Susceptibility of grapevine pruning wounds to infection by Lasiodiplodia theobromae and Neofusicoccum parvum

Authors


E-mail: wdgubler@ucdavis.edu

Abstract

The susceptibility of 1- and 2-year-old grapevine wood to botryosphaeria canker caused by Lasiodiplodia theobromae and Neofusicoccum parvum was evaluated in California in two seasons. In the 2007 dormant season, pruning-wound susceptibility was highest when wounds were inoculated immediately after pruning in December (80% of pruning wounds were infected in Chardonnay for both fungal species and 75% and 98% in Cabernet Sauvignon for N. parvum and L. theobromae, respectively). In the 2008 dormant season, pruning-wound susceptibility was highest in November in Chardonnay (86% and 93% for N. parvum and L. theobromae, respectively) and in December in Cabernet Sauvignon (71% and 75% for N. parvum and L. theobromae, respectively). The lowest infection rates (13–35%) were observed when vines were pruned and inoculated in March in both dormant seasons and for both cultivars. Susceptibility of pruning wounds did not differ significantly (= 0·7612) between 1- and 2-year-old wood and consequently, pruning-wound protection treatments should be applied to all wounds. In conclusion, grapevine pruning wounds were susceptible to infection by L. theobromae and N. parvum to varying extents throughout the dormant season in California (November–March), but, overall, susceptibility of pruning wounds was highest when inoculations were done immediately after pruning and decreased significantly as the interval between pruning and inoculation increased. Results of this study suggest that pruning grapevines in late winter (March) in California would significantly reduce the risk of infection by L. theobromae and N. parvum.

Introduction

Several species in the family Botryosphaeriaceae are important grapevine vascular pathogens causing severe and economically significant decline and dieback symptoms of grapevines worldwide (Phillips, 2002; van Niekerk et al., 2004; Úrbez-Torres et al., 2006; Luque et al., 2009). In California, Botryosphaeriaceae species are responsible for botryosphaeria canker (locally known as bot canker), a grapevine trunk disease that along with eutypa dieback of grapevines is responsible for economic losses estimated at over $260 million annually to the California grapevine industry (Siebert, 2001). Symptoms of botryosphaeria canker are characterized by perennial cankers (wedge-shape) in the vascular tissue of spurs, cordons and trunks that typically begin in pruning wounds (Úrbez-Torres et al., 2006, 2008). To date, 13 species in the Botryosphaeriaceae family have been associated with botryosphaeria canker of grapevines and the presence of the disease has been confirmed in grape-growing areas of both the Northern and Southern Hemispheres (van Niekerk et al., 2004; Úrbez-Torres et al., 2006, 2008; Pitt et al., 2010). Although all 13 Botryosphaeriaceae species have been shown to be pathogenic on grapevines, pathogenicity studies revealed Lasiodiplodia theobromae and Neofusicoccum parvum to be the most aggressive species in California (Úrbez-Torres & Gubler, 2009).

Epidemiology of Botryosphaeriaceae species infecting grapevines has been studied and release of pycnidiospores has been shown to be correlated with rainfall periods (Kuntzmann et al., 2009; van Niekerk et al., 2010; Úrbez-Torres et al., 2010). Pycnidia of different Botryosphaeriaceae anamorphs associated with botryosphaeria canker can be found in old pruning wounds, infected spurs, embedded in the bark of cordons and/or trunks of infected grapevines, and in pruning debris left in the vineyard. Pycnidiospores are primarily dispersed by rain splash and spread the disease by infecting the exposed xylem of pruning wounds, where spore germination and infection occurs (Úrbez-Torres et al., 2010). Studies of the seasonal abundance of Botryosphaeriaceae pycnidiospores in California showed the highest spore release from December to February. This time period correlates with the main rainy season in California (Úrbez-Torres et al., 2010). Since pruning of grapevines in California primarily occurs during these months, new infection sites are available to the botryosphaeriaceous fungi throughout this time. Susceptibility of grapevine pruning wounds has been studied in other grapevine trunk diseases, such as eutypa dieback caused by Eutypa lata (Petzoldt et al., 1981; Munkvold & Marois, 1995; Chapuis et al., 1998) and young esca and esca caused by Phaeomoniella chlamydospora and Phaeoacremonium aleophilum, respectively (Eskalen et al., 2007; Serra et al., 2008). These studies showed that disease incidence of both eutypa dieback and esca decreased as the length of time between pruning and inoculation increased. Additionally, susceptibility studies of grapevine pruning wounds to infection by E. lata in California and France showed a significant reduction in susceptibility when pruning was conducted in late winter (February–March) (Petzoldt et al., 1981; Munkvold & Marois, 1995; Chapuis et al., 1998). In order to determine the periods of low risk of infection it is necessary to not only study the seasonal patterns of spore production, but also to assess the seasonal variation of pruning-wound susceptibility to infection. However, little is known regarding the susceptibility of grapevine pruning wounds to infection by Botryosphaeriaceae species. The objectives of this study were to determine the duration of susceptibility of pruning wounds and the effect of pruning time on susceptibility of fresh wounds to infection by L. theobromae and N. parvum, the most aggressive Botryosphaeriaceae species in California vineyards.

