Bleeding Sap and Old Wood Are the Two Main Sources of Contamination of Merging Organs of Vine Plants by Xylophilus ampelinus, the Causal Agent of Bacterial Necrosis

Abstract

The spatial distribution of vine plants contaminated by Xylophilus ampelinus, the agent responsible for bacterial necrosis, was studied over a 5-year period within two vineyards in the Cognac area. Both vineyards were planted with Vitis vinifera cv. Ugni blanc but were different in age and agronomic location. The emission of X. ampelinus in contaminated bleeding sap was observed during vine sprouting. Contaminated bleeding sap is an important source of inoculum for external contamination due to the high susceptibility of young merging shoots to the pathogen. X. ampelinus emission by bleeding sap was not affected by the age of the plants or the location of the vineyards. However, its emission was irregular with time, and it varied between two fruit canes from individual plants and between plants as well as between years. Moreover, the two vineyards appeared to be entirely contaminated. Consequently, the behavior of the pathogen is not predictable. The distribution of the pathogen inside vine plant organs was analyzed through the four growing seasons. The old wood was contaminated throughout the year and constituted a stock inoculum for endophytic contamination of crude sap during the winter and the spring. Despite the fact that most of the young green shoots were contaminated in May, X.ampelinus was not found in green shoots in June and September, refuting the hypothesis of an epiphytic life of the pathogen under natural conditions. Although all plants were entirely contaminated in both vineyards, symptoms were rare and were observed on different plants each year.

Xylophilus ampelinus (22), formerly called Xanthomonas ampelina (16), is the causal agent of bacterial necrosis of grapevines and is still responsible for important losses in South Africa (Mansvelt, personal communication) and French vineyards. Indeed, in France, outbreaks appeared in 1993 and 1997 (19). In the last two decades, the most severe losses have been observed for cultivars Ugni blanc, Colombard, and Macabeu in the vineyards of Cognac and Armagnac as well as for cultivar Clairette in the vineyards of Die. These cultivars are widely planted and are all very susceptible to bacterial necrosis. Bacterial necrosis of grapevines is characterized by typical symptoms such as cankers on stems and petioles, by necrotic foliar spots, which are often confused with symptoms of other diseases, and by bud death. So far, grapevines are the only known hosts of X. ampelinus. The susceptibility of cultivars varies from tolerant to highly susceptible, depending on the evaluation method and the vineyard (7, 12, 17, 20). Former studies highlighted periods of very high receptivity of grapevines to inoculation with X. ampelinus in November and December and in February and March (20). In vineyards, contamination can occur during prepruning, pruning, and harvesting. Indeed, bacteria have formerly been found on pruning shears and in the blowers of harvesting machines (15, 19). The capability of X.ampelinus to stay inside vine plants in a latent state for a few years has been suggested (20). Recent studies have also shown that contamination can occur without the appearance of symptoms (C. Dupuits, S. Grall, B. Legendre, F. Poliakoff, B.Barthelet, and C. Manceau, Abstr. 5th Rencontres Plantes-Bactéries, Aussois, France, abstr. 81, 2002). Consequently, vine growers have difficulties certifying that plant vines from their nurseries are free of X. ampelinus. To guaranty healthy plant vines, we need an efficient strategy for X. ampelinus detection at all the plant development steps.

X. ampelinus has formerly been found in the bleeding sap of contaminated vine plants (15). Bleeding sap seems to be an important source of contamination. A previous study showed that X. ampelinus survives and multiplies in the xylem vessels as biofilms at very high concentrations after inoculation (8). The pathogen progresses down to the crown after inoculation on the wounded stem of a young plant. When an inoculum was sprayed on the foliage of plants, the bacterium stayed in high concentrations on/in the leaves when cultivated under a humid atmosphere (8). However, we have no information concerning the behavior of this pathogen and its ecology inside vine plants in the vineyard.

The first part of this paper describes the kinetics of X. ampelinus emission by bleeding sap in vineyards planted with Vitis vinifera cv. Ugni blanc. The second part relates to the evolution of grapevine contamination and the typical symptomatic appearance of bacterial necrosis in a 5-year period in a naturally contaminated vineyard. The final part describes X. ampelinus repartition in entire vine stocks at the different steps of grapevine development.

MATERIALS AND METHODS

Plant material.

