Ceratocystis ficicola causing a serious disease of Ficus carica in Greece

Copyright: © 2021 P. Tsopelas, N. Soulioti, M.J. Wingfield, I. Barnes, S. Marincowitz, E.C. Tjamos, E.J. Paplomatas. This is an open access, peer-reviewed article published by Firenze University Press (http://www.fupress.com/pm) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


INTRODUCTION
is one of the oldest domesticated tree crops globally. According to Kislev et al. (2006) this tree was cultivated for its fruit during the Early Neolithic period (11,400 to 11,200 years ago), preceding cereal domestication. Ficus (Moraceae) includes more than 700 species, native to the tropics or sub-tropics. Ficus carica is indigenous to the Middle East and Asia Minor, and is mostly cultivated in warm and dry areas of the Mediterranean region (Condit, 1955;Flaishman et al., 2008). Ficus carica has been cultivated in Greece since ancient times, as mentioned by Homer (8th century BC) The Odyssey, and Theophrastus (4-3rd century BC) described cross-fertilization of wild and domesticated figs (Condit, 1955). Fig continues to be an important crop, in Greece, where production in 2019 was 19,730 t (FAO, 2019). Thus, any negative impact on these trees due to serious disease is considered important from both a cultural and an economic standpoint.
Since the 1970s, a serious disease caused by a Ceratocystis species has been reported in Japanese fig orchards. The causal agent of this disease was initially attributed to Ceratocystis fimbriata sensu lato (Kato et al., 1982). Kajitani and Masuya (2011) described the pathogen as a new species, Ceratocystis ficicola Kajitani and Masuya, based on discernable morphological characteristics and DNA sequence data. Like many other Ceratocystis species, C. ficicola is a soil-borne pathogen that infects roots and main stems of susceptible hosts, causing vascular wilt and eventually tree death. The pathogen is dispersed by human activities, mainly with contaminated plant material, and has spread to many fig-growing areas of Japan (Kajitani and Masuya, 2011;Kajii et al., 2013).
In 2018, a Ceratocystis species was isolated from two neighbouring fig orchards in the Attica region of Greece. The affected trees showed symptoms of wilt and extensive crown defoliation, and some trees had died. The morphological characteristics of the isolated fungus were similar to those of C. ficicola (Tjamos et al., 2018). In 2019 and 2020, the same fungus, based on morphology, was also isolated from diseased fig trees at different localities on Euboea Island, where considerable damage was occurring in fig orchards.
The objective of the present study was to identify Ceratocystis species, having similar morphology to C. ficicola, found causing disease on F. carica in the two regions of Greece. This was achieved based on multilocus phylogenetic analyses and detailed observation of morphological characteristics of the fungus, including comparisons with the ex-type isolate of C. ficicola. Pathways of dispersal of the pathogen were also considered and the extent of damage to fig orchards in Greece is reported. Pathogenicity tests were also conducted with the fungus on plants of F. carica and Ficus benjamina L.  (Table 1). These two regions are more than 200 km apart.

