Experiments were carried out at fields naturally infested by V. dahliae located in the term of Andújar in 2012 and 2013 (Jaén province, Andalucía, southern Spain) planted with different commercial olive cultivars, such as ‘Picual,’ ‘Hojiblanca,’ ‘Cornicabra,’ ‘Frantoio,’ or ‘Arbequina.’ Sampling was made from olive trees when Verticillium wilt symptoms disease symptoms were apparent. In 2012, samples were collected from various cultivars growing in the experimental plot, whereas in 2013 they were collected from 15 year-old ‘Picual’ (susceptible cv.) trees. In 2012, three groups of samples and 10 samples per group were collected, with no consideration of the cultivar: asymptomatic shoots from asymptomatic trees, asymptomatic shoots from affected trees and dead shoots from affected trees. In 2013, the three groups of samples collected were: asymptomatic shoots (n = 7), affected shoots with different degrees of symptom severity (defoliation, chlorosis, and/or wilting, n = 21) and dead shoots (n = 4). For both years, samples were collected during late spring and each sample was collected from just one tree. Starch content and density of occluded vessels (only for year 2013) were determined as explained below.

Plant material used for artificial inoculations consisted of 6 month-old plants of ‘Picual,’ ‘Arbequina,’ and ‘Frantoio’ cultivars, which are considered highly susceptible, moderately susceptible and resistant, respectively, to defoliating isolates of V. dahliae in field conditions (Trapero et al., 2013b). In controlled conditions, both ‘Picual’ and ‘Arbequina’ are considered susceptible (López-Escudero et al., 2004). Plants were inoculated with the V. dahliae defoliating isolate (VCG1A) named “V117,” from the collection of the Agronomy Department, University of Córdoba (Blanco-López et al., 1984). This isolate was collected in southern Spain from a cotton plant. The high virulence in olive of this isolate has been previously reported in several artificial inoculations (López-Escudero et al., 2004, 2007; Martos-Moreno et al., 2006; Trapero et al., 2013a, 2015). Plants were inoculated by dipping their bare roots for 30 min in a 10⁷ conidia/ml suspension of the described V. dahliae isolate. Non-inoculated (control) plants were treated by the same way but using sterilized water. Both inoculated and control plants were then transplanted to a mix of peat, coir, and perlite (55:30:15) in sterile 1.5 L pots. Plants were grown in a greenhouse under day/night temperatures of 22 ± 3 and 18 ± 3°C, arranged in a randomized block design.

Disease severity was evaluated on a weekly basis for 10 weeks, starting 2 weeks after inoculation. Wilt resistance was assessed on a scale from 0 to 4 based on the percentage of plant tissue affected by chlorosis, leaf and shoot necrosis or defoliation (0 = healthy or asymptomatic plant; 1 = plant affected by 1–33%; 2 = 34–66%; 3 = 67–99%; 4 = dead plant) (López-Escudero et al., 2004). For each cultivar, 41 plants were inoculated and 41 treated as a control (non-inoculated). To record the disease severity, 14 of the inoculated plants and 14 of the non-inoculated plants per cultivar were kept until the end of the experiment, evaluating weekly the disease severity from the 4th to the 15th weeks after inoculation. To determine the pathogen DNA amount, starch content and density of occluded vessels, the remaining 27 inoculated plants and 27 non-inoculated plants per cultivar were used. For the pathogen DNA, three inoculated and three control plants per cultivar were collected and sampled at different times: just before inoculation and at 1, 3, 5, 7, 9, 11, 13, and 15 weeks after inoculation. For both starch content and density of occluded vessels the same plants were used, but they were grouped by symptom categories. The determinations were only made from the time at which disease symptoms started, that is at 7, 9, 11, 13, and 15 weeks after inoculation.

In this experiment, only ‘Picual’ plants were used, which were grown under the same greenhouse conditions described before, but without inoculating plants. Instead, plants were subjected to three different irrigation treatments.

One group of five plants was watered daily (control plants) and a second group of six plants remained without irrigation during 10 days and showed wilting symptoms (drought stressed plants), at this time starch content and density of occluded vessels were determined in both groups. A third group of six plants remained without irrigation during 10 days and after that they received daily irrigation during a period of 14 days (post-drought-irrigated plants), at this time starch content and density of occluded vessels were determined as explained below.

For assessing the starch content, from each sampled plant we cut a 2 cm long portion from the middle of the stem, then with a hand microtome three transversal sections, approximately 30 μm thick, were obtained and subsequently stained with Lugol (2 g KI and 1 g I₂ dissolved in 100 ml distilled water) for 3 min. The slides were observed using a Nikon YS100 light microscope and the starch content in the different tissues (pitch, xylem, phloem, and cortex) were assessed on the basis of a 0 to 3 rating scale: 0 = less than 10%; 1 = from 10 to 30%; 2 = from 30 to 60%; 3 = from 60 to 100% stained tissue. To obtain a more objective estimation, the samples were blind-analyzed.

The density of occluded vessels in the xylem of the stem was determined by light microscopy observation following a similar procedure than that for starch content. In this case, sections were stained with toluidine blue at 0.05% (Sakai, 1973) for 3 min. The number of occluded vessels was counted using a 40x magnification objective and referred to the xylem area (n°/mm²). This area was calculated from data of two perpendicular diameters of both the internal and external limits of the xylem. Partially occluded vessels were observed, but only the completely occluded vessels were counted.