The two experimental plots, located in Saint-Preuil and Sigogne, respectively, near Cognac, France, are both planted with Vitis vinifera cv. Ugni blanc, which is very susceptible to X. ampelinus. The Saint-Preuil vineyard, located in the “Grande Champagne” area, was planted in 1996, and symptoms of bacterial necrosis were first observed in 1999. Our study took place with the first nine rows of vine plants (row A to row I; 40 to 111 vine plants in increasing sizes) in this vineyard. The Sigogne vineyard is located in the “Fins Bois” area and was planted in 1990. The Station Viticole of the BNIC (Bureau National Interprofessionnel de Cognac) has monitored this vineyard since 1997, when the first symptoms were noticed. Our study was concerned with only the first 69 vine plants of one row (row H) of this vineyard.

Sampling. (i) Sap sampling.

Sap samples were collected every spring (2001-2004) when vine plants showed an increase in metabolic activity (when the vine plants began to bleed). To collect bleeding sap, the two remaining fruit canes conserved after pruning were cut again a few centimeters into the branch, and a tube was attached to the cut end. Enough sap was collected for analysis. In 2002, the times of tube installation and tube recovery were recorded as well as the sap volume. Bleeding sap samples were stored at 4°C until use.

(ii) Vine plant sampling.

Eleven vine plants from row H in the Sigogne vineyard were cut near the ground level and stored at 4°C for further analysis.

Sample preparation.

Sap samples were ready to use after collection. However, entire vine plants were first cut internode by internode for mature canes and trunks and every three or four internodes for green shoots. Plant samples were cut into small lugs with an axe when necessary and were crushed in a sterile plastic bag with a hammer. Next, crushed samples were soaked in 6 ml of phosphate-buffered saline (NaCl, 4 g; NaH₂PO₄ · 2H₂O, 0.2 g; Na₂HPO₄ · 12H₂O, 2.71 g [per liter]) per gram of plant tissue for 1 to 2 h at room temperature for green shoots and overnight at 4°C for mature canes and trunks. Finally, for immunofluorescence staining, 10-fold dilutions of sap samples or soaked samples were made in phosphate buffer, and 20-μl samples of undiluted and diluted soaking liquids were deposited on glass slides. Then, 1 ml was centrifuged at 4°C for 10 min at 13,000 × g and stored at −20°C for further DNA extraction.

Sample analysis. (i) Immunofluorescence staining.

A polyclonal anti-X. ampelinus antibody raised in a rabbit was used as the primary antibody for indirect immunofluorescence analysis (3). A goat anti-rabbit immunoglobulin G-fluorescein isothiocyanate conjugate was used as the secondary antibody. Microscopic observations were done with an Olympus BH2 microscope under UV light with a 455-nm filter (EY455) at a magnification of ×1,000.

(ii) DNA extraction.

For DNA extraction, the protocol was adapted from the work of Kerkoud et al. (11). Pellets were resuspended in 100 μl of Edwards lysis buffer (Tris-HCl, pH 7.5, 0.2 M; NaCl, 0.25 M; sodium dodecyl sulfate, 0.5% [wt/vol]; and polyvinylpyrrolidone 360, 2% [wt/vol]) and left at room temperature for 10 min. DNAs were extracted from the lysis buffer using a silica-based procedure (Ultraclean 15 DNA purification kit; MoBio Laboratories Inc.) as described by the manufacturer.

(iii) PCR detection.

Before PCR, samples were treated with GeneReleaser (Eurogentec) for purification as described by the manufacturer. Positive controls were generated by suspending X. ampelinus in sterile distilled water. Dilutions of bacterial suspensions were boiled for 10 min and then stored on ice until required. Neither DNA extraction nor GeneReleaser treatment was required for the positive controls. The following two specific primers were designed for the internal transcribed spacer region of the rrn operon of X. ampelinus: XATS1, 5′-TGC GTA GTT CAA CAC CAA AGT G-3′; and XATS2-Biotin, 5′-biotin-TAT GAC CCT CTT TCC ACC AGC-3′. The biotin-conjugated primer was used to analyze the amplified products by a colorimetric technique based on an enzyme-linked immunosorbent assay (ELISA) protocol (14). For amplification, 5 μl of sample or boiled bacterial suspension was added to 45 μl of reaction mixture containing 24.75 μl of ultra-pure water, 5 μl of 10× buffer (Eurogentec), 10 μl of MgCl₂ (25 mM), 2.25 μl of deoxynucleoside triphosphates (20 mM [each]), 1.4 μl of each of the two primers, and 0.2 μl of GoldStar Red DNA polymerase (Eurogentec). The PCR program for the thermocycler (GenAmp system 9700; Applied Biosystems) was 1 cycle of 94°C for 2 min; 35 cycles of 94°C for 45 s, 60°C for 1 min, and 72°C for 1 min; and 1 cycle of 72°C for 2 min.