Sampling and fungus isolations
Trees with symptoms including crown defoliation and leaf wilt were sampled. The lower parts of the main stems were examined for the presence of necrotic bark, which was then removed to expose diseased cambial tissues and discoloured wood ( Figure 1C-Ε). Samples of stained wood were then cut from the lesions using a surface-disinfected chisel. Some of the young plants were uprooted; the root systems examined for disease symptoms, and these were collected for laboratory examination. Soil samples were also collected from a depth of 15-20 cm near the bases of symptomatic trees from most of the surveyed orchards.
For isolation of fungi, PARPH V8 agar medium was used (containing 5% clarified V8 juice, 5 μg mL -1 pimaricin, 250 μg mL -1 ampicillin, 10 μg mL -1 rifamycin, 50 μg mL -1 PCNB, and 50 μg mL -1 hymexazol). This selective medium is widely used for the isolation of Phytophthora species (Ferguson and Jeffers, 1999) but was also effective for the isolation of Ceratocystis species. Direct isolation from wood samples was performed after they had been surface-disinfested with 70% ethanol and air dried in a laminar flow cabinet. Small pieces of stained wood were aseptically removed from the samples with a sterile scalpel and transferred to Petri dishes containing PARPH V8A medium.
A carrot baiting method was also used for a sub-set of samples. Small pieces of stained wood were placed between two slices (each 5 mm thick) of carrot root, as described by Moller and De Vay (1968). The carrot baits were then placed in empty Petri dishes with moistened filter paper and incubated for 7-10 d at 25 o C. The carrot discs were subsequently examined for the presence of ascomata under a dissecting microscope, and when these were present, masses of ascospores were transferred to PARPH V8A medium.
Isolations from soil samples were performed using a modification of the trapping technique described by Grosclaude et al. (1988) for Ceratocystis platani (Walter) Engelbrecht and Harrington. Soil samples (40-50 g) were placed in 250 mL capacity conical flasks, which were filled with 200 mL of sterile deionized water. Freshly cut twigs from fig trees (approx. 10 cm long and 6-10 mm in diameter) with the bark removed were placed in the flasks. Air was continuously pumped into the flasks through a Pasteur pipette, using an aquarium pump (Aqua-Air AP4, Interpet). The flasks were incubated at room temperature (23-25 o C). After 7-10 days, the twigs were examined under a dissecting microscope for the presence of ascomata and, where present, masses of ascospores were aseptically transferred to PARPH V8A medium. In some cases, this trapping technique was also used with pieces of infected wood in the flasks rather than soil. In all cases, resulting, agar medium colonies having Ceratocystis morphology were sub-cultured onto V8A medium.

Growth in culture
Growth rates of fungal cultures were determined on 10% V8A at different temperatures from 10 to 40 °C at 5°C intervals. Two measurements of colony diameters were made perpendicular to each other on each culture plate after 7 and 14 d incubation at 25°C.
Measurements of fungal structures were made from 7 to 10d old pure cultures on V8A and malt extract agar (MEA) with fungal structures mounted in water or 85% lactic acid. A Zeiss Axioskop light microscope and SC30 camera (Olympus) or a Nikon Eclipse Ni microscope and DS-Ri2 camera (Nikon) were used to capture images and to take measurements. Ascomata were examined at 50x or 100× magnification and the conidia and ascospores at 400× or 1000×. For each of the two isolates from Greece (CMW 56935 and CMW 54794), the dimensions of 50 conidia of each type were measured.
Besides the isolates from Greece, the ex-holotype culture of C. ficicola from Japan (MAFF 625119 = CMW 38543) was also examined.

DNA extraction, sequencing and phylogenetic analyses
To confirm the identity of the Ceratocystis species isolated from infected fig trees, two cultures from infected wood and two from the soil baiting (Table 2) were subjected to DNA sequencing. Mycelium from the cultures grown on 2% MEA for approx. 2 weeks was scraped from the surfaces of the agar culture plates and DNA was extracted using the Fungal DNA MiniPrep kit (Zymo Research).
PCR reactions and conditions, including optimal MgCl 2 concentrations and annealing temperature for the three gene regions, were carried out as described by Fourie et al. (2015), with an Expand-PCR programme of 96°C for 10 min, (94°C for 30 s, primer specific anneal- ing temperature for 45 s, 72°C for 1 min) for 10 cycles, (94°C for 30 s, primer specific annealing temperature for 45 s, 72°C for 1 min + 5 s/cycle increase) for 30 cycles, and 72°C for 10 min. All amplicons were purified using 6% Sephadex G-50 columns (Sigma-Aldrich), were sequenced in both directions using the ABI PRISM TM Big DYE Terminator Cycle Sequencing Ready Reaction Kit (Applied BioSystems), and the products were run on an ABI 3100 Genetic Analyzer (Applied BioSystems, Thermo Fisher). The forward and reverse sequence reads were assembled into contigs in CLC Bio Main workbench v.6 (CLC Bio, www.clcbio.com), and the consensus sequences generated were used in BLAST analyses against the NCBI GenBank database (NCBI; http://www.ncbi.nlm.nih.gov) to determine the closest similarity matches with authenticated Ceratocystis species.
Sequences were added to the combined rbp2, bt1 and tef1 Ceratocystis database obtained from Barnes et al. (2018) (TreeBASE No. S22005), with a focus on retaining all the Ceratocystis species in the Asian-Australian Clade (AAC) in the phylogenetic analyses. Ceratocystis albifundus was used as the outgroup taxon. Sequences were re-aligned using the online version of MAFFT v. 7 (http://mafft.cbrc.jp/alignment/server/). Maximum parsimony (MP) analyses performed in PAUP v. 4.0 (Swofford 2002) and Maximum likelihood (ML) analysis in RAxML (Stamatakis 2014) were carried out on the combined dataset with the same parameters used by Barnes et al. (2018).