Materials and methods

Fungal isolates and inoculum preparation

Lasiodiplodia theobromae isolate UCD1255Na and N. parvum isolate UCD1125Na, isolated from grapevines diagnosed with botryosphaeria canker in previous studies (Úrbez-Torres et al., 2006), were obtained from the Plant Pathology Department, University of California, Davis, fungal collection. Isolates were maintained on 4% potato dextrose agar (PDA) (Difco) mycelial plugs in glass vials with sterile water at 4°C until use. Fungal mycelium of each isolate was recovered from the collection after plating of a mycelial plug on PDA amended with tetracycline hydrochloride (0·01%) (Sigma-Aldrich) (PDA-tet) and incubating it at room temperature (23 ± 2°C) for 10 days. Pruning-wound susceptibility studies were conducted using conidial suspension of each isolate. Spore suspensions were prepared as described by Úrbez-Torres & Gubler (2009). For each isolate, a spore suspension was adjusted to a final concentration of approximately 105 conidia mL−1. A conidial suspension of each isolate was prepared 24 h before each inoculation day and stored at 4°C until use. For each inoculation day, germination percentage of each isolate was assessed after incubating 20 μL of each spore suspension on PDA overnight at room temperature. Germination percentage from each suspension was 99–100%.

Effect of pruning date and duration of pruning-wound susceptibility to infection

Field experiments were conducted during two consecutive dormant seasons (2007 and 2008) in a 9- and 7-year-old Chardonnay and Cabernet Sauvignon commercial vineyards in Napa Valley, California, respectively. Grapevine training systems in both vineyards were vertical shoot positioning (VSP), unilateral two-bud spur-pruned cordon with eight to 12 spur positions per cordon (16–24 1-year-old canes per vine). Grapevines were pruned in each month from November to March in both vineyards leaving 30- to 50-cm canes (Table 1). For each pruning date, three vines per fungal isolate were inoculated immediately after pruning and an extra three vines were left uninoculated and exposed to natural infection. For each fungal inoculation treatment all pruning wounds on the vine were inoculated. Inoculated and non-inoculated pruning wounds were left uncovered. Pruning wounds were then inoculated at 12-day intervals until final pruning was conducted in both vineyards (Table 1). Inoculations were done by depositing 50 μL spore suspension on the pruning wounds using a micropipette. Natural infections primarily occur during rain events. Consequently, pruning wounds were slightly wetted by spraying sterile distilled water before inoculation to simulate rain and to assure an even distribution of the pycnidiospores over the wound surface, as described by Petzoldt et al. (1981). In the 2007 trial, pruning-wound susceptibility was assessed up to 48 days after pruning in November, December and January, and up to 24 days after pruning in February. In the 2008 trial, pruning-wound susceptibility was assessed up to 84 days after pruning in November and December, and up to 72, 36, and 12 days after pruning in January, February, and March, respectively.

Table 1.   Pruning dates and total number (in parentheses) of grapevines (Vitis vinifera) pruned in two seasons in a 9-year-old Chardonnay, a 7-year-old Cabernet Sauvignon and a 17-year-old Cabernet Sauvignon vineyard in California
MonthPruning dates
2007–2008a2008–2009a2008–2009b
  1. a9-year-old Chardonnay and 7-year-old Cabernet Sauvignon vineyards.