(iv) Symptom observations.

In June of each year, we checked the two vineyards for the appearance of typical symptoms, such as cankers, foliar spot necrosis, and foliar brick-red circumference.

(v) Data analysis.

We attributed a contamination level index (CI) to each bleeding sap sample according to the results obtained by the different analysis techniques (immunofluorescence and PCR-ELISA), as follows: 0, no bacteria were detected (regardless of the detection technique); 1, <10⁴ bacteria/ml were detected or the sample was positive only by PCR; 2, the bacterial concentration assessed by immunofluorescence was between 10⁴ and 10⁶ bacteria/ml; and 3, the bacterial concentration assessed by immunofluorescence was higher than 10⁶ bacteria/ml.

For statistical analysis, we used the chi-square test or one-way analysis of variance. When differences were significant, we ranged the data using the Duncan test (2). For the study of contamination and canker dispersal, we used the PLSD test of Fisher (Stat Box) and the run analysis described by Madden et al. (13).

RESULTS

Dynamics of X. ampelinus emission in bleeding sap. (i) Bleeding sap production.

In the vineyards of Cognac, the bleeding period starts in the middle of March and lasts about 7 weeks. Bleeding sap samples from 21 vine plants were collected from the two fruit canes remaining after pruning from the middle of March to the end of April in 2001, 2002, 2003, and 2004 in the Sigogne vineyard. In order to assess if there were some differences in the proportions of bleeding fruit canes due to the location of the vineyard, 26 plants from row G of the Saint-Preuil vineyard were studied comparatively in 2003 and 2004. For the two vineyards, we noticed significant differences in the percentages of bleeding vine plants, regardless of the sampling date and year, according to the chi-square test (Fig. 1). In addition, each vine plant bled irregularly during sprouting. Moreover, for a given vine plant, the two fruit canes did not bleed similarly (data not shown).

FIG. 1.

ACI and percentage of bleeding canes for each sampling date in 2001, 2002, 2003, and 2004. Error bars for the ACIs represent standard errors. Letters for the percentages of bleeding canes (a to e for Saint-Preuil and x to z for Sigogne) represent significance according to a chi-square test (P = 0.05).

(ii) Sigogne vineyard.

In 2002, the bleeding sap flow varied from 23 μl/min to 111 μl/min. We observed a significant correlation (r = 0.799, P = 0.05) between the bleeding sap flow and the average temperature of the day before the sampling date; the highest bleeding sap flows corresponded to warmer average temperatures (data not shown).

The proportions of bleeding fruit canes were very high throughout sprouting and varied from one sampling date to another. The percentage of bleeding fruit canes varied from 40% to 88% in 2001, from 81% to 98% in 2002, from 19% to 98% in 2003, and from 50% to 100% in 2004 (Fig. 1). According to a t test on ArcSin (square root of proportion), the average proportion of bleeding fruit canes per sampling date was significantly higher in 2002 (38/52) than in 2001 (29/52) and 2003 (23/52). The average proportion of bleeding fruit canes per sampling date was not significantly different in 2004 (33/40) from those in the other years.

(iii) Saint-Preuil vineyard.

The percentage of bleeding fruit canes varied from 29% to 96% in 2003 and from 42% to 96% in 2004 (Fig. 1). According to a t test on ArcSin (square root of proportion), the average proportion of bleeding fruit canes per sampling date was not significantly higher in 2004 (44/52) than that in 2003 (40/52).

(iv) Comparison of the two vineyards.

The proportions of bleeding fruit canes were compared in 2003 and in 2004, using the chi-square test. In 2003, the proportion of bleeding fruit canes was significantly higher in the Saint-Preuil vineyard thanin the Sigogne vineyard on 19 March (49/52 versus 8/42), 8April (44/52 versus 10/42), and 30 April (33/52 versus 10/42) (Fig. 1). On the other hand, the proportion was significantly higher in the Sigogne vineyard (36/42) than in the Saint-Preuil vineyard (15/52) on 15 April. In 2004, the proportion of bleeding fruit canes was significantly higher in the Saint-Preuil vineyard than in the Sigogne vineyard on 30 March (50/52 versus 32/40) and 14 April (50/52 versus 20/40).

Emission of contaminated bleeding sap.