Inoculation tests
Two isolates of C. ficicola collected from different localities were used for inoculation trials. One of these (CMW 56935) from the Attica region was used in the first test (2018) and the other (CMW 54794) from Euboea Island was used in the second test (2019). The isolates used in the inoculation tests were grown for 2 weeks at 25°C in Petri dishes containing V8A. For inoculations, a 3 mm diam. cork-borer was used to remove the bark from stems or branches of test trees (see below), and a mycelium plug of similar size was inserted into each wound which was then wrapped with Parafilm®. Control plants were inoculated in the same way with sterile V8 agar plugs.
The two inoculation tests were carried out at the Institute of Mediterranean Forest Ecosystems in Athens, Greece. In the first test, 11 branches (diam. 1-4 cm) of a wild F. carica tree were inoculated in June 2018 with isolate CMW 56935. In the second test conducted in September 2019, a total of 13 F. carica saplings (1-2 years-old) of different size and grown in 7.5 L capacity pots, were inoculated with isolate CMW 54794. Stem diameters of these plants at the inoculation points was from 5 to 20 mm. Non-inoculated control plants were not used in the first test, while in the second test six plants were inoculated with sterile V8 agar plugs.
In September 2019, inoculations were also performed on branches of two Ficus benjamina trees, in the same area with isolate CMW 54794. A total of 16 branches on the two trees were inoculated. A third F. benjamina tree was used as a control, by inoculating seven branches with sterile V8 agar plugs.
After inoculation, the plants were monitored regularly for expression of disease symptoms. In the first inoculation test (2018), the inoculated branches were harvested after 5 weeks. In the second test (2019), most of the plants were harvested 6 weeks after inoculation, although some of the plants displaying symptoms of wilt were harvested at an earlier time to allow for lesions to be measured. All inoculated stems and branches were sliced open vertically above and below the inoculation points, and the lengths of internal xylem staining were measured. Re-isolations from these lesions were made at different distances from the inoculation points, by aseptically transferring pieces of stained wood onto the selective V8A medium.
Statistical analyses were performed to compare the means of internal lesion lengths for the inoculations on F. carica and F. benjamina branches or stems of the F. carica sapling plants. In these comparisons, only the saplings of F. carica that were harvested after 6 weeks were included. A Shapiro-Wilk test was used to test normality of the data, and a Levene's test was used to assess equality of variances. A Kruskal-Wallis H test was then applied to compare the means of lesion lengths between F. carica and F. benjamina plants, followed by Dunn-Bonferroni post hoc method.