  2. b17-year-old Cabernet Sauvignon vineyard.

November22/11/2007 (33)04/11/2008 (51) 
December16/12/2007 (33)10/12/2008 (51)18/12/2008 (75)
January09/01/2008 (33)02/01/2009 (45)08/01/2009 (57)
February12/02/2008 (21)07/02/2009 (27)05/02/2009 (33)
March07/03/2008 (9)03/03/2009 (21)05/03/2009 (18)
Final pruning15/03/200820/03/200921/03/2009

Two-year-old versus one-year-old wood pruning-wound susceptibility

In the 2008 dormant season a second trial was conducted in a 17-year-old Cabernet Sauvignon commercial vineyard in Napa Valley, California to determine the pruning-wound susceptibility of 2-year-old grapevine wood to L. theobromae and N. parvum. The grapevine training system was the same as described above. Grapevines were pruned monthly from December to March (Table 1). Pruning was conducted by leaving one 30- to 50-cm cane (1-year-old wood) and one wound (2-year-old wood) per spur position. For each pruning date, three vines per fungal isolate were inoculated immediately after pruning and an extra three vines were left uninoculated and exposed to natural infection. For each fungal inoculation treatment all 1- and 2-year-old wood was inoculated as previously described. Pruning wounds were then inoculated at 7-day intervals until final pruning was conducted (Table 1). Pruning-wound susceptibility was assessed up to 84, 63, 35 and 7 days after pruning in December, January, February and March, respectively.

Assessment of pruning-wound susceptibility to infection

Fungal re-isolation from the inoculated wood was conducted after final pruning for both years. In the field, 1- and 2-year-old canes were excised about 20 and 3 cm below the point of inoculation, respectively. In the laboratory, canes were surface-disinfested in 4% sodium hypochlorite for 5 min. After air drying, the bark around the wound was removed and fragments of necrotic tissue from about 0·5 cm below the point of inoculation were cultured on PDA-tet at room temperature until fungal colonies were observed. For both cultivars in both seasons, 1-year-old canes (15 per vine) were assessed. In the 17-year-old Cabernet Sauvignon vineyard, five 1- and 2-year-old canes per vine were assessed. The recovered fungi were identified based on colony morphology on PDA-tet and conidial characteristics by comparing with those reported in previous studies (Úrbez-Torres et al., 2006).

Data analyses

Inoculated vines were established in a completely randomized design with three replications per treatment. Data were expressed as percentages of infected pruning wounds per inoculation time. The data were subjected to analysis of variance using statistical procedures in sas9·1 (SAS Institute) after an arcsine transformation. A factorial anova was performed for significant effects of four independent variables and their interactions, cultivars, fungal species, age of wood and pruning date on susceptibility of pruning wounds. The significance of differences in wound susceptibility among treatments was tested with anova and means were separated using Fisher’s least significant difference (LSD) test at the 5% significance level. Percentage of infected pruning wounds was correlated with the age of the pruning wound at the time of inoculation by using simple linear regression analysis (proc glm). In addition, percentage of infected pruning wounds for each inoculation time was superimposed with records of average temperature for the same period. Daily average temperature values were collected from weather stations adjacent to vineyard sites (California Irrigation Management Information System, Department of Water Resources, Office of Water Use Efficiency). A correlation analysis was performed to determine the correlation between the percentage of infected pruning wounds and the average temperature.

Results

Effect of pruning date and duration of pruning-wound susceptibility to infection

For both cultivars in both dormant seasons, artificially inoculated pruning wounds were successfully colonized by L. theobromae and N. parvum (Figs 1 and 2). Analysis of the full dataset, following transformation, revealed no significant differences between the 2007 and 2008 trials (= 0·12, = 0·7292) and therefore, data were pooled across years. The analysis of variance indicated that all main effects (cultivar, fungal species and pruning date) were significant (Table 2). The overall infection rate was significantly higher in the 7-year-old Cabernet Sauvignon than in the 9-year-old Chardonnay vineyard in the 2007 dormant season (= 3·9, = 0·0494). However, the overall infection rate was not significantly different between both cultivars in the 2008 dormant season (= 0·23, = 0·8994). Susceptibility of pruning wounds was significantly higher to infection by L. theobromae than N. parvum in both dormant seasons and both cultivars (Table 2).