Bleeding sap samples from the vineyards of Sigogne and Saint-Preuil were analyzed for the presence of X. ampelinus. For a given vine plant, bacterial emission by the bleeding sap was irregular during the sprouting time, regardless of the year and the vineyard (Fig. 1). Bacterial emission by the bleeding sap varied between vine plants, regardless of the sampling date and the year. In addition, we observed that the two fruit canes of a given vine plant did not emit bacteria synchronically (data not shown).

(i) Sigogne vineyard.

In 2001, the vine plants in the Sigogne vineyard produced contaminated bleeding sap throughout the sprouting time (Fig. 1). Only one fruit cane never emitted contaminated sap. In 2002, X. ampelinus was only detected in bleeding sap samples from 3 April to the last sampling date. Only one vine plant and one single fruit cane from a second plant never emitted contaminated bleeding sap. In 2003, no contaminated bleeding sap was detected on the two first sampling dates (Fig. 1). Three vine plants and one single fruit cane from five other plants did not emit contaminated bleeding sap. In 2004, contaminated bleeding sap was detected at every sampling date (Fig. 1). Two vine plants and one single fruit cane from three other plants did not emit contaminated bleeding sap.

Like bacterial emission by the bleeding sap, the CI fluctuated during the sprouting time, regardless of the year (Fig. 1). Thus, the average CI (ACI) varied from 0.8 to 1.7 in 2001, from 0.0 to 1.1 in 2002 and 2003, and from 0.1 to 1.1 in 2004. According to a t test, the ACI was significantly higher in 2001 (1.184) than in 2002 (0.540), 2003 (0.281), or 2004 (0.516) (taking into account all the sampling dates).

(ii) Saint-Preuil vineyard.

In 2003 and 2004, the vine plants produced contaminated bleeding sap during the whole sprouting time, except on 14 April 2003 (Fig. 1). In 2003, all vine plants but four fruit canes never emitted X. ampelinus in the bleeding sap. In 2004, one vine plant never produced contaminated bleeding sap and five vine plants never emitted X. ampelinus in the bleeding sap of one fruit cane out of two. Regardless of the studied year, the ACI fluctuated between vine plants, between the two fruit canes of a vine plant, between the sampling dates, and from one year to the next (data not shown). The ACI varied from 0.0 to 2.1 in 2003 and from 0.1 to 1.0 in 2004 (Fig. 1). According to a t test, the average ACI was not significantly lower in 2004 (0.517) than in 2003 (0.683).

(iii) Comparison of the two vineyards.

In 2003 and 2004, the ACI was higher for the Saint-Preuil vineyard than for the Sigogne vineyard, except for 20 and 23 April 2004 (Fig. 1). According to a t test, the average ACI was significantly higher for the Saint-Preuil vineyard (0.683) than for the Sigogne vineyard (0.321) in 2003 but was not significantly higher (0.517 versus 0.516) in 2004. For the two vineyards, no correlation analysis was found between the contamination index and the maximal/minimal temperature as well as the amount of rainfall of the day before (data not shown).

Symptoms.

We observed three types of symptoms on the studied plants, namely, cankers, foliar spot necrosis, and foliar brick-red circumferences.

(i) Sigogne vineyard.

The typical appearances of symptoms were compared in 2000, 2001, 2002, 2003, and 2004 for 21 vine plants studied. In 2000, 10 plants (H5, H11, H22, H30, H42, H49, H55, H59, H64, and H65) presented typical symptoms. In 2001, we observed typical symptoms on seven vine plants (H1, H2, H12, H13, H15, H20, and H22). Five of them presented cankers (H1, H13, H15, H20, and H22). In 2002, no symptoms were observed. In 2003, six vine plants (H6, H15, H16, H18, H22, and H55) expressed at least typical cankers, two plants (H28 and H55) expressed foliar spot necrosis, and four plants (H6, H27, H55, and H64) presented foliar brick-red circumferences. This last symptom was not observed in the three previous years. It is generally linked to water stress caused by an important aggregation of X. ampelinus cells in xylem vessels, resulting in their obstruction. Two of the vine plants with symptoms (H27 and H55) did not produce contaminated bleeding sap in 2003. In 2004, only four vine plants (H16, H18, H22, and H28) presented typical foliar spot necrosis, but two of them (H16 and H18) did not emit contaminated bleeding sap at sprouting time; no typical cankers were found.

(ii) Vineyard of Saint-Preuil.