Disease symptoms and distribution
In the Markopoulo Mesogaias municipality in East Attica, wilt of fig trees was detected at one location with orchards of mature fig trees (20-30 years-old) of the 'Vasilika honey-white' and 'Vasilika black' cultivars, distributed over approx. 10 ha (Table 1). In the Istiaia-Aidipsos municipality, in the northern part of Euboea Island, the disease was found in fig orchards at four locations (villages) more than 10 km apart, on young as well as mature trees. All the wilt-affected trees on Euboea Island were of the 'Smyrna' cultivar (Table 1).
Wilted mature trees were usually located in one part of each orchard, and the disease appeared to have spread to neighbouring fig trees. In some cases, more than one disease centre was observed. Infected mature trees of all three cultivars showed thinning and yellowing of the foliage, either on some of the branches or over the entire crowns ( Figure 1A). In some trees, most of the leaves had dropped by the end of summer, while some of the unripe fruits were retained on the trees indicating rapid death ( Figure 1B). In many of the affected mature trees, complete defoliation was observed by the end of the summer, although full foliage had been present on these trees early in the spring. The affected trees eventually died, and in many cases, where dead trees had been replaced with new young plants; these were also dying within 1 to 2 years after replacement.
On mature trees, extensive bark cankers were observed at the bases of the trunks. Necrotic dark brown sapwood was obvious after removal of the bark in the lower part of the trunks ( Figure 1C-E), and these lesions extended upwards. Necrotic roots were also observed on affected trees. In some symptomatic young (2-3 yearsold) trees, wood staining was not always present at the trunk bases, although in these cases necrotic roots were observed. Wood staining was not found on the branches of wilted trees with chlorotic leaves, or on defoliated trees. Infestations by wood boring insects were not observed on any of the wilted fig trees examined.
In a newly planted 6 ha orchard on Euboea Island, 3 years after planting, approx. 40-50% of the trees were dead or dying with evident disease symptoms. This land had not previously been used for fig tree cultivation.

Fungus isolations
A Ceratocystis species was consistently isolated directly from stained fig tree wood on the selective PARPH V8A medium. The fungus was also isolated on plain V8A, but in most of these cases it was overgrown by saprotrophic fungi. On selective medium, colonies were suppressed and scarcely formed ascomata, but they resumed normal growth when transferred to plain V8A medium and ascomata developed after 7-10 d.
The Ceratocystis species was also isolated using the carrot baiting technique, but in many cases the carrots were contaminated with bacteria and isolations were inconsistent using this technique. The trapping technique with fig twigs as baits was effective for isolating the Ceratocystis species from soil and from wood samples. For twig baiting, the fungus had colonized the twigs and formed ascomata (Figure 2A) after 7-10 d, at the water level. The fungus was easily isolated from ascospore masses transferred to selective medium. In a few cases, it was not possible to isolate the fungus from wood samples taken from symptomatic trees, especially where these trees had died, but the fungus was readily isolated from soil samples collected near the bases of these trees using the twig baiting technique.

Fungus morphology
Colonies in V8A, after 7-10 d in culture, were circular with slightly undulate margins and were dark olivaceous green. The optimum growth temperature was 25°C, with a radial growth of 20-23 mm week -1 . The fungus grew well at 30°C but no growth was observed at 35 or 40°C. When the cultures were transferred to 25°C after 2 weeks incubation at 35°C, the colonies resumed normal growth, but incubation for 2 weeks at 40°C killed the fungus and no growth was observed on the plates after transferring these to 25°C.
Morphological characteristics of the Greek isolates were identical to those of the Japanese isolate (exholotype MAFF 625119 = CMW 38543), except that the dimensions of the endoconidia were different. The Japanese isolate had cylindrical endoconidia with dimensions of 5-9.5 × 4.5-8 µm, whereas those made in the present study were 12-40 × 4-7 µm.