Figure 1.

 Percentage of Chardonnay and Cabernet Sauvignon pruning wounds infected with Lasiodiplodia theobromae (inline image) and Neofusicoccum parvum (inline image) at different pruning-wound ages during the 2007 dormant season. Each value corresponds to the mean of 45 samples. Bars represent standard error of the mean. Temperature (inline image).

Figure 2.

 Percentage of Chardonnay and Cabernet Sauvignon pruning wounds infected with Lasiodiplodia theobromae (inline image) and Neofusicoccum parvum (inline image) at different pruning-wound ages during the 2008 dormant season. Each value corresponds to the mean of 45 samples. Bars represent standard error of the mean. Temperature (inline image).

Table 2.   Analysis of the fixed effects cultivar, fungal species (Lasiodiplodia theobromae and Neofusicoccum parvum) and pruning date, and their interactions on percentage of infected pruning wounds in 9-year-old Chardonnay and 7-year-old Cabernet Sauvignon grapevines
Effectd.f.F valueP value
Cultivar14·210·0428
Fungal species17·510·0072
Pruning date466·75< 0·0001
Cultivar × fungal species10·460·4993
Cultivar × pruning date42·430·0519
Fungal species × pruning date40·770·5442

In the 2007 dormant season, pruning-wound susceptibility to infection by both fungal species was highest when inoculations were done immediately after pruning (fresh wounds) in December and February in both cultivars, and in January in Cabernet Sauvignon (Fig. 1). Pruning-wound susceptibility was slightly higher in Chardonnay when wounds were inoculated with L. theobromae 24 days after pruning in November and January and when wounds were inoculated with N. parvum 12 days after pruning in November (Fig. 1). Pruning-wound susceptibility was much higher in Cabernet Sauvignon when wounds were inoculated with L. theobromae 24 days after pruning in November (Fig. 1). In the 2008 dormant season, pruning-wound susceptibility was higher in Chardonnay when wounds were inoculated with L. theobromae 12 days after pruning in January (Fig. 2). For the rest of pruning dates, pruning-wound susceptibility to infection by both fungal species was highest when inoculations were done immediately after pruning (Fig. 2).

Overall results showed that susceptibility of fresh pruning wounds was significantly higher (< 0·0001) when vines were pruned in December than when they were pruned in November, January, February or March. Susceptibility of fresh pruning wounds was significantly lower (< 0·0001) in both dormant seasons and both cultivars when vines were pruned in early March compared to pruning in November, December, January or February (Figs 1 and 2). Natural infections in the non-inoculated Chardonnay controls were 0% in November, 5% and 3% in December, 1% and 4% in January, and 0% in both February and March in the 2007 and 2008 dormant seasons, respectively. Natural infections in the non-inoculated Cabernet Sauvignon controls were 0% in November, 2% and 7% in December, 5% and 6% in January, 0% and 1% in February, and 0% in March in the 2007 and 2008 dormant seasons, respectively. Based on morphological characters, the fungal pathogens Diplodia seriata and E. lata were identified from the non-inoculated Chardonnay and Cabernet Sauvignon controls. Lasiodiplodia theobromae and N. parvum were not isolated from the uninoculated controls. Other fungi, such as Alternaria sp., Aspergillus sp., Cladosporium sp., Penicillum sp. and Epicoccum sp. were also periodically isolated from both inoculated and non-inoculated pruning wounds in both seasons and both cultivars.