The 26 vine plants from row G (G16 to G41) were checked for the presence of symptoms from 2000 to 2004. In 2000, typical cankers were found on four vine plants (G32 to G35); the first three of them and plant G36 presented typical foliar spot necrosis. Both vine plants G35 and G36 presented foliar brick-red circumferences. In 2001, nine vine plants (G18, G20, G23, G30, G31, G32, G35, G36, and G41) presented typical cankers.

(iii) Comparison of the two vineyards.

The proportions of vine plants with cankers, with foliar spot necrosis (but no cankers), and with only foliar brick-red circumferences were compared for each year between the studied vine plants of the two vineyards, using the chi-square test. In 2000 only, the proportion of vine plants with typical cankers was significantly higher in the Sigogne vineyard than in the Saint-Preuil vineyard. The other differences between the two vineyards were not significant if we discounted the nil values. In 2002, symptoms were found only in the Saint-Preuil vineyard.

Proportion of canes producing contaminated bleeding sap per individual vine plant.

On 23 March 2001, before pruning, we sampled bleeding sap from all the canes that bled among 11 vine plants in row H of the Sigogne vineyard. All of these vine plants produced contaminated bleeding sap from at least one cane. Two to seven canes per vine plant were bleeding. A total of 61 canes bled. The individual proportions of canes emitting contaminated bleeding sap ranged from 0.43 to 1.00. The total proportion of canes emitting contaminated bleeding sap was 0.74. On the same sampling date, among 42 fruit canes analyzed for this parameter, 38 bled, and the proportion of canes emitting contaminated bleeding sap was 0.63. According to the chi-square test, there was no significant difference between the proportion of canes emitting contaminated bleeding sap within a single vine plant and that within a group of plants.

Spatial distribution of X. ampelinus in vine plants during vine development.

We collected one to three vine plants on five distinct dates from June 2001 to September 2002. We looked for X. ampelinus cells in every single part of the collected vine plants by immunofluorescence and PCR-ELISA (Fig. 2).

FIG. 2.

Schemes representing contamination of entire vine plants by X. ampelinus at the different steps of vine development. Vine plants were sampled in 2001 and 2002 in the studied vineyard in Sigogne. Vine plants are indicated by the letter H (row H) and a number. Light gray lines, X. ampelinus was not detected; gray lines, X. ampelinus was detected only by PCR; black lines, X. ampelinus was detected by immunofluorescence. For May 2002, merging new green shoots are shown with ball-and-stick representation.

(i) June 2001 during the vegetative period.

The vine plants were pruned in March, leaving two fruit canes on each vine. On 6 June 2001, at the vine preflowering stage, one vine plant (H21) was collected from the Sigogne vineyard. According to the chi-square test, the proportion of contaminated samples of old wood (11/13) was significantly higher than the proportion of contaminated samples of 1-year-old canes (3/12). The three contaminated samples of 1-year-old canes were parts of the same single cane (Fig. 2). The proportion of contaminated samples of green shoots (2/56) was significantly lower than the proportion of contaminated samples of 1-year-old canes, and again these two samples were two parts of a single green shoot. Therefore, X. ampelinus was located mostly in the old wood and partially in 1-year-old canes but in very few green shoots in June 2001.

(ii) January 2002 during dormancy.

Three vine plants (H3, H4, and H7) were collected on 7 January 2002 before pruning. For the three vine plants, all samples of old wood were contaminated (49 samples). The proportion of contaminated samples of old wood was not significantly different in January 2002 from that in June 2001. The proportions of contaminated samples of 1-year-old canes (45/77) and mature shoots (78/145) were not significantly different, but they were significantly lower than the proportion of contaminated old wood samples. They were significantly higher than those in June 2001. Therefore, X. ampelinus colonized 1-year-old canes during dormancy. In addition, we noticed that the repartition of X. ampelinus in the mature shoots on fruit canes was heterogeneous. Thus, on one contaminated fruit cane, we could find both contaminated and noncontaminated shoots. The three analyzed vine plants were not contaminated similarly: vine plant H4 was entirely contaminated (data not shown), while one of the fruit canes of vine plant H3 was not contaminated (Fig. 2) and one of the fruit canes of vine plant H7 had some internodes that were not contaminated, despite being bordered by contaminated internodes (data not shown).

(iii) May 2002 when leaves were laid out.