Sequencing and phylogenetic analyses
Amplicons of the rpb2, bt1 and tef1 gene regions produced fragments of approx. 1123, 592 and 754 bp, respectively. For each of the three gene regions, the sequences for all four isolates were 100% identical to each other. Blast searches against the NCBI GenBank database resulted in 100% identity matches to C. ficicola ex-type isolate from Japan [Mycobank MB518749, CMW 38543 for rpb2 (KY685082), bt1 (KY685077) and tef1 (KY316544) gene regions (Table 2)].
The combined dataset (rpb2, bt1 and tef1) of the sequences used for the phylogenetic analyses consisted of 2127 aligned characters. MP analyses resulted in eight most parsimonious trees with TL = 299, CI = 0.863, RI = 0.945 and RC = 0.815, with 238 parsimony-informative characters and six uninformative characters. For the ML analyses, using GTR-GAMMA as the best fit model, the log-likelihood of the most likely tree obtained was -5139.12. MP and ML analyses placed the isolates from Greece, collected from infected F. carica, and those obtained from the soil, in the same clade as those of C. ficicola from Japan with 100% bootstrap support (Figure 3). These isolates form part of the Asian Australian Clade (ACC) of Ceratocystis, as defined by Li et al. (2017).

Inoculation tests
All the plants of F. carica and F. benjamina inoculated with C. ficicola developed symptoms of infections (Figure 4), while no symptoms were observed on any  Figure 4B), which, in some cases, was evident as early as 2 weeks after inoculation. Exposure of the cambium in stems showed brown-coloured streaked discolouration of the woody tissues, extending longitudinally in the xylem beyond the necrotic bark ( Figure 4D, E).

Branch inoculations on Ficus carica
The internal lesions associated with the stained wood on the F. carica branches inoculated in 2018, and measured after 5 weeks, extended up to 12 cm from the inoculation points. The lesion lengths were 10-23 cm (mean = 14.2 cm).

Inoculations on Ficus carica saplings
Three of the F. carica saplings inoculated in 2019 that had small diameters (57 mm), showed stem girdling and symptoms of wilt 2 weeks after inoculation. For these three plants, the lesions were 6-12 cm long. For the other ten saplings, the internal lesion lengths were up to 21 cm long at 6 weeks after inoculation. The total lengths of the stained wood associated with these inoculations was from 2.6-37.8 cm (mean = 12.9 cm) ( Figure 5).

Branch inoculations on Ficus benjamina
Inoculated-branches developed necrotic lesions around the inoculation sites ( Figure 4C). On this host, wood staining was very limited, extending only 0.8-1.5 cm from the inoculation points with total length 1.4-3 cm (mean = 2.17 cm). Symptoms of wilt were not evident on branches 6 weeks after inoculation.
There were no statistically significant differences in the mean internal lesion lengths between F. carica saplings (stems) and branches (twigs) in the two experiments, according to Kruskal-Wallis H test and Dunn-  Bonferroni post hoc method (P ≤ 0.05) ( Figure 5). In contrast, the mean lesion lengths on the F. benjamina branches differed significantly from both of the F. carica inoculations. There were also significant differences between lesion lengths for the control plants and those inoculated with the C. ficicola on both hosts ( Figure  5). For F. carica, wood discolouration was limited to approx. 3 mm around the inoculation sites.
The inoculated fungus was consistently re-isolated from stained wood from F. carica and F. benjamina plants, and at different distances (up to 17 cm) from the inoculation points. The re-isolated fungus had the typical morphology of the inoculated fungus. The fungus could not be isolated from plants inoculated as experimental controls.