For quantifying the amount of pathogen DNA in the experiment of artificial inoculation, one sample was taken per plant. Each sample of 0.5 g shoot tissues was composed by 10 1 cm-long shoot segments randomly taken from a single plant. Samples were immersed in liquid nitrogen for 10 min. Subsequently, the frozen samples were lyophilized for 24 h, ground into a fine powder using a freezer mill (Spex 6770, SPEX SamplePrep, Metuchen, NJ, United States), and stored at -80°C prior to use. The pathogen DNA was quantified using real-time PCR. First, DNA was extracted from the samples using the DNeasy Plant Mini Kit from Qiagen (Hilden, Germany). Primers sequences and amplification conditions were described by Eynck et al. (2007), reaction buffer was modified by adding 0.3 μM bovine serum albumin (Sigma-Aldrich, St. Louis, MO, United States) to the reaction mixture which was 10 μl in volume. Real-time PCR amplification and melting curve analysis were performed using the iCycler System (Bio-Rad, Hercules, CA, United States). The results were analyzed with the Bio-Rad iQ5 program (Bio-Rad, Hercules, CA, United States). The amount of V. dahliae DNA was estimated from a calibration curve using increasing amounts of genomic V. dahliae DNA from 0.5 to 64.0 pg. The concentration of V. dahliae DNA used for the construction of the calibration curve was estimated using densitometry on agarose gels stained with ethidium bromide, with Lambda Phage DNA as the standard.

The starch content and density of the occluded vessels were analyzed using an analysis of variance (ANOVA) performed by Statistix 9.0 (Analytical Software, Tallahassee, FL, United States). The mean values were compared using the Fisher’s protected LSD test at P = 0.05. Due to a lack of homogeneity in the variances, the values for disease severity were analyzed using a non-parametric Kruskal–Wallis test at P = 0.05.

Shoots showing no symptoms, from both asymtomatic or symptomatic trees, contained abundant starch in the stem, which was distributed in the pith, xylem (radial and axial parenchyma), phloem (radial and axial parenchyma) and cortex (Figure 1A). For both years and for most of the tissues studied, the starch content in the symptomatic shoots was significantly lower (P < 0.05) than in the asymptomatic shoots (Figures 2, 3). In the 2012 sampling, no differences were observed between the two groups of non-symptomatic samples with the exception of the cortex (Figure 2). In the 2013 sampling, the symptomatic and dead shoots had significantly lower starch content in all stem tissues compared with asymptomatic shoots (Figure 3). In the same way, the density of occluded vessels in the affected or dead shoots was significantly (P < 0.05) higher than that assessed in the asymptomatic shoots (Figure 3, inside). Figure 1B illustrates the differences between the occluded and non-occluded vessels in the microscopic observations.

The disease progression is shown in Figure 4. The susceptible ‘Arbequina’ and ‘Picual’ cultivars’ showed symptoms since the 5th week after inoculation and rapidly developed symptoms between the 6th and the 11th weeks, reaching a final severity around 3. In contrast, symptoms of the resistant ‘Frantoio’ cultivar started after the 9th week following inoculation and reached a final severity below 0.5. The percentage of plants with symptoms was 100% in ‘Picual,’ 76% in ‘Arbequina,’ and 11% in ‘Frantoio.’

Pathogen DNA in plant tissue was quantified by real-time PCR. Inhibition of PCR by plant matrix was excluded by comparing threshold cycle values for pure fungal DNA with fungal DNA mixed with control plant DNA (data not shown). V. dahliae DNA was only detected in inoculated plants, with DNA amounts that were significantly higher in the susceptible cultivars from the 7th week after inoculation (P = 0.0039). At this point in time, the amount of pathogen DNA in the stems of the ‘Arbequina’ and ‘Picual’ inoculated plants was 0.3 and 10.4 ng/g of plant tissue, respectively, and increased until the 9th week after inoculation in concurrence with a period of rapid symptom development (Figure 4). Afterward, the DNA amount slowly decreased to 10.9 and 20.7 ng/g of plant tissue in ‘Arbequina’ and ‘Picual,’ reaching low values in dead plants (plants with disease severity value = 4). In cultivar ‘Frantoio,’ fungal DNA was only detected in one of all the sampled inoculated plants.

The starch content and the density of occluded vessels were determined in the same plants used for pathogen DNA, but in this case results are presented according to the severity reached by each plant, independently of the time after the inoculation. In the three cultivars, the starch content in the inoculated plants that showed slight disease symptoms (lower than 1 in the rating scale) did not significantly differ from the non-inoculated plants (Figure 5). On the other hand, in those plants with more severe symptoms (severity rating ≥ 1, only found in ‘Picual’ and ‘Arbequina’ cultivars), a significant reduction in starch content was observed.

Density of occluded vessels was also related to disease severity (Figure 6). Non-inoculated plants and those inoculated but asymptomatic (disease severity = 0) showed a significantly (P < 0.05) lower density of occluded vessels than inoculated plants that were more severely affected (disease severity ≥ 1). These plants, from the ‘Picual’ and ‘Arbequina’ cultivars, reached mean values higher than 12 occluded vessels per mm² of xylem, whereas in the resistant ‘Frantoio’ cultivar, the mean values were always below 2.

Plants exposed to 10 days without watering showed wilt symptoms and a significantly (P < 0.05) lower starch content in their stems than plants that had received daily irrigation (Figure 7). This decrease was exhibited in all tissues, and especially in the pith and xylem. In plants exposed to 10 days without watering followed by daily irrigation over a period of 14 days, recovery from wilting symptoms was observed in four out of six plants. The starch content in the plants showing recovery was similar to that in plants with continuous adequate irrigation (Figure 7), whereas in plants that did not show recovery, the starch content was similar to that found in plants that underwent 10 days without irrigation (data no shown). On the other hand, xylem vessel occlusions were not detected in any of the three groups of plants.