bring back the early phase of X. f. subsp. pauca ST53 olive colonization in Salento during 2002–2003.

In any case, at the time of the first official record in October 2013, a well-demarcated area, which involved some municipalities around Gallipoli, was identified as the place of the initial outbreak of OQDS in Salento ^[38][1]. However, the place where the infection started is not known. One striking piece of evidence of the OQDS outbreak is that, at the time of the official record, the disease was already present on about 8000–10,000 ha, which corresponds to about 1 million olive trees ^[38][1], which precluded any attempt of pathogen eradication in the area. One year after the first report, OQDS was reported on about 23,000 ha ^[39][42].

One prevailing aspect of the initial spread of OQDS was that, apart from the Gallipoli area where the disease had the time to expand between the continuum of olive groves due to the feeding activity of the insect vectors repeated over many years, additional initial foci of the disease were observed a few months after October 2013, several kilometers away and across from each other, as in the cases of Trepuzzi, Monteroni, Galatina, and Copertino (Lecce province) along with Oria and San Pietro Vernotico (Brindisi province). Owing to the possibility that infected vectors “hitchhiked” through mobile vehicles (i.e., cars, motorcycles, bikes, tractors, trucks, buses, trains), the long-distance dispersal events of OQDS have indeed remained highly probable and account for a “stratified dispersal” model of the pathogen in Salento that implies a relevant role for the non-olive host plants for the bacterium spreading ^[40][43]. Within this scenario, it seems useful to quote an experience that occurred during a monitoring survey carried out in the Campania region (Southern Italy) for ascertaining the possible presence of X. fastidiosa in that area. During the survey, an adult specimen of Philaenus spumarius, the main insect vector of X. f. subsp. pauca in Salento, adhered to a moving car for more than 40 km, remaining alive after displacement ^[41][44]. Moreover, during summer 2022, the insect vector was also found in the wheels of cars that moved from Salento to Bari through the freeway ^[42][45]. The introduction of infected ornamental coffee plants in other local nurseries or shopping centers of Salento is an alternative hypothesis.

4. Further Spread of “Olive Quick Decline Syndrome” in Salento

The wide and continued extension of the olive groves in the Salento areas where the OQDS outbreaks started were greatly conducive to the expansion of the disease, such that, only a few years after the first record, it was defined as an endemic ^[43][68]. The high abundance of alternative host plants (i.e., natural flora, oleander, and other ornamental shrubs) greatly augments the potential dispersal of the bacterium in the area ^[40][44][43,69]. However, there are different estimates about the advancement of the disease in the affected areas. Some data obtained from the Apulia regional services indicate an expansion front of about 20 km per year ^[37][41]. In another study, through logistic function analyses based on results provided by monitoring survey data, the shape and the rate of the movement of the disease front was assessed ^[34][38]. This study indicated that, by starting from Gallipoli, in the northwest direction, the disease had an estimated mean rate of movement of about 10 km per year, and an invasion front width that spans from about 100 to 150 km characterizes the OQDS spread in Salento ^[34][38]. By contrast, EFSA, by analyzing the disease dispersal, reported that 90% of the newly infected trees for one year were observed within a mean of 5.2 km of a previously infected area ^[45][46]. However, studies on P. spumarius (i.e., the main vector for X. f. subsp. pauca ST53 in Apulia) ascertained that the mean mobility of adults during their peak activity in May and June was about 200 m, with only 2% of the adults moving for about 400 m ^[46][70]. Based on the biological activity of the main insect vector in Salento, which indicates an adult dispersal of less than 0,5 km within a single olive grove during their peak activity in May and June, and jointly to its low possibility to be passively transported by wind ^[46][70], the wider estimates of the rate of OQDS spread in the infected area—calculated at 5 ^[45][46], 10 ^[34][38], and 20 km ^[37][41] per year—should be explained by additional factors. Such factors would have greatly augmented the velocity of the disease spread starting from the Gallipoli area in a northwest dispersal direction ^[34][38]. Among these factors, (a) the feeding activity and the dispersal of the vectors on the weeds and wild plants in autumn could be underestimated, (b) the observed ability to adhere to vehicles shown by the vectors could have highly enhanced the spread of the bacterium in the olive grove continuum, and (c) other phytopathogens (i.e., fungi and bacteria) could have contributed to wilting symptoms, such as those caused by X. f. subsp. pauca ST53. Currently, the front of infection has reached many areas of the Taranto and Brindisi provinces and some municipalities south of Bari. It should also be noted that, in Salento, there is a relevant mobility of vehicles during the period when the feeding activity of the insect vectors is high (end of spring and early summer). During this time period, due to the commercial, working, and tourist activities, all kinds of motor and electric vehicles run across the heavily infected areas of Salento and reach northern areas, such as Basilicata and north of Bari, where olives are also cultivated to a great extent. The possibility that, during the last 15–20 years, no adult of infected P. spumarius, whose density can reach 40–100 nymphs per square meter of weeds ^[44][69], has reached these areas through vehicle “hitchhiking” seems quite low. Another reason for this is that, as stated above, contemporary to the outbreaks observed in the Gallipoli area, other foci of OQDS were observed in other locales of Salento quite far from each other and without a nursery nearby. Some intrinsic characteristics of these nearby still non-infected areas could have impeded the occurrence of OQDS.