Three vine plants (H10, H14, and H17) were collected on 3 May 2002. They were all contaminated. The proportion of contaminated samples of old wood (45/62) was significantly higher than the proportions of contaminated samples of 1-year-old canes (26/70) and young green shoots (30/65). The proportion of contaminated samples of 1-year-old canes was not significantly different from the proportion of contaminated samples of young green shoots. The proportions of contaminated samples of old wood and of 1-year-old canes were significantly lower in May 2002 than in January 2002. The fruit canes (1-year-old canes) of the three vine plants were not similarly contaminated: very few of those of vine plants H10 and H14 (data not shown) were contaminated, while those of vine plant H17 were entirely contaminated (Fig. 2). Many young green shoots were contaminated on vine plants H10 and H14, even though the fruit canes were mostly not contaminated. This was probably due to external contaminations. Almost all of the young green shoots of vine plant H17 were contaminated (Fig. 2).

(iv) June 2002 during flowering.

Two vine plants (H19 and H23) were collected on 18 June 2002. They were both contaminated (Fig. 2). The proportion of contaminated samples of old wood (16/22) was not significantly higher than the proportion of contaminated samples of 1-year-old canes (24/43) but was significantly higher than the proportion of contaminated samples of green shoots (10/98). The proportions of contaminated samples of old wood and of 1-year-old canes were not significantly different in June 2002 from those in June 2001. The proportion of contaminated samples of green shoots was significantly lower in June 2002 than in May 2002 but was not significantly different from that in June 2001.

(v) September 2002 during ripening.

Two vine plants (H24 and H25 [except the original trunk]) were collected on 16 September 2002. The two vine plants were contaminated. The proportion of contaminated samples of old wood (8/9) was significantly higher than the proportions of contaminated samples of 1-year-old canes (16/47) and green shoots (2/168). The proportion of contaminated samples of 1-year-old canes was significantly different from the proportion of contaminated samples of green shoots. The proportions of contaminated samples of old wood and 1-year-old canes were not significantly different in September 2002 from those in June 2002. Similar to the observations made on the other sampling dates, the fruit cane contamination levels between plants were not similar: for vine plant H24, one fruit cane was entirely contaminated while only a few internodes of the other fruit cane were contaminated (Fig. 2), and only a few internodes were contaminated on the two fruit canes of plant H25 (data not shown). Thus, the contamination of the vine plants was heterogeneous and varied from one plant to another.

Spatial distribution of contaminated vine plants and symptoms in Saint-Preuil vineyard. (i) Contamination.

We compared the contamination levels of vine plant sap at the sprouting time, on 15 March 2001 and 25 March 2002, in the first nine rows of the Saint-Preuil vineyard. We collected bleeding sap from one of the two fruit canes of each vine plant and checked the samples for the presence of X. ampelinus by immunofluorescence. Negative samples were checked by PCR-ELISA, which appeared to be more sensitive (14) than immunofluorescence. The PLSD Fisher test was then applied to compare the results.

Regardless of the year, the nine rows of vine plants presented contaminated bleeding sap. The contamination fluctuated from one year to another and from one row to another (Fig. 3). Thus, on 15 March 2001, 264 among 683 vine plants produced contaminated bleeding sap, and on 25 March 2002, only 57 among 673 vine plants produced contaminated bleeding sap. There were many significant differences in bleeding sap contamination between the rows (data not shown). The percentage of vine plants producing contaminated bleeding sap varied from 4.8% to 80.8% in 2001 and from 1.1% to 35.5% in 2002.

FIG. 3.

Spatial distribution of vine plants that emitted contaminated bleeding sap on each sampling date in the nine rows (A to I) of the studied vineyard in Saint-Preuil in 2001 and 2002. White bars, no bleeding sap or plant not sampled; gray bars, X. ampelinus was not detected; black bars, X. ampelinus was detected.

In 2002, no aggregation of plants emitting contaminated bleeding sap was found in the nine rows, while in 2001, eight rows presented aggregation (Fig. 3), according to a statistical analysis of the runs (13). We noticed that the distribution of vine plants producing contaminated bleeding sap changed from one year to another. Thus, only 24 vine plants produced contaminated bleeding sap in both 2001 and 2002 (3.3%).

(ii) Symptoms.

The occurrence of cankers fluctuated between 2000, 2001, 2002, 2003, and 2004 (Fig. 4). The total proportions of vine plants with typical cankers were significantly higher in 2001 (44/720) and 2003 (50/720) than those in 2000 (14/720), 2002 (10/720), and 2004 (4/720). Only 17 vine plants presented typical cankers in two of the five years.

FIG. 4.

Spatial distribution of vine plants with typical cankers (black bars) in June of the 5 years of the study in the nine rows (A to I) of the studied vineyard in Saint-Preuil.