DISCUSSION
This study has provided clear evidence for the presence of C. ficicola in fig orchards in two separate regions (Attica and Euboea Island) of Greece. These two areas separated by large distance. This is the first detailed report of the presence of this fungus outside of Japan, in this case also supported by morphological and DNA sequence comparisons. This study also included inoculation trials that confirmed the pathogenicity of Greek isolates of C. ficicola.
A similar disease to that caused by C. ficicola in Japan and Greece has been reported in Brazil on cultivated fig trees. In that case the pathogen was reported as Ceratocystis fimbriata Ellis and Halsted sensu lato (Pasura and Harrington, 2004). Kajitani and Masuya (2011) showed that C. ficicola is different to the Ceratocystis species from fig trees in Brazil, based on morphological differences and ITS sequences. The distinct nature of the fungus killing fig trees in Brazil and Japan was also confirmed by Harrington et al. (2011). Results of the present study based on morphological comparisons and DNA sequence data for three gene regions also confirmed that C. ficicola from Greece is in the Asian-Australian clade of Ceratocystis (Li et al., 2017). In contrast, isolates from fig trees in Brazil form part of the Latin American Clade (Harrington et al., 2011).
Observations in the present study agree with those of Kajitani and Masuya (2011), confirming that C. ficicola has morphology that is distinct from other species of Ceratocystis. This is relevant as many species of Ceratocystis, as defined by De Beer et al. (2014), cannot easily be distinguished based on morphology. The most characteristic structures in C. ficicola that distinguish it from other Ceratocystis species are the large ascomata with very long necks (>2000 μm). Ascospores are very similar in shape and size to those of Ceratocystis fimbriata sensu stricto (Marincowitz et al., 2020). Doliform endoconidia were not observed in the present study, and these were also absent from observations of Kajitani and Masuya (2011). However, there was a difference in the size of the cylindrical endoconidia in the present study isolates compared with those described by Kajitani and Masuya (2011). This was confirmed in comparisons of the Greek isolates and the Japanese ex-holotype isolate (CMW 38543) made in the present study.
In Japan, the disease of fig caused by C. ficicola is commonly referred to as "Ceratocystis canker", but Kajii et al. (2013) stated that the disease is more typical of vascular wilt. The pathogen invades the roots and the main stems of host plants, causing xylem dysfunction and wilt symptoms on infected fig trees. In cross sections of main stems, dark brown radial discolouration of the sapwood tissues was evident at the bases of affected trees, extending upwards and downwards to the roots (Kajii et al., 2013). Consistent with the view of Kajii et al. (2013) that infection by C. ficicola should be defined as a wilt disease, Morita et al. (2016) and Sumida et al. (2016) in inoculation studies on mature fig trees, as well as young seedlings, reported that xylem discolourations were correlated with xylem dysfunction. The water supply to leaves decreased when infection progressed causing leaf wilting and extensive xylem dysfunction leading to plant death.
Ceratocystis ficicola also kills host cambium and the bark tissues causing cankers, which were present at the bases of the main stems on most symptomatic trees in Greece. In our surveys, wood discolouration was not evident in the branches of trees displaying wilt of the leaves, and similar observations were made in Japan by Kajii et al. (2013). However, the inoculated branch twigs in our experiments developed cankers and extensive wood discoloration. Overall, these symptoms are similar to those of many other tree diseases caused by Ceratocystis species, and the disease is best described as canker-wilt (Tsopelas et al., 2017;Nasution et al., 2019).
In the present study, inoculation experiments, showed that F. benjamina and not only F. carica plants can become infected with C. ficicola. However, this species was less susceptible to the pathogen than F. carica. In Japan, Ficus erecta Thunb., which is an indigenous species, has been proven resistant to C. ficicola and has been used in breeding programmes to develop resistant fig tree rootstocks (Yakushiji et al., 2019).
In several studies carried out in Japan, it has been shown that C. ficicola is a soil-borne pathogen. It has also been suggested that the fungus persists in the soil as thickwalled aleurioconidia, and newly planted trees in contami-nated soil are soon infected by the pathogen (Kajii et al., 2013;Yakushiji et al., 2019). Our field observations, in both regions of Greece considered, also suggest soil-borne disease problems, where trees were planted to replace those that had died also became infected. We consistently isolated C. ficicola from soils associated with dying trees.