5. Factors That Were Conducive to the Spread of “Olive Quick Decline Syndrome”

The non-immediate identification of the causal agent of OQDS was certainly among the main causes of the disease spread for many years in the Salento territory. It is evident that, given that about 1 million olive trees were judged as infected at the time of the official record of October 2013 ^[38][1], the pathogen had found favorable conditions for expansion since its introduction in the area from abroad. Within this scenario, given the consequent identification of the pathogen vector reported the subsequent year ^[47][71], there was a considerable time period during which the bacterium could exponentially spread for many years among the olive groves of Salento, which, in many cases, formed a continuum of trees for many kilometers ^[43][48][68,72]. During this time lapse, some predisposing factors jointly to the bacterium spread through the insect vectors activities and additional phytopathogens that occur in the area could have played a significant role either in augmenting the severity or in the further spread of OQDS in Salento. The high susceptibility of the local olive cultivars used to produce high-quality oil ^[49][50][73,74], namely Ogliarola salentina and Cellina di Nardò, to X. f. subsp. pauca largely contributed to the spread of OQDS in Salento. These cultivars indeed show, upon infection, more xylem occlusions caused by tylose formation than the less susceptible cultivar Leccino ^[51][75]. There is also anatomical evidence that Cellina di Nardò has larger and fewer xylem vessels in comparison with Leccino, resulting in a higher vulnerability to droughts ^[52][76]. These large vessels are more prone to air embolism ^[52][76], tylose formation ^[53][77], and cavitation ^[54][78], possibly conducive to X. f. subsp. pauca infection ^[52][76] or the consequent twig wilting ^[53][54][77,78]. Moreover, both Ogliarola salentina and Cellina di Nardò have a lower level of hydroxytyrosol glucoside in comparison with Leccino, this phenolic compound having a high antioxidant activity directly involved in the defense mechanism against X. f. subsp. pauca ^[55][79]. In addition, when artificially inoculated, the development of visible symptoms occurs in Cellina di Nardò plants before it occurs in Leccino plants ^[56][80], and both Ogliarola salentina and Cellina di Nardò host a significantly higher number of X. f. subsp. pauca cells in the twigs than Leccino and other olive cultivars in naturally infected trees ^[57][81]. Soil characteristics of the areas where OQDS has spread in recent years could also have been conducive to the disease. A study aimed at assessing the ionome (i.e., elemental composition at ion level) content in the soil and leaves in the olive groves of the infected areas of Salento, in comparison with olive groves of Northern Apulia and Basilicata, where OQDS is not reported, ascertained a very low content of certain micronutrients in all of the sampled olive groves of the infected areas, where a significantly lower content of zinc, copper, and manganese was found in both soil and leaves ^[58][82]. It should be noted that zinc is involved as a cofactor in many enzymes, such as alcohol dehydrogenase, RNA polymerase, and carbonic anhydrase ^[59][83], whereas copper is essential for the formation of chlorophyll ^[60][84], and manganese is involved in the photosynthetic machinery and in the detoxification of ROS ^[61][85]. Taken as a whole, these data point to trees that grow in highly depleted soil, and the leaf ionome composition reflects such a condition. Consequently, the combination of intrinsic cultivar susceptibility with a low amount of micronutrients in the soil that are important for overall plant physiology resulted in an environment that is highly conducive to OQDS spread. Given the similar geological substrate (i.e., calcareous) of South and North Apulia, some factors could have determined the relevant decrease in micronutrients found in Salento. A survey on the utilization of herbicides used in Apulia from 2003 to 2014 verified that, in areas where OQDS appeared and spread (i.e., Lecce province), the utilization of glyphosate was used twice more than in the northern areas (i.e., Bari province) ^[62][86], with an average distribution in the olive groves of three times per year ^[63][87]. This relevant distribution of glyphosate that occurred for many years could have altered the chemical and biological equilibrium of the soil. It is known that such a distribution decreases the overall availability of nitrogen and phosphorus along with the organic matter content in the soil ^[64][88], and that there is a high reduction in beneficial pseudomonads that control soil-borne pathogenic fungi, with a consequent increase in agrobacteria that play a role in manganese oxidation ^[65][66][67][89,90,91]. In addition, the prolonged utilization of glyphosate can also induce a reduced assimilation of zinc and copper in the plant ^[68][92]. It seems evident that this agronomical practice, prolonged over many years, helped to weaken the olive trees, thus facilitating disease spread. It should be added that a direct link between glyphosate utilization and increases in plant disease severity has already been verified ^[69][93]. Additional factors, such as improper pruning and extreme climatic events (i.e., droughts, frost, and “water bombs”, an exceptionally large amount of rain falling in a few hours on a relatively small area), could have contributed to the increase in the tree susceptibility to OQDS. Hard pruning carried out in an attempt to reduce the X. f. subsp. pauca inoculum load within the tree has resulted in its subsequent further weakening and death ^[56][70][80,94]. It should be said indeed that in infected young olive trees, pruning has a limited effect in eliminating the bacterium from the tree ^[70][94]. It should also be added that, in Salento, in recent decades, tree pruning was very often performed every 4–5 years, resulting in damage to plant physiology ^[71][95]. In addition to the droughts previously reported that could have been conducive to the initial OQDS outbreaks, other drought events have been further recorded, such as that occurring during the summer of 2017, when 7.8 mm of total rainfall was recorded in Lecce province from 1 June to 30 August ^[71][95]. In the same year, in January, a prolonged (i.e., one week) frost with minimum temperatures that reached on some days from −4 to −5.1 °C was recorded ^[71][95]. Because of climate change, in recent years, the Salento area faced more extreme precipitation events such as the “water bombs” accompanied by strong wind and hail. Apart from the direct damage to vegetation, there are also indirect negative effects on tree physiology due to prolonged water permanence in soil (i.e., waterlogging). When the soil is saturated with water on consecutive days, anaerobic conditions prevail, thus inducing root hypoxia in the olive trees ^[72][73][96,97]. It seems that an environment conducive to the rapid expansion of OQDS, given the combination of highly susceptible olive cultivars, improper agronomical practices, and repeated adverse climatic events, was present in Salento.