In both the 2007 and 2008 dormant seasons, Chardonnay and Cabernet Sauvignon pruning wounds were susceptible to infection by L. theobromae and N. parvum to a varying degree throughout the duration of the experiment (Figs 1 and 2). In the 2007 dormant season, wounds made in November, December and January were susceptible for up to 7 weeks (Fig. 1). In the 2008 dormant season, wounds made in November and December were susceptible for up to 12 weeks (Fig. 2). Overall, susceptibility of pruning wounds decreased significantly as the length of time between pruning and inoculation increased (Table 4). During the first 12 days after pruning, vines pruned in February showed a more rapid decrease in susceptibility than those pruned in November, December or January (Figs 1 and 2). The lowest rate of infection was observed when vines were inoculated 12 days after pruning in March in the 2008 dormant season (< 10% infected wounds); however, susceptibility did not differ significantly (= 0·5421) vines inoculated 5 weeks after pruning in February (Fig. 2). Regression analyses indicated a positive relationship between pruning-wound susceptibility and average temperatures for all pruning dates in both cultivars and in both dormant seasons (Table 4). The correlation between increment of temperature and decline of pruning-wound susceptibility was significantly higher (< 0·0001) for the January and February pruning dates than for other pruning dates in both cultivars and both dormant seasons (Table 4).

Table 4.   Correlations (r2 values) between wound age and average temperature (T) at the time of inoculation with Lasiodiplodia theobromae or Neofusicoccum parvum and infection of pruning wounds on 1- and 2-year-old grapevine wood
SeasonCultivarPruning dateL. theobromaeN. parvumL. theobromaeN. parvum
1-year-old2-year-old1-year-old2-year-old
AgeTAgeTAgeTAgeTAgeTAgeT
  1. a17-year-old Cabernet Sauvignon vineyard.

2007ChardonnayNovember0·650·110·780·04        
Cabernet Sauvignon0·220·150·370·19        
ChardonnayDecember0·960·020·860·05        
Cabernet Sauvignon0·470·070·710·02        
ChardonnayJanuary0·740·580·910·66        
Cabernet Sauvignon0·840·120·880·14        
ChardonnayFebruary0·870·870·890·89        
Cabernet Sauvignon0·540·310·940·78        
2008ChardonnayNovember0·830·210·810·26        
Cabernet Sauvignon0·930·210·760·14        
ChardonnayDecember0·750·230·960·43        
Cabernet Sauvignon0·930·290·940·31        
Cabernet Sauvignona    0·860·350·850·260·910·510·930·42
ChardonnayJanuary0·850·290·960·47        
Cabernet Sauvignon0·930·140·860·08        
Cabernet Sauvignona    0·870·020·930·040·870·020·890·02
ChardonnayFebruary0·850·260·680·13        
Cabernet Sauvignon0·720·030·910·03        
Cabernet Sauvignona    0·880·040·890·010·840·020·710·04

Susceptibility of 2-year-old versus 1-year-old wood

Artificially inoculated 1- and 2-year-old wood was successfully colonized by L. theobromae and N. parvum (Fig. 3). The summary of the analysis of variance presented in Table 3 indicated that the main effects (1- vs. 2-year-old wood and fungal species) and their interactions were not significant. On the other hand, analysis of variance showed pruning date to be significant (Table 3). Susceptibility of 1-year-old wood was slightly higher when wounds were inoculated with N. parvum 7 days after pruning in December (Fig. 3). Similarly, susceptibility of 2-year-old wood was slightly higher when wounds were inoculated with L. theobromae 7 days after pruning in January (Fig. 3). For the rest of pruning dates, susceptibility of 1- and 2-year-old wood to infection by both fungal species was higher when inoculations were done immediately after pruning (Fig. 3).

Figure 3.

 Percentage of 1-year-old (inline image) and 2-year-old (inline image) Cabernet Sauvignon pruning wounds infected with Lasiodiplodia theobromae and Neofusicoccum parvum at different pruning-wound ages. Each value corresponds to the mean of 15 samples. Bars represent standard error of the mean. Temperature (inline image).