The proportion of vine plants with typical cankers varied between the rows (Fig. 4). In 2000, this proportion varied from 0% to 5.1%. The percentage of vine plants with cankers was significantly higher for row G than for rows A, F, and I. In 2001, this proportion varied from 1.4% to 12.2%. The percentage of vine plants with cankers was significantly higher for row G than for rows A, B, D, and E. This percentage varied from 0% to 2.9% in 2002, from 2.4% to 13% in 2003, and from 0% to 2.7% in 2004. There was no significant difference between the rows.

According to statistical analysis, aggregations of plants with cankers were found in rows G and H in 2000, row G in 2001, row H in 2002, and rows C and H in 2003, which indicates the occurrence of small outbreaks dispersed in the vineyard.

DISCUSSION

Bleeding sap is an important source of inoculum for external contaminations of merging shoots.

Bleeding sap leaking from pruning wounds of fruit canes in vineyards is the first external manifestation of the renewal of metabolic activity after the dormancy period. The bleeding period generally begins in March and can last more than 7 weeks in Cognac vineyards. It is caused by an increase in the ground temperature: the increasing ground temperature increases the osmotic pressure in root cells, which causes a high root pressure. Consequently, the water moves into the root cells by osmosis, resulting in increased water movement in the xylem vessels to all the organs of the plant (5). The temperature necessary to trigger high root pressure varies according to the vine cultivar and species (10). We noticed variations in the bleeding sap flow from one sampling date to the other. In 2002, we correlated these variations with variations in the atmospheric temperature the day before. The lag in time that we observed between increasing temperatures and increasing flows of bleeding sap may be due to ground inertia. We found differences in the proportions of bleeding canes per sampling date between the two studied vineyards, which may be due to the nutrition of the plants, as they were from different areas (the “Fins Bois” area in Sigogne and the “Grande Champagne” area in Saint-Preuil) and were not the same age. We observed that all vine plants did not bleed at the same time, even though they all bled at least once during sprouting. The flow of bleeding sap varies according to the vine cultivar, the distance between the wound and the roots, and the vine stock and age (1, 5).We can add that sap bleeding is probably caused by many interactions between the climate, nutrient availability, and genetic traits, which make the forecast of bleeding very difficult.

Bleeding sap is poor in nutrients, but nevertheless it allows the growth and development of X. ampelinus populations. It contains some nitrogen, reductive sugars (mainly glucose and fructose), minerals, and amino acids (5, 6).

Vine plants emitted X. ampelinus in the bleeding sap from the beginning of the bleeding period in 2001, 2002, and 2004 and from about 11 days after the beginning of the bleeding period in 2003. The emission of bacteria by bleeding sap varied between the vine plants and between the canes of a given vine plant, as did bleeding sap production. These differences were probably due to a lack of uniform contamination of the vascular system (8): some canes were irrigated by contaminated xylem vessels, while some were not. Furthermore, no significant difference was observed between the vineyards, meaning that the vineyard location and age have no effect on bacterial emission by bleeding sap. Bacterial emission in the bleeding sap lasted during the whole bleeding period. The bleeding period is associated with bud bursting. By the end of April, young green shoots that already have more than four leaves are generally developed from the initial buds. Merging leaves are very sensitive to X. ampelinus infections. Bleeding sap can fall on merging organs that are close to pruning wounds. Consequently, bleeding sap is an important source of primary and secondary contamination as well. Most foliar plant-pathogenic bacteria have an epiphytic phase, which serves as a source of inoculum for aerial dispersion, including Pseudomonas syringae (9), Xanthomonas axonopodis pv. phaseoli (21), and Xanthomonas axonopodis pv. dieffenbachiae (4). With regard to the results of our previous study (8), we suggested a possible epiphytic lifestyle for X. ampelinus. Our idea was that X. ampelinus survival and development on the leaf surface, followed by symptom development, could be linked to the weather. No epiphytic lifestyle has been observed so far for X. ampelinus in vineyards; indeed, X. ampelinus was not detected on leaves except for those located under the pruning edge (data not shown). Furthermore, in 2002, our analysis of entire vine plants showed that more than one-half of the young shoots were contaminated in May 2002, while in June 2002 very few green branches were contaminated. With regard to these results, we conclude that it is highly probable that X. ampelinus does not develop an epiphytic phase under natural conditions. Typical symptoms were observed in both studied vineyards. There were more vine plants with typical symptoms of bacterial necrosis in 2001 and 2003 than in 2000, 2002, and 2004. In contrast to previous observations (18, 20), we could not clearly correlate the occurrence of symptoms with autumn and winter temperatures and rainfall. All that we observed was that the larger numbers of plants with typical cankers in June 2000, 2001, and 2003 corresponded to a period from September to February with more rainfall than that in 2002 (data not shown). Although all the plants of the two vineyards are entirely contaminated, only a few vine plants developed typical cankers each year. Indeed, all of the studied vine plants emitted bacteria in bleeding sap on at least one sampling date when studied for the parameter of X. ampelinus emission by bleeding sap. The studied vine plants were representative of the whole vineyards. Not all of the vine plants emitted bacteria in the bleeding sap at each sampling time (0% to 84%), even though they were all contaminated. These differences in bleeding sap contamination were not linked to variations in the flow of bleeding sap. In the Saint-Preuil vineyard, we sampled bleeding sap only once a year. Due to the variations in bacterial emission by bleeding sap, the contamination of bleeding sap at a chosen time is not representative of the contamination state of the individual vine plants of the entire vineyard. Our results highlighted differences in bleeding sap contamination among the 3 years of our study. We did not notice any sampling date with a significantly higher contamination index that was common to all 4 years of experimentation. Variations in bacterial emission by the bleeding sap were different each year. Consequently, we cannot predict precisely how X. ampelinus will behave during sprouting in the forthcoming years. Furthermore, contamination cannot be correlated with any climatic event.

