Fig 1.
Characteristic hazelnut decline symptoms observed on hazelnut plants (cv. Tonda Gentile Romana).
Decline began with sudden wilting of a stem (A) while the rest was still intact. Successively, decline affected the whole branch which manifested the dieback while the remaining branch manifested suffering state with pale green foliage (B). When the bark of branch bearing pale green foliage was excised, discoloration of cambium could be observed (C). During the end of vegetative phase complete plant death occurred although leaves were still attached (D). The following year, also the plants present in the proximity of the plant that showed initial decline symptoms appeared completely affected (E).
Fig 2.
A very young hazelnut plantation (4 year old plants, cv. Tonda gentile Romana) completely decimated by hazelnut decline phenomenon.
The young plants were planted in the same plot where severely affected old plants (15-year-old) were removed.
Fig 3.
Incidence of hazelnut decline (%) across the study sites in the province of Viterbo.
The circle size inside the map indicates the different hazelnut decline incidence.
Table 1.
Isolation of Pseudomonas avellanae (Pav) from hazelnut twigs collected from 2010 to 2012.
Samples were collected from March to August, each year. The average percentage of Pav isolation for the three years is calculated by the ratio between the number of Pav-positive samples and the total number of samples used for isolation.
Fig 4.
An example of BOX PCR fingerprint patterns obtained for genomic DNAs of Pseudomonas avellanae strains isolated from Corylus avellana by using the BOX primer set.
Lanes from left to right—M: molecular size marker (1-kb ladder), 1: strain PaVT1, 2: strain PaVT2, 3: strain PaVT5, 4: strain PaVT6, 5: strain PaVT8, 6: strain PaVT10, 7: strain PaVT17, 8: strain PaVT24, 9: strain PaVT25, 10: strain PaVT28, 11: strain PaVT29, 12: strain NCPPB 3489, 13: strain NCPPB 3487 and m: molecular size marker (100bp ladder). All strains isolated in this study were clonal and no diversity was observed among them.
Fig 5.
Bayesian tree constructed on the concatenated sequences of cts, gyrB, rpoD and gapA (1465bp).
Bootstrap values are indicated at each node, while names at the branches indicate the name of the strains. Pathovar affiliations are reported for some of the Pseudomonas syringae strains belonging to different phylogenetic groups. Sequences for the different P. syringae strains and their phylogroup affiliation were described previously by Berge et al. [46]. MLST sequences of the Pav strains isolated previously from Greece and Italy were reported by Wang et al. [18]. Sequences for PavCRAFRUec1, PavCRAFRUec2 and PavCRAFRUec3 are published by Marcelletti and Scortichini [41]. Pseudomonas aeruginona (PAO1) was used as an outgroup. Triangles with different colours are used to label strains in the following way. Blue triangles: strains isolated in this study, red triangles: strains isolated in central Italy during the first epidemic, and green triangles: strains isolated in Greece.
Fig 6.
Length of external and internal necroses developed on hazelnut plants (cv. Tonda Gentile Romana) inoculated through leaf scar.
The length was measured after 4 weeks post infection. The reference strains of Pseudomonas avellanae (NCPPB 3487 (= BPIC631) and NCPPB 3489) previously isolated in Greece from bacterial canker symptoms produced mild external and internal necroses, together with the strain PaVT10 isolated in this study. The rest of the strains tested in this study caused only weak necrosis on the inoculated plants. The error bars represent the standard error of the measurements.