6. The Role of Vectors

One of the most intriguing aspects of OQDS outbreak and spread is the apparent ease with which an insect vector (i.e., P. spumarius) that had never encountered an alien bacterium such as X. fastidiosa in its evolutionary life so rapidly acquired the ability to efficiently spread it in a new host plant for the microbe (i.e., olive), starting its acquisition from a host plant (i.e., coffee) that is a poor source of pathogen acquisition for the insect ^[25][29]. By contrast, it is known that X. fastidiosa switches from a plant to an insect-colonized state only when it reaches a high cell density ^[74][98] and after a complex multistep process that involves, at the same time, plant, insect, and pathogen components (i.e., pectin, cutin, adhesins, and pili) ^[75][99]. Some difficulties have been observed during the probing and feeding of P. spumarius during the transmission of X. fastidiosa ^[76][100]. There might also be an additional possibility not usually considered for explaining the introduction of X. f. subsp. pauca from abroad: the direct introduction of an efficient vector from Central America that, upon its arrival in Salento, could have initially transmitted the pathogen to the olive trees and subsequently became extinct due to an inability to adapt to the new environment. Such a possibility was not completely discarded concerning X. fastidiosa outbreaks in the Balearic Islands, in Spain ^[77][101]. The live adults of Homalodisca vitripennis (“Hemiptera”: “Cicadellidae”), one of the main vectors of Pierce’s diseases caused by X. f. subsp. fastidiosa in California, was indeed intercepted in French Polynesia and Japan from cargo bins, hangars, and planes ^[45][46]. The presumptive fast adaptation of the bacterium to a new vector would require additional studies to determine the mechanisms of their interaction. The polyphagous P. spumarius (“Hemiptera”: “Aphrophoridae”), also known as the “meadow spittlebug”, is the main vector of X. f. subsp. pauca ST53 in Salento ^[78][79][102,103]. Other less efficient vectors include Philaenus italosignus (“Hemiptera”: “Aphrophoridae”) and Neophilaenus campestris (“Hemiptera”: “Aphrophoridae”) ^[79][80][103,104]. P. italosignus is rarely found in olive groves, and the nymphs exclusively colonize Asphodelus spp., which also serves ovideposition, whereas N. campestris was positive for the bacterium only in May due to its non-preference for olive trees and its preference for Poaceae plant species ^[81][105].