Table 3.   Analysis of the fixed effects cultivar, fungal species (Lasiodiplodia theobromae and Neofusicoccum parvum) and pruning date, and their interactions on percentage of infected pruning wounds in 17-year-old Cabernet Sauvignon grapevines
Effectd.f.F valueP value
1- vs.2-year-old wood1 0·090·7612
Fungal species1 2·360·1338
Pruning date315·94< 0·0001
1- vs. 2-year-old × fungal species1 0·010·9369
1- vs. 2-year-old × pruning date3 0·190·9011
Fungal species × pruning date3 0·470·7051

Susceptibility of fresh pruning wounds was significantly higher (< 0·0001) when vines were pruned in December and January compared to pruning in February and March. Susceptibility of fresh pruning wounds was lowest (< 0·0001) when vines were pruned in early March (Fig. 3). Natural infections in the non-inoculated Cabernet Sauvignon controls were 26·6% and 49·8% in December, 16·7% and 39·9% in January, 3% and 6·6% in February and 0% in March in 1- and 2-year-old wood, respectively. Based on morphological characters, E. lata was the only grapevine fungal pathogen identified from both 1- and 2-year-old non-inoculated Cabernet Sauvignon controls. Pruning wounds on 1- and 2-year-old wood were susceptible to infection by L. theobromae and N. parvum for up to 12 weeks after pruning in December (Fig. 3). Susceptibility of both 1- and 2-year-old wood decreased significantly as the length of time between pruning and inoculation increased (Table 4). Correlation between increment of temperature and decline of pruning-wound susceptibility was significantly highest (< 0·0001) for the December pruning date (Table 4).

Discussion

Previous studies showed that as a result of both low inoculum levels in the vineyards and low pruning-wound susceptibility, pruning grapes in late winter can significantly reduce the risk of infection by the canker-causing pathogen E. lata in California and France (Ramos et al., 1975; Petzoldt et al., 1981; Munkvold & Marois, 1995; Chapuis et al., 1998). The present findings show that late pruning can also reduce the risk of infection caused by the botryosphaeriaceous fungi L. theobromae and N. parvum. Pruning-wound susceptibility to infection by L. theobromae and N. parvum was significantly higher in early winter (December–January) than late winter (February–March) in California. The significance of these results is reinforced by the fact that spore release by Botryosphaeriaceae species is significantly lower late in the dormant season throughout California (Úrbez-Torres et al., 2010). Furthermore, a high initial rate of infection was also observed in the present study when vines were pruned and inoculated in November. Spore release of Botryosphaeriaceae species was shown to generally start in California vineyards in October–November, coinciding with the first rains of the season (Úrbez-Torres et al., 2010). As a result, even though pruning grapes in November is unlikely in most of the grapevine-growing regions in California, a risk of infection could also exist in those vineyards that are pruned early in the dormant season. Consequently, it is now clear that pruning grapes in California in late winter (late February to early March) is desirable, not only because inoculum levels are lower, but also because pruning wounds are less susceptible.

Grapevine pruning in California typically occurs during the dormant season from December to early March. This study showed that grapevine pruning wounds remained susceptible to infection by L. theobromae and N. parvum during the entire pruning season in California. However, susceptibility of pruning wounds significantly decreased as the length of time between pruning and inoculation increased. This was particularly true during February and March, when pruning-wound susceptibility was lower than 10% in some cases. In a similar study conducted in Italy, grapevine pruning wounds remained susceptible to infection by the botryosphaeriaceous fungus D. seriata from January to May; however, contrary to the present results, pruning-wound susceptibility remained high, even 8 weeks after pruning in March (Serra et al., 2008).

Decline in susceptibility of pruning wounds over time in perennial plants has been correlated with changes in the environmental conditions that directly affect both biological and physiological wound responses occurring in the bark and wood, including the wound-healing process (Doster & Bostock, 1988; Bostock & Stermer, 1989). However, the reasons why grapevine pruning wounds remained highly susceptible to infection by a Botryosphaeriaceae species for a longer period of time in Italy than in California are still not understood.

Data collected over two seasons showed that grapevine pruning wounds remained susceptible to infection by L. theobromae and N. parvum for at least 12 weeks when vines were pruned early in the dormant season (November–January). Duration of susceptibility was shorter when vines were pruned from late February to early March; however, susceptibility of pruning wounds could only be assessed up to 36 and 12 days after pruning in February and March, respectively. Consequently, the duration of grapevine pruning–wound susceptibility to Botryosphaeriaceae species during late spring and summer is not clear at the present time. However, because very few or no Botryosphaeriaceae spores were trapped in late-spring and summer in California (Úrbez-Torres et al., 2010) and wounds are significantly less susceptible in late-winter, it is reasonable to propose that most of the Botryosphaeriaceae infections occur early in the dormant season in California.