Detection of C. ficicola in the soil at an early stage will be important for pathogen detection and application of canker-wilt control strategies. In the present study, using twig baiting, the fungus was readily isolated from soil, even in cases where the pathogen could not be directly isolated from wood samples from nearby symptomatic trees. This baiting method was originally described by Grosclaude et al. (1988) for detection and isolation of Ceratocystis platani, which causes the canker stain of plane trees (Platanus). The method was also successfully used to isolate C. platani from adult Platypus cylindrus Fab. beetles, a vector of C. platani (Soulioti et al., 2015). The present study is the first to use baiting technique to isolate C. ficicola from soil and wood samples.
According to Kajitani and Masuya (2011), the dispersal biology of C. ficicola is not fully understood. One of the suggested pathways of disease spread in Japan was with infected propagation material (Kajii et al., 2013). This was also evident in one of the orchards examined on Euboea Island. The pathogen was widespread in a new plantation over an area of 6 ha, 3 years after the fig trees had been planted. Since there was no previous use of this land for fig cultivation, the most probable means of pathogen dissemination over such a short period would be with infected nursery stock.
Every effort should be made to avoid further spread of C. ficicola with planting material and via contaminated soil. This is a probable pathway for introduction and spread of the pathogen into other areas and countries. The pathogen can survive in soil associated with plants for planting, when symptoms are not evident. While nursery stock should be carefully examined for the presence of C. ficicola, disease symptoms may not be always evident, especially during winter when new plants are being established. This problem is exacerbated with fig trees having no foliage during this period, and infected plants may escape the attention during inspections.
Ceratocystis ficicola may also spread with tractors used for ploughing that move infested soil and wood debris within or between orchards. Deep ploughing also causes root wounding, allowing pathogen infection sites. The use of infested ploughing machinery is likely to be a major pathway for pathogen dissemination in the fig orchards of Euboea Island. A major pathway of disease spread for C. platani in many areas of Europe has been from terracing machinery that transmit the pathogen over short and long distances (Tsopelas et al., 2017). Some Ceratocystis species can also survive for long periods as aleurioconidia in infected wood. This spore form was also observed by Kajii et al., (2013) in the vessel elements of fig trees infected with C. ficicola. Contaminated pruning and cutting tools may therefore be involved in the spread of the pathogen to healthy trees. Precautionary measures, including disinfestation of pruning tools and the machinery used in fig orchards have been recommended to avoid further spread of C. ficicola in Greece.
Ambrosia beetles (Euwallacea interjectus Blandford) have been suggested as possible vectors of C. ficicola in Japan. These insects also contribute to expansion of the fungus in sapwood of infected trees (Kajii et al., 2013). The frass produced by the insects may also disseminate the pathogen, as with other Ceratocystis species (Harrington, 2013;Tsopelas et al., 2017). Frass can be dispersed by wind, rain splash or running water over short distances, and can contaminate soil or initiate new infections in pruning wounds (Harrington, 2013). In Greece, infestation by wood boring insects in C. ficicola-infected fig trees has not been observed. However, future surveys should include careful inspection for the presence of such insects. The ambrosia beetle Platypus cylindrus has been reported in Greece to attack Platanus orientalis trees already infected by C. platani, and this insect may be involved in the spread of that pathogen (Soulioti et al., 2015).
The origin of C. ficicola is unknown. Kajitani and Masuya (2011) have suggested two possibilities; one that the fungus is invasive in Japan, or alternatively, that it is native to that country and has, over time, come into contact with the highly susceptible F. carica. All evidence emerging from the present study indicates that C. ficicola is an alien invasive pathogen in Greece. This because F. carica has been cultivated in Greece since ancient times, but canker-wilt has only been detected in recent years and at a small scale. An important question remains as to how the pathogen was introduced into this country, either from Japan or from another source in the Mediterranean region where it is yet to be detected.
Ceratocystis ficicola causes a very important disease of fig trees in Japan, resulting in severe outbreaks of canker-wilt (Kajitani and Masuya, 2011;Yakushiji et al., 2019). The presence of the pathogen in Greece is particularly relevant . Fig is a very important crop in Mediterranean countries, because it is adapted to the dry climatic conditions of the region. About 70% of total world fig production occurs in Mediterranean countries (Flaishman et al., 2008). Consequently, the disease could result in severe loses if the pathogen were to be disseminated to other countries, such as Turkey, Egypt or Morocco where fig production is also important.