Old wood is an inoculum source for endogenous contamination of woody canes and bleeding sap in winter.

In the Sigogne vineyard, regardless of the sampling date, almost all of the old wood was contaminated. We found very high bacterial concentrations in the trunk (>10⁸ bacteria/g of vegetal tissue). These concentrations were higher than those in the other parts of the vine plant. As discussed above, external contamination can take place on canes and green branches. The heterogeneous contamination observed early in spring in green shoots between vine plants while woody parts of all plants were similarly contaminated confirmed that the contamination of merging shoots is caused by an external inoculum. Hence, after the artificial contamination of vine plants, X. ampelinus cells progress through the xylem vessels to the trunk, as shown previously (8). Bacteria progress in the vine plant towards the trunk. Taken together, these results suggest that the old wood is where X. ampelinus survives and multiplies throughout the seasons and is a source of inoculum for further cane contamination. This could explain why the procedure of cutting back to the stumps of contaminated vine plants in order to eradicate the disease failed in Die vineyards planted with cultivar Clairette (data not shown).

The X. ampelinus distribution was homogeneous in the trunks of all analyzed vine plants, but it was heterogeneous in the canes and green branches. This heterogeneity is the result of the heterogeneity of contaminated xylem vessel repartition in the wood (8). Thus, some canes were innervated by noncontaminated xylem vessels, while other canes were innervated by contaminated xylem vessels. We suggest that all the xylem vessels do not bleed together, resulting in differences in bleeding sap contamination.

Our results underscored that very few of the newly emerged canes were contaminated from June to November, while in January more than one-half of them were contaminated. We concluded that endogenous contamination of the canes occurred only during dormancy. Following this endogenous contamination, bacteria were then emitted by the bleeding sap during the sprouting time. Former studies showed that apparently healthy canes from cultivar Sultana collected in Greece during dormancy (December to February) were sources of inoculum, since up to 50% of them were contaminated (17). Furthermore, X. ampelinus has been detected in grafts during dormancy (20). Therefore, we suggest that both bacterial multiplication and migration can occur in xylem vessels during dormancy. In contrast to the migration towards the old wood and against the sap flow that follows an external contamination, the migration towards the canes that occurs in winter happens with the sap flow, even though the flow is very slow.

In conclusion, this study highlighted the important role of bleeding sap in bacterial dispersion and in the external contamination of green branches. Such contamination can often be followed by a typical occurrence of symptoms, but symptom development requires specific climatic conditions. Consequently, vine growers should concentrate their actions to control bacterial necrosis just before and during the sprouting time. The heterogeneity of X. ampelinus dispersion stays problematic for the elaboration of a simple sanitary certification of vine plants and multiplication material. The control of bacterial necrosis is confounded by the capability of X. ampelinus to contaminate vine plants without expressing symptoms. The probability of such discrete contamination is increasing for vineyards located in contaminated areas. When symptoms are not expressed in a vineyard, the green branches are generally healthy. Consequently, it is difficult to localize newly contaminated vineyards. Sampling bleeding sap may be an interesting way to investigate whether a vineyard is contaminated or not, but it underestimates the real contamination state of the vine plants and of the entire vineyard.