Grapevine pruning wounds were shown to remain susceptible to infection by E. lata for up to 4 and 7 weeks after pruning early in the dormant season in California (Petzoldt et al., 1981; Munkvold & Marois, 1995) and France (Chapuis et al., 1998), respectively. However, susceptibility to infection by E. lata was shown to significantly decrease (over 75% reduction) 4 weeks after pruning in December and January (Petzoldt et al., 1981; Chapuis et al., 1998). The results of this study showed that grapevine pruning wounds needed a much longer period of time after pruning in December and January to reach a similar reduction in susceptibility to infection by Botryosphaeriaceae species than the period observed for E. lata. This longer interval of susceptibility to infection by Botryosphaeriaceae species, coupled with higher inoculum levels during the pruning season, greater pathogenicity and a faster wood colonization process, may explain why Botryosphaeriaceae species are more frequently associated with grapevine cankers than E. lata in California (Úrbez-Torres et al., 2006).

Previous studies showed that no significant differences in infection by E. lata were observed when 1-, 2- and 3-year-old wood was inoculated with the same inoculum dose (Munkvold & Marois, 1995; Chapuis et al., 1998). Results from the 17-year-old Cabernet Sauvignon vineyard in the present study (1840 inoculated wounds) showed no differences in infection between 1- and 2-year-old wood when inoculated with L. theobromae or N. parvum. Furthermore, the susceptibility of 2-year-old wood to infection declined at a similar rate to that of 1-year-old wood. As a result, protection treatments against Botryosphaeriaceae species should be recommended for all pruning wounds. Additionally, since large wounds have a higher probability of being infected (Munkvold & Marois, 1995), pruning-wound treatments should be strongly recommended for large pruning wounds, such us those left after retraining dead/dying spur positions, cordons and/or trunks.

Pruning grapes in dry and warm weather is known to be critical for reducing the maximum risk of infection by grapevine canker-causing agents as a result of the lack of airborne inoculum and the enhancement of mechanisms that reduce the susceptibility of pruning wounds (Munkvold & Marois, 1995; Rolshausen et al., 2010). However, grapevines need to be pruned during the dormant season, which in California coincides with the main period of rain. According to the results obtained in this study and those previously reported for E. lata (Petzoldt et al., 1981; Munkvold & Marois, 1995; Weber et al., 2007) pruning grapes from late February to mid-March, or as close as possible to budbreak, would be the best cultural practice to significantly reduce infection levels in California vineyards. However, this strategy means a short window of time for pruning, which may not be feasible for growers with large vineyards. Additionally, late-winter precipitation can occur in California, which would increase inoculum levels and delay physiological responses such as wound healing. Consequently, other management techniques need to be applied to reduce the risk of pruning-wound infection. Pruning-wound protectants have been shown to reduce the risk of infection by both Botryosphaeriaceae species and E. lata (Moller & Kasimatis, 1980; Irelan et al., 1999; Rolshausen & Gubler, 2005; Sosnowski et al., 2008; Rolshausen et al., 2010). However, because of the high economic costs of these applications, they are only viable in vineyards with a high cash return. Additionally, it is still not clear how long the pruning wounds can be protected by these treatments, and more than one application may be necessary throughout the dormant season. On the other hand, pruning-wound protection by using spray programmes can reduce costs and make the treatment of large-acreage vineyards in a shorter period of time feasible. Additionally, spray applications could be rapidly applied after pruning late in the season, therefore resulting in an extra protection period if risk of infection occurs. Laboratory studies are currently underway to evaluate both chemical and biological spray programmes, as well as cultural techniques for management of grapevine trunk diseases.

Acknowledgements

This work was funded by grants from the American Vineyard Foundation and the California Competitive Grant Program for Research in Viticulture and Enology. We gratefully acknowledge Dr T. Gordon (Department of Plant Pathology, University of California, Davis) for valuable comments on this manuscript and T. Merrit and R. Sanchez (Walsh Vineyards Management, Napa, CA) for assisting us with vineyard sites to conduct this research.

Ancillary